diff options
author | Igor Tsimbalist <igor.v.tsimbalist@intel.com> | 2017-10-20 23:52:52 +0300 |
---|---|---|
committer | Igor Tsimbalist <igor.v.tsimbalist@intel.com> | 2017-10-23 15:58:18 +0300 |
commit | 8cfcb7659cb844dff00efbbb644c15b650fb7e8b (patch) | |
tree | e9139b9cde353c68cfb9b963743cb73453e6d81c | |
parent | ff1982d53a1fba573e7f9a3b455f7644440cb336 (diff) | |
download | gdb-8cfcb7659cb844dff00efbbb644c15b650fb7e8b.zip gdb-8cfcb7659cb844dff00efbbb644c15b650fb7e8b.tar.gz gdb-8cfcb7659cb844dff00efbbb644c15b650fb7e8b.tar.bz2 |
Enable Intel AVX512_VNNI instructions.
Intel has disclosed a set of new instructions. The spec is
https://software.intel.com/sites/default/files/managed/c5/15/architecture-instruction-set-extensions-programming-reference.pdf
gas/
* config/tc-i386.c (cpu_arch): Add .avx512_vnni.
(cpu_noarch): Add noavx512_vnni.
* doc/c-i386.texi: Document .avx512_vnni.
* testsuite/gas/i386/i386.exp: Add AVX512_VNNI tests.
* testsuite/gas/i386/avx512vnni-intel.d: New test.
* testsuite/gas/i386/avx512vnni.d: Likewise.
* testsuite/gas/i386/avx512vnni.s: Likewise.
* testsuite/gas/i386/avx512vnni_vl-intel.d: Likewise.
* testsuite/gas/i386/avx512vnni_vl.d: Likewise.
* testsuite/gas/i386/avx512vnni_vl.s: Likewise.
* testsuite/gas/i386/x86-64-avx512vnni-intel.d: Likewise.
* testsuite/gas/i386/x86-64-avx512vnni.d: Likewise.
* testsuite/gas/i386/x86-64-avx512vnni.s: Likewise.
* testsuite/gas/i386/x86-64-avx512vnni_vl-intel.d: Likewise.
* testsuite/gas/i386/x86-64-avx512vnni_vl.d: Likewise.
* testsuite/gas/i386/x86-64-avx512vnni_vl.s: Likewise.
opcodes/
* i386-dis.c (enum): Add PREFIX_EVEX_0F3850, PREFIX_EVEX_0F3851.
* i386-dis-evex.h (evex_table): Updated.
* i386-gen.c (cpu_flag_init): Add CPU_AVX512_VNNI,
CPU_ANY_AVX512_VNNI_FLAGS. Update CPU_ANY_AVX512F_FLAGS.
(cpu_flags): Add CpuAVX512_VNNI.
* i386-opc.h (enum): Add CpuAVX512_VNNI.
(i386_cpu_flags): Add cpuavx512_vnni.
* i386-opc.tbl Add Intel AVX512_VNNI instructions.
* i386-init.h: Regenerate.
* i386-tbl.h: Likewise.
-rw-r--r-- | gas/config/tc-i386.c | 3 | ||||
-rw-r--r-- | gas/doc/c-i386.texi | 4 | ||||
-rw-r--r-- | gas/testsuite/gas/i386/avx512vnni-intel.d | 60 | ||||
-rw-r--r-- | gas/testsuite/gas/i386/avx512vnni.d | 60 | ||||
-rw-r--r-- | gas/testsuite/gas/i386/avx512vnni.s | 60 | ||||
-rw-r--r-- | gas/testsuite/gas/i386/avx512vnni_vl-intel.d | 92 | ||||
-rw-r--r-- | gas/testsuite/gas/i386/avx512vnni_vl.d | 92 | ||||
-rw-r--r-- | gas/testsuite/gas/i386/avx512vnni_vl.s | 93 | ||||
-rw-r--r-- | gas/testsuite/gas/i386/i386.exp | 8 | ||||
-rw-r--r-- | gas/testsuite/gas/i386/x86-64-avx512vnni-intel.d | 60 | ||||
-rw-r--r-- | gas/testsuite/gas/i386/x86-64-avx512vnni.d | 60 | ||||
-rw-r--r-- | gas/testsuite/gas/i386/x86-64-avx512vnni.s | 61 | ||||
-rw-r--r-- | gas/testsuite/gas/i386/x86-64-avx512vnni_vl-intel.d | 108 | ||||
-rw-r--r-- | gas/testsuite/gas/i386/x86-64-avx512vnni_vl.d | 108 | ||||
-rw-r--r-- | gas/testsuite/gas/i386/x86-64-avx512vnni_vl.s | 109 | ||||
-rw-r--r-- | opcodes/i386-dis-evex.h | 20 | ||||
-rw-r--r-- | opcodes/i386-dis.c | 2 | ||||
-rw-r--r-- | opcodes/i386-gen.c | 7 | ||||
-rw-r--r-- | opcodes/i386-opc.h | 5 | ||||
-rw-r--r-- | opcodes/i386-opc.tbl | 20 |
20 files changed, 1025 insertions, 7 deletions
diff --git a/gas/config/tc-i386.c b/gas/config/tc-i386.c index f4b86f6..1f2efdc 100644 --- a/gas/config/tc-i386.c +++ b/gas/config/tc-i386.c @@ -984,6 +984,8 @@ static const arch_entry cpu_arch[] = CPU_AVX512_VPOPCNTDQ_FLAGS, 0 }, { STRING_COMMA_LEN (".avx512_vbmi2"), PROCESSOR_UNKNOWN, CPU_AVX512_VBMI2_FLAGS, 0 }, + { STRING_COMMA_LEN (".avx512_vnni"), PROCESSOR_UNKNOWN, + CPU_AVX512_VNNI_FLAGS, 0 }, { STRING_COMMA_LEN (".clzero"), PROCESSOR_UNKNOWN, CPU_CLZERO_FLAGS, 0 }, { STRING_COMMA_LEN (".mwaitx"), PROCESSOR_UNKNOWN, @@ -1033,6 +1035,7 @@ static const noarch_entry cpu_noarch[] = { STRING_COMMA_LEN ("noavx512_4vnniw"), CPU_ANY_AVX512_4VNNIW_FLAGS }, { STRING_COMMA_LEN ("noavx512_vpopcntdq"), CPU_ANY_AVX512_VPOPCNTDQ_FLAGS }, { STRING_COMMA_LEN ("noavx512_vbmi2"), CPU_ANY_AVX512_VBMI2_FLAGS }, + { STRING_COMMA_LEN ("noavx512_vnni"), CPU_ANY_AVX512_VNNI_FLAGS }, }; #ifdef I386COFF diff --git a/gas/doc/c-i386.texi b/gas/doc/c-i386.texi index 44ade83..f2e8f4e 100644 --- a/gas/doc/c-i386.texi +++ b/gas/doc/c-i386.texi @@ -188,6 +188,7 @@ accept various extension mnemonics. For example, @code{avx512_4vnniw}, @code{avx512_vpopcntdq}, @code{avx512_vbmi2}, +@code{avx512_vnni}, @code{noavx512f}, @code{noavx512cd}, @code{noavx512er}, @@ -201,6 +202,7 @@ accept various extension mnemonics. For example, @code{noavx512_4vnniw}, @code{noavx512_vpopcntdq}, @code{noavx512_vbmi2}, +@code{noavx512_vnni}, @code{vmx}, @code{vmfunc}, @code{smx}, @@ -1223,7 +1225,7 @@ supported on the CPU specified. The choices for @var{cpu_type} are: @item @samp{.avx512f} @tab @samp{.avx512cd} @tab @samp{.avx512er} @tab @samp{.avx512pf} @item @samp{.avx512vl} @tab @samp{.avx512bw} @tab @samp{.avx512dq} @tab @samp{.avx512ifma} @item @samp{.avx512vbmi} @tab @samp{.avx512_4fmaps} @tab @samp{.avx512_4vnniw} -@item @samp{.avx512_vpopcntdq} @tab @samp{.avx512_vbmi2} +@item @samp{.avx512_vpopcntdq} @tab @samp{.avx512_vbmi2} @tab @samp{.avx512_vnni} @item @samp{.clwb} @tab @samp{.rdpid} @tab @samp{.ptwrite} @tab @item @samp{.cet} @item @samp{.3dnow} @tab @samp{.3dnowa} @tab @samp{.sse4a} @tab @samp{.sse5} @item @samp{.syscall} @tab @samp{.rdtscp} @tab @samp{.svme} @tab @samp{.abm} diff --git a/gas/testsuite/gas/i386/avx512vnni-intel.d b/gas/testsuite/gas/i386/avx512vnni-intel.d new file mode 100644 index 0000000..b2ac448 --- /dev/null +++ b/gas/testsuite/gas/i386/avx512vnni-intel.d @@ -0,0 +1,60 @@ +#as: +#objdump: -dw -Mintel +#name: i386 AVX512VNNI insns (Intel disassembly) +#source: avx512vnni.s + +.*: +file format .* + + +Disassembly of section \.text: + +00000000 <_start>: +[ ]*[a-f0-9]+:[ ]*62 f2 75 48 52 e3[ ]*vpdpwssd zmm4,zmm1,zmm3 +[ ]*[a-f0-9]+:[ ]*62 f2 75 49 52 e3[ ]*vpdpwssd zmm4\{k1\},zmm1,zmm3 +[ ]*[a-f0-9]+:[ ]*62 f2 75 c9 52 e3[ ]*vpdpwssd zmm4\{k1\}\{z\},zmm1,zmm3 +[ ]*[a-f0-9]+:[ ]*62 f2 75 48 52 a4 f4 c0 1d fe ff[ ]*vpdpwssd zmm4,zmm1,ZMMWORD PTR \[esp\+esi\*8-0x1e240\] +[ ]*[a-f0-9]+:[ ]*62 f2 75 48 52 62 7f[ ]*vpdpwssd zmm4,zmm1,ZMMWORD PTR \[edx\+0x1fc0\] +[ ]*[a-f0-9]+:[ ]*62 f2 75 58 52 62 7f[ ]*vpdpwssd zmm4,zmm1,DWORD PTR \[edx\+0x1fc\]\{1to16\} +[ ]*[a-f0-9]+:[ ]*62 f2 55 48 53 d4[ ]*vpdpwssds zmm2,zmm5,zmm4 +[ ]*[a-f0-9]+:[ ]*62 f2 55 4e 53 d4[ ]*vpdpwssds zmm2\{k6\},zmm5,zmm4 +[ ]*[a-f0-9]+:[ ]*62 f2 55 ce 53 d4[ ]*vpdpwssds zmm2\{k6\}\{z\},zmm5,zmm4 +[ ]*[a-f0-9]+:[ ]*62 f2 55 48 53 94 f4 c0 1d fe ff[ ]*vpdpwssds zmm2,zmm5,ZMMWORD PTR \[esp\+esi\*8-0x1e240\] +[ ]*[a-f0-9]+:[ ]*62 f2 55 48 53 52 7f[ ]*vpdpwssds zmm2,zmm5,ZMMWORD PTR \[edx\+0x1fc0\] +[ ]*[a-f0-9]+:[ ]*62 f2 55 58 53 52 7f[ ]*vpdpwssds zmm2,zmm5,DWORD PTR \[edx\+0x1fc\]\{1to16\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 48 50 eb[ ]*vpdpbusd zmm5,zmm2,zmm3 +[ ]*[a-f0-9]+:[ ]*62 f2 6d 49 50 eb[ ]*vpdpbusd zmm5\{k1\},zmm2,zmm3 +[ ]*[a-f0-9]+:[ ]*62 f2 6d c9 50 eb[ ]*vpdpbusd zmm5\{k1\}\{z\},zmm2,zmm3 +[ ]*[a-f0-9]+:[ ]*62 f2 6d 48 50 ac f4 c0 1d fe ff[ ]*vpdpbusd zmm5,zmm2,ZMMWORD PTR \[esp\+esi\*8-0x1e240\] +[ ]*[a-f0-9]+:[ ]*62 f2 6d 48 50 6a 7f[ ]*vpdpbusd zmm5,zmm2,ZMMWORD PTR \[edx\+0x1fc0\] +[ ]*[a-f0-9]+:[ ]*62 f2 6d 58 50 6a 7f[ ]*vpdpbusd zmm5,zmm2,DWORD PTR \[edx\+0x1fc\]\{1to16\} +[ ]*[a-f0-9]+:[ ]*62 f2 65 48 51 e9[ ]*vpdpbusds zmm5,zmm3,zmm1 +[ ]*[a-f0-9]+:[ ]*62 f2 65 4a 51 e9[ ]*vpdpbusds zmm5\{k2\},zmm3,zmm1 +[ ]*[a-f0-9]+:[ ]*62 f2 65 ca 51 e9[ ]*vpdpbusds zmm5\{k2\}\{z\},zmm3,zmm1 +[ ]*[a-f0-9]+:[ ]*62 f2 65 48 51 ac f4 c0 1d fe ff[ ]*vpdpbusds zmm5,zmm3,ZMMWORD PTR \[esp\+esi\*8-0x1e240\] +[ ]*[a-f0-9]+:[ ]*62 f2 65 48 51 6a 7f[ ]*vpdpbusds zmm5,zmm3,ZMMWORD PTR \[edx\+0x1fc0\] +[ ]*[a-f0-9]+:[ ]*62 f2 65 58 51 6a 7f[ ]*vpdpbusds zmm5,zmm3,DWORD PTR \[edx\+0x1fc\]\{1to16\} +[ ]*[a-f0-9]+:[ ]*62 f2 5d 48 52 d9[ ]*vpdpwssd zmm3,zmm4,zmm1 +[ ]*[a-f0-9]+:[ ]*62 f2 5d 4b 52 d9[ ]*vpdpwssd zmm3\{k3\},zmm4,zmm1 +[ ]*[a-f0-9]+:[ ]*62 f2 5d cb 52 d9[ ]*vpdpwssd zmm3\{k3\}\{z\},zmm4,zmm1 +[ ]*[a-f0-9]+:[ ]*62 f2 5d 48 52 9c f4 c0 1d fe ff[ ]*vpdpwssd zmm3,zmm4,ZMMWORD PTR \[esp\+esi\*8-0x1e240\] +[ ]*[a-f0-9]+:[ ]*62 f2 5d 48 52 5a 7f[ ]*vpdpwssd zmm3,zmm4,ZMMWORD PTR \[edx\+0x1fc0\] +[ ]*[a-f0-9]+:[ ]*62 f2 5d 58 52 5a 7f[ ]*vpdpwssd zmm3,zmm4,DWORD PTR \[edx\+0x1fc\]\{1to16\} +[ ]*[a-f0-9]+:[ ]*62 f2 75 48 53 da[ ]*vpdpwssds zmm3,zmm1,zmm2 +[ ]*[a-f0-9]+:[ ]*62 f2 75 4f 53 da[ ]*vpdpwssds zmm3\{k7\},zmm1,zmm2 +[ ]*[a-f0-9]+:[ ]*62 f2 75 cf 53 da[ ]*vpdpwssds zmm3\{k7\}\{z\},zmm1,zmm2 +[ ]*[a-f0-9]+:[ ]*62 f2 75 48 53 9c f4 c0 1d fe ff[ ]*vpdpwssds zmm3,zmm1,ZMMWORD PTR \[esp\+esi\*8-0x1e240\] +[ ]*[a-f0-9]+:[ ]*62 f2 75 48 53 5a 7f[ ]*vpdpwssds zmm3,zmm1,ZMMWORD PTR \[edx\+0x1fc0\] +[ ]*[a-f0-9]+:[ ]*62 f2 75 58 53 5a 7f[ ]*vpdpwssds zmm3,zmm1,DWORD PTR \[edx\+0x1fc\]\{1to16\} +[ ]*[a-f0-9]+:[ ]*62 f2 5d 48 50 d9[ ]*vpdpbusd zmm3,zmm4,zmm1 +[ ]*[a-f0-9]+:[ ]*62 f2 5d 4e 50 d9[ ]*vpdpbusd zmm3\{k6\},zmm4,zmm1 +[ ]*[a-f0-9]+:[ ]*62 f2 5d ce 50 d9[ ]*vpdpbusd zmm3\{k6\}\{z\},zmm4,zmm1 +[ ]*[a-f0-9]+:[ ]*62 f2 5d 48 50 9c f4 c0 1d fe ff[ ]*vpdpbusd zmm3,zmm4,ZMMWORD PTR \[esp\+esi\*8-0x1e240\] +[ ]*[a-f0-9]+:[ ]*62 f2 5d 48 50 5a 7f[ ]*vpdpbusd zmm3,zmm4,ZMMWORD PTR \[edx\+0x1fc0\] +[ ]*[a-f0-9]+:[ ]*62 f2 5d 58 50 5a 7f[ ]*vpdpbusd zmm3,zmm4,DWORD PTR \[edx\+0x1fc\]\{1to16\} +[ ]*[a-f0-9]+:[ ]*62 f2 75 48 51 c9[ ]*vpdpbusds zmm1,zmm1,zmm1 +[ ]*[a-f0-9]+:[ ]*62 f2 75 49 51 c9[ ]*vpdpbusds zmm1\{k1\},zmm1,zmm1 +[ ]*[a-f0-9]+:[ ]*62 f2 75 c9 51 c9[ ]*vpdpbusds zmm1\{k1\}\{z\},zmm1,zmm1 +[ ]*[a-f0-9]+:[ ]*62 f2 75 48 51 8c f4 c0 1d fe ff[ ]*vpdpbusds zmm1,zmm1,ZMMWORD PTR \[esp\+esi\*8-0x1e240\] +[ ]*[a-f0-9]+:[ ]*62 f2 75 48 51 4a 7f[ ]*vpdpbusds zmm1,zmm1,ZMMWORD PTR \[edx\+0x1fc0\] +[ ]*[a-f0-9]+:[ ]*62 f2 75 58 51 4a 7f[ ]*vpdpbusds zmm1,zmm1,DWORD PTR \[edx\+0x1fc\]\{1to16\} +#pass diff --git a/gas/testsuite/gas/i386/avx512vnni.d b/gas/testsuite/gas/i386/avx512vnni.d new file mode 100644 index 0000000..353dfcf --- /dev/null +++ b/gas/testsuite/gas/i386/avx512vnni.d @@ -0,0 +1,60 @@ +#as: +#objdump: -dw +#name: i386 AVX512VNNI insns +#source: avx512vnni.s + +.*: +file format .* + + +Disassembly of section \.text: + +00000000 <_start>: +[ ]*[a-f0-9]+:[ ]*62 f2 75 48 52 e3[ ]*vpdpwssd %zmm3,%zmm1,%zmm4 +[ ]*[a-f0-9]+:[ ]*62 f2 75 49 52 e3[ ]*vpdpwssd %zmm3,%zmm1,%zmm4\{%k1\} +[ ]*[a-f0-9]+:[ ]*62 f2 75 c9 52 e3[ ]*vpdpwssd %zmm3,%zmm1,%zmm4\{%k1\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 f2 75 48 52 a4 f4 c0 1d fe ff[ ]*vpdpwssd -0x1e240\(%esp,%esi,8\),%zmm1,%zmm4 +[ ]*[a-f0-9]+:[ ]*62 f2 75 48 52 62 7f[ ]*vpdpwssd 0x1fc0\(%edx\),%zmm1,%zmm4 +[ ]*[a-f0-9]+:[ ]*62 f2 75 58 52 62 7f[ ]*vpdpwssd 0x1fc\(%edx\)\{1to16\},%zmm1,%zmm4 +[ ]*[a-f0-9]+:[ ]*62 f2 55 48 53 d4[ ]*vpdpwssds %zmm4,%zmm5,%zmm2 +[ ]*[a-f0-9]+:[ ]*62 f2 55 4e 53 d4[ ]*vpdpwssds %zmm4,%zmm5,%zmm2\{%k6\} +[ ]*[a-f0-9]+:[ ]*62 f2 55 ce 53 d4[ ]*vpdpwssds %zmm4,%zmm5,%zmm2\{%k6\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 f2 55 48 53 94 f4 c0 1d fe ff[ ]*vpdpwssds -0x1e240\(%esp,%esi,8\),%zmm5,%zmm2 +[ ]*[a-f0-9]+:[ ]*62 f2 55 48 53 52 7f[ ]*vpdpwssds 0x1fc0\(%edx\),%zmm5,%zmm2 +[ ]*[a-f0-9]+:[ ]*62 f2 55 58 53 52 7f[ ]*vpdpwssds 0x1fc\(%edx\)\{1to16\},%zmm5,%zmm2 +[ ]*[a-f0-9]+:[ ]*62 f2 6d 48 50 eb[ ]*vpdpbusd %zmm3,%zmm2,%zmm5 +[ ]*[a-f0-9]+:[ ]*62 f2 6d 49 50 eb[ ]*vpdpbusd %zmm3,%zmm2,%zmm5\{%k1\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d c9 50 eb[ ]*vpdpbusd %zmm3,%zmm2,%zmm5\{%k1\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 48 50 ac f4 c0 1d fe ff[ ]*vpdpbusd -0x1e240\(%esp,%esi,8\),%zmm2,%zmm5 +[ ]*[a-f0-9]+:[ ]*62 f2 6d 48 50 6a 7f[ ]*vpdpbusd 0x1fc0\(%edx\),%zmm2,%zmm5 +[ ]*[a-f0-9]+:[ ]*62 f2 6d 58 50 6a 7f[ ]*vpdpbusd 0x1fc\(%edx\)\{1to16\},%zmm2,%zmm5 +[ ]*[a-f0-9]+:[ ]*62 f2 65 48 51 e9[ ]*vpdpbusds %zmm1,%zmm3,%zmm5 +[ ]*[a-f0-9]+:[ ]*62 f2 65 4a 51 e9[ ]*vpdpbusds %zmm1,%zmm3,%zmm5\{%k2\} +[ ]*[a-f0-9]+:[ ]*62 f2 65 ca 51 e9[ ]*vpdpbusds %zmm1,%zmm3,%zmm5\{%k2\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 f2 65 48 51 ac f4 c0 1d fe ff[ ]*vpdpbusds -0x1e240\(%esp,%esi,8\),%zmm3,%zmm5 +[ ]*[a-f0-9]+:[ ]*62 f2 65 48 51 6a 7f[ ]*vpdpbusds 0x1fc0\(%edx\),%zmm3,%zmm5 +[ ]*[a-f0-9]+:[ ]*62 f2 65 58 51 6a 7f[ ]*vpdpbusds 0x1fc\(%edx\)\{1to16\},%zmm3,%zmm5 +[ ]*[a-f0-9]+:[ ]*62 f2 5d 48 52 d9[ ]*vpdpwssd %zmm1,%zmm4,%zmm3 +[ ]*[a-f0-9]+:[ ]*62 f2 5d 4b 52 d9[ ]*vpdpwssd %zmm1,%zmm4,%zmm3\{%k3\} +[ ]*[a-f0-9]+:[ ]*62 f2 5d cb 52 d9[ ]*vpdpwssd %zmm1,%zmm4,%zmm3\{%k3\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 f2 5d 48 52 9c f4 c0 1d fe ff[ ]*vpdpwssd -0x1e240\(%esp,%esi,8\),%zmm4,%zmm3 +[ ]*[a-f0-9]+:[ ]*62 f2 5d 48 52 5a 7f[ ]*vpdpwssd 0x1fc0\(%edx\),%zmm4,%zmm3 +[ ]*[a-f0-9]+:[ ]*62 f2 5d 58 52 5a 7f[ ]*vpdpwssd 0x1fc\(%edx\)\{1to16\},%zmm4,%zmm3 +[ ]*[a-f0-9]+:[ ]*62 f2 75 48 53 da[ ]*vpdpwssds %zmm2,%zmm1,%zmm3 +[ ]*[a-f0-9]+:[ ]*62 f2 75 4f 53 da[ ]*vpdpwssds %zmm2,%zmm1,%zmm3\{%k7\} +[ ]*[a-f0-9]+:[ ]*62 f2 75 cf 53 da[ ]*vpdpwssds %zmm2,%zmm1,%zmm3\{%k7\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 f2 75 48 53 9c f4 c0 1d fe ff[ ]*vpdpwssds -0x1e240\(%esp,%esi,8\),%zmm1,%zmm3 +[ ]*[a-f0-9]+:[ ]*62 f2 75 48 53 5a 7f[ ]*vpdpwssds 0x1fc0\(%edx\),%zmm1,%zmm3 +[ ]*[a-f0-9]+:[ ]*62 f2 75 58 53 5a 7f[ ]*vpdpwssds 0x1fc\(%edx\)\{1to16\},%zmm1,%zmm3 +[ ]*[a-f0-9]+:[ ]*62 f2 5d 48 50 d9[ ]*vpdpbusd %zmm1,%zmm4,%zmm3 +[ ]*[a-f0-9]+:[ ]*62 f2 5d 4e 50 d9[ ]*vpdpbusd %zmm1,%zmm4,%zmm3\{%k6\} +[ ]*[a-f0-9]+:[ ]*62 f2 5d ce 50 d9[ ]*vpdpbusd %zmm1,%zmm4,%zmm3\{%k6\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 f2 5d 48 50 9c f4 c0 1d fe ff[ ]*vpdpbusd -0x1e240\(%esp,%esi,8\),%zmm4,%zmm3 +[ ]*[a-f0-9]+:[ ]*62 f2 5d 48 50 5a 7f[ ]*vpdpbusd 0x1fc0\(%edx\),%zmm4,%zmm3 +[ ]*[a-f0-9]+:[ ]*62 f2 5d 58 50 5a 7f[ ]*vpdpbusd 0x1fc\(%edx\)\{1to16\},%zmm4,%zmm3 +[ ]*[a-f0-9]+:[ ]*62 f2 75 48 51 c9[ ]*vpdpbusds %zmm1,%zmm1,%zmm1 +[ ]*[a-f0-9]+:[ ]*62 f2 75 49 51 c9[ ]*vpdpbusds %zmm1,%zmm1,%zmm1\{%k1\} +[ ]*[a-f0-9]+:[ ]*62 f2 75 c9 51 c9[ ]*vpdpbusds %zmm1,%zmm1,%zmm1\{%k1\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 f2 75 48 51 8c f4 c0 1d fe ff[ ]*vpdpbusds -0x1e240\(%esp,%esi,8\),%zmm1,%zmm1 +[ ]*[a-f0-9]+:[ ]*62 f2 75 48 51 4a 7f[ ]*vpdpbusds 0x1fc0\(%edx\),%zmm1,%zmm1 +[ ]*[a-f0-9]+:[ ]*62 f2 75 58 51 4a 7f[ ]*vpdpbusds 0x1fc\(%edx\)\{1to16\},%zmm1,%zmm1 +#pass diff --git a/gas/testsuite/gas/i386/avx512vnni.s b/gas/testsuite/gas/i386/avx512vnni.s new file mode 100644 index 0000000..b197613 --- /dev/null +++ b/gas/testsuite/gas/i386/avx512vnni.s @@ -0,0 +1,60 @@ +# Check 32bit AVX512VNNI instructions + + .allow_index_reg + .text +_start: + vpdpwssd %zmm3, %zmm1, %zmm4 # AVX512VNNI + vpdpwssd %zmm3, %zmm1, %zmm4{%k1} # AVX512VNNI + vpdpwssd %zmm3, %zmm1, %zmm4{%k1}{z} # AVX512VNNI + vpdpwssd -123456(%esp,%esi,8), %zmm1, %zmm4 # AVX512VNNI + vpdpwssd 8128(%edx), %zmm1, %zmm4 # AVX512VNNI Disp8 + vpdpwssd 508(%edx){1to16}, %zmm1, %zmm4 # AVX512VNNI Disp8 + + vpdpwssds %zmm4, %zmm5, %zmm2 # AVX512VNNI + vpdpwssds %zmm4, %zmm5, %zmm2{%k6} # AVX512VNNI + vpdpwssds %zmm4, %zmm5, %zmm2{%k6}{z} # AVX512VNNI + vpdpwssds -123456(%esp,%esi,8), %zmm5, %zmm2 # AVX512VNNI + vpdpwssds 8128(%edx), %zmm5, %zmm2 # AVX512VNNI Disp8 + vpdpwssds 508(%edx){1to16}, %zmm5, %zmm2 # AVX512VNNI Disp8 + + vpdpbusd %zmm3, %zmm2, %zmm5 # AVX512VNNI + vpdpbusd %zmm3, %zmm2, %zmm5{%k1} # AVX512VNNI + vpdpbusd %zmm3, %zmm2, %zmm5{%k1}{z} # AVX512VNNI + vpdpbusd -123456(%esp,%esi,8), %zmm2, %zmm5 # AVX512VNNI + vpdpbusd 8128(%edx), %zmm2, %zmm5 # AVX512VNNI Disp8 + vpdpbusd 508(%edx){1to16}, %zmm2, %zmm5 # AVX512VNNI Disp8 + + vpdpbusds %zmm1, %zmm3, %zmm5 # AVX512VNNI + vpdpbusds %zmm1, %zmm3, %zmm5{%k2} # AVX512VNNI + vpdpbusds %zmm1, %zmm3, %zmm5{%k2}{z} # AVX512VNNI + vpdpbusds -123456(%esp,%esi,8), %zmm3, %zmm5 # AVX512VNNI + vpdpbusds 8128(%edx), %zmm3, %zmm5 # AVX512VNNI Disp8 + vpdpbusds 508(%edx){1to16}, %zmm3, %zmm5 # AVX512VNNI Disp8 + + .intel_syntax noprefix + vpdpwssd zmm3, zmm4, zmm1 # AVX512VNNI + vpdpwssd zmm3{k3}, zmm4, zmm1 # AVX512VNNI + vpdpwssd zmm3{k3}{z}, zmm4, zmm1 # AVX512VNNI + vpdpwssd zmm3, zmm4, ZMMWORD PTR [esp+esi*8-123456] # AVX512VNNI + vpdpwssd zmm3, zmm4, ZMMWORD PTR [edx+8128] # AVX512VNNI Disp8 + vpdpwssd zmm3, zmm4, [edx+508]{1to16} # AVX512VNNI Disp8 + + vpdpwssds zmm3, zmm1, zmm2 # AVX512VNNI + vpdpwssds zmm3{k7}, zmm1, zmm2 # AVX512VNNI + vpdpwssds zmm3{k7}{z}, zmm1, zmm2 # AVX512VNNI + vpdpwssds zmm3, zmm1, ZMMWORD PTR [esp+esi*8-123456] # AVX512VNNI + vpdpwssds zmm3, zmm1, ZMMWORD PTR [edx+8128] # AVX512VNNI Disp8 + vpdpwssds zmm3, zmm1, [edx+508]{1to16} # AVX512VNNI Disp8 + + vpdpbusd zmm3, zmm4, zmm1 # AVX512VNNI + vpdpbusd zmm3{k6}, zmm4, zmm1 # AVX512VNNI + vpdpbusd zmm3{k6}{z}, zmm4, zmm1 # AVX512VNNI + vpdpbusd zmm3, zmm4, ZMMWORD PTR [esp+esi*8-123456] # AVX512VNNI + vpdpbusd zmm3, zmm4, ZMMWORD PTR [edx+8128] # AVX512VNNI Disp8 + vpdpbusd zmm3, zmm4, [edx+508]{1to16} # AVX512VNNI Disp8 + vpdpbusds zmm1, zmm1, zmm1 # AVX512VNNI + vpdpbusds zmm1{k1}, zmm1, zmm1 # AVX512VNNI + vpdpbusds zmm1{k1}{z}, zmm1, zmm1 # AVX512VNNI + vpdpbusds zmm1, zmm1, ZMMWORD PTR [esp+esi*8-123456] # AVX512VNNI + vpdpbusds zmm1, zmm1, ZMMWORD PTR [edx+8128] # AVX512VNNI Disp8 + vpdpbusds zmm1, zmm1, [edx+508]{1to16} # AVX512VNNI Disp8 diff --git a/gas/testsuite/gas/i386/avx512vnni_vl-intel.d b/gas/testsuite/gas/i386/avx512vnni_vl-intel.d new file mode 100644 index 0000000..5a8c8c4 --- /dev/null +++ b/gas/testsuite/gas/i386/avx512vnni_vl-intel.d @@ -0,0 +1,92 @@ +#as: +#objdump: -dw -Mintel +#name: i386 AVX512VNNI/VL insns (Intel disassembly) +#source: avx512vnni_vl.s + +.*: +file format .* + + +Disassembly of section \.text: + +00000000 <_start>: +[ ]*[a-f0-9]+:[ ]*62 f2 5d 0b 52 d2[ ]*vpdpwssd xmm2\{k3\},xmm4,xmm2 +[ ]*[a-f0-9]+:[ ]*62 f2 5d 8b 52 d2[ ]*vpdpwssd xmm2\{k3\}\{z\},xmm4,xmm2 +[ ]*[a-f0-9]+:[ ]*62 f2 5d 09 52 94 f4 c0 1d fe ff[ ]*vpdpwssd xmm2\{k1\},xmm4,XMMWORD PTR \[esp\+esi\*8-0x1e240\] +[ ]*[a-f0-9]+:[ ]*62 f2 5d 09 52 52 7f[ ]*vpdpwssd xmm2\{k1\},xmm4,XMMWORD PTR \[edx\+0x7f0\] +[ ]*[a-f0-9]+:[ ]*62 f2 5d 19 52 52 7f[ ]*vpdpwssd xmm2\{k1\},xmm4,DWORD PTR \[edx\+0x1fc\]\{1to4\} +[ ]*[a-f0-9]+:[ ]*62 f2 65 29 52 d9[ ]*vpdpwssd ymm3\{k1\},ymm3,ymm1 +[ ]*[a-f0-9]+:[ ]*62 f2 65 a9 52 d9[ ]*vpdpwssd ymm3\{k1\}\{z\},ymm3,ymm1 +[ ]*[a-f0-9]+:[ ]*62 f2 65 2c 52 9c f4 c0 1d fe ff[ ]*vpdpwssd ymm3\{k4\},ymm3,YMMWORD PTR \[esp\+esi\*8-0x1e240\] +[ ]*[a-f0-9]+:[ ]*62 f2 65 2c 52 5a 7f[ ]*vpdpwssd ymm3\{k4\},ymm3,YMMWORD PTR \[edx\+0xfe0\] +[ ]*[a-f0-9]+:[ ]*62 f2 65 3c 52 5a 7f[ ]*vpdpwssd ymm3\{k4\},ymm3,DWORD PTR \[edx\+0x1fc\]\{1to8\} +[ ]*[a-f0-9]+:[ ]*62 f2 5d 09 53 d1[ ]*vpdpwssds xmm2\{k1\},xmm4,xmm1 +[ ]*[a-f0-9]+:[ ]*62 f2 5d 89 53 d1[ ]*vpdpwssds xmm2\{k1\}\{z\},xmm4,xmm1 +[ ]*[a-f0-9]+:[ ]*62 f2 5d 0c 53 94 f4 c0 1d fe ff[ ]*vpdpwssds xmm2\{k4\},xmm4,XMMWORD PTR \[esp\+esi\*8-0x1e240\] +[ ]*[a-f0-9]+:[ ]*62 f2 5d 0c 53 52 7f[ ]*vpdpwssds xmm2\{k4\},xmm4,XMMWORD PTR \[edx\+0x7f0\] +[ ]*[a-f0-9]+:[ ]*62 f2 5d 1c 53 52 7f[ ]*vpdpwssds xmm2\{k4\},xmm4,DWORD PTR \[edx\+0x1fc\]\{1to4\} +[ ]*[a-f0-9]+:[ ]*62 f2 75 2f 53 e4[ ]*vpdpwssds ymm4\{k7\},ymm1,ymm4 +[ ]*[a-f0-9]+:[ ]*62 f2 75 af 53 e4[ ]*vpdpwssds ymm4\{k7\}\{z\},ymm1,ymm4 +[ ]*[a-f0-9]+:[ ]*62 f2 75 2b 53 a4 f4 c0 1d fe ff[ ]*vpdpwssds ymm4\{k3\},ymm1,YMMWORD PTR \[esp\+esi\*8-0x1e240\] +[ ]*[a-f0-9]+:[ ]*62 f2 75 2b 53 62 7f[ ]*vpdpwssds ymm4\{k3\},ymm1,YMMWORD PTR \[edx\+0xfe0\] +[ ]*[a-f0-9]+:[ ]*62 f2 75 3b 53 62 7f[ ]*vpdpwssds ymm4\{k3\},ymm1,DWORD PTR \[edx\+0x1fc\]\{1to8\} +[ ]*[a-f0-9]+:[ ]*62 f2 65 0c 50 d1[ ]*vpdpbusd xmm2\{k4\},xmm3,xmm1 +[ ]*[a-f0-9]+:[ ]*62 f2 65 8c 50 d1[ ]*vpdpbusd xmm2\{k4\}\{z\},xmm3,xmm1 +[ ]*[a-f0-9]+:[ ]*62 f2 65 0a 50 94 f4 c0 1d fe ff[ ]*vpdpbusd xmm2\{k2\},xmm3,XMMWORD PTR \[esp\+esi\*8-0x1e240\] +[ ]*[a-f0-9]+:[ ]*62 f2 65 0a 50 52 7f[ ]*vpdpbusd xmm2\{k2\},xmm3,XMMWORD PTR \[edx\+0x7f0\] +[ ]*[a-f0-9]+:[ ]*62 f2 65 1a 50 52 7f[ ]*vpdpbusd xmm2\{k2\},xmm3,DWORD PTR \[edx\+0x1fc\]\{1to4\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 2d 50 d2[ ]*vpdpbusd ymm2\{k5\},ymm2,ymm2 +[ ]*[a-f0-9]+:[ ]*62 f2 6d ad 50 d2[ ]*vpdpbusd ymm2\{k5\}\{z\},ymm2,ymm2 +[ ]*[a-f0-9]+:[ ]*62 f2 6d 2f 50 94 f4 c0 1d fe ff[ ]*vpdpbusd ymm2\{k7\},ymm2,YMMWORD PTR \[esp\+esi\*8-0x1e240\] +[ ]*[a-f0-9]+:[ ]*62 f2 6d 2f 50 52 7f[ ]*vpdpbusd ymm2\{k7\},ymm2,YMMWORD PTR \[edx\+0xfe0\] +[ ]*[a-f0-9]+:[ ]*62 f2 6d 3f 50 52 7f[ ]*vpdpbusd ymm2\{k7\},ymm2,DWORD PTR \[edx\+0x1fc\]\{1to8\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 0e 51 f4[ ]*vpdpbusds xmm6\{k6\},xmm2,xmm4 +[ ]*[a-f0-9]+:[ ]*62 f2 6d 8e 51 f4[ ]*vpdpbusds xmm6\{k6\}\{z\},xmm2,xmm4 +[ ]*[a-f0-9]+:[ ]*62 f2 6d 0c 51 b4 f4 c0 1d fe ff[ ]*vpdpbusds xmm6\{k4\},xmm2,XMMWORD PTR \[esp\+esi\*8-0x1e240\] +[ ]*[a-f0-9]+:[ ]*62 f2 6d 0c 51 72 7f[ ]*vpdpbusds xmm6\{k4\},xmm2,XMMWORD PTR \[edx\+0x7f0\] +[ ]*[a-f0-9]+:[ ]*62 f2 6d 1c 51 72 7f[ ]*vpdpbusds xmm6\{k4\},xmm2,DWORD PTR \[edx\+0x1fc\]\{1to4\} +[ ]*[a-f0-9]+:[ ]*62 f2 65 2f 51 e1[ ]*vpdpbusds ymm4\{k7\},ymm3,ymm1 +[ ]*[a-f0-9]+:[ ]*62 f2 65 af 51 e1[ ]*vpdpbusds ymm4\{k7\}\{z\},ymm3,ymm1 +[ ]*[a-f0-9]+:[ ]*62 f2 65 29 51 a4 f4 c0 1d fe ff[ ]*vpdpbusds ymm4\{k1\},ymm3,YMMWORD PTR \[esp\+esi\*8-0x1e240\] +[ ]*[a-f0-9]+:[ ]*62 f2 65 29 51 62 7f[ ]*vpdpbusds ymm4\{k1\},ymm3,YMMWORD PTR \[edx\+0xfe0\] +[ ]*[a-f0-9]+:[ ]*62 f2 65 39 51 62 7f[ ]*vpdpbusds ymm4\{k1\},ymm3,DWORD PTR \[edx\+0x1fc\]\{1to8\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 09 52 ea[ ]*vpdpwssd xmm5\{k1\},xmm2,xmm2 +[ ]*[a-f0-9]+:[ ]*62 f2 6d 89 52 ea[ ]*vpdpwssd xmm5\{k1\}\{z\},xmm2,xmm2 +[ ]*[a-f0-9]+:[ ]*62 f2 6d 0e 52 ac f4 c0 1d fe ff[ ]*vpdpwssd xmm5\{k6\},xmm2,XMMWORD PTR \[esp\+esi\*8-0x1e240\] +[ ]*[a-f0-9]+:[ ]*62 f2 6d 0e 52 6a 7f[ ]*vpdpwssd xmm5\{k6\},xmm2,XMMWORD PTR \[edx\+0x7f0\] +[ ]*[a-f0-9]+:[ ]*62 f2 6d 1e 52 6a 7f[ ]*vpdpwssd xmm5\{k6\},xmm2,DWORD PTR \[edx\+0x1fc\]\{1to4\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 2f 52 cc[ ]*vpdpwssd ymm1\{k7\},ymm2,ymm4 +[ ]*[a-f0-9]+:[ ]*62 f2 6d af 52 cc[ ]*vpdpwssd ymm1\{k7\}\{z\},ymm2,ymm4 +[ ]*[a-f0-9]+:[ ]*62 f2 6d 2e 52 8c f4 c0 1d fe ff[ ]*vpdpwssd ymm1\{k6\},ymm2,YMMWORD PTR \[esp\+esi\*8-0x1e240\] +[ ]*[a-f0-9]+:[ ]*62 f2 6d 2e 52 4a 7f[ ]*vpdpwssd ymm1\{k6\},ymm2,YMMWORD PTR \[edx\+0xfe0\] +[ ]*[a-f0-9]+:[ ]*62 f2 6d 3e 52 4a 7f[ ]*vpdpwssd ymm1\{k6\},ymm2,DWORD PTR \[edx\+0x1fc\]\{1to8\} +[ ]*[a-f0-9]+:[ ]*62 f2 5d 0a 53 c9[ ]*vpdpwssds xmm1\{k2\},xmm4,xmm1 +[ ]*[a-f0-9]+:[ ]*62 f2 5d 8a 53 c9[ ]*vpdpwssds xmm1\{k2\}\{z\},xmm4,xmm1 +[ ]*[a-f0-9]+:[ ]*62 f2 5d 0e 53 8c f4 c0 1d fe ff[ ]*vpdpwssds xmm1\{k6\},xmm4,XMMWORD PTR \[esp\+esi\*8-0x1e240\] +[ ]*[a-f0-9]+:[ ]*62 f2 5d 0e 53 4a 7f[ ]*vpdpwssds xmm1\{k6\},xmm4,XMMWORD PTR \[edx\+0x7f0\] +[ ]*[a-f0-9]+:[ ]*62 f2 5d 1e 53 4a 7f[ ]*vpdpwssds xmm1\{k6\},xmm4,DWORD PTR \[edx\+0x1fc\]\{1to4\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 2c 53 dc[ ]*vpdpwssds ymm3\{k4\},ymm2,ymm4 +[ ]*[a-f0-9]+:[ ]*62 f2 6d ac 53 dc[ ]*vpdpwssds ymm3\{k4\}\{z\},ymm2,ymm4 +[ ]*[a-f0-9]+:[ ]*62 f2 6d 2d 53 9c f4 c0 1d fe ff[ ]*vpdpwssds ymm3\{k5\},ymm2,YMMWORD PTR \[esp\+esi\*8-0x1e240\] +[ ]*[a-f0-9]+:[ ]*62 f2 6d 2d 53 5a 7f[ ]*vpdpwssds ymm3\{k5\},ymm2,YMMWORD PTR \[edx\+0xfe0\] +[ ]*[a-f0-9]+:[ ]*62 f2 6d 3d 53 5a 7f[ ]*vpdpwssds ymm3\{k5\},ymm2,DWORD PTR \[edx\+0x1fc\]\{1to8\} +[ ]*[a-f0-9]+:[ ]*62 f2 5d 0f 50 dc[ ]*vpdpbusd xmm3\{k7\},xmm4,xmm4 +[ ]*[a-f0-9]+:[ ]*62 f2 5d 8f 50 dc[ ]*vpdpbusd xmm3\{k7\}\{z\},xmm4,xmm4 +[ ]*[a-f0-9]+:[ ]*62 f2 5d 09 50 9c f4 c0 1d fe ff[ ]*vpdpbusd xmm3\{k1\},xmm4,XMMWORD PTR \[esp\+esi\*8-0x1e240\] +[ ]*[a-f0-9]+:[ ]*62 f2 5d 09 50 5a 7f[ ]*vpdpbusd xmm3\{k1\},xmm4,XMMWORD PTR \[edx\+0x7f0\] +[ ]*[a-f0-9]+:[ ]*62 f2 5d 19 50 5a 7f[ ]*vpdpbusd xmm3\{k1\},xmm4,DWORD PTR \[edx\+0x1fc\]\{1to4\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 2d 50 f4[ ]*vpdpbusd ymm6\{k5\},ymm2,ymm4 +[ ]*[a-f0-9]+:[ ]*62 f2 6d ad 50 f4[ ]*vpdpbusd ymm6\{k5\}\{z\},ymm2,ymm4 +[ ]*[a-f0-9]+:[ ]*62 f2 6d 2d 50 b4 f4 c0 1d fe ff[ ]*vpdpbusd ymm6\{k5\},ymm2,YMMWORD PTR \[esp\+esi\*8-0x1e240\] +[ ]*[a-f0-9]+:[ ]*62 f2 6d 2d 50 72 7f[ ]*vpdpbusd ymm6\{k5\},ymm2,YMMWORD PTR \[edx\+0xfe0\] +[ ]*[a-f0-9]+:[ ]*62 f2 6d 3d 50 72 7f[ ]*vpdpbusd ymm6\{k5\},ymm2,DWORD PTR \[edx\+0x1fc\]\{1to8\} +[ ]*[a-f0-9]+:[ ]*62 f2 65 0d 51 dc[ ]*vpdpbusds xmm3\{k5\},xmm3,xmm4 +[ ]*[a-f0-9]+:[ ]*62 f2 65 8d 51 dc[ ]*vpdpbusds xmm3\{k5\}\{z\},xmm3,xmm4 +[ ]*[a-f0-9]+:[ ]*62 f2 65 0c 51 9c f4 c0 1d fe ff[ ]*vpdpbusds xmm3\{k4\},xmm3,XMMWORD PTR \[esp\+esi\*8-0x1e240\] +[ ]*[a-f0-9]+:[ ]*62 f2 65 0c 51 5a 7f[ ]*vpdpbusds xmm3\{k4\},xmm3,XMMWORD PTR \[edx\+0x7f0\] +[ ]*[a-f0-9]+:[ ]*62 f2 65 1c 51 5a 7f[ ]*vpdpbusds xmm3\{k4\},xmm3,DWORD PTR \[edx\+0x1fc\]\{1to4\} +[ ]*[a-f0-9]+:[ ]*62 f2 65 2c 51 d4[ ]*vpdpbusds ymm2\{k4\},ymm3,ymm4 +[ ]*[a-f0-9]+:[ ]*62 f2 65 ac 51 d4[ ]*vpdpbusds ymm2\{k4\}\{z\},ymm3,ymm4 +[ ]*[a-f0-9]+:[ ]*62 f2 65 29 51 94 f4 c0 1d fe ff[ ]*vpdpbusds ymm2\{k1\},ymm3,YMMWORD PTR \[esp\+esi\*8-0x1e240\] +[ ]*[a-f0-9]+:[ ]*62 f2 65 29 51 52 7f[ ]*vpdpbusds ymm2\{k1\},ymm3,YMMWORD PTR \[edx\+0xfe0\] +[ ]*[a-f0-9]+:[ ]*62 f2 65 39 51 52 7f[ ]*vpdpbusds ymm2\{k1\},ymm3,DWORD PTR \[edx\+0x1fc\]\{1to8\} +#pass diff --git a/gas/testsuite/gas/i386/avx512vnni_vl.d b/gas/testsuite/gas/i386/avx512vnni_vl.d new file mode 100644 index 0000000..f4a096e --- /dev/null +++ b/gas/testsuite/gas/i386/avx512vnni_vl.d @@ -0,0 +1,92 @@ +#as: +#objdump: -dw +#name: i386 AVX512VNNI/VL insns +#source: avx512vnni_vl.s + +.*: +file format .* + + +Disassembly of section \.text: + +00000000 <_start>: +[ ]*[a-f0-9]+:[ ]*62 f2 5d 0b 52 d2[ ]*vpdpwssd %xmm2,%xmm4,%xmm2\{%k3\} +[ ]*[a-f0-9]+:[ ]*62 f2 5d 8b 52 d2[ ]*vpdpwssd %xmm2,%xmm4,%xmm2\{%k3\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 f2 5d 09 52 94 f4 c0 1d fe ff[ ]*vpdpwssd -0x1e240\(%esp,%esi,8\),%xmm4,%xmm2\{%k1\} +[ ]*[a-f0-9]+:[ ]*62 f2 5d 09 52 52 7f[ ]*vpdpwssd 0x7f0\(%edx\),%xmm4,%xmm2\{%k1\} +[ ]*[a-f0-9]+:[ ]*62 f2 5d 19 52 52 7f[ ]*vpdpwssd 0x1fc\(%edx\)\{1to4\},%xmm4,%xmm2\{%k1\} +[ ]*[a-f0-9]+:[ ]*62 f2 65 29 52 d9[ ]*vpdpwssd %ymm1,%ymm3,%ymm3\{%k1\} +[ ]*[a-f0-9]+:[ ]*62 f2 65 a9 52 d9[ ]*vpdpwssd %ymm1,%ymm3,%ymm3\{%k1\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 f2 65 2c 52 9c f4 c0 1d fe ff[ ]*vpdpwssd -0x1e240\(%esp,%esi,8\),%ymm3,%ymm3\{%k4\} +[ ]*[a-f0-9]+:[ ]*62 f2 65 2c 52 5a 7f[ ]*vpdpwssd 0xfe0\(%edx\),%ymm3,%ymm3\{%k4\} +[ ]*[a-f0-9]+:[ ]*62 f2 65 3c 52 5a 7f[ ]*vpdpwssd 0x1fc\(%edx\)\{1to8\},%ymm3,%ymm3\{%k4\} +[ ]*[a-f0-9]+:[ ]*62 f2 5d 09 53 d1[ ]*vpdpwssds %xmm1,%xmm4,%xmm2\{%k1\} +[ ]*[a-f0-9]+:[ ]*62 f2 5d 89 53 d1[ ]*vpdpwssds %xmm1,%xmm4,%xmm2\{%k1\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 f2 5d 0c 53 94 f4 c0 1d fe ff[ ]*vpdpwssds -0x1e240\(%esp,%esi,8\),%xmm4,%xmm2\{%k4\} +[ ]*[a-f0-9]+:[ ]*62 f2 5d 0c 53 52 7f[ ]*vpdpwssds 0x7f0\(%edx\),%xmm4,%xmm2\{%k4\} +[ ]*[a-f0-9]+:[ ]*62 f2 5d 1c 53 52 7f[ ]*vpdpwssds 0x1fc\(%edx\)\{1to4\},%xmm4,%xmm2\{%k4\} +[ ]*[a-f0-9]+:[ ]*62 f2 75 2f 53 e4[ ]*vpdpwssds %ymm4,%ymm1,%ymm4\{%k7\} +[ ]*[a-f0-9]+:[ ]*62 f2 75 af 53 e4[ ]*vpdpwssds %ymm4,%ymm1,%ymm4\{%k7\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 f2 75 2b 53 a4 f4 c0 1d fe ff[ ]*vpdpwssds -0x1e240\(%esp,%esi,8\),%ymm1,%ymm4\{%k3\} +[ ]*[a-f0-9]+:[ ]*62 f2 75 2b 53 62 7f[ ]*vpdpwssds 0xfe0\(%edx\),%ymm1,%ymm4\{%k3\} +[ ]*[a-f0-9]+:[ ]*62 f2 75 3b 53 62 7f[ ]*vpdpwssds 0x1fc\(%edx\)\{1to8\},%ymm1,%ymm4\{%k3\} +[ ]*[a-f0-9]+:[ ]*62 f2 65 0c 50 d1[ ]*vpdpbusd %xmm1,%xmm3,%xmm2\{%k4\} +[ ]*[a-f0-9]+:[ ]*62 f2 65 8c 50 d1[ ]*vpdpbusd %xmm1,%xmm3,%xmm2\{%k4\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 f2 65 0a 50 94 f4 c0 1d fe ff[ ]*vpdpbusd -0x1e240\(%esp,%esi,8\),%xmm3,%xmm2\{%k2\} +[ ]*[a-f0-9]+:[ ]*62 f2 65 0a 50 52 7f[ ]*vpdpbusd 0x7f0\(%edx\),%xmm3,%xmm2\{%k2\} +[ ]*[a-f0-9]+:[ ]*62 f2 65 1a 50 52 7f[ ]*vpdpbusd 0x1fc\(%edx\)\{1to4\},%xmm3,%xmm2\{%k2\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 2d 50 d2[ ]*vpdpbusd %ymm2,%ymm2,%ymm2\{%k5\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d ad 50 d2[ ]*vpdpbusd %ymm2,%ymm2,%ymm2\{%k5\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 2f 50 94 f4 c0 1d fe ff[ ]*vpdpbusd -0x1e240\(%esp,%esi,8\),%ymm2,%ymm2\{%k7\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 2f 50 52 7f[ ]*vpdpbusd 0xfe0\(%edx\),%ymm2,%ymm2\{%k7\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 3f 50 52 7f[ ]*vpdpbusd 0x1fc\(%edx\)\{1to8\},%ymm2,%ymm2\{%k7\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 0e 51 f4[ ]*vpdpbusds %xmm4,%xmm2,%xmm6\{%k6\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 8e 51 f4[ ]*vpdpbusds %xmm4,%xmm2,%xmm6\{%k6\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 0c 51 b4 f4 c0 1d fe ff[ ]*vpdpbusds -0x1e240\(%esp,%esi,8\),%xmm2,%xmm6\{%k4\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 0c 51 72 7f[ ]*vpdpbusds 0x7f0\(%edx\),%xmm2,%xmm6\{%k4\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 1c 51 72 7f[ ]*vpdpbusds 0x1fc\(%edx\)\{1to4\},%xmm2,%xmm6\{%k4\} +[ ]*[a-f0-9]+:[ ]*62 f2 65 2f 51 e1[ ]*vpdpbusds %ymm1,%ymm3,%ymm4\{%k7\} +[ ]*[a-f0-9]+:[ ]*62 f2 65 af 51 e1[ ]*vpdpbusds %ymm1,%ymm3,%ymm4\{%k7\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 f2 65 29 51 a4 f4 c0 1d fe ff[ ]*vpdpbusds -0x1e240\(%esp,%esi,8\),%ymm3,%ymm4\{%k1\} +[ ]*[a-f0-9]+:[ ]*62 f2 65 29 51 62 7f[ ]*vpdpbusds 0xfe0\(%edx\),%ymm3,%ymm4\{%k1\} +[ ]*[a-f0-9]+:[ ]*62 f2 65 39 51 62 7f[ ]*vpdpbusds 0x1fc\(%edx\)\{1to8\},%ymm3,%ymm4\{%k1\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 09 52 ea[ ]*vpdpwssd %xmm2,%xmm2,%xmm5\{%k1\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 89 52 ea[ ]*vpdpwssd %xmm2,%xmm2,%xmm5\{%k1\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 0e 52 ac f4 c0 1d fe ff[ ]*vpdpwssd -0x1e240\(%esp,%esi,8\),%xmm2,%xmm5\{%k6\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 0e 52 6a 7f[ ]*vpdpwssd 0x7f0\(%edx\),%xmm2,%xmm5\{%k6\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 1e 52 6a 7f[ ]*vpdpwssd 0x1fc\(%edx\)\{1to4\},%xmm2,%xmm5\{%k6\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 2f 52 cc[ ]*vpdpwssd %ymm4,%ymm2,%ymm1\{%k7\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d af 52 cc[ ]*vpdpwssd %ymm4,%ymm2,%ymm1\{%k7\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 2e 52 8c f4 c0 1d fe ff[ ]*vpdpwssd -0x1e240\(%esp,%esi,8\),%ymm2,%ymm1\{%k6\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 2e 52 4a 7f[ ]*vpdpwssd 0xfe0\(%edx\),%ymm2,%ymm1\{%k6\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 3e 52 4a 7f[ ]*vpdpwssd 0x1fc\(%edx\)\{1to8\},%ymm2,%ymm1\{%k6\} +[ ]*[a-f0-9]+:[ ]*62 f2 5d 0a 53 c9[ ]*vpdpwssds %xmm1,%xmm4,%xmm1\{%k2\} +[ ]*[a-f0-9]+:[ ]*62 f2 5d 8a 53 c9[ ]*vpdpwssds %xmm1,%xmm4,%xmm1\{%k2\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 f2 5d 0e 53 8c f4 c0 1d fe ff[ ]*vpdpwssds -0x1e240\(%esp,%esi,8\),%xmm4,%xmm1\{%k6\} +[ ]*[a-f0-9]+:[ ]*62 f2 5d 0e 53 4a 7f[ ]*vpdpwssds 0x7f0\(%edx\),%xmm4,%xmm1\{%k6\} +[ ]*[a-f0-9]+:[ ]*62 f2 5d 1e 53 4a 7f[ ]*vpdpwssds 0x1fc\(%edx\)\{1to4\},%xmm4,%xmm1\{%k6\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 2c 53 dc[ ]*vpdpwssds %ymm4,%ymm2,%ymm3\{%k4\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d ac 53 dc[ ]*vpdpwssds %ymm4,%ymm2,%ymm3\{%k4\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 2d 53 9c f4 c0 1d fe ff[ ]*vpdpwssds -0x1e240\(%esp,%esi,8\),%ymm2,%ymm3\{%k5\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 2d 53 5a 7f[ ]*vpdpwssds 0xfe0\(%edx\),%ymm2,%ymm3\{%k5\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 3d 53 5a 7f[ ]*vpdpwssds 0x1fc\(%edx\)\{1to8\},%ymm2,%ymm3\{%k5\} +[ ]*[a-f0-9]+:[ ]*62 f2 5d 0f 50 dc[ ]*vpdpbusd %xmm4,%xmm4,%xmm3\{%k7\} +[ ]*[a-f0-9]+:[ ]*62 f2 5d 8f 50 dc[ ]*vpdpbusd %xmm4,%xmm4,%xmm3\{%k7\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 f2 5d 09 50 9c f4 c0 1d fe ff[ ]*vpdpbusd -0x1e240\(%esp,%esi,8\),%xmm4,%xmm3\{%k1\} +[ ]*[a-f0-9]+:[ ]*62 f2 5d 09 50 5a 7f[ ]*vpdpbusd 0x7f0\(%edx\),%xmm4,%xmm3\{%k1\} +[ ]*[a-f0-9]+:[ ]*62 f2 5d 19 50 5a 7f[ ]*vpdpbusd 0x1fc\(%edx\)\{1to4\},%xmm4,%xmm3\{%k1\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 2d 50 f4[ ]*vpdpbusd %ymm4,%ymm2,%ymm6\{%k5\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d ad 50 f4[ ]*vpdpbusd %ymm4,%ymm2,%ymm6\{%k5\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 2d 50 b4 f4 c0 1d fe ff[ ]*vpdpbusd -0x1e240\(%esp,%esi,8\),%ymm2,%ymm6\{%k5\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 2d 50 72 7f[ ]*vpdpbusd 0xfe0\(%edx\),%ymm2,%ymm6\{%k5\} +[ ]*[a-f0-9]+:[ ]*62 f2 6d 3d 50 72 7f[ ]*vpdpbusd 0x1fc\(%edx\)\{1to8\},%ymm2,%ymm6\{%k5\} +[ ]*[a-f0-9]+:[ ]*62 f2 65 0d 51 dc[ ]*vpdpbusds %xmm4,%xmm3,%xmm3\{%k5\} +[ ]*[a-f0-9]+:[ ]*62 f2 65 8d 51 dc[ ]*vpdpbusds %xmm4,%xmm3,%xmm3\{%k5\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 f2 65 0c 51 9c f4 c0 1d fe ff[ ]*vpdpbusds -0x1e240\(%esp,%esi,8\),%xmm3,%xmm3\{%k4\} +[ ]*[a-f0-9]+:[ ]*62 f2 65 0c 51 5a 7f[ ]*vpdpbusds 0x7f0\(%edx\),%xmm3,%xmm3\{%k4\} +[ ]*[a-f0-9]+:[ ]*62 f2 65 1c 51 5a 7f[ ]*vpdpbusds 0x1fc\(%edx\)\{1to4\},%xmm3,%xmm3\{%k4\} +[ ]*[a-f0-9]+:[ ]*62 f2 65 2c 51 d4[ ]*vpdpbusds %ymm4,%ymm3,%ymm2\{%k4\} +[ ]*[a-f0-9]+:[ ]*62 f2 65 ac 51 d4[ ]*vpdpbusds %ymm4,%ymm3,%ymm2\{%k4\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 f2 65 29 51 94 f4 c0 1d fe ff[ ]*vpdpbusds -0x1e240\(%esp,%esi,8\),%ymm3,%ymm2\{%k1\} +[ ]*[a-f0-9]+:[ ]*62 f2 65 29 51 52 7f[ ]*vpdpbusds 0xfe0\(%edx\),%ymm3,%ymm2\{%k1\} +[ ]*[a-f0-9]+:[ ]*62 f2 65 39 51 52 7f[ ]*vpdpbusds 0x1fc\(%edx\)\{1to8\},%ymm3,%ymm2\{%k1\} +#pass diff --git a/gas/testsuite/gas/i386/avx512vnni_vl.s b/gas/testsuite/gas/i386/avx512vnni_vl.s new file mode 100644 index 0000000..f61e179 --- /dev/null +++ b/gas/testsuite/gas/i386/avx512vnni_vl.s @@ -0,0 +1,93 @@ +# Check 32bit AVX512{VNNI,VL} instructions + + .allow_index_reg + .text +_start: + vpdpwssd %xmm2, %xmm4, %xmm2{%k3} # AVX512{VNNI,VL} + vpdpwssd %xmm2, %xmm4, %xmm2{%k3}{z} # AVX512{VNNI,VL} + vpdpwssd -123456(%esp,%esi,8), %xmm4, %xmm2{%k1} # AVX512{VNNI,VL} + vpdpwssd 2032(%edx), %xmm4, %xmm2{%k1} # AVX512{VNNI,VL} Disp8 + vpdpwssd 508(%edx){1to4}, %xmm4, %xmm2{%k1} # AVX512{VNNI,VL} Disp8 + vpdpwssd %ymm1, %ymm3, %ymm3{%k1} # AVX512{VNNI,VL} + vpdpwssd %ymm1, %ymm3, %ymm3{%k1}{z} # AVX512{VNNI,VL} + vpdpwssd -123456(%esp,%esi,8), %ymm3, %ymm3{%k4} # AVX512{VNNI,VL} + vpdpwssd 4064(%edx), %ymm3, %ymm3{%k4} # AVX512{VNNI,VL} Disp8 + vpdpwssd 508(%edx){1to8}, %ymm3, %ymm3{%k4} # AVX512{VNNI,VL} Disp8 + + vpdpwssds %xmm1, %xmm4, %xmm2{%k1} # AVX512{VNNI,VL} + vpdpwssds %xmm1, %xmm4, %xmm2{%k1}{z} # AVX512{VNNI,VL} + vpdpwssds -123456(%esp,%esi,8), %xmm4, %xmm2{%k4} # AVX512{VNNI,VL} + vpdpwssds 2032(%edx), %xmm4, %xmm2{%k4} # AVX512{VNNI,VL} Disp8 + vpdpwssds 508(%edx){1to4}, %xmm4, %xmm2{%k4} # AVX512{VNNI,VL} Disp8 + vpdpwssds %ymm4, %ymm1, %ymm4{%k7} # AVX512{VNNI,VL} + vpdpwssds %ymm4, %ymm1, %ymm4{%k7}{z} # AVX512{VNNI,VL} + vpdpwssds -123456(%esp,%esi,8), %ymm1, %ymm4{%k3} # AVX512{VNNI,VL} + vpdpwssds 4064(%edx), %ymm1, %ymm4{%k3} # AVX512{VNNI,VL} Disp8 + vpdpwssds 508(%edx){1to8}, %ymm1, %ymm4{%k3} # AVX512{VNNI,VL} Disp8 + + vpdpbusd %xmm1, %xmm3, %xmm2{%k4} # AVX512{VNNI,VL} + vpdpbusd %xmm1, %xmm3, %xmm2{%k4}{z} # AVX512{VNNI,VL} + vpdpbusd -123456(%esp,%esi,8), %xmm3, %xmm2{%k2} # AVX512{VNNI,VL} + vpdpbusd 2032(%edx), %xmm3, %xmm2{%k2} # AVX512{VNNI,VL} Disp8 + vpdpbusd 508(%edx){1to4}, %xmm3, %xmm2{%k2} # AVX512{VNNI,VL} Disp8 + vpdpbusd %ymm2, %ymm2, %ymm2{%k5} # AVX512{VNNI,VL} + vpdpbusd %ymm2, %ymm2, %ymm2{%k5}{z} # AVX512{VNNI,VL} + vpdpbusd -123456(%esp,%esi,8), %ymm2, %ymm2{%k7} # AVX512{VNNI,VL} + vpdpbusd 4064(%edx), %ymm2, %ymm2{%k7} # AVX512{VNNI,VL} Disp8 + vpdpbusd 508(%edx){1to8}, %ymm2, %ymm2{%k7} # AVX512{VNNI,VL} Disp8 + + vpdpbusds %xmm4, %xmm2, %xmm6{%k6} # AVX512{VNNI,VL} + vpdpbusds %xmm4, %xmm2, %xmm6{%k6}{z} # AVX512{VNNI,VL} + vpdpbusds -123456(%esp,%esi,8), %xmm2, %xmm6{%k4} # AVX512{VNNI,VL} + vpdpbusds 2032(%edx), %xmm2, %xmm6{%k4} # AVX512{VNNI,VL} Disp8 + vpdpbusds 508(%edx){1to4}, %xmm2, %xmm6{%k4} # AVX512{VNNI,VL} Disp8 + vpdpbusds %ymm1, %ymm3, %ymm4{%k7} # AVX512{VNNI,VL} + vpdpbusds %ymm1, %ymm3, %ymm4{%k7}{z} # AVX512{VNNI,VL} + vpdpbusds -123456(%esp,%esi,8), %ymm3, %ymm4{%k1} # AVX512{VNNI,VL} + vpdpbusds 4064(%edx), %ymm3, %ymm4{%k1} # AVX512{VNNI,VL} Disp8 + vpdpbusds 508(%edx){1to8}, %ymm3, %ymm4{%k1} # AVX512{VNNI,VL} Disp8 + + .intel_syntax noprefix + vpdpwssd xmm5{k1}, xmm2, xmm2 # AVX512{VNNI,VL} + vpdpwssd xmm5{k1}{z}, xmm2, xmm2 # AVX512{VNNI,VL} + vpdpwssd xmm5{k6}, xmm2, XMMWORD PTR [esp+esi*8-123456] # AVX512{VNNI,VL} + vpdpwssd xmm5{k6}, xmm2, XMMWORD PTR [edx+2032] # AVX512{VNNI,VL} Disp8 + vpdpwssd xmm5{k6}, xmm2, [edx+508]{1to4} # AVX512{VNNI,VL} Disp8 + vpdpwssd ymm1{k7}, ymm2, ymm4 # AVX512{VNNI,VL} + vpdpwssd ymm1{k7}{z}, ymm2, ymm4 # AVX512{VNNI,VL} + vpdpwssd ymm1{k6}, ymm2, YMMWORD PTR [esp+esi*8-123456] # AVX512{VNNI,VL} + vpdpwssd ymm1{k6}, ymm2, YMMWORD PTR [edx+4064] # AVX512{VNNI,VL} Disp8 + vpdpwssd ymm1{k6}, ymm2, [edx+508]{1to8} # AVX512{VNNI,VL} Disp8 + + vpdpwssds xmm1{k2}, xmm4, xmm1 # AVX512{VNNI,VL} + vpdpwssds xmm1{k2}{z}, xmm4, xmm1 # AVX512{VNNI,VL} + vpdpwssds xmm1{k6}, xmm4, XMMWORD PTR [esp+esi*8-123456] # AVX512{VNNI,VL} + vpdpwssds xmm1{k6}, xmm4, XMMWORD PTR [edx+2032] # AVX512{VNNI,VL} Disp8 + vpdpwssds xmm1{k6}, xmm4, [edx+508]{1to4} # AVX512{VNNI,VL} Disp8 + vpdpwssds ymm3{k4}, ymm2, ymm4 # AVX512{VNNI,VL} + vpdpwssds ymm3{k4}{z}, ymm2, ymm4 # AVX512{VNNI,VL} + vpdpwssds ymm3{k5}, ymm2, YMMWORD PTR [esp+esi*8-123456] # AVX512{VNNI,VL} + vpdpwssds ymm3{k5}, ymm2, YMMWORD PTR [edx+4064] # AVX512{VNNI,VL} Disp8 + vpdpwssds ymm3{k5}, ymm2, [edx+508]{1to8} # AVX512{VNNI,VL} Disp8 + + vpdpbusd xmm3{k7}, xmm4, xmm4 # AVX512{VNNI,VL} + vpdpbusd xmm3{k7}{z}, xmm4, xmm4 # AVX512{VNNI,VL} + vpdpbusd xmm3{k1}, xmm4, XMMWORD PTR [esp+esi*8-123456] # AVX512{VNNI,VL} + vpdpbusd xmm3{k1}, xmm4, XMMWORD PTR [edx+2032] # AVX512{VNNI,VL} Disp8 + vpdpbusd xmm3{k1}, xmm4, [edx+508]{1to4} # AVX512{VNNI,VL} Disp8 + vpdpbusd ymm6{k5}, ymm2, ymm4 # AVX512{VNNI,VL} + vpdpbusd ymm6{k5}{z}, ymm2, ymm4 # AVX512{VNNI,VL} + vpdpbusd ymm6{k5}, ymm2, YMMWORD PTR [esp+esi*8-123456] # AVX512{VNNI,VL} + vpdpbusd ymm6{k5}, ymm2, YMMWORD PTR [edx+4064] # AVX512{VNNI,VL} Disp8 + vpdpbusd ymm6{k5}, ymm2, [edx+508]{1to8} # AVX512{VNNI,VL} Disp8 + + vpdpbusds xmm3{k5}, xmm3, xmm4 # AVX512{VNNI,VL} + vpdpbusds xmm3{k5}{z}, xmm3, xmm4 # AVX512{VNNI,VL} + vpdpbusds xmm3{k4}, xmm3, XMMWORD PTR [esp+esi*8-123456] # AVX512{VNNI,VL} + vpdpbusds xmm3{k4}, xmm3, XMMWORD PTR [edx+2032] # AVX512{VNNI,VL} Disp8 + vpdpbusds xmm3{k4}, xmm3, [edx+508]{1to4} # AVX512{VNNI,VL} Disp8 + vpdpbusds ymm2{k4}, ymm3, ymm4 # AVX512{VNNI,VL} + vpdpbusds ymm2{k4}{z}, ymm3, ymm4 # AVX512{VNNI,VL} + vpdpbusds ymm2{k1}, ymm3, YMMWORD PTR [esp+esi*8-123456] # AVX512{VNNI,VL} + vpdpbusds ymm2{k1}, ymm3, YMMWORD PTR [edx+4064] # AVX512{VNNI,VL} Disp8 + vpdpbusds ymm2{k1}, ymm3, [edx+508]{1to8} # AVX512{VNNI,VL} Disp8 diff --git a/gas/testsuite/gas/i386/i386.exp b/gas/testsuite/gas/i386/i386.exp index 78c3562..0c79bbf 100644 --- a/gas/testsuite/gas/i386/i386.exp +++ b/gas/testsuite/gas/i386/i386.exp @@ -394,6 +394,10 @@ if [expr ([istarget "i*86-*-*"] || [istarget "x86_64-*-*"]) && [gas_32_check]] run_dump_test "avx512vl_vpclmulqdq-intel" run_dump_test "avx512vl_vpclmulqdq-wig1" run_dump_test "avx512vl_vpclmulqdq-wig1-intel" + run_dump_test "avx512vnni" + run_dump_test "avx512vnni-intel" + run_dump_test "avx512vnni_vl" + run_dump_test "avx512vnni_vl-intel" run_dump_test "clzero" run_dump_test "disassem" run_dump_test "mwaitx-bdver4" @@ -847,6 +851,10 @@ if [expr ([istarget "i*86-*-*"] || [istarget "x86_64-*-*"]) && [gas_64_check]] t run_dump_test "x86-64-avx512vl_vpclmulqdq-intel" run_dump_test "x86-64-avx512vl_vpclmulqdq-wig1" run_dump_test "x86-64-avx512vl_vpclmulqdq-wig1-intel" + run_dump_test "x86-64-avx512vnni" + run_dump_test "x86-64-avx512vnni-intel" + run_dump_test "x86-64-avx512vnni_vl" + run_dump_test "x86-64-avx512vnni_vl-intel" run_dump_test "x86-64-clzero" run_dump_test "x86-64-mwaitx-bdver4" run_list_test "x86-64-mwaitx-reg" diff --git a/gas/testsuite/gas/i386/x86-64-avx512vnni-intel.d b/gas/testsuite/gas/i386/x86-64-avx512vnni-intel.d new file mode 100644 index 0000000..2a721e7 --- /dev/null +++ b/gas/testsuite/gas/i386/x86-64-avx512vnni-intel.d @@ -0,0 +1,60 @@ +#as: +#objdump: -dw -Mintel +#name: x86_64 AVX512VNNI insns (Intel disassembly) +#source: x86-64-avx512vnni.s + +.*: +file format .* + + +Disassembly of section \.text: + +0+ <_start>: +[ ]*[a-f0-9]+:[ ]*62 a2 6d 40 52 d1[ ]*vpdpwssd zmm18,zmm18,zmm17 +[ ]*[a-f0-9]+:[ ]*62 a2 6d 45 52 d1[ ]*vpdpwssd zmm18\{k5\},zmm18,zmm17 +[ ]*[a-f0-9]+:[ ]*62 a2 6d c5 52 d1[ ]*vpdpwssd zmm18\{k5\}\{z\},zmm18,zmm17 +[ ]*[a-f0-9]+:[ ]*62 a2 6d 40 52 94 f0 23 01 00 00[ ]*vpdpwssd zmm18,zmm18,ZMMWORD PTR \[rax\+r14\*8\+0x123\] +[ ]*[a-f0-9]+:[ ]*62 e2 6d 40 52 52 7f[ ]*vpdpwssd zmm18,zmm18,ZMMWORD PTR \[rdx\+0x1fc0\] +[ ]*[a-f0-9]+:[ ]*62 e2 6d 50 52 52 7f[ ]*vpdpwssd zmm18,zmm18,DWORD PTR \[rdx\+0x1fc\]\{1to16\} +[ ]*[a-f0-9]+:[ ]*62 a2 55 40 53 e9[ ]*vpdpwssds zmm21,zmm21,zmm17 +[ ]*[a-f0-9]+:[ ]*62 a2 55 44 53 e9[ ]*vpdpwssds zmm21\{k4\},zmm21,zmm17 +[ ]*[a-f0-9]+:[ ]*62 a2 55 c4 53 e9[ ]*vpdpwssds zmm21\{k4\}\{z\},zmm21,zmm17 +[ ]*[a-f0-9]+:[ ]*62 a2 55 40 53 ac f0 23 01 00 00[ ]*vpdpwssds zmm21,zmm21,ZMMWORD PTR \[rax\+r14\*8\+0x123\] +[ ]*[a-f0-9]+:[ ]*62 e2 55 40 53 6a 7f[ ]*vpdpwssds zmm21,zmm21,ZMMWORD PTR \[rdx\+0x1fc0\] +[ ]*[a-f0-9]+:[ ]*62 e2 55 50 53 6a 7f[ ]*vpdpwssds zmm21,zmm21,DWORD PTR \[rdx\+0x1fc\]\{1to16\} +[ ]*[a-f0-9]+:[ ]*62 a2 55 40 50 fa[ ]*vpdpbusd zmm23,zmm21,zmm18 +[ ]*[a-f0-9]+:[ ]*62 a2 55 44 50 fa[ ]*vpdpbusd zmm23\{k4\},zmm21,zmm18 +[ ]*[a-f0-9]+:[ ]*62 a2 55 c4 50 fa[ ]*vpdpbusd zmm23\{k4\}\{z\},zmm21,zmm18 +[ ]*[a-f0-9]+:[ ]*62 a2 55 40 50 bc f0 23 01 00 00[ ]*vpdpbusd zmm23,zmm21,ZMMWORD PTR \[rax\+r14\*8\+0x123\] +[ ]*[a-f0-9]+:[ ]*62 e2 55 40 50 7a 7f[ ]*vpdpbusd zmm23,zmm21,ZMMWORD PTR \[rdx\+0x1fc0\] +[ ]*[a-f0-9]+:[ ]*62 e2 55 50 50 7a 7f[ ]*vpdpbusd zmm23,zmm21,DWORD PTR \[rdx\+0x1fc\]\{1to16\} +[ ]*[a-f0-9]+:[ ]*62 02 3d 40 51 c1[ ]*vpdpbusds zmm24,zmm24,zmm25 +[ ]*[a-f0-9]+:[ ]*62 02 3d 47 51 c1[ ]*vpdpbusds zmm24\{k7\},zmm24,zmm25 +[ ]*[a-f0-9]+:[ ]*62 02 3d c7 51 c1[ ]*vpdpbusds zmm24\{k7\}\{z\},zmm24,zmm25 +[ ]*[a-f0-9]+:[ ]*62 22 3d 40 51 84 f0 23 01 00 00[ ]*vpdpbusds zmm24,zmm24,ZMMWORD PTR \[rax\+r14\*8\+0x123\] +[ ]*[a-f0-9]+:[ ]*62 62 3d 40 51 42 7f[ ]*vpdpbusds zmm24,zmm24,ZMMWORD PTR \[rdx\+0x1fc0\] +[ ]*[a-f0-9]+:[ ]*62 62 3d 50 51 42 7f[ ]*vpdpbusds zmm24,zmm24,DWORD PTR \[rdx\+0x1fc\]\{1to16\} +[ ]*[a-f0-9]+:[ ]*62 22 25 40 52 e1[ ]*vpdpwssd zmm28,zmm27,zmm17 +[ ]*[a-f0-9]+:[ ]*62 22 25 47 52 e1[ ]*vpdpwssd zmm28\{k7\},zmm27,zmm17 +[ ]*[a-f0-9]+:[ ]*62 22 25 c7 52 e1[ ]*vpdpwssd zmm28\{k7\}\{z\},zmm27,zmm17 +[ ]*[a-f0-9]+:[ ]*62 22 25 40 52 a4 f0 34 12 00 00[ ]*vpdpwssd zmm28,zmm27,ZMMWORD PTR \[rax\+r14\*8\+0x1234\] +[ ]*[a-f0-9]+:[ ]*62 62 25 40 52 62 7f[ ]*vpdpwssd zmm28,zmm27,ZMMWORD PTR \[rdx\+0x1fc0\] +[ ]*[a-f0-9]+:[ ]*62 62 25 50 52 62 7f[ ]*vpdpwssd zmm28,zmm27,DWORD PTR \[rdx\+0x1fc\]\{1to16\} +[ ]*[a-f0-9]+:[ ]*62 22 1d 40 53 e9[ ]*vpdpwssds zmm29,zmm28,zmm17 +[ ]*[a-f0-9]+:[ ]*62 22 1d 43 53 e9[ ]*vpdpwssds zmm29\{k3\},zmm28,zmm17 +[ ]*[a-f0-9]+:[ ]*62 22 1d c3 53 e9[ ]*vpdpwssds zmm29\{k3\}\{z\},zmm28,zmm17 +[ ]*[a-f0-9]+:[ ]*62 22 1d 40 53 ac f0 34 12 00 00[ ]*vpdpwssds zmm29,zmm28,ZMMWORD PTR \[rax\+r14\*8\+0x1234\] +[ ]*[a-f0-9]+:[ ]*62 62 1d 40 53 6a 7f[ ]*vpdpwssds zmm29,zmm28,ZMMWORD PTR \[rdx\+0x1fc0\] +[ ]*[a-f0-9]+:[ ]*62 62 1d 50 53 6a 7f[ ]*vpdpwssds zmm29,zmm28,DWORD PTR \[rdx\+0x1fc\]\{1to16\} +[ ]*[a-f0-9]+:[ ]*62 22 3d 40 50 e5[ ]*vpdpbusd zmm28,zmm24,zmm21 +[ ]*[a-f0-9]+:[ ]*62 22 3d 46 50 e5[ ]*vpdpbusd zmm28\{k6\},zmm24,zmm21 +[ ]*[a-f0-9]+:[ ]*62 22 3d c6 50 e5[ ]*vpdpbusd zmm28\{k6\}\{z\},zmm24,zmm21 +[ ]*[a-f0-9]+:[ ]*62 22 3d 40 50 a4 f0 34 12 00 00[ ]*vpdpbusd zmm28,zmm24,ZMMWORD PTR \[rax\+r14\*8\+0x1234\] +[ ]*[a-f0-9]+:[ ]*62 62 3d 40 50 62 7f[ ]*vpdpbusd zmm28,zmm24,ZMMWORD PTR \[rdx\+0x1fc0\] +[ ]*[a-f0-9]+:[ ]*62 62 3d 50 50 62 7f[ ]*vpdpbusd zmm28,zmm24,DWORD PTR \[rdx\+0x1fc\]\{1to16\} +[ ]*[a-f0-9]+:[ ]*62 a2 75 40 51 e4[ ]*vpdpbusds zmm20,zmm17,zmm20 +[ ]*[a-f0-9]+:[ ]*62 a2 75 42 51 e4[ ]*vpdpbusds zmm20\{k2\},zmm17,zmm20 +[ ]*[a-f0-9]+:[ ]*62 a2 75 c2 51 e4[ ]*vpdpbusds zmm20\{k2\}\{z\},zmm17,zmm20 +[ ]*[a-f0-9]+:[ ]*62 a2 75 40 51 a4 f0 34 12 00 00[ ]*vpdpbusds zmm20,zmm17,ZMMWORD PTR \[rax\+r14\*8\+0x1234\] +[ ]*[a-f0-9]+:[ ]*62 e2 75 40 51 62 7f[ ]*vpdpbusds zmm20,zmm17,ZMMWORD PTR \[rdx\+0x1fc0\] +[ ]*[a-f0-9]+:[ ]*62 e2 75 50 51 62 7f[ ]*vpdpbusds zmm20,zmm17,DWORD PTR \[rdx\+0x1fc\]\{1to16\} +#pass diff --git a/gas/testsuite/gas/i386/x86-64-avx512vnni.d b/gas/testsuite/gas/i386/x86-64-avx512vnni.d new file mode 100644 index 0000000..bd164bf --- /dev/null +++ b/gas/testsuite/gas/i386/x86-64-avx512vnni.d @@ -0,0 +1,60 @@ +#as: +#objdump: -dw +#name: x86_64 AVX512VNNI insns +#source: x86-64-avx512vnni.s + +.*: +file format .* + + +Disassembly of section \.text: + +0+ <_start>: +[ ]*[a-f0-9]+:[ ]*62 a2 6d 40 52 d1[ ]*vpdpwssd %zmm17,%zmm18,%zmm18 +[ ]*[a-f0-9]+:[ ]*62 a2 6d 45 52 d1[ ]*vpdpwssd %zmm17,%zmm18,%zmm18\{%k5\} +[ ]*[a-f0-9]+:[ ]*62 a2 6d c5 52 d1[ ]*vpdpwssd %zmm17,%zmm18,%zmm18\{%k5\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 a2 6d 40 52 94 f0 23 01 00 00[ ]*vpdpwssd 0x123\(%rax,%r14,8\),%zmm18,%zmm18 +[ ]*[a-f0-9]+:[ ]*62 e2 6d 40 52 52 7f[ ]*vpdpwssd 0x1fc0\(%rdx\),%zmm18,%zmm18 +[ ]*[a-f0-9]+:[ ]*62 e2 6d 50 52 52 7f[ ]*vpdpwssd 0x1fc\(%rdx\)\{1to16\},%zmm18,%zmm18 +[ ]*[a-f0-9]+:[ ]*62 a2 55 40 53 e9[ ]*vpdpwssds %zmm17,%zmm21,%zmm21 +[ ]*[a-f0-9]+:[ ]*62 a2 55 44 53 e9[ ]*vpdpwssds %zmm17,%zmm21,%zmm21\{%k4\} +[ ]*[a-f0-9]+:[ ]*62 a2 55 c4 53 e9[ ]*vpdpwssds %zmm17,%zmm21,%zmm21\{%k4\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 a2 55 40 53 ac f0 23 01 00 00[ ]*vpdpwssds 0x123\(%rax,%r14,8\),%zmm21,%zmm21 +[ ]*[a-f0-9]+:[ ]*62 e2 55 40 53 6a 7f[ ]*vpdpwssds 0x1fc0\(%rdx\),%zmm21,%zmm21 +[ ]*[a-f0-9]+:[ ]*62 e2 55 50 53 6a 7f[ ]*vpdpwssds 0x1fc\(%rdx\)\{1to16\},%zmm21,%zmm21 +[ ]*[a-f0-9]+:[ ]*62 a2 55 40 50 fa[ ]*vpdpbusd %zmm18,%zmm21,%zmm23 +[ ]*[a-f0-9]+:[ ]*62 a2 55 44 50 fa[ ]*vpdpbusd %zmm18,%zmm21,%zmm23\{%k4\} +[ ]*[a-f0-9]+:[ ]*62 a2 55 c4 50 fa[ ]*vpdpbusd %zmm18,%zmm21,%zmm23\{%k4\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 a2 55 40 50 bc f0 23 01 00 00[ ]*vpdpbusd 0x123\(%rax,%r14,8\),%zmm21,%zmm23 +[ ]*[a-f0-9]+:[ ]*62 e2 55 40 50 7a 7f[ ]*vpdpbusd 0x1fc0\(%rdx\),%zmm21,%zmm23 +[ ]*[a-f0-9]+:[ ]*62 e2 55 50 50 7a 7f[ ]*vpdpbusd 0x1fc\(%rdx\)\{1to16\},%zmm21,%zmm23 +[ ]*[a-f0-9]+:[ ]*62 02 3d 40 51 c1[ ]*vpdpbusds %zmm25,%zmm24,%zmm24 +[ ]*[a-f0-9]+:[ ]*62 02 3d 47 51 c1[ ]*vpdpbusds %zmm25,%zmm24,%zmm24\{%k7\} +[ ]*[a-f0-9]+:[ ]*62 02 3d c7 51 c1[ ]*vpdpbusds %zmm25,%zmm24,%zmm24\{%k7\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 22 3d 40 51 84 f0 23 01 00 00[ ]*vpdpbusds 0x123\(%rax,%r14,8\),%zmm24,%zmm24 +[ ]*[a-f0-9]+:[ ]*62 62 3d 40 51 42 7f[ ]*vpdpbusds 0x1fc0\(%rdx\),%zmm24,%zmm24 +[ ]*[a-f0-9]+:[ ]*62 62 3d 50 51 42 7f[ ]*vpdpbusds 0x1fc\(%rdx\)\{1to16\},%zmm24,%zmm24 +[ ]*[a-f0-9]+:[ ]*62 22 25 40 52 e1[ ]*vpdpwssd %zmm17,%zmm27,%zmm28 +[ ]*[a-f0-9]+:[ ]*62 22 25 47 52 e1[ ]*vpdpwssd %zmm17,%zmm27,%zmm28\{%k7\} +[ ]*[a-f0-9]+:[ ]*62 22 25 c7 52 e1[ ]*vpdpwssd %zmm17,%zmm27,%zmm28\{%k7\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 22 25 40 52 a4 f0 34 12 00 00[ ]*vpdpwssd 0x1234\(%rax,%r14,8\),%zmm27,%zmm28 +[ ]*[a-f0-9]+:[ ]*62 62 25 40 52 62 7f[ ]*vpdpwssd 0x1fc0\(%rdx\),%zmm27,%zmm28 +[ ]*[a-f0-9]+:[ ]*62 62 25 50 52 62 7f[ ]*vpdpwssd 0x1fc\(%rdx\)\{1to16\},%zmm27,%zmm28 +[ ]*[a-f0-9]+:[ ]*62 22 1d 40 53 e9[ ]*vpdpwssds %zmm17,%zmm28,%zmm29 +[ ]*[a-f0-9]+:[ ]*62 22 1d 43 53 e9[ ]*vpdpwssds %zmm17,%zmm28,%zmm29\{%k3\} +[ ]*[a-f0-9]+:[ ]*62 22 1d c3 53 e9[ ]*vpdpwssds %zmm17,%zmm28,%zmm29\{%k3\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 22 1d 40 53 ac f0 34 12 00 00[ ]*vpdpwssds 0x1234\(%rax,%r14,8\),%zmm28,%zmm29 +[ ]*[a-f0-9]+:[ ]*62 62 1d 40 53 6a 7f[ ]*vpdpwssds 0x1fc0\(%rdx\),%zmm28,%zmm29 +[ ]*[a-f0-9]+:[ ]*62 62 1d 50 53 6a 7f[ ]*vpdpwssds 0x1fc\(%rdx\)\{1to16\},%zmm28,%zmm29 +[ ]*[a-f0-9]+:[ ]*62 22 3d 40 50 e5[ ]*vpdpbusd %zmm21,%zmm24,%zmm28 +[ ]*[a-f0-9]+:[ ]*62 22 3d 46 50 e5[ ]*vpdpbusd %zmm21,%zmm24,%zmm28\{%k6\} +[ ]*[a-f0-9]+:[ ]*62 22 3d c6 50 e5[ ]*vpdpbusd %zmm21,%zmm24,%zmm28\{%k6\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 22 3d 40 50 a4 f0 34 12 00 00[ ]*vpdpbusd 0x1234\(%rax,%r14,8\),%zmm24,%zmm28 +[ ]*[a-f0-9]+:[ ]*62 62 3d 40 50 62 7f[ ]*vpdpbusd 0x1fc0\(%rdx\),%zmm24,%zmm28 +[ ]*[a-f0-9]+:[ ]*62 62 3d 50 50 62 7f[ ]*vpdpbusd 0x1fc\(%rdx\)\{1to16\},%zmm24,%zmm28 +[ ]*[a-f0-9]+:[ ]*62 a2 75 40 51 e4[ ]*vpdpbusds %zmm20,%zmm17,%zmm20 +[ ]*[a-f0-9]+:[ ]*62 a2 75 42 51 e4[ ]*vpdpbusds %zmm20,%zmm17,%zmm20\{%k2\} +[ ]*[a-f0-9]+:[ ]*62 a2 75 c2 51 e4[ ]*vpdpbusds %zmm20,%zmm17,%zmm20\{%k2\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 a2 75 40 51 a4 f0 34 12 00 00[ ]*vpdpbusds 0x1234\(%rax,%r14,8\),%zmm17,%zmm20 +[ ]*[a-f0-9]+:[ ]*62 e2 75 40 51 62 7f[ ]*vpdpbusds 0x1fc0\(%rdx\),%zmm17,%zmm20 +[ ]*[a-f0-9]+:[ ]*62 e2 75 50 51 62 7f[ ]*vpdpbusds 0x1fc\(%rdx\)\{1to16\},%zmm17,%zmm20 +#pass diff --git a/gas/testsuite/gas/i386/x86-64-avx512vnni.s b/gas/testsuite/gas/i386/x86-64-avx512vnni.s new file mode 100644 index 0000000..8ae051c --- /dev/null +++ b/gas/testsuite/gas/i386/x86-64-avx512vnni.s @@ -0,0 +1,61 @@ +# Check 64bit AVX512VNNI instructions + + .allow_index_reg + .text +_start: + vpdpwssd %zmm17, %zmm18, %zmm18 # AVX512VNNI + vpdpwssd %zmm17, %zmm18, %zmm18{%k5} # AVX512VNNI + vpdpwssd %zmm17, %zmm18, %zmm18{%k5}{z} # AVX512VNNI + vpdpwssd 0x123(%rax,%r14,8), %zmm18, %zmm18 # AVX512VNNI + vpdpwssd 8128(%rdx), %zmm18, %zmm18 # AVX512VNNI Disp8 + vpdpwssd 508(%rdx){1to16}, %zmm18, %zmm18 # AVX512VNNI Disp8 + + vpdpwssds %zmm17, %zmm21, %zmm21 # AVX512VNNI + vpdpwssds %zmm17, %zmm21, %zmm21{%k4} # AVX512VNNI + vpdpwssds %zmm17, %zmm21, %zmm21{%k4}{z} # AVX512VNNI + vpdpwssds 0x123(%rax,%r14,8), %zmm21, %zmm21 # AVX512VNNI + vpdpwssds 8128(%rdx), %zmm21, %zmm21 # AVX512VNNI Disp8 + vpdpwssds 508(%rdx){1to16}, %zmm21, %zmm21 # AVX512VNNI Disp8 + + vpdpbusd %zmm18, %zmm21, %zmm23 # AVX512VNNI + vpdpbusd %zmm18, %zmm21, %zmm23{%k4} # AVX512VNNI + vpdpbusd %zmm18, %zmm21, %zmm23{%k4}{z} # AVX512VNNI + vpdpbusd 0x123(%rax,%r14,8), %zmm21, %zmm23 # AVX512VNNI + vpdpbusd 8128(%rdx), %zmm21, %zmm23 # AVX512VNNI Disp8 + vpdpbusd 508(%rdx){1to16}, %zmm21, %zmm23 # AVX512VNNI Disp8 + + vpdpbusds %zmm25, %zmm24, %zmm24 # AVX512VNNI + vpdpbusds %zmm25, %zmm24, %zmm24{%k7} # AVX512VNNI + vpdpbusds %zmm25, %zmm24, %zmm24{%k7}{z} # AVX512VNNI + vpdpbusds 0x123(%rax,%r14,8), %zmm24, %zmm24 # AVX512VNNI + vpdpbusds 8128(%rdx), %zmm24, %zmm24 # AVX512VNNI Disp8 + vpdpbusds 508(%rdx){1to16}, %zmm24, %zmm24 # AVX512VNNI Disp8 + + .intel_syntax noprefix + vpdpwssd zmm28, zmm27, zmm17 # AVX512VNNI + vpdpwssd zmm28{k7}, zmm27, zmm17 # AVX512VNNI + vpdpwssd zmm28{k7}{z}, zmm27, zmm17 # AVX512VNNI + vpdpwssd zmm28, zmm27, ZMMWORD PTR [rax+r14*8+0x1234] # AVX512VNNI + vpdpwssd zmm28, zmm27, ZMMWORD PTR [rdx+8128] # AVX512VNNI Disp8 + vpdpwssd zmm28, zmm27, [rdx+508]{1to16} # AVX512VNNI Disp8 + + vpdpwssds zmm29, zmm28, zmm17 # AVX512VNNI + vpdpwssds zmm29{k3}, zmm28, zmm17 # AVX512VNNI + vpdpwssds zmm29{k3}{z}, zmm28, zmm17 # AVX512VNNI + vpdpwssds zmm29, zmm28, ZMMWORD PTR [rax+r14*8+0x1234] # AVX512VNNI + vpdpwssds zmm29, zmm28, ZMMWORD PTR [rdx+8128] # AVX512VNNI Disp8 + vpdpwssds zmm29, zmm28, [rdx+508]{1to16} # AVX512VNNI Disp8 + + vpdpbusd zmm28, zmm24, zmm21 # AVX512VNNI + vpdpbusd zmm28{k6}, zmm24, zmm21 # AVX512VNNI + vpdpbusd zmm28{k6}{z}, zmm24, zmm21 # AVX512VNNI + vpdpbusd zmm28, zmm24, ZMMWORD PTR [rax+r14*8+0x1234] # AVX512VNNI + vpdpbusd zmm28, zmm24, ZMMWORD PTR [rdx+8128] # AVX512VNNI Disp8 + vpdpbusd zmm28, zmm24, [rdx+508]{1to16} # AVX512VNNI Disp8 + + vpdpbusds zmm20, zmm17, zmm20 # AVX512VNNI + vpdpbusds zmm20{k2}, zmm17, zmm20 # AVX512VNNI + vpdpbusds zmm20{k2}{z}, zmm17, zmm20 # AVX512VNNI + vpdpbusds zmm20, zmm17, ZMMWORD PTR [rax+r14*8+0x1234] # AVX512VNNI + vpdpbusds zmm20, zmm17, ZMMWORD PTR [rdx+8128] # AVX512VNNI Disp8 + vpdpbusds zmm20, zmm17, [rdx+508]{1to16} # AVX512VNNI Disp8 diff --git a/gas/testsuite/gas/i386/x86-64-avx512vnni_vl-intel.d b/gas/testsuite/gas/i386/x86-64-avx512vnni_vl-intel.d new file mode 100644 index 0000000..1e3fe98 --- /dev/null +++ b/gas/testsuite/gas/i386/x86-64-avx512vnni_vl-intel.d @@ -0,0 +1,108 @@ +#as: +#objdump: -dw -Mintel +#name: x86_64 AVX512VNNI/VL insns (Intel disassembly) +#source: x86-64-avx512vnni_vl.s + +.*: +file format .* + + +Disassembly of section \.text: + +0+ <_start>: +[ ]*[a-f0-9]+:[ ]*62 22 4d 00 52 d4[ ]*vpdpwssd xmm26,xmm22,xmm20 +[ ]*[a-f0-9]+:[ ]*62 22 4d 03 52 d4[ ]*vpdpwssd xmm26\{k3\},xmm22,xmm20 +[ ]*[a-f0-9]+:[ ]*62 22 4d 83 52 d4[ ]*vpdpwssd xmm26\{k3\}\{z\},xmm22,xmm20 +[ ]*[a-f0-9]+:[ ]*62 22 4d 00 52 94 f0 23 01 00 00[ ]*vpdpwssd xmm26,xmm22,XMMWORD PTR \[rax\+r14\*8\+0x123\] +[ ]*[a-f0-9]+:[ ]*62 62 4d 00 52 52 7f[ ]*vpdpwssd xmm26,xmm22,XMMWORD PTR \[rdx\+0x7f0\] +[ ]*[a-f0-9]+:[ ]*62 62 4d 10 52 52 7f[ ]*vpdpwssd xmm26,xmm22,DWORD PTR \[rdx\+0x1fc\]\{1to4\} +[ ]*[a-f0-9]+:[ ]*62 a2 5d 20 52 e2[ ]*vpdpwssd ymm20,ymm20,ymm18 +[ ]*[a-f0-9]+:[ ]*62 a2 5d 25 52 e2[ ]*vpdpwssd ymm20\{k5\},ymm20,ymm18 +[ ]*[a-f0-9]+:[ ]*62 a2 5d a5 52 e2[ ]*vpdpwssd ymm20\{k5\}\{z\},ymm20,ymm18 +[ ]*[a-f0-9]+:[ ]*62 a2 5d 20 52 a4 f0 23 01 00 00[ ]*vpdpwssd ymm20,ymm20,YMMWORD PTR \[rax\+r14\*8\+0x123\] +[ ]*[a-f0-9]+:[ ]*62 e2 5d 20 52 62 7f[ ]*vpdpwssd ymm20,ymm20,YMMWORD PTR \[rdx\+0xfe0\] +[ ]*[a-f0-9]+:[ ]*62 e2 5d 30 52 62 7f[ ]*vpdpwssd ymm20,ymm20,DWORD PTR \[rdx\+0x1fc\]\{1to8\} +[ ]*[a-f0-9]+:[ ]*62 a2 65 00 53 f7[ ]*vpdpwssds xmm22,xmm19,xmm23 +[ ]*[a-f0-9]+:[ ]*62 a2 65 07 53 f7[ ]*vpdpwssds xmm22\{k7\},xmm19,xmm23 +[ ]*[a-f0-9]+:[ ]*62 a2 65 87 53 f7[ ]*vpdpwssds xmm22\{k7\}\{z\},xmm19,xmm23 +[ ]*[a-f0-9]+:[ ]*62 a2 65 00 53 b4 f0 23 01 00 00[ ]*vpdpwssds xmm22,xmm19,XMMWORD PTR \[rax\+r14\*8\+0x123\] +[ ]*[a-f0-9]+:[ ]*62 e2 65 00 53 72 7f[ ]*vpdpwssds xmm22,xmm19,XMMWORD PTR \[rdx\+0x7f0\] +[ ]*[a-f0-9]+:[ ]*62 e2 65 10 53 72 7f[ ]*vpdpwssds xmm22,xmm19,DWORD PTR \[rdx\+0x1fc\]\{1to4\} +[ ]*[a-f0-9]+:[ ]*62 82 45 20 53 fc[ ]*vpdpwssds ymm23,ymm23,ymm28 +[ ]*[a-f0-9]+:[ ]*62 82 45 23 53 fc[ ]*vpdpwssds ymm23\{k3\},ymm23,ymm28 +[ ]*[a-f0-9]+:[ ]*62 82 45 a3 53 fc[ ]*vpdpwssds ymm23\{k3\}\{z\},ymm23,ymm28 +[ ]*[a-f0-9]+:[ ]*62 a2 45 20 53 bc f0 23 01 00 00[ ]*vpdpwssds ymm23,ymm23,YMMWORD PTR \[rax\+r14\*8\+0x123\] +[ ]*[a-f0-9]+:[ ]*62 e2 45 20 53 7a 7f[ ]*vpdpwssds ymm23,ymm23,YMMWORD PTR \[rdx\+0xfe0\] +[ ]*[a-f0-9]+:[ ]*62 e2 45 30 53 7a 7f[ ]*vpdpwssds ymm23,ymm23,DWORD PTR \[rdx\+0x1fc\]\{1to8\} +[ ]*[a-f0-9]+:[ ]*62 82 15 00 50 d4[ ]*vpdpbusd xmm18,xmm29,xmm28 +[ ]*[a-f0-9]+:[ ]*62 82 15 03 50 d4[ ]*vpdpbusd xmm18\{k3\},xmm29,xmm28 +[ ]*[a-f0-9]+:[ ]*62 82 15 83 50 d4[ ]*vpdpbusd xmm18\{k3\}\{z\},xmm29,xmm28 +[ ]*[a-f0-9]+:[ ]*62 a2 15 00 50 94 f0 23 01 00 00[ ]*vpdpbusd xmm18,xmm29,XMMWORD PTR \[rax\+r14\*8\+0x123\] +[ ]*[a-f0-9]+:[ ]*62 e2 15 00 50 52 7f[ ]*vpdpbusd xmm18,xmm29,XMMWORD PTR \[rdx\+0x7f0\] +[ ]*[a-f0-9]+:[ ]*62 e2 15 10 50 52 7f[ ]*vpdpbusd xmm18,xmm29,DWORD PTR \[rdx\+0x1fc\]\{1to4\} +[ ]*[a-f0-9]+:[ ]*62 a2 6d 20 50 e1[ ]*vpdpbusd ymm20,ymm18,ymm17 +[ ]*[a-f0-9]+:[ ]*62 a2 6d 22 50 e1[ ]*vpdpbusd ymm20\{k2\},ymm18,ymm17 +[ ]*[a-f0-9]+:[ ]*62 a2 6d a2 50 e1[ ]*vpdpbusd ymm20\{k2\}\{z\},ymm18,ymm17 +[ ]*[a-f0-9]+:[ ]*62 a2 6d 20 50 a4 f0 23 01 00 00[ ]*vpdpbusd ymm20,ymm18,YMMWORD PTR \[rax\+r14\*8\+0x123\] +[ ]*[a-f0-9]+:[ ]*62 e2 6d 20 50 62 7f[ ]*vpdpbusd ymm20,ymm18,YMMWORD PTR \[rdx\+0xfe0\] +[ ]*[a-f0-9]+:[ ]*62 e2 6d 30 50 62 7f[ ]*vpdpbusd ymm20,ymm18,DWORD PTR \[rdx\+0x1fc\]\{1to8\} +[ ]*[a-f0-9]+:[ ]*62 02 2d 00 51 c3[ ]*vpdpbusds xmm24,xmm26,xmm27 +[ ]*[a-f0-9]+:[ ]*62 02 2d 04 51 c3[ ]*vpdpbusds xmm24\{k4\},xmm26,xmm27 +[ ]*[a-f0-9]+:[ ]*62 02 2d 84 51 c3[ ]*vpdpbusds xmm24\{k4\}\{z\},xmm26,xmm27 +[ ]*[a-f0-9]+:[ ]*62 22 2d 00 51 84 f0 23 01 00 00[ ]*vpdpbusds xmm24,xmm26,XMMWORD PTR \[rax\+r14\*8\+0x123\] +[ ]*[a-f0-9]+:[ ]*62 62 2d 00 51 42 7f[ ]*vpdpbusds xmm24,xmm26,XMMWORD PTR \[rdx\+0x7f0\] +[ ]*[a-f0-9]+:[ ]*62 62 2d 10 51 42 7f[ ]*vpdpbusds xmm24,xmm26,DWORD PTR \[rdx\+0x1fc\]\{1to4\} +[ ]*[a-f0-9]+:[ ]*62 02 15 20 51 f1[ ]*vpdpbusds ymm30,ymm29,ymm25 +[ ]*[a-f0-9]+:[ ]*62 02 15 21 51 f1[ ]*vpdpbusds ymm30\{k1\},ymm29,ymm25 +[ ]*[a-f0-9]+:[ ]*62 02 15 a1 51 f1[ ]*vpdpbusds ymm30\{k1\}\{z\},ymm29,ymm25 +[ ]*[a-f0-9]+:[ ]*62 22 15 20 51 b4 f0 23 01 00 00[ ]*vpdpbusds ymm30,ymm29,YMMWORD PTR \[rax\+r14\*8\+0x123\] +[ ]*[a-f0-9]+:[ ]*62 62 15 20 51 72 7f[ ]*vpdpbusds ymm30,ymm29,YMMWORD PTR \[rdx\+0xfe0\] +[ ]*[a-f0-9]+:[ ]*62 62 15 30 51 72 7f[ ]*vpdpbusds ymm30,ymm29,DWORD PTR \[rdx\+0x1fc\]\{1to8\} +[ ]*[a-f0-9]+:[ ]*62 a2 5d 00 52 ef[ ]*vpdpwssd xmm21,xmm20,xmm23 +[ ]*[a-f0-9]+:[ ]*62 a2 5d 06 52 ef[ ]*vpdpwssd xmm21\{k6\},xmm20,xmm23 +[ ]*[a-f0-9]+:[ ]*62 a2 5d 86 52 ef[ ]*vpdpwssd xmm21\{k6\}\{z\},xmm20,xmm23 +[ ]*[a-f0-9]+:[ ]*62 a2 5d 00 52 ac f0 34 12 00 00[ ]*vpdpwssd xmm21,xmm20,XMMWORD PTR \[rax\+r14\*8\+0x1234\] +[ ]*[a-f0-9]+:[ ]*62 e2 5d 00 52 6a 7f[ ]*vpdpwssd xmm21,xmm20,XMMWORD PTR \[rdx\+0x7f0\] +[ ]*[a-f0-9]+:[ ]*62 e2 5d 10 52 6a 7f[ ]*vpdpwssd xmm21,xmm20,DWORD PTR \[rdx\+0x1fc\]\{1to4\} +[ ]*[a-f0-9]+:[ ]*62 22 25 20 52 c9[ ]*vpdpwssd ymm25,ymm27,ymm17 +[ ]*[a-f0-9]+:[ ]*62 22 25 26 52 c9[ ]*vpdpwssd ymm25\{k6\},ymm27,ymm17 +[ ]*[a-f0-9]+:[ ]*62 22 25 a6 52 c9[ ]*vpdpwssd ymm25\{k6\}\{z\},ymm27,ymm17 +[ ]*[a-f0-9]+:[ ]*62 22 25 20 52 8c f0 34 12 00 00[ ]*vpdpwssd ymm25,ymm27,YMMWORD PTR \[rax\+r14\*8\+0x1234\] +[ ]*[a-f0-9]+:[ ]*62 62 25 20 52 4a 7f[ ]*vpdpwssd ymm25,ymm27,YMMWORD PTR \[rdx\+0xfe0\] +[ ]*[a-f0-9]+:[ ]*62 62 25 30 52 4a 7f[ ]*vpdpwssd ymm25,ymm27,DWORD PTR \[rdx\+0x1fc\]\{1to8\} +[ ]*[a-f0-9]+:[ ]*62 22 35 00 53 f5[ ]*vpdpwssds xmm30,xmm25,xmm21 +[ ]*[a-f0-9]+:[ ]*62 22 35 06 53 f5[ ]*vpdpwssds xmm30\{k6\},xmm25,xmm21 +[ ]*[a-f0-9]+:[ ]*62 22 35 86 53 f5[ ]*vpdpwssds xmm30\{k6\}\{z\},xmm25,xmm21 +[ ]*[a-f0-9]+:[ ]*62 22 35 00 53 b4 f0 34 12 00 00[ ]*vpdpwssds xmm30,xmm25,XMMWORD PTR \[rax\+r14\*8\+0x1234\] +[ ]*[a-f0-9]+:[ ]*62 62 35 00 53 72 7f[ ]*vpdpwssds xmm30,xmm25,XMMWORD PTR \[rdx\+0x7f0\] +[ ]*[a-f0-9]+:[ ]*62 62 35 10 53 72 7f[ ]*vpdpwssds xmm30,xmm25,DWORD PTR \[rdx\+0x1fc\]\{1to4\} +[ ]*[a-f0-9]+:[ ]*62 02 25 20 53 e3[ ]*vpdpwssds ymm28,ymm27,ymm27 +[ ]*[a-f0-9]+:[ ]*62 02 25 27 53 e3[ ]*vpdpwssds ymm28\{k7\},ymm27,ymm27 +[ ]*[a-f0-9]+:[ ]*62 02 25 a7 53 e3[ ]*vpdpwssds ymm28\{k7\}\{z\},ymm27,ymm27 +[ ]*[a-f0-9]+:[ ]*62 22 25 20 53 a4 f0 34 12 00 00[ ]*vpdpwssds ymm28,ymm27,YMMWORD PTR \[rax\+r14\*8\+0x1234\] +[ ]*[a-f0-9]+:[ ]*62 62 25 20 53 62 7f[ ]*vpdpwssds ymm28,ymm27,YMMWORD PTR \[rdx\+0xfe0\] +[ ]*[a-f0-9]+:[ ]*62 62 25 30 53 62 7f[ ]*vpdpwssds ymm28,ymm27,DWORD PTR \[rdx\+0x1fc\]\{1to8\} +[ ]*[a-f0-9]+:[ ]*62 22 6d 00 50 d3[ ]*vpdpbusd xmm26,xmm18,xmm19 +[ ]*[a-f0-9]+:[ ]*62 22 6d 06 50 d3[ ]*vpdpbusd xmm26\{k6\},xmm18,xmm19 +[ ]*[a-f0-9]+:[ ]*62 22 6d 86 50 d3[ ]*vpdpbusd xmm26\{k6\}\{z\},xmm18,xmm19 +[ ]*[a-f0-9]+:[ ]*62 22 6d 00 50 94 f0 34 12 00 00[ ]*vpdpbusd xmm26,xmm18,XMMWORD PTR \[rax\+r14\*8\+0x1234\] +[ ]*[a-f0-9]+:[ ]*62 62 6d 00 50 52 7f[ ]*vpdpbusd xmm26,xmm18,XMMWORD PTR \[rdx\+0x7f0\] +[ ]*[a-f0-9]+:[ ]*62 62 6d 10 50 52 7f[ ]*vpdpbusd xmm26,xmm18,DWORD PTR \[rdx\+0x1fc\]\{1to4\} +[ ]*[a-f0-9]+:[ ]*62 82 75 20 50 eb[ ]*vpdpbusd ymm21,ymm17,ymm27 +[ ]*[a-f0-9]+:[ ]*62 82 75 22 50 eb[ ]*vpdpbusd ymm21\{k2\},ymm17,ymm27 +[ ]*[a-f0-9]+:[ ]*62 82 75 a2 50 eb[ ]*vpdpbusd ymm21\{k2\}\{z\},ymm17,ymm27 +[ ]*[a-f0-9]+:[ ]*62 a2 75 20 50 ac f0 34 12 00 00[ ]*vpdpbusd ymm21,ymm17,YMMWORD PTR \[rax\+r14\*8\+0x1234\] +[ ]*[a-f0-9]+:[ ]*62 e2 75 20 50 6a 7f[ ]*vpdpbusd ymm21,ymm17,YMMWORD PTR \[rdx\+0xfe0\] +[ ]*[a-f0-9]+:[ ]*62 e2 75 30 50 6a 7f[ ]*vpdpbusd ymm21,ymm17,DWORD PTR \[rdx\+0x1fc\]\{1to8\} +[ ]*[a-f0-9]+:[ ]*62 02 2d 00 51 e0[ ]*vpdpbusds xmm28,xmm26,xmm24 +[ ]*[a-f0-9]+:[ ]*62 02 2d 01 51 e0[ ]*vpdpbusds xmm28\{k1\},xmm26,xmm24 +[ ]*[a-f0-9]+:[ ]*62 02 2d 81 51 e0[ ]*vpdpbusds xmm28\{k1\}\{z\},xmm26,xmm24 +[ ]*[a-f0-9]+:[ ]*62 22 2d 00 51 a4 f0 34 12 00 00[ ]*vpdpbusds xmm28,xmm26,XMMWORD PTR \[rax\+r14\*8\+0x1234\] +[ ]*[a-f0-9]+:[ ]*62 62 2d 00 51 62 7f[ ]*vpdpbusds xmm28,xmm26,XMMWORD PTR \[rdx\+0x7f0\] +[ ]*[a-f0-9]+:[ ]*62 62 2d 10 51 62 7f[ ]*vpdpbusds xmm28,xmm26,DWORD PTR \[rdx\+0x1fc\]\{1to4\} +[ ]*[a-f0-9]+:[ ]*62 82 6d 20 51 fb[ ]*vpdpbusds ymm23,ymm18,ymm27 +[ ]*[a-f0-9]+:[ ]*62 82 6d 26 51 fb[ ]*vpdpbusds ymm23\{k6\},ymm18,ymm27 +[ ]*[a-f0-9]+:[ ]*62 82 6d a6 51 fb[ ]*vpdpbusds ymm23\{k6\}\{z\},ymm18,ymm27 +[ ]*[a-f0-9]+:[ ]*62 a2 6d 20 51 bc f0 34 12 00 00[ ]*vpdpbusds ymm23,ymm18,YMMWORD PTR \[rax\+r14\*8\+0x1234\] +[ ]*[a-f0-9]+:[ ]*62 e2 6d 20 51 7a 7f[ ]*vpdpbusds ymm23,ymm18,YMMWORD PTR \[rdx\+0xfe0\] +[ ]*[a-f0-9]+:[ ]*62 e2 6d 30 51 7a 7f[ ]*vpdpbusds ymm23,ymm18,DWORD PTR \[rdx\+0x1fc\]\{1to8\} +#pass diff --git a/gas/testsuite/gas/i386/x86-64-avx512vnni_vl.d b/gas/testsuite/gas/i386/x86-64-avx512vnni_vl.d new file mode 100644 index 0000000..2817562 --- /dev/null +++ b/gas/testsuite/gas/i386/x86-64-avx512vnni_vl.d @@ -0,0 +1,108 @@ +#as: +#objdump: -dw +#name: x86_64 AVX512VNNI/VL insns +#source: x86-64-avx512vnni_vl.s + +.*: +file format .* + + +Disassembly of section \.text: + +0+ <_start>: +[ ]*[a-f0-9]+:[ ]*62 22 4d 00 52 d4[ ]*vpdpwssd %xmm20,%xmm22,%xmm26 +[ ]*[a-f0-9]+:[ ]*62 22 4d 03 52 d4[ ]*vpdpwssd %xmm20,%xmm22,%xmm26\{%k3\} +[ ]*[a-f0-9]+:[ ]*62 22 4d 83 52 d4[ ]*vpdpwssd %xmm20,%xmm22,%xmm26\{%k3\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 22 4d 00 52 94 f0 23 01 00 00[ ]*vpdpwssd 0x123\(%rax,%r14,8\),%xmm22,%xmm26 +[ ]*[a-f0-9]+:[ ]*62 62 4d 00 52 52 7f[ ]*vpdpwssd 0x7f0\(%rdx\),%xmm22,%xmm26 +[ ]*[a-f0-9]+:[ ]*62 62 4d 10 52 52 7f[ ]*vpdpwssd 0x1fc\(%rdx\)\{1to4\},%xmm22,%xmm26 +[ ]*[a-f0-9]+:[ ]*62 a2 5d 20 52 e2[ ]*vpdpwssd %ymm18,%ymm20,%ymm20 +[ ]*[a-f0-9]+:[ ]*62 a2 5d 25 52 e2[ ]*vpdpwssd %ymm18,%ymm20,%ymm20\{%k5\} +[ ]*[a-f0-9]+:[ ]*62 a2 5d a5 52 e2[ ]*vpdpwssd %ymm18,%ymm20,%ymm20\{%k5\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 a2 5d 20 52 a4 f0 23 01 00 00[ ]*vpdpwssd 0x123\(%rax,%r14,8\),%ymm20,%ymm20 +[ ]*[a-f0-9]+:[ ]*62 e2 5d 20 52 62 7f[ ]*vpdpwssd 0xfe0\(%rdx\),%ymm20,%ymm20 +[ ]*[a-f0-9]+:[ ]*62 e2 5d 30 52 62 7f[ ]*vpdpwssd 0x1fc\(%rdx\)\{1to8\},%ymm20,%ymm20 +[ ]*[a-f0-9]+:[ ]*62 a2 65 00 53 f7[ ]*vpdpwssds %xmm23,%xmm19,%xmm22 +[ ]*[a-f0-9]+:[ ]*62 a2 65 07 53 f7[ ]*vpdpwssds %xmm23,%xmm19,%xmm22\{%k7\} +[ ]*[a-f0-9]+:[ ]*62 a2 65 87 53 f7[ ]*vpdpwssds %xmm23,%xmm19,%xmm22\{%k7\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 a2 65 00 53 b4 f0 23 01 00 00[ ]*vpdpwssds 0x123\(%rax,%r14,8\),%xmm19,%xmm22 +[ ]*[a-f0-9]+:[ ]*62 e2 65 00 53 72 7f[ ]*vpdpwssds 0x7f0\(%rdx\),%xmm19,%xmm22 +[ ]*[a-f0-9]+:[ ]*62 e2 65 10 53 72 7f[ ]*vpdpwssds 0x1fc\(%rdx\)\{1to4\},%xmm19,%xmm22 +[ ]*[a-f0-9]+:[ ]*62 82 45 20 53 fc[ ]*vpdpwssds %ymm28,%ymm23,%ymm23 +[ ]*[a-f0-9]+:[ ]*62 82 45 23 53 fc[ ]*vpdpwssds %ymm28,%ymm23,%ymm23\{%k3\} +[ ]*[a-f0-9]+:[ ]*62 82 45 a3 53 fc[ ]*vpdpwssds %ymm28,%ymm23,%ymm23\{%k3\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 a2 45 20 53 bc f0 23 01 00 00[ ]*vpdpwssds 0x123\(%rax,%r14,8\),%ymm23,%ymm23 +[ ]*[a-f0-9]+:[ ]*62 e2 45 20 53 7a 7f[ ]*vpdpwssds 0xfe0\(%rdx\),%ymm23,%ymm23 +[ ]*[a-f0-9]+:[ ]*62 e2 45 30 53 7a 7f[ ]*vpdpwssds 0x1fc\(%rdx\)\{1to8\},%ymm23,%ymm23 +[ ]*[a-f0-9]+:[ ]*62 82 15 00 50 d4[ ]*vpdpbusd %xmm28,%xmm29,%xmm18 +[ ]*[a-f0-9]+:[ ]*62 82 15 03 50 d4[ ]*vpdpbusd %xmm28,%xmm29,%xmm18\{%k3\} +[ ]*[a-f0-9]+:[ ]*62 82 15 83 50 d4[ ]*vpdpbusd %xmm28,%xmm29,%xmm18\{%k3\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 a2 15 00 50 94 f0 23 01 00 00[ ]*vpdpbusd 0x123\(%rax,%r14,8\),%xmm29,%xmm18 +[ ]*[a-f0-9]+:[ ]*62 e2 15 00 50 52 7f[ ]*vpdpbusd 0x7f0\(%rdx\),%xmm29,%xmm18 +[ ]*[a-f0-9]+:[ ]*62 e2 15 10 50 52 7f[ ]*vpdpbusd 0x1fc\(%rdx\)\{1to4\},%xmm29,%xmm18 +[ ]*[a-f0-9]+:[ ]*62 a2 6d 20 50 e1[ ]*vpdpbusd %ymm17,%ymm18,%ymm20 +[ ]*[a-f0-9]+:[ ]*62 a2 6d 22 50 e1[ ]*vpdpbusd %ymm17,%ymm18,%ymm20\{%k2\} +[ ]*[a-f0-9]+:[ ]*62 a2 6d a2 50 e1[ ]*vpdpbusd %ymm17,%ymm18,%ymm20\{%k2\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 a2 6d 20 50 a4 f0 23 01 00 00[ ]*vpdpbusd 0x123\(%rax,%r14,8\),%ymm18,%ymm20 +[ ]*[a-f0-9]+:[ ]*62 e2 6d 20 50 62 7f[ ]*vpdpbusd 0xfe0\(%rdx\),%ymm18,%ymm20 +[ ]*[a-f0-9]+:[ ]*62 e2 6d 30 50 62 7f[ ]*vpdpbusd 0x1fc\(%rdx\)\{1to8\},%ymm18,%ymm20 +[ ]*[a-f0-9]+:[ ]*62 02 2d 00 51 c3[ ]*vpdpbusds %xmm27,%xmm26,%xmm24 +[ ]*[a-f0-9]+:[ ]*62 02 2d 04 51 c3[ ]*vpdpbusds %xmm27,%xmm26,%xmm24\{%k4\} +[ ]*[a-f0-9]+:[ ]*62 02 2d 84 51 c3[ ]*vpdpbusds %xmm27,%xmm26,%xmm24\{%k4\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 22 2d 00 51 84 f0 23 01 00 00[ ]*vpdpbusds 0x123\(%rax,%r14,8\),%xmm26,%xmm24 +[ ]*[a-f0-9]+:[ ]*62 62 2d 00 51 42 7f[ ]*vpdpbusds 0x7f0\(%rdx\),%xmm26,%xmm24 +[ ]*[a-f0-9]+:[ ]*62 62 2d 10 51 42 7f[ ]*vpdpbusds 0x1fc\(%rdx\)\{1to4\},%xmm26,%xmm24 +[ ]*[a-f0-9]+:[ ]*62 02 15 20 51 f1[ ]*vpdpbusds %ymm25,%ymm29,%ymm30 +[ ]*[a-f0-9]+:[ ]*62 02 15 21 51 f1[ ]*vpdpbusds %ymm25,%ymm29,%ymm30\{%k1\} +[ ]*[a-f0-9]+:[ ]*62 02 15 a1 51 f1[ ]*vpdpbusds %ymm25,%ymm29,%ymm30\{%k1\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 22 15 20 51 b4 f0 23 01 00 00[ ]*vpdpbusds 0x123\(%rax,%r14,8\),%ymm29,%ymm30 +[ ]*[a-f0-9]+:[ ]*62 62 15 20 51 72 7f[ ]*vpdpbusds 0xfe0\(%rdx\),%ymm29,%ymm30 +[ ]*[a-f0-9]+:[ ]*62 62 15 30 51 72 7f[ ]*vpdpbusds 0x1fc\(%rdx\)\{1to8\},%ymm29,%ymm30 +[ ]*[a-f0-9]+:[ ]*62 a2 5d 00 52 ef[ ]*vpdpwssd %xmm23,%xmm20,%xmm21 +[ ]*[a-f0-9]+:[ ]*62 a2 5d 06 52 ef[ ]*vpdpwssd %xmm23,%xmm20,%xmm21\{%k6\} +[ ]*[a-f0-9]+:[ ]*62 a2 5d 86 52 ef[ ]*vpdpwssd %xmm23,%xmm20,%xmm21\{%k6\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 a2 5d 00 52 ac f0 34 12 00 00[ ]*vpdpwssd 0x1234\(%rax,%r14,8\),%xmm20,%xmm21 +[ ]*[a-f0-9]+:[ ]*62 e2 5d 00 52 6a 7f[ ]*vpdpwssd 0x7f0\(%rdx\),%xmm20,%xmm21 +[ ]*[a-f0-9]+:[ ]*62 e2 5d 10 52 6a 7f[ ]*vpdpwssd 0x1fc\(%rdx\)\{1to4\},%xmm20,%xmm21 +[ ]*[a-f0-9]+:[ ]*62 22 25 20 52 c9[ ]*vpdpwssd %ymm17,%ymm27,%ymm25 +[ ]*[a-f0-9]+:[ ]*62 22 25 26 52 c9[ ]*vpdpwssd %ymm17,%ymm27,%ymm25\{%k6\} +[ ]*[a-f0-9]+:[ ]*62 22 25 a6 52 c9[ ]*vpdpwssd %ymm17,%ymm27,%ymm25\{%k6\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 22 25 20 52 8c f0 34 12 00 00[ ]*vpdpwssd 0x1234\(%rax,%r14,8\),%ymm27,%ymm25 +[ ]*[a-f0-9]+:[ ]*62 62 25 20 52 4a 7f[ ]*vpdpwssd 0xfe0\(%rdx\),%ymm27,%ymm25 +[ ]*[a-f0-9]+:[ ]*62 62 25 30 52 4a 7f[ ]*vpdpwssd 0x1fc\(%rdx\)\{1to8\},%ymm27,%ymm25 +[ ]*[a-f0-9]+:[ ]*62 22 35 00 53 f5[ ]*vpdpwssds %xmm21,%xmm25,%xmm30 +[ ]*[a-f0-9]+:[ ]*62 22 35 06 53 f5[ ]*vpdpwssds %xmm21,%xmm25,%xmm30\{%k6\} +[ ]*[a-f0-9]+:[ ]*62 22 35 86 53 f5[ ]*vpdpwssds %xmm21,%xmm25,%xmm30\{%k6\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 22 35 00 53 b4 f0 34 12 00 00[ ]*vpdpwssds 0x1234\(%rax,%r14,8\),%xmm25,%xmm30 +[ ]*[a-f0-9]+:[ ]*62 62 35 00 53 72 7f[ ]*vpdpwssds 0x7f0\(%rdx\),%xmm25,%xmm30 +[ ]*[a-f0-9]+:[ ]*62 62 35 10 53 72 7f[ ]*vpdpwssds 0x1fc\(%rdx\)\{1to4\},%xmm25,%xmm30 +[ ]*[a-f0-9]+:[ ]*62 02 25 20 53 e3[ ]*vpdpwssds %ymm27,%ymm27,%ymm28 +[ ]*[a-f0-9]+:[ ]*62 02 25 27 53 e3[ ]*vpdpwssds %ymm27,%ymm27,%ymm28\{%k7\} +[ ]*[a-f0-9]+:[ ]*62 02 25 a7 53 e3[ ]*vpdpwssds %ymm27,%ymm27,%ymm28\{%k7\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 22 25 20 53 a4 f0 34 12 00 00[ ]*vpdpwssds 0x1234\(%rax,%r14,8\),%ymm27,%ymm28 +[ ]*[a-f0-9]+:[ ]*62 62 25 20 53 62 7f[ ]*vpdpwssds 0xfe0\(%rdx\),%ymm27,%ymm28 +[ ]*[a-f0-9]+:[ ]*62 62 25 30 53 62 7f[ ]*vpdpwssds 0x1fc\(%rdx\)\{1to8\},%ymm27,%ymm28 +[ ]*[a-f0-9]+:[ ]*62 22 6d 00 50 d3[ ]*vpdpbusd %xmm19,%xmm18,%xmm26 +[ ]*[a-f0-9]+:[ ]*62 22 6d 06 50 d3[ ]*vpdpbusd %xmm19,%xmm18,%xmm26\{%k6\} +[ ]*[a-f0-9]+:[ ]*62 22 6d 86 50 d3[ ]*vpdpbusd %xmm19,%xmm18,%xmm26\{%k6\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 22 6d 00 50 94 f0 34 12 00 00[ ]*vpdpbusd 0x1234\(%rax,%r14,8\),%xmm18,%xmm26 +[ ]*[a-f0-9]+:[ ]*62 62 6d 00 50 52 7f[ ]*vpdpbusd 0x7f0\(%rdx\),%xmm18,%xmm26 +[ ]*[a-f0-9]+:[ ]*62 62 6d 10 50 52 7f[ ]*vpdpbusd 0x1fc\(%rdx\)\{1to4\},%xmm18,%xmm26 +[ ]*[a-f0-9]+:[ ]*62 82 75 20 50 eb[ ]*vpdpbusd %ymm27,%ymm17,%ymm21 +[ ]*[a-f0-9]+:[ ]*62 82 75 22 50 eb[ ]*vpdpbusd %ymm27,%ymm17,%ymm21\{%k2\} +[ ]*[a-f0-9]+:[ ]*62 82 75 a2 50 eb[ ]*vpdpbusd %ymm27,%ymm17,%ymm21\{%k2\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 a2 75 20 50 ac f0 34 12 00 00[ ]*vpdpbusd 0x1234\(%rax,%r14,8\),%ymm17,%ymm21 +[ ]*[a-f0-9]+:[ ]*62 e2 75 20 50 6a 7f[ ]*vpdpbusd 0xfe0\(%rdx\),%ymm17,%ymm21 +[ ]*[a-f0-9]+:[ ]*62 e2 75 30 50 6a 7f[ ]*vpdpbusd 0x1fc\(%rdx\)\{1to8\},%ymm17,%ymm21 +[ ]*[a-f0-9]+:[ ]*62 02 2d 00 51 e0[ ]*vpdpbusds %xmm24,%xmm26,%xmm28 +[ ]*[a-f0-9]+:[ ]*62 02 2d 01 51 e0[ ]*vpdpbusds %xmm24,%xmm26,%xmm28\{%k1\} +[ ]*[a-f0-9]+:[ ]*62 02 2d 81 51 e0[ ]*vpdpbusds %xmm24,%xmm26,%xmm28\{%k1\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 22 2d 00 51 a4 f0 34 12 00 00[ ]*vpdpbusds 0x1234\(%rax,%r14,8\),%xmm26,%xmm28 +[ ]*[a-f0-9]+:[ ]*62 62 2d 00 51 62 7f[ ]*vpdpbusds 0x7f0\(%rdx\),%xmm26,%xmm28 +[ ]*[a-f0-9]+:[ ]*62 62 2d 10 51 62 7f[ ]*vpdpbusds 0x1fc\(%rdx\)\{1to4\},%xmm26,%xmm28 +[ ]*[a-f0-9]+:[ ]*62 82 6d 20 51 fb[ ]*vpdpbusds %ymm27,%ymm18,%ymm23 +[ ]*[a-f0-9]+:[ ]*62 82 6d 26 51 fb[ ]*vpdpbusds %ymm27,%ymm18,%ymm23\{%k6\} +[ ]*[a-f0-9]+:[ ]*62 82 6d a6 51 fb[ ]*vpdpbusds %ymm27,%ymm18,%ymm23\{%k6\}\{z\} +[ ]*[a-f0-9]+:[ ]*62 a2 6d 20 51 bc f0 34 12 00 00[ ]*vpdpbusds 0x1234\(%rax,%r14,8\),%ymm18,%ymm23 +[ ]*[a-f0-9]+:[ ]*62 e2 6d 20 51 7a 7f[ ]*vpdpbusds 0xfe0\(%rdx\),%ymm18,%ymm23 +[ ]*[a-f0-9]+:[ ]*62 e2 6d 30 51 7a 7f[ ]*vpdpbusds 0x1fc\(%rdx\)\{1to8\},%ymm18,%ymm23 +#pass diff --git a/gas/testsuite/gas/i386/x86-64-avx512vnni_vl.s b/gas/testsuite/gas/i386/x86-64-avx512vnni_vl.s new file mode 100644 index 0000000..9c85ac7 --- /dev/null +++ b/gas/testsuite/gas/i386/x86-64-avx512vnni_vl.s @@ -0,0 +1,109 @@ +# Check 64bit AVX512{VNNI,VL} instructions + + .allow_index_reg + .text +_start: + vpdpwssd %xmm20, %xmm22, %xmm26 # AVX512{VNNI,VL} + vpdpwssd %xmm20, %xmm22, %xmm26{%k3} # AVX512{VNNI,VL} + vpdpwssd %xmm20, %xmm22, %xmm26{%k3}{z} # AVX512{VNNI,VL} + vpdpwssd 0x123(%rax,%r14,8), %xmm22, %xmm26 # AVX512{VNNI,VL} + vpdpwssd 2032(%rdx), %xmm22, %xmm26 # AVX512{VNNI,VL} Disp8 + vpdpwssd 508(%rdx){1to4}, %xmm22, %xmm26 # AVX512{VNNI,VL} Disp8 + vpdpwssd %ymm18, %ymm20, %ymm20 # AVX512{VNNI,VL} + vpdpwssd %ymm18, %ymm20, %ymm20{%k5} # AVX512{VNNI,VL} + vpdpwssd %ymm18, %ymm20, %ymm20{%k5}{z} # AVX512{VNNI,VL} + vpdpwssd 0x123(%rax,%r14,8), %ymm20, %ymm20 # AVX512{VNNI,VL} + vpdpwssd 4064(%rdx), %ymm20, %ymm20 # AVX512{VNNI,VL} Disp8 + vpdpwssd 508(%rdx){1to8}, %ymm20, %ymm20 # AVX512{VNNI,VL} Disp8 + + vpdpwssds %xmm23, %xmm19, %xmm22 # AVX512{VNNI,VL} + vpdpwssds %xmm23, %xmm19, %xmm22{%k7} # AVX512{VNNI,VL} + vpdpwssds %xmm23, %xmm19, %xmm22{%k7}{z} # AVX512{VNNI,VL} + vpdpwssds 0x123(%rax,%r14,8), %xmm19, %xmm22 # AVX512{VNNI,VL} + vpdpwssds 2032(%rdx), %xmm19, %xmm22 # AVX512{VNNI,VL} Disp8 + vpdpwssds 508(%rdx){1to4}, %xmm19, %xmm22 # AVX512{VNNI,VL} Disp8 + vpdpwssds %ymm28, %ymm23, %ymm23 # AVX512{VNNI,VL} + vpdpwssds %ymm28, %ymm23, %ymm23{%k3} # AVX512{VNNI,VL} + vpdpwssds %ymm28, %ymm23, %ymm23{%k3}{z} # AVX512{VNNI,VL} + vpdpwssds 0x123(%rax,%r14,8), %ymm23, %ymm23 # AVX512{VNNI,VL} + vpdpwssds 4064(%rdx), %ymm23, %ymm23 # AVX512{VNNI,VL} Disp8 + vpdpwssds 508(%rdx){1to8}, %ymm23, %ymm23 # AVX512{VNNI,VL} Disp8 + + vpdpbusd %xmm28, %xmm29, %xmm18 # AVX512{VNNI,VL} + vpdpbusd %xmm28, %xmm29, %xmm18{%k3} # AVX512{VNNI,VL} + vpdpbusd %xmm28, %xmm29, %xmm18{%k3}{z} # AVX512{VNNI,VL} + vpdpbusd 0x123(%rax,%r14,8), %xmm29, %xmm18 # AVX512{VNNI,VL} + vpdpbusd 2032(%rdx), %xmm29, %xmm18 # AVX512{VNNI,VL} Disp8 + vpdpbusd 508(%rdx){1to4}, %xmm29, %xmm18 # AVX512{VNNI,VL} Disp8 + vpdpbusd %ymm17, %ymm18, %ymm20 # AVX512{VNNI,VL} + vpdpbusd %ymm17, %ymm18, %ymm20{%k2} # AVX512{VNNI,VL} + vpdpbusd %ymm17, %ymm18, %ymm20{%k2}{z} # AVX512{VNNI,VL} + vpdpbusd 0x123(%rax,%r14,8), %ymm18, %ymm20 # AVX512{VNNI,VL} + vpdpbusd 4064(%rdx), %ymm18, %ymm20 # AVX512{VNNI,VL} Disp8 + vpdpbusd 508(%rdx){1to8}, %ymm18, %ymm20 # AVX512{VNNI,VL} Disp8 + + vpdpbusds %xmm27, %xmm26, %xmm24 # AVX512{VNNI,VL} + vpdpbusds %xmm27, %xmm26, %xmm24{%k4} # AVX512{VNNI,VL} + vpdpbusds %xmm27, %xmm26, %xmm24{%k4}{z} # AVX512{VNNI,VL} + vpdpbusds 0x123(%rax,%r14,8), %xmm26, %xmm24 # AVX512{VNNI,VL} + vpdpbusds 2032(%rdx), %xmm26, %xmm24 # AVX512{VNNI,VL} Disp8 + vpdpbusds 508(%rdx){1to4}, %xmm26, %xmm24 # AVX512{VNNI,VL} Disp8 + vpdpbusds %ymm25, %ymm29, %ymm30 # AVX512{VNNI,VL} + vpdpbusds %ymm25, %ymm29, %ymm30{%k1} # AVX512{VNNI,VL} + vpdpbusds %ymm25, %ymm29, %ymm30{%k1}{z} # AVX512{VNNI,VL} + vpdpbusds 0x123(%rax,%r14,8), %ymm29, %ymm30 # AVX512{VNNI,VL} + vpdpbusds 4064(%rdx), %ymm29, %ymm30 # AVX512{VNNI,VL} Disp8 + vpdpbusds 508(%rdx){1to8}, %ymm29, %ymm30 # AVX512{VNNI,VL} Disp8 + + .intel_syntax noprefix + vpdpwssd xmm21, xmm20, xmm23 # AVX512{VNNI,VL} + vpdpwssd xmm21{k6}, xmm20, xmm23 # AVX512{VNNI,VL} + vpdpwssd xmm21{k6}{z}, xmm20, xmm23 # AVX512{VNNI,VL} + vpdpwssd xmm21, xmm20, XMMWORD PTR [rax+r14*8+0x1234] # AVX512{VNNI,VL} + vpdpwssd xmm21, xmm20, XMMWORD PTR [rdx+2032] # AVX512{VNNI,VL} Disp8 + vpdpwssd xmm21, xmm20, [rdx+508]{1to4} # AVX512{VNNI,VL} Disp8 + vpdpwssd ymm25, ymm27, ymm17 # AVX512{VNNI,VL} + vpdpwssd ymm25{k6}, ymm27, ymm17 # AVX512{VNNI,VL} + vpdpwssd ymm25{k6}{z}, ymm27, ymm17 # AVX512{VNNI,VL} + vpdpwssd ymm25, ymm27, YMMWORD PTR [rax+r14*8+0x1234] # AVX512{VNNI,VL} + vpdpwssd ymm25, ymm27, YMMWORD PTR [rdx+4064] # AVX512{VNNI,VL} Disp8 + vpdpwssd ymm25, ymm27, [rdx+508]{1to8} # AVX512{VNNI,VL} Disp8 + + vpdpwssds xmm30, xmm25, xmm21 # AVX512{VNNI,VL} + vpdpwssds xmm30{k6}, xmm25, xmm21 # AVX512{VNNI,VL} + vpdpwssds xmm30{k6}{z}, xmm25, xmm21 # AVX512{VNNI,VL} + vpdpwssds xmm30, xmm25, XMMWORD PTR [rax+r14*8+0x1234] # AVX512{VNNI,VL} + vpdpwssds xmm30, xmm25, XMMWORD PTR [rdx+2032] # AVX512{VNNI,VL} Disp8 + vpdpwssds xmm30, xmm25, [rdx+508]{1to4} # AVX512{VNNI,VL} Disp8 + vpdpwssds ymm28, ymm27, ymm27 # AVX512{VNNI,VL} + vpdpwssds ymm28{k7}, ymm27, ymm27 # AVX512{VNNI,VL} + vpdpwssds ymm28{k7}{z}, ymm27, ymm27 # AVX512{VNNI,VL} + vpdpwssds ymm28, ymm27, YMMWORD PTR [rax+r14*8+0x1234] # AVX512{VNNI,VL} + vpdpwssds ymm28, ymm27, YMMWORD PTR [rdx+4064] # AVX512{VNNI,VL} Disp8 + vpdpwssds ymm28, ymm27, [rdx+508]{1to8} # AVX512{VNNI,VL} Disp8 + + vpdpbusd xmm26, xmm18, xmm19 # AVX512{VNNI,VL} + vpdpbusd xmm26{k6}, xmm18, xmm19 # AVX512{VNNI,VL} + vpdpbusd xmm26{k6}{z}, xmm18, xmm19 # AVX512{VNNI,VL} + vpdpbusd xmm26, xmm18, XMMWORD PTR [rax+r14*8+0x1234] # AVX512{VNNI,VL} + vpdpbusd xmm26, xmm18, XMMWORD PTR [rdx+2032] # AVX512{VNNI,VL} Disp8 + vpdpbusd xmm26, xmm18, [rdx+508]{1to4} # AVX512{VNNI,VL} Disp8 + vpdpbusd ymm21, ymm17, ymm27 # AVX512{VNNI,VL} + vpdpbusd ymm21{k2}, ymm17, ymm27 # AVX512{VNNI,VL} + vpdpbusd ymm21{k2}{z}, ymm17, ymm27 # AVX512{VNNI,VL} + vpdpbusd ymm21, ymm17, YMMWORD PTR [rax+r14*8+0x1234] # AVX512{VNNI,VL} + vpdpbusd ymm21, ymm17, YMMWORD PTR [rdx+4064] # AVX512{VNNI,VL} Disp8 + vpdpbusd ymm21, ymm17, [rdx+508]{1to8} # AVX512{VNNI,VL} Disp8 + + vpdpbusds xmm28, xmm26, xmm24 # AVX512{VNNI,VL} + vpdpbusds xmm28{k1}, xmm26, xmm24 # AVX512{VNNI,VL} + vpdpbusds xmm28{k1}{z}, xmm26, xmm24 # AVX512{VNNI,VL} + vpdpbusds xmm28, xmm26, XMMWORD PTR [rax+r14*8+0x1234] # AVX512{VNNI,VL} + vpdpbusds xmm28, xmm26, XMMWORD PTR [rdx+2032] # AVX512{VNNI,VL} Disp8 + vpdpbusds xmm28, xmm26, [rdx+508]{1to4} # AVX512{VNNI,VL} Disp8 + vpdpbusds ymm23, ymm18, ymm27 # AVX512{VNNI,VL} + vpdpbusds ymm23{k6}, ymm18, ymm27 # AVX512{VNNI,VL} + vpdpbusds ymm23{k6}{z}, ymm18, ymm27 # AVX512{VNNI,VL} + vpdpbusds ymm23, ymm18, YMMWORD PTR [rax+r14*8+0x1234] # AVX512{VNNI,VL} + vpdpbusds ymm23, ymm18, YMMWORD PTR [rdx+4064] # AVX512{VNNI,VL} Disp8 + vpdpbusds ymm23, ymm18, [rdx+508]{1to8} # AVX512{VNNI,VL} Disp8 diff --git a/opcodes/i386-dis-evex.h b/opcodes/i386-dis-evex.h index ef5c963..467a2d3 100644 --- a/opcodes/i386-dis-evex.h +++ b/opcodes/i386-dis-evex.h @@ -385,8 +385,8 @@ static const struct dis386 evex_table[][256] = { { PREFIX_TABLE (PREFIX_EVEX_0F384E) }, { PREFIX_TABLE (PREFIX_EVEX_0F384F) }, /* 50 */ - { Bad_Opcode }, - { Bad_Opcode }, + { PREFIX_TABLE (PREFIX_EVEX_0F3850) }, + { PREFIX_TABLE (PREFIX_EVEX_0F3851) }, { PREFIX_TABLE (PREFIX_EVEX_0F3852) }, { PREFIX_TABLE (PREFIX_EVEX_0F3853) }, { Bad_Opcode }, @@ -2005,18 +2005,30 @@ static const struct dis386 evex_table[][256] = { { Bad_Opcode }, { "vrsqrt14s%XW", { XMScalar, VexScalar, EXxmm_mdq }, 0 }, }, - /* PREFIX_EVEX_0F3852 */ + /* PREFIX_EVEX_0F3850 */ + { + { Bad_Opcode }, + { Bad_Opcode }, + { "vpdpbusd", { XM, Vex, EXx }, 0 }, + }, + /* PREFIX_EVEX_0F3851 */ { { Bad_Opcode }, { Bad_Opcode }, + { "vpdpbusds", { XM, Vex, EXx }, 0 }, + }, + /* PREFIX_EVEX_0F3852 */ + { + { Bad_Opcode }, { Bad_Opcode }, + { "vpdpwssd", { XM, Vex, EXx }, 0 }, { "vp4dpwssd", { XM, Vex, EXxmm }, 0 }, }, /* PREFIX_EVEX_0F3853 */ { { Bad_Opcode }, { Bad_Opcode }, - { Bad_Opcode }, + { "vpdpwssds", { XM, Vex, EXx }, 0 }, { "vp4dpwssds", { XM, Vex, EXxmm }, 0 }, }, /* PREFIX_EVEX_0F3855 */ diff --git a/opcodes/i386-dis.c b/opcodes/i386-dis.c index 7f3b18f..637fce3 100644 --- a/opcodes/i386-dis.c +++ b/opcodes/i386-dis.c @@ -1572,6 +1572,8 @@ enum PREFIX_EVEX_0F384D, PREFIX_EVEX_0F384E, PREFIX_EVEX_0F384F, + PREFIX_EVEX_0F3850, + PREFIX_EVEX_0F3851, PREFIX_EVEX_0F3852, PREFIX_EVEX_0F3853, PREFIX_EVEX_0F3855, diff --git a/opcodes/i386-gen.c b/opcodes/i386-gen.c index c04b364..1202376 100644 --- a/opcodes/i386-gen.c +++ b/opcodes/i386-gen.c @@ -225,6 +225,8 @@ static initializer cpu_flag_init[] = "CPU_AVX512F_FLAGS|CpuAVX512_VPOPCNTDQ" }, { "CPU_AVX512_VBMI2_FLAGS", "CPU_AVX512F_FLAGS|CpuAVX512_VBMI2" }, + { "CPU_AVX512_VNNI_FLAGS", + "CPU_AVX512F_FLAGS|CpuAVX512_VNNI" }, { "CPU_L1OM_FLAGS", "unknown" }, { "CPU_K1OM_FLAGS", @@ -300,7 +302,7 @@ static initializer cpu_flag_init[] = { "CPU_ANY_AVX2_FLAGS", "CpuAVX2" }, { "CPU_ANY_AVX512F_FLAGS", - "CpuVREX|CpuRegZMM|CpuRegMask|CpuAVX512CD|CpuAVX512ER|CpuAVX512PF|CpuAVX512DQ|CpuAVX512BW|CpuAVX512VL|CpuAVX512IFMA|CpuAVX512VBMI|CpuAVX512_4FMAPS|CpuAVX512_4VNNIW|CpuAVX512_VPOPCNTDQ|CpuAVX512_VBMI2|CpuAVX512F" }, + "CpuVREX|CpuRegZMM|CpuRegMask|CpuAVX512CD|CpuAVX512ER|CpuAVX512PF|CpuAVX512DQ|CpuAVX512BW|CpuAVX512VL|CpuAVX512IFMA|CpuAVX512VBMI|CpuAVX512_4FMAPS|CpuAVX512_4VNNIW|CpuAVX512_VPOPCNTDQ|CpuAVX512_VBMI2|CpuAVX512_VNNI|CpuAVX512F" }, { "CPU_ANY_AVX512CD_FLAGS", "CpuAVX512CD" }, { "CPU_ANY_AVX512ER_FLAGS", @@ -325,6 +327,8 @@ static initializer cpu_flag_init[] = "CpuAVX512_VPOPCNTDQ" }, { "CPU_ANY_AVX512_VBMI2_FLAGS", "CpuAVX512_VBMI2" }, + { "CPU_ANY_AVX512_VNNI_FLAGS", + "CpuAVX512_VNNI" }, }; static initializer operand_type_init[] = @@ -532,6 +536,7 @@ static bitfield cpu_flags[] = BITFIELD (CpuAVX512_4VNNIW), BITFIELD (CpuAVX512_VPOPCNTDQ), BITFIELD (CpuAVX512_VBMI2), + BITFIELD (CpuAVX512_VNNI), BITFIELD (CpuMWAITX), BITFIELD (CpuCLZERO), BITFIELD (CpuOSPKE), diff --git a/opcodes/i386-opc.h b/opcodes/i386-opc.h index a14f66d..34b57f5 100644 --- a/opcodes/i386-opc.h +++ b/opcodes/i386-opc.h @@ -200,6 +200,8 @@ enum CpuAVX512_VPOPCNTDQ, /* Intel AVX-512 VBMI2 Instructions support required. */ CpuAVX512_VBMI2, + /* Intel AVX-512 VNNI Instructions support required. */ + CpuAVX512_VNNI, /* mwaitx instruction required */ CpuMWAITX, /* Clzero instruction required */ @@ -243,7 +245,7 @@ enum /* If you get a compiler error for zero width of the unused field, comment it out. */ - #define CpuUnused (CpuMax + 1) +#define CpuUnused (CpuMax + 1) /* We can check if an instruction is available with array instead of bitfield. */ @@ -335,6 +337,7 @@ typedef union i386_cpu_flags unsigned int cpuavx512_4vnniw:1; unsigned int cpuavx512_vpopcntdq:1; unsigned int cpuavx512_vbmi2:1; + unsigned int cpuavx512_vnni:1; unsigned int cpumwaitx:1; unsigned int cpuclzero:1; unsigned int cpuospke:1; diff --git a/opcodes/i386-opc.tbl b/opcodes/i386-opc.tbl index 0dcc17b..6b7dea7 100644 --- a/opcodes/i386-opc.tbl +++ b/opcodes/i386-opc.tbl @@ -6075,6 +6075,26 @@ vpshrdw, 4, 0x6672, None, 1, CpuAVX512_VBMI2|CpuAVX512VL, Modrm|EVex=3|Masking=3 // AVX512_VBMI2 instructions end +// AVX512_VNNI instructions + +vpdpbusd, 3, 0x6650, None, 1, CpuAVX512_VNNI, Modrm|EVex=1|Masking=3|VexOpcode=1|VexVVVV=1|VexW=1|Broadcast=1|Disp8MemShift=6|IgnoreSize|No_bSuf|No_wSuf|No_lSuf|No_sSuf|No_qSuf|No_ldSuf, { RegZMM|Dword|ZMMword|Unspecified|BaseIndex|Disp8|Disp16|Disp32|Disp32S|Vec_Disp8, RegZMM, RegZMM } +vpdpbusd, 3, 0x6650, None, 1, CpuAVX512_VNNI|CpuAVX512VL, Modrm|EVex=2|Masking=3|VexOpcode=1|VexVVVV=1|VexW=1|Broadcast=3|Disp8MemShift=4|IgnoreSize|No_bSuf|No_wSuf|No_lSuf|No_sSuf|No_qSuf|No_ldSuf, { RegXMM|Dword|XMMword|Unspecified|BaseIndex|Disp8|Disp16|Disp32|Disp32S|Vec_Disp8, RegXMM, RegXMM } +vpdpbusd, 3, 0x6650, None, 1, CpuAVX512_VNNI|CpuAVX512VL, Modrm|EVex=3|Masking=3|VexOpcode=1|VexVVVV=1|VexW=1|Broadcast=2|Disp8MemShift=5|IgnoreSize|No_bSuf|No_wSuf|No_lSuf|No_sSuf|No_qSuf|No_ldSuf, { RegYMM|Dword|YMMword|Unspecified|BaseIndex|Disp8|Disp16|Disp32|Disp32S|Vec_Disp8, RegYMM, RegYMM } + +vpdpbusds, 3, 0x6651, None, 1, CpuAVX512_VNNI, Modrm|EVex=1|Masking=3|VexOpcode=1|VexVVVV=1|VexW=1|Broadcast=1|Disp8MemShift=6|IgnoreSize|No_bSuf|No_wSuf|No_lSuf|No_sSuf|No_qSuf|No_ldSuf, { RegZMM|Dword|ZMMword|Unspecified|BaseIndex|Disp8|Disp16|Disp32|Disp32S|Vec_Disp8, RegZMM, RegZMM } +vpdpbusds, 3, 0x6651, None, 1, CpuAVX512_VNNI|CpuAVX512VL, Modrm|EVex=2|Masking=3|VexOpcode=1|VexVVVV=1|VexW=1|Broadcast=3|Disp8MemShift=4|IgnoreSize|No_bSuf|No_wSuf|No_lSuf|No_sSuf|No_qSuf|No_ldSuf, { RegXMM|Dword|XMMword|Unspecified|BaseIndex|Disp8|Disp16|Disp32|Disp32S|Vec_Disp8, RegXMM, RegXMM } +vpdpbusds, 3, 0x6651, None, 1, CpuAVX512_VNNI|CpuAVX512VL, Modrm|EVex=3|Masking=3|VexOpcode=1|VexVVVV=1|VexW=1|Broadcast=2|Disp8MemShift=5|IgnoreSize|No_bSuf|No_wSuf|No_lSuf|No_sSuf|No_qSuf|No_ldSuf, { RegYMM|Dword|YMMword|Unspecified|BaseIndex|Disp8|Disp16|Disp32|Disp32S|Vec_Disp8, RegYMM, RegYMM } + +vpdpwssd, 3, 0x6652, None, 1, CpuAVX512_VNNI, Modrm|EVex=1|Masking=3|VexOpcode=1|VexVVVV=1|VexW=1|Broadcast=1|Disp8MemShift=6|IgnoreSize|No_bSuf|No_wSuf|No_lSuf|No_sSuf|No_qSuf|No_ldSuf, { RegZMM|Dword|ZMMword|Unspecified|BaseIndex|Disp8|Disp16|Disp32|Disp32S|Vec_Disp8, RegZMM, RegZMM } +vpdpwssd, 3, 0x6652, None, 1, CpuAVX512_VNNI|CpuAVX512VL, Modrm|EVex=2|Masking=3|VexOpcode=1|VexVVVV=1|VexW=1|Broadcast=3|Disp8MemShift=4|IgnoreSize|No_bSuf|No_wSuf|No_lSuf|No_sSuf|No_qSuf|No_ldSuf, { RegXMM|Dword|XMMword|Unspecified|BaseIndex|Disp8|Disp16|Disp32|Disp32S|Vec_Disp8, RegXMM, RegXMM } +vpdpwssd, 3, 0x6652, None, 1, CpuAVX512_VNNI|CpuAVX512VL, Modrm|EVex=3|Masking=3|VexOpcode=1|VexVVVV=1|VexW=1|Broadcast=2|Disp8MemShift=5|IgnoreSize|No_bSuf|No_wSuf|No_lSuf|No_sSuf|No_qSuf|No_ldSuf, { RegYMM|Dword|YMMword|Unspecified|BaseIndex|Disp8|Disp16|Disp32|Disp32S|Vec_Disp8, RegYMM, RegYMM } + +vpdpwssds, 3, 0x6653, None, 1, CpuAVX512_VNNI, Modrm|EVex=1|Masking=3|VexOpcode=1|VexVVVV=1|VexW=1|Broadcast=1|Disp8MemShift=6|IgnoreSize|No_bSuf|No_wSuf|No_lSuf|No_sSuf|No_qSuf|No_ldSuf, { RegZMM|Dword|ZMMword|Unspecified|BaseIndex|Disp8|Disp16|Disp32|Disp32S|Vec_Disp8, RegZMM, RegZMM } +vpdpwssds, 3, 0x6653, None, 1, CpuAVX512_VNNI|CpuAVX512VL, Modrm|EVex=2|Masking=3|VexOpcode=1|VexVVVV=1|VexW=1|Broadcast=3|Disp8MemShift=4|IgnoreSize|No_bSuf|No_wSuf|No_lSuf|No_sSuf|No_qSuf|No_ldSuf, { RegXMM|Dword|XMMword|Unspecified|BaseIndex|Disp8|Disp16|Disp32|Disp32S|Vec_Disp8, RegXMM, RegXMM } +vpdpwssds, 3, 0x6653, None, 1, CpuAVX512_VNNI|CpuAVX512VL, Modrm|EVex=3|Masking=3|VexOpcode=1|VexVVVV=1|VexW=1|Broadcast=2|Disp8MemShift=5|IgnoreSize|No_bSuf|No_wSuf|No_lSuf|No_sSuf|No_qSuf|No_ldSuf, { RegYMM|Dword|YMMword|Unspecified|BaseIndex|Disp8|Disp16|Disp32|Disp32S|Vec_Disp8, RegYMM, RegYMM } + +// AVX512_VNNI instructions end + // AVX512 + GFNI instructions vgf2p8affineinvqb, 4, 0x66cf, None, 1, CpuAVX512F|CpuGFNI, Modrm|EVex=1|Masking=3|VexOpcode=2|VexVVVV=1|VexW=2|VecESize=1|Broadcast=2|Disp8MemShift=6|IgnoreSize|No_bSuf|No_wSuf|No_lSuf|No_sSuf|No_qSuf|No_ldSuf, { Imm8, RegZMM|Qword|ZMMword|Unspecified|BaseIndex|Disp8|Disp16|Disp32|Disp32S|Vec_Disp8, RegZMM, RegZMM } |