Implement SSE4.1, SSE4.2 (x86).

This adds support for CPUID_EXT_SSE41, CPUID_EXT_SSE42, CPUID_EXT_POPCNT
extensions.  Most instructions haven't been tested yet.


git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@5411 c046a42c-6fe2-441c-8c8c-71466251a162

3 files changed, 973 insertions, 33 deletions

diff --git a/target-i386/ops_sse.h b/target-i386/ops_sse.h
index 4fa8e06624..1f1db50241 100644
--- a/target-i386/ops_sse.h
+++ b/target-i386/ops_sse.h
@@ -1,7 +1,8 @@
 /*
- *  MMX/3DNow!/SSE/SSE2/SSE3/SSSE3/PNI support
+ *  MMX/3DNow!/SSE/SSE2/SSE3/SSSE3/SSE4/PNI support
  *
  *  Copyright (c) 2005 Fabrice Bellard
+ *  Copyright (c) 2008 Intel Corporation  <andrew.zaborowski@intel.com>
  *
  * This library is free software; you can redistribute it and/or
  * modify it under the terms of the GNU Lesser General Public
@@ -1420,6 +1421,621 @@ void glue(helper_palignr, SUFFIX) (Reg *d, Reg *s, int32_t shift)
     *d = r;
 }
 
+#define XMM0 env->xmm_regs[0]
+
+#if SHIFT == 1
+#define SSE_HELPER_V(name, elem, num, F)\
+void glue(name, SUFFIX) (Reg *d, Reg *s)\
+{\
+    d->elem(0) = F(d->elem(0), s->elem(0), XMM0.elem(0));\
+    d->elem(1) = F(d->elem(1), s->elem(1), XMM0.elem(1));\
+    if (num > 2) {\
+        d->elem(2) = F(d->elem(2), s->elem(2), XMM0.elem(2));\
+        d->elem(3) = F(d->elem(3), s->elem(3), XMM0.elem(3));\
+        if (num > 4) {\
+            d->elem(4) = F(d->elem(4), s->elem(4), XMM0.elem(4));\
+            d->elem(5) = F(d->elem(5), s->elem(5), XMM0.elem(5));\
+            d->elem(6) = F(d->elem(6), s->elem(6), XMM0.elem(6));\
+            d->elem(7) = F(d->elem(7), s->elem(7), XMM0.elem(7));\
+            if (num > 8) {\
+                d->elem(8) = F(d->elem(8), s->elem(8), XMM0.elem(8));\
+                d->elem(9) = F(d->elem(9), s->elem(9), XMM0.elem(9));\
+                d->elem(10) = F(d->elem(10), s->elem(10), XMM0.elem(10));\
+                d->elem(11) = F(d->elem(11), s->elem(11), XMM0.elem(11));\
+                d->elem(12) = F(d->elem(12), s->elem(12), XMM0.elem(12));\
+                d->elem(13) = F(d->elem(13), s->elem(13), XMM0.elem(13));\
+                d->elem(14) = F(d->elem(14), s->elem(14), XMM0.elem(14));\
+                d->elem(15) = F(d->elem(15), s->elem(15), XMM0.elem(15));\
+            }\
+        }\
+    }\
+}
+
+#define SSE_HELPER_I(name, elem, num, F)\
+void glue(name, SUFFIX) (Reg *d, Reg *s, uint32_t imm)\
+{\
+    d->elem(0) = F(d->elem(0), s->elem(0), ((imm >> 0) & 1));\
+    d->elem(1) = F(d->elem(1), s->elem(1), ((imm >> 1) & 1));\
+    if (num > 2) {\
+        d->elem(2) = F(d->elem(2), s->elem(2), ((imm >> 2) & 1));\
+        d->elem(3) = F(d->elem(3), s->elem(3), ((imm >> 3) & 1));\
+        if (num > 4) {\
+            d->elem(4) = F(d->elem(4), s->elem(4), ((imm >> 4) & 1));\
+            d->elem(5) = F(d->elem(5), s->elem(5), ((imm >> 5) & 1));\
+            d->elem(6) = F(d->elem(6), s->elem(6), ((imm >> 6) & 1));\
+            d->elem(7) = F(d->elem(7), s->elem(7), ((imm >> 7) & 1));\
+            if (num > 8) {\
+                d->elem(8) = F(d->elem(8), s->elem(8), ((imm >> 8) & 1));\
+                d->elem(9) = F(d->elem(9), s->elem(9), ((imm >> 9) & 1));\
+                d->elem(10) = F(d->elem(10), s->elem(10), ((imm >> 10) & 1));\
+                d->elem(11) = F(d->elem(11), s->elem(11), ((imm >> 11) & 1));\
+                d->elem(12) = F(d->elem(12), s->elem(12), ((imm >> 12) & 1));\
+                d->elem(13) = F(d->elem(13), s->elem(13), ((imm >> 13) & 1));\
+                d->elem(14) = F(d->elem(14), s->elem(14), ((imm >> 14) & 1));\
+                d->elem(15) = F(d->elem(15), s->elem(15), ((imm >> 15) & 1));\
+            }\
+        }\
+    }\
+}
+
+/* SSE4.1 op helpers */
+#define FBLENDVB(d, s, m) (m & 0x80) ? s : d
+#define FBLENDVPS(d, s, m) (m & 0x80000000) ? s : d
+#define FBLENDVPD(d, s, m) (m & 0x8000000000000000) ? s : d
+SSE_HELPER_V(helper_pblendvb, B, 16, FBLENDVB)
+SSE_HELPER_V(helper_blendvps, L, 4, FBLENDVPS)
+SSE_HELPER_V(helper_blendvpd, Q, 2, FBLENDVPD)
+
+void glue(helper_ptest, SUFFIX) (Reg *d, Reg *s)
+{
+    uint64_t zf = (s->Q(0) &  d->Q(0)) | (s->Q(1) &  d->Q(1));
+    uint64_t cf = (s->Q(0) & ~d->Q(0)) | (s->Q(1) & ~d->Q(1));
+
+    CC_SRC = (zf ? 0 : CC_Z) | (cf ? 0 : CC_C);
+}
+
+#define SSE_HELPER_F(name, elem, num, F)\
+void glue(name, SUFFIX) (Reg *d, Reg *s)\
+{\
+    d->elem(0) = F(0);\
+    d->elem(1) = F(1);\
+    d->elem(2) = F(2);\
+    d->elem(3) = F(3);\
+    if (num > 3) {\
+        d->elem(4) = F(4);\
+        d->elem(5) = F(5);\
+        if (num > 5) {\
+            d->elem(6) = F(6);\
+            d->elem(7) = F(7);\
+        }\
+    }\
+}
+
+SSE_HELPER_F(helper_pmovsxbw, W, 8, (int8_t) s->B)
+SSE_HELPER_F(helper_pmovsxbd, L, 4, (int8_t) s->B)
+SSE_HELPER_F(helper_pmovsxbq, Q, 2, (int8_t) s->B)
+SSE_HELPER_F(helper_pmovsxwd, L, 4, (int16_t) s->W)
+SSE_HELPER_F(helper_pmovsxwq, Q, 2, (int16_t) s->W)
+SSE_HELPER_F(helper_pmovsxdq, Q, 2, (int32_t) s->L)
+SSE_HELPER_F(helper_pmovzxbw, W, 8, s->B)
+SSE_HELPER_F(helper_pmovzxbd, L, 4, s->B)
+SSE_HELPER_F(helper_pmovzxbq, Q, 2, s->B)
+SSE_HELPER_F(helper_pmovzxwd, L, 4, s->W)
+SSE_HELPER_F(helper_pmovzxwq, Q, 2, s->W)
+SSE_HELPER_F(helper_pmovzxdq, Q, 2, s->L)
+
+void glue(helper_pmuldq, SUFFIX) (Reg *d, Reg *s)
+{
+    d->Q(0) = (int64_t) (int32_t) d->L(0) * (int32_t) s->L(0);
+    d->Q(1) = (int64_t) (int32_t) d->L(2) * (int32_t) s->L(2);
+}
+
+#define FCMPEQQ(d, s) d == s ? -1 : 0
+SSE_HELPER_Q(helper_pcmpeqq, FCMPEQQ)
+
+void glue(helper_packusdw, SUFFIX) (Reg *d, Reg *s)
+{
+    d->W(0) = satuw((int32_t) d->L(0));
+    d->W(1) = satuw((int32_t) d->L(1));
+    d->W(2) = satuw((int32_t) d->L(2));
+    d->W(3) = satuw((int32_t) d->L(3));
+    d->W(4) = satuw((int32_t) s->L(0));
+    d->W(5) = satuw((int32_t) s->L(1));
+    d->W(6) = satuw((int32_t) s->L(2));
+    d->W(7) = satuw((int32_t) s->L(3));
+}
+
+#define FMINSB(d, s) MIN((int8_t) d, (int8_t) s)
+#define FMINSD(d, s) MIN((int32_t) d, (int32_t) s)
+#define FMAXSB(d, s) MAX((int8_t) d, (int8_t) s)
+#define FMAXSD(d, s) MAX((int32_t) d, (int32_t) s)
+SSE_HELPER_B(helper_pminsb, FMINSB)
+SSE_HELPER_L(helper_pminsd, FMINSD)
+SSE_HELPER_W(helper_pminuw, MIN)
+SSE_HELPER_L(helper_pminud, MIN)
+SSE_HELPER_B(helper_pmaxsb, FMAXSB)
+SSE_HELPER_L(helper_pmaxsd, FMAXSD)
+SSE_HELPER_W(helper_pmaxuw, MAX)
+SSE_HELPER_L(helper_pmaxud, MAX)
+
+#define FMULLD(d, s) (int32_t) d * (int32_t) s
+SSE_HELPER_L(helper_pmulld, FMULLD)
+
+void glue(helper_phminposuw, SUFFIX) (Reg *d, Reg *s)
+{
+    int idx = 0;
+
+    if (s->W(1) < s->W(idx))
+        idx = 1;
+    if (s->W(2) < s->W(idx))
+        idx = 2;
+    if (s->W(3) < s->W(idx))
+        idx = 3;
+    if (s->W(4) < s->W(idx))
+        idx = 4;
+    if (s->W(5) < s->W(idx))
+        idx = 5;
+    if (s->W(6) < s->W(idx))
+        idx = 6;
+    if (s->W(7) < s->W(idx))
+        idx = 7;
+
+    d->Q(1) = 0;
+    d->L(1) = 0;
+    d->W(1) = idx;
+    d->W(0) = s->W(idx);
+}
+
+void glue(helper_roundps, SUFFIX) (Reg *d, Reg *s, uint32_t mode)
+{
+    signed char prev_rounding_mode;
+
+    prev_rounding_mode = env->sse_status.float_rounding_mode;
+    if (!(mode & (1 << 2)))
+        switch (mode & 3) {
+        case 0:
+            set_float_rounding_mode(float_round_nearest_even, &env->sse_status);
+            break;
+        case 1:
+            set_float_rounding_mode(float_round_down, &env->sse_status);
+            break;
+        case 2:
+            set_float_rounding_mode(float_round_up, &env->sse_status);
+            break;
+        case 3:
+            set_float_rounding_mode(float_round_to_zero, &env->sse_status);
+            break;
+        }
+
+    d->L(0) = float64_round_to_int(s->L(0), &env->sse_status);
+    d->L(1) = float64_round_to_int(s->L(1), &env->sse_status);
+    d->L(2) = float64_round_to_int(s->L(2), &env->sse_status);
+    d->L(3) = float64_round_to_int(s->L(3), &env->sse_status);
+
+#if 0 /* TODO */
+    if (mode & (1 << 3))
+        set_float_exception_flags(
+                        get_float_exception_flags(&env->sse_status) &
+                        ~float_flag_inexact,
+                        &env->sse_status);
+#endif
+    env->sse_status.float_rounding_mode = prev_rounding_mode;
+}
+
+void glue(helper_roundpd, SUFFIX) (Reg *d, Reg *s, uint32_t mode)
+{
+    signed char prev_rounding_mode;
+
+    prev_rounding_mode = env->sse_status.float_rounding_mode;
+    if (!(mode & (1 << 2)))
+        switch (mode & 3) {
+        case 0:
+            set_float_rounding_mode(float_round_nearest_even, &env->sse_status);
+            break;
+        case 1:
+            set_float_rounding_mode(float_round_down, &env->sse_status);
+            break;
+        case 2:
+            set_float_rounding_mode(float_round_up, &env->sse_status);
+            break;
+        case 3:
+            set_float_rounding_mode(float_round_to_zero, &env->sse_status);
+            break;
+        }
+
+    d->Q(0) = float64_round_to_int(s->Q(0), &env->sse_status);
+    d->Q(1) = float64_round_to_int(s->Q(1), &env->sse_status);
+
+#if 0 /* TODO */
+    if (mode & (1 << 3))
+        set_float_exception_flags(
+                        get_float_exception_flags(&env->sse_status) &
+                        ~float_flag_inexact,
+                        &env->sse_status);
+#endif
+    env->sse_status.float_rounding_mode = prev_rounding_mode;
+}
+
+void glue(helper_roundss, SUFFIX) (Reg *d, Reg *s, uint32_t mode)
+{
+    signed char prev_rounding_mode;
+
+    prev_rounding_mode = env->sse_status.float_rounding_mode;
+    if (!(mode & (1 << 2)))
+        switch (mode & 3) {
+        case 0:
+            set_float_rounding_mode(float_round_nearest_even, &env->sse_status);
+            break;
+        case 1:
+            set_float_rounding_mode(float_round_down, &env->sse_status);
+            break;
+        case 2:
+            set_float_rounding_mode(float_round_up, &env->sse_status);
+            break;
+        case 3:
+            set_float_rounding_mode(float_round_to_zero, &env->sse_status);
+            break;
+        }
+
+    d->L(0) = float64_round_to_int(s->L(0), &env->sse_status);
+
+#if 0 /* TODO */
+    if (mode & (1 << 3))
+        set_float_exception_flags(
+                        get_float_exception_flags(&env->sse_status) &
+                        ~float_flag_inexact,
+                        &env->sse_status);
+#endif
+    env->sse_status.float_rounding_mode = prev_rounding_mode;
+}
+
+void glue(helper_roundsd, SUFFIX) (Reg *d, Reg *s, uint32_t mode)
+{
+    signed char prev_rounding_mode;
+
+    prev_rounding_mode = env->sse_status.float_rounding_mode;
+    if (!(mode & (1 << 2)))
+        switch (mode & 3) {
+        case 0:
+            set_float_rounding_mode(float_round_nearest_even, &env->sse_status);
+            break;
+        case 1:
+            set_float_rounding_mode(float_round_down, &env->sse_status);
+            break;
+        case 2:
+            set_float_rounding_mode(float_round_up, &env->sse_status);
+            break;
+        case 3:
+            set_float_rounding_mode(float_round_to_zero, &env->sse_status);
+            break;
+        }
+
+    d->Q(0) = float64_round_to_int(s->Q(0), &env->sse_status);
+
+#if 0 /* TODO */
+    if (mode & (1 << 3))
+        set_float_exception_flags(
+                        get_float_exception_flags(&env->sse_status) &
+                        ~float_flag_inexact,
+                        &env->sse_status);
+#endif
+    env->sse_status.float_rounding_mode = prev_rounding_mode;
+}
+
+#define FBLENDP(d, s, m) m ? s : d
+SSE_HELPER_I(helper_blendps, L, 4, FBLENDP)
+SSE_HELPER_I(helper_blendpd, Q, 2, FBLENDP)
+SSE_HELPER_I(helper_pblendw, W, 8, FBLENDP)
+
+void glue(helper_dpps, SUFFIX) (Reg *d, Reg *s, uint32_t mask)
+{
+    float32 iresult = 0 /*float32_zero*/;
+
+    if (mask & (1 << 4))
+        iresult = float32_add(iresult,
+                        float32_mul(d->L(0), s->L(0), &env->sse_status),
+                        &env->sse_status);
+    if (mask & (1 << 5))
+        iresult = float32_add(iresult,
+                        float32_mul(d->L(1), s->L(1), &env->sse_status),
+                        &env->sse_status);
+    if (mask & (1 << 6))
+        iresult = float32_add(iresult,
+                        float32_mul(d->L(2), s->L(2), &env->sse_status),
+                        &env->sse_status);
+    if (mask & (1 << 7))
+        iresult = float32_add(iresult,
+                        float32_mul(d->L(3), s->L(3), &env->sse_status),
+                        &env->sse_status);
+    d->L(0) = (mask & (1 << 0)) ? iresult : 0 /*float32_zero*/;
+    d->L(1) = (mask & (1 << 1)) ? iresult : 0 /*float32_zero*/;
+    d->L(2) = (mask & (1 << 2)) ? iresult : 0 /*float32_zero*/;
+    d->L(3) = (mask & (1 << 3)) ? iresult : 0 /*float32_zero*/;
+}
+
+void glue(helper_dppd, SUFFIX) (Reg *d, Reg *s, uint32_t mask)
+{
+    float64 iresult = 0 /*float64_zero*/;
+
+    if (mask & (1 << 4))
+        iresult = float64_add(iresult,
+                        float64_mul(d->Q(0), s->Q(0), &env->sse_status),
+                        &env->sse_status);
+    if (mask & (1 << 5))
+        iresult = float64_add(iresult,
+                        float64_mul(d->Q(1), s->Q(1), &env->sse_status),
+                        &env->sse_status);
+    d->Q(0) = (mask & (1 << 0)) ? iresult : 0 /*float64_zero*/;
+    d->Q(1) = (mask & (1 << 1)) ? iresult : 0 /*float64_zero*/;
+}
+
+void glue(helper_mpsadbw, SUFFIX) (Reg *d, Reg *s, uint32_t offset)
+{
+    int s0 = (offset & 3) << 2;
+    int d0 = (offset & 4) << 0;
+    int i;
+    Reg r;
+
+    for (i = 0; i < 8; i++, d0++) {
+        r.W(i) = 0;
+        r.W(i) += abs1(d->B(d0 + 0) - s->B(s0 + 0));
+        r.W(i) += abs1(d->B(d0 + 1) - s->B(s0 + 1));
+        r.W(i) += abs1(d->B(d0 + 2) - s->B(s0 + 2));
+        r.W(i) += abs1(d->B(d0 + 3) - s->B(s0 + 3));
+    }
+
+    *d = r;
+}
+
+/* SSE4.2 op helpers */
+/* it's unclear whether signed or unsigned */
+#define FCMPGTQ(d, s) d > s ? -1 : 0
+SSE_HELPER_Q(helper_pcmpgtq, FCMPGTQ)
+
+static inline int pcmp_elen(int reg, uint32_t ctrl)
+{
+    int val;
+
+    /* Presence of REX.W is indicated by a bit higher than 7 set */
+    if (ctrl >> 8)
+        val = abs1((int64_t) env->regs[reg]);
+    else
+        val = abs1((int32_t) env->regs[reg]);
+
+    if (ctrl & 1) {
+        if (val > 8)
+            return 8;
+    } else
+        if (val > 16)
+            return 16;
+
+    return val;
+}
+
+static inline int pcmp_ilen(Reg *r, uint8_t ctrl)
+{
+    int val = 0;
+
+    if (ctrl & 1) {
+        while (val < 8 && r->W(val))
+            val++;
+    } else
+        while (val < 16 && r->B(val))
+            val++;
+
+    return val;
+}
+
+static inline int pcmp_val(Reg *r, uint8_t ctrl, int i)
+{
+    switch ((ctrl >> 0) & 3) {
+    case 0:
+        return r->B(i);
+    case 1:
+        return r->W(i);
+    case 2:
+        return (int8_t) r->B(i);
+    case 3:
+    default:
+        return (int16_t) r->W(i);
+    }
+}
+
+static inline unsigned pcmpxstrx(Reg *d, Reg *s,
+                int8_t ctrl, int valids, int validd)
+{
+    unsigned int res = 0;
+    int v;
+    int j, i;
+    int upper = (ctrl & 1) ? 7 : 15;
+
+    valids--;
+    validd--;
+
+    CC_SRC = (valids < upper ? CC_Z : 0) | (validd < upper ? CC_S : 0);
+
+    switch ((ctrl >> 2) & 3) {
+    case 0:
+        for (j = valids; j >= 0; j--) {
+            res <<= 1;
+            v = pcmp_val(s, ctrl, j);
+            for (i = validd; i >= 0; i--)
+                res |= (v == pcmp_val(d, ctrl, i));
+        }
+        break;
+    case 1:
+        for (j = valids; j >= 0; j--) {
+            res <<= 1;
+            v = pcmp_val(s, ctrl, j);
+            for (i = ((validd - 1) | 1); i >= 0; i -= 2)
+                res |= (pcmp_val(d, ctrl, i - 0) <= v &&
+                        pcmp_val(d, ctrl, i - 1) >= v);
+        }
+        break;
+    case 2:
+        res = (2 << (upper - MAX(valids, validd))) - 1;
+        res <<= MAX(valids, validd) - MIN(valids, validd);
+        for (i = MIN(valids, validd); i >= 0; i--) {
+            res <<= 1;
+            v = pcmp_val(s, ctrl, i);
+            res |= (v == pcmp_val(d, ctrl, i));
+        }
+        break;
+    case 3:
+        for (j = valids - validd; j >= 0; j--) {
+            res <<= 1;
+            res |= 1;
+            for (i = MIN(upper - j, validd); i >= 0; i--)
+                res &= (pcmp_val(s, ctrl, i + j) == pcmp_val(d, ctrl, i));
+        }
+        break;
+    }
+
+    switch ((ctrl >> 4) & 3) {
+    case 1:
+        res ^= (2 << upper) - 1;
+        break;
+    case 3:
+        res ^= (2 << valids) - 1;
+        break;
+    }
+
+    if (res)
+       CC_SRC |= CC_C;
+    if (res & 1)
+       CC_SRC |= CC_O;
+
+    return res;
+}
+
+static inline int rffs1(unsigned int val)
+{
+    int ret = 1, hi;
+
+    for (hi = sizeof(val) * 4; hi; hi /= 2)
+        if (val >> hi) {
+            val >>= hi;
+            ret += hi;
+        }
+
+    return ret;
+}
+
+static inline int ffs1(unsigned int val)
+{
+    int ret = 1, hi;
+
+    for (hi = sizeof(val) * 4; hi; hi /= 2)
+        if (val << hi) {
+            val <<= hi;
+            ret += hi;
+        }
+
+    return ret;
+}
+
+void glue(helper_pcmpestri, SUFFIX) (Reg *d, Reg *s, uint32_t ctrl)
+{
+    unsigned int res = pcmpxstrx(d, s, ctrl,
+                    pcmp_elen(R_EDX, ctrl),
+                    pcmp_elen(R_EAX, ctrl));
+
+    if (res)
+        env->regs[R_ECX] = ((ctrl & (1 << 6)) ? rffs1 : ffs1)(res) - 1;
+    else
+        env->regs[R_ECX] = 16 >> (ctrl & (1 << 0));
+}
+
+void glue(helper_pcmpestrm, SUFFIX) (Reg *d, Reg *s, uint32_t ctrl)
+{
+    int i;
+    unsigned int res = pcmpxstrx(d, s, ctrl,
+                    pcmp_elen(R_EDX, ctrl),
+                    pcmp_elen(R_EAX, ctrl));
+
+    if ((ctrl >> 6) & 1) {
+        if (ctrl & 1)
+            for (i = 0; i <= 8; i--, res >>= 1)
+                d->W(i) = (res & 1) ? ~0 : 0;
+        else
+            for (i = 0; i <= 16; i--, res >>= 1)
+                d->B(i) = (res & 1) ? ~0 : 0;
+    } else {
+        d->Q(1) = 0;
+        d->Q(0) = res;
+    }
+}
+
+void glue(helper_pcmpistri, SUFFIX) (Reg *d, Reg *s, uint32_t ctrl)
+{
+    unsigned int res = pcmpxstrx(d, s, ctrl,
+                    pcmp_ilen(s, ctrl),
+                    pcmp_ilen(d, ctrl));
+
+    if (res)
+        env->regs[R_ECX] = ((ctrl & (1 << 6)) ? rffs1 : ffs1)(res) - 1;
+    else
+        env->regs[R_ECX] = 16 >> (ctrl & (1 << 0));
+}
+
+void glue(helper_pcmpistrm, SUFFIX) (Reg *d, Reg *s, uint32_t ctrl)
+{
+    int i;
+    unsigned int res = pcmpxstrx(d, s, ctrl,
+                    pcmp_ilen(s, ctrl),
+                    pcmp_ilen(d, ctrl));
+
+    if ((ctrl >> 6) & 1) {
+        if (ctrl & 1)
+            for (i = 0; i <= 8; i--, res >>= 1)
+                d->W(i) = (res & 1) ? ~0 : 0;
+        else
+            for (i = 0; i <= 16; i--, res >>= 1)
+                d->B(i) = (res & 1) ? ~0 : 0;
+    } else {
+        d->Q(1) = 0;
+        d->Q(0) = res;
+    }
+}
+
+#define CRCPOLY        0x1edc6f41
+#define CRCPOLY_BITREV 0x82f63b78
+target_ulong helper_crc32(uint32_t crc1, target_ulong msg, uint32_t len)
+{
+    target_ulong crc = (msg & ((target_ulong) -1 >>
+                            (TARGET_LONG_BITS - len))) ^ crc1;
+
+    while (len--)
+        crc = (crc >> 1) ^ ((crc & 1) ? CRCPOLY_BITREV : 0);
+
+    return crc;
+}
+
+#define POPMASK(i)     ((target_ulong) -1 / ((1LL << (1 << i)) + 1))
+#define POPCOUNT(n, i) (n & POPMASK(i)) + ((n >> (1 << i)) & POPMASK(i))
+target_ulong helper_popcnt(target_ulong n, uint32_t type)
+{
+    CC_SRC = n ? 0 : CC_Z;
+
+    n = POPCOUNT(n, 0);
+    n = POPCOUNT(n, 1);
+    n = POPCOUNT(n, 2);
+    n = POPCOUNT(n, 3);
+    if (type == 1)
+        return n & 0xff;
+
+    n = POPCOUNT(n, 4);
+#ifndef TARGET_X86_64
+    return n;
+#else
+    if (type == 2)
+        return n & 0xff;
+
+    return POPCOUNT(n, 5);
+#endif
+}
+#endif
+
 #undef SHIFT
 #undef XMM_ONLY
 #undef Reg
diff --git a/target-i386/ops_sse_header.h b/target-i386/ops_sse_header.h
index 22b77b7a79..9066eb01de 100644
--- a/target-i386/ops_sse_header.h
+++ b/target-i386/ops_sse_header.h
@@ -1,5 +1,5 @@
 /*
- *  MMX/3DNow!/SSE/SSE2/SSE3/SSSE3/PNI support
+ *  MMX/3DNow!/SSE/SSE2/SSE3/SSSE3/SSE4/PNI support
  *
  *  Copyright (c) 2005 Fabrice Bellard
  *
@@ -269,6 +269,61 @@ DEF_HELPER(void, glue(helper_psignw, SUFFIX), (Reg *d, Reg *s))
 DEF_HELPER(void, glue(helper_psignd, SUFFIX), (Reg *d, Reg *s))
 DEF_HELPER(void, glue(helper_palignr, SUFFIX), (Reg *d, Reg *s, int32_t shift))
 
+/* SSE4.1 op helpers */
+#if SHIFT == 1
+DEF_HELPER(void, glue(helper_pblendvb, SUFFIX), (Reg *d, Reg *s))
+DEF_HELPER(void, glue(helper_blendvps, SUFFIX), (Reg *d, Reg *s))
+DEF_HELPER(void, glue(helper_blendvpd, SUFFIX), (Reg *d, Reg *s))
+DEF_HELPER(void, glue(helper_ptest, SUFFIX), (Reg *d, Reg *s))
+DEF_HELPER(void, glue(helper_pmovsxbw, SUFFIX), (Reg *d, Reg *s))
+DEF_HELPER(void, glue(helper_pmovsxbd, SUFFIX), (Reg *d, Reg *s))
+DEF_HELPER(void, glue(helper_pmovsxbq, SUFFIX), (Reg *d, Reg *s))
+DEF_HELPER(void, glue(helper_pmovsxwd, SUFFIX), (Reg *d, Reg *s))
+DEF_HELPER(void, glue(helper_pmovsxwq, SUFFIX), (Reg *d, Reg *s))
+DEF_HELPER(void, glue(helper_pmovsxdq, SUFFIX), (Reg *d, Reg *s))
+DEF_HELPER(void, glue(helper_pmovzxbw, SUFFIX), (Reg *d, Reg *s))
+DEF_HELPER(void, glue(helper_pmovzxbd, SUFFIX), (Reg *d, Reg *s))
+DEF_HELPER(void, glue(helper_pmovzxbq, SUFFIX), (Reg *d, Reg *s))
+DEF_HELPER(void, glue(helper_pmovzxwd, SUFFIX), (Reg *d, Reg *s))
+DEF_HELPER(void, glue(helper_pmovzxwq, SUFFIX), (Reg *d, Reg *s))
+DEF_HELPER(void, glue(helper_pmovzxdq, SUFFIX), (Reg *d, Reg *s))
+DEF_HELPER(void, glue(helper_pmuldq, SUFFIX), (Reg *d, Reg *s))
+DEF_HELPER(void, glue(helper_pcmpeqq, SUFFIX), (Reg *d, Reg *s))
+DEF_HELPER(void, glue(helper_packusdw, SUFFIX), (Reg *d, Reg *s))
+DEF_HELPER(void, glue(helper_pminsb, SUFFIX), (Reg *d, Reg *s))
+DEF_HELPER(void, glue(helper_pminsd, SUFFIX), (Reg *d, Reg *s))
+DEF_HELPER(void, glue(helper_pminuw, SUFFIX), (Reg *d, Reg *s))
+DEF_HELPER(void, glue(helper_pminud, SUFFIX), (Reg *d, Reg *s))
+DEF_HELPER(void, glue(helper_pmaxsb, SUFFIX), (Reg *d, Reg *s))
+DEF_HELPER(void, glue(helper_pmaxsd, SUFFIX), (Reg *d, Reg *s))
+DEF_HELPER(void, glue(helper_pmaxuw, SUFFIX), (Reg *d, Reg *s))
+DEF_HELPER(void, glue(helper_pmaxud, SUFFIX), (Reg *d, Reg *s))
+DEF_HELPER(void, glue(helper_pmulld, SUFFIX), (Reg *d, Reg *s))
+DEF_HELPER(void, glue(helper_phminposuw, SUFFIX), (Reg *d, Reg *s))
+DEF_HELPER(void, glue(helper_roundps, SUFFIX), (Reg *d, Reg *s, uint32_t mode))
+DEF_HELPER(void, glue(helper_roundpd, SUFFIX), (Reg *d, Reg *s, uint32_t mode))
+DEF_HELPER(void, glue(helper_roundss, SUFFIX), (Reg *d, Reg *s, uint32_t mode))
+DEF_HELPER(void, glue(helper_roundsd, SUFFIX), (Reg *d, Reg *s, uint32_t mode))
+DEF_HELPER(void, glue(helper_blendps, SUFFIX), (Reg *d, Reg *s, uint32_t imm))
+DEF_HELPER(void, glue(helper_blendpd, SUFFIX), (Reg *d, Reg *s, uint32_t imm))
+DEF_HELPER(void, glue(helper_pblendw, SUFFIX), (Reg *d, Reg *s, uint32_t imm))
+DEF_HELPER(void, glue(helper_dpps, SUFFIX), (Reg *d, Reg *s, uint32_t mask))
+DEF_HELPER(void, glue(helper_dppd, SUFFIX), (Reg *d, Reg *s, uint32_t mask))
+DEF_HELPER(void, glue(helper_mpsadbw, SUFFIX), (Reg *d, Reg *s, uint32_t off))
+#endif
+
+/* SSE4.2 op helpers */
+#if SHIFT == 1
+DEF_HELPER(void, glue(helper_pcmpgtq, SUFFIX), (Reg *d, Reg *s))
+DEF_HELPER(void, glue(helper_pcmpestri, SUFFIX), (Reg *d, Reg *s, uint32_t ctl))
+DEF_HELPER(void, glue(helper_pcmpestrm, SUFFIX), (Reg *d, Reg *s, uint32_t ctl))
+DEF_HELPER(void, glue(helper_pcmpistri, SUFFIX), (Reg *d, Reg *s, uint32_t ctl))
+DEF_HELPER(void, glue(helper_pcmpistrm, SUFFIX), (Reg *d, Reg *s, uint32_t ctl))
+DEF_HELPER(target_ulong, helper_crc32,
+                (uint32_t crc1, target_ulong msg, uint32_t len))
+DEF_HELPER(target_ulong, helper_popcnt, (target_ulong n, uint32_t type))
+#endif
+
 #undef SHIFT
 #undef Reg
 #undef SUFFIX
diff --git a/target-i386/translate.c b/target-i386/translate.c
index 88d0173487..4d1b0c2c7d 100644
--- a/target-i386/translate.c
+++ b/target-i386/translate.c
@@ -2140,7 +2140,7 @@ static void gen_add_A0_ds_seg(DisasContext *s)
     }
 }
 
-/* generate modrm memory load or store of 'reg'. TMP0 is used if reg !=
+/* generate modrm memory load or store of 'reg'. TMP0 is used if reg ==
    OR_TMP0 */
 static void gen_ldst_modrm(DisasContext *s, int modrm, int ot, int reg, int is_store)
 {
@@ -2770,8 +2770,8 @@ static void *sse_op_table1[256][4] = {
     [0xc2] = SSE_FOP(cmpeq),
     [0xc6] = { helper_shufps, helper_shufpd },
 
-    [0x38] = { SSE_SPECIAL, SSE_SPECIAL },  /* SSSE3 */
-    [0x3a] = { SSE_SPECIAL, SSE_SPECIAL },  /* SSSE3 */
+    [0x38] = { SSE_SPECIAL, SSE_SPECIAL, NULL, SSE_SPECIAL }, /* SSSE3/SSE4 */
+    [0x3a] = { SSE_SPECIAL, SSE_SPECIAL }, /* SSSE3/SSE4 */
 
     /* MMX ops and their SSE extensions */
     [0x60] = MMX_OP2(punpcklbw),
@@ -2924,26 +2924,85 @@ static void *sse_op_table5[256] = {
     [0xbf] = helper_pavgb_mmx /* pavgusb */
 };
 
-static void *sse_op_table6[256][2] = {
-    [0x00] = MMX_OP2(pshufb),
-    [0x01] = MMX_OP2(phaddw),
-    [0x02] = MMX_OP2(phaddd),
-    [0x03] = MMX_OP2(phaddsw),
-    [0x04] = MMX_OP2(pmaddubsw),
-    [0x05] = MMX_OP2(phsubw),
-    [0x06] = MMX_OP2(phsubd),
-    [0x07] = MMX_OP2(phsubsw),
-    [0x08] = MMX_OP2(psignb),
-    [0x09] = MMX_OP2(psignw),
-    [0x0a] = MMX_OP2(psignd),
-    [0x0b] = MMX_OP2(pmulhrsw),
-    [0x1c] = MMX_OP2(pabsb),
-    [0x1d] = MMX_OP2(pabsw),
-    [0x1e] = MMX_OP2(pabsd),
+struct sse_op_helper_s {
+    void *op[2]; uint32_t ext_mask;
+};
+#define SSSE3_OP(x) { MMX_OP2(x), CPUID_EXT_SSSE3 }
+#define SSE41_OP(x) { { NULL, helper_ ## x ## _xmm }, CPUID_EXT_SSE41 }
+#define SSE42_OP(x) { { NULL, helper_ ## x ## _xmm }, CPUID_EXT_SSE42 }
+#define SSE41_SPECIAL { { NULL, SSE_SPECIAL }, CPUID_EXT_SSE41 }
+static struct sse_op_helper_s sse_op_table6[256] = {
+    [0x00] = SSSE3_OP(pshufb),
+    [0x01] = SSSE3_OP(phaddw),
+    [0x02] = SSSE3_OP(phaddd),
+    [0x03] = SSSE3_OP(phaddsw),
+    [0x04] = SSSE3_OP(pmaddubsw),
+    [0x05] = SSSE3_OP(phsubw),
+    [0x06] = SSSE3_OP(phsubd),
+    [0x07] = SSSE3_OP(phsubsw),
+    [0x08] = SSSE3_OP(psignb),
+    [0x09] = SSSE3_OP(psignw),
+    [0x0a] = SSSE3_OP(psignd),
+    [0x0b] = SSSE3_OP(pmulhrsw),
+    [0x10] = SSE41_OP(pblendvb),
+    [0x14] = SSE41_OP(blendvps),
+    [0x15] = SSE41_OP(blendvpd),
+    [0x17] = SSE41_OP(ptest),
+    [0x1c] = SSSE3_OP(pabsb),
+    [0x1d] = SSSE3_OP(pabsw),
+    [0x1e] = SSSE3_OP(pabsd),
+    [0x20] = SSE41_OP(pmovsxbw),
+    [0x21] = SSE41_OP(pmovsxbd),
+    [0x22] = SSE41_OP(pmovsxbq),
+    [0x23] = SSE41_OP(pmovsxwd),
+    [0x24] = SSE41_OP(pmovsxwq),
+    [0x25] = SSE41_OP(pmovsxdq),
+    [0x28] = SSE41_OP(pmuldq),
+    [0x29] = SSE41_OP(pcmpeqq),
+    [0x2a] = SSE41_SPECIAL, /* movntqda */
+    [0x2b] = SSE41_OP(packusdw),
+    [0x30] = SSE41_OP(pmovzxbw),
+    [0x31] = SSE41_OP(pmovzxbd),
+    [0x32] = SSE41_OP(pmovzxbq),
+    [0x33] = SSE41_OP(pmovzxwd),
+    [0x34] = SSE41_OP(pmovzxwq),
+    [0x35] = SSE41_OP(pmovzxdq),
+    [0x37] = SSE42_OP(pcmpgtq),
+    [0x38] = SSE41_OP(pminsb),
+    [0x39] = SSE41_OP(pminsd),
+    [0x3a] = SSE41_OP(pminuw),
+    [0x3b] = SSE41_OP(pminud),
+    [0x3c] = SSE41_OP(pmaxsb),
+    [0x3d] = SSE41_OP(pmaxsd),
+    [0x3e] = SSE41_OP(pmaxuw),
+    [0x3f] = SSE41_OP(pmaxud),
+    [0x40] = SSE41_OP(pmulld),
+    [0x41] = SSE41_OP(phminposuw),
 };
 
-static void *sse_op_table7[256][2] = {
-    [0x0f] = MMX_OP2(palignr),
+static struct sse_op_helper_s sse_op_table7[256] = {
+    [0x08] = SSE41_OP(roundps),
+    [0x09] = SSE41_OP(roundpd),
+    [0x0a] = SSE41_OP(roundss),
+    [0x0b] = SSE41_OP(roundsd),
+    [0x0c] = SSE41_OP(blendps),
+    [0x0d] = SSE41_OP(blendpd),
+    [0x0e] = SSE41_OP(pblendw),
+    [0x0f] = SSSE3_OP(palignr),
+    [0x14] = SSE41_SPECIAL, /* pextrb */
+    [0x15] = SSE41_SPECIAL, /* pextrw */
+    [0x16] = SSE41_SPECIAL, /* pextrd/pextrq */
+    [0x17] = SSE41_SPECIAL, /* extractps */
+    [0x20] = SSE41_SPECIAL, /* pinsrb */
+    [0x21] = SSE41_SPECIAL, /* insertps */
+    [0x22] = SSE41_SPECIAL, /* pinsrd/pinsrq */
+    [0x40] = SSE41_OP(dpps),
+    [0x41] = SSE41_OP(dppd),
+    [0x42] = SSE41_OP(mpsadbw),
+    [0x60] = SSE42_OP(pcmpestrm),
+    [0x61] = SSE42_OP(pcmpestri),
+    [0x62] = SSE42_OP(pcmpistrm),
+    [0x63] = SSE42_OP(pcmpistri),
 };
 
 static void gen_sse(DisasContext *s, int b, target_ulong pc_start, int rex_r)
@@ -3511,18 +3570,20 @@ static void gen_sse(DisasContext *s, int b, target_ulong pc_start, int rex_r)
             break;
         case 0x038:
         case 0x138:
-            if (!(s->cpuid_ext_features & CPUID_EXT_SSSE3))
-                goto illegal_op;
-
             b = modrm;
             modrm = ldub_code(s->pc++);
             rm = modrm & 7;
             reg = ((modrm >> 3) & 7) | rex_r;
             mod = (modrm >> 6) & 3;
 
-            sse_op2 = sse_op_table6[b][b1];
+            if (s->prefix & PREFIX_REPNZ)
+                goto crc32;
+
+            sse_op2 = sse_op_table6[b].op[b1];
             if (!sse_op2)
                 goto illegal_op;
+            if (!(s->cpuid_ext_features & sse_op_table6[b].ext_mask))
+                goto illegal_op;
 
             if (b1) {
                 op1_offset = offsetof(CPUX86State,xmm_regs[reg]);
@@ -3531,7 +3592,32 @@ static void gen_sse(DisasContext *s, int b, target_ulong pc_start, int rex_r)
                 } else {
                     op2_offset = offsetof(CPUX86State,xmm_t0);
                     gen_lea_modrm(s, modrm, &reg_addr, &offset_addr);
-                    gen_ldo_env_A0(s->mem_index, op2_offset);
+                    switch (b) {
+                    case 0x20: case 0x30: /* pmovsxbw, pmovzxbw */
+                    case 0x23: case 0x33: /* pmovsxwd, pmovzxwd */
+                    case 0x25: case 0x35: /* pmovsxdq, pmovzxdq */
+                        gen_ldq_env_A0(s->mem_index, op2_offset +
+                                        offsetof(XMMReg, XMM_Q(0)));
+                        break;
+                    case 0x21: case 0x31: /* pmovsxbd, pmovzxbd */
+                    case 0x24: case 0x34: /* pmovsxwq, pmovzxwq */
+                        tcg_gen_qemu_ld32u(cpu_tmp2_i32, cpu_A0,
+                                          (s->mem_index >> 2) - 1);
+                        tcg_gen_st_i32(cpu_tmp2_i32, cpu_env, op2_offset +
+                                        offsetof(XMMReg, XMM_L(0)));
+                        break;
+                    case 0x22: case 0x32: /* pmovsxbq, pmovzxbq */
+                        tcg_gen_qemu_ld16u(cpu_tmp0, cpu_A0,
+                                          (s->mem_index >> 2) - 1);
+                        tcg_gen_st16_tl(cpu_tmp0, cpu_env, op2_offset +
+                                        offsetof(XMMReg, XMM_W(0)));
+                        break;
+                    case 0x2a:            /* movntqda */
+                        gen_ldo_env_A0(s->mem_index, op1_offset);
+                        return;
+                    default:
+                        gen_ldo_env_A0(s->mem_index, op2_offset);
+                    }
                 }
             } else {
                 op1_offset = offsetof(CPUX86State,fpregs[reg].mmx);
@@ -3543,24 +3629,177 @@ static void gen_sse(DisasContext *s, int b, target_ulong pc_start, int rex_r)
                     gen_ldq_env_A0(s->mem_index, op2_offset);
                 }
             }
+            if (sse_op2 == SSE_SPECIAL)
+                goto illegal_op;
+
             tcg_gen_addi_ptr(cpu_ptr0, cpu_env, op1_offset);
             tcg_gen_addi_ptr(cpu_ptr1, cpu_env, op2_offset);
             tcg_gen_helper_0_2(sse_op2, cpu_ptr0, cpu_ptr1);
+
+            if (b == 0x17)
+                s->cc_op = CC_OP_EFLAGS;
             break;
-        case 0x03a:
-        case 0x13a:
-            if (!(s->cpuid_ext_features & CPUID_EXT_SSSE3))
+        case 0x338: /* crc32 */
+        crc32:
+            b = modrm;
+            modrm = ldub_code(s->pc++);
+            reg = ((modrm >> 3) & 7) | rex_r;
+
+            if (b != 0xf0 && b != 0xf1)
+                goto illegal_op;
+            if (!(s->cpuid_ext_features & CPUID_EXT_SSE42))
                 goto illegal_op;
 
+            if (b == 0xf0)
+                ot = OT_BYTE;
+            else if (b == 0xf1 && s->dflag != 2)
+                if (s->prefix & PREFIX_DATA)
+                    ot = OT_WORD;
+                else
+                    ot = OT_LONG;
+            else
+                ot = OT_QUAD;
+
+            gen_op_mov_TN_reg(OT_LONG, 0, reg);
+            tcg_gen_trunc_tl_i32(cpu_tmp2_i32, cpu_T[0]);
+            gen_ldst_modrm(s, modrm, ot, OR_TMP0, 0);
+            tcg_gen_helper_1_3(helper_crc32, cpu_T[0], cpu_tmp2_i32,
+                            cpu_T[0], tcg_const_i32(8 << ot));
+
+            ot = (s->dflag == 2) ? OT_QUAD : OT_LONG;
+            gen_op_mov_reg_T0(ot, reg);
+            break;
+        case 0x03a:
+        case 0x13a:
             b = modrm;
             modrm = ldub_code(s->pc++);
             rm = modrm & 7;
             reg = ((modrm >> 3) & 7) | rex_r;
             mod = (modrm >> 6) & 3;
 
-            sse_op2 = sse_op_table7[b][b1];
+            sse_op2 = sse_op_table7[b].op[b1];
             if (!sse_op2)
                 goto illegal_op;
+            if (!(s->cpuid_ext_features & sse_op_table7[b].ext_mask))
+                goto illegal_op;
+
+            if (sse_op2 == SSE_SPECIAL) {
+                ot = (s->dflag == 2) ? OT_QUAD : OT_LONG;
+                rm = (modrm & 7) | REX_B(s);
+                if (mod != 3)
+                    gen_lea_modrm(s, modrm, &reg_addr, &offset_addr);
+                reg = ((modrm >> 3) & 7) | rex_r;
+                val = ldub_code(s->pc++);
+                switch (b) {
+                case 0x14: /* pextrb */
+                    tcg_gen_ld8u_tl(cpu_T[0], cpu_env, offsetof(CPUX86State,
+                                            xmm_regs[reg].XMM_B(val & 15)));
+                    if (mod == 3)
+                        gen_op_mov_reg_T0(ot, rm);
+                    else
+                        tcg_gen_qemu_st8(cpu_T[0], cpu_A0,
+                                        (s->mem_index >> 2) - 1);
+                    break;
+                case 0x15: /* pextrw */
+                    tcg_gen_ld16u_tl(cpu_T[0], cpu_env, offsetof(CPUX86State,
+                                            xmm_regs[reg].XMM_W(val & 7)));
+                    if (mod == 3)
+                        gen_op_mov_reg_T0(ot, rm);
+                    else
+                        tcg_gen_qemu_st16(cpu_T[0], cpu_A0,
+                                        (s->mem_index >> 2) - 1);
+                    break;
+                case 0x16:
+                    if (ot == OT_LONG) { /* pextrd */
+                        tcg_gen_ld_i32(cpu_tmp2_i32, cpu_env,
+                                        offsetof(CPUX86State,
+                                                xmm_regs[reg].XMM_L(val & 3)));
+                        if (mod == 3)
+                            gen_op_mov_reg_v(ot, rm, cpu_tmp2_i32);
+                        else
+                            tcg_gen_qemu_st32(cpu_tmp2_i32, cpu_A0,
+                                            (s->mem_index >> 2) - 1);
+                    } else { /* pextrq */
+                        tcg_gen_ld_i64(cpu_tmp1_i64, cpu_env,
+                                        offsetof(CPUX86State,
+                                                xmm_regs[reg].XMM_Q(val & 1)));
+                        if (mod == 3)
+                            gen_op_mov_reg_v(ot, rm, cpu_tmp1_i64);
+                        else
+                            tcg_gen_qemu_st64(cpu_tmp1_i64, cpu_A0,
+                                            (s->mem_index >> 2) - 1);
+                    }
+                    break;
+                case 0x17: /* extractps */
+                    tcg_gen_ld32u_tl(cpu_T[0], cpu_env, offsetof(CPUX86State,
+                                            xmm_regs[reg].XMM_L(val & 3)));
+                    if (mod == 3)
+                        gen_op_mov_reg_T0(ot, rm);
+                    else
+                        tcg_gen_qemu_st32(cpu_T[0], cpu_A0,
+                                        (s->mem_index >> 2) - 1);
+                    break;
+                case 0x20: /* pinsrb */
+                    if (mod == 3)
+                        gen_op_mov_TN_reg(OT_LONG, 0, rm);
+                    else
+                        tcg_gen_qemu_ld8u(cpu_T[0], cpu_A0,
+                                        (s->mem_index >> 2) - 1);
+                    tcg_gen_st8_tl(cpu_T[0], cpu_env, offsetof(CPUX86State,
+                                            xmm_regs[reg].XMM_B(val & 15)));
+                    break;
+                case 0x21: /* insertps */
+                    if (mod == 3)
+                        tcg_gen_ld_i32(cpu_tmp2_i32, cpu_env,
+                                        offsetof(CPUX86State,xmm_regs[rm]
+                                                .XMM_L((val >> 6) & 3)));
+                    else
+                        tcg_gen_qemu_ld32u(cpu_tmp2_i32, cpu_A0,
+                                        (s->mem_index >> 2) - 1);
+                    tcg_gen_st_i32(cpu_tmp2_i32, cpu_env,
+                                    offsetof(CPUX86State,xmm_regs[reg]
+                                            .XMM_L((val >> 4) & 3)));
+                    if ((val >> 0) & 1)
+                        tcg_gen_st_i32(tcg_const_i32(0 /*float32_zero*/),
+                                        cpu_env, offsetof(CPUX86State,
+                                                xmm_regs[reg].XMM_L(0)));
+                    if ((val >> 1) & 1)
+                        tcg_gen_st_i32(tcg_const_i32(0 /*float32_zero*/),
+                                        cpu_env, offsetof(CPUX86State,
+                                                xmm_regs[reg].XMM_L(1)));
+                    if ((val >> 2) & 1)
+                        tcg_gen_st_i32(tcg_const_i32(0 /*float32_zero*/),
+                                        cpu_env, offsetof(CPUX86State,
+                                                xmm_regs[reg].XMM_L(2)));
+                    if ((val >> 3) & 1)
+                        tcg_gen_st_i32(tcg_const_i32(0 /*float32_zero*/),
+                                        cpu_env, offsetof(CPUX86State,
+                                                xmm_regs[reg].XMM_L(3)));
+                    break;
+                case 0x22:
+                    if (ot == OT_LONG) { /* pinsrd */
+                        if (mod == 3)
+                            gen_op_mov_v_reg(ot, cpu_tmp2_i32, rm);
+                        else
+                            tcg_gen_qemu_ld32u(cpu_tmp2_i32, cpu_A0,
+                                            (s->mem_index >> 2) - 1);
+                        tcg_gen_st_i32(cpu_tmp2_i32, cpu_env,
+                                        offsetof(CPUX86State,
+                                                xmm_regs[reg].XMM_L(val & 3)));
+                    } else { /* pinsrq */
+                        if (mod == 3)
+                            gen_op_mov_v_reg(ot, cpu_tmp1_i64, rm);
+                        else
+                            tcg_gen_qemu_ld64(cpu_tmp1_i64, cpu_A0,
+                                            (s->mem_index >> 2) - 1);
+                        tcg_gen_st_i64(cpu_tmp1_i64, cpu_env,
+                                        offsetof(CPUX86State,
+                                                xmm_regs[reg].XMM_Q(val & 1)));
+                    }
+                    break;
+                }
+                return;
+            }
 
             if (b1) {
                 op1_offset = offsetof(CPUX86State,xmm_regs[reg]);
@@ -3583,6 +3822,14 @@ static void gen_sse(DisasContext *s, int b, target_ulong pc_start, int rex_r)
             }
             val = ldub_code(s->pc++);
 
+            if ((b & 0xfc) == 0x60) { /* pcmpXstrX */
+                s->cc_op = CC_OP_EFLAGS;
+
+                if (s->dflag == 2)
+                    /* The helper must use entire 64-bit gp registers */
+                    val |= 1 << 8;
+            }
+
             tcg_gen_addi_ptr(cpu_ptr0, cpu_env, op1_offset);
             tcg_gen_addi_ptr(cpu_ptr1, cpu_env, op2_offset);
             tcg_gen_helper_0_3(sse_op2, cpu_ptr0, cpu_ptr1, tcg_const_i32(val));
@@ -7094,7 +7341,7 @@ static target_ulong disas_insn(DisasContext *s, target_ulong pc_start)
             gen_eob(s);
         }
         break;
-    /* MMX/3DNow!/SSE/SSE2/SSE3/SSSE3 support */
+    /* MMX/3DNow!/SSE/SSE2/SSE3/SSSE3/SSE4 support */
     case 0x1c3: /* MOVNTI reg, mem */
         if (!(s->cpuid_features & CPUID_SSE2))
             goto illegal_op;
@@ -7202,6 +7449,28 @@ static target_ulong disas_insn(DisasContext *s, target_ulong pc_start)
         tcg_gen_helper_0_0(helper_rsm);
         gen_eob(s);
         break;
+    case 0x1b8: /* SSE4.2 popcnt */
+        if ((prefixes & (PREFIX_REPZ | PREFIX_LOCK | PREFIX_REPNZ)) !=
+             PREFIX_REPZ)
+            goto illegal_op;
+        if (!(s->cpuid_ext_features & CPUID_EXT_POPCNT))
+            goto illegal_op;
+
+        modrm = ldub_code(s->pc++);
+        reg = ((modrm >> 3) & 7);
+
+        if (s->prefix & PREFIX_DATA)
+            ot = OT_WORD;
+        else if (s->dflag != 2)
+            ot = OT_LONG;
+        else
+            ot = OT_QUAD;
+
+        gen_ldst_modrm(s, modrm, ot, OR_TMP0, 0);
+        tcg_gen_helper_1_2(helper_popcnt,
+                cpu_T[0], cpu_T[0], tcg_const_i32(ot));
+        gen_op_mov_reg_T0(ot, reg);
+        break;
     case 0x10e ... 0x10f:
         /* 3DNow! instructions, ignore prefixes */
         s->prefix &= ~(PREFIX_REPZ | PREFIX_REPNZ | PREFIX_DATA);