Most of this is the Neon decodetree patches, followed by Edgar's versal cleanups. thanks -- PMM The following changes since commit 2ef486e76d64436be90f7359a3071fb2a56ce835: Merge remote-tracking branch 'remotes/marcel/tags/rdma-pull-request' into staging (2020-05-03 14:12:56 +0100) are available in the Git repository at: https://git.linaro.org/people/pmaydell/qemu-arm.git tags/pull-target-arm-20200504 for you to fetch changes up to 9aefc6cf9b73f66062d2f914a0136756e7a28211: target/arm: Move gen_ function typedefs to translate.h (2020-05-04 12:59:26 +0100) ---------------------------------------------------------------- target-arm queue: * Start of conversion of Neon insns to decodetree * versal board: support SD and RTC * Implement ARMv8.2-TTS2UXN * Make VQDMULL undefined when U=1 * Some minor code cleanups ---------------------------------------------------------------- Edgar E. Iglesias (11): hw/arm: versal: Remove inclusion of arm_gicv3_common.h hw/arm: versal: Move misplaced comment hw/arm: versal-virt: Fix typo xlnx-ve -> xlnx-versal hw/arm: versal: Embed the UARTs into the SoC type hw/arm: versal: Embed the GEMs into the SoC type hw/arm: versal: Embed the ADMAs into the SoC type hw/arm: versal: Embed the APUs into the SoC type hw/arm: versal: Add support for SD hw/arm: versal: Add support for the RTC hw/arm: versal-virt: Add support for SD hw/arm: versal-virt: Add support for the RTC Fredrik Strupe (1): target/arm: Make VQDMULL undefined when U=1 Peter Maydell (25): target/arm: Don't use a TLB for ARMMMUIdx_Stage2 target/arm: Use enum constant in get_phys_addr_lpae() call target/arm: Add new 's1_is_el0' argument to get_phys_addr_lpae() target/arm: Implement ARMv8.2-TTS2UXN target/arm: Use correct variable for setting 'max' cpu's ID_AA64DFR0 target/arm/translate-vfp.inc.c: Remove duplicate simd_r32 check target/arm: Don't allow Thumb Neon insns without FEATURE_NEON target/arm: Add stubs for AArch32 Neon decodetree target/arm: Convert VCMLA (vector) to decodetree target/arm: Convert VCADD (vector) to decodetree target/arm: Convert V[US]DOT (vector) to decodetree target/arm: Convert VFM[AS]L (vector) to decodetree target/arm: Convert VCMLA (scalar) to decodetree target/arm: Convert V[US]DOT (scalar) to decodetree target/arm: Convert VFM[AS]L (scalar) to decodetree target/arm: Convert Neon load/store multiple structures to decodetree target/arm: Convert Neon 'load single structure to all lanes' to decodetree target/arm: Convert Neon 'load/store single structure' to decodetree target/arm: Convert Neon 3-reg-same VADD/VSUB to decodetree target/arm: Convert Neon 3-reg-same logic ops to decodetree target/arm: Convert Neon 3-reg-same VMAX/VMIN to decodetree target/arm: Convert Neon 3-reg-same comparisons to decodetree target/arm: Convert Neon 3-reg-same VQADD/VQSUB to decodetree target/arm: Convert Neon 3-reg-same VMUL, VMLA, VMLS, VSHL to decodetree target/arm: Move gen_ function typedefs to translate.h Philippe Mathieu-Daudé (2): hw/arm/mps2-tz: Use TYPE_IOTKIT instead of hardcoded string target/arm: Use uint64_t for midr field in CPU state struct include/hw/arm/xlnx-versal.h | 31 +- target/arm/cpu-param.h | 2 +- target/arm/cpu.h | 38 ++- target/arm/translate-a64.h | 9 - target/arm/translate.h | 26 ++ target/arm/neon-dp.decode | 86 +++++ target/arm/neon-ls.decode | 52 +++ target/arm/neon-shared.decode | 66 ++++ hw/arm/mps2-tz.c | 2 +- hw/arm/xlnx-versal-virt.c | 74 ++++- hw/arm/xlnx-versal.c | 115 +++++-- target/arm/cpu.c | 3 +- target/arm/cpu64.c | 8 +- target/arm/helper.c | 183 ++++------ target/arm/translate-a64.c | 17 - target/arm/translate-neon.inc.c | 714 +++++++++++++++++++++++++++++++++++++++ target/arm/translate-vfp.inc.c | 6 - target/arm/translate.c | 716 +++------------------------------------- target/arm/Makefile.objs | 18 + 19 files changed, 1302 insertions(+), 864 deletions(-) create mode 100644 target/arm/neon-dp.decode create mode 100644 target/arm/neon-ls.decode create mode 100644 target/arm/neon-shared.decode create mode 100644 target/arm/translate-neon.inc.c From: Fredrik Strupe <fredrik@strupe.net> According to Arm ARM, VQDMULL is only valid when U=0, while having U=1 is unallocated. Signed-off-by: Fredrik Strupe <fredrik@strupe.net> Fixes: 695272dcb976 ("target-arm: Handle UNDEF cases for Neon 3-regs-different-widths") Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/translate.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) {0, 0, 0, 0}, /* VMLSL */ {0, 0, 0, 9}, /* VQDMLSL */ {0, 0, 0, 0}, /* Integer VMULL */ - {0, 0, 0, 1}, /* VQDMULL */ + {0, 0, 0, 9}, /* VQDMULL */ {0, 0, 0, 0xa}, /* Polynomial VMULL */ {0, 0, 0, 7}, /* Reserved: always UNDEF */ }; -- 2.20.1 From: Philippe Mathieu-Daudé <f4bug@amsat.org> By using the TYPE_* definitions for devices, we can: - quickly find where devices are used with 'git-grep' - easily rename a device (one-line change). Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20200428154650.21991-1-f4bug@amsat.org Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/mps2-tz.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/hw/arm/mps2-tz.c b/hw/arm/mps2-tz.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/mps2-tz.c +++ b/hw/arm/mps2-tz.c @@ -XXX,XX +XXX,XX @@ static void mps2tz_common_init(MachineState *machine) exit(EXIT_FAILURE); } - sysbus_init_child_obj(OBJECT(machine), "iotkit", &mms->iotkit, + sysbus_init_child_obj(OBJECT(machine), TYPE_IOTKIT, &mms->iotkit, sizeof(mms->iotkit), mmc->armsse_type); iotkitdev = DEVICE(&mms->iotkit); object_property_set_link(OBJECT(&mms->iotkit), OBJECT(system_memory), -- 2.20.1 We define ARMMMUIdx_Stage2 as being an MMU index which uses a QEMU TLB. However we never actually use the TLB -- all stage 2 lookups are done by direct calls to get_phys_addr_lpae() followed by a physical address load via address_space_ld*(). Remove Stage2 from the list of ARM MMU indexes which correspond to real core MMU indexes, and instead put it in the set of "NOTLB" ARM MMU indexes. This allows us to drop NB_MMU_MODES to 11. It also means we can safely add support for the ARMv8.3-TTS2UXN extension, which adds permission bits to the stage 2 descriptors which define execute permission separatel for EL0 and EL1; supporting that while keeping Stage2 in a QEMU TLB would require us to use separate TLBs for "Stage2 for an EL0 access" and "Stage2 for an EL1 access", which is a lot of extra complication given we aren't even using the QEMU TLB. In the process of updating the comment on our MMU index use, fix a couple of other minor errors: * NS EL2 EL2&0 was missing from the list in the comment * some text hadn't been updated from when we bumped NB_MMU_MODES above 8 Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200330210400.11724-2-peter.maydell@linaro.org --- target/arm/cpu-param.h | 2 +- target/arm/cpu.h | 21 +++++--- target/arm/helper.c | 112 ++++------------------------------------- 3 files changed, 27 insertions(+), 108 deletions(-) diff --git a/target/arm/cpu-param.h b/target/arm/cpu-param.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu-param.h +++ b/target/arm/cpu-param.h @@ -XXX,XX +XXX,XX @@ # define TARGET_PAGE_BITS_MIN 10 #endif -#define NB_MMU_MODES 12 +#define NB_MMU_MODES 11 #endif diff --git a/target/arm/cpu.h b/target/arm/cpu.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.h +++ b/target/arm/cpu.h @@ -XXX,XX +XXX,XX @@ bool write_cpustate_to_list(ARMCPU *cpu, bool kvm_sync); * handling via the TLB. The only way to do a stage 1 translation without * the immediate stage 2 translation is via the ATS or AT system insns, * which can be slow-pathed and always do a page table walk. + * The only use of stage 2 translations is either as part of an s1+2 + * lookup or when loading the descriptors during a stage 1 page table walk, + * and in both those cases we don't use the TLB. * 4. we can also safely fold together the "32 bit EL3" and "64 bit EL3" * translation regimes, because they map reasonably well to each other * and they can't both be active at the same time. @@ -XXX,XX +XXX,XX @@ bool write_cpustate_to_list(ARMCPU *cpu, bool kvm_sync); * NS EL1 EL1&0 stage 1+2 (aka NS PL1) * NS EL1 EL1&0 stage 1+2 +PAN * NS EL0 EL2&0 + * NS EL2 EL2&0 * NS EL2 EL2&0 +PAN * NS EL2 (aka NS PL2) * S EL0 EL1&0 (aka S PL0) * S EL1 EL1&0 (not used if EL3 is 32 bit) * S EL1 EL1&0 +PAN * S EL3 (aka S PL1) - * NS EL1&0 stage 2 * - * for a total of 12 different mmu_idx. + * for a total of 11 different mmu_idx. * * R profile CPUs have an MPU, but can use the same set of MMU indexes * as A profile. They only need to distinguish NS EL0 and NS EL1 (and @@ -XXX,XX +XXX,XX @@ bool write_cpustate_to_list(ARMCPU *cpu, bool kvm_sync); * are not quite the same -- different CPU types (most notably M profile * vs A/R profile) would like to use MMU indexes with different semantics, * but since we don't ever need to use all of those in a single CPU we - * can avoid setting NB_MMU_MODES to more than 8. The lower bits of + * can avoid having to set NB_MMU_MODES to "total number of A profile MMU + * modes + total number of M profile MMU modes". The lower bits of * ARMMMUIdx are the core TLB mmu index, and the higher bits are always * the same for any particular CPU. * Variables of type ARMMUIdx are always full values, and the core @@ -XXX,XX +XXX,XX @@ typedef enum ARMMMUIdx { ARMMMUIdx_SE10_1_PAN = 9 | ARM_MMU_IDX_A, ARMMMUIdx_SE3 = 10 | ARM_MMU_IDX_A, - ARMMMUIdx_Stage2 = 11 | ARM_MMU_IDX_A, - /* * These are not allocated TLBs and are used only for AT system * instructions or for the first stage of an S12 page table walk. @@ -XXX,XX +XXX,XX @@ typedef enum ARMMMUIdx { ARMMMUIdx_Stage1_E0 = 0 | ARM_MMU_IDX_NOTLB, ARMMMUIdx_Stage1_E1 = 1 | ARM_MMU_IDX_NOTLB, ARMMMUIdx_Stage1_E1_PAN = 2 | ARM_MMU_IDX_NOTLB, + /* + * Not allocated a TLB: used only for second stage of an S12 page + * table walk, or for descriptor loads during first stage of an S1 + * page table walk. Note that if we ever want to have a TLB for this + * then various TLB flush insns which currently are no-ops or flush + * only stage 1 MMU indexes will need to change to flush stage 2. + */ + ARMMMUIdx_Stage2 = 3 | ARM_MMU_IDX_NOTLB, /* * M-profile. @@ -XXX,XX +XXX,XX @@ typedef enum ARMMMUIdxBit { TO_CORE_BIT(SE10_1), TO_CORE_BIT(SE10_1_PAN), TO_CORE_BIT(SE3), - TO_CORE_BIT(Stage2), TO_CORE_BIT(MUser), TO_CORE_BIT(MPriv), 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 tlbiall_nsnh_write(CPUARMState *env, const ARMCPRegInfo *ri, tlb_flush_by_mmuidx(cs, ARMMMUIdxBit_E10_1 | ARMMMUIdxBit_E10_1_PAN | - ARMMMUIdxBit_E10_0 | - ARMMMUIdxBit_Stage2); + ARMMMUIdxBit_E10_0); } static void tlbiall_nsnh_is_write(CPUARMState *env, const ARMCPRegInfo *ri, @@ -XXX,XX +XXX,XX @@ static void tlbiall_nsnh_is_write(CPUARMState *env, const ARMCPRegInfo *ri, tlb_flush_by_mmuidx_all_cpus_synced(cs, ARMMMUIdxBit_E10_1 | ARMMMUIdxBit_E10_1_PAN | - ARMMMUIdxBit_E10_0 | - ARMMMUIdxBit_Stage2); + ARMMMUIdxBit_E10_0); } -static void tlbiipas2_write(CPUARMState *env, const ARMCPRegInfo *ri, - uint64_t value) -{ - /* Invalidate by IPA. This has to invalidate any structures that - * contain only stage 2 translation information, but does not need - * to apply to structures that contain combined stage 1 and stage 2 - * translation information. - * This must NOP if EL2 isn't implemented or SCR_EL3.NS is zero. - */ - CPUState *cs = env_cpu(env); - uint64_t pageaddr; - - if (!arm_feature(env, ARM_FEATURE_EL2) || !(env->cp15.scr_el3 & SCR_NS)) { - return; - } - - pageaddr = sextract64(value << 12, 0, 40); - - tlb_flush_page_by_mmuidx(cs, pageaddr, ARMMMUIdxBit_Stage2); -} - -static void tlbiipas2_is_write(CPUARMState *env, const ARMCPRegInfo *ri, - uint64_t value) -{ - CPUState *cs = env_cpu(env); - uint64_t pageaddr; - - if (!arm_feature(env, ARM_FEATURE_EL2) || !(env->cp15.scr_el3 & SCR_NS)) { - return; - } - - pageaddr = sextract64(value << 12, 0, 40); - - tlb_flush_page_by_mmuidx_all_cpus_synced(cs, pageaddr, - ARMMMUIdxBit_Stage2); -} static void tlbiall_hyp_write(CPUARMState *env, const ARMCPRegInfo *ri, uint64_t value) @@ -XXX,XX +XXX,XX @@ static void vttbr_write(CPUARMState *env, const ARMCPRegInfo *ri, tlb_flush_by_mmuidx(cs, ARMMMUIdxBit_E10_1 | ARMMMUIdxBit_E10_1_PAN | - ARMMMUIdxBit_E10_0 | - ARMMMUIdxBit_Stage2); + ARMMMUIdxBit_E10_0); raw_write(env, ri, value); } } @@ -XXX,XX +XXX,XX @@ static int alle1_tlbmask(CPUARMState *env) return ARMMMUIdxBit_SE10_1 | ARMMMUIdxBit_SE10_1_PAN | ARMMMUIdxBit_SE10_0; - } else if (arm_feature(env, ARM_FEATURE_EL2)) { - return ARMMMUIdxBit_E10_1 | - ARMMMUIdxBit_E10_1_PAN | - ARMMMUIdxBit_E10_0 | - ARMMMUIdxBit_Stage2; } else { return ARMMMUIdxBit_E10_1 | ARMMMUIdxBit_E10_1_PAN | @@ -XXX,XX +XXX,XX @@ static void tlbi_aa64_vae3is_write(CPUARMState *env, const ARMCPRegInfo *ri, ARMMMUIdxBit_SE3); } -static void tlbi_aa64_ipas2e1_write(CPUARMState *env, const ARMCPRegInfo *ri, - uint64_t value) -{ - /* Invalidate by IPA. This has to invalidate any structures that - * contain only stage 2 translation information, but does not need - * to apply to structures that contain combined stage 1 and stage 2 - * translation information. - * This must NOP if EL2 isn't implemented or SCR_EL3.NS is zero. - */ - ARMCPU *cpu = env_archcpu(env); - CPUState *cs = CPU(cpu); - uint64_t pageaddr; - - if (!arm_feature(env, ARM_FEATURE_EL2) || !(env->cp15.scr_el3 & SCR_NS)) { - return; - } - - pageaddr = sextract64(value << 12, 0, 48); - - tlb_flush_page_by_mmuidx(cs, pageaddr, ARMMMUIdxBit_Stage2); -} - -static void tlbi_aa64_ipas2e1is_write(CPUARMState *env, const ARMCPRegInfo *ri, - uint64_t value) -{ - CPUState *cs = env_cpu(env); - uint64_t pageaddr; - - if (!arm_feature(env, ARM_FEATURE_EL2) || !(env->cp15.scr_el3 & SCR_NS)) { - return; - } - - pageaddr = sextract64(value << 12, 0, 48); - - tlb_flush_page_by_mmuidx_all_cpus_synced(cs, pageaddr, - ARMMMUIdxBit_Stage2); -} - static CPAccessResult aa64_zva_access(CPUARMState *env, const ARMCPRegInfo *ri, bool isread) { @@ -XXX,XX +XXX,XX @@ static const ARMCPRegInfo v8_cp_reginfo[] = { .writefn = tlbi_aa64_vae1_write }, { .name = "TLBI_IPAS2E1IS", .state = ARM_CP_STATE_AA64, .opc0 = 1, .opc1 = 4, .crn = 8, .crm = 0, .opc2 = 1, - .access = PL2_W, .type = ARM_CP_NO_RAW, - .writefn = tlbi_aa64_ipas2e1is_write }, + .access = PL2_W, .type = ARM_CP_NOP }, { .name = "TLBI_IPAS2LE1IS", .state = ARM_CP_STATE_AA64, .opc0 = 1, .opc1 = 4, .crn = 8, .crm = 0, .opc2 = 5, - .access = PL2_W, .type = ARM_CP_NO_RAW, - .writefn = tlbi_aa64_ipas2e1is_write }, + .access = PL2_W, .type = ARM_CP_NOP }, { .name = "TLBI_ALLE1IS", .state = ARM_CP_STATE_AA64, .opc0 = 1, .opc1 = 4, .crn = 8, .crm = 3, .opc2 = 4, .access = PL2_W, .type = ARM_CP_NO_RAW, @@ -XXX,XX +XXX,XX @@ static const ARMCPRegInfo v8_cp_reginfo[] = { .writefn = tlbi_aa64_alle1is_write }, { .name = "TLBI_IPAS2E1", .state = ARM_CP_STATE_AA64, .opc0 = 1, .opc1 = 4, .crn = 8, .crm = 4, .opc2 = 1, - .access = PL2_W, .type = ARM_CP_NO_RAW, - .writefn = tlbi_aa64_ipas2e1_write }, + .access = PL2_W, .type = ARM_CP_NOP }, { .name = "TLBI_IPAS2LE1", .state = ARM_CP_STATE_AA64, .opc0 = 1, .opc1 = 4, .crn = 8, .crm = 4, .opc2 = 5, - .access = PL2_W, .type = ARM_CP_NO_RAW, - .writefn = tlbi_aa64_ipas2e1_write }, + .access = PL2_W, .type = ARM_CP_NOP }, { .name = "TLBI_ALLE1", .state = ARM_CP_STATE_AA64, .opc0 = 1, .opc1 = 4, .crn = 8, .crm = 7, .opc2 = 4, .access = PL2_W, .type = ARM_CP_NO_RAW, @@ -XXX,XX +XXX,XX @@ static const ARMCPRegInfo v8_cp_reginfo[] = { .writefn = tlbimva_hyp_is_write }, { .name = "TLBIIPAS2", .cp = 15, .opc1 = 4, .crn = 8, .crm = 4, .opc2 = 1, - .type = ARM_CP_NO_RAW, .access = PL2_W, - .writefn = tlbiipas2_write }, + .type = ARM_CP_NOP, .access = PL2_W }, { .name = "TLBIIPAS2IS", .cp = 15, .opc1 = 4, .crn = 8, .crm = 0, .opc2 = 1, - .type = ARM_CP_NO_RAW, .access = PL2_W, - .writefn = tlbiipas2_is_write }, + .type = ARM_CP_NOP, .access = PL2_W }, { .name = "TLBIIPAS2L", .cp = 15, .opc1 = 4, .crn = 8, .crm = 4, .opc2 = 5, - .type = ARM_CP_NO_RAW, .access = PL2_W, - .writefn = tlbiipas2_write }, + .type = ARM_CP_NOP, .access = PL2_W }, { .name = "TLBIIPAS2LIS", .cp = 15, .opc1 = 4, .crn = 8, .crm = 0, .opc2 = 5, - .type = ARM_CP_NO_RAW, .access = PL2_W, - .writefn = tlbiipas2_is_write }, + .type = ARM_CP_NOP, .access = PL2_W }, /* 32 bit cache operations */ { .name = "ICIALLUIS", .cp = 15, .opc1 = 0, .crn = 7, .crm = 1, .opc2 = 0, .type = ARM_CP_NOP, .access = PL1_W, .accessfn = aa64_cacheop_pou_access }, -- 2.20.1 The access_type argument to get_phys_addr_lpae() is an MMUAccessType; use the enum constant MMU_DATA_LOAD rather than a literal 0 when we call it in S1_ptw_translate(). Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200330210400.11724-3-peter.maydell@linaro.org --- target/arm/helper.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/target/arm/helper.c b/target/arm/helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper.c +++ b/target/arm/helper.c @@ -XXX,XX +XXX,XX @@ static hwaddr S1_ptw_translate(CPUARMState *env, ARMMMUIdx mmu_idx, pcacheattrs = &cacheattrs; } - ret = get_phys_addr_lpae(env, addr, 0, ARMMMUIdx_Stage2, &s2pa, - &txattrs, &s2prot, &s2size, fi, pcacheattrs); + ret = get_phys_addr_lpae(env, addr, MMU_DATA_LOAD, ARMMMUIdx_Stage2, + &s2pa, &txattrs, &s2prot, &s2size, fi, + pcacheattrs); if (ret) { assert(fi->type != ARMFault_None); fi->s2addr = addr; -- 2.20.1 For ARMv8.2-TTS2UXN, the stage 2 page table walk wants to know whether the stage 1 access is for EL0 or not, because whether exec permission is given can depend on whether this is an EL0 or EL1 access. Add a new argument to get_phys_addr_lpae() so the call sites can pass this information in. Since get_phys_addr_lpae() doesn't already have a doc comment, add one so we have a place to put the documentation of the semantics of the new s1_is_el0 argument. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200330210400.11724-4-peter.maydell@linaro.org --- target/arm/helper.c | 29 ++++++++++++++++++++++++++++- 1 file changed, 28 insertions(+), 1 deletion(-) 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 bool get_phys_addr_lpae(CPUARMState *env, target_ulong address, MMUAccessType access_type, ARMMMUIdx mmu_idx, + bool s1_is_el0, hwaddr *phys_ptr, MemTxAttrs *txattrs, int *prot, target_ulong *page_size_ptr, ARMMMUFaultInfo *fi, ARMCacheAttrs *cacheattrs); @@ -XXX,XX +XXX,XX @@ static hwaddr S1_ptw_translate(CPUARMState *env, ARMMMUIdx mmu_idx, } ret = get_phys_addr_lpae(env, addr, MMU_DATA_LOAD, ARMMMUIdx_Stage2, + false, &s2pa, &txattrs, &s2prot, &s2size, fi, pcacheattrs); if (ret) { @@ -XXX,XX +XXX,XX @@ static ARMVAParameters aa32_va_parameters(CPUARMState *env, uint32_t va, }; } +/** + * get_phys_addr_lpae: perform one stage of page table walk, LPAE format + * + * Returns false if the translation was successful. Otherwise, phys_ptr, attrs, + * prot and page_size may not be filled in, and the populated fsr value provides + * information on why the translation aborted, in the format of a long-format + * DFSR/IFSR fault register, with the following caveats: + * * the WnR bit is never set (the caller must do this). + * + * @env: CPUARMState + * @address: virtual address to get physical address for + * @access_type: MMU_DATA_LOAD, MMU_DATA_STORE or MMU_INST_FETCH + * @mmu_idx: MMU index indicating required translation regime + * @s1_is_el0: if @mmu_idx is ARMMMUIdx_Stage2 (so this is a stage 2 page table + * walk), must be true if this is stage 2 of a stage 1+2 walk for an + * EL0 access). If @mmu_idx is anything else, @s1_is_el0 is ignored. + * @phys_ptr: set to the physical address corresponding to the virtual address + * @attrs: set to the memory transaction attributes to use + * @prot: set to the permissions for the page containing phys_ptr + * @page_size_ptr: set to the size of the page containing phys_ptr + * @fi: set to fault info if the translation fails + * @cacheattrs: (if non-NULL) set to the cacheability/shareability attributes + */ static bool get_phys_addr_lpae(CPUARMState *env, target_ulong address, MMUAccessType access_type, ARMMMUIdx mmu_idx, + bool s1_is_el0, hwaddr *phys_ptr, MemTxAttrs *txattrs, int *prot, target_ulong *page_size_ptr, ARMMMUFaultInfo *fi, ARMCacheAttrs *cacheattrs) @@ -XXX,XX +XXX,XX @@ bool get_phys_addr(CPUARMState *env, target_ulong address, /* S1 is done. Now do S2 translation. */ ret = get_phys_addr_lpae(env, ipa, access_type, ARMMMUIdx_Stage2, + mmu_idx == ARMMMUIdx_E10_0, phys_ptr, attrs, &s2_prot, page_size, fi, cacheattrs != NULL ? &cacheattrs2 : NULL); @@ -XXX,XX +XXX,XX @@ bool get_phys_addr(CPUARMState *env, target_ulong address, } if (regime_using_lpae_format(env, mmu_idx)) { - return get_phys_addr_lpae(env, address, access_type, mmu_idx, + return get_phys_addr_lpae(env, address, access_type, mmu_idx, false, phys_ptr, attrs, prot, page_size, fi, cacheattrs); } else if (regime_sctlr(env, mmu_idx) & SCTLR_XP) { -- 2.20.1 The ARMv8.2-TTS2UXN feature extends the XN field in stage 2 translation table descriptors from just bit [54] to bits [54:53], allowing stage 2 to control execution permissions separately for EL0 and EL1. Implement the new semantics of the XN field and enable the feature for our 'max' CPU. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200330210400.11724-5-peter.maydell@linaro.org --- target/arm/cpu.h | 15 +++++++++++++++ target/arm/cpu.c | 1 + target/arm/cpu64.c | 2 ++ target/arm/helper.c | 37 +++++++++++++++++++++++++++++++------ 4 files changed, 49 insertions(+), 6 deletions(-) diff --git a/target/arm/cpu.h b/target/arm/cpu.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.h +++ b/target/arm/cpu.h @@ -XXX,XX +XXX,XX @@ static inline bool isar_feature_aa32_ccidx(const ARMISARegisters *id) return FIELD_EX32(id->id_mmfr4, ID_MMFR4, CCIDX) != 0; } +static inline bool isar_feature_aa32_tts2uxn(const ARMISARegisters *id) +{ + return FIELD_EX32(id->id_mmfr4, ID_MMFR4, XNX) != 0; +} + /* * 64-bit feature tests via id registers. */ @@ -XXX,XX +XXX,XX @@ static inline bool isar_feature_aa64_ccidx(const ARMISARegisters *id) return FIELD_EX64(id->id_aa64mmfr2, ID_AA64MMFR2, CCIDX) != 0; } +static inline bool isar_feature_aa64_tts2uxn(const ARMISARegisters *id) +{ + return FIELD_EX64(id->id_aa64mmfr1, ID_AA64MMFR1, XNX) != 0; +} + /* * Feature tests for "does this exist in either 32-bit or 64-bit?" */ @@ -XXX,XX +XXX,XX @@ static inline bool isar_feature_any_ccidx(const ARMISARegisters *id) return isar_feature_aa64_ccidx(id) || isar_feature_aa32_ccidx(id); } +static inline bool isar_feature_any_tts2uxn(const ARMISARegisters *id) +{ + return isar_feature_aa64_tts2uxn(id) || isar_feature_aa32_tts2uxn(id); +} + /* * Forward to the above feature tests given an ARMCPU pointer. */ diff --git a/target/arm/cpu.c b/target/arm/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.c +++ b/target/arm/cpu.c @@ -XXX,XX +XXX,XX @@ static void arm_max_initfn(Object *obj) t = FIELD_DP32(t, ID_MMFR4, HPDS, 1); /* AA32HPD */ t = FIELD_DP32(t, ID_MMFR4, AC2, 1); /* ACTLR2, HACTLR2 */ t = FIELD_DP32(t, ID_MMFR4, CNP, 1); /* TTCNP */ + t = FIELD_DP32(t, ID_MMFR4, XNX, 1); /* TTS2UXN */ cpu->isar.id_mmfr4 = t; } #endif diff --git a/target/arm/cpu64.c b/target/arm/cpu64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu64.c +++ b/target/arm/cpu64.c @@ -XXX,XX +XXX,XX @@ static void aarch64_max_initfn(Object *obj) t = FIELD_DP64(t, ID_AA64MMFR1, VH, 1); t = FIELD_DP64(t, ID_AA64MMFR1, PAN, 2); /* ATS1E1 */ t = FIELD_DP64(t, ID_AA64MMFR1, VMIDBITS, 2); /* VMID16 */ + t = FIELD_DP64(t, ID_AA64MMFR1, XNX, 1); /* TTS2UXN */ cpu->isar.id_aa64mmfr1 = t; t = cpu->isar.id_aa64mmfr2; @@ -XXX,XX +XXX,XX @@ static void aarch64_max_initfn(Object *obj) u = FIELD_DP32(u, ID_MMFR4, HPDS, 1); /* AA32HPD */ u = FIELD_DP32(u, ID_MMFR4, AC2, 1); /* ACTLR2, HACTLR2 */ u = FIELD_DP32(u, ID_MMFR4, CNP, 1); /* TTCNP */ + u = FIELD_DP32(u, ID_MMFR4, XNX, 1); /* TTS2UXN */ cpu->isar.id_mmfr4 = u; u = cpu->isar.id_aa64dfr0; 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 @@ simple_ap_to_rw_prot(CPUARMState *env, ARMMMUIdx mmu_idx, int ap) * * @env: CPUARMState * @s2ap: The 2-bit stage2 access permissions (S2AP) - * @xn: XN (execute-never) bit + * @xn: XN (execute-never) bits + * @s1_is_el0: true if this is S2 of an S1+2 walk for EL0 */ -static int get_S2prot(CPUARMState *env, int s2ap, int xn) +static int get_S2prot(CPUARMState *env, int s2ap, int xn, bool s1_is_el0) { int prot = 0; @@ -XXX,XX +XXX,XX @@ static int get_S2prot(CPUARMState *env, int s2ap, int xn) if (s2ap & 2) { prot |= PAGE_WRITE; } - if (!xn) { - if (arm_el_is_aa64(env, 2) || prot & PAGE_READ) { + + if (cpu_isar_feature(any_tts2uxn, env_archcpu(env))) { + switch (xn) { + case 0: prot |= PAGE_EXEC; + break; + case 1: + if (s1_is_el0) { + prot |= PAGE_EXEC; + } + break; + case 2: + break; + case 3: + if (!s1_is_el0) { + prot |= PAGE_EXEC; + } + break; + default: + g_assert_not_reached(); + } + } else { + if (!extract32(xn, 1, 1)) { + if (arm_el_is_aa64(env, 2) || prot & PAGE_READ) { + prot |= PAGE_EXEC; + } } } return prot; @@ -XXX,XX +XXX,XX @@ static bool get_phys_addr_lpae(CPUARMState *env, target_ulong address, } ap = extract32(attrs, 4, 2); - xn = extract32(attrs, 12, 1); if (mmu_idx == ARMMMUIdx_Stage2) { ns = true; - *prot = get_S2prot(env, ap, xn); + xn = extract32(attrs, 11, 2); + *prot = get_S2prot(env, ap, xn, s1_is_el0); } else { ns = extract32(attrs, 3, 1); + xn = extract32(attrs, 12, 1); pxn = extract32(attrs, 11, 1); *prot = get_S1prot(env, mmu_idx, aarch64, ap, ns, xn, pxn); } -- 2.20.1 In aarch64_max_initfn() we update both 32-bit and 64-bit ID registers. The intended pattern is that for 64-bit ID registers we use FIELD_DP64 and the uint64_t 't' register, while 32-bit ID registers use FIELD_DP32 and the uint32_t 'u' register. For ID_AA64DFR0 we accidentally used 'u', meaning that the top 32 bits of this 64-bit ID register would end up always zero. Luckily at the moment that's what they should be anyway, so this bug has no visible effects. Use the right-sized variable. Fixes: 3bec78447a958d481991 Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Laurent Desnogues <laurent.desnogues@gmail.com> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20200423110915.10527-1-peter.maydell@linaro.org --- target/arm/cpu64.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) 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) u = FIELD_DP32(u, ID_MMFR4, XNX, 1); /* TTS2UXN */ cpu->isar.id_mmfr4 = u; - u = cpu->isar.id_aa64dfr0; - u = FIELD_DP64(u, ID_AA64DFR0, PMUVER, 5); /* v8.4-PMU */ - cpu->isar.id_aa64dfr0 = u; + t = cpu->isar.id_aa64dfr0; + t = FIELD_DP64(t, ID_AA64DFR0, PMUVER, 5); /* v8.4-PMU */ + cpu->isar.id_aa64dfr0 = t; u = cpu->isar.id_dfr0; u = FIELD_DP32(u, ID_DFR0, PERFMON, 5); /* v8.4-PMU */ -- 2.20.1 From: Philippe Mathieu-Daudé <f4bug@amsat.org> MIDR_EL1 is a 64-bit system register with the top 32-bit being RES0. Represent it in QEMU's ARMCPU struct with a uint64_t, not a uint32_t. This fixes an error when compiling with -Werror=conversion because we were manipulating the register value using a local uint64_t variable: target/arm/cpu64.c: In function ‘aarch64_max_initfn’: target/arm/cpu64.c:628:21: error: conversion from ‘uint64_t’ {aka ‘long unsigned int’} to ‘uint32_t’ {aka ‘unsigned int’} may change value [-Werror=conversion] 628 | cpu->midr = t; | ^ and future-proofs us against a possible future architecture change using some of the top 32 bits. Suggested-by: Laurent Desnogues <laurent.desnogues@gmail.com> Suggested-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Laurent Desnogues <laurent.desnogues@gmail.com> Message-id: 20200428172634.29707-1-f4bug@amsat.org Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/cpu.h | 2 +- target/arm/cpu.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/target/arm/cpu.h b/target/arm/cpu.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.h +++ b/target/arm/cpu.h @@ -XXX,XX +XXX,XX @@ struct ARMCPU { uint64_t id_aa64dfr0; uint64_t id_aa64dfr1; } isar; - uint32_t midr; + uint64_t midr; uint32_t revidr; uint32_t reset_fpsid; uint32_t ctr; diff --git a/target/arm/cpu.c b/target/arm/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.c +++ b/target/arm/cpu.c @@ -XXX,XX +XXX,XX @@ static const ARMCPUInfo arm_cpus[] = { static Property arm_cpu_properties[] = { DEFINE_PROP_BOOL("start-powered-off", ARMCPU, start_powered_off, false), DEFINE_PROP_UINT32("psci-conduit", ARMCPU, psci_conduit, 0), - DEFINE_PROP_UINT32("midr", ARMCPU, midr, 0), + DEFINE_PROP_UINT64("midr", ARMCPU, midr, 0), DEFINE_PROP_UINT64("mp-affinity", ARMCPU, mp_affinity, ARM64_AFFINITY_INVALID), DEFINE_PROP_INT32("node-id", ARMCPU, node_id, CPU_UNSET_NUMA_NODE_ID), -- 2.20.1 From: "Edgar E. Iglesias" <edgar.iglesias@xilinx.com> Remove inclusion of arm_gicv3_common.h, this already gets included via xlnx-versal.h. Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Luc Michel <luc.michel@greensocs.com> Message-id: 20200427181649.26851-2-edgar.iglesias@gmail.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/xlnx-versal.c | 1 - 1 file changed, 1 deletion(-) diff --git a/hw/arm/xlnx-versal.c b/hw/arm/xlnx-versal.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-versal.c +++ b/hw/arm/xlnx-versal.c @@ -XXX,XX +XXX,XX @@ #include "hw/arm/boot.h" #include "kvm_arm.h" #include "hw/misc/unimp.h" -#include "hw/intc/arm_gicv3_common.h" #include "hw/arm/xlnx-versal.h" #include "hw/char/pl011.h" -- 2.20.1 From: "Edgar E. Iglesias" <edgar.iglesias@xilinx.com> Move misplaced comment. Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Luc Michel <luc.michel@greensocs.com> Message-id: 20200427181649.26851-3-edgar.iglesias@gmail.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/xlnx-versal.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/hw/arm/xlnx-versal.c b/hw/arm/xlnx-versal.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-versal.c +++ b/hw/arm/xlnx-versal.c @@ -XXX,XX +XXX,XX @@ static void versal_create_apu_cpus(Versal *s) obj = object_new(XLNX_VERSAL_ACPU_TYPE); if (!obj) { - /* Secondary CPUs start in PSCI powered-down state */ error_report("Unable to create apu.cpu[%d] of type %s", i, XLNX_VERSAL_ACPU_TYPE); exit(EXIT_FAILURE); @@ -XXX,XX +XXX,XX @@ static void versal_create_apu_cpus(Versal *s) object_property_set_int(obj, s->cfg.psci_conduit, "psci-conduit", &error_abort); if (i) { + /* Secondary CPUs start in PSCI powered-down state */ object_property_set_bool(obj, true, "start-powered-off", &error_abort); } -- 2.20.1 From: "Edgar E. Iglesias" <edgar.iglesias@xilinx.com> Fix typo xlnx-ve -> xlnx-versal. Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Luc Michel <luc.michel@greensocs.com> Message-id: 20200427181649.26851-4-edgar.iglesias@gmail.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/xlnx-versal-virt.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/hw/arm/xlnx-versal-virt.c b/hw/arm/xlnx-versal-virt.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-versal-virt.c +++ b/hw/arm/xlnx-versal-virt.c @@ -XXX,XX +XXX,XX @@ static void versal_virt_init(MachineState *machine) psci_conduit = QEMU_PSCI_CONDUIT_SMC; } - sysbus_init_child_obj(OBJECT(machine), "xlnx-ve", &s->soc, + sysbus_init_child_obj(OBJECT(machine), "xlnx-versal", &s->soc, sizeof(s->soc), TYPE_XLNX_VERSAL); object_property_set_link(OBJECT(&s->soc), OBJECT(machine->ram), "ddr", &error_abort); -- 2.20.1 From: "Edgar E. Iglesias" <edgar.iglesias@xilinx.com> Embed the UARTs into the SoC type. Suggested-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Luc Michel <luc.michel@greensocs.com> Message-id: 20200427181649.26851-5-edgar.iglesias@gmail.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/xlnx-versal.h | 3 ++- hw/arm/xlnx-versal.c | 12 ++++++------ 2 files changed, 8 insertions(+), 7 deletions(-) diff --git a/include/hw/arm/xlnx-versal.h b/include/hw/arm/xlnx-versal.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/xlnx-versal.h +++ b/include/hw/arm/xlnx-versal.h @@ -XXX,XX +XXX,XX @@ #include "hw/sysbus.h" #include "hw/arm/boot.h" #include "hw/intc/arm_gicv3.h" +#include "hw/char/pl011.h" #define TYPE_XLNX_VERSAL "xlnx-versal" #define XLNX_VERSAL(obj) OBJECT_CHECK(Versal, (obj), TYPE_XLNX_VERSAL) @@ -XXX,XX +XXX,XX @@ typedef struct Versal { MemoryRegion mr_ocm; struct { - SysBusDevice *uart[XLNX_VERSAL_NR_UARTS]; + PL011State uart[XLNX_VERSAL_NR_UARTS]; SysBusDevice *gem[XLNX_VERSAL_NR_GEMS]; SysBusDevice *adma[XLNX_VERSAL_NR_ADMAS]; } iou; diff --git a/hw/arm/xlnx-versal.c b/hw/arm/xlnx-versal.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-versal.c +++ b/hw/arm/xlnx-versal.c @@ -XXX,XX +XXX,XX @@ #include "kvm_arm.h" #include "hw/misc/unimp.h" #include "hw/arm/xlnx-versal.h" -#include "hw/char/pl011.h" #define XLNX_VERSAL_ACPU_TYPE ARM_CPU_TYPE_NAME("cortex-a72") #define GEM_REVISION 0x40070106 @@ -XXX,XX +XXX,XX @@ static void versal_create_uarts(Versal *s, qemu_irq *pic) DeviceState *dev; MemoryRegion *mr; - dev = qdev_create(NULL, TYPE_PL011); - s->lpd.iou.uart[i] = SYS_BUS_DEVICE(dev); + sysbus_init_child_obj(OBJECT(s), name, + &s->lpd.iou.uart[i], sizeof(s->lpd.iou.uart[i]), + TYPE_PL011); + dev = DEVICE(&s->lpd.iou.uart[i]); qdev_prop_set_chr(dev, "chardev", serial_hd(i)); - object_property_add_child(OBJECT(s), name, OBJECT(dev), &error_fatal); qdev_init_nofail(dev); - mr = sysbus_mmio_get_region(s->lpd.iou.uart[i], 0); + mr = sysbus_mmio_get_region(SYS_BUS_DEVICE(dev), 0); memory_region_add_subregion(&s->mr_ps, addrs[i], mr); - sysbus_connect_irq(s->lpd.iou.uart[i], 0, pic[irqs[i]]); + sysbus_connect_irq(SYS_BUS_DEVICE(dev), 0, pic[irqs[i]]); g_free(name); } } -- 2.20.1 From: "Edgar E. Iglesias" <edgar.iglesias@xilinx.com> Embed the GEMs into the SoC type. Suggested-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Luc Michel <luc.michel@greensocs.com> Message-id: 20200427181649.26851-6-edgar.iglesias@gmail.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/xlnx-versal.h | 3 ++- hw/arm/xlnx-versal.c | 15 ++++++++------- 2 files changed, 10 insertions(+), 8 deletions(-) diff --git a/include/hw/arm/xlnx-versal.h b/include/hw/arm/xlnx-versal.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/xlnx-versal.h +++ b/include/hw/arm/xlnx-versal.h @@ -XXX,XX +XXX,XX @@ #include "hw/arm/boot.h" #include "hw/intc/arm_gicv3.h" #include "hw/char/pl011.h" +#include "hw/net/cadence_gem.h" #define TYPE_XLNX_VERSAL "xlnx-versal" #define XLNX_VERSAL(obj) OBJECT_CHECK(Versal, (obj), TYPE_XLNX_VERSAL) @@ -XXX,XX +XXX,XX @@ typedef struct Versal { struct { PL011State uart[XLNX_VERSAL_NR_UARTS]; - SysBusDevice *gem[XLNX_VERSAL_NR_GEMS]; + CadenceGEMState gem[XLNX_VERSAL_NR_GEMS]; SysBusDevice *adma[XLNX_VERSAL_NR_ADMAS]; } iou; } lpd; diff --git a/hw/arm/xlnx-versal.c b/hw/arm/xlnx-versal.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-versal.c +++ b/hw/arm/xlnx-versal.c @@ -XXX,XX +XXX,XX @@ static void versal_create_gems(Versal *s, qemu_irq *pic) DeviceState *dev; MemoryRegion *mr; - dev = qdev_create(NULL, "cadence_gem"); - s->lpd.iou.gem[i] = SYS_BUS_DEVICE(dev); - object_property_add_child(OBJECT(s), name, OBJECT(dev), &error_fatal); + sysbus_init_child_obj(OBJECT(s), name, + &s->lpd.iou.gem[i], sizeof(s->lpd.iou.gem[i]), + TYPE_CADENCE_GEM); + dev = DEVICE(&s->lpd.iou.gem[i]); if (nd->used) { qemu_check_nic_model(nd, "cadence_gem"); qdev_set_nic_properties(dev, nd); } - object_property_set_int(OBJECT(s->lpd.iou.gem[i]), + object_property_set_int(OBJECT(dev), 2, "num-priority-queues", &error_abort); - object_property_set_link(OBJECT(s->lpd.iou.gem[i]), + object_property_set_link(OBJECT(dev), OBJECT(&s->mr_ps), "dma", &error_abort); qdev_init_nofail(dev); - mr = sysbus_mmio_get_region(s->lpd.iou.gem[i], 0); + mr = sysbus_mmio_get_region(SYS_BUS_DEVICE(dev), 0); memory_region_add_subregion(&s->mr_ps, addrs[i], mr); - sysbus_connect_irq(s->lpd.iou.gem[i], 0, pic[irqs[i]]); + sysbus_connect_irq(SYS_BUS_DEVICE(dev), 0, pic[irqs[i]]); g_free(name); } } -- 2.20.1 From: "Edgar E. Iglesias" <edgar.iglesias@xilinx.com> Embed the ADMAs into the SoC type. Suggested-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Luc Michel <luc.michel@greensocs.com> Message-id: 20200427181649.26851-7-edgar.iglesias@gmail.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/xlnx-versal.h | 3 ++- hw/arm/xlnx-versal.c | 14 +++++++------- 2 files changed, 9 insertions(+), 8 deletions(-) diff --git a/include/hw/arm/xlnx-versal.h b/include/hw/arm/xlnx-versal.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/xlnx-versal.h +++ b/include/hw/arm/xlnx-versal.h @@ -XXX,XX +XXX,XX @@ #include "hw/arm/boot.h" #include "hw/intc/arm_gicv3.h" #include "hw/char/pl011.h" +#include "hw/dma/xlnx-zdma.h" #include "hw/net/cadence_gem.h" #define TYPE_XLNX_VERSAL "xlnx-versal" @@ -XXX,XX +XXX,XX @@ typedef struct Versal { struct { PL011State uart[XLNX_VERSAL_NR_UARTS]; CadenceGEMState gem[XLNX_VERSAL_NR_GEMS]; - SysBusDevice *adma[XLNX_VERSAL_NR_ADMAS]; + XlnxZDMA adma[XLNX_VERSAL_NR_ADMAS]; } iou; } lpd; diff --git a/hw/arm/xlnx-versal.c b/hw/arm/xlnx-versal.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-versal.c +++ b/hw/arm/xlnx-versal.c @@ -XXX,XX +XXX,XX @@ static void versal_create_admas(Versal *s, qemu_irq *pic) DeviceState *dev; MemoryRegion *mr; - dev = qdev_create(NULL, "xlnx.zdma"); - s->lpd.iou.adma[i] = SYS_BUS_DEVICE(dev); - object_property_set_int(OBJECT(s->lpd.iou.adma[i]), 128, "bus-width", - &error_abort); - object_property_add_child(OBJECT(s), name, OBJECT(dev), &error_fatal); + sysbus_init_child_obj(OBJECT(s), name, + &s->lpd.iou.adma[i], sizeof(s->lpd.iou.adma[i]), + TYPE_XLNX_ZDMA); + dev = DEVICE(&s->lpd.iou.adma[i]); + object_property_set_int(OBJECT(dev), 128, "bus-width", &error_abort); qdev_init_nofail(dev); - mr = sysbus_mmio_get_region(s->lpd.iou.adma[i], 0); + mr = sysbus_mmio_get_region(SYS_BUS_DEVICE(dev), 0); memory_region_add_subregion(&s->mr_ps, MM_ADMA_CH0 + i * MM_ADMA_CH0_SIZE, mr); - sysbus_connect_irq(s->lpd.iou.adma[i], 0, pic[VERSAL_ADMA_IRQ_0 + i]); + sysbus_connect_irq(SYS_BUS_DEVICE(dev), 0, pic[VERSAL_ADMA_IRQ_0 + i]); g_free(name); } } -- 2.20.1 From: "Edgar E. Iglesias" <edgar.iglesias@xilinx.com> Embed the APUs into the SoC type. Suggested-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Luc Michel <luc.michel@greensocs.com> Message-id: 20200427181649.26851-8-edgar.iglesias@gmail.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/xlnx-versal.h | 2 +- hw/arm/xlnx-versal-virt.c | 4 ++-- hw/arm/xlnx-versal.c | 19 +++++-------------- 3 files changed, 8 insertions(+), 17 deletions(-) diff --git a/include/hw/arm/xlnx-versal.h b/include/hw/arm/xlnx-versal.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/xlnx-versal.h +++ b/include/hw/arm/xlnx-versal.h @@ -XXX,XX +XXX,XX @@ typedef struct Versal { struct { struct { MemoryRegion mr; - ARMCPU *cpu[XLNX_VERSAL_NR_ACPUS]; + ARMCPU cpu[XLNX_VERSAL_NR_ACPUS]; GICv3State gic; } apu; } fpd; diff --git a/hw/arm/xlnx-versal-virt.c b/hw/arm/xlnx-versal-virt.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-versal-virt.c +++ b/hw/arm/xlnx-versal-virt.c @@ -XXX,XX +XXX,XX @@ static void versal_virt_init(MachineState *machine) s->binfo.get_dtb = versal_virt_get_dtb; s->binfo.modify_dtb = versal_virt_modify_dtb; if (machine->kernel_filename) { - arm_load_kernel(s->soc.fpd.apu.cpu[0], machine, &s->binfo); + arm_load_kernel(&s->soc.fpd.apu.cpu[0], machine, &s->binfo); } else { - AddressSpace *as = arm_boot_address_space(s->soc.fpd.apu.cpu[0], + AddressSpace *as = arm_boot_address_space(&s->soc.fpd.apu.cpu[0], &s->binfo); /* Some boot-loaders (e.g u-boot) don't like blobs at address 0 (NULL). * Offset things by 4K. */ diff --git a/hw/arm/xlnx-versal.c b/hw/arm/xlnx-versal.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-versal.c +++ b/hw/arm/xlnx-versal.c @@ -XXX,XX +XXX,XX @@ static void versal_create_apu_cpus(Versal *s) for (i = 0; i < ARRAY_SIZE(s->fpd.apu.cpu); i++) { Object *obj; - char *name; - - obj = object_new(XLNX_VERSAL_ACPU_TYPE); - if (!obj) { - error_report("Unable to create apu.cpu[%d] of type %s", - i, XLNX_VERSAL_ACPU_TYPE); - exit(EXIT_FAILURE); - } - - name = g_strdup_printf("apu-cpu[%d]", i); - object_property_add_child(OBJECT(s), name, obj, &error_fatal); - g_free(name); + object_initialize_child(OBJECT(s), "apu-cpu[*]", + &s->fpd.apu.cpu[i], sizeof(s->fpd.apu.cpu[i]), + XLNX_VERSAL_ACPU_TYPE, &error_abort, NULL); + obj = OBJECT(&s->fpd.apu.cpu[i]); object_property_set_int(obj, s->cfg.psci_conduit, "psci-conduit", &error_abort); if (i) { @@ -XXX,XX +XXX,XX @@ static void versal_create_apu_cpus(Versal *s) object_property_set_link(obj, OBJECT(&s->fpd.apu.mr), "memory", &error_abort); object_property_set_bool(obj, true, "realized", &error_fatal); - s->fpd.apu.cpu[i] = ARM_CPU(obj); } } @@ -XXX,XX +XXX,XX @@ static void versal_create_apu_gic(Versal *s, qemu_irq *pic) } for (i = 0; i < nr_apu_cpus; i++) { - DeviceState *cpudev = DEVICE(s->fpd.apu.cpu[i]); + DeviceState *cpudev = DEVICE(&s->fpd.apu.cpu[i]); int ppibase = XLNX_VERSAL_NR_IRQS + i * GIC_INTERNAL + GIC_NR_SGIS; qemu_irq maint_irq; int ti; -- 2.20.1 From: "Edgar E. Iglesias" <edgar.iglesias@xilinx.com> Add support for SD. Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Luc Michel <luc.michel@greensocs.com> Message-id: 20200427181649.26851-9-edgar.iglesias@gmail.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/xlnx-versal.h | 12 ++++++++++++ hw/arm/xlnx-versal.c | 31 +++++++++++++++++++++++++++++++ 2 files changed, 43 insertions(+) diff --git a/include/hw/arm/xlnx-versal.h b/include/hw/arm/xlnx-versal.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/xlnx-versal.h +++ b/include/hw/arm/xlnx-versal.h @@ -XXX,XX +XXX,XX @@ #include "hw/sysbus.h" #include "hw/arm/boot.h" +#include "hw/sd/sdhci.h" #include "hw/intc/arm_gicv3.h" #include "hw/char/pl011.h" #include "hw/dma/xlnx-zdma.h" @@ -XXX,XX +XXX,XX @@ #define XLNX_VERSAL_NR_UARTS 2 #define XLNX_VERSAL_NR_GEMS 2 #define XLNX_VERSAL_NR_ADMAS 8 +#define XLNX_VERSAL_NR_SDS 2 #define XLNX_VERSAL_NR_IRQS 192 typedef struct Versal { @@ -XXX,XX +XXX,XX @@ typedef struct Versal { } iou; } lpd; + /* The Platform Management Controller subsystem. */ + struct { + struct { + SDHCIState sd[XLNX_VERSAL_NR_SDS]; + } iou; + } pmc; + struct { MemoryRegion *mr_ddr; uint32_t psci_conduit; @@ -XXX,XX +XXX,XX @@ typedef struct Versal { #define VERSAL_GEM1_IRQ_0 58 #define VERSAL_GEM1_WAKE_IRQ_0 59 #define VERSAL_ADMA_IRQ_0 60 +#define VERSAL_SD0_IRQ_0 126 /* Architecturally reserved IRQs suitable for virtualization. */ #define VERSAL_RSVD_IRQ_FIRST 111 @@ -XXX,XX +XXX,XX @@ typedef struct Versal { #define MM_FPD_CRF 0xfd1a0000U #define MM_FPD_CRF_SIZE 0x140000 +#define MM_PMC_SD0 0xf1040000U +#define MM_PMC_SD0_SIZE 0x10000 #define MM_PMC_CRP 0xf1260000U #define MM_PMC_CRP_SIZE 0x10000 #endif diff --git a/hw/arm/xlnx-versal.c b/hw/arm/xlnx-versal.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-versal.c +++ b/hw/arm/xlnx-versal.c @@ -XXX,XX +XXX,XX @@ static void versal_create_admas(Versal *s, qemu_irq *pic) } } +#define SDHCI_CAPABILITIES 0x280737ec6481 /* Same as on ZynqMP. */ +static void versal_create_sds(Versal *s, qemu_irq *pic) +{ + int i; + + for (i = 0; i < ARRAY_SIZE(s->pmc.iou.sd); i++) { + DeviceState *dev; + MemoryRegion *mr; + + sysbus_init_child_obj(OBJECT(s), "sd[*]", + &s->pmc.iou.sd[i], sizeof(s->pmc.iou.sd[i]), + TYPE_SYSBUS_SDHCI); + dev = DEVICE(&s->pmc.iou.sd[i]); + + object_property_set_uint(OBJECT(dev), + 3, "sd-spec-version", &error_fatal); + object_property_set_uint(OBJECT(dev), SDHCI_CAPABILITIES, "capareg", + &error_fatal); + object_property_set_uint(OBJECT(dev), UHS_I, "uhs", &error_fatal); + qdev_init_nofail(dev); + + mr = sysbus_mmio_get_region(SYS_BUS_DEVICE(dev), 0); + memory_region_add_subregion(&s->mr_ps, + MM_PMC_SD0 + i * MM_PMC_SD0_SIZE, mr); + + sysbus_connect_irq(SYS_BUS_DEVICE(dev), 0, + pic[VERSAL_SD0_IRQ_0 + i * 2]); + } +} + /* This takes the board allocated linear DDR memory and creates aliases * for each split DDR range/aperture on the Versal address map. */ @@ -XXX,XX +XXX,XX @@ static void versal_realize(DeviceState *dev, Error **errp) versal_create_uarts(s, pic); versal_create_gems(s, pic); versal_create_admas(s, pic); + versal_create_sds(s, pic); versal_map_ddr(s); versal_unimp(s); -- 2.20.1 From: "Edgar E. Iglesias" <edgar.iglesias@xilinx.com> hw/arm: versal: Add support for the RTC. Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Luc Michel <luc.michel@greensocs.com> Message-id: 20200427181649.26851-10-edgar.iglesias@gmail.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/xlnx-versal.h | 8 ++++++++ hw/arm/xlnx-versal.c | 21 +++++++++++++++++++++ 2 files changed, 29 insertions(+) diff --git a/include/hw/arm/xlnx-versal.h b/include/hw/arm/xlnx-versal.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/xlnx-versal.h +++ b/include/hw/arm/xlnx-versal.h @@ -XXX,XX +XXX,XX @@ #include "hw/char/pl011.h" #include "hw/dma/xlnx-zdma.h" #include "hw/net/cadence_gem.h" +#include "hw/rtc/xlnx-zynqmp-rtc.h" #define TYPE_XLNX_VERSAL "xlnx-versal" #define XLNX_VERSAL(obj) OBJECT_CHECK(Versal, (obj), TYPE_XLNX_VERSAL) @@ -XXX,XX +XXX,XX @@ typedef struct Versal { struct { SDHCIState sd[XLNX_VERSAL_NR_SDS]; } iou; + + XlnxZynqMPRTC rtc; } pmc; struct { @@ -XXX,XX +XXX,XX @@ typedef struct Versal { #define VERSAL_GEM1_IRQ_0 58 #define VERSAL_GEM1_WAKE_IRQ_0 59 #define VERSAL_ADMA_IRQ_0 60 +#define VERSAL_RTC_APB_ERR_IRQ 121 #define VERSAL_SD0_IRQ_0 126 +#define VERSAL_RTC_ALARM_IRQ 142 +#define VERSAL_RTC_SECONDS_IRQ 143 /* Architecturally reserved IRQs suitable for virtualization. */ #define VERSAL_RSVD_IRQ_FIRST 111 @@ -XXX,XX +XXX,XX @@ typedef struct Versal { #define MM_PMC_SD0_SIZE 0x10000 #define MM_PMC_CRP 0xf1260000U #define MM_PMC_CRP_SIZE 0x10000 +#define MM_PMC_RTC 0xf12a0000 +#define MM_PMC_RTC_SIZE 0x10000 #endif diff --git a/hw/arm/xlnx-versal.c b/hw/arm/xlnx-versal.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-versal.c +++ b/hw/arm/xlnx-versal.c @@ -XXX,XX +XXX,XX @@ static void versal_create_sds(Versal *s, qemu_irq *pic) } } +static void versal_create_rtc(Versal *s, qemu_irq *pic) +{ + SysBusDevice *sbd; + MemoryRegion *mr; + + sysbus_init_child_obj(OBJECT(s), "rtc", &s->pmc.rtc, sizeof(s->pmc.rtc), + TYPE_XLNX_ZYNQMP_RTC); + sbd = SYS_BUS_DEVICE(&s->pmc.rtc); + qdev_init_nofail(DEVICE(sbd)); + + mr = sysbus_mmio_get_region(sbd, 0); + memory_region_add_subregion(&s->mr_ps, MM_PMC_RTC, mr); + + /* + * TODO: Connect the ALARM and SECONDS interrupts once our RTC model + * supports them. + */ + sysbus_connect_irq(sbd, 1, pic[VERSAL_RTC_APB_ERR_IRQ]); +} + /* This takes the board allocated linear DDR memory and creates aliases * for each split DDR range/aperture on the Versal address map. */ @@ -XXX,XX +XXX,XX @@ static void versal_realize(DeviceState *dev, Error **errp) versal_create_gems(s, pic); versal_create_admas(s, pic); versal_create_sds(s, pic); + versal_create_rtc(s, pic); versal_map_ddr(s); versal_unimp(s); -- 2.20.1 From: "Edgar E. Iglesias" <edgar.iglesias@xilinx.com> Add support for SD. Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Luc Michel <luc.michel@greensocs.com> Message-id: 20200427181649.26851-11-edgar.iglesias@gmail.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/xlnx-versal-virt.c | 46 +++++++++++++++++++++++++++++++++++++++ 1 file changed, 46 insertions(+) diff --git a/hw/arm/xlnx-versal-virt.c b/hw/arm/xlnx-versal-virt.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-versal-virt.c +++ b/hw/arm/xlnx-versal-virt.c @@ -XXX,XX +XXX,XX @@ #include "hw/arm/sysbus-fdt.h" #include "hw/arm/fdt.h" #include "cpu.h" +#include "hw/qdev-properties.h" #include "hw/arm/xlnx-versal.h" #define TYPE_XLNX_VERSAL_VIRT_MACHINE MACHINE_TYPE_NAME("xlnx-versal-virt") @@ -XXX,XX +XXX,XX @@ static void fdt_add_zdma_nodes(VersalVirt *s) } } +static void fdt_add_sd_nodes(VersalVirt *s) +{ + const char clocknames[] = "clk_xin\0clk_ahb"; + const char compat[] = "arasan,sdhci-8.9a"; + int i; + + for (i = ARRAY_SIZE(s->soc.pmc.iou.sd) - 1; i >= 0; i--) { + uint64_t addr = MM_PMC_SD0 + MM_PMC_SD0_SIZE * i; + char *name = g_strdup_printf("/sdhci@%" PRIx64, addr); + + qemu_fdt_add_subnode(s->fdt, name); + + qemu_fdt_setprop_cells(s->fdt, name, "clocks", + s->phandle.clk_25Mhz, s->phandle.clk_25Mhz); + qemu_fdt_setprop(s->fdt, name, "clock-names", + clocknames, sizeof(clocknames)); + qemu_fdt_setprop_cells(s->fdt, name, "interrupts", + GIC_FDT_IRQ_TYPE_SPI, VERSAL_SD0_IRQ_0 + i * 2, + GIC_FDT_IRQ_FLAGS_LEVEL_HI); + qemu_fdt_setprop_sized_cells(s->fdt, name, "reg", + 2, addr, 2, MM_PMC_SD0_SIZE); + qemu_fdt_setprop(s->fdt, name, "compatible", compat, sizeof(compat)); + g_free(name); + } +} + static void fdt_nop_memory_nodes(void *fdt, Error **errp) { Error *err = NULL; @@ -XXX,XX +XXX,XX @@ static void create_virtio_regions(VersalVirt *s) } } +static void sd_plugin_card(SDHCIState *sd, DriveInfo *di) +{ + BlockBackend *blk = di ? blk_by_legacy_dinfo(di) : NULL; + DeviceState *card; + + card = qdev_create(qdev_get_child_bus(DEVICE(sd), "sd-bus"), TYPE_SD_CARD); + object_property_add_child(OBJECT(sd), "card[*]", OBJECT(card), + &error_fatal); + qdev_prop_set_drive(card, "drive", blk, &error_fatal); + object_property_set_bool(OBJECT(card), true, "realized", &error_fatal); +} + static void versal_virt_init(MachineState *machine) { VersalVirt *s = XLNX_VERSAL_VIRT_MACHINE(machine); int psci_conduit = QEMU_PSCI_CONDUIT_DISABLED; + int i; /* * If the user provides an Operating System to be loaded, we expect them @@ -XXX,XX +XXX,XX @@ static void versal_virt_init(MachineState *machine) fdt_add_gic_nodes(s); fdt_add_timer_nodes(s); fdt_add_zdma_nodes(s); + fdt_add_sd_nodes(s); fdt_add_cpu_nodes(s, psci_conduit); fdt_add_clk_node(s, "/clk125", 125000000, s->phandle.clk_125Mhz); fdt_add_clk_node(s, "/clk25", 25000000, s->phandle.clk_25Mhz); @@ -XXX,XX +XXX,XX @@ static void versal_virt_init(MachineState *machine) memory_region_add_subregion_overlap(get_system_memory(), 0, &s->soc.fpd.apu.mr, 0); + /* Plugin SD cards. */ + for (i = 0; i < ARRAY_SIZE(s->soc.pmc.iou.sd); i++) { + sd_plugin_card(&s->soc.pmc.iou.sd[i], drive_get_next(IF_SD)); + } + s->binfo.ram_size = machine->ram_size; s->binfo.loader_start = 0x0; s->binfo.get_dtb = versal_virt_get_dtb; -- 2.20.1 From: "Edgar E. Iglesias" <edgar.iglesias@xilinx.com> Add support for the RTC. Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Luc Michel <luc.michel@greensocs.com> Message-id: 20200427181649.26851-12-edgar.iglesias@gmail.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/xlnx-versal-virt.c | 22 ++++++++++++++++++++++ 1 file changed, 22 insertions(+) diff --git a/hw/arm/xlnx-versal-virt.c b/hw/arm/xlnx-versal-virt.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-versal-virt.c +++ b/hw/arm/xlnx-versal-virt.c @@ -XXX,XX +XXX,XX @@ static void fdt_add_sd_nodes(VersalVirt *s) } } +static void fdt_add_rtc_node(VersalVirt *s) +{ + const char compat[] = "xlnx,zynqmp-rtc"; + const char interrupt_names[] = "alarm\0sec"; + char *name = g_strdup_printf("/rtc@%x", MM_PMC_RTC); + + qemu_fdt_add_subnode(s->fdt, name); + + qemu_fdt_setprop_cells(s->fdt, name, "interrupts", + GIC_FDT_IRQ_TYPE_SPI, VERSAL_RTC_ALARM_IRQ, + GIC_FDT_IRQ_FLAGS_LEVEL_HI, + GIC_FDT_IRQ_TYPE_SPI, VERSAL_RTC_SECONDS_IRQ, + GIC_FDT_IRQ_FLAGS_LEVEL_HI); + qemu_fdt_setprop(s->fdt, name, "interrupt-names", + interrupt_names, sizeof(interrupt_names)); + qemu_fdt_setprop_sized_cells(s->fdt, name, "reg", + 2, MM_PMC_RTC, 2, MM_PMC_RTC_SIZE); + qemu_fdt_setprop(s->fdt, name, "compatible", compat, sizeof(compat)); + g_free(name); +} + static void fdt_nop_memory_nodes(void *fdt, Error **errp) { Error *err = NULL; @@ -XXX,XX +XXX,XX @@ static void versal_virt_init(MachineState *machine) fdt_add_timer_nodes(s); fdt_add_zdma_nodes(s); fdt_add_sd_nodes(s); + fdt_add_rtc_node(s); fdt_add_cpu_nodes(s, psci_conduit); fdt_add_clk_node(s, "/clk125", 125000000, s->phandle.clk_125Mhz); fdt_add_clk_node(s, "/clk25", 25000000, s->phandle.clk_25Mhz); -- 2.20.1 Somewhere along theline we accidentally added a duplicate "using D16-D31 when they don't exist" check to do_vfm_dp() (probably an artifact of a patchseries rebase). Remove it. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20200430181003.21682-2-peter.maydell@linaro.org --- target/arm/translate-vfp.inc.c | 6 ------ 1 file changed, 6 deletions(-) diff --git a/target/arm/translate-vfp.inc.c b/target/arm/translate-vfp.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-vfp.inc.c +++ b/target/arm/translate-vfp.inc.c @@ -XXX,XX +XXX,XX @@ static bool do_vfm_dp(DisasContext *s, arg_VFMA_dp *a, bool neg_n, bool neg_d) return false; } - /* UNDEF accesses to D16-D31 if they don't exist. */ - if (!dc_isar_feature(aa32_simd_r32, s) && - ((a->vd | a->vn | a->vm) & 0x10)) { - return false; - } - if (!vfp_access_check(s)) { return true; } -- 2.20.1 We were accidentally permitting decode of Thumb Neon insns even if the CPU didn't have the FEATURE_NEON bit set, because the feature check was being done before the call to disas_neon_data_insn() and disas_neon_ls_insn() in the Arm decoder but was omitted from the Thumb decoder. Push the feature bit check down into the called functions so it is done for both Arm and Thumb encodings. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20200430181003.21682-3-peter.maydell@linaro.org --- target/arm/translate.c | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static int disas_neon_ls_insn(DisasContext *s, uint32_t insn) TCGv_i32 tmp2; TCGv_i64 tmp64; + if (!arm_dc_feature(s, ARM_FEATURE_NEON)) { + return 1; + } + /* FIXME: this access check should not take precedence over UNDEF * for invalid encodings; we will generate incorrect syndrome information * for attempts to execute invalid vfp/neon encodings with FP disabled. @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) TCGv_ptr ptr1, ptr2, ptr3; TCGv_i64 tmp64; + if (!arm_dc_feature(s, ARM_FEATURE_NEON)) { + return 1; + } + /* FIXME: this access check should not take precedence over UNDEF * for invalid encodings; we will generate incorrect syndrome information * for attempts to execute invalid vfp/neon encodings with FP disabled. @@ -XXX,XX +XXX,XX @@ static void disas_arm_insn(DisasContext *s, unsigned int insn) if (((insn >> 25) & 7) == 1) { /* NEON Data processing. */ - if (!arm_dc_feature(s, ARM_FEATURE_NEON)) { - goto illegal_op; - } - if (disas_neon_data_insn(s, insn)) { goto illegal_op; } @@ -XXX,XX +XXX,XX @@ static void disas_arm_insn(DisasContext *s, unsigned int insn) } if ((insn & 0x0f100000) == 0x04000000) { /* NEON load/store. */ - if (!arm_dc_feature(s, ARM_FEATURE_NEON)) { - goto illegal_op; - } - if (disas_neon_ls_insn(s, insn)) { goto illegal_op; } -- 2.20.1 Add the infrastructure for building and invoking a decodetree decoder for the AArch32 Neon encodings. At the moment the new decoder covers nothing, so we always fall back to the existing hand-written decode. We follow the same pattern we did for the VFP decodetree conversion (commit 78e138bc1f672c145ef6ace74617d and following): code that deals with Neon will be moving gradually out to translate-neon.vfp.inc, which we #include into translate.c. In order to share the decode files between A32 and T32, we split Neon into 3 parts: * data-processing * load-store * 'shared' encodings The first two groups of instructions have similar but not identical A32 and T32 encodings, so we need to manually transform the T32 encoding into the A32 one before calling the decoder; the third group covers the Neon instructions which are identical in A32 and T32. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-4-peter.maydell@linaro.org --- target/arm/neon-dp.decode | 29 ++++++++++++++++++++++++++ target/arm/neon-ls.decode | 29 ++++++++++++++++++++++++++ target/arm/neon-shared.decode | 27 +++++++++++++++++++++++++ target/arm/translate-neon.inc.c | 32 +++++++++++++++++++++++++++++ target/arm/translate.c | 36 +++++++++++++++++++++++++++++++-- target/arm/Makefile.objs | 18 +++++++++++++++++ 6 files changed, 169 insertions(+), 2 deletions(-) create mode 100644 target/arm/neon-dp.decode create mode 100644 target/arm/neon-ls.decode create mode 100644 target/arm/neon-shared.decode create mode 100644 target/arm/translate-neon.inc.c diff --git a/target/arm/neon-dp.decode b/target/arm/neon-dp.decode new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/target/arm/neon-dp.decode @@ -XXX,XX +XXX,XX @@ +# AArch32 Neon data-processing instruction descriptions +# +# Copyright (c) 2020 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 of the License, or (at your option) any later version. +# +# This library is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# Lesser General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public +# License along with this library; if not, see <http://www.gnu.org/licenses/>. + +# +# This file is processed by scripts/decodetree.py +# + +# Encodings for Neon data processing instructions where the T32 encoding +# is a simple transformation of the A32 encoding. +# More specifically, this file covers instructions where the A32 encoding is +# 0b1111_001p_qqqq_qqqq_qqqq_qqqq_qqqq_qqqq +# and the T32 encoding is +# 0b111p_1111_qqqq_qqqq_qqqq_qqqq_qqqq_qqqq +# This file works on the A32 encoding only; calling code for T32 has to +# transform the insn into the A32 version first. diff --git a/target/arm/neon-ls.decode b/target/arm/neon-ls.decode new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/target/arm/neon-ls.decode @@ -XXX,XX +XXX,XX @@ +# AArch32 Neon load/store instruction descriptions +# +# Copyright (c) 2020 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 of the License, or (at your option) any later version. +# +# This library is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# Lesser General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public +# License along with this library; if not, see <http://www.gnu.org/licenses/>. + +# +# This file is processed by scripts/decodetree.py +# + +# Encodings for Neon load/store instructions where the T32 encoding +# is a simple transformation of the A32 encoding. +# More specifically, this file covers instructions where the A32 encoding is +# 0b1111_0100_xxx0_xxxx_xxxx_xxxx_xxxx_xxxx +# and the T32 encoding is +# 0b1111_1001_xxx0_xxxx_xxxx_xxxx_xxxx_xxxx +# This file works on the A32 encoding only; calling code for T32 has to +# transform the insn into the A32 version first. diff --git a/target/arm/neon-shared.decode b/target/arm/neon-shared.decode new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/target/arm/neon-shared.decode @@ -XXX,XX +XXX,XX @@ +# AArch32 Neon instruction descriptions +# +# Copyright (c) 2020 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 of the License, or (at your option) any later version. +# +# This library is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# Lesser General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public +# License along with this library; if not, see <http://www.gnu.org/licenses/>. + +# +# This file is processed by scripts/decodetree.py +# + +# Encodings for Neon instructions whose encoding is the same for +# both A32 and T32. + +# More specifically, this covers: +# 2reg scalar ext: 0b1111_1110_xxxx_xxxx_xxxx_1x0x_xxxx_xxxx +# 3same ext: 0b1111_110x_xxxx_xxxx_xxxx_1x0x_xxxx_xxxx diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ +/* + * ARM translation: AArch32 Neon instructions + * + * Copyright (c) 2003 Fabrice Bellard + * Copyright (c) 2005-2007 CodeSourcery + * Copyright (c) 2007 OpenedHand, Ltd. + * Copyright (c) 2020 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 of the License, or (at your option) any later version. + * + * This library is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this library; if not, see <http://www.gnu.org/licenses/>. + */ + +/* + * This file is intended to be included from translate.c; it uses + * some macros and definitions provided by that file. + * It might be possible to convert it to a standalone .c file eventually. + */ + +/* Include the generated Neon decoder */ +#include "decode-neon-dp.inc.c" +#include "decode-neon-ls.inc.c" +#include "decode-neon-shared.inc.c" diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static TCGv_ptr vfp_reg_ptr(bool dp, int reg) #define ARM_CP_RW_BIT (1 << 20) -/* Include the VFP decoder */ +/* Include the VFP and Neon decoders */ #include "translate-vfp.inc.c" +#include "translate-neon.inc.c" static inline void iwmmxt_load_reg(TCGv_i64 var, int reg) { @@ -XXX,XX +XXX,XX @@ static void disas_arm_insn(DisasContext *s, unsigned int insn) /* Unconditional instructions. */ /* TODO: Perhaps merge these into one decodetree output file. */ if (disas_a32_uncond(s, insn) || - disas_vfp_uncond(s, insn)) { + disas_vfp_uncond(s, insn) || + disas_neon_dp(s, insn) || + disas_neon_ls(s, insn) || + disas_neon_shared(s, insn)) { return; } /* fall back to legacy decoder */ @@ -XXX,XX +XXX,XX @@ static void disas_thumb2_insn(DisasContext *s, uint32_t insn) ARCH(6T2); } + if ((insn & 0xef000000) == 0xef000000) { + /* + * T32 encodings 0b111p_1111_qqqq_qqqq_qqqq_qqqq_qqqq_qqqq + * transform into + * A32 encodings 0b1111_001p_qqqq_qqqq_qqqq_qqqq_qqqq_qqqq + */ + uint32_t a32_insn = (insn & 0xe2ffffff) | + ((insn & (1 << 28)) >> 4) | (1 << 28); + + if (disas_neon_dp(s, a32_insn)) { + return; + } + } + + if ((insn & 0xff100000) == 0xf9000000) { + /* + * T32 encodings 0b1111_1001_ppp0_qqqq_qqqq_qqqq_qqqq_qqqq + * transform into + * A32 encodings 0b1111_0100_ppp0_qqqq_qqqq_qqqq_qqqq_qqqq + */ + uint32_t a32_insn = (insn & 0x00ffffff) | 0xf4000000; + + if (disas_neon_ls(s, a32_insn)) { + return; + } + } + /* * TODO: Perhaps merge these into one decodetree output file. * Note disas_vfp is written for a32 with cond field in the @@ -XXX,XX +XXX,XX @@ static void disas_thumb2_insn(DisasContext *s, uint32_t insn) */ if (disas_t32(s, insn) || disas_vfp_uncond(s, insn) || + disas_neon_shared(s, insn) || ((insn >> 28) == 0xe && disas_vfp(s, insn))) { return; } diff --git a/target/arm/Makefile.objs b/target/arm/Makefile.objs index XXXXXXX..XXXXXXX 100644 --- a/target/arm/Makefile.objs +++ b/target/arm/Makefile.objs @@ -XXX,XX +XXX,XX @@ target/arm/decode-sve.inc.c: $(SRC_PATH)/target/arm/sve.decode $(DECODETREE) $(PYTHON) $(DECODETREE) --decode disas_sve -o $@ $<,\ "GEN", $(TARGET_DIR)$@) +target/arm/decode-neon-shared.inc.c: $(SRC_PATH)/target/arm/neon-shared.decode $(DECODETREE) + $(call quiet-command,\ + $(PYTHON) $(DECODETREE) --static-decode disas_neon_shared -o $@ $<,\ + "GEN", $(TARGET_DIR)$@) + +target/arm/decode-neon-dp.inc.c: $(SRC_PATH)/target/arm/neon-dp.decode $(DECODETREE) + $(call quiet-command,\ + $(PYTHON) $(DECODETREE) --static-decode disas_neon_dp -o $@ $<,\ + "GEN", $(TARGET_DIR)$@) + +target/arm/decode-neon-ls.inc.c: $(SRC_PATH)/target/arm/neon-ls.decode $(DECODETREE) + $(call quiet-command,\ + $(PYTHON) $(DECODETREE) --static-decode disas_neon_ls -o $@ $<,\ + "GEN", $(TARGET_DIR)$@) + target/arm/decode-vfp.inc.c: $(SRC_PATH)/target/arm/vfp.decode $(DECODETREE) $(call quiet-command,\ $(PYTHON) $(DECODETREE) --static-decode disas_vfp -o $@ $<,\ @@ -XXX,XX +XXX,XX @@ target/arm/decode-t16.inc.c: $(SRC_PATH)/target/arm/t16.decode $(DECODETREE) "GEN", $(TARGET_DIR)$@) target/arm/translate-sve.o: target/arm/decode-sve.inc.c +target/arm/translate.o: target/arm/decode-neon-shared.inc.c +target/arm/translate.o: target/arm/decode-neon-dp.inc.c +target/arm/translate.o: target/arm/decode-neon-ls.inc.c target/arm/translate.o: target/arm/decode-vfp.inc.c target/arm/translate.o: target/arm/decode-vfp-uncond.inc.c target/arm/translate.o: target/arm/decode-a32.inc.c -- 2.20.1 Convert the VCMLA (vector) insns in the 3same extension group to decodetree. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-5-peter.maydell@linaro.org --- target/arm/neon-shared.decode | 11 ++++++++++ target/arm/translate-neon.inc.c | 37 +++++++++++++++++++++++++++++++++ target/arm/translate.c | 11 +--------- 3 files changed, 49 insertions(+), 10 deletions(-) diff --git a/target/arm/neon-shared.decode b/target/arm/neon-shared.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-shared.decode +++ b/target/arm/neon-shared.decode @@ -XXX,XX +XXX,XX @@ # More specifically, this covers: # 2reg scalar ext: 0b1111_1110_xxxx_xxxx_xxxx_1x0x_xxxx_xxxx # 3same ext: 0b1111_110x_xxxx_xxxx_xxxx_1x0x_xxxx_xxxx + +# VFP/Neon register fields; same as vfp.decode +%vm_dp 5:1 0:4 +%vm_sp 0:4 5:1 +%vn_dp 7:1 16:4 +%vn_sp 16:4 7:1 +%vd_dp 22:1 12:4 +%vd_sp 12:4 22:1 + +VCMLA 1111 110 rot:2 . 1 size:1 .... .... 1000 . q:1 . 0 .... \ + vm=%vm_dp vn=%vn_dp vd=%vd_dp diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ #include "decode-neon-dp.inc.c" #include "decode-neon-ls.inc.c" #include "decode-neon-shared.inc.c" + +static bool trans_VCMLA(DisasContext *s, arg_VCMLA *a) +{ + int opr_sz; + TCGv_ptr fpst; + gen_helper_gvec_3_ptr *fn_gvec_ptr; + + if (!dc_isar_feature(aa32_vcma, s) + || (!a->size && !dc_isar_feature(aa32_fp16_arith, s))) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist. */ + if (!dc_isar_feature(aa32_simd_r32, s) && + ((a->vd | a->vn | a->vm) & 0x10)) { + return false; + } + + if ((a->vn | a->vm | a->vd) & a->q) { + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + opr_sz = (1 + a->q) * 8; + fpst = get_fpstatus_ptr(1); + fn_gvec_ptr = a->size ? gen_helper_gvec_fcmlas : gen_helper_gvec_fcmlah; + tcg_gen_gvec_3_ptr(vfp_reg_offset(1, a->vd), + vfp_reg_offset(1, a->vn), + vfp_reg_offset(1, a->vm), + fpst, opr_sz, opr_sz, a->rot, + fn_gvec_ptr); + tcg_temp_free_ptr(fpst); + return true; +} diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static int disas_neon_insn_3same_ext(DisasContext *s, uint32_t insn) bool is_long = false, q = extract32(insn, 6, 1); bool ptr_is_env = false; - if ((insn & 0xfe200f10) == 0xfc200800) { - /* VCMLA -- 1111 110R R.1S .... .... 1000 ...0 .... */ - int size = extract32(insn, 20, 1); - data = extract32(insn, 23, 2); /* rot */ - if (!dc_isar_feature(aa32_vcma, s) - || (!size && !dc_isar_feature(aa32_fp16_arith, s))) { - return 1; - } - fn_gvec_ptr = size ? gen_helper_gvec_fcmlas : gen_helper_gvec_fcmlah; - } else if ((insn & 0xfea00f10) == 0xfc800800) { + if ((insn & 0xfea00f10) == 0xfc800800) { /* VCADD -- 1111 110R 1.0S .... .... 1000 ...0 .... */ int size = extract32(insn, 20, 1); data = extract32(insn, 24, 1); /* rot */ -- 2.20.1 Convert the VCADD (vector) insns to decodetree. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-6-peter.maydell@linaro.org --- target/arm/neon-shared.decode | 3 +++ target/arm/translate-neon.inc.c | 37 +++++++++++++++++++++++++++++++++ target/arm/translate.c | 11 +--------- 3 files changed, 41 insertions(+), 10 deletions(-) diff --git a/target/arm/neon-shared.decode b/target/arm/neon-shared.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-shared.decode +++ b/target/arm/neon-shared.decode @@ -XXX,XX +XXX,XX @@ VCMLA 1111 110 rot:2 . 1 size:1 .... .... 1000 . q:1 . 0 .... \ vm=%vm_dp vn=%vn_dp vd=%vd_dp + +VCADD 1111 110 rot:1 1 . 0 size:1 .... .... 1000 . q:1 . 0 .... \ + vm=%vm_dp vn=%vn_dp vd=%vd_dp diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool trans_VCMLA(DisasContext *s, arg_VCMLA *a) tcg_temp_free_ptr(fpst); return true; } + +static bool trans_VCADD(DisasContext *s, arg_VCADD *a) +{ + int opr_sz; + TCGv_ptr fpst; + gen_helper_gvec_3_ptr *fn_gvec_ptr; + + if (!dc_isar_feature(aa32_vcma, s) + || (!a->size && !dc_isar_feature(aa32_fp16_arith, s))) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist. */ + if (!dc_isar_feature(aa32_simd_r32, s) && + ((a->vd | a->vn | a->vm) & 0x10)) { + return false; + } + + if ((a->vn | a->vm | a->vd) & a->q) { + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + opr_sz = (1 + a->q) * 8; + fpst = get_fpstatus_ptr(1); + fn_gvec_ptr = a->size ? gen_helper_gvec_fcadds : gen_helper_gvec_fcaddh; + tcg_gen_gvec_3_ptr(vfp_reg_offset(1, a->vd), + vfp_reg_offset(1, a->vn), + vfp_reg_offset(1, a->vm), + fpst, opr_sz, opr_sz, a->rot, + fn_gvec_ptr); + tcg_temp_free_ptr(fpst); + return true; +} diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static int disas_neon_insn_3same_ext(DisasContext *s, uint32_t insn) bool is_long = false, q = extract32(insn, 6, 1); bool ptr_is_env = false; - if ((insn & 0xfea00f10) == 0xfc800800) { - /* VCADD -- 1111 110R 1.0S .... .... 1000 ...0 .... */ - int size = extract32(insn, 20, 1); - data = extract32(insn, 24, 1); /* rot */ - if (!dc_isar_feature(aa32_vcma, s) - || (!size && !dc_isar_feature(aa32_fp16_arith, s))) { - return 1; - } - fn_gvec_ptr = size ? gen_helper_gvec_fcadds : gen_helper_gvec_fcaddh; - } else if ((insn & 0xfeb00f00) == 0xfc200d00) { + if ((insn & 0xfeb00f00) == 0xfc200d00) { /* V[US]DOT -- 1111 1100 0.10 .... .... 1101 .Q.U .... */ bool u = extract32(insn, 4, 1); if (!dc_isar_feature(aa32_dp, s)) { -- 2.20.1 Convert the V[US]DOT (vector) insns to decodetree. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-7-peter.maydell@linaro.org --- target/arm/neon-shared.decode | 4 ++++ target/arm/translate-neon.inc.c | 32 ++++++++++++++++++++++++++++++++ target/arm/translate.c | 9 +-------- 3 files changed, 37 insertions(+), 8 deletions(-) diff --git a/target/arm/neon-shared.decode b/target/arm/neon-shared.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-shared.decode +++ b/target/arm/neon-shared.decode @@ -XXX,XX +XXX,XX @@ VCMLA 1111 110 rot:2 . 1 size:1 .... .... 1000 . q:1 . 0 .... \ VCADD 1111 110 rot:1 1 . 0 size:1 .... .... 1000 . q:1 . 0 .... \ vm=%vm_dp vn=%vn_dp vd=%vd_dp + +# VUDOT and VSDOT +VDOT 1111 110 00 . 10 .... .... 1101 . q:1 . u:1 .... \ + vm=%vm_dp vn=%vn_dp vd=%vd_dp diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool trans_VCADD(DisasContext *s, arg_VCADD *a) tcg_temp_free_ptr(fpst); return true; } + +static bool trans_VDOT(DisasContext *s, arg_VDOT *a) +{ + int opr_sz; + gen_helper_gvec_3 *fn_gvec; + + if (!dc_isar_feature(aa32_dp, s)) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist. */ + if (!dc_isar_feature(aa32_simd_r32, s) && + ((a->vd | a->vn | a->vm) & 0x10)) { + return false; + } + + if ((a->vn | a->vm | a->vd) & a->q) { + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + opr_sz = (1 + a->q) * 8; + fn_gvec = a->u ? gen_helper_gvec_udot_b : gen_helper_gvec_sdot_b; + tcg_gen_gvec_3_ool(vfp_reg_offset(1, a->vd), + vfp_reg_offset(1, a->vn), + vfp_reg_offset(1, a->vm), + opr_sz, opr_sz, 0, fn_gvec); + return true; +} diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static int disas_neon_insn_3same_ext(DisasContext *s, uint32_t insn) bool is_long = false, q = extract32(insn, 6, 1); bool ptr_is_env = false; - if ((insn & 0xfeb00f00) == 0xfc200d00) { - /* V[US]DOT -- 1111 1100 0.10 .... .... 1101 .Q.U .... */ - bool u = extract32(insn, 4, 1); - if (!dc_isar_feature(aa32_dp, s)) { - return 1; - } - fn_gvec = u ? gen_helper_gvec_udot_b : gen_helper_gvec_sdot_b; - } else if ((insn & 0xff300f10) == 0xfc200810) { + if ((insn & 0xff300f10) == 0xfc200810) { /* VFM[AS]L -- 1111 1100 S.10 .... .... 1000 .Q.1 .... */ int is_s = extract32(insn, 23, 1); if (!dc_isar_feature(aa32_fhm, s)) { -- 2.20.1 Convert the VFM[AS]L (vector) insns to decodetree. This is the last insn in the legacy decoder for the 3same_ext group, so we can delete the legacy decoder function for the group entirely. Note that in disas_thumb2_insn() the parts of this encoding space where the decodetree decoder returns false will correctly be directed to illegal_op by the "(insn & (1 << 28))" check so they won't fall into disas_coproc_insn() by mistake. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-8-peter.maydell@linaro.org --- target/arm/neon-shared.decode | 6 +++ target/arm/translate-neon.inc.c | 31 +++++++++++ target/arm/translate.c | 92 +-------------------------------- 3 files changed, 38 insertions(+), 91 deletions(-) diff --git a/target/arm/neon-shared.decode b/target/arm/neon-shared.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-shared.decode +++ b/target/arm/neon-shared.decode @@ -XXX,XX +XXX,XX @@ VCADD 1111 110 rot:1 1 . 0 size:1 .... .... 1000 . q:1 . 0 .... \ # VUDOT and VSDOT VDOT 1111 110 00 . 10 .... .... 1101 . q:1 . u:1 .... \ vm=%vm_dp vn=%vn_dp vd=%vd_dp + +# VFM[AS]L +VFML 1111 110 0 s:1 . 10 .... .... 1000 . 0 . 1 .... \ + vm=%vm_sp vn=%vn_sp vd=%vd_dp q=0 +VFML 1111 110 0 s:1 . 10 .... .... 1000 . 1 . 1 .... \ + vm=%vm_dp vn=%vn_dp vd=%vd_dp q=1 diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool trans_VDOT(DisasContext *s, arg_VDOT *a) opr_sz, opr_sz, 0, fn_gvec); return true; } + +static bool trans_VFML(DisasContext *s, arg_VFML *a) +{ + int opr_sz; + + if (!dc_isar_feature(aa32_fhm, s)) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist. */ + if (!dc_isar_feature(aa32_simd_r32, s) && + (a->vd & 0x10)) { + return false; + } + + if (a->vd & a->q) { + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + opr_sz = (1 + a->q) * 8; + tcg_gen_gvec_3_ptr(vfp_reg_offset(1, a->vd), + vfp_reg_offset(a->q, a->vn), + vfp_reg_offset(a->q, a->vm), + cpu_env, opr_sz, opr_sz, a->s, /* is_2 == 0 */ + gen_helper_gvec_fmlal_a32); + return true; +} diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) return 0; } -/* Advanced SIMD three registers of the same length extension. - * 31 25 23 22 20 16 12 11 10 9 8 3 0 - * +---------------+-----+---+-----+----+----+---+----+---+----+---------+----+ - * | 1 1 1 1 1 1 0 | op1 | D | op2 | Vn | Vd | 1 | o3 | 0 | o4 | N Q M U | Vm | - * +---------------+-----+---+-----+----+----+---+----+---+----+---------+----+ - */ -static int disas_neon_insn_3same_ext(DisasContext *s, uint32_t insn) -{ - gen_helper_gvec_3 *fn_gvec = NULL; - gen_helper_gvec_3_ptr *fn_gvec_ptr = NULL; - int rd, rn, rm, opr_sz; - int data = 0; - int off_rn, off_rm; - bool is_long = false, q = extract32(insn, 6, 1); - bool ptr_is_env = false; - - if ((insn & 0xff300f10) == 0xfc200810) { - /* VFM[AS]L -- 1111 1100 S.10 .... .... 1000 .Q.1 .... */ - int is_s = extract32(insn, 23, 1); - if (!dc_isar_feature(aa32_fhm, s)) { - return 1; - } - is_long = true; - data = is_s; /* is_2 == 0 */ - fn_gvec_ptr = gen_helper_gvec_fmlal_a32; - ptr_is_env = true; - } else { - return 1; - } - - VFP_DREG_D(rd, insn); - if (rd & q) { - return 1; - } - if (q || !is_long) { - VFP_DREG_N(rn, insn); - VFP_DREG_M(rm, insn); - if ((rn | rm) & q & !is_long) { - return 1; - } - off_rn = vfp_reg_offset(1, rn); - off_rm = vfp_reg_offset(1, rm); - } else { - rn = VFP_SREG_N(insn); - rm = VFP_SREG_M(insn); - off_rn = vfp_reg_offset(0, rn); - off_rm = vfp_reg_offset(0, rm); - } - - if (s->fp_excp_el) { - gen_exception_insn(s, s->pc_curr, EXCP_UDEF, - syn_simd_access_trap(1, 0xe, false), s->fp_excp_el); - return 0; - } - if (!s->vfp_enabled) { - return 1; - } - - opr_sz = (1 + q) * 8; - if (fn_gvec_ptr) { - TCGv_ptr ptr; - if (ptr_is_env) { - ptr = cpu_env; - } else { - ptr = get_fpstatus_ptr(1); - } - tcg_gen_gvec_3_ptr(vfp_reg_offset(1, rd), off_rn, off_rm, ptr, - opr_sz, opr_sz, data, fn_gvec_ptr); - if (!ptr_is_env) { - tcg_temp_free_ptr(ptr); - } - } else { - tcg_gen_gvec_3_ool(vfp_reg_offset(1, rd), off_rn, off_rm, - opr_sz, opr_sz, data, fn_gvec); - } - return 0; -} - /* Advanced SIMD two registers and a scalar extension. * 31 24 23 22 20 16 12 11 10 9 8 3 0 * +-----------------+----+---+----+----+----+---+----+---+----+---------+----+ @@ -XXX,XX +XXX,XX @@ static void disas_arm_insn(DisasContext *s, unsigned int insn) } } } - } else if ((insn & 0x0e000a00) == 0x0c000800 - && arm_dc_feature(s, ARM_FEATURE_V8)) { - if (disas_neon_insn_3same_ext(s, insn)) { - goto illegal_op; - } - return; } else if ((insn & 0x0f000a00) == 0x0e000800 && arm_dc_feature(s, ARM_FEATURE_V8)) { if (disas_neon_insn_2reg_scalar_ext(s, insn)) { @@ -XXX,XX +XXX,XX @@ static void disas_thumb2_insn(DisasContext *s, uint32_t insn) } break; } - if ((insn & 0xfe000a00) == 0xfc000800 + if ((insn & 0xff000a00) == 0xfe000800 && arm_dc_feature(s, ARM_FEATURE_V8)) { /* The Thumb2 and ARM encodings are identical. */ - if (disas_neon_insn_3same_ext(s, insn)) { - goto illegal_op; - } - } else if ((insn & 0xff000a00) == 0xfe000800 - && arm_dc_feature(s, ARM_FEATURE_V8)) { - /* The Thumb2 and ARM encodings are identical. */ if (disas_neon_insn_2reg_scalar_ext(s, insn)) { goto illegal_op; } -- 2.20.1 Convert VCMLA (scalar) in the 2reg-scalar-ext group to decodetree. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-9-peter.maydell@linaro.org --- target/arm/neon-shared.decode | 5 +++++ target/arm/translate-neon.inc.c | 40 +++++++++++++++++++++++++++++++++ target/arm/translate.c | 26 +-------------------- 3 files changed, 46 insertions(+), 25 deletions(-) diff --git a/target/arm/neon-shared.decode b/target/arm/neon-shared.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-shared.decode +++ b/target/arm/neon-shared.decode @@ -XXX,XX +XXX,XX @@ VFML 1111 110 0 s:1 . 10 .... .... 1000 . 0 . 1 .... \ vm=%vm_sp vn=%vn_sp vd=%vd_dp q=0 VFML 1111 110 0 s:1 . 10 .... .... 1000 . 1 . 1 .... \ vm=%vm_dp vn=%vn_dp vd=%vd_dp q=1 + +VCMLA_scalar 1111 1110 0 . rot:2 .... .... 1000 . q:1 index:1 0 vm:4 \ + vn=%vn_dp vd=%vd_dp size=0 +VCMLA_scalar 1111 1110 1 . rot:2 .... .... 1000 . q:1 . 0 .... \ + vm=%vm_dp vn=%vn_dp vd=%vd_dp size=1 index=0 diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool trans_VFML(DisasContext *s, arg_VFML *a) gen_helper_gvec_fmlal_a32); return true; } + +static bool trans_VCMLA_scalar(DisasContext *s, arg_VCMLA_scalar *a) +{ + gen_helper_gvec_3_ptr *fn_gvec_ptr; + int opr_sz; + TCGv_ptr fpst; + + if (!dc_isar_feature(aa32_vcma, s)) { + return false; + } + if (a->size == 0 && !dc_isar_feature(aa32_fp16_arith, s)) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist. */ + if (!dc_isar_feature(aa32_simd_r32, s) && + ((a->vd | a->vn | a->vm) & 0x10)) { + return false; + } + + if ((a->vd | a->vn) & a->q) { + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + fn_gvec_ptr = (a->size ? gen_helper_gvec_fcmlas_idx + : gen_helper_gvec_fcmlah_idx); + opr_sz = (1 + a->q) * 8; + fpst = get_fpstatus_ptr(1); + tcg_gen_gvec_3_ptr(vfp_reg_offset(1, a->vd), + vfp_reg_offset(1, a->vn), + vfp_reg_offset(1, a->vm), + fpst, opr_sz, opr_sz, + (a->index << 2) | a->rot, fn_gvec_ptr); + tcg_temp_free_ptr(fpst); + return true; +} diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static int disas_neon_insn_2reg_scalar_ext(DisasContext *s, uint32_t insn) bool is_long = false, q = extract32(insn, 6, 1); bool ptr_is_env = false; - if ((insn & 0xff000f10) == 0xfe000800) { - /* VCMLA (indexed) -- 1111 1110 S.RR .... .... 1000 ...0 .... */ - int rot = extract32(insn, 20, 2); - int size = extract32(insn, 23, 1); - int index; - - if (!dc_isar_feature(aa32_vcma, s)) { - return 1; - } - if (size == 0) { - if (!dc_isar_feature(aa32_fp16_arith, s)) { - return 1; - } - /* For fp16, rm is just Vm, and index is M. */ - rm = extract32(insn, 0, 4); - index = extract32(insn, 5, 1); - } else { - /* For fp32, rm is the usual M:Vm, and index is 0. */ - VFP_DREG_M(rm, insn); - index = 0; - } - data = (index << 2) | rot; - fn_gvec_ptr = (size ? gen_helper_gvec_fcmlas_idx - : gen_helper_gvec_fcmlah_idx); - } else if ((insn & 0xffb00f00) == 0xfe200d00) { + if ((insn & 0xffb00f00) == 0xfe200d00) { /* V[US]DOT -- 1111 1110 0.10 .... .... 1101 .Q.U .... */ int u = extract32(insn, 4, 1); -- 2.20.1 Convert the V[US]DOT (scalar) insns in the 2reg-scalar-ext group to decodetree. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-10-peter.maydell@linaro.org --- target/arm/neon-shared.decode | 3 +++ target/arm/translate-neon.inc.c | 35 +++++++++++++++++++++++++++++++++ target/arm/translate.c | 13 +----------- 3 files changed, 39 insertions(+), 12 deletions(-) diff --git a/target/arm/neon-shared.decode b/target/arm/neon-shared.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-shared.decode +++ b/target/arm/neon-shared.decode @@ -XXX,XX +XXX,XX @@ VCMLA_scalar 1111 1110 0 . rot:2 .... .... 1000 . q:1 index:1 0 vm:4 \ vn=%vn_dp vd=%vd_dp size=0 VCMLA_scalar 1111 1110 1 . rot:2 .... .... 1000 . q:1 . 0 .... \ vm=%vm_dp vn=%vn_dp vd=%vd_dp size=1 index=0 + +VDOT_scalar 1111 1110 0 . 10 .... .... 1101 . q:1 index:1 u:1 rm:4 \ + vm=%vm_dp vn=%vn_dp vd=%vd_dp diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool trans_VCMLA_scalar(DisasContext *s, arg_VCMLA_scalar *a) tcg_temp_free_ptr(fpst); return true; } + +static bool trans_VDOT_scalar(DisasContext *s, arg_VDOT_scalar *a) +{ + gen_helper_gvec_3 *fn_gvec; + int opr_sz; + TCGv_ptr fpst; + + if (!dc_isar_feature(aa32_dp, s)) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist. */ + if (!dc_isar_feature(aa32_simd_r32, s) && + ((a->vd | a->vn) & 0x10)) { + return false; + } + + if ((a->vd | a->vn) & a->q) { + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + fn_gvec = a->u ? gen_helper_gvec_udot_idx_b : gen_helper_gvec_sdot_idx_b; + opr_sz = (1 + a->q) * 8; + fpst = get_fpstatus_ptr(1); + tcg_gen_gvec_3_ool(vfp_reg_offset(1, a->vd), + vfp_reg_offset(1, a->vn), + vfp_reg_offset(1, a->rm), + opr_sz, opr_sz, a->index, fn_gvec); + tcg_temp_free_ptr(fpst); + return true; +} diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static int disas_neon_insn_2reg_scalar_ext(DisasContext *s, uint32_t insn) bool is_long = false, q = extract32(insn, 6, 1); bool ptr_is_env = false; - if ((insn & 0xffb00f00) == 0xfe200d00) { - /* V[US]DOT -- 1111 1110 0.10 .... .... 1101 .Q.U .... */ - int u = extract32(insn, 4, 1); - - if (!dc_isar_feature(aa32_dp, s)) { - return 1; - } - fn_gvec = u ? gen_helper_gvec_udot_idx_b : gen_helper_gvec_sdot_idx_b; - /* rm is just Vm, and index is M. */ - data = extract32(insn, 5, 1); /* index */ - rm = extract32(insn, 0, 4); - } else if ((insn & 0xffa00f10) == 0xfe000810) { + if ((insn & 0xffa00f10) == 0xfe000810) { /* VFM[AS]L -- 1111 1110 0.0S .... .... 1000 .Q.1 .... */ int is_s = extract32(insn, 20, 1); int vm20 = extract32(insn, 0, 3); -- 2.20.1 Convert the VFM[AS]L (scalar) insns in the 2reg-scalar-ext group to decodetree. These are the last ones in the group so we can remove all the legacy decode for the group. Note that in disas_thumb2_insn() the parts of this encoding space where the decodetree decoder returns false will correctly be directed to illegal_op by the "(insn & (1 << 28))" check so they won't fall into disas_coproc_insn() by mistake. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-11-peter.maydell@linaro.org --- target/arm/neon-shared.decode | 7 +++ target/arm/translate-neon.inc.c | 32 ++++++++++ target/arm/translate.c | 107 +------------------------------- 3 files changed, 40 insertions(+), 106 deletions(-) diff --git a/target/arm/neon-shared.decode b/target/arm/neon-shared.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-shared.decode +++ b/target/arm/neon-shared.decode @@ -XXX,XX +XXX,XX @@ VCMLA_scalar 1111 1110 1 . rot:2 .... .... 1000 . q:1 . 0 .... \ VDOT_scalar 1111 1110 0 . 10 .... .... 1101 . q:1 index:1 u:1 rm:4 \ vm=%vm_dp vn=%vn_dp vd=%vd_dp + +%vfml_scalar_q0_rm 0:3 5:1 +%vfml_scalar_q1_index 5:1 3:1 +VFML_scalar 1111 1110 0 . 0 s:1 .... .... 1000 . 0 . 1 index:1 ... \ + rm=%vfml_scalar_q0_rm vn=%vn_sp vd=%vd_dp q=0 +VFML_scalar 1111 1110 0 . 0 s:1 .... .... 1000 . 1 . 1 . rm:3 \ + index=%vfml_scalar_q1_index vn=%vn_dp vd=%vd_dp q=1 diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool trans_VDOT_scalar(DisasContext *s, arg_VDOT_scalar *a) tcg_temp_free_ptr(fpst); return true; } + +static bool trans_VFML_scalar(DisasContext *s, arg_VFML_scalar *a) +{ + int opr_sz; + + if (!dc_isar_feature(aa32_fhm, s)) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist. */ + if (!dc_isar_feature(aa32_simd_r32, s) && + ((a->vd & 0x10) || (a->q && (a->vn & 0x10)))) { + return false; + } + + if (a->vd & a->q) { + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + opr_sz = (1 + a->q) * 8; + tcg_gen_gvec_3_ptr(vfp_reg_offset(1, a->vd), + vfp_reg_offset(a->q, a->vn), + vfp_reg_offset(a->q, a->rm), + cpu_env, opr_sz, opr_sz, + (a->index << 2) | a->s, /* is_2 == 0 */ + gen_helper_gvec_fmlal_idx_a32); + return true; +} diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static int disas_dsp_insn(DisasContext *s, uint32_t insn) } #define VFP_REG_SHR(x, n) (((n) > 0) ? (x) >> (n) : (x) << -(n)) -#define VFP_SREG(insn, bigbit, smallbit) \ - ((VFP_REG_SHR(insn, bigbit - 1) & 0x1e) | (((insn) >> (smallbit)) & 1)) #define VFP_DREG(reg, insn, bigbit, smallbit) do { \ if (dc_isar_feature(aa32_simd_r32, s)) { \ reg = (((insn) >> (bigbit)) & 0x0f) \ @@ -XXX,XX +XXX,XX @@ static int disas_dsp_insn(DisasContext *s, uint32_t insn) reg = ((insn) >> (bigbit)) & 0x0f; \ }} while (0) -#define VFP_SREG_D(insn) VFP_SREG(insn, 12, 22) #define VFP_DREG_D(reg, insn) VFP_DREG(reg, insn, 12, 22) -#define VFP_SREG_N(insn) VFP_SREG(insn, 16, 7) #define VFP_DREG_N(reg, insn) VFP_DREG(reg, insn, 16, 7) -#define VFP_SREG_M(insn) VFP_SREG(insn, 0, 5) #define VFP_DREG_M(reg, insn) VFP_DREG(reg, insn, 0, 5) static void gen_neon_dup_low16(TCGv_i32 var) @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) return 0; } -/* Advanced SIMD two registers and a scalar extension. - * 31 24 23 22 20 16 12 11 10 9 8 3 0 - * +-----------------+----+---+----+----+----+---+----+---+----+---------+----+ - * | 1 1 1 1 1 1 1 0 | o1 | D | o2 | Vn | Vd | 1 | o3 | 0 | o4 | N Q M U | Vm | - * +-----------------+----+---+----+----+----+---+----+---+----+---------+----+ - * - */ - -static int disas_neon_insn_2reg_scalar_ext(DisasContext *s, uint32_t insn) -{ - gen_helper_gvec_3 *fn_gvec = NULL; - gen_helper_gvec_3_ptr *fn_gvec_ptr = NULL; - int rd, rn, rm, opr_sz, data; - int off_rn, off_rm; - bool is_long = false, q = extract32(insn, 6, 1); - bool ptr_is_env = false; - - if ((insn & 0xffa00f10) == 0xfe000810) { - /* VFM[AS]L -- 1111 1110 0.0S .... .... 1000 .Q.1 .... */ - int is_s = extract32(insn, 20, 1); - int vm20 = extract32(insn, 0, 3); - int vm3 = extract32(insn, 3, 1); - int m = extract32(insn, 5, 1); - int index; - - if (!dc_isar_feature(aa32_fhm, s)) { - return 1; - } - if (q) { - rm = vm20; - index = m * 2 + vm3; - } else { - rm = vm20 * 2 + m; - index = vm3; - } - is_long = true; - data = (index << 2) | is_s; /* is_2 == 0 */ - fn_gvec_ptr = gen_helper_gvec_fmlal_idx_a32; - ptr_is_env = true; - } else { - return 1; - } - - VFP_DREG_D(rd, insn); - if (rd & q) { - return 1; - } - if (q || !is_long) { - VFP_DREG_N(rn, insn); - if (rn & q & !is_long) { - return 1; - } - off_rn = vfp_reg_offset(1, rn); - off_rm = vfp_reg_offset(1, rm); - } else { - rn = VFP_SREG_N(insn); - off_rn = vfp_reg_offset(0, rn); - off_rm = vfp_reg_offset(0, rm); - } - if (s->fp_excp_el) { - gen_exception_insn(s, s->pc_curr, EXCP_UDEF, - syn_simd_access_trap(1, 0xe, false), s->fp_excp_el); - return 0; - } - if (!s->vfp_enabled) { - return 1; - } - - opr_sz = (1 + q) * 8; - if (fn_gvec_ptr) { - TCGv_ptr ptr; - if (ptr_is_env) { - ptr = cpu_env; - } else { - ptr = get_fpstatus_ptr(1); - } - tcg_gen_gvec_3_ptr(vfp_reg_offset(1, rd), off_rn, off_rm, ptr, - opr_sz, opr_sz, data, fn_gvec_ptr); - if (!ptr_is_env) { - tcg_temp_free_ptr(ptr); - } - } else { - tcg_gen_gvec_3_ool(vfp_reg_offset(1, rd), off_rn, off_rm, - opr_sz, opr_sz, data, fn_gvec); - } - return 0; -} - static int disas_coproc_insn(DisasContext *s, uint32_t insn) { int cpnum, is64, crn, crm, opc1, opc2, isread, rt, rt2; @@ -XXX,XX +XXX,XX @@ static void disas_arm_insn(DisasContext *s, unsigned int insn) } } } - } else if ((insn & 0x0f000a00) == 0x0e000800 - && arm_dc_feature(s, ARM_FEATURE_V8)) { - if (disas_neon_insn_2reg_scalar_ext(s, insn)) { - goto illegal_op; - } - return; } goto illegal_op; } @@ -XXX,XX +XXX,XX @@ static void disas_thumb2_insn(DisasContext *s, uint32_t insn) } break; } - if ((insn & 0xff000a00) == 0xfe000800 - && arm_dc_feature(s, ARM_FEATURE_V8)) { - /* The Thumb2 and ARM encodings are identical. */ - if (disas_neon_insn_2reg_scalar_ext(s, insn)) { - goto illegal_op; - } - } else if (((insn >> 24) & 3) == 3) { + if (((insn >> 24) & 3) == 3) { /* Translate into the equivalent ARM encoding. */ insn = (insn & 0xe2ffffff) | ((insn & (1 << 28)) >> 4) | (1 << 28); if (disas_neon_data_insn(s, insn)) { -- 2.20.1 Convert the Neon "load/store multiple structures" insns to decodetree. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-12-peter.maydell@linaro.org --- target/arm/neon-ls.decode | 7 ++ target/arm/translate-neon.inc.c | 124 ++++++++++++++++++++++++++++++++ target/arm/translate.c | 91 +---------------------- 3 files changed, 133 insertions(+), 89 deletions(-) diff --git a/target/arm/neon-ls.decode b/target/arm/neon-ls.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-ls.decode +++ b/target/arm/neon-ls.decode @@ -XXX,XX +XXX,XX @@ # 0b1111_1001_xxx0_xxxx_xxxx_xxxx_xxxx_xxxx # This file works on the A32 encoding only; calling code for T32 has to # transform the insn into the A32 version first. + +%vd_dp 22:1 12:4 + +# Neon load/store multiple structures + +VLDST_multiple 1111 0100 0 . l:1 0 rn:4 .... itype:4 size:2 align:2 rm:4 \ + vd=%vd_dp diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool trans_VFML_scalar(DisasContext *s, arg_VFML_scalar *a) gen_helper_gvec_fmlal_idx_a32); return true; } + +static struct { + int nregs; + int interleave; + int spacing; +} const neon_ls_element_type[11] = { + {1, 4, 1}, + {1, 4, 2}, + {4, 1, 1}, + {2, 2, 2}, + {1, 3, 1}, + {1, 3, 2}, + {3, 1, 1}, + {1, 1, 1}, + {1, 2, 1}, + {1, 2, 2}, + {2, 1, 1} +}; + +static void gen_neon_ldst_base_update(DisasContext *s, int rm, int rn, + int stride) +{ + if (rm != 15) { + TCGv_i32 base; + + base = load_reg(s, rn); + if (rm == 13) { + tcg_gen_addi_i32(base, base, stride); + } else { + TCGv_i32 index; + index = load_reg(s, rm); + tcg_gen_add_i32(base, base, index); + tcg_temp_free_i32(index); + } + store_reg(s, rn, base); + } +} + +static bool trans_VLDST_multiple(DisasContext *s, arg_VLDST_multiple *a) +{ + /* Neon load/store multiple structures */ + int nregs, interleave, spacing, reg, n; + MemOp endian = s->be_data; + int mmu_idx = get_mem_index(s); + int size = a->size; + TCGv_i64 tmp64; + TCGv_i32 addr, tmp; + + if (!arm_dc_feature(s, ARM_FEATURE_NEON)) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist */ + if (!dc_isar_feature(aa32_simd_r32, s) && (a->vd & 0x10)) { + return false; + } + if (a->itype > 10) { + return false; + } + /* Catch UNDEF cases for bad values of align field */ + switch (a->itype & 0xc) { + case 4: + if (a->align >= 2) { + return false; + } + break; + case 8: + if (a->align == 3) { + return false; + } + break; + default: + break; + } + nregs = neon_ls_element_type[a->itype].nregs; + interleave = neon_ls_element_type[a->itype].interleave; + spacing = neon_ls_element_type[a->itype].spacing; + if (size == 3 && (interleave | spacing) != 1) { + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + /* For our purposes, bytes are always little-endian. */ + if (size == 0) { + endian = MO_LE; + } + /* + * Consecutive little-endian elements from a single register + * can be promoted to a larger little-endian operation. + */ + if (interleave == 1 && endian == MO_LE) { + size = 3; + } + tmp64 = tcg_temp_new_i64(); + addr = tcg_temp_new_i32(); + tmp = tcg_const_i32(1 << size); + load_reg_var(s, addr, a->rn); + for (reg = 0; reg < nregs; reg++) { + for (n = 0; n < 8 >> size; n++) { + int xs; + for (xs = 0; xs < interleave; xs++) { + int tt = a->vd + reg + spacing * xs; + + if (a->l) { + gen_aa32_ld_i64(s, tmp64, addr, mmu_idx, endian | size); + neon_store_element64(tt, n, size, tmp64); + } else { + neon_load_element64(tmp64, tt, n, size); + gen_aa32_st_i64(s, tmp64, addr, mmu_idx, endian | size); + } + tcg_gen_add_i32(addr, addr, tmp); + } + } + } + tcg_temp_free_i32(addr); + tcg_temp_free_i32(tmp); + tcg_temp_free_i64(tmp64); + + gen_neon_ldst_base_update(s, a->rm, a->rn, nregs * interleave * 8); + return true; +} diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static void gen_neon_trn_u16(TCGv_i32 t0, TCGv_i32 t1) } -static struct { - int nregs; - int interleave; - int spacing; -} const neon_ls_element_type[11] = { - {1, 4, 1}, - {1, 4, 2}, - {4, 1, 1}, - {2, 2, 2}, - {1, 3, 1}, - {1, 3, 2}, - {3, 1, 1}, - {1, 1, 1}, - {1, 2, 1}, - {1, 2, 2}, - {2, 1, 1} -}; - /* Translate a NEON load/store element instruction. Return nonzero if the instruction is invalid. */ static int disas_neon_ls_insn(DisasContext *s, uint32_t insn) { int rd, rn, rm; - int op; int nregs; - int interleave; - int spacing; int stride; int size; int reg; int load; - int n; int vec_size; - int mmu_idx; - MemOp endian; TCGv_i32 addr; TCGv_i32 tmp; - TCGv_i32 tmp2; - TCGv_i64 tmp64; if (!arm_dc_feature(s, ARM_FEATURE_NEON)) { return 1; @@ -XXX,XX +XXX,XX @@ static int disas_neon_ls_insn(DisasContext *s, uint32_t insn) rn = (insn >> 16) & 0xf; rm = insn & 0xf; load = (insn & (1 << 21)) != 0; - endian = s->be_data; - mmu_idx = get_mem_index(s); if ((insn & (1 << 23)) == 0) { - /* Load store all elements. */ - op = (insn >> 8) & 0xf; - size = (insn >> 6) & 3; - if (op > 10) - return 1; - /* Catch UNDEF cases for bad values of align field */ - switch (op & 0xc) { - case 4: - if (((insn >> 5) & 1) == 1) { - return 1; - } - break; - case 8: - if (((insn >> 4) & 3) == 3) { - return 1; - } - break; - default: - break; - } - nregs = neon_ls_element_type[op].nregs; - interleave = neon_ls_element_type[op].interleave; - spacing = neon_ls_element_type[op].spacing; - if (size == 3 && (interleave | spacing) != 1) { - return 1; - } - /* For our purposes, bytes are always little-endian. */ - if (size == 0) { - endian = MO_LE; - } - /* Consecutive little-endian elements from a single register - * can be promoted to a larger little-endian operation. - */ - if (interleave == 1 && endian == MO_LE) { - size = 3; - } - tmp64 = tcg_temp_new_i64(); - addr = tcg_temp_new_i32(); - tmp2 = tcg_const_i32(1 << size); - load_reg_var(s, addr, rn); - for (reg = 0; reg < nregs; reg++) { - for (n = 0; n < 8 >> size; n++) { - int xs; - for (xs = 0; xs < interleave; xs++) { - int tt = rd + reg + spacing * xs; - - if (load) { - gen_aa32_ld_i64(s, tmp64, addr, mmu_idx, endian | size); - neon_store_element64(tt, n, size, tmp64); - } else { - neon_load_element64(tmp64, tt, n, size); - gen_aa32_st_i64(s, tmp64, addr, mmu_idx, endian | size); - } - tcg_gen_add_i32(addr, addr, tmp2); - } - } - } - tcg_temp_free_i32(addr); - tcg_temp_free_i32(tmp2); - tcg_temp_free_i64(tmp64); - stride = nregs * interleave * 8; + /* Load store all elements -- handled already by decodetree */ + return 1; } else { size = (insn >> 10) & 3; if (size == 3) { -- 2.20.1 Convert the Neon "load single structure to all lanes" insns to decodetree. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-13-peter.maydell@linaro.org --- target/arm/neon-ls.decode | 5 +++ target/arm/translate-neon.inc.c | 73 +++++++++++++++++++++++++++++++++ target/arm/translate.c | 55 +------------------------ 3 files changed, 80 insertions(+), 53 deletions(-) diff --git a/target/arm/neon-ls.decode b/target/arm/neon-ls.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-ls.decode +++ b/target/arm/neon-ls.decode @@ -XXX,XX +XXX,XX @@ VLDST_multiple 1111 0100 0 . l:1 0 rn:4 .... itype:4 size:2 align:2 rm:4 \ vd=%vd_dp + +# Neon load single element to all lanes + +VLD_all_lanes 1111 0100 1 . 1 0 rn:4 .... 11 n:2 size:2 t:1 a:1 rm:4 \ + vd=%vd_dp diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool trans_VLDST_multiple(DisasContext *s, arg_VLDST_multiple *a) gen_neon_ldst_base_update(s, a->rm, a->rn, nregs * interleave * 8); return true; } + +static bool trans_VLD_all_lanes(DisasContext *s, arg_VLD_all_lanes *a) +{ + /* Neon load single structure to all lanes */ + int reg, stride, vec_size; + int vd = a->vd; + int size = a->size; + int nregs = a->n + 1; + TCGv_i32 addr, tmp; + + if (!arm_dc_feature(s, ARM_FEATURE_NEON)) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist */ + if (!dc_isar_feature(aa32_simd_r32, s) && (a->vd & 0x10)) { + return false; + } + + if (size == 3) { + if (nregs != 4 || a->a == 0) { + return false; + } + /* For VLD4 size == 3 a == 1 means 32 bits at 16 byte alignment */ + size = 2; + } + if (nregs == 1 && a->a == 1 && size == 0) { + return false; + } + if (nregs == 3 && a->a == 1) { + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + /* + * VLD1 to all lanes: T bit indicates how many Dregs to write. + * VLD2/3/4 to all lanes: T bit indicates register stride. + */ + stride = a->t ? 2 : 1; + vec_size = nregs == 1 ? stride * 8 : 8; + + tmp = tcg_temp_new_i32(); + addr = tcg_temp_new_i32(); + load_reg_var(s, addr, a->rn); + for (reg = 0; reg < nregs; reg++) { + gen_aa32_ld_i32(s, tmp, addr, get_mem_index(s), + s->be_data | size); + if ((vd & 1) && vec_size == 16) { + /* + * We cannot write 16 bytes at once because the + * destination is unaligned. + */ + tcg_gen_gvec_dup_i32(size, neon_reg_offset(vd, 0), + 8, 8, tmp); + tcg_gen_gvec_mov(0, neon_reg_offset(vd + 1, 0), + neon_reg_offset(vd, 0), 8, 8); + } else { + tcg_gen_gvec_dup_i32(size, neon_reg_offset(vd, 0), + vec_size, vec_size, tmp); + } + tcg_gen_addi_i32(addr, addr, 1 << size); + vd += stride; + } + tcg_temp_free_i32(tmp); + tcg_temp_free_i32(addr); + + gen_neon_ldst_base_update(s, a->rm, a->rn, (1 << size) * nregs); + + return true; +} diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static int disas_neon_ls_insn(DisasContext *s, uint32_t insn) int size; int reg; int load; - int vec_size; TCGv_i32 addr; TCGv_i32 tmp; @@ -XXX,XX +XXX,XX @@ static int disas_neon_ls_insn(DisasContext *s, uint32_t insn) } else { size = (insn >> 10) & 3; if (size == 3) { - /* Load single element to all lanes. */ - int a = (insn >> 4) & 1; - if (!load) { - return 1; - } - size = (insn >> 6) & 3; - nregs = ((insn >> 8) & 3) + 1; - - if (size == 3) { - if (nregs != 4 || a == 0) { - return 1; - } - /* For VLD4 size==3 a == 1 means 32 bits at 16 byte alignment */ - size = 2; - } - if (nregs == 1 && a == 1 && size == 0) { - return 1; - } - if (nregs == 3 && a == 1) { - return 1; - } - addr = tcg_temp_new_i32(); - load_reg_var(s, addr, rn); - - /* VLD1 to all lanes: bit 5 indicates how many Dregs to write. - * VLD2/3/4 to all lanes: bit 5 indicates register stride. - */ - stride = (insn & (1 << 5)) ? 2 : 1; - vec_size = nregs == 1 ? stride * 8 : 8; - - tmp = tcg_temp_new_i32(); - for (reg = 0; reg < nregs; reg++) { - gen_aa32_ld_i32(s, tmp, addr, get_mem_index(s), - s->be_data | size); - if ((rd & 1) && vec_size == 16) { - /* We cannot write 16 bytes at once because the - * destination is unaligned. - */ - tcg_gen_gvec_dup_i32(size, neon_reg_offset(rd, 0), - 8, 8, tmp); - tcg_gen_gvec_mov(0, neon_reg_offset(rd + 1, 0), - neon_reg_offset(rd, 0), 8, 8); - } else { - tcg_gen_gvec_dup_i32(size, neon_reg_offset(rd, 0), - vec_size, vec_size, tmp); - } - tcg_gen_addi_i32(addr, addr, 1 << size); - rd += stride; - } - tcg_temp_free_i32(tmp); - tcg_temp_free_i32(addr); - stride = (1 << size) * nregs; + /* Load single element to all lanes -- handled by decodetree */ + return 1; } else { /* Single element. */ int idx = (insn >> 4) & 0xf; -- 2.20.1 Convert the Neon "load/store single structure to one lane" insns to decodetree. As this is the last set of insns in the neon load/store group, we can remove the whole disas_neon_ls_insn() function. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-14-peter.maydell@linaro.org --- target/arm/neon-ls.decode | 11 +++ target/arm/translate-neon.inc.c | 89 +++++++++++++++++++ target/arm/translate.c | 147 -------------------------------- 3 files changed, 100 insertions(+), 147 deletions(-) diff --git a/target/arm/neon-ls.decode b/target/arm/neon-ls.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-ls.decode +++ b/target/arm/neon-ls.decode @@ -XXX,XX +XXX,XX @@ VLDST_multiple 1111 0100 0 . l:1 0 rn:4 .... itype:4 size:2 align:2 rm:4 \ VLD_all_lanes 1111 0100 1 . 1 0 rn:4 .... 11 n:2 size:2 t:1 a:1 rm:4 \ vd=%vd_dp + +# Neon load/store single structure to one lane +%imm1_5_p1 5:1 !function=plus1 +%imm1_6_p1 6:1 !function=plus1 + +VLDST_single 1111 0100 1 . l:1 0 rn:4 .... 00 n:2 reg_idx:3 align:1 rm:4 \ + vd=%vd_dp size=0 stride=1 +VLDST_single 1111 0100 1 . l:1 0 rn:4 .... 01 n:2 reg_idx:2 align:2 rm:4 \ + vd=%vd_dp size=1 stride=%imm1_5_p1 +VLDST_single 1111 0100 1 . l:1 0 rn:4 .... 10 n:2 reg_idx:1 align:3 rm:4 \ + vd=%vd_dp size=2 stride=%imm1_6_p1 diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ * It might be possible to convert it to a standalone .c file eventually. */ +static inline int plus1(DisasContext *s, int x) +{ + return x + 1; +} + /* Include the generated Neon decoder */ #include "decode-neon-dp.inc.c" #include "decode-neon-ls.inc.c" @@ -XXX,XX +XXX,XX @@ static bool trans_VLD_all_lanes(DisasContext *s, arg_VLD_all_lanes *a) return true; } + +static bool trans_VLDST_single(DisasContext *s, arg_VLDST_single *a) +{ + /* Neon load/store single structure to one lane */ + int reg; + int nregs = a->n + 1; + int vd = a->vd; + TCGv_i32 addr, tmp; + + if (!arm_dc_feature(s, ARM_FEATURE_NEON)) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist */ + if (!dc_isar_feature(aa32_simd_r32, s) && (a->vd & 0x10)) { + return false; + } + + /* Catch the UNDEF cases. This is unavoidably a bit messy. */ + switch (nregs) { + case 1: + if (((a->align & (1 << a->size)) != 0) || + (a->size == 2 && ((a->align & 3) == 1 || (a->align & 3) == 2))) { + return false; + } + break; + case 3: + if ((a->align & 1) != 0) { + return false; + } + /* fall through */ + case 2: + if (a->size == 2 && (a->align & 2) != 0) { + return false; + } + break; + case 4: + if ((a->size == 2) && ((a->align & 3) == 3)) { + return false; + } + break; + default: + abort(); + } + if ((vd + a->stride * (nregs - 1)) > 31) { + /* + * Attempts to write off the end of the register file are + * UNPREDICTABLE; we choose to UNDEF because otherwise we would + * access off the end of the array that holds the register data. + */ + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + tmp = tcg_temp_new_i32(); + addr = tcg_temp_new_i32(); + load_reg_var(s, addr, a->rn); + /* + * TODO: if we implemented alignment exceptions, we should check + * addr against the alignment encoded in a->align here. + */ + for (reg = 0; reg < nregs; reg++) { + if (a->l) { + gen_aa32_ld_i32(s, tmp, addr, get_mem_index(s), + s->be_data | a->size); + neon_store_element(vd, a->reg_idx, a->size, tmp); + } else { /* Store */ + neon_load_element(tmp, vd, a->reg_idx, a->size); + gen_aa32_st_i32(s, tmp, addr, get_mem_index(s), + s->be_data | a->size); + } + vd += a->stride; + tcg_gen_addi_i32(addr, addr, 1 << a->size); + } + tcg_temp_free_i32(addr); + tcg_temp_free_i32(tmp); + + gen_neon_ldst_base_update(s, a->rm, a->rn, (1 << a->size) * nregs); + + return true; +} diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static void gen_neon_trn_u16(TCGv_i32 t0, TCGv_i32 t1) tcg_temp_free_i32(rd); } - -/* Translate a NEON load/store element instruction. Return nonzero if the - instruction is invalid. */ -static int disas_neon_ls_insn(DisasContext *s, uint32_t insn) -{ - int rd, rn, rm; - int nregs; - int stride; - int size; - int reg; - int load; - TCGv_i32 addr; - TCGv_i32 tmp; - - if (!arm_dc_feature(s, ARM_FEATURE_NEON)) { - return 1; - } - - /* FIXME: this access check should not take precedence over UNDEF - * for invalid encodings; we will generate incorrect syndrome information - * for attempts to execute invalid vfp/neon encodings with FP disabled. - */ - if (s->fp_excp_el) { - gen_exception_insn(s, s->pc_curr, EXCP_UDEF, - syn_simd_access_trap(1, 0xe, false), s->fp_excp_el); - return 0; - } - - if (!s->vfp_enabled) - return 1; - VFP_DREG_D(rd, insn); - rn = (insn >> 16) & 0xf; - rm = insn & 0xf; - load = (insn & (1 << 21)) != 0; - if ((insn & (1 << 23)) == 0) { - /* Load store all elements -- handled already by decodetree */ - return 1; - } else { - size = (insn >> 10) & 3; - if (size == 3) { - /* Load single element to all lanes -- handled by decodetree */ - return 1; - } else { - /* Single element. */ - int idx = (insn >> 4) & 0xf; - int reg_idx; - switch (size) { - case 0: - reg_idx = (insn >> 5) & 7; - stride = 1; - break; - case 1: - reg_idx = (insn >> 6) & 3; - stride = (insn & (1 << 5)) ? 2 : 1; - break; - case 2: - reg_idx = (insn >> 7) & 1; - stride = (insn & (1 << 6)) ? 2 : 1; - break; - default: - abort(); - } - nregs = ((insn >> 8) & 3) + 1; - /* Catch the UNDEF cases. This is unavoidably a bit messy. */ - switch (nregs) { - case 1: - if (((idx & (1 << size)) != 0) || - (size == 2 && ((idx & 3) == 1 || (idx & 3) == 2))) { - return 1; - } - break; - case 3: - if ((idx & 1) != 0) { - return 1; - } - /* fall through */ - case 2: - if (size == 2 && (idx & 2) != 0) { - return 1; - } - break; - case 4: - if ((size == 2) && ((idx & 3) == 3)) { - return 1; - } - break; - default: - abort(); - } - if ((rd + stride * (nregs - 1)) > 31) { - /* Attempts to write off the end of the register file - * are UNPREDICTABLE; we choose to UNDEF because otherwise - * the neon_load_reg() would write off the end of the array. - */ - return 1; - } - tmp = tcg_temp_new_i32(); - addr = tcg_temp_new_i32(); - load_reg_var(s, addr, rn); - for (reg = 0; reg < nregs; reg++) { - if (load) { - gen_aa32_ld_i32(s, tmp, addr, get_mem_index(s), - s->be_data | size); - neon_store_element(rd, reg_idx, size, tmp); - } else { /* Store */ - neon_load_element(tmp, rd, reg_idx, size); - gen_aa32_st_i32(s, tmp, addr, get_mem_index(s), - s->be_data | size); - } - rd += stride; - tcg_gen_addi_i32(addr, addr, 1 << size); - } - tcg_temp_free_i32(addr); - tcg_temp_free_i32(tmp); - stride = nregs * (1 << size); - } - } - if (rm != 15) { - TCGv_i32 base; - - base = load_reg(s, rn); - if (rm == 13) { - tcg_gen_addi_i32(base, base, stride); - } else { - TCGv_i32 index; - index = load_reg(s, rm); - tcg_gen_add_i32(base, base, index); - tcg_temp_free_i32(index); - } - store_reg(s, rn, base); - } - return 0; -} - static inline void gen_neon_narrow(int size, TCGv_i32 dest, TCGv_i64 src) { switch (size) { @@ -XXX,XX +XXX,XX @@ static void disas_arm_insn(DisasContext *s, unsigned int insn) } return; } - if ((insn & 0x0f100000) == 0x04000000) { - /* NEON load/store. */ - if (disas_neon_ls_insn(s, insn)) { - goto illegal_op; - } - return; - } if ((insn & 0x0e000f00) == 0x0c000100) { if (arm_dc_feature(s, ARM_FEATURE_IWMMXT)) { /* iWMMXt register transfer. */ @@ -XXX,XX +XXX,XX @@ static void disas_thumb2_insn(DisasContext *s, uint32_t insn) } break; case 12: - if ((insn & 0x01100000) == 0x01000000) { - if (disas_neon_ls_insn(s, insn)) { - goto illegal_op; - } - break; - } goto illegal_op; default: illegal_op: -- 2.20.1 Convert the Neon 3-reg-same VADD and VSUB insns to decodetree. Note that we don't need the neon_3r_sizes[op] check here because all size values are OK for VADD and VSUB; we'll add this when we convert the first insn that has size restrictions. For this we need one of the GVecGen*Fn typedefs currently in translate-a64.h; move them all to translate.h as a block so they are visible to the 32-bit decoder. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-15-peter.maydell@linaro.org --- target/arm/translate-a64.h | 9 -------- target/arm/translate.h | 9 ++++++++ target/arm/neon-dp.decode | 17 +++++++++++++++ target/arm/translate-neon.inc.c | 38 +++++++++++++++++++++++++++++++++ target/arm/translate.c | 14 ++++-------- 5 files changed, 68 insertions(+), 19 deletions(-) diff --git a/target/arm/translate-a64.h b/target/arm/translate-a64.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a64.h +++ b/target/arm/translate-a64.h @@ -XXX,XX +XXX,XX @@ static inline int vec_full_reg_size(DisasContext *s) bool disas_sve(DisasContext *, uint32_t); -/* Note that the gvec expanders operate on offsets + sizes. */ -typedef void GVecGen2Fn(unsigned, uint32_t, uint32_t, uint32_t, uint32_t); -typedef void GVecGen2iFn(unsigned, uint32_t, uint32_t, int64_t, - uint32_t, uint32_t); -typedef void GVecGen3Fn(unsigned, uint32_t, uint32_t, - uint32_t, uint32_t, uint32_t); -typedef void GVecGen4Fn(unsigned, uint32_t, uint32_t, uint32_t, - uint32_t, uint32_t, uint32_t); - #endif /* TARGET_ARM_TRANSLATE_A64_H */ 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 @@ void gen_sshl_i64(TCGv_i64 d, TCGv_i64 a, TCGv_i64 b); #define dc_isar_feature(name, ctx) \ ({ DisasContext *ctx_ = (ctx); isar_feature_##name(ctx_->isar); }) +/* Note that the gvec expanders operate on offsets + sizes. */ +typedef void GVecGen2Fn(unsigned, uint32_t, uint32_t, uint32_t, uint32_t); +typedef void GVecGen2iFn(unsigned, uint32_t, uint32_t, int64_t, + uint32_t, uint32_t); +typedef void GVecGen3Fn(unsigned, uint32_t, uint32_t, + uint32_t, uint32_t, uint32_t); +typedef void GVecGen4Fn(unsigned, uint32_t, uint32_t, uint32_t, + uint32_t, uint32_t, uint32_t); + #endif /* TARGET_ARM_TRANSLATE_H */ diff --git a/target/arm/neon-dp.decode b/target/arm/neon-dp.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-dp.decode +++ b/target/arm/neon-dp.decode @@ -XXX,XX +XXX,XX @@ # # This file is processed by scripts/decodetree.py # +# VFP/Neon register fields; same as vfp.decode +%vm_dp 5:1 0:4 +%vn_dp 7:1 16:4 +%vd_dp 22:1 12:4 # Encodings for Neon data processing instructions where the T32 encoding # is a simple transformation of the A32 encoding. @@ -XXX,XX +XXX,XX @@ # 0b111p_1111_qqqq_qqqq_qqqq_qqqq_qqqq_qqqq # This file works on the A32 encoding only; calling code for T32 has to # transform the insn into the A32 version first. + +###################################################################### +# 3-reg-same grouping: +# 1111 001 U 0 D sz:2 Vn:4 Vd:4 opc:4 N Q M op Vm:4 +###################################################################### + +&3same vm vn vd q size + +@3same .... ... . . . size:2 .... .... .... . q:1 . . .... \ + &3same vm=%vm_dp vn=%vn_dp vd=%vd_dp + +VADD_3s 1111 001 0 0 . .. .... .... 1000 . . . 0 .... @3same +VSUB_3s 1111 001 1 0 . .. .... .... 1000 . . . 0 .... @3same diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool trans_VLDST_single(DisasContext *s, arg_VLDST_single *a) return true; } + +static bool do_3same(DisasContext *s, arg_3same *a, GVecGen3Fn fn) +{ + int vec_size = a->q ? 16 : 8; + int rd_ofs = neon_reg_offset(a->vd, 0); + int rn_ofs = neon_reg_offset(a->vn, 0); + int rm_ofs = neon_reg_offset(a->vm, 0); + + if (!arm_dc_feature(s, ARM_FEATURE_NEON)) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist. */ + if (!dc_isar_feature(aa32_simd_r32, s) && + ((a->vd | a->vn | a->vm) & 0x10)) { + return false; + } + + if ((a->vn | a->vm | a->vd) & a->q) { + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + fn(a->size, rd_ofs, rn_ofs, rm_ofs, vec_size, vec_size); + return true; +} + +#define DO_3SAME(INSN, FUNC) \ + static bool trans_##INSN##_3s(DisasContext *s, arg_3same *a) \ + { \ + return do_3same(s, a, FUNC); \ + } + +DO_3SAME(VADD, tcg_gen_gvec_add) +DO_3SAME(VSUB, tcg_gen_gvec_sub) diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) } return 0; - case NEON_3R_VADD_VSUB: - if (u) { - tcg_gen_gvec_sub(size, rd_ofs, rn_ofs, rm_ofs, - vec_size, vec_size); - } else { - tcg_gen_gvec_add(size, rd_ofs, rn_ofs, rm_ofs, - vec_size, vec_size); - } - return 0; - case NEON_3R_VQADD: tcg_gen_gvec_4(rd_ofs, offsetof(CPUARMState, vfp.qc), rn_ofs, rm_ofs, vec_size, vec_size, @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) tcg_gen_gvec_3(rd_ofs, rm_ofs, rn_ofs, vec_size, vec_size, u ? &ushl_op[size] : &sshl_op[size]); return 0; + + case NEON_3R_VADD_VSUB: + /* Already handled by decodetree */ + return 1; } if (size == 3) { -- 2.20.1 Convert the Neon logic ops in the 3-reg-same grouping to decodetree. Note that for the logic ops the 'size' field forms part of their decode and the actual operations are always bitwise. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-16-peter.maydell@linaro.org --- target/arm/neon-dp.decode | 12 +++++++++++ target/arm/translate-neon.inc.c | 19 +++++++++++++++++ target/arm/translate.c | 38 +-------------------------------- 3 files changed, 32 insertions(+), 37 deletions(-) diff --git a/target/arm/neon-dp.decode b/target/arm/neon-dp.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-dp.decode +++ b/target/arm/neon-dp.decode @@ -XXX,XX +XXX,XX @@ @3same .... ... . . . size:2 .... .... .... . q:1 . . .... \ &3same vm=%vm_dp vn=%vn_dp vd=%vd_dp +@3same_logic .... ... . . . .. .... .... .... . q:1 .. .... \ + &3same vm=%vm_dp vn=%vn_dp vd=%vd_dp size=0 + +VAND_3s 1111 001 0 0 . 00 .... .... 0001 ... 1 .... @3same_logic +VBIC_3s 1111 001 0 0 . 01 .... .... 0001 ... 1 .... @3same_logic +VORR_3s 1111 001 0 0 . 10 .... .... 0001 ... 1 .... @3same_logic +VORN_3s 1111 001 0 0 . 11 .... .... 0001 ... 1 .... @3same_logic +VEOR_3s 1111 001 1 0 . 00 .... .... 0001 ... 1 .... @3same_logic +VBSL_3s 1111 001 1 0 . 01 .... .... 0001 ... 1 .... @3same_logic +VBIT_3s 1111 001 1 0 . 10 .... .... 0001 ... 1 .... @3same_logic +VBIF_3s 1111 001 1 0 . 11 .... .... 0001 ... 1 .... @3same_logic + VADD_3s 1111 001 0 0 . .. .... .... 1000 . . . 0 .... @3same VSUB_3s 1111 001 1 0 . .. .... .... 1000 . . . 0 .... @3same diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool do_3same(DisasContext *s, arg_3same *a, GVecGen3Fn fn) DO_3SAME(VADD, tcg_gen_gvec_add) DO_3SAME(VSUB, tcg_gen_gvec_sub) +DO_3SAME(VAND, tcg_gen_gvec_and) +DO_3SAME(VBIC, tcg_gen_gvec_andc) +DO_3SAME(VORR, tcg_gen_gvec_or) +DO_3SAME(VORN, tcg_gen_gvec_orc) +DO_3SAME(VEOR, tcg_gen_gvec_xor) + +/* These insns are all gvec_bitsel but with the inputs in various orders. */ +#define DO_3SAME_BITSEL(INSN, O1, O2, O3) \ + static void gen_##INSN##_3s(unsigned vece, uint32_t rd_ofs, \ + uint32_t rn_ofs, uint32_t rm_ofs, \ + uint32_t oprsz, uint32_t maxsz) \ + { \ + tcg_gen_gvec_bitsel(vece, rd_ofs, O1, O2, O3, oprsz, maxsz); \ + } \ + DO_3SAME(INSN, gen_##INSN##_3s) + +DO_3SAME_BITSEL(VBSL, rd_ofs, rn_ofs, rm_ofs) +DO_3SAME_BITSEL(VBIT, rm_ofs, rn_ofs, rd_ofs) +DO_3SAME_BITSEL(VBIF, rm_ofs, rd_ofs, rn_ofs) diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) } return 1; - case NEON_3R_LOGIC: /* Logic ops. */ - switch ((u << 2) | size) { - case 0: /* VAND */ - tcg_gen_gvec_and(0, rd_ofs, rn_ofs, rm_ofs, - vec_size, vec_size); - break; - case 1: /* VBIC */ - tcg_gen_gvec_andc(0, rd_ofs, rn_ofs, rm_ofs, - vec_size, vec_size); - break; - case 2: /* VORR */ - tcg_gen_gvec_or(0, rd_ofs, rn_ofs, rm_ofs, - vec_size, vec_size); - break; - case 3: /* VORN */ - tcg_gen_gvec_orc(0, rd_ofs, rn_ofs, rm_ofs, - vec_size, vec_size); - break; - case 4: /* VEOR */ - tcg_gen_gvec_xor(0, rd_ofs, rn_ofs, rm_ofs, - vec_size, vec_size); - break; - case 5: /* VBSL */ - tcg_gen_gvec_bitsel(MO_8, rd_ofs, rd_ofs, rn_ofs, rm_ofs, - vec_size, vec_size); - break; - case 6: /* VBIT */ - tcg_gen_gvec_bitsel(MO_8, rd_ofs, rm_ofs, rn_ofs, rd_ofs, - vec_size, vec_size); - break; - case 7: /* VBIF */ - tcg_gen_gvec_bitsel(MO_8, rd_ofs, rm_ofs, rd_ofs, rn_ofs, - vec_size, vec_size); - break; - } - return 0; - case NEON_3R_VQADD: tcg_gen_gvec_4(rd_ofs, offsetof(CPUARMState, vfp.qc), rn_ofs, rm_ofs, vec_size, vec_size, @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) return 0; case NEON_3R_VADD_VSUB: + case NEON_3R_LOGIC: /* Already handled by decodetree */ return 1; } -- 2.20.1 Convert the Neon 3-reg-same VMAX and VMIN insns to decodetree. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-17-peter.maydell@linaro.org --- target/arm/neon-dp.decode | 5 +++++ target/arm/translate-neon.inc.c | 14 ++++++++++++++ target/arm/translate.c | 21 ++------------------- 3 files changed, 21 insertions(+), 19 deletions(-) diff --git a/target/arm/neon-dp.decode b/target/arm/neon-dp.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-dp.decode +++ b/target/arm/neon-dp.decode @@ -XXX,XX +XXX,XX @@ VBSL_3s 1111 001 1 0 . 01 .... .... 0001 ... 1 .... @3same_logic VBIT_3s 1111 001 1 0 . 10 .... .... 0001 ... 1 .... @3same_logic VBIF_3s 1111 001 1 0 . 11 .... .... 0001 ... 1 .... @3same_logic +VMAX_S_3s 1111 001 0 0 . .. .... .... 0110 . . . 0 .... @3same +VMAX_U_3s 1111 001 1 0 . .. .... .... 0110 . . . 0 .... @3same +VMIN_S_3s 1111 001 0 0 . .. .... .... 0110 . . . 1 .... @3same +VMIN_U_3s 1111 001 1 0 . .. .... .... 0110 . . . 1 .... @3same + VADD_3s 1111 001 0 0 . .. .... .... 1000 . . . 0 .... @3same VSUB_3s 1111 001 1 0 . .. .... .... 1000 . . . 0 .... @3same diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ DO_3SAME(VEOR, tcg_gen_gvec_xor) DO_3SAME_BITSEL(VBSL, rd_ofs, rn_ofs, rm_ofs) DO_3SAME_BITSEL(VBIT, rm_ofs, rn_ofs, rd_ofs) DO_3SAME_BITSEL(VBIF, rm_ofs, rd_ofs, rn_ofs) + +#define DO_3SAME_NO_SZ_3(INSN, FUNC) \ + static bool trans_##INSN##_3s(DisasContext *s, arg_3same *a) \ + { \ + if (a->size == 3) { \ + return false; \ + } \ + return do_3same(s, a, FUNC); \ + } + +DO_3SAME_NO_SZ_3(VMAX_S, tcg_gen_gvec_smax) +DO_3SAME_NO_SZ_3(VMAX_U, tcg_gen_gvec_umax) +DO_3SAME_NO_SZ_3(VMIN_S, tcg_gen_gvec_smin) +DO_3SAME_NO_SZ_3(VMIN_U, tcg_gen_gvec_umin) diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) rd_ofs, rn_ofs, rm_ofs, vec_size, vec_size); return 0; - case NEON_3R_VMAX: - if (u) { - tcg_gen_gvec_umax(size, rd_ofs, rn_ofs, rm_ofs, - vec_size, vec_size); - } else { - tcg_gen_gvec_smax(size, rd_ofs, rn_ofs, rm_ofs, - vec_size, vec_size); - } - return 0; - case NEON_3R_VMIN: - if (u) { - tcg_gen_gvec_umin(size, rd_ofs, rn_ofs, rm_ofs, - vec_size, vec_size); - } else { - tcg_gen_gvec_smin(size, rd_ofs, rn_ofs, rm_ofs, - vec_size, vec_size); - } - return 0; - case NEON_3R_VSHL: /* Note the operation is vshl vd,vm,vn */ tcg_gen_gvec_3(rd_ofs, rm_ofs, rn_ofs, vec_size, vec_size, @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) case NEON_3R_VADD_VSUB: case NEON_3R_LOGIC: + case NEON_3R_VMAX: + case NEON_3R_VMIN: /* Already handled by decodetree */ return 1; } -- 2.20.1 Convert the Neon comparison ops in the 3-reg-same grouping to decodetree. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-18-peter.maydell@linaro.org --- target/arm/neon-dp.decode | 8 ++++++++ target/arm/translate-neon.inc.c | 22 ++++++++++++++++++++++ target/arm/translate.c | 23 +++-------------------- 3 files changed, 33 insertions(+), 20 deletions(-) diff --git a/target/arm/neon-dp.decode b/target/arm/neon-dp.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-dp.decode +++ b/target/arm/neon-dp.decode @@ -XXX,XX +XXX,XX @@ VBSL_3s 1111 001 1 0 . 01 .... .... 0001 ... 1 .... @3same_logic VBIT_3s 1111 001 1 0 . 10 .... .... 0001 ... 1 .... @3same_logic VBIF_3s 1111 001 1 0 . 11 .... .... 0001 ... 1 .... @3same_logic +VCGT_S_3s 1111 001 0 0 . .. .... .... 0011 . . . 0 .... @3same +VCGT_U_3s 1111 001 1 0 . .. .... .... 0011 . . . 0 .... @3same +VCGE_S_3s 1111 001 0 0 . .. .... .... 0011 . . . 1 .... @3same +VCGE_U_3s 1111 001 1 0 . .. .... .... 0011 . . . 1 .... @3same + VMAX_S_3s 1111 001 0 0 . .. .... .... 0110 . . . 0 .... @3same VMAX_U_3s 1111 001 1 0 . .. .... .... 0110 . . . 0 .... @3same VMIN_S_3s 1111 001 0 0 . .. .... .... 0110 . . . 1 .... @3same @@ -XXX,XX +XXX,XX @@ VMIN_U_3s 1111 001 1 0 . .. .... .... 0110 . . . 1 .... @3same VADD_3s 1111 001 0 0 . .. .... .... 1000 . . . 0 .... @3same VSUB_3s 1111 001 1 0 . .. .... .... 1000 . . . 0 .... @3same + +VTST_3s 1111 001 0 0 . .. .... .... 1000 . . . 1 .... @3same +VCEQ_3s 1111 001 1 0 . .. .... .... 1000 . . . 1 .... @3same diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ DO_3SAME_NO_SZ_3(VMAX_S, tcg_gen_gvec_smax) DO_3SAME_NO_SZ_3(VMAX_U, tcg_gen_gvec_umax) DO_3SAME_NO_SZ_3(VMIN_S, tcg_gen_gvec_smin) DO_3SAME_NO_SZ_3(VMIN_U, tcg_gen_gvec_umin) + +#define DO_3SAME_CMP(INSN, COND) \ + static void gen_##INSN##_3s(unsigned vece, uint32_t rd_ofs, \ + uint32_t rn_ofs, uint32_t rm_ofs, \ + uint32_t oprsz, uint32_t maxsz) \ + { \ + tcg_gen_gvec_cmp(COND, vece, rd_ofs, rn_ofs, rm_ofs, oprsz, maxsz); \ + } \ + DO_3SAME_NO_SZ_3(INSN, gen_##INSN##_3s) + +DO_3SAME_CMP(VCGT_S, TCG_COND_GT) +DO_3SAME_CMP(VCGT_U, TCG_COND_GTU) +DO_3SAME_CMP(VCGE_S, TCG_COND_GE) +DO_3SAME_CMP(VCGE_U, TCG_COND_GEU) +DO_3SAME_CMP(VCEQ, TCG_COND_EQ) + +static void gen_VTST_3s(unsigned vece, uint32_t rd_ofs, uint32_t rn_ofs, + uint32_t rm_ofs, uint32_t oprsz, uint32_t maxsz) +{ + tcg_gen_gvec_3(rd_ofs, rn_ofs, rm_ofs, oprsz, maxsz, &cmtst_op[vece]); +} +DO_3SAME_NO_SZ_3(VTST, gen_VTST_3s) diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) u ? &mls_op[size] : &mla_op[size]); return 0; - case NEON_3R_VTST_VCEQ: - if (u) { /* VCEQ */ - tcg_gen_gvec_cmp(TCG_COND_EQ, size, rd_ofs, rn_ofs, rm_ofs, - vec_size, vec_size); - } else { /* VTST */ - tcg_gen_gvec_3(rd_ofs, rn_ofs, rm_ofs, - vec_size, vec_size, &cmtst_op[size]); - } - return 0; - - case NEON_3R_VCGT: - tcg_gen_gvec_cmp(u ? TCG_COND_GTU : TCG_COND_GT, size, - rd_ofs, rn_ofs, rm_ofs, vec_size, vec_size); - return 0; - - case NEON_3R_VCGE: - tcg_gen_gvec_cmp(u ? TCG_COND_GEU : TCG_COND_GE, size, - rd_ofs, rn_ofs, rm_ofs, vec_size, vec_size); - return 0; - case NEON_3R_VSHL: /* Note the operation is vshl vd,vm,vn */ tcg_gen_gvec_3(rd_ofs, rm_ofs, rn_ofs, vec_size, vec_size, @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) case NEON_3R_LOGIC: case NEON_3R_VMAX: case NEON_3R_VMIN: + case NEON_3R_VTST_VCEQ: + case NEON_3R_VCGT: + case NEON_3R_VCGE: /* Already handled by decodetree */ return 1; } -- 2.20.1 Convert the Neon VQADD/VQSUB insns in the 3-reg-same grouping to decodetree. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-19-peter.maydell@linaro.org --- target/arm/neon-dp.decode | 6 ++++++ target/arm/translate-neon.inc.c | 15 +++++++++++++++ target/arm/translate.c | 14 ++------------ 3 files changed, 23 insertions(+), 12 deletions(-) diff --git a/target/arm/neon-dp.decode b/target/arm/neon-dp.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-dp.decode +++ b/target/arm/neon-dp.decode @@ -XXX,XX +XXX,XX @@ @3same .... ... . . . size:2 .... .... .... . q:1 . . .... \ &3same vm=%vm_dp vn=%vn_dp vd=%vd_dp +VQADD_S_3s 1111 001 0 0 . .. .... .... 0000 . . . 1 .... @3same +VQADD_U_3s 1111 001 1 0 . .. .... .... 0000 . . . 1 .... @3same + @3same_logic .... ... . . . .. .... .... .... . q:1 .. .... \ &3same vm=%vm_dp vn=%vn_dp vd=%vd_dp size=0 @@ -XXX,XX +XXX,XX @@ VBSL_3s 1111 001 1 0 . 01 .... .... 0001 ... 1 .... @3same_logic VBIT_3s 1111 001 1 0 . 10 .... .... 0001 ... 1 .... @3same_logic VBIF_3s 1111 001 1 0 . 11 .... .... 0001 ... 1 .... @3same_logic +VQSUB_S_3s 1111 001 0 0 . .. .... .... 0010 . . . 1 .... @3same +VQSUB_U_3s 1111 001 1 0 . .. .... .... 0010 . . . 1 .... @3same + VCGT_S_3s 1111 001 0 0 . .. .... .... 0011 . . . 0 .... @3same VCGT_U_3s 1111 001 1 0 . .. .... .... 0011 . . . 0 .... @3same VCGE_S_3s 1111 001 0 0 . .. .... .... 0011 . . . 1 .... @3same diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static void gen_VTST_3s(unsigned vece, uint32_t rd_ofs, uint32_t rn_ofs, tcg_gen_gvec_3(rd_ofs, rn_ofs, rm_ofs, oprsz, maxsz, &cmtst_op[vece]); } DO_3SAME_NO_SZ_3(VTST, gen_VTST_3s) + +#define DO_3SAME_GVEC4(INSN, OPARRAY) \ + static void gen_##INSN##_3s(unsigned vece, uint32_t rd_ofs, \ + uint32_t rn_ofs, uint32_t rm_ofs, \ + uint32_t oprsz, uint32_t maxsz) \ + { \ + tcg_gen_gvec_4(rd_ofs, offsetof(CPUARMState, vfp.qc), \ + rn_ofs, rm_ofs, oprsz, maxsz, &OPARRAY[vece]); \ + } \ + DO_3SAME(INSN, gen_##INSN##_3s) + +DO_3SAME_GVEC4(VQADD_S, sqadd_op) +DO_3SAME_GVEC4(VQADD_U, uqadd_op) +DO_3SAME_GVEC4(VQSUB_S, sqsub_op) +DO_3SAME_GVEC4(VQSUB_U, uqsub_op) diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) } return 1; - case NEON_3R_VQADD: - tcg_gen_gvec_4(rd_ofs, offsetof(CPUARMState, vfp.qc), - rn_ofs, rm_ofs, vec_size, vec_size, - (u ? uqadd_op : sqadd_op) + size); - return 0; - - case NEON_3R_VQSUB: - tcg_gen_gvec_4(rd_ofs, offsetof(CPUARMState, vfp.qc), - rn_ofs, rm_ofs, vec_size, vec_size, - (u ? uqsub_op : sqsub_op) + size); - return 0; - case NEON_3R_VMUL: /* VMUL */ if (u) { /* Polynomial case allows only P8. */ @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) case NEON_3R_VTST_VCEQ: case NEON_3R_VCGT: case NEON_3R_VCGE: + case NEON_3R_VQADD: + case NEON_3R_VQSUB: /* Already handled by decodetree */ return 1; } -- 2.20.1 Convert the Neon VMUL, VMLA, VMLS and VSHL insns in the 3-reg-same grouping to decodetree. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-20-peter.maydell@linaro.org --- target/arm/neon-dp.decode | 9 +++++++ target/arm/translate-neon.inc.c | 44 +++++++++++++++++++++++++++++++++ target/arm/translate.c | 28 +++------------------ 3 files changed, 56 insertions(+), 25 deletions(-) diff --git a/target/arm/neon-dp.decode b/target/arm/neon-dp.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-dp.decode +++ b/target/arm/neon-dp.decode @@ -XXX,XX +XXX,XX @@ VCGT_U_3s 1111 001 1 0 . .. .... .... 0011 . . . 0 .... @3same VCGE_S_3s 1111 001 0 0 . .. .... .... 0011 . . . 1 .... @3same VCGE_U_3s 1111 001 1 0 . .. .... .... 0011 . . . 1 .... @3same +VSHL_S_3s 1111 001 0 0 . .. .... .... 0100 . . . 0 .... @3same +VSHL_U_3s 1111 001 1 0 . .. .... .... 0100 . . . 0 .... @3same + VMAX_S_3s 1111 001 0 0 . .. .... .... 0110 . . . 0 .... @3same VMAX_U_3s 1111 001 1 0 . .. .... .... 0110 . . . 0 .... @3same VMIN_S_3s 1111 001 0 0 . .. .... .... 0110 . . . 1 .... @3same @@ -XXX,XX +XXX,XX @@ VSUB_3s 1111 001 1 0 . .. .... .... 1000 . . . 0 .... @3same VTST_3s 1111 001 0 0 . .. .... .... 1000 . . . 1 .... @3same VCEQ_3s 1111 001 1 0 . .. .... .... 1000 . . . 1 .... @3same + +VMLA_3s 1111 001 0 0 . .. .... .... 1001 . . . 0 .... @3same +VMLS_3s 1111 001 1 0 . .. .... .... 1001 . . . 0 .... @3same + +VMUL_3s 1111 001 0 0 . .. .... .... 1001 . . . 1 .... @3same +VMUL_p_3s 1111 001 1 0 . .. .... .... 1001 . . . 1 .... @3same diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ DO_3SAME_NO_SZ_3(VMAX_S, tcg_gen_gvec_smax) DO_3SAME_NO_SZ_3(VMAX_U, tcg_gen_gvec_umax) DO_3SAME_NO_SZ_3(VMIN_S, tcg_gen_gvec_smin) DO_3SAME_NO_SZ_3(VMIN_U, tcg_gen_gvec_umin) +DO_3SAME_NO_SZ_3(VMUL, tcg_gen_gvec_mul) #define DO_3SAME_CMP(INSN, COND) \ static void gen_##INSN##_3s(unsigned vece, uint32_t rd_ofs, \ @@ -XXX,XX +XXX,XX @@ DO_3SAME_GVEC4(VQADD_S, sqadd_op) DO_3SAME_GVEC4(VQADD_U, uqadd_op) DO_3SAME_GVEC4(VQSUB_S, sqsub_op) DO_3SAME_GVEC4(VQSUB_U, uqsub_op) + +static void gen_VMUL_p_3s(unsigned vece, uint32_t rd_ofs, uint32_t rn_ofs, + uint32_t rm_ofs, uint32_t oprsz, uint32_t maxsz) +{ + tcg_gen_gvec_3_ool(rd_ofs, rn_ofs, rm_ofs, oprsz, maxsz, + 0, gen_helper_gvec_pmul_b); +} + +static bool trans_VMUL_p_3s(DisasContext *s, arg_3same *a) +{ + if (a->size != 0) { + return false; + } + return do_3same(s, a, gen_VMUL_p_3s); +} + +#define DO_3SAME_GVEC3_NO_SZ_3(INSN, OPARRAY) \ + static void gen_##INSN##_3s(unsigned vece, uint32_t rd_ofs, \ + uint32_t rn_ofs, uint32_t rm_ofs, \ + uint32_t oprsz, uint32_t maxsz) \ + { \ + tcg_gen_gvec_3(rd_ofs, rn_ofs, rm_ofs, \ + oprsz, maxsz, &OPARRAY[vece]); \ + } \ + DO_3SAME_NO_SZ_3(INSN, gen_##INSN##_3s) + + +DO_3SAME_GVEC3_NO_SZ_3(VMLA, mla_op) +DO_3SAME_GVEC3_NO_SZ_3(VMLS, mls_op) + +#define DO_3SAME_GVEC3_SHIFT(INSN, OPARRAY) \ + static void gen_##INSN##_3s(unsigned vece, uint32_t rd_ofs, \ + uint32_t rn_ofs, uint32_t rm_ofs, \ + uint32_t oprsz, uint32_t maxsz) \ + { \ + /* Note the operation is vshl vd,vm,vn */ \ + tcg_gen_gvec_3(rd_ofs, rm_ofs, rn_ofs, \ + oprsz, maxsz, &OPARRAY[vece]); \ + } \ + DO_3SAME(INSN, gen_##INSN##_3s) + +DO_3SAME_GVEC3_SHIFT(VSHL_S, sshl_op) +DO_3SAME_GVEC3_SHIFT(VSHL_U, ushl_op) diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) } return 1; - case NEON_3R_VMUL: /* VMUL */ - if (u) { - /* Polynomial case allows only P8. */ - if (size != 0) { - return 1; - } - tcg_gen_gvec_3_ool(rd_ofs, rn_ofs, rm_ofs, vec_size, vec_size, - 0, gen_helper_gvec_pmul_b); - } else { - tcg_gen_gvec_mul(size, rd_ofs, rn_ofs, rm_ofs, - vec_size, vec_size); - } - return 0; - - case NEON_3R_VML: /* VMLA, VMLS */ - tcg_gen_gvec_3(rd_ofs, rn_ofs, rm_ofs, vec_size, vec_size, - u ? &mls_op[size] : &mla_op[size]); - return 0; - - case NEON_3R_VSHL: - /* Note the operation is vshl vd,vm,vn */ - tcg_gen_gvec_3(rd_ofs, rm_ofs, rn_ofs, vec_size, vec_size, - u ? &ushl_op[size] : &sshl_op[size]); - return 0; - case NEON_3R_VADD_VSUB: case NEON_3R_LOGIC: case NEON_3R_VMAX: @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) case NEON_3R_VCGE: case NEON_3R_VQADD: case NEON_3R_VQSUB: + case NEON_3R_VMUL: + case NEON_3R_VML: + case NEON_3R_VSHL: /* Already handled by decodetree */ return 1; } -- 2.20.1 We're going to want at least some of the NeonGen* typedefs for the refactored 32-bit Neon decoder, so move them all to translate.h since it makes more sense to keep them in one group. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-23-peter.maydell@linaro.org --- target/arm/translate.h | 17 +++++++++++++++++ target/arm/translate-a64.c | 17 ----------------- 2 files changed, 17 insertions(+), 17 deletions(-) diff --git a/target/arm/translate.h b/target/arm/translate.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.h +++ b/target/arm/translate.h @@ -XXX,XX +XXX,XX @@ typedef void GVecGen3Fn(unsigned, uint32_t, uint32_t, typedef void GVecGen4Fn(unsigned, uint32_t, uint32_t, uint32_t, uint32_t, uint32_t, uint32_t); +/* Function prototype for gen_ functions for calling Neon helpers */ +typedef void NeonGenOneOpEnvFn(TCGv_i32, TCGv_ptr, TCGv_i32); +typedef void NeonGenTwoOpFn(TCGv_i32, TCGv_i32, TCGv_i32); +typedef void NeonGenTwoOpEnvFn(TCGv_i32, TCGv_ptr, TCGv_i32, TCGv_i32); +typedef void NeonGenTwo64OpFn(TCGv_i64, TCGv_i64, TCGv_i64); +typedef void NeonGenTwo64OpEnvFn(TCGv_i64, TCGv_ptr, TCGv_i64, TCGv_i64); +typedef void NeonGenNarrowFn(TCGv_i32, TCGv_i64); +typedef void NeonGenNarrowEnvFn(TCGv_i32, TCGv_ptr, TCGv_i64); +typedef void NeonGenWidenFn(TCGv_i64, TCGv_i32); +typedef void NeonGenTwoSingleOPFn(TCGv_i32, TCGv_i32, TCGv_i32, TCGv_ptr); +typedef void NeonGenTwoDoubleOPFn(TCGv_i64, TCGv_i64, TCGv_i64, TCGv_ptr); +typedef void NeonGenOneOpFn(TCGv_i64, TCGv_i64); +typedef void CryptoTwoOpFn(TCGv_ptr, TCGv_ptr); +typedef void CryptoThreeOpIntFn(TCGv_ptr, TCGv_ptr, TCGv_i32); +typedef void CryptoThreeOpFn(TCGv_ptr, TCGv_ptr, TCGv_ptr); +typedef void AtomicThreeOpFn(TCGv_i64, TCGv_i64, TCGv_i64, TCGArg, MemOp); + #endif /* TARGET_ARM_TRANSLATE_H */ diff --git a/target/arm/translate-a64.c b/target/arm/translate-a64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a64.c +++ b/target/arm/translate-a64.c @@ -XXX,XX +XXX,XX @@ typedef struct AArch64DecodeTable { AArch64DecodeFn *disas_fn; } AArch64DecodeTable; -/* Function prototype for gen_ functions for calling Neon helpers */ -typedef void NeonGenOneOpEnvFn(TCGv_i32, TCGv_ptr, TCGv_i32); -typedef void NeonGenTwoOpFn(TCGv_i32, TCGv_i32, TCGv_i32); -typedef void NeonGenTwoOpEnvFn(TCGv_i32, TCGv_ptr, TCGv_i32, TCGv_i32); -typedef void NeonGenTwo64OpFn(TCGv_i64, TCGv_i64, TCGv_i64); -typedef void NeonGenTwo64OpEnvFn(TCGv_i64, TCGv_ptr, TCGv_i64, TCGv_i64); -typedef void NeonGenNarrowFn(TCGv_i32, TCGv_i64); -typedef void NeonGenNarrowEnvFn(TCGv_i32, TCGv_ptr, TCGv_i64); -typedef void NeonGenWidenFn(TCGv_i64, TCGv_i32); -typedef void NeonGenTwoSingleOPFn(TCGv_i32, TCGv_i32, TCGv_i32, TCGv_ptr); -typedef void NeonGenTwoDoubleOPFn(TCGv_i64, TCGv_i64, TCGv_i64, TCGv_ptr); -typedef void NeonGenOneOpFn(TCGv_i64, TCGv_i64); -typedef void CryptoTwoOpFn(TCGv_ptr, TCGv_ptr); -typedef void CryptoThreeOpIntFn(TCGv_ptr, TCGv_ptr, TCGv_i32); -typedef void CryptoThreeOpFn(TCGv_ptr, TCGv_ptr, TCGv_ptr); -typedef void AtomicThreeOpFn(TCGv_i64, TCGv_i64, TCGv_i64, TCGArg, MemOp); - /* initialize TCG globals. */ void a64_translate_init(void) { -- 2.20.1 target-arm queue: mostly aspeed changes from Cédric. thanks -- PMM The following changes since commit 85182c96de61f0b600bbe834d5a23e713162e892: Merge remote-tracking branch 'remotes/dgilbert/tags/pull-migration-20190912a' into staging (2019-09-13 14:37:48 +0100) are available in the Git repository at: https://git.linaro.org/people/pmaydell/qemu-arm.git tags/pull-target-arm-20190913 for you to fetch changes up to 27a296fce9821e3608d537756cffa6e43a46df3b: qemu-ga: Convert invocation documentation to rST (2019-09-13 16:05:01 +0100) ---------------------------------------------------------------- target-arm queue: * aspeed: add a GPIO controller to the SoC * aspeed: Various refactorings * aspeed: Improve DMA controller modelling * atomic_template: fix indentation in GEN_ATOMIC_HELPER * qemu-ga: Convert invocation documentation to rST ---------------------------------------------------------------- Christian Svensson (1): aspeed/smc: Calculate checksum on normal DMA Cédric Le Goater (7): aspeed: Remove unused SoC definitions aspeed: Use consistent typenames aspeed/smc: Add support for DMAs aspeed/smc: Add DMA calibration settings aspeed/smc: Inject errors in DMA checksum aspeed/scu: Introduce per-SoC SCU types aspeed/scu: Introduce a aspeed_scu_get_apb_freq() routine Emilio G. Cota (1): atomic_template: fix indentation in GEN_ATOMIC_HELPER Peter Maydell (1): qemu-ga: Convert invocation documentation to rST Rashmica Gupta (2): hw/gpio: Add basic Aspeed GPIO model for AST2400 and AST2500 aspeed: add a GPIO controller to the SoC Makefile | 24 +- hw/gpio/Makefile.objs | 1 + accel/tcg/atomic_template.h | 2 +- include/hw/arm/aspeed_soc.h | 4 +- include/hw/gpio/aspeed_gpio.h | 100 +++++ include/hw/misc/aspeed_scu.h | 21 +- include/hw/ssi/aspeed_smc.h | 7 + hw/arm/aspeed.c | 2 + hw/arm/aspeed_soc.c | 63 ++- hw/gpio/aspeed_gpio.c | 884 ++++++++++++++++++++++++++++++++++++++++++ hw/misc/aspeed_scu.c | 102 ++--- hw/ssi/aspeed_smc.c | 335 +++++++++++++++- hw/timer/aspeed_timer.c | 3 +- MAINTAINERS | 2 +- docs/conf.py | 18 +- docs/interop/conf.py | 7 + docs/interop/index.rst | 1 + docs/interop/qemu-ga.rst | 133 +++++++ qemu-doc.texi | 5 - qemu-ga.texi | 137 ------- 20 files changed, 1585 insertions(+), 266 deletions(-) create mode 100644 include/hw/gpio/aspeed_gpio.h create mode 100644 hw/gpio/aspeed_gpio.c create mode 100644 docs/interop/qemu-ga.rst delete mode 100644 qemu-ga.texi From: Rashmica Gupta <rashmica.g@gmail.com> GPIO pins are arranged in groups of 8 pins labeled A,B,..,Y,Z,AA,AB,AC. (Note that the ast2400 controller only goes up to group AB). A set has four groups (except set AC which only has one) and is referred to by the groups it is composed of (eg ABCD,EFGH,...,YZAAAB). Each set is accessed and controlled by a bank of 14 registers. These registers operate on a per pin level where each bit in the register corresponds to a pin, except for the command source registers. The command source registers operate on a per group level where bits 24, 16, 8 and 0 correspond to each group in the set. eg. registers for set ABCD: |D7...D0|C7...C0|B7...B0|A7...A0| <- GPIOs |31...24|23...16|15....8|7.....0| <- bit position Note that there are a couple of groups that only have 4 pins. There are two ways that this model deviates from the behaviour of the actual controller: (1) The only control source driving the GPIO pins in the model is the ARM model (as there currently aren't models for the LPC or Coprocessor). (2) None of the registers in the model are reset tolerant (needs integration with the watchdog). Signed-off-by: Rashmica Gupta <rashmica.g@gmail.com> Tested-by: Andrew Jeffery <andrew@aj.id.au> Reviewed-by: Cédric Le Goater <clg@kaod.org> Signed-off-by: Cédric Le Goater <clg@kaod.org> Message-id: 20190904070506.1052-2-clg@kaod.org [clg: fixed missing header files made use of HWADDR_PRIx to fix compilation on windows ] Signed-off-by: Cédric Le Goater <clg@kaod.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/gpio/Makefile.objs | 1 + include/hw/gpio/aspeed_gpio.h | 100 ++++ hw/gpio/aspeed_gpio.c | 884 ++++++++++++++++++++++++++++++++++ 3 files changed, 985 insertions(+) create mode 100644 include/hw/gpio/aspeed_gpio.h create mode 100644 hw/gpio/aspeed_gpio.c diff --git a/hw/gpio/Makefile.objs b/hw/gpio/Makefile.objs index XXXXXXX..XXXXXXX 100644 --- a/hw/gpio/Makefile.objs +++ b/hw/gpio/Makefile.objs @@ -XXX,XX +XXX,XX @@ obj-$(CONFIG_OMAP) += omap_gpio.o obj-$(CONFIG_IMX) += imx_gpio.o obj-$(CONFIG_RASPI) += bcm2835_gpio.o obj-$(CONFIG_NRF51_SOC) += nrf51_gpio.o +obj-$(CONFIG_ASPEED_SOC) += aspeed_gpio.o diff --git a/include/hw/gpio/aspeed_gpio.h b/include/hw/gpio/aspeed_gpio.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/include/hw/gpio/aspeed_gpio.h @@ -XXX,XX +XXX,XX @@ +/* + * ASPEED GPIO Controller + * + * Copyright (C) 2017-2018 IBM Corp. + * + * This code is licensed under the GPL version 2 or later. See + * the COPYING file in the top-level directory. + */ + +#ifndef ASPEED_GPIO_H +#define ASPEED_GPIO_H + +#include "hw/sysbus.h" + +#define TYPE_ASPEED_GPIO "aspeed.gpio" +#define ASPEED_GPIO(obj) OBJECT_CHECK(AspeedGPIOState, (obj), TYPE_ASPEED_GPIO) +#define ASPEED_GPIO_CLASS(klass) \ + OBJECT_CLASS_CHECK(AspeedGPIOClass, (klass), TYPE_ASPEED_GPIO) +#define ASPEED_GPIO_GET_CLASS(obj) \ + OBJECT_GET_CLASS(AspeedGPIOClass, (obj), TYPE_ASPEED_GPIO) + +#define ASPEED_GPIO_MAX_NR_SETS 8 +#define ASPEED_REGS_PER_BANK 14 +#define ASPEED_GPIO_MAX_NR_REGS (ASPEED_REGS_PER_BANK * ASPEED_GPIO_MAX_NR_SETS) +#define ASPEED_GPIO_NR_PINS 228 +#define ASPEED_GROUPS_PER_SET 4 +#define ASPEED_GPIO_NR_DEBOUNCE_REGS 3 +#define ASPEED_CHARS_PER_GROUP_LABEL 4 + +typedef struct GPIOSets GPIOSets; + +typedef struct GPIOSetProperties { + uint32_t input; + uint32_t output; + char group_label[ASPEED_GROUPS_PER_SET][ASPEED_CHARS_PER_GROUP_LABEL]; +} GPIOSetProperties; + +enum GPIORegType { + gpio_not_a_reg, + gpio_reg_data_value, + gpio_reg_direction, + gpio_reg_int_enable, + gpio_reg_int_sens_0, + gpio_reg_int_sens_1, + gpio_reg_int_sens_2, + gpio_reg_int_status, + gpio_reg_reset_tolerant, + gpio_reg_debounce_1, + gpio_reg_debounce_2, + gpio_reg_cmd_source_0, + gpio_reg_cmd_source_1, + gpio_reg_data_read, + gpio_reg_input_mask, +}; + +typedef struct AspeedGPIOReg { + uint16_t set_idx; + enum GPIORegType type; + } AspeedGPIOReg; + +typedef struct AspeedGPIOClass { + SysBusDevice parent_obj; + const GPIOSetProperties *props; + uint32_t nr_gpio_pins; + uint32_t nr_gpio_sets; + uint32_t gap; + const AspeedGPIOReg *reg_table; +} AspeedGPIOClass; + +typedef struct AspeedGPIOState { + /* <private> */ + SysBusDevice parent; + + /*< public >*/ + MemoryRegion iomem; + int pending; + qemu_irq irq; + qemu_irq gpios[ASPEED_GPIO_NR_PINS]; + +/* Parallel GPIO Registers */ + uint32_t debounce_regs[ASPEED_GPIO_NR_DEBOUNCE_REGS]; + struct GPIOSets { + uint32_t data_value; /* Reflects pin values */ + uint32_t data_read; /* Contains last value written to data value */ + uint32_t direction; + uint32_t int_enable; + uint32_t int_sens_0; + uint32_t int_sens_1; + uint32_t int_sens_2; + uint32_t int_status; + uint32_t reset_tol; + uint32_t cmd_source_0; + uint32_t cmd_source_1; + uint32_t debounce_1; + uint32_t debounce_2; + uint32_t input_mask; + } sets[ASPEED_GPIO_MAX_NR_SETS]; +} AspeedGPIOState; + +#endif /* _ASPEED_GPIO_H_ */ diff --git a/hw/gpio/aspeed_gpio.c b/hw/gpio/aspeed_gpio.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/gpio/aspeed_gpio.c @@ -XXX,XX +XXX,XX @@ +/* + * ASPEED GPIO Controller + * + * Copyright (C) 2017-2019 IBM Corp. + * + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#include <assert.h> + +#include "qemu/osdep.h" +#include "qemu/host-utils.h" +#include "qemu/log.h" +#include "hw/gpio/aspeed_gpio.h" +#include "include/hw/misc/aspeed_scu.h" +#include "qapi/error.h" +#include "qapi/visitor.h" +#include "hw/irq.h" +#include "migration/vmstate.h" + +#define GPIOS_PER_REG 32 +#define GPIOS_PER_SET GPIOS_PER_REG +#define GPIO_PIN_GAP_SIZE 4 +#define GPIOS_PER_GROUP 8 +#define GPIO_GROUP_SHIFT 3 + +/* GPIO Source Types */ +#define ASPEED_CMD_SRC_MASK 0x01010101 +#define ASPEED_SOURCE_ARM 0 +#define ASPEED_SOURCE_LPC 1 +#define ASPEED_SOURCE_COPROCESSOR 2 +#define ASPEED_SOURCE_RESERVED 3 + +/* GPIO Interrupt Triggers */ +/* + * For each set of gpios there are three sensitivity registers that control + * the interrupt trigger mode. + * + * | 2 | 1 | 0 | trigger mode + * ----------------------------- + * | 0 | 0 | 0 | falling-edge + * | 0 | 0 | 1 | rising-edge + * | 0 | 1 | 0 | level-low + * | 0 | 1 | 1 | level-high + * | 1 | X | X | dual-edge + */ +#define ASPEED_FALLING_EDGE 0 +#define ASPEED_RISING_EDGE 1 +#define ASPEED_LEVEL_LOW 2 +#define ASPEED_LEVEL_HIGH 3 +#define ASPEED_DUAL_EDGE 4 + +/* GPIO Register Address Offsets */ +#define GPIO_ABCD_DATA_VALUE (0x000 >> 2) +#define GPIO_ABCD_DIRECTION (0x004 >> 2) +#define GPIO_ABCD_INT_ENABLE (0x008 >> 2) +#define GPIO_ABCD_INT_SENS_0 (0x00C >> 2) +#define GPIO_ABCD_INT_SENS_1 (0x010 >> 2) +#define GPIO_ABCD_INT_SENS_2 (0x014 >> 2) +#define GPIO_ABCD_INT_STATUS (0x018 >> 2) +#define GPIO_ABCD_RESET_TOLERANT (0x01C >> 2) +#define GPIO_EFGH_DATA_VALUE (0x020 >> 2) +#define GPIO_EFGH_DIRECTION (0x024 >> 2) +#define GPIO_EFGH_INT_ENABLE (0x028 >> 2) +#define GPIO_EFGH_INT_SENS_0 (0x02C >> 2) +#define GPIO_EFGH_INT_SENS_1 (0x030 >> 2) +#define GPIO_EFGH_INT_SENS_2 (0x034 >> 2) +#define GPIO_EFGH_INT_STATUS (0x038 >> 2) +#define GPIO_EFGH_RESET_TOLERANT (0x03C >> 2) +#define GPIO_ABCD_DEBOUNCE_1 (0x040 >> 2) +#define GPIO_ABCD_DEBOUNCE_2 (0x044 >> 2) +#define GPIO_EFGH_DEBOUNCE_1 (0x048 >> 2) +#define GPIO_EFGH_DEBOUNCE_2 (0x04C >> 2) +#define GPIO_DEBOUNCE_TIME_1 (0x050 >> 2) +#define GPIO_DEBOUNCE_TIME_2 (0x054 >> 2) +#define GPIO_DEBOUNCE_TIME_3 (0x058 >> 2) +#define GPIO_ABCD_COMMAND_SRC_0 (0x060 >> 2) +#define GPIO_ABCD_COMMAND_SRC_1 (0x064 >> 2) +#define GPIO_EFGH_COMMAND_SRC_0 (0x068 >> 2) +#define GPIO_EFGH_COMMAND_SRC_1 (0x06C >> 2) +#define GPIO_IJKL_DATA_VALUE (0x070 >> 2) +#define GPIO_IJKL_DIRECTION (0x074 >> 2) +#define GPIO_MNOP_DATA_VALUE (0x078 >> 2) +#define GPIO_MNOP_DIRECTION (0x07C >> 2) +#define GPIO_QRST_DATA_VALUE (0x080 >> 2) +#define GPIO_QRST_DIRECTION (0x084 >> 2) +#define GPIO_UVWX_DATA_VALUE (0x088 >> 2) +#define GPIO_UVWX_DIRECTION (0x08C >> 2) +#define GPIO_IJKL_COMMAND_SRC_0 (0x090 >> 2) +#define GPIO_IJKL_COMMAND_SRC_1 (0x094 >> 2) +#define GPIO_IJKL_INT_ENABLE (0x098 >> 2) +#define GPIO_IJKL_INT_SENS_0 (0x09C >> 2) +#define GPIO_IJKL_INT_SENS_1 (0x0A0 >> 2) +#define GPIO_IJKL_INT_SENS_2 (0x0A4 >> 2) +#define GPIO_IJKL_INT_STATUS (0x0A8 >> 2) +#define GPIO_IJKL_RESET_TOLERANT (0x0AC >> 2) +#define GPIO_IJKL_DEBOUNCE_1 (0x0B0 >> 2) +#define GPIO_IJKL_DEBOUNCE_2 (0x0B4 >> 2) +#define GPIO_IJKL_INPUT_MASK (0x0B8 >> 2) +#define GPIO_ABCD_DATA_READ (0x0C0 >> 2) +#define GPIO_EFGH_DATA_READ (0x0C4 >> 2) +#define GPIO_IJKL_DATA_READ (0x0C8 >> 2) +#define GPIO_MNOP_DATA_READ (0x0CC >> 2) +#define GPIO_QRST_DATA_READ (0x0D0 >> 2) +#define GPIO_UVWX_DATA_READ (0x0D4 >> 2) +#define GPIO_YZAAAB_DATA_READ (0x0D8 >> 2) +#define GPIO_AC_DATA_READ (0x0DC >> 2) +#define GPIO_MNOP_COMMAND_SRC_0 (0x0E0 >> 2) +#define GPIO_MNOP_COMMAND_SRC_1 (0x0E4 >> 2) +#define GPIO_MNOP_INT_ENABLE (0x0E8 >> 2) +#define GPIO_MNOP_INT_SENS_0 (0x0EC >> 2) +#define GPIO_MNOP_INT_SENS_1 (0x0F0 >> 2) +#define GPIO_MNOP_INT_SENS_2 (0x0F4 >> 2) +#define GPIO_MNOP_INT_STATUS (0x0F8 >> 2) +#define GPIO_MNOP_RESET_TOLERANT (0x0FC >> 2) +#define GPIO_MNOP_DEBOUNCE_1 (0x100 >> 2) +#define GPIO_MNOP_DEBOUNCE_2 (0x104 >> 2) +#define GPIO_MNOP_INPUT_MASK (0x108 >> 2) +#define GPIO_QRST_COMMAND_SRC_0 (0x110 >> 2) +#define GPIO_QRST_COMMAND_SRC_1 (0x114 >> 2) +#define GPIO_QRST_INT_ENABLE (0x118 >> 2) +#define GPIO_QRST_INT_SENS_0 (0x11C >> 2) +#define GPIO_QRST_INT_SENS_1 (0x120 >> 2) +#define GPIO_QRST_INT_SENS_2 (0x124 >> 2) +#define GPIO_QRST_INT_STATUS (0x128 >> 2) +#define GPIO_QRST_RESET_TOLERANT (0x12C >> 2) +#define GPIO_QRST_DEBOUNCE_1 (0x130 >> 2) +#define GPIO_QRST_DEBOUNCE_2 (0x134 >> 2) +#define GPIO_QRST_INPUT_MASK (0x138 >> 2) +#define GPIO_UVWX_COMMAND_SRC_0 (0x140 >> 2) +#define GPIO_UVWX_COMMAND_SRC_1 (0x144 >> 2) +#define GPIO_UVWX_INT_ENABLE (0x148 >> 2) +#define GPIO_UVWX_INT_SENS_0 (0x14C >> 2) +#define GPIO_UVWX_INT_SENS_1 (0x150 >> 2) +#define GPIO_UVWX_INT_SENS_2 (0x154 >> 2) +#define GPIO_UVWX_INT_STATUS (0x158 >> 2) +#define GPIO_UVWX_RESET_TOLERANT (0x15C >> 2) +#define GPIO_UVWX_DEBOUNCE_1 (0x160 >> 2) +#define GPIO_UVWX_DEBOUNCE_2 (0x164 >> 2) +#define GPIO_UVWX_INPUT_MASK (0x168 >> 2) +#define GPIO_YZAAAB_COMMAND_SRC_0 (0x170 >> 2) +#define GPIO_YZAAAB_COMMAND_SRC_1 (0x174 >> 2) +#define GPIO_YZAAAB_INT_ENABLE (0x178 >> 2) +#define GPIO_YZAAAB_INT_SENS_0 (0x17C >> 2) +#define GPIO_YZAAAB_INT_SENS_1 (0x180 >> 2) +#define GPIO_YZAAAB_INT_SENS_2 (0x184 >> 2) +#define GPIO_YZAAAB_INT_STATUS (0x188 >> 2) +#define GPIO_YZAAAB_RESET_TOLERANT (0x18C >> 2) +#define GPIO_YZAAAB_DEBOUNCE_1 (0x190 >> 2) +#define GPIO_YZAAAB_DEBOUNCE_2 (0x194 >> 2) +#define GPIO_YZAAAB_INPUT_MASK (0x198 >> 2) +#define GPIO_AC_COMMAND_SRC_0 (0x1A0 >> 2) +#define GPIO_AC_COMMAND_SRC_1 (0x1A4 >> 2) +#define GPIO_AC_INT_ENABLE (0x1A8 >> 2) +#define GPIO_AC_INT_SENS_0 (0x1AC >> 2) +#define GPIO_AC_INT_SENS_1 (0x1B0 >> 2) +#define GPIO_AC_INT_SENS_2 (0x1B4 >> 2) +#define GPIO_AC_INT_STATUS (0x1B8 >> 2) +#define GPIO_AC_RESET_TOLERANT (0x1BC >> 2) +#define GPIO_AC_DEBOUNCE_1 (0x1C0 >> 2) +#define GPIO_AC_DEBOUNCE_2 (0x1C4 >> 2) +#define GPIO_AC_INPUT_MASK (0x1C8 >> 2) +#define GPIO_ABCD_INPUT_MASK (0x1D0 >> 2) +#define GPIO_EFGH_INPUT_MASK (0x1D4 >> 2) +#define GPIO_YZAAAB_DATA_VALUE (0x1E0 >> 2) +#define GPIO_YZAAAB_DIRECTION (0x1E4 >> 2) +#define GPIO_AC_DATA_VALUE (0x1E8 >> 2) +#define GPIO_AC_DIRECTION (0x1EC >> 2) +#define GPIO_3_6V_MEM_SIZE 0x1F0 +#define GPIO_3_6V_REG_ARRAY_SIZE (GPIO_3_6V_MEM_SIZE >> 2) + +static int aspeed_evaluate_irq(GPIOSets *regs, int gpio_prev_high, int gpio) +{ + uint32_t falling_edge = 0, rising_edge = 0; + uint32_t int_trigger = extract32(regs->int_sens_0, gpio, 1) + | extract32(regs->int_sens_1, gpio, 1) << 1 + | extract32(regs->int_sens_2, gpio, 1) << 2; + uint32_t gpio_curr_high = extract32(regs->data_value, gpio, 1); + uint32_t gpio_int_enabled = extract32(regs->int_enable, gpio, 1); + + if (!gpio_int_enabled) { + return 0; + } + + /* Detect edges */ + if (gpio_curr_high && !gpio_prev_high) { + rising_edge = 1; + } else if (!gpio_curr_high && gpio_prev_high) { + falling_edge = 1; + } + + if (((int_trigger == ASPEED_FALLING_EDGE) && falling_edge) || + ((int_trigger == ASPEED_RISING_EDGE) && rising_edge) || + ((int_trigger == ASPEED_LEVEL_LOW) && !gpio_curr_high) || + ((int_trigger == ASPEED_LEVEL_HIGH) && gpio_curr_high) || + ((int_trigger >= ASPEED_DUAL_EDGE) && (rising_edge || falling_edge))) + { + regs->int_status = deposit32(regs->int_status, gpio, 1, 1); + return 1; + } + return 0; +} + +#define nested_struct_index(ta, pa, m, tb, pb) \ + (pb - ((tb *)(((char *)pa) + offsetof(ta, m)))) + +static ptrdiff_t aspeed_gpio_set_idx(AspeedGPIOState *s, GPIOSets *regs) +{ + return nested_struct_index(AspeedGPIOState, s, sets, GPIOSets, regs); +} + +static void aspeed_gpio_update(AspeedGPIOState *s, GPIOSets *regs, + uint32_t value) +{ + uint32_t input_mask = regs->input_mask; + uint32_t direction = regs->direction; + uint32_t old = regs->data_value; + uint32_t new = value; + uint32_t diff; + int gpio; + + diff = old ^ new; + if (diff) { + for (gpio = 0; gpio < GPIOS_PER_REG; gpio++) { + uint32_t mask = 1 << gpio; + + /* If the gpio needs to be updated... */ + if (!(diff & mask)) { + continue; + } + + /* ...and we're output or not input-masked... */ + if (!(direction & mask) && (input_mask & mask)) { + continue; + } + + /* ...then update the state. */ + if (mask & new) { + regs->data_value |= mask; + } else { + regs->data_value &= ~mask; + } + + /* If the gpio is set to output... */ + if (direction & mask) { + /* ...trigger the line-state IRQ */ + ptrdiff_t set = aspeed_gpio_set_idx(s, regs); + size_t offset = set * GPIOS_PER_SET + gpio; + qemu_set_irq(s->gpios[offset], !!(new & mask)); + } else { + /* ...otherwise if we meet the line's current IRQ policy... */ + if (aspeed_evaluate_irq(regs, old & mask, gpio)) { + /* ...trigger the VIC IRQ */ + s->pending++; + } + } + } + } + qemu_set_irq(s->irq, !!(s->pending)); +} + +static uint32_t aspeed_adjust_pin(AspeedGPIOState *s, uint32_t pin) +{ + AspeedGPIOClass *agc = ASPEED_GPIO_GET_CLASS(s); + /* + * The 2500 has a 4 pin gap in group AB and the 2400 has a 4 pin + * gap in group Y (and only four pins in AB but this is the last group so + * it doesn't matter). + */ + if (agc->gap && pin >= agc->gap) { + pin += GPIO_PIN_GAP_SIZE; + } + + return pin; +} + +static bool aspeed_gpio_get_pin_level(AspeedGPIOState *s, uint32_t set_idx, + uint32_t pin) +{ + uint32_t reg_val; + uint32_t pin_mask = 1 << pin; + + reg_val = s->sets[set_idx].data_value; + + return !!(reg_val & pin_mask); +} + +static void aspeed_gpio_set_pin_level(AspeedGPIOState *s, uint32_t set_idx, + uint32_t pin, bool level) +{ + uint32_t value = s->sets[set_idx].data_value; + uint32_t pin_mask = 1 << pin; + + if (level) { + value |= pin_mask; + } else { + value &= !pin_mask; + } + + aspeed_gpio_update(s, &s->sets[set_idx], value); +} + +/* + * | src_1 | src_2 | source | + * |-----------------------------| + * | 0 | 0 | ARM | + * | 0 | 1 | LPC | + * | 1 | 0 | Coprocessor| + * | 1 | 1 | Reserved | + * + * Once the source of a set is programmed, corresponding bits in the + * data_value, direction, interrupt [enable, sens[0-2]], reset_tol and + * debounce registers can only be written by the source. + * + * Source is ARM by default + * only bits 24, 16, 8, and 0 can be set + * + * we don't currently have a model for the LPC or Coprocessor + */ +static uint32_t update_value_control_source(GPIOSets *regs, uint32_t old_value, + uint32_t value) +{ + int i; + int cmd_source; + + /* assume the source is always ARM for now */ + int source = ASPEED_SOURCE_ARM; + + uint32_t new_value = 0; + + /* for each group in set */ + for (i = 0; i < GPIOS_PER_REG; i += GPIOS_PER_GROUP) { + cmd_source = extract32(regs->cmd_source_0, i, 1) + | (extract32(regs->cmd_source_1, i, 1) << 1); + + if (source == cmd_source) { + new_value |= (0xff << i) & value; + } else { + new_value |= (0xff << i) & old_value; + } + } + return new_value; +} + +static const AspeedGPIOReg aspeed_3_6v_gpios[GPIO_3_6V_REG_ARRAY_SIZE] = { + /* Set ABCD */ + [GPIO_ABCD_DATA_VALUE] = { 0, gpio_reg_data_value }, + [GPIO_ABCD_DIRECTION] = { 0, gpio_reg_direction }, + [GPIO_ABCD_INT_ENABLE] = { 0, gpio_reg_int_enable }, + [GPIO_ABCD_INT_SENS_0] = { 0, gpio_reg_int_sens_0 }, + [GPIO_ABCD_INT_SENS_1] = { 0, gpio_reg_int_sens_1 }, + [GPIO_ABCD_INT_SENS_2] = { 0, gpio_reg_int_sens_2 }, + [GPIO_ABCD_INT_STATUS] = { 0, gpio_reg_int_status }, + [GPIO_ABCD_RESET_TOLERANT] = { 0, gpio_reg_reset_tolerant }, + [GPIO_ABCD_DEBOUNCE_1] = { 0, gpio_reg_debounce_1 }, + [GPIO_ABCD_DEBOUNCE_2] = { 0, gpio_reg_debounce_2 }, + [GPIO_ABCD_COMMAND_SRC_0] = { 0, gpio_reg_cmd_source_0 }, + [GPIO_ABCD_COMMAND_SRC_1] = { 0, gpio_reg_cmd_source_1 }, + [GPIO_ABCD_DATA_READ] = { 0, gpio_reg_data_read }, + [GPIO_ABCD_INPUT_MASK] = { 0, gpio_reg_input_mask }, + /* Set EFGH */ + [GPIO_EFGH_DATA_VALUE] = { 1, gpio_reg_data_value }, + [GPIO_EFGH_DIRECTION] = { 1, gpio_reg_direction }, + [GPIO_EFGH_INT_ENABLE] = { 1, gpio_reg_int_enable }, + [GPIO_EFGH_INT_SENS_0] = { 1, gpio_reg_int_sens_0 }, + [GPIO_EFGH_INT_SENS_1] = { 1, gpio_reg_int_sens_1 }, + [GPIO_EFGH_INT_SENS_2] = { 1, gpio_reg_int_sens_2 }, + [GPIO_EFGH_INT_STATUS] = { 1, gpio_reg_int_status }, + [GPIO_EFGH_RESET_TOLERANT] = { 1, gpio_reg_reset_tolerant }, + [GPIO_EFGH_DEBOUNCE_1] = { 1, gpio_reg_debounce_1 }, + [GPIO_EFGH_DEBOUNCE_2] = { 1, gpio_reg_debounce_2 }, + [GPIO_EFGH_COMMAND_SRC_0] = { 1, gpio_reg_cmd_source_0 }, + [GPIO_EFGH_COMMAND_SRC_1] = { 1, gpio_reg_cmd_source_1 }, + [GPIO_EFGH_DATA_READ] = { 1, gpio_reg_data_read }, + [GPIO_EFGH_INPUT_MASK] = { 1, gpio_reg_input_mask }, + /* Set IJKL */ + [GPIO_IJKL_DATA_VALUE] = { 2, gpio_reg_data_value }, + [GPIO_IJKL_DIRECTION] = { 2, gpio_reg_direction }, + [GPIO_IJKL_INT_ENABLE] = { 2, gpio_reg_int_enable }, + [GPIO_IJKL_INT_SENS_0] = { 2, gpio_reg_int_sens_0 }, + [GPIO_IJKL_INT_SENS_1] = { 2, gpio_reg_int_sens_1 }, + [GPIO_IJKL_INT_SENS_2] = { 2, gpio_reg_int_sens_2 }, + [GPIO_IJKL_INT_STATUS] = { 2, gpio_reg_int_status }, + [GPIO_IJKL_RESET_TOLERANT] = { 2, gpio_reg_reset_tolerant }, + [GPIO_IJKL_DEBOUNCE_1] = { 2, gpio_reg_debounce_1 }, + [GPIO_IJKL_DEBOUNCE_2] = { 2, gpio_reg_debounce_2 }, + [GPIO_IJKL_COMMAND_SRC_0] = { 2, gpio_reg_cmd_source_0 }, + [GPIO_IJKL_COMMAND_SRC_1] = { 2, gpio_reg_cmd_source_1 }, + [GPIO_IJKL_DATA_READ] = { 2, gpio_reg_data_read }, + [GPIO_IJKL_INPUT_MASK] = { 2, gpio_reg_input_mask }, + /* Set MNOP */ + [GPIO_MNOP_DATA_VALUE] = { 3, gpio_reg_data_value }, + [GPIO_MNOP_DIRECTION] = { 3, gpio_reg_direction }, + [GPIO_MNOP_INT_ENABLE] = { 3, gpio_reg_int_enable }, + [GPIO_MNOP_INT_SENS_0] = { 3, gpio_reg_int_sens_0 }, + [GPIO_MNOP_INT_SENS_1] = { 3, gpio_reg_int_sens_1 }, + [GPIO_MNOP_INT_SENS_2] = { 3, gpio_reg_int_sens_2 }, + [GPIO_MNOP_INT_STATUS] = { 3, gpio_reg_int_status }, + [GPIO_MNOP_RESET_TOLERANT] = { 3, gpio_reg_reset_tolerant }, + [GPIO_MNOP_DEBOUNCE_1] = { 3, gpio_reg_debounce_1 }, + [GPIO_MNOP_DEBOUNCE_2] = { 3, gpio_reg_debounce_2 }, + [GPIO_MNOP_COMMAND_SRC_0] = { 3, gpio_reg_cmd_source_0 }, + [GPIO_MNOP_COMMAND_SRC_1] = { 3, gpio_reg_cmd_source_1 }, + [GPIO_MNOP_DATA_READ] = { 3, gpio_reg_data_read }, + [GPIO_MNOP_INPUT_MASK] = { 3, gpio_reg_input_mask }, + /* Set QRST */ + [GPIO_QRST_DATA_VALUE] = { 4, gpio_reg_data_value }, + [GPIO_QRST_DIRECTION] = { 4, gpio_reg_direction }, + [GPIO_QRST_INT_ENABLE] = { 4, gpio_reg_int_enable }, + [GPIO_QRST_INT_SENS_0] = { 4, gpio_reg_int_sens_0 }, + [GPIO_QRST_INT_SENS_1] = { 4, gpio_reg_int_sens_1 }, + [GPIO_QRST_INT_SENS_2] = { 4, gpio_reg_int_sens_2 }, + [GPIO_QRST_INT_STATUS] = { 4, gpio_reg_int_status }, + [GPIO_QRST_RESET_TOLERANT] = { 4, gpio_reg_reset_tolerant }, + [GPIO_QRST_DEBOUNCE_1] = { 4, gpio_reg_debounce_1 }, + [GPIO_QRST_DEBOUNCE_2] = { 4, gpio_reg_debounce_2 }, + [GPIO_QRST_COMMAND_SRC_0] = { 4, gpio_reg_cmd_source_0 }, + [GPIO_QRST_COMMAND_SRC_1] = { 4, gpio_reg_cmd_source_1 }, + [GPIO_QRST_DATA_READ] = { 4, gpio_reg_data_read }, + [GPIO_QRST_INPUT_MASK] = { 4, gpio_reg_input_mask }, + /* Set UVWX */ + [GPIO_UVWX_DATA_VALUE] = { 5, gpio_reg_data_value }, + [GPIO_UVWX_DIRECTION] = { 5, gpio_reg_direction }, + [GPIO_UVWX_INT_ENABLE] = { 5, gpio_reg_int_enable }, + [GPIO_UVWX_INT_SENS_0] = { 5, gpio_reg_int_sens_0 }, + [GPIO_UVWX_INT_SENS_1] = { 5, gpio_reg_int_sens_1 }, + [GPIO_UVWX_INT_SENS_2] = { 5, gpio_reg_int_sens_2 }, + [GPIO_UVWX_INT_STATUS] = { 5, gpio_reg_int_status }, + [GPIO_UVWX_RESET_TOLERANT] = { 5, gpio_reg_reset_tolerant }, + [GPIO_UVWX_DEBOUNCE_1] = { 5, gpio_reg_debounce_1 }, + [GPIO_UVWX_DEBOUNCE_2] = { 5, gpio_reg_debounce_2 }, + [GPIO_UVWX_COMMAND_SRC_0] = { 5, gpio_reg_cmd_source_0 }, + [GPIO_UVWX_COMMAND_SRC_1] = { 5, gpio_reg_cmd_source_1 }, + [GPIO_UVWX_DATA_READ] = { 5, gpio_reg_data_read }, + [GPIO_UVWX_INPUT_MASK] = { 5, gpio_reg_input_mask }, + /* Set YZAAAB */ + [GPIO_YZAAAB_DATA_VALUE] = { 6, gpio_reg_data_value }, + [GPIO_YZAAAB_DIRECTION] = { 6, gpio_reg_direction }, + [GPIO_YZAAAB_INT_ENABLE] = { 6, gpio_reg_int_enable }, + [GPIO_YZAAAB_INT_SENS_0] = { 6, gpio_reg_int_sens_0 }, + [GPIO_YZAAAB_INT_SENS_1] = { 6, gpio_reg_int_sens_1 }, + [GPIO_YZAAAB_INT_SENS_2] = { 6, gpio_reg_int_sens_2 }, + [GPIO_YZAAAB_INT_STATUS] = { 6, gpio_reg_int_status }, + [GPIO_YZAAAB_RESET_TOLERANT] = { 6, gpio_reg_reset_tolerant }, + [GPIO_YZAAAB_DEBOUNCE_1] = { 6, gpio_reg_debounce_1 }, + [GPIO_YZAAAB_DEBOUNCE_2] = { 6, gpio_reg_debounce_2 }, + [GPIO_YZAAAB_COMMAND_SRC_0] = { 6, gpio_reg_cmd_source_0 }, + [GPIO_YZAAAB_COMMAND_SRC_1] = { 6, gpio_reg_cmd_source_1 }, + [GPIO_YZAAAB_DATA_READ] = { 6, gpio_reg_data_read }, + [GPIO_YZAAAB_INPUT_MASK] = { 6, gpio_reg_input_mask }, + /* Set AC (ast2500 only) */ + [GPIO_AC_DATA_VALUE] = { 7, gpio_reg_data_value }, + [GPIO_AC_DIRECTION] = { 7, gpio_reg_direction }, + [GPIO_AC_INT_ENABLE] = { 7, gpio_reg_int_enable }, + [GPIO_AC_INT_SENS_0] = { 7, gpio_reg_int_sens_0 }, + [GPIO_AC_INT_SENS_1] = { 7, gpio_reg_int_sens_1 }, + [GPIO_AC_INT_SENS_2] = { 7, gpio_reg_int_sens_2 }, + [GPIO_AC_INT_STATUS] = { 7, gpio_reg_int_status }, + [GPIO_AC_RESET_TOLERANT] = { 7, gpio_reg_reset_tolerant }, + [GPIO_AC_DEBOUNCE_1] = { 7, gpio_reg_debounce_1 }, + [GPIO_AC_DEBOUNCE_2] = { 7, gpio_reg_debounce_2 }, + [GPIO_AC_COMMAND_SRC_0] = { 7, gpio_reg_cmd_source_0 }, + [GPIO_AC_COMMAND_SRC_1] = { 7, gpio_reg_cmd_source_1 }, + [GPIO_AC_DATA_READ] = { 7, gpio_reg_data_read }, + [GPIO_AC_INPUT_MASK] = { 7, gpio_reg_input_mask }, +}; + +static uint64_t aspeed_gpio_read(void *opaque, hwaddr offset, uint32_t size) +{ + AspeedGPIOState *s = ASPEED_GPIO(opaque); + AspeedGPIOClass *agc = ASPEED_GPIO_GET_CLASS(s); + uint64_t idx = -1; + const AspeedGPIOReg *reg; + GPIOSets *set; + + idx = offset >> 2; + if (idx >= GPIO_DEBOUNCE_TIME_1 && idx <= GPIO_DEBOUNCE_TIME_3) { + idx -= GPIO_DEBOUNCE_TIME_1; + return (uint64_t) s->debounce_regs[idx]; + } + + reg = &agc->reg_table[idx]; + if (reg->set_idx >= agc->nr_gpio_sets) { + qemu_log_mask(LOG_GUEST_ERROR, "%s: no getter for offset 0x%" + HWADDR_PRIx"\n", __func__, offset); + return 0; + } + + set = &s->sets[reg->set_idx]; + switch (reg->type) { + case gpio_reg_data_value: + return set->data_value; + case gpio_reg_direction: + return set->direction; + case gpio_reg_int_enable: + return set->int_enable; + case gpio_reg_int_sens_0: + return set->int_sens_0; + case gpio_reg_int_sens_1: + return set->int_sens_1; + case gpio_reg_int_sens_2: + return set->int_sens_2; + case gpio_reg_int_status: + return set->int_status; + case gpio_reg_reset_tolerant: + return set->reset_tol; + case gpio_reg_debounce_1: + return set->debounce_1; + case gpio_reg_debounce_2: + return set->debounce_2; + case gpio_reg_cmd_source_0: + return set->cmd_source_0; + case gpio_reg_cmd_source_1: + return set->cmd_source_1; + case gpio_reg_data_read: + return set->data_read; + case gpio_reg_input_mask: + return set->input_mask; + default: + qemu_log_mask(LOG_GUEST_ERROR, "%s: no getter for offset 0x%" + HWADDR_PRIx"\n", __func__, offset); + return 0; + }; +} + +static void aspeed_gpio_write(void *opaque, hwaddr offset, uint64_t data, + uint32_t size) +{ + AspeedGPIOState *s = ASPEED_GPIO(opaque); + AspeedGPIOClass *agc = ASPEED_GPIO_GET_CLASS(s); + const GPIOSetProperties *props; + uint64_t idx = -1; + const AspeedGPIOReg *reg; + GPIOSets *set; + uint32_t cleared; + + idx = offset >> 2; + if (idx >= GPIO_DEBOUNCE_TIME_1 && idx <= GPIO_DEBOUNCE_TIME_3) { + idx -= GPIO_DEBOUNCE_TIME_1; + s->debounce_regs[idx] = (uint32_t) data; + return; + } + + reg = &agc->reg_table[idx]; + if (reg->set_idx >= agc->nr_gpio_sets) { + qemu_log_mask(LOG_GUEST_ERROR, "%s: no setter for offset 0x%" + HWADDR_PRIx"\n", __func__, offset); + return; + } + + set = &s->sets[reg->set_idx]; + props = &agc->props[reg->set_idx]; + + switch (reg->type) { + case gpio_reg_data_value: + data &= props->output; + data = update_value_control_source(set, set->data_value, data); + set->data_read = data; + aspeed_gpio_update(s, set, data); + return; + case gpio_reg_direction: + /* + * where data is the value attempted to be written to the pin: + * pin type | input mask | output mask | expected value + * ------------------------------------------------------------ + * bidirectional | 1 | 1 | data + * input only | 1 | 0 | 0 + * output only | 0 | 1 | 1 + * no pin / gap | 0 | 0 | 0 + * + * which is captured by: + * data = ( data | ~input) & output; + */ + data = (data | ~props->input) & props->output; + set->direction = update_value_control_source(set, set->direction, data); + break; + case gpio_reg_int_enable: + set->int_enable = update_value_control_source(set, set->int_enable, + data); + break; + case gpio_reg_int_sens_0: + set->int_sens_0 = update_value_control_source(set, set->int_sens_0, + data); + break; + case gpio_reg_int_sens_1: + set->int_sens_1 = update_value_control_source(set, set->int_sens_1, + data); + break; + case gpio_reg_int_sens_2: + set->int_sens_2 = update_value_control_source(set, set->int_sens_2, + data); + break; + case gpio_reg_int_status: + cleared = ctpop32(data & set->int_status); + if (s->pending && cleared) { + assert(s->pending >= cleared); + s->pending -= cleared; + } + set->int_status &= ~data; + break; + case gpio_reg_reset_tolerant: + set->reset_tol = update_value_control_source(set, set->reset_tol, + data); + return; + case gpio_reg_debounce_1: + set->debounce_1 = update_value_control_source(set, set->debounce_1, + data); + return; + case gpio_reg_debounce_2: + set->debounce_2 = update_value_control_source(set, set->debounce_2, + data); + return; + case gpio_reg_cmd_source_0: + set->cmd_source_0 = data & ASPEED_CMD_SRC_MASK; + return; + case gpio_reg_cmd_source_1: + set->cmd_source_1 = data & ASPEED_CMD_SRC_MASK; + return; + case gpio_reg_data_read: + /* Read only register */ + return; + case gpio_reg_input_mask: + /* + * feeds into interrupt generation + * 0: read from data value reg will be updated + * 1: read from data value reg will not be updated + */ + set->input_mask = data & props->input; + break; + default: + qemu_log_mask(LOG_GUEST_ERROR, "%s: no setter for offset 0x%" + HWADDR_PRIx"\n", __func__, offset); + return; + } + aspeed_gpio_update(s, set, set->data_value); + return; +} + +static int get_set_idx(AspeedGPIOState *s, const char *group, int *group_idx) +{ + AspeedGPIOClass *agc = ASPEED_GPIO_GET_CLASS(s); + int set_idx, g_idx = *group_idx; + + for (set_idx = 0; set_idx < agc->nr_gpio_sets; set_idx++) { + const GPIOSetProperties *set_props = &agc->props[set_idx]; + for (g_idx = 0; g_idx < ASPEED_GROUPS_PER_SET; g_idx++) { + if (!strncmp(group, set_props->group_label[g_idx], strlen(group))) { + *group_idx = g_idx; + return set_idx; + } + } + } + return -1; +} + +static void aspeed_gpio_get_pin(Object *obj, Visitor *v, const char *name, + void *opaque, Error **errp) +{ + int pin = 0xfff; + bool level = true; + char group[3]; + AspeedGPIOState *s = ASPEED_GPIO(obj); + int set_idx, group_idx = 0; + + if (sscanf(name, "gpio%2[A-Z]%1d", group, &pin) != 2) { + error_setg(errp, "%s: error reading %s", __func__, name); + return; + } + set_idx = get_set_idx(s, group, &group_idx); + if (set_idx == -1) { + error_setg(errp, "%s: invalid group %s", __func__, group); + return; + } + pin = pin + group_idx * GPIOS_PER_GROUP; + level = aspeed_gpio_get_pin_level(s, set_idx, pin); + visit_type_bool(v, name, &level, errp); +} + +static void aspeed_gpio_set_pin(Object *obj, Visitor *v, const char *name, + void *opaque, Error **errp) +{ + Error *local_err = NULL; + bool level; + int pin = 0xfff; + char group[3]; + AspeedGPIOState *s = ASPEED_GPIO(obj); + int set_idx, group_idx = 0; + + visit_type_bool(v, name, &level, &local_err); + if (local_err) { + error_propagate(errp, local_err); + return; + } + if (sscanf(name, "gpio%2[A-Z]%1d", group, &pin) != 2) { + error_setg(errp, "%s: error reading %s", __func__, name); + return; + } + set_idx = get_set_idx(s, group, &group_idx); + if (set_idx == -1) { + error_setg(errp, "%s: invalid group %s", __func__, group); + return; + } + pin = pin + group_idx * GPIOS_PER_GROUP; + aspeed_gpio_set_pin_level(s, set_idx, pin, level); +} + +/****************** Setup functions ******************/ +static const GPIOSetProperties ast2400_set_props[] = { + [0] = {0xffffffff, 0xffffffff, {"A", "B", "C", "D"} }, + [1] = {0xffffffff, 0xffffffff, {"E", "F", "G", "H"} }, + [2] = {0xffffffff, 0xffffffff, {"I", "J", "K", "L"} }, + [3] = {0xffffffff, 0xffffffff, {"M", "N", "O", "P"} }, + [4] = {0xffffffff, 0xffffffff, {"Q", "R", "S", "T"} }, + [5] = {0xffffffff, 0x0000ffff, {"U", "V", "W", "X"} }, + [6] = {0x0000000f, 0x0fffff0f, {"Y", "Z", "AA", "AB"} }, +}; + +static const GPIOSetProperties ast2500_set_props[] = { + [0] = {0xffffffff, 0xffffffff, {"A", "B", "C", "D"} }, + [1] = {0xffffffff, 0xffffffff, {"E", "F", "G", "H"} }, + [2] = {0xffffffff, 0xffffffff, {"I", "J", "K", "L"} }, + [3] = {0xffffffff, 0xffffffff, {"M", "N", "O", "P"} }, + [4] = {0xffffffff, 0xffffffff, {"Q", "R", "S", "T"} }, + [5] = {0xffffffff, 0x0000ffff, {"U", "V", "W", "X"} }, + [6] = {0xffffff0f, 0x0fffff0f, {"Y", "Z", "AA", "AB"} }, + [7] = {0x000000ff, 0x000000ff, {"AC"} }, +}; + +static const MemoryRegionOps aspeed_gpio_ops = { + .read = aspeed_gpio_read, + .write = aspeed_gpio_write, + .endianness = DEVICE_LITTLE_ENDIAN, + .valid.min_access_size = 4, + .valid.max_access_size = 4, +}; + +static void aspeed_gpio_reset(DeviceState *dev) +{ + AspeedGPIOState *s = ASPEED_GPIO(dev); + + /* TODO: respect the reset tolerance registers */ + memset(s->sets, 0, sizeof(s->sets)); +} + +static void aspeed_gpio_realize(DeviceState *dev, Error **errp) +{ + AspeedGPIOState *s = ASPEED_GPIO(dev); + SysBusDevice *sbd = SYS_BUS_DEVICE(dev); + AspeedGPIOClass *agc = ASPEED_GPIO_GET_CLASS(s); + int pin; + + /* Interrupt parent line */ + sysbus_init_irq(sbd, &s->irq); + + /* Individual GPIOs */ + for (pin = 0; pin < agc->nr_gpio_pins; pin++) { + sysbus_init_irq(sbd, &s->gpios[pin]); + } + + memory_region_init_io(&s->iomem, OBJECT(s), &aspeed_gpio_ops, s, + TYPE_ASPEED_GPIO, GPIO_3_6V_MEM_SIZE); + + sysbus_init_mmio(sbd, &s->iomem); +} + +static void aspeed_gpio_init(Object *obj) +{ + AspeedGPIOState *s = ASPEED_GPIO(obj); + AspeedGPIOClass *agc = ASPEED_GPIO_GET_CLASS(s); + int pin; + + for (pin = 0; pin < agc->nr_gpio_pins; pin++) { + char *name; + int set_idx = pin / GPIOS_PER_SET; + int pin_idx = aspeed_adjust_pin(s, pin) - (set_idx * GPIOS_PER_SET); + int group_idx = pin_idx >> GPIO_GROUP_SHIFT; + const GPIOSetProperties *props = &agc->props[set_idx]; + + name = g_strdup_printf("gpio%s%d", props->group_label[group_idx], + pin_idx % GPIOS_PER_GROUP); + object_property_add(obj, name, "bool", aspeed_gpio_get_pin, + aspeed_gpio_set_pin, NULL, NULL, NULL); + } +} + +static const VMStateDescription vmstate_gpio_regs = { + .name = TYPE_ASPEED_GPIO"/regs", + .version_id = 1, + .minimum_version_id = 1, + .fields = (VMStateField[]) { + VMSTATE_UINT32(data_value, GPIOSets), + VMSTATE_UINT32(data_read, GPIOSets), + VMSTATE_UINT32(direction, GPIOSets), + VMSTATE_UINT32(int_enable, GPIOSets), + VMSTATE_UINT32(int_sens_0, GPIOSets), + VMSTATE_UINT32(int_sens_1, GPIOSets), + VMSTATE_UINT32(int_sens_2, GPIOSets), + VMSTATE_UINT32(int_status, GPIOSets), + VMSTATE_UINT32(reset_tol, GPIOSets), + VMSTATE_UINT32(cmd_source_0, GPIOSets), + VMSTATE_UINT32(cmd_source_1, GPIOSets), + VMSTATE_UINT32(debounce_1, GPIOSets), + VMSTATE_UINT32(debounce_2, GPIOSets), + VMSTATE_UINT32(input_mask, GPIOSets), + VMSTATE_END_OF_LIST(), + } +}; + +static const VMStateDescription vmstate_aspeed_gpio = { + .name = TYPE_ASPEED_GPIO, + .version_id = 1, + .minimum_version_id = 1, + .fields = (VMStateField[]) { + VMSTATE_STRUCT_ARRAY(sets, AspeedGPIOState, ASPEED_GPIO_MAX_NR_SETS, + 1, vmstate_gpio_regs, GPIOSets), + VMSTATE_UINT32_ARRAY(debounce_regs, AspeedGPIOState, + ASPEED_GPIO_NR_DEBOUNCE_REGS), + VMSTATE_END_OF_LIST(), + } +}; + +static void aspeed_gpio_class_init(ObjectClass *klass, void *data) +{ + DeviceClass *dc = DEVICE_CLASS(klass); + + dc->realize = aspeed_gpio_realize; + dc->reset = aspeed_gpio_reset; + dc->desc = "Aspeed GPIO Controller"; + dc->vmsd = &vmstate_aspeed_gpio; +} + +static void aspeed_gpio_ast2400_class_init(ObjectClass *klass, void *data) +{ + AspeedGPIOClass *agc = ASPEED_GPIO_CLASS(klass); + + agc->props = ast2400_set_props; + agc->nr_gpio_pins = 216; + agc->nr_gpio_sets = 7; + agc->gap = 196; + agc->reg_table = aspeed_3_6v_gpios; +} + +static void aspeed_gpio_2500_class_init(ObjectClass *klass, void *data) +{ + AspeedGPIOClass *agc = ASPEED_GPIO_CLASS(klass); + + agc->props = ast2500_set_props; + agc->nr_gpio_pins = 228; + agc->nr_gpio_sets = 8; + agc->gap = 220; + agc->reg_table = aspeed_3_6v_gpios; +} + +static const TypeInfo aspeed_gpio_info = { + .name = TYPE_ASPEED_GPIO, + .parent = TYPE_SYS_BUS_DEVICE, + .instance_size = sizeof(AspeedGPIOState), + .class_size = sizeof(AspeedGPIOClass), + .class_init = aspeed_gpio_class_init, + .abstract = true, +}; + +static const TypeInfo aspeed_gpio_ast2400_info = { + .name = TYPE_ASPEED_GPIO "-ast2400", + .parent = TYPE_ASPEED_GPIO, + .class_init = aspeed_gpio_ast2400_class_init, + .instance_init = aspeed_gpio_init, +}; + +static const TypeInfo aspeed_gpio_ast2500_info = { + .name = TYPE_ASPEED_GPIO "-ast2500", + .parent = TYPE_ASPEED_GPIO, + .class_init = aspeed_gpio_2500_class_init, + .instance_init = aspeed_gpio_init, +}; + +static void aspeed_gpio_register_types(void) +{ + type_register_static(&aspeed_gpio_info); + type_register_static(&aspeed_gpio_ast2400_info); + type_register_static(&aspeed_gpio_ast2500_info); +} + +type_init(aspeed_gpio_register_types); -- 2.20.1 From: Rashmica Gupta <rashmica.g@gmail.com> Signed-off-by: Rashmica Gupta <rashmica.g@gmail.com> Reviewed-by: Cédric Le Goater <clg@kaod.org> Signed-off-by: Cédric Le Goater <clg@kaod.org> Message-id: 20190904070506.1052-3-clg@kaod.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/aspeed_soc.h | 3 +++ hw/arm/aspeed_soc.c | 17 +++++++++++++++++ 2 files changed, 20 insertions(+) diff --git a/include/hw/arm/aspeed_soc.h b/include/hw/arm/aspeed_soc.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/aspeed_soc.h +++ b/include/hw/arm/aspeed_soc.h @@ -XXX,XX +XXX,XX @@ #include "hw/watchdog/wdt_aspeed.h" #include "hw/net/ftgmac100.h" #include "target/arm/cpu.h" +#include "hw/gpio/aspeed_gpio.h" #define ASPEED_SPIS_NUM 2 #define ASPEED_WDTS_NUM 3 @@ -XXX,XX +XXX,XX @@ typedef struct AspeedSoCState { AspeedSDMCState sdmc; AspeedWDTState wdt[ASPEED_WDTS_NUM]; FTGMAC100State ftgmac100[ASPEED_MACS_NUM]; + AspeedGPIOState gpio; } AspeedSoCState; #define TYPE_ASPEED_SOC "aspeed-soc" @@ -XXX,XX +XXX,XX @@ typedef struct AspeedSoCInfo { int spis_num; const char *fmc_typename; const char **spi_typename; + const char *gpio_typename; int wdts_num; const int *irqmap; const hwaddr *memmap; diff --git a/hw/arm/aspeed_soc.c b/hw/arm/aspeed_soc.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/aspeed_soc.c +++ b/hw/arm/aspeed_soc.c @@ -XXX,XX +XXX,XX @@ static const AspeedSoCInfo aspeed_socs[] = { .spis_num = 1, .fmc_typename = "aspeed.smc.fmc", .spi_typename = aspeed_soc_ast2400_typenames, + .gpio_typename = "aspeed.gpio-ast2400", .wdts_num = 2, .irqmap = aspeed_soc_ast2400_irqmap, .memmap = aspeed_soc_ast2400_memmap, @@ -XXX,XX +XXX,XX @@ static const AspeedSoCInfo aspeed_socs[] = { .spis_num = 1, .fmc_typename = "aspeed.smc.fmc", .spi_typename = aspeed_soc_ast2400_typenames, + .gpio_typename = "aspeed.gpio-ast2400", .wdts_num = 2, .irqmap = aspeed_soc_ast2400_irqmap, .memmap = aspeed_soc_ast2400_memmap, @@ -XXX,XX +XXX,XX @@ static const AspeedSoCInfo aspeed_socs[] = { .spis_num = 1, .fmc_typename = "aspeed.smc.fmc", .spi_typename = aspeed_soc_ast2400_typenames, + .gpio_typename = "aspeed.gpio-ast2400", .wdts_num = 2, .irqmap = aspeed_soc_ast2400_irqmap, .memmap = aspeed_soc_ast2400_memmap, @@ -XXX,XX +XXX,XX @@ static const AspeedSoCInfo aspeed_socs[] = { .spis_num = 2, .fmc_typename = "aspeed.smc.ast2500-fmc", .spi_typename = aspeed_soc_ast2500_typenames, + .gpio_typename = "aspeed.gpio-ast2500", .wdts_num = 3, .irqmap = aspeed_soc_ast2500_irqmap, .memmap = aspeed_soc_ast2500_memmap, @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_init(Object *obj) sysbus_init_child_obj(obj, "xdma", OBJECT(&s->xdma), sizeof(s->xdma), TYPE_ASPEED_XDMA); + + sysbus_init_child_obj(obj, "gpio", OBJECT(&s->gpio), sizeof(s->gpio), + sc->info->gpio_typename); } static void aspeed_soc_realize(DeviceState *dev, Error **errp) @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_realize(DeviceState *dev, Error **errp) sc->info->memmap[ASPEED_XDMA]); sysbus_connect_irq(SYS_BUS_DEVICE(&s->xdma), 0, aspeed_soc_get_irq(s, ASPEED_XDMA)); + + /* GPIO */ + object_property_set_bool(OBJECT(&s->gpio), true, "realized", &err); + if (err) { + error_propagate(errp, err); + return; + } + sysbus_mmio_map(SYS_BUS_DEVICE(&s->gpio), 0, sc->info->memmap[ASPEED_GPIO]); + sysbus_connect_irq(SYS_BUS_DEVICE(&s->gpio), 0, + aspeed_soc_get_irq(s, ASPEED_GPIO)); } static Property aspeed_soc_properties[] = { DEFINE_PROP_UINT32("num-cpus", AspeedSoCState, num_cpus, 0), -- 2.20.1 From: Cédric Le Goater <clg@kaod.org> There are no QEMU Aspeed machines using the SoCs "ast2400-a0" or "ast2400". Signed-off-by: Cédric Le Goater <clg@kaod.org> Message-id: 20190904070506.1052-4-clg@kaod.org Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/aspeed_soc.c | 26 -------------------------- 1 file changed, 26 deletions(-) diff --git a/hw/arm/aspeed_soc.c b/hw/arm/aspeed_soc.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/aspeed_soc.c +++ b/hw/arm/aspeed_soc.c @@ -XXX,XX +XXX,XX @@ static const char *aspeed_soc_ast2500_typenames[] = { static const AspeedSoCInfo aspeed_socs[] = { { - .name = "ast2400-a0", - .cpu_type = ARM_CPU_TYPE_NAME("arm926"), - .silicon_rev = AST2400_A0_SILICON_REV, - .sram_size = 0x8000, - .spis_num = 1, - .fmc_typename = "aspeed.smc.fmc", - .spi_typename = aspeed_soc_ast2400_typenames, - .gpio_typename = "aspeed.gpio-ast2400", - .wdts_num = 2, - .irqmap = aspeed_soc_ast2400_irqmap, - .memmap = aspeed_soc_ast2400_memmap, - .num_cpus = 1, - }, { .name = "ast2400-a1", .cpu_type = ARM_CPU_TYPE_NAME("arm926"), .silicon_rev = AST2400_A1_SILICON_REV, @@ -XXX,XX +XXX,XX @@ static const AspeedSoCInfo aspeed_socs[] = { .irqmap = aspeed_soc_ast2400_irqmap, .memmap = aspeed_soc_ast2400_memmap, .num_cpus = 1, - }, { - .name = "ast2400", - .cpu_type = ARM_CPU_TYPE_NAME("arm926"), - .silicon_rev = AST2400_A0_SILICON_REV, - .sram_size = 0x8000, - .spis_num = 1, - .fmc_typename = "aspeed.smc.fmc", - .spi_typename = aspeed_soc_ast2400_typenames, - .gpio_typename = "aspeed.gpio-ast2400", - .wdts_num = 2, - .irqmap = aspeed_soc_ast2400_irqmap, - .memmap = aspeed_soc_ast2400_memmap, - .num_cpus = 1, }, { .name = "ast2500-a1", .cpu_type = ARM_CPU_TYPE_NAME("arm1176"), -- 2.20.1 From: Cédric Le Goater <clg@kaod.org> Improve the naming of the different controller models to ease their generation when initializing the SoC. The rename of the SMC types is breaking migration compatibility. Signed-off-by: Cédric Le Goater <clg@kaod.org> Message-id: 20190904070506.1052-5-clg@kaod.org Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/aspeed_soc.h | 3 --- hw/arm/aspeed_soc.c | 25 ++++++++++++------------- hw/ssi/aspeed_smc.c | 12 ++++++------ 3 files changed, 18 insertions(+), 22 deletions(-) diff --git a/include/hw/arm/aspeed_soc.h b/include/hw/arm/aspeed_soc.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/aspeed_soc.h +++ b/include/hw/arm/aspeed_soc.h @@ -XXX,XX +XXX,XX @@ typedef struct AspeedSoCInfo { uint32_t silicon_rev; uint64_t sram_size; int spis_num; - const char *fmc_typename; - const char **spi_typename; - const char *gpio_typename; int wdts_num; const int *irqmap; const hwaddr *memmap; diff --git a/hw/arm/aspeed_soc.c b/hw/arm/aspeed_soc.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/aspeed_soc.c +++ b/hw/arm/aspeed_soc.c @@ -XXX,XX +XXX,XX @@ static const int aspeed_soc_ast2400_irqmap[] = { #define aspeed_soc_ast2500_irqmap aspeed_soc_ast2400_irqmap -static const char *aspeed_soc_ast2400_typenames[] = { "aspeed.smc.spi" }; -static const char *aspeed_soc_ast2500_typenames[] = { - "aspeed.smc.ast2500-spi1", "aspeed.smc.ast2500-spi2" }; - static const AspeedSoCInfo aspeed_socs[] = { { .name = "ast2400-a1", @@ -XXX,XX +XXX,XX @@ static const AspeedSoCInfo aspeed_socs[] = { .silicon_rev = AST2400_A1_SILICON_REV, .sram_size = 0x8000, .spis_num = 1, - .fmc_typename = "aspeed.smc.fmc", - .spi_typename = aspeed_soc_ast2400_typenames, - .gpio_typename = "aspeed.gpio-ast2400", .wdts_num = 2, .irqmap = aspeed_soc_ast2400_irqmap, .memmap = aspeed_soc_ast2400_memmap, @@ -XXX,XX +XXX,XX @@ static const AspeedSoCInfo aspeed_socs[] = { .silicon_rev = AST2500_A1_SILICON_REV, .sram_size = 0x9000, .spis_num = 2, - .fmc_typename = "aspeed.smc.ast2500-fmc", - .spi_typename = aspeed_soc_ast2500_typenames, - .gpio_typename = "aspeed.gpio-ast2500", .wdts_num = 3, .irqmap = aspeed_soc_ast2500_irqmap, .memmap = aspeed_soc_ast2500_memmap, @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_init(Object *obj) AspeedSoCState *s = ASPEED_SOC(obj); AspeedSoCClass *sc = ASPEED_SOC_GET_CLASS(s); int i; + char socname[8]; + char typename[64]; + + if (sscanf(sc->info->name, "%7s", socname) != 1) { + g_assert_not_reached(); + } for (i = 0; i < sc->info->num_cpus; i++) { object_initialize_child(obj, "cpu[*]", OBJECT(&s->cpu[i]), @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_init(Object *obj) sysbus_init_child_obj(obj, "i2c", OBJECT(&s->i2c), sizeof(s->i2c), TYPE_ASPEED_I2C); + snprintf(typename, sizeof(typename), "aspeed.fmc-%s", socname); sysbus_init_child_obj(obj, "fmc", OBJECT(&s->fmc), sizeof(s->fmc), - sc->info->fmc_typename); + typename); object_property_add_alias(obj, "num-cs", OBJECT(&s->fmc), "num-cs", &error_abort); for (i = 0; i < sc->info->spis_num; i++) { + snprintf(typename, sizeof(typename), "aspeed.spi%d-%s", i + 1, socname); sysbus_init_child_obj(obj, "spi[*]", OBJECT(&s->spi[i]), - sizeof(s->spi[i]), sc->info->spi_typename[i]); + sizeof(s->spi[i]), typename); } sysbus_init_child_obj(obj, "sdmc", OBJECT(&s->sdmc), sizeof(s->sdmc), @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_init(Object *obj) sysbus_init_child_obj(obj, "xdma", OBJECT(&s->xdma), sizeof(s->xdma), TYPE_ASPEED_XDMA); + snprintf(typename, sizeof(typename), "aspeed.gpio-%s", socname); sysbus_init_child_obj(obj, "gpio", OBJECT(&s->gpio), sizeof(s->gpio), - sc->info->gpio_typename); + typename); } static void aspeed_soc_realize(DeviceState *dev, Error **errp) diff --git a/hw/ssi/aspeed_smc.c b/hw/ssi/aspeed_smc.c index XXXXXXX..XXXXXXX 100644 --- a/hw/ssi/aspeed_smc.c +++ b/hw/ssi/aspeed_smc.c @@ -XXX,XX +XXX,XX @@ static const AspeedSegments aspeed_segments_ast2500_spi2[] = { static const AspeedSMCController controllers[] = { { - .name = "aspeed.smc.smc", + .name = "aspeed.smc-ast2400", .r_conf = R_CONF, .r_ce_ctrl = R_CE_CTRL, .r_ctrl0 = R_CTRL0, @@ -XXX,XX +XXX,XX @@ static const AspeedSMCController controllers[] = { .has_dma = false, .nregs = ASPEED_SMC_R_SMC_MAX, }, { - .name = "aspeed.smc.fmc", + .name = "aspeed.fmc-ast2400", .r_conf = R_CONF, .r_ce_ctrl = R_CE_CTRL, .r_ctrl0 = R_CTRL0, @@ -XXX,XX +XXX,XX @@ static const AspeedSMCController controllers[] = { .has_dma = true, .nregs = ASPEED_SMC_R_MAX, }, { - .name = "aspeed.smc.spi", + .name = "aspeed.spi1-ast2400", .r_conf = R_SPI_CONF, .r_ce_ctrl = 0xff, .r_ctrl0 = R_SPI_CTRL0, @@ -XXX,XX +XXX,XX @@ static const AspeedSMCController controllers[] = { .has_dma = false, .nregs = ASPEED_SMC_R_SPI_MAX, }, { - .name = "aspeed.smc.ast2500-fmc", + .name = "aspeed.fmc-ast2500", .r_conf = R_CONF, .r_ce_ctrl = R_CE_CTRL, .r_ctrl0 = R_CTRL0, @@ -XXX,XX +XXX,XX @@ static const AspeedSMCController controllers[] = { .has_dma = true, .nregs = ASPEED_SMC_R_MAX, }, { - .name = "aspeed.smc.ast2500-spi1", + .name = "aspeed.spi1-ast2500", .r_conf = R_CONF, .r_ce_ctrl = R_CE_CTRL, .r_ctrl0 = R_CTRL0, @@ -XXX,XX +XXX,XX @@ static const AspeedSMCController controllers[] = { .has_dma = false, .nregs = ASPEED_SMC_R_MAX, }, { - .name = "aspeed.smc.ast2500-spi2", + .name = "aspeed.spi2-ast2500", .r_conf = R_CONF, .r_ce_ctrl = R_CE_CTRL, .r_ctrl0 = R_CTRL0, -- 2.20.1 From: Cédric Le Goater <clg@kaod.org> The FMC controller on the Aspeed SoCs support DMA to access the flash modules. It can operate in a normal mode, to copy to or from the flash module mapping window, or in a checksum calculation mode, to evaluate the best clock settings for reads. The model introduces two custom address spaces for DMAs: one for the AHB window of the FMC flash devices and one for the DRAM. The latter is populated using a "dram" link set from the machine with the RAM container region. Signed-off-by: Cédric Le Goater <clg@kaod.org> Acked-by: Joel Stanley <joel@jms.id.au> Message-id: 20190904070506.1052-6-clg@kaod.org Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/ssi/aspeed_smc.h | 6 + hw/arm/aspeed.c | 2 + hw/arm/aspeed_soc.c | 2 + hw/ssi/aspeed_smc.c | 222 +++++++++++++++++++++++++++++++++++- 4 files changed, 226 insertions(+), 6 deletions(-) diff --git a/include/hw/ssi/aspeed_smc.h b/include/hw/ssi/aspeed_smc.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/ssi/aspeed_smc.h +++ b/include/hw/ssi/aspeed_smc.h @@ -XXX,XX +XXX,XX @@ typedef struct AspeedSMCController { hwaddr flash_window_base; uint32_t flash_window_size; bool has_dma; + hwaddr dma_flash_mask; + hwaddr dma_dram_mask; uint32_t nregs; } AspeedSMCController; @@ -XXX,XX +XXX,XX @@ typedef struct AspeedSMCState { /* for DMA support */ uint64_t sdram_base; + AddressSpace flash_as; + MemoryRegion *dram_mr; + AddressSpace dram_as; + AspeedSMCFlash *flashes; uint8_t snoop_index; diff --git a/hw/arm/aspeed.c b/hw/arm/aspeed.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/aspeed.c +++ b/hw/arm/aspeed.c @@ -XXX,XX +XXX,XX @@ static void aspeed_board_init(MachineState *machine, &error_abort); object_property_set_int(OBJECT(&bmc->soc), machine->smp.cpus, "num-cpus", &error_abort); + object_property_set_link(OBJECT(&bmc->soc), OBJECT(&bmc->ram_container), + "dram", &error_abort); if (machine->kernel_filename) { /* * When booting with a -kernel command line there is no u-boot diff --git a/hw/arm/aspeed_soc.c b/hw/arm/aspeed_soc.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/aspeed_soc.c +++ b/hw/arm/aspeed_soc.c @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_init(Object *obj) typename); object_property_add_alias(obj, "num-cs", OBJECT(&s->fmc), "num-cs", &error_abort); + object_property_add_alias(obj, "dram", OBJECT(&s->fmc), "dram", + &error_abort); for (i = 0; i < sc->info->spis_num; i++) { snprintf(typename, sizeof(typename), "aspeed.spi%d-%s", i + 1, socname); diff --git a/hw/ssi/aspeed_smc.c b/hw/ssi/aspeed_smc.c index XXXXXXX..XXXXXXX 100644 --- a/hw/ssi/aspeed_smc.c +++ b/hw/ssi/aspeed_smc.c @@ -XXX,XX +XXX,XX @@ #include "qemu/log.h" #include "qemu/module.h" #include "qemu/error-report.h" +#include "qapi/error.h" +#include "exec/address-spaces.h" #include "hw/irq.h" #include "hw/qdev-properties.h" @@ -XXX,XX +XXX,XX @@ #define DMA_CTRL_FREQ_SHIFT 4 #define DMA_CTRL_MODE (1 << 3) #define DMA_CTRL_CKSUM (1 << 2) -#define DMA_CTRL_DIR (1 << 1) -#define DMA_CTRL_EN (1 << 0) +#define DMA_CTRL_WRITE (1 << 1) +#define DMA_CTRL_ENABLE (1 << 0) /* DMA Flash Side Address */ #define R_DMA_FLASH_ADDR (0x84 / 4) @@ -XXX,XX +XXX,XX @@ #define ASPEED_SOC_SPI_FLASH_BASE 0x30000000 #define ASPEED_SOC_SPI2_FLASH_BASE 0x38000000 +/* + * DMA DRAM addresses should be 4 bytes aligned and the valid address + * range is 0x40000000 - 0x5FFFFFFF (AST2400) + * 0x80000000 - 0xBFFFFFFF (AST2500) + * + * DMA flash addresses should be 4 bytes aligned and the valid address + * range is 0x20000000 - 0x2FFFFFFF. + * + * DMA length is from 4 bytes to 32MB + * 0: 4 bytes + * 0x7FFFFF: 32M bytes + */ +#define DMA_DRAM_ADDR(s, val) ((s)->sdram_base | \ + ((val) & (s)->ctrl->dma_dram_mask)) +#define DMA_FLASH_ADDR(s, val) ((s)->ctrl->flash_window_base | \ + ((val) & (s)->ctrl->dma_flash_mask)) +#define DMA_LENGTH(val) ((val) & 0x01FFFFFC) + /* Flash opcodes. */ #define SPI_OP_READ 0x03 /* Read data bytes (low frequency) */ @@ -XXX,XX +XXX,XX @@ static const AspeedSMCController controllers[] = { .flash_window_base = ASPEED_SOC_FMC_FLASH_BASE, .flash_window_size = 0x10000000, .has_dma = true, + .dma_flash_mask = 0x0FFFFFFC, + .dma_dram_mask = 0x1FFFFFFC, .nregs = ASPEED_SMC_R_MAX, }, { .name = "aspeed.spi1-ast2400", @@ -XXX,XX +XXX,XX @@ static const AspeedSMCController controllers[] = { .flash_window_base = ASPEED_SOC_FMC_FLASH_BASE, .flash_window_size = 0x10000000, .has_dma = true, + .dma_flash_mask = 0x0FFFFFFC, + .dma_dram_mask = 0x3FFFFFFC, .nregs = ASPEED_SMC_R_MAX, }, { .name = "aspeed.spi1-ast2500", @@ -XXX,XX +XXX,XX @@ static void aspeed_smc_reset(DeviceState *d) memset(s->regs, 0, sizeof s->regs); - /* Pretend DMA is done (u-boot initialization) */ - s->regs[R_INTR_CTRL] = INTR_CTRL_DMA_STATUS; - /* Unselect all slaves */ for (i = 0; i < s->num_cs; ++i) { s->regs[s->r_ctrl0 + i] |= CTRL_CE_STOP_ACTIVE; @@ -XXX,XX +XXX,XX @@ static uint64_t aspeed_smc_read(void *opaque, hwaddr addr, unsigned int size) addr == s->r_ce_ctrl || addr == R_INTR_CTRL || addr == R_DUMMY_DATA || + (s->ctrl->has_dma && addr == R_DMA_CTRL) || + (s->ctrl->has_dma && addr == R_DMA_FLASH_ADDR) || + (s->ctrl->has_dma && addr == R_DMA_DRAM_ADDR) || + (s->ctrl->has_dma && addr == R_DMA_LEN) || + (s->ctrl->has_dma && addr == R_DMA_CHECKSUM) || (addr >= R_SEG_ADDR0 && addr < R_SEG_ADDR0 + s->ctrl->max_slaves) || (addr >= s->r_ctrl0 && addr < s->r_ctrl0 + s->ctrl->max_slaves)) { return s->regs[addr]; @@ -XXX,XX +XXX,XX @@ static uint64_t aspeed_smc_read(void *opaque, hwaddr addr, unsigned int size) } } +/* + * Accumulate the result of the reads to provide a checksum that will + * be used to validate the read timing settings. + */ +static void aspeed_smc_dma_checksum(AspeedSMCState *s) +{ + MemTxResult result; + uint32_t data; + + if (s->regs[R_DMA_CTRL] & DMA_CTRL_WRITE) { + qemu_log_mask(LOG_GUEST_ERROR, + "%s: invalid direction for DMA checksum\n", __func__); + return; + } + + while (s->regs[R_DMA_LEN]) { + data = address_space_ldl_le(&s->flash_as, s->regs[R_DMA_FLASH_ADDR], + MEMTXATTRS_UNSPECIFIED, &result); + if (result != MEMTX_OK) { + qemu_log_mask(LOG_GUEST_ERROR, "%s: Flash read failed @%08x\n", + __func__, s->regs[R_DMA_FLASH_ADDR]); + return; + } + + /* + * When the DMA is on-going, the DMA registers are updated + * with the current working addresses and length. + */ + s->regs[R_DMA_CHECKSUM] += data; + s->regs[R_DMA_FLASH_ADDR] += 4; + s->regs[R_DMA_LEN] -= 4; + } +} + +static void aspeed_smc_dma_rw(AspeedSMCState *s) +{ + MemTxResult result; + uint32_t data; + + while (s->regs[R_DMA_LEN]) { + if (s->regs[R_DMA_CTRL] & DMA_CTRL_WRITE) { + data = address_space_ldl_le(&s->dram_as, s->regs[R_DMA_DRAM_ADDR], + MEMTXATTRS_UNSPECIFIED, &result); + if (result != MEMTX_OK) { + qemu_log_mask(LOG_GUEST_ERROR, "%s: DRAM read failed @%08x\n", + __func__, s->regs[R_DMA_DRAM_ADDR]); + return; + } + + address_space_stl_le(&s->flash_as, s->regs[R_DMA_FLASH_ADDR], + data, MEMTXATTRS_UNSPECIFIED, &result); + if (result != MEMTX_OK) { + qemu_log_mask(LOG_GUEST_ERROR, "%s: Flash write failed @%08x\n", + __func__, s->regs[R_DMA_FLASH_ADDR]); + return; + } + } else { + data = address_space_ldl_le(&s->flash_as, s->regs[R_DMA_FLASH_ADDR], + MEMTXATTRS_UNSPECIFIED, &result); + if (result != MEMTX_OK) { + qemu_log_mask(LOG_GUEST_ERROR, "%s: Flash read failed @%08x\n", + __func__, s->regs[R_DMA_FLASH_ADDR]); + return; + } + + address_space_stl_le(&s->dram_as, s->regs[R_DMA_DRAM_ADDR], + data, MEMTXATTRS_UNSPECIFIED, &result); + if (result != MEMTX_OK) { + qemu_log_mask(LOG_GUEST_ERROR, "%s: DRAM write failed @%08x\n", + __func__, s->regs[R_DMA_DRAM_ADDR]); + return; + } + } + + /* + * When the DMA is on-going, the DMA registers are updated + * with the current working addresses and length. + */ + s->regs[R_DMA_FLASH_ADDR] += 4; + s->regs[R_DMA_DRAM_ADDR] += 4; + s->regs[R_DMA_LEN] -= 4; + } +} + +static void aspeed_smc_dma_stop(AspeedSMCState *s) +{ + /* + * When the DMA is disabled, INTR_CTRL_DMA_STATUS=0 means the + * engine is idle + */ + s->regs[R_INTR_CTRL] &= ~INTR_CTRL_DMA_STATUS; + s->regs[R_DMA_CHECKSUM] = 0; + + /* + * Lower the DMA irq in any case. The IRQ control register could + * have been cleared before disabling the DMA. + */ + qemu_irq_lower(s->irq); +} + +/* + * When INTR_CTRL_DMA_STATUS=1, the DMA has completed and a new DMA + * can start even if the result of the previous was not collected. + */ +static bool aspeed_smc_dma_in_progress(AspeedSMCState *s) +{ + return s->regs[R_DMA_CTRL] & DMA_CTRL_ENABLE && + !(s->regs[R_INTR_CTRL] & INTR_CTRL_DMA_STATUS); +} + +static void aspeed_smc_dma_done(AspeedSMCState *s) +{ + s->regs[R_INTR_CTRL] |= INTR_CTRL_DMA_STATUS; + if (s->regs[R_INTR_CTRL] & INTR_CTRL_DMA_EN) { + qemu_irq_raise(s->irq); + } +} + +static void aspeed_smc_dma_ctrl(AspeedSMCState *s, uint64_t dma_ctrl) +{ + if (!(dma_ctrl & DMA_CTRL_ENABLE)) { + s->regs[R_DMA_CTRL] = dma_ctrl; + + aspeed_smc_dma_stop(s); + return; + } + + if (aspeed_smc_dma_in_progress(s)) { + qemu_log_mask(LOG_GUEST_ERROR, "%s: DMA in progress\n", __func__); + return; + } + + s->regs[R_DMA_CTRL] = dma_ctrl; + + if (s->regs[R_DMA_CTRL] & DMA_CTRL_CKSUM) { + aspeed_smc_dma_checksum(s); + } else { + aspeed_smc_dma_rw(s); + } + + aspeed_smc_dma_done(s); +} + static void aspeed_smc_write(void *opaque, hwaddr addr, uint64_t data, unsigned int size) { @@ -XXX,XX +XXX,XX @@ static void aspeed_smc_write(void *opaque, hwaddr addr, uint64_t data, } } else if (addr == R_DUMMY_DATA) { s->regs[addr] = value & 0xff; + } else if (addr == R_INTR_CTRL) { + s->regs[addr] = value; + } else if (s->ctrl->has_dma && addr == R_DMA_CTRL) { + aspeed_smc_dma_ctrl(s, value); + } else if (s->ctrl->has_dma && addr == R_DMA_DRAM_ADDR) { + s->regs[addr] = DMA_DRAM_ADDR(s, value); + } else if (s->ctrl->has_dma && addr == R_DMA_FLASH_ADDR) { + s->regs[addr] = DMA_FLASH_ADDR(s, value); + } else if (s->ctrl->has_dma && addr == R_DMA_LEN) { + s->regs[addr] = DMA_LENGTH(value); } else { qemu_log_mask(LOG_UNIMP, "%s: not implemented: 0x%" HWADDR_PRIx "\n", __func__, addr); @@ -XXX,XX +XXX,XX @@ static const MemoryRegionOps aspeed_smc_ops = { .valid.unaligned = true, }; + +/* + * Initialize the custom address spaces for DMAs + */ +static void aspeed_smc_dma_setup(AspeedSMCState *s, Error **errp) +{ + char *name; + + if (!s->dram_mr) { + error_setg(errp, TYPE_ASPEED_SMC ": 'dram' link not set"); + return; + } + + name = g_strdup_printf("%s-dma-flash", s->ctrl->name); + address_space_init(&s->flash_as, &s->mmio_flash, name); + g_free(name); + + name = g_strdup_printf("%s-dma-dram", s->ctrl->name); + address_space_init(&s->dram_as, s->dram_mr, name); + g_free(name); +} + static void aspeed_smc_realize(DeviceState *dev, Error **errp) { SysBusDevice *sbd = SYS_BUS_DEVICE(dev); @@ -XXX,XX +XXX,XX @@ static void aspeed_smc_realize(DeviceState *dev, Error **errp) s->num_cs = s->ctrl->max_slaves; } + /* DMA irq. Keep it first for the initialization in the SoC */ + sysbus_init_irq(sbd, &s->irq); + s->spi = ssi_create_bus(dev, "spi"); /* Setup cs_lines for slaves */ - sysbus_init_irq(sbd, &s->irq); s->cs_lines = g_new0(qemu_irq, s->num_cs); ssi_auto_connect_slaves(dev, s->cs_lines, s->spi); @@ -XXX,XX +XXX,XX @@ static void aspeed_smc_realize(DeviceState *dev, Error **errp) memory_region_add_subregion(&s->mmio_flash, offset, &fl->mmio); offset += fl->size; } + + /* DMA support */ + if (s->ctrl->has_dma) { + aspeed_smc_dma_setup(s, errp); + } } static const VMStateDescription vmstate_aspeed_smc = { @@ -XXX,XX +XXX,XX @@ static const VMStateDescription vmstate_aspeed_smc = { static Property aspeed_smc_properties[] = { DEFINE_PROP_UINT32("num-cs", AspeedSMCState, num_cs, 1), DEFINE_PROP_UINT64("sdram-base", AspeedSMCState, sdram_base, 0), + DEFINE_PROP_LINK("dram", AspeedSMCState, dram_mr, + TYPE_MEMORY_REGION, MemoryRegion *), DEFINE_PROP_END_OF_LIST(), }; -- 2.20.1 From: Cédric Le Goater <clg@kaod.org> When doing calibration, the SPI clock rate in the CE0 Control Register and the read delay cycles in the Read Timing Compensation Register are set using bit[11:4] of the DMA Control Register. Signed-off-by: Cédric Le Goater <clg@kaod.org> Acked-by: Joel Stanley <joel@jms.id.au> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Message-id: 20190904070506.1052-7-clg@kaod.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/ssi/aspeed_smc.c | 64 ++++++++++++++++++++++++++++++++++++++++++++- 1 file changed, 63 insertions(+), 1 deletion(-) diff --git a/hw/ssi/aspeed_smc.c b/hw/ssi/aspeed_smc.c index XXXXXXX..XXXXXXX 100644 --- a/hw/ssi/aspeed_smc.c +++ b/hw/ssi/aspeed_smc.c @@ -XXX,XX +XXX,XX @@ #define CTRL_CMD_MASK 0xff #define CTRL_DUMMY_HIGH_SHIFT 14 #define CTRL_AST2400_SPI_4BYTE (1 << 13) +#define CE_CTRL_CLOCK_FREQ_SHIFT 8 +#define CE_CTRL_CLOCK_FREQ_MASK 0xf +#define CE_CTRL_CLOCK_FREQ(div) \ + (((div) & CE_CTRL_CLOCK_FREQ_MASK) << CE_CTRL_CLOCK_FREQ_SHIFT) #define CTRL_DUMMY_LOW_SHIFT 6 /* 2 bits [7:6] */ #define CTRL_CE_STOP_ACTIVE (1 << 2) #define CTRL_CMD_MODE_MASK 0x3 @@ -XXX,XX +XXX,XX @@ #define DMA_CTRL_DELAY_SHIFT 8 #define DMA_CTRL_FREQ_MASK 0xf #define DMA_CTRL_FREQ_SHIFT 4 -#define DMA_CTRL_MODE (1 << 3) +#define DMA_CTRL_CALIB (1 << 3) #define DMA_CTRL_CKSUM (1 << 2) #define DMA_CTRL_WRITE (1 << 1) #define DMA_CTRL_ENABLE (1 << 0) @@ -XXX,XX +XXX,XX @@ static uint64_t aspeed_smc_read(void *opaque, hwaddr addr, unsigned int size) } } +static uint8_t aspeed_smc_hclk_divisor(uint8_t hclk_mask) +{ + /* HCLK/1 .. HCLK/16 */ + const uint8_t hclk_divisors[] = { + 15, 7, 14, 6, 13, 5, 12, 4, 11, 3, 10, 2, 9, 1, 8, 0 + }; + int i; + + for (i = 0; i < ARRAY_SIZE(hclk_divisors); i++) { + if (hclk_mask == hclk_divisors[i]) { + return i + 1; + } + } + + qemu_log_mask(LOG_GUEST_ERROR, "invalid HCLK mask %x", hclk_mask); + return 0; +} + +/* + * When doing calibration, the SPI clock rate in the CE0 Control + * Register and the read delay cycles in the Read Timing Compensation + * Register are set using bit[11:4] of the DMA Control Register. + */ +static void aspeed_smc_dma_calibration(AspeedSMCState *s) +{ + uint8_t delay = + (s->regs[R_DMA_CTRL] >> DMA_CTRL_DELAY_SHIFT) & DMA_CTRL_DELAY_MASK; + uint8_t hclk_mask = + (s->regs[R_DMA_CTRL] >> DMA_CTRL_FREQ_SHIFT) & DMA_CTRL_FREQ_MASK; + uint8_t hclk_div = aspeed_smc_hclk_divisor(hclk_mask); + uint32_t hclk_shift = (hclk_div - 1) << 2; + uint8_t cs; + + /* + * The Read Timing Compensation Register values apply to all CS on + * the SPI bus and only HCLK/1 - HCLK/5 can have tunable delays + */ + if (hclk_div && hclk_div < 6) { + s->regs[s->r_timings] &= ~(0xf << hclk_shift); + s->regs[s->r_timings] |= delay << hclk_shift; + } + + /* + * TODO: compute the CS from the DMA address and the segment + * registers. This is not really a problem for now because the + * Timing Register values apply to all CS and software uses CS0 to + * do calibration. + */ + cs = 0; + s->regs[s->r_ctrl0 + cs] &= + ~(CE_CTRL_CLOCK_FREQ_MASK << CE_CTRL_CLOCK_FREQ_SHIFT); + s->regs[s->r_ctrl0 + cs] |= CE_CTRL_CLOCK_FREQ(hclk_div); +} + /* * Accumulate the result of the reads to provide a checksum that will * be used to validate the read timing settings. @@ -XXX,XX +XXX,XX @@ static void aspeed_smc_dma_checksum(AspeedSMCState *s) return; } + if (s->regs[R_DMA_CTRL] & DMA_CTRL_CALIB) { + aspeed_smc_dma_calibration(s); + } + while (s->regs[R_DMA_LEN]) { data = address_space_ldl_le(&s->flash_as, s->regs[R_DMA_FLASH_ADDR], MEMTXATTRS_UNSPECIFIED, &result); -- 2.20.1 From: Cédric Le Goater <clg@kaod.org> Emulate read errors in the DMA Checksum Register for high frequencies and optimistic settings of the Read Timing Compensation Register. This will help in tuning the SPI timing calibration algorithm. Errors are only injected when the property "inject_failure" is set to true as suggested by Philippe. The values below are those to expect from the first flash device of the FMC controller of a palmetto-bmc machine. Cc: Philippe Mathieu-Daudé <philmd@redhat.com> Signed-off-by: Cédric Le Goater <clg@kaod.org> Reviewed-by: Joel Stanley <joel@jms.id.au> Message-id: 20190904070506.1052-8-clg@kaod.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/ssi/aspeed_smc.h | 1 + hw/ssi/aspeed_smc.c | 36 ++++++++++++++++++++++++++++++++++++ 2 files changed, 37 insertions(+) diff --git a/include/hw/ssi/aspeed_smc.h b/include/hw/ssi/aspeed_smc.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/ssi/aspeed_smc.h +++ b/include/hw/ssi/aspeed_smc.h @@ -XXX,XX +XXX,XX @@ typedef struct AspeedSMCState { uint32_t num_cs; qemu_irq *cs_lines; + bool inject_failure; SSIBus *spi; diff --git a/hw/ssi/aspeed_smc.c b/hw/ssi/aspeed_smc.c index XXXXXXX..XXXXXXX 100644 --- a/hw/ssi/aspeed_smc.c +++ b/hw/ssi/aspeed_smc.c @@ -XXX,XX +XXX,XX @@ static void aspeed_smc_dma_calibration(AspeedSMCState *s) s->regs[s->r_ctrl0 + cs] |= CE_CTRL_CLOCK_FREQ(hclk_div); } +/* + * Emulate read errors in the DMA Checksum Register for high + * frequencies and optimistic settings of the Read Timing Compensation + * Register. This will help in tuning the SPI timing calibration + * algorithm. + */ +static bool aspeed_smc_inject_read_failure(AspeedSMCState *s) +{ + uint8_t delay = + (s->regs[R_DMA_CTRL] >> DMA_CTRL_DELAY_SHIFT) & DMA_CTRL_DELAY_MASK; + uint8_t hclk_mask = + (s->regs[R_DMA_CTRL] >> DMA_CTRL_FREQ_SHIFT) & DMA_CTRL_FREQ_MASK; + + /* + * Typical values of a palmetto-bmc machine. + */ + switch (aspeed_smc_hclk_divisor(hclk_mask)) { + case 4 ... 16: + return false; + case 3: /* at least one HCLK cycle delay */ + return (delay & 0x7) < 1; + case 2: /* at least two HCLK cycle delay */ + return (delay & 0x7) < 2; + case 1: /* (> 100MHz) is above the max freq of the controller */ + return true; + default: + g_assert_not_reached(); + } +} + /* * Accumulate the result of the reads to provide a checksum that will * be used to validate the read timing settings. @@ -XXX,XX +XXX,XX @@ static void aspeed_smc_dma_checksum(AspeedSMCState *s) s->regs[R_DMA_FLASH_ADDR] += 4; s->regs[R_DMA_LEN] -= 4; } + + if (s->inject_failure && aspeed_smc_inject_read_failure(s)) { + s->regs[R_DMA_CHECKSUM] = 0xbadc0de; + } + } static void aspeed_smc_dma_rw(AspeedSMCState *s) @@ -XXX,XX +XXX,XX @@ static const VMStateDescription vmstate_aspeed_smc = { static Property aspeed_smc_properties[] = { DEFINE_PROP_UINT32("num-cs", AspeedSMCState, num_cs, 1), + DEFINE_PROP_BOOL("inject-failure", AspeedSMCState, inject_failure, false), DEFINE_PROP_UINT64("sdram-base", AspeedSMCState, sdram_base, 0), DEFINE_PROP_LINK("dram", AspeedSMCState, dram_mr, TYPE_MEMORY_REGION, MemoryRegion *), -- 2.20.1 From: Christian Svensson <bluecmd@google.com> This patch adds the missing checksum calculation on normal DMA transfer. According to the datasheet this is how the SMC should behave. Verified on AST1250 that the hardware matches the behaviour. Signed-off-by: Christian Svensson <bluecmd@google.com> Reviewed-by: Joel Stanley <joel@jms.id.au> Signed-off-by: Cédric Le Goater <clg@kaod.org> Message-id: 20190904070506.1052-9-clg@kaod.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/ssi/aspeed_smc.c | 1 + 1 file changed, 1 insertion(+) diff --git a/hw/ssi/aspeed_smc.c b/hw/ssi/aspeed_smc.c index XXXXXXX..XXXXXXX 100644 --- a/hw/ssi/aspeed_smc.c +++ b/hw/ssi/aspeed_smc.c @@ -XXX,XX +XXX,XX @@ static void aspeed_smc_dma_rw(AspeedSMCState *s) s->regs[R_DMA_FLASH_ADDR] += 4; s->regs[R_DMA_DRAM_ADDR] += 4; s->regs[R_DMA_LEN] -= 4; + s->regs[R_DMA_CHECKSUM] += data; } } -- 2.20.1 From: Cédric Le Goater <clg@kaod.org> and use a class AspeedSCUClass to define each SoC characteristics. Signed-off-by: Cédric Le Goater <clg@kaod.org> Message-id: 20190904070506.1052-10-clg@kaod.org Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/misc/aspeed_scu.h | 15 +++++++ hw/arm/aspeed_soc.c | 3 +- hw/misc/aspeed_scu.c | 83 ++++++++++++++++++++---------------- 3 files changed, 64 insertions(+), 37 deletions(-) diff --git a/include/hw/misc/aspeed_scu.h b/include/hw/misc/aspeed_scu.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/misc/aspeed_scu.h +++ b/include/hw/misc/aspeed_scu.h @@ -XXX,XX +XXX,XX @@ #define TYPE_ASPEED_SCU "aspeed.scu" #define ASPEED_SCU(obj) OBJECT_CHECK(AspeedSCUState, (obj), TYPE_ASPEED_SCU) +#define TYPE_ASPEED_2400_SCU TYPE_ASPEED_SCU "-ast2400" +#define TYPE_ASPEED_2500_SCU TYPE_ASPEED_SCU "-ast2500" #define ASPEED_SCU_NR_REGS (0x1A8 >> 2) @@ -XXX,XX +XXX,XX @@ typedef struct AspeedSCUState { extern bool is_supported_silicon_rev(uint32_t silicon_rev); +#define ASPEED_SCU_CLASS(klass) \ + OBJECT_CLASS_CHECK(AspeedSCUClass, (klass), TYPE_ASPEED_SCU) +#define ASPEED_SCU_GET_CLASS(obj) \ + OBJECT_GET_CLASS(AspeedSCUClass, (obj), TYPE_ASPEED_SCU) + +typedef struct AspeedSCUClass { + SysBusDeviceClass parent_class; + + const uint32_t *resets; + uint32_t (*calc_hpll)(AspeedSCUState *s); + uint32_t apb_divider; +} AspeedSCUClass; + #define ASPEED_SCU_PROT_KEY 0x1688A8A8 /* diff --git a/hw/arm/aspeed_soc.c b/hw/arm/aspeed_soc.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/aspeed_soc.c +++ b/hw/arm/aspeed_soc.c @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_init(Object *obj) &error_abort, NULL); } + snprintf(typename, sizeof(typename), "aspeed.scu-%s", socname); sysbus_init_child_obj(obj, "scu", OBJECT(&s->scu), sizeof(s->scu), - TYPE_ASPEED_SCU); + typename); qdev_prop_set_uint32(DEVICE(&s->scu), "silicon-rev", sc->info->silicon_rev); object_property_add_alias(obj, "hw-strap1", OBJECT(&s->scu), diff --git a/hw/misc/aspeed_scu.c b/hw/misc/aspeed_scu.c index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/aspeed_scu.c +++ b/hw/misc/aspeed_scu.c @@ -XXX,XX +XXX,XX @@ static uint32_t aspeed_scu_get_random(void) static void aspeed_scu_set_apb_freq(AspeedSCUState *s) { - uint32_t apb_divider; - - switch (s->silicon_rev) { - case AST2400_A0_SILICON_REV: - case AST2400_A1_SILICON_REV: - apb_divider = 2; - break; - case AST2500_A0_SILICON_REV: - case AST2500_A1_SILICON_REV: - apb_divider = 4; - break; - default: - g_assert_not_reached(); - } + AspeedSCUClass *asc = ASPEED_SCU_GET_CLASS(s); s->apb_freq = s->hpll / (SCU_CLK_GET_PCLK_DIV(s->regs[CLK_SEL]) + 1) - / apb_divider; + / asc->apb_divider; } static uint64_t aspeed_scu_read(void *opaque, hwaddr offset, unsigned size) @@ -XXX,XX +XXX,XX @@ static const uint32_t hpll_ast2400_freqs[][4] = { { 400, 375, 350, 425 }, /* 25MHz */ }; -static uint32_t aspeed_scu_calc_hpll_ast2400(AspeedSCUState *s) +static uint32_t aspeed_2400_scu_calc_hpll(AspeedSCUState *s) { uint32_t hpll_reg = s->regs[HPLL_PARAM]; uint8_t freq_select; @@ -XXX,XX +XXX,XX @@ static uint32_t aspeed_scu_calc_hpll_ast2400(AspeedSCUState *s) return hpll_ast2400_freqs[clk_25m_in][freq_select] * 1000000; } -static uint32_t aspeed_scu_calc_hpll_ast2500(AspeedSCUState *s) +static uint32_t aspeed_2500_scu_calc_hpll(AspeedSCUState *s) { uint32_t hpll_reg = s->regs[HPLL_PARAM]; uint32_t multiplier = 1; @@ -XXX,XX +XXX,XX @@ static uint32_t aspeed_scu_calc_hpll_ast2500(AspeedSCUState *s) static void aspeed_scu_reset(DeviceState *dev) { AspeedSCUState *s = ASPEED_SCU(dev); - const uint32_t *reset; - uint32_t (*calc_hpll)(AspeedSCUState *s); + AspeedSCUClass *asc = ASPEED_SCU_GET_CLASS(dev); - switch (s->silicon_rev) { - case AST2400_A0_SILICON_REV: - case AST2400_A1_SILICON_REV: - reset = ast2400_a0_resets; - calc_hpll = aspeed_scu_calc_hpll_ast2400; - break; - case AST2500_A0_SILICON_REV: - case AST2500_A1_SILICON_REV: - reset = ast2500_a1_resets; - calc_hpll = aspeed_scu_calc_hpll_ast2500; - break; - default: - g_assert_not_reached(); - } - - memcpy(s->regs, reset, sizeof(s->regs)); + memcpy(s->regs, asc->resets, sizeof(s->regs)); s->regs[SILICON_REV] = s->silicon_rev; s->regs[HW_STRAP1] = s->hw_strap1; s->regs[HW_STRAP2] = s->hw_strap2; @@ -XXX,XX +XXX,XX @@ static void aspeed_scu_reset(DeviceState *dev) * All registers are set. Now compute the frequencies of the main clocks */ s->clkin = aspeed_scu_get_clkin(s); - s->hpll = calc_hpll(s); + s->hpll = asc->calc_hpll(s); aspeed_scu_set_apb_freq(s); } @@ -XXX,XX +XXX,XX @@ static const TypeInfo aspeed_scu_info = { .parent = TYPE_SYS_BUS_DEVICE, .instance_size = sizeof(AspeedSCUState), .class_init = aspeed_scu_class_init, + .class_size = sizeof(AspeedSCUClass), + .abstract = true, +}; + +static void aspeed_2400_scu_class_init(ObjectClass *klass, void *data) +{ + DeviceClass *dc = DEVICE_CLASS(klass); + AspeedSCUClass *asc = ASPEED_SCU_CLASS(klass); + + dc->desc = "ASPEED 2400 System Control Unit"; + asc->resets = ast2400_a0_resets; + asc->calc_hpll = aspeed_2400_scu_calc_hpll; + asc->apb_divider = 2; +} + +static const TypeInfo aspeed_2400_scu_info = { + .name = TYPE_ASPEED_2400_SCU, + .parent = TYPE_ASPEED_SCU, + .instance_size = sizeof(AspeedSCUState), + .class_init = aspeed_2400_scu_class_init, +}; + +static void aspeed_2500_scu_class_init(ObjectClass *klass, void *data) +{ + DeviceClass *dc = DEVICE_CLASS(klass); + AspeedSCUClass *asc = ASPEED_SCU_CLASS(klass); + + dc->desc = "ASPEED 2500 System Control Unit"; + asc->resets = ast2500_a1_resets; + asc->calc_hpll = aspeed_2500_scu_calc_hpll; + asc->apb_divider = 4; +} + +static const TypeInfo aspeed_2500_scu_info = { + .name = TYPE_ASPEED_2500_SCU, + .parent = TYPE_ASPEED_SCU, + .instance_size = sizeof(AspeedSCUState), + .class_init = aspeed_2500_scu_class_init, }; static void aspeed_scu_register_types(void) { type_register_static(&aspeed_scu_info); + type_register_static(&aspeed_2400_scu_info); + type_register_static(&aspeed_2500_scu_info); } type_init(aspeed_scu_register_types); -- 2.20.1 From: Cédric Le Goater <clg@kaod.org> The APB frequency can be calculated directly when needed from the HPLL_PARAM and CLK_SEL register values. This removes useless state in the model. Signed-off-by: Cédric Le Goater <clg@kaod.org> Message-id: 20190904070506.1052-11-clg@kaod.org Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/misc/aspeed_scu.h | 8 +++----- hw/misc/aspeed_scu.c | 25 +++++++++---------------- hw/timer/aspeed_timer.c | 3 ++- 3 files changed, 14 insertions(+), 22 deletions(-) diff --git a/include/hw/misc/aspeed_scu.h b/include/hw/misc/aspeed_scu.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/misc/aspeed_scu.h +++ b/include/hw/misc/aspeed_scu.h @@ -XXX,XX +XXX,XX @@ typedef struct AspeedSCUState { uint32_t hw_strap1; uint32_t hw_strap2; uint32_t hw_prot_key; - - uint32_t clkin; - uint32_t hpll; - uint32_t apb_freq; } AspeedSCUState; #define AST2400_A0_SILICON_REV 0x02000303U @@ -XXX,XX +XXX,XX @@ typedef struct AspeedSCUClass { SysBusDeviceClass parent_class; const uint32_t *resets; - uint32_t (*calc_hpll)(AspeedSCUState *s); + uint32_t (*calc_hpll)(AspeedSCUState *s, uint32_t hpll_reg); uint32_t apb_divider; } AspeedSCUClass; #define ASPEED_SCU_PROT_KEY 0x1688A8A8 +uint32_t aspeed_scu_get_apb_freq(AspeedSCUState *s); + /* * Extracted from Aspeed SDK v00.03.21. Fixes and extra definitions * were added. diff --git a/hw/misc/aspeed_scu.c b/hw/misc/aspeed_scu.c index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/aspeed_scu.c +++ b/hw/misc/aspeed_scu.c @@ -XXX,XX +XXX,XX @@ static uint32_t aspeed_scu_get_random(void) return num; } -static void aspeed_scu_set_apb_freq(AspeedSCUState *s) +uint32_t aspeed_scu_get_apb_freq(AspeedSCUState *s) { AspeedSCUClass *asc = ASPEED_SCU_GET_CLASS(s); + uint32_t hpll = asc->calc_hpll(s, s->regs[HPLL_PARAM]); - s->apb_freq = s->hpll / (SCU_CLK_GET_PCLK_DIV(s->regs[CLK_SEL]) + 1) + return hpll / (SCU_CLK_GET_PCLK_DIV(s->regs[CLK_SEL]) + 1) / asc->apb_divider; } @@ -XXX,XX +XXX,XX @@ static void aspeed_scu_write(void *opaque, hwaddr offset, uint64_t data, return; case CLK_SEL: s->regs[reg] = data; - aspeed_scu_set_apb_freq(s); break; case HW_STRAP1: if (ASPEED_IS_AST2500(s->regs[SILICON_REV])) { @@ -XXX,XX +XXX,XX @@ static const uint32_t hpll_ast2400_freqs[][4] = { { 400, 375, 350, 425 }, /* 25MHz */ }; -static uint32_t aspeed_2400_scu_calc_hpll(AspeedSCUState *s) +static uint32_t aspeed_2400_scu_calc_hpll(AspeedSCUState *s, uint32_t hpll_reg) { - uint32_t hpll_reg = s->regs[HPLL_PARAM]; uint8_t freq_select; bool clk_25m_in; + uint32_t clkin = aspeed_scu_get_clkin(s); if (hpll_reg & SCU_AST2400_H_PLL_OFF) { return 0; @@ -XXX,XX +XXX,XX @@ static uint32_t aspeed_2400_scu_calc_hpll(AspeedSCUState *s) multiplier = (2 - od) * ((n + 2) / (d + 1)); } - return s->clkin * multiplier; + return clkin * multiplier; } /* HW strapping */ @@ -XXX,XX +XXX,XX @@ static uint32_t aspeed_2400_scu_calc_hpll(AspeedSCUState *s) return hpll_ast2400_freqs[clk_25m_in][freq_select] * 1000000; } -static uint32_t aspeed_2500_scu_calc_hpll(AspeedSCUState *s) +static uint32_t aspeed_2500_scu_calc_hpll(AspeedSCUState *s, uint32_t hpll_reg) { - uint32_t hpll_reg = s->regs[HPLL_PARAM]; uint32_t multiplier = 1; + uint32_t clkin = aspeed_scu_get_clkin(s); if (hpll_reg & SCU_H_PLL_OFF) { return 0; @@ -XXX,XX +XXX,XX @@ static uint32_t aspeed_2500_scu_calc_hpll(AspeedSCUState *s) multiplier = ((m + 1) / (n + 1)) / (p + 1); } - return s->clkin * multiplier; + return clkin * multiplier; } static void aspeed_scu_reset(DeviceState *dev) @@ -XXX,XX +XXX,XX @@ static void aspeed_scu_reset(DeviceState *dev) s->regs[HW_STRAP1] = s->hw_strap1; s->regs[HW_STRAP2] = s->hw_strap2; s->regs[PROT_KEY] = s->hw_prot_key; - - /* - * All registers are set. Now compute the frequencies of the main clocks - */ - s->clkin = aspeed_scu_get_clkin(s); - s->hpll = asc->calc_hpll(s); - aspeed_scu_set_apb_freq(s); } static uint32_t aspeed_silicon_revs[] = { diff --git a/hw/timer/aspeed_timer.c b/hw/timer/aspeed_timer.c index XXXXXXX..XXXXXXX 100644 --- a/hw/timer/aspeed_timer.c +++ b/hw/timer/aspeed_timer.c @@ -XXX,XX +XXX,XX @@ static inline uint32_t calculate_rate(struct AspeedTimer *t) { AspeedTimerCtrlState *s = timer_to_ctrl(t); - return timer_external_clock(t) ? TIMER_CLOCK_EXT_HZ : s->scu->apb_freq; + return timer_external_clock(t) ? TIMER_CLOCK_EXT_HZ : + aspeed_scu_get_apb_freq(s->scu); } static inline uint32_t calculate_ticks(struct AspeedTimer *t, uint64_t now_ns) -- 2.20.1 From: "Emilio G. Cota" <cota@braap.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Signed-off-by: Emilio G. Cota <cota@braap.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Alex Bennée <alex.bennee@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- accel/tcg/atomic_template.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/accel/tcg/atomic_template.h b/accel/tcg/atomic_template.h index XXXXXXX..XXXXXXX 100644 --- a/accel/tcg/atomic_template.h +++ b/accel/tcg/atomic_template.h @@ -XXX,XX +XXX,XX @@ ABI_TYPE ATOMIC_NAME(xchg)(CPUArchState *env, target_ulong addr, #define GEN_ATOMIC_HELPER(X) \ ABI_TYPE ATOMIC_NAME(X)(CPUArchState *env, target_ulong addr, \ - ABI_TYPE val EXTRA_ARGS) \ + ABI_TYPE val EXTRA_ARGS) \ { \ ATOMIC_MMU_DECLS; \ DATA_TYPE *haddr = ATOMIC_MMU_LOOKUP; \ -- 2.20.1 The qemu-ga documentation is currently in qemu-ga.texi in Texinfo format, which we present to the user as: * a qemu-ga manpage * a section of the main qemu-doc HTML documentation Convert the documentation to rST format, and present it to the user as: * a qemu-ga manpage * part of the interop/ Sphinx manual Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Michael Roth <mdroth@linux.vnet.ibm.com> Tested-by: Michael Roth <mdroth@linux.vnet.ibm.com> Message-id: 20190905131040.8350-1-peter.maydell@linaro.org --- Makefile | 24 ++++--- MAINTAINERS | 2 +- docs/conf.py | 18 ++--- docs/interop/conf.py | 7 ++ docs/interop/index.rst | 1 + docs/interop/qemu-ga.rst | 133 +++++++++++++++++++++++++++++++++++++ qemu-doc.texi | 5 -- qemu-ga.texi | 137 --------------------------------------- 8 files changed, 166 insertions(+), 161 deletions(-) create mode 100644 docs/interop/qemu-ga.rst delete mode 100644 qemu-ga.texi diff --git a/Makefile b/Makefile index XXXXXXX..XXXXXXX 100644 --- a/Makefile +++ b/Makefile @@ -XXX,XX +XXX,XX @@ endif endif ifdef BUILD_DOCS -DOCS=qemu-doc.html qemu-doc.txt qemu.1 qemu-img.1 qemu-nbd.8 qemu-ga.8 +DOCS=qemu-doc.html qemu-doc.txt qemu.1 qemu-img.1 qemu-nbd.8 docs/interop/qemu-ga.8 DOCS+=docs/interop/qemu-qmp-ref.html docs/interop/qemu-qmp-ref.txt docs/interop/qemu-qmp-ref.7 DOCS+=docs/interop/qemu-ga-ref.html docs/interop/qemu-ga-ref.txt docs/interop/qemu-ga-ref.7 DOCS+=docs/qemu-block-drivers.7 @@ -XXX,XX +XXX,XX @@ DESCS= endif # Note that we manually filter-out the non-Sphinx documentation which -# is currently built into the docs/interop directory in the build tree. +# is currently built into the docs/interop directory in the build tree, +# and also any sphinx-built manpages. define install-manual = for d in $$(cd $(MANUAL_BUILDDIR) && find $1 -type d); do $(INSTALL_DIR) "$(DESTDIR)$(qemu_docdir)/$$d"; done -for f in $$(cd $(MANUAL_BUILDDIR) && find $1 -type f -a '!' '(' -name 'qemu-*-qapi.*' -o -name 'qemu-*-ref.*' ')' ); do $(INSTALL_DATA) "$(MANUAL_BUILDDIR)/$$f" "$(DESTDIR)$(qemu_docdir)/$$f"; done +for f in $$(cd $(MANUAL_BUILDDIR) && find $1 -type f -a '!' '(' -name '*.[0-9]' -o -name 'qemu-*-qapi.*' -o -name 'qemu-*-ref.*' ')' ); do $(INSTALL_DATA) "$(MANUAL_BUILDDIR)/$$f" "$(DESTDIR)$(qemu_docdir)/$$f"; done endef # Note that we deliberately do not install the "devel" manual: it is @@ -XXX,XX +XXX,XX @@ ifdef CONFIG_TRACE_SYSTEMTAP $(INSTALL_DATA) scripts/qemu-trace-stap.1 "$(DESTDIR)$(mandir)/man1" endif ifneq (,$(findstring qemu-ga,$(TOOLS))) - $(INSTALL_DATA) qemu-ga.8 "$(DESTDIR)$(mandir)/man8" + $(INSTALL_DATA) docs/interop/qemu-ga.8 "$(DESTDIR)$(mandir)/man8" $(INSTALL_DATA) docs/interop/qemu-ga-ref.html "$(DESTDIR)$(qemu_docdir)" $(INSTALL_DATA) docs/interop/qemu-ga-ref.txt "$(DESTDIR)$(qemu_docdir)" $(INSTALL_DATA) docs/interop/qemu-ga-ref.7 "$(DESTDIR)$(mandir)/man7" @@ -XXX,XX +XXX,XX @@ docs/version.texi: $(SRC_PATH)/VERSION config-host.mak sphinxdocs: $(MANUAL_BUILDDIR)/devel/index.html $(MANUAL_BUILDDIR)/interop/index.html $(MANUAL_BUILDDIR)/specs/index.html # Canned command to build a single manual -build-manual = $(call quiet-command,sphinx-build $(if $(V),,-q) -W -n -b html -D version=$(VERSION) -D release="$(FULL_VERSION)" -d .doctrees/$1 $(SRC_PATH)/docs/$1 $(MANUAL_BUILDDIR)/$1 ,"SPHINX","$(MANUAL_BUILDDIR)/$1") +# Arguments: $1 = manual name, $2 = Sphinx builder ('html' or 'man') +build-manual = $(call quiet-command,CONFDIR="$(qemu_confdir)" sphinx-build $(if $(V),,-q) -W -n -b $2 -D version=$(VERSION) -D release="$(FULL_VERSION)" -d .doctrees/$1 $(SRC_PATH)/docs/$1 $(MANUAL_BUILDDIR)/$1 ,"SPHINX","$(MANUAL_BUILDDIR)/$1") # We assume all RST files in the manual's directory are used in it manual-deps = $(wildcard $(SRC_PATH)/docs/$1/*.rst) $(SRC_PATH)/docs/$1/conf.py $(SRC_PATH)/docs/conf.py $(MANUAL_BUILDDIR)/devel/index.html: $(call manual-deps,devel) - $(call build-manual,devel) + $(call build-manual,devel,html) $(MANUAL_BUILDDIR)/interop/index.html: $(call manual-deps,interop) - $(call build-manual,interop) + $(call build-manual,interop,html) $(MANUAL_BUILDDIR)/specs/index.html: $(call manual-deps,specs) - $(call build-manual,specs) + $(call build-manual,specs,html) + +$(MANUAL_BUILDDIR)/interop/qemu-ga.8: $(call manual-deps,interop) + $(call build-manual,interop,man) qemu-options.texi: $(SRC_PATH)/qemu-options.hx $(SRC_PATH)/scripts/hxtool $(call quiet-command,sh $(SRC_PATH)/scripts/hxtool -t < $< > $@,"GEN","$@") @@ -XXX,XX +XXX,XX @@ qemu.1: qemu-option-trace.texi qemu-img.1: qemu-img.texi qemu-option-trace.texi qemu-img-cmds.texi fsdev/virtfs-proxy-helper.1: fsdev/virtfs-proxy-helper.texi qemu-nbd.8: qemu-nbd.texi qemu-option-trace.texi -qemu-ga.8: qemu-ga.texi docs/qemu-block-drivers.7: docs/qemu-block-drivers.texi docs/qemu-cpu-models.7: docs/qemu-cpu-models.texi scripts/qemu-trace-stap.1: scripts/qemu-trace-stap.texi @@ -XXX,XX +XXX,XX @@ txt: qemu-doc.txt docs/interop/qemu-qmp-ref.txt docs/interop/qemu-ga-ref.txt qemu-doc.html qemu-doc.info qemu-doc.pdf qemu-doc.txt: \ qemu-img.texi qemu-nbd.texi qemu-options.texi \ qemu-tech.texi qemu-option-trace.texi \ - qemu-deprecated.texi qemu-monitor.texi qemu-img-cmds.texi qemu-ga.texi \ + qemu-deprecated.texi qemu-monitor.texi qemu-img-cmds.texi \ qemu-monitor-info.texi docs/qemu-block-drivers.texi \ docs/qemu-cpu-models.texi docs/security.texi diff --git a/MAINTAINERS b/MAINTAINERS index XXXXXXX..XXXXXXX 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -XXX,XX +XXX,XX @@ QEMU Guest Agent M: Michael Roth <mdroth@linux.vnet.ibm.com> S: Maintained F: qga/ -F: qemu-ga.texi +F: docs/interop/qemu-ga.rst F: scripts/qemu-guest-agent/ F: tests/test-qga.c F: docs/interop/qemu-ga-ref.texi diff --git a/docs/conf.py b/docs/conf.py index XXXXXXX..XXXXXXX 100644 --- a/docs/conf.py +++ b/docs/conf.py @@ -XXX,XX +XXX,XX @@ todo_include_todos = False # with "option::" in the document being processed. Turn that off. suppress_warnings = ["ref.option"] +# The rst_epilog fragment is effectively included in every rST file. +# We use it to define substitutions based on build config that +# can then be used in the documentation. The fallback if the +# environment variable is not set is for the benefit of readthedocs +# style document building; our Makefile always sets the variable. +confdir = os.getenv('CONFDIR', "/etc/qemu") +rst_epilog = ".. |CONFDIR| replace:: ``" + confdir + "``\n" + # -- Options for HTML output ---------------------------------------------- # The theme to use for HTML and HTML Help pages. See the documentation for @@ -XXX,XX +XXX,XX @@ latex_documents = [ # -- Options for manual page output --------------------------------------- - -# One entry per manual page. List of tuples -# (source start file, name, description, authors, manual section). -man_pages = [ - (master_doc, 'qemu', u'QEMU Documentation', - [author], 1) -] - +# Individual manual/conf.py can override this to create man pages +man_pages = [] # -- Options for Texinfo output ------------------------------------------- diff --git a/docs/interop/conf.py b/docs/interop/conf.py index XXXXXXX..XXXXXXX 100644 --- a/docs/interop/conf.py +++ b/docs/interop/conf.py @@ -XXX,XX +XXX,XX @@ exec(compile(open(parent_config, "rb").read(), parent_config, 'exec')) # This slightly misuses the 'description', but is the best way to get # the manual title to appear in the sidebar. html_theme_options['description'] = u'System Emulation Management and Interoperability Guide' + +# One entry per manual page. List of tuples +# (source start file, name, description, authors, manual section). +man_pages = [ + ('qemu-ga', 'qemu-ga', u'QEMU Guest Agent', + ['Michael Roth <mdroth@linux.vnet.ibm.com>'], 8) +] diff --git a/docs/interop/index.rst b/docs/interop/index.rst index XXXXXXX..XXXXXXX 100644 --- a/docs/interop/index.rst +++ b/docs/interop/index.rst @@ -XXX,XX +XXX,XX @@ Contents: bitmaps live-block-operations pr-helper + qemu-ga vhost-user vhost-user-gpu diff --git a/docs/interop/qemu-ga.rst b/docs/interop/qemu-ga.rst new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/docs/interop/qemu-ga.rst @@ -XXX,XX +XXX,XX @@ +QEMU Guest Agent +================ + +Synopsis +-------- + +**qemu-ga** [*OPTIONS*] + +Description +----------- + +The QEMU Guest Agent is a daemon intended to be run within virtual +machines. It allows the hypervisor host to perform various operations +in the guest, such as: + +- get information from the guest +- set the guest's system time +- read/write a file +- sync and freeze the filesystems +- suspend the guest +- reconfigure guest local processors +- set user's password +- ... + +qemu-ga will read a system configuration file on startup (located at +|CONFDIR|\ ``/qemu-ga.conf`` by default), then parse remaining +configuration options on the command line. For the same key, the last +option wins, but the lists accumulate (see below for configuration +file format). + +Options +------- + +.. program:: qemu-ga + +.. option:: -m, --method=METHOD + + Transport method: one of ``unix-listen``, ``virtio-serial``, or + ``isa-serial`` (``virtio-serial`` is the default). + +.. option:: -p, --path=PATH + + Device/socket path (the default for virtio-serial is + ``/dev/virtio-ports/org.qemu.guest_agent.0``, + the default for isa-serial is ``/dev/ttyS0``) + +.. option:: -l, --logfile=PATH + + Set log file path (default is stderr). + +.. option:: -f, --pidfile=PATH + + Specify pid file (default is ``/var/run/qemu-ga.pid``). + +.. option:: -F, --fsfreeze-hook=PATH + + Enable fsfreeze hook. Accepts an optional argument that specifies + script to run on freeze/thaw. Script will be called with + 'freeze'/'thaw' arguments accordingly (default is + |CONFDIR|\ ``/fsfreeze-hook``). If using -F with an argument, do + not follow -F with a space (for example: + ``-F/var/run/fsfreezehook.sh``). + +.. option:: -t, --statedir=PATH + + Specify the directory to store state information (absolute paths only, + default is ``/var/run``). + +.. option:: -v, --verbose + + Log extra debugging information. + +.. option:: -V, --version + + Print version information and exit. + +.. option:: -d, --daemon + + Daemonize after startup (detach from terminal). + +.. option:: -b, --blacklist=LIST + + Comma-separated list of RPCs to disable (no spaces, ``?`` to list + available RPCs). + +.. option:: -D, --dump-conf + + Dump the configuration in a format compatible with ``qemu-ga.conf`` + and exit. + +.. option:: -h, --help + + Display this help and exit. + +Files +----- + + +The syntax of the ``qemu-ga.conf`` configuration file follows the +Desktop Entry Specification, here is a quick summary: it consists of +groups of key-value pairs, interspersed with comments. + +:: + + # qemu-ga configuration sample + [general] + daemonize = 0 + pidfile = /var/run/qemu-ga.pid + verbose = 0 + method = virtio-serial + path = /dev/virtio-ports/org.qemu.guest_agent.0 + statedir = /var/run + +The list of keys follows the command line options: + +============= =========== +Key Key type +============= =========== +daemon boolean +method string +path string +logfile string +pidfile string +fsfreeze-hook string +statedir string +verbose boolean +blacklist string list +============= =========== + +See also +-------- + +:manpage:`qemu(1)` diff --git a/qemu-doc.texi b/qemu-doc.texi index XXXXXXX..XXXXXXX 100644 --- a/qemu-doc.texi +++ b/qemu-doc.texi @@ -XXX,XX +XXX,XX @@ so should only be used with trusted guest OS. @c man end -@node QEMU Guest Agent -@chapter QEMU Guest Agent invocation - -@include qemu-ga.texi - @node QEMU User space emulator @chapter QEMU User space emulator diff --git a/qemu-ga.texi b/qemu-ga.texi deleted file mode 100644 index XXXXXXX..XXXXXXX --- a/qemu-ga.texi +++ /dev/null @@ -XXX,XX +XXX,XX @@ -@example -@c man begin SYNOPSIS -@command{qemu-ga} [@var{OPTIONS}] -@c man end -@end example - -@c man begin DESCRIPTION - -The QEMU Guest Agent is a daemon intended to be run within virtual -machines. It allows the hypervisor host to perform various operations -in the guest, such as: - -@itemize -@item -get information from the guest -@item -set the guest's system time -@item -read/write a file -@item -sync and freeze the filesystems -@item -suspend the guest -@item -reconfigure guest local processors -@item -set user's password -@item -... -@end itemize - -qemu-ga will read a system configuration file on startup (located at -@file{@value{CONFDIR}/qemu-ga.conf} by default), then parse remaining -configuration options on the command line. For the same key, the last -option wins, but the lists accumulate (see below for configuration -file format). - -@c man end - -@c man begin OPTIONS -@table @option -@item -m, --method=@var{method} - Transport method: one of @samp{unix-listen}, @samp{virtio-serial}, or - @samp{isa-serial} (@samp{virtio-serial} is the default). - -@item -p, --path=@var{path} - Device/socket path (the default for virtio-serial is - @samp{/dev/virtio-ports/org.qemu.guest_agent.0}, - the default for isa-serial is @samp{/dev/ttyS0}) - -@item -l, --logfile=@var{path} - Set log file path (default is stderr). - -@item -f, --pidfile=@var{path} - Specify pid file (default is @samp{/var/run/qemu-ga.pid}). - -@item -F, --fsfreeze-hook=@var{path} - Enable fsfreeze hook. Accepts an optional argument that specifies - script to run on freeze/thaw. Script will be called with - 'freeze'/'thaw' arguments accordingly (default is - @samp{@value{CONFDIR}/fsfreeze-hook}). If using -F with an argument, do - not follow -F with a space (for example: - @samp{-F/var/run/fsfreezehook.sh}). - -@item -t, --statedir=@var{path} - Specify the directory to store state information (absolute paths only, - default is @samp{/var/run}). - -@item -v, --verbose - Log extra debugging information. - -@item -V, --version - Print version information and exit. - -@item -d, --daemon - Daemonize after startup (detach from terminal). - -@item -b, --blacklist=@var{list} - Comma-separated list of RPCs to disable (no spaces, @samp{?} to list - available RPCs). - -@item -D, --dump-conf - Dump the configuration in a format compatible with @file{qemu-ga.conf} - and exit. - -@item -h, --help - Display this help and exit. -@end table - -@c man end - -@c man begin FILES - -The syntax of the @file{qemu-ga.conf} configuration file follows the -Desktop Entry Specification, here is a quick summary: it consists of -groups of key-value pairs, interspersed with comments. - -@example -# qemu-ga configuration sample -[general] -daemonize = 0 -pidfile = /var/run/qemu-ga.pid -verbose = 0 -method = virtio-serial -path = /dev/virtio-ports/org.qemu.guest_agent.0 -statedir = /var/run -@end example - -The list of keys follows the command line options: -@table @option -@item daemon= boolean -@item method= string -@item path= string -@item logfile= string -@item pidfile= string -@item fsfreeze-hook= string -@item statedir= string -@item verbose= boolean -@item blacklist= string list -@end table - -@c man end - -@ignore - -@setfilename qemu-ga -@settitle QEMU Guest Agent - -@c man begin AUTHOR -Michael Roth <mdroth@linux.vnet.ibm.com> -@c man end - -@c man begin SEEALSO -qemu(1) -@c man end - -@end ignore -- 2.20.1