Skip to contentSkip to conversation

x86/insn: Fix to decode bsr/bsf/jmpe with operand-size prefix

PATCH In review Masami Hiramatsu 1 participant 2 messages Jun 4, 2012 – Jun 6, 2012 x86

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 x86/insn: Fix to decode bsr/bsf/jmpe with operand-size prefix

Fix x86 instruction decoder to decode bsr/bsf/jmpe with operand-size prefix (66h). This fixes below failure of the test case.

bsf/bsr/jmpe have a special encoding. Opcode map in Intel Software Developers Manual vol2 says they have TZCNT/LZCNT variants if it has F3h prefix. However, there is no information if it has other 66h or F2h prefixes. Current instruction decoder supposes that those are a bad instruction, but it actually accept at least operand-size prefix.

This fixes errors reported by test_get_len as below:

Warning: arch/x86/tools/test_get_len found difference at
<em_bsf>:ffffffff81036d87
Warning: ffffffff81036de5: 66 0f bc c2 bsf %dx,%ax
Warning: objdump says 4 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at
<em_bsr>:ffffffff81036ea6
Warning: ffffffff81036f04: 66 0f bd c2 bsr %dx,%ax
Warning: objdump says 4 bytes, but insn_get_length() says 3
Warning: decoded and checked 13298882 instructions with 2 warnings

---

 arch/x86/lib/x86-opcode-map.txt      |    8 ++++----
 arch/x86/tools/gen-insn-attr-x86.awk |   14 +++++++++-----
 2 files changed, 13 insertions(+), 9 deletions(-)
arch/x86/lib/x86-opcode-map.txt +4 -4
index 8191379..5d7e51f 100644
@@ -28,7 +28,7 @@
 #  - (66): the last prefix is 0x66
 #  - (F3): the last prefix is 0xF3
 #  - (F2): the last prefix is 0xF2
-#
+#  - (!F3) : the last prefix is not 0xF3 (including non-last prefix case)
 
 Table: one byte opcode
 Referrer:
@@ -515,12 +515,12 @@ b4: LFS Gv,Mp
 b5: LGS Gv,Mp
 b6: MOVZX Gv,Eb
 b7: MOVZX Gv,Ew
-b8: JMPE | POPCNT Gv,Ev (F3)
+b8: JMPE (!F3) | POPCNT Gv,Ev (F3)
 b9: Grp10 (1A)
 ba: Grp8 Ev,Ib (1A)
 bb: BTC Ev,Gv
-bc: BSF Gv,Ev | TZCNT Gv,Ev (F3)
-bd: BSR Gv,Ev | LZCNT Gv,Ev (F3)
+bc: BSF Gv,Ev (!F3) | TZCNT Gv,Ev (F3)
+bd: BSR Gv,Ev (!F3) | LZCNT Gv,Ev (F3)
 be: MOVSX Gv,Eb
 bf: MOVSX Gv,Ew
 # 0x0f 0xc0-0xcf
arch/x86/tools/gen-insn-attr-x86.awk +9 -5
index 5f6a5b6..ddcf39b 100644
@@ -66,9 +66,10 @@ BEGIN {
 	rex_expr = "^REX(\\.[XRWB]+)*"
 	fpu_expr = "^ESC" # TODO
 
-	lprefix1_expr = "\\(66\\)"
+	lprefix1_expr = "\\((66|!F3)\\)"
 	lprefix2_expr = "\\(F3\\)"
-	lprefix3_expr = "\\(F2\\)"
+	lprefix3_expr = "\\((F2|!F3)\\)"
+	lprefix_expr = "\\((66|F2|F3)\\)"
 	max_lprefix = 4
 
 	# All opcodes starting with lower-case 'v' or with (v1) superscript
@@ -333,13 +334,16 @@ function convert_operands(count,opnd,       i,j,imm,mod)
 		if (match(ext, lprefix1_expr)) {
 			lptable1[idx] = add_flags(lptable1[idx],flags)
 			variant = "INAT_VARIANT"
-		} else if (match(ext, lprefix2_expr)) {
+		}
+		if (match(ext, lprefix2_expr)) {
 			lptable2[idx] = add_flags(lptable2[idx],flags)
 			variant = "INAT_VARIANT"
-		} else if (match(ext, lprefix3_expr)) {
+		}
+		if (match(ext, lprefix3_expr)) {
 			lptable3[idx] = add_flags(lptable3[idx],flags)
 			variant = "INAT_VARIANT"
-		} else {
+		}
+		if (!match(ext, lprefix_expr)){
 			table[idx] = add_flags(table[idx],flags)
 		}
 	}
#1 lore raw

[tip:perf/core] x86/decoder: Fix bsr/bsf/ jmpe decoding with operand-size prefix

Commit-ID:  436d03faf6961b30e13b2d0967aea9d772d6cf44
Gitweb:     http://git.kernel.org/tip/436d03faf6961b30e13b2d0967aea9d772d6cf44
Author:     Masami Hiramatsu <[email protected]>
AuthorDate: Tue, 5 Jun 2012 00:09:11 +0900
Committer:  Ingo Molnar <[email protected]>
CommitDate: Wed, 6 Jun 2012 08:54:18 +0200

x86/decoder: Fix bsr/bsf/jmpe decoding with operand-size prefix

Fix the x86 instruction decoder to decode bsr/bsf/jmpe with operand-size prefix (66h). This fixes the test case failure reported by Linus, attached below.

bsf/bsr/jmpe have a special encoding. Opcode map in Intel Software Developers Manual vol2 says they have TZCNT/LZCNT variants if it has F3h prefix. However, there is no information if it has other 66h or F2h prefixes. Current instruction decoder supposes that those are bad instructions, but it actually accepts at least operand-size prefixes.

H. Peter Anvin further explains:

" TZCNT/LZCNT are F3 + BSF/BSR exactly because the F2 and
F3 prefixes have historically been no-ops with most instructions. This allows software to unconditionally use the prefixed versions and get TZCNT/LZCNT on the processors that have them if they don't care about the difference. "

This fixes errors reported by test_get_len:

Warning: arch/x86/tools/test_get_len found difference at <em_bsf>:ffffffff81036d87
Warning: ffffffff81036de5: 66 0f bc c2 bsf %dx,%ax
Warning: objdump says 4 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <em_bsr>:ffffffff81036ea6
Warning: ffffffff81036f04: 66 0f bd c2 bsr %dx,%ax
Warning: objdump says 4 bytes, but insn_get_length() says 3
Warning: decoded and checked 13298882 instructions with 2 warnings

---
 arch/x86/lib/x86-opcode-map.txt      |    8 ++++----
 arch/x86/tools/gen-insn-attr-x86.awk |   14 +++++++++-----
 2 files changed, 13 insertions(+), 9 deletions(-)
arch/x86/lib/x86-opcode-map.txt +4 -4
index 8191379..5d7e51f 100644
@@ -28,7 +28,7 @@
 #  - (66): the last prefix is 0x66
 #  - (F3): the last prefix is 0xF3
 #  - (F2): the last prefix is 0xF2
-#
+#  - (!F3) : the last prefix is not 0xF3 (including non-last prefix case)
 
 Table: one byte opcode
 Referrer:
@@ -515,12 +515,12 @@ b4: LFS Gv,Mp
 b5: LGS Gv,Mp
 b6: MOVZX Gv,Eb
 b7: MOVZX Gv,Ew
-b8: JMPE | POPCNT Gv,Ev (F3)
+b8: JMPE (!F3) | POPCNT Gv,Ev (F3)
 b9: Grp10 (1A)
 ba: Grp8 Ev,Ib (1A)
 bb: BTC Ev,Gv
-bc: BSF Gv,Ev | TZCNT Gv,Ev (F3)
-bd: BSR Gv,Ev | LZCNT Gv,Ev (F3)
+bc: BSF Gv,Ev (!F3) | TZCNT Gv,Ev (F3)
+bd: BSR Gv,Ev (!F3) | LZCNT Gv,Ev (F3)
 be: MOVSX Gv,Eb
 bf: MOVSX Gv,Ew
 # 0x0f 0xc0-0xcf
arch/x86/tools/gen-insn-attr-x86.awk +9 -5
index 5f6a5b6..ddcf39b 100644
@@ -66,9 +66,10 @@ BEGIN {
 	rex_expr = "^REX(\\.[XRWB]+)*"
 	fpu_expr = "^ESC" # TODO
 
-	lprefix1_expr = "\\(66\\)"
+	lprefix1_expr = "\\((66|!F3)\\)"
 	lprefix2_expr = "\\(F3\\)"
-	lprefix3_expr = "\\(F2\\)"
+	lprefix3_expr = "\\((F2|!F3)\\)"
+	lprefix_expr = "\\((66|F2|F3)\\)"
 	max_lprefix = 4
 
 	# All opcodes starting with lower-case 'v' or with (v1) superscript
@@ -333,13 +334,16 @@ function convert_operands(count,opnd,       i,j,imm,mod)
 		if (match(ext, lprefix1_expr)) {
 			lptable1[idx] = add_flags(lptable1[idx],flags)
 			variant = "INAT_VARIANT"
-		} else if (match(ext, lprefix2_expr)) {
+		}
+		if (match(ext, lprefix2_expr)) {
 			lptable2[idx] = add_flags(lptable2[idx],flags)
 			variant = "INAT_VARIANT"
-		} else if (match(ext, lprefix3_expr)) {
+		}
+		if (match(ext, lprefix3_expr)) {
 			lptable3[idx] = add_flags(lptable3[idx],flags)
 			variant = "INAT_VARIANT"
-		} else {
+		}
+		if (!match(ext, lprefix_expr)){
 			table[idx] = add_flags(table[idx],flags)
 		}
 	}