The following changes since commit 53f306f316549d20c76886903181413d20842423: Merge remote-tracking branch 'remotes/ehabkost-gl/tags/x86-next-pull-request' into staging (2021-06-21 11:26:04 +0100) are available in the Git repository at: https://git.linaro.org/people/pmaydell/qemu-arm.git tags/pull-target-arm-20210621 for you to fetch changes up to a83f1d9263d281f938a3984cda7104d55affd43a: docs/system: arm: Add nRF boards description (2021-06-21 17:24:33 +0100) ---------------------------------------------------------------- target-arm queue: * Don't require 'virt' board to be compiled in for ACPI GHES code * docs: Document which architecture extensions we emulate * Fix bugs in M-profile FPCXT_NS accesses * First slice of MVE patches * Implement MTE3 * docs/system: arm: Add nRF boards description ---------------------------------------------------------------- Alexandre Iooss (1): docs/system: arm: Add nRF boards description Peter Collingbourne (1): target/arm: Implement MTE3 Peter Maydell (55): hw/acpi: Provide stub version of acpi_ghes_record_errors() hw/acpi: Provide function acpi_ghes_present() target/arm: Use acpi_ghes_present() to see if we report ACPI memory errors docs/system/arm: Document which architecture extensions we emulate target/arm/translate-vfp.c: Whitespace fixes target/arm: Handle FPU being disabled in FPCXT_NS accesses target/arm: Don't NOCP fault for FPCXT_NS accesses target/arm: Handle writeback in VLDR/VSTR sysreg with no memory access target/arm: Factor FP context update code out into helper function target/arm: Split vfp_access_check() into A and M versions target/arm: Handle FPU check for FPCXT_NS insns via vfp_access_check_m() target/arm: Implement MVE VLDR/VSTR (non-widening forms) target/arm: Implement widening/narrowing MVE VLDR/VSTR insns target/arm: Implement MVE VCLZ target/arm: Implement MVE VCLS target/arm: Implement MVE VREV16, VREV32, VREV64 target/arm: Implement MVE VMVN (register) target/arm: Implement MVE VABS target/arm: Implement MVE VNEG tcg: Make gen_dup_i32/i64() public as tcg_gen_dup_i32/i64 target/arm: Implement MVE VDUP target/arm: Implement MVE VAND, VBIC, VORR, VORN, VEOR target/arm: Implement MVE VADD, VSUB, VMUL target/arm: Implement MVE VMULH target/arm: Implement MVE VRMULH target/arm: Implement MVE VMAX, VMIN target/arm: Implement MVE VABD target/arm: Implement MVE VHADD, VHSUB target/arm: Implement MVE VMULL target/arm: Implement MVE VMLALDAV target/arm: Implement MVE VMLSLDAV target/arm: Implement MVE VRMLALDAVH, VRMLSLDAVH target/arm: Implement MVE VADD (scalar) target/arm: Implement MVE VSUB, VMUL (scalar) target/arm: Implement MVE VHADD, VHSUB (scalar) target/arm: Implement MVE VBRSR target/arm: Implement MVE VPST target/arm: Implement MVE VQADD and VQSUB target/arm: Implement MVE VQDMULH and VQRDMULH (scalar) target/arm: Implement MVE VQDMULL scalar target/arm: Implement MVE VQDMULH, VQRDMULH (vector) target/arm: Implement MVE VQADD, VQSUB (vector) target/arm: Implement MVE VQSHL (vector) target/arm: Implement MVE VQRSHL target/arm: Implement MVE VSHL insn target/arm: Implement MVE VRSHL target/arm: Implement MVE VQDMLADH and VQRDMLADH target/arm: Implement MVE VQDMLSDH and VQRDMLSDH target/arm: Implement MVE VQDMULL (vector) target/arm: Implement MVE VRHADD target/arm: Implement MVE VADC, VSBC target/arm: Implement MVE VCADD target/arm: Implement MVE VHCADD target/arm: Implement MVE VADDV target/arm: Make VMOV scalar <-> gpreg beatwise for MVE docs/system/arm/emulation.rst | 103 ++++ docs/system/arm/nrf.rst | 51 ++ docs/system/target-arm.rst | 7 + include/hw/acpi/ghes.h | 9 + include/tcg/tcg-op.h | 8 + include/tcg/tcg.h | 1 - target/arm/helper-mve.h | 357 +++++++++++++ target/arm/helper.h | 2 + target/arm/internals.h | 11 + target/arm/translate-a32.h | 3 + target/arm/translate.h | 10 + target/arm/m-nocp.decode | 24 + target/arm/mve.decode | 240 +++++++++ target/arm/vfp.decode | 14 - hw/acpi/ghes-stub.c | 22 + hw/acpi/ghes.c | 17 + target/arm/cpu64.c | 2 +- target/arm/kvm64.c | 6 +- target/arm/mte_helper.c | 82 +-- target/arm/mve_helper.c | 1160 +++++++++++++++++++++++++++++++++++++++++ target/arm/translate-m-nocp.c | 550 +++++++++++++++++++ target/arm/translate-mve.c | 759 +++++++++++++++++++++++++++ target/arm/translate-vfp.c | 741 +++++++------------------- tcg/tcg-op-gvec.c | 20 +- MAINTAINERS | 1 + hw/acpi/meson.build | 6 +- target/arm/meson.build | 1 + 27 files changed, 3578 insertions(+), 629 deletions(-) create mode 100644 docs/system/arm/emulation.rst create mode 100644 docs/system/arm/nrf.rst create mode 100644 target/arm/helper-mve.h create mode 100644 hw/acpi/ghes-stub.c create mode 100644 target/arm/mve_helper.c Generic code in target/arm wants to call acpi_ghes_record_errors(); provide a stub version so that we don't fail to link when CONFIG_ACPI_APEI is not set. This requires us to add a new ghes-stub.c file to contain it and the meson.build mechanics to use it when appropriate. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Dongjiu Geng <gengdongjiu1@gmail.com> Message-id: 20210603171259.27962-2-peter.maydell@linaro.org --- hw/acpi/ghes-stub.c | 17 +++++++++++++++++ hw/acpi/meson.build | 6 +++--- 2 files changed, 20 insertions(+), 3 deletions(-) create mode 100644 hw/acpi/ghes-stub.c diff --git a/hw/acpi/ghes-stub.c b/hw/acpi/ghes-stub.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/acpi/ghes-stub.c @@ -XXX,XX +XXX,XX @@ +/* + * Support for generating APEI tables and recording CPER for Guests: + * stub functions. + * + * Copyright (c) 2021 Linaro, Ltd + * + * This work is licensed under the terms of the GNU GPL, version 2 or later. + * See the COPYING file in the top-level directory. + */ + +#include "qemu/osdep.h" +#include "hw/acpi/ghes.h" + +int acpi_ghes_record_errors(uint8_t source_id, uint64_t physical_address) +{ + return -1; +} diff --git a/hw/acpi/meson.build b/hw/acpi/meson.build index XXXXXXX..XXXXXXX 100644 --- a/hw/acpi/meson.build +++ b/hw/acpi/meson.build @@ -XXX,XX +XXX,XX @@ acpi_ss.add(when: 'CONFIG_ACPI_PCI', if_true: files('pci.c')) acpi_ss.add(when: 'CONFIG_ACPI_VMGENID', if_true: files('vmgenid.c')) acpi_ss.add(when: 'CONFIG_ACPI_HW_REDUCED', if_true: files('generic_event_device.c')) acpi_ss.add(when: 'CONFIG_ACPI_HMAT', if_true: files('hmat.c')) -acpi_ss.add(when: 'CONFIG_ACPI_APEI', if_true: files('ghes.c')) +acpi_ss.add(when: 'CONFIG_ACPI_APEI', if_true: files('ghes.c'), if_false: files('ghes-stub.c')) acpi_ss.add(when: 'CONFIG_ACPI_X86', if_true: files('core.c', 'piix4.c', 'pcihp.c'), if_false: files('acpi-stub.c')) acpi_ss.add(when: 'CONFIG_ACPI_X86_ICH', if_true: files('ich9.c', 'tco.c')) acpi_ss.add(when: 'CONFIG_IPMI', if_true: files('ipmi.c'), if_false: files('ipmi-stub.c')) acpi_ss.add(when: 'CONFIG_PC', if_false: files('acpi-x86-stub.c')) acpi_ss.add(when: 'CONFIG_TPM', if_true: files('tpm.c')) -softmmu_ss.add(when: 'CONFIG_ACPI', if_false: files('acpi-stub.c', 'aml-build-stub.c')) +softmmu_ss.add(when: 'CONFIG_ACPI', if_false: files('acpi-stub.c', 'aml-build-stub.c', 'ghes-stub.c')) softmmu_ss.add_all(when: 'CONFIG_ACPI', if_true: acpi_ss) softmmu_ss.add(when: 'CONFIG_ALL', if_true: files('acpi-stub.c', 'aml-build-stub.c', - 'acpi-x86-stub.c', 'ipmi-stub.c')) + 'acpi-x86-stub.c', 'ipmi-stub.c', 'ghes-stub.c')) -- 2.20.1 Allow code elsewhere in the system to check whether the ACPI GHES table is present, so it can determine whether it is OK to try to record an error by calling acpi_ghes_record_errors(). (We don't need to migrate the new 'present' field in AcpiGhesState, because it is set once at system initialization and doesn't change.) Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Dongjiu Geng <gengdongjiu1@gmail.com> Message-id: 20210603171259.27962-3-peter.maydell@linaro.org --- include/hw/acpi/ghes.h | 9 +++++++++ hw/acpi/ghes-stub.c | 5 +++++ hw/acpi/ghes.c | 17 +++++++++++++++++ 3 files changed, 31 insertions(+) diff --git a/include/hw/acpi/ghes.h b/include/hw/acpi/ghes.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/acpi/ghes.h +++ b/include/hw/acpi/ghes.h @@ -XXX,XX +XXX,XX @@ enum { typedef struct AcpiGhesState { uint64_t ghes_addr_le; + bool present; /* True if GHES is present at all on this board */ } AcpiGhesState; void build_ghes_error_table(GArray *hardware_errors, BIOSLinker *linker); @@ -XXX,XX +XXX,XX @@ void acpi_build_hest(GArray *table_data, BIOSLinker *linker, void acpi_ghes_add_fw_cfg(AcpiGhesState *vms, FWCfgState *s, GArray *hardware_errors); int acpi_ghes_record_errors(uint8_t notify, uint64_t error_physical_addr); + +/** + * acpi_ghes_present: Report whether ACPI GHES table is present + * + * Returns: true if the system has an ACPI GHES table and it is + * safe to call acpi_ghes_record_errors() to record a memory error. + */ +bool acpi_ghes_present(void); #endif diff --git a/hw/acpi/ghes-stub.c b/hw/acpi/ghes-stub.c index XXXXXXX..XXXXXXX 100644 --- a/hw/acpi/ghes-stub.c +++ b/hw/acpi/ghes-stub.c @@ -XXX,XX +XXX,XX @@ int acpi_ghes_record_errors(uint8_t source_id, uint64_t physical_address) { return -1; } + +bool acpi_ghes_present(void) +{ + return false; +} diff --git a/hw/acpi/ghes.c b/hw/acpi/ghes.c index XXXXXXX..XXXXXXX 100644 --- a/hw/acpi/ghes.c +++ b/hw/acpi/ghes.c @@ -XXX,XX +XXX,XX @@ void acpi_ghes_add_fw_cfg(AcpiGhesState *ags, FWCfgState *s, /* Create a read-write fw_cfg file for Address */ fw_cfg_add_file_callback(s, ACPI_GHES_DATA_ADDR_FW_CFG_FILE, NULL, NULL, NULL, &(ags->ghes_addr_le), sizeof(ags->ghes_addr_le), false); + + ags->present = true; } int acpi_ghes_record_errors(uint8_t source_id, uint64_t physical_address) @@ -XXX,XX +XXX,XX @@ int acpi_ghes_record_errors(uint8_t source_id, uint64_t physical_address) return ret; } + +bool acpi_ghes_present(void) +{ + AcpiGedState *acpi_ged_state; + AcpiGhesState *ags; + + acpi_ged_state = ACPI_GED(object_resolve_path_type("", TYPE_ACPI_GED, + NULL)); + + if (!acpi_ged_state) { + return false; + } + ags = &acpi_ged_state->ghes_state; + return ags->present; +} -- 2.20.1 The virt_is_acpi_enabled() function is specific to the virt board, as is the check for its 'ras' property. Use the new acpi_ghes_present() function to check whether we should report memory errors via acpi_ghes_record_errors(). This avoids a link error if QEMU was built without support for the virt board, and provides a mechanism that can be used by any future board models that want to add ACPI memory error reporting support (they only need to call acpi_ghes_add_fw_cfg()). Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Dongjiu Geng <gengdongjiu1@gmail.com> Message-id: 20210603171259.27962-4-peter.maydell@linaro.org --- target/arm/kvm64.c | 6 +----- 1 file changed, 1 insertion(+), 5 deletions(-) diff --git a/target/arm/kvm64.c b/target/arm/kvm64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/kvm64.c +++ b/target/arm/kvm64.c @@ -XXX,XX +XXX,XX @@ void kvm_arch_on_sigbus_vcpu(CPUState *c, int code, void *addr) { ram_addr_t ram_addr; hwaddr paddr; - Object *obj = qdev_get_machine(); - VirtMachineState *vms = VIRT_MACHINE(obj); - bool acpi_enabled = virt_is_acpi_enabled(vms); assert(code == BUS_MCEERR_AR || code == BUS_MCEERR_AO); - if (acpi_enabled && addr && - object_property_get_bool(obj, "ras", NULL)) { + if (acpi_ghes_present() && addr) { ram_addr = qemu_ram_addr_from_host(addr); if (ram_addr != RAM_ADDR_INVALID && kvm_physical_memory_addr_from_host(c->kvm_state, addr, &paddr)) { -- 2.20.1 These days the Arm architecture has a wide range of fine-grained optional extra architectural features. We implement quite a lot of these but by no means all of them. Document what we do implement, so that users can find out without having to dig through back-issues of our Changelog on the wiki. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Message-id: 20210617140328.28622-1-peter.maydell@linaro.org Reviewed-by: Richard Henderson <richard.henderson@linaro.org> --- docs/system/arm/emulation.rst | 102 ++++++++++++++++++++++++++++++++++ docs/system/target-arm.rst | 6 ++ 2 files changed, 108 insertions(+) create mode 100644 docs/system/arm/emulation.rst diff --git a/docs/system/arm/emulation.rst b/docs/system/arm/emulation.rst new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/docs/system/arm/emulation.rst @@ -XXX,XX +XXX,XX @@ +A-profile CPU architecture support +================================== + +QEMU's TCG emulation includes support for the Armv5, Armv6, Armv7 and +Armv8 versions of the A-profile architecture. It also has support for +the following architecture extensions: + +- FEAT_AA32BF16 (AArch32 BFloat16 instructions) +- FEAT_AA32HPD (AArch32 hierarchical permission disables) +- FEAT_AA32I8MM (AArch32 Int8 matrix multiplication instructions) +- FEAT_AES (AESD and AESE instructions) +- FEAT_BF16 (AArch64 BFloat16 instructions) +- FEAT_BTI (Branch Target Identification) +- FEAT_DIT (Data Independent Timing instructions) +- FEAT_DPB (DC CVAP instruction) +- FEAT_DotProd (Advanced SIMD dot product instructions) +- FEAT_FCMA (Floating-point complex number instructions) +- FEAT_FHM (Floating-point half-precision multiplication instructions) +- FEAT_FP16 (Half-precision floating-point data processing) +- FEAT_FRINTTS (Floating-point to integer instructions) +- FEAT_FlagM (Flag manipulation instructions v2) +- FEAT_FlagM2 (Enhancements to flag manipulation instructions) +- FEAT_HPDS (Hierarchical permission disables) +- FEAT_I8MM (AArch64 Int8 matrix multiplication instructions) +- FEAT_JSCVT (JavaScript conversion instructions) +- FEAT_LOR (Limited ordering regions) +- FEAT_LRCPC (Load-acquire RCpc instructions) +- FEAT_LRCPC2 (Load-acquire RCpc instructions v2) +- FEAT_LSE (Large System Extensions) +- FEAT_MTE (Memory Tagging Extension) +- FEAT_MTE2 (Memory Tagging Extension) +- FEAT_PAN (Privileged access never) +- FEAT_PAN2 (AT S1E1R and AT S1E1W instruction variants affected by PSTATE.PAN) +- FEAT_PAuth (Pointer authentication) +- FEAT_PMULL (PMULL, PMULL2 instructions) +- FEAT_PMUv3p1 (PMU Extensions v3.1) +- FEAT_PMUv3p4 (PMU Extensions v3.4) +- FEAT_RDM (Advanced SIMD rounding double multiply accumulate instructions) +- FEAT_RNG (Random number generator) +- FEAT_SB (Speculation Barrier) +- FEAT_SEL2 (Secure EL2) +- FEAT_SHA1 (SHA1 instructions) +- FEAT_SHA256 (SHA256 instructions) +- FEAT_SHA3 (Advanced SIMD SHA3 instructions) +- FEAT_SHA512 (Advanced SIMD SHA512 instructions) +- FEAT_SM3 (Advanced SIMD SM3 instructions) +- FEAT_SM4 (Advanced SIMD SM4 instructions) +- FEAT_SPECRES (Speculation restriction instructions) +- FEAT_SSBS (Speculative Store Bypass Safe) +- FEAT_TLBIOS (TLB invalidate instructions in Outer Shareable domain) +- FEAT_TLBIRANGE (TLB invalidate range instructions) +- FEAT_TTCNP (Translation table Common not private translations) +- FEAT_TTST (Small translation tables) +- FEAT_UAO (Unprivileged Access Override control) +- FEAT_VHE (Virtualization Host Extensions) +- FEAT_VMID16 (16-bit VMID) +- FEAT_XNX (Translation table stage 2 Unprivileged Execute-never) +- SVE (The Scalable Vector Extension) +- SVE2 (The Scalable Vector Extension v2) + +For information on the specifics of these extensions, please refer +to the `Armv8-A Arm Architecture Reference Manual +<https://developer.arm.com/documentation/ddi0487/latest>`_. + +When a specific named CPU is being emulated, only those features which +are present in hardware for that CPU are emulated. (If a feature is +not in the list above then it is not supported, even if the real +hardware should have it.) The ``max`` CPU enables all features. + +R-profile CPU architecture support +================================== + +QEMU's TCG emulation support for R-profile CPUs is currently limited. +We emulate only the Cortex-R5 and Cortex-R5F CPUs. + +M-profile CPU architecture support +================================== + +QEMU's TCG emulation includes support for Armv6-M, Armv7-M, Armv8-M, and +Armv8.1-M versions of the M-profile architucture. It also has support +for the following architecture extensions: + +- FP (Floating-point Extension) +- FPCXT (FPCXT access instructions) +- HP (Half-precision floating-point instructions) +- LOB (Low Overhead loops and Branch future) +- M (Main Extension) +- MPU (Memory Protection Unit Extension) +- PXN (Privileged Execute Never) +- RAS (Reliability, Serviceability and Availability): "minimum RAS Extension" only +- S (Security Extension) +- ST (System Timer Extension) + +For information on the specifics of these extensions, please refer +to the `Armv8-M Arm Architecture Reference Manual +<https://developer.arm.com/documentation/ddi0553/latest>`_. + +When a specific named CPU is being emulated, only those features which +are present in hardware for that CPU are emulated. (If a feature is +not in the list above then it is not supported, even if the real +hardware should have it.) There is no equivalent of the ``max`` CPU for +M-profile. diff --git a/docs/system/target-arm.rst b/docs/system/target-arm.rst index XXXXXXX..XXXXXXX 100644 --- a/docs/system/target-arm.rst +++ b/docs/system/target-arm.rst @@ -XXX,XX +XXX,XX @@ undocumented; you can get a complete list by running arm/virt arm/xlnx-versal-virt +Emulated CPU architecture support +================================= + +.. toctree:: + arm/emulation + Arm CPU features ================ -- 2.20.1 In the code for handling VFP system register accesses there is some stray whitespace after a unary '-' operator, and also some incorrect indent in a couple of function prototypes. We're about to move this code to another file, so fix the code style issues first so checkpatch doesn't complain about the code-movement patch. Cc: qemu-stable@nongnu.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210618141019.10671-2-peter.maydell@linaro.org --- target/arm/translate-vfp.c | 11 +++++------ 1 file changed, 5 insertions(+), 6 deletions(-) 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 void gen_branch_fpInactive(DisasContext *s, TCGCond cond, } static bool gen_M_fp_sysreg_write(DisasContext *s, int regno, - fp_sysreg_loadfn *loadfn, - void *opaque) + void *opaque) { /* Do a write to an M-profile floating point system register */ TCGv_i32 tmp; @@ -XXX,XX +XXX,XX @@ static bool gen_M_fp_sysreg_write(DisasContext *s, int regno, } static bool gen_M_fp_sysreg_read(DisasContext *s, int regno, - fp_sysreg_storefn *storefn, - void *opaque) + fp_sysreg_storefn *storefn, + void *opaque) { /* Do a read from an M-profile floating point system register */ TCGv_i32 tmp; @@ -XXX,XX +XXX,XX @@ static void fp_sysreg_to_memory(DisasContext *s, void *opaque, TCGv_i32 value) TCGv_i32 addr; if (!a->a) { - offset = - offset; + offset = -offset; } addr = load_reg(s, a->rn); @@ -XXX,XX +XXX,XX @@ static TCGv_i32 memory_to_fp_sysreg(DisasContext *s, void *opaque) TCGv_i32 value = tcg_temp_new_i32(); if (!a->a) { - offset = - offset; + offset = -offset; } addr = load_reg(s, a->rn); -- 2.20.1 If the guest makes an FPCXT_NS access when the FPU is disabled, one of two things happens: * if there is no active FP context, then the insn behaves the same way as if the FPU was enabled: writes ignored, reads same value as FPDSCR_NS * if there is an active FP context, then we take a NOCP exception Add code to the sysreg read/write functions which emits code to take the NOCP exception in the latter case. At the moment this will never be used, because the NOCP checks in m-nocp.decode happen first, and so the trans functions are never called when the FPU is disabled. The code will be needed when we move the sysreg access insns to before the NOCP patterns in the following commit. Cc: qemu-stable@nongnu.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210618141019.10671-3-peter.maydell@linaro.org --- target/arm/translate-vfp.c | 32 ++++++++++++++++++++++++++++++-- 1 file changed, 30 insertions(+), 2 deletions(-) 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 gen_M_fp_sysreg_write(DisasContext *s, int regno, lab_end = gen_new_label(); /* fpInactive case: write is a NOP, so branch to end */ gen_branch_fpInactive(s, TCG_COND_NE, lab_end); - /* !fpInactive: PreserveFPState(), and reads same as FPCXT_S */ + /* + * !fpInactive: if FPU disabled, take NOCP exception; + * otherwise PreserveFPState(), and then FPCXT_NS writes + * behave the same as FPCXT_S writes. + */ + if (s->fp_excp_el) { + gen_exception_insn(s, s->pc_curr, EXCP_NOCP, + syn_uncategorized(), s->fp_excp_el); + /* + * This was only a conditional exception, so override + * gen_exception_insn()'s default to DISAS_NORETURN + */ + s->base.is_jmp = DISAS_NEXT; + break; + } gen_preserve_fp_state(s); /* fall through */ case ARM_VFP_FPCXT_S: @@ -XXX,XX +XXX,XX @@ static bool gen_M_fp_sysreg_read(DisasContext *s, int regno, tcg_gen_br(lab_end); gen_set_label(lab_active); - /* !fpInactive: Reads the same as FPCXT_S, but side effects differ */ + /* + * !fpInactive: if FPU disabled, take NOCP exception; + * otherwise PreserveFPState(), and then FPCXT_NS + * reads the same as FPCXT_S. + */ + if (s->fp_excp_el) { + gen_exception_insn(s, s->pc_curr, EXCP_NOCP, + syn_uncategorized(), s->fp_excp_el); + /* + * This was only a conditional exception, so override + * gen_exception_insn()'s default to DISAS_NORETURN + */ + s->base.is_jmp = DISAS_NEXT; + break; + } gen_preserve_fp_state(s); tmp = tcg_temp_new_i32(); sfpa = tcg_temp_new_i32(); -- 2.20.1 The M-profile architecture requires that accesses to FPCXT_NS when there is no active FP state must not take a NOCP fault even if the FPU is disabled. We were not implementing this correctly, because in our decode we catch the NOCP faults early in m-nocp.decode. Fix this bug by moving all the handling of M-profile FP system register accesses from vfp.decode into m-nocp.decode and putting it above the NOCP blocks. This provides the correct behaviour: * for accesses other than FPCXT_NS the trans functions call vfp_access_check(), which will check for FPU disabled and raise a NOCP exception if necessary * for FPCXT_NS we have the special case code that doesn't call vfp_access_check() * when these trans functions want to raise an UNDEF they return false, so the decoder will fall through into the NOCP blocks. This means that NOCP correctly takes precedence over UNDEF for these insns. (This is a difference from the other insns handled by m-nocp.decode, where UNDEF takes precedence and which we implement by having those trans functions call unallocated_encoding() in the appropriate places.) [Note for backport to stable: this commit has a semantic dependency on commit 9a486856e9173af, which was not marked as cc-stable because we didn't know we'd need it for a for-stable bugfix.] Cc: qemu-stable@nongnu.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210618141019.10671-4-peter.maydell@linaro.org --- target/arm/translate-a32.h | 1 + target/arm/m-nocp.decode | 24 ++ target/arm/vfp.decode | 14 - target/arm/translate-m-nocp.c | 514 +++++++++++++++++++++++++++++++++ target/arm/translate-vfp.c | 517 +--------------------------------- 5 files changed, 542 insertions(+), 528 deletions(-) diff --git a/target/arm/translate-a32.h b/target/arm/translate-a32.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a32.h +++ b/target/arm/translate-a32.h @@ -XXX,XX +XXX,XX @@ bool disas_neon_shared(DisasContext *s, uint32_t insn); void load_reg_var(DisasContext *s, TCGv_i32 var, int reg); void arm_gen_condlabel(DisasContext *s); bool vfp_access_check(DisasContext *s); +void gen_preserve_fp_state(DisasContext *s); void read_neon_element32(TCGv_i32 dest, int reg, int ele, MemOp memop); void read_neon_element64(TCGv_i64 dest, int reg, int ele, MemOp memop); void write_neon_element32(TCGv_i32 src, int reg, int ele, MemOp memop); diff --git a/target/arm/m-nocp.decode b/target/arm/m-nocp.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/m-nocp.decode +++ b/target/arm/m-nocp.decode @@ -XXX,XX +XXX,XX @@ &nocp cp +# M-profile VLDR/VSTR to sysreg +%vldr_sysreg 22:1 13:3 +%imm7_0x4 0:7 !function=times_4 + +&vldr_sysreg rn reg imm a w p +@vldr_sysreg .... ... . a:1 . . . rn:4 ... . ... .. ....... \ + reg=%vldr_sysreg imm=%imm7_0x4 &vldr_sysreg + { # Special cases which do not take an early NOCP: VLLDM and VLSTM VLLDM_VLSTM 1110 1100 001 l:1 rn:4 0000 1010 op:1 000 0000 @@ -XXX,XX +XXX,XX @@ VSCCLRM 1110 1100 1.01 1111 .... 1011 imm:7 0 vd=%vd_dp size=3 VSCCLRM 1110 1100 1.01 1111 .... 1010 imm:8 vd=%vd_sp size=2 + # FP system register accesses: these are a special case because accesses + # to FPCXT_NS succeed even if the FPU is disabled. We therefore need + # to handle them before the big NOCP blocks. Note that within these + # insns NOCP still has higher priority than UNDEFs; this is implemented + # by their returning 'false' for UNDEF so as to fall through into the + # NOCP check (in contrast to VLLDM etc, which call unallocated_encoding() + # for the UNDEFs there that must take precedence over NOCP.) + + VMSR_VMRS ---- 1110 111 l:1 reg:4 rt:4 1010 0001 0000 + + # P=0 W=0 is SEE "Related encodings", so split into two patterns + VLDR_sysreg ---- 110 1 . . w:1 1 .... ... 0 111 11 ....... @vldr_sysreg p=1 + VLDR_sysreg ---- 110 0 . . 1 1 .... ... 0 111 11 ....... @vldr_sysreg p=0 w=1 + VSTR_sysreg ---- 110 1 . . w:1 0 .... ... 0 111 11 ....... @vldr_sysreg p=1 + VSTR_sysreg ---- 110 0 . . 1 0 .... ... 0 111 11 ....... @vldr_sysreg p=0 w=1 + NOCP 111- 1110 ---- ---- ---- cp:4 ---- ---- &nocp NOCP 111- 110- ---- ---- ---- cp:4 ---- ---- &nocp # From v8.1M onwards this range will also NOCP: diff --git a/target/arm/vfp.decode b/target/arm/vfp.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/vfp.decode +++ b/target/arm/vfp.decode @@ -XXX,XX +XXX,XX @@ VLDR_VSTR_hp ---- 1101 u:1 .0 l:1 rn:4 .... 1001 imm:8 vd=%vd_sp VLDR_VSTR_sp ---- 1101 u:1 .0 l:1 rn:4 .... 1010 imm:8 vd=%vd_sp VLDR_VSTR_dp ---- 1101 u:1 .0 l:1 rn:4 .... 1011 imm:8 vd=%vd_dp -# M-profile VLDR/VSTR to sysreg -%vldr_sysreg 22:1 13:3 -%imm7_0x4 0:7 !function=times_4 - -&vldr_sysreg rn reg imm a w p -@vldr_sysreg .... ... . a:1 . . . rn:4 ... . ... .. ....... \ - reg=%vldr_sysreg imm=%imm7_0x4 &vldr_sysreg - -# P=0 W=0 is SEE "Related encodings", so split into two patterns -VLDR_sysreg ---- 110 1 . . w:1 1 .... ... 0 111 11 ....... @vldr_sysreg p=1 -VLDR_sysreg ---- 110 0 . . 1 1 .... ... 0 111 11 ....... @vldr_sysreg p=0 w=1 -VSTR_sysreg ---- 110 1 . . w:1 0 .... ... 0 111 11 ....... @vldr_sysreg p=1 -VSTR_sysreg ---- 110 0 . . 1 0 .... ... 0 111 11 ....... @vldr_sysreg p=0 w=1 - # We split the load/store multiple up into two patterns to avoid # overlap with other insns in the "Advanced SIMD load/store and 64-bit move" # grouping: diff --git a/target/arm/translate-m-nocp.c b/target/arm/translate-m-nocp.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-m-nocp.c +++ b/target/arm/translate-m-nocp.c @@ -XXX,XX +XXX,XX @@ #include "qemu/osdep.h" #include "tcg/tcg-op.h" +#include "tcg/tcg-op-gvec.h" #include "translate.h" #include "translate-a32.h" @@ -XXX,XX +XXX,XX @@ static bool trans_VSCCLRM(DisasContext *s, arg_VSCCLRM *a) return true; } +/* + * M-profile provides two different sets of instructions that can + * access floating point system registers: VMSR/VMRS (which move + * to/from a general purpose register) and VLDR/VSTR sysreg (which + * move directly to/from memory). In some cases there are also side + * effects which must happen after any write to memory (which could + * cause an exception). So we implement the common logic for the + * sysreg access in gen_M_fp_sysreg_write() and gen_M_fp_sysreg_read(), + * which take pointers to callback functions which will perform the + * actual "read/write general purpose register" and "read/write + * memory" operations. + */ + +/* + * Emit code to store the sysreg to its final destination; frees the + * TCG temp 'value' it is passed. + */ +typedef void fp_sysreg_storefn(DisasContext *s, void *opaque, TCGv_i32 value); +/* + * Emit code to load the value to be copied to the sysreg; returns + * a new TCG temporary + */ +typedef TCGv_i32 fp_sysreg_loadfn(DisasContext *s, void *opaque); + +/* Common decode/access checks for fp sysreg read/write */ +typedef enum FPSysRegCheckResult { + FPSysRegCheckFailed, /* caller should return false */ + FPSysRegCheckDone, /* caller should return true */ + FPSysRegCheckContinue, /* caller should continue generating code */ +} FPSysRegCheckResult; + +static FPSysRegCheckResult fp_sysreg_checks(DisasContext *s, int regno) +{ + if (!dc_isar_feature(aa32_fpsp_v2, s) && !dc_isar_feature(aa32_mve, s)) { + return FPSysRegCheckFailed; + } + + switch (regno) { + case ARM_VFP_FPSCR: + case QEMU_VFP_FPSCR_NZCV: + break; + case ARM_VFP_FPSCR_NZCVQC: + if (!arm_dc_feature(s, ARM_FEATURE_V8_1M)) { + return FPSysRegCheckFailed; + } + break; + case ARM_VFP_FPCXT_S: + case ARM_VFP_FPCXT_NS: + if (!arm_dc_feature(s, ARM_FEATURE_V8_1M)) { + return FPSysRegCheckFailed; + } + if (!s->v8m_secure) { + return FPSysRegCheckFailed; + } + break; + case ARM_VFP_VPR: + case ARM_VFP_P0: + if (!dc_isar_feature(aa32_mve, s)) { + return FPSysRegCheckFailed; + } + break; + default: + return FPSysRegCheckFailed; + } + + /* + * FPCXT_NS is a special case: it has specific handling for + * "current FP state is inactive", and must do the PreserveFPState() + * but not the usual full set of actions done by ExecuteFPCheck(). + * So we don't call vfp_access_check() and the callers must handle this. + */ + if (regno != ARM_VFP_FPCXT_NS && !vfp_access_check(s)) { + return FPSysRegCheckDone; + } + return FPSysRegCheckContinue; +} + +static void gen_branch_fpInactive(DisasContext *s, TCGCond cond, + TCGLabel *label) +{ + /* + * FPCXT_NS is a special case: it has specific handling for + * "current FP state is inactive", and must do the PreserveFPState() + * but not the usual full set of actions done by ExecuteFPCheck(). + * We don't have a TB flag that matches the fpInactive check, so we + * do it at runtime as we don't expect FPCXT_NS accesses to be frequent. + * + * Emit code that checks fpInactive and does a conditional + * branch to label based on it: + * if cond is TCG_COND_NE then branch if fpInactive != 0 (ie if inactive) + * if cond is TCG_COND_EQ then branch if fpInactive == 0 (ie if active) + */ + assert(cond == TCG_COND_EQ || cond == TCG_COND_NE); + + /* fpInactive = FPCCR_NS.ASPEN == 1 && CONTROL.FPCA == 0 */ + TCGv_i32 aspen, fpca; + aspen = load_cpu_field(v7m.fpccr[M_REG_NS]); + fpca = load_cpu_field(v7m.control[M_REG_S]); + tcg_gen_andi_i32(aspen, aspen, R_V7M_FPCCR_ASPEN_MASK); + tcg_gen_xori_i32(aspen, aspen, R_V7M_FPCCR_ASPEN_MASK); + tcg_gen_andi_i32(fpca, fpca, R_V7M_CONTROL_FPCA_MASK); + tcg_gen_or_i32(fpca, fpca, aspen); + tcg_gen_brcondi_i32(tcg_invert_cond(cond), fpca, 0, label); + tcg_temp_free_i32(aspen); + tcg_temp_free_i32(fpca); +} + +static bool gen_M_fp_sysreg_write(DisasContext *s, int regno, + fp_sysreg_loadfn *loadfn, + void *opaque) +{ + /* Do a write to an M-profile floating point system register */ + TCGv_i32 tmp; + TCGLabel *lab_end = NULL; + + switch (fp_sysreg_checks(s, regno)) { + case FPSysRegCheckFailed: + return false; + case FPSysRegCheckDone: + return true; + case FPSysRegCheckContinue: + break; + } + + switch (regno) { + case ARM_VFP_FPSCR: + tmp = loadfn(s, opaque); + gen_helper_vfp_set_fpscr(cpu_env, tmp); + tcg_temp_free_i32(tmp); + gen_lookup_tb(s); + break; + case ARM_VFP_FPSCR_NZCVQC: + { + TCGv_i32 fpscr; + tmp = loadfn(s, opaque); + if (dc_isar_feature(aa32_mve, s)) { + /* QC is only present for MVE; otherwise RES0 */ + TCGv_i32 qc = tcg_temp_new_i32(); + tcg_gen_andi_i32(qc, tmp, FPCR_QC); + /* + * The 4 vfp.qc[] fields need only be "zero" vs "non-zero"; + * here writing the same value into all elements is simplest. + */ + tcg_gen_gvec_dup_i32(MO_32, offsetof(CPUARMState, vfp.qc), + 16, 16, qc); + } + tcg_gen_andi_i32(tmp, tmp, FPCR_NZCV_MASK); + fpscr = load_cpu_field(vfp.xregs[ARM_VFP_FPSCR]); + tcg_gen_andi_i32(fpscr, fpscr, ~FPCR_NZCV_MASK); + tcg_gen_or_i32(fpscr, fpscr, tmp); + store_cpu_field(fpscr, vfp.xregs[ARM_VFP_FPSCR]); + tcg_temp_free_i32(tmp); + break; + } + case ARM_VFP_FPCXT_NS: + lab_end = gen_new_label(); + /* fpInactive case: write is a NOP, so branch to end */ + gen_branch_fpInactive(s, TCG_COND_NE, lab_end); + /* + * !fpInactive: if FPU disabled, take NOCP exception; + * otherwise PreserveFPState(), and then FPCXT_NS writes + * behave the same as FPCXT_S writes. + */ + if (s->fp_excp_el) { + gen_exception_insn(s, s->pc_curr, EXCP_NOCP, + syn_uncategorized(), s->fp_excp_el); + /* + * This was only a conditional exception, so override + * gen_exception_insn()'s default to DISAS_NORETURN + */ + s->base.is_jmp = DISAS_NEXT; + break; + } + gen_preserve_fp_state(s); + /* fall through */ + case ARM_VFP_FPCXT_S: + { + TCGv_i32 sfpa, control; + /* + * Set FPSCR and CONTROL.SFPA from value; the new FPSCR takes + * bits [27:0] from value and zeroes bits [31:28]. + */ + tmp = loadfn(s, opaque); + sfpa = tcg_temp_new_i32(); + tcg_gen_shri_i32(sfpa, tmp, 31); + control = load_cpu_field(v7m.control[M_REG_S]); + tcg_gen_deposit_i32(control, control, sfpa, + R_V7M_CONTROL_SFPA_SHIFT, 1); + store_cpu_field(control, v7m.control[M_REG_S]); + tcg_gen_andi_i32(tmp, tmp, ~FPCR_NZCV_MASK); + gen_helper_vfp_set_fpscr(cpu_env, tmp); + tcg_temp_free_i32(tmp); + tcg_temp_free_i32(sfpa); + break; + } + case ARM_VFP_VPR: + /* Behaves as NOP if not privileged */ + if (IS_USER(s)) { + break; + } + tmp = loadfn(s, opaque); + store_cpu_field(tmp, v7m.vpr); + break; + case ARM_VFP_P0: + { + TCGv_i32 vpr; + tmp = loadfn(s, opaque); + vpr = load_cpu_field(v7m.vpr); + tcg_gen_deposit_i32(vpr, vpr, tmp, + R_V7M_VPR_P0_SHIFT, R_V7M_VPR_P0_LENGTH); + store_cpu_field(vpr, v7m.vpr); + tcg_temp_free_i32(tmp); + break; + } + default: + g_assert_not_reached(); + } + if (lab_end) { + gen_set_label(lab_end); + } + return true; +} + +static bool gen_M_fp_sysreg_read(DisasContext *s, int regno, + fp_sysreg_storefn *storefn, + void *opaque) +{ + /* Do a read from an M-profile floating point system register */ + TCGv_i32 tmp; + TCGLabel *lab_end = NULL; + bool lookup_tb = false; + + switch (fp_sysreg_checks(s, regno)) { + case FPSysRegCheckFailed: + return false; + case FPSysRegCheckDone: + return true; + case FPSysRegCheckContinue: + break; + } + + if (regno == ARM_VFP_FPSCR_NZCVQC && !dc_isar_feature(aa32_mve, s)) { + /* QC is RES0 without MVE, so NZCVQC simplifies to NZCV */ + regno = QEMU_VFP_FPSCR_NZCV; + } + + switch (regno) { + case ARM_VFP_FPSCR: + tmp = tcg_temp_new_i32(); + gen_helper_vfp_get_fpscr(tmp, cpu_env); + storefn(s, opaque, tmp); + break; + case ARM_VFP_FPSCR_NZCVQC: + tmp = tcg_temp_new_i32(); + gen_helper_vfp_get_fpscr(tmp, cpu_env); + tcg_gen_andi_i32(tmp, tmp, FPCR_NZCVQC_MASK); + storefn(s, opaque, tmp); + break; + case QEMU_VFP_FPSCR_NZCV: + /* + * Read just NZCV; this is a special case to avoid the + * helper call for the "VMRS to CPSR.NZCV" insn. + */ + tmp = load_cpu_field(vfp.xregs[ARM_VFP_FPSCR]); + tcg_gen_andi_i32(tmp, tmp, FPCR_NZCV_MASK); + storefn(s, opaque, tmp); + break; + case ARM_VFP_FPCXT_S: + { + TCGv_i32 control, sfpa, fpscr; + /* Bits [27:0] from FPSCR, bit [31] from CONTROL.SFPA */ + tmp = tcg_temp_new_i32(); + sfpa = tcg_temp_new_i32(); + gen_helper_vfp_get_fpscr(tmp, cpu_env); + tcg_gen_andi_i32(tmp, tmp, ~FPCR_NZCV_MASK); + control = load_cpu_field(v7m.control[M_REG_S]); + tcg_gen_andi_i32(sfpa, control, R_V7M_CONTROL_SFPA_MASK); + tcg_gen_shli_i32(sfpa, sfpa, 31 - R_V7M_CONTROL_SFPA_SHIFT); + tcg_gen_or_i32(tmp, tmp, sfpa); + tcg_temp_free_i32(sfpa); + /* + * Store result before updating FPSCR etc, in case + * it is a memory write which causes an exception. + */ + storefn(s, opaque, tmp); + /* + * Now we must reset FPSCR from FPDSCR_NS, and clear + * CONTROL.SFPA; so we'll end the TB here. + */ + tcg_gen_andi_i32(control, control, ~R_V7M_CONTROL_SFPA_MASK); + store_cpu_field(control, v7m.control[M_REG_S]); + fpscr = load_cpu_field(v7m.fpdscr[M_REG_NS]); + gen_helper_vfp_set_fpscr(cpu_env, fpscr); + tcg_temp_free_i32(fpscr); + lookup_tb = true; + break; + } + case ARM_VFP_FPCXT_NS: + { + TCGv_i32 control, sfpa, fpscr, fpdscr, zero; + TCGLabel *lab_active = gen_new_label(); + + lookup_tb = true; + + gen_branch_fpInactive(s, TCG_COND_EQ, lab_active); + /* fpInactive case: reads as FPDSCR_NS */ + TCGv_i32 tmp = load_cpu_field(v7m.fpdscr[M_REG_NS]); + storefn(s, opaque, tmp); + lab_end = gen_new_label(); + tcg_gen_br(lab_end); + + gen_set_label(lab_active); + /* + * !fpInactive: if FPU disabled, take NOCP exception; + * otherwise PreserveFPState(), and then FPCXT_NS + * reads the same as FPCXT_S. + */ + if (s->fp_excp_el) { + gen_exception_insn(s, s->pc_curr, EXCP_NOCP, + syn_uncategorized(), s->fp_excp_el); + /* + * This was only a conditional exception, so override + * gen_exception_insn()'s default to DISAS_NORETURN + */ + s->base.is_jmp = DISAS_NEXT; + break; + } + gen_preserve_fp_state(s); + tmp = tcg_temp_new_i32(); + sfpa = tcg_temp_new_i32(); + fpscr = tcg_temp_new_i32(); + gen_helper_vfp_get_fpscr(fpscr, cpu_env); + tcg_gen_andi_i32(tmp, fpscr, ~FPCR_NZCV_MASK); + control = load_cpu_field(v7m.control[M_REG_S]); + tcg_gen_andi_i32(sfpa, control, R_V7M_CONTROL_SFPA_MASK); + tcg_gen_shli_i32(sfpa, sfpa, 31 - R_V7M_CONTROL_SFPA_SHIFT); + tcg_gen_or_i32(tmp, tmp, sfpa); + tcg_temp_free_i32(control); + /* Store result before updating FPSCR, in case it faults */ + storefn(s, opaque, tmp); + /* If SFPA is zero then set FPSCR from FPDSCR_NS */ + fpdscr = load_cpu_field(v7m.fpdscr[M_REG_NS]); + zero = tcg_const_i32(0); + tcg_gen_movcond_i32(TCG_COND_EQ, fpscr, sfpa, zero, fpdscr, fpscr); + gen_helper_vfp_set_fpscr(cpu_env, fpscr); + tcg_temp_free_i32(zero); + tcg_temp_free_i32(sfpa); + tcg_temp_free_i32(fpdscr); + tcg_temp_free_i32(fpscr); + break; + } + case ARM_VFP_VPR: + /* Behaves as NOP if not privileged */ + if (IS_USER(s)) { + break; + } + tmp = load_cpu_field(v7m.vpr); + storefn(s, opaque, tmp); + break; + case ARM_VFP_P0: + tmp = load_cpu_field(v7m.vpr); + tcg_gen_extract_i32(tmp, tmp, R_V7M_VPR_P0_SHIFT, R_V7M_VPR_P0_LENGTH); + storefn(s, opaque, tmp); + break; + default: + g_assert_not_reached(); + } + + if (lab_end) { + gen_set_label(lab_end); + } + if (lookup_tb) { + gen_lookup_tb(s); + } + return true; +} + +static void fp_sysreg_to_gpr(DisasContext *s, void *opaque, TCGv_i32 value) +{ + arg_VMSR_VMRS *a = opaque; + + if (a->rt == 15) { + /* Set the 4 flag bits in the CPSR */ + gen_set_nzcv(value); + tcg_temp_free_i32(value); + } else { + store_reg(s, a->rt, value); + } +} + +static TCGv_i32 gpr_to_fp_sysreg(DisasContext *s, void *opaque) +{ + arg_VMSR_VMRS *a = opaque; + + return load_reg(s, a->rt); +} + +static bool trans_VMSR_VMRS(DisasContext *s, arg_VMSR_VMRS *a) +{ + /* + * Accesses to R15 are UNPREDICTABLE; we choose to undef. + * FPSCR -> r15 is a special case which writes to the PSR flags; + * set a->reg to a special value to tell gen_M_fp_sysreg_read() + * we only care about the top 4 bits of FPSCR there. + */ + if (a->rt == 15) { + if (a->l && a->reg == ARM_VFP_FPSCR) { + a->reg = QEMU_VFP_FPSCR_NZCV; + } else { + return false; + } + } + + if (a->l) { + /* VMRS, move FP system register to gp register */ + return gen_M_fp_sysreg_read(s, a->reg, fp_sysreg_to_gpr, a); + } else { + /* VMSR, move gp register to FP system register */ + return gen_M_fp_sysreg_write(s, a->reg, gpr_to_fp_sysreg, a); + } +} + +static void fp_sysreg_to_memory(DisasContext *s, void *opaque, TCGv_i32 value) +{ + arg_vldr_sysreg *a = opaque; + uint32_t offset = a->imm; + TCGv_i32 addr; + + if (!a->a) { + offset = -offset; + } + + addr = load_reg(s, a->rn); + if (a->p) { + tcg_gen_addi_i32(addr, addr, offset); + } + + if (s->v8m_stackcheck && a->rn == 13 && a->w) { + gen_helper_v8m_stackcheck(cpu_env, addr); + } + + gen_aa32_st_i32(s, value, addr, get_mem_index(s), + MO_UL | MO_ALIGN | s->be_data); + tcg_temp_free_i32(value); + + if (a->w) { + /* writeback */ + if (!a->p) { + tcg_gen_addi_i32(addr, addr, offset); + } + store_reg(s, a->rn, addr); + } else { + tcg_temp_free_i32(addr); + } +} + +static TCGv_i32 memory_to_fp_sysreg(DisasContext *s, void *opaque) +{ + arg_vldr_sysreg *a = opaque; + uint32_t offset = a->imm; + TCGv_i32 addr; + TCGv_i32 value = tcg_temp_new_i32(); + + if (!a->a) { + offset = -offset; + } + + addr = load_reg(s, a->rn); + if (a->p) { + tcg_gen_addi_i32(addr, addr, offset); + } + + if (s->v8m_stackcheck && a->rn == 13 && a->w) { + gen_helper_v8m_stackcheck(cpu_env, addr); + } + + gen_aa32_ld_i32(s, value, addr, get_mem_index(s), + MO_UL | MO_ALIGN | s->be_data); + + if (a->w) { + /* writeback */ + if (!a->p) { + tcg_gen_addi_i32(addr, addr, offset); + } + store_reg(s, a->rn, addr); + } else { + tcg_temp_free_i32(addr); + } + return value; +} + +static bool trans_VLDR_sysreg(DisasContext *s, arg_vldr_sysreg *a) +{ + if (!arm_dc_feature(s, ARM_FEATURE_V8_1M)) { + return false; + } + if (a->rn == 15) { + return false; + } + return gen_M_fp_sysreg_write(s, a->reg, memory_to_fp_sysreg, a); +} + +static bool trans_VSTR_sysreg(DisasContext *s, arg_vldr_sysreg *a) +{ + if (!arm_dc_feature(s, ARM_FEATURE_V8_1M)) { + return false; + } + if (a->rn == 15) { + return false; + } + return gen_M_fp_sysreg_read(s, a->reg, fp_sysreg_to_memory, a); +} + static bool trans_NOCP(DisasContext *s, arg_nocp *a) { /* 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 inline long vfp_f16_offset(unsigned reg, bool top) * Generate code for M-profile lazy FP state preservation if needed; * this corresponds to the pseudocode PreserveFPState() function. */ -static void gen_preserve_fp_state(DisasContext *s) +void gen_preserve_fp_state(DisasContext *s) { if (s->v7m_lspact) { /* @@ -XXX,XX +XXX,XX @@ static bool trans_VDUP(DisasContext *s, arg_VDUP *a) return true; } -/* - * M-profile provides two different sets of instructions that can - * access floating point system registers: VMSR/VMRS (which move - * to/from a general purpose register) and VLDR/VSTR sysreg (which - * move directly to/from memory). In some cases there are also side - * effects which must happen after any write to memory (which could - * cause an exception). So we implement the common logic for the - * sysreg access in gen_M_fp_sysreg_write() and gen_M_fp_sysreg_read(), - * which take pointers to callback functions which will perform the - * actual "read/write general purpose register" and "read/write - * memory" operations. - */ - -/* - * Emit code to store the sysreg to its final destination; frees the - * TCG temp 'value' it is passed. - */ -typedef void fp_sysreg_storefn(DisasContext *s, void *opaque, TCGv_i32 value); -/* - * Emit code to load the value to be copied to the sysreg; returns - * a new TCG temporary - */ -typedef TCGv_i32 fp_sysreg_loadfn(DisasContext *s, void *opaque); - -/* Common decode/access checks for fp sysreg read/write */ -typedef enum FPSysRegCheckResult { - FPSysRegCheckFailed, /* caller should return false */ - FPSysRegCheckDone, /* caller should return true */ - FPSysRegCheckContinue, /* caller should continue generating code */ -} FPSysRegCheckResult; - -static FPSysRegCheckResult fp_sysreg_checks(DisasContext *s, int regno) -{ - if (!dc_isar_feature(aa32_fpsp_v2, s) && !dc_isar_feature(aa32_mve, s)) { - return FPSysRegCheckFailed; - } - - switch (regno) { - case ARM_VFP_FPSCR: - case QEMU_VFP_FPSCR_NZCV: - break; - case ARM_VFP_FPSCR_NZCVQC: - if (!arm_dc_feature(s, ARM_FEATURE_V8_1M)) { - return FPSysRegCheckFailed; - } - break; - case ARM_VFP_FPCXT_S: - case ARM_VFP_FPCXT_NS: - if (!arm_dc_feature(s, ARM_FEATURE_V8_1M)) { - return FPSysRegCheckFailed; - } - if (!s->v8m_secure) { - return FPSysRegCheckFailed; - } - break; - case ARM_VFP_VPR: - case ARM_VFP_P0: - if (!dc_isar_feature(aa32_mve, s)) { - return FPSysRegCheckFailed; - } - break; - default: - return FPSysRegCheckFailed; - } - - /* - * FPCXT_NS is a special case: it has specific handling for - * "current FP state is inactive", and must do the PreserveFPState() - * but not the usual full set of actions done by ExecuteFPCheck(). - * So we don't call vfp_access_check() and the callers must handle this. - */ - if (regno != ARM_VFP_FPCXT_NS && !vfp_access_check(s)) { - return FPSysRegCheckDone; - } - return FPSysRegCheckContinue; -} - -static void gen_branch_fpInactive(DisasContext *s, TCGCond cond, - TCGLabel *label) -{ - /* - * FPCXT_NS is a special case: it has specific handling for - * "current FP state is inactive", and must do the PreserveFPState() - * but not the usual full set of actions done by ExecuteFPCheck(). - * We don't have a TB flag that matches the fpInactive check, so we - * do it at runtime as we don't expect FPCXT_NS accesses to be frequent. - * - * Emit code that checks fpInactive and does a conditional - * branch to label based on it: - * if cond is TCG_COND_NE then branch if fpInactive != 0 (ie if inactive) - * if cond is TCG_COND_EQ then branch if fpInactive == 0 (ie if active) - */ - assert(cond == TCG_COND_EQ || cond == TCG_COND_NE); - - /* fpInactive = FPCCR_NS.ASPEN == 1 && CONTROL.FPCA == 0 */ - TCGv_i32 aspen, fpca; - aspen = load_cpu_field(v7m.fpccr[M_REG_NS]); - fpca = load_cpu_field(v7m.control[M_REG_S]); - tcg_gen_andi_i32(aspen, aspen, R_V7M_FPCCR_ASPEN_MASK); - tcg_gen_xori_i32(aspen, aspen, R_V7M_FPCCR_ASPEN_MASK); - tcg_gen_andi_i32(fpca, fpca, R_V7M_CONTROL_FPCA_MASK); - tcg_gen_or_i32(fpca, fpca, aspen); - tcg_gen_brcondi_i32(tcg_invert_cond(cond), fpca, 0, label); - tcg_temp_free_i32(aspen); - tcg_temp_free_i32(fpca); -} - -static bool gen_M_fp_sysreg_write(DisasContext *s, int regno, - fp_sysreg_loadfn *loadfn, - void *opaque) -{ - /* Do a write to an M-profile floating point system register */ - TCGv_i32 tmp; - TCGLabel *lab_end = NULL; - - switch (fp_sysreg_checks(s, regno)) { - case FPSysRegCheckFailed: - return false; - case FPSysRegCheckDone: - return true; - case FPSysRegCheckContinue: - break; - } - - switch (regno) { - case ARM_VFP_FPSCR: - tmp = loadfn(s, opaque); - gen_helper_vfp_set_fpscr(cpu_env, tmp); - tcg_temp_free_i32(tmp); - gen_lookup_tb(s); - break; - case ARM_VFP_FPSCR_NZCVQC: - { - TCGv_i32 fpscr; - tmp = loadfn(s, opaque); - if (dc_isar_feature(aa32_mve, s)) { - /* QC is only present for MVE; otherwise RES0 */ - TCGv_i32 qc = tcg_temp_new_i32(); - tcg_gen_andi_i32(qc, tmp, FPCR_QC); - /* - * The 4 vfp.qc[] fields need only be "zero" vs "non-zero"; - * here writing the same value into all elements is simplest. - */ - tcg_gen_gvec_dup_i32(MO_32, offsetof(CPUARMState, vfp.qc), - 16, 16, qc); - } - tcg_gen_andi_i32(tmp, tmp, FPCR_NZCV_MASK); - fpscr = load_cpu_field(vfp.xregs[ARM_VFP_FPSCR]); - tcg_gen_andi_i32(fpscr, fpscr, ~FPCR_NZCV_MASK); - tcg_gen_or_i32(fpscr, fpscr, tmp); - store_cpu_field(fpscr, vfp.xregs[ARM_VFP_FPSCR]); - tcg_temp_free_i32(tmp); - break; - } - case ARM_VFP_FPCXT_NS: - lab_end = gen_new_label(); - /* fpInactive case: write is a NOP, so branch to end */ - gen_branch_fpInactive(s, TCG_COND_NE, lab_end); - /* - * !fpInactive: if FPU disabled, take NOCP exception; - * otherwise PreserveFPState(), and then FPCXT_NS writes - * behave the same as FPCXT_S writes. - */ - if (s->fp_excp_el) { - gen_exception_insn(s, s->pc_curr, EXCP_NOCP, - syn_uncategorized(), s->fp_excp_el); - /* - * This was only a conditional exception, so override - * gen_exception_insn()'s default to DISAS_NORETURN - */ - s->base.is_jmp = DISAS_NEXT; - break; - } - gen_preserve_fp_state(s); - /* fall through */ - case ARM_VFP_FPCXT_S: - { - TCGv_i32 sfpa, control; - /* - * Set FPSCR and CONTROL.SFPA from value; the new FPSCR takes - * bits [27:0] from value and zeroes bits [31:28]. - */ - tmp = loadfn(s, opaque); - sfpa = tcg_temp_new_i32(); - tcg_gen_shri_i32(sfpa, tmp, 31); - control = load_cpu_field(v7m.control[M_REG_S]); - tcg_gen_deposit_i32(control, control, sfpa, - R_V7M_CONTROL_SFPA_SHIFT, 1); - store_cpu_field(control, v7m.control[M_REG_S]); - tcg_gen_andi_i32(tmp, tmp, ~FPCR_NZCV_MASK); - gen_helper_vfp_set_fpscr(cpu_env, tmp); - tcg_temp_free_i32(tmp); - tcg_temp_free_i32(sfpa); - break; - } - case ARM_VFP_VPR: - /* Behaves as NOP if not privileged */ - if (IS_USER(s)) { - break; - } - tmp = loadfn(s, opaque); - store_cpu_field(tmp, v7m.vpr); - break; - case ARM_VFP_P0: - { - TCGv_i32 vpr; - tmp = loadfn(s, opaque); - vpr = load_cpu_field(v7m.vpr); - tcg_gen_deposit_i32(vpr, vpr, tmp, - R_V7M_VPR_P0_SHIFT, R_V7M_VPR_P0_LENGTH); - store_cpu_field(vpr, v7m.vpr); - tcg_temp_free_i32(tmp); - break; - } - default: - g_assert_not_reached(); - } - if (lab_end) { - gen_set_label(lab_end); - } - return true; -} - -static bool gen_M_fp_sysreg_read(DisasContext *s, int regno, - fp_sysreg_storefn *storefn, - void *opaque) -{ - /* Do a read from an M-profile floating point system register */ - TCGv_i32 tmp; - TCGLabel *lab_end = NULL; - bool lookup_tb = false; - - switch (fp_sysreg_checks(s, regno)) { - case FPSysRegCheckFailed: - return false; - case FPSysRegCheckDone: - return true; - case FPSysRegCheckContinue: - break; - } - - if (regno == ARM_VFP_FPSCR_NZCVQC && !dc_isar_feature(aa32_mve, s)) { - /* QC is RES0 without MVE, so NZCVQC simplifies to NZCV */ - regno = QEMU_VFP_FPSCR_NZCV; - } - - switch (regno) { - case ARM_VFP_FPSCR: - tmp = tcg_temp_new_i32(); - gen_helper_vfp_get_fpscr(tmp, cpu_env); - storefn(s, opaque, tmp); - break; - case ARM_VFP_FPSCR_NZCVQC: - tmp = tcg_temp_new_i32(); - gen_helper_vfp_get_fpscr(tmp, cpu_env); - tcg_gen_andi_i32(tmp, tmp, FPCR_NZCVQC_MASK); - storefn(s, opaque, tmp); - break; - case QEMU_VFP_FPSCR_NZCV: - /* - * Read just NZCV; this is a special case to avoid the - * helper call for the "VMRS to CPSR.NZCV" insn. - */ - tmp = load_cpu_field(vfp.xregs[ARM_VFP_FPSCR]); - tcg_gen_andi_i32(tmp, tmp, FPCR_NZCV_MASK); - storefn(s, opaque, tmp); - break; - case ARM_VFP_FPCXT_S: - { - TCGv_i32 control, sfpa, fpscr; - /* Bits [27:0] from FPSCR, bit [31] from CONTROL.SFPA */ - tmp = tcg_temp_new_i32(); - sfpa = tcg_temp_new_i32(); - gen_helper_vfp_get_fpscr(tmp, cpu_env); - tcg_gen_andi_i32(tmp, tmp, ~FPCR_NZCV_MASK); - control = load_cpu_field(v7m.control[M_REG_S]); - tcg_gen_andi_i32(sfpa, control, R_V7M_CONTROL_SFPA_MASK); - tcg_gen_shli_i32(sfpa, sfpa, 31 - R_V7M_CONTROL_SFPA_SHIFT); - tcg_gen_or_i32(tmp, tmp, sfpa); - tcg_temp_free_i32(sfpa); - /* - * Store result before updating FPSCR etc, in case - * it is a memory write which causes an exception. - */ - storefn(s, opaque, tmp); - /* - * Now we must reset FPSCR from FPDSCR_NS, and clear - * CONTROL.SFPA; so we'll end the TB here. - */ - tcg_gen_andi_i32(control, control, ~R_V7M_CONTROL_SFPA_MASK); - store_cpu_field(control, v7m.control[M_REG_S]); - fpscr = load_cpu_field(v7m.fpdscr[M_REG_NS]); - gen_helper_vfp_set_fpscr(cpu_env, fpscr); - tcg_temp_free_i32(fpscr); - lookup_tb = true; - break; - } - case ARM_VFP_FPCXT_NS: - { - TCGv_i32 control, sfpa, fpscr, fpdscr, zero; - TCGLabel *lab_active = gen_new_label(); - - lookup_tb = true; - - gen_branch_fpInactive(s, TCG_COND_EQ, lab_active); - /* fpInactive case: reads as FPDSCR_NS */ - TCGv_i32 tmp = load_cpu_field(v7m.fpdscr[M_REG_NS]); - storefn(s, opaque, tmp); - lab_end = gen_new_label(); - tcg_gen_br(lab_end); - - gen_set_label(lab_active); - /* - * !fpInactive: if FPU disabled, take NOCP exception; - * otherwise PreserveFPState(), and then FPCXT_NS - * reads the same as FPCXT_S. - */ - if (s->fp_excp_el) { - gen_exception_insn(s, s->pc_curr, EXCP_NOCP, - syn_uncategorized(), s->fp_excp_el); - /* - * This was only a conditional exception, so override - * gen_exception_insn()'s default to DISAS_NORETURN - */ - s->base.is_jmp = DISAS_NEXT; - break; - } - gen_preserve_fp_state(s); - tmp = tcg_temp_new_i32(); - sfpa = tcg_temp_new_i32(); - fpscr = tcg_temp_new_i32(); - gen_helper_vfp_get_fpscr(fpscr, cpu_env); - tcg_gen_andi_i32(tmp, fpscr, ~FPCR_NZCV_MASK); - control = load_cpu_field(v7m.control[M_REG_S]); - tcg_gen_andi_i32(sfpa, control, R_V7M_CONTROL_SFPA_MASK); - tcg_gen_shli_i32(sfpa, sfpa, 31 - R_V7M_CONTROL_SFPA_SHIFT); - tcg_gen_or_i32(tmp, tmp, sfpa); - tcg_temp_free_i32(control); - /* Store result before updating FPSCR, in case it faults */ - storefn(s, opaque, tmp); - /* If SFPA is zero then set FPSCR from FPDSCR_NS */ - fpdscr = load_cpu_field(v7m.fpdscr[M_REG_NS]); - zero = tcg_const_i32(0); - tcg_gen_movcond_i32(TCG_COND_EQ, fpscr, sfpa, zero, fpdscr, fpscr); - gen_helper_vfp_set_fpscr(cpu_env, fpscr); - tcg_temp_free_i32(zero); - tcg_temp_free_i32(sfpa); - tcg_temp_free_i32(fpdscr); - tcg_temp_free_i32(fpscr); - break; - } - case ARM_VFP_VPR: - /* Behaves as NOP if not privileged */ - if (IS_USER(s)) { - break; - } - tmp = load_cpu_field(v7m.vpr); - storefn(s, opaque, tmp); - break; - case ARM_VFP_P0: - tmp = load_cpu_field(v7m.vpr); - tcg_gen_extract_i32(tmp, tmp, R_V7M_VPR_P0_SHIFT, R_V7M_VPR_P0_LENGTH); - storefn(s, opaque, tmp); - break; - default: - g_assert_not_reached(); - } - - if (lab_end) { - gen_set_label(lab_end); - } - if (lookup_tb) { - gen_lookup_tb(s); - } - return true; -} - -static void fp_sysreg_to_gpr(DisasContext *s, void *opaque, TCGv_i32 value) -{ - arg_VMSR_VMRS *a = opaque; - - if (a->rt == 15) { - /* Set the 4 flag bits in the CPSR */ - gen_set_nzcv(value); - tcg_temp_free_i32(value); - } else { - store_reg(s, a->rt, value); - } -} - -static TCGv_i32 gpr_to_fp_sysreg(DisasContext *s, void *opaque) -{ - arg_VMSR_VMRS *a = opaque; - - return load_reg(s, a->rt); -} - -static bool gen_M_VMSR_VMRS(DisasContext *s, arg_VMSR_VMRS *a) -{ - /* - * Accesses to R15 are UNPREDICTABLE; we choose to undef. - * FPSCR -> r15 is a special case which writes to the PSR flags; - * set a->reg to a special value to tell gen_M_fp_sysreg_read() - * we only care about the top 4 bits of FPSCR there. - */ - if (a->rt == 15) { - if (a->l && a->reg == ARM_VFP_FPSCR) { - a->reg = QEMU_VFP_FPSCR_NZCV; - } else { - return false; - } - } - - if (a->l) { - /* VMRS, move FP system register to gp register */ - return gen_M_fp_sysreg_read(s, a->reg, fp_sysreg_to_gpr, a); - } else { - /* VMSR, move gp register to FP system register */ - return gen_M_fp_sysreg_write(s, a->reg, gpr_to_fp_sysreg, a); - } -} - static bool trans_VMSR_VMRS(DisasContext *s, arg_VMSR_VMRS *a) { TCGv_i32 tmp; bool ignore_vfp_enabled = false; if (arm_dc_feature(s, ARM_FEATURE_M)) { - return gen_M_VMSR_VMRS(s, a); + /* M profile version was already handled in m-nocp.decode */ + return false; } if (!dc_isar_feature(aa32_fpsp_v2, s)) { @@ -XXX,XX +XXX,XX @@ static bool trans_VMSR_VMRS(DisasContext *s, arg_VMSR_VMRS *a) return true; } -static void fp_sysreg_to_memory(DisasContext *s, void *opaque, TCGv_i32 value) -{ - arg_vldr_sysreg *a = opaque; - uint32_t offset = a->imm; - TCGv_i32 addr; - - if (!a->a) { - offset = -offset; - } - - addr = load_reg(s, a->rn); - if (a->p) { - tcg_gen_addi_i32(addr, addr, offset); - } - - if (s->v8m_stackcheck && a->rn == 13 && a->w) { - gen_helper_v8m_stackcheck(cpu_env, addr); - } - - gen_aa32_st_i32(s, value, addr, get_mem_index(s), - MO_UL | MO_ALIGN | s->be_data); - tcg_temp_free_i32(value); - - if (a->w) { - /* writeback */ - if (!a->p) { - tcg_gen_addi_i32(addr, addr, offset); - } - store_reg(s, a->rn, addr); - } else { - tcg_temp_free_i32(addr); - } -} - -static TCGv_i32 memory_to_fp_sysreg(DisasContext *s, void *opaque) -{ - arg_vldr_sysreg *a = opaque; - uint32_t offset = a->imm; - TCGv_i32 addr; - TCGv_i32 value = tcg_temp_new_i32(); - - if (!a->a) { - offset = -offset; - } - - addr = load_reg(s, a->rn); - if (a->p) { - tcg_gen_addi_i32(addr, addr, offset); - } - - if (s->v8m_stackcheck && a->rn == 13 && a->w) { - gen_helper_v8m_stackcheck(cpu_env, addr); - } - - gen_aa32_ld_i32(s, value, addr, get_mem_index(s), - MO_UL | MO_ALIGN | s->be_data); - - if (a->w) { - /* writeback */ - if (!a->p) { - tcg_gen_addi_i32(addr, addr, offset); - } - store_reg(s, a->rn, addr); - } else { - tcg_temp_free_i32(addr); - } - return value; -} - -static bool trans_VLDR_sysreg(DisasContext *s, arg_vldr_sysreg *a) -{ - if (!arm_dc_feature(s, ARM_FEATURE_V8_1M)) { - return false; - } - if (a->rn == 15) { - return false; - } - return gen_M_fp_sysreg_write(s, a->reg, memory_to_fp_sysreg, a); -} - -static bool trans_VSTR_sysreg(DisasContext *s, arg_vldr_sysreg *a) -{ - if (!arm_dc_feature(s, ARM_FEATURE_V8_1M)) { - return false; - } - if (a->rn == 15) { - return false; - } - return gen_M_fp_sysreg_read(s, a->reg, fp_sysreg_to_memory, a); -} static bool trans_VMOV_half(DisasContext *s, arg_VMOV_single *a) { -- 2.20.1 A few subcases of VLDR/VSTR sysreg succeed but do not perform a memory access: * VSTR of VPR when unprivileged * VLDR to VPR when unprivileged * VLDR to FPCXT_NS when fpInactive In these cases, even though we don't do the memory access we should still update the base register and perform the stack limit check if the insn's addressing mode specifies writeback. Our implementation failed to do this, because we handle these side-effects inside the memory_to_fp_sysreg() and fp_sysreg_to_memory() callback functions, which are only called if there's something to load or store. Fix this by adding an extra argument to the callbacks which is set to true to actually perform the access and false to only do side effects like writeback, and calling the callback with do_access = false for the three cases listed above. This produces slightly suboptimal code for the case of a write to FPCXT_NS when the FPU is inactive and the insn didn't have side effects (ie no writeback, or via VMSR), in which case we'll generate a conditional branch over an unconditional branch. But this doesn't seem to be important enough to merit requiring the callback to report back whether it generated any code or not. Cc: qemu-stable@nongnu.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210618141019.10671-5-peter.maydell@linaro.org --- target/arm/translate-m-nocp.c | 102 ++++++++++++++++++++++++---------- 1 file changed, 72 insertions(+), 30 deletions(-) diff --git a/target/arm/translate-m-nocp.c b/target/arm/translate-m-nocp.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-m-nocp.c +++ b/target/arm/translate-m-nocp.c @@ -XXX,XX +XXX,XX @@ static bool trans_VSCCLRM(DisasContext *s, arg_VSCCLRM *a) /* * Emit code to store the sysreg to its final destination; frees the - * TCG temp 'value' it is passed. + * TCG temp 'value' it is passed. do_access is true to do the store, + * and false to skip it and only perform side-effects like base + * register writeback. */ -typedef void fp_sysreg_storefn(DisasContext *s, void *opaque, TCGv_i32 value); +typedef void fp_sysreg_storefn(DisasContext *s, void *opaque, TCGv_i32 value, + bool do_access); /* * Emit code to load the value to be copied to the sysreg; returns - * a new TCG temporary + * a new TCG temporary. do_access is true to do the store, + * and false to skip it and only perform side-effects like base + * register writeback. */ -typedef TCGv_i32 fp_sysreg_loadfn(DisasContext *s, void *opaque); +typedef TCGv_i32 fp_sysreg_loadfn(DisasContext *s, void *opaque, + bool do_access); /* Common decode/access checks for fp sysreg read/write */ typedef enum FPSysRegCheckResult { @@ -XXX,XX +XXX,XX @@ static bool gen_M_fp_sysreg_write(DisasContext *s, int regno, switch (regno) { case ARM_VFP_FPSCR: - tmp = loadfn(s, opaque); + tmp = loadfn(s, opaque, true); gen_helper_vfp_set_fpscr(cpu_env, tmp); tcg_temp_free_i32(tmp); gen_lookup_tb(s); @@ -XXX,XX +XXX,XX @@ static bool gen_M_fp_sysreg_write(DisasContext *s, int regno, case ARM_VFP_FPSCR_NZCVQC: { TCGv_i32 fpscr; - tmp = loadfn(s, opaque); + tmp = loadfn(s, opaque, true); if (dc_isar_feature(aa32_mve, s)) { /* QC is only present for MVE; otherwise RES0 */ TCGv_i32 qc = tcg_temp_new_i32(); @@ -XXX,XX +XXX,XX @@ static bool gen_M_fp_sysreg_write(DisasContext *s, int regno, break; } case ARM_VFP_FPCXT_NS: + { + TCGLabel *lab_active = gen_new_label(); + lab_end = gen_new_label(); - /* fpInactive case: write is a NOP, so branch to end */ - gen_branch_fpInactive(s, TCG_COND_NE, lab_end); + gen_branch_fpInactive(s, TCG_COND_EQ, lab_active); + /* + * fpInactive case: write is a NOP, so only do side effects + * like register writeback before we branch to end + */ + loadfn(s, opaque, false); + tcg_gen_br(lab_end); + + gen_set_label(lab_active); /* * !fpInactive: if FPU disabled, take NOCP exception; * otherwise PreserveFPState(), and then FPCXT_NS writes @@ -XXX,XX +XXX,XX @@ static bool gen_M_fp_sysreg_write(DisasContext *s, int regno, break; } gen_preserve_fp_state(s); - /* fall through */ + } + /* fall through */ case ARM_VFP_FPCXT_S: { TCGv_i32 sfpa, control; @@ -XXX,XX +XXX,XX @@ static bool gen_M_fp_sysreg_write(DisasContext *s, int regno, * Set FPSCR and CONTROL.SFPA from value; the new FPSCR takes * bits [27:0] from value and zeroes bits [31:28]. */ - tmp = loadfn(s, opaque); + tmp = loadfn(s, opaque, true); sfpa = tcg_temp_new_i32(); tcg_gen_shri_i32(sfpa, tmp, 31); control = load_cpu_field(v7m.control[M_REG_S]); @@ -XXX,XX +XXX,XX @@ static bool gen_M_fp_sysreg_write(DisasContext *s, int regno, case ARM_VFP_VPR: /* Behaves as NOP if not privileged */ if (IS_USER(s)) { + loadfn(s, opaque, false); break; } - tmp = loadfn(s, opaque); + tmp = loadfn(s, opaque, true); store_cpu_field(tmp, v7m.vpr); break; case ARM_VFP_P0: { TCGv_i32 vpr; - tmp = loadfn(s, opaque); + tmp = loadfn(s, opaque, true); vpr = load_cpu_field(v7m.vpr); tcg_gen_deposit_i32(vpr, vpr, tmp, R_V7M_VPR_P0_SHIFT, R_V7M_VPR_P0_LENGTH); @@ -XXX,XX +XXX,XX @@ static bool gen_M_fp_sysreg_read(DisasContext *s, int regno, case ARM_VFP_FPSCR: tmp = tcg_temp_new_i32(); gen_helper_vfp_get_fpscr(tmp, cpu_env); - storefn(s, opaque, tmp); + storefn(s, opaque, tmp, true); break; case ARM_VFP_FPSCR_NZCVQC: tmp = tcg_temp_new_i32(); gen_helper_vfp_get_fpscr(tmp, cpu_env); tcg_gen_andi_i32(tmp, tmp, FPCR_NZCVQC_MASK); - storefn(s, opaque, tmp); + storefn(s, opaque, tmp, true); break; case QEMU_VFP_FPSCR_NZCV: /* @@ -XXX,XX +XXX,XX @@ static bool gen_M_fp_sysreg_read(DisasContext *s, int regno, */ tmp = load_cpu_field(vfp.xregs[ARM_VFP_FPSCR]); tcg_gen_andi_i32(tmp, tmp, FPCR_NZCV_MASK); - storefn(s, opaque, tmp); + storefn(s, opaque, tmp, true); break; case ARM_VFP_FPCXT_S: { @@ -XXX,XX +XXX,XX @@ static bool gen_M_fp_sysreg_read(DisasContext *s, int regno, * Store result before updating FPSCR etc, in case * it is a memory write which causes an exception. */ - storefn(s, opaque, tmp); + storefn(s, opaque, tmp, true); /* * Now we must reset FPSCR from FPDSCR_NS, and clear * CONTROL.SFPA; so we'll end the TB here. @@ -XXX,XX +XXX,XX @@ static bool gen_M_fp_sysreg_read(DisasContext *s, int regno, gen_branch_fpInactive(s, TCG_COND_EQ, lab_active); /* fpInactive case: reads as FPDSCR_NS */ TCGv_i32 tmp = load_cpu_field(v7m.fpdscr[M_REG_NS]); - storefn(s, opaque, tmp); + storefn(s, opaque, tmp, true); lab_end = gen_new_label(); tcg_gen_br(lab_end); @@ -XXX,XX +XXX,XX @@ static bool gen_M_fp_sysreg_read(DisasContext *s, int regno, tcg_gen_or_i32(tmp, tmp, sfpa); tcg_temp_free_i32(control); /* Store result before updating FPSCR, in case it faults */ - storefn(s, opaque, tmp); + storefn(s, opaque, tmp, true); /* If SFPA is zero then set FPSCR from FPDSCR_NS */ fpdscr = load_cpu_field(v7m.fpdscr[M_REG_NS]); zero = tcg_const_i32(0); @@ -XXX,XX +XXX,XX @@ static bool gen_M_fp_sysreg_read(DisasContext *s, int regno, case ARM_VFP_VPR: /* Behaves as NOP if not privileged */ if (IS_USER(s)) { + storefn(s, opaque, NULL, false); break; } tmp = load_cpu_field(v7m.vpr); - storefn(s, opaque, tmp); + storefn(s, opaque, tmp, true); break; case ARM_VFP_P0: tmp = load_cpu_field(v7m.vpr); tcg_gen_extract_i32(tmp, tmp, R_V7M_VPR_P0_SHIFT, R_V7M_VPR_P0_LENGTH); - storefn(s, opaque, tmp); + storefn(s, opaque, tmp, true); break; default: g_assert_not_reached(); @@ -XXX,XX +XXX,XX @@ static bool gen_M_fp_sysreg_read(DisasContext *s, int regno, return true; } -static void fp_sysreg_to_gpr(DisasContext *s, void *opaque, TCGv_i32 value) +static void fp_sysreg_to_gpr(DisasContext *s, void *opaque, TCGv_i32 value, + bool do_access) { arg_VMSR_VMRS *a = opaque; + if (!do_access) { + return; + } + if (a->rt == 15) { /* Set the 4 flag bits in the CPSR */ gen_set_nzcv(value); @@ -XXX,XX +XXX,XX @@ static void fp_sysreg_to_gpr(DisasContext *s, void *opaque, TCGv_i32 value) } } -static TCGv_i32 gpr_to_fp_sysreg(DisasContext *s, void *opaque) +static TCGv_i32 gpr_to_fp_sysreg(DisasContext *s, void *opaque, bool do_access) { arg_VMSR_VMRS *a = opaque; + if (!do_access) { + return NULL; + } return load_reg(s, a->rt); } @@ -XXX,XX +XXX,XX @@ static bool trans_VMSR_VMRS(DisasContext *s, arg_VMSR_VMRS *a) } } -static void fp_sysreg_to_memory(DisasContext *s, void *opaque, TCGv_i32 value) +static void fp_sysreg_to_memory(DisasContext *s, void *opaque, TCGv_i32 value, + bool do_access) { arg_vldr_sysreg *a = opaque; uint32_t offset = a->imm; @@ -XXX,XX +XXX,XX @@ static void fp_sysreg_to_memory(DisasContext *s, void *opaque, TCGv_i32 value) offset = -offset; } + if (!do_access && !a->w) { + return; + } + addr = load_reg(s, a->rn); if (a->p) { tcg_gen_addi_i32(addr, addr, offset); @@ -XXX,XX +XXX,XX @@ static void fp_sysreg_to_memory(DisasContext *s, void *opaque, TCGv_i32 value) gen_helper_v8m_stackcheck(cpu_env, addr); } - gen_aa32_st_i32(s, value, addr, get_mem_index(s), - MO_UL | MO_ALIGN | s->be_data); - tcg_temp_free_i32(value); + if (do_access) { + gen_aa32_st_i32(s, value, addr, get_mem_index(s), + MO_UL | MO_ALIGN | s->be_data); + tcg_temp_free_i32(value); + } if (a->w) { /* writeback */ @@ -XXX,XX +XXX,XX @@ static void fp_sysreg_to_memory(DisasContext *s, void *opaque, TCGv_i32 value) } } -static TCGv_i32 memory_to_fp_sysreg(DisasContext *s, void *opaque) +static TCGv_i32 memory_to_fp_sysreg(DisasContext *s, void *opaque, + bool do_access) { arg_vldr_sysreg *a = opaque; uint32_t offset = a->imm; TCGv_i32 addr; - TCGv_i32 value = tcg_temp_new_i32(); + TCGv_i32 value = NULL; if (!a->a) { offset = -offset; } + if (!do_access && !a->w) { + return NULL; + } + addr = load_reg(s, a->rn); if (a->p) { tcg_gen_addi_i32(addr, addr, offset); @@ -XXX,XX +XXX,XX @@ static TCGv_i32 memory_to_fp_sysreg(DisasContext *s, void *opaque) gen_helper_v8m_stackcheck(cpu_env, addr); } - gen_aa32_ld_i32(s, value, addr, get_mem_index(s), - MO_UL | MO_ALIGN | s->be_data); + if (do_access) { + value = tcg_temp_new_i32(); + gen_aa32_ld_i32(s, value, addr, get_mem_index(s), + MO_UL | MO_ALIGN | s->be_data); + } if (a->w) { /* writeback */ -- 2.20.1 Factor the code in full_vfp_access_check() which updates the ownership of the FP context and creates a new FP context out into its own function. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210618141019.10671-6-peter.maydell@linaro.org --- target/arm/translate-vfp.c | 104 +++++++++++++++++++++---------------- 1 file changed, 58 insertions(+), 46 deletions(-) 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 @@ void gen_preserve_fp_state(DisasContext *s) } } +/* + * Generate code for M-profile FP context handling: update the + * ownership of the FP context, and create a new context if + * necessary. This corresponds to the parts of the pseudocode + * ExecuteFPCheck() after the inital PreserveFPState() call. + */ +static void gen_update_fp_context(DisasContext *s) +{ + /* Update ownership of FP context: set FPCCR.S to match current state */ + if (s->v8m_fpccr_s_wrong) { + TCGv_i32 tmp; + + tmp = load_cpu_field(v7m.fpccr[M_REG_S]); + if (s->v8m_secure) { + tcg_gen_ori_i32(tmp, tmp, R_V7M_FPCCR_S_MASK); + } else { + tcg_gen_andi_i32(tmp, tmp, ~R_V7M_FPCCR_S_MASK); + } + store_cpu_field(tmp, v7m.fpccr[M_REG_S]); + /* Don't need to do this for any further FP insns in this TB */ + s->v8m_fpccr_s_wrong = false; + } + + if (s->v7m_new_fp_ctxt_needed) { + /* + * Create new FP context by updating CONTROL.FPCA, CONTROL.SFPA, + * the FPSCR, and VPR. + */ + TCGv_i32 control, fpscr; + uint32_t bits = R_V7M_CONTROL_FPCA_MASK; + + fpscr = load_cpu_field(v7m.fpdscr[s->v8m_secure]); + gen_helper_vfp_set_fpscr(cpu_env, fpscr); + tcg_temp_free_i32(fpscr); + if (dc_isar_feature(aa32_mve, s)) { + TCGv_i32 z32 = tcg_const_i32(0); + store_cpu_field(z32, v7m.vpr); + } + + /* + * We don't need to arrange to end the TB, because the only + * parts of FPSCR which we cache in the TB flags are the VECLEN + * and VECSTRIDE, and those don't exist for M-profile. + */ + + if (s->v8m_secure) { + bits |= R_V7M_CONTROL_SFPA_MASK; + } + control = load_cpu_field(v7m.control[M_REG_S]); + tcg_gen_ori_i32(control, control, bits); + store_cpu_field(control, v7m.control[M_REG_S]); + /* Don't need to do this for any further FP insns in this TB */ + s->v7m_new_fp_ctxt_needed = false; + } +} + /* * Check that VFP access is enabled. If it is, do the necessary * M-profile lazy-FP handling and then return true. @@ -XXX,XX +XXX,XX @@ static bool full_vfp_access_check(DisasContext *s, bool ignore_vfp_enabled) /* Trigger lazy-state preservation if necessary */ gen_preserve_fp_state(s); - /* Update ownership of FP context: set FPCCR.S to match current state */ - if (s->v8m_fpccr_s_wrong) { - TCGv_i32 tmp; - - tmp = load_cpu_field(v7m.fpccr[M_REG_S]); - if (s->v8m_secure) { - tcg_gen_ori_i32(tmp, tmp, R_V7M_FPCCR_S_MASK); - } else { - tcg_gen_andi_i32(tmp, tmp, ~R_V7M_FPCCR_S_MASK); - } - store_cpu_field(tmp, v7m.fpccr[M_REG_S]); - /* Don't need to do this for any further FP insns in this TB */ - s->v8m_fpccr_s_wrong = false; - } - - if (s->v7m_new_fp_ctxt_needed) { - /* - * Create new FP context by updating CONTROL.FPCA, CONTROL.SFPA, - * the FPSCR, and VPR. - */ - TCGv_i32 control, fpscr; - uint32_t bits = R_V7M_CONTROL_FPCA_MASK; - - fpscr = load_cpu_field(v7m.fpdscr[s->v8m_secure]); - gen_helper_vfp_set_fpscr(cpu_env, fpscr); - tcg_temp_free_i32(fpscr); - if (dc_isar_feature(aa32_mve, s)) { - TCGv_i32 z32 = tcg_const_i32(0); - store_cpu_field(z32, v7m.vpr); - } - - /* - * We don't need to arrange to end the TB, because the only - * parts of FPSCR which we cache in the TB flags are the VECLEN - * and VECSTRIDE, and those don't exist for M-profile. - */ - - if (s->v8m_secure) { - bits |= R_V7M_CONTROL_SFPA_MASK; - } - control = load_cpu_field(v7m.control[M_REG_S]); - tcg_gen_ori_i32(control, control, bits); - store_cpu_field(control, v7m.control[M_REG_S]); - /* Don't need to do this for any further FP insns in this TB */ - s->v7m_new_fp_ctxt_needed = false; - } + /* Update ownership of FP context and create new FP context if needed */ + gen_update_fp_context(s); } return true; -- 2.20.1 vfp_access_check and its helper routine full_vfp_access_check() has gradually grown and is now an awkward mix of A-profile only and M-profile only pieces. Refactor it into an A-profile only and an M-profile only version, taking advantage of the fact that now the only direct call to full_vfp_access_check() is in A-profile-only code. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210618141019.10671-7-peter.maydell@linaro.org --- target/arm/translate-vfp.c | 79 +++++++++++++++++++++++--------------- 1 file changed, 48 insertions(+), 31 deletions(-) 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 void gen_update_fp_context(DisasContext *s) } /* - * Check that VFP access is enabled. If it is, do the necessary - * M-profile lazy-FP handling and then return true. - * If not, emit code to generate an appropriate exception and - * return false. + * Check that VFP access is enabled, A-profile specific version. + * + * If VFP is enabled, return true. If not, emit code to generate an + * appropriate exception and return false. * The ignore_vfp_enabled argument specifies that we should ignore - * whether VFP is enabled via FPEXC[EN]: this should be true for FMXR/FMRX + * whether VFP is enabled via FPEXC.EN: this should be true for FMXR/FMRX * accesses to FPSID, FPEXC, MVFR0, MVFR1, MVFR2, and false for all other insns. */ -static bool full_vfp_access_check(DisasContext *s, bool ignore_vfp_enabled) +static bool vfp_access_check_a(DisasContext *s, bool ignore_vfp_enabled) { if (s->fp_excp_el) { - if (arm_dc_feature(s, ARM_FEATURE_M)) { - /* - * M-profile mostly catches the "FPU disabled" case early, in - * disas_m_nocp(), but a few insns (eg LCTP, WLSTP, DLSTP) - * which do coprocessor-checks are outside the large ranges of - * the encoding space handled by the patterns in m-nocp.decode, - * and for them we may need to raise NOCP here. - */ - gen_exception_insn(s, s->pc_curr, EXCP_NOCP, - syn_uncategorized(), s->fp_excp_el); - } else { - gen_exception_insn(s, s->pc_curr, EXCP_UDEF, - syn_fp_access_trap(1, 0xe, false), - s->fp_excp_el); - } + gen_exception_insn(s, s->pc_curr, EXCP_UDEF, + syn_fp_access_trap(1, 0xe, false), s->fp_excp_el); return false; } @@ -XXX,XX +XXX,XX @@ static bool full_vfp_access_check(DisasContext *s, bool ignore_vfp_enabled) unallocated_encoding(s); return false; } + return true; +} - if (arm_dc_feature(s, ARM_FEATURE_M)) { - /* Handle M-profile lazy FP state mechanics */ - - /* Trigger lazy-state preservation if necessary */ - gen_preserve_fp_state(s); - - /* Update ownership of FP context and create new FP context if needed */ - gen_update_fp_context(s); +/* + * Check that VFP access is enabled, M-profile specific version. + * + * If VFP is enabled, do the necessary M-profile lazy-FP handling and then + * return true. If not, emit code to generate an appropriate exception and + * return false. + */ +static bool vfp_access_check_m(DisasContext *s) +{ + if (s->fp_excp_el) { + /* + * M-profile mostly catches the "FPU disabled" case early, in + * disas_m_nocp(), but a few insns (eg LCTP, WLSTP, DLSTP) + * which do coprocessor-checks are outside the large ranges of + * the encoding space handled by the patterns in m-nocp.decode, + * and for them we may need to raise NOCP here. + */ + gen_exception_insn(s, s->pc_curr, EXCP_NOCP, + syn_uncategorized(), s->fp_excp_el); + return false; } + /* Handle M-profile lazy FP state mechanics */ + + /* Trigger lazy-state preservation if necessary */ + gen_preserve_fp_state(s); + + /* Update ownership of FP context and create new FP context if needed */ + gen_update_fp_context(s); + return true; } @@ -XXX,XX +XXX,XX @@ static bool full_vfp_access_check(DisasContext *s, bool ignore_vfp_enabled) */ bool vfp_access_check(DisasContext *s) { - return full_vfp_access_check(s, false); + if (arm_dc_feature(s, ARM_FEATURE_M)) { + return vfp_access_check_m(s); + } else { + return vfp_access_check_a(s, false); + } } static bool trans_VSEL(DisasContext *s, arg_VSEL *a) @@ -XXX,XX +XXX,XX @@ static bool trans_VMSR_VMRS(DisasContext *s, arg_VMSR_VMRS *a) return false; } - if (!full_vfp_access_check(s, ignore_vfp_enabled)) { + /* + * Call vfp_access_check_a() directly, because we need to tell + * it to ignore FPEXC.EN for some register accesses. + */ + if (!vfp_access_check_a(s, ignore_vfp_enabled)) { return true; } -- 2.20.1 Instead of open-coding the "take NOCP exception if FPU disabled, otherwise call gen_preserve_fp_state()" code in the accessors for FPCXT_NS, add an argument to vfp_access_check_m() which tells it to skip the gen_update_fp_context() call, so we can use it for the FPCXT_NS case. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210618141019.10671-8-peter.maydell@linaro.org --- target/arm/translate-a32.h | 2 +- target/arm/translate-m-nocp.c | 10 ++-------- target/arm/translate-vfp.c | 13 ++++++++----- 3 files changed, 11 insertions(+), 14 deletions(-) diff --git a/target/arm/translate-a32.h b/target/arm/translate-a32.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a32.h +++ b/target/arm/translate-a32.h @@ -XXX,XX +XXX,XX @@ bool disas_neon_shared(DisasContext *s, uint32_t insn); void load_reg_var(DisasContext *s, TCGv_i32 var, int reg); void arm_gen_condlabel(DisasContext *s); bool vfp_access_check(DisasContext *s); -void gen_preserve_fp_state(DisasContext *s); +bool vfp_access_check_m(DisasContext *s, bool skip_context_update); void read_neon_element32(TCGv_i32 dest, int reg, int ele, MemOp memop); void read_neon_element64(TCGv_i64 dest, int reg, int ele, MemOp memop); void write_neon_element32(TCGv_i32 src, int reg, int ele, MemOp memop); diff --git a/target/arm/translate-m-nocp.c b/target/arm/translate-m-nocp.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-m-nocp.c +++ b/target/arm/translate-m-nocp.c @@ -XXX,XX +XXX,XX @@ static bool gen_M_fp_sysreg_write(DisasContext *s, int regno, * otherwise PreserveFPState(), and then FPCXT_NS writes * behave the same as FPCXT_S writes. */ - if (s->fp_excp_el) { - gen_exception_insn(s, s->pc_curr, EXCP_NOCP, - syn_uncategorized(), s->fp_excp_el); + if (!vfp_access_check_m(s, true)) { /* * This was only a conditional exception, so override * gen_exception_insn()'s default to DISAS_NORETURN @@ -XXX,XX +XXX,XX @@ static bool gen_M_fp_sysreg_write(DisasContext *s, int regno, s->base.is_jmp = DISAS_NEXT; break; } - gen_preserve_fp_state(s); } /* fall through */ case ARM_VFP_FPCXT_S: @@ -XXX,XX +XXX,XX @@ static bool gen_M_fp_sysreg_read(DisasContext *s, int regno, * otherwise PreserveFPState(), and then FPCXT_NS * reads the same as FPCXT_S. */ - if (s->fp_excp_el) { - gen_exception_insn(s, s->pc_curr, EXCP_NOCP, - syn_uncategorized(), s->fp_excp_el); + if (!vfp_access_check_m(s, true)) { /* * This was only a conditional exception, so override * gen_exception_insn()'s default to DISAS_NORETURN @@ -XXX,XX +XXX,XX @@ static bool gen_M_fp_sysreg_read(DisasContext *s, int regno, s->base.is_jmp = DISAS_NEXT; break; } - gen_preserve_fp_state(s); tmp = tcg_temp_new_i32(); sfpa = tcg_temp_new_i32(); fpscr = tcg_temp_new_i32(); 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 inline long vfp_f16_offset(unsigned reg, bool top) * Generate code for M-profile lazy FP state preservation if needed; * this corresponds to the pseudocode PreserveFPState() function. */ -void gen_preserve_fp_state(DisasContext *s) +static void gen_preserve_fp_state(DisasContext *s) { if (s->v7m_lspact) { /* @@ -XXX,XX +XXX,XX @@ static bool vfp_access_check_a(DisasContext *s, bool ignore_vfp_enabled) * If VFP is enabled, do the necessary M-profile lazy-FP handling and then * return true. If not, emit code to generate an appropriate exception and * return false. + * skip_context_update is true to skip the "update FP context" part of this. */ -static bool vfp_access_check_m(DisasContext *s) +bool vfp_access_check_m(DisasContext *s, bool skip_context_update) { if (s->fp_excp_el) { /* @@ -XXX,XX +XXX,XX @@ static bool vfp_access_check_m(DisasContext *s) /* Trigger lazy-state preservation if necessary */ gen_preserve_fp_state(s); - /* Update ownership of FP context and create new FP context if needed */ - gen_update_fp_context(s); + if (!skip_context_update) { + /* Update ownership of FP context and create new FP context if needed */ + gen_update_fp_context(s); + } return true; } @@ -XXX,XX +XXX,XX @@ static bool vfp_access_check_m(DisasContext *s) bool vfp_access_check(DisasContext *s) { if (arm_dc_feature(s, ARM_FEATURE_M)) { - return vfp_access_check_m(s); + return vfp_access_check_m(s, false); } else { return vfp_access_check_a(s, false); } -- 2.20.1 Implement the forms of the MVE VLDR and VSTR insns which perform non-widening loads of bytes, halfwords or words from memory into vector elements of the same width (encodings T5, T6, T7). (At the moment we know for MVE and M-profile in general that vfp_access_check() can never return false, but we include the conventional return-true-on-failure check for consistency with non-M-profile translation code.) Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-2-peter.maydell@linaro.org --- target/arm/{translate-mve.c => helper-mve.h} | 19 +- target/arm/helper.h | 2 + target/arm/internals.h | 11 ++ target/arm/mve.decode | 22 +++ target/arm/mve_helper.c | 172 +++++++++++++++++++ target/arm/translate-mve.c | 119 +++++++++++++ target/arm/meson.build | 1 + 7 files changed, 334 insertions(+), 12 deletions(-) copy target/arm/{translate-mve.c => helper-mve.h} (61%) create mode 100644 target/arm/mve_helper.c diff --git a/target/arm/translate-mve.c b/target/arm/helper-mve.h similarity index 61% copy from target/arm/translate-mve.c copy to target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ /* - * ARM translation: M-profile MVE instructions + * M-profile MVE specific helper definitions * * Copyright (c) 2021 Linaro, Ltd. * @@ -XXX,XX +XXX,XX @@ * 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 "tcg/tcg-op.h" -#include "tcg/tcg-op-gvec.h" -#include "exec/exec-all.h" -#include "exec/gen-icount.h" -#include "translate.h" -#include "translate-a32.h" - -/* Include the generated decoder */ -#include "decode-mve.c.inc" +DEF_HELPER_FLAGS_3(mve_vldrb, TCG_CALL_NO_WG, void, env, ptr, i32) +DEF_HELPER_FLAGS_3(mve_vldrh, TCG_CALL_NO_WG, void, env, ptr, i32) +DEF_HELPER_FLAGS_3(mve_vldrw, TCG_CALL_NO_WG, void, env, ptr, i32) +DEF_HELPER_FLAGS_3(mve_vstrb, TCG_CALL_NO_WG, void, env, ptr, i32) +DEF_HELPER_FLAGS_3(mve_vstrh, TCG_CALL_NO_WG, void, env, ptr, i32) +DEF_HELPER_FLAGS_3(mve_vstrw, TCG_CALL_NO_WG, void, env, ptr, i32) 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_idx, TCG_CALL_NO_RWG, #include "helper-a64.h" #include "helper-sve.h" #endif + +#include "helper-mve.h" diff --git a/target/arm/internals.h b/target/arm/internals.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/internals.h +++ b/target/arm/internals.h @@ -XXX,XX +XXX,XX @@ static inline uint64_t useronly_maybe_clean_ptr(uint32_t desc, uint64_t ptr) return ptr; } +/* Values for M-profile PSR.ECI for MVE insns */ +enum MVEECIState { + ECI_NONE = 0, /* No completed beats */ + ECI_A0 = 1, /* Completed: A0 */ + ECI_A0A1 = 2, /* Completed: A0, A1 */ + /* 3 is reserved */ + ECI_A0A1A2 = 4, /* Completed: A0, A1, A2 */ + ECI_A0A1A2B0 = 5, /* Completed: A0, A1, A2, B0 */ + /* All other values reserved */ +}; + #endif diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ # # This file is processed by scripts/decodetree.py # + +%qd 22:1 13:3 + +&vldr_vstr rn qd imm p a w size l + +@vldr_vstr ....... . . . . l:1 rn:4 ... ...... imm:7 &vldr_vstr qd=%qd + +# Vector loads and stores + +# Non-widening loads/stores (P=0 W=0 is 'related encoding') +VLDR_VSTR 1110110 0 a:1 . 1 . .... ... 111100 ....... @vldr_vstr \ + size=0 p=0 w=1 +VLDR_VSTR 1110110 0 a:1 . 1 . .... ... 111101 ....... @vldr_vstr \ + size=1 p=0 w=1 +VLDR_VSTR 1110110 0 a:1 . 1 . .... ... 111110 ....... @vldr_vstr \ + size=2 p=0 w=1 +VLDR_VSTR 1110110 1 a:1 . w:1 . .... ... 111100 ....... @vldr_vstr \ + size=0 p=1 +VLDR_VSTR 1110110 1 a:1 . w:1 . .... ... 111101 ....... @vldr_vstr \ + size=1 p=1 +VLDR_VSTR 1110110 1 a:1 . w:1 . .... ... 111110 ....... @vldr_vstr \ + size=2 p=1 diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ +/* + * M-profile MVE Operations + * + * Copyright (c) 2021 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 "internals.h" +#include "vec_internal.h" +#include "exec/helper-proto.h" +#include "exec/cpu_ldst.h" +#include "exec/exec-all.h" + +static uint16_t mve_element_mask(CPUARMState *env) +{ + /* + * Return the mask of which elements in the MVE vector should be + * updated. This is a combination of multiple things: + * (1) by default, we update every lane in the vector + * (2) VPT predication stores its state in the VPR register; + * (3) low-overhead-branch tail predication will mask out part + * the vector on the final iteration of the loop + * (4) if EPSR.ECI is set then we must execute only some beats + * of the insn + * We combine all these into a 16-bit result with the same semantics + * as VPR.P0: 0 to mask the lane, 1 if it is active. + * 8-bit vector ops will look at all bits of the result; + * 16-bit ops will look at bits 0, 2, 4, ...; + * 32-bit ops will look at bits 0, 4, 8 and 12. + * Compare pseudocode GetCurInstrBeat(), though that only returns + * the 4-bit slice of the mask corresponding to a single beat. + */ + uint16_t mask = FIELD_EX32(env->v7m.vpr, V7M_VPR, P0); + + if (!(env->v7m.vpr & R_V7M_VPR_MASK01_MASK)) { + mask |= 0xff; + } + if (!(env->v7m.vpr & R_V7M_VPR_MASK23_MASK)) { + mask |= 0xff00; + } + + if (env->v7m.ltpsize < 4 && + env->regs[14] <= (1 << (4 - env->v7m.ltpsize))) { + /* + * Tail predication active, and this is the last loop iteration. + * The element size is (1 << ltpsize), and we only want to process + * loopcount elements, so we want to retain the least significant + * (loopcount * esize) predicate bits and zero out bits above that. + */ + int masklen = env->regs[14] << env->v7m.ltpsize; + assert(masklen <= 16); + mask &= MAKE_64BIT_MASK(0, masklen); + } + + if ((env->condexec_bits & 0xf) == 0) { + /* + * ECI bits indicate which beats are already executed; + * we handle this by effectively predicating them out. + */ + int eci = env->condexec_bits >> 4; + switch (eci) { + case ECI_NONE: + break; + case ECI_A0: + mask &= 0xfff0; + break; + case ECI_A0A1: + mask &= 0xff00; + break; + case ECI_A0A1A2: + case ECI_A0A1A2B0: + mask &= 0xf000; + break; + default: + g_assert_not_reached(); + } + } + + return mask; +} + +static void mve_advance_vpt(CPUARMState *env) +{ + /* Advance the VPT and ECI state if necessary */ + uint32_t vpr = env->v7m.vpr; + unsigned mask01, mask23; + + if ((env->condexec_bits & 0xf) == 0) { + env->condexec_bits = (env->condexec_bits == (ECI_A0A1A2B0 << 4)) ? + (ECI_A0 << 4) : (ECI_NONE << 4); + } + + if (!(vpr & (R_V7M_VPR_MASK01_MASK | R_V7M_VPR_MASK23_MASK))) { + /* VPT not enabled, nothing to do */ + return; + } + + mask01 = FIELD_EX32(vpr, V7M_VPR, MASK01); + mask23 = FIELD_EX32(vpr, V7M_VPR, MASK23); + if (mask01 > 8) { + /* high bit set, but not 0b1000: invert the relevant half of P0 */ + vpr ^= 0xff; + } + if (mask23 > 8) { + /* high bit set, but not 0b1000: invert the relevant half of P0 */ + vpr ^= 0xff00; + } + vpr = FIELD_DP32(vpr, V7M_VPR, MASK01, mask01 << 1); + vpr = FIELD_DP32(vpr, V7M_VPR, MASK23, mask23 << 1); + env->v7m.vpr = vpr; +} + + +#define DO_VLDR(OP, MSIZE, LDTYPE, ESIZE, TYPE) \ + void HELPER(mve_##OP)(CPUARMState *env, void *vd, uint32_t addr) \ + { \ + TYPE *d = vd; \ + uint16_t mask = mve_element_mask(env); \ + unsigned b, e; \ + /* \ + * R_SXTM allows the dest reg to become UNKNOWN for abandoned \ + * beats so we don't care if we update part of the dest and \ + * then take an exception. \ + */ \ + for (b = 0, e = 0; b < 16; b += ESIZE, e++) { \ + if (mask & (1 << b)) { \ + d[H##ESIZE(e)] = cpu_##LDTYPE##_data_ra(env, addr, GETPC()); \ + } \ + addr += MSIZE; \ + } \ + mve_advance_vpt(env); \ + } + +#define DO_VSTR(OP, MSIZE, STTYPE, ESIZE, TYPE) \ + void HELPER(mve_##OP)(CPUARMState *env, void *vd, uint32_t addr) \ + { \ + TYPE *d = vd; \ + uint16_t mask = mve_element_mask(env); \ + unsigned b, e; \ + for (b = 0, e = 0; b < 16; b += ESIZE, e++) { \ + if (mask & (1 << b)) { \ + cpu_##STTYPE##_data_ra(env, addr, d[H##ESIZE(e)], GETPC()); \ + } \ + addr += MSIZE; \ + } \ + mve_advance_vpt(env); \ + } + +DO_VLDR(vldrb, 1, ldub, 1, uint8_t) +DO_VLDR(vldrh, 2, lduw, 2, uint16_t) +DO_VLDR(vldrw, 4, ldl, 4, uint32_t) + +DO_VSTR(vstrb, 1, stb, 1, uint8_t) +DO_VSTR(vstrh, 2, stw, 2, uint16_t) +DO_VSTR(vstrw, 4, stl, 4, uint32_t) + +#undef DO_VLDR +#undef DO_VSTR diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ /* Include the generated decoder */ #include "decode-mve.c.inc" + +typedef void MVEGenLdStFn(TCGv_ptr, TCGv_ptr, TCGv_i32); + +/* Return the offset of a Qn register (same semantics as aa32_vfp_qreg()) */ +static inline long mve_qreg_offset(unsigned reg) +{ + return offsetof(CPUARMState, vfp.zregs[reg].d[0]); +} + +static TCGv_ptr mve_qreg_ptr(unsigned reg) +{ + TCGv_ptr ret = tcg_temp_new_ptr(); + tcg_gen_addi_ptr(ret, cpu_env, mve_qreg_offset(reg)); + return ret; +} + +static bool mve_check_qreg_bank(DisasContext *s, int qmask) +{ + /* + * Check whether Qregs are in range. For v8.1M only Q0..Q7 + * are supported, see VFPSmallRegisterBank(). + */ + return qmask < 8; +} + +static bool mve_eci_check(DisasContext *s) +{ + /* + * This is a beatwise insn: check that ECI is valid (not a + * reserved value) and note that we are handling it. + * Return true if OK, false if we generated an exception. + */ + s->eci_handled = true; + switch (s->eci) { + case ECI_NONE: + case ECI_A0: + case ECI_A0A1: + case ECI_A0A1A2: + case ECI_A0A1A2B0: + return true; + default: + /* Reserved value: INVSTATE UsageFault */ + gen_exception_insn(s, s->pc_curr, EXCP_INVSTATE, syn_uncategorized(), + default_exception_el(s)); + return false; + } +} + +static void mve_update_eci(DisasContext *s) +{ + /* + * The helper function will always update the CPUState field, + * so we only need to update the DisasContext field. + */ + if (s->eci) { + s->eci = (s->eci == ECI_A0A1A2B0) ? ECI_A0 : ECI_NONE; + } +} + +static bool do_ldst(DisasContext *s, arg_VLDR_VSTR *a, MVEGenLdStFn *fn) +{ + TCGv_i32 addr; + uint32_t offset; + TCGv_ptr qreg; + + if (!dc_isar_feature(aa32_mve, s) || + !mve_check_qreg_bank(s, a->qd) || + !fn) { + return false; + } + + /* CONSTRAINED UNPREDICTABLE: we choose to UNDEF */ + if (a->rn == 15 || (a->rn == 13 && a->w)) { + return false; + } + + if (!mve_eci_check(s) || !vfp_access_check(s)) { + return true; + } + + offset = a->imm << a->size; + if (!a->a) { + offset = -offset; + } + addr = load_reg(s, a->rn); + if (a->p) { + tcg_gen_addi_i32(addr, addr, offset); + } + + qreg = mve_qreg_ptr(a->qd); + fn(cpu_env, qreg, addr); + tcg_temp_free_ptr(qreg); + + /* + * Writeback always happens after the last beat of the insn, + * regardless of predication + */ + if (a->w) { + if (!a->p) { + tcg_gen_addi_i32(addr, addr, offset); + } + store_reg(s, a->rn, addr); + } else { + tcg_temp_free_i32(addr); + } + mve_update_eci(s); + return true; +} + +static bool trans_VLDR_VSTR(DisasContext *s, arg_VLDR_VSTR *a) +{ + static MVEGenLdStFn * const ldstfns[4][2] = { + { gen_helper_mve_vstrb, gen_helper_mve_vldrb }, + { gen_helper_mve_vstrh, gen_helper_mve_vldrh }, + { gen_helper_mve_vstrw, gen_helper_mve_vldrw }, + { NULL, NULL } + }; + return do_ldst(s, a, ldstfns[a->size][a->l]); +} diff --git a/target/arm/meson.build b/target/arm/meson.build index XXXXXXX..XXXXXXX 100644 --- a/target/arm/meson.build +++ b/target/arm/meson.build @@ -XXX,XX +XXX,XX @@ arm_ss.add(files( 'helper.c', 'iwmmxt_helper.c', 'm_helper.c', + 'mve_helper.c', 'neon_helper.c', 'op_helper.c', 'tlb_helper.c', -- 2.20.1 Implement the variants of MVE VLDR (encodings T1, T2) which perform "widening" loads where bytes or halfwords are loaded from memory and zero or sign-extended into halfword or word length vector elements, and the narrowing MVE VSTR (encodings T1, T2) where bytes or halfwords are stored from halfword or word elements. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-3-peter.maydell@linaro.org --- target/arm/helper-mve.h | 10 ++++++++++ target/arm/mve.decode | 25 +++++++++++++++++++++++-- target/arm/mve_helper.c | 11 +++++++++++ target/arm/translate-mve.c | 14 ++++++++++++++ 4 files changed, 58 insertions(+), 2 deletions(-) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_3(mve_vldrw, TCG_CALL_NO_WG, void, env, ptr, i32) DEF_HELPER_FLAGS_3(mve_vstrb, TCG_CALL_NO_WG, void, env, ptr, i32) DEF_HELPER_FLAGS_3(mve_vstrh, TCG_CALL_NO_WG, void, env, ptr, i32) DEF_HELPER_FLAGS_3(mve_vstrw, TCG_CALL_NO_WG, void, env, ptr, i32) + +DEF_HELPER_FLAGS_3(mve_vldrb_sh, TCG_CALL_NO_WG, void, env, ptr, i32) +DEF_HELPER_FLAGS_3(mve_vldrb_sw, TCG_CALL_NO_WG, void, env, ptr, i32) +DEF_HELPER_FLAGS_3(mve_vldrb_uh, TCG_CALL_NO_WG, void, env, ptr, i32) +DEF_HELPER_FLAGS_3(mve_vldrb_uw, TCG_CALL_NO_WG, void, env, ptr, i32) +DEF_HELPER_FLAGS_3(mve_vldrh_sw, TCG_CALL_NO_WG, void, env, ptr, i32) +DEF_HELPER_FLAGS_3(mve_vldrh_uw, TCG_CALL_NO_WG, void, env, ptr, i32) +DEF_HELPER_FLAGS_3(mve_vstrb_h, TCG_CALL_NO_WG, void, env, ptr, i32) +DEF_HELPER_FLAGS_3(mve_vstrb_w, TCG_CALL_NO_WG, void, env, ptr, i32) +DEF_HELPER_FLAGS_3(mve_vstrh_w, TCG_CALL_NO_WG, void, env, ptr, i32) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ %qd 22:1 13:3 -&vldr_vstr rn qd imm p a w size l +&vldr_vstr rn qd imm p a w size l u -@vldr_vstr ....... . . . . l:1 rn:4 ... ...... imm:7 &vldr_vstr qd=%qd +@vldr_vstr ....... . . . . l:1 rn:4 ... ...... imm:7 &vldr_vstr qd=%qd u=0 +# Note that both Rn and Qd are 3 bits only (no D bit) +@vldst_wn ... u:1 ... . . . . l:1 . rn:3 qd:3 . ... .. imm:7 &vldr_vstr # Vector loads and stores +# Widening loads and narrowing stores: +# for these P=0 W=0 is 'related encoding'; sz=11 is 'related encoding' +# This means we need to expand out to multiple patterns for P, W, SZ. +# For stores the U bit must be 0 but we catch that in the trans_ function. +# The naming scheme here is "VLDSTB_H == in-memory byte load/store to/from +# signed halfword element in register", etc. +VLDSTB_H 111 . 110 0 a:1 0 1 . 0 ... ... 0 111 01 ....... @vldst_wn \ + p=0 w=1 size=1 +VLDSTB_H 111 . 110 1 a:1 0 w:1 . 0 ... ... 0 111 01 ....... @vldst_wn \ + p=1 size=1 +VLDSTB_W 111 . 110 0 a:1 0 1 . 0 ... ... 0 111 10 ....... @vldst_wn \ + p=0 w=1 size=2 +VLDSTB_W 111 . 110 1 a:1 0 w:1 . 0 ... ... 0 111 10 ....... @vldst_wn \ + p=1 size=2 +VLDSTH_W 111 . 110 0 a:1 0 1 . 1 ... ... 0 111 10 ....... @vldst_wn \ + p=0 w=1 size=2 +VLDSTH_W 111 . 110 1 a:1 0 w:1 . 1 ... ... 0 111 10 ....... @vldst_wn \ + p=1 size=2 + # Non-widening loads/stores (P=0 W=0 is 'related encoding') VLDR_VSTR 1110110 0 a:1 . 1 . .... ... 111100 ....... @vldr_vstr \ size=0 p=0 w=1 diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ DO_VSTR(vstrb, 1, stb, 1, uint8_t) DO_VSTR(vstrh, 2, stw, 2, uint16_t) DO_VSTR(vstrw, 4, stl, 4, uint32_t) +DO_VLDR(vldrb_sh, 1, ldsb, 2, int16_t) +DO_VLDR(vldrb_sw, 1, ldsb, 4, int32_t) +DO_VLDR(vldrb_uh, 1, ldub, 2, uint16_t) +DO_VLDR(vldrb_uw, 1, ldub, 4, uint32_t) +DO_VLDR(vldrh_sw, 2, ldsw, 4, int32_t) +DO_VLDR(vldrh_uw, 2, lduw, 4, uint32_t) + +DO_VSTR(vstrb_h, 1, stb, 2, int16_t) +DO_VSTR(vstrb_w, 1, stb, 4, int32_t) +DO_VSTR(vstrh_w, 2, stw, 4, int32_t) + #undef DO_VLDR #undef DO_VSTR diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ static bool trans_VLDR_VSTR(DisasContext *s, arg_VLDR_VSTR *a) }; return do_ldst(s, a, ldstfns[a->size][a->l]); } + +#define DO_VLDST_WIDE_NARROW(OP, SLD, ULD, ST) \ + static bool trans_##OP(DisasContext *s, arg_VLDR_VSTR *a) \ + { \ + static MVEGenLdStFn * const ldstfns[2][2] = { \ + { gen_helper_mve_##ST, gen_helper_mve_##SLD }, \ + { NULL, gen_helper_mve_##ULD }, \ + }; \ + return do_ldst(s, a, ldstfns[a->u][a->l]); \ + } + +DO_VLDST_WIDE_NARROW(VLDSTB_H, vldrb_sh, vldrb_uh, vstrb_h) +DO_VLDST_WIDE_NARROW(VLDSTB_W, vldrb_sw, vldrb_uw, vstrb_w) +DO_VLDST_WIDE_NARROW(VLDSTH_W, vldrh_sw, vldrh_uw, vstrh_w) -- 2.20.1 Implement the MVE VCLZ insn (and the necessary machinery for MVE 1-input vector ops). Note that for non-load instructions predication is always performed at a byte level granularity regardless of element size (R_ZLSJ), and so the masking logic here differs from that used in the VLDR and VSTR helpers. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-4-peter.maydell@linaro.org --- target/arm/helper-mve.h | 4 ++ target/arm/mve.decode | 8 ++++ target/arm/mve_helper.c | 82 ++++++++++++++++++++++++++++++++++++++ target/arm/translate-mve.c | 38 ++++++++++++++++++ 4 files changed, 132 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_3(mve_vldrh_uw, TCG_CALL_NO_WG, void, env, ptr, i32) DEF_HELPER_FLAGS_3(mve_vstrb_h, TCG_CALL_NO_WG, void, env, ptr, i32) DEF_HELPER_FLAGS_3(mve_vstrb_w, TCG_CALL_NO_WG, void, env, ptr, i32) DEF_HELPER_FLAGS_3(mve_vstrh_w, TCG_CALL_NO_WG, void, env, ptr, i32) + +DEF_HELPER_FLAGS_3(mve_vclzb, TCG_CALL_NO_WG, void, env, ptr, ptr) +DEF_HELPER_FLAGS_3(mve_vclzh, TCG_CALL_NO_WG, void, env, ptr, ptr) +DEF_HELPER_FLAGS_3(mve_vclzw, TCG_CALL_NO_WG, void, env, ptr, ptr) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ # %qd 22:1 13:3 +%qm 5:1 1:3 &vldr_vstr rn qd imm p a w size l u +&1op qd qm size @vldr_vstr ....... . . . . l:1 rn:4 ... ...... imm:7 &vldr_vstr qd=%qd u=0 # Note that both Rn and Qd are 3 bits only (no D bit) @vldst_wn ... u:1 ... . . . . l:1 . rn:3 qd:3 . ... .. imm:7 &vldr_vstr +@1op .... .... .... size:2 .. .... .... .... .... &1op qd=%qd qm=%qm + # Vector loads and stores # Widening loads and narrowing stores: @@ -XXX,XX +XXX,XX @@ VLDR_VSTR 1110110 1 a:1 . w:1 . .... ... 111101 ....... @vldr_vstr \ size=1 p=1 VLDR_VSTR 1110110 1 a:1 . w:1 . .... ... 111110 ....... @vldr_vstr \ size=2 p=1 + +# Vector miscellaneous + +VCLZ 1111 1111 1 . 11 .. 00 ... 0 0100 11 . 0 ... 0 @1op diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ DO_VSTR(vstrh_w, 2, stw, 4, int32_t) #undef DO_VLDR #undef DO_VSTR + +/* + * The mergemask(D, R, M) macro performs the operation "*D = R" but + * storing only the bytes which correspond to 1 bits in M, + * leaving other bytes in *D unchanged. We use _Generic + * to select the correct implementation based on the type of D. + */ + +static void mergemask_ub(uint8_t *d, uint8_t r, uint16_t mask) +{ + if (mask & 1) { + *d = r; + } +} + +static void mergemask_sb(int8_t *d, int8_t r, uint16_t mask) +{ + mergemask_ub((uint8_t *)d, r, mask); +} + +static void mergemask_uh(uint16_t *d, uint16_t r, uint16_t mask) +{ + uint16_t bmask = expand_pred_b_data[mask & 3]; + *d = (*d & ~bmask) | (r & bmask); +} + +static void mergemask_sh(int16_t *d, int16_t r, uint16_t mask) +{ + mergemask_uh((uint16_t *)d, r, mask); +} + +static void mergemask_uw(uint32_t *d, uint32_t r, uint16_t mask) +{ + uint32_t bmask = expand_pred_b_data[mask & 0xf]; + *d = (*d & ~bmask) | (r & bmask); +} + +static void mergemask_sw(int32_t *d, int32_t r, uint16_t mask) +{ + mergemask_uw((uint32_t *)d, r, mask); +} + +static void mergemask_uq(uint64_t *d, uint64_t r, uint16_t mask) +{ + uint64_t bmask = expand_pred_b_data[mask & 0xff]; + *d = (*d & ~bmask) | (r & bmask); +} + +static void mergemask_sq(int64_t *d, int64_t r, uint16_t mask) +{ + mergemask_uq((uint64_t *)d, r, mask); +} + +#define mergemask(D, R, M) \ + _Generic(D, \ + uint8_t *: mergemask_ub, \ + int8_t *: mergemask_sb, \ + uint16_t *: mergemask_uh, \ + int16_t *: mergemask_sh, \ + uint32_t *: mergemask_uw, \ + int32_t *: mergemask_sw, \ + uint64_t *: mergemask_uq, \ + int64_t *: mergemask_sq)(D, R, M) + +#define DO_1OP(OP, ESIZE, TYPE, FN) \ + void HELPER(mve_##OP)(CPUARMState *env, void *vd, void *vm) \ + { \ + TYPE *d = vd, *m = vm; \ + uint16_t mask = mve_element_mask(env); \ + unsigned e; \ + for (e = 0; e < 16 / ESIZE; e++, mask >>= ESIZE) { \ + mergemask(&d[H##ESIZE(e)], FN(m[H##ESIZE(e)]), mask); \ + } \ + mve_advance_vpt(env); \ + } + +#define DO_CLZ_B(N) (clz32(N) - 24) +#define DO_CLZ_H(N) (clz32(N) - 16) + +DO_1OP(vclzb, 1, uint8_t, DO_CLZ_B) +DO_1OP(vclzh, 2, uint16_t, DO_CLZ_H) +DO_1OP(vclzw, 4, uint32_t, clz32) diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ #include "decode-mve.c.inc" typedef void MVEGenLdStFn(TCGv_ptr, TCGv_ptr, TCGv_i32); +typedef void MVEGenOneOpFn(TCGv_ptr, TCGv_ptr, TCGv_ptr); /* Return the offset of a Qn register (same semantics as aa32_vfp_qreg()) */ static inline long mve_qreg_offset(unsigned reg) @@ -XXX,XX +XXX,XX @@ static bool trans_VLDR_VSTR(DisasContext *s, arg_VLDR_VSTR *a) DO_VLDST_WIDE_NARROW(VLDSTB_H, vldrb_sh, vldrb_uh, vstrb_h) DO_VLDST_WIDE_NARROW(VLDSTB_W, vldrb_sw, vldrb_uw, vstrb_w) DO_VLDST_WIDE_NARROW(VLDSTH_W, vldrh_sw, vldrh_uw, vstrh_w) + +static bool do_1op(DisasContext *s, arg_1op *a, MVEGenOneOpFn fn) +{ + TCGv_ptr qd, qm; + + if (!dc_isar_feature(aa32_mve, s) || + !mve_check_qreg_bank(s, a->qd | a->qm) || + !fn) { + return false; + } + + if (!mve_eci_check(s) || !vfp_access_check(s)) { + return true; + } + + qd = mve_qreg_ptr(a->qd); + qm = mve_qreg_ptr(a->qm); + fn(cpu_env, qd, qm); + tcg_temp_free_ptr(qd); + tcg_temp_free_ptr(qm); + mve_update_eci(s); + return true; +} + +#define DO_1OP(INSN, FN) \ + static bool trans_##INSN(DisasContext *s, arg_1op *a) \ + { \ + static MVEGenOneOpFn * const fns[] = { \ + gen_helper_mve_##FN##b, \ + gen_helper_mve_##FN##h, \ + gen_helper_mve_##FN##w, \ + NULL, \ + }; \ + return do_1op(s, a, fns[a->size]); \ + } + +DO_1OP(VCLZ, vclz) -- 2.20.1 Implement the MVE VCLS insn. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-5-peter.maydell@linaro.org --- target/arm/helper-mve.h | 4 ++++ target/arm/mve.decode | 1 + target/arm/mve_helper.c | 7 +++++++ target/arm/translate-mve.c | 1 + 4 files changed, 13 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_3(mve_vstrb_h, TCG_CALL_NO_WG, void, env, ptr, i32) DEF_HELPER_FLAGS_3(mve_vstrb_w, TCG_CALL_NO_WG, void, env, ptr, i32) DEF_HELPER_FLAGS_3(mve_vstrh_w, TCG_CALL_NO_WG, void, env, ptr, i32) +DEF_HELPER_FLAGS_3(mve_vclsb, TCG_CALL_NO_WG, void, env, ptr, ptr) +DEF_HELPER_FLAGS_3(mve_vclsh, TCG_CALL_NO_WG, void, env, ptr, ptr) +DEF_HELPER_FLAGS_3(mve_vclsw, TCG_CALL_NO_WG, void, env, ptr, ptr) + DEF_HELPER_FLAGS_3(mve_vclzb, TCG_CALL_NO_WG, void, env, ptr, ptr) DEF_HELPER_FLAGS_3(mve_vclzh, TCG_CALL_NO_WG, void, env, ptr, ptr) DEF_HELPER_FLAGS_3(mve_vclzw, TCG_CALL_NO_WG, void, env, ptr, ptr) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ VLDR_VSTR 1110110 1 a:1 . w:1 . .... ... 111110 ....... @vldr_vstr \ # Vector miscellaneous +VCLS 1111 1111 1 . 11 .. 00 ... 0 0100 01 . 0 ... 0 @1op VCLZ 1111 1111 1 . 11 .. 00 ... 0 0100 11 . 0 ... 0 @1op diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ static void mergemask_sq(int64_t *d, int64_t r, uint16_t mask) mve_advance_vpt(env); \ } +#define DO_CLS_B(N) (clrsb32(N) - 24) +#define DO_CLS_H(N) (clrsb32(N) - 16) + +DO_1OP(vclsb, 1, int8_t, DO_CLS_B) +DO_1OP(vclsh, 2, int16_t, DO_CLS_H) +DO_1OP(vclsw, 4, int32_t, clrsb32) + #define DO_CLZ_B(N) (clz32(N) - 24) #define DO_CLZ_H(N) (clz32(N) - 16) diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ static bool do_1op(DisasContext *s, arg_1op *a, MVEGenOneOpFn fn) } DO_1OP(VCLZ, vclz) +DO_1OP(VCLS, vcls) -- 2.20.1 Implement the MVE instructions VREV16, VREV32 and VREV64. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-6-peter.maydell@linaro.org --- target/arm/helper-mve.h | 7 +++++++ target/arm/mve.decode | 4 ++++ target/arm/mve_helper.c | 7 +++++++ target/arm/translate-mve.c | 33 +++++++++++++++++++++++++++++++++ 4 files changed, 51 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_3(mve_vclsw, TCG_CALL_NO_WG, void, env, ptr, ptr) DEF_HELPER_FLAGS_3(mve_vclzb, TCG_CALL_NO_WG, void, env, ptr, ptr) DEF_HELPER_FLAGS_3(mve_vclzh, TCG_CALL_NO_WG, void, env, ptr, ptr) DEF_HELPER_FLAGS_3(mve_vclzw, TCG_CALL_NO_WG, void, env, ptr, ptr) + +DEF_HELPER_FLAGS_3(mve_vrev16b, TCG_CALL_NO_WG, void, env, ptr, ptr) +DEF_HELPER_FLAGS_3(mve_vrev32b, TCG_CALL_NO_WG, void, env, ptr, ptr) +DEF_HELPER_FLAGS_3(mve_vrev32h, TCG_CALL_NO_WG, void, env, ptr, ptr) +DEF_HELPER_FLAGS_3(mve_vrev64b, TCG_CALL_NO_WG, void, env, ptr, ptr) +DEF_HELPER_FLAGS_3(mve_vrev64h, TCG_CALL_NO_WG, void, env, ptr, ptr) +DEF_HELPER_FLAGS_3(mve_vrev64w, TCG_CALL_NO_WG, void, env, ptr, ptr) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ VLDR_VSTR 1110110 1 a:1 . w:1 . .... ... 111110 ....... @vldr_vstr \ VCLS 1111 1111 1 . 11 .. 00 ... 0 0100 01 . 0 ... 0 @1op VCLZ 1111 1111 1 . 11 .. 00 ... 0 0100 11 . 0 ... 0 @1op + +VREV16 1111 1111 1 . 11 .. 00 ... 0 0001 01 . 0 ... 0 @1op +VREV32 1111 1111 1 . 11 .. 00 ... 0 0000 11 . 0 ... 0 @1op +VREV64 1111 1111 1 . 11 .. 00 ... 0 0000 01 . 0 ... 0 @1op diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ DO_1OP(vclsw, 4, int32_t, clrsb32) DO_1OP(vclzb, 1, uint8_t, DO_CLZ_B) DO_1OP(vclzh, 2, uint16_t, DO_CLZ_H) DO_1OP(vclzw, 4, uint32_t, clz32) + +DO_1OP(vrev16b, 2, uint16_t, bswap16) +DO_1OP(vrev32b, 4, uint32_t, bswap32) +DO_1OP(vrev32h, 4, uint32_t, hswap32) +DO_1OP(vrev64b, 8, uint64_t, bswap64) +DO_1OP(vrev64h, 8, uint64_t, hswap64) +DO_1OP(vrev64w, 8, uint64_t, wswap64) diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ static bool do_1op(DisasContext *s, arg_1op *a, MVEGenOneOpFn fn) DO_1OP(VCLZ, vclz) DO_1OP(VCLS, vcls) + +static bool trans_VREV16(DisasContext *s, arg_1op *a) +{ + static MVEGenOneOpFn * const fns[] = { + gen_helper_mve_vrev16b, + NULL, + NULL, + NULL, + }; + return do_1op(s, a, fns[a->size]); +} + +static bool trans_VREV32(DisasContext *s, arg_1op *a) +{ + static MVEGenOneOpFn * const fns[] = { + gen_helper_mve_vrev32b, + gen_helper_mve_vrev32h, + NULL, + NULL, + }; + return do_1op(s, a, fns[a->size]); +} + +static bool trans_VREV64(DisasContext *s, arg_1op *a) +{ + static MVEGenOneOpFn * const fns[] = { + gen_helper_mve_vrev64b, + gen_helper_mve_vrev64h, + gen_helper_mve_vrev64w, + NULL, + }; + return do_1op(s, a, fns[a->size]); +} -- 2.20.1 Implement the MVE VMVN(register) operation. Note that for predication this operation is byte-by-byte. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-7-peter.maydell@linaro.org --- target/arm/helper-mve.h | 2 ++ target/arm/mve.decode | 3 +++ target/arm/mve_helper.c | 4 ++++ target/arm/translate-mve.c | 5 +++++ 4 files changed, 14 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_3(mve_vrev32h, TCG_CALL_NO_WG, void, env, ptr, ptr) DEF_HELPER_FLAGS_3(mve_vrev64b, TCG_CALL_NO_WG, void, env, ptr, ptr) DEF_HELPER_FLAGS_3(mve_vrev64h, TCG_CALL_NO_WG, void, env, ptr, ptr) DEF_HELPER_FLAGS_3(mve_vrev64w, TCG_CALL_NO_WG, void, env, ptr, ptr) + +DEF_HELPER_FLAGS_3(mve_vmvn, TCG_CALL_NO_WG, void, env, ptr, ptr) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ @vldst_wn ... u:1 ... . . . . l:1 . rn:3 qd:3 . ... .. imm:7 &vldr_vstr @1op .... .... .... size:2 .. .... .... .... .... &1op qd=%qd qm=%qm +@1op_nosz .... .... .... .... .... .... .... .... &1op qd=%qd qm=%qm size=0 # Vector loads and stores @@ -XXX,XX +XXX,XX @@ VCLZ 1111 1111 1 . 11 .. 00 ... 0 0100 11 . 0 ... 0 @1op VREV16 1111 1111 1 . 11 .. 00 ... 0 0001 01 . 0 ... 0 @1op VREV32 1111 1111 1 . 11 .. 00 ... 0 0000 11 . 0 ... 0 @1op VREV64 1111 1111 1 . 11 .. 00 ... 0 0000 01 . 0 ... 0 @1op + +VMVN 1111 1111 1 . 11 00 00 ... 0 0101 11 . 0 ... 0 @1op_nosz diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ DO_1OP(vrev32h, 4, uint32_t, hswap32) DO_1OP(vrev64b, 8, uint64_t, bswap64) DO_1OP(vrev64h, 8, uint64_t, hswap64) DO_1OP(vrev64w, 8, uint64_t, wswap64) + +#define DO_NOT(N) (~(N)) + +DO_1OP(vmvn, 8, uint64_t, DO_NOT) diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ static bool trans_VREV64(DisasContext *s, arg_1op *a) }; return do_1op(s, a, fns[a->size]); } + +static bool trans_VMVN(DisasContext *s, arg_1op *a) +{ + return do_1op(s, a, gen_helper_mve_vmvn); +} -- 2.20.1 Implement the MVE VABS functions (both integer and floating point). Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-8-peter.maydell@linaro.org --- target/arm/helper-mve.h | 6 ++++++ target/arm/mve.decode | 3 +++ target/arm/mve_helper.c | 13 +++++++++++++ target/arm/translate-mve.c | 15 +++++++++++++++ 4 files changed, 37 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_3(mve_vrev64h, TCG_CALL_NO_WG, void, env, ptr, ptr) DEF_HELPER_FLAGS_3(mve_vrev64w, TCG_CALL_NO_WG, void, env, ptr, ptr) DEF_HELPER_FLAGS_3(mve_vmvn, TCG_CALL_NO_WG, void, env, ptr, ptr) + +DEF_HELPER_FLAGS_3(mve_vabsb, TCG_CALL_NO_WG, void, env, ptr, ptr) +DEF_HELPER_FLAGS_3(mve_vabsh, TCG_CALL_NO_WG, void, env, ptr, ptr) +DEF_HELPER_FLAGS_3(mve_vabsw, TCG_CALL_NO_WG, void, env, ptr, ptr) +DEF_HELPER_FLAGS_3(mve_vfabsh, TCG_CALL_NO_WG, void, env, ptr, ptr) +DEF_HELPER_FLAGS_3(mve_vfabss, TCG_CALL_NO_WG, void, env, ptr, ptr) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ VREV32 1111 1111 1 . 11 .. 00 ... 0 0000 11 . 0 ... 0 @1op VREV64 1111 1111 1 . 11 .. 00 ... 0 0000 01 . 0 ... 0 @1op VMVN 1111 1111 1 . 11 00 00 ... 0 0101 11 . 0 ... 0 @1op_nosz + +VABS 1111 1111 1 . 11 .. 01 ... 0 0011 01 . 0 ... 0 @1op +VABS_fp 1111 1111 1 . 11 .. 01 ... 0 0111 01 . 0 ... 0 @1op diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ #include "exec/helper-proto.h" #include "exec/cpu_ldst.h" #include "exec/exec-all.h" +#include "tcg/tcg.h" static uint16_t mve_element_mask(CPUARMState *env) { @@ -XXX,XX +XXX,XX @@ DO_1OP(vrev64w, 8, uint64_t, wswap64) #define DO_NOT(N) (~(N)) DO_1OP(vmvn, 8, uint64_t, DO_NOT) + +#define DO_ABS(N) ((N) < 0 ? -(N) : (N)) +#define DO_FABSH(N) ((N) & dup_const(MO_16, 0x7fff)) +#define DO_FABSS(N) ((N) & dup_const(MO_32, 0x7fffffff)) + +DO_1OP(vabsb, 1, int8_t, DO_ABS) +DO_1OP(vabsh, 2, int16_t, DO_ABS) +DO_1OP(vabsw, 4, int32_t, DO_ABS) + +/* We can do these 64 bits at a time */ +DO_1OP(vfabsh, 8, uint64_t, DO_FABSH) +DO_1OP(vfabss, 8, uint64_t, DO_FABSS) diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ static bool do_1op(DisasContext *s, arg_1op *a, MVEGenOneOpFn fn) DO_1OP(VCLZ, vclz) DO_1OP(VCLS, vcls) +DO_1OP(VABS, vabs) static bool trans_VREV16(DisasContext *s, arg_1op *a) { @@ -XXX,XX +XXX,XX @@ static bool trans_VMVN(DisasContext *s, arg_1op *a) { return do_1op(s, a, gen_helper_mve_vmvn); } + +static bool trans_VABS_fp(DisasContext *s, arg_1op *a) +{ + static MVEGenOneOpFn * const fns[] = { + NULL, + gen_helper_mve_vfabsh, + gen_helper_mve_vfabss, + NULL, + }; + if (!dc_isar_feature(aa32_mve_fp, s)) { + return false; + } + return do_1op(s, a, fns[a->size]); +} -- 2.20.1 Implement the MVE VNEG insn (both integer and floating point forms). Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-9-peter.maydell@linaro.org --- target/arm/helper-mve.h | 6 ++++++ target/arm/mve.decode | 2 ++ target/arm/mve_helper.c | 12 ++++++++++++ target/arm/translate-mve.c | 15 +++++++++++++++ 4 files changed, 35 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_3(mve_vabsh, TCG_CALL_NO_WG, void, env, ptr, ptr) DEF_HELPER_FLAGS_3(mve_vabsw, TCG_CALL_NO_WG, void, env, ptr, ptr) DEF_HELPER_FLAGS_3(mve_vfabsh, TCG_CALL_NO_WG, void, env, ptr, ptr) DEF_HELPER_FLAGS_3(mve_vfabss, TCG_CALL_NO_WG, void, env, ptr, ptr) + +DEF_HELPER_FLAGS_3(mve_vnegb, TCG_CALL_NO_WG, void, env, ptr, ptr) +DEF_HELPER_FLAGS_3(mve_vnegh, TCG_CALL_NO_WG, void, env, ptr, ptr) +DEF_HELPER_FLAGS_3(mve_vnegw, TCG_CALL_NO_WG, void, env, ptr, ptr) +DEF_HELPER_FLAGS_3(mve_vfnegh, TCG_CALL_NO_WG, void, env, ptr, ptr) +DEF_HELPER_FLAGS_3(mve_vfnegs, TCG_CALL_NO_WG, void, env, ptr, ptr) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ VMVN 1111 1111 1 . 11 00 00 ... 0 0101 11 . 0 ... 0 @1op_nosz VABS 1111 1111 1 . 11 .. 01 ... 0 0011 01 . 0 ... 0 @1op VABS_fp 1111 1111 1 . 11 .. 01 ... 0 0111 01 . 0 ... 0 @1op +VNEG 1111 1111 1 . 11 .. 01 ... 0 0011 11 . 0 ... 0 @1op +VNEG_fp 1111 1111 1 . 11 .. 01 ... 0 0111 11 . 0 ... 0 @1op diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ DO_1OP(vabsw, 4, int32_t, DO_ABS) /* We can do these 64 bits at a time */ DO_1OP(vfabsh, 8, uint64_t, DO_FABSH) DO_1OP(vfabss, 8, uint64_t, DO_FABSS) + +#define DO_NEG(N) (-(N)) +#define DO_FNEGH(N) ((N) ^ dup_const(MO_16, 0x8000)) +#define DO_FNEGS(N) ((N) ^ dup_const(MO_32, 0x80000000)) + +DO_1OP(vnegb, 1, int8_t, DO_NEG) +DO_1OP(vnegh, 2, int16_t, DO_NEG) +DO_1OP(vnegw, 4, int32_t, DO_NEG) + +/* We can do these 64 bits at a time */ +DO_1OP(vfnegh, 8, uint64_t, DO_FNEGH) +DO_1OP(vfnegs, 8, uint64_t, DO_FNEGS) diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ static bool do_1op(DisasContext *s, arg_1op *a, MVEGenOneOpFn fn) DO_1OP(VCLZ, vclz) DO_1OP(VCLS, vcls) DO_1OP(VABS, vabs) +DO_1OP(VNEG, vneg) static bool trans_VREV16(DisasContext *s, arg_1op *a) { @@ -XXX,XX +XXX,XX @@ static bool trans_VABS_fp(DisasContext *s, arg_1op *a) } return do_1op(s, a, fns[a->size]); } + +static bool trans_VNEG_fp(DisasContext *s, arg_1op *a) +{ + static MVEGenOneOpFn * const fns[] = { + NULL, + gen_helper_mve_vfnegh, + gen_helper_mve_vfnegs, + NULL, + }; + if (!dc_isar_feature(aa32_mve_fp, s)) { + return false; + } + return do_1op(s, a, fns[a->size]); +} -- 2.20.1 The Arm MVE VDUP implementation would like to be able to emit code to duplicate a byte or halfword value into an i32. We have code to do this already in tcg-op-gvec.c, so all we need to do is make the functions global. For consistency with other functions made available to the frontends: * we rename to tcg_gen_dup_* * we expose both the _i32 and _i64 forms * we provide the #define for a _tl form Suggested-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Message-id: 20210617121628.20116-10-peter.maydell@linaro.org --- include/tcg/tcg-op.h | 8 ++++++++ include/tcg/tcg.h | 1 - tcg/tcg-op-gvec.c | 20 ++++++++++---------- 3 files changed, 18 insertions(+), 11 deletions(-) diff --git a/include/tcg/tcg-op.h b/include/tcg/tcg-op.h index XXXXXXX..XXXXXXX 100644 --- a/include/tcg/tcg-op.h +++ b/include/tcg/tcg-op.h @@ -XXX,XX +XXX,XX @@ void tcg_gen_umin_i32(TCGv_i32, TCGv_i32 arg1, TCGv_i32 arg2); void tcg_gen_umax_i32(TCGv_i32, TCGv_i32 arg1, TCGv_i32 arg2); void tcg_gen_abs_i32(TCGv_i32, TCGv_i32); +/* Replicate a value of size @vece from @in to all the lanes in @out */ +void tcg_gen_dup_i32(unsigned vece, TCGv_i32 out, TCGv_i32 in); + static inline void tcg_gen_discard_i32(TCGv_i32 arg) { tcg_gen_op1_i32(INDEX_op_discard, arg); @@ -XXX,XX +XXX,XX @@ void tcg_gen_umin_i64(TCGv_i64, TCGv_i64 arg1, TCGv_i64 arg2); void tcg_gen_umax_i64(TCGv_i64, TCGv_i64 arg1, TCGv_i64 arg2); void tcg_gen_abs_i64(TCGv_i64, TCGv_i64); +/* Replicate a value of size @vece from @in to all the lanes in @out */ +void tcg_gen_dup_i64(unsigned vece, TCGv_i64 out, TCGv_i64 in); + #if TCG_TARGET_REG_BITS == 64 static inline void tcg_gen_discard_i64(TCGv_i64 arg) { @@ -XXX,XX +XXX,XX @@ void tcg_gen_stl_vec(TCGv_vec r, TCGv_ptr base, TCGArg offset, TCGType t); #define tcg_gen_atomic_smax_fetch_tl tcg_gen_atomic_smax_fetch_i64 #define tcg_gen_atomic_umax_fetch_tl tcg_gen_atomic_umax_fetch_i64 #define tcg_gen_dup_tl_vec tcg_gen_dup_i64_vec +#define tcg_gen_dup_tl tcg_gen_dup_i64 #else #define tcg_gen_movi_tl tcg_gen_movi_i32 #define tcg_gen_mov_tl tcg_gen_mov_i32 @@ -XXX,XX +XXX,XX @@ void tcg_gen_stl_vec(TCGv_vec r, TCGv_ptr base, TCGArg offset, TCGType t); #define tcg_gen_atomic_smax_fetch_tl tcg_gen_atomic_smax_fetch_i32 #define tcg_gen_atomic_umax_fetch_tl tcg_gen_atomic_umax_fetch_i32 #define tcg_gen_dup_tl_vec tcg_gen_dup_i32_vec +#define tcg_gen_dup_tl tcg_gen_dup_i32 #endif #if UINTPTR_MAX == UINT32_MAX diff --git a/include/tcg/tcg.h b/include/tcg/tcg.h index XXXXXXX..XXXXXXX 100644 --- a/include/tcg/tcg.h +++ b/include/tcg/tcg.h @@ -XXX,XX +XXX,XX @@ uint64_t dup_const(unsigned vece, uint64_t c); : (qemu_build_not_reached_always(), 0)) \ : dup_const(VECE, C)) - /* * Memory helpers that will be used by TCG generated code. */ diff --git a/tcg/tcg-op-gvec.c b/tcg/tcg-op-gvec.c index XXXXXXX..XXXXXXX 100644 --- a/tcg/tcg-op-gvec.c +++ b/tcg/tcg-op-gvec.c @@ -XXX,XX +XXX,XX @@ uint64_t (dup_const)(unsigned vece, uint64_t c) } /* Duplicate IN into OUT as per VECE. */ -static void gen_dup_i32(unsigned vece, TCGv_i32 out, TCGv_i32 in) +void tcg_gen_dup_i32(unsigned vece, TCGv_i32 out, TCGv_i32 in) { switch (vece) { case MO_8: @@ -XXX,XX +XXX,XX @@ static void gen_dup_i32(unsigned vece, TCGv_i32 out, TCGv_i32 in) } } -static void gen_dup_i64(unsigned vece, TCGv_i64 out, TCGv_i64 in) +void tcg_gen_dup_i64(unsigned vece, TCGv_i64 out, TCGv_i64 in) { switch (vece) { case MO_8: @@ -XXX,XX +XXX,XX @@ static void do_dup(unsigned vece, uint32_t dofs, uint32_t oprsz, && (vece != MO_32 || !check_size_impl(oprsz, 4))) { t_64 = tcg_temp_new_i64(); tcg_gen_extu_i32_i64(t_64, in_32); - gen_dup_i64(vece, t_64, t_64); + tcg_gen_dup_i64(vece, t_64, t_64); } else { t_32 = tcg_temp_new_i32(); - gen_dup_i32(vece, t_32, in_32); + tcg_gen_dup_i32(vece, t_32, in_32); } } else if (in_64) { /* We are given a 64-bit variable input. */ t_64 = tcg_temp_new_i64(); - gen_dup_i64(vece, t_64, in_64); + tcg_gen_dup_i64(vece, t_64, in_64); } else { /* We are given a constant input. */ /* For 64-bit hosts, use 64-bit constants for "simple" constants @@ -XXX,XX +XXX,XX @@ void tcg_gen_gvec_2s(uint32_t dofs, uint32_t aofs, uint32_t oprsz, } else if (g->fni8 && check_size_impl(oprsz, 8)) { TCGv_i64 t64 = tcg_temp_new_i64(); - gen_dup_i64(g->vece, t64, c); + tcg_gen_dup_i64(g->vece, t64, c); expand_2s_i64(dofs, aofs, oprsz, t64, g->scalar_first, g->fni8); tcg_temp_free_i64(t64); } else if (g->fni4 && check_size_impl(oprsz, 4)) { TCGv_i32 t32 = tcg_temp_new_i32(); tcg_gen_extrl_i64_i32(t32, c); - gen_dup_i32(g->vece, t32, t32); + tcg_gen_dup_i32(g->vece, t32, t32); expand_2s_i32(dofs, aofs, oprsz, t32, g->scalar_first, g->fni4); tcg_temp_free_i32(t32); } else { @@ -XXX,XX +XXX,XX @@ void tcg_gen_gvec_ands(unsigned vece, uint32_t dofs, uint32_t aofs, TCGv_i64 c, uint32_t oprsz, uint32_t maxsz) { TCGv_i64 tmp = tcg_temp_new_i64(); - gen_dup_i64(vece, tmp, c); + tcg_gen_dup_i64(vece, tmp, c); tcg_gen_gvec_2s(dofs, aofs, oprsz, maxsz, tmp, &gop_ands); tcg_temp_free_i64(tmp); } @@ -XXX,XX +XXX,XX @@ void tcg_gen_gvec_xors(unsigned vece, uint32_t dofs, uint32_t aofs, TCGv_i64 c, uint32_t oprsz, uint32_t maxsz) { TCGv_i64 tmp = tcg_temp_new_i64(); - gen_dup_i64(vece, tmp, c); + tcg_gen_dup_i64(vece, tmp, c); tcg_gen_gvec_2s(dofs, aofs, oprsz, maxsz, tmp, &gop_xors); tcg_temp_free_i64(tmp); } @@ -XXX,XX +XXX,XX @@ void tcg_gen_gvec_ors(unsigned vece, uint32_t dofs, uint32_t aofs, TCGv_i64 c, uint32_t oprsz, uint32_t maxsz) { TCGv_i64 tmp = tcg_temp_new_i64(); - gen_dup_i64(vece, tmp, c); + tcg_gen_dup_i64(vece, tmp, c); tcg_gen_gvec_2s(dofs, aofs, oprsz, maxsz, tmp, &gop_ors); tcg_temp_free_i64(tmp); } -- 2.20.1 Implement the MVE VDUP insn, which duplicates a value from a general-purpose register into every lane of a vector register (subject to predication). Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-11-peter.maydell@linaro.org --- target/arm/helper-mve.h | 2 ++ target/arm/mve.decode | 10 ++++++++++ target/arm/mve_helper.c | 16 ++++++++++++++++ target/arm/translate-mve.c | 27 +++++++++++++++++++++++++++ 4 files changed, 55 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_3(mve_vstrb_h, TCG_CALL_NO_WG, void, env, ptr, i32) DEF_HELPER_FLAGS_3(mve_vstrb_w, TCG_CALL_NO_WG, void, env, ptr, i32) DEF_HELPER_FLAGS_3(mve_vstrh_w, TCG_CALL_NO_WG, void, env, ptr, i32) +DEF_HELPER_FLAGS_3(mve_vdup, TCG_CALL_NO_WG, void, env, ptr, i32) + DEF_HELPER_FLAGS_3(mve_vclsb, TCG_CALL_NO_WG, void, env, ptr, ptr) DEF_HELPER_FLAGS_3(mve_vclsh, TCG_CALL_NO_WG, void, env, ptr, ptr) DEF_HELPER_FLAGS_3(mve_vclsw, TCG_CALL_NO_WG, void, env, ptr, ptr) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ %qd 22:1 13:3 %qm 5:1 1:3 +%qn 7:1 17:3 &vldr_vstr rn qd imm p a w size l u &1op qd qm size @@ -XXX,XX +XXX,XX @@ VABS 1111 1111 1 . 11 .. 01 ... 0 0011 01 . 0 ... 0 @1op VABS_fp 1111 1111 1 . 11 .. 01 ... 0 0111 01 . 0 ... 0 @1op VNEG 1111 1111 1 . 11 .. 01 ... 0 0011 11 . 0 ... 0 @1op VNEG_fp 1111 1111 1 . 11 .. 01 ... 0 0111 11 . 0 ... 0 @1op + +&vdup qd rt size +# Qd is in the fields usually named Qn +@vdup .... .... . . .. ... . rt:4 .... . . . . .... qd=%qn &vdup + +# B and E bits encode size, which we decode here to the usual size values +VDUP 1110 1110 1 1 10 ... 0 .... 1011 . 0 0 1 0000 @vdup size=0 +VDUP 1110 1110 1 0 10 ... 0 .... 1011 . 0 1 1 0000 @vdup size=1 +VDUP 1110 1110 1 0 10 ... 0 .... 1011 . 0 0 1 0000 @vdup size=2 diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ static void mergemask_sq(int64_t *d, int64_t r, uint16_t mask) uint64_t *: mergemask_uq, \ int64_t *: mergemask_sq)(D, R, M) +void HELPER(mve_vdup)(CPUARMState *env, void *vd, uint32_t val) +{ + /* + * The generated code already replicated an 8 or 16 bit constant + * into the 32-bit value, so we only need to write the 32-bit + * value to all elements of the Qreg, allowing for predication. + */ + uint32_t *d = vd; + uint16_t mask = mve_element_mask(env); + unsigned e; + for (e = 0; e < 16 / 4; e++, mask >>= 4) { + mergemask(&d[H4(e)], val, mask); + } + mve_advance_vpt(env); +} + #define DO_1OP(OP, ESIZE, TYPE, FN) \ void HELPER(mve_##OP)(CPUARMState *env, void *vd, void *vm) \ { \ diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ DO_VLDST_WIDE_NARROW(VLDSTB_H, vldrb_sh, vldrb_uh, vstrb_h) DO_VLDST_WIDE_NARROW(VLDSTB_W, vldrb_sw, vldrb_uw, vstrb_w) DO_VLDST_WIDE_NARROW(VLDSTH_W, vldrh_sw, vldrh_uw, vstrh_w) +static bool trans_VDUP(DisasContext *s, arg_VDUP *a) +{ + TCGv_ptr qd; + TCGv_i32 rt; + + if (!dc_isar_feature(aa32_mve, s) || + !mve_check_qreg_bank(s, a->qd)) { + return false; + } + if (a->rt == 13 || a->rt == 15) { + /* UNPREDICTABLE; we choose to UNDEF */ + return false; + } + if (!mve_eci_check(s) || !vfp_access_check(s)) { + return true; + } + + qd = mve_qreg_ptr(a->qd); + rt = load_reg(s, a->rt); + tcg_gen_dup_i32(a->size, rt, rt); + gen_helper_mve_vdup(cpu_env, qd, rt); + tcg_temp_free_ptr(qd); + tcg_temp_free_i32(rt); + mve_update_eci(s); + return true; +} + static bool do_1op(DisasContext *s, arg_1op *a, MVEGenOneOpFn fn) { TCGv_ptr qd, qm; -- 2.20.1 Implement the MVE vector logical operations operating on two registers. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-12-peter.maydell@linaro.org --- target/arm/helper-mve.h | 6 ++++++ target/arm/mve.decode | 9 +++++++++ target/arm/mve_helper.c | 26 ++++++++++++++++++++++++++ target/arm/translate-mve.c | 37 +++++++++++++++++++++++++++++++++++++ 4 files changed, 78 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_3(mve_vnegh, TCG_CALL_NO_WG, void, env, ptr, ptr) DEF_HELPER_FLAGS_3(mve_vnegw, TCG_CALL_NO_WG, void, env, ptr, ptr) DEF_HELPER_FLAGS_3(mve_vfnegh, TCG_CALL_NO_WG, void, env, ptr, ptr) DEF_HELPER_FLAGS_3(mve_vfnegs, TCG_CALL_NO_WG, void, env, ptr, ptr) + +DEF_HELPER_FLAGS_4(mve_vand, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vbic, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vorr, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vorn, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_veor, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ &vldr_vstr rn qd imm p a w size l u &1op qd qm size +&2op qd qm qn size @vldr_vstr ....... . . . . l:1 rn:4 ... ...... imm:7 &vldr_vstr qd=%qd u=0 # Note that both Rn and Qd are 3 bits only (no D bit) @@ -XXX,XX +XXX,XX @@ @1op .... .... .... size:2 .. .... .... .... .... &1op qd=%qd qm=%qm @1op_nosz .... .... .... .... .... .... .... .... &1op qd=%qd qm=%qm size=0 +@2op_nosz .... .... .... .... .... .... .... .... &2op qd=%qd qm=%qm qn=%qn size=0 # Vector loads and stores @@ -XXX,XX +XXX,XX @@ VLDR_VSTR 1110110 1 a:1 . w:1 . .... ... 111101 ....... @vldr_vstr \ VLDR_VSTR 1110110 1 a:1 . w:1 . .... ... 111110 ....... @vldr_vstr \ size=2 p=1 +# Vector 2-op +VAND 1110 1111 0 . 00 ... 0 ... 0 0001 . 1 . 1 ... 0 @2op_nosz +VBIC 1110 1111 0 . 01 ... 0 ... 0 0001 . 1 . 1 ... 0 @2op_nosz +VORR 1110 1111 0 . 10 ... 0 ... 0 0001 . 1 . 1 ... 0 @2op_nosz +VORN 1110 1111 0 . 11 ... 0 ... 0 0001 . 1 . 1 ... 0 @2op_nosz +VEOR 1111 1111 0 . 00 ... 0 ... 0 0001 . 1 . 1 ... 0 @2op_nosz + # Vector miscellaneous VCLS 1111 1111 1 . 11 .. 00 ... 0 0100 01 . 0 ... 0 @1op diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ DO_1OP(vnegw, 4, int32_t, DO_NEG) /* We can do these 64 bits at a time */ DO_1OP(vfnegh, 8, uint64_t, DO_FNEGH) DO_1OP(vfnegs, 8, uint64_t, DO_FNEGS) + +#define DO_2OP(OP, ESIZE, TYPE, FN) \ + void HELPER(glue(mve_, OP))(CPUARMState *env, \ + void *vd, void *vn, void *vm) \ + { \ + TYPE *d = vd, *n = vn, *m = vm; \ + uint16_t mask = mve_element_mask(env); \ + unsigned e; \ + for (e = 0; e < 16 / ESIZE; e++, mask >>= ESIZE) { \ + mergemask(&d[H##ESIZE(e)], \ + FN(n[H##ESIZE(e)], m[H##ESIZE(e)]), mask); \ + } \ + mve_advance_vpt(env); \ + } + +#define DO_AND(N, M) ((N) & (M)) +#define DO_BIC(N, M) ((N) & ~(M)) +#define DO_ORR(N, M) ((N) | (M)) +#define DO_ORN(N, M) ((N) | ~(M)) +#define DO_EOR(N, M) ((N) ^ (M)) + +DO_2OP(vand, 8, uint64_t, DO_AND) +DO_2OP(vbic, 8, uint64_t, DO_BIC) +DO_2OP(vorr, 8, uint64_t, DO_ORR) +DO_2OP(vorn, 8, uint64_t, DO_ORN) +DO_2OP(veor, 8, uint64_t, DO_EOR) diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ typedef void MVEGenLdStFn(TCGv_ptr, TCGv_ptr, TCGv_i32); typedef void MVEGenOneOpFn(TCGv_ptr, TCGv_ptr, TCGv_ptr); +typedef void MVEGenTwoOpFn(TCGv_ptr, TCGv_ptr, TCGv_ptr, TCGv_ptr); /* Return the offset of a Qn register (same semantics as aa32_vfp_qreg()) */ static inline long mve_qreg_offset(unsigned reg) @@ -XXX,XX +XXX,XX @@ static bool trans_VNEG_fp(DisasContext *s, arg_1op *a) } return do_1op(s, a, fns[a->size]); } + +static bool do_2op(DisasContext *s, arg_2op *a, MVEGenTwoOpFn fn) +{ + TCGv_ptr qd, qn, qm; + + if (!dc_isar_feature(aa32_mve, s) || + !mve_check_qreg_bank(s, a->qd | a->qn | a->qm) || + !fn) { + return false; + } + if (!mve_eci_check(s) || !vfp_access_check(s)) { + return true; + } + + qd = mve_qreg_ptr(a->qd); + qn = mve_qreg_ptr(a->qn); + qm = mve_qreg_ptr(a->qm); + fn(cpu_env, qd, qn, qm); + tcg_temp_free_ptr(qd); + tcg_temp_free_ptr(qn); + tcg_temp_free_ptr(qm); + mve_update_eci(s); + return true; +} + +#define DO_LOGIC(INSN, HELPER) \ + static bool trans_##INSN(DisasContext *s, arg_2op *a) \ + { \ + return do_2op(s, a, HELPER); \ + } + +DO_LOGIC(VAND, gen_helper_mve_vand) +DO_LOGIC(VBIC, gen_helper_mve_vbic) +DO_LOGIC(VORR, gen_helper_mve_vorr) +DO_LOGIC(VORN, gen_helper_mve_vorn) +DO_LOGIC(VEOR, gen_helper_mve_veor) -- 2.20.1 Implement the MVE VADD, VSUB and VMUL insns. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-13-peter.maydell@linaro.org --- target/arm/helper-mve.h | 12 ++++++++++++ target/arm/mve.decode | 5 +++++ target/arm/mve_helper.c | 14 ++++++++++++++ target/arm/translate-mve.c | 16 ++++++++++++++++ 4 files changed, 47 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(mve_vbic, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vorr, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vorn, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_veor, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + +DEF_HELPER_FLAGS_4(mve_vaddb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vaddh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vaddw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + +DEF_HELPER_FLAGS_4(mve_vsubb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vsubh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vsubw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + +DEF_HELPER_FLAGS_4(mve_vmulb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vmulh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vmulw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ @1op .... .... .... size:2 .. .... .... .... .... &1op qd=%qd qm=%qm @1op_nosz .... .... .... .... .... .... .... .... &1op qd=%qd qm=%qm size=0 +@2op .... .... .. size:2 .... .... .... .... .... &2op qd=%qd qm=%qm qn=%qn @2op_nosz .... .... .... .... .... .... .... .... &2op qd=%qd qm=%qm qn=%qn size=0 # Vector loads and stores @@ -XXX,XX +XXX,XX @@ VORR 1110 1111 0 . 10 ... 0 ... 0 0001 . 1 . 1 ... 0 @2op_nosz VORN 1110 1111 0 . 11 ... 0 ... 0 0001 . 1 . 1 ... 0 @2op_nosz VEOR 1111 1111 0 . 00 ... 0 ... 0 0001 . 1 . 1 ... 0 @2op_nosz +VADD 1110 1111 0 . .. ... 0 ... 0 1000 . 1 . 0 ... 0 @2op +VSUB 1111 1111 0 . .. ... 0 ... 0 1000 . 1 . 0 ... 0 @2op +VMUL 1110 1111 0 . .. ... 0 ... 0 1001 . 1 . 1 ... 0 @2op + # Vector miscellaneous VCLS 1111 1111 1 . 11 .. 00 ... 0 0100 01 . 0 ... 0 @1op diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ DO_1OP(vfnegs, 8, uint64_t, DO_FNEGS) mve_advance_vpt(env); \ } +/* provide unsigned 2-op helpers for all sizes */ +#define DO_2OP_U(OP, FN) \ + DO_2OP(OP##b, 1, uint8_t, FN) \ + DO_2OP(OP##h, 2, uint16_t, FN) \ + DO_2OP(OP##w, 4, uint32_t, FN) + #define DO_AND(N, M) ((N) & (M)) #define DO_BIC(N, M) ((N) & ~(M)) #define DO_ORR(N, M) ((N) | (M)) @@ -XXX,XX +XXX,XX @@ DO_2OP(vbic, 8, uint64_t, DO_BIC) DO_2OP(vorr, 8, uint64_t, DO_ORR) DO_2OP(vorn, 8, uint64_t, DO_ORN) DO_2OP(veor, 8, uint64_t, DO_EOR) + +#define DO_ADD(N, M) ((N) + (M)) +#define DO_SUB(N, M) ((N) - (M)) +#define DO_MUL(N, M) ((N) * (M)) + +DO_2OP_U(vadd, DO_ADD) +DO_2OP_U(vsub, DO_SUB) +DO_2OP_U(vmul, DO_MUL) diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ DO_LOGIC(VBIC, gen_helper_mve_vbic) DO_LOGIC(VORR, gen_helper_mve_vorr) DO_LOGIC(VORN, gen_helper_mve_vorn) DO_LOGIC(VEOR, gen_helper_mve_veor) + +#define DO_2OP(INSN, FN) \ + static bool trans_##INSN(DisasContext *s, arg_2op *a) \ + { \ + static MVEGenTwoOpFn * const fns[] = { \ + gen_helper_mve_##FN##b, \ + gen_helper_mve_##FN##h, \ + gen_helper_mve_##FN##w, \ + NULL, \ + }; \ + return do_2op(s, a, fns[a->size]); \ + } + +DO_2OP(VADD, vadd) +DO_2OP(VSUB, vsub) +DO_2OP(VMUL, vmul) -- 2.20.1 Implement the MVE VMULH insn, which performs a vector multiply and returns the high half of the result. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-14-peter.maydell@linaro.org --- target/arm/helper-mve.h | 7 +++++++ target/arm/mve.decode | 3 +++ target/arm/mve_helper.c | 26 ++++++++++++++++++++++++++ target/arm/translate-mve.c | 2 ++ 4 files changed, 38 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(mve_vsubw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vmulb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vmulh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vmulw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + +DEF_HELPER_FLAGS_4(mve_vmulhsb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vmulhsh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vmulhsw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vmulhub, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vmulhuh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vmulhuw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ VADD 1110 1111 0 . .. ... 0 ... 0 1000 . 1 . 0 ... 0 @2op VSUB 1111 1111 0 . .. ... 0 ... 0 1000 . 1 . 0 ... 0 @2op VMUL 1110 1111 0 . .. ... 0 ... 0 1001 . 1 . 1 ... 0 @2op +VMULH_S 111 0 1110 0 . .. ...1 ... 0 1110 . 0 . 0 ... 1 @2op +VMULH_U 111 1 1110 0 . .. ...1 ... 0 1110 . 0 . 0 ... 1 @2op + # Vector miscellaneous VCLS 1111 1111 1 . 11 .. 00 ... 0 0100 01 . 0 ... 0 @1op diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ DO_2OP(veor, 8, uint64_t, DO_EOR) DO_2OP_U(vadd, DO_ADD) DO_2OP_U(vsub, DO_SUB) DO_2OP_U(vmul, DO_MUL) + +/* + * Because the computation type is at least twice as large as required, + * these work for both signed and unsigned source types. + */ +static inline uint8_t do_mulh_b(int32_t n, int32_t m) +{ + return (n * m) >> 8; +} + +static inline uint16_t do_mulh_h(int32_t n, int32_t m) +{ + return (n * m) >> 16; +} + +static inline uint32_t do_mulh_w(int64_t n, int64_t m) +{ + return (n * m) >> 32; +} + +DO_2OP(vmulhsb, 1, int8_t, do_mulh_b) +DO_2OP(vmulhsh, 2, int16_t, do_mulh_h) +DO_2OP(vmulhsw, 4, int32_t, do_mulh_w) +DO_2OP(vmulhub, 1, uint8_t, do_mulh_b) +DO_2OP(vmulhuh, 2, uint16_t, do_mulh_h) +DO_2OP(vmulhuw, 4, uint32_t, do_mulh_w) diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ DO_LOGIC(VEOR, gen_helper_mve_veor) DO_2OP(VADD, vadd) DO_2OP(VSUB, vsub) DO_2OP(VMUL, vmul) +DO_2OP(VMULH_S, vmulhs) +DO_2OP(VMULH_U, vmulhu) -- 2.20.1 Implement the MVE VRMULH insn, which performs a rounding multiply and then returns the high half. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-15-peter.maydell@linaro.org --- target/arm/helper-mve.h | 7 +++++++ target/arm/mve.decode | 3 +++ target/arm/mve_helper.c | 22 ++++++++++++++++++++++ target/arm/translate-mve.c | 2 ++ 4 files changed, 34 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(mve_vmulhsw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vmulhub, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vmulhuh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vmulhuw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + +DEF_HELPER_FLAGS_4(mve_vrmulhsb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vrmulhsh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vrmulhsw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vrmulhub, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vrmulhuh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vrmulhuw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ VMUL 1110 1111 0 . .. ... 0 ... 0 1001 . 1 . 1 ... 0 @2op VMULH_S 111 0 1110 0 . .. ...1 ... 0 1110 . 0 . 0 ... 1 @2op VMULH_U 111 1 1110 0 . .. ...1 ... 0 1110 . 0 . 0 ... 1 @2op +VRMULH_S 111 0 1110 0 . .. ...1 ... 1 1110 . 0 . 0 ... 1 @2op +VRMULH_U 111 1 1110 0 . .. ...1 ... 1 1110 . 0 . 0 ... 1 @2op + # Vector miscellaneous VCLS 1111 1111 1 . 11 .. 00 ... 0 0100 01 . 0 ... 0 @1op diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ static inline uint32_t do_mulh_w(int64_t n, int64_t m) return (n * m) >> 32; } +static inline uint8_t do_rmulh_b(int32_t n, int32_t m) +{ + return (n * m + (1U << 7)) >> 8; +} + +static inline uint16_t do_rmulh_h(int32_t n, int32_t m) +{ + return (n * m + (1U << 15)) >> 16; +} + +static inline uint32_t do_rmulh_w(int64_t n, int64_t m) +{ + return (n * m + (1U << 31)) >> 32; +} + DO_2OP(vmulhsb, 1, int8_t, do_mulh_b) DO_2OP(vmulhsh, 2, int16_t, do_mulh_h) DO_2OP(vmulhsw, 4, int32_t, do_mulh_w) DO_2OP(vmulhub, 1, uint8_t, do_mulh_b) DO_2OP(vmulhuh, 2, uint16_t, do_mulh_h) DO_2OP(vmulhuw, 4, uint32_t, do_mulh_w) + +DO_2OP(vrmulhsb, 1, int8_t, do_rmulh_b) +DO_2OP(vrmulhsh, 2, int16_t, do_rmulh_h) +DO_2OP(vrmulhsw, 4, int32_t, do_rmulh_w) +DO_2OP(vrmulhub, 1, uint8_t, do_rmulh_b) +DO_2OP(vrmulhuh, 2, uint16_t, do_rmulh_h) +DO_2OP(vrmulhuw, 4, uint32_t, do_rmulh_w) diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ DO_2OP(VSUB, vsub) DO_2OP(VMUL, vmul) DO_2OP(VMULH_S, vmulhs) DO_2OP(VMULH_U, vmulhu) +DO_2OP(VRMULH_S, vrmulhs) +DO_2OP(VRMULH_U, vrmulhu) -- 2.20.1 Implement the MVE VMAX and VMIN insns. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-16-peter.maydell@linaro.org --- target/arm/helper-mve.h | 14 ++++++++++++++ target/arm/mve.decode | 5 +++++ target/arm/mve_helper.c | 14 ++++++++++++++ target/arm/translate-mve.c | 4 ++++ 4 files changed, 37 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(mve_vrmulhsw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vrmulhub, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vrmulhuh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vrmulhuw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + +DEF_HELPER_FLAGS_4(mve_vmaxsb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vmaxsh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vmaxsw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vmaxub, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vmaxuh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vmaxuw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + +DEF_HELPER_FLAGS_4(mve_vminsb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vminsh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vminsw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vminub, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vminuh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vminuw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ VMULH_U 111 1 1110 0 . .. ...1 ... 0 1110 . 0 . 0 ... 1 @2op VRMULH_S 111 0 1110 0 . .. ...1 ... 1 1110 . 0 . 0 ... 1 @2op VRMULH_U 111 1 1110 0 . .. ...1 ... 1 1110 . 0 . 0 ... 1 @2op +VMAX_S 111 0 1111 0 . .. ... 0 ... 0 0110 . 1 . 0 ... 0 @2op +VMAX_U 111 1 1111 0 . .. ... 0 ... 0 0110 . 1 . 0 ... 0 @2op +VMIN_S 111 0 1111 0 . .. ... 0 ... 0 0110 . 1 . 1 ... 0 @2op +VMIN_U 111 1 1111 0 . .. ... 0 ... 0 0110 . 1 . 1 ... 0 @2op + # Vector miscellaneous VCLS 1111 1111 1 . 11 .. 00 ... 0 0100 01 . 0 ... 0 @1op diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ DO_1OP(vfnegs, 8, uint64_t, DO_FNEGS) DO_2OP(OP##h, 2, uint16_t, FN) \ DO_2OP(OP##w, 4, uint32_t, FN) +/* provide signed 2-op helpers for all sizes */ +#define DO_2OP_S(OP, FN) \ + DO_2OP(OP##b, 1, int8_t, FN) \ + DO_2OP(OP##h, 2, int16_t, FN) \ + DO_2OP(OP##w, 4, int32_t, FN) + #define DO_AND(N, M) ((N) & (M)) #define DO_BIC(N, M) ((N) & ~(M)) #define DO_ORR(N, M) ((N) | (M)) @@ -XXX,XX +XXX,XX @@ DO_2OP(vrmulhsw, 4, int32_t, do_rmulh_w) DO_2OP(vrmulhub, 1, uint8_t, do_rmulh_b) DO_2OP(vrmulhuh, 2, uint16_t, do_rmulh_h) DO_2OP(vrmulhuw, 4, uint32_t, do_rmulh_w) + +#define DO_MAX(N, M) ((N) >= (M) ? (N) : (M)) +#define DO_MIN(N, M) ((N) >= (M) ? (M) : (N)) + +DO_2OP_S(vmaxs, DO_MAX) +DO_2OP_U(vmaxu, DO_MAX) +DO_2OP_S(vmins, DO_MIN) +DO_2OP_U(vminu, DO_MIN) diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ DO_2OP(VMULH_S, vmulhs) DO_2OP(VMULH_U, vmulhu) DO_2OP(VRMULH_S, vrmulhs) DO_2OP(VRMULH_U, vrmulhu) +DO_2OP(VMAX_S, vmaxs) +DO_2OP(VMAX_U, vmaxu) +DO_2OP(VMIN_S, vmins) +DO_2OP(VMIN_U, vminu) -- 2.20.1 Implement the MVE VABD insn. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-17-peter.maydell@linaro.org --- target/arm/helper-mve.h | 7 +++++++ target/arm/mve.decode | 3 +++ target/arm/mve_helper.c | 5 +++++ target/arm/translate-mve.c | 2 ++ 4 files changed, 17 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(mve_vminsw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vminub, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vminuh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vminuw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + +DEF_HELPER_FLAGS_4(mve_vabdsb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vabdsh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vabdsw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vabdub, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vabduh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vabduw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ VMAX_U 111 1 1111 0 . .. ... 0 ... 0 0110 . 1 . 0 ... 0 @2op VMIN_S 111 0 1111 0 . .. ... 0 ... 0 0110 . 1 . 1 ... 0 @2op VMIN_U 111 1 1111 0 . .. ... 0 ... 0 0110 . 1 . 1 ... 0 @2op +VABD_S 111 0 1111 0 . .. ... 0 ... 0 0111 . 1 . 0 ... 0 @2op +VABD_U 111 1 1111 0 . .. ... 0 ... 0 0111 . 1 . 0 ... 0 @2op + # Vector miscellaneous VCLS 1111 1111 1 . 11 .. 00 ... 0 0100 01 . 0 ... 0 @1op diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ DO_2OP_S(vmaxs, DO_MAX) DO_2OP_U(vmaxu, DO_MAX) DO_2OP_S(vmins, DO_MIN) DO_2OP_U(vminu, DO_MIN) + +#define DO_ABD(N, M) ((N) >= (M) ? (N) - (M) : (M) - (N)) + +DO_2OP_S(vabds, DO_ABD) +DO_2OP_U(vabdu, DO_ABD) diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ DO_2OP(VMAX_S, vmaxs) DO_2OP(VMAX_U, vmaxu) DO_2OP(VMIN_S, vmins) DO_2OP(VMIN_U, vminu) +DO_2OP(VABD_S, vabds) +DO_2OP(VABD_U, vabdu) -- 2.20.1 Implement MVE VHADD and VHSUB insns, which perform an addition or subtraction and then halve the result. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-18-peter.maydell@linaro.org --- target/arm/helper-mve.h | 14 ++++++++++++++ target/arm/mve.decode | 5 +++++ target/arm/mve_helper.c | 25 +++++++++++++++++++++++++ target/arm/translate-mve.c | 4 ++++ 4 files changed, 48 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(mve_vabdsw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vabdub, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vabduh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vabduw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + +DEF_HELPER_FLAGS_4(mve_vhaddsb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vhaddsh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vhaddsw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vhaddub, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vhadduh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vhadduw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + +DEF_HELPER_FLAGS_4(mve_vhsubsb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vhsubsh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vhsubsw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vhsubub, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vhsubuh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vhsubuw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ VMIN_U 111 1 1111 0 . .. ... 0 ... 0 0110 . 1 . 1 ... 0 @2op VABD_S 111 0 1111 0 . .. ... 0 ... 0 0111 . 1 . 0 ... 0 @2op VABD_U 111 1 1111 0 . .. ... 0 ... 0 0111 . 1 . 0 ... 0 @2op +VHADD_S 111 0 1111 0 . .. ... 0 ... 0 0000 . 1 . 0 ... 0 @2op +VHADD_U 111 1 1111 0 . .. ... 0 ... 0 0000 . 1 . 0 ... 0 @2op +VHSUB_S 111 0 1111 0 . .. ... 0 ... 0 0010 . 1 . 0 ... 0 @2op +VHSUB_U 111 1 1111 0 . .. ... 0 ... 0 0010 . 1 . 0 ... 0 @2op + # Vector miscellaneous VCLS 1111 1111 1 . 11 .. 00 ... 0 0100 01 . 0 ... 0 @1op diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ DO_2OP_U(vminu, DO_MIN) DO_2OP_S(vabds, DO_ABD) DO_2OP_U(vabdu, DO_ABD) + +static inline uint32_t do_vhadd_u(uint32_t n, uint32_t m) +{ + return ((uint64_t)n + m) >> 1; +} + +static inline int32_t do_vhadd_s(int32_t n, int32_t m) +{ + return ((int64_t)n + m) >> 1; +} + +static inline uint32_t do_vhsub_u(uint32_t n, uint32_t m) +{ + return ((uint64_t)n - m) >> 1; +} + +static inline int32_t do_vhsub_s(int32_t n, int32_t m) +{ + return ((int64_t)n - m) >> 1; +} + +DO_2OP_S(vhadds, do_vhadd_s) +DO_2OP_U(vhaddu, do_vhadd_u) +DO_2OP_S(vhsubs, do_vhsub_s) +DO_2OP_U(vhsubu, do_vhsub_u) diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ DO_2OP(VMIN_S, vmins) DO_2OP(VMIN_U, vminu) DO_2OP(VABD_S, vabds) DO_2OP(VABD_U, vabdu) +DO_2OP(VHADD_S, vhadds) +DO_2OP(VHADD_U, vhaddu) +DO_2OP(VHSUB_S, vhsubs) +DO_2OP(VHSUB_U, vhsubu) -- 2.20.1 Implement the MVE VMULL insn, which multiplies two single width integer elements to produce a double width result. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-19-peter.maydell@linaro.org --- target/arm/helper-mve.h | 14 ++++++++++++++ target/arm/mve.decode | 5 +++++ target/arm/mve_helper.c | 34 ++++++++++++++++++++++++++++++++++ target/arm/translate-mve.c | 4 ++++ 4 files changed, 57 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(mve_vhsubsw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vhsubub, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vhsubuh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vhsubuw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + +DEF_HELPER_FLAGS_4(mve_vmullbsb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vmullbsh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vmullbsw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vmullbub, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vmullbuh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vmullbuw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + +DEF_HELPER_FLAGS_4(mve_vmulltsb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vmulltsh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vmulltsw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vmulltub, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vmulltuh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vmulltuw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ VHADD_U 111 1 1111 0 . .. ... 0 ... 0 0000 . 1 . 0 ... 0 @2op VHSUB_S 111 0 1111 0 . .. ... 0 ... 0 0010 . 1 . 0 ... 0 @2op VHSUB_U 111 1 1111 0 . .. ... 0 ... 0 0010 . 1 . 0 ... 0 @2op +VMULL_BS 111 0 1110 0 . .. ... 1 ... 0 1110 . 0 . 0 ... 0 @2op +VMULL_BU 111 1 1110 0 . .. ... 1 ... 0 1110 . 0 . 0 ... 0 @2op +VMULL_TS 111 0 1110 0 . .. ... 1 ... 1 1110 . 0 . 0 ... 0 @2op +VMULL_TU 111 1 1110 0 . .. ... 1 ... 1 1110 . 0 . 0 ... 0 @2op + # Vector miscellaneous VCLS 1111 1111 1 . 11 .. 00 ... 0 0100 01 . 0 ... 0 @1op diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ DO_1OP(vfnegs, 8, uint64_t, DO_FNEGS) DO_2OP(OP##h, 2, int16_t, FN) \ DO_2OP(OP##w, 4, int32_t, FN) +/* + * "Long" operations where two half-sized inputs (taken from either the + * top or the bottom of the input vector) produce a double-width result. + * Here ESIZE, TYPE are for the input, and LESIZE, LTYPE for the output. + */ +#define DO_2OP_L(OP, TOP, ESIZE, TYPE, LESIZE, LTYPE, FN) \ + void HELPER(glue(mve_, OP))(CPUARMState *env, void *vd, void *vn, void *vm) \ + { \ + LTYPE *d = vd; \ + TYPE *n = vn, *m = vm; \ + uint16_t mask = mve_element_mask(env); \ + unsigned le; \ + for (le = 0; le < 16 / LESIZE; le++, mask >>= LESIZE) { \ + LTYPE r = FN((LTYPE)n[H##ESIZE(le * 2 + TOP)], \ + m[H##ESIZE(le * 2 + TOP)]); \ + mergemask(&d[H##LESIZE(le)], r, mask); \ + } \ + mve_advance_vpt(env); \ + } + #define DO_AND(N, M) ((N) & (M)) #define DO_BIC(N, M) ((N) & ~(M)) #define DO_ORR(N, M) ((N) | (M)) @@ -XXX,XX +XXX,XX @@ DO_2OP_U(vadd, DO_ADD) DO_2OP_U(vsub, DO_SUB) DO_2OP_U(vmul, DO_MUL) +DO_2OP_L(vmullbsb, 0, 1, int8_t, 2, int16_t, DO_MUL) +DO_2OP_L(vmullbsh, 0, 2, int16_t, 4, int32_t, DO_MUL) +DO_2OP_L(vmullbsw, 0, 4, int32_t, 8, int64_t, DO_MUL) +DO_2OP_L(vmullbub, 0, 1, uint8_t, 2, uint16_t, DO_MUL) +DO_2OP_L(vmullbuh, 0, 2, uint16_t, 4, uint32_t, DO_MUL) +DO_2OP_L(vmullbuw, 0, 4, uint32_t, 8, uint64_t, DO_MUL) + +DO_2OP_L(vmulltsb, 1, 1, int8_t, 2, int16_t, DO_MUL) +DO_2OP_L(vmulltsh, 1, 2, int16_t, 4, int32_t, DO_MUL) +DO_2OP_L(vmulltsw, 1, 4, int32_t, 8, int64_t, DO_MUL) +DO_2OP_L(vmulltub, 1, 1, uint8_t, 2, uint16_t, DO_MUL) +DO_2OP_L(vmulltuh, 1, 2, uint16_t, 4, uint32_t, DO_MUL) +DO_2OP_L(vmulltuw, 1, 4, uint32_t, 8, uint64_t, DO_MUL) + /* * Because the computation type is at least twice as large as required, * these work for both signed and unsigned source types. diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ DO_2OP(VHADD_S, vhadds) DO_2OP(VHADD_U, vhaddu) DO_2OP(VHSUB_S, vhsubs) DO_2OP(VHSUB_U, vhsubu) +DO_2OP(VMULL_BS, vmullbs) +DO_2OP(VMULL_BU, vmullbu) +DO_2OP(VMULL_TS, vmullts) +DO_2OP(VMULL_TU, vmulltu) -- 2.20.1 Implement the MVE VMLALDAV insn, which multiplies pairs of integer elements, accumulating them into a 64-bit result in a pair of general-purpose registers. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-20-peter.maydell@linaro.org --- target/arm/helper-mve.h | 8 ++++ target/arm/translate.h | 10 ++++ target/arm/mve.decode | 15 ++++++ target/arm/mve_helper.c | 34 ++++++++++++++ target/arm/translate-mve.c | 96 ++++++++++++++++++++++++++++++++++++++ 5 files changed, 163 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(mve_vmulltsw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vmulltub, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vmulltuh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vmulltuw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + +DEF_HELPER_FLAGS_4(mve_vmlaldavsh, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) +DEF_HELPER_FLAGS_4(mve_vmlaldavsw, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) +DEF_HELPER_FLAGS_4(mve_vmlaldavxsh, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) +DEF_HELPER_FLAGS_4(mve_vmlaldavxsw, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) + +DEF_HELPER_FLAGS_4(mve_vmlaldavuh, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) +DEF_HELPER_FLAGS_4(mve_vmlaldavuw, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) 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 negate(DisasContext *s, int x) return -x; } +static inline int plus_1(DisasContext *s, int x) +{ + return x + 1; +} + static inline int plus_2(DisasContext *s, int x) { return x + 2; @@ -XXX,XX +XXX,XX @@ static inline int times_4(DisasContext *s, int x) return x * 4; } +static inline int times_2_plus_1(DisasContext *s, int x) +{ + return x * 2 + 1; +} + static inline int arm_dc_feature(DisasContext *dc, int feature) { return (dc->features & (1ULL << feature)) != 0; diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ VNEG_fp 1111 1111 1 . 11 .. 01 ... 0 0111 11 . 0 ... 0 @1op VDUP 1110 1110 1 1 10 ... 0 .... 1011 . 0 0 1 0000 @vdup size=0 VDUP 1110 1110 1 0 10 ... 0 .... 1011 . 0 1 1 0000 @vdup size=1 VDUP 1110 1110 1 0 10 ... 0 .... 1011 . 0 0 1 0000 @vdup size=2 + +# multiply-add long dual accumulate +# rdahi: bits [3:1] from insn, bit 0 is 1 +# rdalo: bits [3:1] from insn, bit 0 is 0 +%rdahi 20:3 !function=times_2_plus_1 +%rdalo 13:3 !function=times_2 +# size bit is 0 for 16 bit, 1 for 32 bit +%size_16 16:1 !function=plus_1 + +&vmlaldav rdahi rdalo size qn qm x a + +@vmlaldav .... .... . ... ... . ... . .... .... qm:3 . \ + qn=%qn rdahi=%rdahi rdalo=%rdalo size=%size_16 &vmlaldav +VMLALDAV_S 1110 1110 1 ... ... . ... x:1 1110 . 0 a:1 0 ... 0 @vmlaldav +VMLALDAV_U 1111 1110 1 ... ... . ... x:1 1110 . 0 a:1 0 ... 0 @vmlaldav diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ DO_2OP_S(vhadds, do_vhadd_s) DO_2OP_U(vhaddu, do_vhadd_u) DO_2OP_S(vhsubs, do_vhsub_s) DO_2OP_U(vhsubu, do_vhsub_u) + + +/* + * Multiply add long dual accumulate ops. + */ +#define DO_LDAV(OP, ESIZE, TYPE, XCHG, EVENACC, ODDACC) \ + uint64_t HELPER(glue(mve_, OP))(CPUARMState *env, void *vn, \ + void *vm, uint64_t a) \ + { \ + uint16_t mask = mve_element_mask(env); \ + unsigned e; \ + TYPE *n = vn, *m = vm; \ + for (e = 0; e < 16 / ESIZE; e++, mask >>= ESIZE) { \ + if (mask & 1) { \ + if (e & 1) { \ + a ODDACC \ + (int64_t)n[H##ESIZE(e - 1 * XCHG)] * m[H##ESIZE(e)]; \ + } else { \ + a EVENACC \ + (int64_t)n[H##ESIZE(e + 1 * XCHG)] * m[H##ESIZE(e)]; \ + } \ + } \ + } \ + mve_advance_vpt(env); \ + return a; \ + } + +DO_LDAV(vmlaldavsh, 2, int16_t, false, +=, +=) +DO_LDAV(vmlaldavxsh, 2, int16_t, true, +=, +=) +DO_LDAV(vmlaldavsw, 4, int32_t, false, +=, +=) +DO_LDAV(vmlaldavxsw, 4, int32_t, true, +=, +=) + +DO_LDAV(vmlaldavuh, 2, uint16_t, false, +=, +=) +DO_LDAV(vmlaldavuw, 4, uint32_t, false, +=, +=) diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ typedef void MVEGenLdStFn(TCGv_ptr, TCGv_ptr, TCGv_i32); typedef void MVEGenOneOpFn(TCGv_ptr, TCGv_ptr, TCGv_ptr); typedef void MVEGenTwoOpFn(TCGv_ptr, TCGv_ptr, TCGv_ptr, TCGv_ptr); +typedef void MVEGenDualAccOpFn(TCGv_i64, TCGv_ptr, TCGv_ptr, TCGv_ptr, TCGv_i64); /* Return the offset of a Qn register (same semantics as aa32_vfp_qreg()) */ static inline long mve_qreg_offset(unsigned reg) @@ -XXX,XX +XXX,XX @@ static void mve_update_eci(DisasContext *s) } } +static bool mve_skip_first_beat(DisasContext *s) +{ + /* Return true if PSR.ECI says we must skip the first beat of this insn */ + switch (s->eci) { + case ECI_NONE: + return false; + case ECI_A0: + case ECI_A0A1: + case ECI_A0A1A2: + case ECI_A0A1A2B0: + return true; + default: + g_assert_not_reached(); + } +} + static bool do_ldst(DisasContext *s, arg_VLDR_VSTR *a, MVEGenLdStFn *fn) { TCGv_i32 addr; @@ -XXX,XX +XXX,XX @@ DO_2OP(VMULL_BS, vmullbs) DO_2OP(VMULL_BU, vmullbu) DO_2OP(VMULL_TS, vmullts) DO_2OP(VMULL_TU, vmulltu) + +static bool do_long_dual_acc(DisasContext *s, arg_vmlaldav *a, + MVEGenDualAccOpFn *fn) +{ + TCGv_ptr qn, qm; + TCGv_i64 rda; + TCGv_i32 rdalo, rdahi; + + if (!dc_isar_feature(aa32_mve, s) || + !mve_check_qreg_bank(s, a->qn | a->qm) || + !fn) { + return false; + } + /* + * rdahi == 13 is UNPREDICTABLE; rdahi == 15 is a related + * encoding; rdalo always has bit 0 clear so cannot be 13 or 15. + */ + if (a->rdahi == 13 || a->rdahi == 15) { + return false; + } + if (!mve_eci_check(s) || !vfp_access_check(s)) { + return true; + } + + qn = mve_qreg_ptr(a->qn); + qm = mve_qreg_ptr(a->qm); + + /* + * This insn is subject to beat-wise execution. Partial execution + * of an A=0 (no-accumulate) insn which does not execute the first + * beat must start with the current rda value, not 0. + */ + if (a->a || mve_skip_first_beat(s)) { + rda = tcg_temp_new_i64(); + rdalo = load_reg(s, a->rdalo); + rdahi = load_reg(s, a->rdahi); + tcg_gen_concat_i32_i64(rda, rdalo, rdahi); + tcg_temp_free_i32(rdalo); + tcg_temp_free_i32(rdahi); + } else { + rda = tcg_const_i64(0); + } + + fn(rda, cpu_env, qn, qm, rda); + tcg_temp_free_ptr(qn); + tcg_temp_free_ptr(qm); + + rdalo = tcg_temp_new_i32(); + rdahi = tcg_temp_new_i32(); + tcg_gen_extrl_i64_i32(rdalo, rda); + tcg_gen_extrh_i64_i32(rdahi, rda); + store_reg(s, a->rdalo, rdalo); + store_reg(s, a->rdahi, rdahi); + tcg_temp_free_i64(rda); + mve_update_eci(s); + return true; +} + +static bool trans_VMLALDAV_S(DisasContext *s, arg_vmlaldav *a) +{ + static MVEGenDualAccOpFn * const fns[4][2] = { + { NULL, NULL }, + { gen_helper_mve_vmlaldavsh, gen_helper_mve_vmlaldavxsh }, + { gen_helper_mve_vmlaldavsw, gen_helper_mve_vmlaldavxsw }, + { NULL, NULL }, + }; + return do_long_dual_acc(s, a, fns[a->size][a->x]); +} + +static bool trans_VMLALDAV_U(DisasContext *s, arg_vmlaldav *a) +{ + static MVEGenDualAccOpFn * const fns[4][2] = { + { NULL, NULL }, + { gen_helper_mve_vmlaldavuh, NULL }, + { gen_helper_mve_vmlaldavuw, NULL }, + { NULL, NULL }, + }; + return do_long_dual_acc(s, a, fns[a->size][a->x]); +} -- 2.20.1 Implement the MVE insn VMLSLDAV, which multiplies source elements, alternately adding and subtracting them, and accumulates into a 64-bit result in a pair of general purpose registers. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-21-peter.maydell@linaro.org --- target/arm/helper-mve.h | 5 +++++ target/arm/mve.decode | 2 ++ target/arm/mve_helper.c | 5 +++++ target/arm/translate-mve.c | 11 +++++++++++ 4 files changed, 23 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(mve_vmlaldavxsw, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) DEF_HELPER_FLAGS_4(mve_vmlaldavuh, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) DEF_HELPER_FLAGS_4(mve_vmlaldavuw, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) + +DEF_HELPER_FLAGS_4(mve_vmlsldavsh, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) +DEF_HELPER_FLAGS_4(mve_vmlsldavsw, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) +DEF_HELPER_FLAGS_4(mve_vmlsldavxsh, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) +DEF_HELPER_FLAGS_4(mve_vmlsldavxsw, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ VDUP 1110 1110 1 0 10 ... 0 .... 1011 . 0 0 1 0000 @vdup size=2 qn=%qn rdahi=%rdahi rdalo=%rdalo size=%size_16 &vmlaldav VMLALDAV_S 1110 1110 1 ... ... . ... x:1 1110 . 0 a:1 0 ... 0 @vmlaldav VMLALDAV_U 1111 1110 1 ... ... . ... x:1 1110 . 0 a:1 0 ... 0 @vmlaldav + +VMLSLDAV 1110 1110 1 ... ... . ... x:1 1110 . 0 a:1 0 ... 1 @vmlaldav diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ DO_LDAV(vmlaldavxsw, 4, int32_t, true, +=, +=) DO_LDAV(vmlaldavuh, 2, uint16_t, false, +=, +=) DO_LDAV(vmlaldavuw, 4, uint32_t, false, +=, +=) + +DO_LDAV(vmlsldavsh, 2, int16_t, false, +=, -=) +DO_LDAV(vmlsldavxsh, 2, int16_t, true, +=, -=) +DO_LDAV(vmlsldavsw, 4, int32_t, false, +=, -=) +DO_LDAV(vmlsldavxsw, 4, int32_t, true, +=, -=) diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ static bool trans_VMLALDAV_U(DisasContext *s, arg_vmlaldav *a) }; return do_long_dual_acc(s, a, fns[a->size][a->x]); } + +static bool trans_VMLSLDAV(DisasContext *s, arg_vmlaldav *a) +{ + static MVEGenDualAccOpFn * const fns[4][2] = { + { NULL, NULL }, + { gen_helper_mve_vmlsldavsh, gen_helper_mve_vmlsldavxsh }, + { gen_helper_mve_vmlsldavsw, gen_helper_mve_vmlsldavxsw }, + { NULL, NULL }, + }; + return do_long_dual_acc(s, a, fns[a->size][a->x]); +} -- 2.20.1 Implement the MVE VRMLALDAVH and VRMLSLDAVH insns, which accumulate the results of a rounded multiply of pairs of elements into a 72-bit accumulator, returning the top 64 bits in a pair of general purpose registers. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-22-peter.maydell@linaro.org --- target/arm/helper-mve.h | 8 ++++++++ target/arm/mve.decode | 7 +++++++ target/arm/mve_helper.c | 37 +++++++++++++++++++++++++++++++++++++ target/arm/translate-mve.c | 24 ++++++++++++++++++++++++ 4 files changed, 76 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(mve_vmlsldavsh, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) DEF_HELPER_FLAGS_4(mve_vmlsldavsw, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) DEF_HELPER_FLAGS_4(mve_vmlsldavxsh, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) DEF_HELPER_FLAGS_4(mve_vmlsldavxsw, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) + +DEF_HELPER_FLAGS_4(mve_vrmlaldavhsw, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) +DEF_HELPER_FLAGS_4(mve_vrmlaldavhxsw, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) + +DEF_HELPER_FLAGS_4(mve_vrmlaldavhuw, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) + +DEF_HELPER_FLAGS_4(mve_vrmlsldavhsw, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) +DEF_HELPER_FLAGS_4(mve_vrmlsldavhxsw, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ VDUP 1110 1110 1 0 10 ... 0 .... 1011 . 0 0 1 0000 @vdup size=2 @vmlaldav .... .... . ... ... . ... . .... .... qm:3 . \ qn=%qn rdahi=%rdahi rdalo=%rdalo size=%size_16 &vmlaldav +@vmlaldav_nosz .... .... . ... ... . ... . .... .... qm:3 . \ + qn=%qn rdahi=%rdahi rdalo=%rdalo size=0 &vmlaldav VMLALDAV_S 1110 1110 1 ... ... . ... x:1 1110 . 0 a:1 0 ... 0 @vmlaldav VMLALDAV_U 1111 1110 1 ... ... . ... x:1 1110 . 0 a:1 0 ... 0 @vmlaldav VMLSLDAV 1110 1110 1 ... ... . ... x:1 1110 . 0 a:1 0 ... 1 @vmlaldav + +VRMLALDAVH_S 1110 1110 1 ... ... 0 ... x:1 1111 . 0 a:1 0 ... 0 @vmlaldav_nosz +VRMLALDAVH_U 1111 1110 1 ... ... 0 ... x:1 1111 . 0 a:1 0 ... 0 @vmlaldav_nosz + +VRMLSLDAVH 1111 1110 1 ... ... 0 ... x:1 1110 . 0 a:1 0 ... 1 @vmlaldav_nosz diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ */ #include "qemu/osdep.h" +#include "qemu/int128.h" #include "cpu.h" #include "internals.h" #include "vec_internal.h" @@ -XXX,XX +XXX,XX @@ DO_LDAV(vmlsldavsh, 2, int16_t, false, +=, -=) DO_LDAV(vmlsldavxsh, 2, int16_t, true, +=, -=) DO_LDAV(vmlsldavsw, 4, int32_t, false, +=, -=) DO_LDAV(vmlsldavxsw, 4, int32_t, true, +=, -=) + +/* + * Rounding multiply add long dual accumulate high: we must keep + * a 72-bit internal accumulator value and return the top 64 bits. + */ +#define DO_LDAVH(OP, ESIZE, TYPE, XCHG, EVENACC, ODDACC, TO128) \ + uint64_t HELPER(glue(mve_, OP))(CPUARMState *env, void *vn, \ + void *vm, uint64_t a) \ + { \ + uint16_t mask = mve_element_mask(env); \ + unsigned e; \ + TYPE *n = vn, *m = vm; \ + Int128 acc = int128_lshift(TO128(a), 8); \ + for (e = 0; e < 16 / ESIZE; e++, mask >>= ESIZE) { \ + if (mask & 1) { \ + if (e & 1) { \ + acc = ODDACC(acc, TO128(n[H##ESIZE(e - 1 * XCHG)] * \ + m[H##ESIZE(e)])); \ + } else { \ + acc = EVENACC(acc, TO128(n[H##ESIZE(e + 1 * XCHG)] * \ + m[H##ESIZE(e)])); \ + } \ + acc = int128_add(acc, 1 << 7); \ + } \ + } \ + mve_advance_vpt(env); \ + return int128_getlo(int128_rshift(acc, 8)); \ + } + +DO_LDAVH(vrmlaldavhsw, 4, int32_t, false, int128_add, int128_add, int128_makes64) +DO_LDAVH(vrmlaldavhxsw, 4, int32_t, true, int128_add, int128_add, int128_makes64) + +DO_LDAVH(vrmlaldavhuw, 4, uint32_t, false, int128_add, int128_add, int128_make64) + +DO_LDAVH(vrmlsldavhsw, 4, int32_t, false, int128_add, int128_sub, int128_makes64) +DO_LDAVH(vrmlsldavhxsw, 4, int32_t, true, int128_add, int128_sub, int128_makes64) diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ static bool trans_VMLSLDAV(DisasContext *s, arg_vmlaldav *a) }; return do_long_dual_acc(s, a, fns[a->size][a->x]); } + +static bool trans_VRMLALDAVH_S(DisasContext *s, arg_vmlaldav *a) +{ + static MVEGenDualAccOpFn * const fns[] = { + gen_helper_mve_vrmlaldavhsw, gen_helper_mve_vrmlaldavhxsw, + }; + return do_long_dual_acc(s, a, fns[a->x]); +} + +static bool trans_VRMLALDAVH_U(DisasContext *s, arg_vmlaldav *a) +{ + static MVEGenDualAccOpFn * const fns[] = { + gen_helper_mve_vrmlaldavhuw, NULL, + }; + return do_long_dual_acc(s, a, fns[a->x]); +} + +static bool trans_VRMLSLDAVH(DisasContext *s, arg_vmlaldav *a) +{ + static MVEGenDualAccOpFn * const fns[] = { + gen_helper_mve_vrmlsldavhsw, gen_helper_mve_vrmlsldavhxsw, + }; + return do_long_dual_acc(s, a, fns[a->x]); +} -- 2.20.1 Implement the scalar form of the MVE VADD insn. This takes the scalar operand from a general purpose register. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-23-peter.maydell@linaro.org --- target/arm/helper-mve.h | 4 ++++ target/arm/mve.decode | 7 ++++++ target/arm/mve_helper.c | 22 +++++++++++++++++++ target/arm/translate-mve.c | 45 ++++++++++++++++++++++++++++++++++++++ 4 files changed, 78 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(mve_vmulltub, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vmulltuh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vmulltuw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vadd_scalarb, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vadd_scalarh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vadd_scalarw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) + DEF_HELPER_FLAGS_4(mve_vmlaldavsh, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) DEF_HELPER_FLAGS_4(mve_vmlaldavsw, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) DEF_HELPER_FLAGS_4(mve_vmlaldavxsh, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ &vldr_vstr rn qd imm p a w size l u &1op qd qm size &2op qd qm qn size +&2scalar qd qn rm size @vldr_vstr ....... . . . . l:1 rn:4 ... ...... imm:7 &vldr_vstr qd=%qd u=0 # Note that both Rn and Qd are 3 bits only (no D bit) @@ -XXX,XX +XXX,XX @@ @2op .... .... .. size:2 .... .... .... .... .... &2op qd=%qd qm=%qm qn=%qn @2op_nosz .... .... .... .... .... .... .... .... &2op qd=%qd qm=%qm qn=%qn size=0 +@2scalar .... .... .. size:2 .... .... .... .... rm:4 &2scalar qd=%qd qn=%qn + # Vector loads and stores # Widening loads and narrowing stores: @@ -XXX,XX +XXX,XX @@ VRMLALDAVH_S 1110 1110 1 ... ... 0 ... x:1 1111 . 0 a:1 0 ... 0 @vmlaldav_no VRMLALDAVH_U 1111 1110 1 ... ... 0 ... x:1 1111 . 0 a:1 0 ... 0 @vmlaldav_nosz VRMLSLDAVH 1111 1110 1 ... ... 0 ... x:1 1110 . 0 a:1 0 ... 1 @vmlaldav_nosz + +# Scalar operations + +VADD_scalar 1110 1110 0 . .. ... 1 ... 0 1111 . 100 .... @2scalar diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ DO_2OP_S(vhsubs, do_vhsub_s) DO_2OP_U(vhsubu, do_vhsub_u) +#define DO_2OP_SCALAR(OP, ESIZE, TYPE, FN) \ + void HELPER(glue(mve_, OP))(CPUARMState *env, void *vd, void *vn, \ + uint32_t rm) \ + { \ + TYPE *d = vd, *n = vn; \ + TYPE m = rm; \ + uint16_t mask = mve_element_mask(env); \ + unsigned e; \ + for (e = 0; e < 16 / ESIZE; e++, mask >>= ESIZE) { \ + mergemask(&d[H##ESIZE(e)], FN(n[H##ESIZE(e)], m), mask); \ + } \ + mve_advance_vpt(env); \ + } + +/* provide unsigned 2-op scalar helpers for all sizes */ +#define DO_2OP_SCALAR_U(OP, FN) \ + DO_2OP_SCALAR(OP##b, 1, uint8_t, FN) \ + DO_2OP_SCALAR(OP##h, 2, uint16_t, FN) \ + DO_2OP_SCALAR(OP##w, 4, uint32_t, FN) + +DO_2OP_SCALAR_U(vadd_scalar, DO_ADD) + /* * Multiply add long dual accumulate ops. */ diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ typedef void MVEGenLdStFn(TCGv_ptr, TCGv_ptr, TCGv_i32); typedef void MVEGenOneOpFn(TCGv_ptr, TCGv_ptr, TCGv_ptr); typedef void MVEGenTwoOpFn(TCGv_ptr, TCGv_ptr, TCGv_ptr, TCGv_ptr); +typedef void MVEGenTwoOpScalarFn(TCGv_ptr, TCGv_ptr, TCGv_ptr, TCGv_i32); typedef void MVEGenDualAccOpFn(TCGv_i64, TCGv_ptr, TCGv_ptr, TCGv_ptr, TCGv_i64); /* Return the offset of a Qn register (same semantics as aa32_vfp_qreg()) */ @@ -XXX,XX +XXX,XX @@ DO_2OP(VMULL_BU, vmullbu) DO_2OP(VMULL_TS, vmullts) DO_2OP(VMULL_TU, vmulltu) +static bool do_2op_scalar(DisasContext *s, arg_2scalar *a, + MVEGenTwoOpScalarFn fn) +{ + TCGv_ptr qd, qn; + TCGv_i32 rm; + + if (!dc_isar_feature(aa32_mve, s) || + !mve_check_qreg_bank(s, a->qd | a->qn) || + !fn) { + return false; + } + if (a->rm == 13 || a->rm == 15) { + /* UNPREDICTABLE */ + return false; + } + if (!mve_eci_check(s) || !vfp_access_check(s)) { + return true; + } + + qd = mve_qreg_ptr(a->qd); + qn = mve_qreg_ptr(a->qn); + rm = load_reg(s, a->rm); + fn(cpu_env, qd, qn, rm); + tcg_temp_free_i32(rm); + tcg_temp_free_ptr(qd); + tcg_temp_free_ptr(qn); + mve_update_eci(s); + return true; +} + +#define DO_2OP_SCALAR(INSN, FN) \ + static bool trans_##INSN(DisasContext *s, arg_2scalar *a) \ + { \ + static MVEGenTwoOpScalarFn * const fns[] = { \ + gen_helper_mve_##FN##b, \ + gen_helper_mve_##FN##h, \ + gen_helper_mve_##FN##w, \ + NULL, \ + }; \ + return do_2op_scalar(s, a, fns[a->size]); \ + } + +DO_2OP_SCALAR(VADD_scalar, vadd_scalar) + static bool do_long_dual_acc(DisasContext *s, arg_vmlaldav *a, MVEGenDualAccOpFn *fn) { -- 2.20.1 Implement the scalar forms of the MVE VSUB and VMUL insns. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-24-peter.maydell@linaro.org --- target/arm/helper-mve.h | 8 ++++++++ target/arm/mve.decode | 2 ++ target/arm/mve_helper.c | 2 ++ target/arm/translate-mve.c | 2 ++ 4 files changed, 14 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(mve_vadd_scalarb, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vadd_scalarh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vadd_scalarw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vsub_scalarb, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vsub_scalarh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vsub_scalarw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) + +DEF_HELPER_FLAGS_4(mve_vmul_scalarb, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vmul_scalarh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vmul_scalarw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) + DEF_HELPER_FLAGS_4(mve_vmlaldavsh, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) DEF_HELPER_FLAGS_4(mve_vmlaldavsw, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) DEF_HELPER_FLAGS_4(mve_vmlaldavxsh, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ VRMLSLDAVH 1111 1110 1 ... ... 0 ... x:1 1110 . 0 a:1 0 ... 1 @vmlaldav_no # Scalar operations VADD_scalar 1110 1110 0 . .. ... 1 ... 0 1111 . 100 .... @2scalar +VSUB_scalar 1110 1110 0 . .. ... 1 ... 1 1111 . 100 .... @2scalar +VMUL_scalar 1110 1110 0 . .. ... 1 ... 1 1110 . 110 .... @2scalar diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ DO_2OP_U(vhsubu, do_vhsub_u) DO_2OP_SCALAR(OP##w, 4, uint32_t, FN) DO_2OP_SCALAR_U(vadd_scalar, DO_ADD) +DO_2OP_SCALAR_U(vsub_scalar, DO_SUB) +DO_2OP_SCALAR_U(vmul_scalar, DO_MUL) /* * Multiply add long dual accumulate ops. diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ static bool do_2op_scalar(DisasContext *s, arg_2scalar *a, } DO_2OP_SCALAR(VADD_scalar, vadd_scalar) +DO_2OP_SCALAR(VSUB_scalar, vsub_scalar) +DO_2OP_SCALAR(VMUL_scalar, vmul_scalar) static bool do_long_dual_acc(DisasContext *s, arg_vmlaldav *a, MVEGenDualAccOpFn *fn) -- 2.20.1 Implement the scalar variants of the MVE VHADD and VHSUB insns. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-25-peter.maydell@linaro.org --- target/arm/helper-mve.h | 16 ++++++++++++++++ target/arm/mve.decode | 4 ++++ target/arm/mve_helper.c | 8 ++++++++ target/arm/translate-mve.c | 4 ++++ 4 files changed, 32 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(mve_vmul_scalarb, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vmul_scalarh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vmul_scalarw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vhadds_scalarb, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vhadds_scalarh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vhadds_scalarw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) + +DEF_HELPER_FLAGS_4(mve_vhaddu_scalarb, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vhaddu_scalarh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vhaddu_scalarw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) + +DEF_HELPER_FLAGS_4(mve_vhsubs_scalarb, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vhsubs_scalarh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vhsubs_scalarw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) + +DEF_HELPER_FLAGS_4(mve_vhsubu_scalarb, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vhsubu_scalarh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vhsubu_scalarw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) + DEF_HELPER_FLAGS_4(mve_vmlaldavsh, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) DEF_HELPER_FLAGS_4(mve_vmlaldavsw, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) DEF_HELPER_FLAGS_4(mve_vmlaldavxsh, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ VRMLSLDAVH 1111 1110 1 ... ... 0 ... x:1 1110 . 0 a:1 0 ... 1 @vmlaldav_no VADD_scalar 1110 1110 0 . .. ... 1 ... 0 1111 . 100 .... @2scalar VSUB_scalar 1110 1110 0 . .. ... 1 ... 1 1111 . 100 .... @2scalar VMUL_scalar 1110 1110 0 . .. ... 1 ... 1 1110 . 110 .... @2scalar +VHADD_S_scalar 1110 1110 0 . .. ... 0 ... 0 1111 . 100 .... @2scalar +VHADD_U_scalar 1111 1110 0 . .. ... 0 ... 0 1111 . 100 .... @2scalar +VHSUB_S_scalar 1110 1110 0 . .. ... 0 ... 1 1111 . 100 .... @2scalar +VHSUB_U_scalar 1111 1110 0 . .. ... 0 ... 1 1111 . 100 .... @2scalar diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ DO_2OP_U(vhsubu, do_vhsub_u) DO_2OP_SCALAR(OP##b, 1, uint8_t, FN) \ DO_2OP_SCALAR(OP##h, 2, uint16_t, FN) \ DO_2OP_SCALAR(OP##w, 4, uint32_t, FN) +#define DO_2OP_SCALAR_S(OP, FN) \ + DO_2OP_SCALAR(OP##b, 1, int8_t, FN) \ + DO_2OP_SCALAR(OP##h, 2, int16_t, FN) \ + DO_2OP_SCALAR(OP##w, 4, int32_t, FN) DO_2OP_SCALAR_U(vadd_scalar, DO_ADD) DO_2OP_SCALAR_U(vsub_scalar, DO_SUB) DO_2OP_SCALAR_U(vmul_scalar, DO_MUL) +DO_2OP_SCALAR_S(vhadds_scalar, do_vhadd_s) +DO_2OP_SCALAR_U(vhaddu_scalar, do_vhadd_u) +DO_2OP_SCALAR_S(vhsubs_scalar, do_vhsub_s) +DO_2OP_SCALAR_U(vhsubu_scalar, do_vhsub_u) /* * Multiply add long dual accumulate ops. diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ static bool do_2op_scalar(DisasContext *s, arg_2scalar *a, DO_2OP_SCALAR(VADD_scalar, vadd_scalar) DO_2OP_SCALAR(VSUB_scalar, vsub_scalar) DO_2OP_SCALAR(VMUL_scalar, vmul_scalar) +DO_2OP_SCALAR(VHADD_S_scalar, vhadds_scalar) +DO_2OP_SCALAR(VHADD_U_scalar, vhaddu_scalar) +DO_2OP_SCALAR(VHSUB_S_scalar, vhsubs_scalar) +DO_2OP_SCALAR(VHSUB_U_scalar, vhsubu_scalar) static bool do_long_dual_acc(DisasContext *s, arg_vmlaldav *a, MVEGenDualAccOpFn *fn) -- 2.20.1 Implement the MVE VBRSR insn, which reverses a specified number of bits in each element, setting the rest to zero. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-26-peter.maydell@linaro.org --- target/arm/helper-mve.h | 4 ++++ target/arm/mve.decode | 1 + target/arm/mve_helper.c | 43 ++++++++++++++++++++++++++++++++++++++ target/arm/translate-mve.c | 1 + 4 files changed, 49 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(mve_vhsubu_scalarb, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vhsubu_scalarh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vhsubu_scalarw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vbrsrb, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vbrsrh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vbrsrw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) + DEF_HELPER_FLAGS_4(mve_vmlaldavsh, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) DEF_HELPER_FLAGS_4(mve_vmlaldavsw, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) DEF_HELPER_FLAGS_4(mve_vmlaldavxsh, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ VHADD_S_scalar 1110 1110 0 . .. ... 0 ... 0 1111 . 100 .... @2scalar VHADD_U_scalar 1111 1110 0 . .. ... 0 ... 0 1111 . 100 .... @2scalar VHSUB_S_scalar 1110 1110 0 . .. ... 0 ... 1 1111 . 100 .... @2scalar VHSUB_U_scalar 1111 1110 0 . .. ... 0 ... 1 1111 . 100 .... @2scalar +VBRSR 1111 1110 0 . .. ... 1 ... 1 1110 . 110 .... @2scalar diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ DO_2OP_SCALAR_U(vhaddu_scalar, do_vhadd_u) DO_2OP_SCALAR_S(vhsubs_scalar, do_vhsub_s) DO_2OP_SCALAR_U(vhsubu_scalar, do_vhsub_u) +static inline uint32_t do_vbrsrb(uint32_t n, uint32_t m) +{ + m &= 0xff; + if (m == 0) { + return 0; + } + n = revbit8(n); + if (m < 8) { + n >>= 8 - m; + } + return n; +} + +static inline uint32_t do_vbrsrh(uint32_t n, uint32_t m) +{ + m &= 0xff; + if (m == 0) { + return 0; + } + n = revbit16(n); + if (m < 16) { + n >>= 16 - m; + } + return n; +} + +static inline uint32_t do_vbrsrw(uint32_t n, uint32_t m) +{ + m &= 0xff; + if (m == 0) { + return 0; + } + n = revbit32(n); + if (m < 32) { + n >>= 32 - m; + } + return n; +} + +DO_2OP_SCALAR(vbrsrb, 1, uint8_t, do_vbrsrb) +DO_2OP_SCALAR(vbrsrh, 2, uint16_t, do_vbrsrh) +DO_2OP_SCALAR(vbrsrw, 4, uint32_t, do_vbrsrw) + /* * Multiply add long dual accumulate ops. */ diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ DO_2OP_SCALAR(VHADD_S_scalar, vhadds_scalar) DO_2OP_SCALAR(VHADD_U_scalar, vhaddu_scalar) DO_2OP_SCALAR(VHSUB_S_scalar, vhsubs_scalar) DO_2OP_SCALAR(VHSUB_U_scalar, vhsubu_scalar) +DO_2OP_SCALAR(VBRSR, vbrsr) static bool do_long_dual_acc(DisasContext *s, arg_vmlaldav *a, MVEGenDualAccOpFn *fn) -- 2.20.1 Implement the MVE VPST insn, which sets the predicate mask fields in the VPR to the immediate value encoded in the insn. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-27-peter.maydell@linaro.org --- target/arm/mve.decode | 4 +++ target/arm/translate-mve.c | 59 ++++++++++++++++++++++++++++++++++++++ 2 files changed, 63 insertions(+) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ VHADD_U_scalar 1111 1110 0 . .. ... 0 ... 0 1111 . 100 .... @2scalar VHSUB_S_scalar 1110 1110 0 . .. ... 0 ... 1 1111 . 100 .... @2scalar VHSUB_U_scalar 1111 1110 0 . .. ... 0 ... 1 1111 . 100 .... @2scalar VBRSR 1111 1110 0 . .. ... 1 ... 1 1110 . 110 .... @2scalar + +# Predicate operations +%mask_22_13 22:1 13:3 +VPST 1111 1110 0 . 11 000 1 ... 0 1111 0100 1101 mask=%mask_22_13 diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ static void mve_update_eci(DisasContext *s) } } +static void mve_update_and_store_eci(DisasContext *s) +{ + /* + * For insns which don't call a helper function that will call + * mve_advance_vpt(), this version updates s->eci and also stores + * it out to the CPUState field. + */ + if (s->eci) { + mve_update_eci(s); + store_cpu_field(tcg_constant_i32(s->eci << 4), condexec_bits); + } +} + static bool mve_skip_first_beat(DisasContext *s) { /* Return true if PSR.ECI says we must skip the first beat of this insn */ @@ -XXX,XX +XXX,XX @@ static bool trans_VRMLSLDAVH(DisasContext *s, arg_vmlaldav *a) }; return do_long_dual_acc(s, a, fns[a->x]); } + +static bool trans_VPST(DisasContext *s, arg_VPST *a) +{ + TCGv_i32 vpr; + + /* mask == 0 is a "related encoding" */ + if (!dc_isar_feature(aa32_mve, s) || !a->mask) { + return false; + } + if (!mve_eci_check(s) || !vfp_access_check(s)) { + return true; + } + /* + * Set the VPR mask fields. We take advantage of MASK01 and MASK23 + * being adjacent fields in the register. + * + * This insn is not predicated, but it is subject to beat-wise + * execution, and the mask is updated on the odd-numbered beats. + * So if PSR.ECI says we should skip beat 1, we mustn't update the + * 01 mask field. + */ + vpr = load_cpu_field(v7m.vpr); + switch (s->eci) { + case ECI_NONE: + case ECI_A0: + /* Update both 01 and 23 fields */ + tcg_gen_deposit_i32(vpr, vpr, + tcg_constant_i32(a->mask | (a->mask << 4)), + R_V7M_VPR_MASK01_SHIFT, + R_V7M_VPR_MASK01_LENGTH + R_V7M_VPR_MASK23_LENGTH); + break; + case ECI_A0A1: + case ECI_A0A1A2: + case ECI_A0A1A2B0: + /* Update only the 23 mask field */ + tcg_gen_deposit_i32(vpr, vpr, + tcg_constant_i32(a->mask), + R_V7M_VPR_MASK23_SHIFT, R_V7M_VPR_MASK23_LENGTH); + break; + default: + g_assert_not_reached(); + } + store_cpu_field(vpr, v7m.vpr); + mve_update_and_store_eci(s); + return true; +} -- 2.20.1 Implement the MVE VQADD and VQSUB insns, which perform saturating addition of a scalar to each element. Note that individual bytes of each result element are used or discarded according to the predicate mask, but FPSCR.QC is only set if the predicate mask for the lowest byte of the element is set. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-28-peter.maydell@linaro.org --- target/arm/helper-mve.h | 16 ++++++++++ target/arm/mve.decode | 5 +++ target/arm/mve_helper.c | 62 ++++++++++++++++++++++++++++++++++++++ target/arm/translate-mve.c | 4 +++ 4 files changed, 87 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(mve_vhsubu_scalarb, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vhsubu_scalarh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vhsubu_scalarw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vqadds_scalarb, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vqadds_scalarh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vqadds_scalarw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) + +DEF_HELPER_FLAGS_4(mve_vqaddu_scalarb, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vqaddu_scalarh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vqaddu_scalarw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) + +DEF_HELPER_FLAGS_4(mve_vqsubs_scalarb, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vqsubs_scalarh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vqsubs_scalarw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) + +DEF_HELPER_FLAGS_4(mve_vqsubu_scalarb, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vqsubu_scalarh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vqsubu_scalarw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) + DEF_HELPER_FLAGS_4(mve_vbrsrb, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vbrsrh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vbrsrw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ VHADD_S_scalar 1110 1110 0 . .. ... 0 ... 0 1111 . 100 .... @2scalar VHADD_U_scalar 1111 1110 0 . .. ... 0 ... 0 1111 . 100 .... @2scalar VHSUB_S_scalar 1110 1110 0 . .. ... 0 ... 1 1111 . 100 .... @2scalar VHSUB_U_scalar 1111 1110 0 . .. ... 0 ... 1 1111 . 100 .... @2scalar + +VQADD_S_scalar 1110 1110 0 . .. ... 0 ... 0 1111 . 110 .... @2scalar +VQADD_U_scalar 1111 1110 0 . .. ... 0 ... 0 1111 . 110 .... @2scalar +VQSUB_S_scalar 1110 1110 0 . .. ... 0 ... 1 1111 . 110 .... @2scalar +VQSUB_U_scalar 1111 1110 0 . .. ... 0 ... 1 1111 . 110 .... @2scalar VBRSR 1111 1110 0 . .. ... 1 ... 1 1110 . 110 .... @2scalar # Predicate operations diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ DO_2OP_U(vhaddu, do_vhadd_u) DO_2OP_S(vhsubs, do_vhsub_s) DO_2OP_U(vhsubu, do_vhsub_u) +static inline int32_t do_sat_bhw(int64_t val, int64_t min, int64_t max, bool *s) +{ + if (val > max) { + *s = true; + return max; + } else if (val < min) { + *s = true; + return min; + } + return val; +} + +#define DO_SQADD_B(n, m, s) do_sat_bhw((int64_t)n + m, INT8_MIN, INT8_MAX, s) +#define DO_SQADD_H(n, m, s) do_sat_bhw((int64_t)n + m, INT16_MIN, INT16_MAX, s) +#define DO_SQADD_W(n, m, s) do_sat_bhw((int64_t)n + m, INT32_MIN, INT32_MAX, s) + +#define DO_UQADD_B(n, m, s) do_sat_bhw((int64_t)n + m, 0, UINT8_MAX, s) +#define DO_UQADD_H(n, m, s) do_sat_bhw((int64_t)n + m, 0, UINT16_MAX, s) +#define DO_UQADD_W(n, m, s) do_sat_bhw((int64_t)n + m, 0, UINT32_MAX, s) + +#define DO_SQSUB_B(n, m, s) do_sat_bhw((int64_t)n - m, INT8_MIN, INT8_MAX, s) +#define DO_SQSUB_H(n, m, s) do_sat_bhw((int64_t)n - m, INT16_MIN, INT16_MAX, s) +#define DO_SQSUB_W(n, m, s) do_sat_bhw((int64_t)n - m, INT32_MIN, INT32_MAX, s) + +#define DO_UQSUB_B(n, m, s) do_sat_bhw((int64_t)n - m, 0, UINT8_MAX, s) +#define DO_UQSUB_H(n, m, s) do_sat_bhw((int64_t)n - m, 0, UINT16_MAX, s) +#define DO_UQSUB_W(n, m, s) do_sat_bhw((int64_t)n - m, 0, UINT32_MAX, s) #define DO_2OP_SCALAR(OP, ESIZE, TYPE, FN) \ void HELPER(glue(mve_, OP))(CPUARMState *env, void *vd, void *vn, \ @@ -XXX,XX +XXX,XX @@ DO_2OP_U(vhsubu, do_vhsub_u) mve_advance_vpt(env); \ } +#define DO_2OP_SAT_SCALAR(OP, ESIZE, TYPE, FN) \ + void HELPER(glue(mve_, OP))(CPUARMState *env, void *vd, void *vn, \ + uint32_t rm) \ + { \ + TYPE *d = vd, *n = vn; \ + TYPE m = rm; \ + uint16_t mask = mve_element_mask(env); \ + unsigned e; \ + bool qc = false; \ + for (e = 0; e < 16 / ESIZE; e++, mask >>= ESIZE) { \ + bool sat = false; \ + mergemask(&d[H##ESIZE(e)], FN(n[H##ESIZE(e)], m, &sat), \ + mask); \ + qc |= sat & mask & 1; \ + } \ + if (qc) { \ + env->vfp.qc[0] = qc; \ + } \ + mve_advance_vpt(env); \ + } + /* provide unsigned 2-op scalar helpers for all sizes */ #define DO_2OP_SCALAR_U(OP, FN) \ DO_2OP_SCALAR(OP##b, 1, uint8_t, FN) \ @@ -XXX,XX +XXX,XX @@ DO_2OP_SCALAR_U(vhaddu_scalar, do_vhadd_u) DO_2OP_SCALAR_S(vhsubs_scalar, do_vhsub_s) DO_2OP_SCALAR_U(vhsubu_scalar, do_vhsub_u) +DO_2OP_SAT_SCALAR(vqaddu_scalarb, 1, uint8_t, DO_UQADD_B) +DO_2OP_SAT_SCALAR(vqaddu_scalarh, 2, uint16_t, DO_UQADD_H) +DO_2OP_SAT_SCALAR(vqaddu_scalarw, 4, uint32_t, DO_UQADD_W) +DO_2OP_SAT_SCALAR(vqadds_scalarb, 1, int8_t, DO_SQADD_B) +DO_2OP_SAT_SCALAR(vqadds_scalarh, 2, int16_t, DO_SQADD_H) +DO_2OP_SAT_SCALAR(vqadds_scalarw, 4, int32_t, DO_SQADD_W) + +DO_2OP_SAT_SCALAR(vqsubu_scalarb, 1, uint8_t, DO_UQSUB_B) +DO_2OP_SAT_SCALAR(vqsubu_scalarh, 2, uint16_t, DO_UQSUB_H) +DO_2OP_SAT_SCALAR(vqsubu_scalarw, 4, uint32_t, DO_UQSUB_W) +DO_2OP_SAT_SCALAR(vqsubs_scalarb, 1, int8_t, DO_SQSUB_B) +DO_2OP_SAT_SCALAR(vqsubs_scalarh, 2, int16_t, DO_SQSUB_H) +DO_2OP_SAT_SCALAR(vqsubs_scalarw, 4, int32_t, DO_SQSUB_W) + static inline uint32_t do_vbrsrb(uint32_t n, uint32_t m) { m &= 0xff; diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ DO_2OP_SCALAR(VHADD_S_scalar, vhadds_scalar) DO_2OP_SCALAR(VHADD_U_scalar, vhaddu_scalar) DO_2OP_SCALAR(VHSUB_S_scalar, vhsubs_scalar) DO_2OP_SCALAR(VHSUB_U_scalar, vhsubu_scalar) +DO_2OP_SCALAR(VQADD_S_scalar, vqadds_scalar) +DO_2OP_SCALAR(VQADD_U_scalar, vqaddu_scalar) +DO_2OP_SCALAR(VQSUB_S_scalar, vqsubs_scalar) +DO_2OP_SCALAR(VQSUB_U_scalar, vqsubu_scalar) DO_2OP_SCALAR(VBRSR, vbrsr) static bool do_long_dual_acc(DisasContext *s, arg_vmlaldav *a, -- 2.20.1 Implement the MVE VQDMULH and VQRDMULH scalar insns, which multiply elements by the scalar, double, possibly round, take the high half and saturate. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-29-peter.maydell@linaro.org --- target/arm/helper-mve.h | 8 ++++++++ target/arm/mve.decode | 3 +++ target/arm/mve_helper.c | 25 +++++++++++++++++++++++++ target/arm/translate-mve.c | 2 ++ 4 files changed, 38 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(mve_vqsubu_scalarb, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vqsubu_scalarh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vqsubu_scalarw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vqdmulh_scalarb, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vqdmulh_scalarh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vqdmulh_scalarw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) + +DEF_HELPER_FLAGS_4(mve_vqrdmulh_scalarb, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vqrdmulh_scalarh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vqrdmulh_scalarw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) + DEF_HELPER_FLAGS_4(mve_vbrsrb, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vbrsrh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vbrsrw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ VQSUB_S_scalar 1110 1110 0 . .. ... 0 ... 1 1111 . 110 .... @2scalar VQSUB_U_scalar 1111 1110 0 . .. ... 0 ... 1 1111 . 110 .... @2scalar VBRSR 1111 1110 0 . .. ... 1 ... 1 1110 . 110 .... @2scalar +VQDMULH_scalar 1110 1110 0 . .. ... 1 ... 0 1110 . 110 .... @2scalar +VQRDMULH_scalar 1111 1110 0 . .. ... 1 ... 0 1110 . 110 .... @2scalar + # Predicate operations %mask_22_13 22:1 13:3 VPST 1111 1110 0 . 11 000 1 ... 0 1111 0100 1101 mask=%mask_22_13 diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ static inline int32_t do_sat_bhw(int64_t val, int64_t min, int64_t max, bool *s) #define DO_UQSUB_H(n, m, s) do_sat_bhw((int64_t)n - m, 0, UINT16_MAX, s) #define DO_UQSUB_W(n, m, s) do_sat_bhw((int64_t)n - m, 0, UINT32_MAX, s) +/* + * For QDMULH and QRDMULH we simplify "double and shift by esize" into + * "shift by esize-1", adjusting the QRDMULH rounding constant to match. + */ +#define DO_QDMULH_B(n, m, s) do_sat_bhw(((int64_t)n * m) >> 7, \ + INT8_MIN, INT8_MAX, s) +#define DO_QDMULH_H(n, m, s) do_sat_bhw(((int64_t)n * m) >> 15, \ + INT16_MIN, INT16_MAX, s) +#define DO_QDMULH_W(n, m, s) do_sat_bhw(((int64_t)n * m) >> 31, \ + INT32_MIN, INT32_MAX, s) + +#define DO_QRDMULH_B(n, m, s) do_sat_bhw(((int64_t)n * m + (1 << 6)) >> 7, \ + INT8_MIN, INT8_MAX, s) +#define DO_QRDMULH_H(n, m, s) do_sat_bhw(((int64_t)n * m + (1 << 14)) >> 15, \ + INT16_MIN, INT16_MAX, s) +#define DO_QRDMULH_W(n, m, s) do_sat_bhw(((int64_t)n * m + (1 << 30)) >> 31, \ + INT32_MIN, INT32_MAX, s) + #define DO_2OP_SCALAR(OP, ESIZE, TYPE, FN) \ void HELPER(glue(mve_, OP))(CPUARMState *env, void *vd, void *vn, \ uint32_t rm) \ @@ -XXX,XX +XXX,XX @@ DO_2OP_SAT_SCALAR(vqsubs_scalarb, 1, int8_t, DO_SQSUB_B) DO_2OP_SAT_SCALAR(vqsubs_scalarh, 2, int16_t, DO_SQSUB_H) DO_2OP_SAT_SCALAR(vqsubs_scalarw, 4, int32_t, DO_SQSUB_W) +DO_2OP_SAT_SCALAR(vqdmulh_scalarb, 1, int8_t, DO_QDMULH_B) +DO_2OP_SAT_SCALAR(vqdmulh_scalarh, 2, int16_t, DO_QDMULH_H) +DO_2OP_SAT_SCALAR(vqdmulh_scalarw, 4, int32_t, DO_QDMULH_W) +DO_2OP_SAT_SCALAR(vqrdmulh_scalarb, 1, int8_t, DO_QRDMULH_B) +DO_2OP_SAT_SCALAR(vqrdmulh_scalarh, 2, int16_t, DO_QRDMULH_H) +DO_2OP_SAT_SCALAR(vqrdmulh_scalarw, 4, int32_t, DO_QRDMULH_W) + static inline uint32_t do_vbrsrb(uint32_t n, uint32_t m) { m &= 0xff; diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ DO_2OP_SCALAR(VQADD_S_scalar, vqadds_scalar) DO_2OP_SCALAR(VQADD_U_scalar, vqaddu_scalar) DO_2OP_SCALAR(VQSUB_S_scalar, vqsubs_scalar) DO_2OP_SCALAR(VQSUB_U_scalar, vqsubu_scalar) +DO_2OP_SCALAR(VQDMULH_scalar, vqdmulh_scalar) +DO_2OP_SCALAR(VQRDMULH_scalar, vqrdmulh_scalar) DO_2OP_SCALAR(VBRSR, vbrsr) static bool do_long_dual_acc(DisasContext *s, arg_vmlaldav *a, -- 2.20.1 Implement the MVE VQDMULL scalar insn. This multiplies the top or bottom half of each element by the scalar, doubles and saturates to a double-width result. Note that this encoding overlaps with VQADD and VQSUB; it uses what in VQADD and VQSUB would be the 'size=0b11' encoding. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-30-peter.maydell@linaro.org --- target/arm/helper-mve.h | 5 +++ target/arm/mve.decode | 23 +++++++++++--- target/arm/mve_helper.c | 65 ++++++++++++++++++++++++++++++++++++++ target/arm/translate-mve.c | 30 ++++++++++++++++++ 4 files changed, 119 insertions(+), 4 deletions(-) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(mve_vbrsrb, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vbrsrh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vbrsrw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vqdmullb_scalarh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vqdmullb_scalarw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vqdmullt_scalarh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(mve_vqdmullt_scalarw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) + DEF_HELPER_FLAGS_4(mve_vmlaldavsh, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) DEF_HELPER_FLAGS_4(mve_vmlaldavsw, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) DEF_HELPER_FLAGS_4(mve_vmlaldavxsh, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ %qm 5:1 1:3 %qn 7:1 17:3 +# VQDMULL has size in bit 28: 0 for 16 bit, 1 for 32 bit +%size_28 28:1 !function=plus_1 + &vldr_vstr rn qd imm p a w size l u &1op qd qm size &2op qd qm qn size @@ -XXX,XX +XXX,XX @@ @2op_nosz .... .... .... .... .... .... .... .... &2op qd=%qd qm=%qm qn=%qn size=0 @2scalar .... .... .. size:2 .... .... .... .... rm:4 &2scalar qd=%qd qn=%qn +@2scalar_nosz .... .... .... .... .... .... .... rm:4 &2scalar qd=%qd qn=%qn # Vector loads and stores @@ -XXX,XX +XXX,XX @@ VHADD_U_scalar 1111 1110 0 . .. ... 0 ... 0 1111 . 100 .... @2scalar VHSUB_S_scalar 1110 1110 0 . .. ... 0 ... 1 1111 . 100 .... @2scalar VHSUB_U_scalar 1111 1110 0 . .. ... 0 ... 1 1111 . 100 .... @2scalar -VQADD_S_scalar 1110 1110 0 . .. ... 0 ... 0 1111 . 110 .... @2scalar -VQADD_U_scalar 1111 1110 0 . .. ... 0 ... 0 1111 . 110 .... @2scalar -VQSUB_S_scalar 1110 1110 0 . .. ... 0 ... 1 1111 . 110 .... @2scalar -VQSUB_U_scalar 1111 1110 0 . .. ... 0 ... 1 1111 . 110 .... @2scalar +{ + VQADD_S_scalar 1110 1110 0 . .. ... 0 ... 0 1111 . 110 .... @2scalar + VQADD_U_scalar 1111 1110 0 . .. ... 0 ... 0 1111 . 110 .... @2scalar + VQDMULLB_scalar 111 . 1110 0 . 11 ... 0 ... 0 1111 . 110 .... @2scalar_nosz \ + size=%size_28 +} + +{ + VQSUB_S_scalar 1110 1110 0 . .. ... 0 ... 1 1111 . 110 .... @2scalar + VQSUB_U_scalar 1111 1110 0 . .. ... 0 ... 1 1111 . 110 .... @2scalar + VQDMULLT_scalar 111 . 1110 0 . 11 ... 0 ... 1 1111 . 110 .... @2scalar_nosz \ + size=%size_28 +} + VBRSR 1111 1110 0 . .. ... 1 ... 1 1110 . 110 .... @2scalar VQDMULH_scalar 1110 1110 0 . .. ... 1 ... 0 1110 . 110 .... @2scalar VQRDMULH_scalar 1111 1110 0 . .. ... 1 ... 0 1110 . 110 .... @2scalar + # Predicate operations %mask_22_13 22:1 13:3 VPST 1111 1110 0 . 11 000 1 ... 0 1111 0100 1101 mask=%mask_22_13 diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ DO_2OP_SAT_SCALAR(vqrdmulh_scalarb, 1, int8_t, DO_QRDMULH_B) DO_2OP_SAT_SCALAR(vqrdmulh_scalarh, 2, int16_t, DO_QRDMULH_H) DO_2OP_SAT_SCALAR(vqrdmulh_scalarw, 4, int32_t, DO_QRDMULH_W) +/* + * Long saturating scalar ops. As with DO_2OP_L, TYPE and H are for the + * input (smaller) type and LESIZE, LTYPE, LH for the output (long) type. + * SATMASK specifies which bits of the predicate mask matter for determining + * whether to propagate a saturation indication into FPSCR.QC -- for + * the 16x16->32 case we must check only the bit corresponding to the T or B + * half that we used, but for the 32x32->64 case we propagate if the mask + * bit is set for either half. + */ +#define DO_2OP_SAT_SCALAR_L(OP, TOP, ESIZE, TYPE, LESIZE, LTYPE, FN, SATMASK) \ + void HELPER(glue(mve_, OP))(CPUARMState *env, void *vd, void *vn, \ + uint32_t rm) \ + { \ + LTYPE *d = vd; \ + TYPE *n = vn; \ + TYPE m = rm; \ + uint16_t mask = mve_element_mask(env); \ + unsigned le; \ + bool qc = false; \ + for (le = 0; le < 16 / LESIZE; le++, mask >>= LESIZE) { \ + bool sat = false; \ + LTYPE r = FN((LTYPE)n[H##ESIZE(le * 2 + TOP)], m, &sat); \ + mergemask(&d[H##LESIZE(le)], r, mask); \ + qc |= sat && (mask & SATMASK); \ + } \ + if (qc) { \ + env->vfp.qc[0] = qc; \ + } \ + mve_advance_vpt(env); \ + } + +static inline int32_t do_qdmullh(int16_t n, int16_t m, bool *sat) +{ + int64_t r = ((int64_t)n * m) * 2; + return do_sat_bhw(r, INT32_MIN, INT32_MAX, sat); +} + +static inline int64_t do_qdmullw(int32_t n, int32_t m, bool *sat) +{ + /* The multiply can't overflow, but the doubling might */ + int64_t r = (int64_t)n * m; + if (r > INT64_MAX / 2) { + *sat = true; + return INT64_MAX; + } else if (r < INT64_MIN / 2) { + *sat = true; + return INT64_MIN; + } else { + return r * 2; + } +} + +#define SATMASK16B 1 +#define SATMASK16T (1 << 2) +#define SATMASK32 ((1 << 4) | 1) + +DO_2OP_SAT_SCALAR_L(vqdmullb_scalarh, 0, 2, int16_t, 4, int32_t, \ + do_qdmullh, SATMASK16B) +DO_2OP_SAT_SCALAR_L(vqdmullb_scalarw, 0, 4, int32_t, 8, int64_t, \ + do_qdmullw, SATMASK32) +DO_2OP_SAT_SCALAR_L(vqdmullt_scalarh, 1, 2, int16_t, 4, int32_t, \ + do_qdmullh, SATMASK16T) +DO_2OP_SAT_SCALAR_L(vqdmullt_scalarw, 1, 4, int32_t, 8, int64_t, \ + do_qdmullw, SATMASK32) + static inline uint32_t do_vbrsrb(uint32_t n, uint32_t m) { m &= 0xff; diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ DO_2OP_SCALAR(VQDMULH_scalar, vqdmulh_scalar) DO_2OP_SCALAR(VQRDMULH_scalar, vqrdmulh_scalar) DO_2OP_SCALAR(VBRSR, vbrsr) +static bool trans_VQDMULLB_scalar(DisasContext *s, arg_2scalar *a) +{ + static MVEGenTwoOpScalarFn * const fns[] = { + NULL, + gen_helper_mve_vqdmullb_scalarh, + gen_helper_mve_vqdmullb_scalarw, + NULL, + }; + if (a->qd == a->qn && a->size == MO_32) { + /* UNPREDICTABLE; we choose to undef */ + return false; + } + return do_2op_scalar(s, a, fns[a->size]); +} + +static bool trans_VQDMULLT_scalar(DisasContext *s, arg_2scalar *a) +{ + static MVEGenTwoOpScalarFn * const fns[] = { + NULL, + gen_helper_mve_vqdmullt_scalarh, + gen_helper_mve_vqdmullt_scalarw, + NULL, + }; + if (a->qd == a->qn && a->size == MO_32) { + /* UNPREDICTABLE; we choose to undef */ + return false; + } + return do_2op_scalar(s, a, fns[a->size]); +} + static bool do_long_dual_acc(DisasContext *s, arg_vmlaldav *a, MVEGenDualAccOpFn *fn) { -- 2.20.1 Implement the vector forms of the MVE VQDMULH and VQRDMULH insns. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-31-peter.maydell@linaro.org --- target/arm/helper-mve.h | 8 ++++++++ target/arm/mve.decode | 3 +++ target/arm/mve_helper.c | 27 +++++++++++++++++++++++++++ target/arm/translate-mve.c | 2 ++ 4 files changed, 40 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(mve_vmulltub, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vmulltuh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vmulltuw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqdmulhb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqdmulhh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqdmulhw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + +DEF_HELPER_FLAGS_4(mve_vqrdmulhb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqrdmulhh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqrdmulhw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + DEF_HELPER_FLAGS_4(mve_vadd_scalarb, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vadd_scalarh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vadd_scalarw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ VMULL_BU 111 1 1110 0 . .. ... 1 ... 0 1110 . 0 . 0 ... 0 @2op VMULL_TS 111 0 1110 0 . .. ... 1 ... 1 1110 . 0 . 0 ... 0 @2op VMULL_TU 111 1 1110 0 . .. ... 1 ... 1 1110 . 0 . 0 ... 0 @2op +VQDMULH 1110 1111 0 . .. ... 0 ... 0 1011 . 1 . 0 ... 0 @2op +VQRDMULH 1111 1111 0 . .. ... 0 ... 0 1011 . 1 . 0 ... 0 @2op + # Vector miscellaneous VCLS 1111 1111 1 . 11 .. 00 ... 0 0100 01 . 0 ... 0 @1op diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ DO_1OP(vfnegs, 8, uint64_t, DO_FNEGS) mve_advance_vpt(env); \ } +#define DO_2OP_SAT(OP, ESIZE, TYPE, FN) \ + void HELPER(glue(mve_, OP))(CPUARMState *env, void *vd, void *vn, void *vm) \ + { \ + TYPE *d = vd, *n = vn, *m = vm; \ + uint16_t mask = mve_element_mask(env); \ + unsigned e; \ + bool qc = false; \ + for (e = 0; e < 16 / ESIZE; e++, mask >>= ESIZE) { \ + bool sat = false; \ + TYPE r = FN(n[H##ESIZE(e)], m[H##ESIZE(e)], &sat); \ + mergemask(&d[H##ESIZE(e)], r, mask); \ + qc |= sat & mask & 1; \ + } \ + if (qc) { \ + env->vfp.qc[0] = qc; \ + } \ + mve_advance_vpt(env); \ + } + #define DO_AND(N, M) ((N) & (M)) #define DO_BIC(N, M) ((N) & ~(M)) #define DO_ORR(N, M) ((N) | (M)) @@ -XXX,XX +XXX,XX @@ static inline int32_t do_sat_bhw(int64_t val, int64_t min, int64_t max, bool *s) #define DO_QRDMULH_W(n, m, s) do_sat_bhw(((int64_t)n * m + (1 << 30)) >> 31, \ INT32_MIN, INT32_MAX, s) +DO_2OP_SAT(vqdmulhb, 1, int8_t, DO_QDMULH_B) +DO_2OP_SAT(vqdmulhh, 2, int16_t, DO_QDMULH_H) +DO_2OP_SAT(vqdmulhw, 4, int32_t, DO_QDMULH_W) + +DO_2OP_SAT(vqrdmulhb, 1, int8_t, DO_QRDMULH_B) +DO_2OP_SAT(vqrdmulhh, 2, int16_t, DO_QRDMULH_H) +DO_2OP_SAT(vqrdmulhw, 4, int32_t, DO_QRDMULH_W) + #define DO_2OP_SCALAR(OP, ESIZE, TYPE, FN) \ void HELPER(glue(mve_, OP))(CPUARMState *env, void *vd, void *vn, \ uint32_t rm) \ diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ DO_2OP(VMULL_BS, vmullbs) DO_2OP(VMULL_BU, vmullbu) DO_2OP(VMULL_TS, vmullts) DO_2OP(VMULL_TU, vmulltu) +DO_2OP(VQDMULH, vqdmulh) +DO_2OP(VQRDMULH, vqrdmulh) static bool do_2op_scalar(DisasContext *s, arg_2scalar *a, MVEGenTwoOpScalarFn fn) -- 2.20.1 Implement the vector forms of the MVE VQADD and VQSUB insns. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-32-peter.maydell@linaro.org --- target/arm/helper-mve.h | 16 ++++++++++++++++ target/arm/mve.decode | 5 +++++ target/arm/mve_helper.c | 14 ++++++++++++++ target/arm/translate-mve.c | 4 ++++ 4 files changed, 39 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(mve_vqrdmulhb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vqrdmulhh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vqrdmulhw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqaddsb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqaddsh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqaddsw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + +DEF_HELPER_FLAGS_4(mve_vqaddub, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqadduh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqadduw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + +DEF_HELPER_FLAGS_4(mve_vqsubsb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqsubsh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqsubsw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + +DEF_HELPER_FLAGS_4(mve_vqsubub, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqsubuh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqsubuw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + DEF_HELPER_FLAGS_4(mve_vadd_scalarb, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vadd_scalarh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vadd_scalarw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ VMULL_TU 111 1 1110 0 . .. ... 1 ... 1 1110 . 0 . 0 ... 0 @2op VQDMULH 1110 1111 0 . .. ... 0 ... 0 1011 . 1 . 0 ... 0 @2op VQRDMULH 1111 1111 0 . .. ... 0 ... 0 1011 . 1 . 0 ... 0 @2op +VQADD_S 111 0 1111 0 . .. ... 0 ... 0 0000 . 1 . 1 ... 0 @2op +VQADD_U 111 1 1111 0 . .. ... 0 ... 0 0000 . 1 . 1 ... 0 @2op +VQSUB_S 111 0 1111 0 . .. ... 0 ... 0 0010 . 1 . 1 ... 0 @2op +VQSUB_U 111 1 1111 0 . .. ... 0 ... 0 0010 . 1 . 1 ... 0 @2op + # Vector miscellaneous VCLS 1111 1111 1 . 11 .. 00 ... 0 0100 01 . 0 ... 0 @1op diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ DO_2OP_SAT(vqrdmulhb, 1, int8_t, DO_QRDMULH_B) DO_2OP_SAT(vqrdmulhh, 2, int16_t, DO_QRDMULH_H) DO_2OP_SAT(vqrdmulhw, 4, int32_t, DO_QRDMULH_W) +DO_2OP_SAT(vqaddub, 1, uint8_t, DO_UQADD_B) +DO_2OP_SAT(vqadduh, 2, uint16_t, DO_UQADD_H) +DO_2OP_SAT(vqadduw, 4, uint32_t, DO_UQADD_W) +DO_2OP_SAT(vqaddsb, 1, int8_t, DO_SQADD_B) +DO_2OP_SAT(vqaddsh, 2, int16_t, DO_SQADD_H) +DO_2OP_SAT(vqaddsw, 4, int32_t, DO_SQADD_W) + +DO_2OP_SAT(vqsubub, 1, uint8_t, DO_UQSUB_B) +DO_2OP_SAT(vqsubuh, 2, uint16_t, DO_UQSUB_H) +DO_2OP_SAT(vqsubuw, 4, uint32_t, DO_UQSUB_W) +DO_2OP_SAT(vqsubsb, 1, int8_t, DO_SQSUB_B) +DO_2OP_SAT(vqsubsh, 2, int16_t, DO_SQSUB_H) +DO_2OP_SAT(vqsubsw, 4, int32_t, DO_SQSUB_W) + #define DO_2OP_SCALAR(OP, ESIZE, TYPE, FN) \ void HELPER(glue(mve_, OP))(CPUARMState *env, void *vd, void *vn, \ uint32_t rm) \ diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ DO_2OP(VMULL_TS, vmullts) DO_2OP(VMULL_TU, vmulltu) DO_2OP(VQDMULH, vqdmulh) DO_2OP(VQRDMULH, vqrdmulh) +DO_2OP(VQADD_S, vqadds) +DO_2OP(VQADD_U, vqaddu) +DO_2OP(VQSUB_S, vqsubs) +DO_2OP(VQSUB_U, vqsubu) static bool do_2op_scalar(DisasContext *s, arg_2scalar *a, MVEGenTwoOpScalarFn fn) -- 2.20.1 Implement the MVE VQSHL insn (encoding T4, which is the vector-shift-by-vector version). The DO_SQSHL_OP and DO_UQSHL_OP macros here are derived from the neon_helper.c code for qshl_u{8,16,32} and qshl_s{8,16,32}. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-33-peter.maydell@linaro.org --- target/arm/helper-mve.h | 8 ++++++++ target/arm/mve.decode | 12 ++++++++++++ target/arm/mve_helper.c | 34 ++++++++++++++++++++++++++++++++++ target/arm/translate-mve.c | 2 ++ 4 files changed, 56 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(mve_vqsubub, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vqsubuh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vqsubuw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqshlsb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqshlsh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqshlsw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + +DEF_HELPER_FLAGS_4(mve_vqshlub, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqshluh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqshluw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + DEF_HELPER_FLAGS_4(mve_vadd_scalarb, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vadd_scalarh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vadd_scalarw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ @2op .... .... .. size:2 .... .... .... .... .... &2op qd=%qd qm=%qm qn=%qn @2op_nosz .... .... .... .... .... .... .... .... &2op qd=%qd qm=%qm qn=%qn size=0 +# The _rev suffix indicates that Vn and Vm are reversed. This is +# the case for shifts. In the Arm ARM these insns are documented +# with the Vm and Vn fields in their usual places, but in the +# assembly the operands are listed "backwards", ie in the order +# Qd, Qm, Qn where other insns use Qd, Qn, Qm. For QEMU we choose +# to consider Vm and Vn as being in different fields in the insn. +# This gives us consistency with A64 and Neon. +@2op_rev .... .... .. size:2 .... .... .... .... .... &2op qd=%qd qm=%qn qn=%qm + @2scalar .... .... .. size:2 .... .... .... .... rm:4 &2scalar qd=%qd qn=%qn @2scalar_nosz .... .... .... .... .... .... .... rm:4 &2scalar qd=%qd qn=%qn @@ -XXX,XX +XXX,XX @@ VQADD_U 111 1 1111 0 . .. ... 0 ... 0 0000 . 1 . 1 ... 0 @2op VQSUB_S 111 0 1111 0 . .. ... 0 ... 0 0010 . 1 . 1 ... 0 @2op VQSUB_U 111 1 1111 0 . .. ... 0 ... 0 0010 . 1 . 1 ... 0 @2op +VQSHL_S 111 0 1111 0 . .. ... 0 ... 0 0100 . 1 . 1 ... 0 @2op_rev +VQSHL_U 111 1 1111 0 . .. ... 0 ... 0 0100 . 1 . 1 ... 0 @2op_rev + # Vector miscellaneous VCLS 1111 1111 1 . 11 .. 00 ... 0 0100 01 . 0 ... 0 @1op diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ DO_1OP(vfnegs, 8, uint64_t, DO_FNEGS) mve_advance_vpt(env); \ } +/* provide unsigned 2-op helpers for all sizes */ +#define DO_2OP_SAT_U(OP, FN) \ + DO_2OP_SAT(OP##b, 1, uint8_t, FN) \ + DO_2OP_SAT(OP##h, 2, uint16_t, FN) \ + DO_2OP_SAT(OP##w, 4, uint32_t, FN) + +/* provide signed 2-op helpers for all sizes */ +#define DO_2OP_SAT_S(OP, FN) \ + DO_2OP_SAT(OP##b, 1, int8_t, FN) \ + DO_2OP_SAT(OP##h, 2, int16_t, FN) \ + DO_2OP_SAT(OP##w, 4, int32_t, FN) + #define DO_AND(N, M) ((N) & (M)) #define DO_BIC(N, M) ((N) & ~(M)) #define DO_ORR(N, M) ((N) | (M)) @@ -XXX,XX +XXX,XX @@ DO_2OP_SAT(vqsubsb, 1, int8_t, DO_SQSUB_B) DO_2OP_SAT(vqsubsh, 2, int16_t, DO_SQSUB_H) DO_2OP_SAT(vqsubsw, 4, int32_t, DO_SQSUB_W) +/* + * This wrapper fixes up the impedance mismatch between do_sqrshl_bhs() + * and friends wanting a uint32_t* sat and our needing a bool*. + */ +#define WRAP_QRSHL_HELPER(FN, N, M, ROUND, satp) \ + ({ \ + uint32_t su32 = 0; \ + typeof(N) r = FN(N, (int8_t)(M), sizeof(N) * 8, ROUND, &su32); \ + if (su32) { \ + *satp = true; \ + } \ + r; \ + }) + +#define DO_SQSHL_OP(N, M, satp) \ + WRAP_QRSHL_HELPER(do_sqrshl_bhs, N, M, false, satp) +#define DO_UQSHL_OP(N, M, satp) \ + WRAP_QRSHL_HELPER(do_uqrshl_bhs, N, M, false, satp) + +DO_2OP_SAT_S(vqshls, DO_SQSHL_OP) +DO_2OP_SAT_U(vqshlu, DO_UQSHL_OP) + #define DO_2OP_SCALAR(OP, ESIZE, TYPE, FN) \ void HELPER(glue(mve_, OP))(CPUARMState *env, void *vd, void *vn, \ uint32_t rm) \ diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ DO_2OP(VQADD_S, vqadds) DO_2OP(VQADD_U, vqaddu) DO_2OP(VQSUB_S, vqsubs) DO_2OP(VQSUB_U, vqsubu) +DO_2OP(VQSHL_S, vqshls) +DO_2OP(VQSHL_U, vqshlu) static bool do_2op_scalar(DisasContext *s, arg_2scalar *a, MVEGenTwoOpScalarFn fn) -- 2.20.1 Implement the MV VQRSHL (vector) insn. Again, the code to perform the actual shifts is borrowed from neon_helper.c. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-34-peter.maydell@linaro.org --- target/arm/helper-mve.h | 8 ++++++++ target/arm/mve.decode | 3 +++ target/arm/mve_helper.c | 6 ++++++ target/arm/translate-mve.c | 2 ++ 4 files changed, 19 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(mve_vqshlub, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vqshluh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vqshluw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqrshlsb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqrshlsh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqrshlsw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + +DEF_HELPER_FLAGS_4(mve_vqrshlub, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqrshluh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqrshluw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + DEF_HELPER_FLAGS_4(mve_vadd_scalarb, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vadd_scalarh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vadd_scalarw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ VQSUB_U 111 1 1111 0 . .. ... 0 ... 0 0010 . 1 . 1 ... 0 @2op VQSHL_S 111 0 1111 0 . .. ... 0 ... 0 0100 . 1 . 1 ... 0 @2op_rev VQSHL_U 111 1 1111 0 . .. ... 0 ... 0 0100 . 1 . 1 ... 0 @2op_rev +VQRSHL_S 111 0 1111 0 . .. ... 0 ... 0 0101 . 1 . 1 ... 0 @2op_rev +VQRSHL_U 111 1 1111 0 . .. ... 0 ... 0 0101 . 1 . 1 ... 0 @2op_rev + # Vector miscellaneous VCLS 1111 1111 1 . 11 .. 00 ... 0 0100 01 . 0 ... 0 @1op diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ DO_2OP_SAT(vqsubsw, 4, int32_t, DO_SQSUB_W) WRAP_QRSHL_HELPER(do_sqrshl_bhs, N, M, false, satp) #define DO_UQSHL_OP(N, M, satp) \ WRAP_QRSHL_HELPER(do_uqrshl_bhs, N, M, false, satp) +#define DO_SQRSHL_OP(N, M, satp) \ + WRAP_QRSHL_HELPER(do_sqrshl_bhs, N, M, true, satp) +#define DO_UQRSHL_OP(N, M, satp) \ + WRAP_QRSHL_HELPER(do_uqrshl_bhs, N, M, true, satp) DO_2OP_SAT_S(vqshls, DO_SQSHL_OP) DO_2OP_SAT_U(vqshlu, DO_UQSHL_OP) +DO_2OP_SAT_S(vqrshls, DO_SQRSHL_OP) +DO_2OP_SAT_U(vqrshlu, DO_UQRSHL_OP) #define DO_2OP_SCALAR(OP, ESIZE, TYPE, FN) \ void HELPER(glue(mve_, OP))(CPUARMState *env, void *vd, void *vn, \ diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ DO_2OP(VQSUB_S, vqsubs) DO_2OP(VQSUB_U, vqsubu) DO_2OP(VQSHL_S, vqshls) DO_2OP(VQSHL_U, vqshlu) +DO_2OP(VQRSHL_S, vqrshls) +DO_2OP(VQRSHL_U, vqrshlu) static bool do_2op_scalar(DisasContext *s, arg_2scalar *a, MVEGenTwoOpScalarFn fn) -- 2.20.1 Implement the MVE VSHL insn (vector form). Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-35-peter.maydell@linaro.org --- target/arm/helper-mve.h | 8 ++++++++ target/arm/mve.decode | 3 +++ target/arm/mve_helper.c | 6 ++++++ target/arm/translate-mve.c | 2 ++ 4 files changed, 19 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(mve_vqsubub, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vqsubuh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vqsubuw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vshlsb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vshlsh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vshlsw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + +DEF_HELPER_FLAGS_4(mve_vshlub, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vshluh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vshluw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + DEF_HELPER_FLAGS_4(mve_vqshlsb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vqshlsh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vqshlsw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ VQADD_U 111 1 1111 0 . .. ... 0 ... 0 0000 . 1 . 1 ... 0 @2op VQSUB_S 111 0 1111 0 . .. ... 0 ... 0 0010 . 1 . 1 ... 0 @2op VQSUB_U 111 1 1111 0 . .. ... 0 ... 0 0010 . 1 . 1 ... 0 @2op +VSHL_S 111 0 1111 0 . .. ... 0 ... 0 0100 . 1 . 0 ... 0 @2op_rev +VSHL_U 111 1 1111 0 . .. ... 0 ... 0 0100 . 1 . 0 ... 0 @2op_rev + VQSHL_S 111 0 1111 0 . .. ... 0 ... 0 0100 . 1 . 1 ... 0 @2op_rev VQSHL_U 111 1 1111 0 . .. ... 0 ... 0 0100 . 1 . 1 ... 0 @2op_rev diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ DO_2OP_U(vhaddu, do_vhadd_u) DO_2OP_S(vhsubs, do_vhsub_s) DO_2OP_U(vhsubu, do_vhsub_u) +#define DO_VSHLS(N, M) do_sqrshl_bhs(N, (int8_t)(M), sizeof(N) * 8, false, NULL) +#define DO_VSHLU(N, M) do_uqrshl_bhs(N, (int8_t)(M), sizeof(N) * 8, false, NULL) + +DO_2OP_S(vshls, DO_VSHLS) +DO_2OP_U(vshlu, DO_VSHLU) + static inline int32_t do_sat_bhw(int64_t val, int64_t min, int64_t max, bool *s) { if (val > max) { diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ DO_2OP(VQADD_S, vqadds) DO_2OP(VQADD_U, vqaddu) DO_2OP(VQSUB_S, vqsubs) DO_2OP(VQSUB_U, vqsubu) +DO_2OP(VSHL_S, vshls) +DO_2OP(VSHL_U, vshlu) DO_2OP(VQSHL_S, vqshls) DO_2OP(VQSHL_U, vqshlu) DO_2OP(VQRSHL_S, vqrshls) -- 2.20.1 Implement the MVE VRSHL insn (vector form). Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-36-peter.maydell@linaro.org --- target/arm/helper-mve.h | 8 ++++++++ target/arm/mve.decode | 3 +++ target/arm/mve_helper.c | 4 ++++ target/arm/translate-mve.c | 2 ++ 4 files changed, 17 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(mve_vshlub, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vshluh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vshluw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vrshlsb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vrshlsh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vrshlsw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + +DEF_HELPER_FLAGS_4(mve_vrshlub, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vrshluh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vrshluw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + DEF_HELPER_FLAGS_4(mve_vqshlsb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vqshlsh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vqshlsw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ VQSUB_U 111 1 1111 0 . .. ... 0 ... 0 0010 . 1 . 1 ... 0 @2op VSHL_S 111 0 1111 0 . .. ... 0 ... 0 0100 . 1 . 0 ... 0 @2op_rev VSHL_U 111 1 1111 0 . .. ... 0 ... 0 0100 . 1 . 0 ... 0 @2op_rev +VRSHL_S 111 0 1111 0 . .. ... 0 ... 0 0101 . 1 . 0 ... 0 @2op_rev +VRSHL_U 111 1 1111 0 . .. ... 0 ... 0 0101 . 1 . 0 ... 0 @2op_rev + VQSHL_S 111 0 1111 0 . .. ... 0 ... 0 0100 . 1 . 1 ... 0 @2op_rev VQSHL_U 111 1 1111 0 . .. ... 0 ... 0 0100 . 1 . 1 ... 0 @2op_rev diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ DO_2OP_U(vhsubu, do_vhsub_u) #define DO_VSHLS(N, M) do_sqrshl_bhs(N, (int8_t)(M), sizeof(N) * 8, false, NULL) #define DO_VSHLU(N, M) do_uqrshl_bhs(N, (int8_t)(M), sizeof(N) * 8, false, NULL) +#define DO_VRSHLS(N, M) do_sqrshl_bhs(N, (int8_t)(M), sizeof(N) * 8, true, NULL) +#define DO_VRSHLU(N, M) do_uqrshl_bhs(N, (int8_t)(M), sizeof(N) * 8, true, NULL) DO_2OP_S(vshls, DO_VSHLS) DO_2OP_U(vshlu, DO_VSHLU) +DO_2OP_S(vrshls, DO_VRSHLS) +DO_2OP_U(vrshlu, DO_VRSHLU) static inline int32_t do_sat_bhw(int64_t val, int64_t min, int64_t max, bool *s) { diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ DO_2OP(VQSUB_S, vqsubs) DO_2OP(VQSUB_U, vqsubu) DO_2OP(VSHL_S, vshls) DO_2OP(VSHL_U, vshlu) +DO_2OP(VRSHL_S, vrshls) +DO_2OP(VRSHL_U, vrshlu) DO_2OP(VQSHL_S, vqshls) DO_2OP(VQSHL_U, vqshlu) DO_2OP(VQRSHL_S, vqrshls) -- 2.20.1 Implement the MVE VQDMLADH and VQRDMLADH insns. These multiply elements, and then add pairs of products, double, possibly round, saturate and return the high half of the result. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-37-peter.maydell@linaro.org --- target/arm/helper-mve.h | 16 +++++++ target/arm/mve.decode | 5 +++ target/arm/mve_helper.c | 89 ++++++++++++++++++++++++++++++++++++++ target/arm/translate-mve.c | 4 ++ 4 files changed, 114 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(mve_vqrshlub, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vqrshluh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vqrshluw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqdmladhb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqdmladhh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqdmladhw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + +DEF_HELPER_FLAGS_4(mve_vqdmladhxb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqdmladhxh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqdmladhxw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + +DEF_HELPER_FLAGS_4(mve_vqrdmladhb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqrdmladhh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqrdmladhw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + +DEF_HELPER_FLAGS_4(mve_vqrdmladhxb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqrdmladhxh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqrdmladhxw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + DEF_HELPER_FLAGS_4(mve_vadd_scalarb, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vadd_scalarh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vadd_scalarw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ VQSHL_U 111 1 1111 0 . .. ... 0 ... 0 0100 . 1 . 1 ... 0 @2op_rev VQRSHL_S 111 0 1111 0 . .. ... 0 ... 0 0101 . 1 . 1 ... 0 @2op_rev VQRSHL_U 111 1 1111 0 . .. ... 0 ... 0 0101 . 1 . 1 ... 0 @2op_rev +VQDMLADH 1110 1110 0 . .. ... 0 ... 0 1110 . 0 . 0 ... 0 @2op +VQDMLADHX 1110 1110 0 . .. ... 0 ... 1 1110 . 0 . 0 ... 0 @2op +VQRDMLADH 1110 1110 0 . .. ... 0 ... 0 1110 . 0 . 0 ... 1 @2op +VQRDMLADHX 1110 1110 0 . .. ... 0 ... 1 1110 . 0 . 0 ... 1 @2op + # Vector miscellaneous VCLS 1111 1111 1 . 11 .. 00 ... 0 0100 01 . 0 ... 0 @1op diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ DO_2OP_SAT_U(vqshlu, DO_UQSHL_OP) DO_2OP_SAT_S(vqrshls, DO_SQRSHL_OP) DO_2OP_SAT_U(vqrshlu, DO_UQRSHL_OP) +/* + * Multiply add dual returning high half + * The 'FN' here takes four inputs A, B, C, D, a 0/1 indicator of + * whether to add the rounding constant, and the pointer to the + * saturation flag, and should do "(A * B + C * D) * 2 + rounding constant", + * saturate to twice the input size and return the high half; or + * (A * B - C * D) etc for VQDMLSDH. + */ +#define DO_VQDMLADH_OP(OP, ESIZE, TYPE, XCHG, ROUND, FN) \ + void HELPER(glue(mve_, OP))(CPUARMState *env, void *vd, void *vn, \ + void *vm) \ + { \ + TYPE *d = vd, *n = vn, *m = vm; \ + uint16_t mask = mve_element_mask(env); \ + unsigned e; \ + bool qc = false; \ + for (e = 0; e < 16 / ESIZE; e++, mask >>= ESIZE) { \ + bool sat = false; \ + if ((e & 1) == XCHG) { \ + TYPE r = FN(n[H##ESIZE(e)], \ + m[H##ESIZE(e - XCHG)], \ + n[H##ESIZE(e + (1 - 2 * XCHG))], \ + m[H##ESIZE(e + (1 - XCHG))], \ + ROUND, &sat); \ + mergemask(&d[H##ESIZE(e)], r, mask); \ + qc |= sat & mask & 1; \ + } \ + } \ + if (qc) { \ + env->vfp.qc[0] = qc; \ + } \ + mve_advance_vpt(env); \ + } + +static int8_t do_vqdmladh_b(int8_t a, int8_t b, int8_t c, int8_t d, + int round, bool *sat) +{ + int64_t r = ((int64_t)a * b + (int64_t)c * d) * 2 + (round << 7); + return do_sat_bhw(r, INT16_MIN, INT16_MAX, sat) >> 8; +} + +static int16_t do_vqdmladh_h(int16_t a, int16_t b, int16_t c, int16_t d, + int round, bool *sat) +{ + int64_t r = ((int64_t)a * b + (int64_t)c * d) * 2 + (round << 15); + return do_sat_bhw(r, INT32_MIN, INT32_MAX, sat) >> 16; +} + +static int32_t do_vqdmladh_w(int32_t a, int32_t b, int32_t c, int32_t d, + int round, bool *sat) +{ + int64_t m1 = (int64_t)a * b; + int64_t m2 = (int64_t)c * d; + int64_t r; + /* + * Architecturally we should do the entire add, double, round + * and then check for saturation. We do three saturating adds, + * but we need to be careful about the order. If the first + * m1 + m2 saturates then it's impossible for the *2+rc to + * bring it back into the non-saturated range. However, if + * m1 + m2 is negative then it's possible that doing the doubling + * would take the intermediate result below INT64_MAX and the + * addition of the rounding constant then brings it back in range. + * So we add half the rounding constant before doubling rather + * than adding the rounding constant after the doubling. + */ + if (sadd64_overflow(m1, m2, &r) || + sadd64_overflow(r, (round << 30), &r) || + sadd64_overflow(r, r, &r)) { + *sat = true; + return r < 0 ? INT32_MAX : INT32_MIN; + } + return r >> 32; +} + +DO_VQDMLADH_OP(vqdmladhb, 1, int8_t, 0, 0, do_vqdmladh_b) +DO_VQDMLADH_OP(vqdmladhh, 2, int16_t, 0, 0, do_vqdmladh_h) +DO_VQDMLADH_OP(vqdmladhw, 4, int32_t, 0, 0, do_vqdmladh_w) +DO_VQDMLADH_OP(vqdmladhxb, 1, int8_t, 1, 0, do_vqdmladh_b) +DO_VQDMLADH_OP(vqdmladhxh, 2, int16_t, 1, 0, do_vqdmladh_h) +DO_VQDMLADH_OP(vqdmladhxw, 4, int32_t, 1, 0, do_vqdmladh_w) + +DO_VQDMLADH_OP(vqrdmladhb, 1, int8_t, 0, 1, do_vqdmladh_b) +DO_VQDMLADH_OP(vqrdmladhh, 2, int16_t, 0, 1, do_vqdmladh_h) +DO_VQDMLADH_OP(vqrdmladhw, 4, int32_t, 0, 1, do_vqdmladh_w) +DO_VQDMLADH_OP(vqrdmladhxb, 1, int8_t, 1, 1, do_vqdmladh_b) +DO_VQDMLADH_OP(vqrdmladhxh, 2, int16_t, 1, 1, do_vqdmladh_h) +DO_VQDMLADH_OP(vqrdmladhxw, 4, int32_t, 1, 1, do_vqdmladh_w) + #define DO_2OP_SCALAR(OP, ESIZE, TYPE, FN) \ void HELPER(glue(mve_, OP))(CPUARMState *env, void *vd, void *vn, \ uint32_t rm) \ diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ DO_2OP(VQSHL_S, vqshls) DO_2OP(VQSHL_U, vqshlu) DO_2OP(VQRSHL_S, vqrshls) DO_2OP(VQRSHL_U, vqrshlu) +DO_2OP(VQDMLADH, vqdmladh) +DO_2OP(VQDMLADHX, vqdmladhx) +DO_2OP(VQRDMLADH, vqrdmladh) +DO_2OP(VQRDMLADHX, vqrdmladhx) static bool do_2op_scalar(DisasContext *s, arg_2scalar *a, MVEGenTwoOpScalarFn fn) -- 2.20.1 Implement the MVE VQDMLSDH and VQRDMLSDH insns, which are like VQDMLADH and VQRDMLADH except that products are subtracted rather than added. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-38-peter.maydell@linaro.org --- target/arm/helper-mve.h | 16 ++++++++++++++ target/arm/mve.decode | 5 +++++ target/arm/mve_helper.c | 44 ++++++++++++++++++++++++++++++++++++++ target/arm/translate-mve.c | 4 ++++ 4 files changed, 69 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(mve_vqrdmladhxb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vqrdmladhxh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vqrdmladhxw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqdmlsdhb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqdmlsdhh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqdmlsdhw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + +DEF_HELPER_FLAGS_4(mve_vqdmlsdhxb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqdmlsdhxh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqdmlsdhxw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + +DEF_HELPER_FLAGS_4(mve_vqrdmlsdhb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqrdmlsdhh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqrdmlsdhw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + +DEF_HELPER_FLAGS_4(mve_vqrdmlsdhxb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqrdmlsdhxh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqrdmlsdhxw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + DEF_HELPER_FLAGS_4(mve_vadd_scalarb, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vadd_scalarh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vadd_scalarw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ VQDMLADHX 1110 1110 0 . .. ... 0 ... 1 1110 . 0 . 0 ... 0 @2op VQRDMLADH 1110 1110 0 . .. ... 0 ... 0 1110 . 0 . 0 ... 1 @2op VQRDMLADHX 1110 1110 0 . .. ... 0 ... 1 1110 . 0 . 0 ... 1 @2op +VQDMLSDH 1111 1110 0 . .. ... 0 ... 0 1110 . 0 . 0 ... 0 @2op +VQDMLSDHX 1111 1110 0 . .. ... 0 ... 1 1110 . 0 . 0 ... 0 @2op +VQRDMLSDH 1111 1110 0 . .. ... 0 ... 0 1110 . 0 . 0 ... 1 @2op +VQRDMLSDHX 1111 1110 0 . .. ... 0 ... 1 1110 . 0 . 0 ... 1 @2op + # Vector miscellaneous VCLS 1111 1111 1 . 11 .. 00 ... 0 0100 01 . 0 ... 0 @1op diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ static int32_t do_vqdmladh_w(int32_t a, int32_t b, int32_t c, int32_t d, return r >> 32; } +static int8_t do_vqdmlsdh_b(int8_t a, int8_t b, int8_t c, int8_t d, + int round, bool *sat) +{ + int64_t r = ((int64_t)a * b - (int64_t)c * d) * 2 + (round << 7); + return do_sat_bhw(r, INT16_MIN, INT16_MAX, sat) >> 8; +} + +static int16_t do_vqdmlsdh_h(int16_t a, int16_t b, int16_t c, int16_t d, + int round, bool *sat) +{ + int64_t r = ((int64_t)a * b - (int64_t)c * d) * 2 + (round << 15); + return do_sat_bhw(r, INT32_MIN, INT32_MAX, sat) >> 16; +} + +static int32_t do_vqdmlsdh_w(int32_t a, int32_t b, int32_t c, int32_t d, + int round, bool *sat) +{ + int64_t m1 = (int64_t)a * b; + int64_t m2 = (int64_t)c * d; + int64_t r; + /* The same ordering issue as in do_vqdmladh_w applies here too */ + if (ssub64_overflow(m1, m2, &r) || + sadd64_overflow(r, (round << 30), &r) || + sadd64_overflow(r, r, &r)) { + *sat = true; + return r < 0 ? INT32_MAX : INT32_MIN; + } + return r >> 32; +} + DO_VQDMLADH_OP(vqdmladhb, 1, int8_t, 0, 0, do_vqdmladh_b) DO_VQDMLADH_OP(vqdmladhh, 2, int16_t, 0, 0, do_vqdmladh_h) DO_VQDMLADH_OP(vqdmladhw, 4, int32_t, 0, 0, do_vqdmladh_w) @@ -XXX,XX +XXX,XX @@ DO_VQDMLADH_OP(vqrdmladhxb, 1, int8_t, 1, 1, do_vqdmladh_b) DO_VQDMLADH_OP(vqrdmladhxh, 2, int16_t, 1, 1, do_vqdmladh_h) DO_VQDMLADH_OP(vqrdmladhxw, 4, int32_t, 1, 1, do_vqdmladh_w) +DO_VQDMLADH_OP(vqdmlsdhb, 1, int8_t, 0, 0, do_vqdmlsdh_b) +DO_VQDMLADH_OP(vqdmlsdhh, 2, int16_t, 0, 0, do_vqdmlsdh_h) +DO_VQDMLADH_OP(vqdmlsdhw, 4, int32_t, 0, 0, do_vqdmlsdh_w) +DO_VQDMLADH_OP(vqdmlsdhxb, 1, int8_t, 1, 0, do_vqdmlsdh_b) +DO_VQDMLADH_OP(vqdmlsdhxh, 2, int16_t, 1, 0, do_vqdmlsdh_h) +DO_VQDMLADH_OP(vqdmlsdhxw, 4, int32_t, 1, 0, do_vqdmlsdh_w) + +DO_VQDMLADH_OP(vqrdmlsdhb, 1, int8_t, 0, 1, do_vqdmlsdh_b) +DO_VQDMLADH_OP(vqrdmlsdhh, 2, int16_t, 0, 1, do_vqdmlsdh_h) +DO_VQDMLADH_OP(vqrdmlsdhw, 4, int32_t, 0, 1, do_vqdmlsdh_w) +DO_VQDMLADH_OP(vqrdmlsdhxb, 1, int8_t, 1, 1, do_vqdmlsdh_b) +DO_VQDMLADH_OP(vqrdmlsdhxh, 2, int16_t, 1, 1, do_vqdmlsdh_h) +DO_VQDMLADH_OP(vqrdmlsdhxw, 4, int32_t, 1, 1, do_vqdmlsdh_w) + #define DO_2OP_SCALAR(OP, ESIZE, TYPE, FN) \ void HELPER(glue(mve_, OP))(CPUARMState *env, void *vd, void *vn, \ uint32_t rm) \ diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ DO_2OP(VQDMLADH, vqdmladh) DO_2OP(VQDMLADHX, vqdmladhx) DO_2OP(VQRDMLADH, vqrdmladh) DO_2OP(VQRDMLADHX, vqrdmladhx) +DO_2OP(VQDMLSDH, vqdmlsdh) +DO_2OP(VQDMLSDHX, vqdmlsdhx) +DO_2OP(VQRDMLSDH, vqrdmlsdh) +DO_2OP(VQRDMLSDHX, vqrdmlsdhx) static bool do_2op_scalar(DisasContext *s, arg_2scalar *a, MVEGenTwoOpScalarFn fn) -- 2.20.1 Implement the vector form of the MVE VQDMULL insn. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-39-peter.maydell@linaro.org --- target/arm/helper-mve.h | 5 +++++ target/arm/mve.decode | 5 +++++ target/arm/mve_helper.c | 30 ++++++++++++++++++++++++++++++ target/arm/translate-mve.c | 30 ++++++++++++++++++++++++++++++ 4 files changed, 70 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(mve_vqrdmlsdhxb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vqrdmlsdhxh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vqrdmlsdhxw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqdmullbh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqdmullbw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqdmullth, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vqdmulltw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + DEF_HELPER_FLAGS_4(mve_vadd_scalarb, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vadd_scalarh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vadd_scalarw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ @1op_nosz .... .... .... .... .... .... .... .... &1op qd=%qd qm=%qm size=0 @2op .... .... .. size:2 .... .... .... .... .... &2op qd=%qd qm=%qm qn=%qn @2op_nosz .... .... .... .... .... .... .... .... &2op qd=%qd qm=%qm qn=%qn size=0 +@2op_sz28 .... .... .... .... .... .... .... .... &2op qd=%qd qm=%qm qn=%qn \ + size=%size_28 # The _rev suffix indicates that Vn and Vm are reversed. This is # the case for shifts. In the Arm ARM these insns are documented @@ -XXX,XX +XXX,XX @@ VQDMLSDHX 1111 1110 0 . .. ... 0 ... 1 1110 . 0 . 0 ... 0 @2op VQRDMLSDH 1111 1110 0 . .. ... 0 ... 0 1110 . 0 . 0 ... 1 @2op VQRDMLSDHX 1111 1110 0 . .. ... 0 ... 1 1110 . 0 . 0 ... 1 @2op +VQDMULLB 111 . 1110 0 . 11 ... 0 ... 0 1111 . 0 . 0 ... 1 @2op_sz28 +VQDMULLT 111 . 1110 0 . 11 ... 0 ... 1 1111 . 0 . 0 ... 1 @2op_sz28 + # Vector miscellaneous VCLS 1111 1111 1 . 11 .. 00 ... 0 0100 01 . 0 ... 0 @1op diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ DO_2OP_SAT_SCALAR_L(vqdmullt_scalarh, 1, 2, int16_t, 4, int32_t, \ DO_2OP_SAT_SCALAR_L(vqdmullt_scalarw, 1, 4, int32_t, 8, int64_t, \ do_qdmullw, SATMASK32) +/* + * Long saturating ops + */ +#define DO_2OP_SAT_L(OP, TOP, ESIZE, TYPE, LESIZE, LTYPE, FN, SATMASK) \ + void HELPER(glue(mve_, OP))(CPUARMState *env, void *vd, void *vn, \ + void *vm) \ + { \ + LTYPE *d = vd; \ + TYPE *n = vn, *m = vm; \ + uint16_t mask = mve_element_mask(env); \ + unsigned le; \ + bool qc = false; \ + for (le = 0; le < 16 / LESIZE; le++, mask >>= LESIZE) { \ + bool sat = false; \ + LTYPE op1 = n[H##ESIZE(le * 2 + TOP)]; \ + LTYPE op2 = m[H##ESIZE(le * 2 + TOP)]; \ + mergemask(&d[H##LESIZE(le)], FN(op1, op2, &sat), mask); \ + qc |= sat && (mask & SATMASK); \ + } \ + if (qc) { \ + env->vfp.qc[0] = qc; \ + } \ + mve_advance_vpt(env); \ + } + +DO_2OP_SAT_L(vqdmullbh, 0, 2, int16_t, 4, int32_t, do_qdmullh, SATMASK16B) +DO_2OP_SAT_L(vqdmullbw, 0, 4, int32_t, 8, int64_t, do_qdmullw, SATMASK32) +DO_2OP_SAT_L(vqdmullth, 1, 2, int16_t, 4, int32_t, do_qdmullh, SATMASK16T) +DO_2OP_SAT_L(vqdmulltw, 1, 4, int32_t, 8, int64_t, do_qdmullw, SATMASK32) + static inline uint32_t do_vbrsrb(uint32_t n, uint32_t m) { m &= 0xff; diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ DO_2OP(VQDMLSDHX, vqdmlsdhx) DO_2OP(VQRDMLSDH, vqrdmlsdh) DO_2OP(VQRDMLSDHX, vqrdmlsdhx) +static bool trans_VQDMULLB(DisasContext *s, arg_2op *a) +{ + static MVEGenTwoOpFn * const fns[] = { + NULL, + gen_helper_mve_vqdmullbh, + gen_helper_mve_vqdmullbw, + NULL, + }; + if (a->size == MO_32 && (a->qd == a->qm || a->qd == a->qn)) { + /* UNPREDICTABLE; we choose to undef */ + return false; + } + return do_2op(s, a, fns[a->size]); +} + +static bool trans_VQDMULLT(DisasContext *s, arg_2op *a) +{ + static MVEGenTwoOpFn * const fns[] = { + NULL, + gen_helper_mve_vqdmullth, + gen_helper_mve_vqdmulltw, + NULL, + }; + if (a->size == MO_32 && (a->qd == a->qm || a->qd == a->qn)) { + /* UNPREDICTABLE; we choose to undef */ + return false; + } + return do_2op(s, a, fns[a->size]); +} + static bool do_2op_scalar(DisasContext *s, arg_2scalar *a, MVEGenTwoOpScalarFn fn) { -- 2.20.1 Implement the MVE VRHADD insn, which performs a rounded halving addition. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-40-peter.maydell@linaro.org --- target/arm/helper-mve.h | 8 ++++++++ target/arm/mve.decode | 3 +++ target/arm/mve_helper.c | 6 ++++++ target/arm/translate-mve.c | 2 ++ 4 files changed, 19 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(mve_vqdmullbw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vqdmullth, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vqdmulltw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vrhaddsb, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vrhaddsh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vrhaddsw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + +DEF_HELPER_FLAGS_4(mve_vrhaddub, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vrhadduh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vrhadduw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + DEF_HELPER_FLAGS_4(mve_vadd_scalarb, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vadd_scalarh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vadd_scalarw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ VQRDMLSDHX 1111 1110 0 . .. ... 0 ... 1 1110 . 0 . 0 ... 1 @2op VQDMULLB 111 . 1110 0 . 11 ... 0 ... 0 1111 . 0 . 0 ... 1 @2op_sz28 VQDMULLT 111 . 1110 0 . 11 ... 0 ... 1 1111 . 0 . 0 ... 1 @2op_sz28 +VRHADD_S 111 0 1111 0 . .. ... 0 ... 0 0001 . 1 . 0 ... 0 @2op +VRHADD_U 111 1 1111 0 . .. ... 0 ... 0 0001 . 1 . 0 ... 0 @2op + # Vector miscellaneous VCLS 1111 1111 1 . 11 .. 00 ... 0 0100 01 . 0 ... 0 @1op diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ DO_2OP_U(vshlu, DO_VSHLU) DO_2OP_S(vrshls, DO_VRSHLS) DO_2OP_U(vrshlu, DO_VRSHLU) +#define DO_RHADD_S(N, M) (((int64_t)(N) + (M) + 1) >> 1) +#define DO_RHADD_U(N, M) (((uint64_t)(N) + (M) + 1) >> 1) + +DO_2OP_S(vrhadds, DO_RHADD_S) +DO_2OP_U(vrhaddu, DO_RHADD_U) + static inline int32_t do_sat_bhw(int64_t val, int64_t min, int64_t max, bool *s) { if (val > max) { diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ DO_2OP(VQDMLSDH, vqdmlsdh) DO_2OP(VQDMLSDHX, vqdmlsdhx) DO_2OP(VQRDMLSDH, vqrdmlsdh) DO_2OP(VQRDMLSDHX, vqrdmlsdhx) +DO_2OP(VRHADD_S, vrhadds) +DO_2OP(VRHADD_U, vrhaddu) static bool trans_VQDMULLB(DisasContext *s, arg_2op *a) { -- 2.20.1 Implement the MVE VADC and VSBC insns. These perform an add-with-carry or subtract-with-carry of the 32-bit elements in each lane of the input vectors, where the carry-out of each add is the carry-in of the next. The initial carry input is either 1 or is from FPSCR.C; the carry out at the end is written back to FPSCR.C. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-41-peter.maydell@linaro.org --- target/arm/helper-mve.h | 5 ++++ target/arm/mve.decode | 5 ++++ target/arm/mve_helper.c | 52 ++++++++++++++++++++++++++++++++++++++ target/arm/translate-mve.c | 37 +++++++++++++++++++++++++++ 4 files changed, 99 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(mve_vrhaddub, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vrhadduh, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vrhadduw, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vadc, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vadci, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vsbc, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vsbci, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + DEF_HELPER_FLAGS_4(mve_vadd_scalarb, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vadd_scalarh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vadd_scalarw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ VQDMULLT 111 . 1110 0 . 11 ... 0 ... 1 1111 . 0 . 0 ... 1 @2op_sz28 VRHADD_S 111 0 1111 0 . .. ... 0 ... 0 0001 . 1 . 0 ... 0 @2op VRHADD_U 111 1 1111 0 . .. ... 0 ... 0 0001 . 1 . 0 ... 0 @2op +VADC 1110 1110 0 . 11 ... 0 ... 0 1111 . 0 . 0 ... 0 @2op_nosz +VSBC 1111 1110 0 . 11 ... 0 ... 0 1111 . 0 . 0 ... 0 @2op_nosz +VADCI 1110 1110 0 . 11 ... 0 ... 1 1111 . 0 . 0 ... 0 @2op_nosz +VSBCI 1111 1110 0 . 11 ... 0 ... 1 1111 . 0 . 0 ... 0 @2op_nosz + # Vector miscellaneous VCLS 1111 1111 1 . 11 .. 00 ... 0 0100 01 . 0 ... 0 @1op diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ DO_2OP_U(vrshlu, DO_VRSHLU) DO_2OP_S(vrhadds, DO_RHADD_S) DO_2OP_U(vrhaddu, DO_RHADD_U) +static void do_vadc(CPUARMState *env, uint32_t *d, uint32_t *n, uint32_t *m, + uint32_t inv, uint32_t carry_in, bool update_flags) +{ + uint16_t mask = mve_element_mask(env); + unsigned e; + + /* If any additions trigger, we will update flags. */ + if (mask & 0x1111) { + update_flags = true; + } + + for (e = 0; e < 16 / 4; e++, mask >>= 4) { + uint64_t r = carry_in; + r += n[H4(e)]; + r += m[H4(e)] ^ inv; + if (mask & 1) { + carry_in = r >> 32; + } + mergemask(&d[H4(e)], r, mask); + } + + if (update_flags) { + /* Store C, clear NZV. */ + env->vfp.xregs[ARM_VFP_FPSCR] &= ~FPCR_NZCV_MASK; + env->vfp.xregs[ARM_VFP_FPSCR] |= carry_in * FPCR_C; + } + mve_advance_vpt(env); +} + +void HELPER(mve_vadc)(CPUARMState *env, void *vd, void *vn, void *vm) +{ + bool carry_in = env->vfp.xregs[ARM_VFP_FPSCR] & FPCR_C; + do_vadc(env, vd, vn, vm, 0, carry_in, false); +} + +void HELPER(mve_vsbc)(CPUARMState *env, void *vd, void *vn, void *vm) +{ + bool carry_in = env->vfp.xregs[ARM_VFP_FPSCR] & FPCR_C; + do_vadc(env, vd, vn, vm, -1, carry_in, false); +} + + +void HELPER(mve_vadci)(CPUARMState *env, void *vd, void *vn, void *vm) +{ + do_vadc(env, vd, vn, vm, 0, 0, true); +} + +void HELPER(mve_vsbci)(CPUARMState *env, void *vd, void *vn, void *vm) +{ + do_vadc(env, vd, vn, vm, -1, 1, true); +} + static inline int32_t do_sat_bhw(int64_t val, int64_t min, int64_t max, bool *s) { if (val > max) { diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ static bool trans_VQDMULLT(DisasContext *s, arg_2op *a) return do_2op(s, a, fns[a->size]); } +/* + * VADC and VSBC: these perform an add-with-carry or subtract-with-carry + * of the 32-bit elements in each lane of the input vectors, where the + * carry-out of each add is the carry-in of the next. The initial carry + * input is either fixed (0 for VADCI, 1 for VSBCI) or is from FPSCR.C + * (for VADC and VSBC); the carry out at the end is written back to FPSCR.C. + * These insns are subject to beat-wise execution. Partial execution + * of an I=1 (initial carry input fixed) insn which does not + * execute the first beat must start with the current FPSCR.NZCV + * value, not the fixed constant input. + */ +static bool trans_VADC(DisasContext *s, arg_2op *a) +{ + return do_2op(s, a, gen_helper_mve_vadc); +} + +static bool trans_VADCI(DisasContext *s, arg_2op *a) +{ + if (mve_skip_first_beat(s)) { + return trans_VADC(s, a); + } + return do_2op(s, a, gen_helper_mve_vadci); +} + +static bool trans_VSBC(DisasContext *s, arg_2op *a) +{ + return do_2op(s, a, gen_helper_mve_vsbc); +} + +static bool trans_VSBCI(DisasContext *s, arg_2op *a) +{ + if (mve_skip_first_beat(s)) { + return trans_VSBC(s, a); + } + return do_2op(s, a, gen_helper_mve_vsbci); +} + static bool do_2op_scalar(DisasContext *s, arg_2scalar *a, MVEGenTwoOpScalarFn fn) { -- 2.20.1 Implement the MVE VCADD insn, which performs a complex add with rotate. Note that the size=0b11 encoding is VSBC. The architecture grants some leeway for the "destination and Vm source overlap" case for the size MO_32 case, but we choose not to make use of it, instead always calculating all 16 bytes worth of results before setting the destination register. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-42-peter.maydell@linaro.org --- target/arm/helper-mve.h | 8 ++++++++ target/arm/mve.decode | 9 +++++++-- target/arm/mve_helper.c | 29 +++++++++++++++++++++++++++++ target/arm/translate-mve.c | 7 +++++++ 4 files changed, 51 insertions(+), 2 deletions(-) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(mve_vadci, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vsbc, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vsbci, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vcadd90b, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vcadd90h, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vcadd90w, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + +DEF_HELPER_FLAGS_4(mve_vcadd270b, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vcadd270h, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vcadd270w, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + DEF_HELPER_FLAGS_4(mve_vadd_scalarb, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vadd_scalarh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vadd_scalarw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ VRHADD_S 111 0 1111 0 . .. ... 0 ... 0 0001 . 1 . 0 ... 0 @2op VRHADD_U 111 1 1111 0 . .. ... 0 ... 0 0001 . 1 . 0 ... 0 @2op VADC 1110 1110 0 . 11 ... 0 ... 0 1111 . 0 . 0 ... 0 @2op_nosz -VSBC 1111 1110 0 . 11 ... 0 ... 0 1111 . 0 . 0 ... 0 @2op_nosz VADCI 1110 1110 0 . 11 ... 0 ... 1 1111 . 0 . 0 ... 0 @2op_nosz -VSBCI 1111 1110 0 . 11 ... 0 ... 1 1111 . 0 . 0 ... 0 @2op_nosz + +{ + VSBC 1111 1110 0 . 11 ... 0 ... 0 1111 . 0 . 0 ... 0 @2op_nosz + VSBCI 1111 1110 0 . 11 ... 0 ... 1 1111 . 0 . 0 ... 0 @2op_nosz + VCADD90 1111 1110 0 . .. ... 0 ... 0 1111 . 0 . 0 ... 0 @2op + VCADD270 1111 1110 0 . .. ... 0 ... 1 1111 . 0 . 0 ... 0 @2op +} # Vector miscellaneous diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ void HELPER(mve_vsbci)(CPUARMState *env, void *vd, void *vn, void *vm) do_vadc(env, vd, vn, vm, -1, 1, true); } +#define DO_VCADD(OP, ESIZE, TYPE, FN0, FN1) \ + void HELPER(glue(mve_, OP))(CPUARMState *env, void *vd, void *vn, void *vm) \ + { \ + TYPE *d = vd, *n = vn, *m = vm; \ + uint16_t mask = mve_element_mask(env); \ + unsigned e; \ + TYPE r[16 / ESIZE]; \ + /* Calculate all results first to avoid overwriting inputs */ \ + for (e = 0; e < 16 / ESIZE; e++) { \ + if (!(e & 1)) { \ + r[e] = FN0(n[H##ESIZE(e)], m[H##ESIZE(e + 1)]); \ + } else { \ + r[e] = FN1(n[H##ESIZE(e)], m[H##ESIZE(e - 1)]); \ + } \ + } \ + for (e = 0; e < 16 / ESIZE; e++, mask >>= ESIZE) { \ + mergemask(&d[H##ESIZE(e)], r[e], mask); \ + } \ + mve_advance_vpt(env); \ + } + +#define DO_VCADD_ALL(OP, FN0, FN1) \ + DO_VCADD(OP##b, 1, int8_t, FN0, FN1) \ + DO_VCADD(OP##h, 2, int16_t, FN0, FN1) \ + DO_VCADD(OP##w, 4, int32_t, FN0, FN1) + +DO_VCADD_ALL(vcadd90, DO_SUB, DO_ADD) +DO_VCADD_ALL(vcadd270, DO_ADD, DO_SUB) + static inline int32_t do_sat_bhw(int64_t val, int64_t min, int64_t max, bool *s) { if (val > max) { diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ DO_2OP(VQRDMLSDH, vqrdmlsdh) DO_2OP(VQRDMLSDHX, vqrdmlsdhx) DO_2OP(VRHADD_S, vrhadds) DO_2OP(VRHADD_U, vrhaddu) +/* + * VCADD Qd == Qm at size MO_32 is UNPREDICTABLE; we choose not to diagnose + * so we can reuse the DO_2OP macro. (Our implementation calculates the + * "expected" results in this case.) + */ +DO_2OP(VCADD90, vcadd90) +DO_2OP(VCADD270, vcadd270) static bool trans_VQDMULLB(DisasContext *s, arg_2op *a) { -- 2.20.1 Implement the MVE VHCADD insn, which is similar to VCADD but performs a halving step. This one overlaps with VADC. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-43-peter.maydell@linaro.org --- target/arm/helper-mve.h | 8 ++++++++ target/arm/mve.decode | 8 ++++++-- target/arm/mve_helper.c | 2 ++ target/arm/translate-mve.c | 4 +++- 4 files changed, 19 insertions(+), 3 deletions(-) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(mve_vcadd270b, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vcadd270h, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) DEF_HELPER_FLAGS_4(mve_vcadd270w, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vhcadd90b, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vhcadd90h, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vhcadd90w, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + +DEF_HELPER_FLAGS_4(mve_vhcadd270b, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vhcadd270h, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) +DEF_HELPER_FLAGS_4(mve_vhcadd270w, TCG_CALL_NO_WG, void, env, ptr, ptr, ptr) + DEF_HELPER_FLAGS_4(mve_vadd_scalarb, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vadd_scalarh, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) DEF_HELPER_FLAGS_4(mve_vadd_scalarw, TCG_CALL_NO_WG, void, env, ptr, ptr, i32) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ VQDMULLT 111 . 1110 0 . 11 ... 0 ... 1 1111 . 0 . 0 ... 1 @2op_sz28 VRHADD_S 111 0 1111 0 . .. ... 0 ... 0 0001 . 1 . 0 ... 0 @2op VRHADD_U 111 1 1111 0 . .. ... 0 ... 0 0001 . 1 . 0 ... 0 @2op -VADC 1110 1110 0 . 11 ... 0 ... 0 1111 . 0 . 0 ... 0 @2op_nosz -VADCI 1110 1110 0 . 11 ... 0 ... 1 1111 . 0 . 0 ... 0 @2op_nosz +{ + VADC 1110 1110 0 . 11 ... 0 ... 0 1111 . 0 . 0 ... 0 @2op_nosz + VADCI 1110 1110 0 . 11 ... 0 ... 1 1111 . 0 . 0 ... 0 @2op_nosz + VHCADD90 1110 1110 0 . .. ... 0 ... 0 1111 . 0 . 0 ... 0 @2op + VHCADD270 1110 1110 0 . .. ... 0 ... 1 1111 . 0 . 0 ... 0 @2op +} { VSBC 1111 1110 0 . 11 ... 0 ... 0 1111 . 0 . 0 ... 0 @2op_nosz diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ void HELPER(mve_vsbci)(CPUARMState *env, void *vd, void *vn, void *vm) DO_VCADD_ALL(vcadd90, DO_SUB, DO_ADD) DO_VCADD_ALL(vcadd270, DO_ADD, DO_SUB) +DO_VCADD_ALL(vhcadd90, do_vhsub_s, do_vhadd_s) +DO_VCADD_ALL(vhcadd270, do_vhadd_s, do_vhsub_s) static inline int32_t do_sat_bhw(int64_t val, int64_t min, int64_t max, bool *s) { diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ DO_2OP(VRHADD_U, vrhaddu) /* * VCADD Qd == Qm at size MO_32 is UNPREDICTABLE; we choose not to diagnose * so we can reuse the DO_2OP macro. (Our implementation calculates the - * "expected" results in this case.) + * "expected" results in this case.) Similarly for VHCADD. */ DO_2OP(VCADD90, vcadd90) DO_2OP(VCADD270, vcadd270) +DO_2OP(VHCADD90, vhcadd90) +DO_2OP(VHCADD270, vhcadd270) static bool trans_VQDMULLB(DisasContext *s, arg_2op *a) { -- 2.20.1 Implement the MVE VADDV insn, which performs an addition across vector lanes. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-44-peter.maydell@linaro.org --- target/arm/helper-mve.h | 7 +++++++ target/arm/mve.decode | 2 ++ target/arm/mve_helper.c | 24 +++++++++++++++++++++ target/arm/translate-mve.c | 43 ++++++++++++++++++++++++++++++++++++++ 4 files changed, 76 insertions(+) diff --git a/target/arm/helper-mve.h b/target/arm/helper-mve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-mve.h +++ b/target/arm/helper-mve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(mve_vrmlaldavhuw, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) DEF_HELPER_FLAGS_4(mve_vrmlsldavhsw, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) DEF_HELPER_FLAGS_4(mve_vrmlsldavhxsw, TCG_CALL_NO_WG, i64, env, ptr, ptr, i64) + +DEF_HELPER_FLAGS_3(mve_vaddvsb, TCG_CALL_NO_WG, i32, env, ptr, i32) +DEF_HELPER_FLAGS_3(mve_vaddvub, TCG_CALL_NO_WG, i32, env, ptr, i32) +DEF_HELPER_FLAGS_3(mve_vaddvsh, TCG_CALL_NO_WG, i32, env, ptr, i32) +DEF_HELPER_FLAGS_3(mve_vaddvuh, TCG_CALL_NO_WG, i32, env, ptr, i32) +DEF_HELPER_FLAGS_3(mve_vaddvsw, TCG_CALL_NO_WG, i32, env, ptr, i32) +DEF_HELPER_FLAGS_3(mve_vaddvuw, TCG_CALL_NO_WG, i32, env, ptr, i32) diff --git a/target/arm/mve.decode b/target/arm/mve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve.decode +++ b/target/arm/mve.decode @@ -XXX,XX +XXX,XX @@ VBRSR 1111 1110 0 . .. ... 1 ... 1 1110 . 110 .... @2scalar VQDMULH_scalar 1110 1110 0 . .. ... 1 ... 0 1110 . 110 .... @2scalar VQRDMULH_scalar 1111 1110 0 . .. ... 1 ... 0 1110 . 110 .... @2scalar +# Vector add across vector +VADDV 111 u:1 1110 1111 size:2 01 ... 0 1111 0 0 a:1 0 qm:3 0 rda=%rdalo # Predicate operations %mask_22_13 22:1 13:3 diff --git a/target/arm/mve_helper.c b/target/arm/mve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mve_helper.c +++ b/target/arm/mve_helper.c @@ -XXX,XX +XXX,XX @@ DO_LDAVH(vrmlaldavhuw, 4, uint32_t, false, int128_add, int128_add, int128_make64 DO_LDAVH(vrmlsldavhsw, 4, int32_t, false, int128_add, int128_sub, int128_makes64) DO_LDAVH(vrmlsldavhxsw, 4, int32_t, true, int128_add, int128_sub, int128_makes64) + +/* Vector add across vector */ +#define DO_VADDV(OP, ESIZE, TYPE) \ + uint32_t HELPER(glue(mve_, OP))(CPUARMState *env, void *vm, \ + uint32_t ra) \ + { \ + uint16_t mask = mve_element_mask(env); \ + unsigned e; \ + TYPE *m = vm; \ + for (e = 0; e < 16 / ESIZE; e++, mask >>= ESIZE) { \ + if (mask & 1) { \ + ra += m[H##ESIZE(e)]; \ + } \ + } \ + mve_advance_vpt(env); \ + return ra; \ + } \ + +DO_VADDV(vaddvsb, 1, uint8_t) +DO_VADDV(vaddvsh, 2, uint16_t) +DO_VADDV(vaddvsw, 4, uint32_t) +DO_VADDV(vaddvub, 1, uint8_t) +DO_VADDV(vaddvuh, 2, uint16_t) +DO_VADDV(vaddvuw, 4, uint32_t) diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ typedef void MVEGenOneOpFn(TCGv_ptr, TCGv_ptr, TCGv_ptr); typedef void MVEGenTwoOpFn(TCGv_ptr, TCGv_ptr, TCGv_ptr, TCGv_ptr); typedef void MVEGenTwoOpScalarFn(TCGv_ptr, TCGv_ptr, TCGv_ptr, TCGv_i32); typedef void MVEGenDualAccOpFn(TCGv_i64, TCGv_ptr, TCGv_ptr, TCGv_ptr, TCGv_i64); +typedef void MVEGenVADDVFn(TCGv_i32, TCGv_ptr, TCGv_ptr, TCGv_i32); /* Return the offset of a Qn register (same semantics as aa32_vfp_qreg()) */ static inline long mve_qreg_offset(unsigned reg) @@ -XXX,XX +XXX,XX @@ static bool trans_VPST(DisasContext *s, arg_VPST *a) mve_update_and_store_eci(s); return true; } + +static bool trans_VADDV(DisasContext *s, arg_VADDV *a) +{ + /* VADDV: vector add across vector */ + static MVEGenVADDVFn * const fns[4][2] = { + { gen_helper_mve_vaddvsb, gen_helper_mve_vaddvub }, + { gen_helper_mve_vaddvsh, gen_helper_mve_vaddvuh }, + { gen_helper_mve_vaddvsw, gen_helper_mve_vaddvuw }, + { NULL, NULL } + }; + TCGv_ptr qm; + TCGv_i32 rda; + + if (!dc_isar_feature(aa32_mve, s) || + a->size == 3) { + return false; + } + if (!mve_eci_check(s) || !vfp_access_check(s)) { + return true; + } + + /* + * This insn is subject to beat-wise execution. Partial execution + * of an A=0 (no-accumulate) insn which does not execute the first + * beat must start with the current value of Rda, not zero. + */ + if (a->a || mve_skip_first_beat(s)) { + /* Accumulate input from Rda */ + rda = load_reg(s, a->rda); + } else { + /* Accumulate starting at zero */ + rda = tcg_const_i32(0); + } + + qm = mve_qreg_ptr(a->qm); + fns[a->size][a->u](rda, cpu_env, qm, rda); + store_reg(s, a->rda, rda); + tcg_temp_free_ptr(qm); + + mve_update_eci(s); + return true; +} -- 2.20.1 In a CPU with MVE, the VMOV (vector lane to general-purpose register) and VMOV (general-purpose register to vector lane) insns are not predicated, but they are subject to beatwise execution if they are not in an IT block. Since our implementation always executes all 4 beats in one tick, this means only that we need to handle PSR.ECI: * we must do the usual check for bad ECI state * we must advance ECI state if the insn succeeds * if ECI says we should not be executing the beat corresponding to the lane of the vector register being accessed then we should skip performing the move Note that if PSR.ECI is non-zero then we cannot be in an IT block. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210617121628.20116-45-peter.maydell@linaro.org --- target/arm/translate-a32.h | 2 + target/arm/translate-mve.c | 4 +- target/arm/translate-vfp.c | 77 +++++++++++++++++++++++++++++++++++--- 3 files changed, 75 insertions(+), 8 deletions(-) diff --git a/target/arm/translate-a32.h b/target/arm/translate-a32.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a32.h +++ b/target/arm/translate-a32.h @@ -XXX,XX +XXX,XX @@ long neon_full_reg_offset(unsigned reg); long neon_element_offset(int reg, int element, MemOp memop); void gen_rev16(TCGv_i32 dest, TCGv_i32 var); void clear_eci_state(DisasContext *s); +bool mve_eci_check(DisasContext *s); +void mve_update_and_store_eci(DisasContext *s); static inline TCGv_i32 load_cpu_offset(int offset) { diff --git a/target/arm/translate-mve.c b/target/arm/translate-mve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-mve.c +++ b/target/arm/translate-mve.c @@ -XXX,XX +XXX,XX @@ static bool mve_check_qreg_bank(DisasContext *s, int qmask) return qmask < 8; } -static bool mve_eci_check(DisasContext *s) +bool mve_eci_check(DisasContext *s) { /* * This is a beatwise insn: check that ECI is valid (not a @@ -XXX,XX +XXX,XX @@ static void mve_update_eci(DisasContext *s) } } -static void mve_update_and_store_eci(DisasContext *s) +void mve_update_and_store_eci(DisasContext *s) { /* * For insns which don't call a helper function that will call 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 trans_VCVT(DisasContext *s, arg_VCVT *a) return true; } +static bool mve_skip_vmov(DisasContext *s, int vn, int index, int size) +{ + /* + * In a CPU with MVE, the VMOV (vector lane to general-purpose register) + * and VMOV (general-purpose register to vector lane) insns are not + * predicated, but they are subject to beatwise execution if they are + * not in an IT block. + * + * Since our implementation always executes all 4 beats in one tick, + * this means only that if PSR.ECI says we should not be executing + * the beat corresponding to the lane of the vector register being + * accessed then we should skip performing the move, and that we need + * to do the usual check for bad ECI state and advance of ECI state. + * + * Note that if PSR.ECI is non-zero then we cannot be in an IT block. + * + * Return true if this VMOV scalar <-> gpreg should be skipped because + * the MVE PSR.ECI state says we skip the beat where the store happens. + */ + + /* Calculate the byte offset into Qn which we're going to access */ + int ofs = (index << size) + ((vn & 1) * 8); + + if (!dc_isar_feature(aa32_mve, s)) { + return false; + } + + switch (s->eci) { + case ECI_NONE: + return false; + case ECI_A0: + return ofs < 4; + case ECI_A0A1: + return ofs < 8; + case ECI_A0A1A2: + case ECI_A0A1A2B0: + return ofs < 12; + default: + g_assert_not_reached(); + } +} + static bool trans_VMOV_to_gp(DisasContext *s, arg_VMOV_to_gp *a) { /* VMOV scalar to general purpose register */ @@ -XXX,XX +XXX,XX @@ static bool trans_VMOV_to_gp(DisasContext *s, arg_VMOV_to_gp *a) return false; } + if (dc_isar_feature(aa32_mve, s)) { + if (!mve_eci_check(s)) { + return true; + } + } + if (!vfp_access_check(s)) { return true; } - tmp = tcg_temp_new_i32(); - read_neon_element32(tmp, a->vn, a->index, a->size | (a->u ? 0 : MO_SIGN)); - store_reg(s, a->rt, tmp); + if (!mve_skip_vmov(s, a->vn, a->index, a->size)) { + tmp = tcg_temp_new_i32(); + read_neon_element32(tmp, a->vn, a->index, + a->size | (a->u ? 0 : MO_SIGN)); + store_reg(s, a->rt, tmp); + } + if (dc_isar_feature(aa32_mve, s)) { + mve_update_and_store_eci(s); + } return true; } @@ -XXX,XX +XXX,XX @@ static bool trans_VMOV_from_gp(DisasContext *s, arg_VMOV_from_gp *a) return false; } + if (dc_isar_feature(aa32_mve, s)) { + if (!mve_eci_check(s)) { + return true; + } + } + if (!vfp_access_check(s)) { return true; } - tmp = load_reg(s, a->rt); - write_neon_element32(tmp, a->vn, a->index, a->size); - tcg_temp_free_i32(tmp); + if (!mve_skip_vmov(s, a->vn, a->index, a->size)) { + tmp = load_reg(s, a->rt); + write_neon_element32(tmp, a->vn, a->index, a->size); + tcg_temp_free_i32(tmp); + } + if (dc_isar_feature(aa32_mve, s)) { + mve_update_and_store_eci(s); + } return true; } -- 2.20.1 From: Peter Collingbourne <pcc@google.com> MTE3 introduces an asymmetric tag checking mode, in which loads are checked synchronously and stores are checked asynchronously. Add support for it. Signed-off-by: Peter Collingbourne <pcc@google.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210616195614.11785-1-pcc@google.com [PMM: Add line to emulation.rst] Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- docs/system/arm/emulation.rst | 1 + target/arm/cpu64.c | 2 +- target/arm/mte_helper.c | 82 ++++++++++++++++++++++------------- 3 files changed, 53 insertions(+), 32 deletions(-) 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_LSE (Large System Extensions) - FEAT_MTE (Memory Tagging Extension) - FEAT_MTE2 (Memory Tagging Extension) +- FEAT_MTE3 (MTE Asymmetric Fault Handling) - FEAT_PAN (Privileged access never) - FEAT_PAN2 (AT S1E1R and AT S1E1W instruction variants affected by PSTATE.PAN) - FEAT_PAuth (Pointer authentication) 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) * during realize if the board provides no tag memory, much like * we do for EL2 with the virtualization=on property. */ - t = FIELD_DP64(t, ID_AA64PFR1, MTE, 2); + t = FIELD_DP64(t, ID_AA64PFR1, MTE, 3); cpu->isar.id_aa64pfr1 = t; t = cpu->isar.id_aa64mmfr0; diff --git a/target/arm/mte_helper.c b/target/arm/mte_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/mte_helper.c +++ b/target/arm/mte_helper.c @@ -XXX,XX +XXX,XX @@ void HELPER(stzgm_tags)(CPUARMState *env, uint64_t ptr, uint64_t val) } } +static void mte_sync_check_fail(CPUARMState *env, uint32_t desc, + uint64_t dirty_ptr, uintptr_t ra) +{ + int is_write, syn; + + env->exception.vaddress = dirty_ptr; + + is_write = FIELD_EX32(desc, MTEDESC, WRITE); + syn = syn_data_abort_no_iss(arm_current_el(env) != 0, 0, 0, 0, 0, is_write, + 0x11); + raise_exception_ra(env, EXCP_DATA_ABORT, syn, exception_target_el(env), ra); + g_assert_not_reached(); +} + +static void mte_async_check_fail(CPUARMState *env, uint64_t dirty_ptr, + uintptr_t ra, ARMMMUIdx arm_mmu_idx, int el) +{ + int select; + + if (regime_has_2_ranges(arm_mmu_idx)) { + select = extract64(dirty_ptr, 55, 1); + } else { + select = 0; + } + env->cp15.tfsr_el[el] |= 1 << select; +#ifdef CONFIG_USER_ONLY + /* + * Stand in for a timer irq, setting _TIF_MTE_ASYNC_FAULT, + * which then sends a SIGSEGV when the thread is next scheduled. + * This cpu will return to the main loop at the end of the TB, + * which is rather sooner than "normal". But the alternative + * is waiting until the next syscall. + */ + qemu_cpu_kick(env_cpu(env)); +#endif +} + /* Record a tag check failure. */ static void mte_check_fail(CPUARMState *env, uint32_t desc, uint64_t dirty_ptr, uintptr_t ra) { int mmu_idx = FIELD_EX32(desc, MTEDESC, MIDX); ARMMMUIdx arm_mmu_idx = core_to_aa64_mmu_idx(mmu_idx); - int el, reg_el, tcf, select, is_write, syn; + int el, reg_el, tcf; uint64_t sctlr; reg_el = regime_el(env, arm_mmu_idx); @@ -XXX,XX +XXX,XX @@ static void mte_check_fail(CPUARMState *env, uint32_t desc, switch (tcf) { case 1: /* Tag check fail causes a synchronous exception. */ - env->exception.vaddress = dirty_ptr; - - is_write = FIELD_EX32(desc, MTEDESC, WRITE); - syn = syn_data_abort_no_iss(arm_current_el(env) != 0, 0, 0, 0, 0, - is_write, 0x11); - raise_exception_ra(env, EXCP_DATA_ABORT, syn, - exception_target_el(env), ra); - /* noreturn, but fall through to the assert anyway */ + mte_sync_check_fail(env, desc, dirty_ptr, ra); + break; case 0: /* @@ -XXX,XX +XXX,XX @@ static void mte_check_fail(CPUARMState *env, uint32_t desc, case 2: /* Tag check fail causes asynchronous flag set. */ - if (regime_has_2_ranges(arm_mmu_idx)) { - select = extract64(dirty_ptr, 55, 1); - } else { - select = 0; - } - env->cp15.tfsr_el[el] |= 1 << select; -#ifdef CONFIG_USER_ONLY - /* - * Stand in for a timer irq, setting _TIF_MTE_ASYNC_FAULT, - * which then sends a SIGSEGV when the thread is next scheduled. - * This cpu will return to the main loop at the end of the TB, - * which is rather sooner than "normal". But the alternative - * is waiting until the next syscall. - */ - qemu_cpu_kick(env_cpu(env)); -#endif + mte_async_check_fail(env, dirty_ptr, ra, arm_mmu_idx, el); break; - default: - /* Case 3: Reserved. */ - qemu_log_mask(LOG_GUEST_ERROR, - "Tag check failure with SCTLR_EL%d.TCF%s " - "set to reserved value %d\n", - reg_el, el ? "" : "0", tcf); + case 3: + /* + * Tag check fail causes asynchronous flag set for stores, or + * a synchronous exception for loads. + */ + if (FIELD_EX32(desc, MTEDESC, WRITE)) { + mte_async_check_fail(env, dirty_ptr, ra, arm_mmu_idx, el); + } else { + mte_sync_check_fail(env, desc, dirty_ptr, ra); + } break; } } -- 2.20.1 From: Alexandre Iooss <erdnaxe@crans.org> This adds the target guide for BBC Micro:bit. Information is taken from https://wiki.qemu.org/Features/MicroBit and from hw/arm/nrf51_soc.c. Signed-off-by: Alexandre Iooss <erdnaxe@crans.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Joel Stanley <joel@jms.id.au> Message-id: 20210621075625.540471-1-erdnaxe@crans.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- docs/system/arm/nrf.rst | 51 ++++++++++++++++++++++++++++++++++++++ docs/system/target-arm.rst | 1 + MAINTAINERS | 1 + 3 files changed, 53 insertions(+) create mode 100644 docs/system/arm/nrf.rst diff --git a/docs/system/arm/nrf.rst b/docs/system/arm/nrf.rst new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/docs/system/arm/nrf.rst @@ -XXX,XX +XXX,XX @@ +Nordic nRF boards (``microbit``) +================================ + +The `Nordic nRF`_ chips are a family of ARM-based System-on-Chip that +are designed to be used for low-power and short-range wireless solutions. + +.. _Nordic nRF: https://www.nordicsemi.com/Products + +The nRF51 series is the first series for short range wireless applications. +It is superseded by the nRF52 series. +The following machines are based on this chip : + +- ``microbit`` BBC micro:bit board with nRF51822 SoC + +There are other series such as nRF52, nRF53 and nRF91 which are currently not +supported by QEMU. + +Supported devices +----------------- + + * ARM Cortex-M0 (ARMv6-M) + * Serial ports (UART) + * Clock controller + * Timers + * Random Number Generator (RNG) + * GPIO controller + * NVMC + * SWI + +Missing devices +--------------- + + * Watchdog + * Real-Time Clock (RTC) controller + * TWI (i2c) + * SPI controller + * Analog to Digital Converter (ADC) + * Quadrature decoder + * Radio + +Boot options +------------ + +The Micro:bit machine can be started using the ``-device`` option to load a +firmware in `ihex format`_. Example: + +.. _ihex format: https://en.wikipedia.org/wiki/Intel_HEX + +.. code-block:: bash + + $ qemu-system-arm -M microbit -device loader,file=test.hex diff --git a/docs/system/target-arm.rst b/docs/system/target-arm.rst index XXXXXXX..XXXXXXX 100644 --- a/docs/system/target-arm.rst +++ b/docs/system/target-arm.rst @@ -XXX,XX +XXX,XX @@ undocumented; you can get a complete list by running arm/digic arm/musicpal arm/gumstix + arm/nrf arm/nseries arm/nuvoton arm/orangepi diff --git a/MAINTAINERS b/MAINTAINERS index XXXXXXX..XXXXXXX 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -XXX,XX +XXX,XX @@ F: hw/*/microbit*.c F: include/hw/*/nrf51*.h F: include/hw/*/microbit*.h F: tests/qtest/microbit-test.c +F: docs/system/arm/nrf.rst AVR Machines ------------- -- 2.20.1 The following changes since commit 3db29dcac23da85486704ef9e7a8e7217f7829cd: Merge tag 'for-upstream' of https://gitlab.com/bonzini/qemu into staging (2023-01-12 13:51:36 +0000) are available in the Git repository at: https://git.linaro.org/people/pmaydell/qemu-arm.git tags/pull-target-arm-20230113 for you to fetch changes up to 08899b5c68a55a3780d707e2464073c8f2670d31: target/arm: allow writes to SCR_EL3.HXEn bit when FEAT_HCX is enabled (2023-01-13 13:19:36 +0000) ---------------------------------------------------------------- target-arm queue: hw/arm/stm32f405: correctly describe the memory layout hw/arm: Add Olimex H405 board cubieboard: Support booting from an SD card image with u-boot on it target/arm: Fix sve_probe_page target/arm: allow writes to SCR_EL3.HXEn bit when FEAT_HCX is enabled various code cleanups ---------------------------------------------------------------- Evgeny Iakovlev (1): target/arm: allow writes to SCR_EL3.HXEn bit when FEAT_HCX is enabled Felipe Balbi (2): hw/arm/stm32f405: correctly describe the memory layout hw/arm: Add Olimex H405 Philippe Mathieu-Daudé (27): hw/arm/pxa2xx: Simplify pxa255_init() hw/arm/pxa2xx: Simplify pxa270_init() hw/arm/collie: Use the IEC binary prefix definitions hw/arm/collie: Simplify flash creation using for() loop hw/arm/gumstix: Improve documentation hw/arm/gumstix: Use the IEC binary prefix definitions hw/arm/mainstone: Use the IEC binary prefix definitions hw/arm/musicpal: Use the IEC binary prefix definitions hw/arm/omap_sx1: Remove unused 'total_ram' definitions hw/arm/omap_sx1: Use the IEC binary prefix definitions hw/arm/z2: Use the IEC binary prefix definitions hw/arm/vexpress: Remove dead code in vexpress_common_init() hw/arm: Remove unreachable code calling pflash_cfi01_register() hw/arm/pxa: Avoid forward-declaring PXA2xxI2CState hw/gpio/omap_gpio: Add local variable to avoid embedded cast hw/arm/omap: Drop useless casts from void * to pointer hw/gpio/omap_gpio: Use CamelCase for TYPE_OMAP1_GPIO type name hw/gpio/omap_gpio: Use CamelCase for TYPE_OMAP2_GPIO type name hw/intc/omap_intc: Use CamelCase for TYPE_OMAP_INTC type name hw/arm/stellaris: Drop useless casts from void * to pointer hw/arm/stellaris: Use CamelCase for STELLARIS_ADC type name hw/arm/bcm2836: Remove definitions generated by OBJECT_DECLARE_TYPE() hw/arm/npcm7xx: Declare QOM macros using OBJECT_DECLARE_SIMPLE_TYPE() hw/misc/sbsa_ec: Rename TYPE_SBSA_EC -> TYPE_SBSA_SECURE_EC hw/misc/sbsa_ec: Declare QOM macros using OBJECT_DECLARE_SIMPLE_TYPE() hw/intc/xilinx_intc: Use 'XpsIntc' typedef instead of 'struct xlx_pic' hw/timer/xilinx_timer: Use XpsTimerState instead of 'struct timerblock' Richard Henderson (1): target/arm: Fix sve_probe_page Strahinja Jankovic (7): hw/misc: Allwinner-A10 Clock Controller Module Emulation hw/misc: Allwinner A10 DRAM Controller Emulation {hw/i2c,docs/system/arm}: Allwinner TWI/I2C Emulation hw/misc: AXP209 PMU Emulation hw/arm: Add AXP209 to Cubieboard hw/arm: Allwinner A10 enable SPL load from MMC tests/avocado: Add SD boot test to Cubieboard docs/system/arm/cubieboard.rst | 1 + docs/system/arm/orangepi.rst | 1 + docs/system/arm/stm32.rst | 1 + configs/devices/arm-softmmu/default.mak | 1 + include/hw/adc/npcm7xx_adc.h | 7 +- include/hw/arm/allwinner-a10.h | 27 ++ include/hw/arm/allwinner-h3.h | 3 + include/hw/arm/npcm7xx.h | 18 +- include/hw/arm/omap.h | 24 +- include/hw/arm/pxa.h | 11 +- include/hw/arm/stm32f405_soc.h | 5 +- include/hw/i2c/allwinner-i2c.h | 55 ++++ include/hw/i2c/npcm7xx_smbus.h | 7 +- include/hw/misc/allwinner-a10-ccm.h | 67 +++++ include/hw/misc/allwinner-a10-dramc.h | 68 +++++ include/hw/misc/npcm7xx_clk.h | 2 +- include/hw/misc/npcm7xx_gcr.h | 6 +- include/hw/misc/npcm7xx_mft.h | 7 +- include/hw/misc/npcm7xx_pwm.h | 3 +- include/hw/misc/npcm7xx_rng.h | 6 +- include/hw/net/npcm7xx_emc.h | 5 +- include/hw/sd/npcm7xx_sdhci.h | 4 +- hw/arm/allwinner-a10.c | 40 +++ hw/arm/allwinner-h3.c | 11 +- hw/arm/bcm2836.c | 9 +- hw/arm/collie.c | 25 +- hw/arm/cubieboard.c | 11 + hw/arm/gumstix.c | 45 ++-- hw/arm/mainstone.c | 37 ++- hw/arm/musicpal.c | 9 +- hw/arm/olimex-stm32-h405.c | 69 +++++ hw/arm/omap1.c | 115 ++++---- hw/arm/omap2.c | 40 ++- hw/arm/omap_sx1.c | 53 ++-- hw/arm/palm.c | 2 +- hw/arm/pxa2xx.c | 8 +- hw/arm/spitz.c | 6 +- hw/arm/stellaris.c | 73 +++-- hw/arm/stm32f405_soc.c | 8 + hw/arm/tosa.c | 2 +- hw/arm/versatilepb.c | 6 +- hw/arm/vexpress.c | 10 +- hw/arm/z2.c | 16 +- hw/char/omap_uart.c | 7 +- hw/display/omap_dss.c | 15 +- hw/display/omap_lcdc.c | 9 +- hw/dma/omap_dma.c | 15 +- hw/gpio/omap_gpio.c | 48 ++-- hw/i2c/allwinner-i2c.c | 459 ++++++++++++++++++++++++++++++++ hw/intc/omap_intc.c | 38 +-- hw/intc/xilinx_intc.c | 28 +- hw/misc/allwinner-a10-ccm.c | 224 ++++++++++++++++ hw/misc/allwinner-a10-dramc.c | 179 +++++++++++++ hw/misc/axp209.c | 238 +++++++++++++++++ hw/misc/omap_gpmc.c | 12 +- hw/misc/omap_l4.c | 7 +- hw/misc/omap_sdrc.c | 7 +- hw/misc/omap_tap.c | 5 +- hw/misc/sbsa_ec.c | 12 +- hw/sd/omap_mmc.c | 9 +- hw/ssi/omap_spi.c | 7 +- hw/timer/omap_gptimer.c | 22 +- hw/timer/omap_synctimer.c | 4 +- hw/timer/xilinx_timer.c | 27 +- target/arm/helper.c | 3 + target/arm/sve_helper.c | 14 +- MAINTAINERS | 8 + hw/arm/Kconfig | 9 + hw/arm/meson.build | 1 + hw/i2c/Kconfig | 4 + hw/i2c/meson.build | 1 + hw/i2c/trace-events | 5 + hw/misc/Kconfig | 10 + hw/misc/meson.build | 3 + hw/misc/trace-events | 5 + tests/avocado/boot_linux_console.py | 47 ++++ 76 files changed, 1951 insertions(+), 455 deletions(-) create mode 100644 include/hw/i2c/allwinner-i2c.h create mode 100644 include/hw/misc/allwinner-a10-ccm.h create mode 100644 include/hw/misc/allwinner-a10-dramc.h create mode 100644 hw/arm/olimex-stm32-h405.c create mode 100644 hw/i2c/allwinner-i2c.c create mode 100644 hw/misc/allwinner-a10-ccm.c create mode 100644 hw/misc/allwinner-a10-dramc.c create mode 100644 hw/misc/axp209.c From: Felipe Balbi <balbi@kernel.org> STM32F405 has 128K of SRAM and another 64K of CCM (Core-coupled Memory) at a different base address. Correctly describe the memory layout to give existing FW images a chance to run unmodified. Reviewed-by: Alistair Francis <alistair@alistair23.me> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Felipe Balbi <balbi@kernel.org> Message-id: 20221230145733.200496-2-balbi@kernel.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/stm32f405_soc.h | 5 ++++- hw/arm/stm32f405_soc.c | 8 ++++++++ 2 files changed, 12 insertions(+), 1 deletion(-) diff --git a/include/hw/arm/stm32f405_soc.h b/include/hw/arm/stm32f405_soc.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/stm32f405_soc.h +++ b/include/hw/arm/stm32f405_soc.h @@ -XXX,XX +XXX,XX @@ OBJECT_DECLARE_SIMPLE_TYPE(STM32F405State, STM32F405_SOC) #define FLASH_BASE_ADDRESS 0x08000000 #define FLASH_SIZE (1024 * 1024) #define SRAM_BASE_ADDRESS 0x20000000 -#define SRAM_SIZE (192 * 1024) +#define SRAM_SIZE (128 * 1024) +#define CCM_BASE_ADDRESS 0x10000000 +#define CCM_SIZE (64 * 1024) struct STM32F405State { /*< private >*/ @@ -XXX,XX +XXX,XX @@ struct STM32F405State { STM32F2XXADCState adc[STM_NUM_ADCS]; STM32F2XXSPIState spi[STM_NUM_SPIS]; + MemoryRegion ccm; MemoryRegion sram; MemoryRegion flash; MemoryRegion flash_alias; diff --git a/hw/arm/stm32f405_soc.c b/hw/arm/stm32f405_soc.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/stm32f405_soc.c +++ b/hw/arm/stm32f405_soc.c @@ -XXX,XX +XXX,XX @@ static void stm32f405_soc_realize(DeviceState *dev_soc, Error **errp) } memory_region_add_subregion(system_memory, SRAM_BASE_ADDRESS, &s->sram); + memory_region_init_ram(&s->ccm, NULL, "STM32F405.ccm", CCM_SIZE, + &err); + if (err != NULL) { + error_propagate(errp, err); + return; + } + memory_region_add_subregion(system_memory, CCM_BASE_ADDRESS, &s->ccm); + armv7m = DEVICE(&s->armv7m); qdev_prop_set_uint32(armv7m, "num-irq", 96); qdev_prop_set_string(armv7m, "cpu-type", s->cpu_type); -- 2.34.1 From: Felipe Balbi <balbi@kernel.org> Olimex makes a series of low-cost STM32 boards. This commit introduces the minimum setup to support SMT32-H405. See [1] for details [1] https://www.olimex.com/Products/ARM/ST/STM32-H405/ Signed-off-by: Felipe Balbi <balbi@kernel.org> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-id: 20221230145733.200496-3-balbi@kernel.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- docs/system/arm/stm32.rst | 1 + configs/devices/arm-softmmu/default.mak | 1 + hw/arm/olimex-stm32-h405.c | 69 +++++++++++++++++++++++++ MAINTAINERS | 6 +++ hw/arm/Kconfig | 4 ++ hw/arm/meson.build | 1 + 6 files changed, 82 insertions(+) create mode 100644 hw/arm/olimex-stm32-h405.c diff --git a/docs/system/arm/stm32.rst b/docs/system/arm/stm32.rst index XXXXXXX..XXXXXXX 100644 --- a/docs/system/arm/stm32.rst +++ b/docs/system/arm/stm32.rst @@ -XXX,XX +XXX,XX @@ The STM32F4 series is based on ARM Cortex-M4F core. This series is pin-to-pin compatible with STM32F2 series. The following machines are based on this chip : - ``netduinoplus2`` Netduino Plus 2 board with STM32F405RGT6 microcontroller +- ``olimex-stm32-h405`` Olimex STM32 H405 board with STM32F405RGT6 microcontroller There are many other STM32 series that are currently not supported by QEMU. diff --git a/configs/devices/arm-softmmu/default.mak b/configs/devices/arm-softmmu/default.mak index XXXXXXX..XXXXXXX 100644 --- a/configs/devices/arm-softmmu/default.mak +++ b/configs/devices/arm-softmmu/default.mak @@ -XXX,XX +XXX,XX @@ CONFIG_COLLIE=y CONFIG_ASPEED_SOC=y CONFIG_NETDUINO2=y CONFIG_NETDUINOPLUS2=y +CONFIG_OLIMEX_STM32_H405=y CONFIG_MPS2=y CONFIG_RASPI=y CONFIG_DIGIC=y diff --git a/hw/arm/olimex-stm32-h405.c b/hw/arm/olimex-stm32-h405.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/arm/olimex-stm32-h405.c @@ -XXX,XX +XXX,XX @@ +/* + * ST STM32VLDISCOVERY machine + * Olimex STM32-H405 machine + * + * Copyright (c) 2022 Felipe Balbi <balbi@kernel.org> + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN + * THE SOFTWARE. + */ + +#include "qemu/osdep.h" +#include "qapi/error.h" +#include "hw/boards.h" +#include "hw/qdev-properties.h" +#include "hw/qdev-clock.h" +#include "qemu/error-report.h" +#include "hw/arm/stm32f405_soc.h" +#include "hw/arm/boot.h" + +/* olimex-stm32-h405 implementation is derived from netduinoplus2 */ + +/* Main SYSCLK frequency in Hz (168MHz) */ +#define SYSCLK_FRQ 168000000ULL + +static void olimex_stm32_h405_init(MachineState *machine) +{ + DeviceState *dev; + Clock *sysclk; + + /* This clock doesn't need migration because it is fixed-frequency */ + sysclk = clock_new(OBJECT(machine), "SYSCLK"); + clock_set_hz(sysclk, SYSCLK_FRQ); + + dev = qdev_new(TYPE_STM32F405_SOC); + qdev_prop_set_string(dev, "cpu-type", ARM_CPU_TYPE_NAME("cortex-m4")); + qdev_connect_clock_in(dev, "sysclk", sysclk); + sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), &error_fatal); + + armv7m_load_kernel(ARM_CPU(first_cpu), + machine->kernel_filename, + 0, FLASH_SIZE); +} + +static void olimex_stm32_h405_machine_init(MachineClass *mc) +{ + mc->desc = "Olimex STM32-H405 (Cortex-M4)"; + mc->init = olimex_stm32_h405_init; + mc->default_cpu_type = ARM_CPU_TYPE_NAME("cortex-m4"); + + /* SRAM pre-allocated as part of the SoC instantiation */ + mc->default_ram_size = 0; +} + +DEFINE_MACHINE("olimex-stm32-h405", olimex_stm32_h405_machine_init) diff --git a/MAINTAINERS b/MAINTAINERS index XXXXXXX..XXXXXXX 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -XXX,XX +XXX,XX @@ L: qemu-arm@nongnu.org S: Maintained F: hw/arm/netduinoplus2.c +Olimex STM32 H405 +M: Felipe Balbi <balbi@kernel.org> +L: qemu-arm@nongnu.org +S: Maintained +F: hw/arm/olimex-stm32-h405.c + SmartFusion2 M: Subbaraya Sundeep <sundeep.lkml@gmail.com> M: Peter Maydell <peter.maydell@linaro.org> diff --git a/hw/arm/Kconfig b/hw/arm/Kconfig index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/Kconfig +++ b/hw/arm/Kconfig @@ -XXX,XX +XXX,XX @@ config NETDUINOPLUS2 bool select STM32F405_SOC +config OLIMEX_STM32_H405 + bool + select STM32F405_SOC + config NSERIES bool select OMAP diff --git a/hw/arm/meson.build b/hw/arm/meson.build index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/meson.build +++ b/hw/arm/meson.build @@ -XXX,XX +XXX,XX @@ arm_ss.add(when: 'CONFIG_MICROBIT', if_true: files('microbit.c')) arm_ss.add(when: 'CONFIG_MUSICPAL', if_true: files('musicpal.c')) arm_ss.add(when: 'CONFIG_NETDUINO2', if_true: files('netduino2.c')) arm_ss.add(when: 'CONFIG_NETDUINOPLUS2', if_true: files('netduinoplus2.c')) +arm_ss.add(when: 'CONFIG_OLIMEX_STM32_H405', if_true: files('olimex-stm32-h405.c')) arm_ss.add(when: 'CONFIG_NPCM7XX', if_true: files('npcm7xx.c', 'npcm7xx_boards.c')) arm_ss.add(when: 'CONFIG_NSERIES', if_true: files('nseries.c')) arm_ss.add(when: 'CONFIG_SX1', if_true: files('omap_sx1.c')) -- 2.34.1 From: Strahinja Jankovic <strahinjapjankovic@gmail.com> During SPL boot several Clock Controller Module (CCM) registers are read, most important are PLL and Tuning, as well as divisor registers. This patch adds these registers and initializes reset values from user's guide. Signed-off-by: Strahinja Jankovic <strahinja.p.jankovic@gmail.com> Reviewed-by: Niek Linnenbank <nieklinnenbank@gmail.com> Message-id: 20221226220303.14420-2-strahinja.p.jankovic@gmail.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/allwinner-a10.h | 2 + include/hw/misc/allwinner-a10-ccm.h | 67 +++++++++ hw/arm/allwinner-a10.c | 7 + hw/misc/allwinner-a10-ccm.c | 224 ++++++++++++++++++++++++++++ hw/arm/Kconfig | 1 + hw/misc/Kconfig | 3 + hw/misc/meson.build | 1 + 7 files changed, 305 insertions(+) create mode 100644 include/hw/misc/allwinner-a10-ccm.h create mode 100644 hw/misc/allwinner-a10-ccm.c diff --git a/include/hw/arm/allwinner-a10.h b/include/hw/arm/allwinner-a10.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/allwinner-a10.h +++ b/include/hw/arm/allwinner-a10.h @@ -XXX,XX +XXX,XX @@ #include "hw/usb/hcd-ohci.h" #include "hw/usb/hcd-ehci.h" #include "hw/rtc/allwinner-rtc.h" +#include "hw/misc/allwinner-a10-ccm.h" #include "target/arm/cpu.h" #include "qom/object.h" @@ -XXX,XX +XXX,XX @@ struct AwA10State { /*< public >*/ ARMCPU cpu; + AwA10ClockCtlState ccm; AwA10PITState timer; AwA10PICState intc; AwEmacState emac; diff --git a/include/hw/misc/allwinner-a10-ccm.h b/include/hw/misc/allwinner-a10-ccm.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/include/hw/misc/allwinner-a10-ccm.h @@ -XXX,XX +XXX,XX @@ +/* + * Allwinner A10 Clock Control Module emulation + * + * Copyright (C) 2022 Strahinja Jankovic <strahinja.p.jankovic@gmail.com> + * + * This file is derived from Allwinner H3 CCU, + * by Niek Linnenbank. + * + * This program is free software: you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation, either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#ifndef HW_MISC_ALLWINNER_A10_CCM_H +#define HW_MISC_ALLWINNER_A10_CCM_H + +#include "qom/object.h" +#include "hw/sysbus.h" + +/** + * @name Constants + * @{ + */ + +/** Size of register I/O address space used by CCM device */ +#define AW_A10_CCM_IOSIZE (0x400) + +/** Total number of known registers */ +#define AW_A10_CCM_REGS_NUM (AW_A10_CCM_IOSIZE / sizeof(uint32_t)) + +/** @} */ + +/** + * @name Object model + * @{ + */ + +#define TYPE_AW_A10_CCM "allwinner-a10-ccm" +OBJECT_DECLARE_SIMPLE_TYPE(AwA10ClockCtlState, AW_A10_CCM) + +/** @} */ + +/** + * Allwinner A10 CCM object instance state. + */ +struct AwA10ClockCtlState { + /*< private >*/ + SysBusDevice parent_obj; + /*< public >*/ + + /** Maps I/O registers in physical memory */ + MemoryRegion iomem; + + /** Array of hardware registers */ + uint32_t regs[AW_A10_CCM_REGS_NUM]; +}; + +#endif /* HW_MISC_ALLWINNER_H3_CCU_H */ diff --git a/hw/arm/allwinner-a10.c b/hw/arm/allwinner-a10.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/allwinner-a10.c +++ b/hw/arm/allwinner-a10.c @@ -XXX,XX +XXX,XX @@ #include "hw/usb/hcd-ohci.h" #define AW_A10_MMC0_BASE 0x01c0f000 +#define AW_A10_CCM_BASE 0x01c20000 #define AW_A10_PIC_REG_BASE 0x01c20400 #define AW_A10_PIT_REG_BASE 0x01c20c00 #define AW_A10_UART0_REG_BASE 0x01c28000 @@ -XXX,XX +XXX,XX @@ static void aw_a10_init(Object *obj) object_initialize_child(obj, "timer", &s->timer, TYPE_AW_A10_PIT); + object_initialize_child(obj, "ccm", &s->ccm, TYPE_AW_A10_CCM); + object_initialize_child(obj, "emac", &s->emac, TYPE_AW_EMAC); object_initialize_child(obj, "sata", &s->sata, TYPE_ALLWINNER_AHCI); @@ -XXX,XX +XXX,XX @@ static void aw_a10_realize(DeviceState *dev, Error **errp) memory_region_add_subregion(get_system_memory(), 0x00000000, &s->sram_a); create_unimplemented_device("a10-sram-ctrl", 0x01c00000, 4 * KiB); + /* Clock Control Module */ + sysbus_realize(SYS_BUS_DEVICE(&s->ccm), &error_fatal); + sysbus_mmio_map(SYS_BUS_DEVICE(&s->ccm), 0, AW_A10_CCM_BASE); + /* FIXME use qdev NIC properties instead of nd_table[] */ if (nd_table[0].used) { qemu_check_nic_model(&nd_table[0], TYPE_AW_EMAC); diff --git a/hw/misc/allwinner-a10-ccm.c b/hw/misc/allwinner-a10-ccm.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/misc/allwinner-a10-ccm.c @@ -XXX,XX +XXX,XX @@ +/* + * Allwinner A10 Clock Control Module emulation + * + * Copyright (C) 2022 Strahinja Jankovic <strahinja.p.jankovic@gmail.com> + * + * This file is derived from Allwinner H3 CCU, + * by Niek Linnenbank. + * + * This program is free software: you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation, either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#include "qemu/osdep.h" +#include "qemu/units.h" +#include "hw/sysbus.h" +#include "migration/vmstate.h" +#include "qemu/log.h" +#include "qemu/module.h" +#include "hw/misc/allwinner-a10-ccm.h" + +/* CCM register offsets */ +enum { + REG_PLL1_CFG = 0x0000, /* PLL1 Control */ + REG_PLL1_TUN = 0x0004, /* PLL1 Tuning */ + REG_PLL2_CFG = 0x0008, /* PLL2 Control */ + REG_PLL2_TUN = 0x000C, /* PLL2 Tuning */ + REG_PLL3_CFG = 0x0010, /* PLL3 Control */ + REG_PLL4_CFG = 0x0018, /* PLL4 Control */ + REG_PLL5_CFG = 0x0020, /* PLL5 Control */ + REG_PLL5_TUN = 0x0024, /* PLL5 Tuning */ + REG_PLL6_CFG = 0x0028, /* PLL6 Control */ + REG_PLL6_TUN = 0x002C, /* PLL6 Tuning */ + REG_PLL7_CFG = 0x0030, /* PLL7 Control */ + REG_PLL1_TUN2 = 0x0038, /* PLL1 Tuning2 */ + REG_PLL5_TUN2 = 0x003C, /* PLL5 Tuning2 */ + REG_PLL8_CFG = 0x0040, /* PLL8 Control */ + REG_OSC24M_CFG = 0x0050, /* OSC24M Control */ + REG_CPU_AHB_APB0_CFG = 0x0054, /* CPU, AHB and APB0 Divide Ratio */ +}; + +#define REG_INDEX(offset) (offset / sizeof(uint32_t)) + +/* CCM register reset values */ +enum { + REG_PLL1_CFG_RST = 0x21005000, + REG_PLL1_TUN_RST = 0x0A101000, + REG_PLL2_CFG_RST = 0x08100010, + REG_PLL2_TUN_RST = 0x00000000, + REG_PLL3_CFG_RST = 0x0010D063, + REG_PLL4_CFG_RST = 0x21009911, + REG_PLL5_CFG_RST = 0x11049280, + REG_PLL5_TUN_RST = 0x14888000, + REG_PLL6_CFG_RST = 0x21009911, + REG_PLL6_TUN_RST = 0x00000000, + REG_PLL7_CFG_RST = 0x0010D063, + REG_PLL1_TUN2_RST = 0x00000000, + REG_PLL5_TUN2_RST = 0x00000000, + REG_PLL8_CFG_RST = 0x21009911, + REG_OSC24M_CFG_RST = 0x00138013, + REG_CPU_AHB_APB0_CFG_RST = 0x00010010, +}; + +static uint64_t allwinner_a10_ccm_read(void *opaque, hwaddr offset, + unsigned size) +{ + const AwA10ClockCtlState *s = AW_A10_CCM(opaque); + const uint32_t idx = REG_INDEX(offset); + + switch (offset) { + case REG_PLL1_CFG: + case REG_PLL1_TUN: + case REG_PLL2_CFG: + case REG_PLL2_TUN: + case REG_PLL3_CFG: + case REG_PLL4_CFG: + case REG_PLL5_CFG: + case REG_PLL5_TUN: + case REG_PLL6_CFG: + case REG_PLL6_TUN: + case REG_PLL7_CFG: + case REG_PLL1_TUN2: + case REG_PLL5_TUN2: + case REG_PLL8_CFG: + case REG_OSC24M_CFG: + case REG_CPU_AHB_APB0_CFG: + break; + case 0x158 ... AW_A10_CCM_IOSIZE: + qemu_log_mask(LOG_GUEST_ERROR, "%s: out-of-bounds offset 0x%04x\n", + __func__, (uint32_t)offset); + return 0; + default: + qemu_log_mask(LOG_UNIMP, "%s: unimplemented read offset 0x%04x\n", + __func__, (uint32_t)offset); + return 0; + } + + return s->regs[idx]; +} + +static void allwinner_a10_ccm_write(void *opaque, hwaddr offset, + uint64_t val, unsigned size) +{ + AwA10ClockCtlState *s = AW_A10_CCM(opaque); + const uint32_t idx = REG_INDEX(offset); + + switch (offset) { + case REG_PLL1_CFG: + case REG_PLL1_TUN: + case REG_PLL2_CFG: + case REG_PLL2_TUN: + case REG_PLL3_CFG: + case REG_PLL4_CFG: + case REG_PLL5_CFG: + case REG_PLL5_TUN: + case REG_PLL6_CFG: + case REG_PLL6_TUN: + case REG_PLL7_CFG: + case REG_PLL1_TUN2: + case REG_PLL5_TUN2: + case REG_PLL8_CFG: + case REG_OSC24M_CFG: + case REG_CPU_AHB_APB0_CFG: + break; + case 0x158 ... AW_A10_CCM_IOSIZE: + qemu_log_mask(LOG_GUEST_ERROR, "%s: out-of-bounds offset 0x%04x\n", + __func__, (uint32_t)offset); + break; + default: + qemu_log_mask(LOG_UNIMP, "%s: unimplemented write offset 0x%04x\n", + __func__, (uint32_t)offset); + break; + } + + s->regs[idx] = (uint32_t) val; +} + +static const MemoryRegionOps allwinner_a10_ccm_ops = { + .read = allwinner_a10_ccm_read, + .write = allwinner_a10_ccm_write, + .endianness = DEVICE_NATIVE_ENDIAN, + .valid = { + .min_access_size = 4, + .max_access_size = 4, + }, + .impl.min_access_size = 4, +}; + +static void allwinner_a10_ccm_reset_enter(Object *obj, ResetType type) +{ + AwA10ClockCtlState *s = AW_A10_CCM(obj); + + /* Set default values for registers */ + s->regs[REG_INDEX(REG_PLL1_CFG)] = REG_PLL1_CFG_RST; + s->regs[REG_INDEX(REG_PLL1_TUN)] = REG_PLL1_TUN_RST; + s->regs[REG_INDEX(REG_PLL2_CFG)] = REG_PLL2_CFG_RST; + s->regs[REG_INDEX(REG_PLL2_TUN)] = REG_PLL2_TUN_RST; + s->regs[REG_INDEX(REG_PLL3_CFG)] = REG_PLL3_CFG_RST; + s->regs[REG_INDEX(REG_PLL4_CFG)] = REG_PLL4_CFG_RST; + s->regs[REG_INDEX(REG_PLL5_CFG)] = REG_PLL5_CFG_RST; + s->regs[REG_INDEX(REG_PLL5_TUN)] = REG_PLL5_TUN_RST; + s->regs[REG_INDEX(REG_PLL6_CFG)] = REG_PLL6_CFG_RST; + s->regs[REG_INDEX(REG_PLL6_TUN)] = REG_PLL6_TUN_RST; + s->regs[REG_INDEX(REG_PLL7_CFG)] = REG_PLL7_CFG_RST; + s->regs[REG_INDEX(REG_PLL1_TUN2)] = REG_PLL1_TUN2_RST; + s->regs[REG_INDEX(REG_PLL5_TUN2)] = REG_PLL5_TUN2_RST; + s->regs[REG_INDEX(REG_PLL8_CFG)] = REG_PLL8_CFG_RST; + s->regs[REG_INDEX(REG_OSC24M_CFG)] = REG_OSC24M_CFG_RST; + s->regs[REG_INDEX(REG_CPU_AHB_APB0_CFG)] = REG_CPU_AHB_APB0_CFG_RST; +} + +static void allwinner_a10_ccm_init(Object *obj) +{ + SysBusDevice *sbd = SYS_BUS_DEVICE(obj); + AwA10ClockCtlState *s = AW_A10_CCM(obj); + + /* Memory mapping */ + memory_region_init_io(&s->iomem, OBJECT(s), &allwinner_a10_ccm_ops, s, + TYPE_AW_A10_CCM, AW_A10_CCM_IOSIZE); + sysbus_init_mmio(sbd, &s->iomem); +} + +static const VMStateDescription allwinner_a10_ccm_vmstate = { + .name = "allwinner-a10-ccm", + .version_id = 1, + .minimum_version_id = 1, + .fields = (VMStateField[]) { + VMSTATE_UINT32_ARRAY(regs, AwA10ClockCtlState, AW_A10_CCM_REGS_NUM), + VMSTATE_END_OF_LIST() + } +}; + +static void allwinner_a10_ccm_class_init(ObjectClass *klass, void *data) +{ + DeviceClass *dc = DEVICE_CLASS(klass); + ResettableClass *rc = RESETTABLE_CLASS(klass); + + rc->phases.enter = allwinner_a10_ccm_reset_enter; + dc->vmsd = &allwinner_a10_ccm_vmstate; +} + +static const TypeInfo allwinner_a10_ccm_info = { + .name = TYPE_AW_A10_CCM, + .parent = TYPE_SYS_BUS_DEVICE, + .instance_init = allwinner_a10_ccm_init, + .instance_size = sizeof(AwA10ClockCtlState), + .class_init = allwinner_a10_ccm_class_init, +}; + +static void allwinner_a10_ccm_register(void) +{ + type_register_static(&allwinner_a10_ccm_info); +} + +type_init(allwinner_a10_ccm_register) diff --git a/hw/arm/Kconfig b/hw/arm/Kconfig index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/Kconfig +++ b/hw/arm/Kconfig @@ -XXX,XX +XXX,XX @@ config ALLWINNER_A10 select AHCI select ALLWINNER_A10_PIT select ALLWINNER_A10_PIC + select ALLWINNER_A10_CCM select ALLWINNER_EMAC select SERIAL select UNIMP diff --git a/hw/misc/Kconfig b/hw/misc/Kconfig index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/Kconfig +++ b/hw/misc/Kconfig @@ -XXX,XX +XXX,XX @@ config VIRT_CTRL config LASI bool +config ALLWINNER_A10_CCM + bool + source macio/Kconfig diff --git a/hw/misc/meson.build b/hw/misc/meson.build index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/meson.build +++ b/hw/misc/meson.build @@ -XXX,XX +XXX,XX @@ subdir('macio') softmmu_ss.add(when: 'CONFIG_IVSHMEM_DEVICE', if_true: files('ivshmem.c')) +softmmu_ss.add(when: 'CONFIG_ALLWINNER_A10_CCM', if_true: files('allwinner-a10-ccm.c')) softmmu_ss.add(when: 'CONFIG_ALLWINNER_H3', if_true: files('allwinner-h3-ccu.c')) specific_ss.add(when: 'CONFIG_ALLWINNER_H3', if_true: files('allwinner-cpucfg.c')) softmmu_ss.add(when: 'CONFIG_ALLWINNER_H3', if_true: files('allwinner-h3-dramc.c')) -- 2.34.1 From: Strahinja Jankovic <strahinjapjankovic@gmail.com> During SPL boot several DRAM Controller registers are used. Most important registers are those related to DRAM initialization and calibration, where SPL initiates process and waits until certain bit is set/cleared. This patch adds these registers, initializes reset values from user's guide and updates state of registers as SPL expects it. Signed-off-by: Strahinja Jankovic <strahinja.p.jankovic@gmail.com> Reviewed-by: Niek Linnenbank <nieklinnenbank@gmail.com> Message-id: 20221226220303.14420-3-strahinja.p.jankovic@gmail.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/allwinner-a10.h | 2 + include/hw/misc/allwinner-a10-dramc.h | 68 ++++++++++ hw/arm/allwinner-a10.c | 7 + hw/misc/allwinner-a10-dramc.c | 179 ++++++++++++++++++++++++++ hw/arm/Kconfig | 1 + hw/misc/Kconfig | 3 + hw/misc/meson.build | 1 + 7 files changed, 261 insertions(+) create mode 100644 include/hw/misc/allwinner-a10-dramc.h create mode 100644 hw/misc/allwinner-a10-dramc.c diff --git a/include/hw/arm/allwinner-a10.h b/include/hw/arm/allwinner-a10.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/allwinner-a10.h +++ b/include/hw/arm/allwinner-a10.h @@ -XXX,XX +XXX,XX @@ #include "hw/usb/hcd-ehci.h" #include "hw/rtc/allwinner-rtc.h" #include "hw/misc/allwinner-a10-ccm.h" +#include "hw/misc/allwinner-a10-dramc.h" #include "target/arm/cpu.h" #include "qom/object.h" @@ -XXX,XX +XXX,XX @@ struct AwA10State { ARMCPU cpu; AwA10ClockCtlState ccm; + AwA10DramControllerState dramc; AwA10PITState timer; AwA10PICState intc; AwEmacState emac; diff --git a/include/hw/misc/allwinner-a10-dramc.h b/include/hw/misc/allwinner-a10-dramc.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/include/hw/misc/allwinner-a10-dramc.h @@ -XXX,XX +XXX,XX @@ +/* + * Allwinner A10 DRAM Controller emulation + * + * Copyright (C) 2022 Strahinja Jankovic <strahinja.p.jankovic@gmail.com> + * + * This file is derived from Allwinner H3 DRAMC, + * by Niek Linnenbank. + * + * This program is free software: you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation, either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#ifndef HW_MISC_ALLWINNER_A10_DRAMC_H +#define HW_MISC_ALLWINNER_A10_DRAMC_H + +#include "qom/object.h" +#include "hw/sysbus.h" +#include "hw/register.h" + +/** + * @name Constants + * @{ + */ + +/** Size of register I/O address space used by DRAMC device */ +#define AW_A10_DRAMC_IOSIZE (0x1000) + +/** Total number of known registers */ +#define AW_A10_DRAMC_REGS_NUM (AW_A10_DRAMC_IOSIZE / sizeof(uint32_t)) + +/** @} */ + +/** + * @name Object model + * @{ + */ + +#define TYPE_AW_A10_DRAMC "allwinner-a10-dramc" +OBJECT_DECLARE_SIMPLE_TYPE(AwA10DramControllerState, AW_A10_DRAMC) + +/** @} */ + +/** + * Allwinner A10 DRAMC object instance state. + */ +struct AwA10DramControllerState { + /*< private >*/ + SysBusDevice parent_obj; + /*< public >*/ + + /** Maps I/O registers in physical memory */ + MemoryRegion iomem; + + /** Array of hardware registers */ + uint32_t regs[AW_A10_DRAMC_REGS_NUM]; +}; + +#endif /* HW_MISC_ALLWINNER_A10_DRAMC_H */ diff --git a/hw/arm/allwinner-a10.c b/hw/arm/allwinner-a10.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/allwinner-a10.c +++ b/hw/arm/allwinner-a10.c @@ -XXX,XX +XXX,XX @@ #include "hw/boards.h" #include "hw/usb/hcd-ohci.h" +#define AW_A10_DRAMC_BASE 0x01c01000 #define AW_A10_MMC0_BASE 0x01c0f000 #define AW_A10_CCM_BASE 0x01c20000 #define AW_A10_PIC_REG_BASE 0x01c20400 @@ -XXX,XX +XXX,XX @@ static void aw_a10_init(Object *obj) object_initialize_child(obj, "ccm", &s->ccm, TYPE_AW_A10_CCM); + object_initialize_child(obj, "dramc", &s->dramc, TYPE_AW_A10_DRAMC); + object_initialize_child(obj, "emac", &s->emac, TYPE_AW_EMAC); object_initialize_child(obj, "sata", &s->sata, TYPE_ALLWINNER_AHCI); @@ -XXX,XX +XXX,XX @@ static void aw_a10_realize(DeviceState *dev, Error **errp) sysbus_realize(SYS_BUS_DEVICE(&s->ccm), &error_fatal); sysbus_mmio_map(SYS_BUS_DEVICE(&s->ccm), 0, AW_A10_CCM_BASE); + /* DRAM Control Module */ + sysbus_realize(SYS_BUS_DEVICE(&s->dramc), &error_fatal); + sysbus_mmio_map(SYS_BUS_DEVICE(&s->dramc), 0, AW_A10_DRAMC_BASE); + /* FIXME use qdev NIC properties instead of nd_table[] */ if (nd_table[0].used) { qemu_check_nic_model(&nd_table[0], TYPE_AW_EMAC); diff --git a/hw/misc/allwinner-a10-dramc.c b/hw/misc/allwinner-a10-dramc.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/misc/allwinner-a10-dramc.c @@ -XXX,XX +XXX,XX @@ +/* + * Allwinner A10 DRAM Controller emulation + * + * Copyright (C) 2022 Strahinja Jankovic <strahinja.p.jankovic@gmail.com> + * + * This file is derived from Allwinner H3 DRAMC, + * by Niek Linnenbank. + * + * This program is free software: you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation, either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#include "qemu/osdep.h" +#include "qemu/units.h" +#include "hw/sysbus.h" +#include "migration/vmstate.h" +#include "qemu/log.h" +#include "qemu/module.h" +#include "hw/misc/allwinner-a10-dramc.h" + +/* DRAMC register offsets */ +enum { + REG_SDR_CCR = 0x0000, + REG_SDR_ZQCR0 = 0x00a8, + REG_SDR_ZQSR = 0x00b0 +}; + +#define REG_INDEX(offset) (offset / sizeof(uint32_t)) + +/* DRAMC register flags */ +enum { + REG_SDR_CCR_DATA_TRAINING = (1 << 30), + REG_SDR_CCR_DRAM_INIT = (1 << 31), +}; +enum { + REG_SDR_ZQSR_ZCAL = (1 << 31), +}; + +/* DRAMC register reset values */ +enum { + REG_SDR_CCR_RESET = 0x80020000, + REG_SDR_ZQCR0_RESET = 0x07b00000, + REG_SDR_ZQSR_RESET = 0x80000000 +}; + +static uint64_t allwinner_a10_dramc_read(void *opaque, hwaddr offset, + unsigned size) +{ + const AwA10DramControllerState *s = AW_A10_DRAMC(opaque); + const uint32_t idx = REG_INDEX(offset); + + switch (offset) { + case REG_SDR_CCR: + case REG_SDR_ZQCR0: + case REG_SDR_ZQSR: + break; + case 0x2e4 ... AW_A10_DRAMC_IOSIZE: + qemu_log_mask(LOG_GUEST_ERROR, "%s: out-of-bounds offset 0x%04x\n", + __func__, (uint32_t)offset); + return 0; + default: + qemu_log_mask(LOG_UNIMP, "%s: unimplemented read offset 0x%04x\n", + __func__, (uint32_t)offset); + return 0; + } + + return s->regs[idx]; +} + +static void allwinner_a10_dramc_write(void *opaque, hwaddr offset, + uint64_t val, unsigned size) +{ + AwA10DramControllerState *s = AW_A10_DRAMC(opaque); + const uint32_t idx = REG_INDEX(offset); + + switch (offset) { + case REG_SDR_CCR: + if (val & REG_SDR_CCR_DRAM_INIT) { + /* Clear DRAM_INIT to indicate process is done. */ + val &= ~REG_SDR_CCR_DRAM_INIT; + } + if (val & REG_SDR_CCR_DATA_TRAINING) { + /* Clear DATA_TRAINING to indicate process is done. */ + val &= ~REG_SDR_CCR_DATA_TRAINING; + } + break; + case REG_SDR_ZQCR0: + /* Set ZCAL in ZQSR to indicate calibration is done. */ + s->regs[REG_INDEX(REG_SDR_ZQSR)] |= REG_SDR_ZQSR_ZCAL; + break; + case 0x2e4 ... AW_A10_DRAMC_IOSIZE: + qemu_log_mask(LOG_GUEST_ERROR, "%s: out-of-bounds offset 0x%04x\n", + __func__, (uint32_t)offset); + break; + default: + qemu_log_mask(LOG_UNIMP, "%s: unimplemented write offset 0x%04x\n", + __func__, (uint32_t)offset); + break; + } + + s->regs[idx] = (uint32_t) val; +} + +static const MemoryRegionOps allwinner_a10_dramc_ops = { + .read = allwinner_a10_dramc_read, + .write = allwinner_a10_dramc_write, + .endianness = DEVICE_NATIVE_ENDIAN, + .valid = { + .min_access_size = 4, + .max_access_size = 4, + }, + .impl.min_access_size = 4, +}; + +static void allwinner_a10_dramc_reset_enter(Object *obj, ResetType type) +{ + AwA10DramControllerState *s = AW_A10_DRAMC(obj); + + /* Set default values for registers */ + s->regs[REG_INDEX(REG_SDR_CCR)] = REG_SDR_CCR_RESET; + s->regs[REG_INDEX(REG_SDR_ZQCR0)] = REG_SDR_ZQCR0_RESET; + s->regs[REG_INDEX(REG_SDR_ZQSR)] = REG_SDR_ZQSR_RESET; +} + +static void allwinner_a10_dramc_init(Object *obj) +{ + SysBusDevice *sbd = SYS_BUS_DEVICE(obj); + AwA10DramControllerState *s = AW_A10_DRAMC(obj); + + /* Memory mapping */ + memory_region_init_io(&s->iomem, OBJECT(s), &allwinner_a10_dramc_ops, s, + TYPE_AW_A10_DRAMC, AW_A10_DRAMC_IOSIZE); + sysbus_init_mmio(sbd, &s->iomem); +} + +static const VMStateDescription allwinner_a10_dramc_vmstate = { + .name = "allwinner-a10-dramc", + .version_id = 1, + .minimum_version_id = 1, + .fields = (VMStateField[]) { + VMSTATE_UINT32_ARRAY(regs, AwA10DramControllerState, + AW_A10_DRAMC_REGS_NUM), + VMSTATE_END_OF_LIST() + } +}; + +static void allwinner_a10_dramc_class_init(ObjectClass *klass, void *data) +{ + DeviceClass *dc = DEVICE_CLASS(klass); + ResettableClass *rc = RESETTABLE_CLASS(klass); + + rc->phases.enter = allwinner_a10_dramc_reset_enter; + dc->vmsd = &allwinner_a10_dramc_vmstate; +} + +static const TypeInfo allwinner_a10_dramc_info = { + .name = TYPE_AW_A10_DRAMC, + .parent = TYPE_SYS_BUS_DEVICE, + .instance_init = allwinner_a10_dramc_init, + .instance_size = sizeof(AwA10DramControllerState), + .class_init = allwinner_a10_dramc_class_init, +}; + +static void allwinner_a10_dramc_register(void) +{ + type_register_static(&allwinner_a10_dramc_info); +} + +type_init(allwinner_a10_dramc_register) diff --git a/hw/arm/Kconfig b/hw/arm/Kconfig index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/Kconfig +++ b/hw/arm/Kconfig @@ -XXX,XX +XXX,XX @@ config ALLWINNER_A10 select ALLWINNER_A10_PIT select ALLWINNER_A10_PIC select ALLWINNER_A10_CCM + select ALLWINNER_A10_DRAMC select ALLWINNER_EMAC select SERIAL select UNIMP diff --git a/hw/misc/Kconfig b/hw/misc/Kconfig index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/Kconfig +++ b/hw/misc/Kconfig @@ -XXX,XX +XXX,XX @@ config LASI config ALLWINNER_A10_CCM bool +config ALLWINNER_A10_DRAMC + bool + source macio/Kconfig diff --git a/hw/misc/meson.build b/hw/misc/meson.build index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/meson.build +++ b/hw/misc/meson.build @@ -XXX,XX +XXX,XX @@ subdir('macio') softmmu_ss.add(when: 'CONFIG_IVSHMEM_DEVICE', if_true: files('ivshmem.c')) softmmu_ss.add(when: 'CONFIG_ALLWINNER_A10_CCM', if_true: files('allwinner-a10-ccm.c')) +softmmu_ss.add(when: 'CONFIG_ALLWINNER_A10_DRAMC', if_true: files('allwinner-a10-dramc.c')) softmmu_ss.add(when: 'CONFIG_ALLWINNER_H3', if_true: files('allwinner-h3-ccu.c')) specific_ss.add(when: 'CONFIG_ALLWINNER_H3', if_true: files('allwinner-cpucfg.c')) softmmu_ss.add(when: 'CONFIG_ALLWINNER_H3', if_true: files('allwinner-h3-dramc.c')) -- 2.34.1 From: Strahinja Jankovic <strahinjapjankovic@gmail.com> This patch implements Allwinner TWI/I2C controller emulation. Only master-mode functionality is implemented. The SPL boot for Cubieboard expects AXP209 PMIC on TWI0/I2C0 bus, so this is first part enabling the TWI/I2C bus operation. Since both Allwinner A10 and H3 use the same module, it is added for both boards. Docs are also updated for Cubieboard and Orangepi-PC board to indicate I2C availability. Signed-off-by: Strahinja Jankovic <strahinja.p.jankovic@gmail.com> Reviewed-by: Niek Linnenbank <nieklinnenbank@gmail.com> Message-id: 20221226220303.14420-4-strahinja.p.jankovic@gmail.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- docs/system/arm/cubieboard.rst | 1 + docs/system/arm/orangepi.rst | 1 + include/hw/arm/allwinner-a10.h | 2 + include/hw/arm/allwinner-h3.h | 3 + include/hw/i2c/allwinner-i2c.h | 55 ++++ hw/arm/allwinner-a10.c | 8 + hw/arm/allwinner-h3.c | 11 +- hw/i2c/allwinner-i2c.c | 459 +++++++++++++++++++++++++++++++++ hw/arm/Kconfig | 2 + hw/i2c/Kconfig | 4 + hw/i2c/meson.build | 1 + hw/i2c/trace-events | 5 + 12 files changed, 551 insertions(+), 1 deletion(-) create mode 100644 include/hw/i2c/allwinner-i2c.h create mode 100644 hw/i2c/allwinner-i2c.c diff --git a/docs/system/arm/cubieboard.rst b/docs/system/arm/cubieboard.rst index XXXXXXX..XXXXXXX 100644 --- a/docs/system/arm/cubieboard.rst +++ b/docs/system/arm/cubieboard.rst @@ -XXX,XX +XXX,XX @@ Emulated devices: - SDHCI - USB controller - SATA controller +- TWI (I2C) controller diff --git a/docs/system/arm/orangepi.rst b/docs/system/arm/orangepi.rst index XXXXXXX..XXXXXXX 100644 --- a/docs/system/arm/orangepi.rst +++ b/docs/system/arm/orangepi.rst @@ -XXX,XX +XXX,XX @@ The Orange Pi PC machine supports the following devices: * Clock Control Unit * System Control module * Security Identifier device + * TWI (I2C) Limitations """"""""""" diff --git a/include/hw/arm/allwinner-a10.h b/include/hw/arm/allwinner-a10.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/allwinner-a10.h +++ b/include/hw/arm/allwinner-a10.h @@ -XXX,XX +XXX,XX @@ #include "hw/rtc/allwinner-rtc.h" #include "hw/misc/allwinner-a10-ccm.h" #include "hw/misc/allwinner-a10-dramc.h" +#include "hw/i2c/allwinner-i2c.h" #include "target/arm/cpu.h" #include "qom/object.h" @@ -XXX,XX +XXX,XX @@ struct AwA10State { AwEmacState emac; AllwinnerAHCIState sata; AwSdHostState mmc0; + AWI2CState i2c0; AwRtcState rtc; MemoryRegion sram_a; EHCISysBusState ehci[AW_A10_NUM_USB]; diff --git a/include/hw/arm/allwinner-h3.h b/include/hw/arm/allwinner-h3.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/allwinner-h3.h +++ b/include/hw/arm/allwinner-h3.h @@ -XXX,XX +XXX,XX @@ #include "hw/sd/allwinner-sdhost.h" #include "hw/net/allwinner-sun8i-emac.h" #include "hw/rtc/allwinner-rtc.h" +#include "hw/i2c/allwinner-i2c.h" #include "target/arm/cpu.h" #include "sysemu/block-backend.h" @@ -XXX,XX +XXX,XX @@ enum { AW_H3_DEV_UART2, AW_H3_DEV_UART3, AW_H3_DEV_EMAC, + AW_H3_DEV_TWI0, AW_H3_DEV_DRAMCOM, AW_H3_DEV_DRAMCTL, AW_H3_DEV_DRAMPHY, @@ -XXX,XX +XXX,XX @@ struct AwH3State { AwH3SysCtrlState sysctrl; AwSidState sid; AwSdHostState mmc0; + AWI2CState i2c0; AwSun8iEmacState emac; AwRtcState rtc; GICState gic; diff --git a/include/hw/i2c/allwinner-i2c.h b/include/hw/i2c/allwinner-i2c.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/include/hw/i2c/allwinner-i2c.h @@ -XXX,XX +XXX,XX @@ +/* + * Allwinner I2C Bus Serial Interface registers definition + * + * Copyright (C) 2022 Strahinja Jankovic. <strahinja.p.jankovic@gmail.com> + * + * This file is derived from IMX I2C controller, + * by Jean-Christophe DUBOIS . + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License + * for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, see <http://www.gnu.org/licenses/>. + * + */ + +#ifndef ALLWINNER_I2C_H +#define ALLWINNER_I2C_H + +#include "hw/sysbus.h" +#include "qom/object.h" + +#define TYPE_AW_I2C "allwinner.i2c" +OBJECT_DECLARE_SIMPLE_TYPE(AWI2CState, AW_I2C) + +#define AW_I2C_MEM_SIZE 0x24 + +struct AWI2CState { + /*< private >*/ + SysBusDevice parent_obj; + + /*< public >*/ + MemoryRegion iomem; + I2CBus *bus; + qemu_irq irq; + + uint8_t addr; + uint8_t xaddr; + uint8_t data; + uint8_t cntr; + uint8_t stat; + uint8_t ccr; + uint8_t srst; + uint8_t efr; + uint8_t lcr; +}; + +#endif /* ALLWINNER_I2C_H */ diff --git a/hw/arm/allwinner-a10.c b/hw/arm/allwinner-a10.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/allwinner-a10.c +++ b/hw/arm/allwinner-a10.c @@ -XXX,XX +XXX,XX @@ #define AW_A10_OHCI_BASE 0x01c14400 #define AW_A10_SATA_BASE 0x01c18000 #define AW_A10_RTC_BASE 0x01c20d00 +#define AW_A10_I2C0_BASE 0x01c2ac00 static void aw_a10_init(Object *obj) { @@ -XXX,XX +XXX,XX @@ static void aw_a10_init(Object *obj) object_initialize_child(obj, "sata", &s->sata, TYPE_ALLWINNER_AHCI); + object_initialize_child(obj, "i2c0", &s->i2c0, TYPE_AW_I2C); + if (machine_usb(current_machine)) { int i; @@ -XXX,XX +XXX,XX @@ static void aw_a10_realize(DeviceState *dev, Error **errp) /* RTC */ sysbus_realize(SYS_BUS_DEVICE(&s->rtc), &error_fatal); sysbus_mmio_map_overlap(SYS_BUS_DEVICE(&s->rtc), 0, AW_A10_RTC_BASE, 10); + + /* I2C */ + sysbus_realize(SYS_BUS_DEVICE(&s->i2c0), &error_fatal); + sysbus_mmio_map(SYS_BUS_DEVICE(&s->i2c0), 0, AW_A10_I2C0_BASE); + sysbus_connect_irq(SYS_BUS_DEVICE(&s->i2c0), 0, qdev_get_gpio_in(dev, 7)); } static void aw_a10_class_init(ObjectClass *oc, void *data) diff --git a/hw/arm/allwinner-h3.c b/hw/arm/allwinner-h3.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/allwinner-h3.c +++ b/hw/arm/allwinner-h3.c @@ -XXX,XX +XXX,XX @@ const hwaddr allwinner_h3_memmap[] = { [AW_H3_DEV_UART1] = 0x01c28400, [AW_H3_DEV_UART2] = 0x01c28800, [AW_H3_DEV_UART3] = 0x01c28c00, + [AW_H3_DEV_TWI0] = 0x01c2ac00, [AW_H3_DEV_EMAC] = 0x01c30000, [AW_H3_DEV_DRAMCOM] = 0x01c62000, [AW_H3_DEV_DRAMCTL] = 0x01c63000, @@ -XXX,XX +XXX,XX @@ struct AwH3Unimplemented { { "uart1", 0x01c28400, 1 * KiB }, { "uart2", 0x01c28800, 1 * KiB }, { "uart3", 0x01c28c00, 1 * KiB }, - { "twi0", 0x01c2ac00, 1 * KiB }, { "twi1", 0x01c2b000, 1 * KiB }, { "twi2", 0x01c2b400, 1 * KiB }, { "scr", 0x01c2c400, 1 * KiB }, @@ -XXX,XX +XXX,XX @@ enum { AW_H3_GIC_SPI_UART1 = 1, AW_H3_GIC_SPI_UART2 = 2, AW_H3_GIC_SPI_UART3 = 3, + AW_H3_GIC_SPI_TWI0 = 6, AW_H3_GIC_SPI_TIMER0 = 18, AW_H3_GIC_SPI_TIMER1 = 19, AW_H3_GIC_SPI_MMC0 = 60, @@ -XXX,XX +XXX,XX @@ static void allwinner_h3_init(Object *obj) "ram-size"); object_initialize_child(obj, "rtc", &s->rtc, TYPE_AW_RTC_SUN6I); + + object_initialize_child(obj, "twi0", &s->i2c0, TYPE_AW_I2C); } static void allwinner_h3_realize(DeviceState *dev, Error **errp) @@ -XXX,XX +XXX,XX @@ static void allwinner_h3_realize(DeviceState *dev, Error **errp) sysbus_realize(SYS_BUS_DEVICE(&s->rtc), &error_fatal); sysbus_mmio_map(SYS_BUS_DEVICE(&s->rtc), 0, s->memmap[AW_H3_DEV_RTC]); + /* I2C */ + sysbus_realize(SYS_BUS_DEVICE(&s->i2c0), &error_fatal); + sysbus_mmio_map(SYS_BUS_DEVICE(&s->i2c0), 0, s->memmap[AW_H3_DEV_TWI0]); + sysbus_connect_irq(SYS_BUS_DEVICE(&s->i2c0), 0, + qdev_get_gpio_in(DEVICE(&s->gic), AW_H3_GIC_SPI_TWI0)); + /* Unimplemented devices */ for (i = 0; i < ARRAY_SIZE(unimplemented); i++) { create_unimplemented_device(unimplemented[i].device_name, diff --git a/hw/i2c/allwinner-i2c.c b/hw/i2c/allwinner-i2c.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/i2c/allwinner-i2c.c @@ -XXX,XX +XXX,XX @@ +/* + * Allwinner I2C Bus Serial Interface Emulation + * + * Copyright (C) 2022 Strahinja Jankovic <strahinja.p.jankovic@gmail.com> + * + * This file is derived from IMX I2C controller, + * by Jean-Christophe DUBOIS . + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License + * for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, see <http://www.gnu.org/licenses/>. + * + * SPDX-License-Identifier: MIT + */ + +#include "qemu/osdep.h" +#include "hw/i2c/allwinner-i2c.h" +#include "hw/irq.h" +#include "migration/vmstate.h" +#include "hw/i2c/i2c.h" +#include "qemu/log.h" +#include "trace.h" +#include "qemu/module.h" + +/* Allwinner I2C memory map */ +#define TWI_ADDR_REG 0x00 /* slave address register */ +#define TWI_XADDR_REG 0x04 /* extended slave address register */ +#define TWI_DATA_REG 0x08 /* data register */ +#define TWI_CNTR_REG 0x0c /* control register */ +#define TWI_STAT_REG 0x10 /* status register */ +#define TWI_CCR_REG 0x14 /* clock control register */ +#define TWI_SRST_REG 0x18 /* software reset register */ +#define TWI_EFR_REG 0x1c /* enhance feature register */ +#define TWI_LCR_REG 0x20 /* line control register */ + +/* Used only in slave mode, do not set */ +#define TWI_ADDR_RESET 0 +#define TWI_XADDR_RESET 0 + +/* Data register */ +#define TWI_DATA_MASK 0xFF +#define TWI_DATA_RESET 0 + +/* Control register */ +#define TWI_CNTR_INT_EN (1 << 7) +#define TWI_CNTR_BUS_EN (1 << 6) +#define TWI_CNTR_M_STA (1 << 5) +#define TWI_CNTR_M_STP (1 << 4) +#define TWI_CNTR_INT_FLAG (1 << 3) +#define TWI_CNTR_A_ACK (1 << 2) +#define TWI_CNTR_MASK 0xFC +#define TWI_CNTR_RESET 0 + +/* Status register */ +#define TWI_STAT_MASK 0xF8 +#define TWI_STAT_RESET 0xF8 + +/* Clock register */ +#define TWI_CCR_CLK_M_MASK 0x78 +#define TWI_CCR_CLK_N_MASK 0x07 +#define TWI_CCR_MASK 0x7F +#define TWI_CCR_RESET 0 + +/* Soft reset */ +#define TWI_SRST_MASK 0x01 +#define TWI_SRST_RESET 0 + +/* Enhance feature */ +#define TWI_EFR_MASK 0x03 +#define TWI_EFR_RESET 0 + +/* Line control */ +#define TWI_LCR_SCL_STATE (1 << 5) +#define TWI_LCR_SDA_STATE (1 << 4) +#define TWI_LCR_SCL_CTL (1 << 3) +#define TWI_LCR_SCL_CTL_EN (1 << 2) +#define TWI_LCR_SDA_CTL (1 << 1) +#define TWI_LCR_SDA_CTL_EN (1 << 0) +#define TWI_LCR_MASK 0x3F +#define TWI_LCR_RESET 0x3A + +/* Status value in STAT register is shifted by 3 bits */ +#define TWI_STAT_SHIFT 3 +#define STAT_FROM_STA(x) ((x) << TWI_STAT_SHIFT) +#define STAT_TO_STA(x) ((x) >> TWI_STAT_SHIFT) + +enum { + STAT_BUS_ERROR = 0, + /* Master mode */ + STAT_M_STA_TX, + STAT_M_RSTA_TX, + STAT_M_ADDR_WR_ACK, + STAT_M_ADDR_WR_NACK, + STAT_M_DATA_TX_ACK, + STAT_M_DATA_TX_NACK, + STAT_M_ARB_LOST, + STAT_M_ADDR_RD_ACK, + STAT_M_ADDR_RD_NACK, + STAT_M_DATA_RX_ACK, + STAT_M_DATA_RX_NACK, + /* Slave mode */ + STAT_S_ADDR_WR_ACK, + STAT_S_ARB_LOST_AW_ACK, + STAT_S_GCA_ACK, + STAT_S_ARB_LOST_GCA_ACK, + STAT_S_DATA_RX_SA_ACK, + STAT_S_DATA_RX_SA_NACK, + STAT_S_DATA_RX_GCA_ACK, + STAT_S_DATA_RX_GCA_NACK, + STAT_S_STP_RSTA, + STAT_S_ADDR_RD_ACK, + STAT_S_ARB_LOST_AR_ACK, + STAT_S_DATA_TX_ACK, + STAT_S_DATA_TX_NACK, + STAT_S_LB_TX_ACK, + /* Master mode, 10-bit */ + STAT_M_2ND_ADDR_WR_ACK, + STAT_M_2ND_ADDR_WR_NACK, + /* Idle */ + STAT_IDLE = 0x1f +} TWI_STAT_STA; + +static const char *allwinner_i2c_get_regname(unsigned offset) +{ + switch (offset) { + case TWI_ADDR_REG: + return "ADDR"; + case TWI_XADDR_REG: + return "XADDR"; + case TWI_DATA_REG: + return "DATA"; + case TWI_CNTR_REG: + return "CNTR"; + case TWI_STAT_REG: + return "STAT"; + case TWI_CCR_REG: + return "CCR"; + case TWI_SRST_REG: + return "SRST"; + case TWI_EFR_REG: + return "EFR"; + case TWI_LCR_REG: + return "LCR"; + default: + return "[?]"; + } +} + +static inline bool allwinner_i2c_is_reset(AWI2CState *s) +{ + return s->srst & TWI_SRST_MASK; +} + +static inline bool allwinner_i2c_bus_is_enabled(AWI2CState *s) +{ + return s->cntr & TWI_CNTR_BUS_EN; +} + +static inline bool allwinner_i2c_interrupt_is_enabled(AWI2CState *s) +{ + return s->cntr & TWI_CNTR_INT_EN; +} + +static void allwinner_i2c_reset_hold(Object *obj) +{ + AWI2CState *s = AW_I2C(obj); + + if (STAT_TO_STA(s->stat) != STAT_IDLE) { + i2c_end_transfer(s->bus); + } + + s->addr = TWI_ADDR_RESET; + s->xaddr = TWI_XADDR_RESET; + s->data = TWI_DATA_RESET; + s->cntr = TWI_CNTR_RESET; + s->stat = TWI_STAT_RESET; + s->ccr = TWI_CCR_RESET; + s->srst = TWI_SRST_RESET; + s->efr = TWI_EFR_RESET; + s->lcr = TWI_LCR_RESET; +} + +static inline void allwinner_i2c_raise_interrupt(AWI2CState *s) +{ + /* + * Raise an interrupt if the device is not reset and it is configured + * to generate some interrupts. + */ + if (!allwinner_i2c_is_reset(s) && allwinner_i2c_bus_is_enabled(s)) { + if (STAT_TO_STA(s->stat) != STAT_IDLE) { + s->cntr |= TWI_CNTR_INT_FLAG; + if (allwinner_i2c_interrupt_is_enabled(s)) { + qemu_irq_raise(s->irq); + } + } + } +} + +static uint64_t allwinner_i2c_read(void *opaque, hwaddr offset, + unsigned size) +{ + uint16_t value; + AWI2CState *s = AW_I2C(opaque); + + switch (offset) { + case TWI_ADDR_REG: + value = s->addr; + break; + case TWI_XADDR_REG: + value = s->xaddr; + break; + case TWI_DATA_REG: + if ((STAT_TO_STA(s->stat) == STAT_M_ADDR_RD_ACK) || + (STAT_TO_STA(s->stat) == STAT_M_DATA_RX_ACK) || + (STAT_TO_STA(s->stat) == STAT_M_DATA_RX_NACK)) { + /* Get the next byte */ + s->data = i2c_recv(s->bus); + + if (s->cntr & TWI_CNTR_A_ACK) { + s->stat = STAT_FROM_STA(STAT_M_DATA_RX_ACK); + } else { + s->stat = STAT_FROM_STA(STAT_M_DATA_RX_NACK); + } + allwinner_i2c_raise_interrupt(s); + } + value = s->data; + break; + case TWI_CNTR_REG: + value = s->cntr; + break; + case TWI_STAT_REG: + value = s->stat; + /* + * If polling when reading then change state to indicate data + * is available + */ + if (STAT_TO_STA(s->stat) == STAT_M_ADDR_RD_ACK) { + if (s->cntr & TWI_CNTR_A_ACK) { + s->stat = STAT_FROM_STA(STAT_M_DATA_RX_ACK); + } else { + s->stat = STAT_FROM_STA(STAT_M_DATA_RX_NACK); + } + allwinner_i2c_raise_interrupt(s); + } + break; + case TWI_CCR_REG: + value = s->ccr; + break; + case TWI_SRST_REG: + value = s->srst; + break; + case TWI_EFR_REG: + value = s->efr; + break; + case TWI_LCR_REG: + value = s->lcr; + break; + default: + qemu_log_mask(LOG_GUEST_ERROR, "[%s]%s: Bad address at offset 0x%" + HWADDR_PRIx "\n", TYPE_AW_I2C, __func__, offset); + value = 0; + break; + } + + trace_allwinner_i2c_read(allwinner_i2c_get_regname(offset), offset, value); + + return (uint64_t)value; +} + +static void allwinner_i2c_write(void *opaque, hwaddr offset, + uint64_t value, unsigned size) +{ + AWI2CState *s = AW_I2C(opaque); + + value &= 0xff; + + trace_allwinner_i2c_write(allwinner_i2c_get_regname(offset), offset, value); + + switch (offset) { + case TWI_ADDR_REG: + s->addr = (uint8_t)value; + break; + case TWI_XADDR_REG: + s->xaddr = (uint8_t)value; + break; + case TWI_DATA_REG: + /* If the device is in reset or not enabled, nothing to do */ + if (allwinner_i2c_is_reset(s) || (!allwinner_i2c_bus_is_enabled(s))) { + break; + } + + s->data = value & TWI_DATA_MASK; + + switch (STAT_TO_STA(s->stat)) { + case STAT_M_STA_TX: + case STAT_M_RSTA_TX: + /* Send address */ + if (i2c_start_transfer(s->bus, extract32(s->data, 1, 7), + extract32(s->data, 0, 1))) { + /* If non zero is returned, the address is not valid */ + s->stat = STAT_FROM_STA(STAT_M_ADDR_WR_NACK); + } else { + /* Determine if read of write */ + if (extract32(s->data, 0, 1)) { + s->stat = STAT_FROM_STA(STAT_M_ADDR_RD_ACK); + } else { + s->stat = STAT_FROM_STA(STAT_M_ADDR_WR_ACK); + } + allwinner_i2c_raise_interrupt(s); + } + break; + case STAT_M_ADDR_WR_ACK: + case STAT_M_DATA_TX_ACK: + if (i2c_send(s->bus, s->data)) { + /* If the target return non zero then end the transfer */ + s->stat = STAT_FROM_STA(STAT_M_DATA_TX_NACK); + i2c_end_transfer(s->bus); + } else { + s->stat = STAT_FROM_STA(STAT_M_DATA_TX_ACK); + allwinner_i2c_raise_interrupt(s); + } + break; + default: + break; + } + break; + case TWI_CNTR_REG: + if (!allwinner_i2c_is_reset(s)) { + /* Do something only if not in software reset */ + s->cntr = value & TWI_CNTR_MASK; + + /* Check if start condition should be sent */ + if (s->cntr & TWI_CNTR_M_STA) { + /* Update status */ + if (STAT_TO_STA(s->stat) == STAT_IDLE) { + /* Send start condition */ + s->stat = STAT_FROM_STA(STAT_M_STA_TX); + } else { + /* Send repeated start condition */ + s->stat = STAT_FROM_STA(STAT_M_RSTA_TX); + } + /* Clear start condition */ + s->cntr &= ~TWI_CNTR_M_STA; + } + if (s->cntr & TWI_CNTR_M_STP) { + /* Update status */ + i2c_end_transfer(s->bus); + s->stat = STAT_FROM_STA(STAT_IDLE); + s->cntr &= ~TWI_CNTR_M_STP; + } + if ((s->cntr & TWI_CNTR_INT_FLAG) == 0) { + /* Interrupt flag cleared */ + qemu_irq_lower(s->irq); + } + if ((s->cntr & TWI_CNTR_A_ACK) == 0) { + if (STAT_TO_STA(s->stat) == STAT_M_DATA_RX_ACK) { + s->stat = STAT_FROM_STA(STAT_M_DATA_RX_NACK); + } + } else { + if (STAT_TO_STA(s->stat) == STAT_M_DATA_RX_NACK) { + s->stat = STAT_FROM_STA(STAT_M_DATA_RX_ACK); + } + } + allwinner_i2c_raise_interrupt(s); + + } + break; + case TWI_CCR_REG: + s->ccr = value & TWI_CCR_MASK; + break; + case TWI_SRST_REG: + if (((value & TWI_SRST_MASK) == 0) && (s->srst & TWI_SRST_MASK)) { + /* Perform reset */ + allwinner_i2c_reset_hold(OBJECT(s)); + } + s->srst = value & TWI_SRST_MASK; + break; + case TWI_EFR_REG: + s->efr = value & TWI_EFR_MASK; + break; + case TWI_LCR_REG: + s->lcr = value & TWI_LCR_MASK; + break; + default: + qemu_log_mask(LOG_GUEST_ERROR, "[%s]%s: Bad address at offset 0x%" + HWADDR_PRIx "\n", TYPE_AW_I2C, __func__, offset); + break; + } +} + +static const MemoryRegionOps allwinner_i2c_ops = { + .read = allwinner_i2c_read, + .write = allwinner_i2c_write, + .valid.min_access_size = 1, + .valid.max_access_size = 4, + .endianness = DEVICE_NATIVE_ENDIAN, +}; + +static const VMStateDescription allwinner_i2c_vmstate = { + .name = TYPE_AW_I2C, + .version_id = 1, + .minimum_version_id = 1, + .fields = (VMStateField[]) { + VMSTATE_UINT8(addr, AWI2CState), + VMSTATE_UINT8(xaddr, AWI2CState), + VMSTATE_UINT8(data, AWI2CState), + VMSTATE_UINT8(cntr, AWI2CState), + VMSTATE_UINT8(ccr, AWI2CState), + VMSTATE_UINT8(srst, AWI2CState), + VMSTATE_UINT8(efr, AWI2CState), + VMSTATE_UINT8(lcr, AWI2CState), + VMSTATE_END_OF_LIST() + } +}; + +static void allwinner_i2c_realize(DeviceState *dev, Error **errp) +{ + AWI2CState *s = AW_I2C(dev); + + memory_region_init_io(&s->iomem, OBJECT(s), &allwinner_i2c_ops, s, + TYPE_AW_I2C, AW_I2C_MEM_SIZE); + sysbus_init_mmio(SYS_BUS_DEVICE(dev), &s->iomem); + sysbus_init_irq(SYS_BUS_DEVICE(dev), &s->irq); + s->bus = i2c_init_bus(dev, "i2c"); +} + +static void allwinner_i2c_class_init(ObjectClass *klass, void *data) +{ + DeviceClass *dc = DEVICE_CLASS(klass); + ResettableClass *rc = RESETTABLE_CLASS(klass); + + rc->phases.hold = allwinner_i2c_reset_hold; + dc->vmsd = &allwinner_i2c_vmstate; + dc->realize = allwinner_i2c_realize; + dc->desc = "Allwinner I2C Controller"; +} + +static const TypeInfo allwinner_i2c_type_info = { + .name = TYPE_AW_I2C, + .parent = TYPE_SYS_BUS_DEVICE, + .instance_size = sizeof(AWI2CState), + .class_init = allwinner_i2c_class_init, +}; + +static void allwinner_i2c_register_types(void) +{ + type_register_static(&allwinner_i2c_type_info); +} + +type_init(allwinner_i2c_register_types) diff --git a/hw/arm/Kconfig b/hw/arm/Kconfig index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/Kconfig +++ b/hw/arm/Kconfig @@ -XXX,XX +XXX,XX @@ config ALLWINNER_A10 select ALLWINNER_A10_CCM select ALLWINNER_A10_DRAMC select ALLWINNER_EMAC + select ALLWINNER_I2C select SERIAL select UNIMP @@ -XXX,XX +XXX,XX @@ config ALLWINNER_H3 bool select ALLWINNER_A10_PIT select ALLWINNER_SUN8I_EMAC + select ALLWINNER_I2C select SERIAL select ARM_TIMER select ARM_GIC diff --git a/hw/i2c/Kconfig b/hw/i2c/Kconfig index XXXXXXX..XXXXXXX 100644 --- a/hw/i2c/Kconfig +++ b/hw/i2c/Kconfig @@ -XXX,XX +XXX,XX @@ config MPC_I2C bool select I2C +config ALLWINNER_I2C + bool + select I2C + config PCA954X bool select I2C diff --git a/hw/i2c/meson.build b/hw/i2c/meson.build index XXXXXXX..XXXXXXX 100644 --- a/hw/i2c/meson.build +++ b/hw/i2c/meson.build @@ -XXX,XX +XXX,XX @@ i2c_ss.add(when: 'CONFIG_BITBANG_I2C', if_true: files('bitbang_i2c.c')) i2c_ss.add(when: 'CONFIG_EXYNOS4', if_true: files('exynos4210_i2c.c')) i2c_ss.add(when: 'CONFIG_IMX_I2C', if_true: files('imx_i2c.c')) i2c_ss.add(when: 'CONFIG_MPC_I2C', if_true: files('mpc_i2c.c')) +i2c_ss.add(when: 'CONFIG_ALLWINNER_I2C', if_true: files('allwinner-i2c.c')) i2c_ss.add(when: 'CONFIG_NRF51_SOC', if_true: files('microbit_i2c.c')) i2c_ss.add(when: 'CONFIG_NPCM7XX', if_true: files('npcm7xx_smbus.c')) i2c_ss.add(when: 'CONFIG_SMBUS_EEPROM', if_true: files('smbus_eeprom.c')) diff --git a/hw/i2c/trace-events b/hw/i2c/trace-events index XXXXXXX..XXXXXXX 100644 --- a/hw/i2c/trace-events +++ b/hw/i2c/trace-events @@ -XXX,XX +XXX,XX @@ i2c_send_async(uint8_t address, uint8_t data) "send_async(addr:0x%02x) data:0x%0 i2c_recv(uint8_t address, uint8_t data) "recv(addr:0x%02x) data:0x%02x" i2c_ack(void) "" +# allwinner_i2c.c + +allwinner_i2c_read(const char* reg_name, uint64_t offset, uint64_t value) "read %s [0x%" PRIx64 "]: -> 0x%" PRIx64 +allwinner_i2c_write(const char* reg_name, uint64_t offset, uint64_t value) "write %s [0x%" PRIx64 "]: <- 0x%" PRIx64 + # aspeed_i2c.c aspeed_i2c_bus_cmd(uint32_t cmd, const char *cmd_flags, uint32_t count, uint32_t intr_status) "handling cmd=0x%x %s count=%d intr=0x%x" -- 2.34.1 From: Strahinja Jankovic <strahinjapjankovic@gmail.com> This patch adds minimal support for AXP-209 PMU. Most important is chip ID since U-Boot SPL expects version 0x1. Besides the chip ID register, reset values for two more registers used by A10 U-Boot SPL are covered. Signed-off-by: Strahinja Jankovic <strahinja.p.jankovic@gmail.com> Message-id: 20221226220303.14420-5-strahinja.p.jankovic@gmail.com Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/misc/axp209.c | 238 +++++++++++++++++++++++++++++++++++++++++++ MAINTAINERS | 2 + hw/misc/Kconfig | 4 + hw/misc/meson.build | 1 + hw/misc/trace-events | 5 + 5 files changed, 250 insertions(+) create mode 100644 hw/misc/axp209.c diff --git a/hw/misc/axp209.c b/hw/misc/axp209.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/misc/axp209.c @@ -XXX,XX +XXX,XX @@ +/* + * AXP-209 PMU Emulation + * + * Copyright (C) 2022 Strahinja Jankovic <strahinja.p.jankovic@gmail.com> + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * SPDX-License-Identifier: MIT + */ + +#include "qemu/osdep.h" +#include "qemu/log.h" +#include "trace.h" +#include "hw/i2c/i2c.h" +#include "migration/vmstate.h" + +#define TYPE_AXP209_PMU "axp209_pmu" + +#define AXP209(obj) \ + OBJECT_CHECK(AXP209I2CState, (obj), TYPE_AXP209_PMU) + +/* registers */ +enum { + REG_POWER_STATUS = 0x0u, + REG_OPERATING_MODE, + REG_OTG_VBUS_STATUS, + REG_CHIP_VERSION, + REG_DATA_CACHE_0, + REG_DATA_CACHE_1, + REG_DATA_CACHE_2, + REG_DATA_CACHE_3, + REG_DATA_CACHE_4, + REG_DATA_CACHE_5, + REG_DATA_CACHE_6, + REG_DATA_CACHE_7, + REG_DATA_CACHE_8, + REG_DATA_CACHE_9, + REG_DATA_CACHE_A, + REG_DATA_CACHE_B, + REG_POWER_OUTPUT_CTRL = 0x12u, + REG_DC_DC2_OUT_V_CTRL = 0x23u, + REG_DC_DC2_DVS_CTRL = 0x25u, + REG_DC_DC3_OUT_V_CTRL = 0x27u, + REG_LDO2_4_OUT_V_CTRL, + REG_LDO3_OUT_V_CTRL, + REG_VBUS_CH_MGMT = 0x30u, + REG_SHUTDOWN_V_CTRL, + REG_SHUTDOWN_CTRL, + REG_CHARGE_CTRL_1, + REG_CHARGE_CTRL_2, + REG_SPARE_CHARGE_CTRL, + REG_PEK_KEY_CTRL, + REG_DC_DC_FREQ_SET, + REG_CHR_TEMP_TH_SET, + REG_CHR_HIGH_TEMP_TH_CTRL, + REG_IPSOUT_WARN_L1, + REG_IPSOUT_WARN_L2, + REG_DISCHR_TEMP_TH_SET, + REG_DISCHR_HIGH_TEMP_TH_CTRL, + REG_IRQ_BANK_1_CTRL = 0x40u, + REG_IRQ_BANK_2_CTRL, + REG_IRQ_BANK_3_CTRL, + REG_IRQ_BANK_4_CTRL, + REG_IRQ_BANK_5_CTRL, + REG_IRQ_BANK_1_STAT = 0x48u, + REG_IRQ_BANK_2_STAT, + REG_IRQ_BANK_3_STAT, + REG_IRQ_BANK_4_STAT, + REG_IRQ_BANK_5_STAT, + REG_ADC_ACIN_V_H = 0x56u, + REG_ADC_ACIN_V_L, + REG_ADC_ACIN_CURR_H, + REG_ADC_ACIN_CURR_L, + REG_ADC_VBUS_V_H, + REG_ADC_VBUS_V_L, + REG_ADC_VBUS_CURR_H, + REG_ADC_VBUS_CURR_L, + REG_ADC_INT_TEMP_H, + REG_ADC_INT_TEMP_L, + REG_ADC_TEMP_SENS_V_H = 0x62u, + REG_ADC_TEMP_SENS_V_L, + REG_ADC_BAT_V_H = 0x78u, + REG_ADC_BAT_V_L, + REG_ADC_BAT_DISCHR_CURR_H, + REG_ADC_BAT_DISCHR_CURR_L, + REG_ADC_BAT_CHR_CURR_H, + REG_ADC_BAT_CHR_CURR_L, + REG_ADC_IPSOUT_V_H, + REG_ADC_IPSOUT_V_L, + REG_DC_DC_MOD_SEL = 0x80u, + REG_ADC_EN_1, + REG_ADC_EN_2, + REG_ADC_SR_CTRL, + REG_ADC_IN_RANGE, + REG_GPIO1_ADC_IRQ_RISING_TH, + REG_GPIO1_ADC_IRQ_FALLING_TH, + REG_TIMER_CTRL = 0x8au, + REG_VBUS_CTRL_MON_SRP, + REG_OVER_TEMP_SHUTDOWN = 0x8fu, + REG_GPIO0_FEAT_SET, + REG_GPIO_OUT_HIGH_SET, + REG_GPIO1_FEAT_SET, + REG_GPIO2_FEAT_SET, + REG_GPIO_SIG_STATE_SET_MON, + REG_GPIO3_SET, + REG_COULOMB_CNTR_CTRL = 0xb8u, + REG_POWER_MEAS_RES, + NR_REGS +}; + +#define AXP209_CHIP_VERSION_ID (0x01) +#define AXP209_DC_DC2_OUT_V_CTRL_RESET (0x16) +#define AXP209_IRQ_BANK_1_CTRL_RESET (0xd8) + +/* A simple I2C slave which returns values of ID or CNT register. */ +typedef struct AXP209I2CState { + /*< private >*/ + I2CSlave i2c; + /*< public >*/ + uint8_t regs[NR_REGS]; /* peripheral registers */ + uint8_t ptr; /* current register index */ + uint8_t count; /* counter used for tx/rx */ +} AXP209I2CState; + +/* Reset all counters and load ID register */ +static void axp209_reset_enter(Object *obj, ResetType type) +{ + AXP209I2CState *s = AXP209(obj); + + memset(s->regs, 0, NR_REGS); + s->ptr = 0; + s->count = 0; + s->regs[REG_CHIP_VERSION] = AXP209_CHIP_VERSION_ID; + s->regs[REG_DC_DC2_OUT_V_CTRL] = AXP209_DC_DC2_OUT_V_CTRL_RESET; + s->regs[REG_IRQ_BANK_1_CTRL] = AXP209_IRQ_BANK_1_CTRL_RESET; +} + +/* Handle events from master. */ +static int axp209_event(I2CSlave *i2c, enum i2c_event event) +{ + AXP209I2CState *s = AXP209(i2c); + + s->count = 0; + + return 0; +} + +/* Called when master requests read */ +static uint8_t axp209_rx(I2CSlave *i2c) +{ + AXP209I2CState *s = AXP209(i2c); + uint8_t ret = 0xff; + + if (s->ptr < NR_REGS) { + ret = s->regs[s->ptr++]; + } + + trace_axp209_rx(s->ptr - 1, ret); + + return ret; +} + +/* + * Called when master sends write. + * Update ptr with byte 0, then perform write with second byte. + */ +static int axp209_tx(I2CSlave *i2c, uint8_t data) +{ + AXP209I2CState *s = AXP209(i2c); + + if (s->count == 0) { + /* Store register address */ + s->ptr = data; + s->count++; + trace_axp209_select(data); + } else { + trace_axp209_tx(s->ptr, data); + if (s->ptr == REG_DC_DC2_OUT_V_CTRL) { + s->regs[s->ptr++] = data; + } + } + + return 0; +} + +static const VMStateDescription vmstate_axp209 = { + .name = TYPE_AXP209_PMU, + .version_id = 1, + .fields = (VMStateField[]) { + VMSTATE_UINT8_ARRAY(regs, AXP209I2CState, NR_REGS), + VMSTATE_UINT8(count, AXP209I2CState), + VMSTATE_UINT8(ptr, AXP209I2CState), + VMSTATE_END_OF_LIST() + } +}; + +static void axp209_class_init(ObjectClass *oc, void *data) +{ + DeviceClass *dc = DEVICE_CLASS(oc); + I2CSlaveClass *isc = I2C_SLAVE_CLASS(oc); + ResettableClass *rc = RESETTABLE_CLASS(oc); + + rc->phases.enter = axp209_reset_enter; + dc->vmsd = &vmstate_axp209; + isc->event = axp209_event; + isc->recv = axp209_rx; + isc->send = axp209_tx; +} + +static const TypeInfo axp209_info = { + .name = TYPE_AXP209_PMU, + .parent = TYPE_I2C_SLAVE, + .instance_size = sizeof(AXP209I2CState), + .class_init = axp209_class_init +}; + +static void axp209_register_devices(void) +{ + type_register_static(&axp209_info); +} + +type_init(axp209_register_devices); diff --git a/MAINTAINERS b/MAINTAINERS index XXXXXXX..XXXXXXX 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -XXX,XX +XXX,XX @@ ARM Machines Allwinner-a10 M: Beniamino Galvani <b.galvani@gmail.com> M: Peter Maydell <peter.maydell@linaro.org> +R: Strahinja Jankovic <strahinja.p.jankovic@gmail.com> L: qemu-arm@nongnu.org S: Odd Fixes F: hw/*/allwinner* F: include/hw/*/allwinner* F: hw/arm/cubieboard.c F: docs/system/arm/cubieboard.rst +F: hw/misc/axp209.c Allwinner-h3 M: Niek Linnenbank <nieklinnenbank@gmail.com> diff --git a/hw/misc/Kconfig b/hw/misc/Kconfig index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/Kconfig +++ b/hw/misc/Kconfig @@ -XXX,XX +XXX,XX @@ config ALLWINNER_A10_CCM config ALLWINNER_A10_DRAMC bool +config AXP209_PMU + bool + depends on I2C + source macio/Kconfig diff --git a/hw/misc/meson.build b/hw/misc/meson.build index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/meson.build +++ b/hw/misc/meson.build @@ -XXX,XX +XXX,XX @@ specific_ss.add(when: 'CONFIG_ALLWINNER_H3', if_true: files('allwinner-cpucfg.c' softmmu_ss.add(when: 'CONFIG_ALLWINNER_H3', if_true: files('allwinner-h3-dramc.c')) softmmu_ss.add(when: 'CONFIG_ALLWINNER_H3', if_true: files('allwinner-h3-sysctrl.c')) softmmu_ss.add(when: 'CONFIG_ALLWINNER_H3', if_true: files('allwinner-sid.c')) +softmmu_ss.add(when: 'CONFIG_AXP209_PMU', if_true: files('axp209.c')) softmmu_ss.add(when: 'CONFIG_REALVIEW', if_true: files('arm_sysctl.c')) softmmu_ss.add(when: 'CONFIG_NSERIES', if_true: files('cbus.c')) softmmu_ss.add(when: 'CONFIG_ECCMEMCTL', if_true: files('eccmemctl.c')) diff --git a/hw/misc/trace-events b/hw/misc/trace-events index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/trace-events +++ b/hw/misc/trace-events @@ -XXX,XX +XXX,XX @@ allwinner_sid_write(uint64_t offset, uint64_t data, unsigned size) "offset 0x%" avr_power_read(uint8_t value) "power_reduc read value:%u" avr_power_write(uint8_t value) "power_reduc write value:%u" +# axp209.c +axp209_rx(uint8_t reg, uint8_t data) "Read reg 0x%" PRIx8 " : 0x%" PRIx8 +axp209_select(uint8_t reg) "Accessing reg 0x%" PRIx8 +axp209_tx(uint8_t reg, uint8_t data) "Write reg 0x%" PRIx8 " : 0x%" PRIx8 + # eccmemctl.c ecc_mem_writel_mer(uint32_t val) "Write memory enable 0x%08x" ecc_mem_writel_mdr(uint32_t val) "Write memory delay 0x%08x" -- 2.34.1 From: Strahinja Jankovic <strahinjapjankovic@gmail.com> SPL Boot for Cubieboard expects AXP209 connected to I2C0 bus. Signed-off-by: Strahinja Jankovic <strahinja.p.jankovic@gmail.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Message-id: 20221226220303.14420-6-strahinja.p.jankovic@gmail.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/cubieboard.c | 6 ++++++ hw/arm/Kconfig | 1 + 2 files changed, 7 insertions(+) diff --git a/hw/arm/cubieboard.c b/hw/arm/cubieboard.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/cubieboard.c +++ b/hw/arm/cubieboard.c @@ -XXX,XX +XXX,XX @@ #include "hw/boards.h" #include "hw/qdev-properties.h" #include "hw/arm/allwinner-a10.h" +#include "hw/i2c/i2c.h" static struct arm_boot_info cubieboard_binfo = { .loader_start = AW_A10_SDRAM_BASE, @@ -XXX,XX +XXX,XX @@ static void cubieboard_init(MachineState *machine) BlockBackend *blk; BusState *bus; DeviceState *carddev; + I2CBus *i2c; /* BIOS is not supported by this board */ if (machine->firmware) { @@ -XXX,XX +XXX,XX @@ static void cubieboard_init(MachineState *machine) exit(1); } + /* Connect AXP 209 */ + i2c = I2C_BUS(qdev_get_child_bus(DEVICE(&a10->i2c0), "i2c")); + i2c_slave_create_simple(i2c, "axp209_pmu", 0x34); + /* Retrieve SD bus */ di = drive_get(IF_SD, 0, 0); blk = di ? blk_by_legacy_dinfo(di) : NULL; diff --git a/hw/arm/Kconfig b/hw/arm/Kconfig index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/Kconfig +++ b/hw/arm/Kconfig @@ -XXX,XX +XXX,XX @@ config ALLWINNER_A10 select ALLWINNER_A10_DRAMC select ALLWINNER_EMAC select ALLWINNER_I2C + select AXP209_PMU select SERIAL select UNIMP -- 2.34.1 From: Strahinja Jankovic <strahinjapjankovic@gmail.com> This patch enables copying of SPL from MMC if `-kernel` parameter is not passed when starting QEMU. SPL is copied to SRAM_A. The approach is reused from Allwinner H3 implementation. Tested with Armbian and custom Yocto image. Signed-off-by: Strahinja Jankovic <strahinja.p.jankovic@gmail.com> Reviewed-by: Niek Linnenbank <nieklinnenbank@gmail.com> Message-id: 20221226220303.14420-7-strahinja.p.jankovic@gmail.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/allwinner-a10.h | 21 +++++++++++++++++++++ hw/arm/allwinner-a10.c | 18 ++++++++++++++++++ hw/arm/cubieboard.c | 5 +++++ 3 files changed, 44 insertions(+) diff --git a/include/hw/arm/allwinner-a10.h b/include/hw/arm/allwinner-a10.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/allwinner-a10.h +++ b/include/hw/arm/allwinner-a10.h @@ -XXX,XX +XXX,XX @@ #include "hw/misc/allwinner-a10-ccm.h" #include "hw/misc/allwinner-a10-dramc.h" #include "hw/i2c/allwinner-i2c.h" +#include "sysemu/block-backend.h" #include "target/arm/cpu.h" #include "qom/object.h" @@ -XXX,XX +XXX,XX @@ struct AwA10State { OHCISysBusState ohci[AW_A10_NUM_USB]; }; +/** + * Emulate Boot ROM firmware setup functionality. + * + * A real Allwinner A10 SoC contains a Boot ROM + * which is the first code that runs right after + * the SoC is powered on. The Boot ROM is responsible + * for loading user code (e.g. a bootloader) from any + * of the supported external devices and writing the + * downloaded code to internal SRAM. After loading the SoC + * begins executing the code written to SRAM. + * + * This function emulates the Boot ROM by copying 32 KiB + * of data at offset 8 KiB from the given block device and writes it to + * the start of the first internal SRAM memory. + * + * @s: Allwinner A10 state object pointer + * @blk: Block backend device object pointer + */ +void allwinner_a10_bootrom_setup(AwA10State *s, BlockBackend *blk); + #endif diff --git a/hw/arm/allwinner-a10.c b/hw/arm/allwinner-a10.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/allwinner-a10.c +++ b/hw/arm/allwinner-a10.c @@ -XXX,XX +XXX,XX @@ #include "sysemu/sysemu.h" #include "hw/boards.h" #include "hw/usb/hcd-ohci.h" +#include "hw/loader.h" +#define AW_A10_SRAM_A_BASE 0x00000000 #define AW_A10_DRAMC_BASE 0x01c01000 #define AW_A10_MMC0_BASE 0x01c0f000 #define AW_A10_CCM_BASE 0x01c20000 @@ -XXX,XX +XXX,XX @@ #define AW_A10_RTC_BASE 0x01c20d00 #define AW_A10_I2C0_BASE 0x01c2ac00 +void allwinner_a10_bootrom_setup(AwA10State *s, BlockBackend *blk) +{ + const int64_t rom_size = 32 * KiB; + g_autofree uint8_t *buffer = g_new0(uint8_t, rom_size); + + if (blk_pread(blk, 8 * KiB, rom_size, buffer, 0) < 0) { + error_setg(&error_fatal, "%s: failed to read BlockBackend data", + __func__); + return; + } + + rom_add_blob("allwinner-a10.bootrom", buffer, rom_size, + rom_size, AW_A10_SRAM_A_BASE, + NULL, NULL, NULL, NULL, false); +} + static void aw_a10_init(Object *obj) { AwA10State *s = AW_A10(obj); diff --git a/hw/arm/cubieboard.c b/hw/arm/cubieboard.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/cubieboard.c +++ b/hw/arm/cubieboard.c @@ -XXX,XX +XXX,XX @@ static void cubieboard_init(MachineState *machine) memory_region_add_subregion(get_system_memory(), AW_A10_SDRAM_BASE, machine->ram); + /* Load target kernel or start using BootROM */ + if (!machine->kernel_filename && blk && blk_is_available(blk)) { + /* Use Boot ROM to copy data from SD card to SRAM */ + allwinner_a10_bootrom_setup(a10, blk); + } /* TODO create and connect IDE devices for ide_drive_get() */ cubieboard_binfo.ram_size = machine->ram_size; -- 2.34.1 From: Strahinja Jankovic <strahinjapjankovic@gmail.com> Cubieboard now can boot directly from SD card, without the need to pass `-kernel` parameter. Update Avocado tests to cover this functionality. Signed-off-by: Strahinja Jankovic <strahinja.p.jankovic@gmail.com> Reviewed-by: Niek Linnenbank <nieklinnenbank@gmail.com> Tested-by: Niek Linnenbank <nieklinnenbank@gmail.com> Message-id: 20221226220303.14420-8-strahinja.p.jankovic@gmail.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- tests/avocado/boot_linux_console.py | 47 +++++++++++++++++++++++++++++ 1 file changed, 47 insertions(+) diff --git a/tests/avocado/boot_linux_console.py b/tests/avocado/boot_linux_console.py index XXXXXXX..XXXXXXX 100644 --- a/tests/avocado/boot_linux_console.py +++ b/tests/avocado/boot_linux_console.py @@ -XXX,XX +XXX,XX @@ def test_arm_cubieboard_sata(self): 'sda') # cubieboard's reboot is not functioning; omit reboot test. + @skipUnless(os.getenv('AVOCADO_ALLOW_LARGE_STORAGE'), 'storage limited') + def test_arm_cubieboard_openwrt_22_03_2(self): + """ + :avocado: tags=arch:arm + :avocado: tags=machine:cubieboard + :avocado: tags=device:sd + """ + + # This test download a 7.5 MiB compressed image and expand it + # to 126 MiB. + image_url = ('https://downloads.openwrt.org/releases/22.03.2/targets/' + 'sunxi/cortexa8/openwrt-22.03.2-sunxi-cortexa8-' + 'cubietech_a10-cubieboard-ext4-sdcard.img.gz') + image_hash = ('94b5ecbfbc0b3b56276e5146b899eafa' + '2ac5dc2d08733d6705af9f144f39f554') + image_path_gz = self.fetch_asset(image_url, asset_hash=image_hash, + algorithm='sha256') + image_path = archive.extract(image_path_gz, self.workdir) + image_pow2ceil_expand(image_path) + + self.vm.set_console() + self.vm.add_args('-drive', 'file=' + image_path + ',if=sd,format=raw', + '-nic', 'user', + '-no-reboot') + self.vm.launch() + + kernel_command_line = (self.KERNEL_COMMON_COMMAND_LINE + + 'usbcore.nousb ' + 'noreboot') + + self.wait_for_console_pattern('U-Boot SPL') + + interrupt_interactive_console_until_pattern( + self, 'Hit any key to stop autoboot:', '=>') + exec_command_and_wait_for_pattern(self, "setenv extraargs '" + + kernel_command_line + "'", '=>') + exec_command_and_wait_for_pattern(self, 'boot', 'Starting kernel ...'); + + self.wait_for_console_pattern( + 'Please press Enter to activate this console.') + + exec_command_and_wait_for_pattern(self, ' ', 'root@') + + exec_command_and_wait_for_pattern(self, 'cat /proc/cpuinfo', + 'Allwinner sun4i/sun5i') + # cubieboard's reboot is not functioning; omit reboot test. + @skipUnless(os.getenv('AVOCADO_TIMEOUT_EXPECTED'), 'Test might timeout') def test_arm_quanta_gsj(self): """ -- 2.34.1 From: Richard Henderson <richard.henderson@linaro.org> Don't dereference CPUTLBEntryFull until we verify that the page is valid. Move the other user-only info field updates after the valid check to match. Cc: qemu-stable@nongnu.org Resolves: https://gitlab.com/qemu-project/qemu/-/issues/1412 Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Message-id: 20230104190056.305143-1-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sve_helper.c | 14 +++++++++----- 1 file changed, 9 insertions(+), 5 deletions(-) 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 @@ bool sve_probe_page(SVEHostPage *info, bool nofault, CPUARMState *env, #ifdef CONFIG_USER_ONLY flags = probe_access_flags(env, addr, access_type, mmu_idx, nofault, &info->host, retaddr); - memset(&info->attrs, 0, sizeof(info->attrs)); - /* Require both ANON and MTE; see allocation_tag_mem(). */ - info->tagged = (flags & PAGE_ANON) && (flags & PAGE_MTE); #else CPUTLBEntryFull *full; flags = probe_access_full(env, addr, access_type, mmu_idx, nofault, &info->host, &full, retaddr); - info->attrs = full->attrs; - info->tagged = full->pte_attrs == 0xf0; #endif info->flags = flags; @@ -XXX,XX +XXX,XX @@ bool sve_probe_page(SVEHostPage *info, bool nofault, CPUARMState *env, return false; } +#ifdef CONFIG_USER_ONLY + memset(&info->attrs, 0, sizeof(info->attrs)); + /* Require both ANON and MTE; see allocation_tag_mem(). */ + info->tagged = (flags & PAGE_ANON) && (flags & PAGE_MTE); +#else + info->attrs = full->attrs; + info->tagged = full->pte_attrs == 0xf0; +#endif + /* Ensure that info->host[] is relative to addr, not addr + mem_off. */ info->host -= mem_off; return true; -- 2.34.1 From: Philippe Mathieu-Daudé <philmd@linaro.org> Since pxa255_init() must map the device in the system memory, there is no point in passing get_system_memory() by argument. Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230109115316.2235-2-philmd@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/pxa.h | 2 +- hw/arm/gumstix.c | 3 +-- hw/arm/pxa2xx.c | 4 +++- hw/arm/tosa.c | 2 +- 4 files changed, 6 insertions(+), 5 deletions(-) diff --git a/include/hw/arm/pxa.h b/include/hw/arm/pxa.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/pxa.h +++ b/include/hw/arm/pxa.h @@ -XXX,XX +XXX,XX @@ struct PXA2xxI2SState { PXA2xxState *pxa270_init(MemoryRegion *address_space, unsigned int sdram_size, const char *revision); -PXA2xxState *pxa255_init(MemoryRegion *address_space, unsigned int sdram_size); +PXA2xxState *pxa255_init(unsigned int sdram_size); #endif /* PXA_H */ diff --git a/hw/arm/gumstix.c b/hw/arm/gumstix.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/gumstix.c +++ b/hw/arm/gumstix.c @@ -XXX,XX +XXX,XX @@ static void connex_init(MachineState *machine) { PXA2xxState *cpu; DriveInfo *dinfo; - MemoryRegion *address_space_mem = get_system_memory(); uint32_t connex_rom = 0x01000000; uint32_t connex_ram = 0x04000000; - cpu = pxa255_init(address_space_mem, connex_ram); + cpu = pxa255_init(connex_ram); dinfo = drive_get(IF_PFLASH, 0, 0); if (!dinfo && !qtest_enabled()) { diff --git a/hw/arm/pxa2xx.c b/hw/arm/pxa2xx.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/pxa2xx.c +++ b/hw/arm/pxa2xx.c @@ -XXX,XX +XXX,XX @@ #include "qemu/error-report.h" #include "qemu/module.h" #include "qapi/error.h" +#include "exec/address-spaces.h" #include "cpu.h" #include "hw/sysbus.h" #include "migration/vmstate.h" @@ -XXX,XX +XXX,XX @@ PXA2xxState *pxa270_init(MemoryRegion *address_space, } /* Initialise a PXA255 integrated chip (ARM based core). */ -PXA2xxState *pxa255_init(MemoryRegion *address_space, unsigned int sdram_size) +PXA2xxState *pxa255_init(unsigned int sdram_size) { + MemoryRegion *address_space = get_system_memory(); PXA2xxState *s; int i; DriveInfo *dinfo; diff --git a/hw/arm/tosa.c b/hw/arm/tosa.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/tosa.c +++ b/hw/arm/tosa.c @@ -XXX,XX +XXX,XX @@ static void tosa_init(MachineState *machine) TC6393xbState *tmio; DeviceState *scp0, *scp1; - mpu = pxa255_init(address_space_mem, tosa_binfo.ram_size); + mpu = pxa255_init(tosa_binfo.ram_size); memory_region_init_rom(rom, NULL, "tosa.rom", TOSA_ROM, &error_fatal); memory_region_add_subregion(address_space_mem, 0, rom); -- 2.34.1 From: Philippe Mathieu-Daudé <philmd@linaro.org> Since pxa270_init() must map the device in the system memory, there is no point in passing get_system_memory() by argument. Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230109115316.2235-3-philmd@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/pxa.h | 3 +-- hw/arm/gumstix.c | 3 +-- hw/arm/mainstone.c | 10 ++++------ hw/arm/pxa2xx.c | 4 ++-- hw/arm/spitz.c | 6 ++---- hw/arm/z2.c | 3 +-- 6 files changed, 11 insertions(+), 18 deletions(-) diff --git a/include/hw/arm/pxa.h b/include/hw/arm/pxa.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/pxa.h +++ b/include/hw/arm/pxa.h @@ -XXX,XX +XXX,XX @@ struct PXA2xxI2SState { # define PA_FMT "0x%08lx" -PXA2xxState *pxa270_init(MemoryRegion *address_space, unsigned int sdram_size, - const char *revision); +PXA2xxState *pxa270_init(unsigned int sdram_size, const char *revision); PXA2xxState *pxa255_init(unsigned int sdram_size); #endif /* PXA_H */ diff --git a/hw/arm/gumstix.c b/hw/arm/gumstix.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/gumstix.c +++ b/hw/arm/gumstix.c @@ -XXX,XX +XXX,XX @@ static void verdex_init(MachineState *machine) { PXA2xxState *cpu; DriveInfo *dinfo; - MemoryRegion *address_space_mem = get_system_memory(); uint32_t verdex_rom = 0x02000000; uint32_t verdex_ram = 0x10000000; - cpu = pxa270_init(address_space_mem, verdex_ram, machine->cpu_type); + cpu = pxa270_init(verdex_ram, machine->cpu_type); dinfo = drive_get(IF_PFLASH, 0, 0); if (!dinfo && !qtest_enabled()) { diff --git a/hw/arm/mainstone.c b/hw/arm/mainstone.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/mainstone.c +++ b/hw/arm/mainstone.c @@ -XXX,XX +XXX,XX @@ static struct arm_boot_info mainstone_binfo = { .ram_size = 0x04000000, }; -static void mainstone_common_init(MemoryRegion *address_space_mem, - MachineState *machine, +static void mainstone_common_init(MachineState *machine, enum mainstone_model_e model, int arm_id) { uint32_t sector_len = 256 * 1024; @@ -XXX,XX +XXX,XX @@ static void mainstone_common_init(MemoryRegion *address_space_mem, MemoryRegion *rom = g_new(MemoryRegion, 1); /* Setup CPU & memory */ - mpu = pxa270_init(address_space_mem, mainstone_binfo.ram_size, - machine->cpu_type); + mpu = pxa270_init(mainstone_binfo.ram_size, machine->cpu_type); memory_region_init_rom(rom, NULL, "mainstone.rom", MAINSTONE_ROM, &error_fatal); - memory_region_add_subregion(address_space_mem, 0, rom); + memory_region_add_subregion(get_system_memory(), 0x00000000, rom); /* There are two 32MiB flash devices on the board */ for (i = 0; i < 2; i ++) { @@ -XXX,XX +XXX,XX @@ static void mainstone_common_init(MemoryRegion *address_space_mem, static void mainstone_init(MachineState *machine) { - mainstone_common_init(get_system_memory(), machine, mainstone, 0x196); + mainstone_common_init(machine, mainstone, 0x196); } static void mainstone2_machine_init(MachineClass *mc) diff --git a/hw/arm/pxa2xx.c b/hw/arm/pxa2xx.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/pxa2xx.c +++ b/hw/arm/pxa2xx.c @@ -XXX,XX +XXX,XX @@ static void pxa2xx_reset(void *opaque, int line, int level) } /* Initialise a PXA270 integrated chip (ARM based core). */ -PXA2xxState *pxa270_init(MemoryRegion *address_space, - unsigned int sdram_size, const char *cpu_type) +PXA2xxState *pxa270_init(unsigned int sdram_size, const char *cpu_type) { + MemoryRegion *address_space = get_system_memory(); PXA2xxState *s; int i; DriveInfo *dinfo; diff --git a/hw/arm/spitz.c b/hw/arm/spitz.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/spitz.c +++ b/hw/arm/spitz.c @@ -XXX,XX +XXX,XX @@ static void spitz_common_init(MachineState *machine) SpitzMachineState *sms = SPITZ_MACHINE(machine); enum spitz_model_e model = smc->model; PXA2xxState *mpu; - MemoryRegion *address_space_mem = get_system_memory(); MemoryRegion *rom = g_new(MemoryRegion, 1); /* Setup CPU & memory */ - mpu = pxa270_init(address_space_mem, spitz_binfo.ram_size, - machine->cpu_type); + mpu = pxa270_init(spitz_binfo.ram_size, machine->cpu_type); sms->mpu = mpu; sl_flash_register(mpu, (model == spitz) ? FLASH_128M : FLASH_1024M); memory_region_init_rom(rom, NULL, "spitz.rom", SPITZ_ROM, &error_fatal); - memory_region_add_subregion(address_space_mem, 0, rom); + memory_region_add_subregion(get_system_memory(), 0, rom); /* Setup peripherals */ spitz_keyboard_register(mpu); diff --git a/hw/arm/z2.c b/hw/arm/z2.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/z2.c +++ b/hw/arm/z2.c @@ -XXX,XX +XXX,XX @@ static const TypeInfo aer915_info = { static void z2_init(MachineState *machine) { - MemoryRegion *address_space_mem = get_system_memory(); uint32_t sector_len = 0x10000; PXA2xxState *mpu; DriveInfo *dinfo; @@ -XXX,XX +XXX,XX @@ static void z2_init(MachineState *machine) DeviceState *wm; /* Setup CPU & memory */ - mpu = pxa270_init(address_space_mem, z2_binfo.ram_size, machine->cpu_type); + mpu = pxa270_init(z2_binfo.ram_size, machine->cpu_type); dinfo = drive_get(IF_PFLASH, 0, 0); if (!pflash_cfi01_register(Z2_FLASH_BASE, "z2.flash0", Z2_FLASH_SIZE, -- 2.34.1 From: Philippe Mathieu-Daudé <philmd@linaro.org> IEC binary prefixes ease code review: the unit is explicit. Add definitions for RAM / Flash / Flash blocksize. Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230109115316.2235-4-philmd@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/collie.c | 16 ++++++++++------ 1 file changed, 10 insertions(+), 6 deletions(-) diff --git a/hw/arm/collie.c b/hw/arm/collie.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/collie.c +++ b/hw/arm/collie.c @@ -XXX,XX +XXX,XX @@ #include "cpu.h" #include "qom/object.h" +#define RAM_SIZE (512 * MiB) +#define FLASH_SIZE (32 * MiB) +#define FLASH_SECTOR_SIZE (64 * KiB) + struct CollieMachineState { MachineState parent; @@ -XXX,XX +XXX,XX @@ OBJECT_DECLARE_SIMPLE_TYPE(CollieMachineState, COLLIE_MACHINE) static struct arm_boot_info collie_binfo = { .loader_start = SA_SDCS0, - .ram_size = 0x20000000, + .ram_size = RAM_SIZE, }; static void collie_init(MachineState *machine) @@ -XXX,XX +XXX,XX @@ static void collie_init(MachineState *machine) memory_region_add_subregion(get_system_memory(), SA_SDCS0, machine->ram); dinfo = drive_get(IF_PFLASH, 0, 0); - pflash_cfi01_register(SA_CS0, "collie.fl1", 0x02000000, + pflash_cfi01_register(SA_CS0, "collie.fl1", FLASH_SIZE, dinfo ? blk_by_legacy_dinfo(dinfo) : NULL, - 64 * KiB, 4, 0x00, 0x00, 0x00, 0x00, 0); + FLASH_SECTOR_SIZE, 4, 0x00, 0x00, 0x00, 0x00, 0); dinfo = drive_get(IF_PFLASH, 0, 1); - pflash_cfi01_register(SA_CS1, "collie.fl2", 0x02000000, + pflash_cfi01_register(SA_CS1, "collie.fl2", FLASH_SIZE, dinfo ? blk_by_legacy_dinfo(dinfo) : NULL, - 64 * KiB, 4, 0x00, 0x00, 0x00, 0x00, 0); + FLASH_SECTOR_SIZE, 4, 0x00, 0x00, 0x00, 0x00, 0); sysbus_create_simple("scoop", 0x40800000, NULL); @@ -XXX,XX +XXX,XX @@ static void collie_machine_class_init(ObjectClass *oc, void *data) mc->init = collie_init; mc->ignore_memory_transaction_failures = true; mc->default_cpu_type = ARM_CPU_TYPE_NAME("sa1110"); - mc->default_ram_size = 0x20000000; + mc->default_ram_size = RAM_SIZE; mc->default_ram_id = "strongarm.sdram"; } -- 2.34.1 From: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230109115316.2235-5-philmd@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/collie.c | 17 +++++++---------- 1 file changed, 7 insertions(+), 10 deletions(-) diff --git a/hw/arm/collie.c b/hw/arm/collie.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/collie.c +++ b/hw/arm/collie.c @@ -XXX,XX +XXX,XX @@ static struct arm_boot_info collie_binfo = { static void collie_init(MachineState *machine) { - DriveInfo *dinfo; MachineClass *mc = MACHINE_GET_CLASS(machine); CollieMachineState *cms = COLLIE_MACHINE(machine); @@ -XXX,XX +XXX,XX @@ static void collie_init(MachineState *machine) memory_region_add_subregion(get_system_memory(), SA_SDCS0, machine->ram); - dinfo = drive_get(IF_PFLASH, 0, 0); - pflash_cfi01_register(SA_CS0, "collie.fl1", FLASH_SIZE, - dinfo ? blk_by_legacy_dinfo(dinfo) : NULL, - FLASH_SECTOR_SIZE, 4, 0x00, 0x00, 0x00, 0x00, 0); - - dinfo = drive_get(IF_PFLASH, 0, 1); - pflash_cfi01_register(SA_CS1, "collie.fl2", FLASH_SIZE, - dinfo ? blk_by_legacy_dinfo(dinfo) : NULL, - FLASH_SECTOR_SIZE, 4, 0x00, 0x00, 0x00, 0x00, 0); + for (unsigned i = 0; i < 2; i++) { + DriveInfo *dinfo = drive_get(IF_PFLASH, 0, i); + pflash_cfi01_register(i ? SA_CS1 : SA_CS0, + i ? "collie.fl2" : "collie.fl1", FLASH_SIZE, + dinfo ? blk_by_legacy_dinfo(dinfo) : NULL, + FLASH_SECTOR_SIZE, 4, 0x00, 0x00, 0x00, 0x00, 0); + } sysbus_create_simple("scoop", 0x40800000, NULL); -- 2.34.1 From: Philippe Mathieu-Daudé <f4bug@amsat.org> Add a comment describing the Connex uses a Numonyx RC28F128J3F75 flash, and the Verdex uses a Micron RC28F256P30TFA. Correct the Verdex machine description (we model the 'Pro' board). Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230109115316.2235-6-philmd@linaro.org Message-Id: <20200223231044.8003-3-philmd@redhat.com> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/gumstix.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/hw/arm/gumstix.c b/hw/arm/gumstix.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/gumstix.c +++ b/hw/arm/gumstix.c @@ -XXX,XX +XXX,XX @@ * Contributions after 2012-01-13 are licensed under the terms of the * GNU GPL, version 2 or (at your option) any later version. */ - + /* * Example usage: * @@ -XXX,XX +XXX,XX @@ static void connex_init(MachineState *machine) exit(1); } + /* Numonyx RC28F128J3F75 */ if (!pflash_cfi01_register(0x00000000, "connext.rom", connex_rom, dinfo ? blk_by_legacy_dinfo(dinfo) : NULL, sector_len, 2, 0, 0, 0, 0, 0)) { @@ -XXX,XX +XXX,XX @@ static void verdex_init(MachineState *machine) exit(1); } + /* Micron RC28F256P30TFA */ if (!pflash_cfi01_register(0x00000000, "verdex.rom", verdex_rom, dinfo ? blk_by_legacy_dinfo(dinfo) : NULL, sector_len, 2, 0, 0, 0, 0, 0)) { @@ -XXX,XX +XXX,XX @@ static void verdex_class_init(ObjectClass *oc, void *data) { MachineClass *mc = MACHINE_CLASS(oc); - mc->desc = "Gumstix Verdex (PXA270)"; + mc->desc = "Gumstix Verdex Pro XL6P COMs (PXA270)"; mc->init = verdex_init; mc->ignore_memory_transaction_failures = true; mc->default_cpu_type = ARM_CPU_TYPE_NAME("pxa270-c0"); -- 2.34.1 From: Philippe Mathieu-Daudé <f4bug@amsat.org> IEC binary prefixes ease code review: the unit is explicit. Add definitions for RAM / Flash / Flash blocksize. Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230109115316.2235-7-philmd@linaro.org Message-Id: <20200223231044.8003-3-philmd@redhat.com> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/gumstix.c | 27 ++++++++++++++------------- 1 file changed, 14 insertions(+), 13 deletions(-) diff --git a/hw/arm/gumstix.c b/hw/arm/gumstix.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/gumstix.c +++ b/hw/arm/gumstix.c @@ -XXX,XX +XXX,XX @@ */ #include "qemu/osdep.h" +#include "qemu/units.h" #include "qemu/error-report.h" #include "hw/arm/pxa.h" #include "net/net.h" @@ -XXX,XX +XXX,XX @@ #include "sysemu/qtest.h" #include "cpu.h" -static const int sector_len = 128 * 1024; +#define CONNEX_FLASH_SIZE (16 * MiB) +#define CONNEX_RAM_SIZE (64 * MiB) + +#define VERDEX_FLASH_SIZE (32 * MiB) +#define VERDEX_RAM_SIZE (256 * MiB) + +#define FLASH_SECTOR_SIZE (128 * KiB) static void connex_init(MachineState *machine) { PXA2xxState *cpu; DriveInfo *dinfo; - uint32_t connex_rom = 0x01000000; - uint32_t connex_ram = 0x04000000; - - cpu = pxa255_init(connex_ram); + cpu = pxa255_init(CONNEX_RAM_SIZE); dinfo = drive_get(IF_PFLASH, 0, 0); if (!dinfo && !qtest_enabled()) { @@ -XXX,XX +XXX,XX @@ static void connex_init(MachineState *machine) } /* Numonyx RC28F128J3F75 */ - if (!pflash_cfi01_register(0x00000000, "connext.rom", connex_rom, + if (!pflash_cfi01_register(0x00000000, "connext.rom", CONNEX_FLASH_SIZE, dinfo ? blk_by_legacy_dinfo(dinfo) : NULL, - sector_len, 2, 0, 0, 0, 0, 0)) { + FLASH_SECTOR_SIZE, 2, 0, 0, 0, 0, 0)) { error_report("Error registering flash memory"); exit(1); } @@ -XXX,XX +XXX,XX @@ static void verdex_init(MachineState *machine) PXA2xxState *cpu; DriveInfo *dinfo; - uint32_t verdex_rom = 0x02000000; - uint32_t verdex_ram = 0x10000000; - - cpu = pxa270_init(verdex_ram, machine->cpu_type); + cpu = pxa270_init(VERDEX_RAM_SIZE, machine->cpu_type); dinfo = drive_get(IF_PFLASH, 0, 0); if (!dinfo && !qtest_enabled()) { @@ -XXX,XX +XXX,XX @@ static void verdex_init(MachineState *machine) } /* Micron RC28F256P30TFA */ - if (!pflash_cfi01_register(0x00000000, "verdex.rom", verdex_rom, + if (!pflash_cfi01_register(0x00000000, "verdex.rom", VERDEX_FLASH_SIZE, dinfo ? blk_by_legacy_dinfo(dinfo) : NULL, - sector_len, 2, 0, 0, 0, 0, 0)) { + FLASH_SECTOR_SIZE, 2, 0, 0, 0, 0, 0)) { error_report("Error registering flash memory"); exit(1); } -- 2.34.1 From: Philippe Mathieu-Daudé <philmd@linaro.org> IEC binary prefixes ease code review: the unit is explicit. Add the FLASH_SECTOR_SIZE definition. Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230109115316.2235-8-philmd@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/mainstone.c | 18 ++++++++++-------- 1 file changed, 10 insertions(+), 8 deletions(-) diff --git a/hw/arm/mainstone.c b/hw/arm/mainstone.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/mainstone.c +++ b/hw/arm/mainstone.c @@ -XXX,XX +XXX,XX @@ * GNU GPL, version 2 or (at your option) any later version. */ #include "qemu/osdep.h" +#include "qemu/units.h" #include "qemu/error-report.h" #include "qapi/error.h" #include "hw/arm/pxa.h" @@ -XXX,XX +XXX,XX @@ static const struct keymap map[0xE0] = { enum mainstone_model_e { mainstone }; -#define MAINSTONE_RAM 0x04000000 -#define MAINSTONE_ROM 0x00800000 -#define MAINSTONE_FLASH 0x02000000 +#define MAINSTONE_RAM_SIZE (64 * MiB) +#define MAINSTONE_ROM_SIZE (8 * MiB) +#define MAINSTONE_FLASH_SIZE (32 * MiB) static struct arm_boot_info mainstone_binfo = { .loader_start = PXA2XX_SDRAM_BASE, - .ram_size = 0x04000000, + .ram_size = MAINSTONE_RAM_SIZE, }; +#define FLASH_SECTOR_SIZE (256 * KiB) + static void mainstone_common_init(MachineState *machine, enum mainstone_model_e model, int arm_id) { - uint32_t sector_len = 256 * 1024; hwaddr mainstone_flash_base[] = { MST_FLASH_0, MST_FLASH_1 }; PXA2xxState *mpu; DeviceState *mst_irq; @@ -XXX,XX +XXX,XX @@ static void mainstone_common_init(MachineState *machine, /* Setup CPU & memory */ mpu = pxa270_init(mainstone_binfo.ram_size, machine->cpu_type); - memory_region_init_rom(rom, NULL, "mainstone.rom", MAINSTONE_ROM, + memory_region_init_rom(rom, NULL, "mainstone.rom", MAINSTONE_ROM_SIZE, &error_fatal); memory_region_add_subregion(get_system_memory(), 0x00000000, rom); @@ -XXX,XX +XXX,XX @@ static void mainstone_common_init(MachineState *machine, dinfo = drive_get(IF_PFLASH, 0, i); if (!pflash_cfi01_register(mainstone_flash_base[i], i ? "mainstone.flash1" : "mainstone.flash0", - MAINSTONE_FLASH, + MAINSTONE_FLASH_SIZE, dinfo ? blk_by_legacy_dinfo(dinfo) : NULL, - sector_len, 4, 0, 0, 0, 0, 0)) { + FLASH_SECTOR_SIZE, 4, 0, 0, 0, 0, 0)) { error_report("Error registering flash memory"); exit(1); } -- 2.34.1 From: Philippe Mathieu-Daudé <philmd@linaro.org> IEC binary prefixes ease code review: the unit is explicit. Add the FLASH_SECTOR_SIZE definition. Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230109115316.2235-9-philmd@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/musicpal.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) diff --git a/hw/arm/musicpal.c b/hw/arm/musicpal.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/musicpal.c +++ b/hw/arm/musicpal.c @@ -XXX,XX +XXX,XX @@ */ #include "qemu/osdep.h" +#include "qemu/units.h" #include "qapi/error.h" #include "cpu.h" #include "hw/sysbus.h" @@ -XXX,XX +XXX,XX @@ static const TypeInfo musicpal_key_info = { .class_init = musicpal_key_class_init, }; +#define FLASH_SECTOR_SIZE (64 * KiB) + static struct arm_boot_info musicpal_binfo = { .loader_start = 0x0, .board_id = 0x20e, @@ -XXX,XX +XXX,XX @@ static void musicpal_init(MachineState *machine) BlockBackend *blk = blk_by_legacy_dinfo(dinfo); flash_size = blk_getlength(blk); - if (flash_size != 8*1024*1024 && flash_size != 16*1024*1024 && - flash_size != 32*1024*1024) { + if (flash_size != 8 * MiB && flash_size != 16 * MiB && + flash_size != 32 * MiB) { error_report("Invalid flash image size"); exit(1); } @@ -XXX,XX +XXX,XX @@ static void musicpal_init(MachineState *machine) */ pflash_cfi02_register(0x100000000ULL - MP_FLASH_SIZE_MAX, "musicpal.flash", flash_size, - blk, 0x10000, + blk, FLASH_SECTOR_SIZE, MP_FLASH_SIZE_MAX / flash_size, 2, 0x00BF, 0x236D, 0x0000, 0x0000, 0x5555, 0x2AAA, 0); -- 2.34.1 From: Philippe Mathieu-Daudé <philmd@linaro.org> The total_ram_v1/total_ram_v2 definitions were never used. Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230109115316.2235-10-philmd@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/omap_sx1.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/hw/arm/omap_sx1.c b/hw/arm/omap_sx1.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/omap_sx1.c +++ b/hw/arm/omap_sx1.c @@ -XXX,XX +XXX,XX @@ static const MemoryRegionOps static_ops = { #define flash0_size (16 * 1024 * 1024) #define flash1_size ( 8 * 1024 * 1024) #define flash2_size (32 * 1024 * 1024) -#define total_ram_v1 (sdram_size + flash0_size + flash1_size + OMAP15XX_SRAM_SIZE) -#define total_ram_v2 (sdram_size + flash2_size + OMAP15XX_SRAM_SIZE) static struct arm_boot_info sx1_binfo = { .loader_start = OMAP_EMIFF_BASE, -- 2.34.1 From: Philippe Mathieu-Daudé <philmd@linaro.org> IEC binary prefixes ease code review: the unit is explicit. Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230109115316.2235-11-philmd@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/omap_sx1.c | 33 +++++++++++++++++---------------- 1 file changed, 17 insertions(+), 16 deletions(-) diff --git a/hw/arm/omap_sx1.c b/hw/arm/omap_sx1.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/omap_sx1.c +++ b/hw/arm/omap_sx1.c @@ -XXX,XX +XXX,XX @@ * with this program; if not, see <http://www.gnu.org/licenses/>. */ #include "qemu/osdep.h" +#include "qemu/units.h" #include "qapi/error.h" #include "ui/console.h" #include "hw/arm/omap.h" @@ -XXX,XX +XXX,XX @@ static const MemoryRegionOps static_ops = { .endianness = DEVICE_NATIVE_ENDIAN, }; -#define sdram_size 0x02000000 -#define sector_size (128 * 1024) -#define flash0_size (16 * 1024 * 1024) -#define flash1_size ( 8 * 1024 * 1024) -#define flash2_size (32 * 1024 * 1024) +#define SDRAM_SIZE (32 * MiB) +#define SECTOR_SIZE (128 * KiB) +#define FLASH0_SIZE (16 * MiB) +#define FLASH1_SIZE (8 * MiB) +#define FLASH2_SIZE (32 * MiB) static struct arm_boot_info sx1_binfo = { .loader_start = OMAP_EMIFF_BASE, - .ram_size = sdram_size, + .ram_size = SDRAM_SIZE, .board_id = 0x265, }; @@ -XXX,XX +XXX,XX @@ static void sx1_init(MachineState *machine, const int version) static uint32_t cs3val = 0x00001139; DriveInfo *dinfo; int fl_idx; - uint32_t flash_size = flash0_size; + uint32_t flash_size = FLASH0_SIZE; if (machine->ram_size != mc->default_ram_size) { char *sz = size_to_str(mc->default_ram_size); @@ -XXX,XX +XXX,XX @@ static void sx1_init(MachineState *machine, const int version) } if (version == 2) { - flash_size = flash2_size; + flash_size = FLASH2_SIZE; } memory_region_add_subregion(address_space, OMAP_EMIFF_BASE, machine->ram); @@ -XXX,XX +XXX,XX @@ static void sx1_init(MachineState *machine, const int version) if (!pflash_cfi01_register(OMAP_CS0_BASE, "omap_sx1.flash0-1", flash_size, blk_by_legacy_dinfo(dinfo), - sector_size, 4, 0, 0, 0, 0, 0)) { + SECTOR_SIZE, 4, 0, 0, 0, 0, 0)) { fprintf(stderr, "qemu: Error registering flash memory %d.\n", fl_idx); } @@ -XXX,XX +XXX,XX @@ static void sx1_init(MachineState *machine, const int version) (dinfo = drive_get(IF_PFLASH, 0, fl_idx)) != NULL) { MemoryRegion *flash_1 = g_new(MemoryRegion, 1); memory_region_init_rom(flash_1, NULL, "omap_sx1.flash1-0", - flash1_size, &error_fatal); + FLASH1_SIZE, &error_fatal); memory_region_add_subregion(address_space, OMAP_CS1_BASE, flash_1); memory_region_init_io(&cs[1], NULL, &static_ops, &cs1val, - "sx1.cs1", OMAP_CS1_SIZE - flash1_size); + "sx1.cs1", OMAP_CS1_SIZE - FLASH1_SIZE); memory_region_add_subregion(address_space, - OMAP_CS1_BASE + flash1_size, &cs[1]); + OMAP_CS1_BASE + FLASH1_SIZE, &cs[1]); if (!pflash_cfi01_register(OMAP_CS1_BASE, - "omap_sx1.flash1-1", flash1_size, + "omap_sx1.flash1-1", FLASH1_SIZE, blk_by_legacy_dinfo(dinfo), - sector_size, 4, 0, 0, 0, 0, 0)) { + SECTOR_SIZE, 4, 0, 0, 0, 0, 0)) { fprintf(stderr, "qemu: Error registering flash memory %d.\n", fl_idx); } @@ -XXX,XX +XXX,XX @@ static void sx1_machine_v2_class_init(ObjectClass *oc, void *data) mc->init = sx1_init_v2; mc->ignore_memory_transaction_failures = true; mc->default_cpu_type = ARM_CPU_TYPE_NAME("ti925t"); - mc->default_ram_size = sdram_size; + mc->default_ram_size = SDRAM_SIZE; mc->default_ram_id = "omap1.dram"; } @@ -XXX,XX +XXX,XX @@ static void sx1_machine_v1_class_init(ObjectClass *oc, void *data) mc->init = sx1_init_v1; mc->ignore_memory_transaction_failures = true; mc->default_cpu_type = ARM_CPU_TYPE_NAME("ti925t"); - mc->default_ram_size = sdram_size; + mc->default_ram_size = SDRAM_SIZE; mc->default_ram_id = "omap1.dram"; } -- 2.34.1 From: Philippe Mathieu-Daudé <philmd@linaro.org> IEC binary prefixes ease code review: the unit is explicit. Add the FLASH_SECTOR_SIZE definition. Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230109115316.2235-12-philmd@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/z2.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/hw/arm/z2.c b/hw/arm/z2.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/z2.c +++ b/hw/arm/z2.c @@ -XXX,XX +XXX,XX @@ */ #include "qemu/osdep.h" +#include "qemu/units.h" #include "hw/arm/pxa.h" #include "hw/arm/boot.h" #include "hw/i2c/i2c.h" @@ -XXX,XX +XXX,XX @@ static const TypeInfo aer915_info = { .class_init = aer915_class_init, }; +#define FLASH_SECTOR_SIZE (64 * KiB) + static void z2_init(MachineState *machine) { - uint32_t sector_len = 0x10000; PXA2xxState *mpu; DriveInfo *dinfo; void *z2_lcd; @@ -XXX,XX +XXX,XX @@ static void z2_init(MachineState *machine) dinfo = drive_get(IF_PFLASH, 0, 0); if (!pflash_cfi01_register(Z2_FLASH_BASE, "z2.flash0", Z2_FLASH_SIZE, dinfo ? blk_by_legacy_dinfo(dinfo) : NULL, - sector_len, 4, 0, 0, 0, 0, 0)) { + FLASH_SECTOR_SIZE, 4, 0, 0, 0, 0, 0)) { error_report("Error registering flash memory"); exit(1); } -- 2.34.1 From: Philippe Mathieu-Daudé <philmd@linaro.org> Upon introduction in commit b8433303fb ("Set proper device-width for vexpress flash"), ve_pflash_cfi01_register() was calling qdev_init_nofail() which can not fail. This call was later converted with a script to use &error_fatal, still unable to fail. Remove the unreachable code. Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230109115316.2235-13-philmd@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/vexpress.c | 10 +--------- 1 file changed, 1 insertion(+), 9 deletions(-) diff --git a/hw/arm/vexpress.c b/hw/arm/vexpress.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/vexpress.c +++ b/hw/arm/vexpress.c @@ -XXX,XX +XXX,XX @@ static void vexpress_common_init(MachineState *machine) dinfo = drive_get(IF_PFLASH, 0, 0); pflash0 = ve_pflash_cfi01_register(map[VE_NORFLASH0], "vexpress.flash0", dinfo); - if (!pflash0) { - error_report("vexpress: error registering flash 0"); - exit(1); - } if (map[VE_NORFLASHALIAS] != -1) { /* Map flash 0 as an alias into low memory */ @@ -XXX,XX +XXX,XX @@ static void vexpress_common_init(MachineState *machine) } dinfo = drive_get(IF_PFLASH, 0, 1); - if (!ve_pflash_cfi01_register(map[VE_NORFLASH1], "vexpress.flash1", - dinfo)) { - error_report("vexpress: error registering flash 1"); - exit(1); - } + ve_pflash_cfi01_register(map[VE_NORFLASH1], "vexpress.flash1", dinfo); sram_size = 0x2000000; memory_region_init_ram(sram, NULL, "vexpress.sram", sram_size, -- 2.34.1 From: Philippe Mathieu-Daudé <philmd@linaro.org> Since its QOM'ification in commit 368a354f02 ("pflash_cfi0x: QOMified") the pflash_cfi01_register() function does not fail. This call was later converted with a script to use &error_fatal, still unable to fail. Remove the unreachable code. Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230109115316.2235-14-philmd@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/gumstix.c | 18 ++++++------------ hw/arm/mainstone.c | 13 +++++-------- hw/arm/omap_sx1.c | 22 ++++++++-------------- hw/arm/versatilepb.c | 6 ++---- hw/arm/z2.c | 9 +++------ 5 files changed, 24 insertions(+), 44 deletions(-) diff --git a/hw/arm/gumstix.c b/hw/arm/gumstix.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/gumstix.c +++ b/hw/arm/gumstix.c @@ -XXX,XX +XXX,XX @@ static void connex_init(MachineState *machine) } /* Numonyx RC28F128J3F75 */ - if (!pflash_cfi01_register(0x00000000, "connext.rom", CONNEX_FLASH_SIZE, - dinfo ? blk_by_legacy_dinfo(dinfo) : NULL, - FLASH_SECTOR_SIZE, 2, 0, 0, 0, 0, 0)) { - error_report("Error registering flash memory"); - exit(1); - } + pflash_cfi01_register(0x00000000, "connext.rom", CONNEX_FLASH_SIZE, + dinfo ? blk_by_legacy_dinfo(dinfo) : NULL, + FLASH_SECTOR_SIZE, 2, 0, 0, 0, 0, 0); /* Interrupt line of NIC is connected to GPIO line 36 */ smc91c111_init(&nd_table[0], 0x04000300, @@ -XXX,XX +XXX,XX @@ static void verdex_init(MachineState *machine) } /* Micron RC28F256P30TFA */ - if (!pflash_cfi01_register(0x00000000, "verdex.rom", VERDEX_FLASH_SIZE, - dinfo ? blk_by_legacy_dinfo(dinfo) : NULL, - FLASH_SECTOR_SIZE, 2, 0, 0, 0, 0, 0)) { - error_report("Error registering flash memory"); - exit(1); - } + pflash_cfi01_register(0x00000000, "verdex.rom", VERDEX_FLASH_SIZE, + dinfo ? blk_by_legacy_dinfo(dinfo) : NULL, + FLASH_SECTOR_SIZE, 2, 0, 0, 0, 0, 0); /* Interrupt line of NIC is connected to GPIO line 99 */ smc91c111_init(&nd_table[0], 0x04000300, diff --git a/hw/arm/mainstone.c b/hw/arm/mainstone.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/mainstone.c +++ b/hw/arm/mainstone.c @@ -XXX,XX +XXX,XX @@ static void mainstone_common_init(MachineState *machine, /* There are two 32MiB flash devices on the board */ for (i = 0; i < 2; i ++) { dinfo = drive_get(IF_PFLASH, 0, i); - if (!pflash_cfi01_register(mainstone_flash_base[i], - i ? "mainstone.flash1" : "mainstone.flash0", - MAINSTONE_FLASH_SIZE, - dinfo ? blk_by_legacy_dinfo(dinfo) : NULL, - FLASH_SECTOR_SIZE, 4, 0, 0, 0, 0, 0)) { - error_report("Error registering flash memory"); - exit(1); - } + pflash_cfi01_register(mainstone_flash_base[i], + i ? "mainstone.flash1" : "mainstone.flash0", + MAINSTONE_FLASH_SIZE, + dinfo ? blk_by_legacy_dinfo(dinfo) : NULL, + FLASH_SECTOR_SIZE, 4, 0, 0, 0, 0, 0); } mst_irq = sysbus_create_simple("mainstone-fpga", MST_FPGA_PHYS, diff --git a/hw/arm/omap_sx1.c b/hw/arm/omap_sx1.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/omap_sx1.c +++ b/hw/arm/omap_sx1.c @@ -XXX,XX +XXX,XX @@ static void sx1_init(MachineState *machine, const int version) fl_idx = 0; if ((dinfo = drive_get(IF_PFLASH, 0, fl_idx)) != NULL) { - if (!pflash_cfi01_register(OMAP_CS0_BASE, - "omap_sx1.flash0-1", flash_size, - blk_by_legacy_dinfo(dinfo), - SECTOR_SIZE, 4, 0, 0, 0, 0, 0)) { - fprintf(stderr, "qemu: Error registering flash memory %d.\n", - fl_idx); - } + pflash_cfi01_register(OMAP_CS0_BASE, + "omap_sx1.flash0-1", flash_size, + blk_by_legacy_dinfo(dinfo), + SECTOR_SIZE, 4, 0, 0, 0, 0, 0); fl_idx++; } @@ -XXX,XX +XXX,XX @@ static void sx1_init(MachineState *machine, const int version) memory_region_add_subregion(address_space, OMAP_CS1_BASE + FLASH1_SIZE, &cs[1]); - if (!pflash_cfi01_register(OMAP_CS1_BASE, - "omap_sx1.flash1-1", FLASH1_SIZE, - blk_by_legacy_dinfo(dinfo), - SECTOR_SIZE, 4, 0, 0, 0, 0, 0)) { - fprintf(stderr, "qemu: Error registering flash memory %d.\n", - fl_idx); - } + pflash_cfi01_register(OMAP_CS1_BASE, + "omap_sx1.flash1-1", FLASH1_SIZE, + blk_by_legacy_dinfo(dinfo), + SECTOR_SIZE, 4, 0, 0, 0, 0, 0); fl_idx++; } else { memory_region_init_io(&cs[1], NULL, &static_ops, &cs1val, diff --git a/hw/arm/versatilepb.c b/hw/arm/versatilepb.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/versatilepb.c +++ b/hw/arm/versatilepb.c @@ -XXX,XX +XXX,XX @@ static void versatile_init(MachineState *machine, int board_id) /* 0x34000000 NOR Flash */ dinfo = drive_get(IF_PFLASH, 0, 0); - if (!pflash_cfi01_register(VERSATILE_FLASH_ADDR, "versatile.flash", + pflash_cfi01_register(VERSATILE_FLASH_ADDR, "versatile.flash", VERSATILE_FLASH_SIZE, dinfo ? blk_by_legacy_dinfo(dinfo) : NULL, VERSATILE_FLASH_SECT_SIZE, - 4, 0x0089, 0x0018, 0x0000, 0x0, 0)) { - fprintf(stderr, "qemu: Error registering flash memory.\n"); - } + 4, 0x0089, 0x0018, 0x0000, 0x0, 0); versatile_binfo.ram_size = machine->ram_size; versatile_binfo.board_id = board_id; diff --git a/hw/arm/z2.c b/hw/arm/z2.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/z2.c +++ b/hw/arm/z2.c @@ -XXX,XX +XXX,XX @@ static void z2_init(MachineState *machine) mpu = pxa270_init(z2_binfo.ram_size, machine->cpu_type); dinfo = drive_get(IF_PFLASH, 0, 0); - if (!pflash_cfi01_register(Z2_FLASH_BASE, "z2.flash0", Z2_FLASH_SIZE, - dinfo ? blk_by_legacy_dinfo(dinfo) : NULL, - FLASH_SECTOR_SIZE, 4, 0, 0, 0, 0, 0)) { - error_report("Error registering flash memory"); - exit(1); - } + pflash_cfi01_register(Z2_FLASH_BASE, "z2.flash0", Z2_FLASH_SIZE, + dinfo ? blk_by_legacy_dinfo(dinfo) : NULL, + FLASH_SECTOR_SIZE, 4, 0, 0, 0, 0, 0); /* setup keypad */ pxa27x_register_keypad(mpu->kp, map, 0x100); -- 2.34.1 From: Philippe Mathieu-Daudé <philmd@linaro.org> To avoid forward-declaring PXA2xxI2CState, declare PXA2XX_I2C before its use in pxa2xx_i2c_init() prototype. Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230109140306.23161-2-philmd@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/pxa.h | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/include/hw/arm/pxa.h b/include/hw/arm/pxa.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/pxa.h +++ b/include/hw/arm/pxa.h @@ -XXX,XX +XXX,XX @@ void pxa27x_register_keypad(PXA2xxKeyPadState *kp, const struct keymap *map, int size); /* pxa2xx.c */ -typedef struct PXA2xxI2CState PXA2xxI2CState; +#define TYPE_PXA2XX_I2C "pxa2xx_i2c" +OBJECT_DECLARE_SIMPLE_TYPE(PXA2xxI2CState, PXA2XX_I2C) + PXA2xxI2CState *pxa2xx_i2c_init(hwaddr base, qemu_irq irq, uint32_t page_size); I2CBus *pxa2xx_i2c_bus(PXA2xxI2CState *s); -#define TYPE_PXA2XX_I2C "pxa2xx_i2c" typedef struct PXA2xxI2SState PXA2xxI2SState; -OBJECT_DECLARE_SIMPLE_TYPE(PXA2xxI2CState, PXA2XX_I2C) #define TYPE_PXA2XX_FIR "pxa2xx-fir" OBJECT_DECLARE_SIMPLE_TYPE(PXA2xxFIrState, PXA2XX_FIR) -- 2.34.1 From: Philippe Mathieu-Daudé <philmd@linaro.org> Add a local 'struct omap_gpif_s *' variable to improve readability. (This also eases next commit conversion). Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230109140306.23161-3-philmd@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/gpio/omap_gpio.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/hw/gpio/omap_gpio.c b/hw/gpio/omap_gpio.c index XXXXXXX..XXXXXXX 100644 --- a/hw/gpio/omap_gpio.c +++ b/hw/gpio/omap_gpio.c @@ -XXX,XX +XXX,XX @@ struct omap_gpif_s { /* General-Purpose I/O of OMAP1 */ static void omap_gpio_set(void *opaque, int line, int level) { - struct omap_gpio_s *s = &((struct omap_gpif_s *) opaque)->omap1; + struct omap_gpif_s *p = opaque; + struct omap_gpio_s *s = &p->omap1; uint16_t prev = s->inputs; if (level) -- 2.34.1 From: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230109140306.23161-4-philmd@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/omap1.c | 115 ++++++++++++++++++-------------------- hw/arm/omap2.c | 40 ++++++------- hw/arm/omap_sx1.c | 2 +- hw/arm/palm.c | 2 +- hw/char/omap_uart.c | 7 +-- hw/display/omap_dss.c | 15 +++-- hw/display/omap_lcdc.c | 9 ++- hw/dma/omap_dma.c | 15 +++-- hw/gpio/omap_gpio.c | 15 +++-- hw/intc/omap_intc.c | 12 ++-- hw/misc/omap_gpmc.c | 12 ++-- hw/misc/omap_l4.c | 7 +-- hw/misc/omap_sdrc.c | 7 +-- hw/misc/omap_tap.c | 5 +- hw/sd/omap_mmc.c | 9 ++- hw/ssi/omap_spi.c | 7 +-- hw/timer/omap_gptimer.c | 22 ++++---- hw/timer/omap_synctimer.c | 4 +- 18 files changed, 142 insertions(+), 163 deletions(-) diff --git a/hw/arm/omap1.c b/hw/arm/omap1.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/omap1.c +++ b/hw/arm/omap1.c @@ -XXX,XX +XXX,XX @@ static void omap_timer_fire(void *opaque) static void omap_timer_tick(void *opaque) { - struct omap_mpu_timer_s *timer = (struct omap_mpu_timer_s *) opaque; + struct omap_mpu_timer_s *timer = opaque; omap_timer_sync(timer); omap_timer_fire(timer); @@ -XXX,XX +XXX,XX @@ static void omap_timer_tick(void *opaque) static void omap_timer_clk_update(void *opaque, int line, int on) { - struct omap_mpu_timer_s *timer = (struct omap_mpu_timer_s *) opaque; + struct omap_mpu_timer_s *timer = opaque; omap_timer_sync(timer); timer->rate = on ? omap_clk_getrate(timer->clk) : 0; @@ -XXX,XX +XXX,XX @@ static void omap_timer_clk_setup(struct omap_mpu_timer_s *timer) static uint64_t omap_mpu_timer_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_mpu_timer_s *s = (struct omap_mpu_timer_s *) opaque; + struct omap_mpu_timer_s *s = opaque; if (size != 4) { return omap_badwidth_read32(opaque, addr); @@ -XXX,XX +XXX,XX @@ static uint64_t omap_mpu_timer_read(void *opaque, hwaddr addr, static void omap_mpu_timer_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_mpu_timer_s *s = (struct omap_mpu_timer_s *) opaque; + struct omap_mpu_timer_s *s = opaque; if (size != 4) { omap_badwidth_write32(opaque, addr, value); @@ -XXX,XX +XXX,XX @@ struct omap_watchdog_timer_s { static uint64_t omap_wd_timer_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_watchdog_timer_s *s = (struct omap_watchdog_timer_s *) opaque; + struct omap_watchdog_timer_s *s = opaque; if (size != 2) { return omap_badwidth_read16(opaque, addr); @@ -XXX,XX +XXX,XX @@ static uint64_t omap_wd_timer_read(void *opaque, hwaddr addr, static void omap_wd_timer_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_watchdog_timer_s *s = (struct omap_watchdog_timer_s *) opaque; + struct omap_watchdog_timer_s *s = opaque; if (size != 2) { omap_badwidth_write16(opaque, addr, value); @@ -XXX,XX +XXX,XX @@ struct omap_32khz_timer_s { static uint64_t omap_os_timer_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_32khz_timer_s *s = (struct omap_32khz_timer_s *) opaque; + struct omap_32khz_timer_s *s = opaque; int offset = addr & OMAP_MPUI_REG_MASK; if (size != 4) { @@ -XXX,XX +XXX,XX @@ static uint64_t omap_os_timer_read(void *opaque, hwaddr addr, static void omap_os_timer_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_32khz_timer_s *s = (struct omap_32khz_timer_s *) opaque; + struct omap_32khz_timer_s *s = opaque; int offset = addr & OMAP_MPUI_REG_MASK; if (size != 4) { @@ -XXX,XX +XXX,XX @@ static struct omap_32khz_timer_s *omap_os_timer_init(MemoryRegion *memory, static uint64_t omap_ulpd_pm_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_mpu_state_s *s = (struct omap_mpu_state_s *) opaque; + struct omap_mpu_state_s *s = opaque; uint16_t ret; if (size != 2) { @@ -XXX,XX +XXX,XX @@ static inline void omap_ulpd_req_update(struct omap_mpu_state_s *s, static void omap_ulpd_pm_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_mpu_state_s *s = (struct omap_mpu_state_s *) opaque; + struct omap_mpu_state_s *s = opaque; int64_t now, ticks; int div, mult; static const int bypass_div[4] = { 1, 2, 4, 4 }; @@ -XXX,XX +XXX,XX @@ static void omap_ulpd_pm_init(MemoryRegion *system_memory, static uint64_t omap_pin_cfg_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_mpu_state_s *s = (struct omap_mpu_state_s *) opaque; + struct omap_mpu_state_s *s = opaque; if (size != 4) { return omap_badwidth_read32(opaque, addr); @@ -XXX,XX +XXX,XX @@ static inline void omap_pin_modconf1_update(struct omap_mpu_state_s *s, static void omap_pin_cfg_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_mpu_state_s *s = (struct omap_mpu_state_s *) opaque; + struct omap_mpu_state_s *s = opaque; uint32_t diff; if (size != 4) { @@ -XXX,XX +XXX,XX @@ static void omap_pin_cfg_init(MemoryRegion *system_memory, static uint64_t omap_id_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_mpu_state_s *s = (struct omap_mpu_state_s *) opaque; + struct omap_mpu_state_s *s = opaque; if (size != 4) { return omap_badwidth_read32(opaque, addr); @@ -XXX,XX +XXX,XX @@ static void omap_id_init(MemoryRegion *memory, struct omap_mpu_state_s *mpu) static uint64_t omap_mpui_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_mpu_state_s *s = (struct omap_mpu_state_s *) opaque; + struct omap_mpu_state_s *s = opaque; if (size != 4) { return omap_badwidth_read32(opaque, addr); @@ -XXX,XX +XXX,XX @@ static uint64_t omap_mpui_read(void *opaque, hwaddr addr, static void omap_mpui_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_mpu_state_s *s = (struct omap_mpu_state_s *) opaque; + struct omap_mpu_state_s *s = opaque; if (size != 4) { omap_badwidth_write32(opaque, addr, value); @@ -XXX,XX +XXX,XX @@ struct omap_tipb_bridge_s { static uint64_t omap_tipb_bridge_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_tipb_bridge_s *s = (struct omap_tipb_bridge_s *) opaque; + struct omap_tipb_bridge_s *s = opaque; if (size < 2) { return omap_badwidth_read16(opaque, addr); @@ -XXX,XX +XXX,XX @@ static uint64_t omap_tipb_bridge_read(void *opaque, hwaddr addr, static void omap_tipb_bridge_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_tipb_bridge_s *s = (struct omap_tipb_bridge_s *) opaque; + struct omap_tipb_bridge_s *s = opaque; if (size < 2) { omap_badwidth_write16(opaque, addr, value); @@ -XXX,XX +XXX,XX @@ static struct omap_tipb_bridge_s *omap_tipb_bridge_init( static uint64_t omap_tcmi_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_mpu_state_s *s = (struct omap_mpu_state_s *) opaque; + struct omap_mpu_state_s *s = opaque; uint32_t ret; if (size != 4) { @@ -XXX,XX +XXX,XX @@ static uint64_t omap_tcmi_read(void *opaque, hwaddr addr, static void omap_tcmi_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_mpu_state_s *s = (struct omap_mpu_state_s *) opaque; + struct omap_mpu_state_s *s = opaque; if (size != 4) { omap_badwidth_write32(opaque, addr, value); @@ -XXX,XX +XXX,XX @@ struct dpll_ctl_s { static uint64_t omap_dpll_read(void *opaque, hwaddr addr, unsigned size) { - struct dpll_ctl_s *s = (struct dpll_ctl_s *) opaque; + struct dpll_ctl_s *s = opaque; if (size != 2) { return omap_badwidth_read16(opaque, addr); @@ -XXX,XX +XXX,XX @@ static uint64_t omap_dpll_read(void *opaque, hwaddr addr, static void omap_dpll_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct dpll_ctl_s *s = (struct dpll_ctl_s *) opaque; + struct dpll_ctl_s *s = opaque; uint16_t diff; static const int bypass_div[4] = { 1, 2, 4, 4 }; int div, mult; @@ -XXX,XX +XXX,XX @@ static struct dpll_ctl_s *omap_dpll_init(MemoryRegion *memory, static uint64_t omap_clkm_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_mpu_state_s *s = (struct omap_mpu_state_s *) opaque; + struct omap_mpu_state_s *s = opaque; if (size != 2) { return omap_badwidth_read16(opaque, addr); @@ -XXX,XX +XXX,XX @@ static inline void omap_clkm_ckout1_update(struct omap_mpu_state_s *s, static void omap_clkm_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_mpu_state_s *s = (struct omap_mpu_state_s *) opaque; + struct omap_mpu_state_s *s = opaque; uint16_t diff; omap_clk clk; static const char *clkschemename[8] = { @@ -XXX,XX +XXX,XX @@ static const MemoryRegionOps omap_clkm_ops = { static uint64_t omap_clkdsp_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_mpu_state_s *s = (struct omap_mpu_state_s *) opaque; + struct omap_mpu_state_s *s = opaque; CPUState *cpu = CPU(s->cpu); if (size != 2) { @@ -XXX,XX +XXX,XX @@ static inline void omap_clkdsp_idlect2_update(struct omap_mpu_state_s *s, static void omap_clkdsp_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_mpu_state_s *s = (struct omap_mpu_state_s *) opaque; + struct omap_mpu_state_s *s = opaque; uint16_t diff; if (size != 2) { @@ -XXX,XX +XXX,XX @@ struct omap_mpuio_s { static void omap_mpuio_set(void *opaque, int line, int level) { - struct omap_mpuio_s *s = (struct omap_mpuio_s *) opaque; + struct omap_mpuio_s *s = opaque; uint16_t prev = s->inputs; if (level) @@ -XXX,XX +XXX,XX @@ static void omap_mpuio_kbd_update(struct omap_mpuio_s *s) static uint64_t omap_mpuio_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_mpuio_s *s = (struct omap_mpuio_s *) opaque; + struct omap_mpuio_s *s = opaque; int offset = addr & OMAP_MPUI_REG_MASK; uint16_t ret; @@ -XXX,XX +XXX,XX @@ static uint64_t omap_mpuio_read(void *opaque, hwaddr addr, static void omap_mpuio_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_mpuio_s *s = (struct omap_mpuio_s *) opaque; + struct omap_mpuio_s *s = opaque; int offset = addr & OMAP_MPUI_REG_MASK; uint16_t diff; int ln; @@ -XXX,XX +XXX,XX @@ static void omap_mpuio_reset(struct omap_mpuio_s *s) static void omap_mpuio_onoff(void *opaque, int line, int on) { - struct omap_mpuio_s *s = (struct omap_mpuio_s *) opaque; + struct omap_mpuio_s *s = opaque; s->clk = on; if (on) @@ -XXX,XX +XXX,XX @@ static void omap_uwire_transfer_start(struct omap_uwire_s *s) } } -static uint64_t omap_uwire_read(void *opaque, hwaddr addr, - unsigned size) +static uint64_t omap_uwire_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_uwire_s *s = (struct omap_uwire_s *) opaque; + struct omap_uwire_s *s = opaque; int offset = addr & OMAP_MPUI_REG_MASK; if (size != 2) { @@ -XXX,XX +XXX,XX @@ static uint64_t omap_uwire_read(void *opaque, hwaddr addr, static void omap_uwire_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_uwire_s *s = (struct omap_uwire_s *) opaque; + struct omap_uwire_s *s = opaque; int offset = addr & OMAP_MPUI_REG_MASK; if (size != 2) { @@ -XXX,XX +XXX,XX @@ static void omap_pwl_update(struct omap_pwl_s *s) } } -static uint64_t omap_pwl_read(void *opaque, hwaddr addr, - unsigned size) +static uint64_t omap_pwl_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_pwl_s *s = (struct omap_pwl_s *) opaque; + struct omap_pwl_s *s = opaque; int offset = addr & OMAP_MPUI_REG_MASK; if (size != 1) { @@ -XXX,XX +XXX,XX @@ static uint64_t omap_pwl_read(void *opaque, hwaddr addr, static void omap_pwl_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_pwl_s *s = (struct omap_pwl_s *) opaque; + struct omap_pwl_s *s = opaque; int offset = addr & OMAP_MPUI_REG_MASK; if (size != 1) { @@ -XXX,XX +XXX,XX @@ static void omap_pwl_reset(struct omap_pwl_s *s) static void omap_pwl_clk_update(void *opaque, int line, int on) { - struct omap_pwl_s *s = (struct omap_pwl_s *) opaque; + struct omap_pwl_s *s = opaque; s->clk = on; omap_pwl_update(s); @@ -XXX,XX +XXX,XX @@ struct omap_pwt_s { omap_clk clk; }; -static uint64_t omap_pwt_read(void *opaque, hwaddr addr, - unsigned size) +static uint64_t omap_pwt_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_pwt_s *s = (struct omap_pwt_s *) opaque; + struct omap_pwt_s *s = opaque; int offset = addr & OMAP_MPUI_REG_MASK; if (size != 1) { @@ -XXX,XX +XXX,XX @@ static uint64_t omap_pwt_read(void *opaque, hwaddr addr, static void omap_pwt_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_pwt_s *s = (struct omap_pwt_s *) opaque; + struct omap_pwt_s *s = opaque; int offset = addr & OMAP_MPUI_REG_MASK; if (size != 1) { @@ -XXX,XX +XXX,XX @@ static void omap_rtc_alarm_update(struct omap_rtc_s *s) printf("%s: conversion failed\n", __func__); } -static uint64_t omap_rtc_read(void *opaque, hwaddr addr, - unsigned size) +static uint64_t omap_rtc_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_rtc_s *s = (struct omap_rtc_s *) opaque; + struct omap_rtc_s *s = opaque; int offset = addr & OMAP_MPUI_REG_MASK; uint8_t i; @@ -XXX,XX +XXX,XX @@ static uint64_t omap_rtc_read(void *opaque, hwaddr addr, static void omap_rtc_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_rtc_s *s = (struct omap_rtc_s *) opaque; + struct omap_rtc_s *s = opaque; int offset = addr & OMAP_MPUI_REG_MASK; struct tm new_tm; time_t ti[2]; @@ -XXX,XX +XXX,XX @@ static void omap_mcbsp_rx_newdata(struct omap_mcbsp_s *s) static void omap_mcbsp_source_tick(void *opaque) { - struct omap_mcbsp_s *s = (struct omap_mcbsp_s *) opaque; + struct omap_mcbsp_s *s = opaque; static const int bps[8] = { 0, 1, 1, 2, 2, 2, -255, -255 }; if (!s->rx_rate) @@ -XXX,XX +XXX,XX @@ static void omap_mcbsp_tx_newdata(struct omap_mcbsp_s *s) static void omap_mcbsp_sink_tick(void *opaque) { - struct omap_mcbsp_s *s = (struct omap_mcbsp_s *) opaque; + struct omap_mcbsp_s *s = opaque; static const int bps[8] = { 0, 1, 1, 2, 2, 2, -255, -255 }; if (!s->tx_rate) @@ -XXX,XX +XXX,XX @@ static void omap_mcbsp_req_update(struct omap_mcbsp_s *s) static uint64_t omap_mcbsp_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_mcbsp_s *s = (struct omap_mcbsp_s *) opaque; + struct omap_mcbsp_s *s = opaque; int offset = addr & OMAP_MPUI_REG_MASK; uint16_t ret; @@ -XXX,XX +XXX,XX @@ static uint64_t omap_mcbsp_read(void *opaque, hwaddr addr, static void omap_mcbsp_writeh(void *opaque, hwaddr addr, uint32_t value) { - struct omap_mcbsp_s *s = (struct omap_mcbsp_s *) opaque; + struct omap_mcbsp_s *s = opaque; int offset = addr & OMAP_MPUI_REG_MASK; switch (offset) { @@ -XXX,XX +XXX,XX @@ static void omap_mcbsp_writeh(void *opaque, hwaddr addr, static void omap_mcbsp_writew(void *opaque, hwaddr addr, uint32_t value) { - struct omap_mcbsp_s *s = (struct omap_mcbsp_s *) opaque; + struct omap_mcbsp_s *s = opaque; int offset = addr & OMAP_MPUI_REG_MASK; if (offset == 0x04) { /* DXR */ @@ -XXX,XX +XXX,XX @@ static struct omap_mcbsp_s *omap_mcbsp_init(MemoryRegion *system_memory, static void omap_mcbsp_i2s_swallow(void *opaque, int line, int level) { - struct omap_mcbsp_s *s = (struct omap_mcbsp_s *) opaque; + struct omap_mcbsp_s *s = opaque; if (s->rx_rate) { s->rx_req = s->codec->in.len; @@ -XXX,XX +XXX,XX @@ static void omap_mcbsp_i2s_swallow(void *opaque, int line, int level) static void omap_mcbsp_i2s_start(void *opaque, int line, int level) { - struct omap_mcbsp_s *s = (struct omap_mcbsp_s *) opaque; + struct omap_mcbsp_s *s = opaque; if (s->tx_rate) { s->tx_req = s->codec->out.size; @@ -XXX,XX +XXX,XX @@ static void omap_lpg_reset(struct omap_lpg_s *s) omap_lpg_update(s); } -static uint64_t omap_lpg_read(void *opaque, hwaddr addr, - unsigned size) +static uint64_t omap_lpg_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_lpg_s *s = (struct omap_lpg_s *) opaque; + struct omap_lpg_s *s = opaque; int offset = addr & OMAP_MPUI_REG_MASK; if (size != 1) { @@ -XXX,XX +XXX,XX @@ static uint64_t omap_lpg_read(void *opaque, hwaddr addr, static void omap_lpg_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_lpg_s *s = (struct omap_lpg_s *) opaque; + struct omap_lpg_s *s = opaque; int offset = addr & OMAP_MPUI_REG_MASK; if (size != 1) { @@ -XXX,XX +XXX,XX @@ static const MemoryRegionOps omap_lpg_ops = { static void omap_lpg_clk_update(void *opaque, int line, int on) { - struct omap_lpg_s *s = (struct omap_lpg_s *) opaque; + struct omap_lpg_s *s = opaque; s->clk = on; omap_lpg_update(s); @@ -XXX,XX +XXX,XX @@ static void omap_setup_mpui_io(MemoryRegion *system_memory, /* General chip reset */ static void omap1_mpu_reset(void *opaque) { - struct omap_mpu_state_s *mpu = (struct omap_mpu_state_s *) opaque; + struct omap_mpu_state_s *mpu = opaque; omap_dma_reset(mpu->dma); omap_mpu_timer_reset(mpu->timer[0]); @@ -XXX,XX +XXX,XX @@ static void omap_setup_dsp_mapping(MemoryRegion *system_memory, void omap_mpu_wakeup(void *opaque, int irq, int req) { - struct omap_mpu_state_s *mpu = (struct omap_mpu_state_s *) opaque; + struct omap_mpu_state_s *mpu = opaque; CPUState *cpu = CPU(mpu->cpu); if (cpu->halted) { diff --git a/hw/arm/omap2.c b/hw/arm/omap2.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/omap2.c +++ b/hw/arm/omap2.c @@ -XXX,XX +XXX,XX @@ static inline void omap_eac_out_empty(struct omap_eac_s *s) static void omap_eac_in_cb(void *opaque, int avail_b) { - struct omap_eac_s *s = (struct omap_eac_s *) opaque; + struct omap_eac_s *s = opaque; s->codec.rxavail = avail_b >> 2; omap_eac_in_refill(s); @@ -XXX,XX +XXX,XX @@ static void omap_eac_in_cb(void *opaque, int avail_b) static void omap_eac_out_cb(void *opaque, int free_b) { - struct omap_eac_s *s = (struct omap_eac_s *) opaque; + struct omap_eac_s *s = opaque; s->codec.txavail = free_b >> 2; if (s->codec.txlen) @@ -XXX,XX +XXX,XX @@ static void omap_eac_reset(struct omap_eac_s *s) omap_eac_interrupt_update(s); } -static uint64_t omap_eac_read(void *opaque, hwaddr addr, - unsigned size) +static uint64_t omap_eac_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_eac_s *s = (struct omap_eac_s *) opaque; + struct omap_eac_s *s = opaque; uint32_t ret; if (size != 2) { @@ -XXX,XX +XXX,XX @@ static uint64_t omap_eac_read(void *opaque, hwaddr addr, static void omap_eac_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_eac_s *s = (struct omap_eac_s *) opaque; + struct omap_eac_s *s = opaque; if (size != 2) { omap_badwidth_write16(opaque, addr, value); @@ -XXX,XX +XXX,XX @@ static void omap_sti_reset(struct omap_sti_s *s) static uint64_t omap_sti_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_sti_s *s = (struct omap_sti_s *) opaque; + struct omap_sti_s *s = opaque; if (size != 4) { return omap_badwidth_read32(opaque, addr); @@ -XXX,XX +XXX,XX @@ static uint64_t omap_sti_read(void *opaque, hwaddr addr, static void omap_sti_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_sti_s *s = (struct omap_sti_s *) opaque; + struct omap_sti_s *s = opaque; if (size != 4) { omap_badwidth_write32(opaque, addr, value); @@ -XXX,XX +XXX,XX @@ static const MemoryRegionOps omap_sti_ops = { .endianness = DEVICE_NATIVE_ENDIAN, }; -static uint64_t omap_sti_fifo_read(void *opaque, hwaddr addr, - unsigned size) +static uint64_t omap_sti_fifo_read(void *opaque, hwaddr addr, unsigned size) { OMAP_BAD_REG(addr); return 0; @@ -XXX,XX +XXX,XX @@ static uint64_t omap_sti_fifo_read(void *opaque, hwaddr addr, static void omap_sti_fifo_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_sti_s *s = (struct omap_sti_s *) opaque; + struct omap_sti_s *s = opaque; int ch = addr >> 6; uint8_t byte = value; @@ -XXX,XX +XXX,XX @@ static void omap_prcm_int_update(struct omap_prcm_s *s, int dom) static uint64_t omap_prcm_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_prcm_s *s = (struct omap_prcm_s *) opaque; + struct omap_prcm_s *s = opaque; uint32_t ret; if (size != 4) { @@ -XXX,XX +XXX,XX @@ static void omap_prcm_dpll_update(struct omap_prcm_s *s) static void omap_prcm_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_prcm_s *s = (struct omap_prcm_s *) opaque; + struct omap_prcm_s *s = opaque; if (size != 4) { omap_badwidth_write32(opaque, addr, value); @@ -XXX,XX +XXX,XX @@ struct omap_sysctl_s { static uint32_t omap_sysctl_read8(void *opaque, hwaddr addr) { - struct omap_sysctl_s *s = (struct omap_sysctl_s *) opaque; + struct omap_sysctl_s *s = opaque; int pad_offset, byte_offset; int value; @@ -XXX,XX +XXX,XX @@ static uint32_t omap_sysctl_read8(void *opaque, hwaddr addr) static uint32_t omap_sysctl_read(void *opaque, hwaddr addr) { - struct omap_sysctl_s *s = (struct omap_sysctl_s *) opaque; + struct omap_sysctl_s *s = opaque; switch (addr) { case 0x000: /* CONTROL_REVISION */ @@ -XXX,XX +XXX,XX @@ static uint32_t omap_sysctl_read(void *opaque, hwaddr addr) return 0; } -static void omap_sysctl_write8(void *opaque, hwaddr addr, - uint32_t value) +static void omap_sysctl_write8(void *opaque, hwaddr addr, uint32_t value) { - struct omap_sysctl_s *s = (struct omap_sysctl_s *) opaque; + struct omap_sysctl_s *s = opaque; int pad_offset, byte_offset; int prev_value; @@ -XXX,XX +XXX,XX @@ static void omap_sysctl_write8(void *opaque, hwaddr addr, } } -static void omap_sysctl_write(void *opaque, hwaddr addr, - uint32_t value) +static void omap_sysctl_write(void *opaque, hwaddr addr, uint32_t value) { - struct omap_sysctl_s *s = (struct omap_sysctl_s *) opaque; + struct omap_sysctl_s *s = opaque; switch (addr) { case 0x000: /* CONTROL_REVISION */ @@ -XXX,XX +XXX,XX @@ static struct omap_sysctl_s *omap_sysctl_init(struct omap_target_agent_s *ta, /* General chip reset */ static void omap2_mpu_reset(void *opaque) { - struct omap_mpu_state_s *mpu = (struct omap_mpu_state_s *) opaque; + struct omap_mpu_state_s *mpu = opaque; omap_dma_reset(mpu->dma); omap_prcm_reset(mpu->prcm); diff --git a/hw/arm/omap_sx1.c b/hw/arm/omap_sx1.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/omap_sx1.c +++ b/hw/arm/omap_sx1.c @@ -XXX,XX +XXX,XX @@ static uint64_t static_read(void *opaque, hwaddr offset, unsigned size) { - uint32_t *val = (uint32_t *) opaque; + uint32_t *val = opaque; uint32_t mask = (4 / size) - 1; return *val >> ((offset & mask) << 3); diff --git a/hw/arm/palm.c b/hw/arm/palm.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/palm.c +++ b/hw/arm/palm.c @@ -XXX,XX +XXX,XX @@ static struct { static void palmte_button_event(void *opaque, int keycode) { - struct omap_mpu_state_s *cpu = (struct omap_mpu_state_s *) opaque; + struct omap_mpu_state_s *cpu = opaque; if (palmte_keymap[keycode & 0x7f].row != -1) omap_mpuio_key(cpu->mpuio, diff --git a/hw/char/omap_uart.c b/hw/char/omap_uart.c index XXXXXXX..XXXXXXX 100644 --- a/hw/char/omap_uart.c +++ b/hw/char/omap_uart.c @@ -XXX,XX +XXX,XX @@ struct omap_uart_s *omap_uart_init(hwaddr base, return s; } -static uint64_t omap_uart_read(void *opaque, hwaddr addr, - unsigned size) +static uint64_t omap_uart_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_uart_s *s = (struct omap_uart_s *) opaque; + struct omap_uart_s *s = opaque; if (size == 4) { return omap_badwidth_read8(opaque, addr); @@ -XXX,XX +XXX,XX @@ static uint64_t omap_uart_read(void *opaque, hwaddr addr, static void omap_uart_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_uart_s *s = (struct omap_uart_s *) opaque; + struct omap_uart_s *s = opaque; if (size == 4) { omap_badwidth_write8(opaque, addr, value); diff --git a/hw/display/omap_dss.c b/hw/display/omap_dss.c index XXXXXXX..XXXXXXX 100644 --- a/hw/display/omap_dss.c +++ b/hw/display/omap_dss.c @@ -XXX,XX +XXX,XX @@ void omap_dss_reset(struct omap_dss_s *s) static uint64_t omap_diss_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_dss_s *s = (struct omap_dss_s *) opaque; + struct omap_dss_s *s = opaque; if (size != 4) { return omap_badwidth_read32(opaque, addr); @@ -XXX,XX +XXX,XX @@ static uint64_t omap_diss_read(void *opaque, hwaddr addr, static void omap_diss_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_dss_s *s = (struct omap_dss_s *) opaque; + struct omap_dss_s *s = opaque; if (size != 4) { omap_badwidth_write32(opaque, addr, value); @@ -XXX,XX +XXX,XX @@ static const MemoryRegionOps omap_diss_ops = { static uint64_t omap_disc_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_dss_s *s = (struct omap_dss_s *) opaque; + struct omap_dss_s *s = opaque; if (size != 4) { return omap_badwidth_read32(opaque, addr); @@ -XXX,XX +XXX,XX @@ static uint64_t omap_disc_read(void *opaque, hwaddr addr, static void omap_disc_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_dss_s *s = (struct omap_dss_s *) opaque; + struct omap_dss_s *s = opaque; if (size != 4) { omap_badwidth_write32(opaque, addr, value); @@ -XXX,XX +XXX,XX @@ static void omap_rfbi_transfer_start(struct omap_dss_s *s) omap_dispc_interrupt_update(s); } -static uint64_t omap_rfbi_read(void *opaque, hwaddr addr, - unsigned size) +static uint64_t omap_rfbi_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_dss_s *s = (struct omap_dss_s *) opaque; + struct omap_dss_s *s = opaque; if (size != 4) { return omap_badwidth_read32(opaque, addr); @@ -XXX,XX +XXX,XX @@ static uint64_t omap_rfbi_read(void *opaque, hwaddr addr, static void omap_rfbi_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_dss_s *s = (struct omap_dss_s *) opaque; + struct omap_dss_s *s = opaque; if (size != 4) { omap_badwidth_write32(opaque, addr, value); diff --git a/hw/display/omap_lcdc.c b/hw/display/omap_lcdc.c index XXXXXXX..XXXXXXX 100644 --- a/hw/display/omap_lcdc.c +++ b/hw/display/omap_lcdc.c @@ -XXX,XX +XXX,XX @@ static void draw_line16_32(void *opaque, uint8_t *d, const uint8_t *s, static void omap_update_display(void *opaque) { - struct omap_lcd_panel_s *omap_lcd = (struct omap_lcd_panel_s *) opaque; + struct omap_lcd_panel_s *omap_lcd = opaque; DisplaySurface *surface; drawfn draw_line; int size, height, first, last; @@ -XXX,XX +XXX,XX @@ static void omap_lcd_update(struct omap_lcd_panel_s *s) { } } -static uint64_t omap_lcdc_read(void *opaque, hwaddr addr, - unsigned size) +static uint64_t omap_lcdc_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_lcd_panel_s *s = (struct omap_lcd_panel_s *) opaque; + struct omap_lcd_panel_s *s = opaque; switch (addr) { case 0x00: /* LCD_CONTROL */ @@ -XXX,XX +XXX,XX @@ static uint64_t omap_lcdc_read(void *opaque, hwaddr addr, static void omap_lcdc_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_lcd_panel_s *s = (struct omap_lcd_panel_s *) opaque; + struct omap_lcd_panel_s *s = opaque; switch (addr) { case 0x00: /* LCD_CONTROL */ diff --git a/hw/dma/omap_dma.c b/hw/dma/omap_dma.c index XXXXXXX..XXXXXXX 100644 --- a/hw/dma/omap_dma.c +++ b/hw/dma/omap_dma.c @@ -XXX,XX +XXX,XX @@ static int omap_dma_sys_read(struct omap_dma_s *s, int offset, return 0; } -static uint64_t omap_dma_read(void *opaque, hwaddr addr, - unsigned size) +static uint64_t omap_dma_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_dma_s *s = (struct omap_dma_s *) opaque; + struct omap_dma_s *s = opaque; int reg, ch; uint16_t ret; @@ -XXX,XX +XXX,XX @@ static uint64_t omap_dma_read(void *opaque, hwaddr addr, static void omap_dma_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_dma_s *s = (struct omap_dma_s *) opaque; + struct omap_dma_s *s = opaque; int reg, ch; if (size != 2) { @@ -XXX,XX +XXX,XX @@ static const MemoryRegionOps omap_dma_ops = { static void omap_dma_request(void *opaque, int drq, int req) { - struct omap_dma_s *s = (struct omap_dma_s *) opaque; + struct omap_dma_s *s = opaque; /* The request pins are level triggered in QEMU. */ if (req) { if (~s->dma->drqbmp & (1ULL << drq)) { @@ -XXX,XX +XXX,XX @@ static void omap_dma_request(void *opaque, int drq, int req) /* XXX: this won't be needed once soc_dma knows about clocks. */ static void omap_dma_clk_update(void *opaque, int line, int on) { - struct omap_dma_s *s = (struct omap_dma_s *) opaque; + struct omap_dma_s *s = opaque; int i; s->dma->freq = omap_clk_getrate(s->clk); @@ -XXX,XX +XXX,XX @@ static void omap_dma_interrupts_4_update(struct omap_dma_s *s) static uint64_t omap_dma4_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_dma_s *s = (struct omap_dma_s *) opaque; + struct omap_dma_s *s = opaque; int irqn = 0, chnum; struct omap_dma_channel_s *ch; @@ -XXX,XX +XXX,XX @@ static uint64_t omap_dma4_read(void *opaque, hwaddr addr, static void omap_dma4_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_dma_s *s = (struct omap_dma_s *) opaque; + struct omap_dma_s *s = opaque; int chnum, irqn = 0; struct omap_dma_channel_s *ch; diff --git a/hw/gpio/omap_gpio.c b/hw/gpio/omap_gpio.c index XXXXXXX..XXXXXXX 100644 --- a/hw/gpio/omap_gpio.c +++ b/hw/gpio/omap_gpio.c @@ -XXX,XX +XXX,XX @@ static void omap_gpio_set(void *opaque, int line, int level) static uint64_t omap_gpio_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_gpio_s *s = (struct omap_gpio_s *) opaque; + struct omap_gpio_s *s = opaque; int offset = addr & OMAP_MPUI_REG_MASK; if (size != 2) { @@ -XXX,XX +XXX,XX @@ static uint64_t omap_gpio_read(void *opaque, hwaddr addr, static void omap_gpio_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_gpio_s *s = (struct omap_gpio_s *) opaque; + struct omap_gpio_s *s = opaque; int offset = addr & OMAP_MPUI_REG_MASK; uint16_t diff; int ln; @@ -XXX,XX +XXX,XX @@ static void omap2_gpio_module_reset(struct omap2_gpio_s *s) static uint32_t omap2_gpio_module_read(void *opaque, hwaddr addr) { - struct omap2_gpio_s *s = (struct omap2_gpio_s *) opaque; + struct omap2_gpio_s *s = opaque; switch (addr) { case 0x00: /* GPIO_REVISION */ @@ -XXX,XX +XXX,XX @@ static uint32_t omap2_gpio_module_read(void *opaque, hwaddr addr) static void omap2_gpio_module_write(void *opaque, hwaddr addr, uint32_t value) { - struct omap2_gpio_s *s = (struct omap2_gpio_s *) opaque; + struct omap2_gpio_s *s = opaque; uint32_t diff; int ln; @@ -XXX,XX +XXX,XX @@ static void omap2_gpif_reset(DeviceState *dev) s->gpo = 0; } -static uint64_t omap2_gpif_top_read(void *opaque, hwaddr addr, - unsigned size) +static uint64_t omap2_gpif_top_read(void *opaque, hwaddr addr, unsigned size) { - struct omap2_gpif_s *s = (struct omap2_gpif_s *) opaque; + struct omap2_gpif_s *s = opaque; switch (addr) { case 0x00: /* IPGENERICOCPSPL_REVISION */ @@ -XXX,XX +XXX,XX @@ static uint64_t omap2_gpif_top_read(void *opaque, hwaddr addr, static void omap2_gpif_top_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap2_gpif_s *s = (struct omap2_gpif_s *) opaque; + struct omap2_gpif_s *s = opaque; switch (addr) { case 0x00: /* IPGENERICOCPSPL_REVISION */ diff --git a/hw/intc/omap_intc.c b/hw/intc/omap_intc.c index XXXXXXX..XXXXXXX 100644 --- a/hw/intc/omap_intc.c +++ b/hw/intc/omap_intc.c @@ -XXX,XX +XXX,XX @@ static inline void omap_inth_update(struct omap_intr_handler_s *s, int is_fiq) static void omap_set_intr(void *opaque, int irq, int req) { - struct omap_intr_handler_s *ih = (struct omap_intr_handler_s *) opaque; + struct omap_intr_handler_s *ih = opaque; uint32_t rise; struct omap_intr_handler_bank_s *bank = &ih->bank[irq >> 5]; @@ -XXX,XX +XXX,XX @@ static void omap_set_intr(void *opaque, int irq, int req) /* Simplified version with no edge detection */ static void omap_set_intr_noedge(void *opaque, int irq, int req) { - struct omap_intr_handler_s *ih = (struct omap_intr_handler_s *) opaque; + struct omap_intr_handler_s *ih = opaque; uint32_t rise; struct omap_intr_handler_bank_s *bank = &ih->bank[irq >> 5]; @@ -XXX,XX +XXX,XX @@ static void omap_set_intr_noedge(void *opaque, int irq, int req) static uint64_t omap_inth_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_intr_handler_s *s = (struct omap_intr_handler_s *) opaque; + struct omap_intr_handler_s *s = opaque; int i, offset = addr; int bank_no = offset >> 8; int line_no; @@ -XXX,XX +XXX,XX @@ static uint64_t omap_inth_read(void *opaque, hwaddr addr, static void omap_inth_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_intr_handler_s *s = (struct omap_intr_handler_s *) opaque; + struct omap_intr_handler_s *s = opaque; int i, offset = addr; int bank_no = offset >> 8; struct omap_intr_handler_bank_s *bank = &s->bank[bank_no]; @@ -XXX,XX +XXX,XX @@ static const TypeInfo omap_intc_info = { static uint64_t omap2_inth_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_intr_handler_s *s = (struct omap_intr_handler_s *) opaque; + struct omap_intr_handler_s *s = opaque; int offset = addr; int bank_no, line_no; struct omap_intr_handler_bank_s *bank = NULL; @@ -XXX,XX +XXX,XX @@ static uint64_t omap2_inth_read(void *opaque, hwaddr addr, static void omap2_inth_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_intr_handler_s *s = (struct omap_intr_handler_s *) opaque; + struct omap_intr_handler_s *s = opaque; int offset = addr; int bank_no, line_no; struct omap_intr_handler_bank_s *bank = NULL; diff --git a/hw/misc/omap_gpmc.c b/hw/misc/omap_gpmc.c index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/omap_gpmc.c +++ b/hw/misc/omap_gpmc.c @@ -XXX,XX +XXX,XX @@ static void omap_gpmc_dma_update(struct omap_gpmc_s *s, int value) static uint64_t omap_nand_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_gpmc_cs_file_s *f = (struct omap_gpmc_cs_file_s *)opaque; + struct omap_gpmc_cs_file_s *f = opaque; uint64_t v; nand_setpins(f->dev, 0, 0, 0, 1, 0); switch (omap_gpmc_devsize(f)) { @@ -XXX,XX +XXX,XX @@ static void omap_nand_setio(DeviceState *dev, uint64_t value, static void omap_nand_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_gpmc_cs_file_s *f = (struct omap_gpmc_cs_file_s *)opaque; + struct omap_gpmc_cs_file_s *f = opaque; nand_setpins(f->dev, 0, 0, 0, 1, 0); omap_nand_setio(f->dev, value, omap_gpmc_devsize(f), size); } @@ -XXX,XX +XXX,XX @@ static void fill_prefetch_fifo(struct omap_gpmc_s *s) static uint64_t omap_gpmc_prefetch_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_gpmc_s *s = (struct omap_gpmc_s *) opaque; + struct omap_gpmc_s *s = opaque; uint32_t data; if (s->prefetch.config1 & 1) { /* The TRM doesn't define the behaviour if you read from the @@ -XXX,XX +XXX,XX @@ static uint64_t omap_gpmc_prefetch_read(void *opaque, hwaddr addr, static void omap_gpmc_prefetch_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_gpmc_s *s = (struct omap_gpmc_s *) opaque; + struct omap_gpmc_s *s = opaque; int cs = prefetch_cs(s->prefetch.config1); if ((s->prefetch.config1 & 1) == 0) { /* The TRM doesn't define the behaviour of writing to the @@ -XXX,XX +XXX,XX @@ static int gpmc_wordaccess_only(hwaddr addr) static uint64_t omap_gpmc_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_gpmc_s *s = (struct omap_gpmc_s *) opaque; + struct omap_gpmc_s *s = opaque; int cs; struct omap_gpmc_cs_file_s *f; @@ -XXX,XX +XXX,XX @@ static uint64_t omap_gpmc_read(void *opaque, hwaddr addr, static void omap_gpmc_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_gpmc_s *s = (struct omap_gpmc_s *) opaque; + struct omap_gpmc_s *s = opaque; int cs; struct omap_gpmc_cs_file_s *f; diff --git a/hw/misc/omap_l4.c b/hw/misc/omap_l4.c index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/omap_l4.c +++ b/hw/misc/omap_l4.c @@ -XXX,XX +XXX,XX @@ hwaddr omap_l4_region_size(struct omap_target_agent_s *ta, return ta->start[region].size; } -static uint64_t omap_l4ta_read(void *opaque, hwaddr addr, - unsigned size) +static uint64_t omap_l4ta_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_target_agent_s *s = (struct omap_target_agent_s *) opaque; + struct omap_target_agent_s *s = opaque; if (size != 2) { return omap_badwidth_read16(opaque, addr); @@ -XXX,XX +XXX,XX @@ static uint64_t omap_l4ta_read(void *opaque, hwaddr addr, static void omap_l4ta_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_target_agent_s *s = (struct omap_target_agent_s *) opaque; + struct omap_target_agent_s *s = opaque; if (size != 4) { omap_badwidth_write32(opaque, addr, value); diff --git a/hw/misc/omap_sdrc.c b/hw/misc/omap_sdrc.c index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/omap_sdrc.c +++ b/hw/misc/omap_sdrc.c @@ -XXX,XX +XXX,XX @@ void omap_sdrc_reset(struct omap_sdrc_s *s) s->config = 0x10; } -static uint64_t omap_sdrc_read(void *opaque, hwaddr addr, - unsigned size) +static uint64_t omap_sdrc_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_sdrc_s *s = (struct omap_sdrc_s *) opaque; + struct omap_sdrc_s *s = opaque; if (size != 4) { return omap_badwidth_read32(opaque, addr); @@ -XXX,XX +XXX,XX @@ static uint64_t omap_sdrc_read(void *opaque, hwaddr addr, static void omap_sdrc_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_sdrc_s *s = (struct omap_sdrc_s *) opaque; + struct omap_sdrc_s *s = opaque; if (size != 4) { omap_badwidth_write32(opaque, addr, value); diff --git a/hw/misc/omap_tap.c b/hw/misc/omap_tap.c index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/omap_tap.c +++ b/hw/misc/omap_tap.c @@ -XXX,XX +XXX,XX @@ #include "hw/arm/omap.h" /* TEST-Chip-level TAP */ -static uint64_t omap_tap_read(void *opaque, hwaddr addr, - unsigned size) +static uint64_t omap_tap_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_mpu_state_s *s = (struct omap_mpu_state_s *) opaque; + struct omap_mpu_state_s *s = opaque; if (size != 4) { return omap_badwidth_read32(opaque, addr); diff --git a/hw/sd/omap_mmc.c b/hw/sd/omap_mmc.c index XXXXXXX..XXXXXXX 100644 --- a/hw/sd/omap_mmc.c +++ b/hw/sd/omap_mmc.c @@ -XXX,XX +XXX,XX @@ void omap_mmc_reset(struct omap_mmc_s *host) device_cold_reset(DEVICE(host->card)); } -static uint64_t omap_mmc_read(void *opaque, hwaddr offset, - unsigned size) +static uint64_t omap_mmc_read(void *opaque, hwaddr offset, unsigned size) { uint16_t i; - struct omap_mmc_s *s = (struct omap_mmc_s *) opaque; + struct omap_mmc_s *s = opaque; if (size != 2) { return omap_badwidth_read16(opaque, offset); @@ -XXX,XX +XXX,XX @@ static void omap_mmc_write(void *opaque, hwaddr offset, uint64_t value, unsigned size) { int i; - struct omap_mmc_s *s = (struct omap_mmc_s *) opaque; + struct omap_mmc_s *s = opaque; if (size != 2) { omap_badwidth_write16(opaque, offset, value); @@ -XXX,XX +XXX,XX @@ static const MemoryRegionOps omap_mmc_ops = { static void omap_mmc_cover_cb(void *opaque, int line, int level) { - struct omap_mmc_s *host = (struct omap_mmc_s *) opaque; + struct omap_mmc_s *host = opaque; if (!host->cdet_state && level) { host->status |= 0x0002; diff --git a/hw/ssi/omap_spi.c b/hw/ssi/omap_spi.c index XXXXXXX..XXXXXXX 100644 --- a/hw/ssi/omap_spi.c +++ b/hw/ssi/omap_spi.c @@ -XXX,XX +XXX,XX @@ void omap_mcspi_reset(struct omap_mcspi_s *s) omap_mcspi_interrupt_update(s); } -static uint64_t omap_mcspi_read(void *opaque, hwaddr addr, - unsigned size) +static uint64_t omap_mcspi_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_mcspi_s *s = (struct omap_mcspi_s *) opaque; + struct omap_mcspi_s *s = opaque; int ch = 0; uint32_t ret; @@ -XXX,XX +XXX,XX @@ static uint64_t omap_mcspi_read(void *opaque, hwaddr addr, static void omap_mcspi_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_mcspi_s *s = (struct omap_mcspi_s *) opaque; + struct omap_mcspi_s *s = opaque; int ch = 0; if (size != 4) { diff --git a/hw/timer/omap_gptimer.c b/hw/timer/omap_gptimer.c index XXXXXXX..XXXXXXX 100644 --- a/hw/timer/omap_gptimer.c +++ b/hw/timer/omap_gptimer.c @@ -XXX,XX +XXX,XX @@ static inline void omap_gp_timer_trigger(struct omap_gp_timer_s *timer) static void omap_gp_timer_tick(void *opaque) { - struct omap_gp_timer_s *timer = (struct omap_gp_timer_s *) opaque; + struct omap_gp_timer_s *timer = opaque; if (!timer->ar) { timer->st = 0; @@ -XXX,XX +XXX,XX @@ static void omap_gp_timer_tick(void *opaque) static void omap_gp_timer_match(void *opaque) { - struct omap_gp_timer_s *timer = (struct omap_gp_timer_s *) opaque; + struct omap_gp_timer_s *timer = opaque; if (timer->trigger == gpt_trigger_both) omap_gp_timer_trigger(timer); @@ -XXX,XX +XXX,XX @@ static void omap_gp_timer_match(void *opaque) static void omap_gp_timer_input(void *opaque, int line, int on) { - struct omap_gp_timer_s *s = (struct omap_gp_timer_s *) opaque; + struct omap_gp_timer_s *s = opaque; int trigger; switch (s->capture) { @@ -XXX,XX +XXX,XX @@ static void omap_gp_timer_input(void *opaque, int line, int on) static void omap_gp_timer_clk_update(void *opaque, int line, int on) { - struct omap_gp_timer_s *timer = (struct omap_gp_timer_s *) opaque; + struct omap_gp_timer_s *timer = opaque; omap_gp_timer_sync(timer); timer->rate = on ? omap_clk_getrate(timer->clk) : 0; @@ -XXX,XX +XXX,XX @@ void omap_gp_timer_reset(struct omap_gp_timer_s *s) static uint32_t omap_gp_timer_readw(void *opaque, hwaddr addr) { - struct omap_gp_timer_s *s = (struct omap_gp_timer_s *) opaque; + struct omap_gp_timer_s *s = opaque; switch (addr) { case 0x00: /* TIDR */ @@ -XXX,XX +XXX,XX @@ static uint32_t omap_gp_timer_readw(void *opaque, hwaddr addr) static uint32_t omap_gp_timer_readh(void *opaque, hwaddr addr) { - struct omap_gp_timer_s *s = (struct omap_gp_timer_s *) opaque; + struct omap_gp_timer_s *s = opaque; uint32_t ret; if (addr & 2) @@ -XXX,XX +XXX,XX @@ static uint32_t omap_gp_timer_readh(void *opaque, hwaddr addr) } } -static void omap_gp_timer_write(void *opaque, hwaddr addr, - uint32_t value) +static void omap_gp_timer_write(void *opaque, hwaddr addr, uint32_t value) { - struct omap_gp_timer_s *s = (struct omap_gp_timer_s *) opaque; + struct omap_gp_timer_s *s = opaque; switch (addr) { case 0x00: /* TIDR */ @@ -XXX,XX +XXX,XX @@ static void omap_gp_timer_write(void *opaque, hwaddr addr, } } -static void omap_gp_timer_writeh(void *opaque, hwaddr addr, - uint32_t value) +static void omap_gp_timer_writeh(void *opaque, hwaddr addr, uint32_t value) { - struct omap_gp_timer_s *s = (struct omap_gp_timer_s *) opaque; + struct omap_gp_timer_s *s = opaque; if (addr & 2) omap_gp_timer_write(opaque, addr, (value << 16) | s->writeh); diff --git a/hw/timer/omap_synctimer.c b/hw/timer/omap_synctimer.c index XXXXXXX..XXXXXXX 100644 --- a/hw/timer/omap_synctimer.c +++ b/hw/timer/omap_synctimer.c @@ -XXX,XX +XXX,XX @@ void omap_synctimer_reset(struct omap_synctimer_s *s) static uint32_t omap_synctimer_readw(void *opaque, hwaddr addr) { - struct omap_synctimer_s *s = (struct omap_synctimer_s *) opaque; + struct omap_synctimer_s *s = opaque; switch (addr) { case 0x00: /* 32KSYNCNT_REV */ @@ -XXX,XX +XXX,XX @@ static uint32_t omap_synctimer_readw(void *opaque, hwaddr addr) static uint32_t omap_synctimer_readh(void *opaque, hwaddr addr) { - struct omap_synctimer_s *s = (struct omap_synctimer_s *) opaque; + struct omap_synctimer_s *s = opaque; uint32_t ret; if (addr & 2) -- 2.34.1 From: Philippe Mathieu-Daudé <philmd@linaro.org> Following docs/devel/style.rst guidelines, rename omap_gpif_s -> Omap1GpioState. This also remove a use of 'struct' in the DECLARE_INSTANCE_CHECKER() macro call. Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230109140306.23161-5-philmd@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/omap.h | 6 +++--- hw/gpio/omap_gpio.c | 16 ++++++++-------- 2 files changed, 11 insertions(+), 11 deletions(-) diff --git a/include/hw/arm/omap.h b/include/hw/arm/omap.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/omap.h +++ b/include/hw/arm/omap.h @@ -XXX,XX +XXX,XX @@ void omap_i2c_set_fclk(OMAPI2CState *i2c, omap_clk clk); /* omap_gpio.c */ #define TYPE_OMAP1_GPIO "omap-gpio" -DECLARE_INSTANCE_CHECKER(struct omap_gpif_s, OMAP1_GPIO, +typedef struct Omap1GpioState Omap1GpioState; +DECLARE_INSTANCE_CHECKER(Omap1GpioState, OMAP1_GPIO, TYPE_OMAP1_GPIO) #define TYPE_OMAP2_GPIO "omap2-gpio" DECLARE_INSTANCE_CHECKER(struct omap2_gpif_s, OMAP2_GPIO, TYPE_OMAP2_GPIO) -typedef struct omap_gpif_s omap_gpif; typedef struct omap2_gpif_s omap2_gpif; /* TODO: clock framework (see above) */ -void omap_gpio_set_clk(omap_gpif *gpio, omap_clk clk); +void omap_gpio_set_clk(Omap1GpioState *gpio, omap_clk clk); void omap2_gpio_set_iclk(omap2_gpif *gpio, omap_clk clk); void omap2_gpio_set_fclk(omap2_gpif *gpio, uint8_t i, omap_clk clk); diff --git a/hw/gpio/omap_gpio.c b/hw/gpio/omap_gpio.c index XXXXXXX..XXXXXXX 100644 --- a/hw/gpio/omap_gpio.c +++ b/hw/gpio/omap_gpio.c @@ -XXX,XX +XXX,XX @@ struct omap_gpio_s { uint16_t pins; }; -struct omap_gpif_s { +struct Omap1GpioState { SysBusDevice parent_obj; MemoryRegion iomem; @@ -XXX,XX +XXX,XX @@ struct omap_gpif_s { /* General-Purpose I/O of OMAP1 */ static void omap_gpio_set(void *opaque, int line, int level) { - struct omap_gpif_s *p = opaque; + Omap1GpioState *p = opaque; struct omap_gpio_s *s = &p->omap1; uint16_t prev = s->inputs; @@ -XXX,XX +XXX,XX @@ static const MemoryRegionOps omap2_gpio_module_ops = { static void omap_gpif_reset(DeviceState *dev) { - struct omap_gpif_s *s = OMAP1_GPIO(dev); + Omap1GpioState *s = OMAP1_GPIO(dev); omap_gpio_reset(&s->omap1); } @@ -XXX,XX +XXX,XX @@ static const MemoryRegionOps omap2_gpif_top_ops = { static void omap_gpio_init(Object *obj) { DeviceState *dev = DEVICE(obj); - struct omap_gpif_s *s = OMAP1_GPIO(obj); + Omap1GpioState *s = OMAP1_GPIO(obj); SysBusDevice *sbd = SYS_BUS_DEVICE(obj); qdev_init_gpio_in(dev, omap_gpio_set, 16); @@ -XXX,XX +XXX,XX @@ static void omap_gpio_init(Object *obj) static void omap_gpio_realize(DeviceState *dev, Error **errp) { - struct omap_gpif_s *s = OMAP1_GPIO(dev); + Omap1GpioState *s = OMAP1_GPIO(dev); if (!s->clk) { error_setg(errp, "omap-gpio: clk not connected"); @@ -XXX,XX +XXX,XX @@ static void omap2_gpio_realize(DeviceState *dev, Error **errp) } } -void omap_gpio_set_clk(omap_gpif *gpio, omap_clk clk) +void omap_gpio_set_clk(Omap1GpioState *gpio, omap_clk clk) { gpio->clk = clk; } static Property omap_gpio_properties[] = { - DEFINE_PROP_INT32("mpu_model", struct omap_gpif_s, mpu_model, 0), + DEFINE_PROP_INT32("mpu_model", Omap1GpioState, mpu_model, 0), DEFINE_PROP_END_OF_LIST(), }; @@ -XXX,XX +XXX,XX @@ static void omap_gpio_class_init(ObjectClass *klass, void *data) static const TypeInfo omap_gpio_info = { .name = TYPE_OMAP1_GPIO, .parent = TYPE_SYS_BUS_DEVICE, - .instance_size = sizeof(struct omap_gpif_s), + .instance_size = sizeof(Omap1GpioState), .instance_init = omap_gpio_init, .class_init = omap_gpio_class_init, }; -- 2.34.1 From: Philippe Mathieu-Daudé <philmd@linaro.org> Following docs/devel/style.rst guidelines, rename omap2_gpif_s -> Omap2GpioState. This also remove a use of 'struct' in the DECLARE_INSTANCE_CHECKER() macro call. Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230109140306.23161-6-philmd@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/omap.h | 9 ++++----- hw/gpio/omap_gpio.c | 20 ++++++++++---------- 2 files changed, 14 insertions(+), 15 deletions(-) diff --git a/include/hw/arm/omap.h b/include/hw/arm/omap.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/omap.h +++ b/include/hw/arm/omap.h @@ -XXX,XX +XXX,XX @@ DECLARE_INSTANCE_CHECKER(Omap1GpioState, OMAP1_GPIO, TYPE_OMAP1_GPIO) #define TYPE_OMAP2_GPIO "omap2-gpio" -DECLARE_INSTANCE_CHECKER(struct omap2_gpif_s, OMAP2_GPIO, +typedef struct Omap2GpioState Omap2GpioState; +DECLARE_INSTANCE_CHECKER(Omap2GpioState, OMAP2_GPIO, TYPE_OMAP2_GPIO) -typedef struct omap2_gpif_s omap2_gpif; - /* TODO: clock framework (see above) */ void omap_gpio_set_clk(Omap1GpioState *gpio, omap_clk clk); -void omap2_gpio_set_iclk(omap2_gpif *gpio, omap_clk clk); -void omap2_gpio_set_fclk(omap2_gpif *gpio, uint8_t i, omap_clk clk); +void omap2_gpio_set_iclk(Omap2GpioState *gpio, omap_clk clk); +void omap2_gpio_set_fclk(Omap2GpioState *gpio, uint8_t i, omap_clk clk); /* OMAP2 l4 Interconnect */ struct omap_l4_s; diff --git a/hw/gpio/omap_gpio.c b/hw/gpio/omap_gpio.c index XXXXXXX..XXXXXXX 100644 --- a/hw/gpio/omap_gpio.c +++ b/hw/gpio/omap_gpio.c @@ -XXX,XX +XXX,XX @@ struct omap2_gpio_s { uint8_t delay; }; -struct omap2_gpif_s { +struct Omap2GpioState { SysBusDevice parent_obj; MemoryRegion iomem; @@ -XXX,XX +XXX,XX @@ static inline void omap2_gpio_module_int(struct omap2_gpio_s *s, int line) static void omap2_gpio_set(void *opaque, int line, int level) { - struct omap2_gpif_s *p = opaque; + Omap2GpioState *p = opaque; struct omap2_gpio_s *s = &p->modules[line >> 5]; line &= 31; @@ -XXX,XX +XXX,XX @@ static void omap_gpif_reset(DeviceState *dev) static void omap2_gpif_reset(DeviceState *dev) { - struct omap2_gpif_s *s = OMAP2_GPIO(dev); + Omap2GpioState *s = OMAP2_GPIO(dev); int i; for (i = 0; i < s->modulecount; i++) { @@ -XXX,XX +XXX,XX @@ static void omap2_gpif_reset(DeviceState *dev) static uint64_t omap2_gpif_top_read(void *opaque, hwaddr addr, unsigned size) { - struct omap2_gpif_s *s = opaque; + Omap2GpioState *s = opaque; switch (addr) { case 0x00: /* IPGENERICOCPSPL_REVISION */ @@ -XXX,XX +XXX,XX @@ static uint64_t omap2_gpif_top_read(void *opaque, hwaddr addr, unsigned size) static void omap2_gpif_top_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap2_gpif_s *s = opaque; + Omap2GpioState *s = opaque; switch (addr) { case 0x00: /* IPGENERICOCPSPL_REVISION */ @@ -XXX,XX +XXX,XX @@ static void omap_gpio_realize(DeviceState *dev, Error **errp) static void omap2_gpio_realize(DeviceState *dev, Error **errp) { - struct omap2_gpif_s *s = OMAP2_GPIO(dev); + Omap2GpioState *s = OMAP2_GPIO(dev); SysBusDevice *sbd = SYS_BUS_DEVICE(dev); int i; @@ -XXX,XX +XXX,XX @@ static const TypeInfo omap_gpio_info = { .class_init = omap_gpio_class_init, }; -void omap2_gpio_set_iclk(omap2_gpif *gpio, omap_clk clk) +void omap2_gpio_set_iclk(Omap2GpioState *gpio, omap_clk clk) { gpio->iclk = clk; } -void omap2_gpio_set_fclk(omap2_gpif *gpio, uint8_t i, omap_clk clk) +void omap2_gpio_set_fclk(Omap2GpioState *gpio, uint8_t i, omap_clk clk) { assert(i <= 5); gpio->fclk[i] = clk; } static Property omap2_gpio_properties[] = { - DEFINE_PROP_INT32("mpu_model", struct omap2_gpif_s, mpu_model, 0), + DEFINE_PROP_INT32("mpu_model", Omap2GpioState, mpu_model, 0), DEFINE_PROP_END_OF_LIST(), }; @@ -XXX,XX +XXX,XX @@ static void omap2_gpio_class_init(ObjectClass *klass, void *data) static const TypeInfo omap2_gpio_info = { .name = TYPE_OMAP2_GPIO, .parent = TYPE_SYS_BUS_DEVICE, - .instance_size = sizeof(struct omap2_gpif_s), + .instance_size = sizeof(Omap2GpioState), .class_init = omap2_gpio_class_init, }; -- 2.34.1 From: Philippe Mathieu-Daudé <philmd@linaro.org> Following docs/devel/style.rst guidelines, rename omap_intr_handler_s -> OMAPIntcState. This also remove a use of 'struct' in the DECLARE_INSTANCE_CHECKER() macro call. Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230109140306.23161-7-philmd@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/omap.h | 9 ++++----- hw/intc/omap_intc.c | 38 +++++++++++++++++++------------------- 2 files changed, 23 insertions(+), 24 deletions(-) diff --git a/include/hw/arm/omap.h b/include/hw/arm/omap.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/omap.h +++ b/include/hw/arm/omap.h @@ -XXX,XX +XXX,XX @@ void omap_clk_reparent(omap_clk clk, omap_clk parent); /* omap_intc.c */ #define TYPE_OMAP_INTC "common-omap-intc" -typedef struct omap_intr_handler_s omap_intr_handler; -DECLARE_INSTANCE_CHECKER(omap_intr_handler, OMAP_INTC, - TYPE_OMAP_INTC) +typedef struct OMAPIntcState OMAPIntcState; +DECLARE_INSTANCE_CHECKER(OMAPIntcState, OMAP_INTC, TYPE_OMAP_INTC) /* @@ -XXX,XX +XXX,XX @@ DECLARE_INSTANCE_CHECKER(omap_intr_handler, OMAP_INTC, * (ie the struct omap_mpu_state_s*) to do the clockname to pointer * translation.) */ -void omap_intc_set_iclk(omap_intr_handler *intc, omap_clk clk); -void omap_intc_set_fclk(omap_intr_handler *intc, omap_clk clk); +void omap_intc_set_iclk(OMAPIntcState *intc, omap_clk clk); +void omap_intc_set_fclk(OMAPIntcState *intc, omap_clk clk); /* omap_i2c.c */ #define TYPE_OMAP_I2C "omap_i2c" diff --git a/hw/intc/omap_intc.c b/hw/intc/omap_intc.c index XXXXXXX..XXXXXXX 100644 --- a/hw/intc/omap_intc.c +++ b/hw/intc/omap_intc.c @@ -XXX,XX +XXX,XX @@ struct omap_intr_handler_bank_s { unsigned char priority[32]; }; -struct omap_intr_handler_s { +struct OMAPIntcState { SysBusDevice parent_obj; qemu_irq *pins; @@ -XXX,XX +XXX,XX @@ struct omap_intr_handler_s { struct omap_intr_handler_bank_s bank[3]; }; -static void omap_inth_sir_update(struct omap_intr_handler_s *s, int is_fiq) +static void omap_inth_sir_update(OMAPIntcState *s, int is_fiq) { int i, j, sir_intr, p_intr, p; uint32_t level; @@ -XXX,XX +XXX,XX @@ static void omap_inth_sir_update(struct omap_intr_handler_s *s, int is_fiq) s->sir_intr[is_fiq] = sir_intr; } -static inline void omap_inth_update(struct omap_intr_handler_s *s, int is_fiq) +static inline void omap_inth_update(OMAPIntcState *s, int is_fiq) { int i; uint32_t has_intr = 0; @@ -XXX,XX +XXX,XX @@ static inline void omap_inth_update(struct omap_intr_handler_s *s, int is_fiq) static void omap_set_intr(void *opaque, int irq, int req) { - struct omap_intr_handler_s *ih = opaque; + OMAPIntcState *ih = opaque; uint32_t rise; struct omap_intr_handler_bank_s *bank = &ih->bank[irq >> 5]; @@ -XXX,XX +XXX,XX @@ static void omap_set_intr(void *opaque, int irq, int req) /* Simplified version with no edge detection */ static void omap_set_intr_noedge(void *opaque, int irq, int req) { - struct omap_intr_handler_s *ih = opaque; + OMAPIntcState *ih = opaque; uint32_t rise; struct omap_intr_handler_bank_s *bank = &ih->bank[irq >> 5]; @@ -XXX,XX +XXX,XX @@ static void omap_set_intr_noedge(void *opaque, int irq, int req) static uint64_t omap_inth_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_intr_handler_s *s = opaque; + OMAPIntcState *s = opaque; int i, offset = addr; int bank_no = offset >> 8; int line_no; @@ -XXX,XX +XXX,XX @@ static uint64_t omap_inth_read(void *opaque, hwaddr addr, static void omap_inth_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_intr_handler_s *s = opaque; + OMAPIntcState *s = opaque; int i, offset = addr; int bank_no = offset >> 8; struct omap_intr_handler_bank_s *bank = &s->bank[bank_no]; @@ -XXX,XX +XXX,XX @@ static const MemoryRegionOps omap_inth_mem_ops = { static void omap_inth_reset(DeviceState *dev) { - struct omap_intr_handler_s *s = OMAP_INTC(dev); + OMAPIntcState *s = OMAP_INTC(dev); int i; for (i = 0; i < s->nbanks; ++i){ @@ -XXX,XX +XXX,XX @@ static void omap_inth_reset(DeviceState *dev) static void omap_intc_init(Object *obj) { DeviceState *dev = DEVICE(obj); - struct omap_intr_handler_s *s = OMAP_INTC(obj); + OMAPIntcState *s = OMAP_INTC(obj); SysBusDevice *sbd = SYS_BUS_DEVICE(obj); s->nbanks = 1; @@ -XXX,XX +XXX,XX @@ static void omap_intc_init(Object *obj) static void omap_intc_realize(DeviceState *dev, Error **errp) { - struct omap_intr_handler_s *s = OMAP_INTC(dev); + OMAPIntcState *s = OMAP_INTC(dev); if (!s->iclk) { error_setg(errp, "omap-intc: clk not connected"); } } -void omap_intc_set_iclk(omap_intr_handler *intc, omap_clk clk) +void omap_intc_set_iclk(OMAPIntcState *intc, omap_clk clk) { intc->iclk = clk; } -void omap_intc_set_fclk(omap_intr_handler *intc, omap_clk clk) +void omap_intc_set_fclk(OMAPIntcState *intc, omap_clk clk) { intc->fclk = clk; } static Property omap_intc_properties[] = { - DEFINE_PROP_UINT32("size", struct omap_intr_handler_s, size, 0x100), + DEFINE_PROP_UINT32("size", OMAPIntcState, size, 0x100), DEFINE_PROP_END_OF_LIST(), }; @@ -XXX,XX +XXX,XX @@ static const TypeInfo omap_intc_info = { static uint64_t omap2_inth_read(void *opaque, hwaddr addr, unsigned size) { - struct omap_intr_handler_s *s = opaque; + OMAPIntcState *s = opaque; int offset = addr; int bank_no, line_no; struct omap_intr_handler_bank_s *bank = NULL; @@ -XXX,XX +XXX,XX @@ static uint64_t omap2_inth_read(void *opaque, hwaddr addr, static void omap2_inth_write(void *opaque, hwaddr addr, uint64_t value, unsigned size) { - struct omap_intr_handler_s *s = opaque; + OMAPIntcState *s = opaque; int offset = addr; int bank_no, line_no; struct omap_intr_handler_bank_s *bank = NULL; @@ -XXX,XX +XXX,XX @@ static const MemoryRegionOps omap2_inth_mem_ops = { static void omap2_intc_init(Object *obj) { DeviceState *dev = DEVICE(obj); - struct omap_intr_handler_s *s = OMAP_INTC(obj); + OMAPIntcState *s = OMAP_INTC(obj); SysBusDevice *sbd = SYS_BUS_DEVICE(obj); s->level_only = 1; @@ -XXX,XX +XXX,XX @@ static void omap2_intc_init(Object *obj) static void omap2_intc_realize(DeviceState *dev, Error **errp) { - struct omap_intr_handler_s *s = OMAP_INTC(dev); + OMAPIntcState *s = OMAP_INTC(dev); if (!s->iclk) { error_setg(errp, "omap2-intc: iclk not connected"); @@ -XXX,XX +XXX,XX @@ static void omap2_intc_realize(DeviceState *dev, Error **errp) } static Property omap2_intc_properties[] = { - DEFINE_PROP_UINT8("revision", struct omap_intr_handler_s, + DEFINE_PROP_UINT8("revision", OMAPIntcState, revision, 0x21), DEFINE_PROP_END_OF_LIST(), }; @@ -XXX,XX +XXX,XX @@ static const TypeInfo omap2_intc_info = { static const TypeInfo omap_intc_type_info = { .name = TYPE_OMAP_INTC, .parent = TYPE_SYS_BUS_DEVICE, - .instance_size = sizeof(omap_intr_handler), + .instance_size = sizeof(OMAPIntcState), .abstract = true, }; -- 2.34.1 From: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230109140306.23161-8-philmd@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/stellaris.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/hw/arm/stellaris.c b/hw/arm/stellaris.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/stellaris.c +++ b/hw/arm/stellaris.c @@ -XXX,XX +XXX,XX @@ static void stellaris_adc_update(stellaris_adc_state *s) static void stellaris_adc_trigger(void *opaque, int irq, int level) { - stellaris_adc_state *s = (stellaris_adc_state *)opaque; + stellaris_adc_state *s = opaque; int n; for (n = 0; n < 4; n++) { @@ -XXX,XX +XXX,XX @@ static void stellaris_adc_reset(stellaris_adc_state *s) static uint64_t stellaris_adc_read(void *opaque, hwaddr offset, unsigned size) { - stellaris_adc_state *s = (stellaris_adc_state *)opaque; + stellaris_adc_state *s = opaque; /* TODO: Implement this. */ if (offset >= 0x40 && offset < 0xc0) { @@ -XXX,XX +XXX,XX @@ static uint64_t stellaris_adc_read(void *opaque, hwaddr offset, static void stellaris_adc_write(void *opaque, hwaddr offset, uint64_t value, unsigned size) { - stellaris_adc_state *s = (stellaris_adc_state *)opaque; + stellaris_adc_state *s = opaque; /* TODO: Implement this. */ if (offset >= 0x40 && offset < 0xc0) { -- 2.34.1 From: Philippe Mathieu-Daudé <philmd@linaro.org> Following docs/devel/style.rst guidelines, rename stellaris_adc_state -> StellarisADCState. This also remove a use of 'struct' in the DECLARE_INSTANCE_CHECKER() macro call. Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230109140306.23161-9-philmd@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/stellaris.c | 73 +++++++++++++++++++++++----------------------- 1 file changed, 36 insertions(+), 37 deletions(-) diff --git a/hw/arm/stellaris.c b/hw/arm/stellaris.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/stellaris.c +++ b/hw/arm/stellaris.c @@ -XXX,XX +XXX,XX @@ static void stellaris_i2c_init(Object *obj) #define STELLARIS_ADC_FIFO_FULL 0x1000 #define TYPE_STELLARIS_ADC "stellaris-adc" -typedef struct StellarisADCState stellaris_adc_state; -DECLARE_INSTANCE_CHECKER(stellaris_adc_state, STELLARIS_ADC, - TYPE_STELLARIS_ADC) +typedef struct StellarisADCState StellarisADCState; +DECLARE_INSTANCE_CHECKER(StellarisADCState, STELLARIS_ADC, TYPE_STELLARIS_ADC) struct StellarisADCState { SysBusDevice parent_obj; @@ -XXX,XX +XXX,XX @@ struct StellarisADCState { qemu_irq irq[4]; }; -static uint32_t stellaris_adc_fifo_read(stellaris_adc_state *s, int n) +static uint32_t stellaris_adc_fifo_read(StellarisADCState *s, int n) { int tail; @@ -XXX,XX +XXX,XX @@ static uint32_t stellaris_adc_fifo_read(stellaris_adc_state *s, int n) return s->fifo[n].data[tail]; } -static void stellaris_adc_fifo_write(stellaris_adc_state *s, int n, +static void stellaris_adc_fifo_write(StellarisADCState *s, int n, uint32_t value) { int head; @@ -XXX,XX +XXX,XX @@ static void stellaris_adc_fifo_write(stellaris_adc_state *s, int n, s->fifo[n].state |= STELLARIS_ADC_FIFO_FULL; } -static void stellaris_adc_update(stellaris_adc_state *s) +static void stellaris_adc_update(StellarisADCState *s) { int level; int n; @@ -XXX,XX +XXX,XX @@ static void stellaris_adc_update(stellaris_adc_state *s) static void stellaris_adc_trigger(void *opaque, int irq, int level) { - stellaris_adc_state *s = opaque; + StellarisADCState *s = opaque; int n; for (n = 0; n < 4; n++) { @@ -XXX,XX +XXX,XX @@ static void stellaris_adc_trigger(void *opaque, int irq, int level) } } -static void stellaris_adc_reset(stellaris_adc_state *s) +static void stellaris_adc_reset(StellarisADCState *s) { int n; @@ -XXX,XX +XXX,XX @@ static void stellaris_adc_reset(stellaris_adc_state *s) static uint64_t stellaris_adc_read(void *opaque, hwaddr offset, unsigned size) { - stellaris_adc_state *s = opaque; + StellarisADCState *s = opaque; /* TODO: Implement this. */ if (offset >= 0x40 && offset < 0xc0) { @@ -XXX,XX +XXX,XX @@ static uint64_t stellaris_adc_read(void *opaque, hwaddr offset, static void stellaris_adc_write(void *opaque, hwaddr offset, uint64_t value, unsigned size) { - stellaris_adc_state *s = opaque; + StellarisADCState *s = opaque; /* TODO: Implement this. */ if (offset >= 0x40 && offset < 0xc0) { @@ -XXX,XX +XXX,XX @@ static const VMStateDescription vmstate_stellaris_adc = { .version_id = 1, .minimum_version_id = 1, .fields = (VMStateField[]) { - VMSTATE_UINT32(actss, stellaris_adc_state), - VMSTATE_UINT32(ris, stellaris_adc_state), - VMSTATE_UINT32(im, stellaris_adc_state), - VMSTATE_UINT32(emux, stellaris_adc_state), - VMSTATE_UINT32(ostat, stellaris_adc_state), - VMSTATE_UINT32(ustat, stellaris_adc_state), - VMSTATE_UINT32(sspri, stellaris_adc_state), - VMSTATE_UINT32(sac, stellaris_adc_state), - VMSTATE_UINT32(fifo[0].state, stellaris_adc_state), - VMSTATE_UINT32_ARRAY(fifo[0].data, stellaris_adc_state, 16), - VMSTATE_UINT32(ssmux[0], stellaris_adc_state), - VMSTATE_UINT32(ssctl[0], stellaris_adc_state), - VMSTATE_UINT32(fifo[1].state, stellaris_adc_state), - VMSTATE_UINT32_ARRAY(fifo[1].data, stellaris_adc_state, 16), - VMSTATE_UINT32(ssmux[1], stellaris_adc_state), - VMSTATE_UINT32(ssctl[1], stellaris_adc_state), - VMSTATE_UINT32(fifo[2].state, stellaris_adc_state), - VMSTATE_UINT32_ARRAY(fifo[2].data, stellaris_adc_state, 16), - VMSTATE_UINT32(ssmux[2], stellaris_adc_state), - VMSTATE_UINT32(ssctl[2], stellaris_adc_state), - VMSTATE_UINT32(fifo[3].state, stellaris_adc_state), - VMSTATE_UINT32_ARRAY(fifo[3].data, stellaris_adc_state, 16), - VMSTATE_UINT32(ssmux[3], stellaris_adc_state), - VMSTATE_UINT32(ssctl[3], stellaris_adc_state), - VMSTATE_UINT32(noise, stellaris_adc_state), + VMSTATE_UINT32(actss, StellarisADCState), + VMSTATE_UINT32(ris, StellarisADCState), + VMSTATE_UINT32(im, StellarisADCState), + VMSTATE_UINT32(emux, StellarisADCState), + VMSTATE_UINT32(ostat, StellarisADCState), + VMSTATE_UINT32(ustat, StellarisADCState), + VMSTATE_UINT32(sspri, StellarisADCState), + VMSTATE_UINT32(sac, StellarisADCState), + VMSTATE_UINT32(fifo[0].state, StellarisADCState), + VMSTATE_UINT32_ARRAY(fifo[0].data, StellarisADCState, 16), + VMSTATE_UINT32(ssmux[0], StellarisADCState), + VMSTATE_UINT32(ssctl[0], StellarisADCState), + VMSTATE_UINT32(fifo[1].state, StellarisADCState), + VMSTATE_UINT32_ARRAY(fifo[1].data, StellarisADCState, 16), + VMSTATE_UINT32(ssmux[1], StellarisADCState), + VMSTATE_UINT32(ssctl[1], StellarisADCState), + VMSTATE_UINT32(fifo[2].state, StellarisADCState), + VMSTATE_UINT32_ARRAY(fifo[2].data, StellarisADCState, 16), + VMSTATE_UINT32(ssmux[2], StellarisADCState), + VMSTATE_UINT32(ssctl[2], StellarisADCState), + VMSTATE_UINT32(fifo[3].state, StellarisADCState), + VMSTATE_UINT32_ARRAY(fifo[3].data, StellarisADCState, 16), + VMSTATE_UINT32(ssmux[3], StellarisADCState), + VMSTATE_UINT32(ssctl[3], StellarisADCState), + VMSTATE_UINT32(noise, StellarisADCState), VMSTATE_END_OF_LIST() } }; @@ -XXX,XX +XXX,XX @@ static const VMStateDescription vmstate_stellaris_adc = { static void stellaris_adc_init(Object *obj) { DeviceState *dev = DEVICE(obj); - stellaris_adc_state *s = STELLARIS_ADC(obj); + StellarisADCState *s = STELLARIS_ADC(obj); SysBusDevice *sbd = SYS_BUS_DEVICE(obj); int n; @@ -XXX,XX +XXX,XX @@ static void stellaris_adc_class_init(ObjectClass *klass, void *data) static const TypeInfo stellaris_adc_info = { .name = TYPE_STELLARIS_ADC, .parent = TYPE_SYS_BUS_DEVICE, - .instance_size = sizeof(stellaris_adc_state), + .instance_size = sizeof(StellarisADCState), .instance_init = stellaris_adc_init, .class_init = stellaris_adc_class_init, }; -- 2.34.1 From: Philippe Mathieu-Daudé <philmd@linaro.org> The typedef and definitions are generated by the OBJECT_DECLARE_TYPE macro in "hw/arm/bcm2836.h": 20 #define TYPE_BCM283X "bcm283x" 21 OBJECT_DECLARE_TYPE(BCM283XState, BCM283XClass, BCM283X) The script ran in commit a489d1951c ("Use OBJECT_DECLARE_TYPE when possible") missed them because they are declared in a different file unit. Remove them. Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230109140306.23161-10-philmd@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/bcm2836.c | 9 ++------- 1 file changed, 2 insertions(+), 7 deletions(-) diff --git a/hw/arm/bcm2836.c b/hw/arm/bcm2836.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/bcm2836.c +++ b/hw/arm/bcm2836.c @@ -XXX,XX +XXX,XX @@ #include "hw/arm/raspi_platform.h" #include "hw/sysbus.h" -typedef struct BCM283XClass { +struct BCM283XClass { /*< private >*/ DeviceClass parent_class; /*< public >*/ @@ -XXX,XX +XXX,XX @@ typedef struct BCM283XClass { hwaddr peri_base; /* Peripheral base address seen by the CPU */ hwaddr ctrl_base; /* Interrupt controller and mailboxes etc. */ int clusterid; -} BCM283XClass; - -#define BCM283X_CLASS(klass) \ - OBJECT_CLASS_CHECK(BCM283XClass, (klass), TYPE_BCM283X) -#define BCM283X_GET_CLASS(obj) \ - OBJECT_GET_CLASS(BCM283XClass, (obj), TYPE_BCM283X) +}; static Property bcm2836_enabled_cores_property = DEFINE_PROP_UINT32("enabled-cpus", BCM283XState, enabled_cpus, 0); -- 2.34.1 From: Philippe Mathieu-Daudé <philmd@linaro.org> NPCM7XX models have been commited after the conversion from commit 8063396bf3 ("Use OBJECT_DECLARE_SIMPLE_TYPE when possible"). Manually convert them. Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230109140306.23161-11-philmd@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/adc/npcm7xx_adc.h | 7 +++---- include/hw/arm/npcm7xx.h | 18 ++++++------------ include/hw/i2c/npcm7xx_smbus.h | 7 +++---- include/hw/misc/npcm7xx_clk.h | 2 +- include/hw/misc/npcm7xx_gcr.h | 6 +++--- include/hw/misc/npcm7xx_mft.h | 7 +++---- include/hw/misc/npcm7xx_pwm.h | 3 +-- include/hw/misc/npcm7xx_rng.h | 6 +++--- include/hw/net/npcm7xx_emc.h | 5 +---- include/hw/sd/npcm7xx_sdhci.h | 4 ++-- 10 files changed, 26 insertions(+), 39 deletions(-) diff --git a/include/hw/adc/npcm7xx_adc.h b/include/hw/adc/npcm7xx_adc.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/adc/npcm7xx_adc.h +++ b/include/hw/adc/npcm7xx_adc.h @@ -XXX,XX +XXX,XX @@ * @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 { +struct NPCM7xxADCState { SysBusDevice parent; MemoryRegion iomem; @@ -XXX,XX +XXX,XX @@ typedef struct { 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) +OBJECT_DECLARE_SIMPLE_TYPE(NPCM7xxADCState, 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_NR_PWM_MODULES 2 -typedef struct NPCM7xxMachine { +struct NPCM7xxMachine { MachineState parent; /* * PWM fan splitter. each splitter connects to one PWM output and @@ -XXX,XX +XXX,XX @@ typedef struct NPCM7xxMachine { */ SplitIRQ fan_splitter[NPCM7XX_NR_PWM_MODULES * NPCM7XX_PWM_PER_MODULE]; -} NPCM7xxMachine; +}; #define TYPE_NPCM7XX_MACHINE MACHINE_TYPE_NAME("npcm7xx") -#define NPCM7XX_MACHINE(obj) \ - OBJECT_CHECK(NPCM7xxMachine, (obj), TYPE_NPCM7XX_MACHINE) +OBJECT_DECLARE_SIMPLE_TYPE(NPCM7xxMachine, NPCM7XX_MACHINE) typedef struct NPCM7xxMachineClass { MachineClass parent; @@ -XXX,XX +XXX,XX @@ typedef struct NPCM7xxMachineClass { #define NPCM7XX_MACHINE_GET_CLASS(obj) \ OBJECT_GET_CLASS(NPCM7xxMachineClass, (obj), TYPE_NPCM7XX_MACHINE) -typedef struct NPCM7xxState { +struct NPCM7xxState { DeviceState parent; ARMCPU cpu[NPCM7XX_MAX_NUM_CPUS]; @@ -XXX,XX +XXX,XX @@ typedef struct NPCM7xxState { NPCM7xxFIUState fiu[2]; NPCM7xxEMCState emc[2]; NPCM7xxSDHCIState mmc; -} NPCM7xxState; +}; #define TYPE_NPCM7XX "npcm7xx" -#define NPCM7XX(obj) OBJECT_CHECK(NPCM7xxState, (obj), TYPE_NPCM7XX) +OBJECT_DECLARE_TYPE(NPCM7xxState, NPCM7xxClass, NPCM7XX) #define TYPE_NPCM730 "npcm730" #define TYPE_NPCM750 "npcm750" @@ -XXX,XX +XXX,XX @@ typedef struct NPCM7xxClass { uint32_t num_cpus; } NPCM7xxClass; -#define NPCM7XX_CLASS(klass) \ - OBJECT_CLASS_CHECK(NPCM7xxClass, (klass), TYPE_NPCM7XX) -#define NPCM7XX_GET_CLASS(obj) \ - OBJECT_GET_CLASS(NPCM7xxClass, (obj), TYPE_NPCM7XX) - /** * npcm7xx_load_kernel - Loads memory with everything needed to boot * @machine - The machine containing the SoC to be booted. diff --git a/include/hw/i2c/npcm7xx_smbus.h b/include/hw/i2c/npcm7xx_smbus.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/i2c/npcm7xx_smbus.h +++ b/include/hw/i2c/npcm7xx_smbus.h @@ -XXX,XX +XXX,XX @@ typedef enum NPCM7xxSMBusStatus { * @rx_cur: The current position of rx_fifo. * @status: The current status of the SMBus. */ -typedef struct NPCM7xxSMBusState { +struct NPCM7xxSMBusState { SysBusDevice parent; MemoryRegion iomem; @@ -XXX,XX +XXX,XX @@ typedef struct NPCM7xxSMBusState { uint8_t rx_cur; NPCM7xxSMBusStatus status; -} NPCM7xxSMBusState; +}; #define TYPE_NPCM7XX_SMBUS "npcm7xx-smbus" -#define NPCM7XX_SMBUS(obj) OBJECT_CHECK(NPCM7xxSMBusState, (obj), \ - TYPE_NPCM7XX_SMBUS) +OBJECT_DECLARE_SIMPLE_TYPE(NPCM7xxSMBusState, NPCM7XX_SMBUS) #endif /* NPCM7XX_SMBUS_H */ 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 @@ struct NPCM7xxCLKState { }; #define TYPE_NPCM7XX_CLK "npcm7xx-clk" -#define NPCM7XX_CLK(obj) OBJECT_CHECK(NPCM7xxCLKState, (obj), TYPE_NPCM7XX_CLK) +OBJECT_DECLARE_SIMPLE_TYPE(NPCM7xxCLKState, NPCM7XX_CLK) #endif /* NPCM7XX_CLK_H */ diff --git a/include/hw/misc/npcm7xx_gcr.h b/include/hw/misc/npcm7xx_gcr.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/misc/npcm7xx_gcr.h +++ b/include/hw/misc/npcm7xx_gcr.h @@ -XXX,XX +XXX,XX @@ */ #define NPCM7XX_GCR_NR_REGS (0x148 / sizeof(uint32_t)) -typedef struct NPCM7xxGCRState { +struct NPCM7xxGCRState { SysBusDevice parent; MemoryRegion iomem; @@ -XXX,XX +XXX,XX @@ typedef struct NPCM7xxGCRState { uint32_t reset_pwron; uint32_t reset_mdlr; uint32_t reset_intcr3; -} NPCM7xxGCRState; +}; #define TYPE_NPCM7XX_GCR "npcm7xx-gcr" -#define NPCM7XX_GCR(obj) OBJECT_CHECK(NPCM7xxGCRState, (obj), TYPE_NPCM7XX_GCR) +OBJECT_DECLARE_SIMPLE_TYPE(NPCM7xxGCRState, NPCM7XX_GCR) #endif /* NPCM7XX_GCR_H */ diff --git a/include/hw/misc/npcm7xx_mft.h b/include/hw/misc/npcm7xx_mft.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/misc/npcm7xx_mft.h +++ b/include/hw/misc/npcm7xx_mft.h @@ -XXX,XX +XXX,XX @@ * @max_rpm: The maximum rpm for fans. Order: A0, B0, A1, B1. * @duty: The duty cycles for fans, relative to NPCM7XX_PWM_MAX_DUTY. */ -typedef struct NPCM7xxMFTState { +struct NPCM7xxMFTState { SysBusDevice parent; MemoryRegion iomem; @@ -XXX,XX +XXX,XX @@ typedef struct NPCM7xxMFTState { uint32_t max_rpm[NPCM7XX_MFT_FANIN_COUNT]; uint32_t duty[NPCM7XX_MFT_FANIN_COUNT]; -} NPCM7xxMFTState; +}; #define TYPE_NPCM7XX_MFT "npcm7xx-mft" -#define NPCM7XX_MFT(obj) \ - OBJECT_CHECK(NPCM7xxMFTState, (obj), TYPE_NPCM7XX_MFT) +OBJECT_DECLARE_SIMPLE_TYPE(NPCM7xxMFTState, NPCM7XX_MFT) #endif /* NPCM7XX_MFT_H */ diff --git a/include/hw/misc/npcm7xx_pwm.h b/include/hw/misc/npcm7xx_pwm.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/misc/npcm7xx_pwm.h +++ b/include/hw/misc/npcm7xx_pwm.h @@ -XXX,XX +XXX,XX @@ struct NPCM7xxPWMState { }; #define TYPE_NPCM7XX_PWM "npcm7xx-pwm" -#define NPCM7XX_PWM(obj) \ - OBJECT_CHECK(NPCM7xxPWMState, (obj), TYPE_NPCM7XX_PWM) +OBJECT_DECLARE_SIMPLE_TYPE(NPCM7xxPWMState, NPCM7XX_PWM) #endif /* NPCM7XX_PWM_H */ diff --git a/include/hw/misc/npcm7xx_rng.h b/include/hw/misc/npcm7xx_rng.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/misc/npcm7xx_rng.h +++ b/include/hw/misc/npcm7xx_rng.h @@ -XXX,XX +XXX,XX @@ #include "hw/sysbus.h" -typedef struct NPCM7xxRNGState { +struct NPCM7xxRNGState { SysBusDevice parent; MemoryRegion iomem; @@ -XXX,XX +XXX,XX @@ typedef struct NPCM7xxRNGState { uint8_t rngcs; uint8_t rngd; uint8_t rngmode; -} NPCM7xxRNGState; +}; #define TYPE_NPCM7XX_RNG "npcm7xx-rng" -#define NPCM7XX_RNG(obj) OBJECT_CHECK(NPCM7xxRNGState, (obj), TYPE_NPCM7XX_RNG) +OBJECT_DECLARE_SIMPLE_TYPE(NPCM7xxRNGState, NPCM7XX_RNG) #endif /* NPCM7XX_RNG_H */ diff --git a/include/hw/net/npcm7xx_emc.h b/include/hw/net/npcm7xx_emc.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/net/npcm7xx_emc.h +++ b/include/hw/net/npcm7xx_emc.h @@ -XXX,XX +XXX,XX @@ struct NPCM7xxEMCState { bool rx_active; }; -typedef struct NPCM7xxEMCState NPCM7xxEMCState; - #define TYPE_NPCM7XX_EMC "npcm7xx-emc" -#define NPCM7XX_EMC(obj) \ - OBJECT_CHECK(NPCM7xxEMCState, (obj), TYPE_NPCM7XX_EMC) +OBJECT_DECLARE_SIMPLE_TYPE(NPCM7xxEMCState, NPCM7XX_EMC) #endif /* NPCM7XX_EMC_H */ diff --git a/include/hw/sd/npcm7xx_sdhci.h b/include/hw/sd/npcm7xx_sdhci.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/sd/npcm7xx_sdhci.h +++ b/include/hw/sd/npcm7xx_sdhci.h @@ -XXX,XX +XXX,XX @@ typedef struct NPCM7xxRegs { uint32_t boottoctrl; } NPCM7xxRegisters; -typedef struct NPCM7xxSDHCIState { +struct NPCM7xxSDHCIState { SysBusDevice parent; MemoryRegion container; @@ -XXX,XX +XXX,XX @@ typedef struct NPCM7xxSDHCIState { NPCM7xxRegisters regs; SDHCIState sdhci; -} NPCM7xxSDHCIState; +}; #endif /* NPCM7XX_SDHCI_H */ -- 2.34.1 From: Philippe Mathieu-Daudé <philmd@linaro.org> The structure is named SECUREECState. Rename the type accordingly. Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230109140306.23161-12-philmd@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/misc/sbsa_ec.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) diff --git a/hw/misc/sbsa_ec.c b/hw/misc/sbsa_ec.c index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/sbsa_ec.c +++ b/hw/misc/sbsa_ec.c @@ -XXX,XX +XXX,XX @@ #include "hw/sysbus.h" #include "sysemu/runstate.h" -typedef struct { +typedef struct SECUREECState { SysBusDevice parent_obj; MemoryRegion iomem; } SECUREECState; -#define TYPE_SBSA_EC "sbsa-ec" -#define SECURE_EC(obj) OBJECT_CHECK(SECUREECState, (obj), TYPE_SBSA_EC) +#define TYPE_SBSA_SECURE_EC "sbsa-ec" +#define SBSA_SECURE_EC(obj) \ + OBJECT_CHECK(SECUREECState, (obj), TYPE_SBSA_SECURE_EC) enum sbsa_ec_powerstates { SBSA_EC_CMD_POWEROFF = 0x01, @@ -XXX,XX +XXX,XX @@ static uint64_t sbsa_ec_read(void *opaque, hwaddr offset, unsigned size) } static void sbsa_ec_write(void *opaque, hwaddr offset, - uint64_t value, unsigned size) + uint64_t value, unsigned size) { if (offset == 0) { /* PSCI machine power command register */ switch (value) { @@ -XXX,XX +XXX,XX @@ static const MemoryRegionOps sbsa_ec_ops = { static void sbsa_ec_init(Object *obj) { - SECUREECState *s = SECURE_EC(obj); + SECUREECState *s = SBSA_SECURE_EC(obj); SysBusDevice *dev = SYS_BUS_DEVICE(obj); memory_region_init_io(&s->iomem, obj, &sbsa_ec_ops, s, "sbsa-ec", @@ -XXX,XX +XXX,XX @@ static void sbsa_ec_class_init(ObjectClass *klass, void *data) } static const TypeInfo sbsa_ec_info = { - .name = TYPE_SBSA_EC, + .name = TYPE_SBSA_SECURE_EC, .parent = TYPE_SYS_BUS_DEVICE, .instance_size = sizeof(SECUREECState), .instance_init = sbsa_ec_init, -- 2.34.1 From: Philippe Mathieu-Daudé <philmd@linaro.org> This model was merged few days before the QOM cleanup from commit 8063396bf3 ("Use OBJECT_DECLARE_SIMPLE_TYPE when possible") was pulled and merged. Manually adapt. Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20230109140306.23161-13-philmd@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/misc/sbsa_ec.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/hw/misc/sbsa_ec.c b/hw/misc/sbsa_ec.c index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/sbsa_ec.c +++ b/hw/misc/sbsa_ec.c @@ -XXX,XX +XXX,XX @@ typedef struct SECUREECState { } SECUREECState; #define TYPE_SBSA_SECURE_EC "sbsa-ec" -#define SBSA_SECURE_EC(obj) \ - OBJECT_CHECK(SECUREECState, (obj), TYPE_SBSA_SECURE_EC) +OBJECT_DECLARE_SIMPLE_TYPE(SECUREECState, SBSA_SECURE_EC) enum sbsa_ec_powerstates { SBSA_EC_CMD_POWEROFF = 0x01, -- 2.34.1 From: Philippe Mathieu-Daudé <philmd@linaro.org> This remove a use of 'struct' in the DECLARE_INSTANCE_CHECKER() macro call, to avoid after a QOM refactor: hw/intc/xilinx_intc.c:45:1: error: declaration of anonymous struct must be a definition DECLARE_INSTANCE_CHECKER(struct xlx_pic, XILINX_INTC, ^ Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Edgar E. Iglesias <edgar@zeroasic.com> Message-id: 20230109140306.23161-14-philmd@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/intc/xilinx_intc.c | 28 +++++++++++++--------------- 1 file changed, 13 insertions(+), 15 deletions(-) diff --git a/hw/intc/xilinx_intc.c b/hw/intc/xilinx_intc.c index XXXXXXX..XXXXXXX 100644 --- a/hw/intc/xilinx_intc.c +++ b/hw/intc/xilinx_intc.c @@ -XXX,XX +XXX,XX @@ #define R_MAX 8 #define TYPE_XILINX_INTC "xlnx.xps-intc" -DECLARE_INSTANCE_CHECKER(struct xlx_pic, XILINX_INTC, - TYPE_XILINX_INTC) +typedef struct XpsIntc XpsIntc; +DECLARE_INSTANCE_CHECKER(XpsIntc, XILINX_INTC, TYPE_XILINX_INTC) -struct xlx_pic +struct XpsIntc { SysBusDevice parent_obj; @@ -XXX,XX +XXX,XX @@ struct xlx_pic uint32_t irq_pin_state; }; -static void update_irq(struct xlx_pic *p) +static void update_irq(XpsIntc *p) { uint32_t i; @@ -XXX,XX +XXX,XX @@ static void update_irq(struct xlx_pic *p) qemu_set_irq(p->parent_irq, (p->regs[R_MER] & 1) && p->regs[R_IPR]); } -static uint64_t -pic_read(void *opaque, hwaddr addr, unsigned int size) +static uint64_t pic_read(void *opaque, hwaddr addr, unsigned int size) { - struct xlx_pic *p = opaque; + XpsIntc *p = opaque; uint32_t r = 0; addr >>= 2; @@ -XXX,XX +XXX,XX @@ pic_read(void *opaque, hwaddr addr, unsigned int size) return r; } -static void -pic_write(void *opaque, hwaddr addr, - uint64_t val64, unsigned int size) +static void pic_write(void *opaque, hwaddr addr, + uint64_t val64, unsigned int size) { - struct xlx_pic *p = opaque; + XpsIntc *p = opaque; uint32_t value = val64; addr >>= 2; @@ -XXX,XX +XXX,XX @@ static const MemoryRegionOps pic_ops = { static void irq_handler(void *opaque, int irq, int level) { - struct xlx_pic *p = opaque; + XpsIntc *p = opaque; /* edge triggered interrupt */ if (p->c_kind_of_intr & (1 << irq) && p->regs[R_MER] & 2) { @@ -XXX,XX +XXX,XX @@ static void irq_handler(void *opaque, int irq, int level) static void xilinx_intc_init(Object *obj) { - struct xlx_pic *p = XILINX_INTC(obj); + XpsIntc *p = XILINX_INTC(obj); qdev_init_gpio_in(DEVICE(obj), irq_handler, 32); sysbus_init_irq(SYS_BUS_DEVICE(obj), &p->parent_irq); @@ -XXX,XX +XXX,XX @@ static void xilinx_intc_init(Object *obj) } static Property xilinx_intc_properties[] = { - DEFINE_PROP_UINT32("kind-of-intr", struct xlx_pic, c_kind_of_intr, 0), + DEFINE_PROP_UINT32("kind-of-intr", XpsIntc, c_kind_of_intr, 0), DEFINE_PROP_END_OF_LIST(), }; @@ -XXX,XX +XXX,XX @@ static void xilinx_intc_class_init(ObjectClass *klass, void *data) static const TypeInfo xilinx_intc_info = { .name = TYPE_XILINX_INTC, .parent = TYPE_SYS_BUS_DEVICE, - .instance_size = sizeof(struct xlx_pic), + .instance_size = sizeof(XpsIntc), .instance_init = xilinx_intc_init, .class_init = xilinx_intc_class_init, }; -- 2.34.1 From: Philippe Mathieu-Daudé <philmd@linaro.org> This remove a use of 'struct' in the DECLARE_INSTANCE_CHECKER() macro call, to avoid after a QOM refactor: hw/timer/xilinx_timer.c:65:1: error: declaration of anonymous struct must be a definition DECLARE_INSTANCE_CHECKER(struct timerblock, XILINX_TIMER, ^ Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Edgar E. Iglesias <edgar@zeroasic.com> Message-id: 20230109140306.23161-15-philmd@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/timer/xilinx_timer.c | 27 +++++++++++++-------------- 1 file changed, 13 insertions(+), 14 deletions(-) diff --git a/hw/timer/xilinx_timer.c b/hw/timer/xilinx_timer.c index XXXXXXX..XXXXXXX 100644 --- a/hw/timer/xilinx_timer.c +++ b/hw/timer/xilinx_timer.c @@ -XXX,XX +XXX,XX @@ struct xlx_timer }; #define TYPE_XILINX_TIMER "xlnx.xps-timer" -DECLARE_INSTANCE_CHECKER(struct timerblock, XILINX_TIMER, - TYPE_XILINX_TIMER) +typedef struct XpsTimerState XpsTimerState; +DECLARE_INSTANCE_CHECKER(XpsTimerState, XILINX_TIMER, TYPE_XILINX_TIMER) -struct timerblock +struct XpsTimerState { SysBusDevice parent_obj; @@ -XXX,XX +XXX,XX @@ struct timerblock struct xlx_timer *timers; }; -static inline unsigned int num_timers(struct timerblock *t) +static inline unsigned int num_timers(XpsTimerState *t) { return 2 - t->one_timer_only; } @@ -XXX,XX +XXX,XX @@ static inline unsigned int timer_from_addr(hwaddr addr) return addr >> 2; } -static void timer_update_irq(struct timerblock *t) +static void timer_update_irq(XpsTimerState *t) { unsigned int i, irq = 0; uint32_t csr; @@ -XXX,XX +XXX,XX @@ static void timer_update_irq(struct timerblock *t) static uint64_t timer_read(void *opaque, hwaddr addr, unsigned int size) { - struct timerblock *t = opaque; + XpsTimerState *t = opaque; struct xlx_timer *xt; uint32_t r = 0; unsigned int timer; @@ -XXX,XX +XXX,XX @@ static void timer_write(void *opaque, hwaddr addr, uint64_t val64, unsigned int size) { - struct timerblock *t = opaque; + XpsTimerState *t = opaque; struct xlx_timer *xt; unsigned int timer; uint32_t value = val64; @@ -XXX,XX +XXX,XX @@ static const MemoryRegionOps timer_ops = { static void timer_hit(void *opaque) { struct xlx_timer *xt = opaque; - struct timerblock *t = xt->parent; + XpsTimerState *t = xt->parent; D(fprintf(stderr, "%s %d\n", __func__, xt->nr)); xt->regs[R_TCSR] |= TCSR_TINT; @@ -XXX,XX +XXX,XX @@ static void timer_hit(void *opaque) static void xilinx_timer_realize(DeviceState *dev, Error **errp) { - struct timerblock *t = XILINX_TIMER(dev); + XpsTimerState *t = XILINX_TIMER(dev); unsigned int i; /* Init all the ptimers. */ @@ -XXX,XX +XXX,XX @@ static void xilinx_timer_realize(DeviceState *dev, Error **errp) static void xilinx_timer_init(Object *obj) { - struct timerblock *t = XILINX_TIMER(obj); + XpsTimerState *t = XILINX_TIMER(obj); /* All timers share a single irq line. */ sysbus_init_irq(SYS_BUS_DEVICE(obj), &t->irq); } static Property xilinx_timer_properties[] = { - DEFINE_PROP_UINT32("clock-frequency", struct timerblock, freq_hz, - 62 * 1000000), - DEFINE_PROP_UINT8("one-timer-only", struct timerblock, one_timer_only, 0), + DEFINE_PROP_UINT32("clock-frequency", XpsTimerState, freq_hz, 62 * 1000000), + DEFINE_PROP_UINT8("one-timer-only", XpsTimerState, one_timer_only, 0), DEFINE_PROP_END_OF_LIST(), }; @@ -XXX,XX +XXX,XX @@ static void xilinx_timer_class_init(ObjectClass *klass, void *data) static const TypeInfo xilinx_timer_info = { .name = TYPE_XILINX_TIMER, .parent = TYPE_SYS_BUS_DEVICE, - .instance_size = sizeof(struct timerblock), + .instance_size = sizeof(XpsTimerState), .instance_init = xilinx_timer_init, .class_init = xilinx_timer_class_init, }; -- 2.34.1 From: Evgeny Iakovlev <eiakovlev@linux.microsoft.com> ARM trusted firmware, when built with FEAT_HCX support, sets SCR_EL3.HXEn bit to allow EL2 to modify HCRX_EL2 register without trapping it in EL3. Qemu uses a valid mask to clear unsupported SCR_EL3 bits when emulating SCR_EL3 write, and that mask doesn't include SCR_EL3.HXEn bit even if FEAT_HCX is enabled and exposed to the guest. As a result EL3 writes of that bit are ignored. Cc: qemu-stable@nongnu.org Signed-off-by: Evgeny Iakovlev <eiakovlev@linux.microsoft.com> Message-id: 20230105221251.17896-4-eiakovlev@linux.microsoft.com Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/target/arm/helper.c b/target/arm/helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper.c +++ b/target/arm/helper.c @@ -XXX,XX +XXX,XX @@ static void scr_write(CPUARMState *env, const ARMCPRegInfo *ri, uint64_t value) if (cpu_isar_feature(aa64_sme, cpu)) { valid_mask |= SCR_ENTP2; } + if (cpu_isar_feature(aa64_hcx, cpu)) { + valid_mask |= SCR_HXEN; + } } else { valid_mask &= ~(SCR_RW | SCR_ST); if (cpu_isar_feature(aa32_ras, cpu)) { -- 2.34.1