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 Arm queue; not huge but I figured I might as well send it out since I've been doing code review today and there's no queue of unprocessed pullreqs... thanks -- PMM The following changes since commit b3f846c59d8405bb87c551187721fc92ff2f1b92: Merge remote-tracking branch 'remotes/huth-gitlab/tags/pull-request-2021-01-11v2' into staging (2021-01-11 15:15:35 +0000) are available in the Git repository at: https://git.linaro.org/people/pmaydell/qemu-arm.git tags/pull-target-arm-20210112 for you to fetch changes up to 19d131395ccaf503db21dadd8257e6dc9fc1d7de: ui/cocoa: Fix openFile: deprecation on Big Sur (2021-01-12 11:38:37 +0000) ---------------------------------------------------------------- target-arm queue: * arm: Support emulation of ARMv8.4-TTST extension * arm: Update cpu.h ID register field definitions * arm: Fix breakage of XScale instruction emulation * hw/net/lan9118: Fix RX Status FIFO PEEK value * npcm7xx: Add ADC and PWM emulation * ui/cocoa: Make "open docs" help menu entry work again when binary is run from the build tree * ui/cocoa: Fix openFile: deprecation on Big Sur * docs: Add qemu-storage-daemon(1) manpage to meson.build * docs: Build and install all the docs in a single manual ---------------------------------------------------------------- Hao Wu (6): hw/misc: Add clock converter in NPCM7XX CLK module hw/timer: Refactor NPCM7XX Timer to use CLK clock hw/adc: Add an ADC module for NPCM7XX hw/misc: Add a PWM module for NPCM7XX hw/misc: Add QTest for NPCM7XX PWM Module hw/*: Use type casting for SysBusDevice in NPCM7XX Leif Lindholm (6): target/arm: fix typo in cpu.h ID_AA64PFR1 field name target/arm: make ARMCPU.clidr 64-bit target/arm: make ARMCPU.ctr 64-bit target/arm: add descriptions of CLIDR_EL1, CCSIDR_EL1, CTR_EL0 to cpu.h target/arm: add aarch64 ID register fields to cpu.h target/arm: add aarch32 ID register fields to cpu.h Peter Maydell (5): docs: Add qemu-storage-daemon(1) manpage to meson.build docs: Build and install all the docs in a single manual target/arm: Don't decode insns in the XScale/iWMMXt space as cp insns hw/net/lan9118: Fix RX Status FIFO PEEK value hw/net/lan9118: Add symbolic constants for register offsets Roman Bolshakov (2): ui/cocoa: Update path to docs in build tree ui/cocoa: Fix openFile: deprecation on Big Sur Rémi Denis-Courmont (2): target/arm: ARMv8.4-TTST extension target/arm: enable Small Translation tables in max CPU docs/conf.py | 46 ++- docs/devel/conf.py | 15 - docs/index.html.in | 17 - docs/interop/conf.py | 28 -- docs/meson.build | 65 ++-- docs/specs/conf.py | 16 - docs/system/arm/nuvoton.rst | 4 +- docs/system/conf.py | 28 -- docs/tools/conf.py | 37 -- docs/user/conf.py | 15 - meson.build | 1 + hw/adc/trace.h | 1 + include/hw/adc/npcm7xx_adc.h | 69 ++++ include/hw/arm/npcm7xx.h | 4 + include/hw/misc/npcm7xx_clk.h | 146 ++++++- include/hw/misc/npcm7xx_pwm.h | 105 +++++ include/hw/timer/npcm7xx_timer.h | 1 + target/arm/cpu.h | 85 ++++- hw/adc/npcm7xx_adc.c | 301 +++++++++++++++ hw/arm/npcm7xx.c | 55 ++- hw/arm/npcm7xx_boards.c | 2 +- hw/mem/npcm7xx_mc.c | 2 +- hw/misc/npcm7xx_clk.c | 807 ++++++++++++++++++++++++++++++++++++++- hw/misc/npcm7xx_gcr.c | 2 +- hw/misc/npcm7xx_pwm.c | 550 ++++++++++++++++++++++++++ hw/misc/npcm7xx_rng.c | 2 +- hw/net/lan9118.c | 26 +- hw/nvram/npcm7xx_otp.c | 2 +- hw/ssi/npcm7xx_fiu.c | 2 +- hw/timer/npcm7xx_timer.c | 39 +- target/arm/cpu64.c | 1 + target/arm/helper.c | 15 +- target/arm/translate.c | 7 + tests/qtest/npcm7xx_adc-test.c | 377 ++++++++++++++++++ tests/qtest/npcm7xx_pwm-test.c | 490 ++++++++++++++++++++++++ hw/adc/meson.build | 1 + hw/adc/trace-events | 5 + hw/misc/meson.build | 1 + hw/misc/trace-events | 6 + tests/qtest/meson.build | 4 +- ui/cocoa.m | 7 +- 41 files changed, 3124 insertions(+), 263 deletions(-) delete mode 100644 docs/devel/conf.py delete mode 100644 docs/index.html.in delete mode 100644 docs/interop/conf.py delete mode 100644 docs/specs/conf.py delete mode 100644 docs/system/conf.py delete mode 100644 docs/tools/conf.py delete mode 100644 docs/user/conf.py create mode 100644 hw/adc/trace.h create mode 100644 include/hw/adc/npcm7xx_adc.h create mode 100644 include/hw/misc/npcm7xx_pwm.h create mode 100644 hw/adc/npcm7xx_adc.c create mode 100644 hw/misc/npcm7xx_pwm.c create mode 100644 tests/qtest/npcm7xx_adc-test.c create mode 100644 tests/qtest/npcm7xx_pwm-test.c create mode 100644 hw/adc/trace-events From: Rémi Denis-Courmont <remi.denis.courmont@huawei.com> This adds for the Small Translation tables extension in AArch64 state. Signed-off-by: Rémi Denis-Courmont <remi.denis.courmont@huawei.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/cpu.h | 5 +++++ target/arm/helper.c | 15 +++++++++++++-- 2 files changed, 18 insertions(+), 2 deletions(-) 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_uao(const ARMISARegisters *id) return FIELD_EX64(id->id_aa64mmfr2, ID_AA64MMFR2, UAO) != 0; } +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_bti(const ARMISARegisters *id) { return FIELD_EX64(id->id_aa64pfr1, ID_AA64PFR1, BT) != 0; 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 @@ ARMVAParameters aa64_va_parameters(CPUARMState *env, uint64_t va, { uint64_t tcr = regime_tcr(env, mmu_idx)->raw_tcr; bool epd, hpd, using16k, using64k; - int select, tsz, tbi; + int select, tsz, tbi, max_tsz; if (!regime_has_2_ranges(mmu_idx)) { select = 0; @@ -XXX,XX +XXX,XX @@ ARMVAParameters aa64_va_parameters(CPUARMState *env, uint64_t va, hpd = extract64(tcr, 42, 1); } } - tsz = MIN(tsz, 39); /* TODO: ARMv8.4-TTST */ + + if (cpu_isar_feature(aa64_st, env_archcpu(env))) { + max_tsz = 48 - using64k; + } else { + max_tsz = 39; + } + + tsz = MIN(tsz, max_tsz); tsz = MAX(tsz, 16); /* TODO: ARMv8.2-LVA */ /* Present TBI as a composite with TBID. */ @@ -XXX,XX +XXX,XX @@ static bool get_phys_addr_lpae(CPUARMState *env, uint64_t address, if (!aarch64 || stride == 9) { /* AArch32 or 4KB pages */ startlevel = 2 - sl0; + + if (cpu_isar_feature(aa64_st, cpu)) { + startlevel &= 3; + } } else { /* 16KB or 64KB pages */ startlevel = 3 - sl0; -- 2.20.1 From: Rémi Denis-Courmont <remi.denis.courmont@huawei.com> Signed-off-by: Rémi Denis-Courmont <remi.denis.courmont@huawei.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/cpu64.c | 1 + 1 file changed, 1 insertion(+) 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 = cpu->isar.id_aa64mmfr2; t = FIELD_DP64(t, ID_AA64MMFR2, UAO, 1); t = FIELD_DP64(t, ID_AA64MMFR2, CNP, 1); /* TTCNP */ + t = FIELD_DP64(t, ID_AA64MMFR2, ST, 1); /* TTST */ cpu->isar.id_aa64mmfr2 = t; /* Replicate the same data to the 32-bit id registers. */ -- 2.20.1 From: Leif Lindholm <leif@nuviainc.com> SBSS -> SSBS Signed-off-by: Leif Lindholm <leif@nuviainc.com> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Laurent Desnogues <laurent.desnogues@gmail.com> Message-id: 20210108185154.8108-2-leif@nuviainc.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/cpu.h | 2 +- 1 file changed, 1 insertion(+), 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 @@ FIELD(ID_AA64PFR0, RAS, 28, 4) FIELD(ID_AA64PFR0, SVE, 32, 4) FIELD(ID_AA64PFR1, BT, 0, 4) -FIELD(ID_AA64PFR1, SBSS, 4, 4) +FIELD(ID_AA64PFR1, SSBS, 4, 4) FIELD(ID_AA64PFR1, MTE, 8, 4) FIELD(ID_AA64PFR1, RAS_FRAC, 12, 4) -- 2.20.1 From: Leif Lindholm <leif@nuviainc.com> The AArch64 view of CLIDR_EL1 extends the ICB field to include also bit 32, as well as adding a Ttype<n> field when FEAT_MTE is implemented. Extend the clidr field to be able to hold this context. Signed-off-by: Leif Lindholm <leif@nuviainc.com> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Laurent Desnogues <laurent.desnogues@gmail.com> Message-id: 20210108185154.8108-3-leif@nuviainc.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/cpu.h | 2 +- 1 file changed, 1 insertion(+), 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 @@ struct ARMCPU { uint32_t id_afr0; uint64_t id_aa64afr0; uint64_t id_aa64afr1; - uint32_t clidr; + uint64_t clidr; uint64_t mp_affinity; /* MP ID without feature bits */ /* The elements of this array are the CCSIDR values for each cache, * in the order L1DCache, L1ICache, L2DCache, L2ICache, etc. -- 2.20.1 From: Leif Lindholm <leif@nuviainc.com> When FEAT_MTE is implemented, the AArch64 view of CTR_EL0 adds the TminLine field in bits [37:32]. Extend the ctr field to be able to hold this context. Signed-off-by: Leif Lindholm <leif@nuviainc.com> Reviewed-by: Hao Wu <wuhaotsh@google.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Laurent Desnogues <laurent.desnogues@gmail.com> Message-id: 20210108185154.8108-4-leif@nuviainc.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/cpu.h | 2 +- 1 file changed, 1 insertion(+), 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 @@ struct ARMCPU { uint64_t midr; uint32_t revidr; uint32_t reset_fpsid; - uint32_t ctr; + uint64_t ctr; uint32_t reset_sctlr; uint64_t pmceid0; uint64_t pmceid1; -- 2.20.1 From: Leif Lindholm <leif@nuviainc.com> Signed-off-by: Leif Lindholm <leif@nuviainc.com> Reviewed-by: Laurent Desnogues <laurent.desnogues@gmail.com> Message-id: 20210108185154.8108-5-leif@nuviainc.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/cpu.h | 31 +++++++++++++++++++++++++++++++ 1 file changed, 31 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 @@ FIELD(V7M_FPCCR, ASPEN, 31, 1) /* * System register ID fields. */ +FIELD(CLIDR_EL1, CTYPE1, 0, 3) +FIELD(CLIDR_EL1, CTYPE2, 3, 3) +FIELD(CLIDR_EL1, CTYPE3, 6, 3) +FIELD(CLIDR_EL1, CTYPE4, 9, 3) +FIELD(CLIDR_EL1, CTYPE5, 12, 3) +FIELD(CLIDR_EL1, CTYPE6, 15, 3) +FIELD(CLIDR_EL1, CTYPE7, 18, 3) +FIELD(CLIDR_EL1, LOUIS, 21, 3) +FIELD(CLIDR_EL1, LOC, 24, 3) +FIELD(CLIDR_EL1, LOUU, 27, 3) +FIELD(CLIDR_EL1, ICB, 30, 3) + +/* When FEAT_CCIDX is implemented */ +FIELD(CCSIDR_EL1, CCIDX_LINESIZE, 0, 3) +FIELD(CCSIDR_EL1, CCIDX_ASSOCIATIVITY, 3, 21) +FIELD(CCSIDR_EL1, CCIDX_NUMSETS, 32, 24) + +/* When FEAT_CCIDX is not implemented */ +FIELD(CCSIDR_EL1, LINESIZE, 0, 3) +FIELD(CCSIDR_EL1, ASSOCIATIVITY, 3, 10) +FIELD(CCSIDR_EL1, NUMSETS, 13, 15) + +FIELD(CTR_EL0, IMINLINE, 0, 4) +FIELD(CTR_EL0, L1IP, 14, 2) +FIELD(CTR_EL0, DMINLINE, 16, 4) +FIELD(CTR_EL0, ERG, 20, 4) +FIELD(CTR_EL0, CWG, 24, 4) +FIELD(CTR_EL0, IDC, 28, 1) +FIELD(CTR_EL0, DIC, 29, 1) +FIELD(CTR_EL0, TMINLINE, 32, 6) + FIELD(MIDR_EL1, REVISION, 0, 4) FIELD(MIDR_EL1, PARTNUM, 4, 12) FIELD(MIDR_EL1, ARCHITECTURE, 16, 4) -- 2.20.1 From: Leif Lindholm <leif@nuviainc.com> Add entries present in ARM DDI 0487F.c (August 2020). Signed-off-by: Leif Lindholm <leif@nuviainc.com> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Laurent Desnogues <laurent.desnogues@gmail.com> Message-id: 20210108185154.8108-6-leif@nuviainc.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/cpu.h | 15 +++++++++++++++ 1 file changed, 15 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 @@ FIELD(ID_AA64ISAR1, GPI, 28, 4) FIELD(ID_AA64ISAR1, FRINTTS, 32, 4) FIELD(ID_AA64ISAR1, SB, 36, 4) FIELD(ID_AA64ISAR1, SPECRES, 40, 4) +FIELD(ID_AA64ISAR1, BF16, 44, 4) +FIELD(ID_AA64ISAR1, DGH, 48, 4) +FIELD(ID_AA64ISAR1, I8MM, 52, 4) FIELD(ID_AA64PFR0, EL0, 0, 4) FIELD(ID_AA64PFR0, EL1, 4, 4) @@ -XXX,XX +XXX,XX @@ FIELD(ID_AA64PFR0, ADVSIMD, 20, 4) FIELD(ID_AA64PFR0, GIC, 24, 4) FIELD(ID_AA64PFR0, RAS, 28, 4) FIELD(ID_AA64PFR0, SVE, 32, 4) +FIELD(ID_AA64PFR0, SEL2, 36, 4) +FIELD(ID_AA64PFR0, MPAM, 40, 4) +FIELD(ID_AA64PFR0, AMU, 44, 4) +FIELD(ID_AA64PFR0, DIT, 48, 4) +FIELD(ID_AA64PFR0, CSV2, 56, 4) +FIELD(ID_AA64PFR0, CSV3, 60, 4) FIELD(ID_AA64PFR1, BT, 0, 4) FIELD(ID_AA64PFR1, SSBS, 4, 4) FIELD(ID_AA64PFR1, MTE, 8, 4) FIELD(ID_AA64PFR1, RAS_FRAC, 12, 4) +FIELD(ID_AA64PFR1, MPAM_FRAC, 16, 4) FIELD(ID_AA64MMFR0, PARANGE, 0, 4) FIELD(ID_AA64MMFR0, ASIDBITS, 4, 4) @@ -XXX,XX +XXX,XX @@ FIELD(ID_AA64MMFR0, TGRAN16_2, 32, 4) FIELD(ID_AA64MMFR0, TGRAN64_2, 36, 4) FIELD(ID_AA64MMFR0, TGRAN4_2, 40, 4) FIELD(ID_AA64MMFR0, EXS, 44, 4) +FIELD(ID_AA64MMFR0, FGT, 56, 4) +FIELD(ID_AA64MMFR0, ECV, 60, 4) FIELD(ID_AA64MMFR1, HAFDBS, 0, 4) FIELD(ID_AA64MMFR1, VMIDBITS, 4, 4) @@ -XXX,XX +XXX,XX @@ FIELD(ID_AA64MMFR1, LO, 16, 4) FIELD(ID_AA64MMFR1, PAN, 20, 4) FIELD(ID_AA64MMFR1, SPECSEI, 24, 4) FIELD(ID_AA64MMFR1, XNX, 28, 4) +FIELD(ID_AA64MMFR1, TWED, 32, 4) +FIELD(ID_AA64MMFR1, ETS, 36, 4) FIELD(ID_AA64MMFR2, CNP, 0, 4) FIELD(ID_AA64MMFR2, UAO, 4, 4) @@ -XXX,XX +XXX,XX @@ FIELD(ID_AA64DFR0, CTX_CMPS, 28, 4) FIELD(ID_AA64DFR0, PMSVER, 32, 4) FIELD(ID_AA64DFR0, DOUBLELOCK, 36, 4) FIELD(ID_AA64DFR0, TRACEFILT, 40, 4) +FIELD(ID_AA64DFR0, MTPMU, 48, 4) FIELD(ID_DFR0, COPDBG, 0, 4) FIELD(ID_DFR0, COPSDBG, 4, 4) -- 2.20.1 From: Leif Lindholm <leif@nuviainc.com> Add entries present in ARM DDI 0487F.c (August 2020). Signed-off-by: Leif Lindholm <leif@nuviainc.com> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Laurent Desnogues <laurent.desnogues@gmail.com> Message-id: 20210108185154.8108-7-leif@nuviainc.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/cpu.h | 28 ++++++++++++++++++++++++++++ 1 file changed, 28 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 @@ FIELD(ID_ISAR6, DP, 4, 4) FIELD(ID_ISAR6, FHM, 8, 4) FIELD(ID_ISAR6, SB, 12, 4) FIELD(ID_ISAR6, SPECRES, 16, 4) +FIELD(ID_ISAR6, BF16, 20, 4) +FIELD(ID_ISAR6, I8MM, 24, 4) FIELD(ID_MMFR0, VMSA, 0, 4) FIELD(ID_MMFR0, PMSA, 4, 4) @@ -XXX,XX +XXX,XX @@ FIELD(ID_MMFR0, AUXREG, 20, 4) FIELD(ID_MMFR0, FCSE, 24, 4) FIELD(ID_MMFR0, INNERSHR, 28, 4) +FIELD(ID_MMFR1, L1HVDVA, 0, 4) +FIELD(ID_MMFR1, L1UNIVA, 4, 4) +FIELD(ID_MMFR1, L1HVDSW, 8, 4) +FIELD(ID_MMFR1, L1UNISW, 12, 4) +FIELD(ID_MMFR1, L1HVD, 16, 4) +FIELD(ID_MMFR1, L1UNI, 20, 4) +FIELD(ID_MMFR1, L1TSTCLN, 24, 4) +FIELD(ID_MMFR1, BPRED, 28, 4) + +FIELD(ID_MMFR2, L1HVDFG, 0, 4) +FIELD(ID_MMFR2, L1HVDBG, 4, 4) +FIELD(ID_MMFR2, L1HVDRNG, 8, 4) +FIELD(ID_MMFR2, HVDTLB, 12, 4) +FIELD(ID_MMFR2, UNITLB, 16, 4) +FIELD(ID_MMFR2, MEMBARR, 20, 4) +FIELD(ID_MMFR2, WFISTALL, 24, 4) +FIELD(ID_MMFR2, HWACCFLG, 28, 4) + FIELD(ID_MMFR3, CMAINTVA, 0, 4) FIELD(ID_MMFR3, CMAINTSW, 4, 4) FIELD(ID_MMFR3, BPMAINT, 8, 4) @@ -XXX,XX +XXX,XX @@ FIELD(ID_MMFR4, LSM, 20, 4) FIELD(ID_MMFR4, CCIDX, 24, 4) FIELD(ID_MMFR4, EVT, 28, 4) +FIELD(ID_MMFR5, ETS, 0, 4) + FIELD(ID_PFR0, STATE0, 0, 4) FIELD(ID_PFR0, STATE1, 4, 4) FIELD(ID_PFR0, STATE2, 8, 4) @@ -XXX,XX +XXX,XX @@ FIELD(ID_PFR1, SEC_FRAC, 20, 4) FIELD(ID_PFR1, VIRT_FRAC, 24, 4) FIELD(ID_PFR1, GIC, 28, 4) +FIELD(ID_PFR2, CSV3, 0, 4) +FIELD(ID_PFR2, SSBS, 4, 4) +FIELD(ID_PFR2, RAS_FRAC, 8, 4) + FIELD(ID_AA64ISAR0, AES, 4, 4) FIELD(ID_AA64ISAR0, SHA1, 8, 4) FIELD(ID_AA64ISAR0, SHA2, 12, 4) @@ -XXX,XX +XXX,XX @@ FIELD(ID_DFR0, MPROFDBG, 20, 4) FIELD(ID_DFR0, PERFMON, 24, 4) FIELD(ID_DFR0, TRACEFILT, 28, 4) +FIELD(ID_DFR1, MTPMU, 0, 4) + FIELD(DBGDIDR, SE_IMP, 12, 1) FIELD(DBGDIDR, NSUHD_IMP, 14, 1) FIELD(DBGDIDR, VERSION, 16, 4) -- 2.20.1 From: Roman Bolshakov <r.bolshakov@yadro.com> QEMU documentation can't be opened if QEMU is run from build tree because executables are placed in the top of build tree after conversion to meson. Signed-off-by: Roman Bolshakov <r.bolshakov@yadro.com> Reported-by: Peter Maydell <peter.maydell@linaro.org> Message-id: 20210108213815.64678-1-r.bolshakov@yadro.com Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- ui/cocoa.m | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ui/cocoa.m b/ui/cocoa.m index XXXXXXX..XXXXXXX 100644 --- a/ui/cocoa.m +++ b/ui/cocoa.m @@ -XXX,XX +XXX,XX @@ QemuCocoaView *cocoaView; - (void) openDocumentation: (NSString *) filename { /* Where to look for local files */ - NSString *path_array[] = {@"../share/doc/qemu/", @"../doc/qemu/", @"../docs/"}; + NSString *path_array[] = {@"../share/doc/qemu/", @"../doc/qemu/", @"docs/"}; NSString *full_file_path; /* iterate thru the possible paths until the file is found */ -- 2.20.1 In commit 1982e1602d15 we added a new qemu-storage-daemon(1) manpage. At the moment new manpages have to be listed both in the conf.py for Sphinx and also in docs/meson.build for Meson. We forgot the second of those -- correct the omission. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Message-id: 20210108161416.21129-2-peter.maydell@linaro.org --- docs/meson.build | 1 + 1 file changed, 1 insertion(+) diff --git a/docs/meson.build b/docs/meson.build index XXXXXXX..XXXXXXX 100644 --- a/docs/meson.build +++ b/docs/meson.build @@ -XXX,XX +XXX,XX @@ if build_docs 'qemu-img.1': (have_tools ? 'man1' : ''), 'qemu-nbd.8': (have_tools ? 'man8' : ''), 'qemu-pr-helper.8': (have_tools ? 'man8' : ''), + 'qemu-storage-daemon.1': (have_tools ? 'man1' : ''), 'qemu-trace-stap.1': (config_host.has_key('CONFIG_TRACE_SYSTEMTAP') ? 'man1' : ''), 'virtfs-proxy-helper.1': (have_virtfs_proxy_helper ? 'man1' : ''), 'virtiofsd.1': (have_virtiofsd ? 'man1' : ''), -- 2.20.1 When we first converted our documentation to Sphinx, we split it into multiple manuals (system, interop, tools, etc), which are all built separately. The primary driver for this was wanting to be able to avoid shipping the 'devel' manual to end-users. However, this is working against the grain of the way Sphinx wants to be used and causes some annoyances: * Cross-references between documents become much harder or possibly impossible * There is no single index to the whole documentation * Within one manual there's no links or table-of-contents info that lets you easily navigate to the others * The devel manual doesn't get published on the QEMU website (it would be nice to able to refer to it there) Merely hiding our developer documentation from end users seems like it's not enough benefit for these costs. Combine all the documentation into a single manual (the same way that the readthedocs site builds it) and install the whole thing. The previous manual divisions remain as the new top level sections in the manual. * The per-manual conf.py files are no longer needed * The man_pages[] specifications previously in each per-manual conf.py move to the top level conf.py * docs/meson.build logic is simplified as we now only need to run Sphinx once for the HTML and then once for the manpages5B * The old index.html.in that produced the top-level page with links to each manual is no longer needed Unfortunately this means that we now have to build the HTML documentation into docs/manual in the build tree rather than directly into docs/; otherwise it is too awkward to ensure we install only the built manual and not also the dependency info, stamp file, etc. The manual still ends up in the same place in the final installed directory, but anybody who was consulting documentation from within the build tree will have to adjust where they're looking. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Paolo Bonzini <pbonzini@redhat.com> Message-id: 20210108161416.21129-3-peter.maydell@linaro.org --- docs/conf.py | 46 ++++++++++++++++++++++++++++++- docs/devel/conf.py | 15 ----------- docs/index.html.in | 17 ------------ docs/interop/conf.py | 28 ------------------- docs/meson.build | 64 +++++++++++++++++--------------------------- docs/specs/conf.py | 16 ----------- docs/system/conf.py | 28 ------------------- docs/tools/conf.py | 37 ------------------------- docs/user/conf.py | 15 ----------- 9 files changed, 70 insertions(+), 196 deletions(-) delete mode 100644 docs/devel/conf.py delete mode 100644 docs/index.html.in delete mode 100644 docs/interop/conf.py delete mode 100644 docs/specs/conf.py delete mode 100644 docs/system/conf.py delete mode 100644 docs/tools/conf.py delete mode 100644 docs/user/conf.py diff --git a/docs/conf.py b/docs/conf.py index XXXXXXX..XXXXXXX 100644 --- a/docs/conf.py +++ b/docs/conf.py @@ -XXX,XX +XXX,XX @@ latex_documents = [ # -- Options for manual page output --------------------------------------- # Individual manual/conf.py can override this to create man pages -man_pages = [] +man_pages = [ + ('interop/qemu-ga', 'qemu-ga', + 'QEMU Guest Agent', + ['Michael Roth <mdroth@linux.vnet.ibm.com>'], 8), + ('interop/qemu-ga-ref', 'qemu-ga-ref', + 'QEMU Guest Agent Protocol Reference', + [], 7), + ('interop/qemu-qmp-ref', 'qemu-qmp-ref', + 'QEMU QMP Reference Manual', + [], 7), + ('interop/qemu-storage-daemon-qmp-ref', 'qemu-storage-daemon-qmp-ref', + 'QEMU Storage Daemon QMP Reference Manual', + [], 7), + ('system/qemu-manpage', 'qemu', + 'QEMU User Documentation', + ['Fabrice Bellard'], 1), + ('system/qemu-block-drivers', 'qemu-block-drivers', + 'QEMU block drivers reference', + ['Fabrice Bellard and the QEMU Project developers'], 7), + ('system/qemu-cpu-models', 'qemu-cpu-models', + 'QEMU CPU Models', + ['The QEMU Project developers'], 7), + ('tools/qemu-img', 'qemu-img', + 'QEMU disk image utility', + ['Fabrice Bellard'], 1), + ('tools/qemu-nbd', 'qemu-nbd', + 'QEMU Disk Network Block Device Server', + ['Anthony Liguori <anthony@codemonkey.ws>'], 8), + ('tools/qemu-pr-helper', 'qemu-pr-helper', + 'QEMU persistent reservation helper', + [], 8), + ('tools/qemu-storage-daemon', 'qemu-storage-daemon', + 'QEMU storage daemon', + [], 1), + ('tools/qemu-trace-stap', 'qemu-trace-stap', + 'QEMU SystemTap trace tool', + [], 1), + ('tools/virtfs-proxy-helper', 'virtfs-proxy-helper', + 'QEMU 9p virtfs proxy filesystem helper', + ['M. Mohan Kumar'], 1), + ('tools/virtiofsd', 'virtiofsd', + 'QEMU virtio-fs shared file system daemon', + ['Stefan Hajnoczi <stefanha@redhat.com>', + 'Masayoshi Mizuma <m.mizuma@jp.fujitsu.com>'], 1), +] # -- Options for Texinfo output ------------------------------------------- diff --git a/docs/devel/conf.py b/docs/devel/conf.py deleted file mode 100644 index XXXXXXX..XXXXXXX --- a/docs/devel/conf.py +++ /dev/null @@ -XXX,XX +XXX,XX @@ -# -*- coding: utf-8 -*- -# -# QEMU documentation build configuration file for the 'devel' manual. -# -# This includes the top level conf file and then makes any necessary tweaks. -import sys -import os - -qemu_docdir = os.path.abspath("..") -parent_config = os.path.join(qemu_docdir, "conf.py") -exec(compile(open(parent_config, "rb").read(), parent_config, 'exec')) - -# This slightly misuses the 'description', but is the best way to get -# the manual title to appear in the sidebar. -html_theme_options['description'] = u'Developer''s Guide' diff --git a/docs/index.html.in b/docs/index.html.in deleted file mode 100644 index XXXXXXX..XXXXXXX --- a/docs/index.html.in +++ /dev/null @@ -XXX,XX +XXX,XX @@ -<!DOCTYPE html> -<html lang="en"> - <head> - <meta charset="UTF-8"> - <title>QEMU @VERSION@ Documentation</title> - </head> - <body> - <h1>QEMU @VERSION@ Documentation</h1> - <ul> - <li><a href="system/index.html">System Emulation User's Guide</a></li> - <li><a href="user/index.html">User Mode Emulation User's Guide</a></li> - <li><a href="tools/index.html">Tools Guide</a></li> - <li><a href="interop/index.html">System Emulation Management and Interoperability Guide</a></li> - <li><a href="specs/index.html">System Emulation Guest Hardware Specifications</a></li> - </ul> - </body> -</html> diff --git a/docs/interop/conf.py b/docs/interop/conf.py deleted file mode 100644 index XXXXXXX..XXXXXXX --- a/docs/interop/conf.py +++ /dev/null @@ -XXX,XX +XXX,XX @@ -# -*- coding: utf-8 -*- -# -# QEMU documentation build configuration file for the 'interop' manual. -# -# This includes the top level conf file and then makes any necessary tweaks. -import sys -import os - -qemu_docdir = os.path.abspath("..") -parent_config = os.path.join(qemu_docdir, "conf.py") -exec(compile(open(parent_config, "rb").read(), parent_config, 'exec')) - -# This slightly misuses the 'description', but is the best way to get -# the manual title to appear in the sidebar. -html_theme_options['description'] = u'System Emulation Management and Interoperability Guide' - -# One entry per manual page. List of tuples -# (source start file, name, description, authors, manual section). -man_pages = [ - ('qemu-ga', 'qemu-ga', u'QEMU Guest Agent', - ['Michael Roth <mdroth@linux.vnet.ibm.com>'], 8), - ('qemu-ga-ref', 'qemu-ga-ref', 'QEMU Guest Agent Protocol Reference', - [], 7), - ('qemu-qmp-ref', 'qemu-qmp-ref', 'QEMU QMP Reference Manual', - [], 7), - ('qemu-storage-daemon-qmp-ref', 'qemu-storage-daemon-qmp-ref', - 'QEMU Storage Daemon QMP Reference Manual', [], 7), -] diff --git a/docs/meson.build b/docs/meson.build index XXXXXXX..XXXXXXX 100644 --- a/docs/meson.build +++ b/docs/meson.build @@ -XXX,XX +XXX,XX @@ if build_docs meson.source_root() / 'docs/sphinx/qmp_lexer.py', qapi_gen_depends ] - configure_file(output: 'index.html', - input: files('index.html.in'), - configuration: {'VERSION': meson.project_version()}, - install_dir: qemu_docdir) - manuals = [ 'devel', 'interop', 'tools', 'specs', 'system', 'user' ] man_pages = { - 'interop' : { 'qemu-ga.8': (have_tools ? 'man8' : ''), 'qemu-ga-ref.7': 'man7', 'qemu-qmp-ref.7': 'man7', 'qemu-storage-daemon-qmp-ref.7': (have_tools ? 'man7' : ''), - }, - 'tools': { 'qemu-img.1': (have_tools ? 'man1' : ''), 'qemu-nbd.8': (have_tools ? 'man8' : ''), 'qemu-pr-helper.8': (have_tools ? 'man8' : ''), @@ -XXX,XX +XXX,XX @@ if build_docs 'qemu-trace-stap.1': (config_host.has_key('CONFIG_TRACE_SYSTEMTAP') ? 'man1' : ''), 'virtfs-proxy-helper.1': (have_virtfs_proxy_helper ? 'man1' : ''), 'virtiofsd.1': (have_virtiofsd ? 'man1' : ''), - }, - 'system': { 'qemu.1': 'man1', 'qemu-block-drivers.7': 'man7', 'qemu-cpu-models.7': 'man7' - }, } sphinxdocs = [] sphinxmans = [] - foreach manual : manuals - private_dir = meson.current_build_dir() / (manual + '.p') - output_dir = meson.current_build_dir() / manual - input_dir = meson.current_source_dir() / manual - this_manual = custom_target(manual + ' manual', + private_dir = meson.current_build_dir() / 'manual.p' + output_dir = meson.current_build_dir() / 'manual' + input_dir = meson.current_source_dir() + + this_manual = custom_target('QEMU manual', build_by_default: build_docs, - output: [manual + '.stamp'], - input: [files('conf.py'), files(manual / 'conf.py')], - depfile: manual + '.d', + output: 'docs.stamp', + input: files('conf.py'), + depfile: 'docs.d', depend_files: sphinx_extn_depends, command: [SPHINX_ARGS, '-Ddepfile=@DEPFILE@', '-Ddepfile_stamp=@OUTPUT0@', '-b', 'html', '-d', private_dir, input_dir, output_dir]) - sphinxdocs += this_manual - if build_docs and manual != 'devel' - install_subdir(output_dir, install_dir: qemu_docdir) - endif + sphinxdocs += this_manual + install_subdir(output_dir, install_dir: qemu_docdir, strip_directory: true) - these_man_pages = [] - install_dirs = [] - foreach page, section : man_pages.get(manual, {}) - these_man_pages += page - install_dirs += section == '' ? false : get_option('mandir') / section - endforeach - if these_man_pages.length() > 0 - sphinxmans += custom_target(manual + ' man pages', - build_by_default: build_docs, - output: these_man_pages, - input: this_manual, - install: build_docs, - install_dir: install_dirs, - command: [SPHINX_ARGS, '-b', 'man', '-d', private_dir, - input_dir, meson.current_build_dir()]) - endif + these_man_pages = [] + install_dirs = [] + foreach page, section : man_pages + these_man_pages += page + install_dirs += section == '' ? false : get_option('mandir') / section endforeach + + sphinxmans += custom_target('QEMU man pages', + build_by_default: build_docs, + output: these_man_pages, + input: this_manual, + install: build_docs, + install_dir: install_dirs, + command: [SPHINX_ARGS, '-b', 'man', '-d', private_dir, + input_dir, meson.current_build_dir()]) + alias_target('sphinxdocs', sphinxdocs) alias_target('html', sphinxdocs) alias_target('man', sphinxmans) diff --git a/docs/specs/conf.py b/docs/specs/conf.py deleted file mode 100644 index XXXXXXX..XXXXXXX --- a/docs/specs/conf.py +++ /dev/null @@ -XXX,XX +XXX,XX @@ -# -*- coding: utf-8 -*- -# -# QEMU documentation build configuration file for the 'specs' manual. -# -# This includes the top level conf file and then makes any necessary tweaks. -import sys -import os - -qemu_docdir = os.path.abspath("..") -parent_config = os.path.join(qemu_docdir, "conf.py") -exec(compile(open(parent_config, "rb").read(), parent_config, 'exec')) - -# This slightly misuses the 'description', but is the best way to get -# the manual title to appear in the sidebar. -html_theme_options['description'] = \ - u'System Emulation Guest Hardware Specifications' diff --git a/docs/system/conf.py b/docs/system/conf.py deleted file mode 100644 index XXXXXXX..XXXXXXX --- a/docs/system/conf.py +++ /dev/null @@ -XXX,XX +XXX,XX @@ -# -*- coding: utf-8 -*- -# -# QEMU documentation build configuration file for the 'system' manual. -# -# This includes the top level conf file and then makes any necessary tweaks. -import sys -import os - -qemu_docdir = os.path.abspath("..") -parent_config = os.path.join(qemu_docdir, "conf.py") -exec(compile(open(parent_config, "rb").read(), parent_config, 'exec')) - -# This slightly misuses the 'description', but is the best way to get -# the manual title to appear in the sidebar. -html_theme_options['description'] = u'System Emulation User''s Guide' - -# One entry per manual page. List of tuples -# (source start file, name, description, authors, manual section). -man_pages = [ - ('qemu-manpage', 'qemu', u'QEMU User Documentation', - ['Fabrice Bellard'], 1), - ('qemu-block-drivers', 'qemu-block-drivers', - u'QEMU block drivers reference', - ['Fabrice Bellard and the QEMU Project developers'], 7), - ('qemu-cpu-models', 'qemu-cpu-models', - u'QEMU CPU Models', - ['The QEMU Project developers'], 7) -] diff --git a/docs/tools/conf.py b/docs/tools/conf.py deleted file mode 100644 index XXXXXXX..XXXXXXX --- a/docs/tools/conf.py +++ /dev/null @@ -XXX,XX +XXX,XX @@ -# -*- coding: utf-8 -*- -# -# QEMU documentation build configuration file for the 'tools' manual. -# -# This includes the top level conf file and then makes any necessary tweaks. -import sys -import os - -qemu_docdir = os.path.abspath("..") -parent_config = os.path.join(qemu_docdir, "conf.py") -exec(compile(open(parent_config, "rb").read(), parent_config, 'exec')) - -# This slightly misuses the 'description', but is the best way to get -# the manual title to appear in the sidebar. -html_theme_options['description'] = \ - u'Tools Guide' - -# One entry per manual page. List of tuples -# (source start file, name, description, authors, manual section). -man_pages = [ - ('qemu-img', 'qemu-img', u'QEMU disk image utility', - ['Fabrice Bellard'], 1), - ('qemu-storage-daemon', 'qemu-storage-daemon', u'QEMU storage daemon', - [], 1), - ('qemu-nbd', 'qemu-nbd', u'QEMU Disk Network Block Device Server', - ['Anthony Liguori <anthony@codemonkey.ws>'], 8), - ('qemu-pr-helper', 'qemu-pr-helper', 'QEMU persistent reservation helper', - [], 8), - ('qemu-trace-stap', 'qemu-trace-stap', u'QEMU SystemTap trace tool', - [], 1), - ('virtfs-proxy-helper', 'virtfs-proxy-helper', - u'QEMU 9p virtfs proxy filesystem helper', - ['M. Mohan Kumar'], 1), - ('virtiofsd', 'virtiofsd', u'QEMU virtio-fs shared file system daemon', - ['Stefan Hajnoczi <stefanha@redhat.com>', - 'Masayoshi Mizuma <m.mizuma@jp.fujitsu.com>'], 1), -] diff --git a/docs/user/conf.py b/docs/user/conf.py deleted file mode 100644 index XXXXXXX..XXXXXXX --- a/docs/user/conf.py +++ /dev/null @@ -XXX,XX +XXX,XX @@ -# -*- coding: utf-8 -*- -# -# QEMU documentation build configuration file for the 'user' manual. -# -# This includes the top level conf file and then makes any necessary tweaks. -import sys -import os - -qemu_docdir = os.path.abspath("..") -parent_config = os.path.join(qemu_docdir, "conf.py") -exec(compile(open(parent_config, "rb").read(), parent_config, 'exec')) - -# This slightly misuses the 'description', but is the best way to get -# the manual title to appear in the sidebar. -html_theme_options['description'] = u'User Mode Emulation User''s Guide' -- 2.20.1 In commit cd8be50e58f63413c0 we converted the A32 coprocessor insns to decodetree. This accidentally broke XScale/iWMMXt insns, because it moved the handling of "cp insns which are handled by looking up the cp register in the hashtable" from after the call to the legacy disas_xscale_insn() decode to before it, with the result that all XScale/iWMMXt insns now UNDEF. Update valid_cp() so that it knows that on XScale cp 0 and 1 are not standard coprocessor instructions; this will cause the decodetree trans_ functions to ignore them, so that execution will correctly get through to the legacy decode again. Cc: qemu-stable@nongnu.org Reported-by: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Tested-by: Guenter Roeck <linux@roeck-us.net> Message-id: 20210108195157.32067-1-peter.maydell@linaro.org --- target/arm/translate.c | 7 +++++++ 1 file changed, 7 insertions(+) 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 bool valid_cp(DisasContext *s, int cp) * only cp14 and cp15 are valid, and other values aren't considered * to be in the coprocessor-instruction space at all. v8M still * permits coprocessors 0..7. + * For XScale, we must not decode the XScale cp0, cp1 space as + * a standard coprocessor insn, because we want to fall through to + * the legacy disas_xscale_insn() decoder after decodetree is done. */ + if (arm_dc_feature(s, ARM_FEATURE_XSCALE) && (cp == 0 || cp == 1)) { + return false; + } + if (arm_dc_feature(s, ARM_FEATURE_V8) && !arm_dc_feature(s, ARM_FEATURE_M)) { return cp >= 14; -- 2.20.1 A copy-and-paste error meant that the return value for register offset 0x44 (the RX Status FIFO PEEK register) returned a byte from a bogus offset in the rx status FIFO. Fix the typo. Cc: qemu-stable@nongnu.org Fixes: https://bugs.launchpad.net/qemu/+bug/1904954 Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20210108180401.2263-2-peter.maydell@linaro.org --- hw/net/lan9118.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/hw/net/lan9118.c b/hw/net/lan9118.c index XXXXXXX..XXXXXXX 100644 --- a/hw/net/lan9118.c +++ b/hw/net/lan9118.c @@ -XXX,XX +XXX,XX @@ static uint64_t lan9118_readl(void *opaque, hwaddr offset, case 0x40: return rx_status_fifo_pop(s); case 0x44: - return s->rx_status_fifo[s->tx_status_fifo_head]; + return s->rx_status_fifo[s->rx_status_fifo_head]; case 0x48: return tx_status_fifo_pop(s); case 0x4c: -- 2.20.1 The lan9118 code mostly uses symbolic constants for register offsets; the exceptions are those which the datasheet doesn't give an official symbolic name to. Add some names for the registers which don't already have them, based on the longer names they are given in the memory map. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20210108180401.2263-3-peter.maydell@linaro.org --- hw/net/lan9118.c | 24 ++++++++++++++++++------ 1 file changed, 18 insertions(+), 6 deletions(-) diff --git a/hw/net/lan9118.c b/hw/net/lan9118.c index XXXXXXX..XXXXXXX 100644 --- a/hw/net/lan9118.c +++ b/hw/net/lan9118.c @@ -XXX,XX +XXX,XX @@ do { hw_error("lan9118: error: " fmt , ## __VA_ARGS__);} while (0) do { fprintf(stderr, "lan9118: error: " fmt , ## __VA_ARGS__);} while (0) #endif +/* The tx and rx fifo ports are a range of aliased 32-bit registers */ +#define RX_DATA_FIFO_PORT_FIRST 0x00 +#define RX_DATA_FIFO_PORT_LAST 0x1f +#define TX_DATA_FIFO_PORT_FIRST 0x20 +#define TX_DATA_FIFO_PORT_LAST 0x3f + +#define RX_STATUS_FIFO_PORT 0x40 +#define RX_STATUS_FIFO_PEEK 0x44 +#define TX_STATUS_FIFO_PORT 0x48 +#define TX_STATUS_FIFO_PEEK 0x4c + #define CSR_ID_REV 0x50 #define CSR_IRQ_CFG 0x54 #define CSR_INT_STS 0x58 @@ -XXX,XX +XXX,XX @@ static void lan9118_writel(void *opaque, hwaddr offset, offset &= 0xff; //DPRINTF("Write reg 0x%02x = 0x%08x\n", (int)offset, val); - if (offset >= 0x20 && offset < 0x40) { + if (offset >= TX_DATA_FIFO_PORT_FIRST && + offset <= TX_DATA_FIFO_PORT_LAST) { /* TX FIFO */ tx_fifo_push(s, val); return; @@ -XXX,XX +XXX,XX @@ static uint64_t lan9118_readl(void *opaque, hwaddr offset, lan9118_state *s = (lan9118_state *)opaque; //DPRINTF("Read reg 0x%02x\n", (int)offset); - if (offset < 0x20) { + if (offset <= RX_DATA_FIFO_PORT_LAST) { /* RX FIFO */ return rx_fifo_pop(s); } switch (offset) { - case 0x40: + case RX_STATUS_FIFO_PORT: return rx_status_fifo_pop(s); - case 0x44: + case RX_STATUS_FIFO_PEEK: return s->rx_status_fifo[s->rx_status_fifo_head]; - case 0x48: + case TX_STATUS_FIFO_PORT: return tx_status_fifo_pop(s); - case 0x4c: + case TX_STATUS_FIFO_PEEK: return s->tx_status_fifo[s->tx_status_fifo_head]; case CSR_ID_REV: return 0x01180001; -- 2.20.1 From: Hao Wu <wuhaotsh@google.com> This patch allows NPCM7XX CLK module to compute clocks that are used by other NPCM7XX modules. Add a new struct NPCM7xxClockConverterState which represents a single converter. Each clock converter in CLK module represents one converter in NPCM7XX CLK Module(PLL, SEL or Divider). Each converter takes one or more input clocks and converts them into one output clock. They form a clock hierarchy in the CLK module and are responsible for outputing clocks for various other modules in an NPCM7XX SoC. Each converter has a function pointer called "convert" which represents the unique logic for that converter. The clock contains two initialization information: ConverterInitInfo and ConverterConnectionInfo. They represent the vertices and edges in the clock diagram respectively. Reviewed-by: Havard Skinnemoen <hskinnemoen@google.com> Reviewed-by: Tyrone Ting <kfting@nuvoton.com> Signed-off-by: Hao Wu <wuhaotsh@google.com> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Message-id: 20210108190945.949196-2-wuhaotsh@google.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/misc/npcm7xx_clk.h | 140 +++++- hw/misc/npcm7xx_clk.c | 805 +++++++++++++++++++++++++++++++++- 2 files changed, 932 insertions(+), 13 deletions(-) diff --git a/include/hw/misc/npcm7xx_clk.h b/include/hw/misc/npcm7xx_clk.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/misc/npcm7xx_clk.h +++ b/include/hw/misc/npcm7xx_clk.h @@ -XXX,XX +XXX,XX @@ #define NPCM7XX_CLK_H #include "exec/memory.h" +#include "hw/clock.h" #include "hw/sysbus.h" /* @@ -XXX,XX +XXX,XX @@ #define NPCM7XX_WATCHDOG_RESET_GPIO_IN "npcm7xx-clk-watchdog-reset-gpio-in" -typedef struct NPCM7xxCLKState { +/* Maximum amount of clock inputs in a SEL module. */ +#define NPCM7XX_CLK_SEL_MAX_INPUT 5 + +/* PLLs in CLK module. */ +typedef enum NPCM7xxClockPLL { + NPCM7XX_CLOCK_PLL0, + NPCM7XX_CLOCK_PLL1, + NPCM7XX_CLOCK_PLL2, + NPCM7XX_CLOCK_PLLG, + NPCM7XX_CLOCK_NR_PLLS, +} NPCM7xxClockPLL; + +/* SEL/MUX in CLK module. */ +typedef enum NPCM7xxClockSEL { + NPCM7XX_CLOCK_PIXCKSEL, + NPCM7XX_CLOCK_MCCKSEL, + NPCM7XX_CLOCK_CPUCKSEL, + NPCM7XX_CLOCK_CLKOUTSEL, + NPCM7XX_CLOCK_UARTCKSEL, + NPCM7XX_CLOCK_TIMCKSEL, + NPCM7XX_CLOCK_SDCKSEL, + NPCM7XX_CLOCK_GFXMSEL, + NPCM7XX_CLOCK_SUCKSEL, + NPCM7XX_CLOCK_NR_SELS, +} NPCM7xxClockSEL; + +/* Dividers in CLK module. */ +typedef enum NPCM7xxClockDivider { + NPCM7XX_CLOCK_PLL1D2, /* PLL1/2 */ + NPCM7XX_CLOCK_PLL2D2, /* PLL2/2 */ + NPCM7XX_CLOCK_MC_DIVIDER, + NPCM7XX_CLOCK_AXI_DIVIDER, + NPCM7XX_CLOCK_AHB_DIVIDER, + NPCM7XX_CLOCK_AHB3_DIVIDER, + NPCM7XX_CLOCK_SPI0_DIVIDER, + NPCM7XX_CLOCK_SPIX_DIVIDER, + NPCM7XX_CLOCK_APB1_DIVIDER, + NPCM7XX_CLOCK_APB2_DIVIDER, + NPCM7XX_CLOCK_APB3_DIVIDER, + NPCM7XX_CLOCK_APB4_DIVIDER, + NPCM7XX_CLOCK_APB5_DIVIDER, + NPCM7XX_CLOCK_CLKOUT_DIVIDER, + NPCM7XX_CLOCK_UART_DIVIDER, + NPCM7XX_CLOCK_TIMER_DIVIDER, + NPCM7XX_CLOCK_ADC_DIVIDER, + NPCM7XX_CLOCK_MMC_DIVIDER, + NPCM7XX_CLOCK_SDHC_DIVIDER, + NPCM7XX_CLOCK_GFXM_DIVIDER, /* divide by 3 */ + NPCM7XX_CLOCK_UTMI_DIVIDER, + NPCM7XX_CLOCK_NR_DIVIDERS, +} NPCM7xxClockConverter; + +typedef struct NPCM7xxCLKState NPCM7xxCLKState; + +/** + * struct NPCM7xxClockPLLState - A PLL module in CLK module. + * @name: The name of the module. + * @clk: The CLK module that owns this module. + * @clock_in: The input clock of this module. + * @clock_out: The output clock of this module. + * @reg: The control registers for this PLL module. + */ +typedef struct NPCM7xxClockPLLState { + DeviceState parent; + + const char *name; + NPCM7xxCLKState *clk; + Clock *clock_in; + Clock *clock_out; + + int reg; +} NPCM7xxClockPLLState; + +/** + * struct NPCM7xxClockSELState - A SEL module in CLK module. + * @name: The name of the module. + * @clk: The CLK module that owns this module. + * @input_size: The size of inputs of this module. + * @clock_in: The input clocks of this module. + * @clock_out: The output clocks of this module. + * @offset: The offset of this module in the control register. + * @len: The length of this module in the control register. + */ +typedef struct NPCM7xxClockSELState { + DeviceState parent; + + const char *name; + NPCM7xxCLKState *clk; + uint8_t input_size; + Clock *clock_in[NPCM7XX_CLK_SEL_MAX_INPUT]; + Clock *clock_out; + + int offset; + int len; +} NPCM7xxClockSELState; + +/** + * struct NPCM7xxClockDividerState - A Divider module in CLK module. + * @name: The name of the module. + * @clk: The CLK module that owns this module. + * @clock_in: The input clock of this module. + * @clock_out: The output clock of this module. + * @divide: The function the divider uses to divide the input. + * @reg: The index of the control register that contains the divisor. + * @offset: The offset of the divisor in the control register. + * @len: The length of the divisor in the control register. + * @divisor: The divisor for a constant divisor + */ +typedef struct NPCM7xxClockDividerState { + DeviceState parent; + + const char *name; + NPCM7xxCLKState *clk; + Clock *clock_in; + Clock *clock_out; + + uint32_t (*divide)(struct NPCM7xxClockDividerState *s); + union { + struct { + int reg; + int offset; + int len; + }; + int divisor; + }; +} NPCM7xxClockDividerState; + +struct NPCM7xxCLKState { SysBusDevice parent; MemoryRegion iomem; + /* Clock converters */ + NPCM7xxClockPLLState plls[NPCM7XX_CLOCK_NR_PLLS]; + NPCM7xxClockSELState sels[NPCM7XX_CLOCK_NR_SELS]; + NPCM7xxClockDividerState dividers[NPCM7XX_CLOCK_NR_DIVIDERS]; + uint32_t regs[NPCM7XX_CLK_NR_REGS]; /* Time reference for SECCNT and CNTR25M, initialized by power on reset */ int64_t ref_ns; -} NPCM7xxCLKState; + + /* The incoming reference clock. */ + Clock *clkref; +}; #define TYPE_NPCM7XX_CLK "npcm7xx-clk" #define NPCM7XX_CLK(obj) OBJECT_CHECK(NPCM7xxCLKState, (obj), TYPE_NPCM7XX_CLK) diff --git a/hw/misc/npcm7xx_clk.c b/hw/misc/npcm7xx_clk.c index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/npcm7xx_clk.c +++ b/hw/misc/npcm7xx_clk.c @@ -XXX,XX +XXX,XX @@ #include "hw/misc/npcm7xx_clk.h" #include "hw/timer/npcm7xx_timer.h" +#include "hw/qdev-clock.h" #include "migration/vmstate.h" #include "qemu/error-report.h" #include "qemu/log.h" @@ -XXX,XX +XXX,XX @@ #include "trace.h" #include "sysemu/watchdog.h" +/* + * The reference clock hz, and the SECCNT and CNTR25M registers in this module, + * is always 25 MHz. + */ +#define NPCM7XX_CLOCK_REF_HZ (25000000) + +/* Register Field Definitions */ +#define NPCM7XX_CLK_WDRCR_CA9C BIT(0) /* Cortex A9 Cores */ + #define PLLCON_LOKI BIT(31) #define PLLCON_LOKS BIT(30) #define PLLCON_PWDEN BIT(12) +#define PLLCON_FBDV(con) extract32((con), 16, 12) +#define PLLCON_OTDV2(con) extract32((con), 13, 3) +#define PLLCON_OTDV1(con) extract32((con), 8, 3) +#define PLLCON_INDV(con) extract32((con), 0, 6) enum NPCM7xxCLKRegisters { NPCM7XX_CLK_CLKEN1, @@ -XXX,XX +XXX,XX @@ static const uint32_t cold_reset_values[NPCM7XX_CLK_NR_REGS] = { [NPCM7XX_CLK_AHBCKFI] = 0x000000c8, }; -/* Register Field Definitions */ -#define NPCM7XX_CLK_WDRCR_CA9C BIT(0) /* Cortex A9 Cores */ - /* The number of watchdogs that can trigger a reset. */ #define NPCM7XX_NR_WATCHDOGS (3) +/* Clock converter functions */ + +#define TYPE_NPCM7XX_CLOCK_PLL "npcm7xx-clock-pll" +#define NPCM7XX_CLOCK_PLL(obj) OBJECT_CHECK(NPCM7xxClockPLLState, \ + (obj), TYPE_NPCM7XX_CLOCK_PLL) +#define TYPE_NPCM7XX_CLOCK_SEL "npcm7xx-clock-sel" +#define NPCM7XX_CLOCK_SEL(obj) OBJECT_CHECK(NPCM7xxClockSELState, \ + (obj), TYPE_NPCM7XX_CLOCK_SEL) +#define TYPE_NPCM7XX_CLOCK_DIVIDER "npcm7xx-clock-divider" +#define NPCM7XX_CLOCK_DIVIDER(obj) OBJECT_CHECK(NPCM7xxClockDividerState, \ + (obj), TYPE_NPCM7XX_CLOCK_DIVIDER) + +static void npcm7xx_clk_update_pll(void *opaque) +{ + NPCM7xxClockPLLState *s = opaque; + uint32_t con = s->clk->regs[s->reg]; + uint64_t freq; + + /* The PLL is grounded if it is not locked yet. */ + if (con & PLLCON_LOKI) { + freq = clock_get_hz(s->clock_in); + freq *= PLLCON_FBDV(con); + freq /= PLLCON_INDV(con) * PLLCON_OTDV1(con) * PLLCON_OTDV2(con); + } else { + freq = 0; + } + + clock_update_hz(s->clock_out, freq); +} + +static void npcm7xx_clk_update_sel(void *opaque) +{ + NPCM7xxClockSELState *s = opaque; + uint32_t index = extract32(s->clk->regs[NPCM7XX_CLK_CLKSEL], s->offset, + s->len); + + if (index >= s->input_size) { + qemu_log_mask(LOG_GUEST_ERROR, + "%s: SEL index: %u out of range\n", + __func__, index); + index = 0; + } + clock_update_hz(s->clock_out, clock_get_hz(s->clock_in[index])); +} + +static void npcm7xx_clk_update_divider(void *opaque) +{ + NPCM7xxClockDividerState *s = opaque; + uint32_t freq; + + freq = s->divide(s); + clock_update_hz(s->clock_out, freq); +} + +static uint32_t divide_by_constant(NPCM7xxClockDividerState *s) +{ + return clock_get_hz(s->clock_in) / s->divisor; +} + +static uint32_t divide_by_reg_divisor(NPCM7xxClockDividerState *s) +{ + return clock_get_hz(s->clock_in) / + (extract32(s->clk->regs[s->reg], s->offset, s->len) + 1); +} + +static uint32_t divide_by_reg_divisor_times_2(NPCM7xxClockDividerState *s) +{ + return divide_by_reg_divisor(s) / 2; +} + +static uint32_t shift_by_reg_divisor(NPCM7xxClockDividerState *s) +{ + return clock_get_hz(s->clock_in) >> + extract32(s->clk->regs[s->reg], s->offset, s->len); +} + +static NPCM7xxClockPLL find_pll_by_reg(enum NPCM7xxCLKRegisters reg) +{ + switch (reg) { + case NPCM7XX_CLK_PLLCON0: + return NPCM7XX_CLOCK_PLL0; + case NPCM7XX_CLK_PLLCON1: + return NPCM7XX_CLOCK_PLL1; + case NPCM7XX_CLK_PLLCON2: + return NPCM7XX_CLOCK_PLL2; + case NPCM7XX_CLK_PLLCONG: + return NPCM7XX_CLOCK_PLLG; + default: + g_assert_not_reached(); + } +} + +static void npcm7xx_clk_update_all_plls(NPCM7xxCLKState *clk) +{ + int i; + + for (i = 0; i < NPCM7XX_CLOCK_NR_PLLS; ++i) { + npcm7xx_clk_update_pll(&clk->plls[i]); + } +} + +static void npcm7xx_clk_update_all_sels(NPCM7xxCLKState *clk) +{ + int i; + + for (i = 0; i < NPCM7XX_CLOCK_NR_SELS; ++i) { + npcm7xx_clk_update_sel(&clk->sels[i]); + } +} + +static void npcm7xx_clk_update_all_dividers(NPCM7xxCLKState *clk) +{ + int i; + + for (i = 0; i < NPCM7XX_CLOCK_NR_DIVIDERS; ++i) { + npcm7xx_clk_update_divider(&clk->dividers[i]); + } +} + +static void npcm7xx_clk_update_all_clocks(NPCM7xxCLKState *clk) +{ + clock_update_hz(clk->clkref, NPCM7XX_CLOCK_REF_HZ); + npcm7xx_clk_update_all_plls(clk); + npcm7xx_clk_update_all_sels(clk); + npcm7xx_clk_update_all_dividers(clk); +} + +/* Types of clock sources. */ +typedef enum ClockSrcType { + CLKSRC_REF, + CLKSRC_PLL, + CLKSRC_SEL, + CLKSRC_DIV, +} ClockSrcType; + +typedef struct PLLInitInfo { + const char *name; + ClockSrcType src_type; + int src_index; + int reg; + const char *public_name; +} PLLInitInfo; + +typedef struct SELInitInfo { + const char *name; + uint8_t input_size; + ClockSrcType src_type[NPCM7XX_CLK_SEL_MAX_INPUT]; + int src_index[NPCM7XX_CLK_SEL_MAX_INPUT]; + int offset; + int len; + const char *public_name; +} SELInitInfo; + +typedef struct DividerInitInfo { + const char *name; + ClockSrcType src_type; + int src_index; + uint32_t (*divide)(NPCM7xxClockDividerState *s); + int reg; /* not used when type == CONSTANT */ + int offset; /* not used when type == CONSTANT */ + int len; /* not used when type == CONSTANT */ + int divisor; /* used only when type == CONSTANT */ + const char *public_name; +} DividerInitInfo; + +static const PLLInitInfo pll_init_info_list[] = { + [NPCM7XX_CLOCK_PLL0] = { + .name = "pll0", + .src_type = CLKSRC_REF, + .reg = NPCM7XX_CLK_PLLCON0, + }, + [NPCM7XX_CLOCK_PLL1] = { + .name = "pll1", + .src_type = CLKSRC_REF, + .reg = NPCM7XX_CLK_PLLCON1, + }, + [NPCM7XX_CLOCK_PLL2] = { + .name = "pll2", + .src_type = CLKSRC_REF, + .reg = NPCM7XX_CLK_PLLCON2, + }, + [NPCM7XX_CLOCK_PLLG] = { + .name = "pllg", + .src_type = CLKSRC_REF, + .reg = NPCM7XX_CLK_PLLCONG, + }, +}; + +static const SELInitInfo sel_init_info_list[] = { + [NPCM7XX_CLOCK_PIXCKSEL] = { + .name = "pixcksel", + .input_size = 2, + .src_type = {CLKSRC_PLL, CLKSRC_REF}, + .src_index = {NPCM7XX_CLOCK_PLLG, 0}, + .offset = 5, + .len = 1, + .public_name = "pixel-clock", + }, + [NPCM7XX_CLOCK_MCCKSEL] = { + .name = "mccksel", + .input_size = 4, + .src_type = {CLKSRC_DIV, CLKSRC_REF, CLKSRC_REF, + /*MCBPCK, shouldn't be used in normal operation*/ + CLKSRC_REF}, + .src_index = {NPCM7XX_CLOCK_PLL1D2, 0, 0, 0}, + .offset = 12, + .len = 2, + .public_name = "mc-phy-clock", + }, + [NPCM7XX_CLOCK_CPUCKSEL] = { + .name = "cpucksel", + .input_size = 4, + .src_type = {CLKSRC_PLL, CLKSRC_DIV, CLKSRC_REF, + /*SYSBPCK, shouldn't be used in normal operation*/ + CLKSRC_REF}, + .src_index = {NPCM7XX_CLOCK_PLL0, NPCM7XX_CLOCK_PLL1D2, 0, 0}, + .offset = 0, + .len = 2, + .public_name = "system-clock", + }, + [NPCM7XX_CLOCK_CLKOUTSEL] = { + .name = "clkoutsel", + .input_size = 5, + .src_type = {CLKSRC_PLL, CLKSRC_DIV, CLKSRC_REF, + CLKSRC_PLL, CLKSRC_DIV}, + .src_index = {NPCM7XX_CLOCK_PLL0, NPCM7XX_CLOCK_PLL1D2, 0, + NPCM7XX_CLOCK_PLLG, NPCM7XX_CLOCK_PLL2D2}, + .offset = 18, + .len = 3, + .public_name = "tock", + }, + [NPCM7XX_CLOCK_UARTCKSEL] = { + .name = "uartcksel", + .input_size = 4, + .src_type = {CLKSRC_PLL, CLKSRC_DIV, CLKSRC_REF, CLKSRC_DIV}, + .src_index = {NPCM7XX_CLOCK_PLL0, NPCM7XX_CLOCK_PLL1D2, 0, + NPCM7XX_CLOCK_PLL2D2}, + .offset = 8, + .len = 2, + }, + [NPCM7XX_CLOCK_TIMCKSEL] = { + .name = "timcksel", + .input_size = 4, + .src_type = {CLKSRC_PLL, CLKSRC_DIV, CLKSRC_REF, CLKSRC_DIV}, + .src_index = {NPCM7XX_CLOCK_PLL0, NPCM7XX_CLOCK_PLL1D2, 0, + NPCM7XX_CLOCK_PLL2D2}, + .offset = 14, + .len = 2, + }, + [NPCM7XX_CLOCK_SDCKSEL] = { + .name = "sdcksel", + .input_size = 4, + .src_type = {CLKSRC_PLL, CLKSRC_DIV, CLKSRC_REF, CLKSRC_DIV}, + .src_index = {NPCM7XX_CLOCK_PLL0, NPCM7XX_CLOCK_PLL1D2, 0, + NPCM7XX_CLOCK_PLL2D2}, + .offset = 6, + .len = 2, + }, + [NPCM7XX_CLOCK_GFXMSEL] = { + .name = "gfxmksel", + .input_size = 2, + .src_type = {CLKSRC_REF, CLKSRC_PLL}, + .src_index = {0, NPCM7XX_CLOCK_PLL2}, + .offset = 21, + .len = 1, + }, + [NPCM7XX_CLOCK_SUCKSEL] = { + .name = "sucksel", + .input_size = 4, + .src_type = {CLKSRC_PLL, CLKSRC_DIV, CLKSRC_REF, CLKSRC_DIV}, + .src_index = {NPCM7XX_CLOCK_PLL0, NPCM7XX_CLOCK_PLL1D2, 0, + NPCM7XX_CLOCK_PLL2D2}, + .offset = 10, + .len = 2, + }, +}; + +static const DividerInitInfo divider_init_info_list[] = { + [NPCM7XX_CLOCK_PLL1D2] = { + .name = "pll1d2", + .src_type = CLKSRC_PLL, + .src_index = NPCM7XX_CLOCK_PLL1, + .divide = divide_by_constant, + .divisor = 2, + }, + [NPCM7XX_CLOCK_PLL2D2] = { + .name = "pll2d2", + .src_type = CLKSRC_PLL, + .src_index = NPCM7XX_CLOCK_PLL2, + .divide = divide_by_constant, + .divisor = 2, + }, + [NPCM7XX_CLOCK_MC_DIVIDER] = { + .name = "mc-divider", + .src_type = CLKSRC_SEL, + .src_index = NPCM7XX_CLOCK_MCCKSEL, + .divide = divide_by_constant, + .divisor = 2, + .public_name = "mc-clock" + }, + [NPCM7XX_CLOCK_AXI_DIVIDER] = { + .name = "axi-divider", + .src_type = CLKSRC_SEL, + .src_index = NPCM7XX_CLOCK_CPUCKSEL, + .divide = shift_by_reg_divisor, + .reg = NPCM7XX_CLK_CLKDIV1, + .offset = 0, + .len = 1, + .public_name = "clk2" + }, + [NPCM7XX_CLOCK_AHB_DIVIDER] = { + .name = "ahb-divider", + .src_type = CLKSRC_DIV, + .src_index = NPCM7XX_CLOCK_AXI_DIVIDER, + .divide = divide_by_reg_divisor, + .reg = NPCM7XX_CLK_CLKDIV1, + .offset = 26, + .len = 2, + .public_name = "clk4" + }, + [NPCM7XX_CLOCK_AHB3_DIVIDER] = { + .name = "ahb3-divider", + .src_type = CLKSRC_DIV, + .src_index = NPCM7XX_CLOCK_AHB_DIVIDER, + .divide = divide_by_reg_divisor, + .reg = NPCM7XX_CLK_CLKDIV1, + .offset = 6, + .len = 5, + .public_name = "ahb3-spi3-clock" + }, + [NPCM7XX_CLOCK_SPI0_DIVIDER] = { + .name = "spi0-divider", + .src_type = CLKSRC_DIV, + .src_index = NPCM7XX_CLOCK_AHB_DIVIDER, + .divide = divide_by_reg_divisor, + .reg = NPCM7XX_CLK_CLKDIV3, + .offset = 6, + .len = 5, + .public_name = "spi0-clock", + }, + [NPCM7XX_CLOCK_SPIX_DIVIDER] = { + .name = "spix-divider", + .src_type = CLKSRC_DIV, + .src_index = NPCM7XX_CLOCK_AHB_DIVIDER, + .divide = divide_by_reg_divisor, + .reg = NPCM7XX_CLK_CLKDIV3, + .offset = 1, + .len = 5, + .public_name = "spix-clock", + }, + [NPCM7XX_CLOCK_APB1_DIVIDER] = { + .name = "apb1-divider", + .src_type = CLKSRC_DIV, + .src_index = NPCM7XX_CLOCK_AHB_DIVIDER, + .divide = shift_by_reg_divisor, + .reg = NPCM7XX_CLK_CLKDIV2, + .offset = 24, + .len = 2, + .public_name = "apb1-clock", + }, + [NPCM7XX_CLOCK_APB2_DIVIDER] = { + .name = "apb2-divider", + .src_type = CLKSRC_DIV, + .src_index = NPCM7XX_CLOCK_AHB_DIVIDER, + .divide = shift_by_reg_divisor, + .reg = NPCM7XX_CLK_CLKDIV2, + .offset = 26, + .len = 2, + .public_name = "apb2-clock", + }, + [NPCM7XX_CLOCK_APB3_DIVIDER] = { + .name = "apb3-divider", + .src_type = CLKSRC_DIV, + .src_index = NPCM7XX_CLOCK_AHB_DIVIDER, + .divide = shift_by_reg_divisor, + .reg = NPCM7XX_CLK_CLKDIV2, + .offset = 28, + .len = 2, + .public_name = "apb3-clock", + }, + [NPCM7XX_CLOCK_APB4_DIVIDER] = { + .name = "apb4-divider", + .src_type = CLKSRC_DIV, + .src_index = NPCM7XX_CLOCK_AHB_DIVIDER, + .divide = shift_by_reg_divisor, + .reg = NPCM7XX_CLK_CLKDIV2, + .offset = 30, + .len = 2, + .public_name = "apb4-clock", + }, + [NPCM7XX_CLOCK_APB5_DIVIDER] = { + .name = "apb5-divider", + .src_type = CLKSRC_DIV, + .src_index = NPCM7XX_CLOCK_AHB_DIVIDER, + .divide = shift_by_reg_divisor, + .reg = NPCM7XX_CLK_CLKDIV2, + .offset = 22, + .len = 2, + .public_name = "apb5-clock", + }, + [NPCM7XX_CLOCK_CLKOUT_DIVIDER] = { + .name = "clkout-divider", + .src_type = CLKSRC_SEL, + .src_index = NPCM7XX_CLOCK_CLKOUTSEL, + .divide = divide_by_reg_divisor, + .reg = NPCM7XX_CLK_CLKDIV2, + .offset = 16, + .len = 5, + .public_name = "clkout", + }, + [NPCM7XX_CLOCK_UART_DIVIDER] = { + .name = "uart-divider", + .src_type = CLKSRC_SEL, + .src_index = NPCM7XX_CLOCK_UARTCKSEL, + .divide = divide_by_reg_divisor, + .reg = NPCM7XX_CLK_CLKDIV1, + .offset = 16, + .len = 5, + .public_name = "uart-clock", + }, + [NPCM7XX_CLOCK_TIMER_DIVIDER] = { + .name = "timer-divider", + .src_type = CLKSRC_SEL, + .src_index = NPCM7XX_CLOCK_TIMCKSEL, + .divide = divide_by_reg_divisor, + .reg = NPCM7XX_CLK_CLKDIV1, + .offset = 21, + .len = 5, + .public_name = "timer-clock", + }, + [NPCM7XX_CLOCK_ADC_DIVIDER] = { + .name = "adc-divider", + .src_type = CLKSRC_DIV, + .src_index = NPCM7XX_CLOCK_TIMER_DIVIDER, + .divide = shift_by_reg_divisor, + .reg = NPCM7XX_CLK_CLKDIV1, + .offset = 28, + .len = 3, + .public_name = "adc-clock", + }, + [NPCM7XX_CLOCK_MMC_DIVIDER] = { + .name = "mmc-divider", + .src_type = CLKSRC_SEL, + .src_index = NPCM7XX_CLOCK_SDCKSEL, + .divide = divide_by_reg_divisor, + .reg = NPCM7XX_CLK_CLKDIV1, + .offset = 11, + .len = 5, + .public_name = "mmc-clock", + }, + [NPCM7XX_CLOCK_SDHC_DIVIDER] = { + .name = "sdhc-divider", + .src_type = CLKSRC_SEL, + .src_index = NPCM7XX_CLOCK_SDCKSEL, + .divide = divide_by_reg_divisor_times_2, + .reg = NPCM7XX_CLK_CLKDIV2, + .offset = 0, + .len = 4, + .public_name = "sdhc-clock", + }, + [NPCM7XX_CLOCK_GFXM_DIVIDER] = { + .name = "gfxm-divider", + .src_type = CLKSRC_SEL, + .src_index = NPCM7XX_CLOCK_GFXMSEL, + .divide = divide_by_constant, + .divisor = 3, + .public_name = "gfxm-clock", + }, + [NPCM7XX_CLOCK_UTMI_DIVIDER] = { + .name = "utmi-divider", + .src_type = CLKSRC_SEL, + .src_index = NPCM7XX_CLOCK_SUCKSEL, + .divide = divide_by_reg_divisor, + .reg = NPCM7XX_CLK_CLKDIV2, + .offset = 8, + .len = 5, + .public_name = "utmi-clock", + }, +}; + +static void npcm7xx_clk_pll_init(Object *obj) +{ + NPCM7xxClockPLLState *pll = NPCM7XX_CLOCK_PLL(obj); + + pll->clock_in = qdev_init_clock_in(DEVICE(pll), "clock-in", + npcm7xx_clk_update_pll, pll); + pll->clock_out = qdev_init_clock_out(DEVICE(pll), "clock-out"); +} + +static void npcm7xx_clk_sel_init(Object *obj) +{ + int i; + NPCM7xxClockSELState *sel = NPCM7XX_CLOCK_SEL(obj); + + for (i = 0; i < NPCM7XX_CLK_SEL_MAX_INPUT; ++i) { + sel->clock_in[i] = qdev_init_clock_in(DEVICE(sel), + g_strdup_printf("clock-in[%d]", i), + npcm7xx_clk_update_sel, sel); + } + sel->clock_out = qdev_init_clock_out(DEVICE(sel), "clock-out"); +} +static void npcm7xx_clk_divider_init(Object *obj) +{ + NPCM7xxClockDividerState *div = NPCM7XX_CLOCK_DIVIDER(obj); + + div->clock_in = qdev_init_clock_in(DEVICE(div), "clock-in", + npcm7xx_clk_update_divider, div); + div->clock_out = qdev_init_clock_out(DEVICE(div), "clock-out"); +} + +static void npcm7xx_init_clock_pll(NPCM7xxClockPLLState *pll, + NPCM7xxCLKState *clk, const PLLInitInfo *init_info) +{ + pll->name = init_info->name; + pll->clk = clk; + pll->reg = init_info->reg; + if (init_info->public_name != NULL) { + qdev_alias_clock(DEVICE(pll), "clock-out", DEVICE(clk), + init_info->public_name); + } +} + +static void npcm7xx_init_clock_sel(NPCM7xxClockSELState *sel, + NPCM7xxCLKState *clk, const SELInitInfo *init_info) +{ + int input_size = init_info->input_size; + + sel->name = init_info->name; + sel->clk = clk; + sel->input_size = init_info->input_size; + g_assert(input_size <= NPCM7XX_CLK_SEL_MAX_INPUT); + sel->offset = init_info->offset; + sel->len = init_info->len; + if (init_info->public_name != NULL) { + qdev_alias_clock(DEVICE(sel), "clock-out", DEVICE(clk), + init_info->public_name); + } +} + +static void npcm7xx_init_clock_divider(NPCM7xxClockDividerState *div, + NPCM7xxCLKState *clk, const DividerInitInfo *init_info) +{ + div->name = init_info->name; + div->clk = clk; + + div->divide = init_info->divide; + if (div->divide == divide_by_constant) { + div->divisor = init_info->divisor; + } else { + div->reg = init_info->reg; + div->offset = init_info->offset; + div->len = init_info->len; + } + if (init_info->public_name != NULL) { + qdev_alias_clock(DEVICE(div), "clock-out", DEVICE(clk), + init_info->public_name); + } +} + +static Clock *npcm7xx_get_clock(NPCM7xxCLKState *clk, ClockSrcType type, + int index) +{ + switch (type) { + case CLKSRC_REF: + return clk->clkref; + case CLKSRC_PLL: + return clk->plls[index].clock_out; + case CLKSRC_SEL: + return clk->sels[index].clock_out; + case CLKSRC_DIV: + return clk->dividers[index].clock_out; + default: + g_assert_not_reached(); + } +} + +static void npcm7xx_connect_clocks(NPCM7xxCLKState *clk) +{ + int i, j; + Clock *src; + + for (i = 0; i < NPCM7XX_CLOCK_NR_PLLS; ++i) { + src = npcm7xx_get_clock(clk, pll_init_info_list[i].src_type, + pll_init_info_list[i].src_index); + clock_set_source(clk->plls[i].clock_in, src); + } + for (i = 0; i < NPCM7XX_CLOCK_NR_SELS; ++i) { + for (j = 0; j < sel_init_info_list[i].input_size; ++j) { + src = npcm7xx_get_clock(clk, sel_init_info_list[i].src_type[j], + sel_init_info_list[i].src_index[j]); + clock_set_source(clk->sels[i].clock_in[j], src); + } + } + for (i = 0; i < NPCM7XX_CLOCK_NR_DIVIDERS; ++i) { + src = npcm7xx_get_clock(clk, divider_init_info_list[i].src_type, + divider_init_info_list[i].src_index); + clock_set_source(clk->dividers[i].clock_in, src); + } +} + static uint64_t npcm7xx_clk_read(void *opaque, hwaddr offset, unsigned size) { uint32_t reg = offset / sizeof(uint32_t); @@ -XXX,XX +XXX,XX @@ static uint64_t npcm7xx_clk_read(void *opaque, hwaddr offset, unsigned size) * * The 4 LSBs are always zero: (1e9 / 640) << 4 = 25000000. */ - value = (((now_ns - s->ref_ns) / 640) << 4) % NPCM7XX_TIMER_REF_HZ; + value = (((now_ns - s->ref_ns) / 640) << 4) % NPCM7XX_CLOCK_REF_HZ; break; default: @@ -XXX,XX +XXX,XX @@ static void npcm7xx_clk_write(void *opaque, hwaddr offset, value |= (value & PLLCON_LOKS); } } + /* Only update PLL when it is locked. */ + if (value & PLLCON_LOKI) { + npcm7xx_clk_update_pll(&s->plls[find_pll_by_reg(reg)]); + } + break; + + case NPCM7XX_CLK_CLKSEL: + npcm7xx_clk_update_all_sels(s); + break; + + case NPCM7XX_CLK_CLKDIV1: + case NPCM7XX_CLK_CLKDIV2: + case NPCM7XX_CLK_CLKDIV3: + npcm7xx_clk_update_all_dividers(s); break; case NPCM7XX_CLK_CNTR25M: @@ -XXX,XX +XXX,XX @@ static void npcm7xx_clk_enter_reset(Object *obj, ResetType type) case RESET_TYPE_COLD: memcpy(s->regs, cold_reset_values, sizeof(cold_reset_values)); s->ref_ns = qemu_clock_get_ns(QEMU_CLOCK_VIRTUAL); + npcm7xx_clk_update_all_clocks(s); return; } @@ -XXX,XX +XXX,XX @@ static void npcm7xx_clk_enter_reset(Object *obj, ResetType type) __func__, type); } +static void npcm7xx_clk_init_clock_hierarchy(NPCM7xxCLKState *s) +{ + int i; + + s->clkref = qdev_init_clock_in(DEVICE(s), "clkref", NULL, NULL); + + /* First pass: init all converter modules */ + QEMU_BUILD_BUG_ON(ARRAY_SIZE(pll_init_info_list) != NPCM7XX_CLOCK_NR_PLLS); + QEMU_BUILD_BUG_ON(ARRAY_SIZE(sel_init_info_list) != NPCM7XX_CLOCK_NR_SELS); + QEMU_BUILD_BUG_ON(ARRAY_SIZE(divider_init_info_list) + != NPCM7XX_CLOCK_NR_DIVIDERS); + for (i = 0; i < NPCM7XX_CLOCK_NR_PLLS; ++i) { + object_initialize_child(OBJECT(s), pll_init_info_list[i].name, + &s->plls[i], TYPE_NPCM7XX_CLOCK_PLL); + npcm7xx_init_clock_pll(&s->plls[i], s, + &pll_init_info_list[i]); + } + for (i = 0; i < NPCM7XX_CLOCK_NR_SELS; ++i) { + object_initialize_child(OBJECT(s), sel_init_info_list[i].name, + &s->sels[i], TYPE_NPCM7XX_CLOCK_SEL); + npcm7xx_init_clock_sel(&s->sels[i], s, + &sel_init_info_list[i]); + } + for (i = 0; i < NPCM7XX_CLOCK_NR_DIVIDERS; ++i) { + object_initialize_child(OBJECT(s), divider_init_info_list[i].name, + &s->dividers[i], TYPE_NPCM7XX_CLOCK_DIVIDER); + npcm7xx_init_clock_divider(&s->dividers[i], s, + &divider_init_info_list[i]); + } + + /* Second pass: connect converter modules */ + npcm7xx_connect_clocks(s); + + clock_update_hz(s->clkref, NPCM7XX_CLOCK_REF_HZ); +} + static void npcm7xx_clk_init(Object *obj) { NPCM7xxCLKState *s = NPCM7XX_CLK(obj); @@ -XXX,XX +XXX,XX @@ static void npcm7xx_clk_init(Object *obj) memory_region_init_io(&s->iomem, obj, &npcm7xx_clk_ops, s, TYPE_NPCM7XX_CLK, 4 * KiB); sysbus_init_mmio(&s->parent, &s->iomem); - qdev_init_gpio_in_named(DEVICE(s), npcm7xx_clk_perform_watchdog_reset, - NPCM7XX_WATCHDOG_RESET_GPIO_IN, NPCM7XX_NR_WATCHDOGS); } -static const VMStateDescription vmstate_npcm7xx_clk = { - .name = "npcm7xx-clk", +static int npcm7xx_clk_post_load(void *opaque, int version_id) +{ + if (version_id >= 1) { + NPCM7xxCLKState *clk = opaque; + + npcm7xx_clk_update_all_clocks(clk); + } + + return 0; +} + +static void npcm7xx_clk_realize(DeviceState *dev, Error **errp) +{ + int i; + NPCM7xxCLKState *s = NPCM7XX_CLK(dev); + + qdev_init_gpio_in_named(DEVICE(s), npcm7xx_clk_perform_watchdog_reset, + NPCM7XX_WATCHDOG_RESET_GPIO_IN, NPCM7XX_NR_WATCHDOGS); + npcm7xx_clk_init_clock_hierarchy(s); + + /* Realize child devices */ + for (i = 0; i < NPCM7XX_CLOCK_NR_PLLS; ++i) { + if (!qdev_realize(DEVICE(&s->plls[i]), NULL, errp)) { + return; + } + } + for (i = 0; i < NPCM7XX_CLOCK_NR_SELS; ++i) { + if (!qdev_realize(DEVICE(&s->sels[i]), NULL, errp)) { + return; + } + } + for (i = 0; i < NPCM7XX_CLOCK_NR_DIVIDERS; ++i) { + if (!qdev_realize(DEVICE(&s->dividers[i]), NULL, errp)) { + return; + } + } +} + +static const VMStateDescription vmstate_npcm7xx_clk_pll = { + .name = "npcm7xx-clock-pll", .version_id = 0, .minimum_version_id = 0, - .fields = (VMStateField[]) { - VMSTATE_UINT32_ARRAY(regs, NPCM7xxCLKState, NPCM7XX_CLK_NR_REGS), - VMSTATE_INT64(ref_ns, NPCM7xxCLKState), + .fields = (VMStateField[]) { + VMSTATE_CLOCK(clock_in, NPCM7xxClockPLLState), VMSTATE_END_OF_LIST(), }, }; +static const VMStateDescription vmstate_npcm7xx_clk_sel = { + .name = "npcm7xx-clock-sel", + .version_id = 0, + .minimum_version_id = 0, + .fields = (VMStateField[]) { + VMSTATE_ARRAY_OF_POINTER_TO_STRUCT(clock_in, NPCM7xxClockSELState, + NPCM7XX_CLK_SEL_MAX_INPUT, 0, vmstate_clock, Clock), + VMSTATE_END_OF_LIST(), + }, +}; + +static const VMStateDescription vmstate_npcm7xx_clk_divider = { + .name = "npcm7xx-clock-divider", + .version_id = 0, + .minimum_version_id = 0, + .fields = (VMStateField[]) { + VMSTATE_CLOCK(clock_in, NPCM7xxClockDividerState), + VMSTATE_END_OF_LIST(), + }, +}; + +static const VMStateDescription vmstate_npcm7xx_clk = { + .name = "npcm7xx-clk", + .version_id = 1, + .minimum_version_id = 1, + .post_load = npcm7xx_clk_post_load, + .fields = (VMStateField[]) { + VMSTATE_UINT32_ARRAY(regs, NPCM7xxCLKState, NPCM7XX_CLK_NR_REGS), + VMSTATE_INT64(ref_ns, NPCM7xxCLKState), + VMSTATE_CLOCK(clkref, NPCM7xxCLKState), + VMSTATE_END_OF_LIST(), + }, +}; + +static void npcm7xx_clk_pll_class_init(ObjectClass *klass, void *data) +{ + DeviceClass *dc = DEVICE_CLASS(klass); + + dc->desc = "NPCM7xx Clock PLL Module"; + dc->vmsd = &vmstate_npcm7xx_clk_pll; +} + +static void npcm7xx_clk_sel_class_init(ObjectClass *klass, void *data) +{ + DeviceClass *dc = DEVICE_CLASS(klass); + + dc->desc = "NPCM7xx Clock SEL Module"; + dc->vmsd = &vmstate_npcm7xx_clk_sel; +} + +static void npcm7xx_clk_divider_class_init(ObjectClass *klass, void *data) +{ + DeviceClass *dc = DEVICE_CLASS(klass); + + dc->desc = "NPCM7xx Clock Divider Module"; + dc->vmsd = &vmstate_npcm7xx_clk_divider; +} + static void npcm7xx_clk_class_init(ObjectClass *klass, void *data) { ResettableClass *rc = RESETTABLE_CLASS(klass); @@ -XXX,XX +XXX,XX @@ static void npcm7xx_clk_class_init(ObjectClass *klass, void *data) dc->desc = "NPCM7xx Clock Control Registers"; dc->vmsd = &vmstate_npcm7xx_clk; + dc->realize = npcm7xx_clk_realize; rc->phases.enter = npcm7xx_clk_enter_reset; } +static const TypeInfo npcm7xx_clk_pll_info = { + .name = TYPE_NPCM7XX_CLOCK_PLL, + .parent = TYPE_DEVICE, + .instance_size = sizeof(NPCM7xxClockPLLState), + .instance_init = npcm7xx_clk_pll_init, + .class_init = npcm7xx_clk_pll_class_init, +}; + +static const TypeInfo npcm7xx_clk_sel_info = { + .name = TYPE_NPCM7XX_CLOCK_SEL, + .parent = TYPE_DEVICE, + .instance_size = sizeof(NPCM7xxClockSELState), + .instance_init = npcm7xx_clk_sel_init, + .class_init = npcm7xx_clk_sel_class_init, +}; + +static const TypeInfo npcm7xx_clk_divider_info = { + .name = TYPE_NPCM7XX_CLOCK_DIVIDER, + .parent = TYPE_DEVICE, + .instance_size = sizeof(NPCM7xxClockDividerState), + .instance_init = npcm7xx_clk_divider_init, + .class_init = npcm7xx_clk_divider_class_init, +}; + static const TypeInfo npcm7xx_clk_info = { .name = TYPE_NPCM7XX_CLK, .parent = TYPE_SYS_BUS_DEVICE, @@ -XXX,XX +XXX,XX @@ static const TypeInfo npcm7xx_clk_info = { static void npcm7xx_clk_register_type(void) { + type_register_static(&npcm7xx_clk_pll_info); + type_register_static(&npcm7xx_clk_sel_info); + type_register_static(&npcm7xx_clk_divider_info); type_register_static(&npcm7xx_clk_info); } type_init(npcm7xx_clk_register_type); -- 2.20.1 From: Hao Wu <wuhaotsh@google.com> This patch makes NPCM7XX Timer to use a the timer clock generated by the CLK module instead of the magic number TIMER_REF_HZ. Reviewed-by: Havard Skinnemoen <hskinnemoen@google.com> Reviewed-by: Tyrone Ting <kfting@nuvoton.com> Signed-off-by: Hao Wu <wuhaotsh@google.com> Message-id: 20210108190945.949196-3-wuhaotsh@google.com Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/misc/npcm7xx_clk.h | 6 ----- include/hw/timer/npcm7xx_timer.h | 1 + hw/arm/npcm7xx.c | 5 ++++ hw/timer/npcm7xx_timer.c | 39 +++++++++++++++----------------- 4 files changed, 24 insertions(+), 27 deletions(-) diff --git a/include/hw/misc/npcm7xx_clk.h b/include/hw/misc/npcm7xx_clk.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/misc/npcm7xx_clk.h +++ b/include/hw/misc/npcm7xx_clk.h @@ -XXX,XX +XXX,XX @@ #include "hw/clock.h" #include "hw/sysbus.h" -/* - * The reference clock frequency for the timer modules, and the SECCNT and - * CNTR25M registers in this module, is always 25 MHz. - */ -#define NPCM7XX_TIMER_REF_HZ (25000000) - /* * Number of registers in our device state structure. Don't change this without * incrementing the version_id in the vmstate. diff --git a/include/hw/timer/npcm7xx_timer.h b/include/hw/timer/npcm7xx_timer.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/timer/npcm7xx_timer.h +++ b/include/hw/timer/npcm7xx_timer.h @@ -XXX,XX +XXX,XX @@ struct NPCM7xxTimerCtrlState { uint32_t tisr; + Clock *clock; NPCM7xxTimer timer[NPCM7XX_TIMERS_PER_CTRL]; NPCM7xxWatchdogTimer watchdog_timer; }; diff --git a/hw/arm/npcm7xx.c b/hw/arm/npcm7xx.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/npcm7xx.c +++ b/hw/arm/npcm7xx.c @@ -XXX,XX +XXX,XX @@ #include "hw/char/serial.h" #include "hw/loader.h" #include "hw/misc/unimp.h" +#include "hw/qdev-clock.h" #include "hw/qdev-properties.h" #include "qapi/error.h" #include "qemu/units.h" @@ -XXX,XX +XXX,XX @@ static void npcm7xx_realize(DeviceState *dev, Error **errp) int first_irq; int j; + /* Connect the timer clock. */ + qdev_connect_clock_in(DEVICE(&s->tim[i]), "clock", qdev_get_clock_out( + DEVICE(&s->clk), "timer-clock")); + sysbus_realize(sbd, &error_abort); sysbus_mmio_map(sbd, 0, npcm7xx_tim_addr[i]); diff --git a/hw/timer/npcm7xx_timer.c b/hw/timer/npcm7xx_timer.c index XXXXXXX..XXXXXXX 100644 --- a/hw/timer/npcm7xx_timer.c +++ b/hw/timer/npcm7xx_timer.c @@ -XXX,XX +XXX,XX @@ #include "qemu/osdep.h" #include "hw/irq.h" +#include "hw/qdev-clock.h" #include "hw/qdev-properties.h" -#include "hw/misc/npcm7xx_clk.h" #include "hw/timer/npcm7xx_timer.h" #include "migration/vmstate.h" #include "qemu/bitops.h" @@ -XXX,XX +XXX,XX @@ static uint32_t npcm7xx_tcsr_prescaler(uint32_t tcsr) /* Convert a timer cycle count to a time interval in nanoseconds. */ static int64_t npcm7xx_timer_count_to_ns(NPCM7xxTimer *t, uint32_t count) { - int64_t ns = count; + int64_t ticks = count; - ns *= NANOSECONDS_PER_SECOND / NPCM7XX_TIMER_REF_HZ; - ns *= npcm7xx_tcsr_prescaler(t->tcsr); + ticks *= npcm7xx_tcsr_prescaler(t->tcsr); - return ns; + return clock_ticks_to_ns(t->ctrl->clock, ticks); } /* Convert a time interval in nanoseconds to a timer cycle count. */ static uint32_t npcm7xx_timer_ns_to_count(NPCM7xxTimer *t, int64_t ns) { - int64_t count; - - count = ns / (NANOSECONDS_PER_SECOND / NPCM7XX_TIMER_REF_HZ); - count /= npcm7xx_tcsr_prescaler(t->tcsr); - - return count; + return ns / clock_ticks_to_ns(t->ctrl->clock, + npcm7xx_tcsr_prescaler(t->tcsr)); } static uint32_t npcm7xx_watchdog_timer_prescaler(const NPCM7xxWatchdogTimer *t) @@ -XXX,XX +XXX,XX @@ static uint32_t npcm7xx_watchdog_timer_prescaler(const NPCM7xxWatchdogTimer *t) static void npcm7xx_watchdog_timer_reset_cycles(NPCM7xxWatchdogTimer *t, int64_t cycles) { - uint32_t prescaler = npcm7xx_watchdog_timer_prescaler(t); - int64_t ns = (NANOSECONDS_PER_SECOND / NPCM7XX_TIMER_REF_HZ) * cycles; + int64_t ticks = cycles * npcm7xx_watchdog_timer_prescaler(t); + int64_t ns = clock_ticks_to_ns(t->ctrl->clock, ticks); /* * The reset function always clears the current timer. The caller of the @@ -XXX,XX +XXX,XX @@ static void npcm7xx_watchdog_timer_reset_cycles(NPCM7xxWatchdogTimer *t, */ npcm7xx_timer_clear(&t->base_timer); - ns *= prescaler; t->base_timer.remaining_ns = ns; } @@ -XXX,XX +XXX,XX @@ static void npcm7xx_timer_hold_reset(Object *obj) qemu_irq_lower(s->watchdog_timer.irq); } -static void npcm7xx_timer_realize(DeviceState *dev, Error **errp) +static void npcm7xx_timer_init(Object *obj) { - NPCM7xxTimerCtrlState *s = NPCM7XX_TIMER(dev); - SysBusDevice *sbd = &s->parent; + NPCM7xxTimerCtrlState *s = NPCM7XX_TIMER(obj); + DeviceState *dev = DEVICE(obj); + SysBusDevice *sbd = SYS_BUS_DEVICE(obj); int i; NPCM7xxWatchdogTimer *w; @@ -XXX,XX +XXX,XX @@ static void npcm7xx_timer_realize(DeviceState *dev, Error **errp) npcm7xx_watchdog_timer_expired, w); sysbus_init_irq(sbd, &w->irq); - memory_region_init_io(&s->iomem, OBJECT(s), &npcm7xx_timer_ops, s, + memory_region_init_io(&s->iomem, obj, &npcm7xx_timer_ops, s, TYPE_NPCM7XX_TIMER, 4 * KiB); sysbus_init_mmio(sbd, &s->iomem); qdev_init_gpio_out_named(dev, &w->reset_signal, NPCM7XX_WATCHDOG_RESET_GPIO_OUT, 1); + s->clock = qdev_init_clock_in(dev, "clock", NULL, NULL); } static const VMStateDescription vmstate_npcm7xx_base_timer = { @@ -XXX,XX +XXX,XX @@ static const VMStateDescription vmstate_npcm7xx_watchdog_timer = { static const VMStateDescription vmstate_npcm7xx_timer_ctrl = { .name = "npcm7xx-timer-ctrl", - .version_id = 1, - .minimum_version_id = 1, + .version_id = 2, + .minimum_version_id = 2, .fields = (VMStateField[]) { VMSTATE_UINT32(tisr, NPCM7xxTimerCtrlState), + VMSTATE_CLOCK(clock, NPCM7xxTimerCtrlState), VMSTATE_STRUCT_ARRAY(timer, NPCM7xxTimerCtrlState, NPCM7XX_TIMERS_PER_CTRL, 0, vmstate_npcm7xx_timer, NPCM7xxTimer), @@ -XXX,XX +XXX,XX @@ static void npcm7xx_timer_class_init(ObjectClass *klass, void *data) QEMU_BUILD_BUG_ON(NPCM7XX_TIMER_REGS_END > NPCM7XX_TIMER_NR_REGS); dc->desc = "NPCM7xx Timer Controller"; - dc->realize = npcm7xx_timer_realize; dc->vmsd = &vmstate_npcm7xx_timer_ctrl; rc->phases.enter = npcm7xx_timer_enter_reset; rc->phases.hold = npcm7xx_timer_hold_reset; @@ -XXX,XX +XXX,XX @@ static const TypeInfo npcm7xx_timer_info = { .parent = TYPE_SYS_BUS_DEVICE, .instance_size = sizeof(NPCM7xxTimerCtrlState), .class_init = npcm7xx_timer_class_init, + .instance_init = npcm7xx_timer_init, }; static void npcm7xx_timer_register_type(void) -- 2.20.1 From: Hao Wu <wuhaotsh@google.com> The ADC is part of NPCM7XX Module. Its behavior is controled by the ADC_CON register. It converts one of the eight analog inputs into a digital input and stores it in the ADC_DATA register when enabled. Users can alter input value by using qom-set QMP command. Reviewed-by: Havard Skinnemoen <hskinnemoen@google.com> Reviewed-by: Tyrone Ting <kfting@nuvoton.com> Signed-off-by: Hao Wu <wuhaotsh@google.com> Message-id: 20210108190945.949196-4-wuhaotsh@google.com [PMM: Added missing hw/adc/trace.h file] Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- docs/system/arm/nuvoton.rst | 2 +- meson.build | 1 + hw/adc/trace.h | 1 + include/hw/adc/npcm7xx_adc.h | 69 ++++++ include/hw/arm/npcm7xx.h | 2 + hw/adc/npcm7xx_adc.c | 301 ++++++++++++++++++++++++++ hw/arm/npcm7xx.c | 24 ++- tests/qtest/npcm7xx_adc-test.c | 377 +++++++++++++++++++++++++++++++++ hw/adc/meson.build | 1 + hw/adc/trace-events | 5 + tests/qtest/meson.build | 3 +- 11 files changed, 783 insertions(+), 3 deletions(-) create mode 100644 hw/adc/trace.h create mode 100644 include/hw/adc/npcm7xx_adc.h create mode 100644 hw/adc/npcm7xx_adc.c create mode 100644 tests/qtest/npcm7xx_adc-test.c create mode 100644 hw/adc/trace-events diff --git a/docs/system/arm/nuvoton.rst b/docs/system/arm/nuvoton.rst index XXXXXXX..XXXXXXX 100644 --- a/docs/system/arm/nuvoton.rst +++ b/docs/system/arm/nuvoton.rst @@ -XXX,XX +XXX,XX @@ Supported devices * Random Number Generator (RNG) * USB host (USBH) * GPIO controller + * Analog to Digital Converter (ADC) Missing devices --------------- @@ -XXX,XX +XXX,XX @@ Missing devices * USB device (USBD) * SMBus controller (SMBF) * Peripheral SPI controller (PSPI) - * Analog to Digital Converter (ADC) * SD/MMC host * PECI interface * Pulse Width Modulation (PWM) diff --git a/meson.build b/meson.build index XXXXXXX..XXXXXXX 100644 --- a/meson.build +++ b/meson.build @@ -XXX,XX +XXX,XX @@ if have_system 'chardev', 'hw/9pfs', 'hw/acpi', + 'hw/adc', 'hw/alpha', 'hw/arm', 'hw/audio', diff --git a/hw/adc/trace.h b/hw/adc/trace.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/adc/trace.h @@ -0,0 +1 @@ +#include "trace/trace-hw_adc.h" diff --git a/include/hw/adc/npcm7xx_adc.h b/include/hw/adc/npcm7xx_adc.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/include/hw/adc/npcm7xx_adc.h @@ -XXX,XX +XXX,XX @@ +/* + * Nuvoton NPCM7xx ADC Module + * + * Copyright 2020 Google LLC + * + * 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. + */ +#ifndef NPCM7XX_ADC_H +#define NPCM7XX_ADC_H + +#include "hw/clock.h" +#include "hw/irq.h" +#include "hw/sysbus.h" +#include "qemu/timer.h" + +#define NPCM7XX_ADC_NUM_INPUTS 8 +/** + * This value should not be changed unless write_adc_calibration function in + * hw/arm/npcm7xx.c is also changed. + */ +#define NPCM7XX_ADC_NUM_CALIB 2 + +/** + * struct NPCM7xxADCState - Analog to Digital Converter Module device state. + * @parent: System bus device. + * @iomem: Memory region through which registers are accessed. + * @conv_timer: The timer counts down remaining cycles for the conversion. + * @irq: GIC interrupt line to fire on expiration (if enabled). + * @con: The Control Register. + * @data: The Data Buffer. + * @clock: The ADC Clock. + * @adci: The input voltage in units of uV. 1uv = 1e-6V. + * @vref: The external reference voltage. + * @iref: The internal reference voltage, initialized at launch time. + * @rv: The calibrated output values of 0.5V and 1.5V for the ADC. + */ +typedef struct { + SysBusDevice parent; + + MemoryRegion iomem; + + QEMUTimer conv_timer; + + qemu_irq irq; + uint32_t con; + uint32_t data; + Clock *clock; + + /* Voltages are in unit of uV. 1V = 1000000uV. */ + uint32_t adci[NPCM7XX_ADC_NUM_INPUTS]; + uint32_t vref; + uint32_t iref; + + uint16_t calibration_r_values[NPCM7XX_ADC_NUM_CALIB]; +} NPCM7xxADCState; + +#define TYPE_NPCM7XX_ADC "npcm7xx-adc" +#define NPCM7XX_ADC(obj) \ + OBJECT_CHECK(NPCM7xxADCState, (obj), TYPE_NPCM7XX_ADC) + +#endif /* NPCM7XX_ADC_H */ diff --git a/include/hw/arm/npcm7xx.h b/include/hw/arm/npcm7xx.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/npcm7xx.h +++ b/include/hw/arm/npcm7xx.h @@ -XXX,XX +XXX,XX @@ #define NPCM7XX_H #include "hw/boards.h" +#include "hw/adc/npcm7xx_adc.h" #include "hw/cpu/a9mpcore.h" #include "hw/gpio/npcm7xx_gpio.h" #include "hw/mem/npcm7xx_mc.h" @@ -XXX,XX +XXX,XX @@ typedef struct NPCM7xxState { NPCM7xxGCRState gcr; NPCM7xxCLKState clk; NPCM7xxTimerCtrlState tim[3]; + NPCM7xxADCState adc; NPCM7xxOTPState key_storage; NPCM7xxOTPState fuse_array; NPCM7xxMCState mc; diff --git a/hw/adc/npcm7xx_adc.c b/hw/adc/npcm7xx_adc.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/adc/npcm7xx_adc.c @@ -XXX,XX +XXX,XX @@ +/* + * Nuvoton NPCM7xx ADC Module + * + * Copyright 2020 Google LLC + * + * 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. + */ + +#include "qemu/osdep.h" +#include "hw/adc/npcm7xx_adc.h" +#include "hw/qdev-clock.h" +#include "hw/qdev-properties.h" +#include "hw/registerfields.h" +#include "migration/vmstate.h" +#include "qemu/log.h" +#include "qemu/module.h" +#include "qemu/timer.h" +#include "qemu/units.h" +#include "trace.h" + +REG32(NPCM7XX_ADC_CON, 0x0) +REG32(NPCM7XX_ADC_DATA, 0x4) + +/* Register field definitions. */ +#define NPCM7XX_ADC_CON_MUX(rv) extract32(rv, 24, 4) +#define NPCM7XX_ADC_CON_INT_EN BIT(21) +#define NPCM7XX_ADC_CON_REFSEL BIT(19) +#define NPCM7XX_ADC_CON_INT BIT(18) +#define NPCM7XX_ADC_CON_EN BIT(17) +#define NPCM7XX_ADC_CON_RST BIT(16) +#define NPCM7XX_ADC_CON_CONV BIT(14) +#define NPCM7XX_ADC_CON_DIV(rv) extract32(rv, 1, 8) + +#define NPCM7XX_ADC_MAX_RESULT 1023 +#define NPCM7XX_ADC_DEFAULT_IREF 2000000 +#define NPCM7XX_ADC_CONV_CYCLES 20 +#define NPCM7XX_ADC_RESET_CYCLES 10 +#define NPCM7XX_ADC_R0_INPUT 500000 +#define NPCM7XX_ADC_R1_INPUT 1500000 + +static void npcm7xx_adc_reset(NPCM7xxADCState *s) +{ + timer_del(&s->conv_timer); + s->con = 0x000c0001; + s->data = 0x00000000; +} + +static uint32_t npcm7xx_adc_convert(uint32_t input, uint32_t ref) +{ + uint32_t result; + + result = input * (NPCM7XX_ADC_MAX_RESULT + 1) / ref; + if (result > NPCM7XX_ADC_MAX_RESULT) { + result = NPCM7XX_ADC_MAX_RESULT; + } + + return result; +} + +static uint32_t npcm7xx_adc_prescaler(NPCM7xxADCState *s) +{ + return 2 * (NPCM7XX_ADC_CON_DIV(s->con) + 1); +} + +static void npcm7xx_adc_start_timer(Clock *clk, QEMUTimer *timer, + uint32_t cycles, uint32_t prescaler) +{ + int64_t now = qemu_clock_get_ns(QEMU_CLOCK_VIRTUAL); + int64_t ticks = cycles; + int64_t ns; + + ticks *= prescaler; + ns = clock_ticks_to_ns(clk, ticks); + ns += now; + timer_mod(timer, ns); +} + +static void npcm7xx_adc_start_convert(NPCM7xxADCState *s) +{ + uint32_t prescaler = npcm7xx_adc_prescaler(s); + + npcm7xx_adc_start_timer(s->clock, &s->conv_timer, NPCM7XX_ADC_CONV_CYCLES, + prescaler); +} + +static void npcm7xx_adc_convert_done(void *opaque) +{ + NPCM7xxADCState *s = opaque; + uint32_t input = NPCM7XX_ADC_CON_MUX(s->con); + uint32_t ref = (s->con & NPCM7XX_ADC_CON_REFSEL) + ? s->iref : s->vref; + + if (input >= NPCM7XX_ADC_NUM_INPUTS) { + qemu_log_mask(LOG_GUEST_ERROR, "%s: invalid input: %u\n", + __func__, input); + return; + } + s->data = npcm7xx_adc_convert(s->adci[input], ref); + if (s->con & NPCM7XX_ADC_CON_INT_EN) { + s->con |= NPCM7XX_ADC_CON_INT; + qemu_irq_raise(s->irq); + } + s->con &= ~NPCM7XX_ADC_CON_CONV; +} + +static void npcm7xx_adc_calibrate(NPCM7xxADCState *adc) +{ + adc->calibration_r_values[0] = npcm7xx_adc_convert(NPCM7XX_ADC_R0_INPUT, + adc->iref); + adc->calibration_r_values[1] = npcm7xx_adc_convert(NPCM7XX_ADC_R1_INPUT, + adc->iref); +} + +static void npcm7xx_adc_write_con(NPCM7xxADCState *s, uint32_t new_con) +{ + uint32_t old_con = s->con; + + /* Write ADC_INT to 1 to clear it */ + if (new_con & NPCM7XX_ADC_CON_INT) { + new_con &= ~NPCM7XX_ADC_CON_INT; + qemu_irq_lower(s->irq); + } else if (old_con & NPCM7XX_ADC_CON_INT) { + new_con |= NPCM7XX_ADC_CON_INT; + } + + s->con = new_con; + + if (s->con & NPCM7XX_ADC_CON_RST) { + npcm7xx_adc_reset(s); + return; + } + + if ((s->con & NPCM7XX_ADC_CON_EN)) { + if (s->con & NPCM7XX_ADC_CON_CONV) { + if (!(old_con & NPCM7XX_ADC_CON_CONV)) { + npcm7xx_adc_start_convert(s); + } + } else { + timer_del(&s->conv_timer); + } + } +} + +static uint64_t npcm7xx_adc_read(void *opaque, hwaddr offset, unsigned size) +{ + uint64_t value = 0; + NPCM7xxADCState *s = opaque; + + switch (offset) { + case A_NPCM7XX_ADC_CON: + value = s->con; + break; + + case A_NPCM7XX_ADC_DATA: + value = s->data; + break; + + default: + qemu_log_mask(LOG_GUEST_ERROR, + "%s: invalid offset 0x%04" HWADDR_PRIx "\n", + __func__, offset); + break; + } + + trace_npcm7xx_adc_read(DEVICE(s)->canonical_path, offset, value); + return value; +} + +static void npcm7xx_adc_write(void *opaque, hwaddr offset, uint64_t v, + unsigned size) +{ + NPCM7xxADCState *s = opaque; + + trace_npcm7xx_adc_write(DEVICE(s)->canonical_path, offset, v); + switch (offset) { + case A_NPCM7XX_ADC_CON: + npcm7xx_adc_write_con(s, v); + break; + + case A_NPCM7XX_ADC_DATA: + qemu_log_mask(LOG_GUEST_ERROR, + "%s: register @ 0x%04" HWADDR_PRIx " is read-only\n", + __func__, offset); + break; + + default: + qemu_log_mask(LOG_GUEST_ERROR, + "%s: invalid offset 0x%04" HWADDR_PRIx "\n", + __func__, offset); + break; + } + +} + +static const struct MemoryRegionOps npcm7xx_adc_ops = { + .read = npcm7xx_adc_read, + .write = npcm7xx_adc_write, + .endianness = DEVICE_LITTLE_ENDIAN, + .valid = { + .min_access_size = 4, + .max_access_size = 4, + .unaligned = false, + }, +}; + +static void npcm7xx_adc_enter_reset(Object *obj, ResetType type) +{ + NPCM7xxADCState *s = NPCM7XX_ADC(obj); + + npcm7xx_adc_reset(s); +} + +static void npcm7xx_adc_hold_reset(Object *obj) +{ + NPCM7xxADCState *s = NPCM7XX_ADC(obj); + + qemu_irq_lower(s->irq); +} + +static void npcm7xx_adc_init(Object *obj) +{ + NPCM7xxADCState *s = NPCM7XX_ADC(obj); + SysBusDevice *sbd = SYS_BUS_DEVICE(obj); + int i; + + sysbus_init_irq(sbd, &s->irq); + + timer_init_ns(&s->conv_timer, QEMU_CLOCK_VIRTUAL, + npcm7xx_adc_convert_done, s); + memory_region_init_io(&s->iomem, obj, &npcm7xx_adc_ops, s, + TYPE_NPCM7XX_ADC, 4 * KiB); + sysbus_init_mmio(sbd, &s->iomem); + s->clock = qdev_init_clock_in(DEVICE(s), "clock", NULL, NULL); + + for (i = 0; i < NPCM7XX_ADC_NUM_INPUTS; ++i) { + object_property_add_uint32_ptr(obj, "adci[*]", + &s->adci[i], OBJ_PROP_FLAG_WRITE); + } + object_property_add_uint32_ptr(obj, "vref", + &s->vref, OBJ_PROP_FLAG_WRITE); + npcm7xx_adc_calibrate(s); +} + +static const VMStateDescription vmstate_npcm7xx_adc = { + .name = "npcm7xx-adc", + .version_id = 0, + .minimum_version_id = 0, + .fields = (VMStateField[]) { + VMSTATE_TIMER(conv_timer, NPCM7xxADCState), + VMSTATE_UINT32(con, NPCM7xxADCState), + VMSTATE_UINT32(data, NPCM7xxADCState), + VMSTATE_CLOCK(clock, NPCM7xxADCState), + VMSTATE_UINT32_ARRAY(adci, NPCM7xxADCState, NPCM7XX_ADC_NUM_INPUTS), + VMSTATE_UINT32(vref, NPCM7xxADCState), + VMSTATE_UINT32(iref, NPCM7xxADCState), + VMSTATE_UINT16_ARRAY(calibration_r_values, NPCM7xxADCState, + NPCM7XX_ADC_NUM_CALIB), + VMSTATE_END_OF_LIST(), + }, +}; + +static Property npcm7xx_timer_properties[] = { + DEFINE_PROP_UINT32("iref", NPCM7xxADCState, iref, NPCM7XX_ADC_DEFAULT_IREF), + DEFINE_PROP_END_OF_LIST(), +}; + +static void npcm7xx_adc_class_init(ObjectClass *klass, void *data) +{ + ResettableClass *rc = RESETTABLE_CLASS(klass); + DeviceClass *dc = DEVICE_CLASS(klass); + + dc->desc = "NPCM7xx ADC Module"; + dc->vmsd = &vmstate_npcm7xx_adc; + rc->phases.enter = npcm7xx_adc_enter_reset; + rc->phases.hold = npcm7xx_adc_hold_reset; + + device_class_set_props(dc, npcm7xx_timer_properties); +} + +static const TypeInfo npcm7xx_adc_info = { + .name = TYPE_NPCM7XX_ADC, + .parent = TYPE_SYS_BUS_DEVICE, + .instance_size = sizeof(NPCM7xxADCState), + .class_init = npcm7xx_adc_class_init, + .instance_init = npcm7xx_adc_init, +}; + +static void npcm7xx_adc_register_types(void) +{ + type_register_static(&npcm7xx_adc_info); +} + +type_init(npcm7xx_adc_register_types); diff --git a/hw/arm/npcm7xx.c b/hw/arm/npcm7xx.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/npcm7xx.c +++ b/hw/arm/npcm7xx.c @@ -XXX,XX +XXX,XX @@ #define NPCM7XX_EHCI_BA (0xf0806000) #define NPCM7XX_OHCI_BA (0xf0807000) +/* ADC Module */ +#define NPCM7XX_ADC_BA (0xf000c000) + /* Internal AHB SRAM */ #define NPCM7XX_RAM3_BA (0xc0008000) #define NPCM7XX_RAM3_SZ (4 * KiB) @@ -XXX,XX +XXX,XX @@ #define NPCM7XX_ROM_BA (0xffff0000) #define NPCM7XX_ROM_SZ (64 * KiB) + /* Clock configuration values to be fixed up when bypassing bootloader */ /* Run PLL1 at 1600 MHz */ @@ -XXX,XX +XXX,XX @@ * interrupts. */ enum NPCM7xxInterrupt { + NPCM7XX_ADC_IRQ = 0, NPCM7XX_UART0_IRQ = 2, NPCM7XX_UART1_IRQ, NPCM7XX_UART2_IRQ, @@ -XXX,XX +XXX,XX @@ static void npcm7xx_init_fuses(NPCM7xxState *s) sizeof(value)); } +static void npcm7xx_write_adc_calibration(NPCM7xxState *s) +{ + /* Both ADC and the fuse array must have realized. */ + QEMU_BUILD_BUG_ON(sizeof(s->adc.calibration_r_values) != 4); + npcm7xx_otp_array_write(&s->fuse_array, s->adc.calibration_r_values, + NPCM7XX_FUSE_ADC_CALIB, sizeof(s->adc.calibration_r_values)); +} + static qemu_irq npcm7xx_irq(NPCM7xxState *s, int n) { return qdev_get_gpio_in(DEVICE(&s->a9mpcore), n); @@ -XXX,XX +XXX,XX @@ static void npcm7xx_init(Object *obj) TYPE_NPCM7XX_FUSE_ARRAY); object_initialize_child(obj, "mc", &s->mc, TYPE_NPCM7XX_MC); object_initialize_child(obj, "rng", &s->rng, TYPE_NPCM7XX_RNG); + object_initialize_child(obj, "adc", &s->adc, TYPE_NPCM7XX_ADC); for (i = 0; i < ARRAY_SIZE(s->tim); i++) { object_initialize_child(obj, "tim[*]", &s->tim[i], TYPE_NPCM7XX_TIMER); @@ -XXX,XX +XXX,XX @@ static void npcm7xx_realize(DeviceState *dev, Error **errp) sysbus_realize(SYS_BUS_DEVICE(&s->mc), &error_abort); sysbus_mmio_map(SYS_BUS_DEVICE(&s->mc), 0, NPCM7XX_MC_BA); + /* ADC Modules. Cannot fail. */ + qdev_connect_clock_in(DEVICE(&s->adc), "clock", qdev_get_clock_out( + DEVICE(&s->clk), "adc-clock")); + sysbus_realize(SYS_BUS_DEVICE(&s->adc), &error_abort); + sysbus_mmio_map(SYS_BUS_DEVICE(&s->adc), 0, NPCM7XX_ADC_BA); + sysbus_connect_irq(SYS_BUS_DEVICE(&s->adc), 0, + npcm7xx_irq(s, NPCM7XX_ADC_IRQ)); + npcm7xx_write_adc_calibration(s); + /* Timer Modules (TIM). Cannot fail. */ QEMU_BUILD_BUG_ON(ARRAY_SIZE(npcm7xx_tim_addr) != ARRAY_SIZE(s->tim)); for (i = 0; i < ARRAY_SIZE(s->tim); i++) { @@ -XXX,XX +XXX,XX @@ static void npcm7xx_realize(DeviceState *dev, Error **errp) create_unimplemented_device("npcm7xx.vdmx", 0xe0800000, 4 * KiB); create_unimplemented_device("npcm7xx.pcierc", 0xe1000000, 64 * KiB); create_unimplemented_device("npcm7xx.kcs", 0xf0007000, 4 * KiB); - create_unimplemented_device("npcm7xx.adc", 0xf000c000, 4 * KiB); create_unimplemented_device("npcm7xx.gfxi", 0xf000e000, 4 * KiB); create_unimplemented_device("npcm7xx.gpio[0]", 0xf0010000, 4 * KiB); create_unimplemented_device("npcm7xx.gpio[1]", 0xf0011000, 4 * KiB); diff --git a/tests/qtest/npcm7xx_adc-test.c b/tests/qtest/npcm7xx_adc-test.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/tests/qtest/npcm7xx_adc-test.c @@ -XXX,XX +XXX,XX @@ +/* + * QTests for Nuvoton NPCM7xx ADCModules. + * + * Copyright 2020 Google LLC + * + * 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. + */ + +#include "qemu/osdep.h" +#include "qemu/bitops.h" +#include "qemu/timer.h" +#include "libqos/libqtest.h" +#include "qapi/qmp/qdict.h" + +#define REF_HZ (25000000) + +#define CON_OFFSET 0x0 +#define DATA_OFFSET 0x4 + +#define NUM_INPUTS 8 +#define DEFAULT_IREF 2000000 +#define CONV_CYCLES 20 +#define RESET_CYCLES 10 +#define R0_INPUT 500000 +#define R1_INPUT 1500000 +#define MAX_RESULT 1023 + +#define DEFAULT_CLKDIV 5 + +#define FUSE_ARRAY_BA 0xf018a000 +#define FCTL_OFFSET 0x14 +#define FST_OFFSET 0x0 +#define FADDR_OFFSET 0x4 +#define FDATA_OFFSET 0x8 +#define ADC_CALIB_ADDR 24 +#define FUSE_READ 0x2 + +/* Register field definitions. */ +#define CON_MUX(rv) ((rv) << 24) +#define CON_INT_EN BIT(21) +#define CON_REFSEL BIT(19) +#define CON_INT BIT(18) +#define CON_EN BIT(17) +#define CON_RST BIT(16) +#define CON_CONV BIT(14) +#define CON_DIV(rv) extract32(rv, 1, 8) + +#define FST_RDST BIT(1) +#define FDATA_MASK 0xff + +#define MAX_ERROR 10000 +#define MIN_CALIB_INPUT 100000 +#define MAX_CALIB_INPUT 1800000 + +static const uint32_t input_list[] = { + 100000, + 500000, + 1000000, + 1500000, + 1800000, + 2000000, +}; + +static const uint32_t vref_list[] = { + 2000000, + 2200000, + 2500000, +}; + +static const uint32_t iref_list[] = { + 1800000, + 1900000, + 2000000, + 2100000, + 2200000, +}; + +static const uint32_t div_list[] = {0, 1, 3, 7, 15}; + +typedef struct ADC { + int irq; + uint64_t base_addr; +} ADC; + +ADC adc = { + .irq = 0, + .base_addr = 0xf000c000 +}; + +static uint32_t adc_read_con(QTestState *qts, const ADC *adc) +{ + return qtest_readl(qts, adc->base_addr + CON_OFFSET); +} + +static void adc_write_con(QTestState *qts, const ADC *adc, uint32_t value) +{ + qtest_writel(qts, adc->base_addr + CON_OFFSET, value); +} + +static uint32_t adc_read_data(QTestState *qts, const ADC *adc) +{ + return qtest_readl(qts, adc->base_addr + DATA_OFFSET); +} + +static uint32_t adc_calibrate(uint32_t measured, uint32_t *rv) +{ + return R0_INPUT + (R1_INPUT - R0_INPUT) * (int32_t)(measured - rv[0]) + / (int32_t)(rv[1] - rv[0]); +} + +static void adc_qom_set(QTestState *qts, const ADC *adc, + const char *name, uint32_t value) +{ + QDict *response; + const char *path = "/machine/soc/adc"; + + g_test_message("Setting properties %s of %s with value %u", + name, path, value); + response = qtest_qmp(qts, "{ 'execute': 'qom-set'," + " 'arguments': { 'path': %s, 'property': %s, 'value': %u}}", + path, name, value); + /* The qom set message returns successfully. */ + g_assert_true(qdict_haskey(response, "return")); +} + +static void adc_write_input(QTestState *qts, const ADC *adc, + uint32_t index, uint32_t value) +{ + char name[100]; + + sprintf(name, "adci[%u]", index); + adc_qom_set(qts, adc, name, value); +} + +static void adc_write_vref(QTestState *qts, const ADC *adc, uint32_t value) +{ + adc_qom_set(qts, adc, "vref", value); +} + +static uint32_t adc_calculate_output(uint32_t input, uint32_t ref) +{ + uint32_t output; + + g_assert_cmpuint(input, <=, ref); + output = (input * (MAX_RESULT + 1)) / ref; + if (output > MAX_RESULT) { + output = MAX_RESULT; + } + + return output; +} + +static uint32_t adc_prescaler(QTestState *qts, const ADC *adc) +{ + uint32_t div = extract32(adc_read_con(qts, adc), 1, 8); + + return 2 * (div + 1); +} + +static int64_t adc_calculate_steps(uint32_t cycles, uint32_t prescale, + uint32_t clkdiv) +{ + return (NANOSECONDS_PER_SECOND / (REF_HZ >> clkdiv)) * cycles * prescale; +} + +static void adc_wait_conv_finished(QTestState *qts, const ADC *adc, + uint32_t clkdiv) +{ + uint32_t prescaler = adc_prescaler(qts, adc); + + /* + * ADC should takes roughly 20 cycles to convert one sample. So we assert it + * should take 10~30 cycles here. + */ + qtest_clock_step(qts, adc_calculate_steps(CONV_CYCLES / 2, prescaler, + clkdiv)); + /* ADC is still converting. */ + g_assert_true(adc_read_con(qts, adc) & CON_CONV); + qtest_clock_step(qts, adc_calculate_steps(CONV_CYCLES, prescaler, clkdiv)); + /* ADC has finished conversion. */ + g_assert_false(adc_read_con(qts, adc) & CON_CONV); +} + +/* Check ADC can be reset to default value. */ +static void test_init(gconstpointer adc_p) +{ + const ADC *adc = adc_p; + + QTestState *qts = qtest_init("-machine quanta-gsj"); + adc_write_con(qts, adc, CON_REFSEL | CON_INT); + g_assert_cmphex(adc_read_con(qts, adc), ==, CON_REFSEL); + qtest_quit(qts); +} + +/* Check ADC can convert from an internal reference. */ +static void test_convert_internal(gconstpointer adc_p) +{ + const ADC *adc = adc_p; + uint32_t index, input, output, expected_output; + QTestState *qts = qtest_init("-machine quanta-gsj"); + qtest_irq_intercept_in(qts, "/machine/soc/a9mpcore/gic"); + + for (index = 0; index < NUM_INPUTS; ++index) { + for (size_t i = 0; i < ARRAY_SIZE(input_list); ++i) { + input = input_list[i]; + expected_output = adc_calculate_output(input, DEFAULT_IREF); + + adc_write_input(qts, adc, index, input); + adc_write_con(qts, adc, CON_MUX(index) | CON_REFSEL | CON_INT | + CON_EN | CON_CONV); + adc_wait_conv_finished(qts, adc, DEFAULT_CLKDIV); + g_assert_cmphex(adc_read_con(qts, adc), ==, CON_MUX(index) | + CON_REFSEL | CON_EN); + g_assert_false(qtest_get_irq(qts, adc->irq)); + output = adc_read_data(qts, adc); + g_assert_cmpuint(output, ==, expected_output); + } + } + + qtest_quit(qts); +} + +/* Check ADC can convert from an external reference. */ +static void test_convert_external(gconstpointer adc_p) +{ + const ADC *adc = adc_p; + uint32_t index, input, vref, output, expected_output; + QTestState *qts = qtest_init("-machine quanta-gsj"); + qtest_irq_intercept_in(qts, "/machine/soc/a9mpcore/gic"); + + for (index = 0; index < NUM_INPUTS; ++index) { + for (size_t i = 0; i < ARRAY_SIZE(input_list); ++i) { + for (size_t j = 0; j < ARRAY_SIZE(vref_list); ++j) { + input = input_list[i]; + vref = vref_list[j]; + expected_output = adc_calculate_output(input, vref); + + adc_write_input(qts, adc, index, input); + adc_write_vref(qts, adc, vref); + adc_write_con(qts, adc, CON_MUX(index) | CON_INT | CON_EN | + CON_CONV); + adc_wait_conv_finished(qts, adc, DEFAULT_CLKDIV); + g_assert_cmphex(adc_read_con(qts, adc), ==, + CON_MUX(index) | CON_EN); + g_assert_false(qtest_get_irq(qts, adc->irq)); + output = adc_read_data(qts, adc); + g_assert_cmpuint(output, ==, expected_output); + } + } + } + + qtest_quit(qts); +} + +/* Check ADC interrupt files if and only if CON_INT_EN is set. */ +static void test_interrupt(gconstpointer adc_p) +{ + const ADC *adc = adc_p; + uint32_t index, input, output, expected_output; + QTestState *qts = qtest_init("-machine quanta-gsj"); + + index = 1; + input = input_list[1]; + expected_output = adc_calculate_output(input, DEFAULT_IREF); + + qtest_irq_intercept_in(qts, "/machine/soc/a9mpcore/gic"); + adc_write_input(qts, adc, index, input); + g_assert_false(qtest_get_irq(qts, adc->irq)); + adc_write_con(qts, adc, CON_MUX(index) | CON_INT_EN | CON_REFSEL | CON_INT + | CON_EN | CON_CONV); + adc_wait_conv_finished(qts, adc, DEFAULT_CLKDIV); + g_assert_cmphex(adc_read_con(qts, adc), ==, CON_MUX(index) | CON_INT_EN + | CON_REFSEL | CON_INT | CON_EN); + g_assert_true(qtest_get_irq(qts, adc->irq)); + output = adc_read_data(qts, adc); + g_assert_cmpuint(output, ==, expected_output); + + qtest_quit(qts); +} + +/* Check ADC is reset after setting ADC_RST for 10 ADC cycles. */ +static void test_reset(gconstpointer adc_p) +{ + const ADC *adc = adc_p; + QTestState *qts = qtest_init("-machine quanta-gsj"); + + for (size_t i = 0; i < ARRAY_SIZE(div_list); ++i) { + uint32_t div = div_list[i]; + + adc_write_con(qts, adc, CON_INT | CON_EN | CON_RST | CON_DIV(div)); + qtest_clock_step(qts, adc_calculate_steps(RESET_CYCLES, + adc_prescaler(qts, adc), DEFAULT_CLKDIV)); + g_assert_false(adc_read_con(qts, adc) & CON_EN); + } + qtest_quit(qts); +} + +/* Check ADC Calibration works as desired. */ +static void test_calibrate(gconstpointer adc_p) +{ + int i, j; + const ADC *adc = adc_p; + + for (j = 0; j < ARRAY_SIZE(iref_list); ++j) { + uint32_t iref = iref_list[j]; + uint32_t expected_rv[] = { + adc_calculate_output(R0_INPUT, iref), + adc_calculate_output(R1_INPUT, iref), + }; + char buf[100]; + QTestState *qts; + + sprintf(buf, "-machine quanta-gsj -global npcm7xx-adc.iref=%u", iref); + qts = qtest_init(buf); + + /* Check the converted value is correct using the calibration value. */ + for (i = 0; i < ARRAY_SIZE(input_list); ++i) { + uint32_t input; + uint32_t output; + uint32_t expected_output; + uint32_t calibrated_voltage; + uint32_t index = 0; + + input = input_list[i]; + /* Calibration only works for input range 0.1V ~ 1.8V. */ + if (input < MIN_CALIB_INPUT || input > MAX_CALIB_INPUT) { + continue; + } + expected_output = adc_calculate_output(input, iref); + + adc_write_input(qts, adc, index, input); + adc_write_con(qts, adc, CON_MUX(index) | CON_REFSEL | CON_INT | + CON_EN | CON_CONV); + adc_wait_conv_finished(qts, adc, DEFAULT_CLKDIV); + g_assert_cmphex(adc_read_con(qts, adc), ==, + CON_REFSEL | CON_MUX(index) | CON_EN); + output = adc_read_data(qts, adc); + g_assert_cmpuint(output, ==, expected_output); + + calibrated_voltage = adc_calibrate(output, expected_rv); + g_assert_cmpuint(calibrated_voltage, >, input - MAX_ERROR); + g_assert_cmpuint(calibrated_voltage, <, input + MAX_ERROR); + } + + qtest_quit(qts); + } +} + +static void adc_add_test(const char *name, const ADC* wd, + GTestDataFunc fn) +{ + g_autofree char *full_name = g_strdup_printf("npcm7xx_adc/%s", name); + qtest_add_data_func(full_name, wd, fn); +} +#define add_test(name, td) adc_add_test(#name, td, test_##name) + +int main(int argc, char **argv) +{ + g_test_init(&argc, &argv, NULL); + + add_test(init, &adc); + add_test(convert_internal, &adc); + add_test(convert_external, &adc); + add_test(interrupt, &adc); + add_test(reset, &adc); + add_test(calibrate, &adc); + + return g_test_run(); +} diff --git a/hw/adc/meson.build b/hw/adc/meson.build index XXXXXXX..XXXXXXX 100644 --- a/hw/adc/meson.build +++ b/hw/adc/meson.build @@ -1 +1,2 @@ softmmu_ss.add(when: 'CONFIG_STM32F2XX_ADC', if_true: files('stm32f2xx_adc.c')) +softmmu_ss.add(when: 'CONFIG_NPCM7XX', if_true: files('npcm7xx_adc.c')) diff --git a/hw/adc/trace-events b/hw/adc/trace-events new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/adc/trace-events @@ -XXX,XX +XXX,XX @@ +# See docs/devel/tracing.txt for syntax documentation. + +# npcm7xx_adc.c +npcm7xx_adc_read(const char *id, uint64_t offset, uint32_t value) " %s offset: 0x%04" PRIx64 " value 0x%04" PRIx32 +npcm7xx_adc_write(const char *id, uint64_t offset, uint32_t value) "%s offset: 0x%04" PRIx64 " value 0x%04" PRIx32 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_sparc64 = \ ['prom-env-test', 'boot-serial-test'] qtests_npcm7xx = \ - ['npcm7xx_gpio-test', + ['npcm7xx_adc-test', + 'npcm7xx_gpio-test', 'npcm7xx_rng-test', 'npcm7xx_timer-test', 'npcm7xx_watchdog_timer-test'] -- 2.20.1 From: Hao Wu <wuhaotsh@google.com> The PWM module is part of NPCM7XX module. Each NPCM7XX module has two identical PWM modules. Each module contains 4 PWM entries. Each PWM has two outputs: frequency and duty_cycle. Both are computed using inputs from software side. This module does not model detail pulse signals since it is expensive. It also does not model interrupts and watchdogs that are dependant on the detail models. The interfaces for these are left in the module so that anyone in need for these functionalities can implement on their own. The user can read the duty cycle and frequency using qom-get command. Reviewed-by: Havard Skinnemoen <hskinnemoen@google.com> Reviewed-by: Tyrone Ting <kfting@nuvoton.com> Signed-off-by: Hao Wu <wuhaotsh@google.com> Message-id: 20210108190945.949196-5-wuhaotsh@google.com Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- docs/system/arm/nuvoton.rst | 2 +- include/hw/arm/npcm7xx.h | 2 + include/hw/misc/npcm7xx_pwm.h | 105 +++++++ hw/arm/npcm7xx.c | 26 +- hw/misc/npcm7xx_pwm.c | 550 ++++++++++++++++++++++++++++++++++ hw/misc/meson.build | 1 + hw/misc/trace-events | 6 + 7 files changed, 689 insertions(+), 3 deletions(-) create mode 100644 include/hw/misc/npcm7xx_pwm.h create mode 100644 hw/misc/npcm7xx_pwm.c diff --git a/docs/system/arm/nuvoton.rst b/docs/system/arm/nuvoton.rst index XXXXXXX..XXXXXXX 100644 --- a/docs/system/arm/nuvoton.rst +++ b/docs/system/arm/nuvoton.rst @@ -XXX,XX +XXX,XX @@ Supported devices * USB host (USBH) * GPIO controller * Analog to Digital Converter (ADC) + * Pulse Width Modulation (PWM) Missing devices --------------- @@ -XXX,XX +XXX,XX @@ Missing devices * Peripheral SPI controller (PSPI) * SD/MMC host * PECI interface - * Pulse Width Modulation (PWM) * Tachometer * PCI and PCIe root complex and bridges * VDM and MCTP support diff --git a/include/hw/arm/npcm7xx.h b/include/hw/arm/npcm7xx.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/npcm7xx.h +++ b/include/hw/arm/npcm7xx.h @@ -XXX,XX +XXX,XX @@ #include "hw/mem/npcm7xx_mc.h" #include "hw/misc/npcm7xx_clk.h" #include "hw/misc/npcm7xx_gcr.h" +#include "hw/misc/npcm7xx_pwm.h" #include "hw/misc/npcm7xx_rng.h" #include "hw/nvram/npcm7xx_otp.h" #include "hw/timer/npcm7xx_timer.h" @@ -XXX,XX +XXX,XX @@ typedef struct NPCM7xxState { NPCM7xxCLKState clk; NPCM7xxTimerCtrlState tim[3]; NPCM7xxADCState adc; + NPCM7xxPWMState pwm[2]; NPCM7xxOTPState key_storage; NPCM7xxOTPState fuse_array; NPCM7xxMCState mc; diff --git a/include/hw/misc/npcm7xx_pwm.h b/include/hw/misc/npcm7xx_pwm.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/include/hw/misc/npcm7xx_pwm.h @@ -XXX,XX +XXX,XX @@ +/* + * Nuvoton NPCM7xx PWM Module + * + * Copyright 2020 Google LLC + * + * 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. + */ +#ifndef NPCM7XX_PWM_H +#define NPCM7XX_PWM_H + +#include "hw/clock.h" +#include "hw/sysbus.h" +#include "hw/irq.h" + +/* Each PWM module holds 4 PWM channels. */ +#define NPCM7XX_PWM_PER_MODULE 4 + +/* + * Number of registers in one pwm module. Don't change this without increasing + * the version_id in vmstate. + */ +#define NPCM7XX_PWM_NR_REGS (0x54 / sizeof(uint32_t)) + +/* + * The maximum duty values. Each duty unit represents 1/NPCM7XX_PWM_MAX_DUTY + * cycles. For example, if NPCM7XX_PWM_MAX_DUTY=1,000,000 and a PWM has a duty + * value of 100,000 the duty cycle for that PWM is 10%. + */ +#define NPCM7XX_PWM_MAX_DUTY 1000000 + +typedef struct NPCM7xxPWMState NPCM7xxPWMState; + +/** + * struct NPCM7xxPWM - The state of a single PWM channel. + * @module: The PWM module that contains this channel. + * @irq: GIC interrupt line to fire on expiration if enabled. + * @running: Whether this PWM channel is generating output. + * @inverted: Whether this PWM channel is inverted. + * @index: The index of this PWM channel. + * @cnr: The counter register. + * @cmr: The comparator register. + * @pdr: The data register. + * @pwdr: The watchdog register. + * @freq: The frequency of this PWM channel. + * @duty: The duty cycle of this PWM channel. One unit represents + * 1/NPCM7XX_MAX_DUTY cycles. + */ +typedef struct NPCM7xxPWM { + NPCM7xxPWMState *module; + + qemu_irq irq; + + bool running; + bool inverted; + + uint8_t index; + uint32_t cnr; + uint32_t cmr; + uint32_t pdr; + uint32_t pwdr; + + uint32_t freq; + uint32_t duty; +} NPCM7xxPWM; + +/** + * struct NPCM7xxPWMState - Pulse Width Modulation device state. + * @parent: System bus device. + * @iomem: Memory region through which registers are accessed. + * @clock: The PWM clock. + * @pwm: The PWM channels owned by this module. + * @ppr: The prescaler register. + * @csr: The clock selector register. + * @pcr: The control register. + * @pier: The interrupt enable register. + * @piir: The interrupt indication register. + */ +struct NPCM7xxPWMState { + SysBusDevice parent; + + MemoryRegion iomem; + + Clock *clock; + NPCM7xxPWM pwm[NPCM7XX_PWM_PER_MODULE]; + + uint32_t ppr; + uint32_t csr; + uint32_t pcr; + uint32_t pier; + uint32_t piir; +}; + +#define TYPE_NPCM7XX_PWM "npcm7xx-pwm" +#define NPCM7XX_PWM(obj) \ + OBJECT_CHECK(NPCM7xxPWMState, (obj), TYPE_NPCM7XX_PWM) + +#endif /* NPCM7XX_PWM_H */ diff --git a/hw/arm/npcm7xx.c b/hw/arm/npcm7xx.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/npcm7xx.c +++ b/hw/arm/npcm7xx.c @@ -XXX,XX +XXX,XX @@ enum NPCM7xxInterrupt { NPCM7XX_WDG2_IRQ, /* Timer Module 2 Watchdog */ NPCM7XX_EHCI_IRQ = 61, NPCM7XX_OHCI_IRQ = 62, + NPCM7XX_PWM0_IRQ = 93, /* PWM module 0 */ + NPCM7XX_PWM1_IRQ, /* PWM module 1 */ NPCM7XX_GPIO0_IRQ = 116, NPCM7XX_GPIO1_IRQ, NPCM7XX_GPIO2_IRQ, @@ -XXX,XX +XXX,XX @@ static const hwaddr npcm7xx_fiu3_flash_addr[] = { 0xb8000000, /* CS3 */ }; +/* Register base address for each PWM Module */ +static const hwaddr npcm7xx_pwm_addr[] = { + 0xf0103000, + 0xf0104000, +}; + static const struct { hwaddr regs_addr; uint32_t unconnected_pins; @@ -XXX,XX +XXX,XX @@ static void npcm7xx_init(Object *obj) object_initialize_child(obj, npcm7xx_fiu[i].name, &s->fiu[i], TYPE_NPCM7XX_FIU); } + + for (i = 0; i < ARRAY_SIZE(s->pwm); i++) { + object_initialize_child(obj, "pwm[*]", &s->pwm[i], TYPE_NPCM7XX_PWM); + } } static void npcm7xx_realize(DeviceState *dev, Error **errp) @@ -XXX,XX +XXX,XX @@ static void npcm7xx_realize(DeviceState *dev, Error **errp) sysbus_connect_irq(SYS_BUS_DEVICE(&s->ohci), 0, npcm7xx_irq(s, NPCM7XX_OHCI_IRQ)); + /* PWM Modules. Cannot fail. */ + QEMU_BUILD_BUG_ON(ARRAY_SIZE(npcm7xx_pwm_addr) != ARRAY_SIZE(s->pwm)); + for (i = 0; i < ARRAY_SIZE(s->pwm); i++) { + SysBusDevice *sbd = SYS_BUS_DEVICE(&s->pwm[i]); + + qdev_connect_clock_in(DEVICE(&s->pwm[i]), "clock", qdev_get_clock_out( + DEVICE(&s->clk), "apb3-clock")); + sysbus_realize(sbd, &error_abort); + sysbus_mmio_map(sbd, 0, npcm7xx_pwm_addr[i]); + sysbus_connect_irq(sbd, i, npcm7xx_irq(s, NPCM7XX_PWM0_IRQ + i)); + } + /* * Flash Interface Unit (FIU). Can fail if incorrect number of chip selects * specified, but this is a programming error. @@ -XXX,XX +XXX,XX @@ static void npcm7xx_realize(DeviceState *dev, Error **errp) create_unimplemented_device("npcm7xx.peci", 0xf0100000, 4 * KiB); create_unimplemented_device("npcm7xx.siox[1]", 0xf0101000, 4 * KiB); create_unimplemented_device("npcm7xx.siox[2]", 0xf0102000, 4 * KiB); - create_unimplemented_device("npcm7xx.pwm[0]", 0xf0103000, 4 * KiB); - create_unimplemented_device("npcm7xx.pwm[1]", 0xf0104000, 4 * KiB); create_unimplemented_device("npcm7xx.mft[0]", 0xf0180000, 4 * KiB); create_unimplemented_device("npcm7xx.mft[1]", 0xf0181000, 4 * KiB); create_unimplemented_device("npcm7xx.mft[2]", 0xf0182000, 4 * KiB); diff --git a/hw/misc/npcm7xx_pwm.c b/hw/misc/npcm7xx_pwm.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/misc/npcm7xx_pwm.c @@ -XXX,XX +XXX,XX @@ +/* + * Nuvoton NPCM7xx PWM Module + * + * Copyright 2020 Google LLC + * + * 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. + */ + +#include "qemu/osdep.h" +#include "hw/irq.h" +#include "hw/qdev-clock.h" +#include "hw/qdev-properties.h" +#include "hw/misc/npcm7xx_pwm.h" +#include "hw/registerfields.h" +#include "migration/vmstate.h" +#include "qemu/bitops.h" +#include "qemu/error-report.h" +#include "qemu/log.h" +#include "qemu/module.h" +#include "qemu/units.h" +#include "trace.h" + +REG32(NPCM7XX_PWM_PPR, 0x00); +REG32(NPCM7XX_PWM_CSR, 0x04); +REG32(NPCM7XX_PWM_PCR, 0x08); +REG32(NPCM7XX_PWM_CNR0, 0x0c); +REG32(NPCM7XX_PWM_CMR0, 0x10); +REG32(NPCM7XX_PWM_PDR0, 0x14); +REG32(NPCM7XX_PWM_CNR1, 0x18); +REG32(NPCM7XX_PWM_CMR1, 0x1c); +REG32(NPCM7XX_PWM_PDR1, 0x20); +REG32(NPCM7XX_PWM_CNR2, 0x24); +REG32(NPCM7XX_PWM_CMR2, 0x28); +REG32(NPCM7XX_PWM_PDR2, 0x2c); +REG32(NPCM7XX_PWM_CNR3, 0x30); +REG32(NPCM7XX_PWM_CMR3, 0x34); +REG32(NPCM7XX_PWM_PDR3, 0x38); +REG32(NPCM7XX_PWM_PIER, 0x3c); +REG32(NPCM7XX_PWM_PIIR, 0x40); +REG32(NPCM7XX_PWM_PWDR0, 0x44); +REG32(NPCM7XX_PWM_PWDR1, 0x48); +REG32(NPCM7XX_PWM_PWDR2, 0x4c); +REG32(NPCM7XX_PWM_PWDR3, 0x50); + +/* Register field definitions. */ +#define NPCM7XX_PPR(rv, index) extract32((rv), npcm7xx_ppr_base[index], 8) +#define NPCM7XX_CSR(rv, index) extract32((rv), npcm7xx_csr_base[index], 3) +#define NPCM7XX_CH(rv, index) extract32((rv), npcm7xx_ch_base[index], 4) +#define NPCM7XX_CH_EN BIT(0) +#define NPCM7XX_CH_INV BIT(2) +#define NPCM7XX_CH_MOD BIT(3) + +/* Offset of each PWM channel's prescaler in the PPR register. */ +static const int npcm7xx_ppr_base[] = { 0, 0, 8, 8 }; +/* Offset of each PWM channel's clock selector in the CSR register. */ +static const int npcm7xx_csr_base[] = { 0, 4, 8, 12 }; +/* Offset of each PWM channel's control variable in the PCR register. */ +static const int npcm7xx_ch_base[] = { 0, 8, 12, 16 }; + +static uint32_t npcm7xx_pwm_calculate_freq(NPCM7xxPWM *p) +{ + uint32_t ppr; + uint32_t csr; + uint32_t freq; + + if (!p->running) { + return 0; + } + + csr = NPCM7XX_CSR(p->module->csr, p->index); + ppr = NPCM7XX_PPR(p->module->ppr, p->index); + freq = clock_get_hz(p->module->clock); + freq /= ppr + 1; + /* csr can only be 0~4 */ + if (csr > 4) { + qemu_log_mask(LOG_GUEST_ERROR, + "%s: invalid csr value %u\n", + __func__, csr); + csr = 4; + } + /* freq won't be changed if csr == 4. */ + if (csr < 4) { + freq >>= csr + 1; + } + + return freq / (p->cnr + 1); +} + +static uint32_t npcm7xx_pwm_calculate_duty(NPCM7xxPWM *p) +{ + uint64_t duty; + + if (p->running) { + if (p->cnr == 0) { + duty = 0; + } else if (p->cmr >= p->cnr) { + duty = NPCM7XX_PWM_MAX_DUTY; + } else { + duty = NPCM7XX_PWM_MAX_DUTY * (p->cmr + 1) / (p->cnr + 1); + } + } else { + duty = 0; + } + + if (p->inverted) { + duty = NPCM7XX_PWM_MAX_DUTY - duty; + } + + return duty; +} + +static void npcm7xx_pwm_update_freq(NPCM7xxPWM *p) +{ + uint32_t freq = npcm7xx_pwm_calculate_freq(p); + + if (freq != p->freq) { + trace_npcm7xx_pwm_update_freq(DEVICE(p->module)->canonical_path, + p->index, p->freq, freq); + p->freq = freq; + } +} + +static void npcm7xx_pwm_update_duty(NPCM7xxPWM *p) +{ + uint32_t duty = npcm7xx_pwm_calculate_duty(p); + + if (duty != p->duty) { + trace_npcm7xx_pwm_update_duty(DEVICE(p->module)->canonical_path, + p->index, p->duty, duty); + p->duty = duty; + } +} + +static void npcm7xx_pwm_update_output(NPCM7xxPWM *p) +{ + npcm7xx_pwm_update_freq(p); + npcm7xx_pwm_update_duty(p); +} + +static void npcm7xx_pwm_write_ppr(NPCM7xxPWMState *s, uint32_t new_ppr) +{ + int i; + uint32_t old_ppr = s->ppr; + + QEMU_BUILD_BUG_ON(ARRAY_SIZE(npcm7xx_ppr_base) != NPCM7XX_PWM_PER_MODULE); + s->ppr = new_ppr; + for (i = 0; i < NPCM7XX_PWM_PER_MODULE; ++i) { + if (NPCM7XX_PPR(old_ppr, i) != NPCM7XX_PPR(new_ppr, i)) { + npcm7xx_pwm_update_freq(&s->pwm[i]); + } + } +} + +static void npcm7xx_pwm_write_csr(NPCM7xxPWMState *s, uint32_t new_csr) +{ + int i; + uint32_t old_csr = s->csr; + + QEMU_BUILD_BUG_ON(ARRAY_SIZE(npcm7xx_csr_base) != NPCM7XX_PWM_PER_MODULE); + s->csr = new_csr; + for (i = 0; i < NPCM7XX_PWM_PER_MODULE; ++i) { + if (NPCM7XX_CSR(old_csr, i) != NPCM7XX_CSR(new_csr, i)) { + npcm7xx_pwm_update_freq(&s->pwm[i]); + } + } +} + +static void npcm7xx_pwm_write_pcr(NPCM7xxPWMState *s, uint32_t new_pcr) +{ + int i; + bool inverted; + uint32_t pcr; + NPCM7xxPWM *p; + + s->pcr = new_pcr; + QEMU_BUILD_BUG_ON(ARRAY_SIZE(npcm7xx_ch_base) != NPCM7XX_PWM_PER_MODULE); + for (i = 0; i < NPCM7XX_PWM_PER_MODULE; ++i) { + p = &s->pwm[i]; + pcr = NPCM7XX_CH(new_pcr, i); + inverted = pcr & NPCM7XX_CH_INV; + + /* + * We only run a PWM channel with toggle mode. Single-shot mode does not + * generate frequency and duty-cycle values. + */ + if ((pcr & NPCM7XX_CH_EN) && (pcr & NPCM7XX_CH_MOD)) { + if (p->running) { + /* Re-run this PWM channel if inverted changed. */ + if (p->inverted ^ inverted) { + p->inverted = inverted; + npcm7xx_pwm_update_duty(p); + } + } else { + /* Run this PWM channel. */ + p->running = true; + p->inverted = inverted; + npcm7xx_pwm_update_output(p); + } + } else { + /* Clear this PWM channel. */ + p->running = false; + p->inverted = inverted; + npcm7xx_pwm_update_output(p); + } + } + +} + +static hwaddr npcm7xx_cnr_index(hwaddr offset) +{ + switch (offset) { + case A_NPCM7XX_PWM_CNR0: + return 0; + case A_NPCM7XX_PWM_CNR1: + return 1; + case A_NPCM7XX_PWM_CNR2: + return 2; + case A_NPCM7XX_PWM_CNR3: + return 3; + default: + g_assert_not_reached(); + } +} + +static hwaddr npcm7xx_cmr_index(hwaddr offset) +{ + switch (offset) { + case A_NPCM7XX_PWM_CMR0: + return 0; + case A_NPCM7XX_PWM_CMR1: + return 1; + case A_NPCM7XX_PWM_CMR2: + return 2; + case A_NPCM7XX_PWM_CMR3: + return 3; + default: + g_assert_not_reached(); + } +} + +static hwaddr npcm7xx_pdr_index(hwaddr offset) +{ + switch (offset) { + case A_NPCM7XX_PWM_PDR0: + return 0; + case A_NPCM7XX_PWM_PDR1: + return 1; + case A_NPCM7XX_PWM_PDR2: + return 2; + case A_NPCM7XX_PWM_PDR3: + return 3; + default: + g_assert_not_reached(); + } +} + +static hwaddr npcm7xx_pwdr_index(hwaddr offset) +{ + switch (offset) { + case A_NPCM7XX_PWM_PWDR0: + return 0; + case A_NPCM7XX_PWM_PWDR1: + return 1; + case A_NPCM7XX_PWM_PWDR2: + return 2; + case A_NPCM7XX_PWM_PWDR3: + return 3; + default: + g_assert_not_reached(); + } +} + +static uint64_t npcm7xx_pwm_read(void *opaque, hwaddr offset, unsigned size) +{ + NPCM7xxPWMState *s = opaque; + uint64_t value = 0; + + switch (offset) { + case A_NPCM7XX_PWM_CNR0: + case A_NPCM7XX_PWM_CNR1: + case A_NPCM7XX_PWM_CNR2: + case A_NPCM7XX_PWM_CNR3: + value = s->pwm[npcm7xx_cnr_index(offset)].cnr; + break; + + case A_NPCM7XX_PWM_CMR0: + case A_NPCM7XX_PWM_CMR1: + case A_NPCM7XX_PWM_CMR2: + case A_NPCM7XX_PWM_CMR3: + value = s->pwm[npcm7xx_cmr_index(offset)].cmr; + break; + + case A_NPCM7XX_PWM_PDR0: + case A_NPCM7XX_PWM_PDR1: + case A_NPCM7XX_PWM_PDR2: + case A_NPCM7XX_PWM_PDR3: + value = s->pwm[npcm7xx_pdr_index(offset)].pdr; + break; + + case A_NPCM7XX_PWM_PWDR0: + case A_NPCM7XX_PWM_PWDR1: + case A_NPCM7XX_PWM_PWDR2: + case A_NPCM7XX_PWM_PWDR3: + value = s->pwm[npcm7xx_pwdr_index(offset)].pwdr; + break; + + case A_NPCM7XX_PWM_PPR: + value = s->ppr; + break; + + case A_NPCM7XX_PWM_CSR: + value = s->csr; + break; + + case A_NPCM7XX_PWM_PCR: + value = s->pcr; + break; + + case A_NPCM7XX_PWM_PIER: + value = s->pier; + break; + + case A_NPCM7XX_PWM_PIIR: + value = s->piir; + break; + + default: + qemu_log_mask(LOG_GUEST_ERROR, + "%s: invalid offset 0x%04" HWADDR_PRIx "\n", + __func__, offset); + break; + } + + trace_npcm7xx_pwm_read(DEVICE(s)->canonical_path, offset, value); + return value; +} + +static void npcm7xx_pwm_write(void *opaque, hwaddr offset, + uint64_t v, unsigned size) +{ + NPCM7xxPWMState *s = opaque; + NPCM7xxPWM *p; + uint32_t value = v; + + trace_npcm7xx_pwm_write(DEVICE(s)->canonical_path, offset, value); + switch (offset) { + case A_NPCM7XX_PWM_CNR0: + case A_NPCM7XX_PWM_CNR1: + case A_NPCM7XX_PWM_CNR2: + case A_NPCM7XX_PWM_CNR3: + p = &s->pwm[npcm7xx_cnr_index(offset)]; + p->cnr = value; + npcm7xx_pwm_update_output(p); + break; + + case A_NPCM7XX_PWM_CMR0: + case A_NPCM7XX_PWM_CMR1: + case A_NPCM7XX_PWM_CMR2: + case A_NPCM7XX_PWM_CMR3: + p = &s->pwm[npcm7xx_cmr_index(offset)]; + p->cmr = value; + npcm7xx_pwm_update_output(p); + break; + + case A_NPCM7XX_PWM_PDR0: + case A_NPCM7XX_PWM_PDR1: + case A_NPCM7XX_PWM_PDR2: + case A_NPCM7XX_PWM_PDR3: + qemu_log_mask(LOG_GUEST_ERROR, + "%s: register @ 0x%04" HWADDR_PRIx " is read-only\n", + __func__, offset); + break; + + case A_NPCM7XX_PWM_PWDR0: + case A_NPCM7XX_PWM_PWDR1: + case A_NPCM7XX_PWM_PWDR2: + case A_NPCM7XX_PWM_PWDR3: + qemu_log_mask(LOG_UNIMP, + "%s: register @ 0x%04" HWADDR_PRIx " is not implemented\n", + __func__, offset); + break; + + case A_NPCM7XX_PWM_PPR: + npcm7xx_pwm_write_ppr(s, value); + break; + + case A_NPCM7XX_PWM_CSR: + npcm7xx_pwm_write_csr(s, value); + break; + + case A_NPCM7XX_PWM_PCR: + npcm7xx_pwm_write_pcr(s, value); + break; + + case A_NPCM7XX_PWM_PIER: + qemu_log_mask(LOG_UNIMP, + "%s: register @ 0x%04" HWADDR_PRIx " is not implemented\n", + __func__, offset); + break; + + case A_NPCM7XX_PWM_PIIR: + qemu_log_mask(LOG_UNIMP, + "%s: register @ 0x%04" HWADDR_PRIx " is not implemented\n", + __func__, offset); + break; + + default: + qemu_log_mask(LOG_GUEST_ERROR, + "%s: invalid offset 0x%04" HWADDR_PRIx "\n", + __func__, offset); + break; + } +} + +static const struct MemoryRegionOps npcm7xx_pwm_ops = { + .read = npcm7xx_pwm_read, + .write = npcm7xx_pwm_write, + .endianness = DEVICE_LITTLE_ENDIAN, + .valid = { + .min_access_size = 4, + .max_access_size = 4, + .unaligned = false, + }, +}; + +static void npcm7xx_pwm_enter_reset(Object *obj, ResetType type) +{ + NPCM7xxPWMState *s = NPCM7XX_PWM(obj); + int i; + + for (i = 0; i < NPCM7XX_PWM_PER_MODULE; i++) { + NPCM7xxPWM *p = &s->pwm[i]; + + p->cnr = 0x00000000; + p->cmr = 0x00000000; + p->pdr = 0x00000000; + p->pwdr = 0x00000000; + } + + s->ppr = 0x00000000; + s->csr = 0x00000000; + s->pcr = 0x00000000; + s->pier = 0x00000000; + s->piir = 0x00000000; +} + +static void npcm7xx_pwm_hold_reset(Object *obj) +{ + NPCM7xxPWMState *s = NPCM7XX_PWM(obj); + int i; + + for (i = 0; i < NPCM7XX_PWM_PER_MODULE; i++) { + qemu_irq_lower(s->pwm[i].irq); + } +} + +static void npcm7xx_pwm_init(Object *obj) +{ + NPCM7xxPWMState *s = NPCM7XX_PWM(obj); + SysBusDevice *sbd = SYS_BUS_DEVICE(obj); + int i; + + for (i = 0; i < NPCM7XX_PWM_PER_MODULE; i++) { + NPCM7xxPWM *p = &s->pwm[i]; + p->module = s; + p->index = i; + sysbus_init_irq(sbd, &p->irq); + } + + memory_region_init_io(&s->iomem, obj, &npcm7xx_pwm_ops, s, + TYPE_NPCM7XX_PWM, 4 * KiB); + sysbus_init_mmio(sbd, &s->iomem); + s->clock = qdev_init_clock_in(DEVICE(s), "clock", NULL, NULL); + + for (i = 0; i < NPCM7XX_PWM_PER_MODULE; ++i) { + object_property_add_uint32_ptr(obj, "freq[*]", + &s->pwm[i].freq, OBJ_PROP_FLAG_READ); + object_property_add_uint32_ptr(obj, "duty[*]", + &s->pwm[i].duty, OBJ_PROP_FLAG_READ); + } +} + +static const VMStateDescription vmstate_npcm7xx_pwm = { + .name = "npcm7xx-pwm", + .version_id = 0, + .minimum_version_id = 0, + .fields = (VMStateField[]) { + VMSTATE_BOOL(running, NPCM7xxPWM), + VMSTATE_BOOL(inverted, NPCM7xxPWM), + VMSTATE_UINT8(index, NPCM7xxPWM), + VMSTATE_UINT32(cnr, NPCM7xxPWM), + VMSTATE_UINT32(cmr, NPCM7xxPWM), + VMSTATE_UINT32(pdr, NPCM7xxPWM), + VMSTATE_UINT32(pwdr, NPCM7xxPWM), + VMSTATE_UINT32(freq, NPCM7xxPWM), + VMSTATE_UINT32(duty, NPCM7xxPWM), + VMSTATE_END_OF_LIST(), + }, +}; + +static const VMStateDescription vmstate_npcm7xx_pwm_module = { + .name = "npcm7xx-pwm-module", + .version_id = 0, + .minimum_version_id = 0, + .fields = (VMStateField[]) { + VMSTATE_CLOCK(clock, NPCM7xxPWMState), + VMSTATE_STRUCT_ARRAY(pwm, NPCM7xxPWMState, + NPCM7XX_PWM_PER_MODULE, 0, vmstate_npcm7xx_pwm, + NPCM7xxPWM), + VMSTATE_UINT32(ppr, NPCM7xxPWMState), + VMSTATE_UINT32(csr, NPCM7xxPWMState), + VMSTATE_UINT32(pcr, NPCM7xxPWMState), + VMSTATE_UINT32(pier, NPCM7xxPWMState), + VMSTATE_UINT32(piir, NPCM7xxPWMState), + VMSTATE_END_OF_LIST(), + }, +}; + +static void npcm7xx_pwm_class_init(ObjectClass *klass, void *data) +{ + ResettableClass *rc = RESETTABLE_CLASS(klass); + DeviceClass *dc = DEVICE_CLASS(klass); + + dc->desc = "NPCM7xx PWM Controller"; + dc->vmsd = &vmstate_npcm7xx_pwm_module; + rc->phases.enter = npcm7xx_pwm_enter_reset; + rc->phases.hold = npcm7xx_pwm_hold_reset; +} + +static const TypeInfo npcm7xx_pwm_info = { + .name = TYPE_NPCM7XX_PWM, + .parent = TYPE_SYS_BUS_DEVICE, + .instance_size = sizeof(NPCM7xxPWMState), + .class_init = npcm7xx_pwm_class_init, + .instance_init = npcm7xx_pwm_init, +}; + +static void npcm7xx_pwm_register_type(void) +{ + type_register_static(&npcm7xx_pwm_info); +} +type_init(npcm7xx_pwm_register_type); 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_MAINSTONE', if_true: files('mst_fpga.c')) softmmu_ss.add(when: 'CONFIG_NPCM7XX', if_true: files( 'npcm7xx_clk.c', 'npcm7xx_gcr.c', + 'npcm7xx_pwm.c', 'npcm7xx_rng.c', )) softmmu_ss.add(when: 'CONFIG_OMAP', if_true: files( 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 @@ npcm7xx_gcr_write(uint64_t offset, uint32_t value) "offset: 0x%04" PRIx64 " valu npcm7xx_rng_read(uint64_t offset, uint64_t value, unsigned size) "offset: 0x%04" PRIx64 " value: 0x%02" PRIx64 " size: %u" npcm7xx_rng_write(uint64_t offset, uint64_t value, unsigned size) "offset: 0x%04" PRIx64 " value: 0x%02" PRIx64 " size: %u" +# npcm7xx_pwm.c +npcm7xx_pwm_read(const char *id, uint64_t offset, uint32_t value) "%s offset: 0x%04" PRIx64 " value: 0x%08" PRIx32 +npcm7xx_pwm_write(const char *id, uint64_t offset, uint32_t value) "%s offset: 0x%04" PRIx64 " value: 0x%08" PRIx32 +npcm7xx_pwm_update_freq(const char *id, uint8_t index, uint32_t old_value, uint32_t new_value) "%s pwm[%u] Update Freq: old_freq: %u, new_freq: %u" +npcm7xx_pwm_update_duty(const char *id, uint8_t index, uint32_t old_value, uint32_t new_value) "%s pwm[%u] Update Duty: old_duty: %u, new_duty: %u" + # stm32f4xx_syscfg.c stm32f4xx_syscfg_set_irq(int gpio, int line, int level) "Interupt: GPIO: %d, Line: %d; Level: %d" stm32f4xx_pulse_exti(int irq) "Pulse EXTI: %d" -- 2.20.1 From: Hao Wu <wuhaotsh@google.com> We add a qtest for the PWM in the previous patch. It proves it works as expected. Reviewed-by: Havard Skinnemoen <hskinnemoen@google.com> Reviewed-by: Tyrone Ting <kfting@nuvoton.com> Signed-off-by: Hao Wu <wuhaotsh@google.com> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Message-id: 20210108190945.949196-6-wuhaotsh@google.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- tests/qtest/npcm7xx_pwm-test.c | 490 +++++++++++++++++++++++++++++++++ tests/qtest/meson.build | 1 + 2 files changed, 491 insertions(+) create mode 100644 tests/qtest/npcm7xx_pwm-test.c diff --git a/tests/qtest/npcm7xx_pwm-test.c b/tests/qtest/npcm7xx_pwm-test.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/tests/qtest/npcm7xx_pwm-test.c @@ -XXX,XX +XXX,XX @@ +/* + * QTests for Nuvoton NPCM7xx PWM Modules. + * + * Copyright 2020 Google LLC + * + * 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. + */ + +#include "qemu/osdep.h" +#include "qemu/bitops.h" +#include "libqos/libqtest.h" +#include "qapi/qmp/qdict.h" +#include "qapi/qmp/qnum.h" + +#define REF_HZ 25000000 + +/* Register field definitions. */ +#define CH_EN BIT(0) +#define CH_INV BIT(2) +#define CH_MOD BIT(3) + +/* Registers shared between all PWMs in a module */ +#define PPR 0x00 +#define CSR 0x04 +#define PCR 0x08 +#define PIER 0x3c +#define PIIR 0x40 + +/* CLK module related */ +#define CLK_BA 0xf0801000 +#define CLKSEL 0x04 +#define CLKDIV1 0x08 +#define CLKDIV2 0x2c +#define PLLCON0 0x0c +#define PLLCON1 0x10 +#define PLL_INDV(rv) extract32((rv), 0, 6) +#define PLL_FBDV(rv) extract32((rv), 16, 12) +#define PLL_OTDV1(rv) extract32((rv), 8, 3) +#define PLL_OTDV2(rv) extract32((rv), 13, 3) +#define APB3CKDIV(rv) extract32((rv), 28, 2) +#define CLK2CKDIV(rv) extract32((rv), 0, 1) +#define CLK4CKDIV(rv) extract32((rv), 26, 2) +#define CPUCKSEL(rv) extract32((rv), 0, 2) + +#define MAX_DUTY 1000000 + +typedef struct PWMModule { + int irq; + uint64_t base_addr; +} PWMModule; + +typedef struct PWM { + uint32_t cnr_offset; + uint32_t cmr_offset; + uint32_t pdr_offset; + uint32_t pwdr_offset; +} PWM; + +typedef struct TestData { + const PWMModule *module; + const PWM *pwm; +} TestData; + +static const PWMModule pwm_module_list[] = { + { + .irq = 93, + .base_addr = 0xf0103000 + }, + { + .irq = 94, + .base_addr = 0xf0104000 + } +}; + +static const PWM pwm_list[] = { + { + .cnr_offset = 0x0c, + .cmr_offset = 0x10, + .pdr_offset = 0x14, + .pwdr_offset = 0x44, + }, + { + .cnr_offset = 0x18, + .cmr_offset = 0x1c, + .pdr_offset = 0x20, + .pwdr_offset = 0x48, + }, + { + .cnr_offset = 0x24, + .cmr_offset = 0x28, + .pdr_offset = 0x2c, + .pwdr_offset = 0x4c, + }, + { + .cnr_offset = 0x30, + .cmr_offset = 0x34, + .pdr_offset = 0x38, + .pwdr_offset = 0x50, + }, +}; + +static const int ppr_base[] = { 0, 0, 8, 8 }; +static const int csr_base[] = { 0, 4, 8, 12 }; +static const int pcr_base[] = { 0, 8, 12, 16 }; + +static const uint32_t ppr_list[] = { + 0, + 1, + 10, + 100, + 255, /* Max possible value. */ +}; + +static const uint32_t csr_list[] = { + 0, + 1, + 2, + 3, + 4, /* Max possible value. */ +}; + +static const uint32_t cnr_list[] = { + 0, + 1, + 50, + 100, + 150, + 200, + 1000, + 10000, + 65535, /* Max possible value. */ +}; + +static const uint32_t cmr_list[] = { + 0, + 1, + 10, + 50, + 100, + 150, + 200, + 1000, + 10000, + 65535, /* Max possible value. */ +}; + +/* Returns the index of the PWM module. */ +static int pwm_module_index(const PWMModule *module) +{ + ptrdiff_t diff = module - pwm_module_list; + + g_assert_true(diff >= 0 && diff < ARRAY_SIZE(pwm_module_list)); + + return diff; +} + +/* Returns the index of the PWM entry. */ +static int pwm_index(const PWM *pwm) +{ + ptrdiff_t diff = pwm - pwm_list; + + g_assert_true(diff >= 0 && diff < ARRAY_SIZE(pwm_list)); + + return diff; +} + +static uint64_t pwm_qom_get(QTestState *qts, const char *path, const char *name) +{ + QDict *response; + + g_test_message("Getting properties %s from %s", name, path); + response = qtest_qmp(qts, "{ 'execute': 'qom-get'," + " 'arguments': { 'path': %s, 'property': %s}}", + path, name); + /* The qom set message returns successfully. */ + g_assert_true(qdict_haskey(response, "return")); + return qnum_get_uint(qobject_to(QNum, qdict_get(response, "return"))); +} + +static uint64_t pwm_get_freq(QTestState *qts, int module_index, int pwm_index) +{ + char path[100]; + char name[100]; + + sprintf(path, "/machine/soc/pwm[%d]", module_index); + sprintf(name, "freq[%d]", pwm_index); + + return pwm_qom_get(qts, path, name); +} + +static uint64_t pwm_get_duty(QTestState *qts, int module_index, int pwm_index) +{ + char path[100]; + char name[100]; + + sprintf(path, "/machine/soc/pwm[%d]", module_index); + sprintf(name, "duty[%d]", pwm_index); + + return pwm_qom_get(qts, path, name); +} + +static uint32_t get_pll(uint32_t con) +{ + return REF_HZ * PLL_FBDV(con) / (PLL_INDV(con) * PLL_OTDV1(con) + * PLL_OTDV2(con)); +} + +static uint64_t read_pclk(QTestState *qts) +{ + uint64_t freq = REF_HZ; + uint32_t clksel = qtest_readl(qts, CLK_BA + CLKSEL); + uint32_t pllcon; + uint32_t clkdiv1 = qtest_readl(qts, CLK_BA + CLKDIV1); + uint32_t clkdiv2 = qtest_readl(qts, CLK_BA + CLKDIV2); + + switch (CPUCKSEL(clksel)) { + case 0: + pllcon = qtest_readl(qts, CLK_BA + PLLCON0); + freq = get_pll(pllcon); + break; + case 1: + pllcon = qtest_readl(qts, CLK_BA + PLLCON1); + freq = get_pll(pllcon); + break; + case 2: + break; + case 3: + break; + default: + g_assert_not_reached(); + } + + freq >>= (CLK2CKDIV(clkdiv1) + CLK4CKDIV(clkdiv1) + APB3CKDIV(clkdiv2)); + + return freq; +} + +static uint32_t pwm_selector(uint32_t csr) +{ + switch (csr) { + case 0: + return 2; + case 1: + return 4; + case 2: + return 8; + case 3: + return 16; + case 4: + return 1; + default: + g_assert_not_reached(); + } +} + +static uint64_t pwm_compute_freq(QTestState *qts, uint32_t ppr, uint32_t csr, + uint32_t cnr) +{ + return read_pclk(qts) / ((ppr + 1) * pwm_selector(csr) * (cnr + 1)); +} + +static uint64_t pwm_compute_duty(uint32_t cnr, uint32_t cmr, bool inverted) +{ + uint64_t duty; + + if (cnr == 0) { + /* PWM is stopped. */ + duty = 0; + } else if (cmr >= cnr) { + duty = MAX_DUTY; + } else { + duty = MAX_DUTY * (cmr + 1) / (cnr + 1); + } + + if (inverted) { + duty = MAX_DUTY - duty; + } + + return duty; +} + +static uint32_t pwm_read(QTestState *qts, const TestData *td, unsigned offset) +{ + return qtest_readl(qts, td->module->base_addr + offset); +} + +static void pwm_write(QTestState *qts, const TestData *td, unsigned offset, + uint32_t value) +{ + qtest_writel(qts, td->module->base_addr + offset, value); +} + +static uint32_t pwm_read_ppr(QTestState *qts, const TestData *td) +{ + return extract32(pwm_read(qts, td, PPR), ppr_base[pwm_index(td->pwm)], 8); +} + +static void pwm_write_ppr(QTestState *qts, const TestData *td, uint32_t value) +{ + pwm_write(qts, td, PPR, value << ppr_base[pwm_index(td->pwm)]); +} + +static uint32_t pwm_read_csr(QTestState *qts, const TestData *td) +{ + return extract32(pwm_read(qts, td, CSR), csr_base[pwm_index(td->pwm)], 3); +} + +static void pwm_write_csr(QTestState *qts, const TestData *td, uint32_t value) +{ + pwm_write(qts, td, CSR, value << csr_base[pwm_index(td->pwm)]); +} + +static uint32_t pwm_read_pcr(QTestState *qts, const TestData *td) +{ + return extract32(pwm_read(qts, td, PCR), pcr_base[pwm_index(td->pwm)], 4); +} + +static void pwm_write_pcr(QTestState *qts, const TestData *td, uint32_t value) +{ + pwm_write(qts, td, PCR, value << pcr_base[pwm_index(td->pwm)]); +} + +static uint32_t pwm_read_cnr(QTestState *qts, const TestData *td) +{ + return pwm_read(qts, td, td->pwm->cnr_offset); +} + +static void pwm_write_cnr(QTestState *qts, const TestData *td, uint32_t value) +{ + pwm_write(qts, td, td->pwm->cnr_offset, value); +} + +static uint32_t pwm_read_cmr(QTestState *qts, const TestData *td) +{ + return pwm_read(qts, td, td->pwm->cmr_offset); +} + +static void pwm_write_cmr(QTestState *qts, const TestData *td, uint32_t value) +{ + pwm_write(qts, td, td->pwm->cmr_offset, value); +} + +/* Check pwm registers can be reset to default value */ +static void test_init(gconstpointer test_data) +{ + const TestData *td = test_data; + QTestState *qts = qtest_init("-machine quanta-gsj"); + int module = pwm_module_index(td->module); + int pwm = pwm_index(td->pwm); + + g_assert_cmpuint(pwm_get_freq(qts, module, pwm), ==, 0); + g_assert_cmpuint(pwm_get_duty(qts, module, pwm), ==, 0); + + qtest_quit(qts); +} + +/* One-shot mode should not change frequency and duty cycle. */ +static void test_oneshot(gconstpointer test_data) +{ + const TestData *td = test_data; + QTestState *qts = qtest_init("-machine quanta-gsj"); + int module = pwm_module_index(td->module); + int pwm = pwm_index(td->pwm); + uint32_t ppr, csr, pcr; + int i, j; + + pcr = CH_EN; + for (i = 0; i < ARRAY_SIZE(ppr_list); ++i) { + ppr = ppr_list[i]; + pwm_write_ppr(qts, td, ppr); + + for (j = 0; j < ARRAY_SIZE(csr_list); ++j) { + csr = csr_list[j]; + pwm_write_csr(qts, td, csr); + pwm_write_pcr(qts, td, pcr); + + g_assert_cmpuint(pwm_read_ppr(qts, td), ==, ppr); + g_assert_cmpuint(pwm_read_csr(qts, td), ==, csr); + g_assert_cmpuint(pwm_read_pcr(qts, td), ==, pcr); + g_assert_cmpuint(pwm_get_freq(qts, module, pwm), ==, 0); + g_assert_cmpuint(pwm_get_duty(qts, module, pwm), ==, 0); + } + } + + qtest_quit(qts); +} + +/* In toggle mode, the PWM generates correct outputs. */ +static void test_toggle(gconstpointer test_data) +{ + const TestData *td = test_data; + QTestState *qts = qtest_init("-machine quanta-gsj"); + int module = pwm_module_index(td->module); + int pwm = pwm_index(td->pwm); + uint32_t ppr, csr, pcr, cnr, cmr; + int i, j, k, l; + uint64_t expected_freq, expected_duty; + + pcr = CH_EN | CH_MOD; + for (i = 0; i < ARRAY_SIZE(ppr_list); ++i) { + ppr = ppr_list[i]; + pwm_write_ppr(qts, td, ppr); + + for (j = 0; j < ARRAY_SIZE(csr_list); ++j) { + csr = csr_list[j]; + pwm_write_csr(qts, td, csr); + + for (k = 0; k < ARRAY_SIZE(cnr_list); ++k) { + cnr = cnr_list[k]; + pwm_write_cnr(qts, td, cnr); + + for (l = 0; l < ARRAY_SIZE(cmr_list); ++l) { + cmr = cmr_list[l]; + pwm_write_cmr(qts, td, cmr); + expected_freq = pwm_compute_freq(qts, ppr, csr, cnr); + expected_duty = pwm_compute_duty(cnr, cmr, false); + + pwm_write_pcr(qts, td, pcr); + g_assert_cmpuint(pwm_read_ppr(qts, td), ==, ppr); + g_assert_cmpuint(pwm_read_csr(qts, td), ==, csr); + g_assert_cmpuint(pwm_read_pcr(qts, td), ==, pcr); + g_assert_cmpuint(pwm_read_cnr(qts, td), ==, cnr); + g_assert_cmpuint(pwm_read_cmr(qts, td), ==, cmr); + g_assert_cmpuint(pwm_get_duty(qts, module, pwm), + ==, expected_duty); + if (expected_duty != 0 && expected_duty != 100) { + /* Duty cycle with 0 or 100 doesn't need frequency. */ + g_assert_cmpuint(pwm_get_freq(qts, module, pwm), + ==, expected_freq); + } + + /* Test inverted mode */ + expected_duty = pwm_compute_duty(cnr, cmr, true); + pwm_write_pcr(qts, td, pcr | CH_INV); + g_assert_cmpuint(pwm_read_pcr(qts, td), ==, pcr | CH_INV); + g_assert_cmpuint(pwm_get_duty(qts, module, pwm), + ==, expected_duty); + if (expected_duty != 0 && expected_duty != 100) { + /* Duty cycle with 0 or 100 doesn't need frequency. */ + g_assert_cmpuint(pwm_get_freq(qts, module, pwm), + ==, expected_freq); + } + + } + } + } + } + + qtest_quit(qts); +} + +static void pwm_add_test(const char *name, const TestData* td, + GTestDataFunc fn) +{ + g_autofree char *full_name = g_strdup_printf( + "npcm7xx_pwm/module[%d]/pwm[%d]/%s", pwm_module_index(td->module), + pwm_index(td->pwm), name); + qtest_add_data_func(full_name, td, fn); +} +#define add_test(name, td) pwm_add_test(#name, td, test_##name) + +int main(int argc, char **argv) +{ + TestData test_data_list[ARRAY_SIZE(pwm_module_list) * ARRAY_SIZE(pwm_list)]; + + g_test_init(&argc, &argv, NULL); + + for (int i = 0; i < ARRAY_SIZE(pwm_module_list); ++i) { + for (int j = 0; j < ARRAY_SIZE(pwm_list); ++j) { + TestData *td = &test_data_list[i * ARRAY_SIZE(pwm_list) + j]; + + td->module = &pwm_module_list[i]; + td->pwm = &pwm_list[j]; + + add_test(init, td); + add_test(oneshot, td); + add_test(toggle, td); + } + } + + 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_sparc64 = \ qtests_npcm7xx = \ ['npcm7xx_adc-test', 'npcm7xx_gpio-test', + 'npcm7xx_pwm-test', 'npcm7xx_rng-test', 'npcm7xx_timer-test', 'npcm7xx_watchdog_timer-test'] -- 2.20.1 From: Hao Wu <wuhaotsh@google.com> A device shouldn't access its parent object which is QOM internal. Instead it should use type cast for this purporse. This patch fixes this issue for all NPCM7XX Devices. Signed-off-by: Hao Wu <wuhaotsh@google.com> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Message-id: 20210108190945.949196-7-wuhaotsh@google.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/npcm7xx_boards.c | 2 +- hw/mem/npcm7xx_mc.c | 2 +- hw/misc/npcm7xx_clk.c | 2 +- hw/misc/npcm7xx_gcr.c | 2 +- hw/misc/npcm7xx_rng.c | 2 +- hw/nvram/npcm7xx_otp.c | 2 +- hw/ssi/npcm7xx_fiu.c | 2 +- 7 files changed, 7 insertions(+), 7 deletions(-) diff --git a/hw/arm/npcm7xx_boards.c b/hw/arm/npcm7xx_boards.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/npcm7xx_boards.c +++ b/hw/arm/npcm7xx_boards.c @@ -XXX,XX +XXX,XX @@ static NPCM7xxState *npcm7xx_create_soc(MachineState *machine, uint32_t hw_straps) { NPCM7xxMachineClass *nmc = NPCM7XX_MACHINE_GET_CLASS(machine); - MachineClass *mc = &nmc->parent; + MachineClass *mc = MACHINE_CLASS(nmc); Object *obj; if (strcmp(machine->cpu_type, mc->default_cpu_type) != 0) { diff --git a/hw/mem/npcm7xx_mc.c b/hw/mem/npcm7xx_mc.c index XXXXXXX..XXXXXXX 100644 --- a/hw/mem/npcm7xx_mc.c +++ b/hw/mem/npcm7xx_mc.c @@ -XXX,XX +XXX,XX @@ static void npcm7xx_mc_realize(DeviceState *dev, Error **errp) memory_region_init_io(&s->mmio, OBJECT(s), &npcm7xx_mc_ops, s, "regs", NPCM7XX_MC_REGS_SIZE); - sysbus_init_mmio(&s->parent, &s->mmio); + sysbus_init_mmio(SYS_BUS_DEVICE(s), &s->mmio); } static void npcm7xx_mc_class_init(ObjectClass *klass, void *data) diff --git a/hw/misc/npcm7xx_clk.c b/hw/misc/npcm7xx_clk.c index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/npcm7xx_clk.c +++ b/hw/misc/npcm7xx_clk.c @@ -XXX,XX +XXX,XX @@ static void npcm7xx_clk_init(Object *obj) memory_region_init_io(&s->iomem, obj, &npcm7xx_clk_ops, s, TYPE_NPCM7XX_CLK, 4 * KiB); - sysbus_init_mmio(&s->parent, &s->iomem); + sysbus_init_mmio(SYS_BUS_DEVICE(s), &s->iomem); } static int npcm7xx_clk_post_load(void *opaque, int version_id) diff --git a/hw/misc/npcm7xx_gcr.c b/hw/misc/npcm7xx_gcr.c index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/npcm7xx_gcr.c +++ b/hw/misc/npcm7xx_gcr.c @@ -XXX,XX +XXX,XX @@ static void npcm7xx_gcr_init(Object *obj) memory_region_init_io(&s->iomem, obj, &npcm7xx_gcr_ops, s, TYPE_NPCM7XX_GCR, 4 * KiB); - sysbus_init_mmio(&s->parent, &s->iomem); + sysbus_init_mmio(SYS_BUS_DEVICE(s), &s->iomem); } static const VMStateDescription vmstate_npcm7xx_gcr = { diff --git a/hw/misc/npcm7xx_rng.c b/hw/misc/npcm7xx_rng.c index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/npcm7xx_rng.c +++ b/hw/misc/npcm7xx_rng.c @@ -XXX,XX +XXX,XX @@ static void npcm7xx_rng_init(Object *obj) memory_region_init_io(&s->iomem, obj, &npcm7xx_rng_ops, s, "regs", NPCM7XX_RNG_REGS_SIZE); - sysbus_init_mmio(&s->parent, &s->iomem); + sysbus_init_mmio(SYS_BUS_DEVICE(s), &s->iomem); } static const VMStateDescription vmstate_npcm7xx_rng = { diff --git a/hw/nvram/npcm7xx_otp.c b/hw/nvram/npcm7xx_otp.c index XXXXXXX..XXXXXXX 100644 --- a/hw/nvram/npcm7xx_otp.c +++ b/hw/nvram/npcm7xx_otp.c @@ -XXX,XX +XXX,XX @@ static void npcm7xx_otp_realize(DeviceState *dev, Error **errp) { NPCM7xxOTPClass *oc = NPCM7XX_OTP_GET_CLASS(dev); NPCM7xxOTPState *s = NPCM7XX_OTP(dev); - SysBusDevice *sbd = &s->parent; + SysBusDevice *sbd = SYS_BUS_DEVICE(dev); memset(s->array, 0, sizeof(s->array)); diff --git a/hw/ssi/npcm7xx_fiu.c b/hw/ssi/npcm7xx_fiu.c index XXXXXXX..XXXXXXX 100644 --- a/hw/ssi/npcm7xx_fiu.c +++ b/hw/ssi/npcm7xx_fiu.c @@ -XXX,XX +XXX,XX @@ static void npcm7xx_fiu_hold_reset(Object *obj) static void npcm7xx_fiu_realize(DeviceState *dev, Error **errp) { NPCM7xxFIUState *s = NPCM7XX_FIU(dev); - SysBusDevice *sbd = &s->parent; + SysBusDevice *sbd = SYS_BUS_DEVICE(dev); int i; if (s->cs_count <= 0) { -- 2.20.1 From: Roman Bolshakov <r.bolshakov@yadro.com> ui/cocoa.m:1188:44: warning: 'openFile:' is deprecated: first deprecated in macOS 11.0 - Use -[NSWorkspace openURL:] instead. [-Wdeprecated-declarations] if ([[NSWorkspace sharedWorkspace] openFile: full_file_path] == YES) { ^ /Library/Developer/CommandLineTools/SDKs/MacOSX.sdk/System/Library/Frameworks/AppKit.framework/Headers/NSWorkspace.h:350:1: note: 'openFile:' has been explicitly marked deprecated here - (BOOL)openFile:(NSString *)fullPath API_DEPRECATED("Use -[NSWorkspace openURL:] instead.", macos(10.0, 11.0)); ^ Signed-off-by: Roman Bolshakov <r.bolshakov@yadro.com> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Message-id: 20210102150718.47618-1-r.bolshakov@yadro.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- ui/cocoa.m | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/ui/cocoa.m b/ui/cocoa.m index XXXXXXX..XXXXXXX 100644 --- a/ui/cocoa.m +++ b/ui/cocoa.m @@ -XXX,XX +XXX,XX @@ QemuCocoaView *cocoaView; /* Where to look for local files */ NSString *path_array[] = {@"../share/doc/qemu/", @"../doc/qemu/", @"docs/"}; NSString *full_file_path; + NSURL *full_file_url; /* iterate thru the possible paths until the file is found */ int index; @@ -XXX,XX +XXX,XX @@ QemuCocoaView *cocoaView; full_file_path = [full_file_path stringByDeletingLastPathComponent]; full_file_path = [NSString stringWithFormat: @"%@/%@%@", full_file_path, path_array[index], filename]; - if ([[NSWorkspace sharedWorkspace] openFile: full_file_path] == YES) { + full_file_url = [NSURL fileURLWithPath: full_file_path + isDirectory: false]; + if ([[NSWorkspace sharedWorkspace] openURL: full_file_url] == YES) { return; } } -- 2.20.1