http://llvm.org/bugs/show_bug.cgi?id=10851
Summary: [x86 disassembler] vmovmskpd disassembled when
reserved bits not set
Product: libraries
Version: trunk
Platform: PC
OS/Version: All
Status: NEW
Severity: normal
Priority: P
Component: Backend: X86
AssignedTo: [email protected]
ReportedBy: [email protected]
CC: [email protected]
From the "Intel® 64 and IA-32 Architectures Software Developer’s Manual
Volume 2 (2A & 2B): Instruction Set Reference, A-Z", p. 3-691:
VEX.128.66.0F.WIG 50 /r
VMOVMSKPD reg, xmm2
VEX.256.66.0F.WIG 50 /r
VMOVMSKPD reg, ymm2
Note: In VEX-encoded versions, VEX.vvvv is reserved and must be 1111b,
otherwise
instructions will #UD.
Clang on OSX assembles the following:
vmovmskpd %xmm0, %eax
to these bytes:
C5 F9 50 C0
But using llvm-mc built from trunk r139013, this byte sequence disassembles
even though the VEX.vvvv field is not 1111b (it's 0000b):
$ echo '0xc5 0x81 0x50 0xc0'| ./llvm-mc -disassemble -triple="x86_64"
vmovmskpd %xmm0, %eax
The 3-byte VEX prefix form is also incorrect:
$ echo '0xc4 0xe1 0x81 0x50 0xc0'| ./llvm-mc -disassemble -triple="x86_64"
vmovmskpd %xmm0, %eax
--
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.
_______________________________________________
LLVMbugs mailing list
[email protected]
http://lists.cs.uiuc.edu/mailman/listinfo/llvmbugs
