cmd/asm: add most SSE4 missing instructions

Instructions added:
  INSERTPS immb, r/m, xmm
  MPSADBW immb, r/m, xmm
  BLENDPD immb, r/m, xmm
  BLENDPS immb, r/m, xmm
  DPPD immb, r/m, xmm
  DPPS immb, r/m, xmm
  MOVNTDQA r/m, xmm
  PACKUSDW r/m, xmm
  PBLENDW immb, r/m, xmm
  PCMPEQQ r/m, xmm
  PCMPGTQ r/m, xmm
  PCMPISTRI immb, r/m, xmm
  PCMPISTRM immb, r/m, xmm
  PMAXSB r/m, xmm
  PMAXSD r/m, xmm
  PMAXUD r/m, xmm
  PMAXUW r/m, xmm
  PMINSB r/m, xmm
  PMINSD r/m, xmm
  PMINUD r/m, xmm
  PMINUW r/m, xmm
  PTEST r/m, xmm
  PCMPESTRM immb, r/m, xmm

Note: only 'optab' table is extended.

`EXTRACTPS immb, xmm, r/m` is not included in this
change due to new ytab set 'yextractps'. This should simplify
code review.

4-operand instructions are a subject of upcoming changes that
make 4-th (and so on) operands explicit.
Related TODO note in asm6.go:
"dont't hide 4op, some version have xmm version".

Part of the mission to add missing amd64 SSE4 instructions to Go asm.

Change-Id: I71716df14a8a5332e866dd0f0d52d43d7714872f
Reviewed-on: https://go-review.googlesource.com/57470
Run-TryBot: Iskander Sharipov <iskander.sharipov@intel.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ilya Tocar <ilya.tocar@intel.com>

src/cmd/internal/obj/x86/a.out.go

@@ -116,6 +116,7 @@ const (
 	AINCW
 	AINSB
 	AINSL
+	AINSERTPS
 	AINSW
 	AINT
 	AINTO
@@ -171,6 +172,7 @@ const (
 	AMOVSB
 	AMOVSL
 	AMOVSW
+	AMPSADBW
 	AMULB
 	AMULL
 	AMULW
@@ -530,6 +532,8 @@ const (
 	AANDPS
 	ABEXTRL
 	ABEXTRQ
+	ABLENDPD
+	ABLENDPS
 	ABLSIL
 	ABLSIQ
 	ABLSMSKL
@@ -570,6 +574,8 @@ const (
 	ADIVPS
 	ADIVSD
 	ADIVSS
+	ADPPD
+	ADPPS
 	AEMMS
 	AFXRSTOR
 	AFXRSTOR64
@@ -599,6 +605,7 @@ const (
 	AMOVMSKPD
 	AMOVMSKPS
 	AMOVNTO
+	AMOVNTDQA
 	AMOVNTPD
 	AMOVNTPS
 	AMOVNTQ
@@ -618,6 +625,7 @@ const (
 	AORPS
 	APACKSSLW
 	APACKSSWB
+	APACKUSDW
 	APACKUSWB
 	APADDB
 	APADDL
@@ -632,12 +640,17 @@ const (
 	APANDN
 	APAVGB
 	APAVGW
+	APBLENDW
 	APCMPEQB
 	APCMPEQL
+	APCMPEQQ
 	APCMPEQW
 	APCMPGTB
 	APCMPGTL
+	APCMPGTQ
 	APCMPGTW
+	APCMPISTRI
+	APCMPISTRM
 	APDEPL
 	APDEPQ
 	APEXTL
@@ -659,10 +672,18 @@ const (
 	APINSRW
 	APMADDUBSW
 	APMADDWL
+	APMAXSB
+	APMAXSD
 	APMAXSW
 	APMAXUB
+	APMAXUD
+	APMAXUW
+	APMINSB
+	APMINSD
 	APMINSW
 	APMINUB
+	APMINUD
+	APMINUW
 	APMOVMSKB
 	APMOVSXBD
 	APMOVSXBQ
@@ -711,6 +732,7 @@ const (
 	APSUBUSB
 	APSUBUSW
 	APSUBW
+	APTEST
 	APUNPCKHBW
 	APUNPCKHLQ
 	APUNPCKHQDQ
@@ -750,6 +772,7 @@ const (
 	AXORPD
 	AXORPS
 	APCMPESTRI
+	APCMPESTRM
 
 	ARETFW
 	ARETFL

src/cmd/internal/obj/x86/anames.go

@@ -78,6 +78,7 @@ var Anames = []string{
 	"INCW",
 	"INSB",
 	"INSL",
+	"INSERTPS",
 	"INSW",
 	"INT",
 	"INTO",
@@ -133,6 +134,7 @@ var Anames = []string{
 	"MOVSB",
 	"MOVSL",
 	"MOVSW",
+	"MPSADBW",
 	"MULB",
 	"MULL",
 	"MULW",
@@ -473,6 +475,8 @@ var Anames = []string{
 	"ANDPS",
 	"BEXTRL",
 	"BEXTRQ",
+	"BLENDPD",
+	"BLENDPS",
 	"BLSIL",
 	"BLSIQ",
 	"BLSMSKL",
@@ -513,6 +517,8 @@ var Anames = []string{
 	"DIVPS",
 	"DIVSD",
 	"DIVSS",
+	"DPPD",
+	"DPPS",
 	"EMMS",
 	"FXRSTOR",
 	"FXRSTOR64",
@@ -542,6 +548,7 @@ var Anames = []string{
 	"MOVMSKPD",
 	"MOVMSKPS",
 	"MOVNTO",
+	"MOVNTDQA",
 	"MOVNTPD",
 	"MOVNTPS",
 	"MOVNTQ",
@@ -561,6 +568,7 @@ var Anames = []string{
 	"ORPS",
 	"PACKSSLW",
 	"PACKSSWB",
+	"PACKUSDW",
 	"PACKUSWB",
 	"PADDB",
 	"PADDL",
@@ -575,12 +583,17 @@ var Anames = []string{
 	"PANDN",
 	"PAVGB",
 	"PAVGW",
+	"PBLENDW",
 	"PCMPEQB",
 	"PCMPEQL",
+	"PCMPEQQ",
 	"PCMPEQW",
 	"PCMPGTB",
 	"PCMPGTL",
+	"PCMPGTQ",
 	"PCMPGTW",
+	"PCMPISTRI",
+	"PCMPISTRM",
 	"PDEPL",
 	"PDEPQ",
 	"PEXTL",
@@ -602,10 +615,18 @@ var Anames = []string{
 	"PINSRW",
 	"PMADDUBSW",
 	"PMADDWL",
+	"PMAXSB",
+	"PMAXSD",
 	"PMAXSW",
 	"PMAXUB",
+	"PMAXUD",
+	"PMAXUW",
+	"PMINSB",
+	"PMINSD",
 	"PMINSW",
 	"PMINUB",
+	"PMINUD",
+	"PMINUW",
 	"PMOVMSKB",
 	"PMOVSXBD",
 	"PMOVSXBQ",
@@ -654,6 +675,7 @@ var Anames = []string{
 	"PSUBUSB",
 	"PSUBUSW",
 	"PSUBW",
+	"PTEST",
 	"PUNPCKHBW",
 	"PUNPCKHLQ",
 	"PUNPCKHQDQ",
@@ -693,6 +715,7 @@ var Anames = []string{
 	"XORPD",
 	"XORPS",
 	"PCMPESTRI",
+	"PCMPESTRM",
 	"RETFW",
 	"RETFL",
 	"RETFQ",

src/cmd/internal/obj/x86/asm6.go

@@ -733,6 +733,11 @@ var ymshufb = []ytab{
 	{Yxm, Ynone, Yxr, Zm2_r, 2},
 }
 
+// It should never have more than 1 entry,
+// because some optab entries you opcode secuences that
+// are longer than 2 bytes (zoffset=2 here),
+// ROUNDPD and ROUNDPS and recently added BLENDPD,
+// to name a few.
 var yxshuf = []ytab{
 	{Yu8, Yxm, Yxr, Zibm_r, 2},
 }
@@ -1107,6 +1112,8 @@ var optab =
 	{ADIVSD, yxm, Pf2, [23]uint8{0x5e}},
 	{ADIVSS, yxm, Pf3, [23]uint8{0x5e}},
 	{ADIVW, ydivl, Pe, [23]uint8{0xf7, 06}},
+	{ADPPD, yxshuf, Pq, [23]uint8{0x3a, 0x41, 0}},
+	{ADPPS, yxshuf, Pq, [23]uint8{0x3a, 0x40, 0}},
 	{AEMMS, ynone, Pm, [23]uint8{0x77}},
 	{AENTER, nil, 0, [23]uint8{}}, /* botch */
 	{AFXRSTOR, ysvrs, Pm, [23]uint8{0xae, 01, 0xae, 01}},
@@ -1131,6 +1138,7 @@ var optab =
 	{AINL, yin, Px, [23]uint8{0xe5, 0xed}},
 	{AINSB, ynone, Pb, [23]uint8{0x6c}},
 	{AINSL, ynone, Px, [23]uint8{0x6d}},
+	{AINSERTPS, yxshuf, Pq, [23]uint8{0x3a, 0x21, 0}},
 	{AINSW, ynone, Pe, [23]uint8{0x6d}},
 	{AINT, yint, Px, [23]uint8{0xcd}},
 	{AINTO, ynone, P32, [23]uint8{0xce}},
@@ -1217,6 +1225,7 @@ var optab =
 	{AMOVMSKPD, yxrrl, Pq, [23]uint8{0x50}},
 	{AMOVMSKPS, yxrrl, Pm, [23]uint8{0x50}},
 	{AMOVNTO, yxr_ml, Pe, [23]uint8{0xe7}},
+	{AMOVNTDQA, ylddqu, Pq4, [23]uint8{0x2a}},
 	{AMOVNTPD, yxr_ml, Pe, [23]uint8{0x2b}},
 	{AMOVNTPS, yxr_ml, Pm, [23]uint8{0x2b}},
 	{AMOVNTQ, ymr_ml, Pm, [23]uint8{0xe7}},
@@ -1235,6 +1244,7 @@ var optab =
 	{AMOVWLZX, yml_rl, Pm, [23]uint8{0xb7}},
 	{AMOVWQSX, yml_rl, Pw, [23]uint8{0x0f, 0xbf}},
 	{AMOVWQZX, yml_rl, Pw, [23]uint8{0x0f, 0xb7}},
+	{AMPSADBW, yxshuf, Pq, [23]uint8{0x3a, 0x42, 0}},
 	{AMULB, ydivb, Pb, [23]uint8{0xf6, 04}},
 	{AMULL, ydivl, Px, [23]uint8{0xf7, 04}},
 	{AMULPD, yxm, Pe, [23]uint8{0x59}},
@@ -1269,6 +1279,7 @@ var optab =
 	{APABSW, yxm_q4, Pq4, [23]uint8{0x1d}},
 	{APACKSSLW, ymm, Py1, [23]uint8{0x6b, Pe, 0x6b}},
 	{APACKSSWB, ymm, Py1, [23]uint8{0x63, Pe, 0x63}},
+	{APACKUSDW, yxm_q4, Pq4, [23]uint8{0x2b}},
 	{APACKUSWB, ymm, Py1, [23]uint8{0x67, Pe, 0x67}},
 	{APADDB, ymm, Py1, [23]uint8{0xfc, Pe, 0xfc}},
 	{APADDL, ymm, Py1, [23]uint8{0xfe, Pe, 0xfe}},
@@ -1284,12 +1295,17 @@ var optab =
 	{APAUSE, ynone, Px, [23]uint8{0xf3, 0x90}},
 	{APAVGB, ymm, Py1, [23]uint8{0xe0, Pe, 0xe0}},
 	{APAVGW, ymm, Py1, [23]uint8{0xe3, Pe, 0xe3}},
+	{APBLENDW, yxshuf, Pq, [23]uint8{0x3a, 0x0e, 0}},
 	{APCMPEQB, ymm, Py1, [23]uint8{0x74, Pe, 0x74}},
 	{APCMPEQL, ymm, Py1, [23]uint8{0x76, Pe, 0x76}},
+	{APCMPEQQ, yxm_q4, Pq4, [23]uint8{0x29}},
 	{APCMPEQW, ymm, Py1, [23]uint8{0x75, Pe, 0x75}},
 	{APCMPGTB, ymm, Py1, [23]uint8{0x64, Pe, 0x64}},
 	{APCMPGTL, ymm, Py1, [23]uint8{0x66, Pe, 0x66}},
+	{APCMPGTQ, yxm_q4, Pq4, [23]uint8{0x37}},
 	{APCMPGTW, ymm, Py1, [23]uint8{0x65, Pe, 0x65}},
+	{APCMPISTRI, yxshuf, Pq, [23]uint8{0x3a, 0x63, 0}},
+	{APCMPISTRM, yxshuf, Pq, [23]uint8{0x3a, 0x62, 0}},
 	{APEXTRW, yextrw, Pq, [23]uint8{0xc5, 00}},
 	{APEXTRB, yextr, Pq, [23]uint8{0x3a, 0x14, 00}},
 	{APEXTRD, yextr, Pq, [23]uint8{0x3a, 0x16, 00}},
@@ -1307,10 +1323,18 @@ var optab =
 	{APINSRQ, yinsr, Pq3, [23]uint8{0x3a, 0x22, 00}},
 	{APMADDUBSW, yxm_q4, Pq4, [23]uint8{0x04}},
 	{APMADDWL, ymm, Py1, [23]uint8{0xf5, Pe, 0xf5}},
+	{APMAXSB, yxm_q4, Pq4, [23]uint8{0x3c}},
+	{APMAXSD, yxm_q4, Pq4, [23]uint8{0x3d}},
 	{APMAXSW, yxm, Pe, [23]uint8{0xee}},
 	{APMAXUB, yxm, Pe, [23]uint8{0xde}},
+	{APMAXUD, yxm_q4, Pq4, [23]uint8{0x3f}},
+	{APMAXUW, yxm_q4, Pq4, [23]uint8{0x3e}},
+	{APMINSB, yxm_q4, Pq4, [23]uint8{0x38}},
+	{APMINSD, yxm_q4, Pq4, [23]uint8{0x39}},
 	{APMINSW, yxm, Pe, [23]uint8{0xea}},
 	{APMINUB, yxm, Pe, [23]uint8{0xda}},
+	{APMINUD, yxm_q4, Pq4, [23]uint8{0x3b}},
+	{APMINUW, yxm_q4, Pq4, [23]uint8{0x3a}},
 	{APMOVMSKB, ymskb, Px, [23]uint8{Pe, 0xd7, 0xd7}},
 	{APMOVSXBD, yxm_q4, Pq4, [23]uint8{0x21}},
 	{APMOVSXBQ, yxm_q4, Pq4, [23]uint8{0x22}},
@@ -1370,6 +1394,7 @@ var optab =
 	{APSUBUSB, yxm, Pe, [23]uint8{0xd8}},
 	{APSUBUSW, yxm, Pe, [23]uint8{0xd9}},
 	{APSUBW, yxm, Pe, [23]uint8{0xf9}},
+	{APTEST, yxm_q4, Pq4, [23]uint8{0x17}},
 	{APUNPCKHBW, ymm, Py1, [23]uint8{0x68, Pe, 0x68}},
 	{APUNPCKHLQ, ymm, Py1, [23]uint8{0x6a, Pe, 0x6a}},
 	{APUNPCKHQDQ, yxm, Pe, [23]uint8{0x6d}},
@@ -1656,6 +1681,7 @@ var optab =
 	{APSHUFD, yxshuf, Pq, [23]uint8{0x70, 0}},
 	{APCLMULQDQ, yxshuf, Pq, [23]uint8{0x3a, 0x44, 0}},
 	{APCMPESTRI, yxshuf, Pq, [23]uint8{0x3a, 0x61, 0}},
+	{APCMPESTRM, yxshuf, Pq, [23]uint8{0x3a, 0x60, 0}},
 	{AMOVDDUP, yxm, Pf2, [23]uint8{0x12}},
 	{AMOVSHDUP, yxm, Pf3, [23]uint8{0x16}},
 	{AMOVSLDUP, yxm, Pf3, [23]uint8{0x12}},
@@ -1664,6 +1690,8 @@ var optab =
 	{AANDNQ, yvex_r3, Pvex, [23]uint8{VEX_LZ_0F38_W1, 0xF2}},
 	{ABEXTRL, yvex_vmr3, Pvex, [23]uint8{VEX_LZ_0F38_W0, 0xF7}},
 	{ABEXTRQ, yvex_vmr3, Pvex, [23]uint8{VEX_LZ_0F38_W1, 0xF7}},
+	{ABLENDPD, yxshuf, Pq, [23]uint8{0x3a, 0x0d, 0}},
+	{ABLENDPS, yxshuf, Pq, [23]uint8{0x3a, 0x0c, 0}},
 	{ABZHIL, yvex_vmr3, Pvex, [23]uint8{VEX_LZ_0F38_W0, 0xF5}},
 	{ABZHIQ, yvex_vmr3, Pvex, [23]uint8{VEX_LZ_0F38_W1, 0xF5}},
 	{AMULXL, yvex_r3, Pvex, [23]uint8{VEX_LZ_F2_0F38_W0, 0xF6}},