[compiler-rt] 5f40570 - [win/asan] GetInstructionSize: Support some more 5 byte instructions. (#123844)

[via llvm-commits]

Author: bernhardu
Date: 2025-01-22T10:22:02+01:00
New Revision: 5f40570714b355a0ce10ef3e137dc0dc540611d6

URL: https://github.com/llvm/llvm-project/commit/5f40570714b355a0ce10ef3e137dc0dc540611d6
DIFF: https://github.com/llvm/llvm-project/commit/5f40570714b355a0ce10ef3e137dc0dc540611d6.diff

LOG: [win/asan] GetInstructionSize: Support some more 5 byte instructions. (#123844)

This patch adds several instructions seen when trying to run a
executable built with ASan with llvm-mingw.
(x86 and x86_64, using the git tip in llvm-project).

Also includes instructions collected by
Roman Pišl and Eric Pouech in the Wine bug reports below.
```
Related: https://github.com/llvm/llvm-project/issues/96270

Co-authored-by: Roman Pišl <rpisl at seznam.cz>
                https://bugs.winehq.org/show_bug.cgi?id=50993
                https://bugs.winehq.org/attachment.cgi?id=70233
Co-authored-by: Eric Pouech <eric.pouech at gmail.com>
                https://bugs.winehq.org/show_bug.cgi?id=52386
                https://bugs.winehq.org/attachment.cgi?id=71626
```

Added: 
    

Modified: 
    compiler-rt/lib/interception/interception_win.cpp
    compiler-rt/lib/interception/tests/interception_win_test.cpp

Removed: 
    


################################################################################
diff  --git a/compiler-rt/lib/interception/interception_win.cpp b/compiler-rt/lib/interception/interception_win.cpp
index 0841161ee5c428..f42bec0d3e6817 100644
--- a/compiler-rt/lib/interception/interception_win.cpp
+++ b/compiler-rt/lib/interception/interception_win.cpp
@@ -606,6 +606,7 @@ static size_t GetInstructionSize(uptr address, size_t* rel_offset = nullptr) {
 
     case 0xb8:  // b8 XX XX XX XX : mov eax, XX XX XX XX
     case 0xB9:  // b9 XX XX XX XX : mov ecx, XX XX XX XX
+    case 0xBA:  // ba XX XX XX XX : mov edx, XX XX XX XX
       return 5;
 
     // Cannot overwrite control-instruction. Return 0 to indicate failure.
@@ -831,6 +832,7 @@ static size_t GetInstructionSize(uptr address, size_t* rel_offset = nullptr) {
     case 0xf88349:    // 49 83 f8 XX : cmp r8, XX
       return 4;
 
+    case 0x441F0F:  // 0F 1F 44 XX XX :   nop DWORD PTR [...]
     case 0x246483:  // 83 64 24 XX YY :   and    DWORD PTR [rsp+XX], YY
       return 5;
 
@@ -902,6 +904,11 @@ static size_t GetInstructionSize(uptr address, size_t* rel_offset = nullptr) {
       return 6;
   }
 
+  switch (0xFFFFFFFFFFULL & *(u64 *)(address)) {
+    case 0xC07E0F4866:  // 66 48 0F 7E C0 : movq rax, xmm0
+      return 5;
+  }
+
 #else
 
   switch (*(u8*)address) {

diff  --git a/compiler-rt/lib/interception/tests/interception_win_test.cpp b/compiler-rt/lib/interception/tests/interception_win_test.cpp
index 9d8dbfcb4285fb..a0a35d085f85b0 100644
--- a/compiler-rt/lib/interception/tests/interception_win_test.cpp
+++ b/compiler-rt/lib/interception/tests/interception_win_test.cpp
@@ -867,6 +867,7 @@ const struct InstructionSizeData {
     { 5, {0x68, 0x71, 0x72, 0x73, 0x74}, 0, "68 XX XX XX XX : push imm32"},
     { 5, {0xb8, 0x71, 0x72, 0x73, 0x74}, 0, "b8 XX XX XX XX : mov eax, XX XX XX XX"},
     { 5, {0xB9, 0x71, 0x72, 0x73, 0x74}, 0, "b9 XX XX XX XX : mov ecx, XX XX XX XX"},
+    { 5, {0xBA, 0x71, 0x72, 0x73, 0x74}, 0, "ba XX XX XX XX : mov edx, XX XX XX XX"},
     { 7, {0x8D, 0xA4, 0x24, 0x73, 0x74, 0x75, 0x76}, 0, "8D A4 24 XX XX XX XX : lea esp, [esp + XX XX XX XX]"},
 #if SANITIZER_WINDOWS_x64
     // sorted list
@@ -973,6 +974,7 @@ const struct InstructionSizeData {
     { 4, {0x80, 0x7D, 0x72, 0x73}, 0, "80 7D YY XX : cmp BYTE PTR [rbp+YY], XX"},
     { 4, {0x80, 0x7E, 0x72, 0x73}, 0, "80 7E YY XX : cmp BYTE PTR [rsi+YY], XX"},
     { 4, {0x89, 0x54, 0x24, 0x73}, 0, "89 54 24 XX : mov DWORD PTR[rsp + XX], edx"},
+    { 5, {0x0F, 0x1F, 0x44, 0x73, 0x74}, 0, "0F 1F 44 XX XX : nop DWORD PTR [...]"},
     { 5, {0x44, 0x89, 0x44, 0x24, 0x74}, 0, "44 89 44 24 XX : mov DWORD PTR [rsp + XX], r8d"},
     { 5, {0x44, 0x89, 0x4c, 0x24, 0x74}, 0, "44 89 4c 24 XX : mov DWORD PTR [rsp + XX], r9d"},
     { 5, {0x48, 0x89, 0x4C, 0x24, 0x74}, 0, "48 89 4C 24 XX : mov QWORD PTR [rsp + XX], rcx"},
@@ -985,6 +987,7 @@ const struct InstructionSizeData {
     { 5, {0x48, 0x8d, 0x6c, 0x24, 0x74}, 0, "48 8d 6c 24 XX : lea rbp, [rsp + XX]"},
     { 5, {0x4c, 0x89, 0x44, 0x24, 0x74}, 0, "4c 89 44 24 XX : mov QWORD PTR [rsp + XX], r8"},
     { 5, {0x4c, 0x89, 0x4c, 0x24, 0x74}, 0, "4c 89 4c 24 XX : mov QWORD PTR [rsp + XX], r9"},
+    { 5, {0x66, 0x48, 0x0F, 0x7E, 0xC0}, 0, "66 48 0F 7E C0 : movq rax, xmm0"},
     { 5, {0x83, 0x44, 0x72, 0x73, 0x74}, 0, "83 44 72 XX YY : add DWORD PTR [rdx+rsi*2+XX],YY"},
     { 5, {0x83, 0x64, 0x24, 0x73, 0x74}, 0, "83 64 24 XX YY : and DWORD PTR [rsp+XX], YY"},
     { 6, {0x48, 0x83, 0x64, 0x24, 0x74, 0x75}, 0, "48 83 64 24 XX YY : and QWORD PTR [rsp + XX], YY"},