------- Comment #7 from hjl dot tools at gmail dot com 2009-11-20 04:00 -------
(In reply to comment #6)
> The good ones produce:
>
> 650: 55 push %ebp
> 651: 89 e5 mov %esp,%ebp
> 653: 83 e4 f0 and $0xfffffff0,%esp
>
> The bad one:
>
> 000005f0 <timer_stats_update_stats>:
> 5f0: 57 push %edi
> 5f1: 8d 7c 24 08 lea 0x8(%esp),%edi
> 5f5: 83 e4 f0 and $0xfffffff0,%esp
> 5f8: ff 77 fc pushl -0x4(%edi)
> 5fb: 55 push %ebp
> 5fc: 89 e5 mov %esp,%ebp
>
> It's worse code for no reason and breaks the kernel assumption of ebp + 4
> pointing to the real return address on the stack.
I think the difference comes from DRAP:
/* Nonzero if function being compiled needs dynamic realigned
argument pointer (drap) if stack needs realigning. */
bool need_drap;
It may be triggered by -mno-accumulate-outgoing-args, alloca,
long jump, ...
--
hjl dot tools at gmail dot com changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |hjl dot tools at gmail dot
| |com
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42109
