Re: [PATCH v2 0/3] target-i386: Use C struct for xsave area layout, offsets & sizes

[Paolo Bonzini <pbonzini@redhat.com>]

On 30/11/2015 18:34, Eduardo Habkost wrote:
> target-i386/cpu.c:ext_save_area uses magic numbers for the xsave
> area offets and sizes, and target-i386/kvm.c:kvm_{put,get}_xsave()
> uses offset macros and bit manipulation to access the xsave area.
> This series changes both to use C structs for those operations.
> 
> I still need to figure out a way to write unit tests for the new
> code. Maybe I will just copy and paste the new and old functions,
> and test them locally (checking if they give the same results
> when translating blobs of random bytes).
> 
> Changes v1 -> v2:
> * Use uint8_t[8*n] instead of uint64_t[n] for register data
> * Keep the QEMU_BUILD_BUG_ON lines
> 
> v1 -> v2 diff below:
> 
>   diff --git a/target-i386/cpu.h b/target-i386/cpu.h
>   index 3d1d01e..41f55ef 100644
>   --- a/target-i386/cpu.h
>   +++ b/target-i386/cpu.h
>   @@ -818,7 +818,7 @@ typedef union X86LegacyXSaveArea {
>            uint32_t mxcsr;
>            uint32_t mxcsr_mask;
>            FPReg fpregs[8];
>   -        uint64_t xmm_regs[16][2];
>   +        uint8_t xmm_regs[16][16];
>        };
>        uint8_t data[512];
>    } X86LegacyXSaveArea;
>   @@ -831,7 +831,7 @@ typedef struct X86XSaveHeader {
> 
>    /* Ext. save area 2: AVX State */
>    typedef struct XSaveAVX {
>   -    uint64_t ymmh[16][2];
>   +    uint8_t ymmh[16][16];
>    } XSaveAVX;
> 
>    /* Ext. save area 3: BNDREG */
>   @@ -852,12 +852,12 @@ typedef struct XSaveOpmask {
> 
>    /* Ext. save area 6: ZMM_Hi256 */
>    typedef struct XSaveZMM_Hi256 {
>   -    uint64_t zmm_hi256[16][4];
>   +    uint8_t zmm_hi256[16][32];
>    } XSaveZMM_Hi256;
> 
>    /* Ext. save area 7: Hi16_ZMM */
>    typedef struct XSaveHi16_ZMM {
>   -    XMMReg hi16_zmm[16];
>   +    uint8_t hi16_zmm[16][64];
>    } XSaveHi16_ZMM;
> 
>    typedef struct X86XSaveArea {
>   diff --git a/target-i386/kvm.c b/target-i386/kvm.c
>   index 5e7ec70..98249e4 100644
>   --- a/target-i386/kvm.c
>   +++ b/target-i386/kvm.c
>   @@ -1203,6 +1203,43 @@ static int kvm_put_fpu(X86CPU *cpu)
>        return kvm_vcpu_ioctl(CPU(cpu), KVM_SET_FPU, &fpu);
>    }
> 
>   +#define XSAVE_FCW_FSW     0
>   +#define XSAVE_FTW_FOP     1
>   +#define XSAVE_CWD_RIP     2
>   +#define XSAVE_CWD_RDP     4
>   +#define XSAVE_MXCSR       6
>   +#define XSAVE_ST_SPACE    8
>   +#define XSAVE_XMM_SPACE   40
>   +#define XSAVE_XSTATE_BV   128
>   +#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
>   +
>   +#define XSAVE_BYTE_OFFSET(word_offset) \
>   +    ((word_offset)*sizeof(((struct kvm_xsave*)0)->region[0]))
>   +
>   +#define ASSERT_OFFSET(word_offset, field) \
>   +    QEMU_BUILD_BUG_ON(XSAVE_BYTE_OFFSET(word_offset) != \
>   +                      offsetof(X86XSaveArea, field))
>   +
>   +ASSERT_OFFSET(XSAVE_FCW_FSW, legacy.fcw);
>   +ASSERT_OFFSET(XSAVE_FTW_FOP, legacy.ftw);
>   +ASSERT_OFFSET(XSAVE_CWD_RIP, legacy.fpip);
>   +ASSERT_OFFSET(XSAVE_CWD_RDP, legacy.fpdp);
>   +ASSERT_OFFSET(XSAVE_MXCSR, legacy.mxcsr);
>   +ASSERT_OFFSET(XSAVE_ST_SPACE, legacy.fpregs);
>   +ASSERT_OFFSET(XSAVE_XMM_SPACE, legacy.xmm_regs);
>   +ASSERT_OFFSET(XSAVE_XSTATE_BV, header.xstate_bv);
>   +ASSERT_OFFSET(XSAVE_YMMH_SPACE, avx_state);
>   +ASSERT_OFFSET(XSAVE_BNDREGS, bndreg_state);
>   +ASSERT_OFFSET(XSAVE_BNDCSR, bndcsr_state);
>   +ASSERT_OFFSET(XSAVE_OPMASK, opmask_state);
>   +ASSERT_OFFSET(XSAVE_ZMM_Hi256, zmm_hi256_state);
>   +ASSERT_OFFSET(XSAVE_Hi16_ZMM, hi16_zmm_state);
>   +
>    static int kvm_put_xsave(X86CPU *cpu)
>    {
>        CPUX86State *env = &cpu->env;
>   @@ -1239,17 +1276,17 @@ static int kvm_put_xsave(X86CPU *cpu)
>                sizeof env->opmask_regs);
> 
>        for (i = 0; i < CPU_NB_REGS; i++) {
>   -        X86LegacyXSaveArea *legacy = &xsave->legacy;
>   -        XSaveAVX *avx = &xsave->avx_state;
>   -        XSaveZMM_Hi256 *zmm_hi256 = &xsave->zmm_hi256_state;
>   -        stq_p(&legacy->xmm_regs[i][0],     env->xmm_regs[i].XMM_Q(0));
>   -        stq_p(&legacy->xmm_regs[i][1],     env->xmm_regs[i].XMM_Q(1));
>   -        stq_p(&avx->ymmh[i][0],            env->xmm_regs[i].XMM_Q(2));
>   -        stq_p(&avx->ymmh[i][1],            env->xmm_regs[i].XMM_Q(3));
>   -        stq_p(&zmm_hi256->zmm_hi256[i][0], env->xmm_regs[i].XMM_Q(4));
>   -        stq_p(&zmm_hi256->zmm_hi256[i][1], env->xmm_regs[i].XMM_Q(5));
>   -        stq_p(&zmm_hi256->zmm_hi256[i][2], env->xmm_regs[i].XMM_Q(6));
>   -        stq_p(&zmm_hi256->zmm_hi256[i][3], env->xmm_regs[i].XMM_Q(7));
>   +        uint8_t *xmm = xsave->legacy.xmm_regs[i];
>   +        uint8_t *ymmh = xsave->avx_state.ymmh[i];
>   +        uint8_t *zmmh = xsave->zmm_hi256_state.zmm_hi256[i];
>   +        stq_p(xmm,     env->xmm_regs[i].XMM_Q(0));
>   +        stq_p(xmm+8,   env->xmm_regs[i].XMM_Q(1));
>   +        stq_p(ymmh,    env->xmm_regs[i].XMM_Q(2));
>   +        stq_p(ymmh+8,  env->xmm_regs[i].XMM_Q(3));
>   +        stq_p(zmmh,    env->xmm_regs[i].XMM_Q(4));
>   +        stq_p(zmmh+8,  env->xmm_regs[i].XMM_Q(5));
>   +        stq_p(zmmh+16, env->xmm_regs[i].XMM_Q(6));
>   +        stq_p(zmmh+24, env->xmm_regs[i].XMM_Q(7));
>        }
> 
>    #ifdef TARGET_X86_64
>   @@ -1625,17 +1662,17 @@ static int kvm_get_xsave(X86CPU *cpu)
>                sizeof env->opmask_regs);
> 
>        for (i = 0; i < CPU_NB_REGS; i++) {
>   -        X86LegacyXSaveArea *legacy = &xsave->legacy;
>   -        XSaveAVX *avx = &xsave->avx_state;
>   -        XSaveZMM_Hi256 *zmm_hi256 = &xsave->zmm_hi256_state;
>   -        env->xmm_regs[i].XMM_Q(0) = ldq_p(&legacy->xmm_regs[i][0]);
>   -        env->xmm_regs[i].XMM_Q(1) = ldq_p(&legacy->xmm_regs[i][1]);
>   -        env->xmm_regs[i].XMM_Q(2) = ldq_p(&avx->ymmh[i][0]);
>   -        env->xmm_regs[i].XMM_Q(3) = ldq_p(&avx->ymmh[i][1]);
>   -        env->xmm_regs[i].XMM_Q(4) = ldq_p(&zmm_hi256->zmm_hi256[i][0]);
>   -        env->xmm_regs[i].XMM_Q(5) = ldq_p(&zmm_hi256->zmm_hi256[i][1]);
>   -        env->xmm_regs[i].XMM_Q(6) = ldq_p(&zmm_hi256->zmm_hi256[i][2]);
>   -        env->xmm_regs[i].XMM_Q(7) = ldq_p(&zmm_hi256->zmm_hi256[i][3]);
>   +        uint8_t *xmm = xsave->legacy.xmm_regs[i];
>   +        uint8_t *ymmh = xsave->avx_state.ymmh[i];
>   +        uint8_t *zmmh = xsave->zmm_hi256_state.zmm_hi256[i];
>   +        env->xmm_regs[i].XMM_Q(0) = ldq_p(xmm);
>   +        env->xmm_regs[i].XMM_Q(1) = ldq_p(xmm+8);
>   +        env->xmm_regs[i].XMM_Q(2) = ldq_p(ymmh);
>   +        env->xmm_regs[i].XMM_Q(3) = ldq_p(ymmh+8);
>   +        env->xmm_regs[i].XMM_Q(4) = ldq_p(zmmh);
>   +        env->xmm_regs[i].XMM_Q(5) = ldq_p(zmmh+8);
>   +        env->xmm_regs[i].XMM_Q(6) = ldq_p(zmmh+16);
>   +        env->xmm_regs[i].XMM_Q(7) = ldq_p(zmmh+24);
>        }
> 
>    #ifdef TARGET_X86_64
> 
> Eduardo Habkost (3):
>   target-i386: Define structs for layout of xsave area
>   target-i386: Use xsave structs for ext_save_area
>   target-i386: kvm: Use X86XSaveArea struct for xsave save/load
> 
>  target-i386/cpu.c | 18 +++++++----
>  target-i386/cpu.h | 85 ++++++++++++++++++++++++++++++++++++++++++++++++
>  target-i386/kvm.c | 96 +++++++++++++++++++++++++++++++++----------------------
>  3 files changed, 155 insertions(+), 44 deletions(-)
> 

The patches are okay, are you going to rebase them on top of the PKRU
patches?

Paolo