http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53355
Richard Guenther changed:
What|Removed |Added
CC||falk at debian dot org
--- Comment #6
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53355
--- Comment #5 from Richard Guenther 2012-05-18
12:32:11 UTC ---
The following testcase (extracted from PR53346, which stores zero though) shows
another case of slowness
void foo (int *p, int n)
{
int i;
for (i = 0; i < n; ++i)
p[i] = 1;
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53355
--- Comment #4 from Richard Guenther 2012-05-16
14:46:33 UTC ---
One major remaining issue is that the entry checks for the versioned loops
all base on the number of iterations of said loop instead of on the
feature (in the case of peeling for al
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53355
--- Comment #3 from Richard Guenther 2012-05-15
13:19:06 UTC ---
With the patch we now emit
foo:
.LFB0:
.cfi_startproc
movq%rdi, %rax
salq$60, %rax
sarq$63, %rax
movq%rax, %rdx
andl
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53355
--- Comment #2 from Richard Guenther 2012-05-15
13:18:42 UTC ---
Author: rguenth
Date: Tue May 15 13:18:32 2012
New Revision: 187535
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=187535
Log:
2012-05-15 Richard Guenther
PR tree-op
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53355
Richard Guenther changed:
What|Removed |Added
Status|UNCONFIRMED |ASSIGNED
Last reconfirmed|
