[PATCH] x86: fix VMOVNTDQA placement and tagging
Jan Beulich
JBeulich@suse.com
Tue Jul 24 14:20:00 GMT 2012
VMOVNTDQA was both misplaced and improperly tagged as being an AVX
instruction (instead of AVX2).
2012-07-24 Jan Beulich <jbeulich@suse.com>
* i386-opc.tbl (vmovntdqa): Move up into 256-bit integer AVX2
instruction group. Mark as requiring AVX2.
* i386-tbl.h: Re-generate.
--- a/opcodes/i386-opc.tbl
+++ b/opcodes/i386-opc.tbl
@@ -2386,6 +2386,7 @@ vzeroupper, 0, 0x77, None, 1, CpuAVX, Ve
// 256bit integer AVX2 instructions.
+vmovntdqa, 2, 0x662a, None, 1, CpuAVX2, Modrm|Vex=2|VexOpcode=1|VexW=1|IgnoreSize|No_bSuf|No_wSuf|No_lSuf|No_sSuf|No_qSuf|No_ldSuf, { Ymmword|Unspecified|BaseIndex|Disp8|Disp16|Disp32|Disp32S, RegYMM }
vmpsadbw, 4, 0x6642, None, 1, CpuAVX2, Modrm|Vex=2|VexOpcode=2|VexVVVV=1|VexW=1|IgnoreSize|No_bSuf|No_wSuf|No_lSuf|No_sSuf|No_qSuf|No_ldSuf, { Imm8, Ymmword|Unspecified|BaseIndex|Disp8|Disp16|Disp32|Disp32S|RegYMM, RegYMM, RegYMM }
vpabsb, 2, 0x661c, None, 1, CpuAVX2, Modrm|Vex=2|VexOpcode=1|VexW=1|IgnoreSize|No_bSuf|No_wSuf|No_lSuf|No_sSuf|No_qSuf|No_ldSuf, { Ymmword|Unspecified|BaseIndex|Disp8|Disp16|Disp32|Disp32S|RegYMM, RegYMM }
vpabsd, 2, 0x661e, None, 1, CpuAVX2, Modrm|Vex=2|VexOpcode=1|VexW=1|IgnoreSize|No_bSuf|No_wSuf|No_lSuf|No_sSuf|No_qSuf|No_ldSuf, { Ymmword|Unspecified|BaseIndex|Disp8|Disp16|Disp32|Disp32S|RegYMM, RegYMM }
@@ -2508,7 +2509,6 @@ vbroadcasti128, 2, 0x665A, None, 1, CpuA
vbroadcastsd, 2, 0x6619, None, 1, CpuAVX2, Modrm|Vex=2|VexOpcode=1|VexW=1|IgnoreSize|No_bSuf|No_wSuf|No_lSuf|No_sSuf|No_qSuf|No_ldSuf, { RegXMM, RegYMM }
vbroadcastss, 2, 0x6618, None, 1, CpuAVX2, Modrm|Vex|VexOpcode=1|VexW=1|IgnoreSize|No_bSuf|No_wSuf|No_lSuf|No_sSuf|No_qSuf|No_ldSuf, { RegXMM, RegXMM }
vbroadcastss, 2, 0x6618, None, 1, CpuAVX2, Modrm|Vex=2|VexOpcode=1|VexW=1|IgnoreSize|No_bSuf|No_wSuf|No_lSuf|No_sSuf|No_qSuf|No_ldSuf, { RegXMM, RegYMM }
-vmovntdqa, 2, 0x662a, None, 1, CpuAVX, Modrm|Vex=2|VexOpcode=1|VexW=1|IgnoreSize|No_bSuf|No_wSuf|No_lSuf|No_sSuf|No_qSuf|No_ldSuf, { Ymmword|Unspecified|BaseIndex|Disp8|Disp16|Disp32|Disp32S, RegYMM }
vpblendd, 4, 0x6602, None, 1, CpuAVX2, Modrm|Vex|VexOpcode=2|VexVVVV=1|VexW=1|IgnoreSize|No_bSuf|No_wSuf|No_lSuf|No_sSuf|No_qSuf|No_ldSuf, { Imm8, Xmmword|Unspecified|BaseIndex|Disp8|Disp16|Disp32|Disp32S|RegXMM, RegXMM, RegXMM }
vpblendd, 4, 0x6602, None, 1, CpuAVX2, Modrm|Vex=2|VexOpcode=2|VexVVVV=1|VexW=1|IgnoreSize|No_bSuf|No_wSuf|No_lSuf|No_sSuf|No_qSuf|No_ldSuf, { Imm8, Ymmword|Unspecified|BaseIndex|Disp8|Disp16|Disp32|Disp32S|RegYMM, RegYMM, RegYMM }
vpbroadcastb, 2, 0x6678, None, 1, CpuAVX2, Modrm|Vex=1|VexOpcode=1|VexW=1|IgnoreSize|No_bSuf|No_wSuf|No_lSuf|No_sSuf|No_qSuf|No_ldSuf, { Byte|Unspecified|BaseIndex|Disp8|Disp16|Disp32|Disp32S|RegXMM, RegXMM }
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: binutils-mainline-x86-vmovntdqa-256bit-AVX2.patch
URL: <https://sourceware.org/pipermail/binutils/attachments/20120724/d342b37b/attachment.ksh>
More information about the Binutils
mailing list