I don't have anything else queued up at the moment, so this is just Richard's SME patches. -- PMM The following changes since commit 63b38f6c85acd312c2cab68554abf33adf4ee2b3: Merge tag 'pull-target-arm-20220707' of https://git.linaro.org/people/pmaydell/qemu-arm into staging (2022-07-08 06:17:11 +0530) are available in the Git repository at: https://git.linaro.org/people/pmaydell/qemu-arm.git tags/pull-target-arm-20220711 for you to fetch changes up to f9982ceaf26df27d15547a3a7990a95019e9e3a8: linux-user/aarch64: Add SME related hwcap entries (2022-07-11 13:43:52 +0100) ---------------------------------------------------------------- target-arm: * Implement SME emulation, for both system and linux-user ---------------------------------------------------------------- Richard Henderson (45): target/arm: Handle SME in aarch64_cpu_dump_state target/arm: Add infrastructure for disas_sme target/arm: Trap non-streaming usage when Streaming SVE is active target/arm: Mark ADR as non-streaming target/arm: Mark RDFFR, WRFFR, SETFFR as non-streaming target/arm: Mark BDEP, BEXT, BGRP, COMPACT, FEXPA, FTSSEL as non-streaming target/arm: Mark PMULL, FMMLA as non-streaming target/arm: Mark FTSMUL, FTMAD, FADDA as non-streaming target/arm: Mark SMMLA, UMMLA, USMMLA as non-streaming target/arm: Mark string/histo/crypto as non-streaming target/arm: Mark gather/scatter load/store as non-streaming target/arm: Mark gather prefetch as non-streaming target/arm: Mark LDFF1 and LDNF1 as non-streaming target/arm: Mark LD1RO as non-streaming target/arm: Add SME enablement checks target/arm: Handle SME in sve_access_check target/arm: Implement SME RDSVL, ADDSVL, ADDSPL target/arm: Implement SME ZERO target/arm: Implement SME MOVA target/arm: Implement SME LD1, ST1 target/arm: Export unpredicated ld/st from translate-sve.c target/arm: Implement SME LDR, STR target/arm: Implement SME ADDHA, ADDVA target/arm: Implement FMOPA, FMOPS (non-widening) target/arm: Implement BFMOPA, BFMOPS target/arm: Implement FMOPA, FMOPS (widening) target/arm: Implement SME integer outer product target/arm: Implement PSEL target/arm: Implement REVD target/arm: Implement SCLAMP, UCLAMP target/arm: Reset streaming sve state on exception boundaries target/arm: Enable SME for -cpu max linux-user/aarch64: Clear tpidr2_el0 if CLONE_SETTLS linux-user/aarch64: Reset PSTATE.SM on syscalls linux-user/aarch64: Add SM bit to SVE signal context linux-user/aarch64: Tidy target_restore_sigframe error return linux-user/aarch64: Do not allow duplicate or short sve records linux-user/aarch64: Verify extra record lock succeeded linux-user/aarch64: Move sve record checks into restore linux-user/aarch64: Implement SME signal handling linux-user: Rename sve prctls linux-user/aarch64: Implement PR_SME_GET_VL, PR_SME_SET_VL target/arm: Only set ZEN in reset if SVE present target/arm: Enable SME for user-only linux-user/aarch64: Add SME related hwcap entries docs/system/arm/emulation.rst | 4 + linux-user/aarch64/target_cpu.h | 5 +- linux-user/aarch64/target_prctl.h | 62 +- target/arm/cpu.h | 7 + target/arm/helper-sme.h | 126 ++++ target/arm/helper-sve.h | 4 + target/arm/helper.h | 18 + target/arm/translate-a64.h | 45 ++ target/arm/translate.h | 16 + target/arm/sme-fa64.decode | 60 ++ target/arm/sme.decode | 88 +++ target/arm/sve.decode | 41 +- linux-user/aarch64/cpu_loop.c | 9 + linux-user/aarch64/signal.c | 243 ++++++-- linux-user/elfload.c | 20 + linux-user/syscall.c | 28 +- target/arm/cpu.c | 35 +- target/arm/cpu64.c | 11 + target/arm/helper.c | 56 +- target/arm/sme_helper.c | 1140 +++++++++++++++++++++++++++++++++++++ target/arm/sve_helper.c | 28 + target/arm/translate-a64.c | 103 +++- target/arm/translate-sme.c | 373 ++++++++++++ target/arm/translate-sve.c | 393 ++++++++++--- target/arm/translate-vfp.c | 12 + target/arm/translate.c | 2 + target/arm/vec_helper.c | 24 + target/arm/meson.build | 3 + 28 files changed, 2821 insertions(+), 135 deletions(-) create mode 100644 target/arm/sme-fa64.decode create mode 100644 target/arm/sme.decode create mode 100644 target/arm/translate-sme.c From: Richard Henderson <richard.henderson@linaro.org> Dump SVCR, plus use the correct access check for Streaming Mode. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-2-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/cpu.c | 17 ++++++++++++++++- 1 file changed, 16 insertions(+), 1 deletion(-) diff --git a/target/arm/cpu.c b/target/arm/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.c +++ b/target/arm/cpu.c @@ -XXX,XX +XXX,XX @@ static void aarch64_cpu_dump_state(CPUState *cs, FILE *f, int flags) int i; int el = arm_current_el(env); const char *ns_status; + bool sve; qemu_fprintf(f, " PC=%016" PRIx64 " ", env->pc); for (i = 0; i < 32; i++) { @@ -XXX,XX +XXX,XX @@ static void aarch64_cpu_dump_state(CPUState *cs, FILE *f, int flags) el, psr & PSTATE_SP ? 'h' : 't'); + if (cpu_isar_feature(aa64_sme, cpu)) { + qemu_fprintf(f, " SVCR=%08" PRIx64 " %c%c", + env->svcr, + (FIELD_EX64(env->svcr, SVCR, ZA) ? 'Z' : '-'), + (FIELD_EX64(env->svcr, SVCR, SM) ? 'S' : '-')); + } if (cpu_isar_feature(aa64_bti, cpu)) { qemu_fprintf(f, " BTYPE=%d", (psr & PSTATE_BTYPE) >> 10); } @@ -XXX,XX +XXX,XX @@ static void aarch64_cpu_dump_state(CPUState *cs, FILE *f, int flags) qemu_fprintf(f, " FPCR=%08x FPSR=%08x\n", vfp_get_fpcr(env), vfp_get_fpsr(env)); - if (cpu_isar_feature(aa64_sve, cpu) && sve_exception_el(env, el) == 0) { + if (cpu_isar_feature(aa64_sme, cpu) && FIELD_EX64(env->svcr, SVCR, SM)) { + sve = sme_exception_el(env, el) == 0; + } else if (cpu_isar_feature(aa64_sve, cpu)) { + sve = sve_exception_el(env, el) == 0; + } else { + sve = false; + } + + if (sve) { int j, zcr_len = sve_vqm1_for_el(env, el); for (i = 0; i <= FFR_PRED_NUM; i++) { -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> This includes the build rules for the decoder, and the new file for translation, but excludes any instructions. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-3-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/translate-a64.h | 1 + target/arm/sme.decode | 20 ++++++++++++++++++++ target/arm/translate-a64.c | 7 ++++++- target/arm/translate-sme.c | 35 +++++++++++++++++++++++++++++++++++ target/arm/meson.build | 2 ++ 5 files changed, 64 insertions(+), 1 deletion(-) create mode 100644 target/arm/sme.decode create mode 100644 target/arm/translate-sme.c diff --git a/target/arm/translate-a64.h b/target/arm/translate-a64.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a64.h +++ b/target/arm/translate-a64.h @@ -XXX,XX +XXX,XX @@ static inline int pred_gvec_reg_size(DisasContext *s) } bool disas_sve(DisasContext *, uint32_t); +bool disas_sme(DisasContext *, uint32_t); void gen_gvec_rax1(unsigned vece, uint32_t rd_ofs, uint32_t rn_ofs, uint32_t rm_ofs, uint32_t opr_sz, uint32_t max_sz); diff --git a/target/arm/sme.decode b/target/arm/sme.decode new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/target/arm/sme.decode @@ -XXX,XX +XXX,XX @@ +# AArch64 SME instruction descriptions +# +# Copyright (c) 2022 Linaro, Ltd +# +# This library is free software; you can redistribute it and/or +# modify it under the terms of the GNU Lesser General Public +# License as published by the Free Software Foundation; either +# version 2.1 of the License, or (at your option) any later version. +# +# This library is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# Lesser General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public +# License along with this library; if not, see <http://www.gnu.org/licenses/>. + +# +# This file is processed by scripts/decodetree.py +# diff --git a/target/arm/translate-a64.c b/target/arm/translate-a64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a64.c +++ b/target/arm/translate-a64.c @@ -XXX,XX +XXX,XX @@ static void aarch64_tr_translate_insn(DisasContextBase *dcbase, CPUState *cpu) } switch (extract32(insn, 25, 4)) { - case 0x0: case 0x1: case 0x3: /* UNALLOCATED */ + case 0x0: + if (!extract32(insn, 31, 1) || !disas_sme(s, insn)) { + unallocated_encoding(s); + } + break; + case 0x1: case 0x3: /* UNALLOCATED */ unallocated_encoding(s); break; case 0x2: diff --git a/target/arm/translate-sme.c b/target/arm/translate-sme.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/target/arm/translate-sme.c @@ -XXX,XX +XXX,XX @@ +/* + * AArch64 SME translation + * + * Copyright (c) 2022 Linaro, Ltd + * + * This library is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; either + * version 2.1 of the License, or (at your option) any later version. + * + * This library is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this library; if not, see <http://www.gnu.org/licenses/>. + */ + +#include "qemu/osdep.h" +#include "cpu.h" +#include "tcg/tcg-op.h" +#include "tcg/tcg-op-gvec.h" +#include "tcg/tcg-gvec-desc.h" +#include "translate.h" +#include "exec/helper-gen.h" +#include "translate-a64.h" +#include "fpu/softfloat.h" + + +/* + * Include the generated decoder. + */ + +#include "decode-sme.c.inc" diff --git a/target/arm/meson.build b/target/arm/meson.build index XXXXXXX..XXXXXXX 100644 --- a/target/arm/meson.build +++ b/target/arm/meson.build @@ -XXX,XX +XXX,XX @@ gen = [ decodetree.process('sve.decode', extra_args: '--decode=disas_sve'), + decodetree.process('sme.decode', extra_args: '--decode=disas_sme'), decodetree.process('neon-shared.decode', extra_args: '--decode=disas_neon_shared'), decodetree.process('neon-dp.decode', extra_args: '--decode=disas_neon_dp'), decodetree.process('neon-ls.decode', extra_args: '--decode=disas_neon_ls'), @@ -XXX,XX +XXX,XX @@ arm_ss.add(when: 'TARGET_AARCH64', if_true: files( 'sme_helper.c', 'translate-a64.c', 'translate-sve.c', + 'translate-sme.c', )) arm_softmmu_ss = ss.source_set() -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> This new behaviour is in the ARM pseudocode function AArch64.CheckFPAdvSIMDEnabled, which applies to AArch32 via AArch32.CheckAdvSIMDOrFPEnabled when the EL to which the trap would be delivered is in AArch64 mode. Given that ARMv9 drops support for AArch32 outside EL0, the trap EL detection ought to be trivially true, but the pseudocode still contains a number of conditions, and QEMU has not yet committed to dropping A32 support for EL[12] when v9 features are present. Since the computation of SME_TRAP_NONSTREAMING is necessarily different for the two modes, we might as well preserve bits within TBFLAG_ANY and allocate separate bits within TBFLAG_A32 and TBFLAG_A64 instead. Note that DDI0616A.a has typos for bits [22:21] of LD1RO in the table of instructions illegal in streaming mode. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-4-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/cpu.h | 7 +++ target/arm/translate.h | 4 ++ target/arm/sme-fa64.decode | 90 ++++++++++++++++++++++++++++++++++++++ target/arm/helper.c | 41 +++++++++++++++++ target/arm/translate-a64.c | 40 ++++++++++++++++- target/arm/translate-vfp.c | 12 +++++ target/arm/translate.c | 2 + target/arm/meson.build | 1 + 8 files changed, 195 insertions(+), 2 deletions(-) create mode 100644 target/arm/sme-fa64.decode diff --git a/target/arm/cpu.h b/target/arm/cpu.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.h +++ b/target/arm/cpu.h @@ -XXX,XX +XXX,XX @@ FIELD(TBFLAG_A32, HSTR_ACTIVE, 9, 1) * the same thing as the current security state of the processor! */ FIELD(TBFLAG_A32, NS, 10, 1) +/* + * Indicates that SME Streaming mode is active, and SMCR_ELx.FA64 is not. + * This requires an SME trap from AArch32 mode when using NEON. + */ +FIELD(TBFLAG_A32, SME_TRAP_NONSTREAMING, 11, 1) /* * Bit usage when in AArch32 state, for M-profile only. @@ -XXX,XX +XXX,XX @@ FIELD(TBFLAG_A64, SMEEXC_EL, 20, 2) FIELD(TBFLAG_A64, PSTATE_SM, 22, 1) FIELD(TBFLAG_A64, PSTATE_ZA, 23, 1) FIELD(TBFLAG_A64, SVL, 24, 4) +/* Indicates that SME Streaming mode is active, and SMCR_ELx.FA64 is not. */ +FIELD(TBFLAG_A64, SME_TRAP_NONSTREAMING, 28, 1) /* * Helpers for using the above. diff --git a/target/arm/translate.h b/target/arm/translate.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.h +++ b/target/arm/translate.h @@ -XXX,XX +XXX,XX @@ typedef struct DisasContext { bool pstate_sm; /* True if PSTATE.ZA is set. */ bool pstate_za; + /* True if non-streaming insns should raise an SME Streaming exception. */ + bool sme_trap_nonstreaming; + /* True if the current instruction is non-streaming. */ + bool is_nonstreaming; /* True if MVE insns are definitely not predicated by VPR or LTPSIZE */ bool mve_no_pred; /* diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ +# AArch64 SME allowed instruction decoding +# +# Copyright (c) 2022 Linaro, Ltd +# +# This library is free software; you can redistribute it and/or +# modify it under the terms of the GNU Lesser General Public +# License as published by the Free Software Foundation; either +# version 2.1 of the License, or (at your option) any later version. +# +# This library is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# Lesser General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public +# License along with this library; if not, see <http://www.gnu.org/licenses/>. + +# +# This file is processed by scripts/decodetree.py +# + +# These patterns are taken from Appendix E1.1 of DDI0616 A.a, +# Arm Architecture Reference Manual Supplement, +# The Scalable Matrix Extension (SME), for Armv9-A + +{ + [ + OK 0-00 1110 0000 0001 0010 11-- ---- ---- # SMOV W|Xd,Vn.B[0] + OK 0-00 1110 0000 0010 0010 11-- ---- ---- # SMOV W|Xd,Vn.H[0] + OK 0100 1110 0000 0100 0010 11-- ---- ---- # SMOV Xd,Vn.S[0] + OK 0000 1110 0000 0001 0011 11-- ---- ---- # UMOV Wd,Vn.B[0] + OK 0000 1110 0000 0010 0011 11-- ---- ---- # UMOV Wd,Vn.H[0] + OK 0000 1110 0000 0100 0011 11-- ---- ---- # UMOV Wd,Vn.S[0] + OK 0100 1110 0000 1000 0011 11-- ---- ---- # UMOV Xd,Vn.D[0] + ] + FAIL 0--0 111- ---- ---- ---- ---- ---- ---- # Advanced SIMD vector operations +} + +{ + [ + OK 0101 1110 --1- ---- 11-1 11-- ---- ---- # FMULX/FRECPS/FRSQRTS (scalar) + OK 0101 1110 -10- ---- 00-1 11-- ---- ---- # FMULX/FRECPS/FRSQRTS (scalar, FP16) + OK 01-1 1110 1-10 0001 11-1 10-- ---- ---- # FRECPE/FRSQRTE/FRECPX (scalar) + OK 01-1 1110 1111 1001 11-1 10-- ---- ---- # FRECPE/FRSQRTE/FRECPX (scalar, FP16) + ] + FAIL 01-1 111- ---- ---- ---- ---- ---- ---- # Advanced SIMD single-element operations +} + +FAIL 0-00 110- ---- ---- ---- ---- ---- ---- # Advanced SIMD structure load/store +FAIL 1100 1110 ---- ---- ---- ---- ---- ---- # Advanced SIMD cryptography extensions +FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS + +# These are the "avoidance of doubt" final table of Illegal Advanced SIMD instructions +# We don't actually need to include these, as the default is OK. +# -001 111- ---- ---- ---- ---- ---- ---- # Scalar floating-point operations +# --10 110- ---- ---- ---- ---- ---- ---- # Load/store pair of FP registers +# --01 1100 ---- ---- ---- ---- ---- ---- # Load FP register (PC-relative literal) +# --11 1100 --0- ---- ---- ---- ---- ---- # Load/store FP register (unscaled imm) +# --11 1100 --1- ---- ---- ---- ---- --10 # Load/store FP register (register offset) +# --11 1101 ---- ---- ---- ---- ---- ---- # Load/store FP register (scaled imm) + +FAIL 0000 0100 --1- ---- 1010 ---- ---- ---- # ADR +FAIL 0000 0100 --1- ---- 1011 -0-- ---- ---- # FTSSEL, FEXPA +FAIL 0000 0101 --10 0001 100- ---- ---- ---- # COMPACT +FAIL 0010 0101 --01 100- 1111 000- ---0 ---- # RDFFR, RDFFRS +FAIL 0010 0101 --10 1--- 1001 ---- ---- ---- # WRFFR, SETFFR +FAIL 0100 0101 --0- ---- 1011 ---- ---- ---- # BDEP, BEXT, BGRP +FAIL 0100 0101 000- ---- 0110 1--- ---- ---- # PMULLB, PMULLT (128b result) +FAIL 0110 0100 --1- ---- 1110 01-- ---- ---- # FMMLA, BFMMLA +FAIL 0110 0101 --0- ---- 0000 11-- ---- ---- # FTSMUL +FAIL 0110 0101 --01 0--- 100- ---- ---- ---- # FTMAD +FAIL 0110 0101 --01 1--- 001- ---- ---- ---- # FADDA +FAIL 0100 0101 --0- ---- 1001 10-- ---- ---- # SMMLA, UMMLA, USMMLA +FAIL 0100 0101 --1- ---- 1--- ---- ---- ---- # SVE2 string/histo/crypto instructions +FAIL 1000 010- -00- ---- 10-- ---- ---- ---- # SVE2 32-bit gather NT load (vector+scalar) +FAIL 1000 010- -00- ---- 111- ---- ---- ---- # SVE 32-bit gather prefetch (vector+imm) +FAIL 1000 0100 0-1- ---- 0--- ---- ---- ---- # SVE 32-bit gather prefetch (scalar+vector) +FAIL 1000 010- -01- ---- 1--- ---- ---- ---- # SVE 32-bit gather load (vector+imm) +FAIL 1000 0100 0-0- ---- 0--- ---- ---- ---- # SVE 32-bit gather load byte (scalar+vector) +FAIL 1000 0100 1--- ---- 0--- ---- ---- ---- # SVE 32-bit gather load half (scalar+vector) +FAIL 1000 0101 0--- ---- 0--- ---- ---- ---- # SVE 32-bit gather load word (scalar+vector) +FAIL 1010 010- ---- ---- 011- ---- ---- ---- # SVE contiguous FF load (scalar+scalar) +FAIL 1010 010- ---1 ---- 101- ---- ---- ---- # SVE contiguous NF load (scalar+imm) +FAIL 1010 010- -01- ---- 000- ---- ---- ---- # SVE load & replicate 32 bytes (scalar+scalar) +FAIL 1010 010- -010 ---- 001- ---- ---- ---- # SVE load & replicate 32 bytes (scalar+imm) +FAIL 1100 010- ---- ---- ---- ---- ---- ---- # SVE 64-bit gather load/prefetch +FAIL 1110 010- -00- ---- 001- ---- ---- ---- # SVE2 64-bit scatter NT store (vector+scalar) +FAIL 1110 010- -10- ---- 001- ---- ---- ---- # SVE2 32-bit scatter NT store (vector+scalar) +FAIL 1110 010- ---- ---- 1-0- ---- ---- ---- # SVE scatter store (scalar+32-bit vector) +FAIL 1110 010- ---- ---- 101- ---- ---- ---- # SVE scatter store (misc) diff --git a/target/arm/helper.c b/target/arm/helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper.c +++ b/target/arm/helper.c @@ -XXX,XX +XXX,XX @@ int sme_exception_el(CPUARMState *env, int el) return 0; } +/* This corresponds to the ARM pseudocode function IsFullA64Enabled(). */ +static bool sme_fa64(CPUARMState *env, int el) +{ + if (!cpu_isar_feature(aa64_sme_fa64, env_archcpu(env))) { + return false; + } + + if (el <= 1 && !el_is_in_host(env, el)) { + if (!FIELD_EX64(env->vfp.smcr_el[1], SMCR, FA64)) { + return false; + } + } + if (el <= 2 && arm_is_el2_enabled(env)) { + if (!FIELD_EX64(env->vfp.smcr_el[2], SMCR, FA64)) { + return false; + } + } + if (arm_feature(env, ARM_FEATURE_EL3)) { + if (!FIELD_EX64(env->vfp.smcr_el[3], SMCR, FA64)) { + return false; + } + } + + return true; +} + /* * Given that SVE is enabled, return the vector length for EL. */ @@ -XXX,XX +XXX,XX @@ static CPUARMTBFlags rebuild_hflags_a32(CPUARMState *env, int fp_el, DP_TBFLAG_ANY(flags, PSTATE__IL, 1); } + /* + * The SME exception we are testing for is raised via + * AArch64.CheckFPAdvSIMDEnabled(), as called from + * AArch32.CheckAdvSIMDOrFPEnabled(). + */ + if (el == 0 + && FIELD_EX64(env->svcr, SVCR, SM) + && (!arm_is_el2_enabled(env) + || (arm_el_is_aa64(env, 2) && !(env->cp15.hcr_el2 & HCR_TGE))) + && arm_el_is_aa64(env, 1) + && !sme_fa64(env, el)) { + DP_TBFLAG_A32(flags, SME_TRAP_NONSTREAMING, 1); + } + return rebuild_hflags_common_32(env, fp_el, mmu_idx, flags); } @@ -XXX,XX +XXX,XX @@ static CPUARMTBFlags rebuild_hflags_a64(CPUARMState *env, int el, int fp_el, } if (FIELD_EX64(env->svcr, SVCR, SM)) { DP_TBFLAG_A64(flags, PSTATE_SM, 1); + DP_TBFLAG_A64(flags, SME_TRAP_NONSTREAMING, !sme_fa64(env, el)); } DP_TBFLAG_A64(flags, PSTATE_ZA, FIELD_EX64(env->svcr, SVCR, ZA)); } diff --git a/target/arm/translate-a64.c b/target/arm/translate-a64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a64.c +++ b/target/arm/translate-a64.c @@ -XXX,XX +XXX,XX @@ static void do_vec_ld(DisasContext *s, int destidx, int element, * unallocated-encoding checks (otherwise the syndrome information * for the resulting exception will be incorrect). */ -static bool fp_access_check(DisasContext *s) +static bool fp_access_check_only(DisasContext *s) { if (s->fp_excp_el) { assert(!s->fp_access_checked); @@ -XXX,XX +XXX,XX @@ static bool fp_access_check(DisasContext *s) return true; } +static bool fp_access_check(DisasContext *s) +{ + if (!fp_access_check_only(s)) { + return false; + } + if (s->sme_trap_nonstreaming && s->is_nonstreaming) { + gen_exception_insn(s, s->pc_curr, EXCP_UDEF, + syn_smetrap(SME_ET_Streaming, false)); + return false; + } + return true; +} + /* Check that SVE access is enabled. If it is, return true. * If not, emit code to generate an appropriate exception and return false. */ @@ -XXX,XX +XXX,XX @@ static void handle_sys(DisasContext *s, uint32_t insn, bool isread, default: g_assert_not_reached(); } - if ((ri->type & ARM_CP_FPU) && !fp_access_check(s)) { + if ((ri->type & ARM_CP_FPU) && !fp_access_check_only(s)) { return; } else if ((ri->type & ARM_CP_SVE) && !sve_access_check(s)) { return; @@ -XXX,XX +XXX,XX @@ static void disas_data_proc_simd_fp(DisasContext *s, uint32_t insn) } } +/* + * Include the generated SME FA64 decoder. + */ + +#include "decode-sme-fa64.c.inc" + +static bool trans_OK(DisasContext *s, arg_OK *a) +{ + return true; +} + +static bool trans_FAIL(DisasContext *s, arg_OK *a) +{ + s->is_nonstreaming = true; + return true; +} + /** * is_guarded_page: * @env: The cpu environment @@ -XXX,XX +XXX,XX @@ static void aarch64_tr_init_disas_context(DisasContextBase *dcbase, dc->mte_active[1] = EX_TBFLAG_A64(tb_flags, MTE0_ACTIVE); dc->pstate_sm = EX_TBFLAG_A64(tb_flags, PSTATE_SM); dc->pstate_za = EX_TBFLAG_A64(tb_flags, PSTATE_ZA); + dc->sme_trap_nonstreaming = EX_TBFLAG_A64(tb_flags, SME_TRAP_NONSTREAMING); dc->vec_len = 0; dc->vec_stride = 0; dc->cp_regs = arm_cpu->cp_regs; @@ -XXX,XX +XXX,XX @@ static void aarch64_tr_translate_insn(DisasContextBase *dcbase, CPUState *cpu) } } + s->is_nonstreaming = false; + if (s->sme_trap_nonstreaming) { + disas_sme_fa64(s, insn); + } + switch (extract32(insn, 25, 4)) { case 0x0: if (!extract32(insn, 31, 1) || !disas_sme(s, insn)) { diff --git a/target/arm/translate-vfp.c b/target/arm/translate-vfp.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-vfp.c +++ b/target/arm/translate-vfp.c @@ -XXX,XX +XXX,XX @@ static bool vfp_access_check_a(DisasContext *s, bool ignore_vfp_enabled) return false; } + /* + * Note that rebuild_hflags_a32 has already accounted for being in EL0 + * and the higher EL in A64 mode, etc. Unlike A64 mode, there do not + * appear to be any insns which touch VFP which are allowed. + */ + if (s->sme_trap_nonstreaming) { + gen_exception_insn(s, s->pc_curr, EXCP_UDEF, + syn_smetrap(SME_ET_Streaming, + s->base.pc_next - s->pc_curr == 2)); + return false; + } + if (!s->vfp_enabled && !ignore_vfp_enabled) { assert(!arm_dc_feature(s, ARM_FEATURE_M)); unallocated_encoding(s); diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static void arm_tr_init_disas_context(DisasContextBase *dcbase, CPUState *cs) dc->vec_len = EX_TBFLAG_A32(tb_flags, VECLEN); dc->vec_stride = EX_TBFLAG_A32(tb_flags, VECSTRIDE); } + dc->sme_trap_nonstreaming = + EX_TBFLAG_A32(tb_flags, SME_TRAP_NONSTREAMING); } dc->cp_regs = cpu->cp_regs; dc->features = env->features; diff --git a/target/arm/meson.build b/target/arm/meson.build index XXXXXXX..XXXXXXX 100644 --- a/target/arm/meson.build +++ b/target/arm/meson.build @@ -XXX,XX +XXX,XX @@ gen = [ decodetree.process('sve.decode', extra_args: '--decode=disas_sve'), decodetree.process('sme.decode', extra_args: '--decode=disas_sme'), + decodetree.process('sme-fa64.decode', extra_args: '--static-decode=disas_sme_fa64'), decodetree.process('neon-shared.decode', extra_args: '--decode=disas_neon_shared'), decodetree.process('neon-dp.decode', extra_args: '--decode=disas_neon_dp'), decodetree.process('neon-ls.decode', extra_args: '--decode=disas_neon_ls'), -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Mark ADR as a non-streaming instruction, which should trap if full a64 support is not enabled in streaming mode. Removing entries from sme-fa64.decode is an easy way to see what remains to be done. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-5-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/translate.h | 7 +++++++ target/arm/sme-fa64.decode | 1 - target/arm/translate-sve.c | 8 ++++---- 3 files changed, 11 insertions(+), 5 deletions(-) diff --git a/target/arm/translate.h b/target/arm/translate.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.h +++ b/target/arm/translate.h @@ -XXX,XX +XXX,XX @@ uint64_t asimd_imm_const(uint32_t imm, int cmode, int op); static bool trans_##NAME(DisasContext *s, arg_##NAME *a) \ { return dc_isar_feature(FEAT, s) && FUNC(s, __VA_ARGS__); } +#define TRANS_FEAT_NONSTREAMING(NAME, FEAT, FUNC, ...) \ + static bool trans_##NAME(DisasContext *s, arg_##NAME *a) \ + { \ + s->is_nonstreaming = true; \ + return dc_isar_feature(FEAT, s) && FUNC(s, __VA_ARGS__); \ + } + #endif /* TARGET_ARM_TRANSLATE_H */ diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme-fa64.decode +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS # --11 1100 --1- ---- ---- ---- ---- --10 # Load/store FP register (register offset) # --11 1101 ---- ---- ---- ---- ---- ---- # Load/store FP register (scaled imm) -FAIL 0000 0100 --1- ---- 1010 ---- ---- ---- # ADR FAIL 0000 0100 --1- ---- 1011 -0-- ---- ---- # FTSSEL, FEXPA FAIL 0000 0101 --10 0001 100- ---- ---- ---- # COMPACT FAIL 0010 0101 --01 100- 1111 000- ---0 ---- # RDFFR, RDFFRS diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static bool do_adr(DisasContext *s, arg_rrri *a, gen_helper_gvec_3 *fn) return gen_gvec_ool_zzz(s, fn, a->rd, a->rn, a->rm, a->imm); } -TRANS_FEAT(ADR_p32, aa64_sve, do_adr, a, gen_helper_sve_adr_p32) -TRANS_FEAT(ADR_p64, aa64_sve, do_adr, a, gen_helper_sve_adr_p64) -TRANS_FEAT(ADR_s32, aa64_sve, do_adr, a, gen_helper_sve_adr_s32) -TRANS_FEAT(ADR_u32, aa64_sve, do_adr, a, gen_helper_sve_adr_u32) +TRANS_FEAT_NONSTREAMING(ADR_p32, aa64_sve, do_adr, a, gen_helper_sve_adr_p32) +TRANS_FEAT_NONSTREAMING(ADR_p64, aa64_sve, do_adr, a, gen_helper_sve_adr_p64) +TRANS_FEAT_NONSTREAMING(ADR_s32, aa64_sve, do_adr, a, gen_helper_sve_adr_s32) +TRANS_FEAT_NONSTREAMING(ADR_u32, aa64_sve, do_adr, a, gen_helper_sve_adr_u32) /* *** SVE Integer Misc - Unpredicated Group -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Mark these as a non-streaming instructions, which should trap if full a64 support is not enabled in streaming mode. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-6-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sme-fa64.decode | 2 -- target/arm/translate-sve.c | 9 ++++++--- 2 files changed, 6 insertions(+), 5 deletions(-) diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme-fa64.decode +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS FAIL 0000 0100 --1- ---- 1011 -0-- ---- ---- # FTSSEL, FEXPA FAIL 0000 0101 --10 0001 100- ---- ---- ---- # COMPACT -FAIL 0010 0101 --01 100- 1111 000- ---0 ---- # RDFFR, RDFFRS -FAIL 0010 0101 --10 1--- 1001 ---- ---- ---- # WRFFR, SETFFR FAIL 0100 0101 --0- ---- 1011 ---- ---- ---- # BDEP, BEXT, BGRP FAIL 0100 0101 000- ---- 0110 1--- ---- ---- # PMULLB, PMULLT (128b result) FAIL 0110 0100 --1- ---- 1110 01-- ---- ---- # FMMLA, BFMMLA diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static bool do_predset(DisasContext *s, int esz, int rd, int pat, bool setflag) TRANS_FEAT(PTRUE, aa64_sve, do_predset, a->esz, a->rd, a->pat, a->s) /* Note pat == 31 is #all, to set all elements. */ -TRANS_FEAT(SETFFR, aa64_sve, do_predset, 0, FFR_PRED_NUM, 31, false) +TRANS_FEAT_NONSTREAMING(SETFFR, aa64_sve, + do_predset, 0, FFR_PRED_NUM, 31, false) /* Note pat == 32 is #unimp, to set no elements. */ TRANS_FEAT(PFALSE, aa64_sve, do_predset, 0, a->rd, 32, false) @@ -XXX,XX +XXX,XX @@ static bool trans_RDFFR_p(DisasContext *s, arg_RDFFR_p *a) .rd = a->rd, .pg = a->pg, .s = a->s, .rn = FFR_PRED_NUM, .rm = FFR_PRED_NUM, }; + + s->is_nonstreaming = true; return trans_AND_pppp(s, &alt_a); } -TRANS_FEAT(RDFFR, aa64_sve, do_mov_p, a->rd, FFR_PRED_NUM) -TRANS_FEAT(WRFFR, aa64_sve, do_mov_p, FFR_PRED_NUM, a->rn) +TRANS_FEAT_NONSTREAMING(RDFFR, aa64_sve, do_mov_p, a->rd, FFR_PRED_NUM) +TRANS_FEAT_NONSTREAMING(WRFFR, aa64_sve, do_mov_p, FFR_PRED_NUM, a->rn) static bool do_pfirst_pnext(DisasContext *s, arg_rr_esz *a, void (*gen_fn)(TCGv_i32, TCGv_ptr, -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Mark these as a non-streaming instructions, which should trap if full a64 support is not enabled in streaming mode. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-7-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sme-fa64.decode | 3 --- target/arm/translate-sve.c | 22 ++++++++++++---------- 2 files changed, 12 insertions(+), 13 deletions(-) diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme-fa64.decode +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS # --11 1100 --1- ---- ---- ---- ---- --10 # Load/store FP register (register offset) # --11 1101 ---- ---- ---- ---- ---- ---- # Load/store FP register (scaled imm) -FAIL 0000 0100 --1- ---- 1011 -0-- ---- ---- # FTSSEL, FEXPA -FAIL 0000 0101 --10 0001 100- ---- ---- ---- # COMPACT -FAIL 0100 0101 --0- ---- 1011 ---- ---- ---- # BDEP, BEXT, BGRP FAIL 0100 0101 000- ---- 0110 1--- ---- ---- # PMULLB, PMULLT (128b result) FAIL 0110 0100 --1- ---- 1110 01-- ---- ---- # FMMLA, BFMMLA FAIL 0110 0101 --0- ---- 0000 11-- ---- ---- # FTSMUL diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static gen_helper_gvec_2 * const fexpa_fns[4] = { NULL, gen_helper_sve_fexpa_h, gen_helper_sve_fexpa_s, gen_helper_sve_fexpa_d, }; -TRANS_FEAT(FEXPA, aa64_sve, gen_gvec_ool_zz, - fexpa_fns[a->esz], a->rd, a->rn, 0) +TRANS_FEAT_NONSTREAMING(FEXPA, aa64_sve, gen_gvec_ool_zz, + fexpa_fns[a->esz], a->rd, a->rn, 0) static gen_helper_gvec_3 * const ftssel_fns[4] = { NULL, gen_helper_sve_ftssel_h, gen_helper_sve_ftssel_s, gen_helper_sve_ftssel_d, }; -TRANS_FEAT(FTSSEL, aa64_sve, gen_gvec_ool_arg_zzz, ftssel_fns[a->esz], a, 0) +TRANS_FEAT_NONSTREAMING(FTSSEL, aa64_sve, gen_gvec_ool_arg_zzz, + ftssel_fns[a->esz], a, 0) /* *** SVE Predicate Logical Operations Group @@ -XXX,XX +XXX,XX @@ TRANS_FEAT(TRN2_q, aa64_sve_f64mm, gen_gvec_ool_arg_zzz, static gen_helper_gvec_3 * const compact_fns[4] = { NULL, NULL, gen_helper_sve_compact_s, gen_helper_sve_compact_d }; -TRANS_FEAT(COMPACT, aa64_sve, gen_gvec_ool_arg_zpz, compact_fns[a->esz], a, 0) +TRANS_FEAT_NONSTREAMING(COMPACT, aa64_sve, gen_gvec_ool_arg_zpz, + compact_fns[a->esz], a, 0) /* Call the helper that computes the ARM LastActiveElement pseudocode * function, scaled by the element size. This includes the not found @@ -XXX,XX +XXX,XX @@ static gen_helper_gvec_3 * const bext_fns[4] = { gen_helper_sve2_bext_b, gen_helper_sve2_bext_h, gen_helper_sve2_bext_s, gen_helper_sve2_bext_d, }; -TRANS_FEAT(BEXT, aa64_sve2_bitperm, gen_gvec_ool_arg_zzz, - bext_fns[a->esz], a, 0) +TRANS_FEAT_NONSTREAMING(BEXT, aa64_sve2_bitperm, gen_gvec_ool_arg_zzz, + bext_fns[a->esz], a, 0) static gen_helper_gvec_3 * const bdep_fns[4] = { gen_helper_sve2_bdep_b, gen_helper_sve2_bdep_h, gen_helper_sve2_bdep_s, gen_helper_sve2_bdep_d, }; -TRANS_FEAT(BDEP, aa64_sve2_bitperm, gen_gvec_ool_arg_zzz, - bdep_fns[a->esz], a, 0) +TRANS_FEAT_NONSTREAMING(BDEP, aa64_sve2_bitperm, gen_gvec_ool_arg_zzz, + bdep_fns[a->esz], a, 0) static gen_helper_gvec_3 * const bgrp_fns[4] = { gen_helper_sve2_bgrp_b, gen_helper_sve2_bgrp_h, gen_helper_sve2_bgrp_s, gen_helper_sve2_bgrp_d, }; -TRANS_FEAT(BGRP, aa64_sve2_bitperm, gen_gvec_ool_arg_zzz, - bgrp_fns[a->esz], a, 0) +TRANS_FEAT_NONSTREAMING(BGRP, aa64_sve2_bitperm, gen_gvec_ool_arg_zzz, + bgrp_fns[a->esz], a, 0) static gen_helper_gvec_3 * const cadd_fns[4] = { gen_helper_sve2_cadd_b, gen_helper_sve2_cadd_h, -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Mark these as a non-streaming instructions, which should trap if full a64 support is not enabled in streaming mode. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-8-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sme-fa64.decode | 2 -- target/arm/translate-sve.c | 24 +++++++++++++++--------- 2 files changed, 15 insertions(+), 11 deletions(-) diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme-fa64.decode +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS # --11 1100 --1- ---- ---- ---- ---- --10 # Load/store FP register (register offset) # --11 1101 ---- ---- ---- ---- ---- ---- # Load/store FP register (scaled imm) -FAIL 0100 0101 000- ---- 0110 1--- ---- ---- # PMULLB, PMULLT (128b result) -FAIL 0110 0100 --1- ---- 1110 01-- ---- ---- # FMMLA, BFMMLA FAIL 0110 0101 --0- ---- 0000 11-- ---- ---- # FTSMUL FAIL 0110 0101 --01 0--- 100- ---- ---- ---- # FTMAD FAIL 0110 0101 --01 1--- 001- ---- ---- ---- # FADDA diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static bool do_trans_pmull(DisasContext *s, arg_rrr_esz *a, bool sel) gen_helper_gvec_pmull_q, gen_helper_sve2_pmull_h, NULL, gen_helper_sve2_pmull_d, }; - if (a->esz == 0 - ? !dc_isar_feature(aa64_sve2_pmull128, s) - : !dc_isar_feature(aa64_sve, s)) { + + if (a->esz == 0) { + if (!dc_isar_feature(aa64_sve2_pmull128, s)) { + return false; + } + s->is_nonstreaming = true; + } else if (!dc_isar_feature(aa64_sve, s)) { return false; } return gen_gvec_ool_arg_zzz(s, fns[a->esz], a, sel); @@ -XXX,XX +XXX,XX @@ DO_ZPZZ_FP(FMINP, aa64_sve2, sve2_fminp_zpzz) * SVE Integer Multiply-Add (unpredicated) */ -TRANS_FEAT(FMMLA_s, aa64_sve_f32mm, gen_gvec_fpst_zzzz, gen_helper_fmmla_s, - a->rd, a->rn, a->rm, a->ra, 0, FPST_FPCR) -TRANS_FEAT(FMMLA_d, aa64_sve_f64mm, gen_gvec_fpst_zzzz, gen_helper_fmmla_d, - a->rd, a->rn, a->rm, a->ra, 0, FPST_FPCR) +TRANS_FEAT_NONSTREAMING(FMMLA_s, aa64_sve_f32mm, gen_gvec_fpst_zzzz, + gen_helper_fmmla_s, a->rd, a->rn, a->rm, a->ra, + 0, FPST_FPCR) +TRANS_FEAT_NONSTREAMING(FMMLA_d, aa64_sve_f64mm, gen_gvec_fpst_zzzz, + gen_helper_fmmla_d, a->rd, a->rn, a->rm, a->ra, + 0, FPST_FPCR) static gen_helper_gvec_4 * const sqdmlal_zzzw_fns[] = { NULL, gen_helper_sve2_sqdmlal_zzzw_h, @@ -XXX,XX +XXX,XX @@ TRANS_FEAT(BFDOT_zzzz, aa64_sve_bf16, gen_gvec_ool_arg_zzzz, TRANS_FEAT(BFDOT_zzxz, aa64_sve_bf16, gen_gvec_ool_arg_zzxz, gen_helper_gvec_bfdot_idx, a) -TRANS_FEAT(BFMMLA, aa64_sve_bf16, gen_gvec_ool_arg_zzzz, - gen_helper_gvec_bfmmla, a, 0) +TRANS_FEAT_NONSTREAMING(BFMMLA, aa64_sve_bf16, gen_gvec_ool_arg_zzzz, + gen_helper_gvec_bfmmla, a, 0) static bool do_BFMLAL_zzzw(DisasContext *s, arg_rrrr_esz *a, bool sel) { -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Mark these as a non-streaming instructions, which should trap if full a64 support is not enabled in streaming mode. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-9-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sme-fa64.decode | 3 --- target/arm/translate-sve.c | 15 +++++++++++---- 2 files changed, 11 insertions(+), 7 deletions(-) diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme-fa64.decode +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS # --11 1100 --1- ---- ---- ---- ---- --10 # Load/store FP register (register offset) # --11 1101 ---- ---- ---- ---- ---- ---- # Load/store FP register (scaled imm) -FAIL 0110 0101 --0- ---- 0000 11-- ---- ---- # FTSMUL -FAIL 0110 0101 --01 0--- 100- ---- ---- ---- # FTMAD -FAIL 0110 0101 --01 1--- 001- ---- ---- ---- # FADDA FAIL 0100 0101 --0- ---- 1001 10-- ---- ---- # SMMLA, UMMLA, USMMLA FAIL 0100 0101 --1- ---- 1--- ---- ---- ---- # SVE2 string/histo/crypto instructions FAIL 1000 010- -00- ---- 10-- ---- ---- ---- # SVE2 32-bit gather NT load (vector+scalar) diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static gen_helper_gvec_3_ptr * const ftmad_fns[4] = { NULL, gen_helper_sve_ftmad_h, gen_helper_sve_ftmad_s, gen_helper_sve_ftmad_d, }; -TRANS_FEAT(FTMAD, aa64_sve, gen_gvec_fpst_zzz, - ftmad_fns[a->esz], a->rd, a->rn, a->rm, a->imm, - a->esz == MO_16 ? FPST_FPCR_F16 : FPST_FPCR) +TRANS_FEAT_NONSTREAMING(FTMAD, aa64_sve, gen_gvec_fpst_zzz, + ftmad_fns[a->esz], a->rd, a->rn, a->rm, a->imm, + a->esz == MO_16 ? FPST_FPCR_F16 : FPST_FPCR) /* *** SVE Floating Point Accumulating Reduction Group @@ -XXX,XX +XXX,XX @@ static bool trans_FADDA(DisasContext *s, arg_rprr_esz *a) if (a->esz == 0 || !dc_isar_feature(aa64_sve, s)) { return false; } + s->is_nonstreaming = true; if (!sve_access_check(s)) { return true; } @@ -XXX,XX +XXX,XX @@ static bool trans_FADDA(DisasContext *s, arg_rprr_esz *a) DO_FP3(FADD_zzz, fadd) DO_FP3(FSUB_zzz, fsub) DO_FP3(FMUL_zzz, fmul) -DO_FP3(FTSMUL, ftsmul) DO_FP3(FRECPS, recps) DO_FP3(FRSQRTS, rsqrts) #undef DO_FP3 +static gen_helper_gvec_3_ptr * const ftsmul_fns[4] = { + NULL, gen_helper_gvec_ftsmul_h, + gen_helper_gvec_ftsmul_s, gen_helper_gvec_ftsmul_d +}; +TRANS_FEAT_NONSTREAMING(FTSMUL, aa64_sve, gen_gvec_fpst_arg_zzz, + ftsmul_fns[a->esz], a, 0) + /* *** SVE Floating Point Arithmetic - Predicated Group */ -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Mark these as a non-streaming instructions, which should trap if full a64 support is not enabled in streaming mode. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-10-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sme-fa64.decode | 1 - target/arm/translate-sve.c | 12 ++++++------ 2 files changed, 6 insertions(+), 7 deletions(-) diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme-fa64.decode +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS # --11 1100 --1- ---- ---- ---- ---- --10 # Load/store FP register (register offset) # --11 1101 ---- ---- ---- ---- ---- ---- # Load/store FP register (scaled imm) -FAIL 0100 0101 --0- ---- 1001 10-- ---- ---- # SMMLA, UMMLA, USMMLA FAIL 0100 0101 --1- ---- 1--- ---- ---- ---- # SVE2 string/histo/crypto instructions FAIL 1000 010- -00- ---- 10-- ---- ---- ---- # SVE2 32-bit gather NT load (vector+scalar) FAIL 1000 010- -00- ---- 111- ---- ---- ---- # SVE 32-bit gather prefetch (vector+imm) diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ TRANS_FEAT(FMLALT_zzxw, aa64_sve2, do_FMLAL_zzxw, a, false, true) TRANS_FEAT(FMLSLB_zzxw, aa64_sve2, do_FMLAL_zzxw, a, true, false) TRANS_FEAT(FMLSLT_zzxw, aa64_sve2, do_FMLAL_zzxw, a, true, true) -TRANS_FEAT(SMMLA, aa64_sve_i8mm, gen_gvec_ool_arg_zzzz, - gen_helper_gvec_smmla_b, a, 0) -TRANS_FEAT(USMMLA, aa64_sve_i8mm, gen_gvec_ool_arg_zzzz, - gen_helper_gvec_usmmla_b, a, 0) -TRANS_FEAT(UMMLA, aa64_sve_i8mm, gen_gvec_ool_arg_zzzz, - gen_helper_gvec_ummla_b, a, 0) +TRANS_FEAT_NONSTREAMING(SMMLA, aa64_sve_i8mm, gen_gvec_ool_arg_zzzz, + gen_helper_gvec_smmla_b, a, 0) +TRANS_FEAT_NONSTREAMING(USMMLA, aa64_sve_i8mm, gen_gvec_ool_arg_zzzz, + gen_helper_gvec_usmmla_b, a, 0) +TRANS_FEAT_NONSTREAMING(UMMLA, aa64_sve_i8mm, gen_gvec_ool_arg_zzzz, + gen_helper_gvec_ummla_b, a, 0) TRANS_FEAT(BFDOT_zzzz, aa64_sve_bf16, gen_gvec_ool_arg_zzzz, gen_helper_gvec_bfdot, a, 0) -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Mark these as non-streaming instructions, which should trap if full a64 support is not enabled in streaming mode. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-11-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sme-fa64.decode | 1 - target/arm/translate-sve.c | 35 ++++++++++++++++++----------------- 2 files changed, 18 insertions(+), 18 deletions(-) diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme-fa64.decode +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS # --11 1100 --1- ---- ---- ---- ---- --10 # Load/store FP register (register offset) # --11 1101 ---- ---- ---- ---- ---- ---- # Load/store FP register (scaled imm) -FAIL 0100 0101 --1- ---- 1--- ---- ---- ---- # SVE2 string/histo/crypto instructions FAIL 1000 010- -00- ---- 10-- ---- ---- ---- # SVE2 32-bit gather NT load (vector+scalar) FAIL 1000 010- -00- ---- 111- ---- ---- ---- # SVE 32-bit gather prefetch (vector+imm) FAIL 1000 0100 0-1- ---- 0--- ---- ---- ---- # SVE 32-bit gather prefetch (scalar+vector) diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ DO_SVE2_ZZZ_NARROW(RSUBHNT, rsubhnt) static gen_helper_gvec_flags_4 * const match_fns[4] = { gen_helper_sve2_match_ppzz_b, gen_helper_sve2_match_ppzz_h, NULL, NULL }; -TRANS_FEAT(MATCH, aa64_sve2, do_ppzz_flags, a, match_fns[a->esz]) +TRANS_FEAT_NONSTREAMING(MATCH, aa64_sve2, do_ppzz_flags, a, match_fns[a->esz]) static gen_helper_gvec_flags_4 * const nmatch_fns[4] = { gen_helper_sve2_nmatch_ppzz_b, gen_helper_sve2_nmatch_ppzz_h, NULL, NULL }; -TRANS_FEAT(NMATCH, aa64_sve2, do_ppzz_flags, a, nmatch_fns[a->esz]) +TRANS_FEAT_NONSTREAMING(NMATCH, aa64_sve2, do_ppzz_flags, a, nmatch_fns[a->esz]) static gen_helper_gvec_4 * const histcnt_fns[4] = { NULL, NULL, gen_helper_sve2_histcnt_s, gen_helper_sve2_histcnt_d }; -TRANS_FEAT(HISTCNT, aa64_sve2, gen_gvec_ool_arg_zpzz, - histcnt_fns[a->esz], a, 0) +TRANS_FEAT_NONSTREAMING(HISTCNT, aa64_sve2, gen_gvec_ool_arg_zpzz, + histcnt_fns[a->esz], a, 0) -TRANS_FEAT(HISTSEG, aa64_sve2, gen_gvec_ool_arg_zzz, - a->esz == 0 ? gen_helper_sve2_histseg : NULL, a, 0) +TRANS_FEAT_NONSTREAMING(HISTSEG, aa64_sve2, gen_gvec_ool_arg_zzz, + a->esz == 0 ? gen_helper_sve2_histseg : NULL, a, 0) DO_ZPZZ_FP(FADDP, aa64_sve2, sve2_faddp_zpzz) DO_ZPZZ_FP(FMAXNMP, aa64_sve2, sve2_fmaxnmp_zpzz) @@ -XXX,XX +XXX,XX @@ TRANS_FEAT(SQRDCMLAH_zzzz, aa64_sve2, gen_gvec_ool_zzzz, TRANS_FEAT(USDOT_zzzz, aa64_sve_i8mm, gen_gvec_ool_arg_zzzz, a->esz == 2 ? gen_helper_gvec_usdot_b : NULL, a, 0) -TRANS_FEAT(AESMC, aa64_sve2_aes, gen_gvec_ool_zz, - gen_helper_crypto_aesmc, a->rd, a->rd, a->decrypt) +TRANS_FEAT_NONSTREAMING(AESMC, aa64_sve2_aes, gen_gvec_ool_zz, + gen_helper_crypto_aesmc, a->rd, a->rd, a->decrypt) -TRANS_FEAT(AESE, aa64_sve2_aes, gen_gvec_ool_arg_zzz, - gen_helper_crypto_aese, a, false) -TRANS_FEAT(AESD, aa64_sve2_aes, gen_gvec_ool_arg_zzz, - gen_helper_crypto_aese, a, true) +TRANS_FEAT_NONSTREAMING(AESE, aa64_sve2_aes, gen_gvec_ool_arg_zzz, + gen_helper_crypto_aese, a, false) +TRANS_FEAT_NONSTREAMING(AESD, aa64_sve2_aes, gen_gvec_ool_arg_zzz, + gen_helper_crypto_aese, a, true) -TRANS_FEAT(SM4E, aa64_sve2_sm4, gen_gvec_ool_arg_zzz, - gen_helper_crypto_sm4e, a, 0) -TRANS_FEAT(SM4EKEY, aa64_sve2_sm4, gen_gvec_ool_arg_zzz, - gen_helper_crypto_sm4ekey, a, 0) +TRANS_FEAT_NONSTREAMING(SM4E, aa64_sve2_sm4, gen_gvec_ool_arg_zzz, + gen_helper_crypto_sm4e, a, 0) +TRANS_FEAT_NONSTREAMING(SM4EKEY, aa64_sve2_sm4, gen_gvec_ool_arg_zzz, + gen_helper_crypto_sm4ekey, a, 0) -TRANS_FEAT(RAX1, aa64_sve2_sha3, gen_gvec_fn_arg_zzz, gen_gvec_rax1, a) +TRANS_FEAT_NONSTREAMING(RAX1, aa64_sve2_sha3, gen_gvec_fn_arg_zzz, + gen_gvec_rax1, a) TRANS_FEAT(FCVTNT_sh, aa64_sve2, gen_gvec_fpst_arg_zpz, gen_helper_sve2_fcvtnt_sh, a, 0, FPST_FPCR) -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Mark these as a non-streaming instructions, which should trap if full a64 support is not enabled in streaming mode. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-12-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sme-fa64.decode | 9 --------- target/arm/translate-sve.c | 6 ++++++ 2 files changed, 6 insertions(+), 9 deletions(-) diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme-fa64.decode +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS # --11 1100 --1- ---- ---- ---- ---- --10 # Load/store FP register (register offset) # --11 1101 ---- ---- ---- ---- ---- ---- # Load/store FP register (scaled imm) -FAIL 1000 010- -00- ---- 10-- ---- ---- ---- # SVE2 32-bit gather NT load (vector+scalar) FAIL 1000 010- -00- ---- 111- ---- ---- ---- # SVE 32-bit gather prefetch (vector+imm) FAIL 1000 0100 0-1- ---- 0--- ---- ---- ---- # SVE 32-bit gather prefetch (scalar+vector) -FAIL 1000 010- -01- ---- 1--- ---- ---- ---- # SVE 32-bit gather load (vector+imm) -FAIL 1000 0100 0-0- ---- 0--- ---- ---- ---- # SVE 32-bit gather load byte (scalar+vector) -FAIL 1000 0100 1--- ---- 0--- ---- ---- ---- # SVE 32-bit gather load half (scalar+vector) -FAIL 1000 0101 0--- ---- 0--- ---- ---- ---- # SVE 32-bit gather load word (scalar+vector) FAIL 1010 010- ---- ---- 011- ---- ---- ---- # SVE contiguous FF load (scalar+scalar) FAIL 1010 010- ---1 ---- 101- ---- ---- ---- # SVE contiguous NF load (scalar+imm) FAIL 1010 010- -01- ---- 000- ---- ---- ---- # SVE load & replicate 32 bytes (scalar+scalar) FAIL 1010 010- -010 ---- 001- ---- ---- ---- # SVE load & replicate 32 bytes (scalar+imm) FAIL 1100 010- ---- ---- ---- ---- ---- ---- # SVE 64-bit gather load/prefetch -FAIL 1110 010- -00- ---- 001- ---- ---- ---- # SVE2 64-bit scatter NT store (vector+scalar) -FAIL 1110 010- -10- ---- 001- ---- ---- ---- # SVE2 32-bit scatter NT store (vector+scalar) -FAIL 1110 010- ---- ---- 1-0- ---- ---- ---- # SVE scatter store (scalar+32-bit vector) -FAIL 1110 010- ---- ---- 101- ---- ---- ---- # SVE scatter store (misc) diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static bool trans_LD1_zprz(DisasContext *s, arg_LD1_zprz *a) if (!dc_isar_feature(aa64_sve, s)) { return false; } + s->is_nonstreaming = true; if (!sve_access_check(s)) { return true; } @@ -XXX,XX +XXX,XX @@ static bool trans_LD1_zpiz(DisasContext *s, arg_LD1_zpiz *a) if (!dc_isar_feature(aa64_sve, s)) { return false; } + s->is_nonstreaming = true; if (!sve_access_check(s)) { return true; } @@ -XXX,XX +XXX,XX @@ static bool trans_LDNT1_zprz(DisasContext *s, arg_LD1_zprz *a) if (!dc_isar_feature(aa64_sve2, s)) { return false; } + s->is_nonstreaming = true; if (!sve_access_check(s)) { return true; } @@ -XXX,XX +XXX,XX @@ static bool trans_ST1_zprz(DisasContext *s, arg_ST1_zprz *a) if (!dc_isar_feature(aa64_sve, s)) { return false; } + s->is_nonstreaming = true; if (!sve_access_check(s)) { return true; } @@ -XXX,XX +XXX,XX @@ static bool trans_ST1_zpiz(DisasContext *s, arg_ST1_zpiz *a) if (!dc_isar_feature(aa64_sve, s)) { return false; } + s->is_nonstreaming = true; if (!sve_access_check(s)) { return true; } @@ -XXX,XX +XXX,XX @@ static bool trans_STNT1_zprz(DisasContext *s, arg_ST1_zprz *a) if (!dc_isar_feature(aa64_sve2, s)) { return false; } + s->is_nonstreaming = true; if (!sve_access_check(s)) { return true; } -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Mark these as a non-streaming instructions, which should trap if full a64 support is not enabled in streaming mode. In this case, introduce PRF_ns (prefetch non-streaming) to handle the checks. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-13-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sme-fa64.decode | 3 --- target/arm/sve.decode | 10 +++++----- target/arm/translate-sve.c | 11 +++++++++++ 3 files changed, 16 insertions(+), 8 deletions(-) diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme-fa64.decode +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS # --11 1100 --1- ---- ---- ---- ---- --10 # Load/store FP register (register offset) # --11 1101 ---- ---- ---- ---- ---- ---- # Load/store FP register (scaled imm) -FAIL 1000 010- -00- ---- 111- ---- ---- ---- # SVE 32-bit gather prefetch (vector+imm) -FAIL 1000 0100 0-1- ---- 0--- ---- ---- ---- # SVE 32-bit gather prefetch (scalar+vector) FAIL 1010 010- ---- ---- 011- ---- ---- ---- # SVE contiguous FF load (scalar+scalar) FAIL 1010 010- ---1 ---- 101- ---- ---- ---- # SVE contiguous NF load (scalar+imm) FAIL 1010 010- -01- ---- 000- ---- ---- ---- # SVE load & replicate 32 bytes (scalar+scalar) FAIL 1010 010- -010 ---- 001- ---- ---- ---- # SVE load & replicate 32 bytes (scalar+imm) -FAIL 1100 010- ---- ---- ---- ---- ---- ---- # SVE 64-bit gather load/prefetch diff --git a/target/arm/sve.decode b/target/arm/sve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sve.decode +++ b/target/arm/sve.decode @@ -XXX,XX +XXX,XX @@ LD1RO_zpri 1010010 .. 01 0.... 001 ... ..... ..... \ @rpri_load_msz nreg=0 # SVE 32-bit gather prefetch (scalar plus 32-bit scaled offsets) -PRF 1000010 00 -1 ----- 0-- --- ----- 0 ---- +PRF_ns 1000010 00 -1 ----- 0-- --- ----- 0 ---- # SVE 32-bit gather prefetch (vector plus immediate) -PRF 1000010 -- 00 ----- 111 --- ----- 0 ---- +PRF_ns 1000010 -- 00 ----- 111 --- ----- 0 ---- # SVE contiguous prefetch (scalar plus immediate) PRF 1000010 11 1- ----- 0-- --- ----- 0 ---- @@ -XXX,XX +XXX,XX @@ LD1_zpiz 1100010 .. 01 ..... 1.. ... ..... ..... \ @rpri_g_load esz=3 # SVE 64-bit gather prefetch (scalar plus 64-bit scaled offsets) -PRF 1100010 00 11 ----- 1-- --- ----- 0 ---- +PRF_ns 1100010 00 11 ----- 1-- --- ----- 0 ---- # SVE 64-bit gather prefetch (scalar plus unpacked 32-bit scaled offsets) -PRF 1100010 00 -1 ----- 0-- --- ----- 0 ---- +PRF_ns 1100010 00 -1 ----- 0-- --- ----- 0 ---- # SVE 64-bit gather prefetch (vector plus immediate) -PRF 1100010 -- 00 ----- 111 --- ----- 0 ---- +PRF_ns 1100010 -- 00 ----- 111 --- ----- 0 ---- ### SVE Memory Store Group diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static bool trans_PRF_rr(DisasContext *s, arg_PRF_rr *a) return true; } +static bool trans_PRF_ns(DisasContext *s, arg_PRF_ns *a) +{ + if (!dc_isar_feature(aa64_sve, s)) { + return false; + } + /* Prefetch is a nop within QEMU. */ + s->is_nonstreaming = true; + (void)sve_access_check(s); + return true; +} + /* * Move Prefix * -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Mark these as a non-streaming instructions, which should trap if full a64 support is not enabled in streaming mode. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-14-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sme-fa64.decode | 2 -- target/arm/translate-sve.c | 2 ++ 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme-fa64.decode +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS # --11 1100 --1- ---- ---- ---- ---- --10 # Load/store FP register (register offset) # --11 1101 ---- ---- ---- ---- ---- ---- # Load/store FP register (scaled imm) -FAIL 1010 010- ---- ---- 011- ---- ---- ---- # SVE contiguous FF load (scalar+scalar) -FAIL 1010 010- ---1 ---- 101- ---- ---- ---- # SVE contiguous NF load (scalar+imm) FAIL 1010 010- -01- ---- 000- ---- ---- ---- # SVE load & replicate 32 bytes (scalar+scalar) FAIL 1010 010- -010 ---- 001- ---- ---- ---- # SVE load & replicate 32 bytes (scalar+imm) diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static bool trans_LDFF1_zprr(DisasContext *s, arg_rprr_load *a) if (!dc_isar_feature(aa64_sve, s)) { return false; } + s->is_nonstreaming = true; if (sve_access_check(s)) { TCGv_i64 addr = new_tmp_a64(s); tcg_gen_shli_i64(addr, cpu_reg(s, a->rm), dtype_msz(a->dtype)); @@ -XXX,XX +XXX,XX @@ static bool trans_LDNF1_zpri(DisasContext *s, arg_rpri_load *a) if (!dc_isar_feature(aa64_sve, s)) { return false; } + s->is_nonstreaming = true; if (sve_access_check(s)) { int vsz = vec_full_reg_size(s); int elements = vsz >> dtype_esz[a->dtype]; -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Mark these as a non-streaming instructions, which should trap if full a64 support is not enabled in streaming mode. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-15-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sme-fa64.decode | 3 --- target/arm/translate-sve.c | 2 ++ 2 files changed, 2 insertions(+), 3 deletions(-) diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme-fa64.decode +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS # --11 1100 --0- ---- ---- ---- ---- ---- # Load/store FP register (unscaled imm) # --11 1100 --1- ---- ---- ---- ---- --10 # Load/store FP register (register offset) # --11 1101 ---- ---- ---- ---- ---- ---- # Load/store FP register (scaled imm) - -FAIL 1010 010- -01- ---- 000- ---- ---- ---- # SVE load & replicate 32 bytes (scalar+scalar) -FAIL 1010 010- -010 ---- 001- ---- ---- ---- # SVE load & replicate 32 bytes (scalar+imm) diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static bool trans_LD1RO_zprr(DisasContext *s, arg_rprr_load *a) if (a->rm == 31) { return false; } + s->is_nonstreaming = true; if (sve_access_check(s)) { TCGv_i64 addr = new_tmp_a64(s); tcg_gen_shli_i64(addr, cpu_reg(s, a->rm), dtype_msz(a->dtype)); @@ -XXX,XX +XXX,XX @@ static bool trans_LD1RO_zpri(DisasContext *s, arg_rpri_load *a) if (!dc_isar_feature(aa64_sve_f64mm, s)) { return false; } + s->is_nonstreaming = true; if (sve_access_check(s)) { TCGv_i64 addr = new_tmp_a64(s); tcg_gen_addi_i64(addr, cpu_reg_sp(s, a->rn), a->imm * 32); -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> These functions will be used to verify that the cpu is in the correct state for a given instruction. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-16-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/translate-a64.h | 21 +++++++++++++++++++++ target/arm/translate-a64.c | 34 ++++++++++++++++++++++++++++++++++ 2 files changed, 55 insertions(+) diff --git a/target/arm/translate-a64.h b/target/arm/translate-a64.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a64.h +++ b/target/arm/translate-a64.h @@ -XXX,XX +XXX,XX @@ void write_fp_dreg(DisasContext *s, int reg, TCGv_i64 v); bool logic_imm_decode_wmask(uint64_t *result, unsigned int immn, unsigned int imms, unsigned int immr); bool sve_access_check(DisasContext *s); +bool sme_enabled_check(DisasContext *s); +bool sme_enabled_check_with_svcr(DisasContext *s, unsigned); + +/* This function corresponds to CheckStreamingSVEEnabled. */ +static inline bool sme_sm_enabled_check(DisasContext *s) +{ + return sme_enabled_check_with_svcr(s, R_SVCR_SM_MASK); +} + +/* This function corresponds to CheckSMEAndZAEnabled. */ +static inline bool sme_za_enabled_check(DisasContext *s) +{ + return sme_enabled_check_with_svcr(s, R_SVCR_ZA_MASK); +} + +/* Note that this function corresponds to CheckStreamingSVEAndZAEnabled. */ +static inline bool sme_smza_enabled_check(DisasContext *s) +{ + return sme_enabled_check_with_svcr(s, R_SVCR_SM_MASK | R_SVCR_ZA_MASK); +} + TCGv_i64 clean_data_tbi(DisasContext *s, TCGv_i64 addr); TCGv_i64 gen_mte_check1(DisasContext *s, TCGv_i64 addr, bool is_write, bool tag_checked, int log2_size); diff --git a/target/arm/translate-a64.c b/target/arm/translate-a64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a64.c +++ b/target/arm/translate-a64.c @@ -XXX,XX +XXX,XX @@ static bool sme_access_check(DisasContext *s) return true; } +/* This function corresponds to CheckSMEEnabled. */ +bool sme_enabled_check(DisasContext *s) +{ + /* + * Note that unlike sve_excp_el, we have not constrained sme_excp_el + * to be zero when fp_excp_el has priority. This is because we need + * sme_excp_el by itself for cpregs access checks. + */ + if (!s->fp_excp_el || s->sme_excp_el < s->fp_excp_el) { + s->fp_access_checked = true; + return sme_access_check(s); + } + return fp_access_check_only(s); +} + +/* Common subroutine for CheckSMEAnd*Enabled. */ +bool sme_enabled_check_with_svcr(DisasContext *s, unsigned req) +{ + if (!sme_enabled_check(s)) { + return false; + } + if (FIELD_EX64(req, SVCR, SM) && !s->pstate_sm) { + gen_exception_insn(s, s->pc_curr, EXCP_UDEF, + syn_smetrap(SME_ET_NotStreaming, false)); + return false; + } + if (FIELD_EX64(req, SVCR, ZA) && !s->pstate_za) { + gen_exception_insn(s, s->pc_curr, EXCP_UDEF, + syn_smetrap(SME_ET_InactiveZA, false)); + return false; + } + return true; +} + /* * This utility function is for doing register extension with an * optional shift. You will likely want to pass a temporary for the -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> The pseudocode for CheckSVEEnabled gains a check for Streaming SVE mode, and for SME present but SVE absent. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-17-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/translate-a64.c | 22 ++++++++++++++++------ 1 file changed, 16 insertions(+), 6 deletions(-) diff --git a/target/arm/translate-a64.c b/target/arm/translate-a64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a64.c +++ b/target/arm/translate-a64.c @@ -XXX,XX +XXX,XX @@ static bool fp_access_check(DisasContext *s) return true; } -/* Check that SVE access is enabled. If it is, return true. +/* + * Check that SVE access is enabled. If it is, return true. * If not, emit code to generate an appropriate exception and return false. + * This function corresponds to CheckSVEEnabled(). */ bool sve_access_check(DisasContext *s) { - if (s->sve_excp_el) { - assert(!s->sve_access_checked); - s->sve_access_checked = true; - + if (s->pstate_sm || !dc_isar_feature(aa64_sve, s)) { + assert(dc_isar_feature(aa64_sme, s)); + if (!sme_sm_enabled_check(s)) { + goto fail_exit; + } + } else if (s->sve_excp_el) { gen_exception_insn_el(s, s->pc_curr, EXCP_UDEF, syn_sve_access_trap(), s->sve_excp_el); - return false; + goto fail_exit; } s->sve_access_checked = true; return fp_access_check(s); + + fail_exit: + /* Assert that we only raise one exception per instruction. */ + assert(!s->sve_access_checked); + s->sve_access_checked = true; + return false; } /* -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> These SME instructions are nominally within the SVE decode space, so we add them to sve.decode and translate-sve.c. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-18-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/translate-a64.h | 12 ++++++++++++ target/arm/sve.decode | 5 ++++- target/arm/translate-sve.c | 38 ++++++++++++++++++++++++++++++++++++++ 3 files changed, 54 insertions(+), 1 deletion(-) diff --git a/target/arm/translate-a64.h b/target/arm/translate-a64.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a64.h +++ b/target/arm/translate-a64.h @@ -XXX,XX +XXX,XX @@ static inline int vec_full_reg_size(DisasContext *s) return s->vl; } +/* Return the byte size of the vector register, SVL / 8. */ +static inline int streaming_vec_reg_size(DisasContext *s) +{ + return s->svl; +} + /* * Return the offset info CPUARMState of the predicate vector register Pn. * Note for this purpose, FFR is P16. @@ -XXX,XX +XXX,XX @@ static inline int pred_full_reg_size(DisasContext *s) return s->vl >> 3; } +/* Return the byte size of the predicate register, SVL / 64. */ +static inline int streaming_pred_reg_size(DisasContext *s) +{ + return s->svl >> 3; +} + /* * Round up the size of a register to a size allowed by * the tcg vector infrastructure. Any operation which uses this diff --git a/target/arm/sve.decode b/target/arm/sve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sve.decode +++ b/target/arm/sve.decode @@ -XXX,XX +XXX,XX @@ INDEX_ri 00000100 esz:2 1 imm:s5 010001 rn:5 rd:5 # SVE index generation (register start, register increment) INDEX_rr 00000100 .. 1 ..... 010011 ..... ..... @rd_rn_rm -### SVE Stack Allocation Group +### SVE / Streaming SVE Stack Allocation Group # SVE stack frame adjustment ADDVL 00000100 001 ..... 01010 ...... ..... @rd_rn_i6 +ADDSVL 00000100 001 ..... 01011 ...... ..... @rd_rn_i6 ADDPL 00000100 011 ..... 01010 ...... ..... @rd_rn_i6 +ADDSPL 00000100 011 ..... 01011 ...... ..... @rd_rn_i6 # SVE stack frame size RDVL 00000100 101 11111 01010 imm:s6 rd:5 +RDSVL 00000100 101 11111 01011 imm:s6 rd:5 ### SVE Bitwise Shift - Unpredicated Group diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static bool trans_ADDVL(DisasContext *s, arg_ADDVL *a) return true; } +static bool trans_ADDSVL(DisasContext *s, arg_ADDSVL *a) +{ + if (!dc_isar_feature(aa64_sme, s)) { + return false; + } + if (sme_enabled_check(s)) { + TCGv_i64 rd = cpu_reg_sp(s, a->rd); + TCGv_i64 rn = cpu_reg_sp(s, a->rn); + tcg_gen_addi_i64(rd, rn, a->imm * streaming_vec_reg_size(s)); + } + return true; +} + static bool trans_ADDPL(DisasContext *s, arg_ADDPL *a) { if (!dc_isar_feature(aa64_sve, s)) { @@ -XXX,XX +XXX,XX @@ static bool trans_ADDPL(DisasContext *s, arg_ADDPL *a) return true; } +static bool trans_ADDSPL(DisasContext *s, arg_ADDSPL *a) +{ + if (!dc_isar_feature(aa64_sme, s)) { + return false; + } + if (sme_enabled_check(s)) { + TCGv_i64 rd = cpu_reg_sp(s, a->rd); + TCGv_i64 rn = cpu_reg_sp(s, a->rn); + tcg_gen_addi_i64(rd, rn, a->imm * streaming_pred_reg_size(s)); + } + return true; +} + static bool trans_RDVL(DisasContext *s, arg_RDVL *a) { if (!dc_isar_feature(aa64_sve, s)) { @@ -XXX,XX +XXX,XX @@ static bool trans_RDVL(DisasContext *s, arg_RDVL *a) return true; } +static bool trans_RDSVL(DisasContext *s, arg_RDSVL *a) +{ + if (!dc_isar_feature(aa64_sme, s)) { + return false; + } + if (sme_enabled_check(s)) { + TCGv_i64 reg = cpu_reg(s, a->rd); + tcg_gen_movi_i64(reg, a->imm * streaming_vec_reg_size(s)); + } + return true; +} + /* *** SVE Compute Vector Address Group */ -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-19-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper-sme.h | 2 ++ target/arm/sme.decode | 4 ++++ target/arm/sme_helper.c | 25 +++++++++++++++++++++++++ target/arm/translate-sme.c | 13 +++++++++++++ 4 files changed, 44 insertions(+) diff --git a/target/arm/helper-sme.h b/target/arm/helper-sme.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-sme.h +++ b/target/arm/helper-sme.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_2(set_pstate_sm, TCG_CALL_NO_RWG, void, env, i32) DEF_HELPER_FLAGS_2(set_pstate_za, TCG_CALL_NO_RWG, void, env, i32) + +DEF_HELPER_FLAGS_3(sme_zero, TCG_CALL_NO_RWG, void, env, i32, i32) diff --git a/target/arm/sme.decode b/target/arm/sme.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme.decode +++ b/target/arm/sme.decode @@ -XXX,XX +XXX,XX @@ # # This file is processed by scripts/decodetree.py # + +### SME Misc + +ZERO 11000000 00 001 00000000000 imm:8 diff --git a/target/arm/sme_helper.c b/target/arm/sme_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme_helper.c +++ b/target/arm/sme_helper.c @@ -XXX,XX +XXX,XX @@ void helper_set_pstate_za(CPUARMState *env, uint32_t i) memset(env->zarray, 0, sizeof(env->zarray)); } } + +void helper_sme_zero(CPUARMState *env, uint32_t imm, uint32_t svl) +{ + uint32_t i; + + /* + * Special case clearing the entire ZA space. + * This falls into the CONSTRAINED UNPREDICTABLE zeroing of any + * parts of the ZA storage outside of SVL. + */ + if (imm == 0xff) { + memset(env->zarray, 0, sizeof(env->zarray)); + return; + } + + /* + * Recall that ZAnH.D[m] is spread across ZA[n+8*m], + * so each row is discontiguous within ZA[]. + */ + for (i = 0; i < svl; i++) { + if (imm & (1 << (i % 8))) { + memset(&env->zarray[i], 0, svl); + } + } +} diff --git a/target/arm/translate-sme.c b/target/arm/translate-sme.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sme.c +++ b/target/arm/translate-sme.c @@ -XXX,XX +XXX,XX @@ */ #include "decode-sme.c.inc" + + +static bool trans_ZERO(DisasContext *s, arg_ZERO *a) +{ + if (!dc_isar_feature(aa64_sme, s)) { + return false; + } + if (sme_za_enabled_check(s)) { + gen_helper_sme_zero(cpu_env, tcg_constant_i32(a->imm), + tcg_constant_i32(streaming_vec_reg_size(s))); + } + return true; +} -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> We can reuse the SVE functions for implementing moves to/from horizontal tile slices, but we need new ones for moves to/from vertical tile slices. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-20-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper-sme.h | 12 +++ target/arm/helper-sve.h | 2 + target/arm/translate-a64.h | 8 ++ target/arm/translate.h | 5 ++ target/arm/sme.decode | 15 ++++ target/arm/sme_helper.c | 151 ++++++++++++++++++++++++++++++++++++- target/arm/sve_helper.c | 12 +++ target/arm/translate-sme.c | 127 +++++++++++++++++++++++++++++++ 8 files changed, 331 insertions(+), 1 deletion(-) diff --git a/target/arm/helper-sme.h b/target/arm/helper-sme.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-sme.h +++ b/target/arm/helper-sme.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_2(set_pstate_sm, TCG_CALL_NO_RWG, void, env, i32) DEF_HELPER_FLAGS_2(set_pstate_za, TCG_CALL_NO_RWG, void, env, i32) DEF_HELPER_FLAGS_3(sme_zero, TCG_CALL_NO_RWG, void, env, i32, i32) + +/* Move to/from vertical array slices, i.e. columns, so 'c'. */ +DEF_HELPER_FLAGS_4(sme_mova_cz_b, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(sme_mova_zc_b, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(sme_mova_cz_h, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(sme_mova_zc_h, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(sme_mova_cz_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(sme_mova_zc_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(sme_mova_cz_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(sme_mova_zc_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(sme_mova_cz_q, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(sme_mova_zc_q, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) diff --git a/target/arm/helper-sve.h b/target/arm/helper-sve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-sve.h +++ b/target/arm/helper-sve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_5(sve_sel_zpzz_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_5(sve_sel_zpzz_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_5(sve_sel_zpzz_q, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_5(sve2_addp_zpzz_b, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) diff --git a/target/arm/translate-a64.h b/target/arm/translate-a64.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a64.h +++ b/target/arm/translate-a64.h @@ -XXX,XX +XXX,XX @@ static inline int pred_gvec_reg_size(DisasContext *s) return size_for_gvec(pred_full_reg_size(s)); } +/* Return a newly allocated pointer to the predicate register. */ +static inline TCGv_ptr pred_full_reg_ptr(DisasContext *s, int regno) +{ + TCGv_ptr ret = tcg_temp_new_ptr(); + tcg_gen_addi_ptr(ret, cpu_env, pred_full_reg_offset(s, regno)); + return ret; +} + bool disas_sve(DisasContext *, uint32_t); bool disas_sme(DisasContext *, uint32_t); diff --git a/target/arm/translate.h b/target/arm/translate.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.h +++ b/target/arm/translate.h @@ -XXX,XX +XXX,XX @@ static inline int plus_2(DisasContext *s, int x) return x + 2; } +static inline int plus_12(DisasContext *s, int x) +{ + return x + 12; +} + static inline int times_2(DisasContext *s, int x) { return x * 2; diff --git a/target/arm/sme.decode b/target/arm/sme.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme.decode +++ b/target/arm/sme.decode @@ -XXX,XX +XXX,XX @@ ### SME Misc ZERO 11000000 00 001 00000000000 imm:8 + +### SME Move into/from Array + +%mova_rs 13:2 !function=plus_12 +&mova esz rs pg zr za_imm v:bool to_vec:bool + +MOVA 11000000 esz:2 00000 0 v:1 .. pg:3 zr:5 0 za_imm:4 \ + &mova to_vec=0 rs=%mova_rs +MOVA 11000000 11 00000 1 v:1 .. pg:3 zr:5 0 za_imm:4 \ + &mova to_vec=0 rs=%mova_rs esz=4 + +MOVA 11000000 esz:2 00001 0 v:1 .. pg:3 0 za_imm:4 zr:5 \ + &mova to_vec=1 rs=%mova_rs +MOVA 11000000 11 00001 1 v:1 .. pg:3 0 za_imm:4 zr:5 \ + &mova to_vec=1 rs=%mova_rs esz=4 diff --git a/target/arm/sme_helper.c b/target/arm/sme_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme_helper.c +++ b/target/arm/sme_helper.c @@ -XXX,XX +XXX,XX @@ #include "qemu/osdep.h" #include "cpu.h" -#include "internals.h" +#include "tcg/tcg-gvec-desc.h" #include "exec/helper-proto.h" +#include "qemu/int128.h" +#include "vec_internal.h" /* ResetSVEState */ void arm_reset_sve_state(CPUARMState *env) @@ -XXX,XX +XXX,XX @@ void helper_sme_zero(CPUARMState *env, uint32_t imm, uint32_t svl) } } } + + +/* + * When considering the ZA storage as an array of elements of + * type T, the index within that array of the Nth element of + * a vertical slice of a tile can be calculated like this, + * regardless of the size of type T. This is because the tiles + * are interleaved, so if type T is size N bytes then row 1 of + * the tile is N rows away from row 0. The division by N to + * convert a byte offset into an array index and the multiplication + * by N to convert from vslice-index-within-the-tile to + * the index within the ZA storage cancel out. + */ +#define tile_vslice_index(i) ((i) * sizeof(ARMVectorReg)) + +/* + * When doing byte arithmetic on the ZA storage, the element + * byteoff bytes away in a tile vertical slice is always this + * many bytes away in the ZA storage, regardless of the + * size of the tile element, assuming that byteoff is a multiple + * of the element size. Again this is because of the interleaving + * of the tiles. For instance if we have 1 byte per element then + * each row of the ZA storage has one byte of the vslice data, + * and (counting from 0) byte 8 goes in row 8 of the storage + * at offset (8 * row-size-in-bytes). + * If we have 8 bytes per element then each row of the ZA storage + * has 8 bytes of the data, but there are 8 interleaved tiles and + * so byte 8 of the data goes into row 1 of the tile, + * which is again row 8 of the storage, so the offset is still + * (8 * row-size-in-bytes). Similarly for other element sizes. + */ +#define tile_vslice_offset(byteoff) ((byteoff) * sizeof(ARMVectorReg)) + + +/* + * Move Zreg vector to ZArray column. + */ +#define DO_MOVA_C(NAME, TYPE, H) \ +void HELPER(NAME)(void *za, void *vn, void *vg, uint32_t desc) \ +{ \ + int i, oprsz = simd_oprsz(desc); \ + for (i = 0; i < oprsz; ) { \ + uint16_t pg = *(uint16_t *)(vg + H1_2(i >> 3)); \ + do { \ + if (pg & 1) { \ + *(TYPE *)(za + tile_vslice_offset(i)) = *(TYPE *)(vn + H(i)); \ + } \ + i += sizeof(TYPE); \ + pg >>= sizeof(TYPE); \ + } while (i & 15); \ + } \ +} + +DO_MOVA_C(sme_mova_cz_b, uint8_t, H1) +DO_MOVA_C(sme_mova_cz_h, uint16_t, H1_2) +DO_MOVA_C(sme_mova_cz_s, uint32_t, H1_4) + +void HELPER(sme_mova_cz_d)(void *za, void *vn, void *vg, uint32_t desc) +{ + int i, oprsz = simd_oprsz(desc) / 8; + uint8_t *pg = vg; + uint64_t *n = vn; + uint64_t *a = za; + + for (i = 0; i < oprsz; i++) { + if (pg[H1(i)] & 1) { + a[tile_vslice_index(i)] = n[i]; + } + } +} + +void HELPER(sme_mova_cz_q)(void *za, void *vn, void *vg, uint32_t desc) +{ + int i, oprsz = simd_oprsz(desc) / 16; + uint16_t *pg = vg; + Int128 *n = vn; + Int128 *a = za; + + /* + * Int128 is used here simply to copy 16 bytes, and to simplify + * the address arithmetic. + */ + for (i = 0; i < oprsz; i++) { + if (pg[H2(i)] & 1) { + a[tile_vslice_index(i)] = n[i]; + } + } +} + +#undef DO_MOVA_C + +/* + * Move ZArray column to Zreg vector. + */ +#define DO_MOVA_Z(NAME, TYPE, H) \ +void HELPER(NAME)(void *vd, void *za, void *vg, uint32_t desc) \ +{ \ + int i, oprsz = simd_oprsz(desc); \ + for (i = 0; i < oprsz; ) { \ + uint16_t pg = *(uint16_t *)(vg + H1_2(i >> 3)); \ + do { \ + if (pg & 1) { \ + *(TYPE *)(vd + H(i)) = *(TYPE *)(za + tile_vslice_offset(i)); \ + } \ + i += sizeof(TYPE); \ + pg >>= sizeof(TYPE); \ + } while (i & 15); \ + } \ +} + +DO_MOVA_Z(sme_mova_zc_b, uint8_t, H1) +DO_MOVA_Z(sme_mova_zc_h, uint16_t, H1_2) +DO_MOVA_Z(sme_mova_zc_s, uint32_t, H1_4) + +void HELPER(sme_mova_zc_d)(void *vd, void *za, void *vg, uint32_t desc) +{ + int i, oprsz = simd_oprsz(desc) / 8; + uint8_t *pg = vg; + uint64_t *d = vd; + uint64_t *a = za; + + for (i = 0; i < oprsz; i++) { + if (pg[H1(i)] & 1) { + d[i] = a[tile_vslice_index(i)]; + } + } +} + +void HELPER(sme_mova_zc_q)(void *vd, void *za, void *vg, uint32_t desc) +{ + int i, oprsz = simd_oprsz(desc) / 16; + uint16_t *pg = vg; + Int128 *d = vd; + Int128 *a = za; + + /* + * Int128 is used here simply to copy 16 bytes, and to simplify + * the address arithmetic. + */ + for (i = 0; i < oprsz; i++, za += sizeof(ARMVectorReg)) { + if (pg[H2(i)] & 1) { + d[i] = a[tile_vslice_index(i)]; + } + } +} + +#undef DO_MOVA_Z diff --git a/target/arm/sve_helper.c b/target/arm/sve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sve_helper.c +++ b/target/arm/sve_helper.c @@ -XXX,XX +XXX,XX @@ void HELPER(sve_sel_zpzz_d)(void *vd, void *vn, void *vm, } } +void HELPER(sve_sel_zpzz_q)(void *vd, void *vn, void *vm, + void *vg, uint32_t desc) +{ + intptr_t i, opr_sz = simd_oprsz(desc) / 16; + Int128 *d = vd, *n = vn, *m = vm; + uint16_t *pg = vg; + + for (i = 0; i < opr_sz; i += 1) { + d[i] = (pg[H2(i)] & 1 ? n : m)[i]; + } +} + /* Two operand comparison controlled by a predicate. * ??? It is very tempting to want to be able to expand this inline * with x86 instructions, e.g. diff --git a/target/arm/translate-sme.c b/target/arm/translate-sme.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sme.c +++ b/target/arm/translate-sme.c @@ -XXX,XX +XXX,XX @@ #include "decode-sme.c.inc" +/* + * Resolve tile.size[index] to a host pointer, where tile and index + * are always decoded together, dependent on the element size. + */ +static TCGv_ptr get_tile_rowcol(DisasContext *s, int esz, int rs, + int tile_index, bool vertical) +{ + int tile = tile_index >> (4 - esz); + int index = esz == MO_128 ? 0 : extract32(tile_index, 0, 4 - esz); + int pos, len, offset; + TCGv_i32 tmp; + TCGv_ptr addr; + + /* Compute the final index, which is Rs+imm. */ + tmp = tcg_temp_new_i32(); + tcg_gen_trunc_tl_i32(tmp, cpu_reg(s, rs)); + tcg_gen_addi_i32(tmp, tmp, index); + + /* Prepare a power-of-two modulo via extraction of @len bits. */ + len = ctz32(streaming_vec_reg_size(s)) - esz; + + if (vertical) { + /* + * Compute the byte offset of the index within the tile: + * (index % (svl / size)) * size + * = (index % (svl >> esz)) << esz + * Perform the power-of-two modulo via extraction of the low @len bits. + * Perform the multiply by shifting left by @pos bits. + * Perform these operations simultaneously via deposit into zero. + */ + pos = esz; + tcg_gen_deposit_z_i32(tmp, tmp, pos, len); + + /* + * For big-endian, adjust the indexed column byte offset within + * the uint64_t host words that make up env->zarray[]. + */ + if (HOST_BIG_ENDIAN && esz < MO_64) { + tcg_gen_xori_i32(tmp, tmp, 8 - (1 << esz)); + } + } else { + /* + * Compute the byte offset of the index within the tile: + * (index % (svl / size)) * (size * sizeof(row)) + * = (index % (svl >> esz)) << (esz + log2(sizeof(row))) + */ + pos = esz + ctz32(sizeof(ARMVectorReg)); + tcg_gen_deposit_z_i32(tmp, tmp, pos, len); + + /* Row slices are always aligned and need no endian adjustment. */ + } + + /* The tile byte offset within env->zarray is the row. */ + offset = tile * sizeof(ARMVectorReg); + + /* Include the byte offset of zarray to make this relative to env. */ + offset += offsetof(CPUARMState, zarray); + tcg_gen_addi_i32(tmp, tmp, offset); + + /* Add the byte offset to env to produce the final pointer. */ + addr = tcg_temp_new_ptr(); + tcg_gen_ext_i32_ptr(addr, tmp); + tcg_temp_free_i32(tmp); + tcg_gen_add_ptr(addr, addr, cpu_env); + + return addr; +} + static bool trans_ZERO(DisasContext *s, arg_ZERO *a) { if (!dc_isar_feature(aa64_sme, s)) { @@ -XXX,XX +XXX,XX @@ static bool trans_ZERO(DisasContext *s, arg_ZERO *a) } return true; } + +static bool trans_MOVA(DisasContext *s, arg_MOVA *a) +{ + static gen_helper_gvec_4 * const h_fns[5] = { + gen_helper_sve_sel_zpzz_b, gen_helper_sve_sel_zpzz_h, + gen_helper_sve_sel_zpzz_s, gen_helper_sve_sel_zpzz_d, + gen_helper_sve_sel_zpzz_q + }; + static gen_helper_gvec_3 * const cz_fns[5] = { + gen_helper_sme_mova_cz_b, gen_helper_sme_mova_cz_h, + gen_helper_sme_mova_cz_s, gen_helper_sme_mova_cz_d, + gen_helper_sme_mova_cz_q, + }; + static gen_helper_gvec_3 * const zc_fns[5] = { + gen_helper_sme_mova_zc_b, gen_helper_sme_mova_zc_h, + gen_helper_sme_mova_zc_s, gen_helper_sme_mova_zc_d, + gen_helper_sme_mova_zc_q, + }; + + TCGv_ptr t_za, t_zr, t_pg; + TCGv_i32 t_desc; + int svl; + + if (!dc_isar_feature(aa64_sme, s)) { + return false; + } + if (!sme_smza_enabled_check(s)) { + return true; + } + + t_za = get_tile_rowcol(s, a->esz, a->rs, a->za_imm, a->v); + t_zr = vec_full_reg_ptr(s, a->zr); + t_pg = pred_full_reg_ptr(s, a->pg); + + svl = streaming_vec_reg_size(s); + t_desc = tcg_constant_i32(simd_desc(svl, svl, 0)); + + if (a->v) { + /* Vertical slice -- use sme mova helpers. */ + if (a->to_vec) { + zc_fns[a->esz](t_zr, t_za, t_pg, t_desc); + } else { + cz_fns[a->esz](t_za, t_zr, t_pg, t_desc); + } + } else { + /* Horizontal slice -- reuse sve sel helpers. */ + if (a->to_vec) { + h_fns[a->esz](t_zr, t_za, t_zr, t_pg, t_desc); + } else { + h_fns[a->esz](t_za, t_zr, t_za, t_pg, t_desc); + } + } + + tcg_temp_free_ptr(t_za); + tcg_temp_free_ptr(t_zr); + tcg_temp_free_ptr(t_pg); + + return true; +} -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> We cannot reuse the SVE functions for LD[1-4] and ST[1-4], because those functions accept only a Zreg register number. For SME, we want to pass a pointer into ZA storage. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-21-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper-sme.h | 82 +++++ target/arm/sme.decode | 9 + target/arm/sme_helper.c | 595 +++++++++++++++++++++++++++++++++++++ target/arm/translate-sme.c | 70 +++++ 4 files changed, 756 insertions(+) diff --git a/target/arm/helper-sme.h b/target/arm/helper-sme.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-sme.h +++ b/target/arm/helper-sme.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(sme_mova_cz_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_4(sme_mova_zc_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_4(sme_mova_cz_q, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_4(sme_mova_zc_q, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) + +DEF_HELPER_FLAGS_5(sme_ld1b_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1b_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1b_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1b_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) + +DEF_HELPER_FLAGS_5(sme_ld1h_be_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1h_le_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1h_be_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1h_le_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1h_be_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1h_le_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1h_be_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1h_le_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) + +DEF_HELPER_FLAGS_5(sme_ld1s_be_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1s_le_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1s_be_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1s_le_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1s_be_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1s_le_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1s_be_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1s_le_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) + +DEF_HELPER_FLAGS_5(sme_ld1d_be_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1d_le_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1d_be_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1d_le_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1d_be_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1d_le_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1d_be_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1d_le_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) + +DEF_HELPER_FLAGS_5(sme_ld1q_be_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1q_le_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1q_be_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1q_le_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1q_be_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1q_le_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1q_be_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1q_le_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) + +DEF_HELPER_FLAGS_5(sme_st1b_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1b_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1b_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1b_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) + +DEF_HELPER_FLAGS_5(sme_st1h_be_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1h_le_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1h_be_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1h_le_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1h_be_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1h_le_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1h_be_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1h_le_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) + +DEF_HELPER_FLAGS_5(sme_st1s_be_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1s_le_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1s_be_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1s_le_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1s_be_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1s_le_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1s_be_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1s_le_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) + +DEF_HELPER_FLAGS_5(sme_st1d_be_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1d_le_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1d_be_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1d_le_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1d_be_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1d_le_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1d_be_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1d_le_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) + +DEF_HELPER_FLAGS_5(sme_st1q_be_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1q_le_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1q_be_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1q_le_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1q_be_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1q_le_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1q_be_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1q_le_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) diff --git a/target/arm/sme.decode b/target/arm/sme.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme.decode +++ b/target/arm/sme.decode @@ -XXX,XX +XXX,XX @@ MOVA 11000000 esz:2 00001 0 v:1 .. pg:3 0 za_imm:4 zr:5 \ &mova to_vec=1 rs=%mova_rs MOVA 11000000 11 00001 1 v:1 .. pg:3 0 za_imm:4 zr:5 \ &mova to_vec=1 rs=%mova_rs esz=4 + +### SME Memory + +&ldst esz rs pg rn rm za_imm v:bool st:bool + +LDST1 1110000 0 esz:2 st:1 rm:5 v:1 .. pg:3 rn:5 0 za_imm:4 \ + &ldst rs=%mova_rs +LDST1 1110000 111 st:1 rm:5 v:1 .. pg:3 rn:5 0 za_imm:4 \ + &ldst esz=4 rs=%mova_rs diff --git a/target/arm/sme_helper.c b/target/arm/sme_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme_helper.c +++ b/target/arm/sme_helper.c @@ -XXX,XX +XXX,XX @@ #include "qemu/osdep.h" #include "cpu.h" +#include "internals.h" #include "tcg/tcg-gvec-desc.h" #include "exec/helper-proto.h" +#include "exec/cpu_ldst.h" +#include "exec/exec-all.h" #include "qemu/int128.h" #include "vec_internal.h" +#include "sve_ldst_internal.h" /* ResetSVEState */ void arm_reset_sve_state(CPUARMState *env) @@ -XXX,XX +XXX,XX @@ void HELPER(sme_mova_zc_q)(void *vd, void *za, void *vg, uint32_t desc) } #undef DO_MOVA_Z + +/* + * Clear elements in a tile slice comprising len bytes. + */ + +typedef void ClearFn(void *ptr, size_t off, size_t len); + +static void clear_horizontal(void *ptr, size_t off, size_t len) +{ + memset(ptr + off, 0, len); +} + +static void clear_vertical_b(void *vptr, size_t off, size_t len) +{ + for (size_t i = 0; i < len; ++i) { + *(uint8_t *)(vptr + tile_vslice_offset(i + off)) = 0; + } +} + +static void clear_vertical_h(void *vptr, size_t off, size_t len) +{ + for (size_t i = 0; i < len; i += 2) { + *(uint16_t *)(vptr + tile_vslice_offset(i + off)) = 0; + } +} + +static void clear_vertical_s(void *vptr, size_t off, size_t len) +{ + for (size_t i = 0; i < len; i += 4) { + *(uint32_t *)(vptr + tile_vslice_offset(i + off)) = 0; + } +} + +static void clear_vertical_d(void *vptr, size_t off, size_t len) +{ + for (size_t i = 0; i < len; i += 8) { + *(uint64_t *)(vptr + tile_vslice_offset(i + off)) = 0; + } +} + +static void clear_vertical_q(void *vptr, size_t off, size_t len) +{ + for (size_t i = 0; i < len; i += 16) { + memset(vptr + tile_vslice_offset(i + off), 0, 16); + } +} + +/* + * Copy elements from an array into a tile slice comprising len bytes. + */ + +typedef void CopyFn(void *dst, const void *src, size_t len); + +static void copy_horizontal(void *dst, const void *src, size_t len) +{ + memcpy(dst, src, len); +} + +static void copy_vertical_b(void *vdst, const void *vsrc, size_t len) +{ + const uint8_t *src = vsrc; + uint8_t *dst = vdst; + size_t i; + + for (i = 0; i < len; ++i) { + dst[tile_vslice_index(i)] = src[i]; + } +} + +static void copy_vertical_h(void *vdst, const void *vsrc, size_t len) +{ + const uint16_t *src = vsrc; + uint16_t *dst = vdst; + size_t i; + + for (i = 0; i < len / 2; ++i) { + dst[tile_vslice_index(i)] = src[i]; + } +} + +static void copy_vertical_s(void *vdst, const void *vsrc, size_t len) +{ + const uint32_t *src = vsrc; + uint32_t *dst = vdst; + size_t i; + + for (i = 0; i < len / 4; ++i) { + dst[tile_vslice_index(i)] = src[i]; + } +} + +static void copy_vertical_d(void *vdst, const void *vsrc, size_t len) +{ + const uint64_t *src = vsrc; + uint64_t *dst = vdst; + size_t i; + + for (i = 0; i < len / 8; ++i) { + dst[tile_vslice_index(i)] = src[i]; + } +} + +static void copy_vertical_q(void *vdst, const void *vsrc, size_t len) +{ + for (size_t i = 0; i < len; i += 16) { + memcpy(vdst + tile_vslice_offset(i), vsrc + i, 16); + } +} + +/* + * Host and TLB primitives for vertical tile slice addressing. + */ + +#define DO_LD(NAME, TYPE, HOST, TLB) \ +static inline void sme_##NAME##_v_host(void *za, intptr_t off, void *host) \ +{ \ + TYPE val = HOST(host); \ + *(TYPE *)(za + tile_vslice_offset(off)) = val; \ +} \ +static inline void sme_##NAME##_v_tlb(CPUARMState *env, void *za, \ + intptr_t off, target_ulong addr, uintptr_t ra) \ +{ \ + TYPE val = TLB(env, useronly_clean_ptr(addr), ra); \ + *(TYPE *)(za + tile_vslice_offset(off)) = val; \ +} + +#define DO_ST(NAME, TYPE, HOST, TLB) \ +static inline void sme_##NAME##_v_host(void *za, intptr_t off, void *host) \ +{ \ + TYPE val = *(TYPE *)(za + tile_vslice_offset(off)); \ + HOST(host, val); \ +} \ +static inline void sme_##NAME##_v_tlb(CPUARMState *env, void *za, \ + intptr_t off, target_ulong addr, uintptr_t ra) \ +{ \ + TYPE val = *(TYPE *)(za + tile_vslice_offset(off)); \ + TLB(env, useronly_clean_ptr(addr), val, ra); \ +} + +/* + * The ARMVectorReg elements are stored in host-endian 64-bit units. + * For 128-bit quantities, the sequence defined by the Elem[] pseudocode + * corresponds to storing the two 64-bit pieces in little-endian order. + */ +#define DO_LDQ(HNAME, VNAME, BE, HOST, TLB) \ +static inline void HNAME##_host(void *za, intptr_t off, void *host) \ +{ \ + uint64_t val0 = HOST(host), val1 = HOST(host + 8); \ + uint64_t *ptr = za + off; \ + ptr[0] = BE ? val1 : val0, ptr[1] = BE ? val0 : val1; \ +} \ +static inline void VNAME##_v_host(void *za, intptr_t off, void *host) \ +{ \ + HNAME##_host(za, tile_vslice_offset(off), host); \ +} \ +static inline void HNAME##_tlb(CPUARMState *env, void *za, intptr_t off, \ + target_ulong addr, uintptr_t ra) \ +{ \ + uint64_t val0 = TLB(env, useronly_clean_ptr(addr), ra); \ + uint64_t val1 = TLB(env, useronly_clean_ptr(addr + 8), ra); \ + uint64_t *ptr = za + off; \ + ptr[0] = BE ? val1 : val0, ptr[1] = BE ? val0 : val1; \ +} \ +static inline void VNAME##_v_tlb(CPUARMState *env, void *za, intptr_t off, \ + target_ulong addr, uintptr_t ra) \ +{ \ + HNAME##_tlb(env, za, tile_vslice_offset(off), addr, ra); \ +} + +#define DO_STQ(HNAME, VNAME, BE, HOST, TLB) \ +static inline void HNAME##_host(void *za, intptr_t off, void *host) \ +{ \ + uint64_t *ptr = za + off; \ + HOST(host, ptr[BE]); \ + HOST(host + 1, ptr[!BE]); \ +} \ +static inline void VNAME##_v_host(void *za, intptr_t off, void *host) \ +{ \ + HNAME##_host(za, tile_vslice_offset(off), host); \ +} \ +static inline void HNAME##_tlb(CPUARMState *env, void *za, intptr_t off, \ + target_ulong addr, uintptr_t ra) \ +{ \ + uint64_t *ptr = za + off; \ + TLB(env, useronly_clean_ptr(addr), ptr[BE], ra); \ + TLB(env, useronly_clean_ptr(addr + 8), ptr[!BE], ra); \ +} \ +static inline void VNAME##_v_tlb(CPUARMState *env, void *za, intptr_t off, \ + target_ulong addr, uintptr_t ra) \ +{ \ + HNAME##_tlb(env, za, tile_vslice_offset(off), addr, ra); \ +} + +DO_LD(ld1b, uint8_t, ldub_p, cpu_ldub_data_ra) +DO_LD(ld1h_be, uint16_t, lduw_be_p, cpu_lduw_be_data_ra) +DO_LD(ld1h_le, uint16_t, lduw_le_p, cpu_lduw_le_data_ra) +DO_LD(ld1s_be, uint32_t, ldl_be_p, cpu_ldl_be_data_ra) +DO_LD(ld1s_le, uint32_t, ldl_le_p, cpu_ldl_le_data_ra) +DO_LD(ld1d_be, uint64_t, ldq_be_p, cpu_ldq_be_data_ra) +DO_LD(ld1d_le, uint64_t, ldq_le_p, cpu_ldq_le_data_ra) + +DO_LDQ(sve_ld1qq_be, sme_ld1q_be, 1, ldq_be_p, cpu_ldq_be_data_ra) +DO_LDQ(sve_ld1qq_le, sme_ld1q_le, 0, ldq_le_p, cpu_ldq_le_data_ra) + +DO_ST(st1b, uint8_t, stb_p, cpu_stb_data_ra) +DO_ST(st1h_be, uint16_t, stw_be_p, cpu_stw_be_data_ra) +DO_ST(st1h_le, uint16_t, stw_le_p, cpu_stw_le_data_ra) +DO_ST(st1s_be, uint32_t, stl_be_p, cpu_stl_be_data_ra) +DO_ST(st1s_le, uint32_t, stl_le_p, cpu_stl_le_data_ra) +DO_ST(st1d_be, uint64_t, stq_be_p, cpu_stq_be_data_ra) +DO_ST(st1d_le, uint64_t, stq_le_p, cpu_stq_le_data_ra) + +DO_STQ(sve_st1qq_be, sme_st1q_be, 1, stq_be_p, cpu_stq_be_data_ra) +DO_STQ(sve_st1qq_le, sme_st1q_le, 0, stq_le_p, cpu_stq_le_data_ra) + +#undef DO_LD +#undef DO_ST +#undef DO_LDQ +#undef DO_STQ + +/* + * Common helper for all contiguous predicated loads. + */ + +static inline QEMU_ALWAYS_INLINE +void sme_ld1(CPUARMState *env, void *za, uint64_t *vg, + const target_ulong addr, uint32_t desc, const uintptr_t ra, + const int esz, uint32_t mtedesc, bool vertical, + sve_ldst1_host_fn *host_fn, + sve_ldst1_tlb_fn *tlb_fn, + ClearFn *clr_fn, + CopyFn *cpy_fn) +{ + const intptr_t reg_max = simd_oprsz(desc); + const intptr_t esize = 1 << esz; + intptr_t reg_off, reg_last; + SVEContLdSt info; + void *host; + int flags; + + /* Find the active elements. */ + if (!sve_cont_ldst_elements(&info, addr, vg, reg_max, esz, esize)) { + /* The entire predicate was false; no load occurs. */ + clr_fn(za, 0, reg_max); + return; + } + + /* Probe the page(s). Exit with exception for any invalid page. */ + sve_cont_ldst_pages(&info, FAULT_ALL, env, addr, MMU_DATA_LOAD, ra); + + /* Handle watchpoints for all active elements. */ + sve_cont_ldst_watchpoints(&info, env, vg, addr, esize, esize, + BP_MEM_READ, ra); + + /* + * Handle mte checks for all active elements. + * Since TBI must be set for MTE, !mtedesc => !mte_active. + */ + if (mtedesc) { + sve_cont_ldst_mte_check(&info, env, vg, addr, esize, esize, + mtedesc, ra); + } + + flags = info.page[0].flags | info.page[1].flags; + if (unlikely(flags != 0)) { +#ifdef CONFIG_USER_ONLY + g_assert_not_reached(); +#else + /* + * At least one page includes MMIO. + * Any bus operation can fail with cpu_transaction_failed, + * which for ARM will raise SyncExternal. Perform the load + * into scratch memory to preserve register state until the end. + */ + ARMVectorReg scratch = { }; + + reg_off = info.reg_off_first[0]; + reg_last = info.reg_off_last[1]; + if (reg_last < 0) { + reg_last = info.reg_off_split; + if (reg_last < 0) { + reg_last = info.reg_off_last[0]; + } + } + + do { + uint64_t pg = vg[reg_off >> 6]; + do { + if ((pg >> (reg_off & 63)) & 1) { + tlb_fn(env, &scratch, reg_off, addr + reg_off, ra); + } + reg_off += esize; + } while (reg_off & 63); + } while (reg_off <= reg_last); + + cpy_fn(za, &scratch, reg_max); + return; +#endif + } + + /* The entire operation is in RAM, on valid pages. */ + + reg_off = info.reg_off_first[0]; + reg_last = info.reg_off_last[0]; + host = info.page[0].host; + + if (!vertical) { + memset(za, 0, reg_max); + } else if (reg_off) { + clr_fn(za, 0, reg_off); + } + + while (reg_off <= reg_last) { + uint64_t pg = vg[reg_off >> 6]; + do { + if ((pg >> (reg_off & 63)) & 1) { + host_fn(za, reg_off, host + reg_off); + } else if (vertical) { + clr_fn(za, reg_off, esize); + } + reg_off += esize; + } while (reg_off <= reg_last && (reg_off & 63)); + } + + /* + * Use the slow path to manage the cross-page misalignment. + * But we know this is RAM and cannot trap. + */ + reg_off = info.reg_off_split; + if (unlikely(reg_off >= 0)) { + tlb_fn(env, za, reg_off, addr + reg_off, ra); + } + + reg_off = info.reg_off_first[1]; + if (unlikely(reg_off >= 0)) { + reg_last = info.reg_off_last[1]; + host = info.page[1].host; + + do { + uint64_t pg = vg[reg_off >> 6]; + do { + if ((pg >> (reg_off & 63)) & 1) { + host_fn(za, reg_off, host + reg_off); + } else if (vertical) { + clr_fn(za, reg_off, esize); + } + reg_off += esize; + } while (reg_off & 63); + } while (reg_off <= reg_last); + } +} + +static inline QEMU_ALWAYS_INLINE +void sme_ld1_mte(CPUARMState *env, void *za, uint64_t *vg, + target_ulong addr, uint32_t desc, uintptr_t ra, + const int esz, bool vertical, + sve_ldst1_host_fn *host_fn, + sve_ldst1_tlb_fn *tlb_fn, + ClearFn *clr_fn, + CopyFn *cpy_fn) +{ + uint32_t mtedesc = desc >> (SIMD_DATA_SHIFT + SVE_MTEDESC_SHIFT); + int bit55 = extract64(addr, 55, 1); + + /* Remove mtedesc from the normal sve descriptor. */ + desc = extract32(desc, 0, SIMD_DATA_SHIFT + SVE_MTEDESC_SHIFT); + + /* Perform gross MTE suppression early. */ + if (!tbi_check(desc, bit55) || + tcma_check(desc, bit55, allocation_tag_from_addr(addr))) { + mtedesc = 0; + } + + sme_ld1(env, za, vg, addr, desc, ra, esz, mtedesc, vertical, + host_fn, tlb_fn, clr_fn, cpy_fn); +} + +#define DO_LD(L, END, ESZ) \ +void HELPER(sme_ld1##L##END##_h)(CPUARMState *env, void *za, void *vg, \ + target_ulong addr, uint32_t desc) \ +{ \ + sme_ld1(env, za, vg, addr, desc, GETPC(), ESZ, 0, false, \ + sve_ld1##L##L##END##_host, sve_ld1##L##L##END##_tlb, \ + clear_horizontal, copy_horizontal); \ +} \ +void HELPER(sme_ld1##L##END##_v)(CPUARMState *env, void *za, void *vg, \ + target_ulong addr, uint32_t desc) \ +{ \ + sme_ld1(env, za, vg, addr, desc, GETPC(), ESZ, 0, true, \ + sme_ld1##L##END##_v_host, sme_ld1##L##END##_v_tlb, \ + clear_vertical_##L, copy_vertical_##L); \ +} \ +void HELPER(sme_ld1##L##END##_h_mte)(CPUARMState *env, void *za, void *vg, \ + target_ulong addr, uint32_t desc) \ +{ \ + sme_ld1_mte(env, za, vg, addr, desc, GETPC(), ESZ, false, \ + sve_ld1##L##L##END##_host, sve_ld1##L##L##END##_tlb, \ + clear_horizontal, copy_horizontal); \ +} \ +void HELPER(sme_ld1##L##END##_v_mte)(CPUARMState *env, void *za, void *vg, \ + target_ulong addr, uint32_t desc) \ +{ \ + sme_ld1_mte(env, za, vg, addr, desc, GETPC(), ESZ, true, \ + sme_ld1##L##END##_v_host, sme_ld1##L##END##_v_tlb, \ + clear_vertical_##L, copy_vertical_##L); \ +} + +DO_LD(b, , MO_8) +DO_LD(h, _be, MO_16) +DO_LD(h, _le, MO_16) +DO_LD(s, _be, MO_32) +DO_LD(s, _le, MO_32) +DO_LD(d, _be, MO_64) +DO_LD(d, _le, MO_64) +DO_LD(q, _be, MO_128) +DO_LD(q, _le, MO_128) + +#undef DO_LD + +/* + * Common helper for all contiguous predicated stores. + */ + +static inline QEMU_ALWAYS_INLINE +void sme_st1(CPUARMState *env, void *za, uint64_t *vg, + const target_ulong addr, uint32_t desc, const uintptr_t ra, + const int esz, uint32_t mtedesc, bool vertical, + sve_ldst1_host_fn *host_fn, + sve_ldst1_tlb_fn *tlb_fn) +{ + const intptr_t reg_max = simd_oprsz(desc); + const intptr_t esize = 1 << esz; + intptr_t reg_off, reg_last; + SVEContLdSt info; + void *host; + int flags; + + /* Find the active elements. */ + if (!sve_cont_ldst_elements(&info, addr, vg, reg_max, esz, esize)) { + /* The entire predicate was false; no store occurs. */ + return; + } + + /* Probe the page(s). Exit with exception for any invalid page. */ + sve_cont_ldst_pages(&info, FAULT_ALL, env, addr, MMU_DATA_STORE, ra); + + /* Handle watchpoints for all active elements. */ + sve_cont_ldst_watchpoints(&info, env, vg, addr, esize, esize, + BP_MEM_WRITE, ra); + + /* + * Handle mte checks for all active elements. + * Since TBI must be set for MTE, !mtedesc => !mte_active. + */ + if (mtedesc) { + sve_cont_ldst_mte_check(&info, env, vg, addr, esize, esize, + mtedesc, ra); + } + + flags = info.page[0].flags | info.page[1].flags; + if (unlikely(flags != 0)) { +#ifdef CONFIG_USER_ONLY + g_assert_not_reached(); +#else + /* + * At least one page includes MMIO. + * Any bus operation can fail with cpu_transaction_failed, + * which for ARM will raise SyncExternal. We cannot avoid + * this fault and will leave with the store incomplete. + */ + reg_off = info.reg_off_first[0]; + reg_last = info.reg_off_last[1]; + if (reg_last < 0) { + reg_last = info.reg_off_split; + if (reg_last < 0) { + reg_last = info.reg_off_last[0]; + } + } + + do { + uint64_t pg = vg[reg_off >> 6]; + do { + if ((pg >> (reg_off & 63)) & 1) { + tlb_fn(env, za, reg_off, addr + reg_off, ra); + } + reg_off += esize; + } while (reg_off & 63); + } while (reg_off <= reg_last); + return; +#endif + } + + reg_off = info.reg_off_first[0]; + reg_last = info.reg_off_last[0]; + host = info.page[0].host; + + while (reg_off <= reg_last) { + uint64_t pg = vg[reg_off >> 6]; + do { + if ((pg >> (reg_off & 63)) & 1) { + host_fn(za, reg_off, host + reg_off); + } + reg_off += 1 << esz; + } while (reg_off <= reg_last && (reg_off & 63)); + } + + /* + * Use the slow path to manage the cross-page misalignment. + * But we know this is RAM and cannot trap. + */ + reg_off = info.reg_off_split; + if (unlikely(reg_off >= 0)) { + tlb_fn(env, za, reg_off, addr + reg_off, ra); + } + + reg_off = info.reg_off_first[1]; + if (unlikely(reg_off >= 0)) { + reg_last = info.reg_off_last[1]; + host = info.page[1].host; + + do { + uint64_t pg = vg[reg_off >> 6]; + do { + if ((pg >> (reg_off & 63)) & 1) { + host_fn(za, reg_off, host + reg_off); + } + reg_off += 1 << esz; + } while (reg_off & 63); + } while (reg_off <= reg_last); + } +} + +static inline QEMU_ALWAYS_INLINE +void sme_st1_mte(CPUARMState *env, void *za, uint64_t *vg, target_ulong addr, + uint32_t desc, uintptr_t ra, int esz, bool vertical, + sve_ldst1_host_fn *host_fn, + sve_ldst1_tlb_fn *tlb_fn) +{ + uint32_t mtedesc = desc >> (SIMD_DATA_SHIFT + SVE_MTEDESC_SHIFT); + int bit55 = extract64(addr, 55, 1); + + /* Remove mtedesc from the normal sve descriptor. */ + desc = extract32(desc, 0, SIMD_DATA_SHIFT + SVE_MTEDESC_SHIFT); + + /* Perform gross MTE suppression early. */ + if (!tbi_check(desc, bit55) || + tcma_check(desc, bit55, allocation_tag_from_addr(addr))) { + mtedesc = 0; + } + + sme_st1(env, za, vg, addr, desc, ra, esz, mtedesc, + vertical, host_fn, tlb_fn); +} + +#define DO_ST(L, END, ESZ) \ +void HELPER(sme_st1##L##END##_h)(CPUARMState *env, void *za, void *vg, \ + target_ulong addr, uint32_t desc) \ +{ \ + sme_st1(env, za, vg, addr, desc, GETPC(), ESZ, 0, false, \ + sve_st1##L##L##END##_host, sve_st1##L##L##END##_tlb); \ +} \ +void HELPER(sme_st1##L##END##_v)(CPUARMState *env, void *za, void *vg, \ + target_ulong addr, uint32_t desc) \ +{ \ + sme_st1(env, za, vg, addr, desc, GETPC(), ESZ, 0, true, \ + sme_st1##L##END##_v_host, sme_st1##L##END##_v_tlb); \ +} \ +void HELPER(sme_st1##L##END##_h_mte)(CPUARMState *env, void *za, void *vg, \ + target_ulong addr, uint32_t desc) \ +{ \ + sme_st1_mte(env, za, vg, addr, desc, GETPC(), ESZ, false, \ + sve_st1##L##L##END##_host, sve_st1##L##L##END##_tlb); \ +} \ +void HELPER(sme_st1##L##END##_v_mte)(CPUARMState *env, void *za, void *vg, \ + target_ulong addr, uint32_t desc) \ +{ \ + sme_st1_mte(env, za, vg, addr, desc, GETPC(), ESZ, true, \ + sme_st1##L##END##_v_host, sme_st1##L##END##_v_tlb); \ +} + +DO_ST(b, , MO_8) +DO_ST(h, _be, MO_16) +DO_ST(h, _le, MO_16) +DO_ST(s, _be, MO_32) +DO_ST(s, _le, MO_32) +DO_ST(d, _be, MO_64) +DO_ST(d, _le, MO_64) +DO_ST(q, _be, MO_128) +DO_ST(q, _le, MO_128) + +#undef DO_ST diff --git a/target/arm/translate-sme.c b/target/arm/translate-sme.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sme.c +++ b/target/arm/translate-sme.c @@ -XXX,XX +XXX,XX @@ static bool trans_MOVA(DisasContext *s, arg_MOVA *a) return true; } + +static bool trans_LDST1(DisasContext *s, arg_LDST1 *a) +{ + typedef void GenLdSt1(TCGv_env, TCGv_ptr, TCGv_ptr, TCGv, TCGv_i32); + + /* + * Indexed by [esz][be][v][mte][st], which is (except for load/store) + * also the order in which the elements appear in the function names, + * and so how we must concatenate the pieces. + */ + +#define FN_LS(F) { gen_helper_sme_ld1##F, gen_helper_sme_st1##F } +#define FN_MTE(F) { FN_LS(F), FN_LS(F##_mte) } +#define FN_HV(F) { FN_MTE(F##_h), FN_MTE(F##_v) } +#define FN_END(L, B) { FN_HV(L), FN_HV(B) } + + static GenLdSt1 * const fns[5][2][2][2][2] = { + FN_END(b, b), + FN_END(h_le, h_be), + FN_END(s_le, s_be), + FN_END(d_le, d_be), + FN_END(q_le, q_be), + }; + +#undef FN_LS +#undef FN_MTE +#undef FN_HV +#undef FN_END + + TCGv_ptr t_za, t_pg; + TCGv_i64 addr; + int svl, desc = 0; + bool be = s->be_data == MO_BE; + bool mte = s->mte_active[0]; + + if (!dc_isar_feature(aa64_sme, s)) { + return false; + } + if (!sme_smza_enabled_check(s)) { + return true; + } + + t_za = get_tile_rowcol(s, a->esz, a->rs, a->za_imm, a->v); + t_pg = pred_full_reg_ptr(s, a->pg); + addr = tcg_temp_new_i64(); + + tcg_gen_shli_i64(addr, cpu_reg(s, a->rm), a->esz); + tcg_gen_add_i64(addr, addr, cpu_reg_sp(s, a->rn)); + + if (mte) { + desc = FIELD_DP32(desc, MTEDESC, MIDX, get_mem_index(s)); + desc = FIELD_DP32(desc, MTEDESC, TBI, s->tbid); + desc = FIELD_DP32(desc, MTEDESC, TCMA, s->tcma); + desc = FIELD_DP32(desc, MTEDESC, WRITE, a->st); + desc = FIELD_DP32(desc, MTEDESC, SIZEM1, (1 << a->esz) - 1); + desc <<= SVE_MTEDESC_SHIFT; + } else { + addr = clean_data_tbi(s, addr); + } + svl = streaming_vec_reg_size(s); + desc = simd_desc(svl, svl, desc); + + fns[a->esz][be][a->v][mte][a->st](cpu_env, t_za, t_pg, addr, + tcg_constant_i32(desc)); + + tcg_temp_free_ptr(t_za); + tcg_temp_free_ptr(t_pg); + tcg_temp_free_i64(addr); + return true; +} -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Add a TCGv_ptr base argument, which will be cpu_env for SVE. We will reuse this for SME save and restore array insns. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-22-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/translate-a64.h | 3 +++ target/arm/translate-sve.c | 48 ++++++++++++++++++++++++++++---------- 2 files changed, 39 insertions(+), 12 deletions(-) diff --git a/target/arm/translate-a64.h b/target/arm/translate-a64.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a64.h +++ b/target/arm/translate-a64.h @@ -XXX,XX +XXX,XX @@ void gen_gvec_xar(unsigned vece, uint32_t rd_ofs, uint32_t rn_ofs, uint32_t rm_ofs, int64_t shift, uint32_t opr_sz, uint32_t max_sz); +void gen_sve_ldr(DisasContext *s, TCGv_ptr, int vofs, int len, int rn, int imm); +void gen_sve_str(DisasContext *s, TCGv_ptr, int vofs, int len, int rn, int imm); + #endif /* TARGET_ARM_TRANSLATE_A64_H */ diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ TRANS_FEAT(UCVTF_dd, aa64_sve, gen_gvec_fpst_arg_zpz, * The load should begin at the address Rn + IMM. */ -static void do_ldr(DisasContext *s, uint32_t vofs, int len, int rn, int imm) +void gen_sve_ldr(DisasContext *s, TCGv_ptr base, int vofs, + int len, int rn, int imm) { int len_align = QEMU_ALIGN_DOWN(len, 8); int len_remain = len % 8; @@ -XXX,XX +XXX,XX @@ static void do_ldr(DisasContext *s, uint32_t vofs, int len, int rn, int imm) t0 = tcg_temp_new_i64(); for (i = 0; i < len_align; i += 8) { tcg_gen_qemu_ld_i64(t0, clean_addr, midx, MO_LEUQ); - tcg_gen_st_i64(t0, cpu_env, vofs + i); + tcg_gen_st_i64(t0, base, vofs + i); tcg_gen_addi_i64(clean_addr, clean_addr, 8); } tcg_temp_free_i64(t0); @@ -XXX,XX +XXX,XX @@ static void do_ldr(DisasContext *s, uint32_t vofs, int len, int rn, int imm) clean_addr = new_tmp_a64_local(s); tcg_gen_mov_i64(clean_addr, t0); + if (base != cpu_env) { + TCGv_ptr b = tcg_temp_local_new_ptr(); + tcg_gen_mov_ptr(b, base); + base = b; + } + gen_set_label(loop); t0 = tcg_temp_new_i64(); @@ -XXX,XX +XXX,XX @@ static void do_ldr(DisasContext *s, uint32_t vofs, int len, int rn, int imm) tcg_gen_addi_i64(clean_addr, clean_addr, 8); tp = tcg_temp_new_ptr(); - tcg_gen_add_ptr(tp, cpu_env, i); + tcg_gen_add_ptr(tp, base, i); tcg_gen_addi_ptr(i, i, 8); tcg_gen_st_i64(t0, tp, vofs); tcg_temp_free_ptr(tp); @@ -XXX,XX +XXX,XX @@ static void do_ldr(DisasContext *s, uint32_t vofs, int len, int rn, int imm) tcg_gen_brcondi_ptr(TCG_COND_LTU, i, len_align, loop); tcg_temp_free_ptr(i); + + if (base != cpu_env) { + tcg_temp_free_ptr(base); + assert(len_remain == 0); + } } /* @@ -XXX,XX +XXX,XX @@ static void do_ldr(DisasContext *s, uint32_t vofs, int len, int rn, int imm) default: g_assert_not_reached(); } - tcg_gen_st_i64(t0, cpu_env, vofs + len_align); + tcg_gen_st_i64(t0, base, vofs + len_align); tcg_temp_free_i64(t0); } } /* Similarly for stores. */ -static void do_str(DisasContext *s, uint32_t vofs, int len, int rn, int imm) +void gen_sve_str(DisasContext *s, TCGv_ptr base, int vofs, + int len, int rn, int imm) { int len_align = QEMU_ALIGN_DOWN(len, 8); int len_remain = len % 8; @@ -XXX,XX +XXX,XX @@ static void do_str(DisasContext *s, uint32_t vofs, int len, int rn, int imm) t0 = tcg_temp_new_i64(); for (i = 0; i < len_align; i += 8) { - tcg_gen_ld_i64(t0, cpu_env, vofs + i); + tcg_gen_ld_i64(t0, base, vofs + i); tcg_gen_qemu_st_i64(t0, clean_addr, midx, MO_LEUQ); tcg_gen_addi_i64(clean_addr, clean_addr, 8); } @@ -XXX,XX +XXX,XX @@ static void do_str(DisasContext *s, uint32_t vofs, int len, int rn, int imm) clean_addr = new_tmp_a64_local(s); tcg_gen_mov_i64(clean_addr, t0); + if (base != cpu_env) { + TCGv_ptr b = tcg_temp_local_new_ptr(); + tcg_gen_mov_ptr(b, base); + base = b; + } + gen_set_label(loop); t0 = tcg_temp_new_i64(); tp = tcg_temp_new_ptr(); - tcg_gen_add_ptr(tp, cpu_env, i); + tcg_gen_add_ptr(tp, base, i); tcg_gen_ld_i64(t0, tp, vofs); tcg_gen_addi_ptr(i, i, 8); tcg_temp_free_ptr(tp); @@ -XXX,XX +XXX,XX @@ static void do_str(DisasContext *s, uint32_t vofs, int len, int rn, int imm) tcg_gen_brcondi_ptr(TCG_COND_LTU, i, len_align, loop); tcg_temp_free_ptr(i); + + if (base != cpu_env) { + tcg_temp_free_ptr(base); + assert(len_remain == 0); + } } /* Predicate register stores can be any multiple of 2. */ if (len_remain) { t0 = tcg_temp_new_i64(); - tcg_gen_ld_i64(t0, cpu_env, vofs + len_align); + tcg_gen_ld_i64(t0, base, vofs + len_align); switch (len_remain) { case 2: @@ -XXX,XX +XXX,XX @@ static bool trans_LDR_zri(DisasContext *s, arg_rri *a) if (sve_access_check(s)) { int size = vec_full_reg_size(s); int off = vec_full_reg_offset(s, a->rd); - do_ldr(s, off, size, a->rn, a->imm * size); + gen_sve_ldr(s, cpu_env, off, size, a->rn, a->imm * size); } return true; } @@ -XXX,XX +XXX,XX @@ static bool trans_LDR_pri(DisasContext *s, arg_rri *a) if (sve_access_check(s)) { int size = pred_full_reg_size(s); int off = pred_full_reg_offset(s, a->rd); - do_ldr(s, off, size, a->rn, a->imm * size); + gen_sve_ldr(s, cpu_env, off, size, a->rn, a->imm * size); } return true; } @@ -XXX,XX +XXX,XX @@ static bool trans_STR_zri(DisasContext *s, arg_rri *a) if (sve_access_check(s)) { int size = vec_full_reg_size(s); int off = vec_full_reg_offset(s, a->rd); - do_str(s, off, size, a->rn, a->imm * size); + gen_sve_str(s, cpu_env, off, size, a->rn, a->imm * size); } return true; } @@ -XXX,XX +XXX,XX @@ static bool trans_STR_pri(DisasContext *s, arg_rri *a) if (sve_access_check(s)) { int size = pred_full_reg_size(s); int off = pred_full_reg_offset(s, a->rd); - do_str(s, off, size, a->rn, a->imm * size); + gen_sve_str(s, cpu_env, off, size, a->rn, a->imm * size); } return true; } -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> We can reuse the SVE functions for LDR and STR, passing in the base of the ZA vector and a zero offset. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-23-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sme.decode | 7 +++++++ target/arm/translate-sme.c | 24 ++++++++++++++++++++++++ 2 files changed, 31 insertions(+) diff --git a/target/arm/sme.decode b/target/arm/sme.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme.decode +++ b/target/arm/sme.decode @@ -XXX,XX +XXX,XX @@ LDST1 1110000 0 esz:2 st:1 rm:5 v:1 .. pg:3 rn:5 0 za_imm:4 \ &ldst rs=%mova_rs LDST1 1110000 111 st:1 rm:5 v:1 .. pg:3 rn:5 0 za_imm:4 \ &ldst esz=4 rs=%mova_rs + +&ldstr rv rn imm +@ldstr ....... ... . ...... .. ... rn:5 . imm:4 \ + &ldstr rv=%mova_rs + +LDR 1110000 100 0 000000 .. 000 ..... 0 .... @ldstr +STR 1110000 100 1 000000 .. 000 ..... 0 .... @ldstr diff --git a/target/arm/translate-sme.c b/target/arm/translate-sme.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sme.c +++ b/target/arm/translate-sme.c @@ -XXX,XX +XXX,XX @@ static bool trans_LDST1(DisasContext *s, arg_LDST1 *a) tcg_temp_free_i64(addr); return true; } + +typedef void GenLdStR(DisasContext *, TCGv_ptr, int, int, int, int); + +static bool do_ldst_r(DisasContext *s, arg_ldstr *a, GenLdStR *fn) +{ + int svl = streaming_vec_reg_size(s); + int imm = a->imm; + TCGv_ptr base; + + if (!sme_za_enabled_check(s)) { + return true; + } + + /* ZA[n] equates to ZA0H.B[n]. */ + base = get_tile_rowcol(s, MO_8, a->rv, imm, false); + + fn(s, base, 0, svl, a->rn, imm * svl); + + tcg_temp_free_ptr(base); + return true; +} + +TRANS_FEAT(LDR, aa64_sme, do_ldst_r, a, gen_sve_ldr) +TRANS_FEAT(STR, aa64_sme, do_ldst_r, a, gen_sve_str) -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-24-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper-sme.h | 5 +++ target/arm/sme.decode | 11 +++++ target/arm/sme_helper.c | 90 ++++++++++++++++++++++++++++++++++++++ target/arm/translate-sme.c | 31 +++++++++++++ 4 files changed, 137 insertions(+) diff --git a/target/arm/helper-sme.h b/target/arm/helper-sme.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-sme.h +++ b/target/arm/helper-sme.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_5(sme_st1q_be_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i DEF_HELPER_FLAGS_5(sme_st1q_le_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) DEF_HELPER_FLAGS_5(sme_st1q_be_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) DEF_HELPER_FLAGS_5(sme_st1q_le_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) + +DEF_HELPER_FLAGS_5(sme_addha_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_5(sme_addva_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_5(sme_addha_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_5(sme_addva_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) diff --git a/target/arm/sme.decode b/target/arm/sme.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme.decode +++ b/target/arm/sme.decode @@ -XXX,XX +XXX,XX @@ LDST1 1110000 111 st:1 rm:5 v:1 .. pg:3 rn:5 0 za_imm:4 \ LDR 1110000 100 0 000000 .. 000 ..... 0 .... @ldstr STR 1110000 100 1 000000 .. 000 ..... 0 .... @ldstr + +### SME Add Vector to Array + +&adda zad zn pm pn +@adda_32 ........ .. ..... . pm:3 pn:3 zn:5 ... zad:2 &adda +@adda_64 ........ .. ..... . pm:3 pn:3 zn:5 .. zad:3 &adda + +ADDHA_s 11000000 10 01000 0 ... ... ..... 000 .. @adda_32 +ADDVA_s 11000000 10 01000 1 ... ... ..... 000 .. @adda_32 +ADDHA_d 11000000 11 01000 0 ... ... ..... 00 ... @adda_64 +ADDVA_d 11000000 11 01000 1 ... ... ..... 00 ... @adda_64 diff --git a/target/arm/sme_helper.c b/target/arm/sme_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme_helper.c +++ b/target/arm/sme_helper.c @@ -XXX,XX +XXX,XX @@ DO_ST(q, _be, MO_128) DO_ST(q, _le, MO_128) #undef DO_ST + +void HELPER(sme_addha_s)(void *vzda, void *vzn, void *vpn, + void *vpm, uint32_t desc) +{ + intptr_t row, col, oprsz = simd_oprsz(desc) / 4; + uint64_t *pn = vpn, *pm = vpm; + uint32_t *zda = vzda, *zn = vzn; + + for (row = 0; row < oprsz; ) { + uint64_t pa = pn[row >> 4]; + do { + if (pa & 1) { + for (col = 0; col < oprsz; ) { + uint64_t pb = pm[col >> 4]; + do { + if (pb & 1) { + zda[tile_vslice_index(row) + H4(col)] += zn[H4(col)]; + } + pb >>= 4; + } while (++col & 15); + } + } + pa >>= 4; + } while (++row & 15); + } +} + +void HELPER(sme_addha_d)(void *vzda, void *vzn, void *vpn, + void *vpm, uint32_t desc) +{ + intptr_t row, col, oprsz = simd_oprsz(desc) / 8; + uint8_t *pn = vpn, *pm = vpm; + uint64_t *zda = vzda, *zn = vzn; + + for (row = 0; row < oprsz; ++row) { + if (pn[H1(row)] & 1) { + for (col = 0; col < oprsz; ++col) { + if (pm[H1(col)] & 1) { + zda[tile_vslice_index(row) + col] += zn[col]; + } + } + } + } +} + +void HELPER(sme_addva_s)(void *vzda, void *vzn, void *vpn, + void *vpm, uint32_t desc) +{ + intptr_t row, col, oprsz = simd_oprsz(desc) / 4; + uint64_t *pn = vpn, *pm = vpm; + uint32_t *zda = vzda, *zn = vzn; + + for (row = 0; row < oprsz; ) { + uint64_t pa = pn[row >> 4]; + do { + if (pa & 1) { + uint32_t zn_row = zn[H4(row)]; + for (col = 0; col < oprsz; ) { + uint64_t pb = pm[col >> 4]; + do { + if (pb & 1) { + zda[tile_vslice_index(row) + H4(col)] += zn_row; + } + pb >>= 4; + } while (++col & 15); + } + } + pa >>= 4; + } while (++row & 15); + } +} + +void HELPER(sme_addva_d)(void *vzda, void *vzn, void *vpn, + void *vpm, uint32_t desc) +{ + intptr_t row, col, oprsz = simd_oprsz(desc) / 8; + uint8_t *pn = vpn, *pm = vpm; + uint64_t *zda = vzda, *zn = vzn; + + for (row = 0; row < oprsz; ++row) { + if (pn[H1(row)] & 1) { + uint64_t zn_row = zn[row]; + for (col = 0; col < oprsz; ++col) { + if (pm[H1(col)] & 1) { + zda[tile_vslice_index(row) + col] += zn_row; + } + } + } + } +} diff --git a/target/arm/translate-sme.c b/target/arm/translate-sme.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sme.c +++ b/target/arm/translate-sme.c @@ -XXX,XX +XXX,XX @@ static bool do_ldst_r(DisasContext *s, arg_ldstr *a, GenLdStR *fn) TRANS_FEAT(LDR, aa64_sme, do_ldst_r, a, gen_sve_ldr) TRANS_FEAT(STR, aa64_sme, do_ldst_r, a, gen_sve_str) + +static bool do_adda(DisasContext *s, arg_adda *a, MemOp esz, + gen_helper_gvec_4 *fn) +{ + int svl = streaming_vec_reg_size(s); + uint32_t desc = simd_desc(svl, svl, 0); + TCGv_ptr za, zn, pn, pm; + + if (!sme_smza_enabled_check(s)) { + return true; + } + + /* Sum XZR+zad to find ZAd. */ + za = get_tile_rowcol(s, esz, 31, a->zad, false); + zn = vec_full_reg_ptr(s, a->zn); + pn = pred_full_reg_ptr(s, a->pn); + pm = pred_full_reg_ptr(s, a->pm); + + fn(za, zn, pn, pm, tcg_constant_i32(desc)); + + tcg_temp_free_ptr(za); + tcg_temp_free_ptr(zn); + tcg_temp_free_ptr(pn); + tcg_temp_free_ptr(pm); + return true; +} + +TRANS_FEAT(ADDHA_s, aa64_sme, do_adda, a, MO_32, gen_helper_sme_addha_s) +TRANS_FEAT(ADDVA_s, aa64_sme, do_adda, a, MO_32, gen_helper_sme_addva_s) +TRANS_FEAT(ADDHA_d, aa64_sme_i16i64, do_adda, a, MO_64, gen_helper_sme_addha_d) +TRANS_FEAT(ADDVA_d, aa64_sme_i16i64, do_adda, a, MO_64, gen_helper_sme_addva_d) -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-25-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper-sme.h | 5 +++ target/arm/sme.decode | 9 +++++ target/arm/sme_helper.c | 69 ++++++++++++++++++++++++++++++++++++++ target/arm/translate-sme.c | 32 ++++++++++++++++++ 4 files changed, 115 insertions(+) diff --git a/target/arm/helper-sme.h b/target/arm/helper-sme.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-sme.h +++ b/target/arm/helper-sme.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_5(sme_addha_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_5(sme_addva_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_5(sme_addha_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_5(sme_addva_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) + +DEF_HELPER_FLAGS_7(sme_fmopa_s, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_7(sme_fmopa_d, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, ptr, i32) diff --git a/target/arm/sme.decode b/target/arm/sme.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme.decode +++ b/target/arm/sme.decode @@ -XXX,XX +XXX,XX @@ ADDHA_s 11000000 10 01000 0 ... ... ..... 000 .. @adda_32 ADDVA_s 11000000 10 01000 1 ... ... ..... 000 .. @adda_32 ADDHA_d 11000000 11 01000 0 ... ... ..... 00 ... @adda_64 ADDVA_d 11000000 11 01000 1 ... ... ..... 00 ... @adda_64 + +### SME Outer Product + +&op zad zn zm pm pn sub:bool +@op_32 ........ ... zm:5 pm:3 pn:3 zn:5 sub:1 .. zad:2 &op +@op_64 ........ ... zm:5 pm:3 pn:3 zn:5 sub:1 . zad:3 &op + +FMOPA_s 10000000 100 ..... ... ... ..... . 00 .. @op_32 +FMOPA_d 10000000 110 ..... ... ... ..... . 0 ... @op_64 diff --git a/target/arm/sme_helper.c b/target/arm/sme_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme_helper.c +++ b/target/arm/sme_helper.c @@ -XXX,XX +XXX,XX @@ #include "exec/cpu_ldst.h" #include "exec/exec-all.h" #include "qemu/int128.h" +#include "fpu/softfloat.h" #include "vec_internal.h" #include "sve_ldst_internal.h" @@ -XXX,XX +XXX,XX @@ void HELPER(sme_addva_d)(void *vzda, void *vzn, void *vpn, } } } + +void HELPER(sme_fmopa_s)(void *vza, void *vzn, void *vzm, void *vpn, + void *vpm, void *vst, uint32_t desc) +{ + intptr_t row, col, oprsz = simd_maxsz(desc); + uint32_t neg = simd_data(desc) << 31; + uint16_t *pn = vpn, *pm = vpm; + float_status fpst; + + /* + * Make a copy of float_status because this operation does not + * update the cumulative fp exception status. It also produces + * default nans. + */ + fpst = *(float_status *)vst; + set_default_nan_mode(true, &fpst); + + for (row = 0; row < oprsz; ) { + uint16_t pa = pn[H2(row >> 4)]; + do { + if (pa & 1) { + void *vza_row = vza + tile_vslice_offset(row); + uint32_t n = *(uint32_t *)(vzn + H1_4(row)) ^ neg; + + for (col = 0; col < oprsz; ) { + uint16_t pb = pm[H2(col >> 4)]; + do { + if (pb & 1) { + uint32_t *a = vza_row + H1_4(col); + uint32_t *m = vzm + H1_4(col); + *a = float32_muladd(n, *m, *a, 0, vst); + } + col += 4; + pb >>= 4; + } while (col & 15); + } + } + row += 4; + pa >>= 4; + } while (row & 15); + } +} + +void HELPER(sme_fmopa_d)(void *vza, void *vzn, void *vzm, void *vpn, + void *vpm, void *vst, uint32_t desc) +{ + intptr_t row, col, oprsz = simd_oprsz(desc) / 8; + uint64_t neg = (uint64_t)simd_data(desc) << 63; + uint64_t *za = vza, *zn = vzn, *zm = vzm; + uint8_t *pn = vpn, *pm = vpm; + float_status fpst = *(float_status *)vst; + + set_default_nan_mode(true, &fpst); + + for (row = 0; row < oprsz; ++row) { + if (pn[H1(row)] & 1) { + uint64_t *za_row = &za[tile_vslice_index(row)]; + uint64_t n = zn[row] ^ neg; + + for (col = 0; col < oprsz; ++col) { + if (pm[H1(col)] & 1) { + uint64_t *a = &za_row[col]; + *a = float64_muladd(n, zm[col], *a, 0, &fpst); + } + } + } + } +} diff --git a/target/arm/translate-sme.c b/target/arm/translate-sme.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sme.c +++ b/target/arm/translate-sme.c @@ -XXX,XX +XXX,XX @@ TRANS_FEAT(ADDHA_s, aa64_sme, do_adda, a, MO_32, gen_helper_sme_addha_s) TRANS_FEAT(ADDVA_s, aa64_sme, do_adda, a, MO_32, gen_helper_sme_addva_s) TRANS_FEAT(ADDHA_d, aa64_sme_i16i64, do_adda, a, MO_64, gen_helper_sme_addha_d) TRANS_FEAT(ADDVA_d, aa64_sme_i16i64, do_adda, a, MO_64, gen_helper_sme_addva_d) + +static bool do_outprod_fpst(DisasContext *s, arg_op *a, MemOp esz, + gen_helper_gvec_5_ptr *fn) +{ + int svl = streaming_vec_reg_size(s); + uint32_t desc = simd_desc(svl, svl, a->sub); + TCGv_ptr za, zn, zm, pn, pm, fpst; + + if (!sme_smza_enabled_check(s)) { + return true; + } + + /* Sum XZR+zad to find ZAd. */ + za = get_tile_rowcol(s, esz, 31, a->zad, false); + zn = vec_full_reg_ptr(s, a->zn); + zm = vec_full_reg_ptr(s, a->zm); + pn = pred_full_reg_ptr(s, a->pn); + pm = pred_full_reg_ptr(s, a->pm); + fpst = fpstatus_ptr(FPST_FPCR); + + fn(za, zn, zm, pn, pm, fpst, tcg_constant_i32(desc)); + + tcg_temp_free_ptr(za); + tcg_temp_free_ptr(zn); + tcg_temp_free_ptr(pn); + tcg_temp_free_ptr(pm); + tcg_temp_free_ptr(fpst); + return true; +} + +TRANS_FEAT(FMOPA_s, aa64_sme, do_outprod_fpst, a, MO_32, gen_helper_sme_fmopa_s) +TRANS_FEAT(FMOPA_d, aa64_sme_f64f64, do_outprod_fpst, a, MO_64, gen_helper_sme_fmopa_d) -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-26-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper-sme.h | 2 ++ target/arm/sme.decode | 2 ++ target/arm/sme_helper.c | 56 ++++++++++++++++++++++++++++++++++++++ target/arm/translate-sme.c | 30 ++++++++++++++++++++ 4 files changed, 90 insertions(+) diff --git a/target/arm/helper-sme.h b/target/arm/helper-sme.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-sme.h +++ b/target/arm/helper-sme.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_7(sme_fmopa_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_7(sme_fmopa_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_6(sme_bfmopa, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, i32) diff --git a/target/arm/sme.decode b/target/arm/sme.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme.decode +++ b/target/arm/sme.decode @@ -XXX,XX +XXX,XX @@ ADDVA_d 11000000 11 01000 1 ... ... ..... 00 ... @adda_64 FMOPA_s 10000000 100 ..... ... ... ..... . 00 .. @op_32 FMOPA_d 10000000 110 ..... ... ... ..... . 0 ... @op_64 + +BFMOPA 10000001 100 ..... ... ... ..... . 00 .. @op_32 diff --git a/target/arm/sme_helper.c b/target/arm/sme_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme_helper.c +++ b/target/arm/sme_helper.c @@ -XXX,XX +XXX,XX @@ void HELPER(sme_fmopa_d)(void *vza, void *vzn, void *vzm, void *vpn, } } } + +/* + * Alter PAIR as needed for controlling predicates being false, + * and for NEG on an enabled row element. + */ +static inline uint32_t f16mop_adj_pair(uint32_t pair, uint32_t pg, uint32_t neg) +{ + /* + * The pseudocode uses a conditional negate after the conditional zero. + * It is simpler here to unconditionally negate before conditional zero. + */ + pair ^= neg; + if (!(pg & 1)) { + pair &= 0xffff0000u; + } + if (!(pg & 4)) { + pair &= 0x0000ffffu; + } + return pair; +} + +void HELPER(sme_bfmopa)(void *vza, void *vzn, void *vzm, void *vpn, + void *vpm, uint32_t desc) +{ + intptr_t row, col, oprsz = simd_maxsz(desc); + uint32_t neg = simd_data(desc) * 0x80008000u; + uint16_t *pn = vpn, *pm = vpm; + + for (row = 0; row < oprsz; ) { + uint16_t prow = pn[H2(row >> 4)]; + do { + void *vza_row = vza + tile_vslice_offset(row); + uint32_t n = *(uint32_t *)(vzn + H1_4(row)); + + n = f16mop_adj_pair(n, prow, neg); + + for (col = 0; col < oprsz; ) { + uint16_t pcol = pm[H2(col >> 4)]; + do { + if (prow & pcol & 0b0101) { + uint32_t *a = vza_row + H1_4(col); + uint32_t m = *(uint32_t *)(vzm + H1_4(col)); + + m = f16mop_adj_pair(m, pcol, 0); + *a = bfdotadd(*a, n, m); + + col += 4; + pcol >>= 4; + } + } while (col & 15); + } + row += 4; + prow >>= 4; + } while (row & 15); + } +} diff --git a/target/arm/translate-sme.c b/target/arm/translate-sme.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sme.c +++ b/target/arm/translate-sme.c @@ -XXX,XX +XXX,XX @@ TRANS_FEAT(ADDVA_s, aa64_sme, do_adda, a, MO_32, gen_helper_sme_addva_s) TRANS_FEAT(ADDHA_d, aa64_sme_i16i64, do_adda, a, MO_64, gen_helper_sme_addha_d) TRANS_FEAT(ADDVA_d, aa64_sme_i16i64, do_adda, a, MO_64, gen_helper_sme_addva_d) +static bool do_outprod(DisasContext *s, arg_op *a, MemOp esz, + gen_helper_gvec_5 *fn) +{ + int svl = streaming_vec_reg_size(s); + uint32_t desc = simd_desc(svl, svl, a->sub); + TCGv_ptr za, zn, zm, pn, pm; + + if (!sme_smza_enabled_check(s)) { + return true; + } + + /* Sum XZR+zad to find ZAd. */ + za = get_tile_rowcol(s, esz, 31, a->zad, false); + zn = vec_full_reg_ptr(s, a->zn); + zm = vec_full_reg_ptr(s, a->zm); + pn = pred_full_reg_ptr(s, a->pn); + pm = pred_full_reg_ptr(s, a->pm); + + fn(za, zn, zm, pn, pm, tcg_constant_i32(desc)); + + tcg_temp_free_ptr(za); + tcg_temp_free_ptr(zn); + tcg_temp_free_ptr(pn); + tcg_temp_free_ptr(pm); + return true; +} + static bool do_outprod_fpst(DisasContext *s, arg_op *a, MemOp esz, gen_helper_gvec_5_ptr *fn) { @@ -XXX,XX +XXX,XX @@ static bool do_outprod_fpst(DisasContext *s, arg_op *a, MemOp esz, TRANS_FEAT(FMOPA_s, aa64_sme, do_outprod_fpst, a, MO_32, gen_helper_sme_fmopa_s) TRANS_FEAT(FMOPA_d, aa64_sme_f64f64, do_outprod_fpst, a, MO_64, gen_helper_sme_fmopa_d) + +/* TODO: FEAT_EBF16 */ +TRANS_FEAT(BFMOPA, aa64_sme, do_outprod, a, MO_32, gen_helper_sme_bfmopa) -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-27-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper-sme.h | 2 ++ target/arm/sme.decode | 1 + target/arm/sme_helper.c | 74 ++++++++++++++++++++++++++++++++++++++ target/arm/translate-sme.c | 1 + 4 files changed, 78 insertions(+) diff --git a/target/arm/helper-sme.h b/target/arm/helper-sme.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-sme.h +++ b/target/arm/helper-sme.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_5(sme_addva_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_5(sme_addha_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_5(sme_addva_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_7(sme_fmopa_h, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_7(sme_fmopa_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_7(sme_fmopa_d, TCG_CALL_NO_RWG, diff --git a/target/arm/sme.decode b/target/arm/sme.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme.decode +++ b/target/arm/sme.decode @@ -XXX,XX +XXX,XX @@ FMOPA_s 10000000 100 ..... ... ... ..... . 00 .. @op_32 FMOPA_d 10000000 110 ..... ... ... ..... . 0 ... @op_64 BFMOPA 10000001 100 ..... ... ... ..... . 00 .. @op_32 +FMOPA_h 10000001 101 ..... ... ... ..... . 00 .. @op_32 diff --git a/target/arm/sme_helper.c b/target/arm/sme_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme_helper.c +++ b/target/arm/sme_helper.c @@ -XXX,XX +XXX,XX @@ static inline uint32_t f16mop_adj_pair(uint32_t pair, uint32_t pg, uint32_t neg) return pair; } +static float32 f16_dotadd(float32 sum, uint32_t e1, uint32_t e2, + float_status *s_std, float_status *s_odd) +{ + float64 e1r = float16_to_float64(e1 & 0xffff, true, s_std); + float64 e1c = float16_to_float64(e1 >> 16, true, s_std); + float64 e2r = float16_to_float64(e2 & 0xffff, true, s_std); + float64 e2c = float16_to_float64(e2 >> 16, true, s_std); + float64 t64; + float32 t32; + + /* + * The ARM pseudocode function FPDot performs both multiplies + * and the add with a single rounding operation. Emulate this + * by performing the first multiply in round-to-odd, then doing + * the second multiply as fused multiply-add, and rounding to + * float32 all in one step. + */ + t64 = float64_mul(e1r, e2r, s_odd); + t64 = float64r32_muladd(e1c, e2c, t64, 0, s_std); + + /* This conversion is exact, because we've already rounded. */ + t32 = float64_to_float32(t64, s_std); + + /* The final accumulation step is not fused. */ + return float32_add(sum, t32, s_std); +} + +void HELPER(sme_fmopa_h)(void *vza, void *vzn, void *vzm, void *vpn, + void *vpm, void *vst, uint32_t desc) +{ + intptr_t row, col, oprsz = simd_maxsz(desc); + uint32_t neg = simd_data(desc) * 0x80008000u; + uint16_t *pn = vpn, *pm = vpm; + float_status fpst_odd, fpst_std; + + /* + * Make a copy of float_status because this operation does not + * update the cumulative fp exception status. It also produces + * default nans. Make a second copy with round-to-odd -- see above. + */ + fpst_std = *(float_status *)vst; + set_default_nan_mode(true, &fpst_std); + fpst_odd = fpst_std; + set_float_rounding_mode(float_round_to_odd, &fpst_odd); + + for (row = 0; row < oprsz; ) { + uint16_t prow = pn[H2(row >> 4)]; + do { + void *vza_row = vza + tile_vslice_offset(row); + uint32_t n = *(uint32_t *)(vzn + H1_4(row)); + + n = f16mop_adj_pair(n, prow, neg); + + for (col = 0; col < oprsz; ) { + uint16_t pcol = pm[H2(col >> 4)]; + do { + if (prow & pcol & 0b0101) { + uint32_t *a = vza_row + H1_4(col); + uint32_t m = *(uint32_t *)(vzm + H1_4(col)); + + m = f16mop_adj_pair(m, pcol, 0); + *a = f16_dotadd(*a, n, m, &fpst_std, &fpst_odd); + + col += 4; + pcol >>= 4; + } + } while (col & 15); + } + row += 4; + prow >>= 4; + } while (row & 15); + } +} + void HELPER(sme_bfmopa)(void *vza, void *vzn, void *vzm, void *vpn, void *vpm, uint32_t desc) { diff --git a/target/arm/translate-sme.c b/target/arm/translate-sme.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sme.c +++ b/target/arm/translate-sme.c @@ -XXX,XX +XXX,XX @@ static bool do_outprod_fpst(DisasContext *s, arg_op *a, MemOp esz, return true; } +TRANS_FEAT(FMOPA_h, aa64_sme, do_outprod_fpst, a, MO_32, gen_helper_sme_fmopa_h) TRANS_FEAT(FMOPA_s, aa64_sme, do_outprod_fpst, a, MO_32, gen_helper_sme_fmopa_s) TRANS_FEAT(FMOPA_d, aa64_sme_f64f64, do_outprod_fpst, a, MO_64, gen_helper_sme_fmopa_d) -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> This is SMOPA, SUMOPA, USMOPA_s, UMOPA, for both Int8 and Int16. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-28-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper-sme.h | 16 ++++++++ target/arm/sme.decode | 10 +++++ target/arm/sme_helper.c | 82 ++++++++++++++++++++++++++++++++++++++ target/arm/translate-sme.c | 10 +++++ 4 files changed, 118 insertions(+) diff --git a/target/arm/helper-sme.h b/target/arm/helper-sme.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-sme.h +++ b/target/arm/helper-sme.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_7(sme_fmopa_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_6(sme_bfmopa, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_6(sme_smopa_s, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_6(sme_umopa_s, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_6(sme_sumopa_s, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_6(sme_usmopa_s, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_6(sme_smopa_d, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_6(sme_umopa_d, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_6(sme_sumopa_d, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_6(sme_usmopa_d, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, i32) diff --git a/target/arm/sme.decode b/target/arm/sme.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme.decode +++ b/target/arm/sme.decode @@ -XXX,XX +XXX,XX @@ FMOPA_d 10000000 110 ..... ... ... ..... . 0 ... @op_64 BFMOPA 10000001 100 ..... ... ... ..... . 00 .. @op_32 FMOPA_h 10000001 101 ..... ... ... ..... . 00 .. @op_32 + +SMOPA_s 1010000 0 10 0 ..... ... ... ..... . 00 .. @op_32 +SUMOPA_s 1010000 0 10 1 ..... ... ... ..... . 00 .. @op_32 +USMOPA_s 1010000 1 10 0 ..... ... ... ..... . 00 .. @op_32 +UMOPA_s 1010000 1 10 1 ..... ... ... ..... . 00 .. @op_32 + +SMOPA_d 1010000 0 11 0 ..... ... ... ..... . 0 ... @op_64 +SUMOPA_d 1010000 0 11 1 ..... ... ... ..... . 0 ... @op_64 +USMOPA_d 1010000 1 11 0 ..... ... ... ..... . 0 ... @op_64 +UMOPA_d 1010000 1 11 1 ..... ... ... ..... . 0 ... @op_64 diff --git a/target/arm/sme_helper.c b/target/arm/sme_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme_helper.c +++ b/target/arm/sme_helper.c @@ -XXX,XX +XXX,XX @@ void HELPER(sme_bfmopa)(void *vza, void *vzn, void *vzm, void *vpn, } while (row & 15); } } + +typedef uint64_t IMOPFn(uint64_t, uint64_t, uint64_t, uint8_t, bool); + +static inline void do_imopa(uint64_t *za, uint64_t *zn, uint64_t *zm, + uint8_t *pn, uint8_t *pm, + uint32_t desc, IMOPFn *fn) +{ + intptr_t row, col, oprsz = simd_oprsz(desc) / 8; + bool neg = simd_data(desc); + + for (row = 0; row < oprsz; ++row) { + uint8_t pa = pn[H1(row)]; + uint64_t *za_row = &za[tile_vslice_index(row)]; + uint64_t n = zn[row]; + + for (col = 0; col < oprsz; ++col) { + uint8_t pb = pm[H1(col)]; + uint64_t *a = &za_row[col]; + + *a = fn(n, zm[col], *a, pa & pb, neg); + } + } +} + +#define DEF_IMOP_32(NAME, NTYPE, MTYPE) \ +static uint64_t NAME(uint64_t n, uint64_t m, uint64_t a, uint8_t p, bool neg) \ +{ \ + uint32_t sum0 = 0, sum1 = 0; \ + /* Apply P to N as a mask, making the inactive elements 0. */ \ + n &= expand_pred_b(p); \ + sum0 += (NTYPE)(n >> 0) * (MTYPE)(m >> 0); \ + sum0 += (NTYPE)(n >> 8) * (MTYPE)(m >> 8); \ + sum0 += (NTYPE)(n >> 16) * (MTYPE)(m >> 16); \ + sum0 += (NTYPE)(n >> 24) * (MTYPE)(m >> 24); \ + sum1 += (NTYPE)(n >> 32) * (MTYPE)(m >> 32); \ + sum1 += (NTYPE)(n >> 40) * (MTYPE)(m >> 40); \ + sum1 += (NTYPE)(n >> 48) * (MTYPE)(m >> 48); \ + sum1 += (NTYPE)(n >> 56) * (MTYPE)(m >> 56); \ + if (neg) { \ + sum0 = (uint32_t)a - sum0, sum1 = (uint32_t)(a >> 32) - sum1; \ + } else { \ + sum0 = (uint32_t)a + sum0, sum1 = (uint32_t)(a >> 32) + sum1; \ + } \ + return ((uint64_t)sum1 << 32) | sum0; \ +} + +#define DEF_IMOP_64(NAME, NTYPE, MTYPE) \ +static uint64_t NAME(uint64_t n, uint64_t m, uint64_t a, uint8_t p, bool neg) \ +{ \ + uint64_t sum = 0; \ + /* Apply P to N as a mask, making the inactive elements 0. */ \ + n &= expand_pred_h(p); \ + sum += (NTYPE)(n >> 0) * (MTYPE)(m >> 0); \ + sum += (NTYPE)(n >> 16) * (MTYPE)(m >> 16); \ + sum += (NTYPE)(n >> 32) * (MTYPE)(m >> 32); \ + sum += (NTYPE)(n >> 48) * (MTYPE)(m >> 48); \ + return neg ? a - sum : a + sum; \ +} + +DEF_IMOP_32(smopa_s, int8_t, int8_t) +DEF_IMOP_32(umopa_s, uint8_t, uint8_t) +DEF_IMOP_32(sumopa_s, int8_t, uint8_t) +DEF_IMOP_32(usmopa_s, uint8_t, int8_t) + +DEF_IMOP_64(smopa_d, int16_t, int16_t) +DEF_IMOP_64(umopa_d, uint16_t, uint16_t) +DEF_IMOP_64(sumopa_d, int16_t, uint16_t) +DEF_IMOP_64(usmopa_d, uint16_t, int16_t) + +#define DEF_IMOPH(NAME) \ + void HELPER(sme_##NAME)(void *vza, void *vzn, void *vzm, void *vpn, \ + void *vpm, uint32_t desc) \ + { do_imopa(vza, vzn, vzm, vpn, vpm, desc, NAME); } + +DEF_IMOPH(smopa_s) +DEF_IMOPH(umopa_s) +DEF_IMOPH(sumopa_s) +DEF_IMOPH(usmopa_s) +DEF_IMOPH(smopa_d) +DEF_IMOPH(umopa_d) +DEF_IMOPH(sumopa_d) +DEF_IMOPH(usmopa_d) diff --git a/target/arm/translate-sme.c b/target/arm/translate-sme.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sme.c +++ b/target/arm/translate-sme.c @@ -XXX,XX +XXX,XX @@ TRANS_FEAT(FMOPA_d, aa64_sme_f64f64, do_outprod_fpst, a, MO_64, gen_helper_sme_f /* TODO: FEAT_EBF16 */ TRANS_FEAT(BFMOPA, aa64_sme, do_outprod, a, MO_32, gen_helper_sme_bfmopa) + +TRANS_FEAT(SMOPA_s, aa64_sme, do_outprod, a, MO_32, gen_helper_sme_smopa_s) +TRANS_FEAT(UMOPA_s, aa64_sme, do_outprod, a, MO_32, gen_helper_sme_umopa_s) +TRANS_FEAT(SUMOPA_s, aa64_sme, do_outprod, a, MO_32, gen_helper_sme_sumopa_s) +TRANS_FEAT(USMOPA_s, aa64_sme, do_outprod, a, MO_32, gen_helper_sme_usmopa_s) + +TRANS_FEAT(SMOPA_d, aa64_sme_i16i64, do_outprod, a, MO_64, gen_helper_sme_smopa_d) +TRANS_FEAT(UMOPA_d, aa64_sme_i16i64, do_outprod, a, MO_64, gen_helper_sme_umopa_d) +TRANS_FEAT(SUMOPA_d, aa64_sme_i16i64, do_outprod, a, MO_64, gen_helper_sme_sumopa_d) +TRANS_FEAT(USMOPA_d, aa64_sme_i16i64, do_outprod, a, MO_64, gen_helper_sme_usmopa_d) -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> This is an SVE instruction that operates using the SVE vector length but that it is present only if SME is implemented. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-29-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sve.decode | 20 +++++++++++++ target/arm/translate-sve.c | 57 ++++++++++++++++++++++++++++++++++++++ 2 files changed, 77 insertions(+) diff --git a/target/arm/sve.decode b/target/arm/sve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sve.decode +++ b/target/arm/sve.decode @@ -XXX,XX +XXX,XX @@ BFMLALT_zzxw 01100100 11 1 ..... 0100.1 ..... ..... @rrxr_3a esz=2 ### SVE2 floating-point bfloat16 dot-product (indexed) BFDOT_zzxz 01100100 01 1 ..... 010000 ..... ..... @rrxr_2 esz=2 + +### SVE broadcast predicate element + +&psel esz pd pn pm rv imm +%psel_rv 16:2 !function=plus_12 +%psel_imm_b 22:2 19:2 +%psel_imm_h 22:2 20:1 +%psel_imm_s 22:2 +%psel_imm_d 23:1 +@psel ........ .. . ... .. .. pn:4 . pm:4 . pd:4 \ + &psel rv=%psel_rv + +PSEL 00100101 .. 1 ..1 .. 01 .... 0 .... 0 .... \ + @psel esz=0 imm=%psel_imm_b +PSEL 00100101 .. 1 .10 .. 01 .... 0 .... 0 .... \ + @psel esz=1 imm=%psel_imm_h +PSEL 00100101 .. 1 100 .. 01 .... 0 .... 0 .... \ + @psel esz=2 imm=%psel_imm_s +PSEL 00100101 .1 1 000 .. 01 .... 0 .... 0 .... \ + @psel esz=3 imm=%psel_imm_d diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static bool do_BFMLAL_zzxw(DisasContext *s, arg_rrxr_esz *a, bool sel) TRANS_FEAT(BFMLALB_zzxw, aa64_sve_bf16, do_BFMLAL_zzxw, a, false) TRANS_FEAT(BFMLALT_zzxw, aa64_sve_bf16, do_BFMLAL_zzxw, a, true) + +static bool trans_PSEL(DisasContext *s, arg_psel *a) +{ + int vl = vec_full_reg_size(s); + int pl = pred_gvec_reg_size(s); + int elements = vl >> a->esz; + TCGv_i64 tmp, didx, dbit; + TCGv_ptr ptr; + + if (!dc_isar_feature(aa64_sme, s)) { + return false; + } + if (!sve_access_check(s)) { + return true; + } + + tmp = tcg_temp_new_i64(); + dbit = tcg_temp_new_i64(); + didx = tcg_temp_new_i64(); + ptr = tcg_temp_new_ptr(); + + /* Compute the predicate element. */ + tcg_gen_addi_i64(tmp, cpu_reg(s, a->rv), a->imm); + if (is_power_of_2(elements)) { + tcg_gen_andi_i64(tmp, tmp, elements - 1); + } else { + tcg_gen_remu_i64(tmp, tmp, tcg_constant_i64(elements)); + } + + /* Extract the predicate byte and bit indices. */ + tcg_gen_shli_i64(tmp, tmp, a->esz); + tcg_gen_andi_i64(dbit, tmp, 7); + tcg_gen_shri_i64(didx, tmp, 3); + if (HOST_BIG_ENDIAN) { + tcg_gen_xori_i64(didx, didx, 7); + } + + /* Load the predicate word. */ + tcg_gen_trunc_i64_ptr(ptr, didx); + tcg_gen_add_ptr(ptr, ptr, cpu_env); + tcg_gen_ld8u_i64(tmp, ptr, pred_full_reg_offset(s, a->pm)); + + /* Extract the predicate bit and replicate to MO_64. */ + tcg_gen_shr_i64(tmp, tmp, dbit); + tcg_gen_andi_i64(tmp, tmp, 1); + tcg_gen_neg_i64(tmp, tmp); + + /* Apply to either copy the source, or write zeros. */ + tcg_gen_gvec_ands(MO_64, pred_full_reg_offset(s, a->pd), + pred_full_reg_offset(s, a->pn), tmp, pl, pl); + + tcg_temp_free_i64(tmp); + tcg_temp_free_i64(dbit); + tcg_temp_free_i64(didx); + tcg_temp_free_ptr(ptr); + return true; +} -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> This is an SVE instruction that operates using the SVE vector length but that it is present only if SME is implemented. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-30-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper-sve.h | 2 ++ target/arm/sve.decode | 1 + target/arm/sve_helper.c | 16 ++++++++++++++++ target/arm/translate-sve.c | 2 ++ 4 files changed, 21 insertions(+) diff --git a/target/arm/helper-sve.h b/target/arm/helper-sve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-sve.h +++ b/target/arm/helper-sve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(sve_revh_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_4(sve_revw_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(sme_revd_q, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) + DEF_HELPER_FLAGS_4(sve_rbit_b, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_4(sve_rbit_h, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_4(sve_rbit_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) diff --git a/target/arm/sve.decode b/target/arm/sve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sve.decode +++ b/target/arm/sve.decode @@ -XXX,XX +XXX,XX @@ REVB 00000101 .. 1001 00 100 ... ..... ..... @rd_pg_rn REVH 00000101 .. 1001 01 100 ... ..... ..... @rd_pg_rn REVW 00000101 .. 1001 10 100 ... ..... ..... @rd_pg_rn RBIT 00000101 .. 1001 11 100 ... ..... ..... @rd_pg_rn +REVD 00000101 00 1011 10 100 ... ..... ..... @rd_pg_rn_e0 # SVE vector splice (predicated, destructive) SPLICE 00000101 .. 101 100 100 ... ..... ..... @rdn_pg_rm diff --git a/target/arm/sve_helper.c b/target/arm/sve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sve_helper.c +++ b/target/arm/sve_helper.c @@ -XXX,XX +XXX,XX @@ DO_ZPZ_D(sve_revh_d, uint64_t, hswap64) DO_ZPZ_D(sve_revw_d, uint64_t, wswap64) +void HELPER(sme_revd_q)(void *vd, void *vn, void *vg, uint32_t desc) +{ + intptr_t i, opr_sz = simd_oprsz(desc) / 8; + uint64_t *d = vd, *n = vn; + uint8_t *pg = vg; + + for (i = 0; i < opr_sz; i += 2) { + if (pg[H1(i)] & 1) { + uint64_t n0 = n[i + 0]; + uint64_t n1 = n[i + 1]; + d[i + 0] = n1; + d[i + 1] = n0; + } + } +} + DO_ZPZ(sve_rbit_b, uint8_t, H1, revbit8) DO_ZPZ(sve_rbit_h, uint16_t, H1_2, revbit16) DO_ZPZ(sve_rbit_s, uint32_t, H1_4, revbit32) diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ TRANS_FEAT(REVH, aa64_sve, gen_gvec_ool_arg_zpz, revh_fns[a->esz], a, 0) TRANS_FEAT(REVW, aa64_sve, gen_gvec_ool_arg_zpz, a->esz == 3 ? gen_helper_sve_revw_d : NULL, a, 0) +TRANS_FEAT(REVD, aa64_sme, gen_gvec_ool_arg_zpz, gen_helper_sme_revd_q, a, 0) + TRANS_FEAT(SPLICE, aa64_sve, gen_gvec_ool_arg_zpzz, gen_helper_sve_splice, a, a->esz) -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> This is an SVE instruction that operates using the SVE vector length but that it is present only if SME is implemented. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-31-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper.h | 18 +++++++ target/arm/sve.decode | 5 ++ target/arm/translate-sve.c | 102 +++++++++++++++++++++++++++++++++++++ target/arm/vec_helper.c | 24 +++++++++ 4 files changed, 149 insertions(+) diff --git a/target/arm/helper.h b/target/arm/helper.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper.h +++ b/target/arm/helper.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_6(gvec_bfmlal, TCG_CALL_NO_RWG, DEF_HELPER_FLAGS_6(gvec_bfmlal_idx, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_5(gvec_sclamp_b, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_5(gvec_sclamp_h, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_5(gvec_sclamp_s, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_5(gvec_sclamp_d, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, i32) + +DEF_HELPER_FLAGS_5(gvec_uclamp_b, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_5(gvec_uclamp_h, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_5(gvec_uclamp_s, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_5(gvec_uclamp_d, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, i32) + #ifdef TARGET_AARCH64 #include "helper-a64.h" #include "helper-sve.h" diff --git a/target/arm/sve.decode b/target/arm/sve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sve.decode +++ b/target/arm/sve.decode @@ -XXX,XX +XXX,XX @@ PSEL 00100101 .. 1 100 .. 01 .... 0 .... 0 .... \ @psel esz=2 imm=%psel_imm_s PSEL 00100101 .1 1 000 .. 01 .... 0 .... 0 .... \ @psel esz=3 imm=%psel_imm_d + +### SVE clamp + +SCLAMP 01000100 .. 0 ..... 110000 ..... ..... @rda_rn_rm +UCLAMP 01000100 .. 0 ..... 110001 ..... ..... @rda_rn_rm diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static bool trans_PSEL(DisasContext *s, arg_psel *a) tcg_temp_free_ptr(ptr); return true; } + +static void gen_sclamp_i32(TCGv_i32 d, TCGv_i32 n, TCGv_i32 m, TCGv_i32 a) +{ + tcg_gen_smax_i32(d, a, n); + tcg_gen_smin_i32(d, d, m); +} + +static void gen_sclamp_i64(TCGv_i64 d, TCGv_i64 n, TCGv_i64 m, TCGv_i64 a) +{ + tcg_gen_smax_i64(d, a, n); + tcg_gen_smin_i64(d, d, m); +} + +static void gen_sclamp_vec(unsigned vece, TCGv_vec d, TCGv_vec n, + TCGv_vec m, TCGv_vec a) +{ + tcg_gen_smax_vec(vece, d, a, n); + tcg_gen_smin_vec(vece, d, d, m); +} + +static void gen_sclamp(unsigned vece, uint32_t d, uint32_t n, uint32_t m, + uint32_t a, uint32_t oprsz, uint32_t maxsz) +{ + static const TCGOpcode vecop[] = { + INDEX_op_smin_vec, INDEX_op_smax_vec, 0 + }; + static const GVecGen4 ops[4] = { + { .fniv = gen_sclamp_vec, + .fno = gen_helper_gvec_sclamp_b, + .opt_opc = vecop, + .vece = MO_8 }, + { .fniv = gen_sclamp_vec, + .fno = gen_helper_gvec_sclamp_h, + .opt_opc = vecop, + .vece = MO_16 }, + { .fni4 = gen_sclamp_i32, + .fniv = gen_sclamp_vec, + .fno = gen_helper_gvec_sclamp_s, + .opt_opc = vecop, + .vece = MO_32 }, + { .fni8 = gen_sclamp_i64, + .fniv = gen_sclamp_vec, + .fno = gen_helper_gvec_sclamp_d, + .opt_opc = vecop, + .vece = MO_64, + .prefer_i64 = TCG_TARGET_REG_BITS == 64 } + }; + tcg_gen_gvec_4(d, n, m, a, oprsz, maxsz, &ops[vece]); +} + +TRANS_FEAT(SCLAMP, aa64_sme, gen_gvec_fn_arg_zzzz, gen_sclamp, a) + +static void gen_uclamp_i32(TCGv_i32 d, TCGv_i32 n, TCGv_i32 m, TCGv_i32 a) +{ + tcg_gen_umax_i32(d, a, n); + tcg_gen_umin_i32(d, d, m); +} + +static void gen_uclamp_i64(TCGv_i64 d, TCGv_i64 n, TCGv_i64 m, TCGv_i64 a) +{ + tcg_gen_umax_i64(d, a, n); + tcg_gen_umin_i64(d, d, m); +} + +static void gen_uclamp_vec(unsigned vece, TCGv_vec d, TCGv_vec n, + TCGv_vec m, TCGv_vec a) +{ + tcg_gen_umax_vec(vece, d, a, n); + tcg_gen_umin_vec(vece, d, d, m); +} + +static void gen_uclamp(unsigned vece, uint32_t d, uint32_t n, uint32_t m, + uint32_t a, uint32_t oprsz, uint32_t maxsz) +{ + static const TCGOpcode vecop[] = { + INDEX_op_umin_vec, INDEX_op_umax_vec, 0 + }; + static const GVecGen4 ops[4] = { + { .fniv = gen_uclamp_vec, + .fno = gen_helper_gvec_uclamp_b, + .opt_opc = vecop, + .vece = MO_8 }, + { .fniv = gen_uclamp_vec, + .fno = gen_helper_gvec_uclamp_h, + .opt_opc = vecop, + .vece = MO_16 }, + { .fni4 = gen_uclamp_i32, + .fniv = gen_uclamp_vec, + .fno = gen_helper_gvec_uclamp_s, + .opt_opc = vecop, + .vece = MO_32 }, + { .fni8 = gen_uclamp_i64, + .fniv = gen_uclamp_vec, + .fno = gen_helper_gvec_uclamp_d, + .opt_opc = vecop, + .vece = MO_64, + .prefer_i64 = TCG_TARGET_REG_BITS == 64 } + }; + tcg_gen_gvec_4(d, n, m, a, oprsz, maxsz, &ops[vece]); +} + +TRANS_FEAT(UCLAMP, aa64_sme, gen_gvec_fn_arg_zzzz, gen_uclamp, a) diff --git a/target/arm/vec_helper.c b/target/arm/vec_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/vec_helper.c +++ b/target/arm/vec_helper.c @@ -XXX,XX +XXX,XX @@ void HELPER(gvec_bfmlal_idx)(void *vd, void *vn, void *vm, } clear_tail(d, opr_sz, simd_maxsz(desc)); } + +#define DO_CLAMP(NAME, TYPE) \ +void HELPER(NAME)(void *d, void *n, void *m, void *a, uint32_t desc) \ +{ \ + intptr_t i, opr_sz = simd_oprsz(desc); \ + for (i = 0; i < opr_sz; i += sizeof(TYPE)) { \ + TYPE aa = *(TYPE *)(a + i); \ + TYPE nn = *(TYPE *)(n + i); \ + TYPE mm = *(TYPE *)(m + i); \ + TYPE dd = MIN(MAX(aa, nn), mm); \ + *(TYPE *)(d + i) = dd; \ + } \ + clear_tail(d, opr_sz, simd_maxsz(desc)); \ +} + +DO_CLAMP(gvec_sclamp_b, int8_t) +DO_CLAMP(gvec_sclamp_h, int16_t) +DO_CLAMP(gvec_sclamp_s, int32_t) +DO_CLAMP(gvec_sclamp_d, int64_t) + +DO_CLAMP(gvec_uclamp_b, uint8_t) +DO_CLAMP(gvec_uclamp_h, uint16_t) +DO_CLAMP(gvec_uclamp_s, uint32_t) +DO_CLAMP(gvec_uclamp_d, uint64_t) -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> We can handle both exception entry and exception return by hooking into aarch64_sve_change_el. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-32-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper.c | 15 +++++++++++++-- 1 file changed, 13 insertions(+), 2 deletions(-) diff --git a/target/arm/helper.c b/target/arm/helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper.c +++ b/target/arm/helper.c @@ -XXX,XX +XXX,XX @@ void aarch64_sve_change_el(CPUARMState *env, int old_el, return; } + old_a64 = old_el ? arm_el_is_aa64(env, old_el) : el0_a64; + new_a64 = new_el ? arm_el_is_aa64(env, new_el) : el0_a64; + + /* + * Both AArch64.TakeException and AArch64.ExceptionReturn + * invoke ResetSVEState when taking an exception from, or + * returning to, AArch32 state when PSTATE.SM is enabled. + */ + if (old_a64 != new_a64 && FIELD_EX64(env->svcr, SVCR, SM)) { + arm_reset_sve_state(env); + return; + } + /* * DDI0584A.d sec 3.2: "If SVE instructions are disabled or trapped * at ELx, or not available because the EL is in AArch32 state, then @@ -XXX,XX +XXX,XX @@ void aarch64_sve_change_el(CPUARMState *env, int old_el, * we already have the correct register contents when encountering the * vq0->vq0 transition between EL0->EL1. */ - old_a64 = old_el ? arm_el_is_aa64(env, old_el) : el0_a64; old_len = (old_a64 && !sve_exception_el(env, old_el) ? sve_vqm1_for_el(env, old_el) : 0); - new_a64 = new_el ? arm_el_is_aa64(env, new_el) : el0_a64; new_len = (new_a64 && !sve_exception_el(env, new_el) ? sve_vqm1_for_el(env, new_el) : 0); -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Note that SME remains effectively disabled for user-only, because we do not yet set CPACR_EL1.SMEN. This needs to wait until the kernel ABI is implemented. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-33-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- docs/system/arm/emulation.rst | 4 ++++ target/arm/cpu64.c | 11 +++++++++++ 2 files changed, 15 insertions(+) diff --git a/docs/system/arm/emulation.rst b/docs/system/arm/emulation.rst index XXXXXXX..XXXXXXX 100644 --- a/docs/system/arm/emulation.rst +++ b/docs/system/arm/emulation.rst @@ -XXX,XX +XXX,XX @@ the following architecture extensions: - FEAT_SHA512 (Advanced SIMD SHA512 instructions) - FEAT_SM3 (Advanced SIMD SM3 instructions) - FEAT_SM4 (Advanced SIMD SM4 instructions) +- FEAT_SME (Scalable Matrix Extension) +- FEAT_SME_FA64 (Full A64 instruction set in Streaming SVE mode) +- FEAT_SME_F64F64 (Double-precision floating-point outer product instructions) +- FEAT_SME_I16I64 (16-bit to 64-bit integer widening outer product instructions) - FEAT_SPECRES (Speculation restriction instructions) - FEAT_SSBS (Speculative Store Bypass Safe) - FEAT_TLBIOS (TLB invalidate instructions in Outer Shareable domain) diff --git a/target/arm/cpu64.c b/target/arm/cpu64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu64.c +++ b/target/arm/cpu64.c @@ -XXX,XX +XXX,XX @@ static void aarch64_max_initfn(Object *obj) */ t = FIELD_DP64(t, ID_AA64PFR1, MTE, 3); /* FEAT_MTE3 */ t = FIELD_DP64(t, ID_AA64PFR1, RAS_FRAC, 0); /* FEAT_RASv1p1 + FEAT_DoubleFault */ + t = FIELD_DP64(t, ID_AA64PFR1, SME, 1); /* FEAT_SME */ t = FIELD_DP64(t, ID_AA64PFR1, CSV2_FRAC, 0); /* FEAT_CSV2_2 */ cpu->isar.id_aa64pfr1 = t; @@ -XXX,XX +XXX,XX @@ static void aarch64_max_initfn(Object *obj) t = FIELD_DP64(t, ID_AA64DFR0, PMUVER, 5); /* FEAT_PMUv3p4 */ cpu->isar.id_aa64dfr0 = t; + t = cpu->isar.id_aa64smfr0; + t = FIELD_DP64(t, ID_AA64SMFR0, F32F32, 1); /* FEAT_SME */ + t = FIELD_DP64(t, ID_AA64SMFR0, B16F32, 1); /* FEAT_SME */ + t = FIELD_DP64(t, ID_AA64SMFR0, F16F32, 1); /* FEAT_SME */ + t = FIELD_DP64(t, ID_AA64SMFR0, I8I32, 0xf); /* FEAT_SME */ + t = FIELD_DP64(t, ID_AA64SMFR0, F64F64, 1); /* FEAT_SME_F64F64 */ + t = FIELD_DP64(t, ID_AA64SMFR0, I16I64, 0xf); /* FEAT_SME_I16I64 */ + t = FIELD_DP64(t, ID_AA64SMFR0, FA64, 1); /* FEAT_SME_FA64 */ + cpu->isar.id_aa64smfr0 = t; + /* Replicate the same data to the 32-bit id registers. */ aa32_max_features(cpu); -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-34-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/aarch64/target_cpu.h | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/linux-user/aarch64/target_cpu.h b/linux-user/aarch64/target_cpu.h index XXXXXXX..XXXXXXX 100644 --- a/linux-user/aarch64/target_cpu.h +++ b/linux-user/aarch64/target_cpu.h @@ -XXX,XX +XXX,XX @@ static inline void cpu_clone_regs_parent(CPUARMState *env, unsigned flags) static inline void cpu_set_tls(CPUARMState *env, target_ulong newtls) { - /* Note that AArch64 Linux keeps the TLS pointer in TPIDR; this is + /* + * Note that AArch64 Linux keeps the TLS pointer in TPIDR; this is * different from AArch32 Linux, which uses TPIDRRO. */ env->cp15.tpidr_el[0] = newtls; + /* TPIDR2_EL0 is cleared with CLONE_SETTLS. */ + env->cp15.tpidr2_el0 = 0; } static inline abi_ulong get_sp_from_cpustate(CPUARMState *state) -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-35-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/aarch64/cpu_loop.c | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/linux-user/aarch64/cpu_loop.c b/linux-user/aarch64/cpu_loop.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/aarch64/cpu_loop.c +++ b/linux-user/aarch64/cpu_loop.c @@ -XXX,XX +XXX,XX @@ void cpu_loop(CPUARMState *env) switch (trapnr) { case EXCP_SWI: + /* + * On syscall, PSTATE.ZA is preserved, along with the ZA matrix. + * PSTATE.SM is cleared, per SMSTOP, which does ResetSVEState. + */ + if (FIELD_EX64(env->svcr, SVCR, SM)) { + env->svcr = FIELD_DP64(env->svcr, SVCR, SM, 0); + arm_rebuild_hflags(env); + arm_reset_sve_state(env); + } ret = do_syscall(env, env->xregs[8], env->xregs[0], -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Make sure to zero the currently reserved fields. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-36-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/aarch64/signal.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) diff --git a/linux-user/aarch64/signal.c b/linux-user/aarch64/signal.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/aarch64/signal.c +++ b/linux-user/aarch64/signal.c @@ -XXX,XX +XXX,XX @@ struct target_extra_context { struct target_sve_context { struct target_aarch64_ctx head; uint16_t vl; - uint16_t reserved[3]; + uint16_t flags; + uint16_t reserved[2]; /* The actual SVE data immediately follows. It is laid out * according to TARGET_SVE_SIG_{Z,P}REG_OFFSET, based off of * the original struct pointer. @@ -XXX,XX +XXX,XX @@ struct target_sve_context { #define TARGET_SVE_SIG_CONTEXT_SIZE(VQ) \ (TARGET_SVE_SIG_PREG_OFFSET(VQ, 17)) +#define TARGET_SVE_SIG_FLAG_SM 1 + struct target_rt_sigframe { struct target_siginfo info; struct target_ucontext uc; @@ -XXX,XX +XXX,XX @@ static void target_setup_sve_record(struct target_sve_context *sve, { int i, j; + memset(sve, 0, sizeof(*sve)); __put_user(TARGET_SVE_MAGIC, &sve->head.magic); __put_user(size, &sve->head.size); __put_user(vq * TARGET_SVE_VQ_BYTES, &sve->vl); + if (FIELD_EX64(env->svcr, SVCR, SM)) { + __put_user(TARGET_SVE_SIG_FLAG_SM, &sve->flags); + } /* Note that SVE regs are stored as a byte stream, with each byte element * at a subsequent address. This corresponds to a little-endian store -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Fold the return value setting into the goto, so each point of failure need not do both. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-37-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/aarch64/signal.c | 26 +++++++++++--------------- 1 file changed, 11 insertions(+), 15 deletions(-) diff --git a/linux-user/aarch64/signal.c b/linux-user/aarch64/signal.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/aarch64/signal.c +++ b/linux-user/aarch64/signal.c @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, struct target_sve_context *sve = NULL; uint64_t extra_datap = 0; bool used_extra = false; - bool err = false; int vq = 0, sve_size = 0; target_restore_general_frame(env, sf); @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, switch (magic) { case 0: if (size != 0) { - err = true; - goto exit; + goto err; } if (used_extra) { ctx = NULL; @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, case TARGET_FPSIMD_MAGIC: if (fpsimd || size != sizeof(struct target_fpsimd_context)) { - err = true; - goto exit; + goto err; } fpsimd = (struct target_fpsimd_context *)ctx; break; @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, break; } } - err = true; - goto exit; + goto err; case TARGET_EXTRA_MAGIC: if (extra || size != sizeof(struct target_extra_context)) { - err = true; - goto exit; + goto err; } __get_user(extra_datap, &((struct target_extra_context *)ctx)->datap); @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, /* Unknown record -- we certainly didn't generate it. * Did we in fact get out of sync? */ - err = true; - goto exit; + goto err; } ctx = (void *)ctx + size; } @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, if (fpsimd) { target_restore_fpsimd_record(env, fpsimd); } else { - err = true; + goto err; } /* SVE data, if present, overwrites FPSIMD data. */ if (sve) { target_restore_sve_record(env, sve, vq); } - - exit: unlock_user(extra, extra_datap, 0); - return err; + return 0; + + err: + unlock_user(extra, extra_datap, 0); + return 1; } static abi_ulong get_sigframe(struct target_sigaction *ka, -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> In parse_user_sigframe, the kernel rejects duplicate sve records, or records that are smaller than the header. We were silently allowing these cases to pass, dropping the record. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-38-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/aarch64/signal.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/linux-user/aarch64/signal.c b/linux-user/aarch64/signal.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/aarch64/signal.c +++ b/linux-user/aarch64/signal.c @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, break; case TARGET_SVE_MAGIC: + if (sve || size < sizeof(struct target_sve_context)) { + goto err; + } if (cpu_isar_feature(aa64_sve, env_archcpu(env))) { vq = sve_vq(env); sve_size = QEMU_ALIGN_UP(TARGET_SVE_SIG_CONTEXT_SIZE(vq), 16); - if (!sve && size == sve_size) { + if (size == sve_size) { sve = (struct target_sve_context *)ctx; break; } -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-39-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/aarch64/signal.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/linux-user/aarch64/signal.c b/linux-user/aarch64/signal.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/aarch64/signal.c +++ b/linux-user/aarch64/signal.c @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, __get_user(extra_size, &((struct target_extra_context *)ctx)->size); extra = lock_user(VERIFY_READ, extra_datap, extra_size, 0); + if (!extra) { + return 1; + } break; default: -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Move the checks out of the parsing loop and into the restore function. This more closely mirrors the code structure in the kernel, and is slightly clearer. Reject rather than silently skip incorrect VL and SVE record sizes, bringing our checks in to line with those the kernel does. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-40-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/aarch64/signal.c | 51 +++++++++++++++++++++++++------------ 1 file changed, 35 insertions(+), 16 deletions(-) diff --git a/linux-user/aarch64/signal.c b/linux-user/aarch64/signal.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/aarch64/signal.c +++ b/linux-user/aarch64/signal.c @@ -XXX,XX +XXX,XX @@ static void target_restore_fpsimd_record(CPUARMState *env, } } -static void target_restore_sve_record(CPUARMState *env, - struct target_sve_context *sve, int vq) +static bool target_restore_sve_record(CPUARMState *env, + struct target_sve_context *sve, + int size) { - int i, j; + int i, j, vl, vq; - /* Note that SVE regs are stored as a byte stream, with each byte element + if (!cpu_isar_feature(aa64_sve, env_archcpu(env))) { + return false; + } + + __get_user(vl, &sve->vl); + vq = sve_vq(env); + + /* Reject mismatched VL. */ + if (vl != vq * TARGET_SVE_VQ_BYTES) { + return false; + } + + /* Accept empty record -- used to clear PSTATE.SM. */ + if (size <= sizeof(*sve)) { + return true; + } + + /* Reject non-empty but incomplete record. */ + if (size < TARGET_SVE_SIG_CONTEXT_SIZE(vq)) { + return false; + } + + /* + * Note that SVE regs are stored as a byte stream, with each byte element * at a subsequent address. This corresponds to a little-endian load * of our 64-bit hunks. */ @@ -XXX,XX +XXX,XX @@ static void target_restore_sve_record(CPUARMState *env, } } } + return true; } static int target_restore_sigframe(CPUARMState *env, @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, struct target_sve_context *sve = NULL; uint64_t extra_datap = 0; bool used_extra = false; - int vq = 0, sve_size = 0; + int sve_size = 0; target_restore_general_frame(env, sf); @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, if (sve || size < sizeof(struct target_sve_context)) { goto err; } - if (cpu_isar_feature(aa64_sve, env_archcpu(env))) { - vq = sve_vq(env); - sve_size = QEMU_ALIGN_UP(TARGET_SVE_SIG_CONTEXT_SIZE(vq), 16); - if (size == sve_size) { - sve = (struct target_sve_context *)ctx; - break; - } - } - goto err; + sve = (struct target_sve_context *)ctx; + sve_size = size; + break; case TARGET_EXTRA_MAGIC: if (extra || size != sizeof(struct target_extra_context)) { @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, } /* SVE data, if present, overwrites FPSIMD data. */ - if (sve) { - target_restore_sve_record(env, sve, vq); + if (sve && !target_restore_sve_record(env, sve, sve_size)) { + goto err; } unlock_user(extra, extra_datap, 0); return 0; -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Set the SM bit in the SVE record on signal delivery, create the ZA record. Restore SM and ZA state according to the records present on return. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-41-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/aarch64/signal.c | 167 +++++++++++++++++++++++++++++++++--- 1 file changed, 154 insertions(+), 13 deletions(-) diff --git a/linux-user/aarch64/signal.c b/linux-user/aarch64/signal.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/aarch64/signal.c +++ b/linux-user/aarch64/signal.c @@ -XXX,XX +XXX,XX @@ struct target_sve_context { #define TARGET_SVE_SIG_FLAG_SM 1 +#define TARGET_ZA_MAGIC 0x54366345 + +struct target_za_context { + struct target_aarch64_ctx head; + uint16_t vl; + uint16_t reserved[3]; + /* The actual ZA data immediately follows. */ +}; + +#define TARGET_ZA_SIG_REGS_OFFSET \ + QEMU_ALIGN_UP(sizeof(struct target_za_context), TARGET_SVE_VQ_BYTES) +#define TARGET_ZA_SIG_ZAV_OFFSET(VQ, N) \ + (TARGET_ZA_SIG_REGS_OFFSET + (VQ) * TARGET_SVE_VQ_BYTES * (N)) +#define TARGET_ZA_SIG_CONTEXT_SIZE(VQ) \ + TARGET_ZA_SIG_ZAV_OFFSET(VQ, VQ * TARGET_SVE_VQ_BYTES) + struct target_rt_sigframe { struct target_siginfo info; struct target_ucontext uc; @@ -XXX,XX +XXX,XX @@ static void target_setup_end_record(struct target_aarch64_ctx *end) } static void target_setup_sve_record(struct target_sve_context *sve, - CPUARMState *env, int vq, int size) + CPUARMState *env, int size) { - int i, j; + int i, j, vq = sve_vq(env); memset(sve, 0, sizeof(*sve)); __put_user(TARGET_SVE_MAGIC, &sve->head.magic); @@ -XXX,XX +XXX,XX @@ static void target_setup_sve_record(struct target_sve_context *sve, } } +static void target_setup_za_record(struct target_za_context *za, + CPUARMState *env, int size) +{ + int vq = sme_vq(env); + int vl = vq * TARGET_SVE_VQ_BYTES; + int i, j; + + memset(za, 0, sizeof(*za)); + __put_user(TARGET_ZA_MAGIC, &za->head.magic); + __put_user(size, &za->head.size); + __put_user(vl, &za->vl); + + if (size == TARGET_ZA_SIG_CONTEXT_SIZE(0)) { + return; + } + assert(size == TARGET_ZA_SIG_CONTEXT_SIZE(vq)); + + /* + * Note that ZA vectors are stored as a byte stream, + * with each byte element at a subsequent address. + */ + for (i = 0; i < vl; ++i) { + uint64_t *z = (void *)za + TARGET_ZA_SIG_ZAV_OFFSET(vq, i); + for (j = 0; j < vq * 2; ++j) { + __put_user_e(env->zarray[i].d[j], z + j, le); + } + } +} + static void target_restore_general_frame(CPUARMState *env, struct target_rt_sigframe *sf) { @@ -XXX,XX +XXX,XX @@ static void target_restore_fpsimd_record(CPUARMState *env, static bool target_restore_sve_record(CPUARMState *env, struct target_sve_context *sve, - int size) + int size, int *svcr) { - int i, j, vl, vq; + int i, j, vl, vq, flags; + bool sm; - if (!cpu_isar_feature(aa64_sve, env_archcpu(env))) { + __get_user(vl, &sve->vl); + __get_user(flags, &sve->flags); + + sm = flags & TARGET_SVE_SIG_FLAG_SM; + + /* The cpu must support Streaming or Non-streaming SVE. */ + if (sm + ? !cpu_isar_feature(aa64_sme, env_archcpu(env)) + : !cpu_isar_feature(aa64_sve, env_archcpu(env))) { return false; } - __get_user(vl, &sve->vl); - vq = sve_vq(env); + /* + * Note that we cannot use sve_vq() because that depends on the + * current setting of PSTATE.SM, not the state to be restored. + */ + vq = sve_vqm1_for_el_sm(env, 0, sm) + 1; /* Reject mismatched VL. */ if (vl != vq * TARGET_SVE_VQ_BYTES) { @@ -XXX,XX +XXX,XX @@ static bool target_restore_sve_record(CPUARMState *env, return false; } + *svcr = FIELD_DP64(*svcr, SVCR, SM, sm); + /* * Note that SVE regs are stored as a byte stream, with each byte element * at a subsequent address. This corresponds to a little-endian load @@ -XXX,XX +XXX,XX @@ static bool target_restore_sve_record(CPUARMState *env, return true; } +static bool target_restore_za_record(CPUARMState *env, + struct target_za_context *za, + int size, int *svcr) +{ + int i, j, vl, vq; + + if (!cpu_isar_feature(aa64_sme, env_archcpu(env))) { + return false; + } + + __get_user(vl, &za->vl); + vq = sme_vq(env); + + /* Reject mismatched VL. */ + if (vl != vq * TARGET_SVE_VQ_BYTES) { + return false; + } + + /* Accept empty record -- used to clear PSTATE.ZA. */ + if (size <= TARGET_ZA_SIG_CONTEXT_SIZE(0)) { + return true; + } + + /* Reject non-empty but incomplete record. */ + if (size < TARGET_ZA_SIG_CONTEXT_SIZE(vq)) { + return false; + } + + *svcr = FIELD_DP64(*svcr, SVCR, ZA, 1); + + for (i = 0; i < vl; ++i) { + uint64_t *z = (void *)za + TARGET_ZA_SIG_ZAV_OFFSET(vq, i); + for (j = 0; j < vq * 2; ++j) { + __get_user_e(env->zarray[i].d[j], z + j, le); + } + } + return true; +} + static int target_restore_sigframe(CPUARMState *env, struct target_rt_sigframe *sf) { struct target_aarch64_ctx *ctx, *extra = NULL; struct target_fpsimd_context *fpsimd = NULL; struct target_sve_context *sve = NULL; + struct target_za_context *za = NULL; uint64_t extra_datap = 0; bool used_extra = false; int sve_size = 0; + int za_size = 0; + int svcr = 0; target_restore_general_frame(env, sf); @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, sve_size = size; break; + case TARGET_ZA_MAGIC: + if (za || size < sizeof(struct target_za_context)) { + goto err; + } + za = (struct target_za_context *)ctx; + za_size = size; + break; + case TARGET_EXTRA_MAGIC: if (extra || size != sizeof(struct target_extra_context)) { goto err; @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, } /* SVE data, if present, overwrites FPSIMD data. */ - if (sve && !target_restore_sve_record(env, sve, sve_size)) { + if (sve && !target_restore_sve_record(env, sve, sve_size, &svcr)) { goto err; } + if (za && !target_restore_za_record(env, za, za_size, &svcr)) { + goto err; + } + if (env->svcr != svcr) { + env->svcr = svcr; + arm_rebuild_hflags(env); + } unlock_user(extra, extra_datap, 0); return 0; @@ -XXX,XX +XXX,XX @@ static void target_setup_frame(int usig, struct target_sigaction *ka, .total_size = offsetof(struct target_rt_sigframe, uc.tuc_mcontext.__reserved), }; - int fpsimd_ofs, fr_ofs, sve_ofs = 0, vq = 0, sve_size = 0; + int fpsimd_ofs, fr_ofs, sve_ofs = 0, za_ofs = 0; + int sve_size = 0, za_size = 0; struct target_rt_sigframe *frame; struct target_rt_frame_record *fr; abi_ulong frame_addr, return_addr; @@ -XXX,XX +XXX,XX @@ static void target_setup_frame(int usig, struct target_sigaction *ka, &layout); /* SVE state needs saving only if it exists. */ - if (cpu_isar_feature(aa64_sve, env_archcpu(env))) { - vq = sve_vq(env); - sve_size = QEMU_ALIGN_UP(TARGET_SVE_SIG_CONTEXT_SIZE(vq), 16); + if (cpu_isar_feature(aa64_sve, env_archcpu(env)) || + cpu_isar_feature(aa64_sme, env_archcpu(env))) { + sve_size = QEMU_ALIGN_UP(TARGET_SVE_SIG_CONTEXT_SIZE(sve_vq(env)), 16); sve_ofs = alloc_sigframe_space(sve_size, &layout); } + if (cpu_isar_feature(aa64_sme, env_archcpu(env))) { + /* ZA state needs saving only if it is enabled. */ + if (FIELD_EX64(env->svcr, SVCR, ZA)) { + za_size = TARGET_ZA_SIG_CONTEXT_SIZE(sme_vq(env)); + } else { + za_size = TARGET_ZA_SIG_CONTEXT_SIZE(0); + } + za_ofs = alloc_sigframe_space(za_size, &layout); + } if (layout.extra_ofs) { /* Reserve space for the extra end marker. The standard end marker @@ -XXX,XX +XXX,XX @@ static void target_setup_frame(int usig, struct target_sigaction *ka, target_setup_end_record((void *)frame + layout.extra_end_ofs); } if (sve_ofs) { - target_setup_sve_record((void *)frame + sve_ofs, env, vq, sve_size); + target_setup_sve_record((void *)frame + sve_ofs, env, sve_size); + } + if (za_ofs) { + target_setup_za_record((void *)frame + za_ofs, env, za_size); } /* Set up the stack frame for unwinding. */ @@ -XXX,XX +XXX,XX @@ static void target_setup_frame(int usig, struct target_sigaction *ka, env->btype = 2; } + /* + * Invoke the signal handler with both SM and ZA disabled. + * When clearing SM, ResetSVEState, per SMSTOP. + */ + if (FIELD_EX64(env->svcr, SVCR, SM)) { + arm_reset_sve_state(env); + } + if (env->svcr) { + env->svcr = 0; + arm_rebuild_hflags(env); + } + if (info) { tswap_siginfo(&frame->info, info); env->xregs[1] = frame_addr + offsetof(struct target_rt_sigframe, info); -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Add "sve" to the sve prctl functions, to distinguish them from the coming "sme" prctls with similar names. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-42-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/aarch64/target_prctl.h | 8 ++++---- linux-user/syscall.c | 12 ++++++------ 2 files changed, 10 insertions(+), 10 deletions(-) diff --git a/linux-user/aarch64/target_prctl.h b/linux-user/aarch64/target_prctl.h index XXXXXXX..XXXXXXX 100644 --- a/linux-user/aarch64/target_prctl.h +++ b/linux-user/aarch64/target_prctl.h @@ -XXX,XX +XXX,XX @@ #ifndef AARCH64_TARGET_PRCTL_H #define AARCH64_TARGET_PRCTL_H -static abi_long do_prctl_get_vl(CPUArchState *env) +static abi_long do_prctl_sve_get_vl(CPUArchState *env) { ARMCPU *cpu = env_archcpu(env); if (cpu_isar_feature(aa64_sve, cpu)) { @@ -XXX,XX +XXX,XX @@ static abi_long do_prctl_get_vl(CPUArchState *env) } return -TARGET_EINVAL; } -#define do_prctl_get_vl do_prctl_get_vl +#define do_prctl_sve_get_vl do_prctl_sve_get_vl -static abi_long do_prctl_set_vl(CPUArchState *env, abi_long arg2) +static abi_long do_prctl_sve_set_vl(CPUArchState *env, abi_long arg2) { /* * We cannot support either PR_SVE_SET_VL_ONEXEC or PR_SVE_VL_INHERIT. @@ -XXX,XX +XXX,XX @@ static abi_long do_prctl_set_vl(CPUArchState *env, abi_long arg2) } return -TARGET_EINVAL; } -#define do_prctl_set_vl do_prctl_set_vl +#define do_prctl_sve_set_vl do_prctl_sve_set_vl static abi_long do_prctl_reset_keys(CPUArchState *env, abi_long arg2) { diff --git a/linux-user/syscall.c b/linux-user/syscall.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/syscall.c +++ b/linux-user/syscall.c @@ -XXX,XX +XXX,XX @@ static abi_long do_prctl_inval1(CPUArchState *env, abi_long arg2) #ifndef do_prctl_set_fp_mode #define do_prctl_set_fp_mode do_prctl_inval1 #endif -#ifndef do_prctl_get_vl -#define do_prctl_get_vl do_prctl_inval0 +#ifndef do_prctl_sve_get_vl +#define do_prctl_sve_get_vl do_prctl_inval0 #endif -#ifndef do_prctl_set_vl -#define do_prctl_set_vl do_prctl_inval1 +#ifndef do_prctl_sve_set_vl +#define do_prctl_sve_set_vl do_prctl_inval1 #endif #ifndef do_prctl_reset_keys #define do_prctl_reset_keys do_prctl_inval1 @@ -XXX,XX +XXX,XX @@ static abi_long do_prctl(CPUArchState *env, abi_long option, abi_long arg2, case PR_SET_FP_MODE: return do_prctl_set_fp_mode(env, arg2); case PR_SVE_GET_VL: - return do_prctl_get_vl(env); + return do_prctl_sve_get_vl(env); case PR_SVE_SET_VL: - return do_prctl_set_vl(env, arg2); + return do_prctl_sve_set_vl(env, arg2); case PR_PAC_RESET_KEYS: if (arg3 || arg4 || arg5) { return -TARGET_EINVAL; -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> These prctl set the Streaming SVE vector length, which may be completely different from the Normal SVE vector length. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-43-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/aarch64/target_prctl.h | 54 +++++++++++++++++++++++++++++++ linux-user/syscall.c | 16 +++++++++ 2 files changed, 70 insertions(+) diff --git a/linux-user/aarch64/target_prctl.h b/linux-user/aarch64/target_prctl.h index XXXXXXX..XXXXXXX 100644 --- a/linux-user/aarch64/target_prctl.h +++ b/linux-user/aarch64/target_prctl.h @@ -XXX,XX +XXX,XX @@ static abi_long do_prctl_sve_get_vl(CPUArchState *env) { ARMCPU *cpu = env_archcpu(env); if (cpu_isar_feature(aa64_sve, cpu)) { + /* PSTATE.SM is always unset on syscall entry. */ return sve_vq(env) * 16; } return -TARGET_EINVAL; @@ -XXX,XX +XXX,XX @@ static abi_long do_prctl_sve_set_vl(CPUArchState *env, abi_long arg2) && arg2 >= 0 && arg2 <= 512 * 16 && !(arg2 & 15)) { uint32_t vq, old_vq; + /* PSTATE.SM is always unset on syscall entry. */ old_vq = sve_vq(env); /* @@ -XXX,XX +XXX,XX @@ static abi_long do_prctl_sve_set_vl(CPUArchState *env, abi_long arg2) } #define do_prctl_sve_set_vl do_prctl_sve_set_vl +static abi_long do_prctl_sme_get_vl(CPUArchState *env) +{ + ARMCPU *cpu = env_archcpu(env); + if (cpu_isar_feature(aa64_sme, cpu)) { + return sme_vq(env) * 16; + } + return -TARGET_EINVAL; +} +#define do_prctl_sme_get_vl do_prctl_sme_get_vl + +static abi_long do_prctl_sme_set_vl(CPUArchState *env, abi_long arg2) +{ + /* + * We cannot support either PR_SME_SET_VL_ONEXEC or PR_SME_VL_INHERIT. + * Note the kernel definition of sve_vl_valid allows for VQ=512, + * i.e. VL=8192, even though the architectural maximum is VQ=16. + */ + if (cpu_isar_feature(aa64_sme, env_archcpu(env)) + && arg2 >= 0 && arg2 <= 512 * 16 && !(arg2 & 15)) { + int vq, old_vq; + + old_vq = sme_vq(env); + + /* + * Bound the value of vq, so that we know that it fits into + * the 4-bit field in SMCR_EL1. Because PSTATE.SM is cleared + * on syscall entry, we are not modifying the current SVE + * vector length. + */ + vq = MAX(arg2 / 16, 1); + vq = MIN(vq, 16); + env->vfp.smcr_el[1] = + FIELD_DP64(env->vfp.smcr_el[1], SMCR, LEN, vq - 1); + + /* Delay rebuilding hflags until we know if ZA must change. */ + vq = sve_vqm1_for_el_sm(env, 0, true) + 1; + + if (vq != old_vq) { + /* + * PSTATE.ZA state is cleared on any change to SVL. + * We need not call arm_rebuild_hflags because PSTATE.SM was + * cleared on syscall entry, so this hasn't changed VL. + */ + env->svcr = FIELD_DP64(env->svcr, SVCR, ZA, 0); + arm_rebuild_hflags(env); + } + return vq * 16; + } + return -TARGET_EINVAL; +} +#define do_prctl_sme_set_vl do_prctl_sme_set_vl + static abi_long do_prctl_reset_keys(CPUArchState *env, abi_long arg2) { ARMCPU *cpu = env_archcpu(env); diff --git a/linux-user/syscall.c b/linux-user/syscall.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/syscall.c +++ b/linux-user/syscall.c @@ -XXX,XX +XXX,XX @@ abi_long do_arch_prctl(CPUX86State *env, int code, abi_ulong addr) #ifndef PR_SET_SYSCALL_USER_DISPATCH # define PR_SET_SYSCALL_USER_DISPATCH 59 #endif +#ifndef PR_SME_SET_VL +# define PR_SME_SET_VL 63 +# define PR_SME_GET_VL 64 +# define PR_SME_VL_LEN_MASK 0xffff +# define PR_SME_VL_INHERIT (1 << 17) +#endif #include "target_prctl.h" @@ -XXX,XX +XXX,XX @@ static abi_long do_prctl_inval1(CPUArchState *env, abi_long arg2) #ifndef do_prctl_set_unalign #define do_prctl_set_unalign do_prctl_inval1 #endif +#ifndef do_prctl_sme_get_vl +#define do_prctl_sme_get_vl do_prctl_inval0 +#endif +#ifndef do_prctl_sme_set_vl +#define do_prctl_sme_set_vl do_prctl_inval1 +#endif static abi_long do_prctl(CPUArchState *env, abi_long option, abi_long arg2, abi_long arg3, abi_long arg4, abi_long arg5) @@ -XXX,XX +XXX,XX @@ static abi_long do_prctl(CPUArchState *env, abi_long option, abi_long arg2, return do_prctl_sve_get_vl(env); case PR_SVE_SET_VL: return do_prctl_sve_set_vl(env, arg2); + case PR_SME_GET_VL: + return do_prctl_sme_get_vl(env); + case PR_SME_SET_VL: + return do_prctl_sme_set_vl(env, arg2); case PR_PAC_RESET_KEYS: if (arg3 || arg4 || arg5) { return -TARGET_EINVAL; -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> There's no reason to set CPACR_EL1.ZEN if SVE disabled. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-44-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/cpu.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) diff --git a/target/arm/cpu.c b/target/arm/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.c +++ b/target/arm/cpu.c @@ -XXX,XX +XXX,XX @@ static void arm_cpu_reset(DeviceState *dev) /* and to the FP/Neon instructions */ env->cp15.cpacr_el1 = FIELD_DP64(env->cp15.cpacr_el1, CPACR_EL1, FPEN, 3); - /* and to the SVE instructions */ - env->cp15.cpacr_el1 = FIELD_DP64(env->cp15.cpacr_el1, - CPACR_EL1, ZEN, 3); - /* with reasonable vector length */ + /* and to the SVE instructions, with default vector length */ if (cpu_isar_feature(aa64_sve, cpu)) { + env->cp15.cpacr_el1 = FIELD_DP64(env->cp15.cpacr_el1, + CPACR_EL1, ZEN, 3); env->vfp.zcr_el[1] = cpu->sve_default_vq - 1; } /* -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Enable SME, TPIDR2_EL0, and FA64 if supported by the cpu. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-45-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/cpu.c | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/target/arm/cpu.c b/target/arm/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.c +++ b/target/arm/cpu.c @@ -XXX,XX +XXX,XX @@ static void arm_cpu_reset(DeviceState *dev) CPACR_EL1, ZEN, 3); env->vfp.zcr_el[1] = cpu->sve_default_vq - 1; } + /* and for SME instructions, with default vector length, and TPIDR2 */ + if (cpu_isar_feature(aa64_sme, cpu)) { + env->cp15.sctlr_el[1] |= SCTLR_EnTP2; + env->cp15.cpacr_el1 = FIELD_DP64(env->cp15.cpacr_el1, + CPACR_EL1, SMEN, 3); + env->vfp.smcr_el[1] = cpu->sme_default_vq - 1; + if (cpu_isar_feature(aa64_sme_fa64, cpu)) { + env->vfp.smcr_el[1] = FIELD_DP64(env->vfp.smcr_el[1], + SMCR, FA64, 1); + } + } /* * Enable 48-bit address space (TODO: take reserved_va into account). * Enable TBI0 but not TBI1. -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-46-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/elfload.c | 20 ++++++++++++++++++++ 1 file changed, 20 insertions(+) diff --git a/linux-user/elfload.c b/linux-user/elfload.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/elfload.c +++ b/linux-user/elfload.c @@ -XXX,XX +XXX,XX @@ enum { ARM_HWCAP2_A64_RNG = 1 << 16, ARM_HWCAP2_A64_BTI = 1 << 17, ARM_HWCAP2_A64_MTE = 1 << 18, + ARM_HWCAP2_A64_ECV = 1 << 19, + ARM_HWCAP2_A64_AFP = 1 << 20, + ARM_HWCAP2_A64_RPRES = 1 << 21, + ARM_HWCAP2_A64_MTE3 = 1 << 22, + ARM_HWCAP2_A64_SME = 1 << 23, + ARM_HWCAP2_A64_SME_I16I64 = 1 << 24, + ARM_HWCAP2_A64_SME_F64F64 = 1 << 25, + ARM_HWCAP2_A64_SME_I8I32 = 1 << 26, + ARM_HWCAP2_A64_SME_F16F32 = 1 << 27, + ARM_HWCAP2_A64_SME_B16F32 = 1 << 28, + ARM_HWCAP2_A64_SME_F32F32 = 1 << 29, + ARM_HWCAP2_A64_SME_FA64 = 1 << 30, }; #define ELF_HWCAP get_elf_hwcap() @@ -XXX,XX +XXX,XX @@ static uint32_t get_elf_hwcap2(void) GET_FEATURE_ID(aa64_rndr, ARM_HWCAP2_A64_RNG); GET_FEATURE_ID(aa64_bti, ARM_HWCAP2_A64_BTI); GET_FEATURE_ID(aa64_mte, ARM_HWCAP2_A64_MTE); + GET_FEATURE_ID(aa64_sme, (ARM_HWCAP2_A64_SME | + ARM_HWCAP2_A64_SME_F32F32 | + ARM_HWCAP2_A64_SME_B16F32 | + ARM_HWCAP2_A64_SME_F16F32 | + ARM_HWCAP2_A64_SME_I8I32)); + GET_FEATURE_ID(aa64_sme_f64f64, ARM_HWCAP2_A64_SME_F64F64); + GET_FEATURE_ID(aa64_sme_i16i64, ARM_HWCAP2_A64_SME_I16I64); + GET_FEATURE_ID(aa64_sme_fa64, ARM_HWCAP2_A64_SME_FA64); return hwcaps; } -- 2.25.1 Hi; this pullreq includes FEAT_LSE2 support, the new bpim2u board, and some other smaller patchsets. thanks -- PMM The following changes since commit 369081c4558e7e940fa36ce59bf17b2e390f55d3: Merge tag 'pull-tcg-20230605' of https://gitlab.com/rth7680/qemu into staging (2023-06-05 13:16:56 -0700) are available in the Git repository at: https://git.linaro.org/people/pmaydell/qemu-arm.git tags/pull-target-arm-20230606 for you to fetch changes up to f9ac778898cb28307e0f91421aba34d43c34b679: target/arm: trap DCC access in user mode emulation (2023-06-06 10:19:40 +0100) ---------------------------------------------------------------- target-arm queue: * Support gdbstub (guest debug) in HVF * xnlx-versal: Support CANFD controller * bpim2u: New board model: Banana Pi BPI-M2 Ultra * Emulate FEAT_LSE2 * allow DC CVA[D]P in user mode emulation * trap DCC access in user mode emulation ---------------------------------------------------------------- Francesco Cagnin (4): arm: move KVM breakpoints helpers hvf: handle access for more registers hvf: add breakpoint handlers hvf: add guest debugging handlers for Apple Silicon hosts Richard Henderson (20): target/arm: Add commentary for CPUARMState.exclusive_high target/arm: Add feature test for FEAT_LSE2 target/arm: Introduce finalize_memop_{atom,pair} target/arm: Use tcg_gen_qemu_ld_i128 for LDXP target/arm: Use tcg_gen_qemu_{st, ld}_i128 for do_fp_{st, ld} target/arm: Use tcg_gen_qemu_st_i128 for STZG, STZ2G target/arm: Use tcg_gen_qemu_{ld, st}_i128 in gen_sve_{ld, st}r target/arm: Sink gen_mte_check1 into load/store_exclusive target/arm: Load/store integer pair with one tcg operation target/arm: Hoist finalize_memop out of do_gpr_{ld, st} target/arm: Hoist finalize_memop out of do_fp_{ld, st} target/arm: Pass memop to gen_mte_check1* target/arm: Pass single_memop to gen_mte_checkN target/arm: Check alignment in helper_mte_check target/arm: Add SCTLR.nAA to TBFLAG_A64 target/arm: Relax ordered/atomic alignment checks for LSE2 target/arm: Move mte check for store-exclusive tests/tcg/aarch64: Use stz2g in mte-7.c tests/tcg/multiarch: Adjust sigbus.c target/arm: Enable FEAT_LSE2 for -cpu max Vikram Garhwal (4): hw/net/can: Introduce Xilinx Versal CANFD controller xlnx-versal: Connect Xilinx VERSAL CANFD controllers MAINTAINERS: Include canfd tests under Xilinx CAN tests/qtest: Introduce tests for Xilinx VERSAL CANFD controller Zhuojia Shen (3): target/arm: allow DC CVA[D]P in user mode emulation tests/tcg/aarch64: add DC CVA[D]P tests target/arm: trap DCC access in user mode emulation qianfan Zhao (11): hw: arm: Add bananapi M2-Ultra and allwinner-r40 support hw/arm/allwinner-r40: add Clock Control Unit hw: allwinner-r40: Complete uart devices hw: arm: allwinner-r40: Add i2c0 device hw/misc: Rename axp209 to axp22x and add support AXP221 PMU hw/arm/allwinner-r40: add SDRAM controller device hw: sd: allwinner-sdhost: Add sun50i-a64 SoC support hw: arm: allwinner-r40: Add emac and gmac support hw: arm: allwinner-sramc: Add SRAM Controller support for R40 tests: avocado: boot_linux_console: Add test case for bpim2u docs: system: arm: Introduce bananapi_m2u MAINTAINERS | 2 +- docs/system/arm/bananapi_m2u.rst | 139 +++ docs/system/arm/emulation.rst | 1 + docs/system/arm/xlnx-versal-virt.rst | 31 + docs/system/target-arm.rst | 1 + include/hw/arm/allwinner-r40.h | 143 +++ include/hw/arm/xlnx-versal.h | 12 + include/hw/misc/allwinner-r40-ccu.h | 65 + include/hw/misc/allwinner-r40-dramc.h | 108 ++ include/hw/misc/allwinner-sramc.h | 69 ++ include/hw/net/xlnx-versal-canfd.h | 87 ++ include/hw/sd/allwinner-sdhost.h | 9 + include/sysemu/hvf.h | 37 + include/sysemu/hvf_int.h | 2 + target/arm/cpu.h | 16 +- target/arm/hvf_arm.h | 7 + target/arm/internals.h | 53 +- target/arm/tcg/helper-a64.h | 3 + target/arm/tcg/translate-a64.h | 4 +- target/arm/tcg/translate.h | 65 +- accel/hvf/hvf-accel-ops.c | 119 ++ accel/hvf/hvf-all.c | 23 + hw/arm/allwinner-r40.c | 526 ++++++++ hw/arm/bananapi_m2u.c | 145 +++ hw/arm/xlnx-versal-virt.c | 53 + hw/arm/xlnx-versal.c | 37 + hw/misc/allwinner-r40-ccu.c | 209 ++++ hw/misc/allwinner-r40-dramc.c | 513 ++++++++ hw/misc/allwinner-sramc.c | 184 +++ hw/misc/axp209.c | 238 ---- hw/misc/axp2xx.c | 283 +++++ hw/net/can/xlnx-versal-canfd.c | 2107 +++++++++++++++++++++++++++++++++ hw/sd/allwinner-sdhost.c | 72 +- target/arm/cpu.c | 2 + target/arm/debug_helper.c | 5 + target/arm/helper.c | 6 +- target/arm/hvf/hvf.c | 750 +++++++++++- target/arm/hyp_gdbstub.c | 253 ++++ target/arm/kvm64.c | 276 ----- target/arm/tcg/cpu64.c | 1 + target/arm/tcg/helper-a64.c | 7 + target/arm/tcg/hflags.c | 6 + target/arm/tcg/mte_helper.c | 18 + target/arm/tcg/translate-a64.c | 477 +++++--- target/arm/tcg/translate-sve.c | 106 +- target/arm/tcg/translate.c | 1 + target/i386/hvf/hvf.c | 33 + tests/qtest/xlnx-canfd-test.c | 423 +++++++ tests/tcg/aarch64/dcpodp.c | 63 + tests/tcg/aarch64/dcpop.c | 63 + tests/tcg/aarch64/mte-7.c | 3 +- tests/tcg/multiarch/sigbus.c | 13 +- hw/arm/Kconfig | 14 +- hw/arm/meson.build | 1 + hw/misc/Kconfig | 5 +- hw/misc/meson.build | 5 +- hw/misc/trace-events | 26 +- hw/net/can/meson.build | 1 + hw/net/can/trace-events | 7 + target/arm/meson.build | 3 +- tests/avocado/boot_linux_console.py | 176 +++ tests/qtest/meson.build | 1 + tests/tcg/aarch64/Makefile.target | 11 + 63 files changed, 7386 insertions(+), 733 deletions(-) create mode 100644 docs/system/arm/bananapi_m2u.rst create mode 100644 include/hw/arm/allwinner-r40.h create mode 100644 include/hw/misc/allwinner-r40-ccu.h create mode 100644 include/hw/misc/allwinner-r40-dramc.h create mode 100644 include/hw/misc/allwinner-sramc.h create mode 100644 include/hw/net/xlnx-versal-canfd.h create mode 100644 hw/arm/allwinner-r40.c create mode 100644 hw/arm/bananapi_m2u.c create mode 100644 hw/misc/allwinner-r40-ccu.c create mode 100644 hw/misc/allwinner-r40-dramc.c create mode 100644 hw/misc/allwinner-sramc.c delete mode 100644 hw/misc/axp209.c create mode 100644 hw/misc/axp2xx.c create mode 100644 hw/net/can/xlnx-versal-canfd.c create mode 100644 target/arm/hyp_gdbstub.c create mode 100644 tests/qtest/xlnx-canfd-test.c create mode 100644 tests/tcg/aarch64/dcpodp.c create mode 100644 tests/tcg/aarch64/dcpop.c From: Francesco Cagnin <fcagnin@quarkslab.com> These helpers will be also used for HVF. Aside from reformatting a couple of comments for 'checkpatch.pl' and updating meson to compile 'hyp_gdbstub.c', this is just code motion. Signed-off-by: Francesco Cagnin <fcagnin@quarkslab.com> Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Message-id: 20230601153107.81955-2-fcagnin@quarkslab.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/internals.h | 50 +++++++ target/arm/hyp_gdbstub.c | 253 +++++++++++++++++++++++++++++++++++ target/arm/kvm64.c | 276 --------------------------------------- target/arm/meson.build | 3 +- 4 files changed, 305 insertions(+), 277 deletions(-) create mode 100644 target/arm/hyp_gdbstub.c diff --git a/target/arm/internals.h b/target/arm/internals.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/internals.h +++ b/target/arm/internals.h @@ -XXX,XX +XXX,XX @@ static inline bool arm_fgt_active(CPUARMState *env, int el) } void assert_hflags_rebuild_correctly(CPUARMState *env); + +/* + * Although the ARM implementation of hardware assisted debugging + * allows for different breakpoints per-core, the current GDB + * interface treats them as a global pool of registers (which seems to + * be the case for x86, ppc and s390). As a result we store one copy + * of registers which is used for all active cores. + * + * Write access is serialised by virtue of the GDB protocol which + * updates things. Read access (i.e. when the values are copied to the + * vCPU) is also gated by GDB's run control. + * + * This is not unreasonable as most of the time debugging kernels you + * never know which core will eventually execute your function. + */ + +typedef struct { + uint64_t bcr; + uint64_t bvr; +} HWBreakpoint; + +/* + * The watchpoint registers can cover more area than the requested + * watchpoint so we need to store the additional information + * somewhere. We also need to supply a CPUWatchpoint to the GDB stub + * when the watchpoint is hit. + */ +typedef struct { + uint64_t wcr; + uint64_t wvr; + CPUWatchpoint details; +} HWWatchpoint; + +/* Maximum and current break/watch point counts */ +extern int max_hw_bps, max_hw_wps; +extern GArray *hw_breakpoints, *hw_watchpoints; + +#define cur_hw_wps (hw_watchpoints->len) +#define cur_hw_bps (hw_breakpoints->len) +#define get_hw_bp(i) (&g_array_index(hw_breakpoints, HWBreakpoint, i)) +#define get_hw_wp(i) (&g_array_index(hw_watchpoints, HWWatchpoint, i)) + +bool find_hw_breakpoint(CPUState *cpu, target_ulong pc); +int insert_hw_breakpoint(target_ulong pc); +int delete_hw_breakpoint(target_ulong pc); + +bool check_watchpoint_in_range(int i, target_ulong addr); +CPUWatchpoint *find_hw_watchpoint(CPUState *cpu, target_ulong addr); +int insert_hw_watchpoint(target_ulong addr, target_ulong len, int type); +int delete_hw_watchpoint(target_ulong addr, target_ulong len, int type); #endif diff --git a/target/arm/hyp_gdbstub.c b/target/arm/hyp_gdbstub.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/target/arm/hyp_gdbstub.c @@ -XXX,XX +XXX,XX @@ +/* + * ARM implementation of KVM and HVF hooks, 64 bit specific code + * + * Copyright Mian-M. Hamayun 2013, Virtual Open Systems + * Copyright Alex Bennée 2014, Linaro + * + * This work is licensed under the terms of the GNU GPL, version 2 or later. + * See the COPYING file in the top-level directory. + * + */ + +#include "qemu/osdep.h" +#include "cpu.h" +#include "internals.h" +#include "exec/gdbstub.h" + +/* Maximum and current break/watch point counts */ +int max_hw_bps, max_hw_wps; +GArray *hw_breakpoints, *hw_watchpoints; + +/** + * insert_hw_breakpoint() + * @addr: address of breakpoint + * + * See ARM ARM D2.9.1 for details but here we are only going to create + * simple un-linked breakpoints (i.e. we don't chain breakpoints + * together to match address and context or vmid). The hardware is + * capable of fancier matching but that will require exposing that + * fanciness to GDB's interface + * + * DBGBCR<n>_EL1, Debug Breakpoint Control Registers + * + * 31 24 23 20 19 16 15 14 13 12 9 8 5 4 3 2 1 0 + * +------+------+-------+-----+----+------+-----+------+-----+---+ + * | RES0 | BT | LBN | SSC | HMC| RES0 | BAS | RES0 | PMC | E | + * +------+------+-------+-----+----+------+-----+------+-----+---+ + * + * BT: Breakpoint type (0 = unlinked address match) + * LBN: Linked BP number (0 = unused) + * SSC/HMC/PMC: Security, Higher and Priv access control (Table D-12) + * BAS: Byte Address Select (RES1 for AArch64) + * E: Enable bit + * + * DBGBVR<n>_EL1, Debug Breakpoint Value Registers + * + * 63 53 52 49 48 2 1 0 + * +------+-----------+----------+-----+ + * | RESS | VA[52:49] | VA[48:2] | 0 0 | + * +------+-----------+----------+-----+ + * + * Depending on the addressing mode bits the top bits of the register + * are a sign extension of the highest applicable VA bit. Some + * versions of GDB don't do it correctly so we ensure they are correct + * here so future PC comparisons will work properly. + */ + +int insert_hw_breakpoint(target_ulong addr) +{ + HWBreakpoint brk = { + .bcr = 0x1, /* BCR E=1, enable */ + .bvr = sextract64(addr, 0, 53) + }; + + if (cur_hw_bps >= max_hw_bps) { + return -ENOBUFS; + } + + brk.bcr = deposit32(brk.bcr, 1, 2, 0x3); /* PMC = 11 */ + brk.bcr = deposit32(brk.bcr, 5, 4, 0xf); /* BAS = RES1 */ + + g_array_append_val(hw_breakpoints, brk); + + return 0; +} + +/** + * delete_hw_breakpoint() + * @pc: address of breakpoint + * + * Delete a breakpoint and shuffle any above down + */ + +int delete_hw_breakpoint(target_ulong pc) +{ + int i; + for (i = 0; i < hw_breakpoints->len; i++) { + HWBreakpoint *brk = get_hw_bp(i); + if (brk->bvr == pc) { + g_array_remove_index(hw_breakpoints, i); + return 0; + } + } + return -ENOENT; +} + +/** + * insert_hw_watchpoint() + * @addr: address of watch point + * @len: size of area + * @type: type of watch point + * + * See ARM ARM D2.10. As with the breakpoints we can do some advanced + * stuff if we want to. The watch points can be linked with the break + * points above to make them context aware. However for simplicity + * currently we only deal with simple read/write watch points. + * + * D7.3.11 DBGWCR<n>_EL1, Debug Watchpoint Control Registers + * + * 31 29 28 24 23 21 20 19 16 15 14 13 12 5 4 3 2 1 0 + * +------+-------+------+----+-----+-----+-----+-----+-----+-----+---+ + * | RES0 | MASK | RES0 | WT | LBN | SSC | HMC | BAS | LSC | PAC | E | + * +------+-------+------+----+-----+-----+-----+-----+-----+-----+---+ + * + * MASK: num bits addr mask (0=none,01/10=res,11=3 bits (8 bytes)) + * WT: 0 - unlinked, 1 - linked (not currently used) + * LBN: Linked BP number (not currently used) + * SSC/HMC/PAC: Security, Higher and Priv access control (Table D2-11) + * BAS: Byte Address Select + * LSC: Load/Store control (01: load, 10: store, 11: both) + * E: Enable + * + * The bottom 2 bits of the value register are masked. Therefore to + * break on any sizes smaller than an unaligned word you need to set + * MASK=0, BAS=bit per byte in question. For larger regions (^2) you + * need to ensure you mask the address as required and set BAS=0xff + */ + +int insert_hw_watchpoint(target_ulong addr, target_ulong len, int type) +{ + HWWatchpoint wp = { + .wcr = R_DBGWCR_E_MASK, /* E=1, enable */ + .wvr = addr & (~0x7ULL), + .details = { .vaddr = addr, .len = len } + }; + + if (cur_hw_wps >= max_hw_wps) { + return -ENOBUFS; + } + + /* + * HMC=0 SSC=0 PAC=3 will hit EL0 or EL1, any security state, + * valid whether EL3 is implemented or not + */ + wp.wcr = FIELD_DP64(wp.wcr, DBGWCR, PAC, 3); + + switch (type) { + case GDB_WATCHPOINT_READ: + wp.wcr = FIELD_DP64(wp.wcr, DBGWCR, LSC, 1); + wp.details.flags = BP_MEM_READ; + break; + case GDB_WATCHPOINT_WRITE: + wp.wcr = FIELD_DP64(wp.wcr, DBGWCR, LSC, 2); + wp.details.flags = BP_MEM_WRITE; + break; + case GDB_WATCHPOINT_ACCESS: + wp.wcr = FIELD_DP64(wp.wcr, DBGWCR, LSC, 3); + wp.details.flags = BP_MEM_ACCESS; + break; + default: + g_assert_not_reached(); + break; + } + if (len <= 8) { + /* we align the address and set the bits in BAS */ + int off = addr & 0x7; + int bas = (1 << len) - 1; + + wp.wcr = deposit32(wp.wcr, 5 + off, 8 - off, bas); + } else { + /* For ranges above 8 bytes we need to be a power of 2 */ + if (is_power_of_2(len)) { + int bits = ctz64(len); + + wp.wvr &= ~((1 << bits) - 1); + wp.wcr = FIELD_DP64(wp.wcr, DBGWCR, MASK, bits); + wp.wcr = FIELD_DP64(wp.wcr, DBGWCR, BAS, 0xff); + } else { + return -ENOBUFS; + } + } + + g_array_append_val(hw_watchpoints, wp); + return 0; +} + +bool check_watchpoint_in_range(int i, target_ulong addr) +{ + HWWatchpoint *wp = get_hw_wp(i); + uint64_t addr_top, addr_bottom = wp->wvr; + int bas = extract32(wp->wcr, 5, 8); + int mask = extract32(wp->wcr, 24, 4); + + if (mask) { + addr_top = addr_bottom + (1 << mask); + } else { + /* + * BAS must be contiguous but can offset against the base + * address in DBGWVR + */ + addr_bottom = addr_bottom + ctz32(bas); + addr_top = addr_bottom + clo32(bas); + } + + if (addr >= addr_bottom && addr <= addr_top) { + return true; + } + + return false; +} + +/** + * delete_hw_watchpoint() + * @addr: address of breakpoint + * + * Delete a breakpoint and shuffle any above down + */ + +int delete_hw_watchpoint(target_ulong addr, target_ulong len, int type) +{ + int i; + for (i = 0; i < cur_hw_wps; i++) { + if (check_watchpoint_in_range(i, addr)) { + g_array_remove_index(hw_watchpoints, i); + return 0; + } + } + return -ENOENT; +} + +bool find_hw_breakpoint(CPUState *cpu, target_ulong pc) +{ + int i; + + for (i = 0; i < cur_hw_bps; i++) { + HWBreakpoint *bp = get_hw_bp(i); + if (bp->bvr == pc) { + return true; + } + } + return false; +} + +CPUWatchpoint *find_hw_watchpoint(CPUState *cpu, target_ulong addr) +{ + int i; + + for (i = 0; i < cur_hw_wps; i++) { + if (check_watchpoint_in_range(i, addr)) { + return &get_hw_wp(i)->details; + } + } + return NULL; +} diff --git a/target/arm/kvm64.c b/target/arm/kvm64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/kvm64.c +++ b/target/arm/kvm64.c @@ -XXX,XX +XXX,XX @@ static bool have_guest_debug; -/* - * Although the ARM implementation of hardware assisted debugging - * allows for different breakpoints per-core, the current GDB - * interface treats them as a global pool of registers (which seems to - * be the case for x86, ppc and s390). As a result we store one copy - * of registers which is used for all active cores. - * - * Write access is serialised by virtue of the GDB protocol which - * updates things. Read access (i.e. when the values are copied to the - * vCPU) is also gated by GDB's run control. - * - * This is not unreasonable as most of the time debugging kernels you - * never know which core will eventually execute your function. - */ - -typedef struct { - uint64_t bcr; - uint64_t bvr; -} HWBreakpoint; - -/* The watchpoint registers can cover more area than the requested - * watchpoint so we need to store the additional information - * somewhere. We also need to supply a CPUWatchpoint to the GDB stub - * when the watchpoint is hit. - */ -typedef struct { - uint64_t wcr; - uint64_t wvr; - CPUWatchpoint details; -} HWWatchpoint; - -/* Maximum and current break/watch point counts */ -int max_hw_bps, max_hw_wps; -GArray *hw_breakpoints, *hw_watchpoints; - -#define cur_hw_wps (hw_watchpoints->len) -#define cur_hw_bps (hw_breakpoints->len) -#define get_hw_bp(i) (&g_array_index(hw_breakpoints, HWBreakpoint, i)) -#define get_hw_wp(i) (&g_array_index(hw_watchpoints, HWWatchpoint, i)) - void kvm_arm_init_debug(KVMState *s) { have_guest_debug = kvm_check_extension(s, @@ -XXX,XX +XXX,XX @@ void kvm_arm_init_debug(KVMState *s) return; } -/** - * insert_hw_breakpoint() - * @addr: address of breakpoint - * - * See ARM ARM D2.9.1 for details but here we are only going to create - * simple un-linked breakpoints (i.e. we don't chain breakpoints - * together to match address and context or vmid). The hardware is - * capable of fancier matching but that will require exposing that - * fanciness to GDB's interface - * - * DBGBCR<n>_EL1, Debug Breakpoint Control Registers - * - * 31 24 23 20 19 16 15 14 13 12 9 8 5 4 3 2 1 0 - * +------+------+-------+-----+----+------+-----+------+-----+---+ - * | RES0 | BT | LBN | SSC | HMC| RES0 | BAS | RES0 | PMC | E | - * +------+------+-------+-----+----+------+-----+------+-----+---+ - * - * BT: Breakpoint type (0 = unlinked address match) - * LBN: Linked BP number (0 = unused) - * SSC/HMC/PMC: Security, Higher and Priv access control (Table D-12) - * BAS: Byte Address Select (RES1 for AArch64) - * E: Enable bit - * - * DBGBVR<n>_EL1, Debug Breakpoint Value Registers - * - * 63 53 52 49 48 2 1 0 - * +------+-----------+----------+-----+ - * | RESS | VA[52:49] | VA[48:2] | 0 0 | - * +------+-----------+----------+-----+ - * - * Depending on the addressing mode bits the top bits of the register - * are a sign extension of the highest applicable VA bit. Some - * versions of GDB don't do it correctly so we ensure they are correct - * here so future PC comparisons will work properly. - */ - -static int insert_hw_breakpoint(target_ulong addr) -{ - HWBreakpoint brk = { - .bcr = 0x1, /* BCR E=1, enable */ - .bvr = sextract64(addr, 0, 53) - }; - - if (cur_hw_bps >= max_hw_bps) { - return -ENOBUFS; - } - - brk.bcr = deposit32(brk.bcr, 1, 2, 0x3); /* PMC = 11 */ - brk.bcr = deposit32(brk.bcr, 5, 4, 0xf); /* BAS = RES1 */ - - g_array_append_val(hw_breakpoints, brk); - - return 0; -} - -/** - * delete_hw_breakpoint() - * @pc: address of breakpoint - * - * Delete a breakpoint and shuffle any above down - */ - -static int delete_hw_breakpoint(target_ulong pc) -{ - int i; - for (i = 0; i < hw_breakpoints->len; i++) { - HWBreakpoint *brk = get_hw_bp(i); - if (brk->bvr == pc) { - g_array_remove_index(hw_breakpoints, i); - return 0; - } - } - return -ENOENT; -} - -/** - * insert_hw_watchpoint() - * @addr: address of watch point - * @len: size of area - * @type: type of watch point - * - * See ARM ARM D2.10. As with the breakpoints we can do some advanced - * stuff if we want to. The watch points can be linked with the break - * points above to make them context aware. However for simplicity - * currently we only deal with simple read/write watch points. - * - * D7.3.11 DBGWCR<n>_EL1, Debug Watchpoint Control Registers - * - * 31 29 28 24 23 21 20 19 16 15 14 13 12 5 4 3 2 1 0 - * +------+-------+------+----+-----+-----+-----+-----+-----+-----+---+ - * | RES0 | MASK | RES0 | WT | LBN | SSC | HMC | BAS | LSC | PAC | E | - * +------+-------+------+----+-----+-----+-----+-----+-----+-----+---+ - * - * MASK: num bits addr mask (0=none,01/10=res,11=3 bits (8 bytes)) - * WT: 0 - unlinked, 1 - linked (not currently used) - * LBN: Linked BP number (not currently used) - * SSC/HMC/PAC: Security, Higher and Priv access control (Table D2-11) - * BAS: Byte Address Select - * LSC: Load/Store control (01: load, 10: store, 11: both) - * E: Enable - * - * The bottom 2 bits of the value register are masked. Therefore to - * break on any sizes smaller than an unaligned word you need to set - * MASK=0, BAS=bit per byte in question. For larger regions (^2) you - * need to ensure you mask the address as required and set BAS=0xff - */ - -static int insert_hw_watchpoint(target_ulong addr, - target_ulong len, int type) -{ - HWWatchpoint wp = { - .wcr = R_DBGWCR_E_MASK, /* E=1, enable */ - .wvr = addr & (~0x7ULL), - .details = { .vaddr = addr, .len = len } - }; - - if (cur_hw_wps >= max_hw_wps) { - return -ENOBUFS; - } - - /* - * HMC=0 SSC=0 PAC=3 will hit EL0 or EL1, any security state, - * valid whether EL3 is implemented or not - */ - wp.wcr = FIELD_DP64(wp.wcr, DBGWCR, PAC, 3); - - switch (type) { - case GDB_WATCHPOINT_READ: - wp.wcr = FIELD_DP64(wp.wcr, DBGWCR, LSC, 1); - wp.details.flags = BP_MEM_READ; - break; - case GDB_WATCHPOINT_WRITE: - wp.wcr = FIELD_DP64(wp.wcr, DBGWCR, LSC, 2); - wp.details.flags = BP_MEM_WRITE; - break; - case GDB_WATCHPOINT_ACCESS: - wp.wcr = FIELD_DP64(wp.wcr, DBGWCR, LSC, 3); - wp.details.flags = BP_MEM_ACCESS; - break; - default: - g_assert_not_reached(); - break; - } - if (len <= 8) { - /* we align the address and set the bits in BAS */ - int off = addr & 0x7; - int bas = (1 << len) - 1; - - wp.wcr = deposit32(wp.wcr, 5 + off, 8 - off, bas); - } else { - /* For ranges above 8 bytes we need to be a power of 2 */ - if (is_power_of_2(len)) { - int bits = ctz64(len); - - wp.wvr &= ~((1 << bits) - 1); - wp.wcr = FIELD_DP64(wp.wcr, DBGWCR, MASK, bits); - wp.wcr = FIELD_DP64(wp.wcr, DBGWCR, BAS, 0xff); - } else { - return -ENOBUFS; - } - } - - g_array_append_val(hw_watchpoints, wp); - return 0; -} - - -static bool check_watchpoint_in_range(int i, target_ulong addr) -{ - HWWatchpoint *wp = get_hw_wp(i); - uint64_t addr_top, addr_bottom = wp->wvr; - int bas = extract32(wp->wcr, 5, 8); - int mask = extract32(wp->wcr, 24, 4); - - if (mask) { - addr_top = addr_bottom + (1 << mask); - } else { - /* BAS must be contiguous but can offset against the base - * address in DBGWVR */ - addr_bottom = addr_bottom + ctz32(bas); - addr_top = addr_bottom + clo32(bas); - } - - if (addr >= addr_bottom && addr <= addr_top) { - return true; - } - - return false; -} - -/** - * delete_hw_watchpoint() - * @addr: address of breakpoint - * - * Delete a breakpoint and shuffle any above down - */ - -static int delete_hw_watchpoint(target_ulong addr, - target_ulong len, int type) -{ - int i; - for (i = 0; i < cur_hw_wps; i++) { - if (check_watchpoint_in_range(i, addr)) { - g_array_remove_index(hw_watchpoints, i); - return 0; - } - } - return -ENOENT; -} - - int kvm_arch_insert_hw_breakpoint(target_ulong addr, target_ulong len, int type) { @@ -XXX,XX +XXX,XX @@ bool kvm_arm_hw_debug_active(CPUState *cs) return ((cur_hw_wps > 0) || (cur_hw_bps > 0)); } -static bool find_hw_breakpoint(CPUState *cpu, target_ulong pc) -{ - int i; - - for (i = 0; i < cur_hw_bps; i++) { - HWBreakpoint *bp = get_hw_bp(i); - if (bp->bvr == pc) { - return true; - } - } - return false; -} - -static CPUWatchpoint *find_hw_watchpoint(CPUState *cpu, target_ulong addr) -{ - int i; - - for (i = 0; i < cur_hw_wps; i++) { - if (check_watchpoint_in_range(i, addr)) { - return &get_hw_wp(i)->details; - } - } - return NULL; -} - static bool kvm_arm_set_device_attr(CPUState *cs, struct kvm_device_attr *attr, const char *name) { diff --git a/target/arm/meson.build b/target/arm/meson.build index XXXXXXX..XXXXXXX 100644 --- a/target/arm/meson.build +++ b/target/arm/meson.build @@ -XXX,XX +XXX,XX @@ arm_ss.add(files( )) arm_ss.add(zlib) -arm_ss.add(when: 'CONFIG_KVM', if_true: files('kvm.c', 'kvm64.c'), if_false: files('kvm-stub.c')) +arm_ss.add(when: 'CONFIG_KVM', if_true: files('hyp_gdbstub.c', 'kvm.c', 'kvm64.c'), if_false: files('kvm-stub.c')) +arm_ss.add(when: 'CONFIG_HVF', if_true: files('hyp_gdbstub.c')) arm_ss.add(when: 'TARGET_AARCH64', if_true: files( 'cpu64.c', -- 2.34.1 From: Francesco Cagnin <fcagnin@quarkslab.com> Required for guest debugging. Signed-off-by: Francesco Cagnin <fcagnin@quarkslab.com> Message-id: 20230601153107.81955-3-fcagnin@quarkslab.com Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/hvf/hvf.c | 213 +++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 213 insertions(+) diff --git a/target/arm/hvf/hvf.c b/target/arm/hvf/hvf.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/hvf/hvf.c +++ b/target/arm/hvf/hvf.c @@ -XXX,XX +XXX,XX @@ #define SYSREG_ICC_SGI1R_EL1 SYSREG(3, 0, 12, 11, 5) #define SYSREG_ICC_SRE_EL1 SYSREG(3, 0, 12, 12, 5) +#define SYSREG_MDSCR_EL1 SYSREG(2, 0, 0, 2, 2) +#define SYSREG_DBGBVR0_EL1 SYSREG(2, 0, 0, 0, 4) +#define SYSREG_DBGBCR0_EL1 SYSREG(2, 0, 0, 0, 5) +#define SYSREG_DBGWVR0_EL1 SYSREG(2, 0, 0, 0, 6) +#define SYSREG_DBGWCR0_EL1 SYSREG(2, 0, 0, 0, 7) +#define SYSREG_DBGBVR1_EL1 SYSREG(2, 0, 0, 1, 4) +#define SYSREG_DBGBCR1_EL1 SYSREG(2, 0, 0, 1, 5) +#define SYSREG_DBGWVR1_EL1 SYSREG(2, 0, 0, 1, 6) +#define SYSREG_DBGWCR1_EL1 SYSREG(2, 0, 0, 1, 7) +#define SYSREG_DBGBVR2_EL1 SYSREG(2, 0, 0, 2, 4) +#define SYSREG_DBGBCR2_EL1 SYSREG(2, 0, 0, 2, 5) +#define SYSREG_DBGWVR2_EL1 SYSREG(2, 0, 0, 2, 6) +#define SYSREG_DBGWCR2_EL1 SYSREG(2, 0, 0, 2, 7) +#define SYSREG_DBGBVR3_EL1 SYSREG(2, 0, 0, 3, 4) +#define SYSREG_DBGBCR3_EL1 SYSREG(2, 0, 0, 3, 5) +#define SYSREG_DBGWVR3_EL1 SYSREG(2, 0, 0, 3, 6) +#define SYSREG_DBGWCR3_EL1 SYSREG(2, 0, 0, 3, 7) +#define SYSREG_DBGBVR4_EL1 SYSREG(2, 0, 0, 4, 4) +#define SYSREG_DBGBCR4_EL1 SYSREG(2, 0, 0, 4, 5) +#define SYSREG_DBGWVR4_EL1 SYSREG(2, 0, 0, 4, 6) +#define SYSREG_DBGWCR4_EL1 SYSREG(2, 0, 0, 4, 7) +#define SYSREG_DBGBVR5_EL1 SYSREG(2, 0, 0, 5, 4) +#define SYSREG_DBGBCR5_EL1 SYSREG(2, 0, 0, 5, 5) +#define SYSREG_DBGWVR5_EL1 SYSREG(2, 0, 0, 5, 6) +#define SYSREG_DBGWCR5_EL1 SYSREG(2, 0, 0, 5, 7) +#define SYSREG_DBGBVR6_EL1 SYSREG(2, 0, 0, 6, 4) +#define SYSREG_DBGBCR6_EL1 SYSREG(2, 0, 0, 6, 5) +#define SYSREG_DBGWVR6_EL1 SYSREG(2, 0, 0, 6, 6) +#define SYSREG_DBGWCR6_EL1 SYSREG(2, 0, 0, 6, 7) +#define SYSREG_DBGBVR7_EL1 SYSREG(2, 0, 0, 7, 4) +#define SYSREG_DBGBCR7_EL1 SYSREG(2, 0, 0, 7, 5) +#define SYSREG_DBGWVR7_EL1 SYSREG(2, 0, 0, 7, 6) +#define SYSREG_DBGWCR7_EL1 SYSREG(2, 0, 0, 7, 7) +#define SYSREG_DBGBVR8_EL1 SYSREG(2, 0, 0, 8, 4) +#define SYSREG_DBGBCR8_EL1 SYSREG(2, 0, 0, 8, 5) +#define SYSREG_DBGWVR8_EL1 SYSREG(2, 0, 0, 8, 6) +#define SYSREG_DBGWCR8_EL1 SYSREG(2, 0, 0, 8, 7) +#define SYSREG_DBGBVR9_EL1 SYSREG(2, 0, 0, 9, 4) +#define SYSREG_DBGBCR9_EL1 SYSREG(2, 0, 0, 9, 5) +#define SYSREG_DBGWVR9_EL1 SYSREG(2, 0, 0, 9, 6) +#define SYSREG_DBGWCR9_EL1 SYSREG(2, 0, 0, 9, 7) +#define SYSREG_DBGBVR10_EL1 SYSREG(2, 0, 0, 10, 4) +#define SYSREG_DBGBCR10_EL1 SYSREG(2, 0, 0, 10, 5) +#define SYSREG_DBGWVR10_EL1 SYSREG(2, 0, 0, 10, 6) +#define SYSREG_DBGWCR10_EL1 SYSREG(2, 0, 0, 10, 7) +#define SYSREG_DBGBVR11_EL1 SYSREG(2, 0, 0, 11, 4) +#define SYSREG_DBGBCR11_EL1 SYSREG(2, 0, 0, 11, 5) +#define SYSREG_DBGWVR11_EL1 SYSREG(2, 0, 0, 11, 6) +#define SYSREG_DBGWCR11_EL1 SYSREG(2, 0, 0, 11, 7) +#define SYSREG_DBGBVR12_EL1 SYSREG(2, 0, 0, 12, 4) +#define SYSREG_DBGBCR12_EL1 SYSREG(2, 0, 0, 12, 5) +#define SYSREG_DBGWVR12_EL1 SYSREG(2, 0, 0, 12, 6) +#define SYSREG_DBGWCR12_EL1 SYSREG(2, 0, 0, 12, 7) +#define SYSREG_DBGBVR13_EL1 SYSREG(2, 0, 0, 13, 4) +#define SYSREG_DBGBCR13_EL1 SYSREG(2, 0, 0, 13, 5) +#define SYSREG_DBGWVR13_EL1 SYSREG(2, 0, 0, 13, 6) +#define SYSREG_DBGWCR13_EL1 SYSREG(2, 0, 0, 13, 7) +#define SYSREG_DBGBVR14_EL1 SYSREG(2, 0, 0, 14, 4) +#define SYSREG_DBGBCR14_EL1 SYSREG(2, 0, 0, 14, 5) +#define SYSREG_DBGWVR14_EL1 SYSREG(2, 0, 0, 14, 6) +#define SYSREG_DBGWCR14_EL1 SYSREG(2, 0, 0, 14, 7) +#define SYSREG_DBGBVR15_EL1 SYSREG(2, 0, 0, 15, 4) +#define SYSREG_DBGBCR15_EL1 SYSREG(2, 0, 0, 15, 5) +#define SYSREG_DBGWVR15_EL1 SYSREG(2, 0, 0, 15, 6) +#define SYSREG_DBGWCR15_EL1 SYSREG(2, 0, 0, 15, 7) + #define WFX_IS_WFE (1 << 0) #define TMR_CTL_ENABLE (1 << 0) @@ -XXX,XX +XXX,XX @@ static int hvf_sysreg_read(CPUState *cpu, uint32_t reg, uint32_t rt) hvf_raise_exception(cpu, EXCP_UDEF, syn_uncategorized()); } break; + case SYSREG_DBGBVR0_EL1: + case SYSREG_DBGBVR1_EL1: + case SYSREG_DBGBVR2_EL1: + case SYSREG_DBGBVR3_EL1: + case SYSREG_DBGBVR4_EL1: + case SYSREG_DBGBVR5_EL1: + case SYSREG_DBGBVR6_EL1: + case SYSREG_DBGBVR7_EL1: + case SYSREG_DBGBVR8_EL1: + case SYSREG_DBGBVR9_EL1: + case SYSREG_DBGBVR10_EL1: + case SYSREG_DBGBVR11_EL1: + case SYSREG_DBGBVR12_EL1: + case SYSREG_DBGBVR13_EL1: + case SYSREG_DBGBVR14_EL1: + case SYSREG_DBGBVR15_EL1: + val = env->cp15.dbgbvr[SYSREG_CRM(reg)]; + break; + case SYSREG_DBGBCR0_EL1: + case SYSREG_DBGBCR1_EL1: + case SYSREG_DBGBCR2_EL1: + case SYSREG_DBGBCR3_EL1: + case SYSREG_DBGBCR4_EL1: + case SYSREG_DBGBCR5_EL1: + case SYSREG_DBGBCR6_EL1: + case SYSREG_DBGBCR7_EL1: + case SYSREG_DBGBCR8_EL1: + case SYSREG_DBGBCR9_EL1: + case SYSREG_DBGBCR10_EL1: + case SYSREG_DBGBCR11_EL1: + case SYSREG_DBGBCR12_EL1: + case SYSREG_DBGBCR13_EL1: + case SYSREG_DBGBCR14_EL1: + case SYSREG_DBGBCR15_EL1: + val = env->cp15.dbgbcr[SYSREG_CRM(reg)]; + break; + case SYSREG_DBGWVR0_EL1: + case SYSREG_DBGWVR1_EL1: + case SYSREG_DBGWVR2_EL1: + case SYSREG_DBGWVR3_EL1: + case SYSREG_DBGWVR4_EL1: + case SYSREG_DBGWVR5_EL1: + case SYSREG_DBGWVR6_EL1: + case SYSREG_DBGWVR7_EL1: + case SYSREG_DBGWVR8_EL1: + case SYSREG_DBGWVR9_EL1: + case SYSREG_DBGWVR10_EL1: + case SYSREG_DBGWVR11_EL1: + case SYSREG_DBGWVR12_EL1: + case SYSREG_DBGWVR13_EL1: + case SYSREG_DBGWVR14_EL1: + case SYSREG_DBGWVR15_EL1: + val = env->cp15.dbgwvr[SYSREG_CRM(reg)]; + break; + case SYSREG_DBGWCR0_EL1: + case SYSREG_DBGWCR1_EL1: + case SYSREG_DBGWCR2_EL1: + case SYSREG_DBGWCR3_EL1: + case SYSREG_DBGWCR4_EL1: + case SYSREG_DBGWCR5_EL1: + case SYSREG_DBGWCR6_EL1: + case SYSREG_DBGWCR7_EL1: + case SYSREG_DBGWCR8_EL1: + case SYSREG_DBGWCR9_EL1: + case SYSREG_DBGWCR10_EL1: + case SYSREG_DBGWCR11_EL1: + case SYSREG_DBGWCR12_EL1: + case SYSREG_DBGWCR13_EL1: + case SYSREG_DBGWCR14_EL1: + case SYSREG_DBGWCR15_EL1: + val = env->cp15.dbgwcr[SYSREG_CRM(reg)]; + break; default: if (is_id_sysreg(reg)) { /* ID system registers read as RES0 */ @@ -XXX,XX +XXX,XX @@ static int hvf_sysreg_write(CPUState *cpu, uint32_t reg, uint64_t val) hvf_raise_exception(cpu, EXCP_UDEF, syn_uncategorized()); } break; + case SYSREG_MDSCR_EL1: + env->cp15.mdscr_el1 = val; + break; + case SYSREG_DBGBVR0_EL1: + case SYSREG_DBGBVR1_EL1: + case SYSREG_DBGBVR2_EL1: + case SYSREG_DBGBVR3_EL1: + case SYSREG_DBGBVR4_EL1: + case SYSREG_DBGBVR5_EL1: + case SYSREG_DBGBVR6_EL1: + case SYSREG_DBGBVR7_EL1: + case SYSREG_DBGBVR8_EL1: + case SYSREG_DBGBVR9_EL1: + case SYSREG_DBGBVR10_EL1: + case SYSREG_DBGBVR11_EL1: + case SYSREG_DBGBVR12_EL1: + case SYSREG_DBGBVR13_EL1: + case SYSREG_DBGBVR14_EL1: + case SYSREG_DBGBVR15_EL1: + env->cp15.dbgbvr[SYSREG_CRM(reg)] = val; + break; + case SYSREG_DBGBCR0_EL1: + case SYSREG_DBGBCR1_EL1: + case SYSREG_DBGBCR2_EL1: + case SYSREG_DBGBCR3_EL1: + case SYSREG_DBGBCR4_EL1: + case SYSREG_DBGBCR5_EL1: + case SYSREG_DBGBCR6_EL1: + case SYSREG_DBGBCR7_EL1: + case SYSREG_DBGBCR8_EL1: + case SYSREG_DBGBCR9_EL1: + case SYSREG_DBGBCR10_EL1: + case SYSREG_DBGBCR11_EL1: + case SYSREG_DBGBCR12_EL1: + case SYSREG_DBGBCR13_EL1: + case SYSREG_DBGBCR14_EL1: + case SYSREG_DBGBCR15_EL1: + env->cp15.dbgbcr[SYSREG_CRM(reg)] = val; + break; + case SYSREG_DBGWVR0_EL1: + case SYSREG_DBGWVR1_EL1: + case SYSREG_DBGWVR2_EL1: + case SYSREG_DBGWVR3_EL1: + case SYSREG_DBGWVR4_EL1: + case SYSREG_DBGWVR5_EL1: + case SYSREG_DBGWVR6_EL1: + case SYSREG_DBGWVR7_EL1: + case SYSREG_DBGWVR8_EL1: + case SYSREG_DBGWVR9_EL1: + case SYSREG_DBGWVR10_EL1: + case SYSREG_DBGWVR11_EL1: + case SYSREG_DBGWVR12_EL1: + case SYSREG_DBGWVR13_EL1: + case SYSREG_DBGWVR14_EL1: + case SYSREG_DBGWVR15_EL1: + env->cp15.dbgwvr[SYSREG_CRM(reg)] = val; + break; + case SYSREG_DBGWCR0_EL1: + case SYSREG_DBGWCR1_EL1: + case SYSREG_DBGWCR2_EL1: + case SYSREG_DBGWCR3_EL1: + case SYSREG_DBGWCR4_EL1: + case SYSREG_DBGWCR5_EL1: + case SYSREG_DBGWCR6_EL1: + case SYSREG_DBGWCR7_EL1: + case SYSREG_DBGWCR8_EL1: + case SYSREG_DBGWCR9_EL1: + case SYSREG_DBGWCR10_EL1: + case SYSREG_DBGWCR11_EL1: + case SYSREG_DBGWCR12_EL1: + case SYSREG_DBGWCR13_EL1: + case SYSREG_DBGWCR14_EL1: + case SYSREG_DBGWCR15_EL1: + env->cp15.dbgwcr[SYSREG_CRM(reg)] = val; + break; default: cpu_synchronize_state(cpu); trace_hvf_unhandled_sysreg_write(env->pc, reg, -- 2.34.1 From: Francesco Cagnin <fcagnin@quarkslab.com> Required for guest debugging. The code has been structured like the KVM counterpart. Signed-off-by: Francesco Cagnin <fcagnin@quarkslab.com> Message-id: 20230601153107.81955-4-fcagnin@quarkslab.com Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/sysemu/hvf.h | 22 ++++++++ include/sysemu/hvf_int.h | 1 + accel/hvf/hvf-accel-ops.c | 109 ++++++++++++++++++++++++++++++++++++++ accel/hvf/hvf-all.c | 17 ++++++ target/arm/hvf/hvf.c | 63 ++++++++++++++++++++++ target/i386/hvf/hvf.c | 24 +++++++++ 6 files changed, 236 insertions(+) diff --git a/include/sysemu/hvf.h b/include/sysemu/hvf.h index XXXXXXX..XXXXXXX 100644 --- a/include/sysemu/hvf.h +++ b/include/sysemu/hvf.h @@ -XXX,XX +XXX,XX @@ #include "qom/object.h" #ifdef NEED_CPU_H +#include "cpu.h" #ifdef CONFIG_HVF uint32_t hvf_get_supported_cpuid(uint32_t func, uint32_t idx, @@ -XXX,XX +XXX,XX @@ typedef struct HVFState HVFState; DECLARE_INSTANCE_CHECKER(HVFState, HVF_STATE, TYPE_HVF_ACCEL) +#ifdef NEED_CPU_H +struct hvf_sw_breakpoint { + target_ulong pc; + target_ulong saved_insn; + int use_count; + QTAILQ_ENTRY(hvf_sw_breakpoint) entry; +}; + +struct hvf_sw_breakpoint *hvf_find_sw_breakpoint(CPUState *cpu, + target_ulong pc); +int hvf_sw_breakpoints_active(CPUState *cpu); + +int hvf_arch_insert_sw_breakpoint(CPUState *cpu, struct hvf_sw_breakpoint *bp); +int hvf_arch_remove_sw_breakpoint(CPUState *cpu, struct hvf_sw_breakpoint *bp); +int hvf_arch_insert_hw_breakpoint(target_ulong addr, target_ulong len, + int type); +int hvf_arch_remove_hw_breakpoint(target_ulong addr, target_ulong len, + int type); +void hvf_arch_remove_all_hw_breakpoints(void); +#endif /* NEED_CPU_H */ + #endif diff --git a/include/sysemu/hvf_int.h b/include/sysemu/hvf_int.h index XXXXXXX..XXXXXXX 100644 --- a/include/sysemu/hvf_int.h +++ b/include/sysemu/hvf_int.h @@ -XXX,XX +XXX,XX @@ struct HVFState { hvf_vcpu_caps *hvf_caps; uint64_t vtimer_offset; + QTAILQ_HEAD(, hvf_sw_breakpoint) hvf_sw_breakpoints; }; extern HVFState *hvf_state; diff --git a/accel/hvf/hvf-accel-ops.c b/accel/hvf/hvf-accel-ops.c index XXXXXXX..XXXXXXX 100644 --- a/accel/hvf/hvf-accel-ops.c +++ b/accel/hvf/hvf-accel-ops.c @@ -XXX,XX +XXX,XX @@ #include "qemu/main-loop.h" #include "exec/address-spaces.h" #include "exec/exec-all.h" +#include "exec/gdbstub.h" #include "sysemu/cpus.h" #include "sysemu/hvf.h" #include "sysemu/hvf_int.h" @@ -XXX,XX +XXX,XX @@ static int hvf_accel_init(MachineState *ms) s->slots[x].slot_id = x; } + QTAILQ_INIT(&s->hvf_sw_breakpoints); + hvf_state = s; memory_listener_register(&hvf_memory_listener, &address_space_memory); @@ -XXX,XX +XXX,XX @@ static void hvf_start_vcpu_thread(CPUState *cpu) cpu, QEMU_THREAD_JOINABLE); } +static int hvf_insert_breakpoint(CPUState *cpu, int type, hwaddr addr, hwaddr len) +{ + struct hvf_sw_breakpoint *bp; + int err; + + if (type == GDB_BREAKPOINT_SW) { + bp = hvf_find_sw_breakpoint(cpu, addr); + if (bp) { + bp->use_count++; + return 0; + } + + bp = g_new(struct hvf_sw_breakpoint, 1); + bp->pc = addr; + bp->use_count = 1; + err = hvf_arch_insert_sw_breakpoint(cpu, bp); + if (err) { + g_free(bp); + return err; + } + + QTAILQ_INSERT_HEAD(&hvf_state->hvf_sw_breakpoints, bp, entry); + } else { + err = hvf_arch_insert_hw_breakpoint(addr, len, type); + if (err) { + return err; + } + } + + CPU_FOREACH(cpu) { + err = hvf_update_guest_debug(cpu); + if (err) { + return err; + } + } + return 0; +} + +static int hvf_remove_breakpoint(CPUState *cpu, int type, hwaddr addr, hwaddr len) +{ + struct hvf_sw_breakpoint *bp; + int err; + + if (type == GDB_BREAKPOINT_SW) { + bp = hvf_find_sw_breakpoint(cpu, addr); + if (!bp) { + return -ENOENT; + } + + if (bp->use_count > 1) { + bp->use_count--; + return 0; + } + + err = hvf_arch_remove_sw_breakpoint(cpu, bp); + if (err) { + return err; + } + + QTAILQ_REMOVE(&hvf_state->hvf_sw_breakpoints, bp, entry); + g_free(bp); + } else { + err = hvf_arch_remove_hw_breakpoint(addr, len, type); + if (err) { + return err; + } + } + + CPU_FOREACH(cpu) { + err = hvf_update_guest_debug(cpu); + if (err) { + return err; + } + } + return 0; +} + +static void hvf_remove_all_breakpoints(CPUState *cpu) +{ + struct hvf_sw_breakpoint *bp, *next; + CPUState *tmpcpu; + + QTAILQ_FOREACH_SAFE(bp, &hvf_state->hvf_sw_breakpoints, entry, next) { + if (hvf_arch_remove_sw_breakpoint(cpu, bp) != 0) { + /* Try harder to find a CPU that currently sees the breakpoint. */ + CPU_FOREACH(tmpcpu) + { + if (hvf_arch_remove_sw_breakpoint(tmpcpu, bp) == 0) { + break; + } + } + } + QTAILQ_REMOVE(&hvf_state->hvf_sw_breakpoints, bp, entry); + g_free(bp); + } + hvf_arch_remove_all_hw_breakpoints(); + + CPU_FOREACH(cpu) { + hvf_update_guest_debug(cpu); + } +} + static void hvf_accel_ops_class_init(ObjectClass *oc, void *data) { AccelOpsClass *ops = ACCEL_OPS_CLASS(oc); @@ -XXX,XX +XXX,XX @@ static void hvf_accel_ops_class_init(ObjectClass *oc, void *data) ops->synchronize_post_init = hvf_cpu_synchronize_post_init; ops->synchronize_state = hvf_cpu_synchronize_state; ops->synchronize_pre_loadvm = hvf_cpu_synchronize_pre_loadvm; + + ops->insert_breakpoint = hvf_insert_breakpoint; + ops->remove_breakpoint = hvf_remove_breakpoint; + ops->remove_all_breakpoints = hvf_remove_all_breakpoints; }; static const TypeInfo hvf_accel_ops_type = { .name = ACCEL_OPS_NAME("hvf"), diff --git a/accel/hvf/hvf-all.c b/accel/hvf/hvf-all.c index XXXXXXX..XXXXXXX 100644 --- a/accel/hvf/hvf-all.c +++ b/accel/hvf/hvf-all.c @@ -XXX,XX +XXX,XX @@ void assert_hvf_ok(hv_return_t ret) abort(); } + +struct hvf_sw_breakpoint *hvf_find_sw_breakpoint(CPUState *cpu, target_ulong pc) +{ + struct hvf_sw_breakpoint *bp; + + QTAILQ_FOREACH(bp, &hvf_state->hvf_sw_breakpoints, entry) { + if (bp->pc == pc) { + return bp; + } + } + return NULL; +} + +int hvf_sw_breakpoints_active(CPUState *cpu) +{ + return !QTAILQ_EMPTY(&hvf_state->hvf_sw_breakpoints); +} diff --git a/target/arm/hvf/hvf.c b/target/arm/hvf/hvf.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/hvf/hvf.c +++ b/target/arm/hvf/hvf.c @@ -XXX,XX +XXX,XX @@ #include "trace/trace-target_arm_hvf.h" #include "migration/vmstate.h" +#include "exec/gdbstub.h" + #define HVF_SYSREG(crn, crm, op0, op1, op2) \ ENCODE_AA64_CP_REG(CP_REG_ARM64_SYSREG_CP, crn, crm, op0, op1, op2) #define PL1_WRITE_MASK 0x4 @@ -XXX,XX +XXX,XX @@ int hvf_arch_init(void) qemu_add_vm_change_state_handler(hvf_vm_state_change, &vtimer); return 0; } + +static const uint32_t brk_insn = 0xd4200000; + +int hvf_arch_insert_sw_breakpoint(CPUState *cpu, struct hvf_sw_breakpoint *bp) +{ + if (cpu_memory_rw_debug(cpu, bp->pc, (uint8_t *)&bp->saved_insn, 4, 0) || + cpu_memory_rw_debug(cpu, bp->pc, (uint8_t *)&brk_insn, 4, 1)) { + return -EINVAL; + } + return 0; +} + +int hvf_arch_remove_sw_breakpoint(CPUState *cpu, struct hvf_sw_breakpoint *bp) +{ + static uint32_t brk; + + if (cpu_memory_rw_debug(cpu, bp->pc, (uint8_t *)&brk, 4, 0) || + brk != brk_insn || + cpu_memory_rw_debug(cpu, bp->pc, (uint8_t *)&bp->saved_insn, 4, 1)) { + return -EINVAL; + } + return 0; +} + +int hvf_arch_insert_hw_breakpoint(target_ulong addr, target_ulong len, int type) +{ + switch (type) { + case GDB_BREAKPOINT_HW: + return insert_hw_breakpoint(addr); + case GDB_WATCHPOINT_READ: + case GDB_WATCHPOINT_WRITE: + case GDB_WATCHPOINT_ACCESS: + return insert_hw_watchpoint(addr, len, type); + default: + return -ENOSYS; + } +} + +int hvf_arch_remove_hw_breakpoint(target_ulong addr, target_ulong len, int type) +{ + switch (type) { + case GDB_BREAKPOINT_HW: + return delete_hw_breakpoint(addr); + case GDB_WATCHPOINT_READ: + case GDB_WATCHPOINT_WRITE: + case GDB_WATCHPOINT_ACCESS: + return delete_hw_watchpoint(addr, len, type); + default: + return -ENOSYS; + } +} + +void hvf_arch_remove_all_hw_breakpoints(void) +{ + if (cur_hw_wps > 0) { + g_array_remove_range(hw_watchpoints, 0, cur_hw_wps); + } + if (cur_hw_bps > 0) { + g_array_remove_range(hw_breakpoints, 0, cur_hw_bps); + } +} diff --git a/target/i386/hvf/hvf.c b/target/i386/hvf/hvf.c index XXXXXXX..XXXXXXX 100644 --- a/target/i386/hvf/hvf.c +++ b/target/i386/hvf/hvf.c @@ -XXX,XX +XXX,XX @@ int hvf_vcpu_exec(CPUState *cpu) return ret; } + +int hvf_arch_insert_sw_breakpoint(CPUState *cpu, struct hvf_sw_breakpoint *bp) +{ + return -ENOSYS; +} + +int hvf_arch_remove_sw_breakpoint(CPUState *cpu, struct hvf_sw_breakpoint *bp) +{ + return -ENOSYS; +} + +int hvf_arch_insert_hw_breakpoint(target_ulong addr, target_ulong len, int type) +{ + return -ENOSYS; +} + +int hvf_arch_remove_hw_breakpoint(target_ulong addr, target_ulong len, int type) +{ + return -ENOSYS; +} + +void hvf_arch_remove_all_hw_breakpoints(void) +{ +} -- 2.34.1 From: Francesco Cagnin <fcagnin@quarkslab.com> Guests can now be debugged through the gdbstub. Support is added for single-stepping, software breakpoints, hardware breakpoints and watchpoints. The code has been structured like the KVM counterpart. While guest debugging is enabled, the guest can still read and write the DBG*_EL1 registers but they don't have any effect. Signed-off-by: Francesco Cagnin <fcagnin@quarkslab.com> Message-id: 20230601153107.81955-5-fcagnin@quarkslab.com Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/sysemu/hvf.h | 15 ++ include/sysemu/hvf_int.h | 1 + target/arm/hvf_arm.h | 7 + accel/hvf/hvf-accel-ops.c | 10 + accel/hvf/hvf-all.c | 6 + target/arm/hvf/hvf.c | 474 +++++++++++++++++++++++++++++++++++++- target/i386/hvf/hvf.c | 9 + 7 files changed, 520 insertions(+), 2 deletions(-) diff --git a/include/sysemu/hvf.h b/include/sysemu/hvf.h index XXXXXXX..XXXXXXX 100644 --- a/include/sysemu/hvf.h +++ b/include/sysemu/hvf.h @@ -XXX,XX +XXX,XX @@ int hvf_arch_insert_hw_breakpoint(target_ulong addr, target_ulong len, int hvf_arch_remove_hw_breakpoint(target_ulong addr, target_ulong len, int type); void hvf_arch_remove_all_hw_breakpoints(void); + +/* + * hvf_update_guest_debug: + * @cs: CPUState for the CPU to update + * + * Update guest to enable or disable debugging. Per-arch specifics will be + * handled by calling down to hvf_arch_update_guest_debug. + */ +int hvf_update_guest_debug(CPUState *cpu); +void hvf_arch_update_guest_debug(CPUState *cpu); + +/* + * Return whether the guest supports debugging. + */ +bool hvf_arch_supports_guest_debug(void); #endif /* NEED_CPU_H */ #endif diff --git a/include/sysemu/hvf_int.h b/include/sysemu/hvf_int.h index XXXXXXX..XXXXXXX 100644 --- a/include/sysemu/hvf_int.h +++ b/include/sysemu/hvf_int.h @@ -XXX,XX +XXX,XX @@ struct hvf_vcpu_state { void *exit; bool vtimer_masked; sigset_t unblock_ipi_mask; + bool guest_debug_enabled; }; void assert_hvf_ok(hv_return_t ret); diff --git a/target/arm/hvf_arm.h b/target/arm/hvf_arm.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/hvf_arm.h +++ b/target/arm/hvf_arm.h @@ -XXX,XX +XXX,XX @@ #include "cpu.h" +/** + * hvf_arm_init_debug() - initialize guest debug capabilities + * + * Should be called only once before using guest debug capabilities. + */ +void hvf_arm_init_debug(void); + void hvf_arm_set_cpu_features_from_host(ARMCPU *cpu); #endif diff --git a/accel/hvf/hvf-accel-ops.c b/accel/hvf/hvf-accel-ops.c index XXXXXXX..XXXXXXX 100644 --- a/accel/hvf/hvf-accel-ops.c +++ b/accel/hvf/hvf-accel-ops.c @@ -XXX,XX +XXX,XX @@ static int hvf_accel_init(MachineState *ms) return hvf_arch_init(); } +static inline int hvf_gdbstub_sstep_flags(void) +{ + return SSTEP_ENABLE | SSTEP_NOIRQ; +} + static void hvf_accel_class_init(ObjectClass *oc, void *data) { AccelClass *ac = ACCEL_CLASS(oc); ac->name = "HVF"; ac->init_machine = hvf_accel_init; ac->allowed = &hvf_allowed; + ac->gdbstub_supported_sstep_flags = hvf_gdbstub_sstep_flags; } static const TypeInfo hvf_accel_type = { @@ -XXX,XX +XXX,XX @@ static int hvf_init_vcpu(CPUState *cpu) cpu->vcpu_dirty = 1; assert_hvf_ok(r); + cpu->hvf->guest_debug_enabled = false; + return hvf_arch_init_vcpu(cpu); } @@ -XXX,XX +XXX,XX @@ static void hvf_accel_ops_class_init(ObjectClass *oc, void *data) ops->insert_breakpoint = hvf_insert_breakpoint; ops->remove_breakpoint = hvf_remove_breakpoint; ops->remove_all_breakpoints = hvf_remove_all_breakpoints; + ops->update_guest_debug = hvf_update_guest_debug; + ops->supports_guest_debug = hvf_arch_supports_guest_debug; }; static const TypeInfo hvf_accel_ops_type = { .name = ACCEL_OPS_NAME("hvf"), diff --git a/accel/hvf/hvf-all.c b/accel/hvf/hvf-all.c index XXXXXXX..XXXXXXX 100644 --- a/accel/hvf/hvf-all.c +++ b/accel/hvf/hvf-all.c @@ -XXX,XX +XXX,XX @@ int hvf_sw_breakpoints_active(CPUState *cpu) { return !QTAILQ_EMPTY(&hvf_state->hvf_sw_breakpoints); } + +int hvf_update_guest_debug(CPUState *cpu) +{ + hvf_arch_update_guest_debug(cpu); + return 0; +} diff --git a/target/arm/hvf/hvf.c b/target/arm/hvf/hvf.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/hvf/hvf.c +++ b/target/arm/hvf/hvf.c @@ -XXX,XX +XXX,XX @@ #include "exec/gdbstub.h" +#define MDSCR_EL1_SS_SHIFT 0 +#define MDSCR_EL1_MDE_SHIFT 15 + +static uint16_t dbgbcr_regs[] = { + HV_SYS_REG_DBGBCR0_EL1, + HV_SYS_REG_DBGBCR1_EL1, + HV_SYS_REG_DBGBCR2_EL1, + HV_SYS_REG_DBGBCR3_EL1, + HV_SYS_REG_DBGBCR4_EL1, + HV_SYS_REG_DBGBCR5_EL1, + HV_SYS_REG_DBGBCR6_EL1, + HV_SYS_REG_DBGBCR7_EL1, + HV_SYS_REG_DBGBCR8_EL1, + HV_SYS_REG_DBGBCR9_EL1, + HV_SYS_REG_DBGBCR10_EL1, + HV_SYS_REG_DBGBCR11_EL1, + HV_SYS_REG_DBGBCR12_EL1, + HV_SYS_REG_DBGBCR13_EL1, + HV_SYS_REG_DBGBCR14_EL1, + HV_SYS_REG_DBGBCR15_EL1, +}; +static uint16_t dbgbvr_regs[] = { + HV_SYS_REG_DBGBVR0_EL1, + HV_SYS_REG_DBGBVR1_EL1, + HV_SYS_REG_DBGBVR2_EL1, + HV_SYS_REG_DBGBVR3_EL1, + HV_SYS_REG_DBGBVR4_EL1, + HV_SYS_REG_DBGBVR5_EL1, + HV_SYS_REG_DBGBVR6_EL1, + HV_SYS_REG_DBGBVR7_EL1, + HV_SYS_REG_DBGBVR8_EL1, + HV_SYS_REG_DBGBVR9_EL1, + HV_SYS_REG_DBGBVR10_EL1, + HV_SYS_REG_DBGBVR11_EL1, + HV_SYS_REG_DBGBVR12_EL1, + HV_SYS_REG_DBGBVR13_EL1, + HV_SYS_REG_DBGBVR14_EL1, + HV_SYS_REG_DBGBVR15_EL1, +}; +static uint16_t dbgwcr_regs[] = { + HV_SYS_REG_DBGWCR0_EL1, + HV_SYS_REG_DBGWCR1_EL1, + HV_SYS_REG_DBGWCR2_EL1, + HV_SYS_REG_DBGWCR3_EL1, + HV_SYS_REG_DBGWCR4_EL1, + HV_SYS_REG_DBGWCR5_EL1, + HV_SYS_REG_DBGWCR6_EL1, + HV_SYS_REG_DBGWCR7_EL1, + HV_SYS_REG_DBGWCR8_EL1, + HV_SYS_REG_DBGWCR9_EL1, + HV_SYS_REG_DBGWCR10_EL1, + HV_SYS_REG_DBGWCR11_EL1, + HV_SYS_REG_DBGWCR12_EL1, + HV_SYS_REG_DBGWCR13_EL1, + HV_SYS_REG_DBGWCR14_EL1, + HV_SYS_REG_DBGWCR15_EL1, +}; +static uint16_t dbgwvr_regs[] = { + HV_SYS_REG_DBGWVR0_EL1, + HV_SYS_REG_DBGWVR1_EL1, + HV_SYS_REG_DBGWVR2_EL1, + HV_SYS_REG_DBGWVR3_EL1, + HV_SYS_REG_DBGWVR4_EL1, + HV_SYS_REG_DBGWVR5_EL1, + HV_SYS_REG_DBGWVR6_EL1, + HV_SYS_REG_DBGWVR7_EL1, + HV_SYS_REG_DBGWVR8_EL1, + HV_SYS_REG_DBGWVR9_EL1, + HV_SYS_REG_DBGWVR10_EL1, + HV_SYS_REG_DBGWVR11_EL1, + HV_SYS_REG_DBGWVR12_EL1, + HV_SYS_REG_DBGWVR13_EL1, + HV_SYS_REG_DBGWVR14_EL1, + HV_SYS_REG_DBGWVR15_EL1, +}; + +static inline int hvf_arm_num_brps(hv_vcpu_config_t config) +{ + uint64_t val; + hv_return_t ret; + ret = hv_vcpu_config_get_feature_reg(config, HV_FEATURE_REG_ID_AA64DFR0_EL1, + &val); + assert_hvf_ok(ret); + return FIELD_EX64(val, ID_AA64DFR0, BRPS) + 1; +} + +static inline int hvf_arm_num_wrps(hv_vcpu_config_t config) +{ + uint64_t val; + hv_return_t ret; + ret = hv_vcpu_config_get_feature_reg(config, HV_FEATURE_REG_ID_AA64DFR0_EL1, + &val); + assert_hvf_ok(ret); + return FIELD_EX64(val, ID_AA64DFR0, WRPS) + 1; +} + +void hvf_arm_init_debug(void) +{ + hv_vcpu_config_t config; + config = hv_vcpu_config_create(); + + max_hw_bps = hvf_arm_num_brps(config); + hw_breakpoints = + g_array_sized_new(true, true, sizeof(HWBreakpoint), max_hw_bps); + + max_hw_wps = hvf_arm_num_wrps(config); + hw_watchpoints = + g_array_sized_new(true, true, sizeof(HWWatchpoint), max_hw_wps); +} + #define HVF_SYSREG(crn, crm, op0, op1, op2) \ ENCODE_AA64_CP_REG(CP_REG_ARM64_SYSREG_CP, crn, crm, op0, op1, op2) #define PL1_WRITE_MASK 0x4 @@ -XXX,XX +XXX,XX @@ int hvf_get_registers(CPUState *cpu) continue; } + if (cpu->hvf->guest_debug_enabled) { + /* Handle debug registers */ + switch (hvf_sreg_match[i].reg) { + case HV_SYS_REG_DBGBVR0_EL1: + case HV_SYS_REG_DBGBCR0_EL1: + case HV_SYS_REG_DBGWVR0_EL1: + case HV_SYS_REG_DBGWCR0_EL1: + case HV_SYS_REG_DBGBVR1_EL1: + case HV_SYS_REG_DBGBCR1_EL1: + case HV_SYS_REG_DBGWVR1_EL1: + case HV_SYS_REG_DBGWCR1_EL1: + case HV_SYS_REG_DBGBVR2_EL1: + case HV_SYS_REG_DBGBCR2_EL1: + case HV_SYS_REG_DBGWVR2_EL1: + case HV_SYS_REG_DBGWCR2_EL1: + case HV_SYS_REG_DBGBVR3_EL1: + case HV_SYS_REG_DBGBCR3_EL1: + case HV_SYS_REG_DBGWVR3_EL1: + case HV_SYS_REG_DBGWCR3_EL1: + case HV_SYS_REG_DBGBVR4_EL1: + case HV_SYS_REG_DBGBCR4_EL1: + case HV_SYS_REG_DBGWVR4_EL1: + case HV_SYS_REG_DBGWCR4_EL1: + case HV_SYS_REG_DBGBVR5_EL1: + case HV_SYS_REG_DBGBCR5_EL1: + case HV_SYS_REG_DBGWVR5_EL1: + case HV_SYS_REG_DBGWCR5_EL1: + case HV_SYS_REG_DBGBVR6_EL1: + case HV_SYS_REG_DBGBCR6_EL1: + case HV_SYS_REG_DBGWVR6_EL1: + case HV_SYS_REG_DBGWCR6_EL1: + case HV_SYS_REG_DBGBVR7_EL1: + case HV_SYS_REG_DBGBCR7_EL1: + case HV_SYS_REG_DBGWVR7_EL1: + case HV_SYS_REG_DBGWCR7_EL1: + case HV_SYS_REG_DBGBVR8_EL1: + case HV_SYS_REG_DBGBCR8_EL1: + case HV_SYS_REG_DBGWVR8_EL1: + case HV_SYS_REG_DBGWCR8_EL1: + case HV_SYS_REG_DBGBVR9_EL1: + case HV_SYS_REG_DBGBCR9_EL1: + case HV_SYS_REG_DBGWVR9_EL1: + case HV_SYS_REG_DBGWCR9_EL1: + case HV_SYS_REG_DBGBVR10_EL1: + case HV_SYS_REG_DBGBCR10_EL1: + case HV_SYS_REG_DBGWVR10_EL1: + case HV_SYS_REG_DBGWCR10_EL1: + case HV_SYS_REG_DBGBVR11_EL1: + case HV_SYS_REG_DBGBCR11_EL1: + case HV_SYS_REG_DBGWVR11_EL1: + case HV_SYS_REG_DBGWCR11_EL1: + case HV_SYS_REG_DBGBVR12_EL1: + case HV_SYS_REG_DBGBCR12_EL1: + case HV_SYS_REG_DBGWVR12_EL1: + case HV_SYS_REG_DBGWCR12_EL1: + case HV_SYS_REG_DBGBVR13_EL1: + case HV_SYS_REG_DBGBCR13_EL1: + case HV_SYS_REG_DBGWVR13_EL1: + case HV_SYS_REG_DBGWCR13_EL1: + case HV_SYS_REG_DBGBVR14_EL1: + case HV_SYS_REG_DBGBCR14_EL1: + case HV_SYS_REG_DBGWVR14_EL1: + case HV_SYS_REG_DBGWCR14_EL1: + case HV_SYS_REG_DBGBVR15_EL1: + case HV_SYS_REG_DBGBCR15_EL1: + case HV_SYS_REG_DBGWVR15_EL1: + case HV_SYS_REG_DBGWCR15_EL1: { + /* + * If the guest is being debugged, the vCPU's debug registers + * are holding the gdbstub's view of the registers (set in + * hvf_arch_update_guest_debug()). + * Since the environment is used to store only the guest's view + * of the registers, don't update it with the values from the + * vCPU but simply keep the values from the previous + * environment. + */ + const ARMCPRegInfo *ri; + ri = get_arm_cp_reginfo(arm_cpu->cp_regs, hvf_sreg_match[i].key); + val = read_raw_cp_reg(env, ri); + + arm_cpu->cpreg_values[hvf_sreg_match[i].cp_idx] = val; + continue; + } + } + } + ret = hv_vcpu_get_sys_reg(cpu->hvf->fd, hvf_sreg_match[i].reg, &val); assert_hvf_ok(ret); @@ -XXX,XX +XXX,XX @@ int hvf_put_registers(CPUState *cpu) continue; } + if (cpu->hvf->guest_debug_enabled) { + /* Handle debug registers */ + switch (hvf_sreg_match[i].reg) { + case HV_SYS_REG_DBGBVR0_EL1: + case HV_SYS_REG_DBGBCR0_EL1: + case HV_SYS_REG_DBGWVR0_EL1: + case HV_SYS_REG_DBGWCR0_EL1: + case HV_SYS_REG_DBGBVR1_EL1: + case HV_SYS_REG_DBGBCR1_EL1: + case HV_SYS_REG_DBGWVR1_EL1: + case HV_SYS_REG_DBGWCR1_EL1: + case HV_SYS_REG_DBGBVR2_EL1: + case HV_SYS_REG_DBGBCR2_EL1: + case HV_SYS_REG_DBGWVR2_EL1: + case HV_SYS_REG_DBGWCR2_EL1: + case HV_SYS_REG_DBGBVR3_EL1: + case HV_SYS_REG_DBGBCR3_EL1: + case HV_SYS_REG_DBGWVR3_EL1: + case HV_SYS_REG_DBGWCR3_EL1: + case HV_SYS_REG_DBGBVR4_EL1: + case HV_SYS_REG_DBGBCR4_EL1: + case HV_SYS_REG_DBGWVR4_EL1: + case HV_SYS_REG_DBGWCR4_EL1: + case HV_SYS_REG_DBGBVR5_EL1: + case HV_SYS_REG_DBGBCR5_EL1: + case HV_SYS_REG_DBGWVR5_EL1: + case HV_SYS_REG_DBGWCR5_EL1: + case HV_SYS_REG_DBGBVR6_EL1: + case HV_SYS_REG_DBGBCR6_EL1: + case HV_SYS_REG_DBGWVR6_EL1: + case HV_SYS_REG_DBGWCR6_EL1: + case HV_SYS_REG_DBGBVR7_EL1: + case HV_SYS_REG_DBGBCR7_EL1: + case HV_SYS_REG_DBGWVR7_EL1: + case HV_SYS_REG_DBGWCR7_EL1: + case HV_SYS_REG_DBGBVR8_EL1: + case HV_SYS_REG_DBGBCR8_EL1: + case HV_SYS_REG_DBGWVR8_EL1: + case HV_SYS_REG_DBGWCR8_EL1: + case HV_SYS_REG_DBGBVR9_EL1: + case HV_SYS_REG_DBGBCR9_EL1: + case HV_SYS_REG_DBGWVR9_EL1: + case HV_SYS_REG_DBGWCR9_EL1: + case HV_SYS_REG_DBGBVR10_EL1: + case HV_SYS_REG_DBGBCR10_EL1: + case HV_SYS_REG_DBGWVR10_EL1: + case HV_SYS_REG_DBGWCR10_EL1: + case HV_SYS_REG_DBGBVR11_EL1: + case HV_SYS_REG_DBGBCR11_EL1: + case HV_SYS_REG_DBGWVR11_EL1: + case HV_SYS_REG_DBGWCR11_EL1: + case HV_SYS_REG_DBGBVR12_EL1: + case HV_SYS_REG_DBGBCR12_EL1: + case HV_SYS_REG_DBGWVR12_EL1: + case HV_SYS_REG_DBGWCR12_EL1: + case HV_SYS_REG_DBGBVR13_EL1: + case HV_SYS_REG_DBGBCR13_EL1: + case HV_SYS_REG_DBGWVR13_EL1: + case HV_SYS_REG_DBGWCR13_EL1: + case HV_SYS_REG_DBGBVR14_EL1: + case HV_SYS_REG_DBGBCR14_EL1: + case HV_SYS_REG_DBGWVR14_EL1: + case HV_SYS_REG_DBGWCR14_EL1: + case HV_SYS_REG_DBGBVR15_EL1: + case HV_SYS_REG_DBGBCR15_EL1: + case HV_SYS_REG_DBGWVR15_EL1: + case HV_SYS_REG_DBGWCR15_EL1: + /* + * If the guest is being debugged, the vCPU's debug registers + * are already holding the gdbstub's view of the registers (set + * in hvf_arch_update_guest_debug()). + */ + continue; + } + } + val = arm_cpu->cpreg_values[hvf_sreg_match[i].cp_idx]; ret = hv_vcpu_set_sys_reg(cpu->hvf->fd, hvf_sreg_match[i].reg, val); assert_hvf_ok(ret); @@ -XXX,XX +XXX,XX @@ int hvf_vcpu_exec(CPUState *cpu) { ARMCPU *arm_cpu = ARM_CPU(cpu); CPUARMState *env = &arm_cpu->env; + int ret; hv_vcpu_exit_t *hvf_exit = cpu->hvf->exit; hv_return_t r; bool advance_pc = false; - if (hvf_inject_interrupts(cpu)) { + if (!(cpu->singlestep_enabled & SSTEP_NOIRQ) && + hvf_inject_interrupts(cpu)) { return EXCP_INTERRUPT; } @@ -XXX,XX +XXX,XX @@ int hvf_vcpu_exec(CPUState *cpu) uint64_t syndrome = hvf_exit->exception.syndrome; uint32_t ec = syn_get_ec(syndrome); + ret = 0; qemu_mutex_lock_iothread(); switch (exit_reason) { case HV_EXIT_REASON_EXCEPTION: @@ -XXX,XX +XXX,XX @@ int hvf_vcpu_exec(CPUState *cpu) hvf_sync_vtimer(cpu); switch (ec) { + case EC_SOFTWARESTEP: { + ret = EXCP_DEBUG; + + if (!cpu->singlestep_enabled) { + error_report("EC_SOFTWARESTEP but single-stepping not enabled"); + } + break; + } + case EC_AA64_BKPT: { + ret = EXCP_DEBUG; + + cpu_synchronize_state(cpu); + + if (!hvf_find_sw_breakpoint(cpu, env->pc)) { + /* Re-inject into the guest */ + ret = 0; + hvf_raise_exception(cpu, EXCP_BKPT, syn_aa64_bkpt(0)); + } + break; + } + case EC_BREAKPOINT: { + ret = EXCP_DEBUG; + + cpu_synchronize_state(cpu); + + if (!find_hw_breakpoint(cpu, env->pc)) { + error_report("EC_BREAKPOINT but unknown hw breakpoint"); + } + break; + } + case EC_WATCHPOINT: { + ret = EXCP_DEBUG; + + cpu_synchronize_state(cpu); + + CPUWatchpoint *wp = + find_hw_watchpoint(cpu, hvf_exit->exception.virtual_address); + if (!wp) { + error_report("EXCP_DEBUG but unknown hw watchpoint"); + } + cpu->watchpoint_hit = wp; + break; + } case EC_DATAABORT: { bool isv = syndrome & ARM_EL_ISV; bool iswrite = (syndrome >> 6) & 1; @@ -XXX,XX +XXX,XX @@ int hvf_vcpu_exec(CPUState *cpu) pc += 4; r = hv_vcpu_set_reg(cpu->hvf->fd, HV_REG_PC, pc); assert_hvf_ok(r); + + /* Handle single-stepping over instructions which trigger a VM exit */ + if (cpu->singlestep_enabled) { + ret = EXCP_DEBUG; + } } - return 0; + return ret; } static const VMStateDescription vmstate_hvf_vtimer = { @@ -XXX,XX +XXX,XX @@ int hvf_arch_init(void) hvf_state->vtimer_offset = mach_absolute_time(); vmstate_register(NULL, 0, &vmstate_hvf_vtimer, &vtimer); qemu_add_vm_change_state_handler(hvf_vm_state_change, &vtimer); + + hvf_arm_init_debug(); + return 0; } @@ -XXX,XX +XXX,XX @@ void hvf_arch_remove_all_hw_breakpoints(void) g_array_remove_range(hw_breakpoints, 0, cur_hw_bps); } } + +/* + * Update the vCPU with the gdbstub's view of debug registers. This view + * consists of all hardware breakpoints and watchpoints inserted so far while + * debugging the guest. + */ +static void hvf_put_gdbstub_debug_registers(CPUState *cpu) +{ + hv_return_t r = HV_SUCCESS; + int i; + + for (i = 0; i < cur_hw_bps; i++) { + HWBreakpoint *bp = get_hw_bp(i); + r = hv_vcpu_set_sys_reg(cpu->hvf->fd, dbgbcr_regs[i], bp->bcr); + assert_hvf_ok(r); + r = hv_vcpu_set_sys_reg(cpu->hvf->fd, dbgbvr_regs[i], bp->bvr); + assert_hvf_ok(r); + } + for (i = cur_hw_bps; i < max_hw_bps; i++) { + r = hv_vcpu_set_sys_reg(cpu->hvf->fd, dbgbcr_regs[i], 0); + assert_hvf_ok(r); + r = hv_vcpu_set_sys_reg(cpu->hvf->fd, dbgbvr_regs[i], 0); + assert_hvf_ok(r); + } + + for (i = 0; i < cur_hw_wps; i++) { + HWWatchpoint *wp = get_hw_wp(i); + r = hv_vcpu_set_sys_reg(cpu->hvf->fd, dbgwcr_regs[i], wp->wcr); + assert_hvf_ok(r); + r = hv_vcpu_set_sys_reg(cpu->hvf->fd, dbgwvr_regs[i], wp->wvr); + assert_hvf_ok(r); + } + for (i = cur_hw_wps; i < max_hw_wps; i++) { + r = hv_vcpu_set_sys_reg(cpu->hvf->fd, dbgwcr_regs[i], 0); + assert_hvf_ok(r); + r = hv_vcpu_set_sys_reg(cpu->hvf->fd, dbgwvr_regs[i], 0); + assert_hvf_ok(r); + } +} + +/* + * Update the vCPU with the guest's view of debug registers. This view is kept + * in the environment at all times. + */ +static void hvf_put_guest_debug_registers(CPUState *cpu) +{ + ARMCPU *arm_cpu = ARM_CPU(cpu); + CPUARMState *env = &arm_cpu->env; + hv_return_t r = HV_SUCCESS; + int i; + + for (i = 0; i < max_hw_bps; i++) { + r = hv_vcpu_set_sys_reg(cpu->hvf->fd, dbgbcr_regs[i], + env->cp15.dbgbcr[i]); + assert_hvf_ok(r); + r = hv_vcpu_set_sys_reg(cpu->hvf->fd, dbgbvr_regs[i], + env->cp15.dbgbvr[i]); + assert_hvf_ok(r); + } + + for (i = 0; i < max_hw_wps; i++) { + r = hv_vcpu_set_sys_reg(cpu->hvf->fd, dbgwcr_regs[i], + env->cp15.dbgwcr[i]); + assert_hvf_ok(r); + r = hv_vcpu_set_sys_reg(cpu->hvf->fd, dbgwvr_regs[i], + env->cp15.dbgwvr[i]); + assert_hvf_ok(r); + } +} + +static inline bool hvf_arm_hw_debug_active(CPUState *cpu) +{ + return ((cur_hw_wps > 0) || (cur_hw_bps > 0)); +} + +static void hvf_arch_set_traps(void) +{ + CPUState *cpu; + bool should_enable_traps = false; + hv_return_t r = HV_SUCCESS; + + /* Check whether guest debugging is enabled for at least one vCPU; if it + * is, enable exiting the guest on all vCPUs */ + CPU_FOREACH(cpu) { + should_enable_traps |= cpu->hvf->guest_debug_enabled; + } + CPU_FOREACH(cpu) { + /* Set whether debug exceptions exit the guest */ + r = hv_vcpu_set_trap_debug_exceptions(cpu->hvf->fd, + should_enable_traps); + assert_hvf_ok(r); + + /* Set whether accesses to debug registers exit the guest */ + r = hv_vcpu_set_trap_debug_reg_accesses(cpu->hvf->fd, + should_enable_traps); + assert_hvf_ok(r); + } +} + +void hvf_arch_update_guest_debug(CPUState *cpu) +{ + ARMCPU *arm_cpu = ARM_CPU(cpu); + CPUARMState *env = &arm_cpu->env; + + /* Check whether guest debugging is enabled */ + cpu->hvf->guest_debug_enabled = cpu->singlestep_enabled || + hvf_sw_breakpoints_active(cpu) || + hvf_arm_hw_debug_active(cpu); + + /* Update debug registers */ + if (cpu->hvf->guest_debug_enabled) { + hvf_put_gdbstub_debug_registers(cpu); + } else { + hvf_put_guest_debug_registers(cpu); + } + + cpu_synchronize_state(cpu); + + /* Enable/disable single-stepping */ + if (cpu->singlestep_enabled) { + env->cp15.mdscr_el1 = + deposit64(env->cp15.mdscr_el1, MDSCR_EL1_SS_SHIFT, 1, 1); + pstate_write(env, pstate_read(env) | PSTATE_SS); + } else { + env->cp15.mdscr_el1 = + deposit64(env->cp15.mdscr_el1, MDSCR_EL1_SS_SHIFT, 1, 0); + } + + /* Enable/disable Breakpoint exceptions */ + if (hvf_arm_hw_debug_active(cpu)) { + env->cp15.mdscr_el1 = + deposit64(env->cp15.mdscr_el1, MDSCR_EL1_MDE_SHIFT, 1, 1); + } else { + env->cp15.mdscr_el1 = + deposit64(env->cp15.mdscr_el1, MDSCR_EL1_MDE_SHIFT, 1, 0); + } + + hvf_arch_set_traps(); +} + +inline bool hvf_arch_supports_guest_debug(void) +{ + return true; +} diff --git a/target/i386/hvf/hvf.c b/target/i386/hvf/hvf.c index XXXXXXX..XXXXXXX 100644 --- a/target/i386/hvf/hvf.c +++ b/target/i386/hvf/hvf.c @@ -XXX,XX +XXX,XX @@ int hvf_arch_remove_hw_breakpoint(target_ulong addr, target_ulong len, int type) void hvf_arch_remove_all_hw_breakpoints(void) { } + +void hvf_arch_update_guest_debug(CPUState *cpu) +{ +} + +inline bool hvf_arch_supports_guest_debug(void) +{ + return false; +} -- 2.34.1 From: Vikram Garhwal <vikram.garhwal@amd.com> The Xilinx Versal CANFD controller is developed based on SocketCAN, QEMU CAN bus implementation. Bus connection and socketCAN connection for each CAN module can be set through command lines. Signed-off-by: Vikram Garhwal <vikram.garhwal@amd.com> Reviewed-by: Francisco Iglesias <frasse.iglesias@gmail.com> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/net/xlnx-versal-canfd.h | 87 ++ hw/net/can/xlnx-versal-canfd.c | 2107 ++++++++++++++++++++++++++++ hw/net/can/meson.build | 1 + hw/net/can/trace-events | 7 + 4 files changed, 2202 insertions(+) create mode 100644 include/hw/net/xlnx-versal-canfd.h create mode 100644 hw/net/can/xlnx-versal-canfd.c diff --git a/include/hw/net/xlnx-versal-canfd.h b/include/hw/net/xlnx-versal-canfd.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/include/hw/net/xlnx-versal-canfd.h @@ -XXX,XX +XXX,XX @@ +/* + * QEMU model of the Xilinx Versal CANFD Controller. + * + * Copyright (c) 2023 Advanced Micro Devices, Inc. + * + * Written-by: Vikram Garhwal<vikram.garhwal@amd.com> + * Based on QEMU CANFD Device emulation implemented by Jin Yang, Deniz Eren and + * Pavel Pisa. + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN + * THE SOFTWARE. + */ + +#ifndef HW_CANFD_XILINX_H +#define HW_CANFD_XILINX_H + +#include "hw/register.h" +#include "hw/ptimer.h" +#include "net/can_emu.h" +#include "hw/qdev-clock.h" + +#define TYPE_XILINX_CANFD "xlnx.versal-canfd" + +OBJECT_DECLARE_SIMPLE_TYPE(XlnxVersalCANFDState, XILINX_CANFD) + +#define NUM_REGS_PER_MSG_SPACE 18 /* 1 ID + 1 DLC + 16 Data(DW0 - DW15) regs. */ +#define MAX_NUM_RX 64 +#define OFFSET_RX1_DW15 (0x4144 / 4) +#define CANFD_TIMER_MAX 0xFFFFUL +#define CANFD_DEFAULT_CLOCK (25 * 1000 * 1000) + +#define XLNX_VERSAL_CANFD_R_MAX (OFFSET_RX1_DW15 + \ + ((MAX_NUM_RX - 1) * NUM_REGS_PER_MSG_SPACE) + 1) + +typedef struct XlnxVersalCANFDState { + SysBusDevice parent_obj; + MemoryRegion iomem; + + qemu_irq irq_canfd_int; + qemu_irq irq_addr_err; + + RegisterInfo reg_info[XLNX_VERSAL_CANFD_R_MAX]; + RegisterAccessInfo *tx_regs; + RegisterAccessInfo *rx0_regs; + RegisterAccessInfo *rx1_regs; + RegisterAccessInfo *af_regs; + RegisterAccessInfo *txe_regs; + RegisterAccessInfo *rx_mailbox_regs; + RegisterAccessInfo *af_mask_regs_mailbox; + + uint32_t regs[XLNX_VERSAL_CANFD_R_MAX]; + + ptimer_state *canfd_timer; + + CanBusClientState bus_client; + CanBusState *canfdbus; + + struct { + uint8_t rx0_fifo; + uint8_t rx1_fifo; + uint8_t tx_fifo; + bool enable_rx_fifo1; + uint32_t ext_clk_freq; + } cfg; + +} XlnxVersalCANFDState; + +typedef struct tx_ready_reg_info { + uint32_t can_id; + uint32_t reg_num; +} tx_ready_reg_info; + +#endif diff --git a/hw/net/can/xlnx-versal-canfd.c b/hw/net/can/xlnx-versal-canfd.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/net/can/xlnx-versal-canfd.c @@ -XXX,XX +XXX,XX @@ +/* + * QEMU model of the Xilinx Versal CANFD device. + * + * This implementation is based on the following datasheet: + * https://docs.xilinx.com/v/u/2.0-English/pg223-canfd + * + * Copyright (c) 2023 Advanced Micro Devices, Inc. + * + * Written-by: Vikram Garhwal <vikram.garhwal@amd.com> + * + * Based on QEMU CANFD Device emulation implemented by Jin Yang, Deniz Eren and + * Pavel Pisa + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN + * THE SOFTWARE. + */ + +#include "qemu/osdep.h" +#include "hw/sysbus.h" +#include "hw/irq.h" +#include "hw/register.h" +#include "qapi/error.h" +#include "qemu/bitops.h" +#include "qemu/log.h" +#include "qemu/cutils.h" +#include "qemu/event_notifier.h" +#include "hw/qdev-properties.h" +#include "qom/object_interfaces.h" +#include "migration/vmstate.h" +#include "hw/net/xlnx-versal-canfd.h" +#include "trace.h" + +REG32(SOFTWARE_RESET_REGISTER, 0x0) + FIELD(SOFTWARE_RESET_REGISTER, CEN, 1, 1) + FIELD(SOFTWARE_RESET_REGISTER, SRST, 0, 1) +REG32(MODE_SELECT_REGISTER, 0x4) + FIELD(MODE_SELECT_REGISTER, ITO, 8, 8) + FIELD(MODE_SELECT_REGISTER, ABR, 7, 1) + FIELD(MODE_SELECT_REGISTER, SBR, 6, 1) + FIELD(MODE_SELECT_REGISTER, DPEE, 5, 1) + FIELD(MODE_SELECT_REGISTER, DAR, 4, 1) + FIELD(MODE_SELECT_REGISTER, BRSD, 3, 1) + FIELD(MODE_SELECT_REGISTER, SNOOP, 2, 1) + FIELD(MODE_SELECT_REGISTER, LBACK, 1, 1) + FIELD(MODE_SELECT_REGISTER, SLEEP, 0, 1) +REG32(ARBITRATION_PHASE_BAUD_RATE_PRESCALER_REGISTER, 0x8) + FIELD(ARBITRATION_PHASE_BAUD_RATE_PRESCALER_REGISTER, BRP, 0, 8) +REG32(ARBITRATION_PHASE_BIT_TIMING_REGISTER, 0xc) + FIELD(ARBITRATION_PHASE_BIT_TIMING_REGISTER, SJW, 16, 7) + FIELD(ARBITRATION_PHASE_BIT_TIMING_REGISTER, TS2, 8, 7) + FIELD(ARBITRATION_PHASE_BIT_TIMING_REGISTER, TS1, 0, 8) +REG32(ERROR_COUNTER_REGISTER, 0x10) + FIELD(ERROR_COUNTER_REGISTER, REC, 8, 8) + FIELD(ERROR_COUNTER_REGISTER, TEC, 0, 8) +REG32(ERROR_STATUS_REGISTER, 0x14) + FIELD(ERROR_STATUS_REGISTER, F_BERR, 11, 1) + FIELD(ERROR_STATUS_REGISTER, F_STER, 10, 1) + FIELD(ERROR_STATUS_REGISTER, F_FMER, 9, 1) + FIELD(ERROR_STATUS_REGISTER, F_CRCER, 8, 1) + FIELD(ERROR_STATUS_REGISTER, ACKER, 4, 1) + FIELD(ERROR_STATUS_REGISTER, BERR, 3, 1) + FIELD(ERROR_STATUS_REGISTER, STER, 2, 1) + FIELD(ERROR_STATUS_REGISTER, FMER, 1, 1) + FIELD(ERROR_STATUS_REGISTER, CRCER, 0, 1) +REG32(STATUS_REGISTER, 0x18) + FIELD(STATUS_REGISTER, TDCV, 16, 7) + FIELD(STATUS_REGISTER, SNOOP, 12, 1) + FIELD(STATUS_REGISTER, BSFR_CONFIG, 10, 1) + FIELD(STATUS_REGISTER, PEE_CONFIG, 9, 1) + FIELD(STATUS_REGISTER, ESTAT, 7, 2) + FIELD(STATUS_REGISTER, ERRWRN, 6, 1) + FIELD(STATUS_REGISTER, BBSY, 5, 1) + FIELD(STATUS_REGISTER, BIDLE, 4, 1) + FIELD(STATUS_REGISTER, NORMAL, 3, 1) + FIELD(STATUS_REGISTER, SLEEP, 2, 1) + FIELD(STATUS_REGISTER, LBACK, 1, 1) + FIELD(STATUS_REGISTER, CONFIG, 0, 1) +REG32(INTERRUPT_STATUS_REGISTER, 0x1c) + FIELD(INTERRUPT_STATUS_REGISTER, TXEWMFLL, 31, 1) + FIELD(INTERRUPT_STATUS_REGISTER, TXEOFLW, 30, 1) + FIELD(INTERRUPT_STATUS_REGISTER, RXBOFLW_BI, 24, 6) + FIELD(INTERRUPT_STATUS_REGISTER, RXLRM_BI, 18, 6) + FIELD(INTERRUPT_STATUS_REGISTER, RXMNF, 17, 1) + FIELD(INTERRUPT_STATUS_REGISTER, RXFWMFLL_1, 16, 1) + FIELD(INTERRUPT_STATUS_REGISTER, RXFOFLW_1, 15, 1) + FIELD(INTERRUPT_STATUS_REGISTER, TXCRS, 14, 1) + FIELD(INTERRUPT_STATUS_REGISTER, TXRRS, 13, 1) + FIELD(INTERRUPT_STATUS_REGISTER, RXFWMFLL, 12, 1) + FIELD(INTERRUPT_STATUS_REGISTER, WKUP, 11, 1) + FIELD(INTERRUPT_STATUS_REGISTER, SLP, 10, 1) + FIELD(INTERRUPT_STATUS_REGISTER, BSOFF, 9, 1) + /* + * In the original HW description below bit is named as ERROR but an ERROR + * field name collides with a macro in Windows build. To avoid Windows build + * failures, the bit is renamed to ERROR_BIT. + */ + FIELD(INTERRUPT_STATUS_REGISTER, ERROR_BIT, 8, 1) + FIELD(INTERRUPT_STATUS_REGISTER, RXFOFLW, 6, 1) + FIELD(INTERRUPT_STATUS_REGISTER, TSCNT_OFLW, 5, 1) + FIELD(INTERRUPT_STATUS_REGISTER, RXOK, 4, 1) + FIELD(INTERRUPT_STATUS_REGISTER, BSFRD, 3, 1) + FIELD(INTERRUPT_STATUS_REGISTER, PEE, 2, 1) + FIELD(INTERRUPT_STATUS_REGISTER, TXOK, 1, 1) + FIELD(INTERRUPT_STATUS_REGISTER, ARBLST, 0, 1) +REG32(INTERRUPT_ENABLE_REGISTER, 0x20) + FIELD(INTERRUPT_ENABLE_REGISTER, ETXEWMFLL, 31, 1) + FIELD(INTERRUPT_ENABLE_REGISTER, ETXEOFLW, 30, 1) + FIELD(INTERRUPT_ENABLE_REGISTER, ERXMNF, 17, 1) + FIELD(INTERRUPT_ENABLE_REGISTER, ERXFWMFLL_1, 16, 1) + FIELD(INTERRUPT_ENABLE_REGISTER, ERXFOFLW_1, 15, 1) + FIELD(INTERRUPT_ENABLE_REGISTER, ETXCRS, 14, 1) + FIELD(INTERRUPT_ENABLE_REGISTER, ETXRRS, 13, 1) + FIELD(INTERRUPT_ENABLE_REGISTER, ERXFWMFLL, 12, 1) + FIELD(INTERRUPT_ENABLE_REGISTER, EWKUP, 11, 1) + FIELD(INTERRUPT_ENABLE_REGISTER, ESLP, 10, 1) + FIELD(INTERRUPT_ENABLE_REGISTER, EBSOFF, 9, 1) + FIELD(INTERRUPT_ENABLE_REGISTER, EERROR, 8, 1) + FIELD(INTERRUPT_ENABLE_REGISTER, ERFXOFLW, 6, 1) + FIELD(INTERRUPT_ENABLE_REGISTER, ETSCNT_OFLW, 5, 1) + FIELD(INTERRUPT_ENABLE_REGISTER, ERXOK, 4, 1) + FIELD(INTERRUPT_ENABLE_REGISTER, EBSFRD, 3, 1) + FIELD(INTERRUPT_ENABLE_REGISTER, EPEE, 2, 1) + FIELD(INTERRUPT_ENABLE_REGISTER, ETXOK, 1, 1) + FIELD(INTERRUPT_ENABLE_REGISTER, EARBLOST, 0, 1) +REG32(INTERRUPT_CLEAR_REGISTER, 0x24) + FIELD(INTERRUPT_CLEAR_REGISTER, CTXEWMFLL, 31, 1) + FIELD(INTERRUPT_CLEAR_REGISTER, CTXEOFLW, 30, 1) + FIELD(INTERRUPT_CLEAR_REGISTER, CRXMNF, 17, 1) + FIELD(INTERRUPT_CLEAR_REGISTER, CRXFWMFLL_1, 16, 1) + FIELD(INTERRUPT_CLEAR_REGISTER, CRXFOFLW_1, 15, 1) + FIELD(INTERRUPT_CLEAR_REGISTER, CTXCRS, 14, 1) + FIELD(INTERRUPT_CLEAR_REGISTER, CTXRRS, 13, 1) + FIELD(INTERRUPT_CLEAR_REGISTER, CRXFWMFLL, 12, 1) + FIELD(INTERRUPT_CLEAR_REGISTER, CWKUP, 11, 1) + FIELD(INTERRUPT_CLEAR_REGISTER, CSLP, 10, 1) + FIELD(INTERRUPT_CLEAR_REGISTER, CBSOFF, 9, 1) + FIELD(INTERRUPT_CLEAR_REGISTER, CERROR, 8, 1) + FIELD(INTERRUPT_CLEAR_REGISTER, CRFXOFLW, 6, 1) + FIELD(INTERRUPT_CLEAR_REGISTER, CTSCNT_OFLW, 5, 1) + FIELD(INTERRUPT_CLEAR_REGISTER, CRXOK, 4, 1) + FIELD(INTERRUPT_CLEAR_REGISTER, CBSFRD, 3, 1) + FIELD(INTERRUPT_CLEAR_REGISTER, CPEE, 2, 1) + FIELD(INTERRUPT_CLEAR_REGISTER, CTXOK, 1, 1) + FIELD(INTERRUPT_CLEAR_REGISTER, CARBLOST, 0, 1) +REG32(TIMESTAMP_REGISTER, 0x28) + FIELD(TIMESTAMP_REGISTER, TIMESTAMP_CNT, 16, 16) + FIELD(TIMESTAMP_REGISTER, CTS, 0, 1) +REG32(DATA_PHASE_BAUD_RATE_PRESCALER_REGISTER, 0x88) + FIELD(DATA_PHASE_BAUD_RATE_PRESCALER_REGISTER, TDC, 16, 1) + FIELD(DATA_PHASE_BAUD_RATE_PRESCALER_REGISTER, TDCOFF, 8, 6) + FIELD(DATA_PHASE_BAUD_RATE_PRESCALER_REGISTER, DP_BRP, 0, 8) +REG32(DATA_PHASE_BIT_TIMING_REGISTER, 0x8c) + FIELD(DATA_PHASE_BIT_TIMING_REGISTER, DP_SJW, 16, 4) + FIELD(DATA_PHASE_BIT_TIMING_REGISTER, DP_TS2, 8, 4) + FIELD(DATA_PHASE_BIT_TIMING_REGISTER, DP_TS1, 0, 5) +REG32(TX_BUFFER_READY_REQUEST_REGISTER, 0x90) + FIELD(TX_BUFFER_READY_REQUEST_REGISTER, RR31, 31, 1) + FIELD(TX_BUFFER_READY_REQUEST_REGISTER, RR30, 30, 1) + FIELD(TX_BUFFER_READY_REQUEST_REGISTER, RR29, 29, 1) + FIELD(TX_BUFFER_READY_REQUEST_REGISTER, RR28, 28, 1) + FIELD(TX_BUFFER_READY_REQUEST_REGISTER, RR27, 27, 1) + FIELD(TX_BUFFER_READY_REQUEST_REGISTER, RR26, 26, 1) + FIELD(TX_BUFFER_READY_REQUEST_REGISTER, RR25, 25, 1) + FIELD(TX_BUFFER_READY_REQUEST_REGISTER, RR24, 24, 1) + FIELD(TX_BUFFER_READY_REQUEST_REGISTER, RR23, 23, 1) + FIELD(TX_BUFFER_READY_REQUEST_REGISTER, RR22, 22, 1) + FIELD(TX_BUFFER_READY_REQUEST_REGISTER, RR21, 21, 1) + FIELD(TX_BUFFER_READY_REQUEST_REGISTER, RR20, 20, 1) + FIELD(TX_BUFFER_READY_REQUEST_REGISTER, RR19, 19, 1) + FIELD(TX_BUFFER_READY_REQUEST_REGISTER, RR18, 18, 1) + FIELD(TX_BUFFER_READY_REQUEST_REGISTER, RR17, 17, 1) + FIELD(TX_BUFFER_READY_REQUEST_REGISTER, RR16, 16, 1) + FIELD(TX_BUFFER_READY_REQUEST_REGISTER, RR15, 15, 1) + FIELD(TX_BUFFER_READY_REQUEST_REGISTER, RR14, 14, 1) + FIELD(TX_BUFFER_READY_REQUEST_REGISTER, RR13, 13, 1) + FIELD(TX_BUFFER_READY_REQUEST_REGISTER, RR12, 12, 1) + FIELD(TX_BUFFER_READY_REQUEST_REGISTER, RR11, 11, 1) + FIELD(TX_BUFFER_READY_REQUEST_REGISTER, RR10, 10, 1) + FIELD(TX_BUFFER_READY_REQUEST_REGISTER, RR9, 9, 1) + FIELD(TX_BUFFER_READY_REQUEST_REGISTER, RR8, 8, 1) + FIELD(TX_BUFFER_READY_REQUEST_REGISTER, RR7, 7, 1) + FIELD(TX_BUFFER_READY_REQUEST_REGISTER, RR6, 6, 1) + FIELD(TX_BUFFER_READY_REQUEST_REGISTER, RR5, 5, 1) + FIELD(TX_BUFFER_READY_REQUEST_REGISTER, RR4, 4, 1) + FIELD(TX_BUFFER_READY_REQUEST_REGISTER, RR3, 3, 1) + FIELD(TX_BUFFER_READY_REQUEST_REGISTER, RR2, 2, 1) + FIELD(TX_BUFFER_READY_REQUEST_REGISTER, RR1, 1, 1) + FIELD(TX_BUFFER_READY_REQUEST_REGISTER, RR0, 0, 1) +REG32(INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, 0x94) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, ERRS31, 31, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, ERRS30, 30, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, ERRS29, 29, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, ERRS28, 28, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, ERRS27, 27, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, ERRS26, 26, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, ERRS25, 25, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, ERRS24, 24, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, ERRS23, 23, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, ERRS22, 22, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, ERRS21, 21, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, ERRS20, 20, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, ERRS19, 19, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, ERRS18, 18, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, ERRS17, 17, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, ERRS16, 16, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, ERRS15, 15, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, ERRS14, 14, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, ERRS13, 13, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, ERRS12, 12, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, ERRS11, 11, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, ERRS10, 10, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, ERRS9, 9, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, ERRS8, 8, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, ERRS7, 7, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, ERRS6, 6, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, ERRS5, 5, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, ERRS4, 4, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, ERRS3, 3, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, ERRS2, 2, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, ERRS1, 1, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, ERRS0, 0, 1) +REG32(TX_BUFFER_CANCEL_REQUEST_REGISTER, 0x98) + FIELD(TX_BUFFER_CANCEL_REQUEST_REGISTER, CR31, 31, 1) + FIELD(TX_BUFFER_CANCEL_REQUEST_REGISTER, CR30, 30, 1) + FIELD(TX_BUFFER_CANCEL_REQUEST_REGISTER, CR29, 29, 1) + FIELD(TX_BUFFER_CANCEL_REQUEST_REGISTER, CR28, 28, 1) + FIELD(TX_BUFFER_CANCEL_REQUEST_REGISTER, CR27, 27, 1) + FIELD(TX_BUFFER_CANCEL_REQUEST_REGISTER, CR26, 26, 1) + FIELD(TX_BUFFER_CANCEL_REQUEST_REGISTER, CR25, 25, 1) + FIELD(TX_BUFFER_CANCEL_REQUEST_REGISTER, CR24, 24, 1) + FIELD(TX_BUFFER_CANCEL_REQUEST_REGISTER, CR23, 23, 1) + FIELD(TX_BUFFER_CANCEL_REQUEST_REGISTER, CR22, 22, 1) + FIELD(TX_BUFFER_CANCEL_REQUEST_REGISTER, CR21, 21, 1) + FIELD(TX_BUFFER_CANCEL_REQUEST_REGISTER, CR20, 20, 1) + FIELD(TX_BUFFER_CANCEL_REQUEST_REGISTER, CR19, 19, 1) + FIELD(TX_BUFFER_CANCEL_REQUEST_REGISTER, CR18, 18, 1) + FIELD(TX_BUFFER_CANCEL_REQUEST_REGISTER, CR17, 17, 1) + FIELD(TX_BUFFER_CANCEL_REQUEST_REGISTER, CR16, 16, 1) + FIELD(TX_BUFFER_CANCEL_REQUEST_REGISTER, CR15, 15, 1) + FIELD(TX_BUFFER_CANCEL_REQUEST_REGISTER, CR14, 14, 1) + FIELD(TX_BUFFER_CANCEL_REQUEST_REGISTER, CR13, 13, 1) + FIELD(TX_BUFFER_CANCEL_REQUEST_REGISTER, CR12, 12, 1) + FIELD(TX_BUFFER_CANCEL_REQUEST_REGISTER, CR11, 11, 1) + FIELD(TX_BUFFER_CANCEL_REQUEST_REGISTER, CR10, 10, 1) + FIELD(TX_BUFFER_CANCEL_REQUEST_REGISTER, CR9, 9, 1) + FIELD(TX_BUFFER_CANCEL_REQUEST_REGISTER, CR8, 8, 1) + FIELD(TX_BUFFER_CANCEL_REQUEST_REGISTER, CR7, 7, 1) + FIELD(TX_BUFFER_CANCEL_REQUEST_REGISTER, CR6, 6, 1) + FIELD(TX_BUFFER_CANCEL_REQUEST_REGISTER, CR5, 5, 1) + FIELD(TX_BUFFER_CANCEL_REQUEST_REGISTER, CR4, 4, 1) + FIELD(TX_BUFFER_CANCEL_REQUEST_REGISTER, CR3, 3, 1) + FIELD(TX_BUFFER_CANCEL_REQUEST_REGISTER, CR2, 2, 1) + FIELD(TX_BUFFER_CANCEL_REQUEST_REGISTER, CR1, 1, 1) + FIELD(TX_BUFFER_CANCEL_REQUEST_REGISTER, CR0, 0, 1) +REG32(INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, 0x9c) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, ECRS31, 31, + 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, ECRS30, 30, + 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, ECRS29, 29, + 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, ECRS28, 28, + 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, ECRS27, 27, + 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, ECRS26, 26, + 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, ECRS25, 25, + 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, ECRS24, 24, + 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, ECRS23, 23, + 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, ECRS22, 22, + 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, ECRS21, 21, + 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, ECRS20, 20, + 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, ECRS19, 19, + 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, ECRS18, 18, + 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, ECRS17, 17, + 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, ECRS16, 16, + 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, ECRS15, 15, + 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, ECRS14, 14, + 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, ECRS13, 13, + 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, ECRS12, 12, + 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, ECRS11, 11, + 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, ECRS10, 10, + 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, ECRS9, 9, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, ECRS8, 8, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, ECRS7, 7, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, ECRS6, 6, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, ECRS5, 5, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, ECRS4, 4, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, ECRS3, 3, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, ECRS2, 2, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, ECRS1, 1, 1) + FIELD(INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, ECRS0, 0, 1) +REG32(TX_EVENT_FIFO_STATUS_REGISTER, 0xa0) + FIELD(TX_EVENT_FIFO_STATUS_REGISTER, TXE_FL, 8, 6) + FIELD(TX_EVENT_FIFO_STATUS_REGISTER, TXE_IRI, 7, 1) + FIELD(TX_EVENT_FIFO_STATUS_REGISTER, TXE_RI, 0, 5) +REG32(TX_EVENT_FIFO_WATERMARK_REGISTER, 0xa4) + FIELD(TX_EVENT_FIFO_WATERMARK_REGISTER, TXE_FWM, 0, 5) +REG32(ACCEPTANCE_FILTER_CONTROL_REGISTER, 0xe0) + FIELD(ACCEPTANCE_FILTER_CONTROL_REGISTER, UAF31, 31, 1) + FIELD(ACCEPTANCE_FILTER_CONTROL_REGISTER, UAF30, 30, 1) + FIELD(ACCEPTANCE_FILTER_CONTROL_REGISTER, UAF29, 29, 1) + FIELD(ACCEPTANCE_FILTER_CONTROL_REGISTER, UAF28, 28, 1) + FIELD(ACCEPTANCE_FILTER_CONTROL_REGISTER, UAF27, 27, 1) + FIELD(ACCEPTANCE_FILTER_CONTROL_REGISTER, UAF26, 26, 1) + FIELD(ACCEPTANCE_FILTER_CONTROL_REGISTER, UAF25, 25, 1) + FIELD(ACCEPTANCE_FILTER_CONTROL_REGISTER, UAF24, 24, 1) + FIELD(ACCEPTANCE_FILTER_CONTROL_REGISTER, UAF23, 23, 1) + FIELD(ACCEPTANCE_FILTER_CONTROL_REGISTER, UAF22, 22, 1) + FIELD(ACCEPTANCE_FILTER_CONTROL_REGISTER, UAF21, 21, 1) + FIELD(ACCEPTANCE_FILTER_CONTROL_REGISTER, UAF20, 20, 1) + FIELD(ACCEPTANCE_FILTER_CONTROL_REGISTER, UAF19, 19, 1) + FIELD(ACCEPTANCE_FILTER_CONTROL_REGISTER, UAF18, 18, 1) + FIELD(ACCEPTANCE_FILTER_CONTROL_REGISTER, UAF17, 17, 1) + FIELD(ACCEPTANCE_FILTER_CONTROL_REGISTER, UAF16, 16, 1) + FIELD(ACCEPTANCE_FILTER_CONTROL_REGISTER, UAF15, 15, 1) + FIELD(ACCEPTANCE_FILTER_CONTROL_REGISTER, UAF14, 14, 1) + FIELD(ACCEPTANCE_FILTER_CONTROL_REGISTER, UAF13, 13, 1) + FIELD(ACCEPTANCE_FILTER_CONTROL_REGISTER, UAF12, 12, 1) + FIELD(ACCEPTANCE_FILTER_CONTROL_REGISTER, UAF11, 11, 1) + FIELD(ACCEPTANCE_FILTER_CONTROL_REGISTER, UAF10, 10, 1) + FIELD(ACCEPTANCE_FILTER_CONTROL_REGISTER, UAF9, 9, 1) + FIELD(ACCEPTANCE_FILTER_CONTROL_REGISTER, UAF8, 8, 1) + FIELD(ACCEPTANCE_FILTER_CONTROL_REGISTER, UAF7, 7, 1) + FIELD(ACCEPTANCE_FILTER_CONTROL_REGISTER, UAF6, 6, 1) + FIELD(ACCEPTANCE_FILTER_CONTROL_REGISTER, UAF5, 5, 1) + FIELD(ACCEPTANCE_FILTER_CONTROL_REGISTER, UAF4, 4, 1) + FIELD(ACCEPTANCE_FILTER_CONTROL_REGISTER, UAF3, 3, 1) + FIELD(ACCEPTANCE_FILTER_CONTROL_REGISTER, UAF2, 2, 1) + FIELD(ACCEPTANCE_FILTER_CONTROL_REGISTER, UAF1, 1, 1) + FIELD(ACCEPTANCE_FILTER_CONTROL_REGISTER, UAF0, 0, 1) +REG32(RX_FIFO_STATUS_REGISTER, 0xe8) + FIELD(RX_FIFO_STATUS_REGISTER, FL_1, 24, 7) + FIELD(RX_FIFO_STATUS_REGISTER, IRI_1, 23, 1) + FIELD(RX_FIFO_STATUS_REGISTER, RI_1, 16, 6) + FIELD(RX_FIFO_STATUS_REGISTER, FL, 8, 7) + FIELD(RX_FIFO_STATUS_REGISTER, IRI, 7, 1) + FIELD(RX_FIFO_STATUS_REGISTER, RI, 0, 6) +REG32(RX_FIFO_WATERMARK_REGISTER, 0xec) + FIELD(RX_FIFO_WATERMARK_REGISTER, RXFP, 16, 5) + FIELD(RX_FIFO_WATERMARK_REGISTER, RXFWM_1, 8, 6) + FIELD(RX_FIFO_WATERMARK_REGISTER, RXFWM, 0, 6) +REG32(TB_ID_REGISTER, 0x100) + FIELD(TB_ID_REGISTER, ID, 21, 11) + FIELD(TB_ID_REGISTER, SRR_RTR_RRS, 20, 1) + FIELD(TB_ID_REGISTER, IDE, 19, 1) + FIELD(TB_ID_REGISTER, ID_EXT, 1, 18) + FIELD(TB_ID_REGISTER, RTR_RRS, 0, 1) +REG32(TB0_DLC_REGISTER, 0x104) + FIELD(TB0_DLC_REGISTER, DLC, 28, 4) + FIELD(TB0_DLC_REGISTER, FDF, 27, 1) + FIELD(TB0_DLC_REGISTER, BRS, 26, 1) + FIELD(TB0_DLC_REGISTER, RSVD2, 25, 1) + FIELD(TB0_DLC_REGISTER, EFC, 24, 1) + FIELD(TB0_DLC_REGISTER, MM, 16, 8) + FIELD(TB0_DLC_REGISTER, RSVD1, 0, 16) +REG32(TB_DW0_REGISTER, 0x108) + FIELD(TB_DW0_REGISTER, DATA_BYTES0, 24, 8) + FIELD(TB_DW0_REGISTER, DATA_BYTES1, 16, 8) + FIELD(TB_DW0_REGISTER, DATA_BYTES2, 8, 8) + FIELD(TB_DW0_REGISTER, DATA_BYTES3, 0, 8) +REG32(TB_DW1_REGISTER, 0x10c) + FIELD(TB_DW1_REGISTER, DATA_BYTES4, 24, 8) + FIELD(TB_DW1_REGISTER, DATA_BYTES5, 16, 8) + FIELD(TB_DW1_REGISTER, DATA_BYTES6, 8, 8) + FIELD(TB_DW1_REGISTER, DATA_BYTES7, 0, 8) +REG32(TB_DW2_REGISTER, 0x110) + FIELD(TB_DW2_REGISTER, DATA_BYTES8, 24, 8) + FIELD(TB_DW2_REGISTER, DATA_BYTES9, 16, 8) + FIELD(TB_DW2_REGISTER, DATA_BYTES10, 8, 8) + FIELD(TB_DW2_REGISTER, DATA_BYTES11, 0, 8) +REG32(TB_DW3_REGISTER, 0x114) + FIELD(TB_DW3_REGISTER, DATA_BYTES12, 24, 8) + FIELD(TB_DW3_REGISTER, DATA_BYTES13, 16, 8) + FIELD(TB_DW3_REGISTER, DATA_BYTES14, 8, 8) + FIELD(TB_DW3_REGISTER, DATA_BYTES15, 0, 8) +REG32(TB_DW4_REGISTER, 0x118) + FIELD(TB_DW4_REGISTER, DATA_BYTES16, 24, 8) + FIELD(TB_DW4_REGISTER, DATA_BYTES17, 16, 8) + FIELD(TB_DW4_REGISTER, DATA_BYTES18, 8, 8) + FIELD(TB_DW4_REGISTER, DATA_BYTES19, 0, 8) +REG32(TB_DW5_REGISTER, 0x11c) + FIELD(TB_DW5_REGISTER, DATA_BYTES20, 24, 8) + FIELD(TB_DW5_REGISTER, DATA_BYTES21, 16, 8) + FIELD(TB_DW5_REGISTER, DATA_BYTES22, 8, 8) + FIELD(TB_DW5_REGISTER, DATA_BYTES23, 0, 8) +REG32(TB_DW6_REGISTER, 0x120) + FIELD(TB_DW6_REGISTER, DATA_BYTES24, 24, 8) + FIELD(TB_DW6_REGISTER, DATA_BYTES25, 16, 8) + FIELD(TB_DW6_REGISTER, DATA_BYTES26, 8, 8) + FIELD(TB_DW6_REGISTER, DATA_BYTES27, 0, 8) +REG32(TB_DW7_REGISTER, 0x124) + FIELD(TB_DW7_REGISTER, DATA_BYTES28, 24, 8) + FIELD(TB_DW7_REGISTER, DATA_BYTES29, 16, 8) + FIELD(TB_DW7_REGISTER, DATA_BYTES30, 8, 8) + FIELD(TB_DW7_REGISTER, DATA_BYTES31, 0, 8) +REG32(TB_DW8_REGISTER, 0x128) + FIELD(TB_DW8_REGISTER, DATA_BYTES32, 24, 8) + FIELD(TB_DW8_REGISTER, DATA_BYTES33, 16, 8) + FIELD(TB_DW8_REGISTER, DATA_BYTES34, 8, 8) + FIELD(TB_DW8_REGISTER, DATA_BYTES35, 0, 8) +REG32(TB_DW9_REGISTER, 0x12c) + FIELD(TB_DW9_REGISTER, DATA_BYTES36, 24, 8) + FIELD(TB_DW9_REGISTER, DATA_BYTES37, 16, 8) + FIELD(TB_DW9_REGISTER, DATA_BYTES38, 8, 8) + FIELD(TB_DW9_REGISTER, DATA_BYTES39, 0, 8) +REG32(TB_DW10_REGISTER, 0x130) + FIELD(TB_DW10_REGISTER, DATA_BYTES40, 24, 8) + FIELD(TB_DW10_REGISTER, DATA_BYTES41, 16, 8) + FIELD(TB_DW10_REGISTER, DATA_BYTES42, 8, 8) + FIELD(TB_DW10_REGISTER, DATA_BYTES43, 0, 8) +REG32(TB_DW11_REGISTER, 0x134) + FIELD(TB_DW11_REGISTER, DATA_BYTES44, 24, 8) + FIELD(TB_DW11_REGISTER, DATA_BYTES45, 16, 8) + FIELD(TB_DW11_REGISTER, DATA_BYTES46, 8, 8) + FIELD(TB_DW11_REGISTER, DATA_BYTES47, 0, 8) +REG32(TB_DW12_REGISTER, 0x138) + FIELD(TB_DW12_REGISTER, DATA_BYTES48, 24, 8) + FIELD(TB_DW12_REGISTER, DATA_BYTES49, 16, 8) + FIELD(TB_DW12_REGISTER, DATA_BYTES50, 8, 8) + FIELD(TB_DW12_REGISTER, DATA_BYTES51, 0, 8) +REG32(TB_DW13_REGISTER, 0x13c) + FIELD(TB_DW13_REGISTER, DATA_BYTES52, 24, 8) + FIELD(TB_DW13_REGISTER, DATA_BYTES53, 16, 8) + FIELD(TB_DW13_REGISTER, DATA_BYTES54, 8, 8) + FIELD(TB_DW13_REGISTER, DATA_BYTES55, 0, 8) +REG32(TB_DW14_REGISTER, 0x140) + FIELD(TB_DW14_REGISTER, DATA_BYTES56, 24, 8) + FIELD(TB_DW14_REGISTER, DATA_BYTES57, 16, 8) + FIELD(TB_DW14_REGISTER, DATA_BYTES58, 8, 8) + FIELD(TB_DW14_REGISTER, DATA_BYTES59, 0, 8) +REG32(TB_DW15_REGISTER, 0x144) + FIELD(TB_DW15_REGISTER, DATA_BYTES60, 24, 8) + FIELD(TB_DW15_REGISTER, DATA_BYTES61, 16, 8) + FIELD(TB_DW15_REGISTER, DATA_BYTES62, 8, 8) + FIELD(TB_DW15_REGISTER, DATA_BYTES63, 0, 8) +REG32(AFMR_REGISTER, 0xa00) + FIELD(AFMR_REGISTER, AMID, 21, 11) + FIELD(AFMR_REGISTER, AMSRR, 20, 1) + FIELD(AFMR_REGISTER, AMIDE, 19, 1) + FIELD(AFMR_REGISTER, AMID_EXT, 1, 18) + FIELD(AFMR_REGISTER, AMRTR, 0, 1) +REG32(AFIR_REGISTER, 0xa04) + FIELD(AFIR_REGISTER, AIID, 21, 11) + FIELD(AFIR_REGISTER, AISRR, 20, 1) + FIELD(AFIR_REGISTER, AIIDE, 19, 1) + FIELD(AFIR_REGISTER, AIID_EXT, 1, 18) + FIELD(AFIR_REGISTER, AIRTR, 0, 1) +REG32(TXE_FIFO_TB_ID_REGISTER, 0x2000) + FIELD(TXE_FIFO_TB_ID_REGISTER, ID, 21, 11) + FIELD(TXE_FIFO_TB_ID_REGISTER, SRR_RTR_RRS, 20, 1) + FIELD(TXE_FIFO_TB_ID_REGISTER, IDE, 19, 1) + FIELD(TXE_FIFO_TB_ID_REGISTER, ID_EXT, 1, 18) + FIELD(TXE_FIFO_TB_ID_REGISTER, RTR_RRS, 0, 1) +REG32(TXE_FIFO_TB_DLC_REGISTER, 0x2004) + FIELD(TXE_FIFO_TB_DLC_REGISTER, DLC, 28, 4) + FIELD(TXE_FIFO_TB_DLC_REGISTER, FDF, 27, 1) + FIELD(TXE_FIFO_TB_DLC_REGISTER, BRS, 26, 1) + FIELD(TXE_FIFO_TB_DLC_REGISTER, ET, 24, 2) + FIELD(TXE_FIFO_TB_DLC_REGISTER, MM, 16, 8) + FIELD(TXE_FIFO_TB_DLC_REGISTER, TIMESTAMP, 0, 16) +REG32(RB_ID_REGISTER, 0x2100) + FIELD(RB_ID_REGISTER, ID, 21, 11) + FIELD(RB_ID_REGISTER, SRR_RTR_RRS, 20, 1) + FIELD(RB_ID_REGISTER, IDE, 19, 1) + FIELD(RB_ID_REGISTER, ID_EXT, 1, 18) + FIELD(RB_ID_REGISTER, RTR_RRS, 0, 1) +REG32(RB_DLC_REGISTER, 0x2104) + FIELD(RB_DLC_REGISTER, DLC, 28, 4) + FIELD(RB_DLC_REGISTER, FDF, 27, 1) + FIELD(RB_DLC_REGISTER, BRS, 26, 1) + FIELD(RB_DLC_REGISTER, ESI, 25, 1) + FIELD(RB_DLC_REGISTER, MATCHED_FILTER_INDEX, 16, 5) + FIELD(RB_DLC_REGISTER, TIMESTAMP, 0, 16) +REG32(RB_DW0_REGISTER, 0x2108) + FIELD(RB_DW0_REGISTER, DATA_BYTES0, 24, 8) + FIELD(RB_DW0_REGISTER, DATA_BYTES1, 16, 8) + FIELD(RB_DW0_REGISTER, DATA_BYTES2, 8, 8) + FIELD(RB_DW0_REGISTER, DATA_BYTES3, 0, 8) +REG32(RB_DW1_REGISTER, 0x210c) + FIELD(RB_DW1_REGISTER, DATA_BYTES4, 24, 8) + FIELD(RB_DW1_REGISTER, DATA_BYTES5, 16, 8) + FIELD(RB_DW1_REGISTER, DATA_BYTES6, 8, 8) + FIELD(RB_DW1_REGISTER, DATA_BYTES7, 0, 8) +REG32(RB_DW2_REGISTER, 0x2110) + FIELD(RB_DW2_REGISTER, DATA_BYTES8, 24, 8) + FIELD(RB_DW2_REGISTER, DATA_BYTES9, 16, 8) + FIELD(RB_DW2_REGISTER, DATA_BYTES10, 8, 8) + FIELD(RB_DW2_REGISTER, DATA_BYTES11, 0, 8) +REG32(RB_DW3_REGISTER, 0x2114) + FIELD(RB_DW3_REGISTER, DATA_BYTES12, 24, 8) + FIELD(RB_DW3_REGISTER, DATA_BYTES13, 16, 8) + FIELD(RB_DW3_REGISTER, DATA_BYTES14, 8, 8) + FIELD(RB_DW3_REGISTER, DATA_BYTES15, 0, 8) +REG32(RB_DW4_REGISTER, 0x2118) + FIELD(RB_DW4_REGISTER, DATA_BYTES16, 24, 8) + FIELD(RB_DW4_REGISTER, DATA_BYTES17, 16, 8) + FIELD(RB_DW4_REGISTER, DATA_BYTES18, 8, 8) + FIELD(RB_DW4_REGISTER, DATA_BYTES19, 0, 8) +REG32(RB_DW5_REGISTER, 0x211c) + FIELD(RB_DW5_REGISTER, DATA_BYTES20, 24, 8) + FIELD(RB_DW5_REGISTER, DATA_BYTES21, 16, 8) + FIELD(RB_DW5_REGISTER, DATA_BYTES22, 8, 8) + FIELD(RB_DW5_REGISTER, DATA_BYTES23, 0, 8) +REG32(RB_DW6_REGISTER, 0x2120) + FIELD(RB_DW6_REGISTER, DATA_BYTES24, 24, 8) + FIELD(RB_DW6_REGISTER, DATA_BYTES25, 16, 8) + FIELD(RB_DW6_REGISTER, DATA_BYTES26, 8, 8) + FIELD(RB_DW6_REGISTER, DATA_BYTES27, 0, 8) +REG32(RB_DW7_REGISTER, 0x2124) + FIELD(RB_DW7_REGISTER, DATA_BYTES28, 24, 8) + FIELD(RB_DW7_REGISTER, DATA_BYTES29, 16, 8) + FIELD(RB_DW7_REGISTER, DATA_BYTES30, 8, 8) + FIELD(RB_DW7_REGISTER, DATA_BYTES31, 0, 8) +REG32(RB_DW8_REGISTER, 0x2128) + FIELD(RB_DW8_REGISTER, DATA_BYTES32, 24, 8) + FIELD(RB_DW8_REGISTER, DATA_BYTES33, 16, 8) + FIELD(RB_DW8_REGISTER, DATA_BYTES34, 8, 8) + FIELD(RB_DW8_REGISTER, DATA_BYTES35, 0, 8) +REG32(RB_DW9_REGISTER, 0x212c) + FIELD(RB_DW9_REGISTER, DATA_BYTES36, 24, 8) + FIELD(RB_DW9_REGISTER, DATA_BYTES37, 16, 8) + FIELD(RB_DW9_REGISTER, DATA_BYTES38, 8, 8) + FIELD(RB_DW9_REGISTER, DATA_BYTES39, 0, 8) +REG32(RB_DW10_REGISTER, 0x2130) + FIELD(RB_DW10_REGISTER, DATA_BYTES40, 24, 8) + FIELD(RB_DW10_REGISTER, DATA_BYTES41, 16, 8) + FIELD(RB_DW10_REGISTER, DATA_BYTES42, 8, 8) + FIELD(RB_DW10_REGISTER, DATA_BYTES43, 0, 8) +REG32(RB_DW11_REGISTER, 0x2134) + FIELD(RB_DW11_REGISTER, DATA_BYTES44, 24, 8) + FIELD(RB_DW11_REGISTER, DATA_BYTES45, 16, 8) + FIELD(RB_DW11_REGISTER, DATA_BYTES46, 8, 8) + FIELD(RB_DW11_REGISTER, DATA_BYTES47, 0, 8) +REG32(RB_DW12_REGISTER, 0x2138) + FIELD(RB_DW12_REGISTER, DATA_BYTES48, 24, 8) + FIELD(RB_DW12_REGISTER, DATA_BYTES49, 16, 8) + FIELD(RB_DW12_REGISTER, DATA_BYTES50, 8, 8) + FIELD(RB_DW12_REGISTER, DATA_BYTES51, 0, 8) +REG32(RB_DW13_REGISTER, 0x213c) + FIELD(RB_DW13_REGISTER, DATA_BYTES52, 24, 8) + FIELD(RB_DW13_REGISTER, DATA_BYTES53, 16, 8) + FIELD(RB_DW13_REGISTER, DATA_BYTES54, 8, 8) + FIELD(RB_DW13_REGISTER, DATA_BYTES55, 0, 8) +REG32(RB_DW14_REGISTER, 0x2140) + FIELD(RB_DW14_REGISTER, DATA_BYTES56, 24, 8) + FIELD(RB_DW14_REGISTER, DATA_BYTES57, 16, 8) + FIELD(RB_DW14_REGISTER, DATA_BYTES58, 8, 8) + FIELD(RB_DW14_REGISTER, DATA_BYTES59, 0, 8) +REG32(RB_DW15_REGISTER, 0x2144) + FIELD(RB_DW15_REGISTER, DATA_BYTES60, 24, 8) + FIELD(RB_DW15_REGISTER, DATA_BYTES61, 16, 8) + FIELD(RB_DW15_REGISTER, DATA_BYTES62, 8, 8) + FIELD(RB_DW15_REGISTER, DATA_BYTES63, 0, 8) +REG32(RB_ID_REGISTER_1, 0x4100) + FIELD(RB_ID_REGISTER_1, ID, 21, 11) + FIELD(RB_ID_REGISTER_1, SRR_RTR_RRS, 20, 1) + FIELD(RB_ID_REGISTER_1, IDE, 19, 1) + FIELD(RB_ID_REGISTER_1, ID_EXT, 1, 18) + FIELD(RB_ID_REGISTER_1, RTR_RRS, 0, 1) +REG32(RB_DLC_REGISTER_1, 0x4104) + FIELD(RB_DLC_REGISTER_1, DLC, 28, 4) + FIELD(RB_DLC_REGISTER_1, FDF, 27, 1) + FIELD(RB_DLC_REGISTER_1, BRS, 26, 1) + FIELD(RB_DLC_REGISTER_1, ESI, 25, 1) + FIELD(RB_DLC_REGISTER_1, MATCHED_FILTER_INDEX, 16, 5) + FIELD(RB_DLC_REGISTER_1, TIMESTAMP, 0, 16) +REG32(RB0_DW0_REGISTER_1, 0x4108) + FIELD(RB0_DW0_REGISTER_1, DATA_BYTES0, 24, 8) + FIELD(RB0_DW0_REGISTER_1, DATA_BYTES1, 16, 8) + FIELD(RB0_DW0_REGISTER_1, DATA_BYTES2, 8, 8) + FIELD(RB0_DW0_REGISTER_1, DATA_BYTES3, 0, 8) +REG32(RB_DW1_REGISTER_1, 0x410c) + FIELD(RB_DW1_REGISTER_1, DATA_BYTES4, 24, 8) + FIELD(RB_DW1_REGISTER_1, DATA_BYTES5, 16, 8) + FIELD(RB_DW1_REGISTER_1, DATA_BYTES6, 8, 8) + FIELD(RB_DW1_REGISTER_1, DATA_BYTES7, 0, 8) +REG32(RB_DW2_REGISTER_1, 0x4110) + FIELD(RB_DW2_REGISTER_1, DATA_BYTES8, 24, 8) + FIELD(RB_DW2_REGISTER_1, DATA_BYTES9, 16, 8) + FIELD(RB_DW2_REGISTER_1, DATA_BYTES10, 8, 8) + FIELD(RB_DW2_REGISTER_1, DATA_BYTES11, 0, 8) +REG32(RB_DW3_REGISTER_1, 0x4114) + FIELD(RB_DW3_REGISTER_1, DATA_BYTES12, 24, 8) + FIELD(RB_DW3_REGISTER_1, DATA_BYTES13, 16, 8) + FIELD(RB_DW3_REGISTER_1, DATA_BYTES14, 8, 8) + FIELD(RB_DW3_REGISTER_1, DATA_BYTES15, 0, 8) +REG32(RB_DW4_REGISTER_1, 0x4118) + FIELD(RB_DW4_REGISTER_1, DATA_BYTES16, 24, 8) + FIELD(RB_DW4_REGISTER_1, DATA_BYTES17, 16, 8) + FIELD(RB_DW4_REGISTER_1, DATA_BYTES18, 8, 8) + FIELD(RB_DW4_REGISTER_1, DATA_BYTES19, 0, 8) +REG32(RB_DW5_REGISTER_1, 0x411c) + FIELD(RB_DW5_REGISTER_1, DATA_BYTES20, 24, 8) + FIELD(RB_DW5_REGISTER_1, DATA_BYTES21, 16, 8) + FIELD(RB_DW5_REGISTER_1, DATA_BYTES22, 8, 8) + FIELD(RB_DW5_REGISTER_1, DATA_BYTES23, 0, 8) +REG32(RB_DW6_REGISTER_1, 0x4120) + FIELD(RB_DW6_REGISTER_1, DATA_BYTES24, 24, 8) + FIELD(RB_DW6_REGISTER_1, DATA_BYTES25, 16, 8) + FIELD(RB_DW6_REGISTER_1, DATA_BYTES26, 8, 8) + FIELD(RB_DW6_REGISTER_1, DATA_BYTES27, 0, 8) +REG32(RB_DW7_REGISTER_1, 0x4124) + FIELD(RB_DW7_REGISTER_1, DATA_BYTES28, 24, 8) + FIELD(RB_DW7_REGISTER_1, DATA_BYTES29, 16, 8) + FIELD(RB_DW7_REGISTER_1, DATA_BYTES30, 8, 8) + FIELD(RB_DW7_REGISTER_1, DATA_BYTES31, 0, 8) +REG32(RB_DW8_REGISTER_1, 0x4128) + FIELD(RB_DW8_REGISTER_1, DATA_BYTES32, 24, 8) + FIELD(RB_DW8_REGISTER_1, DATA_BYTES33, 16, 8) + FIELD(RB_DW8_REGISTER_1, DATA_BYTES34, 8, 8) + FIELD(RB_DW8_REGISTER_1, DATA_BYTES35, 0, 8) +REG32(RB_DW9_REGISTER_1, 0x412c) + FIELD(RB_DW9_REGISTER_1, DATA_BYTES36, 24, 8) + FIELD(RB_DW9_REGISTER_1, DATA_BYTES37, 16, 8) + FIELD(RB_DW9_REGISTER_1, DATA_BYTES38, 8, 8) + FIELD(RB_DW9_REGISTER_1, DATA_BYTES39, 0, 8) +REG32(RB_DW10_REGISTER_1, 0x4130) + FIELD(RB_DW10_REGISTER_1, DATA_BYTES40, 24, 8) + FIELD(RB_DW10_REGISTER_1, DATA_BYTES41, 16, 8) + FIELD(RB_DW10_REGISTER_1, DATA_BYTES42, 8, 8) + FIELD(RB_DW10_REGISTER_1, DATA_BYTES43, 0, 8) +REG32(RB_DW11_REGISTER_1, 0x4134) + FIELD(RB_DW11_REGISTER_1, DATA_BYTES44, 24, 8) + FIELD(RB_DW11_REGISTER_1, DATA_BYTES45, 16, 8) + FIELD(RB_DW11_REGISTER_1, DATA_BYTES46, 8, 8) + FIELD(RB_DW11_REGISTER_1, DATA_BYTES47, 0, 8) +REG32(RB_DW12_REGISTER_1, 0x4138) + FIELD(RB_DW12_REGISTER_1, DATA_BYTES48, 24, 8) + FIELD(RB_DW12_REGISTER_1, DATA_BYTES49, 16, 8) + FIELD(RB_DW12_REGISTER_1, DATA_BYTES50, 8, 8) + FIELD(RB_DW12_REGISTER_1, DATA_BYTES51, 0, 8) +REG32(RB_DW13_REGISTER_1, 0x413c) + FIELD(RB_DW13_REGISTER_1, DATA_BYTES52, 24, 8) + FIELD(RB_DW13_REGISTER_1, DATA_BYTES53, 16, 8) + FIELD(RB_DW13_REGISTER_1, DATA_BYTES54, 8, 8) + FIELD(RB_DW13_REGISTER_1, DATA_BYTES55, 0, 8) +REG32(RB_DW14_REGISTER_1, 0x4140) + FIELD(RB_DW14_REGISTER_1, DATA_BYTES56, 24, 8) + FIELD(RB_DW14_REGISTER_1, DATA_BYTES57, 16, 8) + FIELD(RB_DW14_REGISTER_1, DATA_BYTES58, 8, 8) + FIELD(RB_DW14_REGISTER_1, DATA_BYTES59, 0, 8) +REG32(RB_DW15_REGISTER_1, 0x4144) + FIELD(RB_DW15_REGISTER_1, DATA_BYTES60, 24, 8) + FIELD(RB_DW15_REGISTER_1, DATA_BYTES61, 16, 8) + FIELD(RB_DW15_REGISTER_1, DATA_BYTES62, 8, 8) + FIELD(RB_DW15_REGISTER_1, DATA_BYTES63, 0, 8) + +static uint8_t canfd_dlc_array[8] = {8, 12, 16, 20, 24, 32, 48, 64}; + +static void canfd_update_irq(XlnxVersalCANFDState *s) +{ + unsigned int irq = s->regs[R_INTERRUPT_STATUS_REGISTER] & + s->regs[R_INTERRUPT_ENABLE_REGISTER]; + g_autofree char *path = object_get_canonical_path(OBJECT(s)); + + /* RX watermark interrupts. */ + if (ARRAY_FIELD_EX32(s->regs, RX_FIFO_STATUS_REGISTER, FL) > + ARRAY_FIELD_EX32(s->regs, RX_FIFO_WATERMARK_REGISTER, RXFWM)) { + ARRAY_FIELD_DP32(s->regs, INTERRUPT_STATUS_REGISTER, RXFWMFLL, 1); + } + + if (ARRAY_FIELD_EX32(s->regs, RX_FIFO_STATUS_REGISTER, FL_1) > + ARRAY_FIELD_EX32(s->regs, RX_FIFO_WATERMARK_REGISTER, RXFWM_1)) { + ARRAY_FIELD_DP32(s->regs, INTERRUPT_STATUS_REGISTER, RXFWMFLL_1, 1); + } + + /* TX watermark interrupt. */ + if (ARRAY_FIELD_EX32(s->regs, TX_EVENT_FIFO_STATUS_REGISTER, TXE_FL) > + ARRAY_FIELD_EX32(s->regs, TX_EVENT_FIFO_WATERMARK_REGISTER, TXE_FWM)) { + ARRAY_FIELD_DP32(s->regs, INTERRUPT_STATUS_REGISTER, TXEWMFLL, 1); + } + + trace_xlnx_canfd_update_irq(path, s->regs[R_INTERRUPT_STATUS_REGISTER], + s->regs[R_INTERRUPT_ENABLE_REGISTER], irq); + + qemu_set_irq(s->irq_canfd_int, irq); +} + +static void canfd_ier_post_write(RegisterInfo *reg, uint64_t val64) +{ + XlnxVersalCANFDState *s = XILINX_CANFD(reg->opaque); + + canfd_update_irq(s); +} + +static uint64_t canfd_icr_pre_write(RegisterInfo *reg, uint64_t val64) +{ + XlnxVersalCANFDState *s = XILINX_CANFD(reg->opaque); + uint32_t val = val64; + + s->regs[R_INTERRUPT_STATUS_REGISTER] &= ~val; + + /* + * RXBOFLW_BI field is automatically cleared to default if RXBOFLW bit is + * cleared in ISR. + */ + if (ARRAY_FIELD_EX32(s->regs, INTERRUPT_STATUS_REGISTER, RXFWMFLL_1)) { + ARRAY_FIELD_DP32(s->regs, INTERRUPT_STATUS_REGISTER, RXBOFLW_BI, 0); + } + + canfd_update_irq(s); + + return 0; +} + +static void canfd_config_reset(XlnxVersalCANFDState *s) +{ + + unsigned int i; + + /* Reset all the configuration registers. */ + for (i = 0; i < R_RX_FIFO_WATERMARK_REGISTER; ++i) { + register_reset(&s->reg_info[i]); + } + + canfd_update_irq(s); +} + +static void canfd_config_mode(XlnxVersalCANFDState *s) +{ + register_reset(&s->reg_info[R_ERROR_COUNTER_REGISTER]); + register_reset(&s->reg_info[R_ERROR_STATUS_REGISTER]); + register_reset(&s->reg_info[R_STATUS_REGISTER]); + + /* Put XlnxVersalCANFDState in configuration mode. */ + ARRAY_FIELD_DP32(s->regs, STATUS_REGISTER, CONFIG, 1); + ARRAY_FIELD_DP32(s->regs, INTERRUPT_STATUS_REGISTER, WKUP, 0); + ARRAY_FIELD_DP32(s->regs, INTERRUPT_STATUS_REGISTER, SLP, 0); + ARRAY_FIELD_DP32(s->regs, INTERRUPT_STATUS_REGISTER, BSOFF, 0); + ARRAY_FIELD_DP32(s->regs, INTERRUPT_STATUS_REGISTER, ERROR_BIT, 0); + ARRAY_FIELD_DP32(s->regs, INTERRUPT_STATUS_REGISTER, RXFOFLW, 0); + ARRAY_FIELD_DP32(s->regs, INTERRUPT_STATUS_REGISTER, RXFOFLW_1, 0); + ARRAY_FIELD_DP32(s->regs, INTERRUPT_STATUS_REGISTER, RXOK, 0); + ARRAY_FIELD_DP32(s->regs, INTERRUPT_STATUS_REGISTER, TXOK, 0); + ARRAY_FIELD_DP32(s->regs, INTERRUPT_STATUS_REGISTER, ARBLST, 0); + + /* Clear the time stamp. */ + ptimer_transaction_begin(s->canfd_timer); + ptimer_set_count(s->canfd_timer, 0); + ptimer_transaction_commit(s->canfd_timer); + + canfd_update_irq(s); +} + +static void update_status_register_mode_bits(XlnxVersalCANFDState *s) +{ + bool sleep_status = ARRAY_FIELD_EX32(s->regs, STATUS_REGISTER, SLEEP); + bool sleep_mode = ARRAY_FIELD_EX32(s->regs, MODE_SELECT_REGISTER, SLEEP); + /* Wake up interrupt bit. */ + bool wakeup_irq_val = !sleep_mode && sleep_status; + /* Sleep interrupt bit. */ + bool sleep_irq_val = sleep_mode && !sleep_status; + + /* Clear previous core mode status bits. */ + ARRAY_FIELD_DP32(s->regs, STATUS_REGISTER, LBACK, 0); + ARRAY_FIELD_DP32(s->regs, STATUS_REGISTER, SLEEP, 0); + ARRAY_FIELD_DP32(s->regs, STATUS_REGISTER, SNOOP, 0); + ARRAY_FIELD_DP32(s->regs, STATUS_REGISTER, NORMAL, 0); + + /* set current mode bit and generate irqs accordingly. */ + if (ARRAY_FIELD_EX32(s->regs, MODE_SELECT_REGISTER, LBACK)) { + ARRAY_FIELD_DP32(s->regs, STATUS_REGISTER, LBACK, 1); + } else if (ARRAY_FIELD_EX32(s->regs, MODE_SELECT_REGISTER, SLEEP)) { + ARRAY_FIELD_DP32(s->regs, STATUS_REGISTER, SLEEP, 1); + ARRAY_FIELD_DP32(s->regs, INTERRUPT_STATUS_REGISTER, SLP, + sleep_irq_val); + } else if (ARRAY_FIELD_EX32(s->regs, MODE_SELECT_REGISTER, SNOOP)) { + ARRAY_FIELD_DP32(s->regs, STATUS_REGISTER, SNOOP, 1); + } else { + /* If all bits are zero, XlnxVersalCANFDState is set in normal mode. */ + ARRAY_FIELD_DP32(s->regs, STATUS_REGISTER, NORMAL, 1); + /* Set wakeup interrupt bit. */ + ARRAY_FIELD_DP32(s->regs, INTERRUPT_STATUS_REGISTER, WKUP, + wakeup_irq_val); + } + + /* Put the CANFD in error active state. */ + ARRAY_FIELD_DP32(s->regs, STATUS_REGISTER, ESTAT, 1); + + canfd_update_irq(s); +} + +static uint64_t canfd_msr_pre_write(RegisterInfo *reg, uint64_t val64) +{ + XlnxVersalCANFDState *s = XILINX_CANFD(reg->opaque); + uint32_t val = val64; + uint8_t multi_mode = 0; + + /* + * Multiple mode set check. This is done to make sure user doesn't set + * multiple modes. + */ + multi_mode = FIELD_EX32(val, MODE_SELECT_REGISTER, LBACK) + + FIELD_EX32(val, MODE_SELECT_REGISTER, SLEEP) + + FIELD_EX32(val, MODE_SELECT_REGISTER, SNOOP); + + if (multi_mode > 1) { + qemu_log_mask(LOG_GUEST_ERROR, "Attempting to configure several modes" + " simultaneously. One mode will be selected according to" + " their priority: LBACK > SLEEP > SNOOP.\n"); + } + + if (ARRAY_FIELD_EX32(s->regs, SOFTWARE_RESET_REGISTER, CEN) == 0) { + /* In configuration mode, any mode can be selected. */ + s->regs[R_MODE_SELECT_REGISTER] = val; + } else { + bool sleep_mode_bit = FIELD_EX32(val, MODE_SELECT_REGISTER, SLEEP); + + ARRAY_FIELD_DP32(s->regs, MODE_SELECT_REGISTER, SLEEP, sleep_mode_bit); + + if (FIELD_EX32(val, MODE_SELECT_REGISTER, LBACK)) { + qemu_log_mask(LOG_GUEST_ERROR, "Attempting to set LBACK mode" + " without setting CEN bit as 0\n"); + } else if (FIELD_EX32(val, MODE_SELECT_REGISTER, SNOOP)) { + qemu_log_mask(LOG_GUEST_ERROR, "Attempting to set SNOOP mode" + " without setting CEN bit as 0\n"); + } + + update_status_register_mode_bits(s); + } + + return s->regs[R_MODE_SELECT_REGISTER]; +} + +static void canfd_exit_sleep_mode(XlnxVersalCANFDState *s) +{ + ARRAY_FIELD_DP32(s->regs, MODE_SELECT_REGISTER, SLEEP, 0); + update_status_register_mode_bits(s); +} + +static void regs2frame(XlnxVersalCANFDState *s, qemu_can_frame *frame, + uint32_t reg_num) +{ + uint32_t i = 0; + uint32_t j = 0; + uint32_t val = 0; + uint32_t dlc_reg_val = 0; + uint32_t dlc_value = 0; + + /* Check that reg_num should be within TX register space. */ + assert(reg_num <= R_TB_ID_REGISTER + (NUM_REGS_PER_MSG_SPACE * + s->cfg.tx_fifo)); + + dlc_reg_val = s->regs[reg_num + 1]; + dlc_value = FIELD_EX32(dlc_reg_val, TB0_DLC_REGISTER, DLC); + + frame->can_id = s->regs[reg_num]; + + if (FIELD_EX32(dlc_reg_val, TB0_DLC_REGISTER, FDF)) { + /* + * CANFD frame. + * Converting dlc(0 to 15) 4 Byte data to plain length(i.e. 0 to 64) + * 1 Byte data. This is done to make it work with SocketCAN. + * On actual CANFD frame, this value can't be more than 0xF. + * Conversion table for DLC to plain length: + * + * DLC Plain Length + * 0 - 8 0 - 8 + * 9 9 - 12 + * 10 13 - 16 + * 11 17 - 20 + * 12 21 - 24 + * 13 25 - 32 + * 14 33 - 48 + * 15 49 - 64 + */ + + frame->flags = QEMU_CAN_FRMF_TYPE_FD; + + if (dlc_value < 8) { + frame->can_dlc = dlc_value; + } else { + assert((dlc_value - 8) < ARRAY_SIZE(canfd_dlc_array)); + frame->can_dlc = canfd_dlc_array[dlc_value - 8]; + } + } else { + /* + * FD Format bit not set that means it is a CAN Frame. + * Conversion table for classic CAN: + * + * DLC Plain Length + * 0 - 7 0 - 7 + * 8 - 15 8 + */ + + if (dlc_value > 8) { + frame->can_dlc = 8; + qemu_log_mask(LOG_GUEST_ERROR, "Maximum DLC value for Classic CAN" + " frame is 8. Only 8 byte data will be sent.\n"); + } else { + frame->can_dlc = dlc_value; + } + } + + for (j = 0; j < frame->can_dlc; j++) { + val = 8 * i; + + frame->data[j] = extract32(s->regs[reg_num + 2 + (j / 4)], val, 8); + i++; + + if (i % 4 == 0) { + i = 0; + } + } +} + +static void process_cancellation_requests(XlnxVersalCANFDState *s) +{ + uint32_t clear_mask = s->regs[R_TX_BUFFER_READY_REQUEST_REGISTER] & + s->regs[R_TX_BUFFER_CANCEL_REQUEST_REGISTER]; + + s->regs[R_TX_BUFFER_READY_REQUEST_REGISTER] &= ~clear_mask; + s->regs[R_TX_BUFFER_CANCEL_REQUEST_REGISTER] &= ~clear_mask; + + canfd_update_irq(s); +} + +static void store_rx_sequential(XlnxVersalCANFDState *s, + const qemu_can_frame *frame, + uint32_t fill_level, uint32_t read_index, + uint32_t store_location, uint8_t rx_fifo, + bool rx_fifo_id, uint8_t filter_index) +{ + int i; + bool is_canfd_frame; + uint8_t dlc = frame->can_dlc; + uint8_t rx_reg_num = 0; + uint32_t dlc_reg_val = 0; + uint32_t data_reg_val = 0; + + /* Getting RX0/1 fill level */ + if ((fill_level) > rx_fifo - 1) { + g_autofree char *path = object_get_canonical_path(OBJECT(s)); + + qemu_log_mask(LOG_GUEST_ERROR, "%s: RX%d Buffer is full. Discarding the" + " message\n", path, rx_fifo_id); + + /* Set the corresponding RF buffer overflow interrupt. */ + if (rx_fifo_id == 0) { + ARRAY_FIELD_DP32(s->regs, INTERRUPT_STATUS_REGISTER, RXFOFLW, 1); + } else { + ARRAY_FIELD_DP32(s->regs, INTERRUPT_STATUS_REGISTER, RXFOFLW_1, 1); + } + } else { + uint16_t rx_timestamp = CANFD_TIMER_MAX - + ptimer_get_count(s->canfd_timer); + + if (rx_timestamp == 0xFFFF) { + ARRAY_FIELD_DP32(s->regs, INTERRUPT_STATUS_REGISTER, TSCNT_OFLW, 1); + } else { + ARRAY_FIELD_DP32(s->regs, TIMESTAMP_REGISTER, TIMESTAMP_CNT, + rx_timestamp); + } + + if (rx_fifo_id == 0) { + ARRAY_FIELD_DP32(s->regs, RX_FIFO_STATUS_REGISTER, FL, + fill_level + 1); + assert(store_location <= + R_RB_ID_REGISTER + (s->cfg.rx0_fifo * + NUM_REGS_PER_MSG_SPACE)); + } else { + ARRAY_FIELD_DP32(s->regs, RX_FIFO_STATUS_REGISTER, FL_1, + fill_level + 1); + assert(store_location <= + R_RB_ID_REGISTER_1 + (s->cfg.rx1_fifo * + NUM_REGS_PER_MSG_SPACE)); + } + + s->regs[store_location] = frame->can_id; + + dlc = frame->can_dlc; + + if (frame->flags == QEMU_CAN_FRMF_TYPE_FD) { + is_canfd_frame = true; + + /* Store dlc value in Xilinx specific format. */ + for (i = 0; i < ARRAY_SIZE(canfd_dlc_array); i++) { + if (canfd_dlc_array[i] == frame->can_dlc) { + dlc_reg_val = FIELD_DP32(0, RB_DLC_REGISTER, DLC, 8 + i); + } + } + } else { + is_canfd_frame = false; + + if (frame->can_dlc > 8) { + dlc = 8; + } + + dlc_reg_val = FIELD_DP32(0, RB_DLC_REGISTER, DLC, dlc); + } + + dlc_reg_val |= FIELD_DP32(0, RB_DLC_REGISTER, FDF, is_canfd_frame); + dlc_reg_val |= FIELD_DP32(0, RB_DLC_REGISTER, TIMESTAMP, rx_timestamp); + dlc_reg_val |= FIELD_DP32(0, RB_DLC_REGISTER, MATCHED_FILTER_INDEX, + filter_index); + s->regs[store_location + 1] = dlc_reg_val; + + for (i = 0; i < dlc; i++) { + /* Register size is 4 byte but frame->data each is 1 byte. */ + switch (i % 4) { + case 0: + rx_reg_num = i / 4; + + data_reg_val = FIELD_DP32(0, RB_DW0_REGISTER, DATA_BYTES3, + frame->data[i]); + break; + case 1: + data_reg_val |= FIELD_DP32(0, RB_DW0_REGISTER, DATA_BYTES2, + frame->data[i]); + break; + case 2: + data_reg_val |= FIELD_DP32(0, RB_DW0_REGISTER, DATA_BYTES1, + frame->data[i]); + break; + case 3: + data_reg_val |= FIELD_DP32(0, RB_DW0_REGISTER, DATA_BYTES0, + frame->data[i]); + /* + * Last Bytes data which means we have all 4 bytes ready to + * store in one rx regs. + */ + s->regs[store_location + rx_reg_num + 2] = data_reg_val; + break; + } + } + + if (i % 4) { + /* + * In case DLC is not multiplier of 4, data is not saved to RX FIFO + * in above switch case. Store the remaining bytes here. + */ + s->regs[store_location + rx_reg_num + 2] = data_reg_val; + } + + /* set the interrupt as RXOK. */ + ARRAY_FIELD_DP32(s->regs, INTERRUPT_STATUS_REGISTER, RXOK, 1); + } +} + +static void update_rx_sequential(XlnxVersalCANFDState *s, + const qemu_can_frame *frame) +{ + bool filter_pass = false; + uint8_t filter_index = 0; + int i; + int filter_partition = ARRAY_FIELD_EX32(s->regs, + RX_FIFO_WATERMARK_REGISTER, RXFP); + uint32_t store_location; + uint32_t fill_level; + uint32_t read_index; + uint8_t store_index = 0; + g_autofree char *path = NULL; + /* + * If all UAF bits are set to 0, then received messages are not stored + * in the RX buffers. + */ + if (s->regs[R_ACCEPTANCE_FILTER_CONTROL_REGISTER]) { + uint32_t acceptance_filter_status = + s->regs[R_ACCEPTANCE_FILTER_CONTROL_REGISTER]; + + for (i = 0; i < 32; i++) { + if (acceptance_filter_status & 0x1) { + uint32_t msg_id_masked = s->regs[R_AFMR_REGISTER + 2 * i] & + frame->can_id; + uint32_t afir_id_masked = s->regs[R_AFIR_REGISTER + 2 * i] & + s->regs[R_AFMR_REGISTER + 2 * i]; + uint16_t std_msg_id_masked = FIELD_EX32(msg_id_masked, + AFIR_REGISTER, AIID); + uint16_t std_afir_id_masked = FIELD_EX32(afir_id_masked, + AFIR_REGISTER, AIID); + uint32_t ext_msg_id_masked = FIELD_EX32(msg_id_masked, + AFIR_REGISTER, + AIID_EXT); + uint32_t ext_afir_id_masked = FIELD_EX32(afir_id_masked, + AFIR_REGISTER, + AIID_EXT); + bool ext_ide = FIELD_EX32(s->regs[R_AFMR_REGISTER + 2 * i], + AFMR_REGISTER, AMIDE); + + if (std_msg_id_masked == std_afir_id_masked) { + if (ext_ide) { + /* Extended message ID message. */ + if (ext_msg_id_masked == ext_afir_id_masked) { + filter_pass = true; + filter_index = i; + + break; + } + } else { + /* Standard message ID. */ + filter_pass = true; + filter_index = i; + + break; + } + } + } + acceptance_filter_status >>= 1; + } + } + + if (!filter_pass) { + path = object_get_canonical_path(OBJECT(s)); + + trace_xlnx_canfd_rx_fifo_filter_reject(path, frame->can_id, + frame->can_dlc); + } else { + if (filter_index <= filter_partition) { + fill_level = ARRAY_FIELD_EX32(s->regs, RX_FIFO_STATUS_REGISTER, FL); + read_index = ARRAY_FIELD_EX32(s->regs, RX_FIFO_STATUS_REGISTER, RI); + store_index = read_index + fill_level; + + if (read_index == s->cfg.rx0_fifo - 1) { + /* + * When ri is s->cfg.rx0_fifo - 1 i.e. max, it goes cyclic that + * means we reset the ri to 0x0. + */ + read_index = 0; + ARRAY_FIELD_DP32(s->regs, RX_FIFO_STATUS_REGISTER, RI, + read_index); + } + + if (store_index > s->cfg.rx0_fifo - 1) { + store_index -= s->cfg.rx0_fifo - 1; + } + + store_location = R_RB_ID_REGISTER + + (store_index * NUM_REGS_PER_MSG_SPACE); + + store_rx_sequential(s, frame, fill_level, read_index, + store_location, s->cfg.rx0_fifo, 0, + filter_index); + } else { + /* RX 1 fill level message */ + fill_level = ARRAY_FIELD_EX32(s->regs, RX_FIFO_STATUS_REGISTER, + FL_1); + read_index = ARRAY_FIELD_EX32(s->regs, RX_FIFO_STATUS_REGISTER, + RI_1); + store_index = read_index + fill_level; + + if (read_index == s->cfg.rx1_fifo - 1) { + /* + * When ri is s->cfg.rx1_fifo - 1 i.e. max, it goes cyclic that + * means we reset the ri to 0x0. + */ + read_index = 0; + ARRAY_FIELD_DP32(s->regs, RX_FIFO_STATUS_REGISTER, RI_1, + read_index); + } + + if (store_index > s->cfg.rx1_fifo - 1) { + store_index -= s->cfg.rx1_fifo - 1; + } + + store_location = R_RB_ID_REGISTER_1 + + (store_index * NUM_REGS_PER_MSG_SPACE); + + store_rx_sequential(s, frame, fill_level, read_index, + store_location, s->cfg.rx1_fifo, 1, + filter_index); + } + + path = object_get_canonical_path(OBJECT(s)); + + trace_xlnx_canfd_rx_data(path, frame->can_id, frame->can_dlc, + frame->flags); + canfd_update_irq(s); + } +} + +static bool tx_ready_check(XlnxVersalCANFDState *s) +{ + if (ARRAY_FIELD_EX32(s->regs, SOFTWARE_RESET_REGISTER, SRST)) { + g_autofree char *path = object_get_canonical_path(OBJECT(s)); + + qemu_log_mask(LOG_GUEST_ERROR, "%s: Attempting to transfer data while" + " XlnxVersalCANFDState is in reset mode\n", path); + + return false; + } + + if (ARRAY_FIELD_EX32(s->regs, SOFTWARE_RESET_REGISTER, CEN) == 0) { + g_autofree char *path = object_get_canonical_path(OBJECT(s)); + + qemu_log_mask(LOG_GUEST_ERROR, "%s: Attempting to transfer data while" + " XlnxVersalCANFDState is in configuration mode." + " Reset the core so operations can start fresh\n", + path); + return false; + } + + if (ARRAY_FIELD_EX32(s->regs, MODE_SELECT_REGISTER, SNOOP)) { + g_autofree char *path = object_get_canonical_path(OBJECT(s)); + + qemu_log_mask(LOG_GUEST_ERROR, "%s: Attempting to transfer data while" + " XlnxVersalCANFDState is in SNOOP MODE\n", + path); + return false; + } + + return true; +} + +static void tx_fifo_stamp(XlnxVersalCANFDState *s, uint32_t tb0_regid) +{ + /* + * If EFC bit in DLC message is set, this means we will store the + * event of this transmitted message with time stamp. + */ + uint32_t dlc_reg_val = 0; + + if (FIELD_EX32(s->regs[tb0_regid + 1], TB0_DLC_REGISTER, EFC)) { + uint8_t dlc_val = FIELD_EX32(s->regs[tb0_regid + 1], TB0_DLC_REGISTER, + DLC); + bool fdf_val = FIELD_EX32(s->regs[tb0_regid + 1], TB0_DLC_REGISTER, + FDF); + bool brs_val = FIELD_EX32(s->regs[tb0_regid + 1], TB0_DLC_REGISTER, + BRS); + uint8_t mm_val = FIELD_EX32(s->regs[tb0_regid + 1], TB0_DLC_REGISTER, + MM); + uint8_t fill_level = ARRAY_FIELD_EX32(s->regs, + TX_EVENT_FIFO_STATUS_REGISTER, + TXE_FL); + uint8_t read_index = ARRAY_FIELD_EX32(s->regs, + TX_EVENT_FIFO_STATUS_REGISTER, + TXE_RI); + uint8_t store_index = fill_level + read_index; + + if ((fill_level) > s->cfg.tx_fifo - 1) { + qemu_log_mask(LOG_GUEST_ERROR, "TX Event Buffer is full." + " Discarding the message\n"); + ARRAY_FIELD_DP32(s->regs, INTERRUPT_STATUS_REGISTER, TXEOFLW, 1); + } else { + if (read_index == s->cfg.tx_fifo - 1) { + /* + * When ri is s->cfg.tx_fifo - 1 i.e. max, it goes cyclic that + * means we reset the ri to 0x0. + */ + read_index = 0; + ARRAY_FIELD_DP32(s->regs, TX_EVENT_FIFO_STATUS_REGISTER, TXE_RI, + read_index); + } + + if (store_index > s->cfg.tx_fifo - 1) { + store_index -= s->cfg.tx_fifo - 1; + } + + assert(store_index < s->cfg.tx_fifo); + + uint32_t tx_event_reg0_id = R_TXE_FIFO_TB_ID_REGISTER + + (store_index * 2); + + /* Store message ID in TX event register. */ + s->regs[tx_event_reg0_id] = s->regs[tb0_regid]; + + uint16_t tx_timestamp = CANFD_TIMER_MAX - + ptimer_get_count(s->canfd_timer); + + /* Store DLC with time stamp in DLC regs. */ + dlc_reg_val = FIELD_DP32(0, TXE_FIFO_TB_DLC_REGISTER, DLC, dlc_val); + dlc_reg_val |= FIELD_DP32(0, TXE_FIFO_TB_DLC_REGISTER, FDF, + fdf_val); + dlc_reg_val |= FIELD_DP32(0, TXE_FIFO_TB_DLC_REGISTER, BRS, + brs_val); + dlc_reg_val |= FIELD_DP32(0, TXE_FIFO_TB_DLC_REGISTER, ET, 0x3); + dlc_reg_val |= FIELD_DP32(0, TXE_FIFO_TB_DLC_REGISTER, MM, mm_val); + dlc_reg_val |= FIELD_DP32(0, TXE_FIFO_TB_DLC_REGISTER, TIMESTAMP, + tx_timestamp); + s->regs[tx_event_reg0_id + 1] = dlc_reg_val; + + ARRAY_FIELD_DP32(s->regs, TX_EVENT_FIFO_STATUS_REGISTER, TXE_FL, + fill_level + 1); + } + } +} + +static gint g_cmp_ids(gconstpointer data1, gconstpointer data2) +{ + tx_ready_reg_info *tx_reg_1 = (tx_ready_reg_info *) data1; + tx_ready_reg_info *tx_reg_2 = (tx_ready_reg_info *) data2; + + return tx_reg_1->can_id - tx_reg_2->can_id; +} + +static void free_list(GSList *list) +{ + GSList *iterator = NULL; + + for (iterator = list; iterator != NULL; iterator = iterator->next) { + g_free((tx_ready_reg_info *)iterator->data); + } + + g_slist_free(list); + + return; +} + +static GSList *prepare_tx_data(XlnxVersalCANFDState *s) +{ + uint8_t i = 0; + GSList *list = NULL; + uint32_t reg_num = 0; + uint32_t reg_ready = s->regs[R_TX_BUFFER_READY_REQUEST_REGISTER]; + + /* First find the messages which are ready for transmission. */ + for (i = 0; i < s->cfg.tx_fifo; i++) { + if (reg_ready & 1) { + reg_num = R_TB_ID_REGISTER + (NUM_REGS_PER_MSG_SPACE * i); + tx_ready_reg_info *temp = g_new(tx_ready_reg_info, 1); + + temp->can_id = s->regs[reg_num]; + temp->reg_num = reg_num; + list = g_slist_prepend(list, temp); + list = g_slist_sort(list, g_cmp_ids); + } + + reg_ready >>= 1; + } + + s->regs[R_TX_BUFFER_READY_REQUEST_REGISTER] = 0; + s->regs[R_TX_BUFFER_CANCEL_REQUEST_REGISTER] = 0; + + return list; +} + +static void transfer_data(XlnxVersalCANFDState *s) +{ + bool canfd_tx = tx_ready_check(s); + GSList *list, *iterator = NULL; + qemu_can_frame frame; + + if (!canfd_tx) { + g_autofree char *path = object_get_canonical_path(OBJECT(s)); + + qemu_log_mask(LOG_GUEST_ERROR, "%s: Controller not enabled for data" + " transfer\n", path); + return; + } + + list = prepare_tx_data(s); + if (list == NULL) { + return; + } + + for (iterator = list; iterator != NULL; iterator = iterator->next) { + regs2frame(s, &frame, + ((tx_ready_reg_info *)iterator->data)->reg_num); + + if (ARRAY_FIELD_EX32(s->regs, STATUS_REGISTER, LBACK)) { + update_rx_sequential(s, &frame); + tx_fifo_stamp(s, ((tx_ready_reg_info *)iterator->data)->reg_num); + + ARRAY_FIELD_DP32(s->regs, INTERRUPT_STATUS_REGISTER, RXOK, 1); + } else { + g_autofree char *path = object_get_canonical_path(OBJECT(s)); + + trace_xlnx_canfd_tx_data(path, frame.can_id, frame.can_dlc, + frame.flags); + can_bus_client_send(&s->bus_client, &frame, 1); + tx_fifo_stamp(s, + ((tx_ready_reg_info *)iterator->data)->reg_num); + + ARRAY_FIELD_DP32(s->regs, INTERRUPT_STATUS_REGISTER, TXRRS, 1); + + if (ARRAY_FIELD_EX32(s->regs, STATUS_REGISTER, SLEEP)) { + canfd_exit_sleep_mode(s); + } + } + } + + ARRAY_FIELD_DP32(s->regs, INTERRUPT_STATUS_REGISTER, TXOK, 1); + free_list(list); + + canfd_update_irq(s); +} + +static uint64_t canfd_srr_pre_write(RegisterInfo *reg, uint64_t val64) +{ + XlnxVersalCANFDState *s = XILINX_CANFD(reg->opaque); + uint32_t val = val64; + + ARRAY_FIELD_DP32(s->regs, SOFTWARE_RESET_REGISTER, CEN, + FIELD_EX32(val, SOFTWARE_RESET_REGISTER, CEN)); + + if (FIELD_EX32(val, SOFTWARE_RESET_REGISTER, SRST)) { + g_autofree char *path = object_get_canonical_path(OBJECT(s)); + + trace_xlnx_canfd_reset(path, val64); + + /* First, core will do software reset then will enter in config mode. */ + canfd_config_reset(s); + } else if (ARRAY_FIELD_EX32(s->regs, SOFTWARE_RESET_REGISTER, CEN) == 0) { + canfd_config_mode(s); + } else { + /* + * Leave config mode. Now XlnxVersalCANFD core will enter Normal, Sleep, + * snoop or Loopback mode depending upon LBACK, SLEEP, SNOOP register + * states. + */ + ARRAY_FIELD_DP32(s->regs, STATUS_REGISTER, CONFIG, 0); + + ptimer_transaction_begin(s->canfd_timer); + ptimer_set_count(s->canfd_timer, 0); + ptimer_transaction_commit(s->canfd_timer); + update_status_register_mode_bits(s); + transfer_data(s); + } + + return s->regs[R_SOFTWARE_RESET_REGISTER]; +} + +static uint64_t filter_mask(RegisterInfo *reg, uint64_t val64) +{ + XlnxVersalCANFDState *s = XILINX_CANFD(reg->opaque); + uint32_t reg_idx = (reg->access->addr) / 4; + uint32_t val = val64; + uint32_t filter_offset = (reg_idx - R_AFMR_REGISTER) / 2; + + if (!(s->regs[R_ACCEPTANCE_FILTER_CONTROL_REGISTER] & + (1 << filter_offset))) { + s->regs[reg_idx] = val; + } else { + g_autofree char *path = object_get_canonical_path(OBJECT(s)); + + qemu_log_mask(LOG_GUEST_ERROR, "%s: Acceptance filter %d not enabled\n", + path, filter_offset + 1); + } + + return s->regs[reg_idx]; +} + +static uint64_t filter_id(RegisterInfo *reg, uint64_t val64) +{ + XlnxVersalCANFDState *s = XILINX_CANFD(reg->opaque); + hwaddr reg_idx = (reg->access->addr) / 4; + uint32_t val = val64; + uint32_t filter_offset = (reg_idx - R_AFIR_REGISTER) / 2; + + if (!(s->regs[R_ACCEPTANCE_FILTER_CONTROL_REGISTER] & + (1 << filter_offset))) { + s->regs[reg_idx] = val; + } else { + g_autofree char *path = object_get_canonical_path(OBJECT(s)); + + qemu_log_mask(LOG_GUEST_ERROR, "%s: Acceptance filter %d not enabled\n", + path, filter_offset + 1); + } + + return s->regs[reg_idx]; +} + +static uint64_t canfd_tx_fifo_status_prew(RegisterInfo *reg, uint64_t val64) +{ + XlnxVersalCANFDState *s = XILINX_CANFD(reg->opaque); + uint32_t val = val64; + uint8_t read_ind = 0; + uint8_t fill_ind = ARRAY_FIELD_EX32(s->regs, TX_EVENT_FIFO_STATUS_REGISTER, + TXE_FL); + + if (FIELD_EX32(val, TX_EVENT_FIFO_STATUS_REGISTER, TXE_IRI) && fill_ind) { + read_ind = ARRAY_FIELD_EX32(s->regs, TX_EVENT_FIFO_STATUS_REGISTER, + TXE_RI) + 1; + + if (read_ind > s->cfg.tx_fifo - 1) { + read_ind = 0; + } + + /* + * Increase the read index by 1 and decrease the fill level by 1. + */ + ARRAY_FIELD_DP32(s->regs, TX_EVENT_FIFO_STATUS_REGISTER, TXE_RI, + read_ind); + ARRAY_FIELD_DP32(s->regs, TX_EVENT_FIFO_STATUS_REGISTER, TXE_FL, + fill_ind - 1); + } + + return s->regs[R_TX_EVENT_FIFO_STATUS_REGISTER]; +} + +static uint64_t canfd_rx_fifo_status_prew(RegisterInfo *reg, uint64_t val64) +{ + XlnxVersalCANFDState *s = XILINX_CANFD(reg->opaque); + uint32_t val = val64; + uint8_t read_ind = 0; + uint8_t fill_ind = 0; + + if (FIELD_EX32(val, RX_FIFO_STATUS_REGISTER, IRI)) { + /* FL index is zero, setting IRI bit has no effect. */ + if (FIELD_EX32(val, RX_FIFO_STATUS_REGISTER, FL) != 0) { + read_ind = FIELD_EX32(val, RX_FIFO_STATUS_REGISTER, RI) + 1; + + if (read_ind > s->cfg.rx0_fifo - 1) { + read_ind = 0; + } + + fill_ind = FIELD_EX32(val, RX_FIFO_STATUS_REGISTER, FL) - 1; + + ARRAY_FIELD_DP32(s->regs, RX_FIFO_STATUS_REGISTER, RI, read_ind); + ARRAY_FIELD_DP32(s->regs, RX_FIFO_STATUS_REGISTER, FL, fill_ind); + } + } + + if (FIELD_EX32(val, RX_FIFO_STATUS_REGISTER, IRI_1)) { + /* FL_1 index is zero, setting IRI_1 bit has no effect. */ + if (FIELD_EX32(val, RX_FIFO_STATUS_REGISTER, FL_1) != 0) { + read_ind = FIELD_EX32(val, RX_FIFO_STATUS_REGISTER, RI_1) + 1; + + if (read_ind > s->cfg.rx1_fifo - 1) { + read_ind = 0; + } + + fill_ind = FIELD_EX32(val, RX_FIFO_STATUS_REGISTER, FL_1) - 1; + + ARRAY_FIELD_DP32(s->regs, RX_FIFO_STATUS_REGISTER, RI_1, read_ind); + ARRAY_FIELD_DP32(s->regs, RX_FIFO_STATUS_REGISTER, FL_1, fill_ind); + } + } + + return s->regs[R_RX_FIFO_STATUS_REGISTER]; +} + +static uint64_t canfd_tsr_pre_write(RegisterInfo *reg, uint64_t val64) +{ + XlnxVersalCANFDState *s = XILINX_CANFD(reg->opaque); + uint32_t val = val64; + + if (FIELD_EX32(val, TIMESTAMP_REGISTER, CTS)) { + ARRAY_FIELD_DP32(s->regs, TIMESTAMP_REGISTER, TIMESTAMP_CNT, 0); + ptimer_transaction_begin(s->canfd_timer); + ptimer_set_count(s->canfd_timer, 0); + ptimer_transaction_commit(s->canfd_timer); + } + + return 0; +} + +static uint64_t canfd_trr_reg_prew(RegisterInfo *reg, uint64_t val64) +{ + XlnxVersalCANFDState *s = XILINX_CANFD(reg->opaque); + + if (ARRAY_FIELD_EX32(s->regs, MODE_SELECT_REGISTER, SNOOP)) { + g_autofree char *path = object_get_canonical_path(OBJECT(s)); + + qemu_log_mask(LOG_GUEST_ERROR, "%s: Controller is in SNOOP mode." + " tx_ready_register will stay in reset mode\n", path); + return 0; + } else { + return val64; + } +} + +static void canfd_trr_reg_postw(RegisterInfo *reg, uint64_t val64) +{ + XlnxVersalCANFDState *s = XILINX_CANFD(reg->opaque); + + transfer_data(s); +} + +static void canfd_cancel_reg_postw(RegisterInfo *reg, uint64_t val64) +{ + XlnxVersalCANFDState *s = XILINX_CANFD(reg->opaque); + + process_cancellation_requests(s); +} + +static uint64_t canfd_write_check_prew(RegisterInfo *reg, uint64_t val64) +{ + XlnxVersalCANFDState *s = XILINX_CANFD(reg->opaque); + uint32_t val = val64; + + if (ARRAY_FIELD_EX32(s->regs, SOFTWARE_RESET_REGISTER, CEN) == 0) { + return val; + } + return 0; +} + +static const RegisterAccessInfo canfd_tx_regs[] = { + { .name = "TB_ID_REGISTER", .addr = A_TB_ID_REGISTER, + },{ .name = "TB0_DLC_REGISTER", .addr = A_TB0_DLC_REGISTER, + },{ .name = "TB_DW0_REGISTER", .addr = A_TB_DW0_REGISTER, + },{ .name = "TB_DW1_REGISTER", .addr = A_TB_DW1_REGISTER, + },{ .name = "TB_DW2_REGISTER", .addr = A_TB_DW2_REGISTER, + },{ .name = "TB_DW3_REGISTER", .addr = A_TB_DW3_REGISTER, + },{ .name = "TB_DW4_REGISTER", .addr = A_TB_DW4_REGISTER, + },{ .name = "TB_DW5_REGISTER", .addr = A_TB_DW5_REGISTER, + },{ .name = "TB_DW6_REGISTER", .addr = A_TB_DW6_REGISTER, + },{ .name = "TB_DW7_REGISTER", .addr = A_TB_DW7_REGISTER, + },{ .name = "TB_DW8_REGISTER", .addr = A_TB_DW8_REGISTER, + },{ .name = "TB_DW9_REGISTER", .addr = A_TB_DW9_REGISTER, + },{ .name = "TB_DW10_REGISTER", .addr = A_TB_DW10_REGISTER, + },{ .name = "TB_DW11_REGISTER", .addr = A_TB_DW11_REGISTER, + },{ .name = "TB_DW12_REGISTER", .addr = A_TB_DW12_REGISTER, + },{ .name = "TB_DW13_REGISTER", .addr = A_TB_DW13_REGISTER, + },{ .name = "TB_DW14_REGISTER", .addr = A_TB_DW14_REGISTER, + },{ .name = "TB_DW15_REGISTER", .addr = A_TB_DW15_REGISTER, + } +}; + +static const RegisterAccessInfo canfd_rx0_regs[] = { + { .name = "RB_ID_REGISTER", .addr = A_RB_ID_REGISTER, + .ro = 0xffffffff, + },{ .name = "RB_DLC_REGISTER", .addr = A_RB_DLC_REGISTER, + .ro = 0xfe1fffff, + },{ .name = "RB_DW0_REGISTER", .addr = A_RB_DW0_REGISTER, + .ro = 0xffffffff, + },{ .name = "RB_DW1_REGISTER", .addr = A_RB_DW1_REGISTER, + .ro = 0xffffffff, + },{ .name = "RB_DW2_REGISTER", .addr = A_RB_DW2_REGISTER, + .ro = 0xffffffff, + },{ .name = "RB_DW3_REGISTER", .addr = A_RB_DW3_REGISTER, + .ro = 0xffffffff, + },{ .name = "RB_DW4_REGISTER", .addr = A_RB_DW4_REGISTER, + .ro = 0xffffffff, + },{ .name = "RB_DW5_REGISTER", .addr = A_RB_DW5_REGISTER, + .ro = 0xffffffff, + },{ .name = "RB_DW6_REGISTER", .addr = A_RB_DW6_REGISTER, + .ro = 0xffffffff, + },{ .name = "RB_DW7_REGISTER", .addr = A_RB_DW7_REGISTER, + .ro = 0xffffffff, + },{ .name = "RB_DW8_REGISTER", .addr = A_RB_DW8_REGISTER, + .ro = 0xffffffff, + },{ .name = "RB_DW9_REGISTER", .addr = A_RB_DW9_REGISTER, + .ro = 0xffffffff, + },{ .name = "RB_DW10_REGISTER", .addr = A_RB_DW10_REGISTER, + .ro = 0xffffffff, + },{ .name = "RB_DW11_REGISTER", .addr = A_RB_DW11_REGISTER, + .ro = 0xffffffff, + },{ .name = "RB_DW12_REGISTER", .addr = A_RB_DW12_REGISTER, + .ro = 0xffffffff, + },{ .name = "RB_DW13_REGISTER", .addr = A_RB_DW13_REGISTER, + .ro = 0xffffffff, + },{ .name = "RB_DW14_REGISTER", .addr = A_RB_DW14_REGISTER, + .ro = 0xffffffff, + },{ .name = "RB_DW15_REGISTER", .addr = A_RB_DW15_REGISTER, + .ro = 0xffffffff, + } +}; + +static const RegisterAccessInfo canfd_rx1_regs[] = { + { .name = "RB_ID_REGISTER_1", .addr = A_RB_ID_REGISTER_1, + .ro = 0xffffffff, + },{ .name = "RB_DLC_REGISTER_1", .addr = A_RB_DLC_REGISTER_1, + .ro = 0xfe1fffff, + },{ .name = "RB0_DW0_REGISTER_1", .addr = A_RB0_DW0_REGISTER_1, + .ro = 0xffffffff, + },{ .name = "RB_DW1_REGISTER_1", .addr = A_RB_DW1_REGISTER_1, + .ro = 0xffffffff, + },{ .name = "RB_DW2_REGISTER_1", .addr = A_RB_DW2_REGISTER_1, + .ro = 0xffffffff, + },{ .name = "RB_DW3_REGISTER_1", .addr = A_RB_DW3_REGISTER_1, + .ro = 0xffffffff, + },{ .name = "RB_DW4_REGISTER_1", .addr = A_RB_DW4_REGISTER_1, + .ro = 0xffffffff, + },{ .name = "RB_DW5_REGISTER_1", .addr = A_RB_DW5_REGISTER_1, + .ro = 0xffffffff, + },{ .name = "RB_DW6_REGISTER_1", .addr = A_RB_DW6_REGISTER_1, + .ro = 0xffffffff, + },{ .name = "RB_DW7_REGISTER_1", .addr = A_RB_DW7_REGISTER_1, + .ro = 0xffffffff, + },{ .name = "RB_DW8_REGISTER_1", .addr = A_RB_DW8_REGISTER_1, + .ro = 0xffffffff, + },{ .name = "RB_DW9_REGISTER_1", .addr = A_RB_DW9_REGISTER_1, + .ro = 0xffffffff, + },{ .name = "RB_DW10_REGISTER_1", .addr = A_RB_DW10_REGISTER_1, + .ro = 0xffffffff, + },{ .name = "RB_DW11_REGISTER_1", .addr = A_RB_DW11_REGISTER_1, + .ro = 0xffffffff, + },{ .name = "RB_DW12_REGISTER_1", .addr = A_RB_DW12_REGISTER_1, + .ro = 0xffffffff, + },{ .name = "RB_DW13_REGISTER_1", .addr = A_RB_DW13_REGISTER_1, + .ro = 0xffffffff, + },{ .name = "RB_DW14_REGISTER_1", .addr = A_RB_DW14_REGISTER_1, + .ro = 0xffffffff, + },{ .name = "RB_DW15_REGISTER_1", .addr = A_RB_DW15_REGISTER_1, + .ro = 0xffffffff, + } +}; + +/* Acceptance filter registers. */ +static const RegisterAccessInfo canfd_af_regs[] = { + { .name = "AFMR_REGISTER", .addr = A_AFMR_REGISTER, + .pre_write = filter_mask, + },{ .name = "AFIR_REGISTER", .addr = A_AFIR_REGISTER, + .pre_write = filter_id, + } +}; + +static const RegisterAccessInfo canfd_txe_regs[] = { + { .name = "TXE_FIFO_TB_ID_REGISTER", .addr = A_TXE_FIFO_TB_ID_REGISTER, + .ro = 0xffffffff, + },{ .name = "TXE_FIFO_TB_DLC_REGISTER", .addr = A_TXE_FIFO_TB_DLC_REGISTER, + .ro = 0xffffffff, + } +}; + +static const RegisterAccessInfo canfd_regs_info[] = { + { .name = "SOFTWARE_RESET_REGISTER", .addr = A_SOFTWARE_RESET_REGISTER, + .pre_write = canfd_srr_pre_write, + },{ .name = "MODE_SELECT_REGISTER", .addr = A_MODE_SELECT_REGISTER, + .pre_write = canfd_msr_pre_write, + },{ .name = "ARBITRATION_PHASE_BAUD_RATE_PRESCALER_REGISTER", + .addr = A_ARBITRATION_PHASE_BAUD_RATE_PRESCALER_REGISTER, + .pre_write = canfd_write_check_prew, + },{ .name = "ARBITRATION_PHASE_BIT_TIMING_REGISTER", + .addr = A_ARBITRATION_PHASE_BIT_TIMING_REGISTER, + .pre_write = canfd_write_check_prew, + },{ .name = "ERROR_COUNTER_REGISTER", .addr = A_ERROR_COUNTER_REGISTER, + .ro = 0xffff, + },{ .name = "ERROR_STATUS_REGISTER", .addr = A_ERROR_STATUS_REGISTER, + .w1c = 0xf1f, + },{ .name = "STATUS_REGISTER", .addr = A_STATUS_REGISTER, + .reset = 0x1, + .ro = 0x7f17ff, + },{ .name = "INTERRUPT_STATUS_REGISTER", + .addr = A_INTERRUPT_STATUS_REGISTER, + .ro = 0xffffff7f, + },{ .name = "INTERRUPT_ENABLE_REGISTER", + .addr = A_INTERRUPT_ENABLE_REGISTER, + .post_write = canfd_ier_post_write, + },{ .name = "INTERRUPT_CLEAR_REGISTER", + .addr = A_INTERRUPT_CLEAR_REGISTER, .pre_write = canfd_icr_pre_write, + },{ .name = "TIMESTAMP_REGISTER", .addr = A_TIMESTAMP_REGISTER, + .ro = 0xffff0000, + .pre_write = canfd_tsr_pre_write, + },{ .name = "DATA_PHASE_BAUD_RATE_PRESCALER_REGISTER", + .addr = A_DATA_PHASE_BAUD_RATE_PRESCALER_REGISTER, + .pre_write = canfd_write_check_prew, + },{ .name = "DATA_PHASE_BIT_TIMING_REGISTER", + .addr = A_DATA_PHASE_BIT_TIMING_REGISTER, + .pre_write = canfd_write_check_prew, + },{ .name = "TX_BUFFER_READY_REQUEST_REGISTER", + .addr = A_TX_BUFFER_READY_REQUEST_REGISTER, + .pre_write = canfd_trr_reg_prew, + .post_write = canfd_trr_reg_postw, + },{ .name = "INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER", + .addr = A_INTERRUPT_ENABLE_TX_BUFFER_READY_REQUEST_REGISTER, + },{ .name = "TX_BUFFER_CANCEL_REQUEST_REGISTER", + .addr = A_TX_BUFFER_CANCEL_REQUEST_REGISTER, + .post_write = canfd_cancel_reg_postw, + },{ .name = "INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER", + .addr = A_INTERRUPT_ENABLE_TX_BUFFER_CANCELLATION_REQUEST_REGISTER, + },{ .name = "TX_EVENT_FIFO_STATUS_REGISTER", + .addr = A_TX_EVENT_FIFO_STATUS_REGISTER, + .ro = 0x3f1f, .pre_write = canfd_tx_fifo_status_prew, + },{ .name = "TX_EVENT_FIFO_WATERMARK_REGISTER", + .addr = A_TX_EVENT_FIFO_WATERMARK_REGISTER, + .reset = 0xf, + .pre_write = canfd_write_check_prew, + },{ .name = "ACCEPTANCE_FILTER_CONTROL_REGISTER", + .addr = A_ACCEPTANCE_FILTER_CONTROL_REGISTER, + },{ .name = "RX_FIFO_STATUS_REGISTER", .addr = A_RX_FIFO_STATUS_REGISTER, + .ro = 0x7f3f7f3f, .pre_write = canfd_rx_fifo_status_prew, + },{ .name = "RX_FIFO_WATERMARK_REGISTER", + .addr = A_RX_FIFO_WATERMARK_REGISTER, + .reset = 0x1f0f0f, + .pre_write = canfd_write_check_prew, + } +}; + +static void xlnx_versal_canfd_ptimer_cb(void *opaque) +{ + /* No action required on the timer rollover. */ +} + +static const MemoryRegionOps canfd_ops = { + .read = register_read_memory, + .write = register_write_memory, + .endianness = DEVICE_LITTLE_ENDIAN, + .valid = { + .min_access_size = 4, + .max_access_size = 4, + }, +}; + +static void canfd_reset(DeviceState *dev) +{ + XlnxVersalCANFDState *s = XILINX_CANFD(dev); + unsigned int i; + + for (i = 0; i < ARRAY_SIZE(s->reg_info); ++i) { + register_reset(&s->reg_info[i]); + } + + ptimer_transaction_begin(s->canfd_timer); + ptimer_set_count(s->canfd_timer, 0); + ptimer_transaction_commit(s->canfd_timer); +} + +static bool can_xilinx_canfd_receive(CanBusClientState *client) +{ + XlnxVersalCANFDState *s = container_of(client, XlnxVersalCANFDState, + bus_client); + + bool reset_state = ARRAY_FIELD_EX32(s->regs, SOFTWARE_RESET_REGISTER, SRST); + bool can_enabled = ARRAY_FIELD_EX32(s->regs, SOFTWARE_RESET_REGISTER, CEN); + + return !reset_state && can_enabled; +} + +static ssize_t canfd_xilinx_receive(CanBusClientState *client, + const qemu_can_frame *buf, + size_t buf_size) +{ + XlnxVersalCANFDState *s = container_of(client, XlnxVersalCANFDState, + bus_client); + const qemu_can_frame *frame = buf; + + assert(buf_size > 0); + + if (ARRAY_FIELD_EX32(s->regs, STATUS_REGISTER, LBACK)) { + /* + * XlnxVersalCANFDState will not participate in normal bus communication + * and does not receive any messages transmitted by other CAN nodes. + */ + return 1; + } + + /* Update the status register that we are receiving message. */ + ARRAY_FIELD_DP32(s->regs, STATUS_REGISTER, BBSY, 1); + + if (ARRAY_FIELD_EX32(s->regs, STATUS_REGISTER, SNOOP)) { + /* Snoop Mode: Just keep the data. no response back. */ + update_rx_sequential(s, frame); + } else { + if ((ARRAY_FIELD_EX32(s->regs, STATUS_REGISTER, SLEEP))) { + /* + * XlnxVersalCANFDState is in sleep mode. Any data on bus will bring + * it to the wake up state. + */ + canfd_exit_sleep_mode(s); + } + + update_rx_sequential(s, frame); + } + + /* Message processing done. Update the status back to !busy */ + ARRAY_FIELD_DP32(s->regs, STATUS_REGISTER, BBSY, 0); + return 1; +} + +static CanBusClientInfo canfd_xilinx_bus_client_info = { + .can_receive = can_xilinx_canfd_receive, + .receive = canfd_xilinx_receive, +}; + +static int xlnx_canfd_connect_to_bus(XlnxVersalCANFDState *s, + CanBusState *bus) +{ + s->bus_client.info = &canfd_xilinx_bus_client_info; + + return can_bus_insert_client(bus, &s->bus_client); +} + +#define NUM_REG_PER_AF ARRAY_SIZE(canfd_af_regs) +#define NUM_AF 32 +#define NUM_REG_PER_TXE ARRAY_SIZE(canfd_txe_regs) +#define NUM_TXE 32 + +static int canfd_populate_regarray(XlnxVersalCANFDState *s, + RegisterInfoArray *r_array, int pos, + const RegisterAccessInfo *rae, + int num_rae) +{ + int i; + + for (i = 0; i < num_rae; i++) { + int index = rae[i].addr / 4; + RegisterInfo *r = &s->reg_info[index]; + + object_initialize(r, sizeof(*r), TYPE_REGISTER); + + *r = (RegisterInfo) { + .data = &s->regs[index], + .data_size = sizeof(uint32_t), + .access = &rae[i], + .opaque = OBJECT(s), + }; + + r_array->r[i + pos] = r; + } + return i + pos; +} + +static void canfd_create_rai(RegisterAccessInfo *rai_array, + const RegisterAccessInfo *canfd_regs, + int template_rai_array_sz, + int num_template_to_copy) +{ + int i; + int reg_num; + + for (reg_num = 0; reg_num < num_template_to_copy; reg_num++) { + int pos = reg_num * template_rai_array_sz; + + memcpy(rai_array + pos, canfd_regs, + template_rai_array_sz * sizeof(RegisterAccessInfo)); + + for (i = 0; i < template_rai_array_sz; i++) { + const char *name = canfd_regs[i].name; + uint64_t addr = canfd_regs[i].addr; + rai_array[i + pos].name = g_strdup_printf("%s%d", name, reg_num); + rai_array[i + pos].addr = addr + pos * 4; + } + } +} + +static RegisterInfoArray *canfd_create_regarray(XlnxVersalCANFDState *s) +{ + const char *device_prefix = object_get_typename(OBJECT(s)); + uint64_t memory_size = XLNX_VERSAL_CANFD_R_MAX * 4; + int num_regs; + int pos = 0; + RegisterInfoArray *r_array; + + num_regs = ARRAY_SIZE(canfd_regs_info) + + s->cfg.tx_fifo * NUM_REGS_PER_MSG_SPACE + + s->cfg.rx0_fifo * NUM_REGS_PER_MSG_SPACE + + NUM_AF * NUM_REG_PER_AF + + NUM_TXE * NUM_REG_PER_TXE; + + s->tx_regs = g_new0(RegisterAccessInfo, + s->cfg.tx_fifo * ARRAY_SIZE(canfd_tx_regs)); + + canfd_create_rai(s->tx_regs, canfd_tx_regs, + ARRAY_SIZE(canfd_tx_regs), s->cfg.tx_fifo); + + s->rx0_regs = g_new0(RegisterAccessInfo, + s->cfg.rx0_fifo * ARRAY_SIZE(canfd_rx0_regs)); + + canfd_create_rai(s->rx0_regs, canfd_rx0_regs, + ARRAY_SIZE(canfd_rx0_regs), s->cfg.rx0_fifo); + + s->af_regs = g_new0(RegisterAccessInfo, + NUM_AF * ARRAY_SIZE(canfd_af_regs)); + + canfd_create_rai(s->af_regs, canfd_af_regs, + ARRAY_SIZE(canfd_af_regs), NUM_AF); + + s->txe_regs = g_new0(RegisterAccessInfo, + NUM_TXE * ARRAY_SIZE(canfd_txe_regs)); + + canfd_create_rai(s->txe_regs, canfd_txe_regs, + ARRAY_SIZE(canfd_txe_regs), NUM_TXE); + + if (s->cfg.enable_rx_fifo1) { + num_regs += s->cfg.rx1_fifo * NUM_REGS_PER_MSG_SPACE; + + s->rx1_regs = g_new0(RegisterAccessInfo, + s->cfg.rx1_fifo * ARRAY_SIZE(canfd_rx1_regs)); + + canfd_create_rai(s->rx1_regs, canfd_rx1_regs, + ARRAY_SIZE(canfd_rx1_regs), s->cfg.rx1_fifo); + } + + r_array = g_new0(RegisterInfoArray, 1); + r_array->r = g_new0(RegisterInfo * , num_regs); + r_array->num_elements = num_regs; + r_array->prefix = device_prefix; + + pos = canfd_populate_regarray(s, r_array, pos, + canfd_regs_info, + ARRAY_SIZE(canfd_regs_info)); + pos = canfd_populate_regarray(s, r_array, pos, + s->tx_regs, s->cfg.tx_fifo * + NUM_REGS_PER_MSG_SPACE); + pos = canfd_populate_regarray(s, r_array, pos, + s->rx0_regs, s->cfg.rx0_fifo * + NUM_REGS_PER_MSG_SPACE); + if (s->cfg.enable_rx_fifo1) { + pos = canfd_populate_regarray(s, r_array, pos, + s->rx1_regs, s->cfg.rx1_fifo * + NUM_REGS_PER_MSG_SPACE); + } + pos = canfd_populate_regarray(s, r_array, pos, + s->af_regs, NUM_AF * NUM_REG_PER_AF); + pos = canfd_populate_regarray(s, r_array, pos, + s->txe_regs, NUM_TXE * NUM_REG_PER_TXE); + + memory_region_init_io(&r_array->mem, OBJECT(s), &canfd_ops, r_array, + device_prefix, memory_size); + return r_array; +} + +static void canfd_realize(DeviceState *dev, Error **errp) +{ + XlnxVersalCANFDState *s = XILINX_CANFD(dev); + RegisterInfoArray *reg_array; + + reg_array = canfd_create_regarray(s); + memory_region_add_subregion(&s->iomem, 0x00, &reg_array->mem); + sysbus_init_mmio(SYS_BUS_DEVICE(dev), &s->iomem); + sysbus_init_irq(SYS_BUS_DEVICE(dev), &s->irq_canfd_int); + + if (s->canfdbus) { + if (xlnx_canfd_connect_to_bus(s, s->canfdbus) < 0) { + g_autofree char *path = object_get_canonical_path(OBJECT(s)); + + error_setg(errp, "%s: xlnx_canfd_connect_to_bus failed", path); + return; + } + + } + + /* Allocate a new timer. */ + s->canfd_timer = ptimer_init(xlnx_versal_canfd_ptimer_cb, s, + PTIMER_POLICY_WRAP_AFTER_ONE_PERIOD | + PTIMER_POLICY_TRIGGER_ONLY_ON_DECREMENT | + PTIMER_POLICY_NO_IMMEDIATE_RELOAD); + + ptimer_transaction_begin(s->canfd_timer); + + ptimer_set_freq(s->canfd_timer, s->cfg.ext_clk_freq); + ptimer_set_limit(s->canfd_timer, CANFD_TIMER_MAX, 1); + ptimer_run(s->canfd_timer, 0); + ptimer_transaction_commit(s->canfd_timer); +} + +static void canfd_init(Object *obj) +{ + XlnxVersalCANFDState *s = XILINX_CANFD(obj); + + memory_region_init(&s->iomem, obj, TYPE_XILINX_CANFD, + XLNX_VERSAL_CANFD_R_MAX * 4); +} + +static const VMStateDescription vmstate_canfd = { + .name = TYPE_XILINX_CANFD, + .version_id = 1, + .minimum_version_id = 1, + .fields = (VMStateField[]) { + VMSTATE_UINT32_ARRAY(regs, XlnxVersalCANFDState, + XLNX_VERSAL_CANFD_R_MAX), + VMSTATE_PTIMER(canfd_timer, XlnxVersalCANFDState), + VMSTATE_END_OF_LIST(), + } +}; + +static Property canfd_core_properties[] = { + DEFINE_PROP_UINT8("rx-fifo0", XlnxVersalCANFDState, cfg.rx0_fifo, 0x40), + DEFINE_PROP_UINT8("rx-fifo1", XlnxVersalCANFDState, cfg.rx1_fifo, 0x40), + DEFINE_PROP_UINT8("tx-fifo", XlnxVersalCANFDState, cfg.tx_fifo, 0x20), + DEFINE_PROP_BOOL("enable-rx-fifo1", XlnxVersalCANFDState, + cfg.enable_rx_fifo1, true), + DEFINE_PROP_UINT32("ext_clk_freq", XlnxVersalCANFDState, cfg.ext_clk_freq, + CANFD_DEFAULT_CLOCK), + DEFINE_PROP_LINK("canfdbus", XlnxVersalCANFDState, canfdbus, TYPE_CAN_BUS, + CanBusState *), + DEFINE_PROP_END_OF_LIST(), +}; + +static void canfd_class_init(ObjectClass *klass, void *data) +{ + DeviceClass *dc = DEVICE_CLASS(klass); + + dc->reset = canfd_reset; + dc->realize = canfd_realize; + device_class_set_props(dc, canfd_core_properties); + dc->vmsd = &vmstate_canfd; +} + +static const TypeInfo canfd_info = { + .name = TYPE_XILINX_CANFD, + .parent = TYPE_SYS_BUS_DEVICE, + .instance_size = sizeof(XlnxVersalCANFDState), + .class_init = canfd_class_init, + .instance_init = canfd_init, +}; + +static void canfd_register_types(void) +{ + type_register_static(&canfd_info); +} + +type_init(canfd_register_types) diff --git a/hw/net/can/meson.build b/hw/net/can/meson.build index XXXXXXX..XXXXXXX 100644 --- a/hw/net/can/meson.build +++ b/hw/net/can/meson.build @@ -XXX,XX +XXX,XX @@ softmmu_ss.add(when: 'CONFIG_CAN_PCI', if_true: files('can_mioe3680_pci.c')) softmmu_ss.add(when: 'CONFIG_CAN_CTUCANFD', if_true: files('ctucan_core.c')) softmmu_ss.add(when: 'CONFIG_CAN_CTUCANFD_PCI', if_true: files('ctucan_pci.c')) softmmu_ss.add(when: 'CONFIG_XLNX_ZYNQMP', if_true: files('xlnx-zynqmp-can.c')) +softmmu_ss.add(when: 'CONFIG_XLNX_VERSAL', if_true: files('xlnx-versal-canfd.c')) diff --git a/hw/net/can/trace-events b/hw/net/can/trace-events index XXXXXXX..XXXXXXX 100644 --- a/hw/net/can/trace-events +++ b/hw/net/can/trace-events @@ -XXX,XX +XXX,XX @@ xlnx_can_filter_mask_pre_write(uint8_t filter_num, uint32_t value) "Filter%d MAS xlnx_can_tx_data(uint32_t id, uint8_t dlc, uint8_t db0, uint8_t db1, uint8_t db2, uint8_t db3, uint8_t db4, uint8_t db5, uint8_t db6, uint8_t db7) "Frame: ID: 0x%08x DLC: 0x%02x DATA: 0x%02x 0x%02x 0x%02x 0x%02x 0x%02x 0x%02x 0x%02x 0x%02x" xlnx_can_rx_data(uint32_t id, uint32_t dlc, uint8_t db0, uint8_t db1, uint8_t db2, uint8_t db3, uint8_t db4, uint8_t db5, uint8_t db6, uint8_t db7) "Frame: ID: 0x%08x DLC: 0x%02x DATA: 0x%02x 0x%02x 0x%02x 0x%02x 0x%02x 0x%02x 0x%02x 0x%02x" xlnx_can_rx_discard(uint32_t status) "Controller is not enabled for bus communication. Status Register: 0x%08x" + +# xlnx-versal-canfd.c +xlnx_canfd_update_irq(char *path, uint32_t isr, uint32_t ier, uint32_t irq) "%s: ISR: 0x%08x IER: 0x%08x IRQ: 0x%08x" +xlnx_canfd_rx_fifo_filter_reject(char *path, uint32_t id, uint8_t dlc) "%s: Frame: ID: 0x%08x DLC: 0x%02x" +xlnx_canfd_rx_data(char *path, uint32_t id, uint8_t dlc, uint8_t flags) "%s: Frame: ID: 0x%08x DLC: 0x%02x CANFD Flag: 0x%02x" +xlnx_canfd_tx_data(char *path, uint32_t id, uint8_t dlc, uint8_t flgas) "%s: Frame: ID: 0x%08x DLC: 0x%02x CANFD Flag: 0x%02x" +xlnx_canfd_reset(char *path, uint32_t val) "%s: Resetting controller with value = 0x%08x" -- 2.34.1 From: Vikram Garhwal <vikram.garhwal@amd.com> Connect CANFD0 and CANFD1 on the Versal-virt machine and update xlnx-versal-virt document with CANFD command line examples. Signed-off-by: Vikram Garhwal <vikram.garhwal@amd.com> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Francisco Iglesias <frasse.iglesias@gmail.com> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- docs/system/arm/xlnx-versal-virt.rst | 31 ++++++++++++++++ include/hw/arm/xlnx-versal.h | 12 +++++++ hw/arm/xlnx-versal-virt.c | 53 ++++++++++++++++++++++++++++ hw/arm/xlnx-versal.c | 37 +++++++++++++++++++ 4 files changed, 133 insertions(+) diff --git a/docs/system/arm/xlnx-versal-virt.rst b/docs/system/arm/xlnx-versal-virt.rst index XXXXXXX..XXXXXXX 100644 --- a/docs/system/arm/xlnx-versal-virt.rst +++ b/docs/system/arm/xlnx-versal-virt.rst @@ -XXX,XX +XXX,XX @@ Implemented devices: - DDR memory - BBRAM (36 bytes of Battery-backed RAM) - eFUSE (3072 bytes of one-time field-programmable bit array) +- 2 CANFDs QEMU does not yet model any other devices, including the PL and the AI Engine. @@ -XXX,XX +XXX,XX @@ To use a different index value, N, from default of 1, add: Better yet, do not use actual product data when running guest image on this Xilinx Versal Virt board. + +Using CANFDs for Versal Virt +"""""""""""""""""""""""""""" +Versal CANFD controller is developed based on SocketCAN and QEMU CAN bus +implementation. Bus connection and socketCAN connection for each CAN module +can be set through command lines. + +To connect both CANFD0 and CANFD1 on the same bus: + +.. code-block:: bash + + -object can-bus,id=canbus -machine canbus0=canbus -machine canbus1=canbus + +To connect CANFD0 and CANFD1 to separate buses: + +.. code-block:: bash + + -object can-bus,id=canbus0 -object can-bus,id=canbus1 \ + -machine canbus0=canbus0 -machine canbus1=canbus1 + +The SocketCAN interface can connect to a Physical or a Virtual CAN interfaces on +the host machine. Please check this document to learn about CAN interface on +Linux: docs/system/devices/can.rst + +To connect CANFD0 and CANFD1 to host machine's CAN interface can0: + +.. code-block:: bash + + -object can-bus,id=canbus -machine canbus0=canbus -machine canbus1=canbus + -object can-host-socketcan,id=canhost0,if=can0,canbus=canbus diff --git a/include/hw/arm/xlnx-versal.h b/include/hw/arm/xlnx-versal.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/xlnx-versal.h +++ b/include/hw/arm/xlnx-versal.h @@ -XXX,XX +XXX,XX @@ #include "hw/dma/xlnx_csu_dma.h" #include "hw/misc/xlnx-versal-crl.h" #include "hw/misc/xlnx-versal-pmc-iou-slcr.h" +#include "hw/net/xlnx-versal-canfd.h" #define TYPE_XLNX_VERSAL "xlnx-versal" OBJECT_DECLARE_SIMPLE_TYPE(Versal, XLNX_VERSAL) @@ -XXX,XX +XXX,XX @@ OBJECT_DECLARE_SIMPLE_TYPE(Versal, XLNX_VERSAL) #define XLNX_VERSAL_NR_SDS 2 #define XLNX_VERSAL_NR_XRAM 4 #define XLNX_VERSAL_NR_IRQS 192 +#define XLNX_VERSAL_NR_CANFD 2 +#define XLNX_VERSAL_CANFD_REF_CLK (24 * 1000 * 1000) struct Versal { /*< private >*/ @@ -XXX,XX +XXX,XX @@ struct Versal { CadenceGEMState gem[XLNX_VERSAL_NR_GEMS]; XlnxZDMA adma[XLNX_VERSAL_NR_ADMAS]; VersalUsb2 usb; + CanBusState *canbus[XLNX_VERSAL_NR_CANFD]; + XlnxVersalCANFDState canfd[XLNX_VERSAL_NR_CANFD]; } iou; /* Real-time Processing Unit. */ @@ -XXX,XX +XXX,XX @@ struct Versal { #define VERSAL_CRL_IRQ 10 #define VERSAL_UART0_IRQ_0 18 #define VERSAL_UART1_IRQ_0 19 +#define VERSAL_CANFD0_IRQ_0 20 +#define VERSAL_CANFD1_IRQ_0 21 #define VERSAL_USB0_IRQ_0 22 #define VERSAL_GEM0_IRQ_0 56 #define VERSAL_GEM0_WAKE_IRQ_0 57 @@ -XXX,XX +XXX,XX @@ struct Versal { #define MM_UART1 0xff010000U #define MM_UART1_SIZE 0x10000 +#define MM_CANFD0 0xff060000U +#define MM_CANFD0_SIZE 0x10000 +#define MM_CANFD1 0xff070000U +#define MM_CANFD1_SIZE 0x10000 + #define MM_GEM0 0xff0c0000U #define MM_GEM0_SIZE 0x10000 #define MM_GEM1 0xff0d0000U diff --git a/hw/arm/xlnx-versal-virt.c b/hw/arm/xlnx-versal-virt.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-versal-virt.c +++ b/hw/arm/xlnx-versal-virt.c @@ -XXX,XX +XXX,XX @@ struct VersalVirt { uint32_t clk_25Mhz; uint32_t usb; uint32_t dwc; + uint32_t canfd[2]; } phandle; struct arm_boot_info binfo; + CanBusState *canbus[XLNX_VERSAL_NR_CANFD]; struct { bool secure; } cfg; @@ -XXX,XX +XXX,XX @@ static void fdt_add_uart_nodes(VersalVirt *s) } } +static void fdt_add_canfd_nodes(VersalVirt *s) +{ + uint64_t addrs[] = { MM_CANFD1, MM_CANFD0 }; + uint32_t size[] = { MM_CANFD1_SIZE, MM_CANFD0_SIZE }; + unsigned int irqs[] = { VERSAL_CANFD1_IRQ_0, VERSAL_CANFD0_IRQ_0 }; + const char clocknames[] = "can_clk\0s_axi_aclk"; + int i; + + /* Create and connect CANFD0 and CANFD1 nodes to canbus0. */ + for (i = 0; i < ARRAY_SIZE(addrs); i++) { + char *name = g_strdup_printf("/canfd@%" PRIx64, addrs[i]); + qemu_fdt_add_subnode(s->fdt, name); + + qemu_fdt_setprop_cell(s->fdt, name, "rx-fifo-depth", 0x40); + qemu_fdt_setprop_cell(s->fdt, name, "tx-mailbox-count", 0x20); + + qemu_fdt_setprop_cells(s->fdt, name, "clocks", + s->phandle.clk_25Mhz, s->phandle.clk_25Mhz); + qemu_fdt_setprop(s->fdt, name, "clock-names", + clocknames, sizeof(clocknames)); + qemu_fdt_setprop_cells(s->fdt, name, "interrupts", + GIC_FDT_IRQ_TYPE_SPI, irqs[i], + GIC_FDT_IRQ_FLAGS_LEVEL_HI); + qemu_fdt_setprop_sized_cells(s->fdt, name, "reg", + 2, addrs[i], 2, size[i]); + qemu_fdt_setprop_string(s->fdt, name, "compatible", + "xlnx,canfd-2.0"); + + g_free(name); + } +} + static void fdt_add_fixed_link_nodes(VersalVirt *s, char *gemname, uint32_t phandle) { @@ -XXX,XX +XXX,XX @@ static void versal_virt_init(MachineState *machine) TYPE_XLNX_VERSAL); object_property_set_link(OBJECT(&s->soc), "ddr", OBJECT(machine->ram), &error_abort); + object_property_set_link(OBJECT(&s->soc), "canbus0", OBJECT(s->canbus[0]), + &error_abort); + object_property_set_link(OBJECT(&s->soc), "canbus1", OBJECT(s->canbus[1]), + &error_abort); sysbus_realize(SYS_BUS_DEVICE(&s->soc), &error_fatal); fdt_create(s); create_virtio_regions(s); fdt_add_gem_nodes(s); fdt_add_uart_nodes(s); + fdt_add_canfd_nodes(s); fdt_add_gic_nodes(s); fdt_add_timer_nodes(s); fdt_add_zdma_nodes(s); @@ -XXX,XX +XXX,XX @@ static void versal_virt_init(MachineState *machine) static void versal_virt_machine_instance_init(Object *obj) { + VersalVirt *s = XLNX_VERSAL_VIRT_MACHINE(obj); + + /* + * User can set canbus0 and canbus1 properties to can-bus object and connect + * to socketcan(optional) interface via command line. + */ + object_property_add_link(obj, "canbus0", TYPE_CAN_BUS, + (Object **)&s->canbus[0], + object_property_allow_set_link, + 0); + object_property_add_link(obj, "canbus1", TYPE_CAN_BUS, + (Object **)&s->canbus[1], + object_property_allow_set_link, + 0); } static void versal_virt_machine_class_init(ObjectClass *oc, void *data) diff --git a/hw/arm/xlnx-versal.c b/hw/arm/xlnx-versal.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-versal.c +++ b/hw/arm/xlnx-versal.c @@ -XXX,XX +XXX,XX @@ static void versal_create_uarts(Versal *s, qemu_irq *pic) } } +static void versal_create_canfds(Versal *s, qemu_irq *pic) +{ + int i; + uint32_t irqs[] = { VERSAL_CANFD0_IRQ_0, VERSAL_CANFD1_IRQ_0}; + uint64_t addrs[] = { MM_CANFD0, MM_CANFD1 }; + + for (i = 0; i < ARRAY_SIZE(s->lpd.iou.canfd); i++) { + char *name = g_strdup_printf("canfd%d", i); + SysBusDevice *sbd; + MemoryRegion *mr; + + object_initialize_child(OBJECT(s), name, &s->lpd.iou.canfd[i], + TYPE_XILINX_CANFD); + sbd = SYS_BUS_DEVICE(&s->lpd.iou.canfd[i]); + + object_property_set_int(OBJECT(&s->lpd.iou.canfd[i]), "ext_clk_freq", + XLNX_VERSAL_CANFD_REF_CLK , &error_abort); + + object_property_set_link(OBJECT(&s->lpd.iou.canfd[i]), "canfdbus", + OBJECT(s->lpd.iou.canbus[i]), + &error_abort); + + sysbus_realize(sbd, &error_fatal); + + mr = sysbus_mmio_get_region(sbd, 0); + memory_region_add_subregion(&s->mr_ps, addrs[i], mr); + + sysbus_connect_irq(sbd, 0, pic[irqs[i]]); + g_free(name); + } +} + static void versal_create_usbs(Versal *s, qemu_irq *pic) { DeviceState *dev; @@ -XXX,XX +XXX,XX @@ static void versal_realize(DeviceState *dev, Error **errp) versal_create_apu_gic(s, pic); versal_create_rpu_cpus(s); versal_create_uarts(s, pic); + versal_create_canfds(s, pic); versal_create_usbs(s, pic); versal_create_gems(s, pic); versal_create_admas(s, pic); @@ -XXX,XX +XXX,XX @@ static void versal_init(Object *obj) static Property versal_properties[] = { DEFINE_PROP_LINK("ddr", Versal, cfg.mr_ddr, TYPE_MEMORY_REGION, MemoryRegion *), + DEFINE_PROP_LINK("canbus0", Versal, lpd.iou.canbus[0], + TYPE_CAN_BUS, CanBusState *), + DEFINE_PROP_LINK("canbus1", Versal, lpd.iou.canbus[1], + TYPE_CAN_BUS, CanBusState *), DEFINE_PROP_END_OF_LIST() }; -- 2.34.1 From: Vikram Garhwal <vikram.garhwal@amd.com> Signed-off-by: Vikram Garhwal <vikram.garhwal@amd.com> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Francisco Iglesias <frasse.iglesias@gmail.com> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- MAINTAINERS | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/MAINTAINERS b/MAINTAINERS index XXXXXXX..XXXXXXX 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -XXX,XX +XXX,XX @@ M: Francisco Iglesias <francisco.iglesias@amd.com> S: Maintained F: hw/net/can/xlnx-* F: include/hw/net/xlnx-* -F: tests/qtest/xlnx-can-test* +F: tests/qtest/xlnx-can*-test* EDU M: Jiri Slaby <jslaby@suse.cz> -- 2.34.1 From: Vikram Garhwal <vikram.garhwal@amd.com> The QTests perform three tests on the Xilinx VERSAL CANFD controller: Tests the CANFD controllers in loopback. Tests the CANFD controllers in normal mode with CAN frame. Tests the CANFD controllers in normal mode with CANFD frame. Signed-off-by: Vikram Garhwal <vikram.garhwal@amd.com> Acked-by: Thomas Huth <thuth@redhat.com> Reviewed-by: Francisco Iglesias <francisco.iglesias@amd.com> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- tests/qtest/xlnx-canfd-test.c | 423 ++++++++++++++++++++++++++++++++++ tests/qtest/meson.build | 1 + 2 files changed, 424 insertions(+) create mode 100644 tests/qtest/xlnx-canfd-test.c diff --git a/tests/qtest/xlnx-canfd-test.c b/tests/qtest/xlnx-canfd-test.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/tests/qtest/xlnx-canfd-test.c @@ -XXX,XX +XXX,XX @@ +/* + * SPDX-License-Identifier: MIT + * + * QTests for the Xilinx Versal CANFD controller. + * + * Copyright (c) 2022 AMD Inc. + * + * Written-by: Vikram Garhwal<vikram.garhwal@amd.com> + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN + * THE SOFTWARE. + */ + +#include "qemu/osdep.h" +#include "libqtest.h" + +/* Base address. */ +#define CANFD0_BASE_ADDR 0xff060000 +#define CANFD1_BASE_ADDR 0xff070000 + +/* Register addresses. */ +#define R_SRR_OFFSET 0x00 +#define R_MSR_OFFSET 0x04 +#define R_FILTER_CONTROL_REGISTER 0xe0 +#define R_SR_OFFSET 0x18 +#define R_ISR_OFFSET 0x1c +#define R_IER_OFFSET 0x20 +#define R_ICR_OFFSET 0x24 +#define R_TX_READY_REQ_REGISTER 0x90 +#define RX_FIFO_STATUS_REGISTER 0xe8 +#define R_TXID_OFFSET 0x100 +#define R_TXDLC_OFFSET 0x104 +#define R_TXDATA1_OFFSET 0x108 +#define R_TXDATA2_OFFSET 0x10c +#define R_AFMR_REGISTER0 0xa00 +#define R_AFIR_REGISTER0 0xa04 +#define R_RX0_ID_OFFSET 0x2100 +#define R_RX0_DLC_OFFSET 0x2104 +#define R_RX0_DATA1_OFFSET 0x2108 +#define R_RX0_DATA2_OFFSET 0x210c + +/* CANFD modes. */ +#define SRR_CONFIG_MODE 0x00 +#define MSR_NORMAL_MODE 0x00 +#define MSR_LOOPBACK_MODE (1 << 1) +#define ENABLE_CANFD (1 << 1) + +/* CANFD status. */ +#define STATUS_CONFIG_MODE (1 << 0) +#define STATUS_NORMAL_MODE (1 << 3) +#define STATUS_LOOPBACK_MODE (1 << 1) +#define ISR_TXOK (1 << 1) +#define ISR_RXOK (1 << 4) + +#define ENABLE_ALL_FILTERS 0xffffffff +#define ENABLE_ALL_INTERRUPTS 0xffffffff + +/* We are sending one canfd message. */ +#define TX_READY_REG_VAL 0x1 + +#define FIRST_RX_STORE_INDEX 0x1 +#define STATUS_REG_MASK 0xf +#define DLC_FD_BIT_SHIFT 0x1b +#define DLC_FD_BIT_MASK 0xf8000000 +#define FIFO_STATUS_READ_INDEX_MASK 0x3f +#define FIFO_STATUS_FILL_LEVEL_MASK 0x7f00 +#define FILL_LEVEL_SHIFT 0x8 + +/* CANFD frame size ID, DLC and 16 DATA word. */ +#define CANFD_FRAME_SIZE 18 +/* CAN frame size ID, DLC and 2 DATA word. */ +#define CAN_FRAME_SIZE 4 + +/* Set the filters for CANFD controller. */ +static void enable_filters(QTestState *qts) +{ + const uint32_t arr_afmr[32] = { 0xb423deaa, 0xa2a40bdc, 0x1b64f486, + 0x95c0d4ee, 0xe0c44528, 0x4b407904, + 0xd2673f46, 0x9fc638d6, 0x8844f3d8, + 0xa607d1e8, 0x67871bf4, 0xc2557dc, + 0x9ea5b53e, 0x3643c0cc, 0x5a05ea8e, + 0x83a46d84, 0x4a25c2b8, 0x93a66008, + 0x2e467470, 0xedc66118, 0x9086f9f2, + 0xfa23dd36, 0xb6654b90, 0xb221b8ca, + 0x3467d1e2, 0xa3a55542, 0x5b26a012, + 0x2281ea7e, 0xcea0ece8, 0xdc61e588, + 0x2e5676a, 0x16821320 }; + + const uint32_t arr_afir[32] = { 0xa833dfa1, 0x255a477e, 0x3a4bb1c5, + 0x8f560a6c, 0x27f38903, 0x2fecec4d, + 0xa014c66d, 0xec289b8, 0x7e52dead, + 0x82e94f3c, 0xcf3e3c5c, 0x66059871, + 0x3f213df4, 0x25ac3959, 0xa12e9bef, + 0xa3ad3af, 0xbafd7fe, 0xb3cb40fd, + 0x5d9caa81, 0x2ed61902, 0x7cd64a0, + 0x4b1fa538, 0x9b5ced8c, 0x150de059, + 0xd2794227, 0x635e820a, 0xbb6b02cf, + 0xbb58176, 0x570025bb, 0xa78d9658, + 0x49d735df, 0xe5399d2f }; + + /* Passing the respective array values to all the AFMR and AFIR pairs. */ + for (int i = 0; i < 32; i++) { + /* For CANFD0. */ + qtest_writel(qts, CANFD0_BASE_ADDR + R_AFMR_REGISTER0 + 8 * i, + arr_afmr[i]); + qtest_writel(qts, CANFD0_BASE_ADDR + R_AFIR_REGISTER0 + 8 * i, + arr_afir[i]); + + /* For CANFD1. */ + qtest_writel(qts, CANFD1_BASE_ADDR + R_AFMR_REGISTER0 + 8 * i, + arr_afmr[i]); + qtest_writel(qts, CANFD1_BASE_ADDR + R_AFIR_REGISTER0 + 8 * i, + arr_afir[i]); + } + + /* Enable all the pairs from AFR register. */ + qtest_writel(qts, CANFD0_BASE_ADDR + R_FILTER_CONTROL_REGISTER, + ENABLE_ALL_FILTERS); + qtest_writel(qts, CANFD1_BASE_ADDR + R_FILTER_CONTROL_REGISTER, + ENABLE_ALL_FILTERS); +} + +static void configure_canfd(QTestState *qts, uint8_t mode) +{ + uint32_t status = 0; + + /* Put CANFD0 and CANFD1 in config mode. */ + qtest_writel(qts, CANFD0_BASE_ADDR + R_SRR_OFFSET, SRR_CONFIG_MODE); + qtest_writel(qts, CANFD1_BASE_ADDR + R_SRR_OFFSET, SRR_CONFIG_MODE); + + /* Write mode of operation in Mode select register. */ + qtest_writel(qts, CANFD0_BASE_ADDR + R_MSR_OFFSET, mode); + qtest_writel(qts, CANFD1_BASE_ADDR + R_MSR_OFFSET, mode); + + enable_filters(qts); + + /* Check here if CANFD0 and CANFD1 are in config mode. */ + status = qtest_readl(qts, CANFD0_BASE_ADDR + R_SR_OFFSET); + status = status & STATUS_REG_MASK; + g_assert_cmpint(status, ==, STATUS_CONFIG_MODE); + + status = qtest_readl(qts, CANFD1_BASE_ADDR + R_SR_OFFSET); + status = status & STATUS_REG_MASK; + g_assert_cmpint(status, ==, STATUS_CONFIG_MODE); + + qtest_writel(qts, CANFD1_BASE_ADDR + R_IER_OFFSET, ENABLE_ALL_INTERRUPTS); + qtest_writel(qts, CANFD1_BASE_ADDR + R_IER_OFFSET, ENABLE_ALL_INTERRUPTS); + + qtest_writel(qts, CANFD0_BASE_ADDR + R_SRR_OFFSET, ENABLE_CANFD); + qtest_writel(qts, CANFD1_BASE_ADDR + R_SRR_OFFSET, ENABLE_CANFD); +} + +static void generate_random_data(uint32_t *buf_tx, bool is_canfd_frame) +{ + /* Generate random TX data for CANFD frame. */ + if (is_canfd_frame) { + for (int i = 0; i < CANFD_FRAME_SIZE - 2; i++) { + buf_tx[2 + i] = rand(); + } + } else { + /* Generate random TX data for CAN frame. */ + for (int i = 0; i < CAN_FRAME_SIZE - 2; i++) { + buf_tx[2 + i] = rand(); + } + } +} + +static void read_data(QTestState *qts, uint64_t can_base_addr, uint32_t *buf_rx) +{ + uint32_t int_status; + uint32_t fifo_status_reg_value; + /* At which RX FIFO the received data is stored. */ + uint8_t store_ind = 0; + bool is_canfd_frame = false; + + /* Read the interrupt on CANFD rx. */ + int_status = qtest_readl(qts, can_base_addr + R_ISR_OFFSET) & ISR_RXOK; + + g_assert_cmpint(int_status, ==, ISR_RXOK); + + /* Find the fill level and read index. */ + fifo_status_reg_value = qtest_readl(qts, can_base_addr + + RX_FIFO_STATUS_REGISTER); + + store_ind = (fifo_status_reg_value & FIFO_STATUS_READ_INDEX_MASK) + + ((fifo_status_reg_value & FIFO_STATUS_FILL_LEVEL_MASK) >> + FILL_LEVEL_SHIFT); + + g_assert_cmpint(store_ind, ==, FIRST_RX_STORE_INDEX); + + /* Read the RX register data for CANFD. */ + buf_rx[0] = qtest_readl(qts, can_base_addr + R_RX0_ID_OFFSET); + buf_rx[1] = qtest_readl(qts, can_base_addr + R_RX0_DLC_OFFSET); + + is_canfd_frame = (buf_rx[1] >> DLC_FD_BIT_SHIFT) & 1; + + if (is_canfd_frame) { + for (int i = 0; i < CANFD_FRAME_SIZE - 2; i++) { + buf_rx[i + 2] = qtest_readl(qts, + can_base_addr + R_RX0_DATA1_OFFSET + 4 * i); + } + } else { + buf_rx[2] = qtest_readl(qts, can_base_addr + R_RX0_DATA1_OFFSET); + buf_rx[3] = qtest_readl(qts, can_base_addr + R_RX0_DATA2_OFFSET); + } + + /* Clear the RX interrupt. */ + qtest_writel(qts, CANFD1_BASE_ADDR + R_ICR_OFFSET, ISR_RXOK); +} + +static void write_data(QTestState *qts, uint64_t can_base_addr, + const uint32_t *buf_tx, bool is_canfd_frame) +{ + /* Write the TX register data for CANFD. */ + qtest_writel(qts, can_base_addr + R_TXID_OFFSET, buf_tx[0]); + qtest_writel(qts, can_base_addr + R_TXDLC_OFFSET, buf_tx[1]); + + if (is_canfd_frame) { + for (int i = 0; i < CANFD_FRAME_SIZE - 2; i++) { + qtest_writel(qts, can_base_addr + R_TXDATA1_OFFSET + 4 * i, + buf_tx[2 + i]); + } + } else { + qtest_writel(qts, can_base_addr + R_TXDATA1_OFFSET, buf_tx[2]); + qtest_writel(qts, can_base_addr + R_TXDATA2_OFFSET, buf_tx[3]); + } +} + +static void send_data(QTestState *qts, uint64_t can_base_addr) +{ + uint32_t int_status; + + qtest_writel(qts, can_base_addr + R_TX_READY_REQ_REGISTER, + TX_READY_REG_VAL); + + /* Read the interrupt on CANFD for tx. */ + int_status = qtest_readl(qts, can_base_addr + R_ISR_OFFSET) & ISR_TXOK; + + g_assert_cmpint(int_status, ==, ISR_TXOK); + + /* Clear the interrupt for tx. */ + qtest_writel(qts, CANFD0_BASE_ADDR + R_ICR_OFFSET, ISR_TXOK); +} + +static void match_rx_tx_data(const uint32_t *buf_tx, const uint32_t *buf_rx, + bool is_canfd_frame) +{ + uint16_t size = 0; + uint8_t len = CAN_FRAME_SIZE; + + if (is_canfd_frame) { + len = CANFD_FRAME_SIZE; + } + + while (size < len) { + if (R_RX0_ID_OFFSET + 4 * size == R_RX0_DLC_OFFSET) { + g_assert_cmpint((buf_rx[size] & DLC_FD_BIT_MASK), ==, + (buf_tx[size] & DLC_FD_BIT_MASK)); + } else { + if (!is_canfd_frame && size == 4) { + break; + } + + g_assert_cmpint(buf_rx[size], ==, buf_tx[size]); + } + + size++; + } +} +/* + * Xilinx CANFD supports both CAN and CANFD frames. This test will be + * transferring CAN frame i.e. 8 bytes of data from CANFD0 and CANFD1 through + * canbus. CANFD0 initiate the data transfer to can-bus, CANFD1 receives the + * data. Test compares the can frame data sent from CANFD0 and received on + * CANFD1. + */ +static void test_can_data_transfer(void) +{ + uint32_t buf_tx[CAN_FRAME_SIZE] = { 0x5a5bb9a4, 0x80000000, + 0x12345678, 0x87654321 }; + uint32_t buf_rx[CAN_FRAME_SIZE] = { 0x00, 0x00, 0x00, 0x00 }; + uint32_t status = 0; + + generate_random_data(buf_tx, false); + + QTestState *qts = qtest_init("-machine xlnx-versal-virt" + " -object can-bus,id=canbus" + " -machine canbus0=canbus" + " -machine canbus1=canbus" + ); + + configure_canfd(qts, MSR_NORMAL_MODE); + + /* Check if CANFD0 and CANFD1 are in Normal mode. */ + status = qtest_readl(qts, CANFD0_BASE_ADDR + R_SR_OFFSET); + status = status & STATUS_REG_MASK; + g_assert_cmpint(status, ==, STATUS_NORMAL_MODE); + + status = qtest_readl(qts, CANFD1_BASE_ADDR + R_SR_OFFSET); + status = status & STATUS_REG_MASK; + g_assert_cmpint(status, ==, STATUS_NORMAL_MODE); + + write_data(qts, CANFD0_BASE_ADDR, buf_tx, false); + + send_data(qts, CANFD0_BASE_ADDR); + read_data(qts, CANFD1_BASE_ADDR, buf_rx); + match_rx_tx_data(buf_tx, buf_rx, false); + + qtest_quit(qts); +} + +/* + * This test will be transferring CANFD frame i.e. 64 bytes of data from CANFD0 + * and CANFD1 through canbus. CANFD0 initiate the data transfer to can-bus, + * CANFD1 receives the data. Test compares the CANFD frame data sent from CANFD0 + * with received on CANFD1. + */ +static void test_canfd_data_transfer(void) +{ + uint32_t buf_tx[CANFD_FRAME_SIZE] = { 0x5a5bb9a4, 0xf8000000 }; + uint32_t buf_rx[CANFD_FRAME_SIZE] = { 0x00, 0x00, 0x00, 0x00 }; + uint32_t status = 0; + + generate_random_data(buf_tx, true); + + QTestState *qts = qtest_init("-machine xlnx-versal-virt" + " -object can-bus,id=canbus" + " -machine canbus0=canbus" + " -machine canbus1=canbus" + ); + + configure_canfd(qts, MSR_NORMAL_MODE); + + /* Check if CANFD0 and CANFD1 are in Normal mode. */ + status = qtest_readl(qts, CANFD0_BASE_ADDR + R_SR_OFFSET); + status = status & STATUS_REG_MASK; + g_assert_cmpint(status, ==, STATUS_NORMAL_MODE); + + status = qtest_readl(qts, CANFD1_BASE_ADDR + R_SR_OFFSET); + status = status & STATUS_REG_MASK; + g_assert_cmpint(status, ==, STATUS_NORMAL_MODE); + + write_data(qts, CANFD0_BASE_ADDR, buf_tx, true); + + send_data(qts, CANFD0_BASE_ADDR); + read_data(qts, CANFD1_BASE_ADDR, buf_rx); + match_rx_tx_data(buf_tx, buf_rx, true); + + qtest_quit(qts); +} + +/* + * This test is performing loopback mode on CANFD0 and CANFD1. Data sent from + * TX of each CANFD0 and CANFD1 are compared with RX register data for + * respective CANFD Controller. + */ +static void test_can_loopback(void) +{ + uint32_t buf_tx[CANFD_FRAME_SIZE] = { 0x5a5bb9a4, 0xf8000000 }; + uint32_t buf_rx[CANFD_FRAME_SIZE] = { 0x00, 0x00, 0x00, 0x00 }; + uint32_t status = 0; + + generate_random_data(buf_tx, true); + + QTestState *qts = qtest_init("-machine xlnx-versal-virt" + " -object can-bus,id=canbus" + " -machine canbus0=canbus" + " -machine canbus1=canbus" + ); + + configure_canfd(qts, MSR_LOOPBACK_MODE); + + /* Check if CANFD0 and CANFD1 are set in correct loopback mode. */ + status = qtest_readl(qts, CANFD0_BASE_ADDR + R_SR_OFFSET); + status = status & STATUS_REG_MASK; + g_assert_cmpint(status, ==, STATUS_LOOPBACK_MODE); + + status = qtest_readl(qts, CANFD1_BASE_ADDR + R_SR_OFFSET); + status = status & STATUS_REG_MASK; + g_assert_cmpint(status, ==, STATUS_LOOPBACK_MODE); + + write_data(qts, CANFD0_BASE_ADDR, buf_tx, true); + + send_data(qts, CANFD0_BASE_ADDR); + read_data(qts, CANFD0_BASE_ADDR, buf_rx); + match_rx_tx_data(buf_tx, buf_rx, true); + + generate_random_data(buf_tx, true); + + write_data(qts, CANFD1_BASE_ADDR, buf_tx, true); + + send_data(qts, CANFD1_BASE_ADDR); + read_data(qts, CANFD1_BASE_ADDR, buf_rx); + match_rx_tx_data(buf_tx, buf_rx, true); + + qtest_quit(qts); +} + +int main(int argc, char **argv) +{ + g_test_init(&argc, &argv, NULL); + + qtest_add_func("/net/canfd/can_data_transfer", test_can_data_transfer); + qtest_add_func("/net/canfd/canfd_data_transfer", test_canfd_data_transfer); + qtest_add_func("/net/canfd/can_loopback", test_can_loopback); + + return g_test_run(); +} diff --git a/tests/qtest/meson.build b/tests/qtest/meson.build index XXXXXXX..XXXXXXX 100644 --- a/tests/qtest/meson.build +++ b/tests/qtest/meson.build @@ -XXX,XX +XXX,XX @@ qtests_aarch64 = \ (config_all.has_key('CONFIG_TCG') and config_all_devices.has_key('CONFIG_TPM_TIS_SYSBUS') ? \ ['tpm-tis-device-test', 'tpm-tis-device-swtpm-test'] : []) + \ (config_all_devices.has_key('CONFIG_XLNX_ZYNQMP_ARM') ? ['xlnx-can-test', 'fuzz-xlnx-dp-test'] : []) + \ + (config_all_devices.has_key('CONFIG_XLNX_VERSAL') ? ['xlnx-canfd-test'] : []) + \ (config_all_devices.has_key('CONFIG_RASPI') ? ['bcm2835-dma-test'] : []) + \ (config_all.has_key('CONFIG_TCG') and \ config_all_devices.has_key('CONFIG_TPM_TIS_I2C') ? ['tpm-tis-i2c-test'] : []) + \ -- 2.34.1 From: qianfan Zhao <qianfanguijin@163.com> Allwinner R40 (sun8i) SoC features a Quad-Core Cortex-A7 ARM CPU, and a Mali400 MP2 GPU from ARM. It's also known as the Allwinner T3 for In-Car Entertainment usage, A40i and A40pro are variants that differ in applicable temperatures range (industrial and military). Signed-off-by: qianfan Zhao <qianfanguijin@163.com> Reviewed-by: Niek Linnenbank <nieklinnenbank@gmail.com> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/allwinner-r40.h | 110 +++++++++ hw/arm/allwinner-r40.c | 415 +++++++++++++++++++++++++++++++++ hw/arm/bananapi_m2u.c | 129 ++++++++++ hw/arm/Kconfig | 10 + hw/arm/meson.build | 1 + 5 files changed, 665 insertions(+) create mode 100644 include/hw/arm/allwinner-r40.h create mode 100644 hw/arm/allwinner-r40.c create mode 100644 hw/arm/bananapi_m2u.c diff --git a/include/hw/arm/allwinner-r40.h b/include/hw/arm/allwinner-r40.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/include/hw/arm/allwinner-r40.h @@ -XXX,XX +XXX,XX @@ +/* + * Allwinner R40/A40i/T3 System on Chip emulation + * + * Copyright (C) 2023 qianfan Zhao <qianfanguijin@163.com> + * + * This program is free software: you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation, either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#ifndef HW_ARM_ALLWINNER_R40_H +#define HW_ARM_ALLWINNER_R40_H + +#include "qom/object.h" +#include "hw/arm/boot.h" +#include "hw/timer/allwinner-a10-pit.h" +#include "hw/intc/arm_gic.h" +#include "hw/sd/allwinner-sdhost.h" +#include "target/arm/cpu.h" +#include "sysemu/block-backend.h" + +enum { + AW_R40_DEV_SRAM_A1, + AW_R40_DEV_SRAM_A2, + AW_R40_DEV_SRAM_A3, + AW_R40_DEV_SRAM_A4, + AW_R40_DEV_MMC0, + AW_R40_DEV_MMC1, + AW_R40_DEV_MMC2, + AW_R40_DEV_MMC3, + AW_R40_DEV_CCU, + AW_R40_DEV_PIT, + AW_R40_DEV_UART0, + AW_R40_DEV_GIC_DIST, + AW_R40_DEV_GIC_CPU, + AW_R40_DEV_GIC_HYP, + AW_R40_DEV_GIC_VCPU, + AW_R40_DEV_SDRAM +}; + +#define AW_R40_NUM_CPUS (4) + +/** + * Allwinner R40 object model + * @{ + */ + +/** Object type for the Allwinner R40 SoC */ +#define TYPE_AW_R40 "allwinner-r40" + +/** Convert input object to Allwinner R40 state object */ +OBJECT_DECLARE_SIMPLE_TYPE(AwR40State, AW_R40) + +/** @} */ + +/** + * Allwinner R40 object + * + * This struct contains the state of all the devices + * which are currently emulated by the R40 SoC code. + */ +#define AW_R40_NUM_MMCS 4 + +struct AwR40State { + /*< private >*/ + DeviceState parent_obj; + /*< public >*/ + + ARMCPU cpus[AW_R40_NUM_CPUS]; + const hwaddr *memmap; + AwA10PITState timer; + AwSdHostState mmc[AW_R40_NUM_MMCS]; + GICState gic; + MemoryRegion sram_a1; + MemoryRegion sram_a2; + MemoryRegion sram_a3; + MemoryRegion sram_a4; +}; + +/** + * Emulate Boot ROM firmware setup functionality. + * + * A real Allwinner R40 SoC contains a Boot ROM + * which is the first code that runs right after + * the SoC is powered on. The Boot ROM is responsible + * for loading user code (e.g. a bootloader) from any + * of the supported external devices and writing the + * downloaded code to internal SRAM. After loading the SoC + * begins executing the code written to SRAM. + * + * This function emulates the Boot ROM by copying 32 KiB + * of data from the given block device and writes it to + * the start of the first internal SRAM memory. + * + * @s: Allwinner R40 state object pointer + * @blk: Block backend device object pointer + * @unit: the mmc control's unit + */ +bool allwinner_r40_bootrom_setup(AwR40State *s, BlockBackend *blk, int unit); + +#endif /* HW_ARM_ALLWINNER_R40_H */ diff --git a/hw/arm/allwinner-r40.c b/hw/arm/allwinner-r40.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/arm/allwinner-r40.c @@ -XXX,XX +XXX,XX @@ +/* + * Allwinner R40/A40i/T3 System on Chip emulation + * + * Copyright (C) 2023 qianfan Zhao <qianfanguijin@163.com> + * + * This program is free software: you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation, either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#include "qemu/osdep.h" +#include "qapi/error.h" +#include "qemu/error-report.h" +#include "qemu/bswap.h" +#include "qemu/module.h" +#include "qemu/units.h" +#include "hw/qdev-core.h" +#include "hw/sysbus.h" +#include "hw/char/serial.h" +#include "hw/misc/unimp.h" +#include "hw/usb/hcd-ehci.h" +#include "hw/loader.h" +#include "sysemu/sysemu.h" +#include "hw/arm/allwinner-r40.h" + +/* Memory map */ +const hwaddr allwinner_r40_memmap[] = { + [AW_R40_DEV_SRAM_A1] = 0x00000000, + [AW_R40_DEV_SRAM_A2] = 0x00004000, + [AW_R40_DEV_SRAM_A3] = 0x00008000, + [AW_R40_DEV_SRAM_A4] = 0x0000b400, + [AW_R40_DEV_MMC0] = 0x01c0f000, + [AW_R40_DEV_MMC1] = 0x01c10000, + [AW_R40_DEV_MMC2] = 0x01c11000, + [AW_R40_DEV_MMC3] = 0x01c12000, + [AW_R40_DEV_PIT] = 0x01c20c00, + [AW_R40_DEV_UART0] = 0x01c28000, + [AW_R40_DEV_GIC_DIST] = 0x01c81000, + [AW_R40_DEV_GIC_CPU] = 0x01c82000, + [AW_R40_DEV_GIC_HYP] = 0x01c84000, + [AW_R40_DEV_GIC_VCPU] = 0x01c86000, + [AW_R40_DEV_SDRAM] = 0x40000000 +}; + +/* List of unimplemented devices */ +struct AwR40Unimplemented { + const char *device_name; + hwaddr base; + hwaddr size; +}; + +static struct AwR40Unimplemented r40_unimplemented[] = { + { "d-engine", 0x01000000, 4 * MiB }, + { "d-inter", 0x01400000, 128 * KiB }, + { "sram-c", 0x01c00000, 4 * KiB }, + { "dma", 0x01c02000, 4 * KiB }, + { "nfdc", 0x01c03000, 4 * KiB }, + { "ts", 0x01c04000, 4 * KiB }, + { "spi0", 0x01c05000, 4 * KiB }, + { "spi1", 0x01c06000, 4 * KiB }, + { "cs0", 0x01c09000, 4 * KiB }, + { "keymem", 0x01c0a000, 4 * KiB }, + { "emac", 0x01c0b000, 4 * KiB }, + { "usb0-otg", 0x01c13000, 4 * KiB }, + { "usb0-host", 0x01c14000, 4 * KiB }, + { "crypto", 0x01c15000, 4 * KiB }, + { "spi2", 0x01c17000, 4 * KiB }, + { "sata", 0x01c18000, 4 * KiB }, + { "usb1-host", 0x01c19000, 4 * KiB }, + { "sid", 0x01c1b000, 4 * KiB }, + { "usb2-host", 0x01c1c000, 4 * KiB }, + { "cs1", 0x01c1d000, 4 * KiB }, + { "spi3", 0x01c1f000, 4 * KiB }, + { "ccu", 0x01c20000, 1 * KiB }, + { "rtc", 0x01c20400, 1 * KiB }, + { "pio", 0x01c20800, 1 * KiB }, + { "owa", 0x01c21000, 1 * KiB }, + { "ac97", 0x01c21400, 1 * KiB }, + { "cir0", 0x01c21800, 1 * KiB }, + { "cir1", 0x01c21c00, 1 * KiB }, + { "pcm0", 0x01c22000, 1 * KiB }, + { "pcm1", 0x01c22400, 1 * KiB }, + { "pcm2", 0x01c22800, 1 * KiB }, + { "audio", 0x01c22c00, 1 * KiB }, + { "keypad", 0x01c23000, 1 * KiB }, + { "pwm", 0x01c23400, 1 * KiB }, + { "keyadc", 0x01c24400, 1 * KiB }, + { "ths", 0x01c24c00, 1 * KiB }, + { "rtp", 0x01c25000, 1 * KiB }, + { "pmu", 0x01c25400, 1 * KiB }, + { "cpu-cfg", 0x01c25c00, 1 * KiB }, + { "uart0", 0x01c28000, 1 * KiB }, + { "uart1", 0x01c28400, 1 * KiB }, + { "uart2", 0x01c28800, 1 * KiB }, + { "uart3", 0x01c28c00, 1 * KiB }, + { "uart4", 0x01c29000, 1 * KiB }, + { "uart5", 0x01c29400, 1 * KiB }, + { "uart6", 0x01c29800, 1 * KiB }, + { "uart7", 0x01c29c00, 1 * KiB }, + { "ps20", 0x01c2a000, 1 * KiB }, + { "ps21", 0x01c2a400, 1 * KiB }, + { "twi0", 0x01c2ac00, 1 * KiB }, + { "twi1", 0x01c2b000, 1 * KiB }, + { "twi2", 0x01c2b400, 1 * KiB }, + { "twi3", 0x01c2b800, 1 * KiB }, + { "twi4", 0x01c2c000, 1 * KiB }, + { "scr", 0x01c2c400, 1 * KiB }, + { "tvd-top", 0x01c30000, 4 * KiB }, + { "tvd0", 0x01c31000, 4 * KiB }, + { "tvd1", 0x01c32000, 4 * KiB }, + { "tvd2", 0x01c33000, 4 * KiB }, + { "tvd3", 0x01c34000, 4 * KiB }, + { "gpu", 0x01c40000, 64 * KiB }, + { "gmac", 0x01c50000, 64 * KiB }, + { "hstmr", 0x01c60000, 4 * KiB }, + { "dram-com", 0x01c62000, 4 * KiB }, + { "dram-ctl", 0x01c63000, 4 * KiB }, + { "tcon-top", 0x01c70000, 4 * KiB }, + { "lcd0", 0x01c71000, 4 * KiB }, + { "lcd1", 0x01c72000, 4 * KiB }, + { "tv0", 0x01c73000, 4 * KiB }, + { "tv1", 0x01c74000, 4 * KiB }, + { "tve-top", 0x01c90000, 16 * KiB }, + { "tve0", 0x01c94000, 16 * KiB }, + { "tve1", 0x01c98000, 16 * KiB }, + { "mipi_dsi", 0x01ca0000, 4 * KiB }, + { "mipi_dphy", 0x01ca1000, 4 * KiB }, + { "ve", 0x01d00000, 1024 * KiB }, + { "mp", 0x01e80000, 128 * KiB }, + { "hdmi", 0x01ee0000, 128 * KiB }, + { "prcm", 0x01f01400, 1 * KiB }, + { "debug", 0x3f500000, 64 * KiB }, + { "cpubist", 0x3f501000, 4 * KiB }, + { "dcu", 0x3fff0000, 64 * KiB }, + { "hstmr", 0x01c60000, 4 * KiB }, + { "brom", 0xffff0000, 36 * KiB } +}; + +/* Per Processor Interrupts */ +enum { + AW_R40_GIC_PPI_MAINT = 9, + AW_R40_GIC_PPI_HYPTIMER = 10, + AW_R40_GIC_PPI_VIRTTIMER = 11, + AW_R40_GIC_PPI_SECTIMER = 13, + AW_R40_GIC_PPI_PHYSTIMER = 14 +}; + +/* Shared Processor Interrupts */ +enum { + AW_R40_GIC_SPI_UART0 = 1, + AW_R40_GIC_SPI_TIMER0 = 22, + AW_R40_GIC_SPI_TIMER1 = 23, + AW_R40_GIC_SPI_MMC0 = 32, + AW_R40_GIC_SPI_MMC1 = 33, + AW_R40_GIC_SPI_MMC2 = 34, + AW_R40_GIC_SPI_MMC3 = 35, +}; + +/* Allwinner R40 general constants */ +enum { + AW_R40_GIC_NUM_SPI = 128 +}; + +#define BOOT0_MAGIC "eGON.BT0" + +/* The low 8-bits of the 'boot_media' field in the SPL header */ +#define SUNXI_BOOTED_FROM_MMC0 0 +#define SUNXI_BOOTED_FROM_NAND 1 +#define SUNXI_BOOTED_FROM_MMC2 2 +#define SUNXI_BOOTED_FROM_SPI 3 + +struct boot_file_head { + uint32_t b_instruction; + uint8_t magic[8]; + uint32_t check_sum; + uint32_t length; + uint32_t pub_head_size; + uint32_t fel_script_address; + uint32_t fel_uEnv_length; + uint32_t dt_name_offset; + uint32_t dram_size; + uint32_t boot_media; + uint32_t string_pool[13]; +}; + +bool allwinner_r40_bootrom_setup(AwR40State *s, BlockBackend *blk, int unit) +{ + const int64_t rom_size = 32 * KiB; + g_autofree uint8_t *buffer = g_new0(uint8_t, rom_size); + struct boot_file_head *head = (struct boot_file_head *)buffer; + + if (blk_pread(blk, 8 * KiB, rom_size, buffer, 0) < 0) { + error_setg(&error_fatal, "%s: failed to read BlockBackend data", + __func__); + return false; + } + + /* we only check the magic string here. */ + if (memcmp(head->magic, BOOT0_MAGIC, sizeof(head->magic))) { + return false; + } + + /* + * Simulate the behavior of the bootROM, it will change the boot_media + * flag to indicate where the chip is booting from. R40 can boot from + * mmc0 or mmc2, the default value of boot_media is zero + * (SUNXI_BOOTED_FROM_MMC0), let's fix this flag when it is booting from + * the others. + */ + if (unit == 2) { + head->boot_media = cpu_to_le32(SUNXI_BOOTED_FROM_MMC2); + } else { + head->boot_media = cpu_to_le32(SUNXI_BOOTED_FROM_MMC0); + } + + rom_add_blob("allwinner-r40.bootrom", buffer, rom_size, + rom_size, s->memmap[AW_R40_DEV_SRAM_A1], + NULL, NULL, NULL, NULL, false); + return true; +} + +static void allwinner_r40_init(Object *obj) +{ + static const char *mmc_names[AW_R40_NUM_MMCS] = { + "mmc0", "mmc1", "mmc2", "mmc3" + }; + AwR40State *s = AW_R40(obj); + + s->memmap = allwinner_r40_memmap; + + for (int i = 0; i < AW_R40_NUM_CPUS; i++) { + object_initialize_child(obj, "cpu[*]", &s->cpus[i], + ARM_CPU_TYPE_NAME("cortex-a7")); + } + + object_initialize_child(obj, "gic", &s->gic, TYPE_ARM_GIC); + + object_initialize_child(obj, "timer", &s->timer, TYPE_AW_A10_PIT); + object_property_add_alias(obj, "clk0-freq", OBJECT(&s->timer), + "clk0-freq"); + object_property_add_alias(obj, "clk1-freq", OBJECT(&s->timer), + "clk1-freq"); + + for (int i = 0; i < AW_R40_NUM_MMCS; i++) { + object_initialize_child(obj, mmc_names[i], &s->mmc[i], + TYPE_AW_SDHOST_SUN5I); + } +} + +static void allwinner_r40_realize(DeviceState *dev, Error **errp) +{ + AwR40State *s = AW_R40(dev); + unsigned i; + + /* CPUs */ + for (i = 0; i < AW_R40_NUM_CPUS; i++) { + + /* + * Disable secondary CPUs. Guest EL3 firmware will start + * them via CPU reset control registers. + */ + qdev_prop_set_bit(DEVICE(&s->cpus[i]), "start-powered-off", + i > 0); + + /* All exception levels required */ + qdev_prop_set_bit(DEVICE(&s->cpus[i]), "has_el3", true); + qdev_prop_set_bit(DEVICE(&s->cpus[i]), "has_el2", true); + + /* Mark realized */ + qdev_realize(DEVICE(&s->cpus[i]), NULL, &error_fatal); + } + + /* Generic Interrupt Controller */ + qdev_prop_set_uint32(DEVICE(&s->gic), "num-irq", AW_R40_GIC_NUM_SPI + + GIC_INTERNAL); + qdev_prop_set_uint32(DEVICE(&s->gic), "revision", 2); + qdev_prop_set_uint32(DEVICE(&s->gic), "num-cpu", AW_R40_NUM_CPUS); + qdev_prop_set_bit(DEVICE(&s->gic), "has-security-extensions", false); + qdev_prop_set_bit(DEVICE(&s->gic), "has-virtualization-extensions", true); + sysbus_realize(SYS_BUS_DEVICE(&s->gic), &error_fatal); + + sysbus_mmio_map(SYS_BUS_DEVICE(&s->gic), 0, s->memmap[AW_R40_DEV_GIC_DIST]); + sysbus_mmio_map(SYS_BUS_DEVICE(&s->gic), 1, s->memmap[AW_R40_DEV_GIC_CPU]); + sysbus_mmio_map(SYS_BUS_DEVICE(&s->gic), 2, s->memmap[AW_R40_DEV_GIC_HYP]); + sysbus_mmio_map(SYS_BUS_DEVICE(&s->gic), 3, s->memmap[AW_R40_DEV_GIC_VCPU]); + + /* + * Wire the outputs from each CPU's generic timer and the GICv2 + * maintenance interrupt signal to the appropriate GIC PPI inputs, + * and the GIC's IRQ/FIQ/VIRQ/VFIQ interrupt outputs to the CPU's inputs. + */ + for (i = 0; i < AW_R40_NUM_CPUS; i++) { + DeviceState *cpudev = DEVICE(&s->cpus[i]); + int ppibase = AW_R40_GIC_NUM_SPI + i * GIC_INTERNAL + GIC_NR_SGIS; + int irq; + /* + * Mapping from the output timer irq lines from the CPU to the + * GIC PPI inputs used for this board. + */ + const int timer_irq[] = { + [GTIMER_PHYS] = AW_R40_GIC_PPI_PHYSTIMER, + [GTIMER_VIRT] = AW_R40_GIC_PPI_VIRTTIMER, + [GTIMER_HYP] = AW_R40_GIC_PPI_HYPTIMER, + [GTIMER_SEC] = AW_R40_GIC_PPI_SECTIMER, + }; + + /* Connect CPU timer outputs to GIC PPI inputs */ + for (irq = 0; irq < ARRAY_SIZE(timer_irq); irq++) { + qdev_connect_gpio_out(cpudev, irq, + qdev_get_gpio_in(DEVICE(&s->gic), + ppibase + timer_irq[irq])); + } + + /* Connect GIC outputs to CPU interrupt inputs */ + sysbus_connect_irq(SYS_BUS_DEVICE(&s->gic), i, + qdev_get_gpio_in(cpudev, ARM_CPU_IRQ)); + sysbus_connect_irq(SYS_BUS_DEVICE(&s->gic), i + AW_R40_NUM_CPUS, + qdev_get_gpio_in(cpudev, ARM_CPU_FIQ)); + sysbus_connect_irq(SYS_BUS_DEVICE(&s->gic), i + (2 * AW_R40_NUM_CPUS), + qdev_get_gpio_in(cpudev, ARM_CPU_VIRQ)); + sysbus_connect_irq(SYS_BUS_DEVICE(&s->gic), i + (3 * AW_R40_NUM_CPUS), + qdev_get_gpio_in(cpudev, ARM_CPU_VFIQ)); + + /* GIC maintenance signal */ + sysbus_connect_irq(SYS_BUS_DEVICE(&s->gic), i + (4 * AW_R40_NUM_CPUS), + qdev_get_gpio_in(DEVICE(&s->gic), + ppibase + AW_R40_GIC_PPI_MAINT)); + } + + /* Timer */ + sysbus_realize(SYS_BUS_DEVICE(&s->timer), &error_fatal); + sysbus_mmio_map(SYS_BUS_DEVICE(&s->timer), 0, s->memmap[AW_R40_DEV_PIT]); + sysbus_connect_irq(SYS_BUS_DEVICE(&s->timer), 0, + qdev_get_gpio_in(DEVICE(&s->gic), + AW_R40_GIC_SPI_TIMER0)); + sysbus_connect_irq(SYS_BUS_DEVICE(&s->timer), 1, + qdev_get_gpio_in(DEVICE(&s->gic), + AW_R40_GIC_SPI_TIMER1)); + + /* SRAM */ + memory_region_init_ram(&s->sram_a1, OBJECT(dev), "sram A1", + 16 * KiB, &error_abort); + memory_region_init_ram(&s->sram_a2, OBJECT(dev), "sram A2", + 16 * KiB, &error_abort); + memory_region_init_ram(&s->sram_a3, OBJECT(dev), "sram A3", + 13 * KiB, &error_abort); + memory_region_init_ram(&s->sram_a4, OBJECT(dev), "sram A4", + 3 * KiB, &error_abort); + memory_region_add_subregion(get_system_memory(), + s->memmap[AW_R40_DEV_SRAM_A1], &s->sram_a1); + memory_region_add_subregion(get_system_memory(), + s->memmap[AW_R40_DEV_SRAM_A2], &s->sram_a2); + memory_region_add_subregion(get_system_memory(), + s->memmap[AW_R40_DEV_SRAM_A3], &s->sram_a3); + memory_region_add_subregion(get_system_memory(), + s->memmap[AW_R40_DEV_SRAM_A4], &s->sram_a4); + + /* SD/MMC */ + for (int i = 0; i < AW_R40_NUM_MMCS; i++) { + qemu_irq irq = qdev_get_gpio_in(DEVICE(&s->gic), + AW_R40_GIC_SPI_MMC0 + i); + const hwaddr addr = s->memmap[AW_R40_DEV_MMC0 + i]; + + object_property_set_link(OBJECT(&s->mmc[i]), "dma-memory", + OBJECT(get_system_memory()), &error_fatal); + sysbus_realize(SYS_BUS_DEVICE(&s->mmc[i]), &error_fatal); + sysbus_mmio_map(SYS_BUS_DEVICE(&s->mmc[i]), 0, addr); + sysbus_connect_irq(SYS_BUS_DEVICE(&s->mmc[i]), 0, irq); + } + + /* UART0. For future clocktree API: All UARTS are connected to APB2_CLK. */ + serial_mm_init(get_system_memory(), s->memmap[AW_R40_DEV_UART0], 2, + qdev_get_gpio_in(DEVICE(&s->gic), AW_R40_GIC_SPI_UART0), + 115200, serial_hd(0), DEVICE_NATIVE_ENDIAN); + + /* Unimplemented devices */ + for (i = 0; i < ARRAY_SIZE(r40_unimplemented); i++) { + create_unimplemented_device(r40_unimplemented[i].device_name, + r40_unimplemented[i].base, + r40_unimplemented[i].size); + } +} + +static void allwinner_r40_class_init(ObjectClass *oc, void *data) +{ + DeviceClass *dc = DEVICE_CLASS(oc); + + dc->realize = allwinner_r40_realize; + /* Reason: uses serial_hd() in realize function */ + dc->user_creatable = false; +} + +static const TypeInfo allwinner_r40_type_info = { + .name = TYPE_AW_R40, + .parent = TYPE_DEVICE, + .instance_size = sizeof(AwR40State), + .instance_init = allwinner_r40_init, + .class_init = allwinner_r40_class_init, +}; + +static void allwinner_r40_register_types(void) +{ + type_register_static(&allwinner_r40_type_info); +} + +type_init(allwinner_r40_register_types) diff --git a/hw/arm/bananapi_m2u.c b/hw/arm/bananapi_m2u.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/arm/bananapi_m2u.c @@ -XXX,XX +XXX,XX @@ +/* + * Bananapi M2U emulation + * + * Copyright (C) 2023 qianfan Zhao <qianfanguijin@163.com> + * + * This program is free software: you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation, either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#include "qemu/osdep.h" +#include "qemu/units.h" +#include "exec/address-spaces.h" +#include "qapi/error.h" +#include "qemu/error-report.h" +#include "hw/boards.h" +#include "hw/qdev-properties.h" +#include "hw/arm/allwinner-r40.h" + +static struct arm_boot_info bpim2u_binfo; + +/* + * R40 can boot from mmc0 and mmc2, and bpim2u has two mmc interface, one is + * connected to sdcard and another mount an emmc media. + * Attach the mmc driver and try loading bootloader. + */ +static void mmc_attach_drive(AwR40State *s, AwSdHostState *mmc, int unit, + bool load_bootroom, bool *bootroom_loaded) +{ + DriveInfo *di = drive_get(IF_SD, 0, unit); + BlockBackend *blk = di ? blk_by_legacy_dinfo(di) : NULL; + BusState *bus; + DeviceState *carddev; + + bus = qdev_get_child_bus(DEVICE(mmc), "sd-bus"); + if (bus == NULL) { + error_report("No SD bus found in SOC object"); + exit(1); + } + + carddev = qdev_new(TYPE_SD_CARD); + qdev_prop_set_drive_err(carddev, "drive", blk, &error_fatal); + qdev_realize_and_unref(carddev, bus, &error_fatal); + + if (load_bootroom && blk && blk_is_available(blk)) { + /* Use Boot ROM to copy data from SD card to SRAM */ + *bootroom_loaded = allwinner_r40_bootrom_setup(s, blk, unit); + } +} + +static void bpim2u_init(MachineState *machine) +{ + bool bootroom_loaded = false; + AwR40State *r40; + + /* BIOS is not supported by this board */ + if (machine->firmware) { + error_report("BIOS not supported for this machine"); + exit(1); + } + + /* Only allow Cortex-A7 for this board */ + if (strcmp(machine->cpu_type, ARM_CPU_TYPE_NAME("cortex-a7")) != 0) { + error_report("This board can only be used with cortex-a7 CPU"); + exit(1); + } + + r40 = AW_R40(object_new(TYPE_AW_R40)); + object_property_add_child(OBJECT(machine), "soc", OBJECT(r40)); + object_unref(OBJECT(r40)); + + /* Setup timer properties */ + object_property_set_int(OBJECT(r40), "clk0-freq", 32768, &error_abort); + object_property_set_int(OBJECT(r40), "clk1-freq", 24 * 1000 * 1000, + &error_abort); + + /* Mark R40 object realized */ + qdev_realize(DEVICE(r40), NULL, &error_abort); + + /* + * Plug in SD card and try load bootrom, R40 has 4 mmc controllers but can + * only booting from mmc0 and mmc2. + */ + for (int i = 0; i < AW_R40_NUM_MMCS; i++) { + switch (i) { + case 0: + case 2: + mmc_attach_drive(r40, &r40->mmc[i], i, + !machine->kernel_filename && !bootroom_loaded, + &bootroom_loaded); + break; + default: + mmc_attach_drive(r40, &r40->mmc[i], i, false, NULL); + break; + } + } + + /* SDRAM */ + memory_region_add_subregion(get_system_memory(), + r40->memmap[AW_R40_DEV_SDRAM], machine->ram); + + bpim2u_binfo.loader_start = r40->memmap[AW_R40_DEV_SDRAM]; + bpim2u_binfo.ram_size = machine->ram_size; + bpim2u_binfo.psci_conduit = QEMU_PSCI_CONDUIT_SMC; + arm_load_kernel(ARM_CPU(first_cpu), machine, &bpim2u_binfo); +} + +static void bpim2u_machine_init(MachineClass *mc) +{ + mc->desc = "Bananapi M2U (Cortex-A7)"; + mc->init = bpim2u_init; + mc->min_cpus = AW_R40_NUM_CPUS; + mc->max_cpus = AW_R40_NUM_CPUS; + mc->default_cpus = AW_R40_NUM_CPUS; + mc->default_cpu_type = ARM_CPU_TYPE_NAME("cortex-a7"); + mc->default_ram_size = 1 * GiB; + mc->default_ram_id = "bpim2u.ram"; +} + +DEFINE_MACHINE("bpim2u", bpim2u_machine_init) diff --git a/hw/arm/Kconfig b/hw/arm/Kconfig index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/Kconfig +++ b/hw/arm/Kconfig @@ -XXX,XX +XXX,XX @@ config ALLWINNER_H3 select USB_EHCI_SYSBUS select SD +config ALLWINNER_R40 + bool + default y if TCG && ARM + select ALLWINNER_A10_PIT + select SERIAL + select ARM_TIMER + select ARM_GIC + select UNIMP + select SD + config RASPI bool default y diff --git a/hw/arm/meson.build b/hw/arm/meson.build index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/meson.build +++ b/hw/arm/meson.build @@ -XXX,XX +XXX,XX @@ arm_ss.add(when: 'CONFIG_OMAP', if_true: files('omap1.c', 'omap2.c')) arm_ss.add(when: 'CONFIG_STRONGARM', if_true: files('strongarm.c')) arm_ss.add(when: 'CONFIG_ALLWINNER_A10', if_true: files('allwinner-a10.c', 'cubieboard.c')) arm_ss.add(when: 'CONFIG_ALLWINNER_H3', if_true: files('allwinner-h3.c', 'orangepi.c')) +arm_ss.add(when: 'CONFIG_ALLWINNER_R40', if_true: files('allwinner-r40.c', 'bananapi_m2u.c')) arm_ss.add(when: 'CONFIG_RASPI', if_true: files('bcm2836.c', 'raspi.c')) arm_ss.add(when: 'CONFIG_STM32F100_SOC', if_true: files('stm32f100_soc.c')) arm_ss.add(when: 'CONFIG_STM32F205_SOC', if_true: files('stm32f205_soc.c')) -- 2.34.1 From: qianfan Zhao <qianfanguijin@163.com> The CCU provides the registers to program the PLLs and the controls most of the clock generation, division, distribution, synchronization and gating. This commit adds support for the Clock Control Unit which emulates a simple read/write register interface. Signed-off-by: qianfan Zhao <qianfanguijin@163.com> Reviewed-by: Niek Linnenbank <nieklinnenbank@gmail.com> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/allwinner-r40.h | 2 + include/hw/misc/allwinner-r40-ccu.h | 65 +++++++++ hw/arm/allwinner-r40.c | 8 +- hw/misc/allwinner-r40-ccu.c | 209 ++++++++++++++++++++++++++++ hw/misc/meson.build | 1 + 5 files changed, 284 insertions(+), 1 deletion(-) create mode 100644 include/hw/misc/allwinner-r40-ccu.h create mode 100644 hw/misc/allwinner-r40-ccu.c diff --git a/include/hw/arm/allwinner-r40.h b/include/hw/arm/allwinner-r40.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/allwinner-r40.h +++ b/include/hw/arm/allwinner-r40.h @@ -XXX,XX +XXX,XX @@ #include "hw/timer/allwinner-a10-pit.h" #include "hw/intc/arm_gic.h" #include "hw/sd/allwinner-sdhost.h" +#include "hw/misc/allwinner-r40-ccu.h" #include "target/arm/cpu.h" #include "sysemu/block-backend.h" @@ -XXX,XX +XXX,XX @@ struct AwR40State { const hwaddr *memmap; AwA10PITState timer; AwSdHostState mmc[AW_R40_NUM_MMCS]; + AwR40ClockCtlState ccu; GICState gic; MemoryRegion sram_a1; MemoryRegion sram_a2; diff --git a/include/hw/misc/allwinner-r40-ccu.h b/include/hw/misc/allwinner-r40-ccu.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/include/hw/misc/allwinner-r40-ccu.h @@ -XXX,XX +XXX,XX @@ +/* + * Allwinner R40 Clock Control Unit emulation + * + * Copyright (C) 2023 qianfan Zhao <qianfanguijin@163.com> + * + * This program is free software: you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation, either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#ifndef HW_MISC_ALLWINNER_R40_CCU_H +#define HW_MISC_ALLWINNER_R40_CCU_H + +#include "qom/object.h" +#include "hw/sysbus.h" + +/** + * @name Constants + * @{ + */ + +/** Size of register I/O address space used by CCU device */ +#define AW_R40_CCU_IOSIZE (0x400) + +/** Total number of known registers */ +#define AW_R40_CCU_REGS_NUM (AW_R40_CCU_IOSIZE / sizeof(uint32_t)) + +/** @} */ + +/** + * @name Object model + * @{ + */ + +#define TYPE_AW_R40_CCU "allwinner-r40-ccu" +OBJECT_DECLARE_SIMPLE_TYPE(AwR40ClockCtlState, AW_R40_CCU) + +/** @} */ + +/** + * Allwinner R40 CCU object instance state. + */ +struct AwR40ClockCtlState { + /*< private >*/ + SysBusDevice parent_obj; + /*< public >*/ + + /** Maps I/O registers in physical memory */ + MemoryRegion iomem; + + /** Array of hardware registers */ + uint32_t regs[AW_R40_CCU_REGS_NUM]; + +}; + +#endif /* HW_MISC_ALLWINNER_R40_CCU_H */ diff --git a/hw/arm/allwinner-r40.c b/hw/arm/allwinner-r40.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/allwinner-r40.c +++ b/hw/arm/allwinner-r40.c @@ -XXX,XX +XXX,XX @@ const hwaddr allwinner_r40_memmap[] = { [AW_R40_DEV_MMC1] = 0x01c10000, [AW_R40_DEV_MMC2] = 0x01c11000, [AW_R40_DEV_MMC3] = 0x01c12000, + [AW_R40_DEV_CCU] = 0x01c20000, [AW_R40_DEV_PIT] = 0x01c20c00, [AW_R40_DEV_UART0] = 0x01c28000, [AW_R40_DEV_GIC_DIST] = 0x01c81000, @@ -XXX,XX +XXX,XX @@ static struct AwR40Unimplemented r40_unimplemented[] = { { "usb2-host", 0x01c1c000, 4 * KiB }, { "cs1", 0x01c1d000, 4 * KiB }, { "spi3", 0x01c1f000, 4 * KiB }, - { "ccu", 0x01c20000, 1 * KiB }, { "rtc", 0x01c20400, 1 * KiB }, { "pio", 0x01c20800, 1 * KiB }, { "owa", 0x01c21000, 1 * KiB }, @@ -XXX,XX +XXX,XX @@ static void allwinner_r40_init(Object *obj) object_property_add_alias(obj, "clk1-freq", OBJECT(&s->timer), "clk1-freq"); + object_initialize_child(obj, "ccu", &s->ccu, TYPE_AW_R40_CCU); + for (int i = 0; i < AW_R40_NUM_MMCS; i++) { object_initialize_child(obj, mmc_names[i], &s->mmc[i], TYPE_AW_SDHOST_SUN5I); @@ -XXX,XX +XXX,XX @@ static void allwinner_r40_realize(DeviceState *dev, Error **errp) memory_region_add_subregion(get_system_memory(), s->memmap[AW_R40_DEV_SRAM_A4], &s->sram_a4); + /* Clock Control Unit */ + sysbus_realize(SYS_BUS_DEVICE(&s->ccu), &error_fatal); + sysbus_mmio_map(SYS_BUS_DEVICE(&s->ccu), 0, s->memmap[AW_R40_DEV_CCU]); + /* SD/MMC */ for (int i = 0; i < AW_R40_NUM_MMCS; i++) { qemu_irq irq = qdev_get_gpio_in(DEVICE(&s->gic), diff --git a/hw/misc/allwinner-r40-ccu.c b/hw/misc/allwinner-r40-ccu.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/misc/allwinner-r40-ccu.c @@ -XXX,XX +XXX,XX @@ +/* + * Allwinner R40 Clock Control Unit emulation + * + * Copyright (C) 2023 qianfan Zhao <qianfanguijin@163.com> + * + * This program is free software: you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation, either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#include "qemu/osdep.h" +#include "qemu/units.h" +#include "hw/sysbus.h" +#include "migration/vmstate.h" +#include "qemu/log.h" +#include "qemu/module.h" +#include "hw/misc/allwinner-r40-ccu.h" + +/* CCU register offsets */ +enum { + REG_PLL_CPUX_CTRL = 0x0000, + REG_PLL_AUDIO_CTRL = 0x0008, + REG_PLL_VIDEO0_CTRL = 0x0010, + REG_PLL_VE_CTRL = 0x0018, + REG_PLL_DDR0_CTRL = 0x0020, + REG_PLL_PERIPH0_CTRL = 0x0028, + REG_PLL_PERIPH1_CTRL = 0x002c, + REG_PLL_VIDEO1_CTRL = 0x0030, + REG_PLL_SATA_CTRL = 0x0034, + REG_PLL_GPU_CTRL = 0x0038, + REG_PLL_MIPI_CTRL = 0x0040, + REG_PLL_DE_CTRL = 0x0048, + REG_PLL_DDR1_CTRL = 0x004c, + REG_AHB1_APB1_CFG = 0x0054, + REG_APB2_CFG = 0x0058, + REG_MMC0_CLK = 0x0088, + REG_MMC1_CLK = 0x008c, + REG_MMC2_CLK = 0x0090, + REG_MMC3_CLK = 0x0094, + REG_USBPHY_CFG = 0x00cc, + REG_PLL_DDR_AUX = 0x00f0, + REG_DRAM_CFG = 0x00f4, + REG_PLL_DDR1_CFG = 0x00f8, + REG_DRAM_CLK_GATING = 0x0100, + REG_GMAC_CLK = 0x0164, + REG_SYS_32K_CLK = 0x0310, + REG_PLL_LOCK_CTRL = 0x0320, +}; + +#define REG_INDEX(offset) (offset / sizeof(uint32_t)) + +/* CCU register flags */ +enum { + REG_PLL_ENABLE = (1 << 31), + REG_PLL_LOCK = (1 << 28), +}; + +static uint64_t allwinner_r40_ccu_read(void *opaque, hwaddr offset, + unsigned size) +{ + const AwR40ClockCtlState *s = AW_R40_CCU(opaque); + const uint32_t idx = REG_INDEX(offset); + + switch (offset) { + case 0x324 ... AW_R40_CCU_IOSIZE: + qemu_log_mask(LOG_GUEST_ERROR, "%s: out-of-bounds offset 0x%04x\n", + __func__, (uint32_t)offset); + return 0; + } + + return s->regs[idx]; +} + +static void allwinner_r40_ccu_write(void *opaque, hwaddr offset, + uint64_t val, unsigned size) +{ + AwR40ClockCtlState *s = AW_R40_CCU(opaque); + + switch (offset) { + case REG_DRAM_CFG: /* DRAM Configuration(for DDR0) */ + /* bit16: SDRCLK_UPD (SDRCLK configuration 0 update) */ + val &= ~(1 << 16); + break; + case REG_PLL_DDR1_CTRL: /* DDR1 Control register */ + /* bit30: SDRPLL_UPD */ + val &= ~(1 << 30); + if (val & REG_PLL_ENABLE) { + val |= REG_PLL_LOCK; + } + break; + case REG_PLL_CPUX_CTRL: + case REG_PLL_AUDIO_CTRL: + case REG_PLL_VE_CTRL: + case REG_PLL_VIDEO0_CTRL: + case REG_PLL_DDR0_CTRL: + case REG_PLL_PERIPH0_CTRL: + case REG_PLL_PERIPH1_CTRL: + case REG_PLL_VIDEO1_CTRL: + case REG_PLL_SATA_CTRL: + case REG_PLL_GPU_CTRL: + case REG_PLL_MIPI_CTRL: + case REG_PLL_DE_CTRL: + if (val & REG_PLL_ENABLE) { + val |= REG_PLL_LOCK; + } + break; + case 0x324 ... AW_R40_CCU_IOSIZE: + qemu_log_mask(LOG_GUEST_ERROR, "%s: out-of-bounds offset 0x%04x\n", + __func__, (uint32_t)offset); + break; + default: + qemu_log_mask(LOG_UNIMP, "%s: unimplemented write offset 0x%04x\n", + __func__, (uint32_t)offset); + break; + } + + s->regs[REG_INDEX(offset)] = (uint32_t) val; +} + +static const MemoryRegionOps allwinner_r40_ccu_ops = { + .read = allwinner_r40_ccu_read, + .write = allwinner_r40_ccu_write, + .endianness = DEVICE_NATIVE_ENDIAN, + .valid = { + .min_access_size = 4, + .max_access_size = 4, + }, + .impl.min_access_size = 4, +}; + +static void allwinner_r40_ccu_reset(DeviceState *dev) +{ + AwR40ClockCtlState *s = AW_R40_CCU(dev); + + memset(s->regs, 0, sizeof(s->regs)); + + /* Set default values for registers */ + s->regs[REG_INDEX(REG_PLL_CPUX_CTRL)] = 0x00001000; + s->regs[REG_INDEX(REG_PLL_AUDIO_CTRL)] = 0x00035514; + s->regs[REG_INDEX(REG_PLL_VIDEO0_CTRL)] = 0x03006207; + s->regs[REG_INDEX(REG_PLL_VE_CTRL)] = 0x03006207; + s->regs[REG_INDEX(REG_PLL_DDR0_CTRL)] = 0x00001000, + s->regs[REG_INDEX(REG_PLL_PERIPH0_CTRL)] = 0x00041811; + s->regs[REG_INDEX(REG_PLL_PERIPH1_CTRL)] = 0x00041811; + s->regs[REG_INDEX(REG_PLL_VIDEO1_CTRL)] = 0x03006207; + s->regs[REG_INDEX(REG_PLL_SATA_CTRL)] = 0x00001811; + s->regs[REG_INDEX(REG_PLL_GPU_CTRL)] = 0x03006207; + s->regs[REG_INDEX(REG_PLL_MIPI_CTRL)] = 0x00000515; + s->regs[REG_INDEX(REG_PLL_DE_CTRL)] = 0x03006207; + s->regs[REG_INDEX(REG_PLL_DDR1_CTRL)] = 0x00001800; + s->regs[REG_INDEX(REG_AHB1_APB1_CFG)] = 0x00001010; + s->regs[REG_INDEX(REG_APB2_CFG)] = 0x01000000; + s->regs[REG_INDEX(REG_PLL_DDR_AUX)] = 0x00000001; + s->regs[REG_INDEX(REG_PLL_DDR1_CFG)] = 0x0ccca000; + s->regs[REG_INDEX(REG_SYS_32K_CLK)] = 0x0000000f; +} + +static void allwinner_r40_ccu_init(Object *obj) +{ + SysBusDevice *sbd = SYS_BUS_DEVICE(obj); + AwR40ClockCtlState *s = AW_R40_CCU(obj); + + /* Memory mapping */ + memory_region_init_io(&s->iomem, OBJECT(s), &allwinner_r40_ccu_ops, s, + TYPE_AW_R40_CCU, AW_R40_CCU_IOSIZE); + sysbus_init_mmio(sbd, &s->iomem); +} + +static const VMStateDescription allwinner_r40_ccu_vmstate = { + .name = "allwinner-r40-ccu", + .version_id = 1, + .minimum_version_id = 1, + .fields = (VMStateField[]) { + VMSTATE_UINT32_ARRAY(regs, AwR40ClockCtlState, AW_R40_CCU_REGS_NUM), + VMSTATE_END_OF_LIST() + } +}; + +static void allwinner_r40_ccu_class_init(ObjectClass *klass, void *data) +{ + DeviceClass *dc = DEVICE_CLASS(klass); + + dc->reset = allwinner_r40_ccu_reset; + dc->vmsd = &allwinner_r40_ccu_vmstate; +} + +static const TypeInfo allwinner_r40_ccu_info = { + .name = TYPE_AW_R40_CCU, + .parent = TYPE_SYS_BUS_DEVICE, + .instance_init = allwinner_r40_ccu_init, + .instance_size = sizeof(AwR40ClockCtlState), + .class_init = allwinner_r40_ccu_class_init, +}; + +static void allwinner_r40_ccu_register(void) +{ + type_register_static(&allwinner_r40_ccu_info); +} + +type_init(allwinner_r40_ccu_register) diff --git a/hw/misc/meson.build b/hw/misc/meson.build index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/meson.build +++ b/hw/misc/meson.build @@ -XXX,XX +XXX,XX @@ specific_ss.add(when: 'CONFIG_ALLWINNER_H3', if_true: files('allwinner-cpucfg.c' softmmu_ss.add(when: 'CONFIG_ALLWINNER_H3', if_true: files('allwinner-h3-dramc.c')) softmmu_ss.add(when: 'CONFIG_ALLWINNER_H3', if_true: files('allwinner-h3-sysctrl.c')) softmmu_ss.add(when: 'CONFIG_ALLWINNER_H3', if_true: files('allwinner-sid.c')) +softmmu_ss.add(when: 'CONFIG_ALLWINNER_R40', if_true: files('allwinner-r40-ccu.c')) softmmu_ss.add(when: 'CONFIG_AXP209_PMU', if_true: files('axp209.c')) softmmu_ss.add(when: 'CONFIG_REALVIEW', if_true: files('arm_sysctl.c')) softmmu_ss.add(when: 'CONFIG_NSERIES', if_true: files('cbus.c')) -- 2.34.1 From: qianfan Zhao <qianfanguijin@163.com> R40 has eight UARTs, support both 16450 and 16550 compatible modes. Signed-off-by: qianfan Zhao <qianfanguijin@163.com> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/allwinner-r40.h | 8 ++++++++ hw/arm/allwinner-r40.c | 34 +++++++++++++++++++++++++++++++--- 2 files changed, 39 insertions(+), 3 deletions(-) diff --git a/include/hw/arm/allwinner-r40.h b/include/hw/arm/allwinner-r40.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/allwinner-r40.h +++ b/include/hw/arm/allwinner-r40.h @@ -XXX,XX +XXX,XX @@ enum { AW_R40_DEV_CCU, AW_R40_DEV_PIT, AW_R40_DEV_UART0, + AW_R40_DEV_UART1, + AW_R40_DEV_UART2, + AW_R40_DEV_UART3, + AW_R40_DEV_UART4, + AW_R40_DEV_UART5, + AW_R40_DEV_UART6, + AW_R40_DEV_UART7, AW_R40_DEV_GIC_DIST, AW_R40_DEV_GIC_CPU, AW_R40_DEV_GIC_HYP, @@ -XXX,XX +XXX,XX @@ OBJECT_DECLARE_SIMPLE_TYPE(AwR40State, AW_R40) * which are currently emulated by the R40 SoC code. */ #define AW_R40_NUM_MMCS 4 +#define AW_R40_NUM_UARTS 8 struct AwR40State { /*< private >*/ diff --git a/hw/arm/allwinner-r40.c b/hw/arm/allwinner-r40.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/allwinner-r40.c +++ b/hw/arm/allwinner-r40.c @@ -XXX,XX +XXX,XX @@ const hwaddr allwinner_r40_memmap[] = { [AW_R40_DEV_CCU] = 0x01c20000, [AW_R40_DEV_PIT] = 0x01c20c00, [AW_R40_DEV_UART0] = 0x01c28000, + [AW_R40_DEV_UART1] = 0x01c28400, + [AW_R40_DEV_UART2] = 0x01c28800, + [AW_R40_DEV_UART3] = 0x01c28c00, + [AW_R40_DEV_UART4] = 0x01c29000, + [AW_R40_DEV_UART5] = 0x01c29400, + [AW_R40_DEV_UART6] = 0x01c29800, + [AW_R40_DEV_UART7] = 0x01c29c00, [AW_R40_DEV_GIC_DIST] = 0x01c81000, [AW_R40_DEV_GIC_CPU] = 0x01c82000, [AW_R40_DEV_GIC_HYP] = 0x01c84000, @@ -XXX,XX +XXX,XX @@ enum { /* Shared Processor Interrupts */ enum { AW_R40_GIC_SPI_UART0 = 1, + AW_R40_GIC_SPI_UART1 = 2, + AW_R40_GIC_SPI_UART2 = 3, + AW_R40_GIC_SPI_UART3 = 4, + AW_R40_GIC_SPI_UART4 = 17, + AW_R40_GIC_SPI_UART5 = 18, + AW_R40_GIC_SPI_UART6 = 19, + AW_R40_GIC_SPI_UART7 = 20, AW_R40_GIC_SPI_TIMER0 = 22, AW_R40_GIC_SPI_TIMER1 = 23, AW_R40_GIC_SPI_MMC0 = 32, @@ -XXX,XX +XXX,XX @@ static void allwinner_r40_realize(DeviceState *dev, Error **errp) } /* UART0. For future clocktree API: All UARTS are connected to APB2_CLK. */ - serial_mm_init(get_system_memory(), s->memmap[AW_R40_DEV_UART0], 2, - qdev_get_gpio_in(DEVICE(&s->gic), AW_R40_GIC_SPI_UART0), - 115200, serial_hd(0), DEVICE_NATIVE_ENDIAN); + for (int i = 0; i < AW_R40_NUM_UARTS; i++) { + static const int uart_irqs[AW_R40_NUM_UARTS] = { + AW_R40_GIC_SPI_UART0, + AW_R40_GIC_SPI_UART1, + AW_R40_GIC_SPI_UART2, + AW_R40_GIC_SPI_UART3, + AW_R40_GIC_SPI_UART4, + AW_R40_GIC_SPI_UART5, + AW_R40_GIC_SPI_UART6, + AW_R40_GIC_SPI_UART7, + }; + const hwaddr addr = s->memmap[AW_R40_DEV_UART0 + i]; + + serial_mm_init(get_system_memory(), addr, 2, + qdev_get_gpio_in(DEVICE(&s->gic), uart_irqs[i]), + 115200, serial_hd(i), DEVICE_NATIVE_ENDIAN); + } /* Unimplemented devices */ for (i = 0; i < ARRAY_SIZE(r40_unimplemented); i++) { -- 2.34.1 From: qianfan Zhao <qianfanguijin@163.com> TWI(i2c) is designed to be used as an interface between CPU host and the serial 2-Wire bus. It can support all standard 2-Wire transfer, can be operated in standard mode(100kbit/s) or fast-mode, supporting data rate up to 400kbit/s. Signed-off-by: qianfan Zhao <qianfanguijin@163.com> Reviewed-by: Niek Linnenbank <nieklinnenbank@gmail.com> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/allwinner-r40.h | 3 +++ hw/arm/allwinner-r40.c | 11 ++++++++++- 2 files changed, 13 insertions(+), 1 deletion(-) diff --git a/include/hw/arm/allwinner-r40.h b/include/hw/arm/allwinner-r40.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/allwinner-r40.h +++ b/include/hw/arm/allwinner-r40.h @@ -XXX,XX +XXX,XX @@ #include "hw/intc/arm_gic.h" #include "hw/sd/allwinner-sdhost.h" #include "hw/misc/allwinner-r40-ccu.h" +#include "hw/i2c/allwinner-i2c.h" #include "target/arm/cpu.h" #include "sysemu/block-backend.h" @@ -XXX,XX +XXX,XX @@ enum { AW_R40_DEV_UART5, AW_R40_DEV_UART6, AW_R40_DEV_UART7, + AW_R40_DEV_TWI0, AW_R40_DEV_GIC_DIST, AW_R40_DEV_GIC_CPU, AW_R40_DEV_GIC_HYP, @@ -XXX,XX +XXX,XX @@ struct AwR40State { AwA10PITState timer; AwSdHostState mmc[AW_R40_NUM_MMCS]; AwR40ClockCtlState ccu; + AWI2CState i2c0; GICState gic; MemoryRegion sram_a1; MemoryRegion sram_a2; diff --git a/hw/arm/allwinner-r40.c b/hw/arm/allwinner-r40.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/allwinner-r40.c +++ b/hw/arm/allwinner-r40.c @@ -XXX,XX +XXX,XX @@ const hwaddr allwinner_r40_memmap[] = { [AW_R40_DEV_UART5] = 0x01c29400, [AW_R40_DEV_UART6] = 0x01c29800, [AW_R40_DEV_UART7] = 0x01c29c00, + [AW_R40_DEV_TWI0] = 0x01c2ac00, [AW_R40_DEV_GIC_DIST] = 0x01c81000, [AW_R40_DEV_GIC_CPU] = 0x01c82000, [AW_R40_DEV_GIC_HYP] = 0x01c84000, @@ -XXX,XX +XXX,XX @@ static struct AwR40Unimplemented r40_unimplemented[] = { { "uart7", 0x01c29c00, 1 * KiB }, { "ps20", 0x01c2a000, 1 * KiB }, { "ps21", 0x01c2a400, 1 * KiB }, - { "twi0", 0x01c2ac00, 1 * KiB }, { "twi1", 0x01c2b000, 1 * KiB }, { "twi2", 0x01c2b400, 1 * KiB }, { "twi3", 0x01c2b800, 1 * KiB }, @@ -XXX,XX +XXX,XX @@ enum { AW_R40_GIC_SPI_UART1 = 2, AW_R40_GIC_SPI_UART2 = 3, AW_R40_GIC_SPI_UART3 = 4, + AW_R40_GIC_SPI_TWI0 = 7, AW_R40_GIC_SPI_UART4 = 17, AW_R40_GIC_SPI_UART5 = 18, AW_R40_GIC_SPI_UART6 = 19, @@ -XXX,XX +XXX,XX @@ static void allwinner_r40_init(Object *obj) object_initialize_child(obj, mmc_names[i], &s->mmc[i], TYPE_AW_SDHOST_SUN5I); } + + object_initialize_child(obj, "twi0", &s->i2c0, TYPE_AW_I2C_SUN6I); } static void allwinner_r40_realize(DeviceState *dev, Error **errp) @@ -XXX,XX +XXX,XX @@ static void allwinner_r40_realize(DeviceState *dev, Error **errp) 115200, serial_hd(i), DEVICE_NATIVE_ENDIAN); } + /* I2C */ + sysbus_realize(SYS_BUS_DEVICE(&s->i2c0), &error_fatal); + sysbus_mmio_map(SYS_BUS_DEVICE(&s->i2c0), 0, s->memmap[AW_R40_DEV_TWI0]); + sysbus_connect_irq(SYS_BUS_DEVICE(&s->i2c0), 0, + qdev_get_gpio_in(DEVICE(&s->gic), AW_R40_GIC_SPI_TWI0)); + /* Unimplemented devices */ for (i = 0; i < ARRAY_SIZE(r40_unimplemented); i++) { create_unimplemented_device(r40_unimplemented[i].device_name, -- 2.34.1 From: qianfan Zhao <qianfanguijin@163.com> This patch adds minimal support for AXP-221 PMU and connect it to bananapi M2U board. Signed-off-by: qianfan Zhao <qianfanguijin@163.com> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/bananapi_m2u.c | 6 + hw/misc/axp209.c | 238 ----------------------------------- hw/misc/axp2xx.c | 283 ++++++++++++++++++++++++++++++++++++++++++ hw/arm/Kconfig | 3 +- hw/misc/Kconfig | 2 +- hw/misc/meson.build | 2 +- hw/misc/trace-events | 8 +- 7 files changed, 297 insertions(+), 245 deletions(-) delete mode 100644 hw/misc/axp209.c create mode 100644 hw/misc/axp2xx.c diff --git a/hw/arm/bananapi_m2u.c b/hw/arm/bananapi_m2u.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/bananapi_m2u.c +++ b/hw/arm/bananapi_m2u.c @@ -XXX,XX +XXX,XX @@ #include "qapi/error.h" #include "qemu/error-report.h" #include "hw/boards.h" +#include "hw/i2c/i2c.h" #include "hw/qdev-properties.h" #include "hw/arm/allwinner-r40.h" @@ -XXX,XX +XXX,XX @@ static void bpim2u_init(MachineState *machine) { bool bootroom_loaded = false; AwR40State *r40; + I2CBus *i2c; /* BIOS is not supported by this board */ if (machine->firmware) { @@ -XXX,XX +XXX,XX @@ static void bpim2u_init(MachineState *machine) } } + /* Connect AXP221 */ + i2c = I2C_BUS(qdev_get_child_bus(DEVICE(&r40->i2c0), "i2c")); + i2c_slave_create_simple(i2c, "axp221_pmu", 0x34); + /* SDRAM */ memory_region_add_subregion(get_system_memory(), r40->memmap[AW_R40_DEV_SDRAM], machine->ram); diff --git a/hw/misc/axp209.c b/hw/misc/axp209.c deleted file mode 100644 index XXXXXXX..XXXXXXX --- a/hw/misc/axp209.c +++ /dev/null @@ -XXX,XX +XXX,XX @@ -/* - * AXP-209 PMU Emulation - * - * Copyright (C) 2022 Strahinja Jankovic <strahinja.p.jankovic@gmail.com> - * - * Permission is hereby granted, free of charge, to any person obtaining a - * copy of this software and associated documentation files (the "Software"), - * to deal in the Software without restriction, including without limitation - * the rights to use, copy, modify, merge, publish, distribute, sublicense, - * and/or sell copies of the Software, and to permit persons to whom the - * Software is furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice shall be included in - * all copies or substantial portions of the Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR - * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE - * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER - * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING - * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER - * DEALINGS IN THE SOFTWARE. - * - * SPDX-License-Identifier: MIT - */ - -#include "qemu/osdep.h" -#include "qemu/log.h" -#include "trace.h" -#include "hw/i2c/i2c.h" -#include "migration/vmstate.h" - -#define TYPE_AXP209_PMU "axp209_pmu" - -#define AXP209(obj) \ - OBJECT_CHECK(AXP209I2CState, (obj), TYPE_AXP209_PMU) - -/* registers */ -enum { - REG_POWER_STATUS = 0x0u, - REG_OPERATING_MODE, - REG_OTG_VBUS_STATUS, - REG_CHIP_VERSION, - REG_DATA_CACHE_0, - REG_DATA_CACHE_1, - REG_DATA_CACHE_2, - REG_DATA_CACHE_3, - REG_DATA_CACHE_4, - REG_DATA_CACHE_5, - REG_DATA_CACHE_6, - REG_DATA_CACHE_7, - REG_DATA_CACHE_8, - REG_DATA_CACHE_9, - REG_DATA_CACHE_A, - REG_DATA_CACHE_B, - REG_POWER_OUTPUT_CTRL = 0x12u, - REG_DC_DC2_OUT_V_CTRL = 0x23u, - REG_DC_DC2_DVS_CTRL = 0x25u, - REG_DC_DC3_OUT_V_CTRL = 0x27u, - REG_LDO2_4_OUT_V_CTRL, - REG_LDO3_OUT_V_CTRL, - REG_VBUS_CH_MGMT = 0x30u, - REG_SHUTDOWN_V_CTRL, - REG_SHUTDOWN_CTRL, - REG_CHARGE_CTRL_1, - REG_CHARGE_CTRL_2, - REG_SPARE_CHARGE_CTRL, - REG_PEK_KEY_CTRL, - REG_DC_DC_FREQ_SET, - REG_CHR_TEMP_TH_SET, - REG_CHR_HIGH_TEMP_TH_CTRL, - REG_IPSOUT_WARN_L1, - REG_IPSOUT_WARN_L2, - REG_DISCHR_TEMP_TH_SET, - REG_DISCHR_HIGH_TEMP_TH_CTRL, - REG_IRQ_BANK_1_CTRL = 0x40u, - REG_IRQ_BANK_2_CTRL, - REG_IRQ_BANK_3_CTRL, - REG_IRQ_BANK_4_CTRL, - REG_IRQ_BANK_5_CTRL, - REG_IRQ_BANK_1_STAT = 0x48u, - REG_IRQ_BANK_2_STAT, - REG_IRQ_BANK_3_STAT, - REG_IRQ_BANK_4_STAT, - REG_IRQ_BANK_5_STAT, - REG_ADC_ACIN_V_H = 0x56u, - REG_ADC_ACIN_V_L, - REG_ADC_ACIN_CURR_H, - REG_ADC_ACIN_CURR_L, - REG_ADC_VBUS_V_H, - REG_ADC_VBUS_V_L, - REG_ADC_VBUS_CURR_H, - REG_ADC_VBUS_CURR_L, - REG_ADC_INT_TEMP_H, - REG_ADC_INT_TEMP_L, - REG_ADC_TEMP_SENS_V_H = 0x62u, - REG_ADC_TEMP_SENS_V_L, - REG_ADC_BAT_V_H = 0x78u, - REG_ADC_BAT_V_L, - REG_ADC_BAT_DISCHR_CURR_H, - REG_ADC_BAT_DISCHR_CURR_L, - REG_ADC_BAT_CHR_CURR_H, - REG_ADC_BAT_CHR_CURR_L, - REG_ADC_IPSOUT_V_H, - REG_ADC_IPSOUT_V_L, - REG_DC_DC_MOD_SEL = 0x80u, - REG_ADC_EN_1, - REG_ADC_EN_2, - REG_ADC_SR_CTRL, - REG_ADC_IN_RANGE, - REG_GPIO1_ADC_IRQ_RISING_TH, - REG_GPIO1_ADC_IRQ_FALLING_TH, - REG_TIMER_CTRL = 0x8au, - REG_VBUS_CTRL_MON_SRP, - REG_OVER_TEMP_SHUTDOWN = 0x8fu, - REG_GPIO0_FEAT_SET, - REG_GPIO_OUT_HIGH_SET, - REG_GPIO1_FEAT_SET, - REG_GPIO2_FEAT_SET, - REG_GPIO_SIG_STATE_SET_MON, - REG_GPIO3_SET, - REG_COULOMB_CNTR_CTRL = 0xb8u, - REG_POWER_MEAS_RES, - NR_REGS -}; - -#define AXP209_CHIP_VERSION_ID (0x01) -#define AXP209_DC_DC2_OUT_V_CTRL_RESET (0x16) -#define AXP209_IRQ_BANK_1_CTRL_RESET (0xd8) - -/* A simple I2C slave which returns values of ID or CNT register. */ -typedef struct AXP209I2CState { - /*< private >*/ - I2CSlave i2c; - /*< public >*/ - uint8_t regs[NR_REGS]; /* peripheral registers */ - uint8_t ptr; /* current register index */ - uint8_t count; /* counter used for tx/rx */ -} AXP209I2CState; - -/* Reset all counters and load ID register */ -static void axp209_reset_enter(Object *obj, ResetType type) -{ - AXP209I2CState *s = AXP209(obj); - - memset(s->regs, 0, NR_REGS); - s->ptr = 0; - s->count = 0; - s->regs[REG_CHIP_VERSION] = AXP209_CHIP_VERSION_ID; - s->regs[REG_DC_DC2_OUT_V_CTRL] = AXP209_DC_DC2_OUT_V_CTRL_RESET; - s->regs[REG_IRQ_BANK_1_CTRL] = AXP209_IRQ_BANK_1_CTRL_RESET; -} - -/* Handle events from master. */ -static int axp209_event(I2CSlave *i2c, enum i2c_event event) -{ - AXP209I2CState *s = AXP209(i2c); - - s->count = 0; - - return 0; -} - -/* Called when master requests read */ -static uint8_t axp209_rx(I2CSlave *i2c) -{ - AXP209I2CState *s = AXP209(i2c); - uint8_t ret = 0xff; - - if (s->ptr < NR_REGS) { - ret = s->regs[s->ptr++]; - } - - trace_axp209_rx(s->ptr - 1, ret); - - return ret; -} - -/* - * Called when master sends write. - * Update ptr with byte 0, then perform write with second byte. - */ -static int axp209_tx(I2CSlave *i2c, uint8_t data) -{ - AXP209I2CState *s = AXP209(i2c); - - if (s->count == 0) { - /* Store register address */ - s->ptr = data; - s->count++; - trace_axp209_select(data); - } else { - trace_axp209_tx(s->ptr, data); - if (s->ptr == REG_DC_DC2_OUT_V_CTRL) { - s->regs[s->ptr++] = data; - } - } - - return 0; -} - -static const VMStateDescription vmstate_axp209 = { - .name = TYPE_AXP209_PMU, - .version_id = 1, - .fields = (VMStateField[]) { - VMSTATE_UINT8_ARRAY(regs, AXP209I2CState, NR_REGS), - VMSTATE_UINT8(count, AXP209I2CState), - VMSTATE_UINT8(ptr, AXP209I2CState), - VMSTATE_END_OF_LIST() - } -}; - -static void axp209_class_init(ObjectClass *oc, void *data) -{ - DeviceClass *dc = DEVICE_CLASS(oc); - I2CSlaveClass *isc = I2C_SLAVE_CLASS(oc); - ResettableClass *rc = RESETTABLE_CLASS(oc); - - rc->phases.enter = axp209_reset_enter; - dc->vmsd = &vmstate_axp209; - isc->event = axp209_event; - isc->recv = axp209_rx; - isc->send = axp209_tx; -} - -static const TypeInfo axp209_info = { - .name = TYPE_AXP209_PMU, - .parent = TYPE_I2C_SLAVE, - .instance_size = sizeof(AXP209I2CState), - .class_init = axp209_class_init -}; - -static void axp209_register_devices(void) -{ - type_register_static(&axp209_info); -} - -type_init(axp209_register_devices); diff --git a/hw/misc/axp2xx.c b/hw/misc/axp2xx.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/misc/axp2xx.c @@ -XXX,XX +XXX,XX @@ +/* + * AXP-2XX PMU Emulation, supported lists: + * AXP209 + * AXP221 + * + * Copyright (C) 2022 Strahinja Jankovic <strahinja.p.jankovic@gmail.com> + * Copyright (C) 2023 qianfan Zhao <qianfanguijin@163.com> + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * SPDX-License-Identifier: MIT + */ + +#include "qemu/osdep.h" +#include "qemu/log.h" +#include "qom/object.h" +#include "trace.h" +#include "hw/i2c/i2c.h" +#include "migration/vmstate.h" + +#define TYPE_AXP2XX "axp2xx_pmu" +#define TYPE_AXP209_PMU "axp209_pmu" +#define TYPE_AXP221_PMU "axp221_pmu" + +OBJECT_DECLARE_TYPE(AXP2xxI2CState, AXP2xxClass, AXP2XX) + +#define NR_REGS (0xff) + +/* A simple I2C slave which returns values of ID or CNT register. */ +typedef struct AXP2xxI2CState { + /*< private >*/ + I2CSlave i2c; + /*< public >*/ + uint8_t regs[NR_REGS]; /* peripheral registers */ + uint8_t ptr; /* current register index */ + uint8_t count; /* counter used for tx/rx */ +} AXP2xxI2CState; + +typedef struct AXP2xxClass { + /*< private >*/ + I2CSlaveClass parent_class; + /*< public >*/ + void (*reset_enter)(AXP2xxI2CState *s, ResetType type); +} AXP2xxClass; + +#define AXP209_CHIP_VERSION_ID (0x01) +#define AXP209_DC_DC2_OUT_V_CTRL_RESET (0x16) + +/* Reset all counters and load ID register */ +static void axp209_reset_enter(AXP2xxI2CState *s, ResetType type) +{ + memset(s->regs, 0, NR_REGS); + s->ptr = 0; + s->count = 0; + + s->regs[0x03] = AXP209_CHIP_VERSION_ID; + s->regs[0x23] = AXP209_DC_DC2_OUT_V_CTRL_RESET; + + s->regs[0x30] = 0x60; + s->regs[0x32] = 0x46; + s->regs[0x34] = 0x41; + s->regs[0x35] = 0x22; + s->regs[0x36] = 0x5d; + s->regs[0x37] = 0x08; + s->regs[0x38] = 0xa5; + s->regs[0x39] = 0x1f; + s->regs[0x3a] = 0x68; + s->regs[0x3b] = 0x5f; + s->regs[0x3c] = 0xfc; + s->regs[0x3d] = 0x16; + s->regs[0x40] = 0xd8; + s->regs[0x42] = 0xff; + s->regs[0x43] = 0x3b; + s->regs[0x80] = 0xe0; + s->regs[0x82] = 0x83; + s->regs[0x83] = 0x80; + s->regs[0x84] = 0x32; + s->regs[0x86] = 0xff; + s->regs[0x90] = 0x07; + s->regs[0x91] = 0xa0; + s->regs[0x92] = 0x07; + s->regs[0x93] = 0x07; +} + +#define AXP221_PWR_STATUS_ACIN_PRESENT BIT(7) +#define AXP221_PWR_STATUS_ACIN_AVAIL BIT(6) +#define AXP221_PWR_STATUS_VBUS_PRESENT BIT(5) +#define AXP221_PWR_STATUS_VBUS_USED BIT(4) +#define AXP221_PWR_STATUS_BAT_CHARGING BIT(2) +#define AXP221_PWR_STATUS_ACIN_VBUS_POWERED BIT(1) + +/* Reset all counters and load ID register */ +static void axp221_reset_enter(AXP2xxI2CState *s, ResetType type) +{ + memset(s->regs, 0, NR_REGS); + s->ptr = 0; + s->count = 0; + + /* input power status register */ + s->regs[0x00] = AXP221_PWR_STATUS_ACIN_PRESENT + | AXP221_PWR_STATUS_ACIN_AVAIL + | AXP221_PWR_STATUS_ACIN_VBUS_POWERED; + + s->regs[0x01] = 0x00; /* no battery is connected */ + + /* + * CHIPID register, no documented on datasheet, but it is checked in + * u-boot spl. I had read it from AXP221s and got 0x06 value. + * So leave 06h here. + */ + s->regs[0x03] = 0x06; + + s->regs[0x10] = 0xbf; + s->regs[0x13] = 0x01; + s->regs[0x30] = 0x60; + s->regs[0x31] = 0x03; + s->regs[0x32] = 0x43; + s->regs[0x33] = 0xc6; + s->regs[0x34] = 0x45; + s->regs[0x35] = 0x0e; + s->regs[0x36] = 0x5d; + s->regs[0x37] = 0x08; + s->regs[0x38] = 0xa5; + s->regs[0x39] = 0x1f; + s->regs[0x3c] = 0xfc; + s->regs[0x3d] = 0x16; + s->regs[0x80] = 0x80; + s->regs[0x82] = 0xe0; + s->regs[0x84] = 0x32; + s->regs[0x8f] = 0x01; + + s->regs[0x90] = 0x07; + s->regs[0x91] = 0x1f; + s->regs[0x92] = 0x07; + s->regs[0x93] = 0x1f; + + s->regs[0x40] = 0xd8; + s->regs[0x41] = 0xff; + s->regs[0x42] = 0x03; + s->regs[0x43] = 0x03; + + s->regs[0xb8] = 0xc0; + s->regs[0xb9] = 0x64; + s->regs[0xe6] = 0xa0; +} + +static void axp2xx_reset_enter(Object *obj, ResetType type) +{ + AXP2xxI2CState *s = AXP2XX(obj); + AXP2xxClass *sc = AXP2XX_GET_CLASS(s); + + sc->reset_enter(s, type); +} + +/* Handle events from master. */ +static int axp2xx_event(I2CSlave *i2c, enum i2c_event event) +{ + AXP2xxI2CState *s = AXP2XX(i2c); + + s->count = 0; + + return 0; +} + +/* Called when master requests read */ +static uint8_t axp2xx_rx(I2CSlave *i2c) +{ + AXP2xxI2CState *s = AXP2XX(i2c); + uint8_t ret = 0xff; + + if (s->ptr < NR_REGS) { + ret = s->regs[s->ptr++]; + } + + trace_axp2xx_rx(s->ptr - 1, ret); + + return ret; +} + +/* + * Called when master sends write. + * Update ptr with byte 0, then perform write with second byte. + */ +static int axp2xx_tx(I2CSlave *i2c, uint8_t data) +{ + AXP2xxI2CState *s = AXP2XX(i2c); + + if (s->count == 0) { + /* Store register address */ + s->ptr = data; + s->count++; + trace_axp2xx_select(data); + } else { + trace_axp2xx_tx(s->ptr, data); + s->regs[s->ptr++] = data; + } + + return 0; +} + +static const VMStateDescription vmstate_axp2xx = { + .name = TYPE_AXP2XX, + .version_id = 1, + .fields = (VMStateField[]) { + VMSTATE_UINT8_ARRAY(regs, AXP2xxI2CState, NR_REGS), + VMSTATE_UINT8(ptr, AXP2xxI2CState), + VMSTATE_UINT8(count, AXP2xxI2CState), + VMSTATE_END_OF_LIST() + } +}; + +static void axp2xx_class_init(ObjectClass *oc, void *data) +{ + DeviceClass *dc = DEVICE_CLASS(oc); + I2CSlaveClass *isc = I2C_SLAVE_CLASS(oc); + ResettableClass *rc = RESETTABLE_CLASS(oc); + + rc->phases.enter = axp2xx_reset_enter; + dc->vmsd = &vmstate_axp2xx; + isc->event = axp2xx_event; + isc->recv = axp2xx_rx; + isc->send = axp2xx_tx; +} + +static const TypeInfo axp2xx_info = { + .name = TYPE_AXP2XX, + .parent = TYPE_I2C_SLAVE, + .instance_size = sizeof(AXP2xxI2CState), + .class_size = sizeof(AXP2xxClass), + .class_init = axp2xx_class_init, + .abstract = true, +}; + +static void axp209_class_init(ObjectClass *oc, void *data) +{ + AXP2xxClass *sc = AXP2XX_CLASS(oc); + + sc->reset_enter = axp209_reset_enter; +} + +static const TypeInfo axp209_info = { + .name = TYPE_AXP209_PMU, + .parent = TYPE_AXP2XX, + .class_init = axp209_class_init +}; + +static void axp221_class_init(ObjectClass *oc, void *data) +{ + AXP2xxClass *sc = AXP2XX_CLASS(oc); + + sc->reset_enter = axp221_reset_enter; +} + +static const TypeInfo axp221_info = { + .name = TYPE_AXP221_PMU, + .parent = TYPE_AXP2XX, + .class_init = axp221_class_init, +}; + +static void axp2xx_register_devices(void) +{ + type_register_static(&axp2xx_info); + type_register_static(&axp209_info); + type_register_static(&axp221_info); +} + +type_init(axp2xx_register_devices); diff --git a/hw/arm/Kconfig b/hw/arm/Kconfig index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/Kconfig +++ b/hw/arm/Kconfig @@ -XXX,XX +XXX,XX @@ config ALLWINNER_A10 select ALLWINNER_WDT select ALLWINNER_EMAC select ALLWINNER_I2C - select AXP209_PMU + select AXP2XX_PMU select SERIAL select UNIMP @@ -XXX,XX +XXX,XX @@ config ALLWINNER_R40 bool default y if TCG && ARM select ALLWINNER_A10_PIT + select AXP2XX_PMU select SERIAL select ARM_TIMER select ARM_GIC diff --git a/hw/misc/Kconfig b/hw/misc/Kconfig index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/Kconfig +++ b/hw/misc/Kconfig @@ -XXX,XX +XXX,XX @@ config ALLWINNER_A10_CCM config ALLWINNER_A10_DRAMC bool -config AXP209_PMU +config AXP2XX_PMU bool depends on I2C diff --git a/hw/misc/meson.build b/hw/misc/meson.build index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/meson.build +++ b/hw/misc/meson.build @@ -XXX,XX +XXX,XX @@ softmmu_ss.add(when: 'CONFIG_ALLWINNER_H3', if_true: files('allwinner-h3-dramc.c softmmu_ss.add(when: 'CONFIG_ALLWINNER_H3', if_true: files('allwinner-h3-sysctrl.c')) softmmu_ss.add(when: 'CONFIG_ALLWINNER_H3', if_true: files('allwinner-sid.c')) softmmu_ss.add(when: 'CONFIG_ALLWINNER_R40', if_true: files('allwinner-r40-ccu.c')) -softmmu_ss.add(when: 'CONFIG_AXP209_PMU', if_true: files('axp209.c')) +softmmu_ss.add(when: 'CONFIG_AXP2XX_PMU', if_true: files('axp2xx.c')) softmmu_ss.add(when: 'CONFIG_REALVIEW', if_true: files('arm_sysctl.c')) softmmu_ss.add(when: 'CONFIG_NSERIES', if_true: files('cbus.c')) softmmu_ss.add(when: 'CONFIG_ECCMEMCTL', if_true: files('eccmemctl.c')) diff --git a/hw/misc/trace-events b/hw/misc/trace-events index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/trace-events +++ b/hw/misc/trace-events @@ -XXX,XX +XXX,XX @@ allwinner_sid_write(uint64_t offset, uint64_t data, unsigned size) "offset 0x%" avr_power_read(uint8_t value) "power_reduc read value:%u" avr_power_write(uint8_t value) "power_reduc write value:%u" -# axp209.c -axp209_rx(uint8_t reg, uint8_t data) "Read reg 0x%" PRIx8 " : 0x%" PRIx8 -axp209_select(uint8_t reg) "Accessing reg 0x%" PRIx8 -axp209_tx(uint8_t reg, uint8_t data) "Write reg 0x%" PRIx8 " : 0x%" PRIx8 +# axp2xx +axp2xx_rx(uint8_t reg, uint8_t data) "Read reg 0x%" PRIx8 " : 0x%" PRIx8 +axp2xx_select(uint8_t reg) "Accessing reg 0x%" PRIx8 +axp2xx_tx(uint8_t reg, uint8_t data) "Write reg 0x%" PRIx8 " : 0x%" PRIx8 # eccmemctl.c ecc_mem_writel_mer(uint32_t val) "Write memory enable 0x%08x" -- 2.34.1 From: qianfan Zhao <qianfanguijin@163.com> Types of memory that the SDRAM controller supports are DDR2/DDR3 and capacities of up to 2GiB. This commit adds emulation support of the Allwinner R40 SDRAM controller. This driver only support 256M, 512M and 1024M memory now. Signed-off-by: qianfan Zhao <qianfanguijin@163.com> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/allwinner-r40.h | 13 +- include/hw/misc/allwinner-r40-dramc.h | 108 ++++++ hw/arm/allwinner-r40.c | 21 +- hw/arm/bananapi_m2u.c | 7 + hw/misc/allwinner-r40-dramc.c | 513 ++++++++++++++++++++++++++ hw/misc/meson.build | 1 + hw/misc/trace-events | 14 + 7 files changed, 674 insertions(+), 3 deletions(-) create mode 100644 include/hw/misc/allwinner-r40-dramc.h create mode 100644 hw/misc/allwinner-r40-dramc.c diff --git a/include/hw/arm/allwinner-r40.h b/include/hw/arm/allwinner-r40.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/allwinner-r40.h +++ b/include/hw/arm/allwinner-r40.h @@ -XXX,XX +XXX,XX @@ #include "hw/intc/arm_gic.h" #include "hw/sd/allwinner-sdhost.h" #include "hw/misc/allwinner-r40-ccu.h" +#include "hw/misc/allwinner-r40-dramc.h" #include "hw/i2c/allwinner-i2c.h" #include "target/arm/cpu.h" #include "sysemu/block-backend.h" @@ -XXX,XX +XXX,XX @@ enum { AW_R40_DEV_GIC_CPU, AW_R40_DEV_GIC_HYP, AW_R40_DEV_GIC_VCPU, - AW_R40_DEV_SDRAM + AW_R40_DEV_SDRAM, + AW_R40_DEV_DRAMCOM, + AW_R40_DEV_DRAMCTL, + AW_R40_DEV_DRAMPHY, }; #define AW_R40_NUM_CPUS (4) @@ -XXX,XX +XXX,XX @@ struct AwR40State { DeviceState parent_obj; /*< public >*/ + /** Physical base address for start of RAM */ + hwaddr ram_addr; + + /** Total RAM size in megabytes */ + uint32_t ram_size; + ARMCPU cpus[AW_R40_NUM_CPUS]; const hwaddr *memmap; AwA10PITState timer; AwSdHostState mmc[AW_R40_NUM_MMCS]; AwR40ClockCtlState ccu; + AwR40DramCtlState dramc; AWI2CState i2c0; GICState gic; MemoryRegion sram_a1; diff --git a/include/hw/misc/allwinner-r40-dramc.h b/include/hw/misc/allwinner-r40-dramc.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/include/hw/misc/allwinner-r40-dramc.h @@ -XXX,XX +XXX,XX @@ +/* + * Allwinner R40 SDRAM Controller emulation + * + * Copyright (C) 2023 qianfan Zhao <qianfanguijin@163.com> + * + * This program is free software: you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation, either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#ifndef HW_MISC_ALLWINNER_R40_DRAMC_H +#define HW_MISC_ALLWINNER_R40_DRAMC_H + +#include "qom/object.h" +#include "hw/sysbus.h" +#include "exec/hwaddr.h" + +/** + * Constants + * @{ + */ + +/** Highest register address used by DRAMCOM module */ +#define AW_R40_DRAMCOM_REGS_MAXADDR (0x804) + +/** Total number of known DRAMCOM registers */ +#define AW_R40_DRAMCOM_REGS_NUM (AW_R40_DRAMCOM_REGS_MAXADDR / \ + sizeof(uint32_t)) + +/** Highest register address used by DRAMCTL module */ +#define AW_R40_DRAMCTL_REGS_MAXADDR (0x88c) + +/** Total number of known DRAMCTL registers */ +#define AW_R40_DRAMCTL_REGS_NUM (AW_R40_DRAMCTL_REGS_MAXADDR / \ + sizeof(uint32_t)) + +/** Highest register address used by DRAMPHY module */ +#define AW_R40_DRAMPHY_REGS_MAXADDR (0x4) + +/** Total number of known DRAMPHY registers */ +#define AW_R40_DRAMPHY_REGS_NUM (AW_R40_DRAMPHY_REGS_MAXADDR / \ + sizeof(uint32_t)) + +/** @} */ + +/** + * Object model + * @{ + */ + +#define TYPE_AW_R40_DRAMC "allwinner-r40-dramc" +OBJECT_DECLARE_SIMPLE_TYPE(AwR40DramCtlState, AW_R40_DRAMC) + +/** @} */ + +/** + * Allwinner R40 SDRAM Controller object instance state. + */ +struct AwR40DramCtlState { + /*< private >*/ + SysBusDevice parent_obj; + /*< public >*/ + + /** Physical base address for start of RAM */ + hwaddr ram_addr; + + /** Total RAM size in megabytes */ + uint32_t ram_size; + + uint8_t set_row_bits; + uint8_t set_bank_bits; + uint8_t set_col_bits; + + /** + * @name Memory Regions + * @{ + */ + MemoryRegion dramcom_iomem; /**< DRAMCOM module I/O registers */ + MemoryRegion dramctl_iomem; /**< DRAMCTL module I/O registers */ + MemoryRegion dramphy_iomem; /**< DRAMPHY module I/O registers */ + MemoryRegion dram_high; /**< The high 1G dram for dualrank detect */ + MemoryRegion detect_cells; /**< DRAM memory cells for auto detect */ + + /** @} */ + + /** + * @name Hardware Registers + * @{ + */ + + uint32_t dramcom[AW_R40_DRAMCOM_REGS_NUM]; /**< DRAMCOM registers */ + uint32_t dramctl[AW_R40_DRAMCTL_REGS_NUM]; /**< DRAMCTL registers */ + uint32_t dramphy[AW_R40_DRAMPHY_REGS_NUM] ;/**< DRAMPHY registers */ + + /** @} */ + +}; + +#endif /* HW_MISC_ALLWINNER_R40_DRAMC_H */ diff --git a/hw/arm/allwinner-r40.c b/hw/arm/allwinner-r40.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/allwinner-r40.c +++ b/hw/arm/allwinner-r40.c @@ -XXX,XX +XXX,XX @@ #include "hw/loader.h" #include "sysemu/sysemu.h" #include "hw/arm/allwinner-r40.h" +#include "hw/misc/allwinner-r40-dramc.h" /* Memory map */ const hwaddr allwinner_r40_memmap[] = { @@ -XXX,XX +XXX,XX @@ const hwaddr allwinner_r40_memmap[] = { [AW_R40_DEV_UART6] = 0x01c29800, [AW_R40_DEV_UART7] = 0x01c29c00, [AW_R40_DEV_TWI0] = 0x01c2ac00, + [AW_R40_DEV_DRAMCOM] = 0x01c62000, + [AW_R40_DEV_DRAMCTL] = 0x01c63000, + [AW_R40_DEV_DRAMPHY] = 0x01c65000, [AW_R40_DEV_GIC_DIST] = 0x01c81000, [AW_R40_DEV_GIC_CPU] = 0x01c82000, [AW_R40_DEV_GIC_HYP] = 0x01c84000, @@ -XXX,XX +XXX,XX @@ static struct AwR40Unimplemented r40_unimplemented[] = { { "gpu", 0x01c40000, 64 * KiB }, { "gmac", 0x01c50000, 64 * KiB }, { "hstmr", 0x01c60000, 4 * KiB }, - { "dram-com", 0x01c62000, 4 * KiB }, - { "dram-ctl", 0x01c63000, 4 * KiB }, { "tcon-top", 0x01c70000, 4 * KiB }, { "lcd0", 0x01c71000, 4 * KiB }, { "lcd1", 0x01c72000, 4 * KiB }, @@ -XXX,XX +XXX,XX @@ static void allwinner_r40_init(Object *obj) } object_initialize_child(obj, "twi0", &s->i2c0, TYPE_AW_I2C_SUN6I); + + object_initialize_child(obj, "dramc", &s->dramc, TYPE_AW_R40_DRAMC); + object_property_add_alias(obj, "ram-addr", OBJECT(&s->dramc), + "ram-addr"); + object_property_add_alias(obj, "ram-size", OBJECT(&s->dramc), + "ram-size"); } static void allwinner_r40_realize(DeviceState *dev, Error **errp) @@ -XXX,XX +XXX,XX @@ static void allwinner_r40_realize(DeviceState *dev, Error **errp) sysbus_connect_irq(SYS_BUS_DEVICE(&s->i2c0), 0, qdev_get_gpio_in(DEVICE(&s->gic), AW_R40_GIC_SPI_TWI0)); + /* DRAMC */ + sysbus_realize(SYS_BUS_DEVICE(&s->dramc), &error_fatal); + sysbus_mmio_map(SYS_BUS_DEVICE(&s->dramc), 0, + s->memmap[AW_R40_DEV_DRAMCOM]); + sysbus_mmio_map(SYS_BUS_DEVICE(&s->dramc), 1, + s->memmap[AW_R40_DEV_DRAMCTL]); + sysbus_mmio_map(SYS_BUS_DEVICE(&s->dramc), 2, + s->memmap[AW_R40_DEV_DRAMPHY]); + /* Unimplemented devices */ for (i = 0; i < ARRAY_SIZE(r40_unimplemented); i++) { create_unimplemented_device(r40_unimplemented[i].device_name, diff --git a/hw/arm/bananapi_m2u.c b/hw/arm/bananapi_m2u.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/bananapi_m2u.c +++ b/hw/arm/bananapi_m2u.c @@ -XXX,XX +XXX,XX @@ static void bpim2u_init(MachineState *machine) object_property_set_int(OBJECT(r40), "clk1-freq", 24 * 1000 * 1000, &error_abort); + /* DRAMC */ + r40->ram_size = machine->ram_size / MiB; + object_property_set_uint(OBJECT(r40), "ram-addr", + r40->memmap[AW_R40_DEV_SDRAM], &error_abort); + object_property_set_int(OBJECT(r40), "ram-size", + r40->ram_size, &error_abort); + /* Mark R40 object realized */ qdev_realize(DEVICE(r40), NULL, &error_abort); diff --git a/hw/misc/allwinner-r40-dramc.c b/hw/misc/allwinner-r40-dramc.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/misc/allwinner-r40-dramc.c @@ -XXX,XX +XXX,XX @@ +/* + * Allwinner R40 SDRAM Controller emulation + * + * CCopyright (C) 2023 qianfan Zhao <qianfanguijin@163.com> + * + * This program is free software: you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation, either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#include "qemu/osdep.h" +#include "qemu/units.h" +#include "qemu/error-report.h" +#include "hw/sysbus.h" +#include "migration/vmstate.h" +#include "qemu/log.h" +#include "qemu/module.h" +#include "exec/address-spaces.h" +#include "hw/qdev-properties.h" +#include "qapi/error.h" +#include "qemu/bitops.h" +#include "hw/misc/allwinner-r40-dramc.h" +#include "trace.h" + +#define REG_INDEX(offset) (offset / sizeof(uint32_t)) + +/* DRAMCOM register offsets */ +enum { + REG_DRAMCOM_CR = 0x0000, /* Control Register */ +}; + +/* DRAMCOMM register flags */ +enum { + REG_DRAMCOM_CR_DUAL_RANK = (1 << 0), +}; + +/* DRAMCTL register offsets */ +enum { + REG_DRAMCTL_PIR = 0x0000, /* PHY Initialization Register */ + REG_DRAMCTL_PGSR = 0x0010, /* PHY General Status Register */ + REG_DRAMCTL_STATR = 0x0018, /* Status Register */ + REG_DRAMCTL_PGCR = 0x0100, /* PHY general configuration registers */ +}; + +/* DRAMCTL register flags */ +enum { + REG_DRAMCTL_PGSR_INITDONE = (1 << 0), + REG_DRAMCTL_PGSR_READ_TIMEOUT = (1 << 13), + REG_DRAMCTL_PGCR_ENABLE_READ_TIMEOUT = (1 << 25), +}; + +enum { + REG_DRAMCTL_STATR_ACTIVE = (1 << 0), +}; + +#define DRAM_MAX_ROW_BITS 16 +#define DRAM_MAX_COL_BITS 13 /* 8192 */ +#define DRAM_MAX_BANK 3 + +static uint64_t dram_autodetect_cells[DRAM_MAX_ROW_BITS] + [DRAM_MAX_BANK] + [DRAM_MAX_COL_BITS]; +struct VirtualDDRChip { + uint32_t ram_size; + uint8_t bank_bits; + uint8_t row_bits; + uint8_t col_bits; +}; + +/* + * Only power of 2 RAM sizes from 256MiB up to 2048MiB are supported, + * 2GiB memory is not supported due to dual rank feature. + */ +static const struct VirtualDDRChip dummy_ddr_chips[] = { + { + .ram_size = 256, + .bank_bits = 3, + .row_bits = 12, + .col_bits = 13, + }, { + .ram_size = 512, + .bank_bits = 3, + .row_bits = 13, + .col_bits = 13, + }, { + .ram_size = 1024, + .bank_bits = 3, + .row_bits = 14, + .col_bits = 13, + }, { + 0 + } +}; + +static const struct VirtualDDRChip *get_match_ddr(uint32_t ram_size) +{ + const struct VirtualDDRChip *ddr; + + for (ddr = &dummy_ddr_chips[0]; ddr->ram_size; ddr++) { + if (ddr->ram_size == ram_size) { + return ddr; + } + } + + return NULL; +} + +static uint64_t *address_to_autodetect_cells(AwR40DramCtlState *s, + const struct VirtualDDRChip *ddr, + uint32_t offset) +{ + int row_index = 0, bank_index = 0, col_index = 0; + uint32_t row_addr, bank_addr, col_addr; + + row_addr = extract32(offset, s->set_col_bits + s->set_bank_bits, + s->set_row_bits); + bank_addr = extract32(offset, s->set_col_bits, s->set_bank_bits); + col_addr = extract32(offset, 0, s->set_col_bits); + + for (int i = 0; i < ddr->row_bits; i++) { + if (row_addr & BIT(i)) { + row_index = i; + } + } + + for (int i = 0; i < ddr->bank_bits; i++) { + if (bank_addr & BIT(i)) { + bank_index = i; + } + } + + for (int i = 0; i < ddr->col_bits; i++) { + if (col_addr & BIT(i)) { + col_index = i; + } + } + + trace_allwinner_r40_dramc_offset_to_cell(offset, row_index, bank_index, + col_index); + return &dram_autodetect_cells[row_index][bank_index][col_index]; +} + +static void allwinner_r40_dramc_map_rows(AwR40DramCtlState *s, uint8_t row_bits, + uint8_t bank_bits, uint8_t col_bits) +{ + const struct VirtualDDRChip *ddr = get_match_ddr(s->ram_size); + bool enable_detect_cells; + + trace_allwinner_r40_dramc_map_rows(row_bits, bank_bits, col_bits); + + if (!ddr) { + return; + } + + s->set_row_bits = row_bits; + s->set_bank_bits = bank_bits; + s->set_col_bits = col_bits; + + enable_detect_cells = ddr->bank_bits != bank_bits + || ddr->row_bits != row_bits + || ddr->col_bits != col_bits; + + if (enable_detect_cells) { + trace_allwinner_r40_dramc_detect_cells_enable(); + } else { + trace_allwinner_r40_dramc_detect_cells_disable(); + } + + memory_region_set_enabled(&s->detect_cells, enable_detect_cells); +} + +static uint64_t allwinner_r40_dramcom_read(void *opaque, hwaddr offset, + unsigned size) +{ + const AwR40DramCtlState *s = AW_R40_DRAMC(opaque); + const uint32_t idx = REG_INDEX(offset); + + if (idx >= AW_R40_DRAMCOM_REGS_NUM) { + qemu_log_mask(LOG_GUEST_ERROR, "%s: out-of-bounds offset 0x%04x\n", + __func__, (uint32_t)offset); + return 0; + } + + trace_allwinner_r40_dramcom_read(offset, s->dramcom[idx], size); + return s->dramcom[idx]; +} + +static void allwinner_r40_dramcom_write(void *opaque, hwaddr offset, + uint64_t val, unsigned size) +{ + AwR40DramCtlState *s = AW_R40_DRAMC(opaque); + const uint32_t idx = REG_INDEX(offset); + + trace_allwinner_r40_dramcom_write(offset, val, size); + + if (idx >= AW_R40_DRAMCOM_REGS_NUM) { + qemu_log_mask(LOG_GUEST_ERROR, "%s: out-of-bounds offset 0x%04x\n", + __func__, (uint32_t)offset); + return; + } + + switch (offset) { + case REG_DRAMCOM_CR: /* Control Register */ + if (!(val & REG_DRAMCOM_CR_DUAL_RANK)) { + allwinner_r40_dramc_map_rows(s, ((val >> 4) & 0xf) + 1, + ((val >> 2) & 0x1) + 2, + (((val >> 8) & 0xf) + 3)); + } + break; + }; + + s->dramcom[idx] = (uint32_t) val; +} + +static uint64_t allwinner_r40_dramctl_read(void *opaque, hwaddr offset, + unsigned size) +{ + const AwR40DramCtlState *s = AW_R40_DRAMC(opaque); + const uint32_t idx = REG_INDEX(offset); + + if (idx >= AW_R40_DRAMCTL_REGS_NUM) { + qemu_log_mask(LOG_GUEST_ERROR, "%s: out-of-bounds offset 0x%04x\n", + __func__, (uint32_t)offset); + return 0; + } + + trace_allwinner_r40_dramctl_read(offset, s->dramctl[idx], size); + return s->dramctl[idx]; +} + +static void allwinner_r40_dramctl_write(void *opaque, hwaddr offset, + uint64_t val, unsigned size) +{ + AwR40DramCtlState *s = AW_R40_DRAMC(opaque); + const uint32_t idx = REG_INDEX(offset); + + trace_allwinner_r40_dramctl_write(offset, val, size); + + if (idx >= AW_R40_DRAMCTL_REGS_NUM) { + qemu_log_mask(LOG_GUEST_ERROR, "%s: out-of-bounds offset 0x%04x\n", + __func__, (uint32_t)offset); + return; + } + + switch (offset) { + case REG_DRAMCTL_PIR: /* PHY Initialization Register */ + s->dramctl[REG_INDEX(REG_DRAMCTL_PGSR)] |= REG_DRAMCTL_PGSR_INITDONE; + s->dramctl[REG_INDEX(REG_DRAMCTL_STATR)] |= REG_DRAMCTL_STATR_ACTIVE; + break; + } + + s->dramctl[idx] = (uint32_t) val; +} + +static uint64_t allwinner_r40_dramphy_read(void *opaque, hwaddr offset, + unsigned size) +{ + const AwR40DramCtlState *s = AW_R40_DRAMC(opaque); + const uint32_t idx = REG_INDEX(offset); + + if (idx >= AW_R40_DRAMPHY_REGS_NUM) { + qemu_log_mask(LOG_GUEST_ERROR, "%s: out-of-bounds offset 0x%04x\n", + __func__, (uint32_t)offset); + return 0; + } + + trace_allwinner_r40_dramphy_read(offset, s->dramphy[idx], size); + return s->dramphy[idx]; +} + +static void allwinner_r40_dramphy_write(void *opaque, hwaddr offset, + uint64_t val, unsigned size) +{ + AwR40DramCtlState *s = AW_R40_DRAMC(opaque); + const uint32_t idx = REG_INDEX(offset); + + trace_allwinner_r40_dramphy_write(offset, val, size); + + if (idx >= AW_R40_DRAMPHY_REGS_NUM) { + qemu_log_mask(LOG_GUEST_ERROR, "%s: out-of-bounds offset 0x%04x\n", + __func__, (uint32_t)offset); + return; + } + + s->dramphy[idx] = (uint32_t) val; +} + +static const MemoryRegionOps allwinner_r40_dramcom_ops = { + .read = allwinner_r40_dramcom_read, + .write = allwinner_r40_dramcom_write, + .endianness = DEVICE_NATIVE_ENDIAN, + .valid = { + .min_access_size = 4, + .max_access_size = 4, + }, + .impl.min_access_size = 4, +}; + +static const MemoryRegionOps allwinner_r40_dramctl_ops = { + .read = allwinner_r40_dramctl_read, + .write = allwinner_r40_dramctl_write, + .endianness = DEVICE_NATIVE_ENDIAN, + .valid = { + .min_access_size = 4, + .max_access_size = 4, + }, + .impl.min_access_size = 4, +}; + +static const MemoryRegionOps allwinner_r40_dramphy_ops = { + .read = allwinner_r40_dramphy_read, + .write = allwinner_r40_dramphy_write, + .endianness = DEVICE_NATIVE_ENDIAN, + .valid = { + .min_access_size = 4, + .max_access_size = 4, + }, + .impl.min_access_size = 4, +}; + +static uint64_t allwinner_r40_detect_read(void *opaque, hwaddr offset, + unsigned size) +{ + AwR40DramCtlState *s = AW_R40_DRAMC(opaque); + const struct VirtualDDRChip *ddr = get_match_ddr(s->ram_size); + uint64_t data = 0; + + if (ddr) { + data = *address_to_autodetect_cells(s, ddr, (uint32_t)offset); + } + + trace_allwinner_r40_dramc_detect_cell_read(offset, data); + return data; +} + +static void allwinner_r40_detect_write(void *opaque, hwaddr offset, + uint64_t data, unsigned size) +{ + AwR40DramCtlState *s = AW_R40_DRAMC(opaque); + const struct VirtualDDRChip *ddr = get_match_ddr(s->ram_size); + + if (ddr) { + uint64_t *cell = address_to_autodetect_cells(s, ddr, (uint32_t)offset); + trace_allwinner_r40_dramc_detect_cell_write(offset, data); + *cell = data; + } +} + +static const MemoryRegionOps allwinner_r40_detect_ops = { + .read = allwinner_r40_detect_read, + .write = allwinner_r40_detect_write, + .endianness = DEVICE_NATIVE_ENDIAN, + .valid = { + .min_access_size = 4, + .max_access_size = 4, + }, + .impl.min_access_size = 4, +}; + +/* + * mctl_r40_detect_rank_count in u-boot will write the high 1G of DDR + * to detect wether the board support dual_rank or not. Create a virtual memory + * if the board's ram_size less or equal than 1G, and set read time out flag of + * REG_DRAMCTL_PGSR when the user touch this high dram. + */ +static uint64_t allwinner_r40_dualrank_detect_read(void *opaque, hwaddr offset, + unsigned size) +{ + AwR40DramCtlState *s = AW_R40_DRAMC(opaque); + uint32_t reg; + + reg = s->dramctl[REG_INDEX(REG_DRAMCTL_PGCR)]; + if (reg & REG_DRAMCTL_PGCR_ENABLE_READ_TIMEOUT) { /* Enable read time out */ + /* + * this driver only support one rank, mark READ_TIMEOUT when try + * read the second rank. + */ + s->dramctl[REG_INDEX(REG_DRAMCTL_PGSR)] + |= REG_DRAMCTL_PGSR_READ_TIMEOUT; + } + + return 0; +} + +static const MemoryRegionOps allwinner_r40_dualrank_detect_ops = { + .read = allwinner_r40_dualrank_detect_read, + .endianness = DEVICE_NATIVE_ENDIAN, + .valid = { + .min_access_size = 4, + .max_access_size = 4, + }, + .impl.min_access_size = 4, +}; + +static void allwinner_r40_dramc_reset(DeviceState *dev) +{ + AwR40DramCtlState *s = AW_R40_DRAMC(dev); + + /* Set default values for registers */ + memset(&s->dramcom, 0, sizeof(s->dramcom)); + memset(&s->dramctl, 0, sizeof(s->dramctl)); + memset(&s->dramphy, 0, sizeof(s->dramphy)); +} + +static void allwinner_r40_dramc_realize(DeviceState *dev, Error **errp) +{ + AwR40DramCtlState *s = AW_R40_DRAMC(dev); + + if (!get_match_ddr(s->ram_size)) { + error_report("%s: ram-size %u MiB is not supported", + __func__, s->ram_size); + exit(1); + } + + /* detect_cells */ + sysbus_mmio_map_overlap(SYS_BUS_DEVICE(s), 3, s->ram_addr, 10); + memory_region_set_enabled(&s->detect_cells, false); + + /* + * We only support DRAM size up to 1G now, so prepare a high memory page + * after 1G for dualrank detect. index = 4 + */ + memory_region_init_io(&s->dram_high, OBJECT(s), + &allwinner_r40_dualrank_detect_ops, s, + "DRAMHIGH", KiB); + sysbus_init_mmio(SYS_BUS_DEVICE(s), &s->dram_high); + sysbus_mmio_map(SYS_BUS_DEVICE(s), 4, s->ram_addr + GiB); +} + +static void allwinner_r40_dramc_init(Object *obj) +{ + SysBusDevice *sbd = SYS_BUS_DEVICE(obj); + AwR40DramCtlState *s = AW_R40_DRAMC(obj); + + /* DRAMCOM registers, index 0 */ + memory_region_init_io(&s->dramcom_iomem, OBJECT(s), + &allwinner_r40_dramcom_ops, s, + "DRAMCOM", 4 * KiB); + sysbus_init_mmio(sbd, &s->dramcom_iomem); + + /* DRAMCTL registers, index 1 */ + memory_region_init_io(&s->dramctl_iomem, OBJECT(s), + &allwinner_r40_dramctl_ops, s, + "DRAMCTL", 4 * KiB); + sysbus_init_mmio(sbd, &s->dramctl_iomem); + + /* DRAMPHY registers. index 2 */ + memory_region_init_io(&s->dramphy_iomem, OBJECT(s), + &allwinner_r40_dramphy_ops, s, + "DRAMPHY", 4 * KiB); + sysbus_init_mmio(sbd, &s->dramphy_iomem); + + /* R40 support max 2G memory but we only support up to 1G now. index 3 */ + memory_region_init_io(&s->detect_cells, OBJECT(s), + &allwinner_r40_detect_ops, s, + "DRAMCELLS", 1 * GiB); + sysbus_init_mmio(sbd, &s->detect_cells); +} + +static Property allwinner_r40_dramc_properties[] = { + DEFINE_PROP_UINT64("ram-addr", AwR40DramCtlState, ram_addr, 0x0), + DEFINE_PROP_UINT32("ram-size", AwR40DramCtlState, ram_size, 256), /* MiB */ + DEFINE_PROP_END_OF_LIST() +}; + +static const VMStateDescription allwinner_r40_dramc_vmstate = { + .name = "allwinner-r40-dramc", + .version_id = 1, + .minimum_version_id = 1, + .fields = (VMStateField[]) { + VMSTATE_UINT32_ARRAY(dramcom, AwR40DramCtlState, + AW_R40_DRAMCOM_REGS_NUM), + VMSTATE_UINT32_ARRAY(dramctl, AwR40DramCtlState, + AW_R40_DRAMCTL_REGS_NUM), + VMSTATE_UINT32_ARRAY(dramphy, AwR40DramCtlState, + AW_R40_DRAMPHY_REGS_NUM), + VMSTATE_END_OF_LIST() + } +}; + +static void allwinner_r40_dramc_class_init(ObjectClass *klass, void *data) +{ + DeviceClass *dc = DEVICE_CLASS(klass); + + dc->reset = allwinner_r40_dramc_reset; + dc->vmsd = &allwinner_r40_dramc_vmstate; + dc->realize = allwinner_r40_dramc_realize; + device_class_set_props(dc, allwinner_r40_dramc_properties); +} + +static const TypeInfo allwinner_r40_dramc_info = { + .name = TYPE_AW_R40_DRAMC, + .parent = TYPE_SYS_BUS_DEVICE, + .instance_init = allwinner_r40_dramc_init, + .instance_size = sizeof(AwR40DramCtlState), + .class_init = allwinner_r40_dramc_class_init, +}; + +static void allwinner_r40_dramc_register(void) +{ + type_register_static(&allwinner_r40_dramc_info); +} + +type_init(allwinner_r40_dramc_register) diff --git a/hw/misc/meson.build b/hw/misc/meson.build index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/meson.build +++ b/hw/misc/meson.build @@ -XXX,XX +XXX,XX @@ softmmu_ss.add(when: 'CONFIG_ALLWINNER_H3', if_true: files('allwinner-h3-dramc.c softmmu_ss.add(when: 'CONFIG_ALLWINNER_H3', if_true: files('allwinner-h3-sysctrl.c')) softmmu_ss.add(when: 'CONFIG_ALLWINNER_H3', if_true: files('allwinner-sid.c')) softmmu_ss.add(when: 'CONFIG_ALLWINNER_R40', if_true: files('allwinner-r40-ccu.c')) +softmmu_ss.add(when: 'CONFIG_ALLWINNER_R40', if_true: files('allwinner-r40-dramc.c')) softmmu_ss.add(when: 'CONFIG_AXP2XX_PMU', if_true: files('axp2xx.c')) softmmu_ss.add(when: 'CONFIG_REALVIEW', if_true: files('arm_sysctl.c')) softmmu_ss.add(when: 'CONFIG_NSERIES', if_true: files('cbus.c')) diff --git a/hw/misc/trace-events b/hw/misc/trace-events index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/trace-events +++ b/hw/misc/trace-events @@ -XXX,XX +XXX,XX @@ allwinner_h3_dramctl_write(uint64_t offset, uint64_t data, unsigned size) "Write allwinner_h3_dramphy_read(uint64_t offset, uint64_t data, unsigned size) "Read: offset 0x%" PRIx64 " data 0x%" PRIx64 " size %" PRIu32 allwinner_h3_dramphy_write(uint64_t offset, uint64_t data, unsigned size) "write: offset 0x%" PRIx64 " data 0x%" PRIx64 " size %" PRIu32 +# allwinner-r40-dramc.c +allwinner_r40_dramc_detect_cells_disable(void) "Disable detect cells" +allwinner_r40_dramc_detect_cells_enable(void) "Enable detect cells" +allwinner_r40_dramc_map_rows(uint8_t row_bits, uint8_t bank_bits, uint8_t col_bits) "DRAM layout: row_bits %d, bank_bits %d, col_bits %d" +allwinner_r40_dramc_offset_to_cell(uint64_t offset, int row, int bank, int col) "offset 0x%" PRIx64 " row %d bank %d col %d" +allwinner_r40_dramc_detect_cell_write(uint64_t offset, uint64_t data) "offset 0x%" PRIx64 " data 0x%" PRIx64 "" +allwinner_r40_dramc_detect_cell_read(uint64_t offset, uint64_t data) "offset 0x%" PRIx64 " data 0x%" PRIx64 "" +allwinner_r40_dramcom_read(uint64_t offset, uint64_t data, unsigned size) "Read: offset 0x%" PRIx64 " data 0x%" PRIx64 " size %" PRIu32 +allwinner_r40_dramcom_write(uint64_t offset, uint64_t data, unsigned size) "Write: offset 0x%" PRIx64 " data 0x%" PRIx64 " size %" PRIu32 +allwinner_r40_dramctl_read(uint64_t offset, uint64_t data, unsigned size) "Read: offset 0x%" PRIx64 " data 0x%" PRIx64 " size %" PRIu32 +allwinner_r40_dramctl_write(uint64_t offset, uint64_t data, unsigned size) "Write: offset 0x%" PRIx64 " data 0x%" PRIx64 " size %" PRIu32 +allwinner_r40_dramphy_read(uint64_t offset, uint64_t data, unsigned size) "Read: offset 0x%" PRIx64 " data 0x%" PRIx64 " size %" PRIu32 +allwinner_r40_dramphy_write(uint64_t offset, uint64_t data, unsigned size) "write: offset 0x%" PRIx64 " data 0x%" PRIx64 " size %" PRIu32 + # allwinner-sid.c allwinner_sid_read(uint64_t offset, uint64_t data, unsigned size) "offset 0x%" PRIx64 " data 0x%" PRIx64 " size %" PRIu32 allwinner_sid_write(uint64_t offset, uint64_t data, unsigned size) "offset 0x%" PRIx64 " data 0x%" PRIx64 " size %" PRIu32 -- 2.34.1 From: qianfan Zhao <qianfanguijin@163.com> A64's sd register was similar to H3, and it introduced a new register named SAMP_DL_REG location at 0x144. The dma descriptor buffer size of mmc2 is only 8K and the other mmc controllers has 64K. Also fix allwinner-r40's mmc controller type. Signed-off-by: qianfan Zhao <qianfanguijin@163.com> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/sd/allwinner-sdhost.h | 9 ++++ hw/arm/allwinner-r40.c | 2 +- hw/sd/allwinner-sdhost.c | 72 ++++++++++++++++++++++++++++++-- 3 files changed, 79 insertions(+), 4 deletions(-) diff --git a/include/hw/sd/allwinner-sdhost.h b/include/hw/sd/allwinner-sdhost.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/sd/allwinner-sdhost.h +++ b/include/hw/sd/allwinner-sdhost.h @@ -XXX,XX +XXX,XX @@ /** Allwinner sun5i family and newer (A13, H2+, H3, etc) */ #define TYPE_AW_SDHOST_SUN5I TYPE_AW_SDHOST "-sun5i" +/** Allwinner sun50i-a64 */ +#define TYPE_AW_SDHOST_SUN50I_A64 TYPE_AW_SDHOST "-sun50i-a64" + +/** Allwinner sun50i-a64 emmc */ +#define TYPE_AW_SDHOST_SUN50I_A64_EMMC TYPE_AW_SDHOST "-sun50i-a64-emmc" + /** @} */ /** @@ -XXX,XX +XXX,XX @@ struct AwSdHostState { uint32_t startbit_detect; /**< eMMC DDR Start Bit Detection Control */ uint32_t response_crc; /**< Response CRC */ uint32_t data_crc[8]; /**< Data CRC */ + uint32_t sample_delay; /**< Sample delay control */ uint32_t status_crc; /**< Status CRC */ /** @} */ @@ -XXX,XX +XXX,XX @@ struct AwSdHostClass { size_t max_desc_size; bool is_sun4i; + /** does the IP block support autocalibration? */ + bool can_calibrate; }; #endif /* HW_SD_ALLWINNER_SDHOST_H */ diff --git a/hw/arm/allwinner-r40.c b/hw/arm/allwinner-r40.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/allwinner-r40.c +++ b/hw/arm/allwinner-r40.c @@ -XXX,XX +XXX,XX @@ static void allwinner_r40_init(Object *obj) for (int i = 0; i < AW_R40_NUM_MMCS; i++) { object_initialize_child(obj, mmc_names[i], &s->mmc[i], - TYPE_AW_SDHOST_SUN5I); + TYPE_AW_SDHOST_SUN50I_A64); } object_initialize_child(obj, "twi0", &s->i2c0, TYPE_AW_I2C_SUN6I); diff --git a/hw/sd/allwinner-sdhost.c b/hw/sd/allwinner-sdhost.c index XXXXXXX..XXXXXXX 100644 --- a/hw/sd/allwinner-sdhost.c +++ b/hw/sd/allwinner-sdhost.c @@ -XXX,XX +XXX,XX @@ enum { REG_SD_DATA1_CRC = 0x12C, /* CRC Data 1 from card/eMMC */ REG_SD_DATA0_CRC = 0x130, /* CRC Data 0 from card/eMMC */ REG_SD_CRC_STA = 0x134, /* CRC status from card/eMMC during write */ + REG_SD_SAMP_DL = 0x144, /* Sample Delay Control (sun50i-a64) */ REG_SD_FIFO = 0x200, /* Read/Write FIFO */ }; @@ -XXX,XX +XXX,XX @@ enum { REG_SD_RES_CRC_RST = 0x0, REG_SD_DATA_CRC_RST = 0x0, REG_SD_CRC_STA_RST = 0x0, + REG_SD_SAMPLE_DL_RST = 0x00002000, REG_SD_FIFO_RST = 0x0, }; @@ -XXX,XX +XXX,XX @@ static uint64_t allwinner_sdhost_read(void *opaque, hwaddr offset, { AwSdHostState *s = AW_SDHOST(opaque); AwSdHostClass *sc = AW_SDHOST_GET_CLASS(s); + bool out_of_bounds = false; uint32_t res = 0; switch (offset) { @@ -XXX,XX +XXX,XX @@ static uint64_t allwinner_sdhost_read(void *opaque, hwaddr offset, case REG_SD_FIFO: /* Read/Write FIFO */ res = allwinner_sdhost_fifo_read(s); break; + case REG_SD_SAMP_DL: /* Sample Delay */ + if (sc->can_calibrate) { + res = s->sample_delay; + } else { + out_of_bounds = true; + } + break; default: - qemu_log_mask(LOG_GUEST_ERROR, "%s: out-of-bounds offset %" - HWADDR_PRIx"\n", __func__, offset); + out_of_bounds = true; res = 0; break; } + if (out_of_bounds) { + qemu_log_mask(LOG_GUEST_ERROR, "%s: out-of-bounds offset %" + HWADDR_PRIx"\n", __func__, offset); + } + trace_allwinner_sdhost_read(offset, res, size); return res; } @@ -XXX,XX +XXX,XX @@ static void allwinner_sdhost_write(void *opaque, hwaddr offset, { AwSdHostState *s = AW_SDHOST(opaque); AwSdHostClass *sc = AW_SDHOST_GET_CLASS(s); + bool out_of_bounds = false; trace_allwinner_sdhost_write(offset, value, size); @@ -XXX,XX +XXX,XX @@ static void allwinner_sdhost_write(void *opaque, hwaddr offset, case REG_SD_DATA0_CRC: /* CRC Data 0 from card/eMMC */ case REG_SD_CRC_STA: /* CRC status from card/eMMC in write operation */ break; + case REG_SD_SAMP_DL: /* Sample delay control */ + if (sc->can_calibrate) { + s->sample_delay = value; + } else { + out_of_bounds = true; + } + break; default: + out_of_bounds = true; + break; + } + + if (out_of_bounds) { qemu_log_mask(LOG_GUEST_ERROR, "%s: out-of-bounds offset %" HWADDR_PRIx"\n", __func__, offset); - break; } } @@ -XXX,XX +XXX,XX @@ static const VMStateDescription vmstate_allwinner_sdhost = { VMSTATE_UINT32(response_crc, AwSdHostState), VMSTATE_UINT32_ARRAY(data_crc, AwSdHostState, 8), VMSTATE_UINT32(status_crc, AwSdHostState), + VMSTATE_UINT32(sample_delay, AwSdHostState), VMSTATE_END_OF_LIST() } }; @@ -XXX,XX +XXX,XX @@ static void allwinner_sdhost_realize(DeviceState *dev, Error **errp) static void allwinner_sdhost_reset(DeviceState *dev) { AwSdHostState *s = AW_SDHOST(dev); + AwSdHostClass *sc = AW_SDHOST_GET_CLASS(s); s->global_ctl = REG_SD_GCTL_RST; s->clock_ctl = REG_SD_CKCR_RST; @@ -XXX,XX +XXX,XX @@ static void allwinner_sdhost_reset(DeviceState *dev) } s->status_crc = REG_SD_CRC_STA_RST; + + if (sc->can_calibrate) { + s->sample_delay = REG_SD_SAMPLE_DL_RST; + } } static void allwinner_sdhost_bus_class_init(ObjectClass *klass, void *data) @@ -XXX,XX +XXX,XX @@ static void allwinner_sdhost_sun4i_class_init(ObjectClass *klass, void *data) AwSdHostClass *sc = AW_SDHOST_CLASS(klass); sc->max_desc_size = 8 * KiB; sc->is_sun4i = true; + sc->can_calibrate = false; } static void allwinner_sdhost_sun5i_class_init(ObjectClass *klass, void *data) @@ -XXX,XX +XXX,XX @@ static void allwinner_sdhost_sun5i_class_init(ObjectClass *klass, void *data) AwSdHostClass *sc = AW_SDHOST_CLASS(klass); sc->max_desc_size = 64 * KiB; sc->is_sun4i = false; + sc->can_calibrate = false; +} + +static void allwinner_sdhost_sun50i_a64_class_init(ObjectClass *klass, + void *data) +{ + AwSdHostClass *sc = AW_SDHOST_CLASS(klass); + sc->max_desc_size = 64 * KiB; + sc->is_sun4i = false; + sc->can_calibrate = true; +} + +static void allwinner_sdhost_sun50i_a64_emmc_class_init(ObjectClass *klass, + void *data) +{ + AwSdHostClass *sc = AW_SDHOST_CLASS(klass); + sc->max_desc_size = 8 * KiB; + sc->is_sun4i = false; + sc->can_calibrate = true; } static const TypeInfo allwinner_sdhost_info = { @@ -XXX,XX +XXX,XX @@ static const TypeInfo allwinner_sdhost_sun5i_info = { .class_init = allwinner_sdhost_sun5i_class_init, }; +static const TypeInfo allwinner_sdhost_sun50i_a64_info = { + .name = TYPE_AW_SDHOST_SUN50I_A64, + .parent = TYPE_AW_SDHOST, + .class_init = allwinner_sdhost_sun50i_a64_class_init, +}; + +static const TypeInfo allwinner_sdhost_sun50i_a64_emmc_info = { + .name = TYPE_AW_SDHOST_SUN50I_A64_EMMC, + .parent = TYPE_AW_SDHOST, + .class_init = allwinner_sdhost_sun50i_a64_emmc_class_init, +}; + static const TypeInfo allwinner_sdhost_bus_info = { .name = TYPE_AW_SDHOST_BUS, .parent = TYPE_SD_BUS, @@ -XXX,XX +XXX,XX @@ static void allwinner_sdhost_register_types(void) type_register_static(&allwinner_sdhost_info); type_register_static(&allwinner_sdhost_sun4i_info); type_register_static(&allwinner_sdhost_sun5i_info); + type_register_static(&allwinner_sdhost_sun50i_a64_info); + type_register_static(&allwinner_sdhost_sun50i_a64_emmc_info); type_register_static(&allwinner_sdhost_bus_info); } -- 2.34.1 From: qianfan Zhao <qianfanguijin@163.com> R40 has two ethernet controllers named as emac and gmac. The emac is compatibled with A10, and the GMAC is compatibled with H3. Signed-off-by: qianfan Zhao <qianfanguijin@163.com> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/allwinner-r40.h | 6 ++++ hw/arm/allwinner-r40.c | 50 ++++++++++++++++++++++++++++++++-- hw/arm/bananapi_m2u.c | 3 ++ 3 files changed, 57 insertions(+), 2 deletions(-) diff --git a/include/hw/arm/allwinner-r40.h b/include/hw/arm/allwinner-r40.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/allwinner-r40.h +++ b/include/hw/arm/allwinner-r40.h @@ -XXX,XX +XXX,XX @@ #include "hw/misc/allwinner-r40-ccu.h" #include "hw/misc/allwinner-r40-dramc.h" #include "hw/i2c/allwinner-i2c.h" +#include "hw/net/allwinner_emac.h" +#include "hw/net/allwinner-sun8i-emac.h" #include "target/arm/cpu.h" #include "sysemu/block-backend.h" @@ -XXX,XX +XXX,XX @@ enum { AW_R40_DEV_SRAM_A2, AW_R40_DEV_SRAM_A3, AW_R40_DEV_SRAM_A4, + AW_R40_DEV_EMAC, AW_R40_DEV_MMC0, AW_R40_DEV_MMC1, AW_R40_DEV_MMC2, @@ -XXX,XX +XXX,XX @@ enum { AW_R40_DEV_UART6, AW_R40_DEV_UART7, AW_R40_DEV_TWI0, + AW_R40_DEV_GMAC, AW_R40_DEV_GIC_DIST, AW_R40_DEV_GIC_CPU, AW_R40_DEV_GIC_HYP, @@ -XXX,XX +XXX,XX @@ struct AwR40State { AwR40ClockCtlState ccu; AwR40DramCtlState dramc; AWI2CState i2c0; + AwEmacState emac; + AwSun8iEmacState gmac; GICState gic; MemoryRegion sram_a1; MemoryRegion sram_a2; diff --git a/hw/arm/allwinner-r40.c b/hw/arm/allwinner-r40.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/allwinner-r40.c +++ b/hw/arm/allwinner-r40.c @@ -XXX,XX +XXX,XX @@ const hwaddr allwinner_r40_memmap[] = { [AW_R40_DEV_SRAM_A2] = 0x00004000, [AW_R40_DEV_SRAM_A3] = 0x00008000, [AW_R40_DEV_SRAM_A4] = 0x0000b400, + [AW_R40_DEV_EMAC] = 0x01c0b000, [AW_R40_DEV_MMC0] = 0x01c0f000, [AW_R40_DEV_MMC1] = 0x01c10000, [AW_R40_DEV_MMC2] = 0x01c11000, @@ -XXX,XX +XXX,XX @@ const hwaddr allwinner_r40_memmap[] = { [AW_R40_DEV_UART6] = 0x01c29800, [AW_R40_DEV_UART7] = 0x01c29c00, [AW_R40_DEV_TWI0] = 0x01c2ac00, + [AW_R40_DEV_GMAC] = 0x01c50000, [AW_R40_DEV_DRAMCOM] = 0x01c62000, [AW_R40_DEV_DRAMCTL] = 0x01c63000, [AW_R40_DEV_DRAMPHY] = 0x01c65000, @@ -XXX,XX +XXX,XX @@ static struct AwR40Unimplemented r40_unimplemented[] = { { "spi1", 0x01c06000, 4 * KiB }, { "cs0", 0x01c09000, 4 * KiB }, { "keymem", 0x01c0a000, 4 * KiB }, - { "emac", 0x01c0b000, 4 * KiB }, { "usb0-otg", 0x01c13000, 4 * KiB }, { "usb0-host", 0x01c14000, 4 * KiB }, { "crypto", 0x01c15000, 4 * KiB }, @@ -XXX,XX +XXX,XX @@ static struct AwR40Unimplemented r40_unimplemented[] = { { "tvd2", 0x01c33000, 4 * KiB }, { "tvd3", 0x01c34000, 4 * KiB }, { "gpu", 0x01c40000, 64 * KiB }, - { "gmac", 0x01c50000, 64 * KiB }, { "hstmr", 0x01c60000, 4 * KiB }, { "tcon-top", 0x01c70000, 4 * KiB }, { "lcd0", 0x01c71000, 4 * KiB }, @@ -XXX,XX +XXX,XX @@ enum { AW_R40_GIC_SPI_MMC1 = 33, AW_R40_GIC_SPI_MMC2 = 34, AW_R40_GIC_SPI_MMC3 = 35, + AW_R40_GIC_SPI_EMAC = 55, + AW_R40_GIC_SPI_GMAC = 85, }; /* Allwinner R40 general constants */ @@ -XXX,XX +XXX,XX @@ static void allwinner_r40_init(Object *obj) object_initialize_child(obj, "twi0", &s->i2c0, TYPE_AW_I2C_SUN6I); + object_initialize_child(obj, "emac", &s->emac, TYPE_AW_EMAC); + object_initialize_child(obj, "gmac", &s->gmac, TYPE_AW_SUN8I_EMAC); + object_property_add_alias(obj, "gmac-phy-addr", + OBJECT(&s->gmac), "phy-addr"); + object_initialize_child(obj, "dramc", &s->dramc, TYPE_AW_R40_DRAMC); object_property_add_alias(obj, "ram-addr", OBJECT(&s->dramc), "ram-addr"); @@ -XXX,XX +XXX,XX @@ static void allwinner_r40_init(Object *obj) static void allwinner_r40_realize(DeviceState *dev, Error **errp) { + const char *r40_nic_models[] = { "gmac", "emac", NULL }; AwR40State *s = AW_R40(dev); unsigned i; @@ -XXX,XX +XXX,XX @@ static void allwinner_r40_realize(DeviceState *dev, Error **errp) sysbus_mmio_map(SYS_BUS_DEVICE(&s->dramc), 2, s->memmap[AW_R40_DEV_DRAMPHY]); + /* nic support gmac and emac */ + for (int i = 0; i < ARRAY_SIZE(r40_nic_models) - 1; i++) { + NICInfo *nic = &nd_table[i]; + + if (!nic->used) { + continue; + } + if (qemu_show_nic_models(nic->model, r40_nic_models)) { + exit(0); + } + + switch (qemu_find_nic_model(nic, r40_nic_models, r40_nic_models[0])) { + case 0: /* gmac */ + qdev_set_nic_properties(DEVICE(&s->gmac), nic); + break; + case 1: /* emac */ + qdev_set_nic_properties(DEVICE(&s->emac), nic); + break; + default: + exit(1); + break; + } + } + + /* GMAC */ + object_property_set_link(OBJECT(&s->gmac), "dma-memory", + OBJECT(get_system_memory()), &error_fatal); + sysbus_realize(SYS_BUS_DEVICE(&s->gmac), &error_fatal); + sysbus_mmio_map(SYS_BUS_DEVICE(&s->gmac), 0, s->memmap[AW_R40_DEV_GMAC]); + sysbus_connect_irq(SYS_BUS_DEVICE(&s->gmac), 0, + qdev_get_gpio_in(DEVICE(&s->gic), AW_R40_GIC_SPI_GMAC)); + + /* EMAC */ + sysbus_realize(SYS_BUS_DEVICE(&s->emac), &error_fatal); + sysbus_mmio_map(SYS_BUS_DEVICE(&s->emac), 0, s->memmap[AW_R40_DEV_EMAC]); + sysbus_connect_irq(SYS_BUS_DEVICE(&s->emac), 0, + qdev_get_gpio_in(DEVICE(&s->gic), AW_R40_GIC_SPI_EMAC)); + /* Unimplemented devices */ for (i = 0; i < ARRAY_SIZE(r40_unimplemented); i++) { create_unimplemented_device(r40_unimplemented[i].device_name, diff --git a/hw/arm/bananapi_m2u.c b/hw/arm/bananapi_m2u.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/bananapi_m2u.c +++ b/hw/arm/bananapi_m2u.c @@ -XXX,XX +XXX,XX @@ static void bpim2u_init(MachineState *machine) object_property_set_int(OBJECT(r40), "ram-size", r40->ram_size, &error_abort); + /* GMAC PHY */ + object_property_set_uint(OBJECT(r40), "gmac-phy-addr", 1, &error_abort); + /* Mark R40 object realized */ qdev_realize(DEVICE(r40), NULL, &error_abort); -- 2.34.1 From: qianfan Zhao <qianfanguijin@163.com> Only a few important registers are added, especially the SRAM_VER register. Signed-off-by: qianfan Zhao <qianfanguijin@163.com> Reviewed-by: Niek Linnenbank <nieklinnenbank@gmail.com> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/allwinner-r40.h | 3 + include/hw/misc/allwinner-sramc.h | 69 +++++++++++ hw/arm/allwinner-r40.c | 7 +- hw/misc/allwinner-sramc.c | 184 ++++++++++++++++++++++++++++++ hw/arm/Kconfig | 1 + hw/misc/Kconfig | 3 + hw/misc/meson.build | 1 + hw/misc/trace-events | 4 + 8 files changed, 271 insertions(+), 1 deletion(-) create mode 100644 include/hw/misc/allwinner-sramc.h create mode 100644 hw/misc/allwinner-sramc.c diff --git a/include/hw/arm/allwinner-r40.h b/include/hw/arm/allwinner-r40.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/allwinner-r40.h +++ b/include/hw/arm/allwinner-r40.h @@ -XXX,XX +XXX,XX @@ #include "hw/sd/allwinner-sdhost.h" #include "hw/misc/allwinner-r40-ccu.h" #include "hw/misc/allwinner-r40-dramc.h" +#include "hw/misc/allwinner-sramc.h" #include "hw/i2c/allwinner-i2c.h" #include "hw/net/allwinner_emac.h" #include "hw/net/allwinner-sun8i-emac.h" @@ -XXX,XX +XXX,XX @@ enum { AW_R40_DEV_SRAM_A2, AW_R40_DEV_SRAM_A3, AW_R40_DEV_SRAM_A4, + AW_R40_DEV_SRAMC, AW_R40_DEV_EMAC, AW_R40_DEV_MMC0, AW_R40_DEV_MMC1, @@ -XXX,XX +XXX,XX @@ struct AwR40State { ARMCPU cpus[AW_R40_NUM_CPUS]; const hwaddr *memmap; + AwSRAMCState sramc; AwA10PITState timer; AwSdHostState mmc[AW_R40_NUM_MMCS]; AwR40ClockCtlState ccu; diff --git a/include/hw/misc/allwinner-sramc.h b/include/hw/misc/allwinner-sramc.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/include/hw/misc/allwinner-sramc.h @@ -XXX,XX +XXX,XX @@ +/* + * Allwinner SRAM controller emulation + * + * Copyright (C) 2023 qianfan Zhao <qianfanguijin@163.com> + * + * This program is free software: you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation, either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#ifndef HW_MISC_ALLWINNER_SRAMC_H +#define HW_MISC_ALLWINNER_SRAMC_H + +#include "qom/object.h" +#include "hw/sysbus.h" +#include "qemu/uuid.h" + +/** + * Object model + * @{ + */ +#define TYPE_AW_SRAMC "allwinner-sramc" +#define TYPE_AW_SRAMC_SUN8I_R40 TYPE_AW_SRAMC "-sun8i-r40" +OBJECT_DECLARE_TYPE(AwSRAMCState, AwSRAMCClass, AW_SRAMC) + +/** @} */ + +/** + * Allwinner SRAMC object instance state + */ +struct AwSRAMCState { + /*< private >*/ + SysBusDevice parent_obj; + /*< public >*/ + + /** Maps I/O registers in physical memory */ + MemoryRegion iomem; + + /* registers */ + uint32_t sram_ctl1; + uint32_t sram_ver; + uint32_t sram_soft_entry_reg0; +}; + +/** + * Allwinner SRAM Controller class-level struct. + * + * This struct is filled by each sunxi device specific code + * such that the generic code can use this struct to support + * all devices. + */ +struct AwSRAMCClass { + /*< private >*/ + SysBusDeviceClass parent_class; + /*< public >*/ + + uint32_t sram_version_code; +}; + +#endif /* HW_MISC_ALLWINNER_SRAMC_H */ diff --git a/hw/arm/allwinner-r40.c b/hw/arm/allwinner-r40.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/allwinner-r40.c +++ b/hw/arm/allwinner-r40.c @@ -XXX,XX +XXX,XX @@ const hwaddr allwinner_r40_memmap[] = { [AW_R40_DEV_SRAM_A2] = 0x00004000, [AW_R40_DEV_SRAM_A3] = 0x00008000, [AW_R40_DEV_SRAM_A4] = 0x0000b400, + [AW_R40_DEV_SRAMC] = 0x01c00000, [AW_R40_DEV_EMAC] = 0x01c0b000, [AW_R40_DEV_MMC0] = 0x01c0f000, [AW_R40_DEV_MMC1] = 0x01c10000, @@ -XXX,XX +XXX,XX @@ struct AwR40Unimplemented { static struct AwR40Unimplemented r40_unimplemented[] = { { "d-engine", 0x01000000, 4 * MiB }, { "d-inter", 0x01400000, 128 * KiB }, - { "sram-c", 0x01c00000, 4 * KiB }, { "dma", 0x01c02000, 4 * KiB }, { "nfdc", 0x01c03000, 4 * KiB }, { "ts", 0x01c04000, 4 * KiB }, @@ -XXX,XX +XXX,XX @@ static void allwinner_r40_init(Object *obj) "ram-addr"); object_property_add_alias(obj, "ram-size", OBJECT(&s->dramc), "ram-size"); + + object_initialize_child(obj, "sramc", &s->sramc, TYPE_AW_SRAMC_SUN8I_R40); } static void allwinner_r40_realize(DeviceState *dev, Error **errp) @@ -XXX,XX +XXX,XX @@ static void allwinner_r40_realize(DeviceState *dev, Error **errp) AW_R40_GIC_SPI_TIMER1)); /* SRAM */ + sysbus_realize(SYS_BUS_DEVICE(&s->sramc), &error_fatal); + sysbus_mmio_map(SYS_BUS_DEVICE(&s->sramc), 0, s->memmap[AW_R40_DEV_SRAMC]); + memory_region_init_ram(&s->sram_a1, OBJECT(dev), "sram A1", 16 * KiB, &error_abort); memory_region_init_ram(&s->sram_a2, OBJECT(dev), "sram A2", diff --git a/hw/misc/allwinner-sramc.c b/hw/misc/allwinner-sramc.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/misc/allwinner-sramc.c @@ -XXX,XX +XXX,XX @@ +/* + * Allwinner R40 SRAM controller emulation + * + * Copyright (C) 2023 qianfan Zhao <qianfanguijin@163.com> + * + * This program is free software: you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation, either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#include "qemu/osdep.h" +#include "qemu/units.h" +#include "hw/sysbus.h" +#include "migration/vmstate.h" +#include "qemu/log.h" +#include "qemu/module.h" +#include "qapi/error.h" +#include "hw/qdev-properties.h" +#include "hw/qdev-properties-system.h" +#include "hw/misc/allwinner-sramc.h" +#include "trace.h" + +/* + * register offsets + * https://linux-sunxi.org/SRAM_Controller_Register_Guide + */ +enum { + REG_SRAM_CTL1_CFG = 0x04, /* SRAM Control register 1 */ + REG_SRAM_VER = 0x24, /* SRAM Version register */ + REG_SRAM_R40_SOFT_ENTRY_REG0 = 0xbc, +}; + +/* REG_SRAMC_VERSION bit defines */ +#define SRAM_VER_READ_ENABLE (1 << 15) +#define SRAM_VER_VERSION_SHIFT 16 +#define SRAM_VERSION_SUN8I_R40 0x1701 + +static uint64_t allwinner_sramc_read(void *opaque, hwaddr offset, + unsigned size) +{ + AwSRAMCState *s = AW_SRAMC(opaque); + AwSRAMCClass *sc = AW_SRAMC_GET_CLASS(s); + uint64_t val = 0; + + switch (offset) { + case REG_SRAM_CTL1_CFG: + val = s->sram_ctl1; + break; + case REG_SRAM_VER: + /* bit15: lock bit, set this bit before reading this register */ + if (s->sram_ver & SRAM_VER_READ_ENABLE) { + val = SRAM_VER_READ_ENABLE | + (sc->sram_version_code << SRAM_VER_VERSION_SHIFT); + } + break; + case REG_SRAM_R40_SOFT_ENTRY_REG0: + val = s->sram_soft_entry_reg0; + break; + default: + qemu_log_mask(LOG_GUEST_ERROR, "%s: out-of-bounds offset 0x%04x\n", + __func__, (uint32_t)offset); + return 0; + } + + trace_allwinner_sramc_read(offset, val); + + return val; +} + +static void allwinner_sramc_write(void *opaque, hwaddr offset, + uint64_t val, unsigned size) +{ + AwSRAMCState *s = AW_SRAMC(opaque); + + trace_allwinner_sramc_write(offset, val); + + switch (offset) { + case REG_SRAM_CTL1_CFG: + s->sram_ctl1 = val; + break; + case REG_SRAM_VER: + /* Only the READ_ENABLE bit is writeable */ + s->sram_ver = val & SRAM_VER_READ_ENABLE; + break; + case REG_SRAM_R40_SOFT_ENTRY_REG0: + s->sram_soft_entry_reg0 = val; + break; + default: + qemu_log_mask(LOG_GUEST_ERROR, "%s: out-of-bounds offset 0x%04x\n", + __func__, (uint32_t)offset); + break; + } +} + +static const MemoryRegionOps allwinner_sramc_ops = { + .read = allwinner_sramc_read, + .write = allwinner_sramc_write, + .endianness = DEVICE_NATIVE_ENDIAN, + .valid = { + .min_access_size = 4, + .max_access_size = 4, + }, + .impl.min_access_size = 4, +}; + +static const VMStateDescription allwinner_sramc_vmstate = { + .name = "allwinner-sramc", + .version_id = 1, + .minimum_version_id = 1, + .fields = (VMStateField[]) { + VMSTATE_UINT32(sram_ver, AwSRAMCState), + VMSTATE_UINT32(sram_soft_entry_reg0, AwSRAMCState), + VMSTATE_END_OF_LIST() + } +}; + +static void allwinner_sramc_reset(DeviceState *dev) +{ + AwSRAMCState *s = AW_SRAMC(dev); + AwSRAMCClass *sc = AW_SRAMC_GET_CLASS(s); + + switch (sc->sram_version_code) { + case SRAM_VERSION_SUN8I_R40: + s->sram_ctl1 = 0x1300; + break; + } +} + +static void allwinner_sramc_class_init(ObjectClass *klass, void *data) +{ + DeviceClass *dc = DEVICE_CLASS(klass); + + dc->reset = allwinner_sramc_reset; + dc->vmsd = &allwinner_sramc_vmstate; +} + +static void allwinner_sramc_init(Object *obj) +{ + SysBusDevice *sbd = SYS_BUS_DEVICE(obj); + AwSRAMCState *s = AW_SRAMC(obj); + + /* Memory mapping */ + memory_region_init_io(&s->iomem, OBJECT(s), &allwinner_sramc_ops, s, + TYPE_AW_SRAMC, 1 * KiB); + sysbus_init_mmio(sbd, &s->iomem); +} + +static const TypeInfo allwinner_sramc_info = { + .name = TYPE_AW_SRAMC, + .parent = TYPE_SYS_BUS_DEVICE, + .instance_init = allwinner_sramc_init, + .instance_size = sizeof(AwSRAMCState), + .class_init = allwinner_sramc_class_init, +}; + +static void allwinner_r40_sramc_class_init(ObjectClass *klass, void *data) +{ + AwSRAMCClass *sc = AW_SRAMC_CLASS(klass); + + sc->sram_version_code = SRAM_VERSION_SUN8I_R40; +} + +static const TypeInfo allwinner_r40_sramc_info = { + .name = TYPE_AW_SRAMC_SUN8I_R40, + .parent = TYPE_AW_SRAMC, + .class_init = allwinner_r40_sramc_class_init, +}; + +static void allwinner_sramc_register(void) +{ + type_register_static(&allwinner_sramc_info); + type_register_static(&allwinner_r40_sramc_info); +} + +type_init(allwinner_sramc_register) diff --git a/hw/arm/Kconfig b/hw/arm/Kconfig index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/Kconfig +++ b/hw/arm/Kconfig @@ -XXX,XX +XXX,XX @@ config ALLWINNER_H3 config ALLWINNER_R40 bool default y if TCG && ARM + select ALLWINNER_SRAMC select ALLWINNER_A10_PIT select AXP2XX_PMU select SERIAL diff --git a/hw/misc/Kconfig b/hw/misc/Kconfig index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/Kconfig +++ b/hw/misc/Kconfig @@ -XXX,XX +XXX,XX @@ config VIRT_CTRL config LASI bool +config ALLWINNER_SRAMC + bool + config ALLWINNER_A10_CCM bool diff --git a/hw/misc/meson.build b/hw/misc/meson.build index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/meson.build +++ b/hw/misc/meson.build @@ -XXX,XX +XXX,XX @@ subdir('macio') softmmu_ss.add(when: 'CONFIG_IVSHMEM_DEVICE', if_true: files('ivshmem.c')) +softmmu_ss.add(when: 'CONFIG_ALLWINNER_SRAMC', if_true: files('allwinner-sramc.c')) softmmu_ss.add(when: 'CONFIG_ALLWINNER_A10_CCM', if_true: files('allwinner-a10-ccm.c')) softmmu_ss.add(when: 'CONFIG_ALLWINNER_A10_DRAMC', if_true: files('allwinner-a10-dramc.c')) softmmu_ss.add(when: 'CONFIG_ALLWINNER_H3', if_true: files('allwinner-h3-ccu.c')) diff --git a/hw/misc/trace-events b/hw/misc/trace-events index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/trace-events +++ b/hw/misc/trace-events @@ -XXX,XX +XXX,XX @@ allwinner_r40_dramphy_write(uint64_t offset, uint64_t data, unsigned size) "writ allwinner_sid_read(uint64_t offset, uint64_t data, unsigned size) "offset 0x%" PRIx64 " data 0x%" PRIx64 " size %" PRIu32 allwinner_sid_write(uint64_t offset, uint64_t data, unsigned size) "offset 0x%" PRIx64 " data 0x%" PRIx64 " size %" PRIu32 +# allwinner-sramc.c +allwinner_sramc_read(uint64_t offset, uint64_t data) "offset 0x%" PRIx64 " data 0x%" PRIx64 +allwinner_sramc_write(uint64_t offset, uint64_t data) "offset 0x%" PRIx64 " data 0x%" PRIx64 + # avr_power.c avr_power_read(uint8_t value) "power_reduc read value:%u" avr_power_write(uint8_t value) "power_reduc write value:%u" -- 2.34.1 From: qianfan Zhao <qianfanguijin@163.com> Add test case for booting from initrd and sd card. Signed-off-by: qianfan Zhao <qianfanguijin@163.com> Reviewed-by: Niek Linnenbank <nieklinnenbank@gmail.com> Tested-by: Niek Linnenbank <nieklinnenbank@gmail.com> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- tests/avocado/boot_linux_console.py | 176 ++++++++++++++++++++++++++++ 1 file changed, 176 insertions(+) diff --git a/tests/avocado/boot_linux_console.py b/tests/avocado/boot_linux_console.py index XXXXXXX..XXXXXXX 100644 --- a/tests/avocado/boot_linux_console.py +++ b/tests/avocado/boot_linux_console.py @@ -XXX,XX +XXX,XX @@ def test_arm_quanta_gsj_initrd(self): self.wait_for_console_pattern( 'Give root password for system maintenance') + def test_arm_bpim2u(self): + """ + :avocado: tags=arch:arm + :avocado: tags=machine:bpim2u + :avocado: tags=accel:tcg + """ + deb_url = ('https://apt.armbian.com/pool/main/l/linux-5.10.16-sunxi/' + 'linux-image-current-sunxi_21.02.2_armhf.deb') + deb_hash = '9fa84beda245cabf0b4fa84cf6eaa7738ead1da0' + deb_path = self.fetch_asset(deb_url, asset_hash=deb_hash) + kernel_path = self.extract_from_deb(deb_path, + '/boot/vmlinuz-5.10.16-sunxi') + dtb_path = ('/usr/lib/linux-image-current-sunxi/' + 'sun8i-r40-bananapi-m2-ultra.dtb') + dtb_path = self.extract_from_deb(deb_path, dtb_path) + + self.vm.set_console() + kernel_command_line = (self.KERNEL_COMMON_COMMAND_LINE + + 'console=ttyS0,115200n8 ' + 'earlycon=uart,mmio32,0x1c28000') + self.vm.add_args('-kernel', kernel_path, + '-dtb', dtb_path, + '-append', kernel_command_line) + self.vm.launch() + console_pattern = 'Kernel command line: %s' % kernel_command_line + self.wait_for_console_pattern(console_pattern) + + def test_arm_bpim2u_initrd(self): + """ + :avocado: tags=arch:arm + :avocado: tags=accel:tcg + :avocado: tags=machine:bpim2u + """ + deb_url = ('https://apt.armbian.com/pool/main/l/linux-5.10.16-sunxi/' + 'linux-image-current-sunxi_21.02.2_armhf.deb') + deb_hash = '9fa84beda245cabf0b4fa84cf6eaa7738ead1da0' + deb_path = self.fetch_asset(deb_url, asset_hash=deb_hash) + kernel_path = self.extract_from_deb(deb_path, + '/boot/vmlinuz-5.10.16-sunxi') + dtb_path = ('/usr/lib/linux-image-current-sunxi/' + 'sun8i-r40-bananapi-m2-ultra.dtb') + dtb_path = self.extract_from_deb(deb_path, dtb_path) + initrd_url = ('https://github.com/groeck/linux-build-test/raw/' + '2eb0a73b5d5a28df3170c546ddaaa9757e1e0848/rootfs/' + 'arm/rootfs-armv7a.cpio.gz') + initrd_hash = '604b2e45cdf35045846b8bbfbf2129b1891bdc9c' + initrd_path_gz = self.fetch_asset(initrd_url, asset_hash=initrd_hash) + initrd_path = os.path.join(self.workdir, 'rootfs.cpio') + archive.gzip_uncompress(initrd_path_gz, initrd_path) + + self.vm.set_console() + kernel_command_line = (self.KERNEL_COMMON_COMMAND_LINE + + 'console=ttyS0,115200 ' + 'panic=-1 noreboot') + self.vm.add_args('-kernel', kernel_path, + '-dtb', dtb_path, + '-initrd', initrd_path, + '-append', kernel_command_line, + '-no-reboot') + self.vm.launch() + self.wait_for_console_pattern('Boot successful.') + + exec_command_and_wait_for_pattern(self, 'cat /proc/cpuinfo', + 'Allwinner sun8i Family') + exec_command_and_wait_for_pattern(self, 'cat /proc/iomem', + 'system-control@1c00000') + exec_command_and_wait_for_pattern(self, 'reboot', + 'reboot: Restarting system') + # Wait for VM to shut down gracefully + self.vm.wait() + + def test_arm_bpim2u_gmac(self): + """ + :avocado: tags=arch:arm + :avocado: tags=accel:tcg + :avocado: tags=machine:bpim2u + :avocado: tags=device:sd + """ + self.require_netdev('user') + + deb_url = ('https://apt.armbian.com/pool/main/l/linux-5.10.16-sunxi/' + 'linux-image-current-sunxi_21.02.2_armhf.deb') + deb_hash = '9fa84beda245cabf0b4fa84cf6eaa7738ead1da0' + deb_path = self.fetch_asset(deb_url, asset_hash=deb_hash) + kernel_path = self.extract_from_deb(deb_path, + '/boot/vmlinuz-5.10.16-sunxi') + dtb_path = ('/usr/lib/linux-image-current-sunxi/' + 'sun8i-r40-bananapi-m2-ultra.dtb') + dtb_path = self.extract_from_deb(deb_path, dtb_path) + rootfs_url = ('http://storage.kernelci.org/images/rootfs/buildroot/' + 'buildroot-baseline/20221116.0/armel/rootfs.ext2.xz') + rootfs_hash = 'fae32f337c7b87547b10f42599acf109da8b6d9a' + rootfs_path_xz = self.fetch_asset(rootfs_url, asset_hash=rootfs_hash) + rootfs_path = os.path.join(self.workdir, 'rootfs.cpio') + archive.lzma_uncompress(rootfs_path_xz, rootfs_path) + image_pow2ceil_expand(rootfs_path) + + self.vm.set_console() + kernel_command_line = (self.KERNEL_COMMON_COMMAND_LINE + + 'console=ttyS0,115200 ' + 'root=/dev/mmcblk0 rootwait rw ' + 'panic=-1 noreboot') + self.vm.add_args('-kernel', kernel_path, + '-dtb', dtb_path, + '-drive', 'file=' + rootfs_path + ',if=sd,format=raw', + '-net', 'nic,model=gmac,netdev=host_gmac', + '-netdev', 'user,id=host_gmac', + '-append', kernel_command_line, + '-no-reboot') + self.vm.launch() + shell_ready = "/bin/sh: can't access tty; job control turned off" + self.wait_for_console_pattern(shell_ready) + + exec_command_and_wait_for_pattern(self, 'cat /proc/cpuinfo', + 'Allwinner sun8i Family') + exec_command_and_wait_for_pattern(self, 'cat /proc/partitions', + 'mmcblk0') + exec_command_and_wait_for_pattern(self, 'ifconfig eth0 up', + 'eth0: Link is Up') + exec_command_and_wait_for_pattern(self, 'udhcpc eth0', + 'udhcpc: lease of 10.0.2.15 obtained') + exec_command_and_wait_for_pattern(self, 'ping -c 3 10.0.2.2', + '3 packets transmitted, 3 packets received, 0% packet loss') + exec_command_and_wait_for_pattern(self, 'reboot', + 'reboot: Restarting system') + # Wait for VM to shut down gracefully + self.vm.wait() + + @skipUnless(os.getenv('AVOCADO_ALLOW_LARGE_STORAGE'), 'storage limited') + def test_arm_bpim2u_openwrt_22_03_3(self): + """ + :avocado: tags=arch:arm + :avocado: tags=machine:bpim2u + :avocado: tags=device:sd + """ + + # This test download a 8.9 MiB compressed image and expand it + # to 127 MiB. + image_url = ('https://downloads.openwrt.org/releases/22.03.3/targets/' + 'sunxi/cortexa7/openwrt-22.03.3-sunxi-cortexa7-' + 'sinovoip_bananapi-m2-ultra-ext4-sdcard.img.gz') + image_hash = ('5b41b4e11423e562c6011640f9a7cd3b' + 'dd0a3d42b83430f7caa70a432e6cd82c') + image_path_gz = self.fetch_asset(image_url, asset_hash=image_hash, + algorithm='sha256') + image_path = archive.extract(image_path_gz, self.workdir) + image_pow2ceil_expand(image_path) + + self.vm.set_console() + self.vm.add_args('-drive', 'file=' + image_path + ',if=sd,format=raw', + '-nic', 'user', + '-no-reboot') + self.vm.launch() + + kernel_command_line = (self.KERNEL_COMMON_COMMAND_LINE + + 'usbcore.nousb ' + 'noreboot') + + self.wait_for_console_pattern('U-Boot SPL') + + interrupt_interactive_console_until_pattern( + self, 'Hit any key to stop autoboot:', '=>') + exec_command_and_wait_for_pattern(self, "setenv extraargs '" + + kernel_command_line + "'", '=>') + exec_command_and_wait_for_pattern(self, 'boot', 'Starting kernel ...'); + + self.wait_for_console_pattern( + 'Please press Enter to activate this console.') + + exec_command_and_wait_for_pattern(self, ' ', 'root@') + + exec_command_and_wait_for_pattern(self, 'cat /proc/cpuinfo', + 'Allwinner sun8i Family') + exec_command_and_wait_for_pattern(self, 'cat /proc/iomem', + 'system-control@1c00000') + def test_arm_orangepi(self): """ :avocado: tags=arch:arm -- 2.34.1 From: qianfan Zhao <qianfanguijin@163.com> Add documents for Banana Pi M2U Signed-off-by: qianfan Zhao <qianfanguijin@163.com> Reviewed-by: Niek Linnenbank <nieklinnenbank@gmail.com> [PMM: Minor format fixes to correct sphinx errors] Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- docs/system/arm/bananapi_m2u.rst | 139 +++++++++++++++++++++++++++++++ docs/system/target-arm.rst | 1 + 2 files changed, 140 insertions(+) create mode 100644 docs/system/arm/bananapi_m2u.rst diff --git a/docs/system/arm/bananapi_m2u.rst b/docs/system/arm/bananapi_m2u.rst new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/docs/system/arm/bananapi_m2u.rst @@ -XXX,XX +XXX,XX @@ +Banana Pi BPI-M2U (``bpim2u``) +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + +Banana Pi BPI-M2 Ultra is a quad-core mini single board computer built with +Allwinner A40i/R40/V40 SoC. It features 2GB of RAM and 8GB eMMC. It also +has onboard WiFi and BT. On the ports side, the BPI-M2 Ultra has 2 USB A +2.0 ports, 1 USB OTG port, 1 HDMI port, 1 audio jack, a DC power port, +and last but not least, a SATA port. + +Supported devices +""""""""""""""""" + +The Banana Pi M2U machine supports the following devices: + + * SMP (Quad Core Cortex-A7) + * Generic Interrupt Controller configuration + * SRAM mappings + * SDRAM controller + * Timer device (re-used from Allwinner A10) + * UART + * SD/MMC storage controller + * EMAC ethernet + * GMAC ethernet + * Clock Control Unit + * TWI (I2C) + +Limitations +""""""""""" + +Currently, Banana Pi M2U does *not* support the following features: + +- Graphical output via HDMI, GPU and/or the Display Engine +- Audio output +- Hardware Watchdog +- Real Time Clock +- USB 2.0 interfaces + +Also see the 'unimplemented' array in the Allwinner R40 SoC module +for a complete list of unimplemented I/O devices: ``./hw/arm/allwinner-r40.c`` + +Boot options +"""""""""""" + +The Banana Pi M2U machine can start using the standard -kernel functionality +for loading a Linux kernel or ELF executable. Additionally, the Banana Pi M2U +machine can also emulate the BootROM which is present on an actual Allwinner R40 +based SoC, which loads the bootloader from a SD card, specified via the -sd +argument to qemu-system-arm. + +Running mainline Linux +"""""""""""""""""""""" + +To build a Linux mainline kernel that can be booted by the Banana Pi M2U machine, +simply configure the kernel using the sunxi_defconfig configuration: + +.. code-block:: bash + + $ ARCH=arm CROSS_COMPILE=arm-linux-gnueabi- make mrproper + $ ARCH=arm CROSS_COMPILE=arm-linux-gnueabi- make sunxi_defconfig + +To boot the newly build linux kernel in QEMU with the Banana Pi M2U machine, use: + +.. code-block:: bash + + $ qemu-system-arm -M bpim2u -nographic \ + -kernel /path/to/linux/arch/arm/boot/zImage \ + -append 'console=ttyS0,115200' \ + -dtb /path/to/linux/arch/arm/boot/dts/sun8i-r40-bananapi-m2-ultra.dtb + +Banana Pi M2U images +"""""""""""""""""""" + +Note that the mainline kernel does not have a root filesystem. You can choose +to build you own image with buildroot using the bananapi_m2_ultra_defconfig. +Also see https://buildroot.org for more information. + +Another possibility is to run an OpenWrt image for Banana Pi M2U which +can be downloaded from: + + https://downloads.openwrt.org/releases/22.03.3/targets/sunxi/cortexa7/ + +When using an image as an SD card, it must be resized to a power of two. This can be +done with the ``qemu-img`` command. It is recommended to only increase the image size +instead of shrinking it to a power of two, to avoid loss of data. For example, +to prepare a downloaded Armbian image, first extract it and then increase +its size to one gigabyte as follows: + +.. code-block:: bash + + $ qemu-img resize \ + openwrt-22.03.3-sunxi-cortexa7-sinovoip_bananapi-m2-ultra-ext4-sdcard.img \ + 1G + +Instead of providing a custom Linux kernel via the -kernel command you may also +choose to let the Banana Pi M2U machine load the bootloader from SD card, just like +a real board would do using the BootROM. Simply pass the selected image via the -sd +argument and remove the -kernel, -append, -dbt and -initrd arguments: + +.. code-block:: bash + + $ qemu-system-arm -M bpim2u -nic user -nographic \ + -sd openwrt-22.03.3-sunxi-cortexa7-sinovoip_bananapi-m2-ultra-ext4-sdcard.img + +Running U-Boot +"""""""""""""" + +U-Boot mainline can be build and configured using the Bananapi_M2_Ultra_defconfig +using similar commands as describe above for Linux. Note that it is recommended +for development/testing to select the following configuration setting in U-Boot: + + Device Tree Control > Provider for DTB for DT Control > Embedded DTB + +The BootROM of allwinner R40 loading u-boot from the 8KiB offset of sdcard. +Let's create an bootable disk image: + +.. code-block:: bash + + $ dd if=/dev/zero of=sd.img bs=32M count=1 + $ dd if=u-boot-sunxi-with-spl.bin of=sd.img bs=1k seek=8 conv=notrunc + +And then boot it. + +.. code-block:: bash + + $ qemu-system-arm -M bpim2u -nographic -sd sd.img + +Banana Pi M2U integration tests +""""""""""""""""""""""""""""""" + +The Banana Pi M2U machine has several integration tests included. +To run the whole set of tests, build QEMU from source and simply +provide the following command: + +.. code-block:: bash + + $ cd qemu-build-dir + $ AVOCADO_ALLOW_LARGE_STORAGE=yes tests/venv/bin/avocado \ + --verbose --show=app,console run -t machine:bpim2u \ + ../tests/avocado/boot_linux_console.py diff --git a/docs/system/target-arm.rst b/docs/system/target-arm.rst index XXXXXXX..XXXXXXX 100644 --- a/docs/system/target-arm.rst +++ b/docs/system/target-arm.rst @@ -XXX,XX +XXX,XX @@ undocumented; you can get a complete list by running arm/versatile arm/vexpress arm/aspeed + arm/bananapi_m2u.rst arm/sabrelite arm/digic arm/cubieboard -- 2.34.1 From: Richard Henderson <richard.henderson@linaro.org> Document the meaning of exclusive_high in a big-endian context, and why we can't change it now. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230530191438.411344-2-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/cpu.h | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/target/arm/cpu.h b/target/arm/cpu.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.h +++ b/target/arm/cpu.h @@ -XXX,XX +XXX,XX @@ typedef struct CPUArchState { uint64_t zcr_el[4]; /* ZCR_EL[1-3] */ uint64_t smcr_el[4]; /* SMCR_EL[1-3] */ } vfp; + uint64_t exclusive_addr; uint64_t exclusive_val; + /* + * Contains the 'val' for the second 64-bit register of LDXP, which comes + * from the higher address, not the high part of a complete 128-bit value. + * In some ways it might be more convenient to record the exclusive value + * as the low and high halves of a 128 bit data value, but the current + * semantics of these fields are baked into the migration format. + */ uint64_t exclusive_high; /* iwMMXt coprocessor state. */ -- 2.34.1 From: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230530191438.411344-3-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/cpu.h | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/target/arm/cpu.h b/target/arm/cpu.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.h +++ b/target/arm/cpu.h @@ -XXX,XX +XXX,XX @@ static inline bool isar_feature_aa64_st(const ARMISARegisters *id) return FIELD_EX64(id->id_aa64mmfr2, ID_AA64MMFR2, ST) != 0; } +static inline bool isar_feature_aa64_lse2(const ARMISARegisters *id) +{ + return FIELD_EX64(id->id_aa64mmfr2, ID_AA64MMFR2, AT) != 0; +} + static inline bool isar_feature_aa64_fwb(const ARMISARegisters *id) { return FIELD_EX64(id->id_aa64mmfr2, ID_AA64MMFR2, FWB) != 0; -- 2.34.1 From: Richard Henderson <richard.henderson@linaro.org> Let finalize_memop_atom be the new basic function, with finalize_memop and finalize_memop_pair testing FEAT_LSE2 to apply the appropriate atomicity. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230530191438.411344-4-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/tcg/translate.h | 39 +++++++++++++++++++++++++++++----- target/arm/tcg/translate-a64.c | 2 ++ target/arm/tcg/translate.c | 1 + 3 files changed, 37 insertions(+), 5 deletions(-) diff --git a/target/arm/tcg/translate.h b/target/arm/tcg/translate.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/tcg/translate.h +++ b/target/arm/tcg/translate.h @@ -XXX,XX +XXX,XX @@ typedef struct DisasContext { uint64_t features; /* CPU features bits */ bool aarch64; bool thumb; + bool lse2; /* Because unallocated encodings generate different exception syndrome * information from traps due to FP being disabled, we can't do a single * "is fp access disabled" check at a high level in the decode tree. @@ -XXX,XX +XXX,XX @@ static inline TCGv_ptr fpstatus_ptr(ARMFPStatusFlavour flavour) } /** - * finalize_memop: + * finalize_memop_atom: * @s: DisasContext * @opc: size+sign+align of the memory operation + * @atom: atomicity of the memory operation * - * Build the complete MemOp for a memory operation, including alignment - * and endianness. + * Build the complete MemOp for a memory operation, including alignment, + * endianness, and atomicity. * * If (op & MO_AMASK) then the operation already contains the required * alignment, e.g. for AccType_ATOMIC. Otherwise, this an optionally @@ -XXX,XX +XXX,XX @@ static inline TCGv_ptr fpstatus_ptr(ARMFPStatusFlavour flavour) * and this is applied here. Note that there is no way to indicate that * no alignment should ever be enforced; this must be handled manually. */ -static inline MemOp finalize_memop(DisasContext *s, MemOp opc) +static inline MemOp finalize_memop_atom(DisasContext *s, MemOp opc, MemOp atom) { if (s->align_mem && !(opc & MO_AMASK)) { opc |= MO_ALIGN; } - return opc | s->be_data; + return opc | atom | s->be_data; +} + +/** + * finalize_memop: + * @s: DisasContext + * @opc: size+sign+align of the memory operation + * + * Like finalize_memop_atom, but with default atomicity. + */ +static inline MemOp finalize_memop(DisasContext *s, MemOp opc) +{ + MemOp atom = s->lse2 ? MO_ATOM_WITHIN16 : MO_ATOM_IFALIGN; + return finalize_memop_atom(s, opc, atom); +} + +/** + * finalize_memop_pair: + * @s: DisasContext + * @opc: size+sign+align of the memory operation + * + * Like finalize_memop_atom, but with atomicity for a pair. + * C.f. Pseudocode for Mem[], operand ispair. + */ +static inline MemOp finalize_memop_pair(DisasContext *s, MemOp opc) +{ + MemOp atom = s->lse2 ? MO_ATOM_WITHIN16_PAIR : MO_ATOM_IFALIGN_PAIR; + return finalize_memop_atom(s, opc, atom); } /** diff --git a/target/arm/tcg/translate-a64.c b/target/arm/tcg/translate-a64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/tcg/translate-a64.c +++ b/target/arm/tcg/translate-a64.c @@ -XXX,XX +XXX,XX @@ static void aarch64_tr_init_disas_context(DisasContextBase *dcbase, tcg_debug_assert(dc->tbid & 1); #endif + dc->lse2 = dc_isar_feature(aa64_lse2, dc); + /* Single step state. The code-generation logic here is: * SS_ACTIVE == 0: * generate code with no special handling for single-stepping (except diff --git a/target/arm/tcg/translate.c b/target/arm/tcg/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/tcg/translate.c +++ b/target/arm/tcg/translate.c @@ -XXX,XX +XXX,XX @@ static void arm_tr_init_disas_context(DisasContextBase *dcbase, CPUState *cs) dc->sme_trap_nonstreaming = EX_TBFLAG_A32(tb_flags, SME_TRAP_NONSTREAMING); } + dc->lse2 = false; /* applies only to aarch64 */ dc->cp_regs = cpu->cp_regs; dc->features = env->features; -- 2.34.1 From: Richard Henderson <richard.henderson@linaro.org> While we don't require 16-byte atomicity here, using a single larger load simplifies the code, and makes it a closer match to STXP. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230530191438.411344-5-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/tcg/translate-a64.c | 31 ++++++++++++++++++++----------- 1 file changed, 20 insertions(+), 11 deletions(-) diff --git a/target/arm/tcg/translate-a64.c b/target/arm/tcg/translate-a64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/tcg/translate-a64.c +++ b/target/arm/tcg/translate-a64.c @@ -XXX,XX +XXX,XX @@ static void gen_load_exclusive(DisasContext *s, int rt, int rt2, TCGv_i64 addr, int size, bool is_pair) { int idx = get_mem_index(s); - MemOp memop = s->be_data; + MemOp memop; g_assert(size <= 3); if (is_pair) { g_assert(size >= 2); if (size == 2) { /* The pair must be single-copy atomic for the doubleword. */ - memop |= MO_64 | MO_ALIGN; + memop = finalize_memop(s, MO_64 | MO_ALIGN); tcg_gen_qemu_ld_i64(cpu_exclusive_val, addr, idx, memop); if (s->be_data == MO_LE) { tcg_gen_extract_i64(cpu_reg(s, rt), cpu_exclusive_val, 0, 32); @@ -XXX,XX +XXX,XX @@ static void gen_load_exclusive(DisasContext *s, int rt, int rt2, tcg_gen_extract_i64(cpu_reg(s, rt2), cpu_exclusive_val, 0, 32); } } else { - /* The pair must be single-copy atomic for *each* doubleword, not - the entire quadword, however it must be quadword aligned. */ - memop |= MO_64; - tcg_gen_qemu_ld_i64(cpu_exclusive_val, addr, idx, - memop | MO_ALIGN_16); + /* + * The pair must be single-copy atomic for *each* doubleword, not + * the entire quadword, however it must be quadword aligned. + * Expose the complete load to tcg, for ease of tlb lookup, + * but indicate that only 8-byte atomicity is required. + */ + TCGv_i128 t16 = tcg_temp_new_i128(); - TCGv_i64 addr2 = tcg_temp_new_i64(); - tcg_gen_addi_i64(addr2, addr, 8); - tcg_gen_qemu_ld_i64(cpu_exclusive_high, addr2, idx, memop); + memop = finalize_memop_atom(s, MO_128 | MO_ALIGN_16, + MO_ATOM_IFALIGN_PAIR); + tcg_gen_qemu_ld_i128(t16, addr, idx, memop); + if (s->be_data == MO_LE) { + tcg_gen_extr_i128_i64(cpu_exclusive_val, + cpu_exclusive_high, t16); + } else { + tcg_gen_extr_i128_i64(cpu_exclusive_high, + cpu_exclusive_val, t16); + } tcg_gen_mov_i64(cpu_reg(s, rt), cpu_exclusive_val); tcg_gen_mov_i64(cpu_reg(s, rt2), cpu_exclusive_high); } } else { - memop |= size | MO_ALIGN; + memop = finalize_memop(s, size | MO_ALIGN); tcg_gen_qemu_ld_i64(cpu_exclusive_val, addr, idx, memop); tcg_gen_mov_i64(cpu_reg(s, rt), cpu_exclusive_val); } -- 2.34.1 From: Richard Henderson <richard.henderson@linaro.org> While we don't require 16-byte atomicity here, using a single larger operation simplifies the code. Introduce finalize_memop_asimd for this. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230530191438.411344-6-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/tcg/translate.h | 24 +++++++++++++++++++++++ target/arm/tcg/translate-a64.c | 35 +++++++++++----------------------- 2 files changed, 35 insertions(+), 24 deletions(-) diff --git a/target/arm/tcg/translate.h b/target/arm/tcg/translate.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/tcg/translate.h +++ b/target/arm/tcg/translate.h @@ -XXX,XX +XXX,XX @@ static inline MemOp finalize_memop_pair(DisasContext *s, MemOp opc) return finalize_memop_atom(s, opc, atom); } +/** + * finalize_memop_asimd: + * @s: DisasContext + * @opc: size+sign+align of the memory operation + * + * Like finalize_memop_atom, but with atomicity of AccessType_ASIMD. + */ +static inline MemOp finalize_memop_asimd(DisasContext *s, MemOp opc) +{ + /* + * In the pseudocode for Mem[], with AccessType_ASIMD, size == 16, + * if IsAligned(8), the first case provides separate atomicity for + * the pair of 64-bit accesses. If !IsAligned(8), the middle cases + * do not apply, and we're left with the final case of no atomicity. + * Thus MO_ATOM_IFALIGN_PAIR. + * + * For other sizes, normal LSE2 rules apply. + */ + if ((opc & MO_SIZE) == MO_128) { + return finalize_memop_atom(s, opc, MO_ATOM_IFALIGN_PAIR); + } + return finalize_memop(s, opc); +} + /** * asimd_imm_const: Expand an encoded SIMD constant value * diff --git a/target/arm/tcg/translate-a64.c b/target/arm/tcg/translate-a64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/tcg/translate-a64.c +++ b/target/arm/tcg/translate-a64.c @@ -XXX,XX +XXX,XX @@ static void do_fp_st(DisasContext *s, int srcidx, TCGv_i64 tcg_addr, int size) { /* This writes the bottom N bits of a 128 bit wide vector to memory */ TCGv_i64 tmplo = tcg_temp_new_i64(); - MemOp mop; + MemOp mop = finalize_memop_asimd(s, size); tcg_gen_ld_i64(tmplo, cpu_env, fp_reg_offset(s, srcidx, MO_64)); - if (size < 4) { - mop = finalize_memop(s, size); + if (size < MO_128) { tcg_gen_qemu_st_i64(tmplo, tcg_addr, get_mem_index(s), mop); } else { - bool be = s->be_data == MO_BE; - TCGv_i64 tcg_hiaddr = tcg_temp_new_i64(); TCGv_i64 tmphi = tcg_temp_new_i64(); + TCGv_i128 t16 = tcg_temp_new_i128(); tcg_gen_ld_i64(tmphi, cpu_env, fp_reg_hi_offset(s, srcidx)); + tcg_gen_concat_i64_i128(t16, tmplo, tmphi); - mop = s->be_data | MO_UQ; - tcg_gen_qemu_st_i64(be ? tmphi : tmplo, tcg_addr, get_mem_index(s), - mop | (s->align_mem ? MO_ALIGN_16 : 0)); - tcg_gen_addi_i64(tcg_hiaddr, tcg_addr, 8); - tcg_gen_qemu_st_i64(be ? tmplo : tmphi, tcg_hiaddr, - get_mem_index(s), mop); + tcg_gen_qemu_st_i128(t16, tcg_addr, get_mem_index(s), mop); } } @@ -XXX,XX +XXX,XX @@ static void do_fp_ld(DisasContext *s, int destidx, TCGv_i64 tcg_addr, int size) /* This always zero-extends and writes to a full 128 bit wide vector */ TCGv_i64 tmplo = tcg_temp_new_i64(); TCGv_i64 tmphi = NULL; - MemOp mop; + MemOp mop = finalize_memop_asimd(s, size); - if (size < 4) { - mop = finalize_memop(s, size); + if (size < MO_128) { tcg_gen_qemu_ld_i64(tmplo, tcg_addr, get_mem_index(s), mop); } else { - bool be = s->be_data == MO_BE; - TCGv_i64 tcg_hiaddr; + TCGv_i128 t16 = tcg_temp_new_i128(); + + tcg_gen_qemu_ld_i128(t16, tcg_addr, get_mem_index(s), mop); tmphi = tcg_temp_new_i64(); - tcg_hiaddr = tcg_temp_new_i64(); - - mop = s->be_data | MO_UQ; - tcg_gen_qemu_ld_i64(be ? tmphi : tmplo, tcg_addr, get_mem_index(s), - mop | (s->align_mem ? MO_ALIGN_16 : 0)); - tcg_gen_addi_i64(tcg_hiaddr, tcg_addr, 8); - tcg_gen_qemu_ld_i64(be ? tmplo : tmphi, tcg_hiaddr, - get_mem_index(s), mop); + tcg_gen_extr_i128_i64(tmplo, tmphi, t16); } tcg_gen_st_i64(tmplo, cpu_env, fp_reg_offset(s, destidx, MO_64)); -- 2.34.1 From: Richard Henderson <richard.henderson@linaro.org> This fixes a bug in that these two insns should have been using atomic 16-byte stores, since MTE is ARMv8.5 and LSE2 is mandatory from ARMv8.4. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230530191438.411344-7-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/tcg/translate-a64.c | 17 ++++++++++------- 1 file changed, 10 insertions(+), 7 deletions(-) diff --git a/target/arm/tcg/translate-a64.c b/target/arm/tcg/translate-a64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/tcg/translate-a64.c +++ b/target/arm/tcg/translate-a64.c @@ -XXX,XX +XXX,XX @@ static void disas_ldst_tag(DisasContext *s, uint32_t insn) if (is_zero) { TCGv_i64 clean_addr = clean_data_tbi(s, addr); - TCGv_i64 tcg_zero = tcg_constant_i64(0); + TCGv_i64 zero64 = tcg_constant_i64(0); + TCGv_i128 zero128 = tcg_temp_new_i128(); int mem_index = get_mem_index(s); - int i, n = (1 + is_pair) << LOG2_TAG_GRANULE; + MemOp mop = finalize_memop(s, MO_128 | MO_ALIGN); - tcg_gen_qemu_st_i64(tcg_zero, clean_addr, mem_index, - MO_UQ | MO_ALIGN_16); - for (i = 8; i < n; i += 8) { - tcg_gen_addi_i64(clean_addr, clean_addr, 8); - tcg_gen_qemu_st_i64(tcg_zero, clean_addr, mem_index, MO_UQ); + tcg_gen_concat_i64_i128(zero128, zero64, zero64); + + /* This is 1 or 2 atomic 16-byte operations. */ + tcg_gen_qemu_st_i128(zero128, clean_addr, mem_index, mop); + if (is_pair) { + tcg_gen_addi_i64(clean_addr, clean_addr, 16); + tcg_gen_qemu_st_i128(zero128, clean_addr, mem_index, mop); } } -- 2.34.1 From: Richard Henderson <richard.henderson@linaro.org> Round len_align to 16 instead of 8, handling an odd 8-byte as part of the tail. Use MO_ATOM_NONE to indicate that all of these memory ops have only byte atomicity. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230530191438.411344-8-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/tcg/translate-sve.c | 95 +++++++++++++++++++++++++--------- 1 file changed, 70 insertions(+), 25 deletions(-) diff --git a/target/arm/tcg/translate-sve.c b/target/arm/tcg/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/tcg/translate-sve.c +++ b/target/arm/tcg/translate-sve.c @@ -XXX,XX +XXX,XX @@ TRANS_FEAT(UCVTF_dd, aa64_sve, gen_gvec_fpst_arg_zpz, void gen_sve_ldr(DisasContext *s, TCGv_ptr base, int vofs, int len, int rn, int imm) { - int len_align = QEMU_ALIGN_DOWN(len, 8); - int len_remain = len % 8; - int nparts = len / 8 + ctpop8(len_remain); + int len_align = QEMU_ALIGN_DOWN(len, 16); + int len_remain = len % 16; + int nparts = len / 16 + ctpop8(len_remain); int midx = get_mem_index(s); TCGv_i64 dirty_addr, clean_addr, t0, t1; + TCGv_i128 t16; dirty_addr = tcg_temp_new_i64(); tcg_gen_addi_i64(dirty_addr, cpu_reg_sp(s, rn), imm); @@ -XXX,XX +XXX,XX @@ void gen_sve_ldr(DisasContext *s, TCGv_ptr base, int vofs, int i; t0 = tcg_temp_new_i64(); - for (i = 0; i < len_align; i += 8) { - tcg_gen_qemu_ld_i64(t0, clean_addr, midx, MO_LEUQ); + t1 = tcg_temp_new_i64(); + t16 = tcg_temp_new_i128(); + + for (i = 0; i < len_align; i += 16) { + tcg_gen_qemu_ld_i128(t16, clean_addr, midx, + MO_LE | MO_128 | MO_ATOM_NONE); + tcg_gen_extr_i128_i64(t0, t1, t16); tcg_gen_st_i64(t0, base, vofs + i); - tcg_gen_addi_i64(clean_addr, clean_addr, 8); + tcg_gen_st_i64(t1, base, vofs + i + 8); + tcg_gen_addi_i64(clean_addr, clean_addr, 16); } } else { TCGLabel *loop = gen_new_label(); @@ -XXX,XX +XXX,XX @@ void gen_sve_ldr(DisasContext *s, TCGv_ptr base, int vofs, tcg_gen_movi_ptr(i, 0); gen_set_label(loop); - t0 = tcg_temp_new_i64(); - tcg_gen_qemu_ld_i64(t0, clean_addr, midx, MO_LEUQ); - tcg_gen_addi_i64(clean_addr, clean_addr, 8); + t16 = tcg_temp_new_i128(); + tcg_gen_qemu_ld_i128(t16, clean_addr, midx, + MO_LE | MO_128 | MO_ATOM_NONE); + tcg_gen_addi_i64(clean_addr, clean_addr, 16); tp = tcg_temp_new_ptr(); tcg_gen_add_ptr(tp, base, i); - tcg_gen_addi_ptr(i, i, 8); + tcg_gen_addi_ptr(i, i, 16); + + t0 = tcg_temp_new_i64(); + t1 = tcg_temp_new_i64(); + tcg_gen_extr_i128_i64(t0, t1, t16); + tcg_gen_st_i64(t0, tp, vofs); + tcg_gen_st_i64(t1, tp, vofs + 8); tcg_gen_brcondi_ptr(TCG_COND_LTU, i, len_align, loop); } @@ -XXX,XX +XXX,XX @@ void gen_sve_ldr(DisasContext *s, TCGv_ptr base, int vofs, * Predicate register loads can be any multiple of 2. * Note that we still store the entire 64-bit unit into cpu_env. */ + if (len_remain >= 8) { + t0 = tcg_temp_new_i64(); + tcg_gen_qemu_ld_i64(t0, clean_addr, midx, MO_LEUQ | MO_ATOM_NONE); + tcg_gen_st_i64(t0, base, vofs + len_align); + len_remain -= 8; + len_align += 8; + if (len_remain) { + tcg_gen_addi_i64(clean_addr, clean_addr, 8); + } + } if (len_remain) { t0 = tcg_temp_new_i64(); switch (len_remain) { @@ -XXX,XX +XXX,XX @@ void gen_sve_ldr(DisasContext *s, TCGv_ptr base, int vofs, case 4: case 8: tcg_gen_qemu_ld_i64(t0, clean_addr, midx, - MO_LE | ctz32(len_remain)); + MO_LE | ctz32(len_remain) | MO_ATOM_NONE); break; case 6: t1 = tcg_temp_new_i64(); - tcg_gen_qemu_ld_i64(t0, clean_addr, midx, MO_LEUL); + tcg_gen_qemu_ld_i64(t0, clean_addr, midx, MO_LEUL | MO_ATOM_NONE); tcg_gen_addi_i64(clean_addr, clean_addr, 4); - tcg_gen_qemu_ld_i64(t1, clean_addr, midx, MO_LEUW); + tcg_gen_qemu_ld_i64(t1, clean_addr, midx, MO_LEUW | MO_ATOM_NONE); tcg_gen_deposit_i64(t0, t0, t1, 32, 32); break; @@ -XXX,XX +XXX,XX @@ void gen_sve_ldr(DisasContext *s, TCGv_ptr base, int vofs, void gen_sve_str(DisasContext *s, TCGv_ptr base, int vofs, int len, int rn, int imm) { - int len_align = QEMU_ALIGN_DOWN(len, 8); - int len_remain = len % 8; - int nparts = len / 8 + ctpop8(len_remain); + int len_align = QEMU_ALIGN_DOWN(len, 16); + int len_remain = len % 16; + int nparts = len / 16 + ctpop8(len_remain); int midx = get_mem_index(s); - TCGv_i64 dirty_addr, clean_addr, t0; + TCGv_i64 dirty_addr, clean_addr, t0, t1; + TCGv_i128 t16; dirty_addr = tcg_temp_new_i64(); tcg_gen_addi_i64(dirty_addr, cpu_reg_sp(s, rn), imm); @@ -XXX,XX +XXX,XX @@ void gen_sve_str(DisasContext *s, TCGv_ptr base, int vofs, int i; t0 = tcg_temp_new_i64(); + t1 = tcg_temp_new_i64(); + t16 = tcg_temp_new_i128(); for (i = 0; i < len_align; i += 8) { tcg_gen_ld_i64(t0, base, vofs + i); - tcg_gen_qemu_st_i64(t0, clean_addr, midx, MO_LEUQ); - tcg_gen_addi_i64(clean_addr, clean_addr, 8); + tcg_gen_ld_i64(t1, base, vofs + i + 8); + tcg_gen_concat_i64_i128(t16, t0, t1); + tcg_gen_qemu_st_i128(t16, clean_addr, midx, + MO_LE | MO_128 | MO_ATOM_NONE); + tcg_gen_addi_i64(clean_addr, clean_addr, 16); } } else { TCGLabel *loop = gen_new_label(); @@ -XXX,XX +XXX,XX @@ void gen_sve_str(DisasContext *s, TCGv_ptr base, int vofs, gen_set_label(loop); t0 = tcg_temp_new_i64(); + t1 = tcg_temp_new_i64(); tp = tcg_temp_new_ptr(); tcg_gen_add_ptr(tp, base, i); tcg_gen_ld_i64(t0, tp, vofs); - tcg_gen_addi_ptr(i, i, 8); + tcg_gen_ld_i64(t1, tp, vofs + 8); + tcg_gen_addi_ptr(i, i, 16); - tcg_gen_qemu_st_i64(t0, clean_addr, midx, MO_LEUQ); - tcg_gen_addi_i64(clean_addr, clean_addr, 8); + t16 = tcg_temp_new_i128(); + tcg_gen_concat_i64_i128(t16, t0, t1); + + tcg_gen_qemu_st_i128(t16, clean_addr, midx, MO_LEUQ); + tcg_gen_addi_i64(clean_addr, clean_addr, 16); tcg_gen_brcondi_ptr(TCG_COND_LTU, i, len_align, loop); } /* Predicate register stores can be any multiple of 2. */ + if (len_remain >= 8) { + t0 = tcg_temp_new_i64(); + tcg_gen_st_i64(t0, base, vofs + len_align); + tcg_gen_qemu_st_i64(t0, clean_addr, midx, MO_LEUQ | MO_ATOM_NONE); + len_remain -= 8; + len_align += 8; + if (len_remain) { + tcg_gen_addi_i64(clean_addr, clean_addr, 8); + } + } if (len_remain) { t0 = tcg_temp_new_i64(); tcg_gen_ld_i64(t0, base, vofs + len_align); @@ -XXX,XX +XXX,XX @@ void gen_sve_str(DisasContext *s, TCGv_ptr base, int vofs, case 4: case 8: tcg_gen_qemu_st_i64(t0, clean_addr, midx, - MO_LE | ctz32(len_remain)); + MO_LE | ctz32(len_remain) | MO_ATOM_NONE); break; case 6: - tcg_gen_qemu_st_i64(t0, clean_addr, midx, MO_LEUL); + tcg_gen_qemu_st_i64(t0, clean_addr, midx, MO_LEUL | MO_ATOM_NONE); tcg_gen_addi_i64(clean_addr, clean_addr, 4); tcg_gen_shri_i64(t0, t0, 32); - tcg_gen_qemu_st_i64(t0, clean_addr, midx, MO_LEUW); + tcg_gen_qemu_st_i64(t0, clean_addr, midx, MO_LEUW | MO_ATOM_NONE); break; default: -- 2.34.1 From: Richard Henderson <richard.henderson@linaro.org> No need to duplicate this check across multiple call sites. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230530191438.411344-9-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/tcg/translate-a64.c | 44 ++++++++++++++++------------------ 1 file changed, 21 insertions(+), 23 deletions(-) diff --git a/target/arm/tcg/translate-a64.c b/target/arm/tcg/translate-a64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/tcg/translate-a64.c +++ b/target/arm/tcg/translate-a64.c @@ -XXX,XX +XXX,XX @@ static void disas_b_exc_sys(DisasContext *s, uint32_t insn) * races in multi-threaded linux-user and when MTTCG softmmu is * enabled. */ -static void gen_load_exclusive(DisasContext *s, int rt, int rt2, - TCGv_i64 addr, int size, bool is_pair) +static void gen_load_exclusive(DisasContext *s, int rt, int rt2, int rn, + int size, bool is_pair) { int idx = get_mem_index(s); MemOp memop; + TCGv_i64 dirty_addr, clean_addr; + + s->is_ldex = true; + dirty_addr = cpu_reg_sp(s, rn); + clean_addr = gen_mte_check1(s, dirty_addr, false, rn != 31, size); g_assert(size <= 3); if (is_pair) { @@ -XXX,XX +XXX,XX @@ static void gen_load_exclusive(DisasContext *s, int rt, int rt2, if (size == 2) { /* The pair must be single-copy atomic for the doubleword. */ memop = finalize_memop(s, MO_64 | MO_ALIGN); - tcg_gen_qemu_ld_i64(cpu_exclusive_val, addr, idx, memop); + tcg_gen_qemu_ld_i64(cpu_exclusive_val, clean_addr, idx, memop); if (s->be_data == MO_LE) { tcg_gen_extract_i64(cpu_reg(s, rt), cpu_exclusive_val, 0, 32); tcg_gen_extract_i64(cpu_reg(s, rt2), cpu_exclusive_val, 32, 32); @@ -XXX,XX +XXX,XX @@ static void gen_load_exclusive(DisasContext *s, int rt, int rt2, memop = finalize_memop_atom(s, MO_128 | MO_ALIGN_16, MO_ATOM_IFALIGN_PAIR); - tcg_gen_qemu_ld_i128(t16, addr, idx, memop); + tcg_gen_qemu_ld_i128(t16, clean_addr, idx, memop); if (s->be_data == MO_LE) { tcg_gen_extr_i128_i64(cpu_exclusive_val, @@ -XXX,XX +XXX,XX @@ static void gen_load_exclusive(DisasContext *s, int rt, int rt2, } } else { memop = finalize_memop(s, size | MO_ALIGN); - tcg_gen_qemu_ld_i64(cpu_exclusive_val, addr, idx, memop); + tcg_gen_qemu_ld_i64(cpu_exclusive_val, clean_addr, idx, memop); tcg_gen_mov_i64(cpu_reg(s, rt), cpu_exclusive_val); } - tcg_gen_mov_i64(cpu_exclusive_addr, addr); + tcg_gen_mov_i64(cpu_exclusive_addr, clean_addr); } static void gen_store_exclusive(DisasContext *s, int rd, int rt, int rt2, - TCGv_i64 addr, int size, int is_pair) + int rn, int size, int is_pair) { /* if (env->exclusive_addr == addr && env->exclusive_val == [addr] * && (!is_pair || env->exclusive_high == [addr + datasize])) { @@ -XXX,XX +XXX,XX @@ static void gen_store_exclusive(DisasContext *s, int rd, int rt, int rt2, */ TCGLabel *fail_label = gen_new_label(); TCGLabel *done_label = gen_new_label(); - TCGv_i64 tmp; + TCGv_i64 tmp, dirty_addr, clean_addr; - tcg_gen_brcond_i64(TCG_COND_NE, addr, cpu_exclusive_addr, fail_label); + dirty_addr = cpu_reg_sp(s, rn); + clean_addr = gen_mte_check1(s, dirty_addr, true, rn != 31, size); + + tcg_gen_brcond_i64(TCG_COND_NE, clean_addr, cpu_exclusive_addr, fail_label); tmp = tcg_temp_new_i64(); if (is_pair) { @@ -XXX,XX +XXX,XX @@ static void disas_ldst_excl(DisasContext *s, uint32_t insn) if (is_lasr) { tcg_gen_mb(TCG_MO_ALL | TCG_BAR_STRL); } - clean_addr = gen_mte_check1(s, cpu_reg_sp(s, rn), - true, rn != 31, size); - gen_store_exclusive(s, rs, rt, rt2, clean_addr, size, false); + gen_store_exclusive(s, rs, rt, rt2, rn, size, false); return; case 0x4: /* LDXR */ @@ -XXX,XX +XXX,XX @@ static void disas_ldst_excl(DisasContext *s, uint32_t insn) if (rn == 31) { gen_check_sp_alignment(s); } - clean_addr = gen_mte_check1(s, cpu_reg_sp(s, rn), - false, rn != 31, size); - s->is_ldex = true; - gen_load_exclusive(s, rt, rt2, clean_addr, size, false); + gen_load_exclusive(s, rt, rt2, rn, size, false); if (is_lasr) { tcg_gen_mb(TCG_MO_ALL | TCG_BAR_LDAQ); } @@ -XXX,XX +XXX,XX @@ static void disas_ldst_excl(DisasContext *s, uint32_t insn) if (is_lasr) { tcg_gen_mb(TCG_MO_ALL | TCG_BAR_STRL); } - clean_addr = gen_mte_check1(s, cpu_reg_sp(s, rn), - true, rn != 31, size); - gen_store_exclusive(s, rs, rt, rt2, clean_addr, size, true); + gen_store_exclusive(s, rs, rt, rt2, rn, size, true); return; } if (rt2 == 31 @@ -XXX,XX +XXX,XX @@ static void disas_ldst_excl(DisasContext *s, uint32_t insn) if (rn == 31) { gen_check_sp_alignment(s); } - clean_addr = gen_mte_check1(s, cpu_reg_sp(s, rn), - false, rn != 31, size); - s->is_ldex = true; - gen_load_exclusive(s, rt, rt2, clean_addr, size, true); + gen_load_exclusive(s, rt, rt2, rn, size, true); if (is_lasr) { tcg_gen_mb(TCG_MO_ALL | TCG_BAR_LDAQ); } -- 2.34.1 From: Richard Henderson <richard.henderson@linaro.org> This is required for LSE2, where the pair must be treated atomically if it does not cross a 16-byte boundary. But it simplifies the code to do this always. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230530191438.411344-10-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/tcg/translate-a64.c | 70 ++++++++++++++++++++++++++-------- 1 file changed, 55 insertions(+), 15 deletions(-) diff --git a/target/arm/tcg/translate-a64.c b/target/arm/tcg/translate-a64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/tcg/translate-a64.c +++ b/target/arm/tcg/translate-a64.c @@ -XXX,XX +XXX,XX @@ static void disas_ldst_pair(DisasContext *s, uint32_t insn) } else { TCGv_i64 tcg_rt = cpu_reg(s, rt); TCGv_i64 tcg_rt2 = cpu_reg(s, rt2); + MemOp mop = size + 1; + + /* + * With LSE2, non-sign-extending pairs are treated atomically if + * aligned, and if unaligned one of the pair will be completely + * within a 16-byte block and that element will be atomic. + * Otherwise each element is separately atomic. + * In all cases, issue one operation with the correct atomicity. + * + * This treats sign-extending loads like zero-extending loads, + * since that reuses the most code below. + */ + if (s->align_mem) { + mop |= (size == 2 ? MO_ALIGN_4 : MO_ALIGN_8); + } + mop = finalize_memop_pair(s, mop); if (is_load) { - TCGv_i64 tmp = tcg_temp_new_i64(); + if (size == 2) { + int o2 = s->be_data == MO_LE ? 32 : 0; + int o1 = o2 ^ 32; - /* Do not modify tcg_rt before recognizing any exception - * from the second load. - */ - do_gpr_ld(s, tmp, clean_addr, size + is_signed * MO_SIGN, - false, false, 0, false, false); - tcg_gen_addi_i64(clean_addr, clean_addr, 1 << size); - do_gpr_ld(s, tcg_rt2, clean_addr, size + is_signed * MO_SIGN, - false, false, 0, false, false); + tcg_gen_qemu_ld_i64(tcg_rt, clean_addr, get_mem_index(s), mop); + if (is_signed) { + tcg_gen_sextract_i64(tcg_rt2, tcg_rt, o2, 32); + tcg_gen_sextract_i64(tcg_rt, tcg_rt, o1, 32); + } else { + tcg_gen_extract_i64(tcg_rt2, tcg_rt, o2, 32); + tcg_gen_extract_i64(tcg_rt, tcg_rt, o1, 32); + } + } else { + TCGv_i128 tmp = tcg_temp_new_i128(); - tcg_gen_mov_i64(tcg_rt, tmp); + tcg_gen_qemu_ld_i128(tmp, clean_addr, get_mem_index(s), mop); + if (s->be_data == MO_LE) { + tcg_gen_extr_i128_i64(tcg_rt, tcg_rt2, tmp); + } else { + tcg_gen_extr_i128_i64(tcg_rt2, tcg_rt, tmp); + } + } } else { - do_gpr_st(s, tcg_rt, clean_addr, size, - false, 0, false, false); - tcg_gen_addi_i64(clean_addr, clean_addr, 1 << size); - do_gpr_st(s, tcg_rt2, clean_addr, size, - false, 0, false, false); + if (size == 2) { + TCGv_i64 tmp = tcg_temp_new_i64(); + + if (s->be_data == MO_LE) { + tcg_gen_concat32_i64(tmp, tcg_rt, tcg_rt2); + } else { + tcg_gen_concat32_i64(tmp, tcg_rt2, tcg_rt); + } + tcg_gen_qemu_st_i64(tmp, clean_addr, get_mem_index(s), mop); + } else { + TCGv_i128 tmp = tcg_temp_new_i128(); + + if (s->be_data == MO_LE) { + tcg_gen_concat_i64_i128(tmp, tcg_rt, tcg_rt2); + } else { + tcg_gen_concat_i64_i128(tmp, tcg_rt2, tcg_rt); + } + tcg_gen_qemu_st_i128(tmp, clean_addr, get_mem_index(s), mop); + } } } -- 2.34.1 From: Richard Henderson <richard.henderson@linaro.org> We are going to need the complete memop beforehand, so let's not compute it twice. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230530191438.411344-11-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/tcg/translate-a64.c | 61 +++++++++++++++++++--------------- 1 file changed, 35 insertions(+), 26 deletions(-) diff --git a/target/arm/tcg/translate-a64.c b/target/arm/tcg/translate-a64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/tcg/translate-a64.c +++ b/target/arm/tcg/translate-a64.c @@ -XXX,XX +XXX,XX @@ static void do_gpr_st_memidx(DisasContext *s, TCGv_i64 source, unsigned int iss_srt, bool iss_sf, bool iss_ar) { - memop = finalize_memop(s, memop); tcg_gen_qemu_st_i64(source, tcg_addr, memidx, memop); if (iss_valid) { @@ -XXX,XX +XXX,XX @@ static void do_gpr_ld_memidx(DisasContext *s, TCGv_i64 dest, TCGv_i64 tcg_addr, bool iss_valid, unsigned int iss_srt, bool iss_sf, bool iss_ar) { - memop = finalize_memop(s, memop); tcg_gen_qemu_ld_i64(dest, tcg_addr, memidx, memop); if (extend && (memop & MO_SIGN)) { @@ -XXX,XX +XXX,XX @@ static void disas_ldst_excl(DisasContext *s, uint32_t insn) int o2_L_o1_o0 = extract32(insn, 21, 3) * 2 | is_lasr; int size = extract32(insn, 30, 2); TCGv_i64 clean_addr; + MemOp memop; switch (o2_L_o1_o0) { case 0x0: /* STXR */ @@ -XXX,XX +XXX,XX @@ static void disas_ldst_excl(DisasContext *s, uint32_t insn) gen_check_sp_alignment(s); } tcg_gen_mb(TCG_MO_ALL | TCG_BAR_STRL); + /* TODO: ARMv8.4-LSE SCTLR.nAA */ + memop = finalize_memop(s, size | MO_ALIGN); clean_addr = gen_mte_check1(s, cpu_reg_sp(s, rn), true, rn != 31, size); - /* TODO: ARMv8.4-LSE SCTLR.nAA */ - do_gpr_st(s, cpu_reg(s, rt), clean_addr, size | MO_ALIGN, true, rt, + do_gpr_st(s, cpu_reg(s, rt), clean_addr, memop, true, rt, disas_ldst_compute_iss_sf(size, false, 0), is_lasr); return; @@ -XXX,XX +XXX,XX @@ static void disas_ldst_excl(DisasContext *s, uint32_t insn) if (rn == 31) { gen_check_sp_alignment(s); } + /* TODO: ARMv8.4-LSE SCTLR.nAA */ + memop = finalize_memop(s, size | MO_ALIGN); clean_addr = gen_mte_check1(s, cpu_reg_sp(s, rn), false, rn != 31, size); - /* TODO: ARMv8.4-LSE SCTLR.nAA */ - do_gpr_ld(s, cpu_reg(s, rt), clean_addr, size | MO_ALIGN, false, true, + do_gpr_ld(s, cpu_reg(s, rt), clean_addr, memop, false, true, rt, disas_ldst_compute_iss_sf(size, false, 0), is_lasr); tcg_gen_mb(TCG_MO_ALL | TCG_BAR_LDAQ); return; @@ -XXX,XX +XXX,XX @@ static void disas_ld_lit(DisasContext *s, uint32_t insn) } else { /* Only unsigned 32bit loads target 32bit registers. */ bool iss_sf = opc != 0; + MemOp memop = finalize_memop(s, size + is_signed * MO_SIGN); - do_gpr_ld(s, tcg_rt, clean_addr, size + is_signed * MO_SIGN, - false, true, rt, iss_sf, false); + do_gpr_ld(s, tcg_rt, clean_addr, memop, false, true, rt, iss_sf, false); } } @@ -XXX,XX +XXX,XX @@ static void disas_ldst_reg_imm9(DisasContext *s, uint32_t insn, bool post_index; bool writeback; int memidx; - + MemOp memop; TCGv_i64 clean_addr, dirty_addr; if (is_vector) { @@ -XXX,XX +XXX,XX @@ static void disas_ldst_reg_imm9(DisasContext *s, uint32_t insn, return; } is_store = (opc == 0); - is_signed = extract32(opc, 1, 1); + is_signed = !is_store && extract32(opc, 1, 1); is_extended = (size < 3) && extract32(opc, 0, 1); } @@ -XXX,XX +XXX,XX @@ static void disas_ldst_reg_imm9(DisasContext *s, uint32_t insn, } memidx = is_unpriv ? get_a64_user_mem_index(s) : get_mem_index(s); + memop = finalize_memop(s, size + is_signed * MO_SIGN); + clean_addr = gen_mte_check1_mmuidx(s, dirty_addr, is_store, writeback || rn != 31, size, is_unpriv, memidx); @@ -XXX,XX +XXX,XX @@ static void disas_ldst_reg_imm9(DisasContext *s, uint32_t insn, bool iss_sf = disas_ldst_compute_iss_sf(size, is_signed, opc); if (is_store) { - do_gpr_st_memidx(s, tcg_rt, clean_addr, size, memidx, + do_gpr_st_memidx(s, tcg_rt, clean_addr, memop, memidx, iss_valid, rt, iss_sf, false); } else { - do_gpr_ld_memidx(s, tcg_rt, clean_addr, size + is_signed * MO_SIGN, + do_gpr_ld_memidx(s, tcg_rt, clean_addr, memop, is_extended, memidx, iss_valid, rt, iss_sf, false); } @@ -XXX,XX +XXX,XX @@ static void disas_ldst_reg_roffset(DisasContext *s, uint32_t insn, bool is_signed = false; bool is_store = false; bool is_extended = false; - TCGv_i64 tcg_rm, clean_addr, dirty_addr; + MemOp memop; if (extract32(opt, 1, 1) == 0) { unallocated_encoding(s); @@ -XXX,XX +XXX,XX @@ static void disas_ldst_reg_roffset(DisasContext *s, uint32_t insn, return; } is_store = (opc == 0); - is_signed = extract32(opc, 1, 1); + is_signed = !is_store && extract32(opc, 1, 1); is_extended = (size < 3) && extract32(opc, 0, 1); } @@ -XXX,XX +XXX,XX @@ static void disas_ldst_reg_roffset(DisasContext *s, uint32_t insn, ext_and_shift_reg(tcg_rm, tcg_rm, opt, shift ? size : 0); tcg_gen_add_i64(dirty_addr, dirty_addr, tcg_rm); + + memop = finalize_memop(s, size + is_signed * MO_SIGN); clean_addr = gen_mte_check1(s, dirty_addr, is_store, true, size); if (is_vector) { @@ -XXX,XX +XXX,XX @@ static void disas_ldst_reg_roffset(DisasContext *s, uint32_t insn, } else { TCGv_i64 tcg_rt = cpu_reg(s, rt); bool iss_sf = disas_ldst_compute_iss_sf(size, is_signed, opc); + if (is_store) { - do_gpr_st(s, tcg_rt, clean_addr, size, + do_gpr_st(s, tcg_rt, clean_addr, memop, true, rt, iss_sf, false); } else { - do_gpr_ld(s, tcg_rt, clean_addr, size + is_signed * MO_SIGN, + do_gpr_ld(s, tcg_rt, clean_addr, memop, is_extended, true, rt, iss_sf, false); } } @@ -XXX,XX +XXX,XX @@ static void disas_ldst_reg_unsigned_imm(DisasContext *s, uint32_t insn, int rn = extract32(insn, 5, 5); unsigned int imm12 = extract32(insn, 10, 12); unsigned int offset; - TCGv_i64 clean_addr, dirty_addr; - bool is_store; bool is_signed = false; bool is_extended = false; + MemOp memop; if (is_vector) { size |= (opc & 2) << 1; @@ -XXX,XX +XXX,XX @@ static void disas_ldst_reg_unsigned_imm(DisasContext *s, uint32_t insn, return; } is_store = (opc == 0); - is_signed = extract32(opc, 1, 1); + is_signed = !is_store && extract32(opc, 1, 1); is_extended = (size < 3) && extract32(opc, 0, 1); } @@ -XXX,XX +XXX,XX @@ static void disas_ldst_reg_unsigned_imm(DisasContext *s, uint32_t insn, dirty_addr = read_cpu_reg_sp(s, rn, 1); offset = imm12 << size; tcg_gen_addi_i64(dirty_addr, dirty_addr, offset); + + memop = finalize_memop(s, size + is_signed * MO_SIGN); clean_addr = gen_mte_check1(s, dirty_addr, is_store, rn != 31, size); if (is_vector) { @@ -XXX,XX +XXX,XX @@ static void disas_ldst_reg_unsigned_imm(DisasContext *s, uint32_t insn, TCGv_i64 tcg_rt = cpu_reg(s, rt); bool iss_sf = disas_ldst_compute_iss_sf(size, is_signed, opc); if (is_store) { - do_gpr_st(s, tcg_rt, clean_addr, size, - true, rt, iss_sf, false); + do_gpr_st(s, tcg_rt, clean_addr, memop, true, rt, iss_sf, false); } else { - do_gpr_ld(s, tcg_rt, clean_addr, size + is_signed * MO_SIGN, + do_gpr_ld(s, tcg_rt, clean_addr, memop, is_extended, true, rt, iss_sf, false); } } @@ -XXX,XX +XXX,XX @@ static void disas_ldst_atomic(DisasContext *s, uint32_t insn, bool a = extract32(insn, 23, 1); TCGv_i64 tcg_rs, tcg_rt, clean_addr; AtomicThreeOpFn *fn = NULL; - MemOp mop = s->be_data | size | MO_ALIGN; + MemOp mop = finalize_memop(s, size | MO_ALIGN); if (is_vector || !dc_isar_feature(aa64_atomics, s)) { unallocated_encoding(s); @@ -XXX,XX +XXX,XX @@ static void disas_ldst_atomic(DisasContext *s, uint32_t insn, * full load-acquire (we only need "load-acquire processor consistent"), * but we choose to implement them as full LDAQ. */ - do_gpr_ld(s, cpu_reg(s, rt), clean_addr, size, false, + do_gpr_ld(s, cpu_reg(s, rt), clean_addr, mop, false, true, rt, disas_ldst_compute_iss_sf(size, false, 0), true); tcg_gen_mb(TCG_MO_ALL | TCG_BAR_LDAQ); return; @@ -XXX,XX +XXX,XX @@ static void disas_ldst_pac(DisasContext *s, uint32_t insn, bool use_key_a = !extract32(insn, 23, 1); int offset; TCGv_i64 clean_addr, dirty_addr, tcg_rt; + MemOp memop; if (size != 3 || is_vector || !dc_isar_feature(aa64_pauth, s)) { unallocated_encoding(s); @@ -XXX,XX +XXX,XX @@ static void disas_ldst_pac(DisasContext *s, uint32_t insn, offset = sextract32(offset << size, 0, 10 + size); tcg_gen_addi_i64(dirty_addr, dirty_addr, offset); + memop = finalize_memop(s, size); + /* Note that "clean" and "dirty" here refer to TBI not PAC. */ clean_addr = gen_mte_check1(s, dirty_addr, false, is_wback || rn != 31, size); tcg_rt = cpu_reg(s, rt); - do_gpr_ld(s, tcg_rt, clean_addr, size, + do_gpr_ld(s, tcg_rt, clean_addr, memop, /* extend */ false, /* iss_valid */ !is_wback, /* iss_srt */ rt, /* iss_sf */ true, /* iss_ar */ false); @@ -XXX,XX +XXX,XX @@ static void disas_ldst_ldapr_stlr(DisasContext *s, uint32_t insn) } /* TODO: ARMv8.4-LSE SCTLR.nAA */ - mop = size | MO_ALIGN; + mop = finalize_memop(s, size | MO_ALIGN); switch (opc) { case 0: /* STLURB */ -- 2.34.1 From: Richard Henderson <richard.henderson@linaro.org> We are going to need the complete memop beforehand, so let's not compute it twice. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230530191438.411344-12-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/tcg/translate-a64.c | 43 ++++++++++++++++++---------------- 1 file changed, 23 insertions(+), 20 deletions(-) diff --git a/target/arm/tcg/translate-a64.c b/target/arm/tcg/translate-a64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/tcg/translate-a64.c +++ b/target/arm/tcg/translate-a64.c @@ -XXX,XX +XXX,XX @@ static void do_gpr_ld(DisasContext *s, TCGv_i64 dest, TCGv_i64 tcg_addr, /* * Store from FP register to memory */ -static void do_fp_st(DisasContext *s, int srcidx, TCGv_i64 tcg_addr, int size) +static void do_fp_st(DisasContext *s, int srcidx, TCGv_i64 tcg_addr, MemOp mop) { /* This writes the bottom N bits of a 128 bit wide vector to memory */ TCGv_i64 tmplo = tcg_temp_new_i64(); - MemOp mop = finalize_memop_asimd(s, size); tcg_gen_ld_i64(tmplo, cpu_env, fp_reg_offset(s, srcidx, MO_64)); - if (size < MO_128) { + if ((mop & MO_SIZE) < MO_128) { tcg_gen_qemu_st_i64(tmplo, tcg_addr, get_mem_index(s), mop); } else { TCGv_i64 tmphi = tcg_temp_new_i64(); @@ -XXX,XX +XXX,XX @@ static void do_fp_st(DisasContext *s, int srcidx, TCGv_i64 tcg_addr, int size) /* * Load from memory to FP register */ -static void do_fp_ld(DisasContext *s, int destidx, TCGv_i64 tcg_addr, int size) +static void do_fp_ld(DisasContext *s, int destidx, TCGv_i64 tcg_addr, MemOp mop) { /* This always zero-extends and writes to a full 128 bit wide vector */ TCGv_i64 tmplo = tcg_temp_new_i64(); TCGv_i64 tmphi = NULL; - MemOp mop = finalize_memop_asimd(s, size); - if (size < MO_128) { + if ((mop & MO_SIZE) < MO_128) { tcg_gen_qemu_ld_i64(tmplo, tcg_addr, get_mem_index(s), mop); } else { TCGv_i128 t16 = tcg_temp_new_i128(); @@ -XXX,XX +XXX,XX @@ static void disas_ld_lit(DisasContext *s, uint32_t insn) bool is_signed = false; int size = 2; TCGv_i64 tcg_rt, clean_addr; + MemOp memop; if (is_vector) { if (opc == 3) { @@ -XXX,XX +XXX,XX @@ static void disas_ld_lit(DisasContext *s, uint32_t insn) if (!fp_access_check(s)) { return; } + memop = finalize_memop_asimd(s, size); } else { if (opc == 3) { /* PRFM (literal) : prefetch */ @@ -XXX,XX +XXX,XX @@ static void disas_ld_lit(DisasContext *s, uint32_t insn) } size = 2 + extract32(opc, 0, 1); is_signed = extract32(opc, 1, 1); + memop = finalize_memop(s, size + is_signed * MO_SIGN); } tcg_rt = cpu_reg(s, rt); clean_addr = tcg_temp_new_i64(); gen_pc_plus_diff(s, clean_addr, imm); + if (is_vector) { - do_fp_ld(s, rt, clean_addr, size); + do_fp_ld(s, rt, clean_addr, memop); } else { /* Only unsigned 32bit loads target 32bit registers. */ bool iss_sf = opc != 0; - MemOp memop = finalize_memop(s, size + is_signed * MO_SIGN); - do_gpr_ld(s, tcg_rt, clean_addr, memop, false, true, rt, iss_sf, false); } } @@ -XXX,XX +XXX,XX @@ static void disas_ldst_pair(DisasContext *s, uint32_t insn) (wback || rn != 31) && !set_tag, 2 << size); if (is_vector) { + MemOp mop = finalize_memop_asimd(s, size); + if (is_load) { - do_fp_ld(s, rt, clean_addr, size); + do_fp_ld(s, rt, clean_addr, mop); } else { - do_fp_st(s, rt, clean_addr, size); + do_fp_st(s, rt, clean_addr, mop); } tcg_gen_addi_i64(clean_addr, clean_addr, 1 << size); if (is_load) { - do_fp_ld(s, rt2, clean_addr, size); + do_fp_ld(s, rt2, clean_addr, mop); } else { - do_fp_st(s, rt2, clean_addr, size); + do_fp_st(s, rt2, clean_addr, mop); } } else { TCGv_i64 tcg_rt = cpu_reg(s, rt); @@ -XXX,XX +XXX,XX @@ static void disas_ldst_reg_imm9(DisasContext *s, uint32_t insn, if (!fp_access_check(s)) { return; } + memop = finalize_memop_asimd(s, size); } else { if (size == 3 && opc == 2) { /* PRFM - prefetch */ @@ -XXX,XX +XXX,XX @@ static void disas_ldst_reg_imm9(DisasContext *s, uint32_t insn, is_store = (opc == 0); is_signed = !is_store && extract32(opc, 1, 1); is_extended = (size < 3) && extract32(opc, 0, 1); + memop = finalize_memop(s, size + is_signed * MO_SIGN); } switch (idx) { @@ -XXX,XX +XXX,XX @@ static void disas_ldst_reg_imm9(DisasContext *s, uint32_t insn, } memidx = is_unpriv ? get_a64_user_mem_index(s) : get_mem_index(s); - memop = finalize_memop(s, size + is_signed * MO_SIGN); clean_addr = gen_mte_check1_mmuidx(s, dirty_addr, is_store, writeback || rn != 31, @@ -XXX,XX +XXX,XX @@ static void disas_ldst_reg_imm9(DisasContext *s, uint32_t insn, if (is_vector) { if (is_store) { - do_fp_st(s, rt, clean_addr, size); + do_fp_st(s, rt, clean_addr, memop); } else { - do_fp_ld(s, rt, clean_addr, size); + do_fp_ld(s, rt, clean_addr, memop); } } else { TCGv_i64 tcg_rt = cpu_reg(s, rt); @@ -XXX,XX +XXX,XX @@ static void disas_ldst_reg_roffset(DisasContext *s, uint32_t insn, if (is_vector) { if (is_store) { - do_fp_st(s, rt, clean_addr, size); + do_fp_st(s, rt, clean_addr, memop); } else { - do_fp_ld(s, rt, clean_addr, size); + do_fp_ld(s, rt, clean_addr, memop); } } else { TCGv_i64 tcg_rt = cpu_reg(s, rt); @@ -XXX,XX +XXX,XX @@ static void disas_ldst_reg_unsigned_imm(DisasContext *s, uint32_t insn, if (is_vector) { if (is_store) { - do_fp_st(s, rt, clean_addr, size); + do_fp_st(s, rt, clean_addr, memop); } else { - do_fp_ld(s, rt, clean_addr, size); + do_fp_ld(s, rt, clean_addr, memop); } } else { TCGv_i64 tcg_rt = cpu_reg(s, rt); -- 2.34.1 From: Richard Henderson <richard.henderson@linaro.org> Pass the completed memop to gen_mte_check1_mmuidx. For the moment, do nothing more than extract the size. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230530191438.411344-13-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/tcg/translate-a64.h | 2 +- target/arm/tcg/translate-a64.c | 82 ++++++++++++++++++---------------- target/arm/tcg/translate-sve.c | 7 +-- 3 files changed, 49 insertions(+), 42 deletions(-) diff --git a/target/arm/tcg/translate-a64.h b/target/arm/tcg/translate-a64.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/tcg/translate-a64.h +++ b/target/arm/tcg/translate-a64.h @@ -XXX,XX +XXX,XX @@ static inline bool sme_smza_enabled_check(DisasContext *s) TCGv_i64 clean_data_tbi(DisasContext *s, TCGv_i64 addr); TCGv_i64 gen_mte_check1(DisasContext *s, TCGv_i64 addr, bool is_write, - bool tag_checked, int log2_size); + bool tag_checked, MemOp memop); TCGv_i64 gen_mte_checkN(DisasContext *s, TCGv_i64 addr, bool is_write, bool tag_checked, int size); diff --git a/target/arm/tcg/translate-a64.c b/target/arm/tcg/translate-a64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/tcg/translate-a64.c +++ b/target/arm/tcg/translate-a64.c @@ -XXX,XX +XXX,XX @@ static void gen_probe_access(DisasContext *s, TCGv_i64 ptr, */ static TCGv_i64 gen_mte_check1_mmuidx(DisasContext *s, TCGv_i64 addr, bool is_write, bool tag_checked, - int log2_size, bool is_unpriv, + MemOp memop, bool is_unpriv, int core_idx) { if (tag_checked && s->mte_active[is_unpriv]) { @@ -XXX,XX +XXX,XX @@ static TCGv_i64 gen_mte_check1_mmuidx(DisasContext *s, TCGv_i64 addr, desc = FIELD_DP32(desc, MTEDESC, TBI, s->tbid); desc = FIELD_DP32(desc, MTEDESC, TCMA, s->tcma); desc = FIELD_DP32(desc, MTEDESC, WRITE, is_write); - desc = FIELD_DP32(desc, MTEDESC, SIZEM1, (1 << log2_size) - 1); + desc = FIELD_DP32(desc, MTEDESC, SIZEM1, memop_size(memop) - 1); ret = tcg_temp_new_i64(); gen_helper_mte_check(ret, cpu_env, tcg_constant_i32(desc), addr); @@ -XXX,XX +XXX,XX @@ static TCGv_i64 gen_mte_check1_mmuidx(DisasContext *s, TCGv_i64 addr, } TCGv_i64 gen_mte_check1(DisasContext *s, TCGv_i64 addr, bool is_write, - bool tag_checked, int log2_size) + bool tag_checked, MemOp memop) { - return gen_mte_check1_mmuidx(s, addr, is_write, tag_checked, log2_size, + return gen_mte_check1_mmuidx(s, addr, is_write, tag_checked, memop, false, get_mem_index(s)); } @@ -XXX,XX +XXX,XX @@ static void gen_load_exclusive(DisasContext *s, int rt, int rt2, int rn, int size, bool is_pair) { int idx = get_mem_index(s); - MemOp memop; TCGv_i64 dirty_addr, clean_addr; + MemOp memop; + + /* + * For pairs: + * if size == 2, the operation is single-copy atomic for the doubleword. + * if size == 3, the operation is single-copy atomic for *each* doubleword, + * not the entire quadword, however it must be quadword aligned. + */ + memop = size + is_pair; + if (memop == MO_128) { + memop = finalize_memop_atom(s, MO_128 | MO_ALIGN, + MO_ATOM_IFALIGN_PAIR); + } else { + memop = finalize_memop(s, memop | MO_ALIGN); + } s->is_ldex = true; dirty_addr = cpu_reg_sp(s, rn); - clean_addr = gen_mte_check1(s, dirty_addr, false, rn != 31, size); + clean_addr = gen_mte_check1(s, dirty_addr, false, rn != 31, memop); g_assert(size <= 3); if (is_pair) { g_assert(size >= 2); if (size == 2) { - /* The pair must be single-copy atomic for the doubleword. */ - memop = finalize_memop(s, MO_64 | MO_ALIGN); tcg_gen_qemu_ld_i64(cpu_exclusive_val, clean_addr, idx, memop); if (s->be_data == MO_LE) { tcg_gen_extract_i64(cpu_reg(s, rt), cpu_exclusive_val, 0, 32); @@ -XXX,XX +XXX,XX @@ static void gen_load_exclusive(DisasContext *s, int rt, int rt2, int rn, tcg_gen_extract_i64(cpu_reg(s, rt2), cpu_exclusive_val, 0, 32); } } else { - /* - * The pair must be single-copy atomic for *each* doubleword, not - * the entire quadword, however it must be quadword aligned. - * Expose the complete load to tcg, for ease of tlb lookup, - * but indicate that only 8-byte atomicity is required. - */ TCGv_i128 t16 = tcg_temp_new_i128(); - memop = finalize_memop_atom(s, MO_128 | MO_ALIGN_16, - MO_ATOM_IFALIGN_PAIR); tcg_gen_qemu_ld_i128(t16, clean_addr, idx, memop); if (s->be_data == MO_LE) { @@ -XXX,XX +XXX,XX @@ static void gen_load_exclusive(DisasContext *s, int rt, int rt2, int rn, tcg_gen_mov_i64(cpu_reg(s, rt2), cpu_exclusive_high); } } else { - memop = finalize_memop(s, size | MO_ALIGN); tcg_gen_qemu_ld_i64(cpu_exclusive_val, clean_addr, idx, memop); tcg_gen_mov_i64(cpu_reg(s, rt), cpu_exclusive_val); } @@ -XXX,XX +XXX,XX @@ static void gen_store_exclusive(DisasContext *s, int rd, int rt, int rt2, TCGLabel *fail_label = gen_new_label(); TCGLabel *done_label = gen_new_label(); TCGv_i64 tmp, dirty_addr, clean_addr; + MemOp memop; + + memop = (size + is_pair) | MO_ALIGN; + memop = finalize_memop(s, memop); dirty_addr = cpu_reg_sp(s, rn); - clean_addr = gen_mte_check1(s, dirty_addr, true, rn != 31, size); + clean_addr = gen_mte_check1(s, dirty_addr, true, rn != 31, memop); tcg_gen_brcond_i64(TCG_COND_NE, clean_addr, cpu_exclusive_addr, fail_label); @@ -XXX,XX +XXX,XX @@ static void gen_store_exclusive(DisasContext *s, int rd, int rt, int rt2, } tcg_gen_atomic_cmpxchg_i64(tmp, cpu_exclusive_addr, cpu_exclusive_val, tmp, - get_mem_index(s), - MO_64 | MO_ALIGN | s->be_data); + get_mem_index(s), memop); tcg_gen_setcond_i64(TCG_COND_NE, tmp, tmp, cpu_exclusive_val); } else { TCGv_i128 t16 = tcg_temp_new_i128(); @@ -XXX,XX +XXX,XX @@ static void gen_store_exclusive(DisasContext *s, int rd, int rt, int rt2, } tcg_gen_atomic_cmpxchg_i128(t16, cpu_exclusive_addr, c16, t16, - get_mem_index(s), - MO_128 | MO_ALIGN | s->be_data); + get_mem_index(s), memop); a = tcg_temp_new_i64(); b = tcg_temp_new_i64(); @@ -XXX,XX +XXX,XX @@ static void gen_store_exclusive(DisasContext *s, int rd, int rt, int rt2, } } else { tcg_gen_atomic_cmpxchg_i64(tmp, cpu_exclusive_addr, cpu_exclusive_val, - cpu_reg(s, rt), get_mem_index(s), - size | MO_ALIGN | s->be_data); + cpu_reg(s, rt), get_mem_index(s), memop); tcg_gen_setcond_i64(TCG_COND_NE, tmp, tmp, cpu_exclusive_val); } tcg_gen_mov_i64(cpu_reg(s, rd), tmp); @@ -XXX,XX +XXX,XX @@ static void gen_compare_and_swap(DisasContext *s, int rs, int rt, TCGv_i64 tcg_rt = cpu_reg(s, rt); int memidx = get_mem_index(s); TCGv_i64 clean_addr; + MemOp memop; if (rn == 31) { gen_check_sp_alignment(s); } - clean_addr = gen_mte_check1(s, cpu_reg_sp(s, rn), true, rn != 31, size); - tcg_gen_atomic_cmpxchg_i64(tcg_rs, clean_addr, tcg_rs, tcg_rt, memidx, - size | MO_ALIGN | s->be_data); + memop = finalize_memop(s, size | MO_ALIGN); + clean_addr = gen_mte_check1(s, cpu_reg_sp(s, rn), true, rn != 31, memop); + tcg_gen_atomic_cmpxchg_i64(tcg_rs, clean_addr, tcg_rs, tcg_rt, + memidx, memop); } static void gen_compare_and_swap_pair(DisasContext *s, int rs, int rt, @@ -XXX,XX +XXX,XX @@ static void gen_compare_and_swap_pair(DisasContext *s, int rs, int rt, TCGv_i64 t2 = cpu_reg(s, rt + 1); TCGv_i64 clean_addr; int memidx = get_mem_index(s); + MemOp memop; if (rn == 31) { gen_check_sp_alignment(s); } /* This is a single atomic access, despite the "pair". */ - clean_addr = gen_mte_check1(s, cpu_reg_sp(s, rn), true, rn != 31, size + 1); + memop = finalize_memop(s, (size + 1) | MO_ALIGN); + clean_addr = gen_mte_check1(s, cpu_reg_sp(s, rn), true, rn != 31, memop); if (size == 2) { TCGv_i64 cmp = tcg_temp_new_i64(); @@ -XXX,XX +XXX,XX @@ static void gen_compare_and_swap_pair(DisasContext *s, int rs, int rt, tcg_gen_concat32_i64(cmp, s2, s1); } - tcg_gen_atomic_cmpxchg_i64(cmp, clean_addr, cmp, val, memidx, - MO_64 | MO_ALIGN | s->be_data); + tcg_gen_atomic_cmpxchg_i64(cmp, clean_addr, cmp, val, memidx, memop); if (s->be_data == MO_LE) { tcg_gen_extr32_i64(s1, s2, cmp); @@ -XXX,XX +XXX,XX @@ static void gen_compare_and_swap_pair(DisasContext *s, int rs, int rt, tcg_gen_concat_i64_i128(cmp, s2, s1); } - tcg_gen_atomic_cmpxchg_i128(cmp, clean_addr, cmp, val, memidx, - MO_128 | MO_ALIGN | s->be_data); + tcg_gen_atomic_cmpxchg_i128(cmp, clean_addr, cmp, val, memidx, memop); if (s->be_data == MO_LE) { tcg_gen_extr_i128_i64(s1, s2, cmp); @@ -XXX,XX +XXX,XX @@ static void disas_ldst_excl(DisasContext *s, uint32_t insn) /* TODO: ARMv8.4-LSE SCTLR.nAA */ memop = finalize_memop(s, size | MO_ALIGN); clean_addr = gen_mte_check1(s, cpu_reg_sp(s, rn), - true, rn != 31, size); + true, rn != 31, memop); do_gpr_st(s, cpu_reg(s, rt), clean_addr, memop, true, rt, disas_ldst_compute_iss_sf(size, false, 0), is_lasr); return; @@ -XXX,XX +XXX,XX @@ static void disas_ldst_excl(DisasContext *s, uint32_t insn) /* TODO: ARMv8.4-LSE SCTLR.nAA */ memop = finalize_memop(s, size | MO_ALIGN); clean_addr = gen_mte_check1(s, cpu_reg_sp(s, rn), - false, rn != 31, size); + false, rn != 31, memop); do_gpr_ld(s, cpu_reg(s, rt), clean_addr, memop, false, true, rt, disas_ldst_compute_iss_sf(size, false, 0), is_lasr); tcg_gen_mb(TCG_MO_ALL | TCG_BAR_LDAQ); @@ -XXX,XX +XXX,XX @@ static void disas_ldst_reg_roffset(DisasContext *s, uint32_t insn, tcg_gen_add_i64(dirty_addr, dirty_addr, tcg_rm); memop = finalize_memop(s, size + is_signed * MO_SIGN); - clean_addr = gen_mte_check1(s, dirty_addr, is_store, true, size); + clean_addr = gen_mte_check1(s, dirty_addr, is_store, true, memop); if (is_vector) { if (is_store) { @@ -XXX,XX +XXX,XX @@ static void disas_ldst_reg_unsigned_imm(DisasContext *s, uint32_t insn, tcg_gen_addi_i64(dirty_addr, dirty_addr, offset); memop = finalize_memop(s, size + is_signed * MO_SIGN); - clean_addr = gen_mte_check1(s, dirty_addr, is_store, rn != 31, size); + clean_addr = gen_mte_check1(s, dirty_addr, is_store, rn != 31, memop); if (is_vector) { if (is_store) { @@ -XXX,XX +XXX,XX @@ static void disas_ldst_atomic(DisasContext *s, uint32_t insn, if (rn == 31) { gen_check_sp_alignment(s); } - clean_addr = gen_mte_check1(s, cpu_reg_sp(s, rn), false, rn != 31, size); + clean_addr = gen_mte_check1(s, cpu_reg_sp(s, rn), false, rn != 31, mop); if (o3_opc == 014) { /* @@ -XXX,XX +XXX,XX @@ static void disas_ldst_pac(DisasContext *s, uint32_t insn, /* Note that "clean" and "dirty" here refer to TBI not PAC. */ clean_addr = gen_mte_check1(s, dirty_addr, false, - is_wback || rn != 31, size); + is_wback || rn != 31, memop); tcg_rt = cpu_reg(s, rt); do_gpr_ld(s, tcg_rt, clean_addr, memop, diff --git a/target/arm/tcg/translate-sve.c b/target/arm/tcg/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/tcg/translate-sve.c +++ b/target/arm/tcg/translate-sve.c @@ -XXX,XX +XXX,XX @@ static bool trans_LD1R_zpri(DisasContext *s, arg_rpri_load *a) unsigned msz = dtype_msz(a->dtype); TCGLabel *over; TCGv_i64 temp, clean_addr; + MemOp memop; if (!dc_isar_feature(aa64_sve, s)) { return false; @@ -XXX,XX +XXX,XX @@ static bool trans_LD1R_zpri(DisasContext *s, arg_rpri_load *a) /* Load the data. */ temp = tcg_temp_new_i64(); tcg_gen_addi_i64(temp, cpu_reg_sp(s, a->rn), a->imm << msz); - clean_addr = gen_mte_check1(s, temp, false, true, msz); - tcg_gen_qemu_ld_i64(temp, clean_addr, get_mem_index(s), - finalize_memop(s, dtype_mop[a->dtype])); + memop = finalize_memop(s, dtype_mop[a->dtype]); + clean_addr = gen_mte_check1(s, temp, false, true, memop); + tcg_gen_qemu_ld_i64(temp, clean_addr, get_mem_index(s), memop); /* Broadcast to *all* elements. */ tcg_gen_gvec_dup_i64(esz, vec_full_reg_offset(s, a->rd), -- 2.34.1 From: Richard Henderson <richard.henderson@linaro.org> Pass the individual memop to gen_mte_checkN. For the moment, do nothing with it. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230530191438.411344-14-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/tcg/translate-a64.h | 2 +- target/arm/tcg/translate-a64.c | 31 +++++++++++++++++++------------ target/arm/tcg/translate-sve.c | 4 ++-- 3 files changed, 22 insertions(+), 15 deletions(-) diff --git a/target/arm/tcg/translate-a64.h b/target/arm/tcg/translate-a64.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/tcg/translate-a64.h +++ b/target/arm/tcg/translate-a64.h @@ -XXX,XX +XXX,XX @@ TCGv_i64 clean_data_tbi(DisasContext *s, TCGv_i64 addr); TCGv_i64 gen_mte_check1(DisasContext *s, TCGv_i64 addr, bool is_write, bool tag_checked, MemOp memop); TCGv_i64 gen_mte_checkN(DisasContext *s, TCGv_i64 addr, bool is_write, - bool tag_checked, int size); + bool tag_checked, int total_size, MemOp memop); /* We should have at some point before trying to access an FP register * done the necessary access check, so assert that diff --git a/target/arm/tcg/translate-a64.c b/target/arm/tcg/translate-a64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/tcg/translate-a64.c +++ b/target/arm/tcg/translate-a64.c @@ -XXX,XX +XXX,XX @@ TCGv_i64 gen_mte_check1(DisasContext *s, TCGv_i64 addr, bool is_write, * For MTE, check multiple logical sequential accesses. */ TCGv_i64 gen_mte_checkN(DisasContext *s, TCGv_i64 addr, bool is_write, - bool tag_checked, int size) + bool tag_checked, int total_size, MemOp single_mop) { if (tag_checked && s->mte_active[0]) { TCGv_i64 ret; @@ -XXX,XX +XXX,XX @@ TCGv_i64 gen_mte_checkN(DisasContext *s, TCGv_i64 addr, bool is_write, desc = FIELD_DP32(desc, MTEDESC, TBI, s->tbid); desc = FIELD_DP32(desc, MTEDESC, TCMA, s->tcma); desc = FIELD_DP32(desc, MTEDESC, WRITE, is_write); - desc = FIELD_DP32(desc, MTEDESC, SIZEM1, size - 1); + desc = FIELD_DP32(desc, MTEDESC, SIZEM1, total_size - 1); ret = tcg_temp_new_i64(); gen_helper_mte_check(ret, cpu_env, tcg_constant_i32(desc), addr); @@ -XXX,XX +XXX,XX @@ static void disas_ldst_pair(DisasContext *s, uint32_t insn) bool is_vector = extract32(insn, 26, 1); bool is_load = extract32(insn, 22, 1); int opc = extract32(insn, 30, 2); - bool is_signed = false; bool postindex = false; bool wback = false; bool set_tag = false; - TCGv_i64 clean_addr, dirty_addr; - + MemOp mop; int size; if (opc == 3) { @@ -XXX,XX +XXX,XX @@ static void disas_ldst_pair(DisasContext *s, uint32_t insn) } } + if (is_vector) { + mop = finalize_memop_asimd(s, size); + } else { + mop = finalize_memop(s, size); + } clean_addr = gen_mte_checkN(s, dirty_addr, !is_load, - (wback || rn != 31) && !set_tag, 2 << size); + (wback || rn != 31) && !set_tag, + 2 << size, mop); if (is_vector) { - MemOp mop = finalize_memop_asimd(s, size); - + /* LSE2 does not merge FP pairs; leave these as separate operations. */ if (is_load) { do_fp_ld(s, rt, clean_addr, mop); } else { @@ -XXX,XX +XXX,XX @@ static void disas_ldst_pair(DisasContext *s, uint32_t insn) } else { TCGv_i64 tcg_rt = cpu_reg(s, rt); TCGv_i64 tcg_rt2 = cpu_reg(s, rt2); - MemOp mop = size + 1; /* + * We built mop above for the single logical access -- rebuild it + * now for the paired operation. + * * With LSE2, non-sign-extending pairs are treated atomically if * aligned, and if unaligned one of the pair will be completely * within a 16-byte block and that element will be atomic. @@ -XXX,XX +XXX,XX @@ static void disas_ldst_pair(DisasContext *s, uint32_t insn) * This treats sign-extending loads like zero-extending loads, * since that reuses the most code below. */ + mop = size + 1; if (s->align_mem) { mop |= (size == 2 ? MO_ALIGN_4 : MO_ALIGN_8); } @@ -XXX,XX +XXX,XX @@ static void disas_ldst_multiple_struct(DisasContext *s, uint32_t insn) * promote consecutive little-endian elements below. */ clean_addr = gen_mte_checkN(s, tcg_rn, is_store, is_postidx || rn != 31, - total); + total, finalize_memop(s, size)); /* * Consecutive little-endian elements from a single register @@ -XXX,XX +XXX,XX @@ static void disas_ldst_single_struct(DisasContext *s, uint32_t insn) total = selem << scale; tcg_rn = cpu_reg_sp(s, rn); - clean_addr = gen_mte_checkN(s, tcg_rn, !is_load, is_postidx || rn != 31, - total); mop = finalize_memop(s, scale); + clean_addr = gen_mte_checkN(s, tcg_rn, !is_load, is_postidx || rn != 31, + total, mop); + tcg_ebytes = tcg_constant_i64(1 << scale); for (xs = 0; xs < selem; xs++) { if (replicate) { diff --git a/target/arm/tcg/translate-sve.c b/target/arm/tcg/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/tcg/translate-sve.c +++ b/target/arm/tcg/translate-sve.c @@ -XXX,XX +XXX,XX @@ void gen_sve_ldr(DisasContext *s, TCGv_ptr base, int vofs, dirty_addr = tcg_temp_new_i64(); tcg_gen_addi_i64(dirty_addr, cpu_reg_sp(s, rn), imm); - clean_addr = gen_mte_checkN(s, dirty_addr, false, rn != 31, len); + clean_addr = gen_mte_checkN(s, dirty_addr, false, rn != 31, len, MO_8); /* * Note that unpredicated load/store of vector/predicate registers @@ -XXX,XX +XXX,XX @@ void gen_sve_str(DisasContext *s, TCGv_ptr base, int vofs, dirty_addr = tcg_temp_new_i64(); tcg_gen_addi_i64(dirty_addr, cpu_reg_sp(s, rn), imm); - clean_addr = gen_mte_checkN(s, dirty_addr, false, rn != 31, len); + clean_addr = gen_mte_checkN(s, dirty_addr, false, rn != 31, len, MO_8); /* Note that unpredicated load/store of vector/predicate registers * are defined as a stream of bytes, which equates to little-endian -- 2.34.1 From: Richard Henderson <richard.henderson@linaro.org> Fixes a bug in that with SCTLR.A set, we should raise any alignment fault before raising any MTE check fault. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230530191438.411344-15-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/internals.h | 3 ++- target/arm/tcg/mte_helper.c | 18 ++++++++++++++++++ target/arm/tcg/translate-a64.c | 2 ++ 3 files changed, 22 insertions(+), 1 deletion(-) diff --git a/target/arm/internals.h b/target/arm/internals.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/internals.h +++ b/target/arm/internals.h @@ -XXX,XX +XXX,XX @@ FIELD(MTEDESC, MIDX, 0, 4) FIELD(MTEDESC, TBI, 4, 2) FIELD(MTEDESC, TCMA, 6, 2) FIELD(MTEDESC, WRITE, 8, 1) -FIELD(MTEDESC, SIZEM1, 9, SIMD_DATA_BITS - 9) /* size - 1 */ +FIELD(MTEDESC, ALIGN, 9, 3) +FIELD(MTEDESC, SIZEM1, 12, SIMD_DATA_BITS - 12) /* size - 1 */ bool mte_probe(CPUARMState *env, uint32_t desc, uint64_t ptr); uint64_t mte_check(CPUARMState *env, uint32_t desc, uint64_t ptr, uintptr_t ra); diff --git a/target/arm/tcg/mte_helper.c b/target/arm/tcg/mte_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/tcg/mte_helper.c +++ b/target/arm/tcg/mte_helper.c @@ -XXX,XX +XXX,XX @@ uint64_t mte_check(CPUARMState *env, uint32_t desc, uint64_t ptr, uintptr_t ra) uint64_t HELPER(mte_check)(CPUARMState *env, uint32_t desc, uint64_t ptr) { + /* + * R_XCHFJ: Alignment check not caused by memory type is priority 1, + * higher than any translation fault. When MTE is disabled, tcg + * performs the alignment check during the code generated for the + * memory access. With MTE enabled, we must check this here before + * raising any translation fault in allocation_tag_mem. + */ + unsigned align = FIELD_EX32(desc, MTEDESC, ALIGN); + if (unlikely(align)) { + align = (1u << align) - 1; + if (unlikely(ptr & align)) { + int idx = FIELD_EX32(desc, MTEDESC, MIDX); + bool w = FIELD_EX32(desc, MTEDESC, WRITE); + MMUAccessType type = w ? MMU_DATA_STORE : MMU_DATA_LOAD; + arm_cpu_do_unaligned_access(env_cpu(env), ptr, type, idx, GETPC()); + } + } + return mte_check(env, desc, ptr, GETPC()); } diff --git a/target/arm/tcg/translate-a64.c b/target/arm/tcg/translate-a64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/tcg/translate-a64.c +++ b/target/arm/tcg/translate-a64.c @@ -XXX,XX +XXX,XX @@ static TCGv_i64 gen_mte_check1_mmuidx(DisasContext *s, TCGv_i64 addr, desc = FIELD_DP32(desc, MTEDESC, TBI, s->tbid); desc = FIELD_DP32(desc, MTEDESC, TCMA, s->tcma); desc = FIELD_DP32(desc, MTEDESC, WRITE, is_write); + desc = FIELD_DP32(desc, MTEDESC, ALIGN, get_alignment_bits(memop)); desc = FIELD_DP32(desc, MTEDESC, SIZEM1, memop_size(memop) - 1); ret = tcg_temp_new_i64(); @@ -XXX,XX +XXX,XX @@ TCGv_i64 gen_mte_checkN(DisasContext *s, TCGv_i64 addr, bool is_write, desc = FIELD_DP32(desc, MTEDESC, TBI, s->tbid); desc = FIELD_DP32(desc, MTEDESC, TCMA, s->tcma); desc = FIELD_DP32(desc, MTEDESC, WRITE, is_write); + desc = FIELD_DP32(desc, MTEDESC, ALIGN, get_alignment_bits(single_mop)); desc = FIELD_DP32(desc, MTEDESC, SIZEM1, total_size - 1); ret = tcg_temp_new_i64(); -- 2.34.1 From: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230530191438.411344-16-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/cpu.h | 3 ++- target/arm/tcg/translate.h | 2 ++ target/arm/tcg/hflags.c | 6 ++++++ target/arm/tcg/translate-a64.c | 1 + 4 files changed, 11 insertions(+), 1 deletion(-) diff --git a/target/arm/cpu.h b/target/arm/cpu.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.h +++ b/target/arm/cpu.h @@ -XXX,XX +XXX,XX @@ void pmu_init(ARMCPU *cpu); #define SCTLR_D (1U << 5) /* up to v5; RAO in v6 */ #define SCTLR_CP15BEN (1U << 5) /* v7 onward */ #define SCTLR_L (1U << 6) /* up to v5; RAO in v6 and v7; RAZ in v8 */ -#define SCTLR_nAA (1U << 6) /* when v8.4-LSE is implemented */ +#define SCTLR_nAA (1U << 6) /* when FEAT_LSE2 is implemented */ #define SCTLR_B (1U << 7) /* up to v6; RAZ in v7 */ #define SCTLR_ITD (1U << 7) /* v8 onward */ #define SCTLR_S (1U << 8) /* up to v6; RAZ in v7 */ @@ -XXX,XX +XXX,XX @@ FIELD(TBFLAG_A64, SVL, 24, 4) /* Indicates that SME Streaming mode is active, and SMCR_ELx.FA64 is not. */ FIELD(TBFLAG_A64, SME_TRAP_NONSTREAMING, 28, 1) FIELD(TBFLAG_A64, FGT_ERET, 29, 1) +FIELD(TBFLAG_A64, NAA, 30, 1) /* * Helpers for using the above. diff --git a/target/arm/tcg/translate.h b/target/arm/tcg/translate.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/tcg/translate.h +++ b/target/arm/tcg/translate.h @@ -XXX,XX +XXX,XX @@ typedef struct DisasContext { bool fgt_eret; /* True if fine-grained trap on SVC is enabled */ bool fgt_svc; + /* True if FEAT_LSE2 SCTLR_ELx.nAA is set */ + bool naa; /* * >= 0, a copy of PSTATE.BTYPE, which will be 0 without v8.5-BTI. * < 0, set by the current instruction. diff --git a/target/arm/tcg/hflags.c b/target/arm/tcg/hflags.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/tcg/hflags.c +++ b/target/arm/tcg/hflags.c @@ -XXX,XX +XXX,XX @@ static CPUARMTBFlags rebuild_hflags_a64(CPUARMState *env, int el, int fp_el, } } + if (cpu_isar_feature(aa64_lse2, env_archcpu(env))) { + if (sctlr & SCTLR_nAA) { + DP_TBFLAG_A64(flags, NAA, 1); + } + } + /* Compute the condition for using AccType_UNPRIV for LDTR et al. */ if (!(env->pstate & PSTATE_UAO)) { switch (mmu_idx) { diff --git a/target/arm/tcg/translate-a64.c b/target/arm/tcg/translate-a64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/tcg/translate-a64.c +++ b/target/arm/tcg/translate-a64.c @@ -XXX,XX +XXX,XX @@ static void aarch64_tr_init_disas_context(DisasContextBase *dcbase, dc->pstate_sm = EX_TBFLAG_A64(tb_flags, PSTATE_SM); dc->pstate_za = EX_TBFLAG_A64(tb_flags, PSTATE_ZA); dc->sme_trap_nonstreaming = EX_TBFLAG_A64(tb_flags, SME_TRAP_NONSTREAMING); + dc->naa = EX_TBFLAG_A64(tb_flags, NAA); dc->vec_len = 0; dc->vec_stride = 0; dc->cp_regs = arm_cpu->cp_regs; -- 2.34.1 From: Richard Henderson <richard.henderson@linaro.org> FEAT_LSE2 only requires that atomic operations not cross a 16-byte boundary. Ordered operations may be completely unaligned if SCTLR.nAA is set. Because this alignment check is so special, do it by hand. Make sure not to keep TCG temps live across the branch. Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230530191438.411344-17-richard.henderson@linaro.org Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/tcg/helper-a64.h | 3 + target/arm/tcg/helper-a64.c | 7 ++ target/arm/tcg/translate-a64.c | 120 ++++++++++++++++++++++++++------- 3 files changed, 104 insertions(+), 26 deletions(-) diff --git a/target/arm/tcg/helper-a64.h b/target/arm/tcg/helper-a64.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/tcg/helper-a64.h +++ b/target/arm/tcg/helper-a64.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_2(st2g_stub, TCG_CALL_NO_WG, void, env, i64) DEF_HELPER_FLAGS_2(ldgm, TCG_CALL_NO_WG, i64, env, i64) DEF_HELPER_FLAGS_3(stgm, TCG_CALL_NO_WG, void, env, i64, i64) DEF_HELPER_FLAGS_3(stzgm_tags, TCG_CALL_NO_WG, void, env, i64, i64) + +DEF_HELPER_FLAGS_4(unaligned_access, TCG_CALL_NO_WG, + noreturn, env, i64, i32, i32) diff --git a/target/arm/tcg/helper-a64.c b/target/arm/tcg/helper-a64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/tcg/helper-a64.c +++ b/target/arm/tcg/helper-a64.c @@ -XXX,XX +XXX,XX @@ void HELPER(dc_zva)(CPUARMState *env, uint64_t vaddr_in) memset(mem, 0, blocklen); } + +void HELPER(unaligned_access)(CPUARMState *env, uint64_t addr, + uint32_t access_type, uint32_t mmu_idx) +{ + arm_cpu_do_unaligned_access(env_cpu(env), addr, access_type, + mmu_idx, GETPC()); +} diff --git a/target/arm/tcg/translate-a64.c b/target/arm/tcg/translate-a64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/tcg/translate-a64.c +++ b/target/arm/tcg/translate-a64.c @@ -XXX,XX +XXX,XX @@ TCGv_i64 gen_mte_checkN(DisasContext *s, TCGv_i64 addr, bool is_write, return clean_data_tbi(s, addr); } +/* + * Generate the special alignment check that applies to AccType_ATOMIC + * and AccType_ORDERED insns under FEAT_LSE2: the access need not be + * naturally aligned, but it must not cross a 16-byte boundary. + * See AArch64.CheckAlignment(). + */ +static void check_lse2_align(DisasContext *s, int rn, int imm, + bool is_write, MemOp mop) +{ + TCGv_i32 tmp; + TCGv_i64 addr; + TCGLabel *over_label; + MMUAccessType type; + int mmu_idx; + + tmp = tcg_temp_new_i32(); + tcg_gen_extrl_i64_i32(tmp, cpu_reg_sp(s, rn)); + tcg_gen_addi_i32(tmp, tmp, imm & 15); + tcg_gen_andi_i32(tmp, tmp, 15); + tcg_gen_addi_i32(tmp, tmp, memop_size(mop)); + + over_label = gen_new_label(); + tcg_gen_brcondi_i32(TCG_COND_LEU, tmp, 16, over_label); + + addr = tcg_temp_new_i64(); + tcg_gen_addi_i64(addr, cpu_reg_sp(s, rn), imm); + + type = is_write ? MMU_DATA_STORE : MMU_DATA_LOAD, + mmu_idx = get_mem_index(s); + gen_helper_unaligned_access(cpu_env, addr, tcg_constant_i32(type), + tcg_constant_i32(mmu_idx)); + + gen_set_label(over_label); + +} + +/* Handle the alignment check for AccType_ATOMIC instructions. */ +static MemOp check_atomic_align(DisasContext *s, int rn, MemOp mop) +{ + MemOp size = mop & MO_SIZE; + + if (size == MO_8) { + return mop; + } + + /* + * If size == MO_128, this is a LDXP, and the operation is single-copy + * atomic for each doubleword, not the entire quadword; it still must + * be quadword aligned. + */ + if (size == MO_128) { + return finalize_memop_atom(s, MO_128 | MO_ALIGN, + MO_ATOM_IFALIGN_PAIR); + } + if (dc_isar_feature(aa64_lse2, s)) { + check_lse2_align(s, rn, 0, true, mop); + } else { + mop |= MO_ALIGN; + } + return finalize_memop(s, mop); +} + +/* Handle the alignment check for AccType_ORDERED instructions. */ +static MemOp check_ordered_align(DisasContext *s, int rn, int imm, + bool is_write, MemOp mop) +{ + MemOp size = mop & MO_SIZE; + + if (size == MO_8) { + return mop; + } + if (size == MO_128) { + return finalize_memop_atom(s, MO_128 | MO_ALIGN, + MO_ATOM_IFALIGN_PAIR); + } + if (!dc_isar_feature(aa64_lse2, s)) { + mop |= MO_ALIGN; + } else if (!s->naa) { + check_lse2_align(s, rn, imm, is_write, mop); + } + return finalize_memop(s, mop); +} + typedef struct DisasCompare64 { TCGCond cond; TCGv_i64 value; @@ -XXX,XX +XXX,XX @@ static void gen_load_exclusive(DisasContext *s, int rt, int rt2, int rn, { int idx = get_mem_index(s); TCGv_i64 dirty_addr, clean_addr; - MemOp memop; - - /* - * For pairs: - * if size == 2, the operation is single-copy atomic for the doubleword. - * if size == 3, the operation is single-copy atomic for *each* doubleword, - * not the entire quadword, however it must be quadword aligned. - */ - memop = size + is_pair; - if (memop == MO_128) { - memop = finalize_memop_atom(s, MO_128 | MO_ALIGN, - MO_ATOM_IFALIGN_PAIR); - } else { - memop = finalize_memop(s, memop | MO_ALIGN); - } + MemOp memop = check_atomic_align(s, rn, size + is_pair); s->is_ldex = true; dirty_addr = cpu_reg_sp(s, rn); @@ -XXX,XX +XXX,XX @@ static void gen_compare_and_swap(DisasContext *s, int rs, int rt, if (rn == 31) { gen_check_sp_alignment(s); } - memop = finalize_memop(s, size | MO_ALIGN); + memop = check_atomic_align(s, rn, size); clean_addr = gen_mte_check1(s, cpu_reg_sp(s, rn), true, rn != 31, memop); tcg_gen_atomic_cmpxchg_i64(tcg_rs, clean_addr, tcg_rs, tcg_rt, memidx, memop); @@ -XXX,XX +XXX,XX @@ static void gen_compare_and_swap_pair(DisasContext *s, int rs, int rt, } /* This is a single atomic access, despite the "pair". */ - memop = finalize_memop(s, (size + 1) | MO_ALIGN); + memop = check_atomic_align(s, rn, size + 1); clean_addr = gen_mte_check1(s, cpu_reg_sp(s, rn), true, rn != 31, memop); if (size == 2) { @@ -XXX,XX +XXX,XX @@ static void disas_ldst_excl(DisasContext *s, uint32_t insn) gen_check_sp_alignment(s); } tcg_gen_mb(TCG_MO_ALL | TCG_BAR_STRL); - /* TODO: ARMv8.4-LSE SCTLR.nAA */ - memop = finalize_memop(s, size | MO_ALIGN); + memop = check_ordered_align(s, rn, 0, true, size); clean_addr = gen_mte_check1(s, cpu_reg_sp(s, rn), true, rn != 31, memop); do_gpr_st(s, cpu_reg(s, rt), clean_addr, memop, true, rt, @@ -XXX,XX +XXX,XX @@ static void disas_ldst_excl(DisasContext *s, uint32_t insn) if (rn == 31) { gen_check_sp_alignment(s); } - /* TODO: ARMv8.4-LSE SCTLR.nAA */ - memop = finalize_memop(s, size | MO_ALIGN); + memop = check_ordered_align(s, rn, 0, false, size); clean_addr = gen_mte_check1(s, cpu_reg_sp(s, rn), false, rn != 31, memop); do_gpr_ld(s, cpu_reg(s, rt), clean_addr, memop, false, true, @@ -XXX,XX +XXX,XX @@ static void disas_ldst_atomic(DisasContext *s, uint32_t insn, bool a = extract32(insn, 23, 1); TCGv_i64 tcg_rs, tcg_rt, clean_addr; AtomicThreeOpFn *fn = NULL; - MemOp mop = finalize_memop(s, size | MO_ALIGN); + MemOp mop = size; if (is_vector || !dc_isar_feature(aa64_atomics, s)) { unallocated_encoding(s); @@ -XXX,XX +XXX,XX @@ static void disas_ldst_atomic(DisasContext *s, uint32_t insn, if (rn == 31) { gen_check_sp_alignment(s); } + + mop = check_atomic_align(s, rn, mop); clean_addr = gen_mte_check1(s, cpu_reg_sp(s, rn), false, rn != 31, mop); if (o3_opc == 014) { @@ -XXX,XX +XXX,XX @@ static void disas_ldst_ldapr_stlr(DisasContext *s, uint32_t insn) bool is_store = false; bool extend = false; bool iss_sf; - MemOp mop; + MemOp mop = size; if (!dc_isar_feature(aa64_rcpc_8_4, s)) { unallocated_encoding(s); return; } - /* TODO: ARMv8.4-LSE SCTLR.nAA */ - mop = finalize_memop(s, size | MO_ALIGN); - switch (opc) { case 0: /* STLURB */ is_store = true; @@ -XXX,XX +XXX,XX @@ static void disas_ldst_ldapr_stlr(DisasContext *s, uint32_t insn) gen_check_sp_alignment(s); } + mop = check_ordered_align(s, rn, offset, is_store, mop); + dirty_addr = read_cpu_reg_sp(s, rn, 1); tcg_gen_addi_i64(dirty_addr, dirty_addr, offset); clean_addr = clean_data_tbi(s, dirty_addr); -- 2.34.1 From: Richard Henderson <richard.henderson@linaro.org> Push the mte check behind the exclusive_addr check. Document the several ways that we are still out of spec with this implementation. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230530191438.411344-18-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/tcg/translate-a64.c | 42 +++++++++++++++++++++++++++++----- 1 file changed, 36 insertions(+), 6 deletions(-) diff --git a/target/arm/tcg/translate-a64.c b/target/arm/tcg/translate-a64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/tcg/translate-a64.c +++ b/target/arm/tcg/translate-a64.c @@ -XXX,XX +XXX,XX @@ static void gen_store_exclusive(DisasContext *s, int rd, int rt, int rt2, */ TCGLabel *fail_label = gen_new_label(); TCGLabel *done_label = gen_new_label(); - TCGv_i64 tmp, dirty_addr, clean_addr; + TCGv_i64 tmp, clean_addr; MemOp memop; - memop = (size + is_pair) | MO_ALIGN; - memop = finalize_memop(s, memop); - - dirty_addr = cpu_reg_sp(s, rn); - clean_addr = gen_mte_check1(s, dirty_addr, true, rn != 31, memop); + /* + * FIXME: We are out of spec here. We have recorded only the address + * from load_exclusive, not the entire range, and we assume that the + * size of the access on both sides match. The architecture allows the + * store to be smaller than the load, so long as the stored bytes are + * within the range recorded by the load. + */ + /* See AArch64.ExclusiveMonitorsPass() and AArch64.IsExclusiveVA(). */ + clean_addr = clean_data_tbi(s, cpu_reg_sp(s, rn)); tcg_gen_brcond_i64(TCG_COND_NE, clean_addr, cpu_exclusive_addr, fail_label); + /* + * The write, and any associated faults, only happen if the virtual + * and physical addresses pass the exclusive monitor check. These + * faults are exceedingly unlikely, because normally the guest uses + * the exact same address register for the load_exclusive, and we + * would have recognized these faults there. + * + * It is possible to trigger an alignment fault pre-LSE2, e.g. with an + * unaligned 4-byte write within the range of an aligned 8-byte load. + * With LSE2, the store would need to cross a 16-byte boundary when the + * load did not, which would mean the store is outside the range + * recorded for the monitor, which would have failed a corrected monitor + * check above. For now, we assume no size change and retain the + * MO_ALIGN to let tcg know what we checked in the load_exclusive. + * + * It is possible to trigger an MTE fault, by performing the load with + * a virtual address with a valid tag and performing the store with the + * same virtual address and a different invalid tag. + */ + memop = size + is_pair; + if (memop == MO_128 || !dc_isar_feature(aa64_lse2, s)) { + memop |= MO_ALIGN; + } + memop = finalize_memop(s, memop); + gen_mte_check1(s, cpu_reg_sp(s, rn), true, rn != 31, memop); + tmp = tcg_temp_new_i64(); if (is_pair) { if (size == 2) { -- 2.34.1 From: Richard Henderson <richard.henderson@linaro.org> We have many other instances of stg in the testsuite; change these to provide an instance of stz2g. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230530191438.411344-19-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- tests/tcg/aarch64/mte-7.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/tests/tcg/aarch64/mte-7.c b/tests/tcg/aarch64/mte-7.c index XXXXXXX..XXXXXXX 100644 --- a/tests/tcg/aarch64/mte-7.c +++ b/tests/tcg/aarch64/mte-7.c @@ -XXX,XX +XXX,XX @@ int main(int ac, char **av) p = (void *)((unsigned long)p | (1ul << 56)); /* Store tag in sequential granules. */ - asm("stg %0, [%0]" : : "r"(p + 0x0ff0)); - asm("stg %0, [%0]" : : "r"(p + 0x1000)); + asm("stz2g %0, [%0]" : : "r"(p + 0x0ff0)); /* * Perform an unaligned store with tag 1 crossing the pages. -- 2.34.1 From: Richard Henderson <richard.henderson@linaro.org> With -cpu max and FEAT_LSE2, the __aarch64__ section will only raise an alignment exception when the load crosses a 16-byte boundary. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230530191438.411344-20-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- tests/tcg/multiarch/sigbus.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) diff --git a/tests/tcg/multiarch/sigbus.c b/tests/tcg/multiarch/sigbus.c index XXXXXXX..XXXXXXX 100644 --- a/tests/tcg/multiarch/sigbus.c +++ b/tests/tcg/multiarch/sigbus.c @@ -XXX,XX +XXX,XX @@ #include <endian.h> -unsigned long long x = 0x8877665544332211ull; -void * volatile p = (void *)&x + 1; +char x[32] __attribute__((aligned(16))) = { + 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07, 0x08, + 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f, 0x10, + 0x11, 0x12, 0x13, 0x14, 0x15, 0x16, 0x17, 0x18, + 0x19, 0x1a, 0x1b, 0x1c, 0x1d, 0x1e, 0x1f, 0x20, +}; +void * volatile p = (void *)&x + 15; void sigbus(int sig, siginfo_t *info, void *uc) { @@ -XXX,XX +XXX,XX @@ int main() * We might as well validate the unaligned load worked. */ if (BYTE_ORDER == LITTLE_ENDIAN) { - assert(tmp == 0x55443322); + assert(tmp == 0x13121110); } else { - assert(tmp == 0x77665544); + assert(tmp == 0x10111213); } return EXIT_SUCCESS; } -- 2.34.1 From: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230530191438.411344-21-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- docs/system/arm/emulation.rst | 1 + target/arm/tcg/cpu64.c | 1 + 2 files changed, 2 insertions(+) diff --git a/docs/system/arm/emulation.rst b/docs/system/arm/emulation.rst index XXXXXXX..XXXXXXX 100644 --- a/docs/system/arm/emulation.rst +++ b/docs/system/arm/emulation.rst @@ -XXX,XX +XXX,XX @@ the following architecture extensions: - FEAT_LRCPC (Load-acquire RCpc instructions) - FEAT_LRCPC2 (Load-acquire RCpc instructions v2) - FEAT_LSE (Large System Extensions) +- FEAT_LSE2 (Large System Extensions v2) - FEAT_LVA (Large Virtual Address space) - FEAT_MTE (Memory Tagging Extension) - FEAT_MTE2 (Memory Tagging Extension) diff --git a/target/arm/tcg/cpu64.c b/target/arm/tcg/cpu64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/tcg/cpu64.c +++ b/target/arm/tcg/cpu64.c @@ -XXX,XX +XXX,XX @@ void aarch64_max_tcg_initfn(Object *obj) t = FIELD_DP64(t, ID_AA64MMFR2, IESB, 1); /* FEAT_IESB */ t = FIELD_DP64(t, ID_AA64MMFR2, VARANGE, 1); /* FEAT_LVA */ t = FIELD_DP64(t, ID_AA64MMFR2, ST, 1); /* FEAT_TTST */ + t = FIELD_DP64(t, ID_AA64MMFR2, AT, 1); /* FEAT_LSE2 */ t = FIELD_DP64(t, ID_AA64MMFR2, IDS, 1); /* FEAT_IDST */ t = FIELD_DP64(t, ID_AA64MMFR2, FWB, 1); /* FEAT_S2FWB */ t = FIELD_DP64(t, ID_AA64MMFR2, TTL, 1); /* FEAT_TTL */ -- 2.34.1 From: Zhuojia Shen <chaosdefinition@hotmail.com> DC CVAP and DC CVADP instructions can be executed in EL0 on Linux, either directly when SCTLR_EL1.UCI == 1 or emulated by the kernel (see user_cache_maint_handler() in arch/arm64/kernel/traps.c). This patch enables execution of the two instructions in user mode emulation. Signed-off-by: Zhuojia Shen <chaosdefinition@hotmail.com> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) diff --git a/target/arm/helper.c b/target/arm/helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper.c +++ b/target/arm/helper.c @@ -XXX,XX +XXX,XX @@ static const ARMCPRegInfo rndr_reginfo[] = { .access = PL0_R, .readfn = rndr_readfn }, }; -#ifndef CONFIG_USER_ONLY static void dccvap_writefn(CPUARMState *env, const ARMCPRegInfo *opaque, uint64_t value) { @@ -XXX,XX +XXX,XX @@ static void dccvap_writefn(CPUARMState *env, const ARMCPRegInfo *opaque, /* This won't be crossing page boundaries */ haddr = probe_read(env, vaddr, dline_size, mem_idx, GETPC()); if (haddr) { +#ifndef CONFIG_USER_ONLY ram_addr_t offset; MemoryRegion *mr; @@ -XXX,XX +XXX,XX @@ static void dccvap_writefn(CPUARMState *env, const ARMCPRegInfo *opaque, if (mr) { memory_region_writeback(mr, offset, dline_size); } +#endif /*CONFIG_USER_ONLY*/ } } @@ -XXX,XX +XXX,XX @@ static const ARMCPRegInfo dcpodp_reg[] = { .fgt = FGT_DCCVADP, .accessfn = aa64_cacheop_poc_access, .writefn = dccvap_writefn }, }; -#endif /*CONFIG_USER_ONLY*/ static CPAccessResult access_aa64_tid5(CPUARMState *env, const ARMCPRegInfo *ri, bool isread) @@ -XXX,XX +XXX,XX @@ void register_cp_regs_for_features(ARMCPU *cpu) if (cpu_isar_feature(aa64_tlbios, cpu)) { define_arm_cp_regs(cpu, tlbios_reginfo); } -#ifndef CONFIG_USER_ONLY /* Data Cache clean instructions up to PoP */ if (cpu_isar_feature(aa64_dcpop, cpu)) { define_one_arm_cp_reg(cpu, dcpop_reg); @@ -XXX,XX +XXX,XX @@ void register_cp_regs_for_features(ARMCPU *cpu) define_one_arm_cp_reg(cpu, dcpodp_reg); } } -#endif /*CONFIG_USER_ONLY*/ /* * If full MTE is enabled, add all of the system registers. -- 2.34.1 From: Zhuojia Shen <chaosdefinition@hotmail.com> Test execution of DC CVAP and DC CVADP instructions under user mode emulation. Signed-off-by: Zhuojia Shen <chaosdefinition@hotmail.com> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- tests/tcg/aarch64/dcpodp.c | 63 +++++++++++++++++++++++++++++++ tests/tcg/aarch64/dcpop.c | 63 +++++++++++++++++++++++++++++++ tests/tcg/aarch64/Makefile.target | 11 ++++++ 3 files changed, 137 insertions(+) create mode 100644 tests/tcg/aarch64/dcpodp.c create mode 100644 tests/tcg/aarch64/dcpop.c diff --git a/tests/tcg/aarch64/dcpodp.c b/tests/tcg/aarch64/dcpodp.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/tests/tcg/aarch64/dcpodp.c @@ -XXX,XX +XXX,XX @@ +/* + * Test execution of DC CVADP instruction. + * + * Copyright (c) 2023 Zhuojia Shen <chaosdefinition@hotmail.com> + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#include <asm/hwcap.h> +#include <sys/auxv.h> + +#include <signal.h> +#include <stdbool.h> +#include <stdio.h> +#include <stdlib.h> + +#ifndef HWCAP2_DCPODP +#define HWCAP2_DCPODP (1 << 0) +#endif + +bool should_fail = false; + +static void signal_handler(int sig, siginfo_t *si, void *data) +{ + ucontext_t *uc = (ucontext_t *)data; + + if (should_fail) { + uc->uc_mcontext.pc += 4; + } else { + exit(EXIT_FAILURE); + } +} + +static int do_dc_cvadp(void) +{ + struct sigaction sa = { + .sa_flags = SA_SIGINFO, + .sa_sigaction = signal_handler, + }; + + sigemptyset(&sa.sa_mask); + if (sigaction(SIGSEGV, &sa, NULL) < 0) { + perror("sigaction"); + return EXIT_FAILURE; + } + + asm volatile("dc cvadp, %0\n\t" :: "r"(&sa)); + + should_fail = true; + asm volatile("dc cvadp, %0\n\t" :: "r"(NULL)); + should_fail = false; + + return EXIT_SUCCESS; +} + +int main(void) +{ + if (getauxval(AT_HWCAP2) & HWCAP2_DCPODP) { + return do_dc_cvadp(); + } else { + printf("SKIP: no HWCAP2_DCPODP on this system\n"); + return EXIT_SUCCESS; + } +} diff --git a/tests/tcg/aarch64/dcpop.c b/tests/tcg/aarch64/dcpop.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/tests/tcg/aarch64/dcpop.c @@ -XXX,XX +XXX,XX @@ +/* + * Test execution of DC CVAP instruction. + * + * Copyright (c) 2023 Zhuojia Shen <chaosdefinition@hotmail.com> + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#include <asm/hwcap.h> +#include <sys/auxv.h> + +#include <signal.h> +#include <stdbool.h> +#include <stdio.h> +#include <stdlib.h> + +#ifndef HWCAP_DCPOP +#define HWCAP_DCPOP (1 << 16) +#endif + +bool should_fail = false; + +static void signal_handler(int sig, siginfo_t *si, void *data) +{ + ucontext_t *uc = (ucontext_t *)data; + + if (should_fail) { + uc->uc_mcontext.pc += 4; + } else { + exit(EXIT_FAILURE); + } +} + +static int do_dc_cvap(void) +{ + struct sigaction sa = { + .sa_flags = SA_SIGINFO, + .sa_sigaction = signal_handler, + }; + + sigemptyset(&sa.sa_mask); + if (sigaction(SIGSEGV, &sa, NULL) < 0) { + perror("sigaction"); + return EXIT_FAILURE; + } + + asm volatile("dc cvap, %0\n\t" :: "r"(&sa)); + + should_fail = true; + asm volatile("dc cvap, %0\n\t" :: "r"(NULL)); + should_fail = false; + + return EXIT_SUCCESS; +} + +int main(void) +{ + if (getauxval(AT_HWCAP) & HWCAP_DCPOP) { + return do_dc_cvap(); + } else { + printf("SKIP: no HWCAP_DCPOP on this system\n"); + return EXIT_SUCCESS; + } +} diff --git a/tests/tcg/aarch64/Makefile.target b/tests/tcg/aarch64/Makefile.target index XXXXXXX..XXXXXXX 100644 --- a/tests/tcg/aarch64/Makefile.target +++ b/tests/tcg/aarch64/Makefile.target @@ -XXX,XX +XXX,XX @@ config-cc.mak: Makefile $(quiet-@)( \ $(call cc-option,-march=armv8.1-a+sve, CROSS_CC_HAS_SVE); \ $(call cc-option,-march=armv8.1-a+sve2, CROSS_CC_HAS_SVE2); \ + $(call cc-option,-march=armv8.2-a, CROSS_CC_HAS_ARMV8_2); \ $(call cc-option,-march=armv8.3-a, CROSS_CC_HAS_ARMV8_3); \ + $(call cc-option,-march=armv8.5-a, CROSS_CC_HAS_ARMV8_5); \ $(call cc-option,-mbranch-protection=standard, CROSS_CC_HAS_ARMV8_BTI); \ $(call cc-option,-march=armv8.5-a+memtag, CROSS_CC_HAS_ARMV8_MTE); \ $(call cc-option,-march=armv9-a+sme, CROSS_CC_HAS_ARMV9_SME)) 3> config-cc.mak -include config-cc.mak +ifneq ($(CROSS_CC_HAS_ARMV8_2),) +AARCH64_TESTS += dcpop +dcpop: CFLAGS += -march=armv8.2-a +endif +ifneq ($(CROSS_CC_HAS_ARMV8_5),) +AARCH64_TESTS += dcpodp +dcpodp: CFLAGS += -march=armv8.5-a +endif + # Pauth Tests ifneq ($(CROSS_CC_HAS_ARMV8_3),) AARCH64_TESTS += pauth-1 pauth-2 pauth-4 pauth-5 -- 2.34.1 From: Zhuojia Shen <chaosdefinition@hotmail.com> Accessing EL0-accessible Debug Communication Channel (DCC) registers in user mode emulation is currently enabled. However, it does not match Linux behavior as Linux sets MDSCR_EL1.TDCC on startup to disable EL0 access to DCC (see __cpu_setup() in arch/arm64/mm/proc.S). This patch fixes access_tdcc() to check MDSCR_EL1.TDCC for EL0 and sets MDSCR_EL1.TDCC for user mode emulation to match Linux. Signed-off-by: Zhuojia Shen <chaosdefinition@hotmail.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: DS7PR12MB630905198DD8E69F6817544CAC4EA@DS7PR12MB6309.namprd12.prod.outlook.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/cpu.c | 2 ++ target/arm/debug_helper.c | 5 +++++ 2 files changed, 7 insertions(+) diff --git a/target/arm/cpu.c b/target/arm/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.c +++ b/target/arm/cpu.c @@ -XXX,XX +XXX,XX @@ static void arm_cpu_reset_hold(Object *obj) * This is not yet exposed from the Linux kernel in any way. */ env->cp15.sctlr_el[1] |= SCTLR_TSCXT; + /* Disable access to Debug Communication Channel (DCC). */ + env->cp15.mdscr_el1 |= 1 << 12; #else /* Reset into the highest available EL */ if (arm_feature(env, ARM_FEATURE_EL3)) { diff --git a/target/arm/debug_helper.c b/target/arm/debug_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/debug_helper.c +++ b/target/arm/debug_helper.c @@ -XXX,XX +XXX,XX @@ static CPAccessResult access_tda(CPUARMState *env, const ARMCPRegInfo *ri, * is implemented then these are controlled by MDCR_EL2.TDCC for * EL2 and MDCR_EL3.TDCC for EL3. They are also controlled by * the general debug access trap bits MDCR_EL2.TDA and MDCR_EL3.TDA. + * For EL0, they are also controlled by MDSCR_EL1.TDCC. */ static CPAccessResult access_tdcc(CPUARMState *env, const ARMCPRegInfo *ri, bool isread) { int el = arm_current_el(env); uint64_t mdcr_el2 = arm_mdcr_el2_eff(env); + bool mdscr_el1_tdcc = extract32(env->cp15.mdscr_el1, 12, 1); bool mdcr_el2_tda = (mdcr_el2 & MDCR_TDA) || (mdcr_el2 & MDCR_TDE) || (arm_hcr_el2_eff(env) & HCR_TGE); bool mdcr_el2_tdcc = cpu_isar_feature(aa64_fgt, env_archcpu(env)) && @@ -XXX,XX +XXX,XX @@ static CPAccessResult access_tdcc(CPUARMState *env, const ARMCPRegInfo *ri, bool mdcr_el3_tdcc = cpu_isar_feature(aa64_fgt, env_archcpu(env)) && (env->cp15.mdcr_el3 & MDCR_TDCC); + if (el < 1 && mdscr_el1_tdcc) { + return CP_ACCESS_TRAP; + } if (el < 2 && (mdcr_el2_tda || mdcr_el2_tdcc)) { return CP_ACCESS_TRAP_EL2; } -- 2.34.1