Fix the x86_64 implementation of call_rwsem_wait()
PATCH 1/1 Fix the x86_64 implementation of call_rwsem_wait()
The x86_64 call_rwsem_wait() treats the active state counter part of the R/W semaphore state as being 16-bit when it's actually 32-bit (it's half of the 64-bit state). It should do "decl %edx" not "decw %dx".
---
arch/x86/lib/rwsem_64.S | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-)
arch/x86/lib/rwsem_64.S +1 -1
@@ -60,7 +60,7 @@ ENTRY(call_rwsem_down_write_failed) ENDPROC(call_rwsem_down_write_failed) ENTRY(call_rwsem_wake) - decw %dx /* do nothing if still outstanding active readers */ + decl %edx /* do nothing if still outstanding active readers */ jnz 1f save_common_regs movq %rax,%rdi
[tip:x86/urgent] x86-64: Use the proper register width in call_rwsem_wait()
Commit-ID: 50db45f2f010e9251dae644c5fb86b321d9ac815 Gitweb: http://git.kernel.org/tip/50db45f2f010e9251dae644c5fb86b321d9ac815 Author: David Howells <[email protected]> AuthorDate: Tue, 4 May 2010 13:42:53 +0100 Committer: H. Peter Anvin <[email protected]> CommitDate: Tue, 4 May 2010 10:15:54 -0700
x86-64: Use the proper register width in call_rwsem_wait()
The x86_64 call_rwsem_wait() treats the active state counter part of the R/W semaphore state as being 16-bit when it's actually 32-bit (it's half of the 64-bit state). It should do "decl %edx" not "decw %dx".
[ hpa note: decl %edx will clear the upper half of %rdx, but the value of
%rdx is dead after this point so that is OK. ]
LKML-Reference: <[email protected]>
--- arch/x86/lib/rwsem_64.S | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-)
arch/x86/lib/rwsem_64.S +1 -1
@@ -60,7 +60,7 @@ ENTRY(call_rwsem_down_write_failed) ENDPROC(call_rwsem_down_write_failed) ENTRY(call_rwsem_wake) - decw %dx /* do nothing if still outstanding active readers */ + decl %edx /* do nothing if still outstanding active readers */ jnz 1f save_common_regs movq %rax,%rdi