From 5567938d29bf8a0ce451e59a8db009c12ab3444c Mon Sep 17 00:00:00 2001 From: Eduardo Habkost Date: Thu, 25 Jun 2015 19:31:29 +0200 Subject: [PATCH 09/10] target-i386: add Intel AVX-512 support Message-id: <1435260689-9556-9-git-send-email-ehabkost@redhat.com> Patchwork-id: 66506 O-Subject: [RHEL-7.2 qemu-kvm PATCH 8/8] target-i386: add Intel AVX-512 support Bugzilla: 1233350 RH-Acked-by: Igor Mammedov RH-Acked-by: Bandan Das RH-Acked-by: Paolo Bonzini Add AVX512 feature bits, register definition and corresponding xsave/vmstate support. Reviewed-by: Eduardo Habkost Signed-off-by: Chao Peng Signed-off-by: Paolo Bonzini (cherry picked from commit 9aecd6f8aef653cea58932f06a2740299dbe5fd3) Signed-off-by: Eduardo Habkost Signed-off-by: Miroslav Rezanina Conflicts: target-i386/cpu.h target-i386/machine.c --- target-i386/cpu.c | 10 ++++-- target-i386/cpu.h | 61 ++++++++++++++++++++++++++++++++++++ target-i386/kvm.c | 19 +++++++++++ target-i386/machine.c | 87 +++++++++++++++++++++++++++++++++++++++++++++++++++ 4 files changed, 175 insertions(+), 2 deletions(-) diff --git a/target-i386/cpu.c b/target-i386/cpu.c index fe60c33..06efe17 100644 --- a/target-i386/cpu.c +++ b/target-i386/cpu.c @@ -146,8 +146,8 @@ static const char *svm_feature_name[] = { static const char *cpuid_7_0_ebx_feature_name[] = { "fsgsbase", NULL, NULL, "bmi1", "hle", "avx2", NULL, "smep", "bmi2", "erms", "invpcid", "rtm", NULL, NULL, "mpx", NULL, - NULL, NULL, "rdseed", "adx", "smap", NULL, NULL, NULL, - NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, + "avx512f", NULL, "rdseed", "adx", "smap", NULL, NULL, NULL, + NULL, NULL, "avx512pf", "avx512er", "avx512cd", NULL, NULL, NULL, }; typedef struct FeatureWordInfo { @@ -228,6 +228,12 @@ static const ExtSaveArea ext_save_areas[] = { .offset = 0x3c0, .size = 0x40 }, [4] = { .feature = FEAT_7_0_EBX, .bits = CPUID_7_0_EBX_MPX, .offset = 0x400, .size = 0x40 }, + [5] = { .feature = FEAT_7_0_EBX, .bits = CPUID_7_0_EBX_AVX512F, + .offset = 0x440, .size = 0x40 }, + [6] = { .feature = FEAT_7_0_EBX, .bits = CPUID_7_0_EBX_AVX512F, + .offset = 0x480, .size = 0x200 }, + [7] = { .feature = FEAT_7_0_EBX, .bits = CPUID_7_0_EBX_AVX512F, + .offset = 0x680, .size = 0x400 }, }; const char *get_register_name_32(unsigned int reg) diff --git a/target-i386/cpu.h b/target-i386/cpu.h index fc1e42e..61e9b86 100644 --- a/target-i386/cpu.h +++ b/target-i386/cpu.h @@ -390,6 +390,9 @@ #define XSTATE_YMM (1ULL << 2) #define XSTATE_BNDREGS (1ULL << 3) #define XSTATE_BNDCSR (1ULL << 4) +#define XSTATE_OPMASK (1ULL << 5) +#define XSTATE_ZMM_Hi256 (1ULL << 6) +#define XSTATE_Hi16_ZMM (1ULL << 7) /* CPUID feature words */ @@ -554,9 +557,13 @@ typedef uint32_t FeatureWordArray[FEATURE_WORDS]; #define CPUID_7_0_EBX_INVPCID (1U << 10) #define CPUID_7_0_EBX_RTM (1U << 11) #define CPUID_7_0_EBX_MPX (1U << 14) +#define CPUID_7_0_EBX_AVX512F (1U << 16) /* AVX-512 Foundation */ #define CPUID_7_0_EBX_RDSEED (1U << 18) #define CPUID_7_0_EBX_ADX (1U << 19) #define CPUID_7_0_EBX_SMAP (1U << 20) +#define CPUID_7_0_EBX_AVX512PF (1U << 26) /* AVX-512 Prefetch */ +#define CPUID_7_0_EBX_AVX512ER (1U << 27) /* AVX-512 Exponential and Reciprocal */ +#define CPUID_7_0_EBX_AVX512CD (1U << 28) /* AVX-512 Conflict Detection */ #define CPUID_VENDOR_SZ 12 @@ -697,6 +704,24 @@ typedef union { } XMMReg; typedef union { + uint8_t _b[32]; + uint16_t _w[16]; + uint32_t _l[8]; + uint64_t _q[4]; + float32 _s[8]; + float64 _d[4]; +} YMMReg; + +typedef union { + uint8_t _b[64]; + uint16_t _w[32]; + uint32_t _l[16]; + uint64_t _q[8]; + float32 _s[16]; + float64 _d[8]; +} ZMMReg; + +typedef union { uint8_t _b[8]; uint16_t _w[4]; uint32_t _l[2]; @@ -715,6 +740,20 @@ typedef struct BNDCSReg { } BNDCSReg; #ifdef HOST_WORDS_BIGENDIAN +#define ZMM_B(n) _b[63 - (n)] +#define ZMM_W(n) _w[31 - (n)] +#define ZMM_L(n) _l[15 - (n)] +#define ZMM_S(n) _s[15 - (n)] +#define ZMM_Q(n) _q[7 - (n)] +#define ZMM_D(n) _d[7 - (n)] + +#define YMM_B(n) _b[31 - (n)] +#define YMM_W(n) _w[15 - (n)] +#define YMM_L(n) _l[7 - (n)] +#define YMM_S(n) _s[7 - (n)] +#define YMM_Q(n) _q[3 - (n)] +#define YMM_D(n) _d[3 - (n)] + #define XMM_B(n) _b[15 - (n)] #define XMM_W(n) _w[7 - (n)] #define XMM_L(n) _l[3 - (n)] @@ -727,6 +766,20 @@ typedef struct BNDCSReg { #define MMX_L(n) _l[1 - (n)] #define MMX_S(n) _s[1 - (n)] #else +#define ZMM_B(n) _b[n] +#define ZMM_W(n) _w[n] +#define ZMM_L(n) _l[n] +#define ZMM_S(n) _s[n] +#define ZMM_Q(n) _q[n] +#define ZMM_D(n) _d[n] + +#define YMM_B(n) _b[n] +#define YMM_W(n) _w[n] +#define YMM_L(n) _l[n] +#define YMM_S(n) _s[n] +#define YMM_Q(n) _q[n] +#define YMM_D(n) _d[n] + #define XMM_B(n) _b[n] #define XMM_W(n) _w[n] #define XMM_L(n) _l[n] @@ -765,6 +818,8 @@ typedef struct { #define NB_MMU_MODES 3 +#define NB_OPMASK_REGS 8 + typedef enum TPRAccess { TPR_ACCESS_READ, TPR_ACCESS_WRITE, @@ -826,6 +881,12 @@ typedef struct CPUX86State { XMMReg ymmh_regs[CPU_NB_REGS]; + uint64_t opmask_regs[NB_OPMASK_REGS]; + YMMReg zmmh_regs[CPU_NB_REGS]; +#ifdef TARGET_X86_64 + ZMMReg hi16_zmm_regs[CPU_NB_REGS]; +#endif + /* sysenter registers */ uint32_t sysenter_cs; target_ulong sysenter_esp; diff --git a/target-i386/kvm.c b/target-i386/kvm.c index bbbbac0..c91bfb8 100644 --- a/target-i386/kvm.c +++ b/target-i386/kvm.c @@ -995,6 +995,9 @@ static int kvm_put_fpu(X86CPU *cpu) #define XSAVE_YMMH_SPACE 144 #define XSAVE_BNDREGS 240 #define XSAVE_BNDCSR 256 +#define XSAVE_OPMASK 272 +#define XSAVE_ZMM_Hi256 288 +#define XSAVE_Hi16_ZMM 416 static int kvm_put_xsave(X86CPU *cpu) { @@ -1031,6 +1034,14 @@ static int kvm_put_xsave(X86CPU *cpu) sizeof env->bnd_regs); memcpy(&xsave->region[XSAVE_BNDCSR], &env->bndcs_regs, sizeof(env->bndcs_regs)); + memcpy(&xsave->region[XSAVE_OPMASK], env->opmask_regs, + sizeof env->opmask_regs); + memcpy(&xsave->region[XSAVE_ZMM_Hi256], env->zmmh_regs, + sizeof env->zmmh_regs); +#ifdef TARGET_X86_64 + memcpy(&xsave->region[XSAVE_Hi16_ZMM], env->hi16_zmm_regs, + sizeof env->hi16_zmm_regs); +#endif r = kvm_vcpu_ioctl(CPU(cpu), KVM_SET_XSAVE, xsave); return r; } @@ -1354,6 +1365,14 @@ static int kvm_get_xsave(X86CPU *cpu) sizeof env->bnd_regs); memcpy(&env->bndcs_regs, &xsave->region[XSAVE_BNDCSR], sizeof(env->bndcs_regs)); + memcpy(env->opmask_regs, &xsave->region[XSAVE_OPMASK], + sizeof env->opmask_regs); + memcpy(env->zmmh_regs, &xsave->region[XSAVE_ZMM_Hi256], + sizeof env->zmmh_regs); +#ifdef TARGET_X86_64 + memcpy(env->hi16_zmm_regs, &xsave->region[XSAVE_Hi16_ZMM], + sizeof env->hi16_zmm_regs); +#endif return 0; } diff --git a/target-i386/machine.c b/target-i386/machine.c index b6be606..2c97002 100644 --- a/target-i386/machine.c +++ b/target-i386/machine.c @@ -63,6 +63,44 @@ static const VMStateDescription vmstate_ymmh_reg = { #define VMSTATE_YMMH_REGS_VARS(_field, _state, _n, _v) \ VMSTATE_STRUCT_ARRAY(_field, _state, _n, _v, vmstate_ymmh_reg, XMMReg) +static const VMStateDescription vmstate_zmmh_reg = { + .name = "zmmh_reg", + .version_id = 1, + .minimum_version_id = 1, + .fields = (VMStateField[]) { + VMSTATE_UINT64(YMM_Q(0), YMMReg), + VMSTATE_UINT64(YMM_Q(1), YMMReg), + VMSTATE_UINT64(YMM_Q(2), YMMReg), + VMSTATE_UINT64(YMM_Q(3), YMMReg), + VMSTATE_END_OF_LIST() + } +}; + +#define VMSTATE_ZMMH_REGS_VARS(_field, _state, _n) \ + VMSTATE_STRUCT_ARRAY(_field, _state, _n, 0, vmstate_zmmh_reg, YMMReg) + +#ifdef TARGET_X86_64 +static const VMStateDescription vmstate_hi16_zmm_reg = { + .name = "hi16_zmm_reg", + .version_id = 1, + .minimum_version_id = 1, + .fields = (VMStateField[]) { + VMSTATE_UINT64(ZMM_Q(0), ZMMReg), + VMSTATE_UINT64(ZMM_Q(1), ZMMReg), + VMSTATE_UINT64(ZMM_Q(2), ZMMReg), + VMSTATE_UINT64(ZMM_Q(3), ZMMReg), + VMSTATE_UINT64(ZMM_Q(4), ZMMReg), + VMSTATE_UINT64(ZMM_Q(5), ZMMReg), + VMSTATE_UINT64(ZMM_Q(6), ZMMReg), + VMSTATE_UINT64(ZMM_Q(7), ZMMReg), + VMSTATE_END_OF_LIST() + } +}; + +#define VMSTATE_Hi16_ZMM_REGS_VARS(_field, _state, _n) \ + VMSTATE_STRUCT_ARRAY(_field, _state, _n, 0, vmstate_hi16_zmm_reg, ZMMReg) +#endif + static const VMStateDescription vmstate_bnd_regs = { .name = "bnd_regs", .version_id = 1, @@ -620,6 +658,52 @@ static const VMStateDescription vmstate_mpx = { } }; +static bool avx512_needed(void *opaque) +{ + X86CPU *cpu = opaque; + CPUX86State *env = &cpu->env; + unsigned int i; + + for (i = 0; i < NB_OPMASK_REGS; i++) { + if (env->opmask_regs[i]) { + return true; + } + } + + for (i = 0; i < CPU_NB_REGS; i++) { +#define ENV_ZMMH(reg, field) (env->zmmh_regs[reg].YMM_Q(field)) + if (ENV_ZMMH(i, 0) || ENV_ZMMH(i, 1) || + ENV_ZMMH(i, 2) || ENV_ZMMH(i, 3)) { + return true; + } +#ifdef TARGET_X86_64 +#define ENV_Hi16_ZMM(reg, field) (env->hi16_zmm_regs[reg].ZMM_Q(field)) + if (ENV_Hi16_ZMM(i, 0) || ENV_Hi16_ZMM(i, 1) || + ENV_Hi16_ZMM(i, 2) || ENV_Hi16_ZMM(i, 3) || + ENV_Hi16_ZMM(i, 4) || ENV_Hi16_ZMM(i, 5) || + ENV_Hi16_ZMM(i, 6) || ENV_Hi16_ZMM(i, 7)) { + return true; + } +#endif + } + + return false; +} + +static const VMStateDescription vmstate_avx512 = { + .name = "cpu/avx512", + .version_id = 1, + .minimum_version_id = 1, + .fields = (VMStateField[]) { + VMSTATE_UINT64_ARRAY(env.opmask_regs, X86CPU, NB_OPMASK_REGS), + VMSTATE_ZMMH_REGS_VARS(env.zmmh_regs, X86CPU, CPU_NB_REGS), +#ifdef TARGET_X86_64 + VMSTATE_Hi16_ZMM_REGS_VARS(env.hi16_zmm_regs, X86CPU, CPU_NB_REGS), +#endif + VMSTATE_END_OF_LIST() + } +}; + const VMStateDescription vmstate_x86_cpu = { .name = "cpu", .version_id = 12, @@ -763,6 +847,9 @@ const VMStateDescription vmstate_x86_cpu = { } , { .vmsd = &vmstate_mpx, .needed = mpx_needed, + }, { + .vmsd = &vmstate_avx512, + .needed = avx512_needed, } , { /* empty */ } -- 1.8.3.1