Landing: 0be368a40aea
Project / Subsystem
binutils / x86
Date
2026-08-14
Author
Jan Beulich <jbeulich@sourceware.org>
Commit
0be368a40aeaaa937faa66424793fcb5024ea3a8
Source
public_inbox
Perf win
No
Breaking
No
All attributes
- project
- binutils
- subsystem
- x86
- patch_id
- —
- commit_hash
- 0be368a40aeaaa937faa66424793fcb5024ea3a8
- source_type
- public_inbox
- headline
- x86: Correct operand type usage in VEX prefix generation
- tldr
- Ensures the correct operand types are used when building VEX prefixes in the x86 assembler.
- author
- Jan Beulich <jbeulich@sourceware.org>
- outcome
- committed
- performance_win
- false
- breaking_change
- false
- series_id
- —
- series_parts
- []
- tags
-
- • x86
- • assembler
- • bugfix
- discussion_id_link
- —
- bugzilla_pr
- —
- date
- 2026-08-14T00:00:00.000Z
This commit corrects how operand types are accessed when building VEX prefixes in the x86 assembler. Previously, the code might have incorrectly used the operand_types from the currently installed template (t->operand_types[]) when it should have referred to the types associated with the instruction instance being processed (i.tm_types[]). This change ensures the VEX prefix logic correctly uses the operand types of the intended instruction, preventing potential issues.