public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
* [Bug tree-optimization/46969] New: -O -ftree-vectorize -ftree-parallelize-loops=2 -fcompare-debug
@ 2010-12-15 19:27 zsojka at seznam dot cz
  2010-12-15 19:30 ` [Bug tree-optimization/46969] -fcompare-debug failure with -O -ftree-vectorize -ftree-parallelize-loops=2 zsojka at seznam dot cz
                   ` (5 more replies)
  0 siblings, 6 replies; 7+ messages in thread
From: zsojka at seznam dot cz @ 2010-12-15 19:27 UTC (permalink / raw)
  To: gcc-bugs

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46969

           Summary: -O -ftree-vectorize -ftree-parallelize-loops=2
                    -fcompare-debug
           Product: gcc
           Version: 4.6.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: tree-optimization
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: zsojka@seznam.cz
                CC: aoliva@gcc.gnu.org
              Host: x86_64-pc-linux-gnu
            Target: x86_64-pc-linux-gnu


Created attachment 22772
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=22772
reduced testcase (from gcc.dg/vect/pr44507.c)

This problem seems to be quite fragile.

$ gcc -O -ftree-vectoriztree-parallelize-loops=2 -fcompare-debug pr46969.c 
gcc: error: pr46969.c: -fcompare-debug failure (length)

Tested revisions:
r167809 - fail
r165699 - fail
r161659 - OK
4.5 r166509 - OK


^ permalink raw reply	[flat|nested] 7+ messages in thread

* [Bug tree-optimization/46969] -fcompare-debug failure with -O -ftree-vectorize -ftree-parallelize-loops=2
  2010-12-15 19:27 [Bug tree-optimization/46969] New: -O -ftree-vectorize -ftree-parallelize-loops=2 -fcompare-debug zsojka at seznam dot cz
@ 2010-12-15 19:30 ` zsojka at seznam dot cz
  2010-12-15 20:15 ` [Bug tree-optimization/46969] [4.6 Regression] " hjl.tools at gmail dot com
                   ` (4 subsequent siblings)
  5 siblings, 0 replies; 7+ messages in thread
From: zsojka at seznam dot cz @ 2010-12-15 19:30 UTC (permalink / raw)
  To: gcc-bugs

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46969

--- Comment #1 from Zdenek Sojka <zsojka at seznam dot cz> 2010-12-15 19:29:53 UTC ---
Created attachment 22773
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=22773
dumps - pr46969.*gkd


^ permalink raw reply	[flat|nested] 7+ messages in thread

* [Bug tree-optimization/46969] [4.6 Regression] -fcompare-debug failure with -O -ftree-vectorize -ftree-parallelize-loops=2
  2010-12-15 19:27 [Bug tree-optimization/46969] New: -O -ftree-vectorize -ftree-parallelize-loops=2 -fcompare-debug zsojka at seznam dot cz
  2010-12-15 19:30 ` [Bug tree-optimization/46969] -fcompare-debug failure with -O -ftree-vectorize -ftree-parallelize-loops=2 zsojka at seznam dot cz
@ 2010-12-15 20:15 ` hjl.tools at gmail dot com
  2010-12-16 11:35 ` irar at il dot ibm.com
                   ` (3 subsequent siblings)
  5 siblings, 0 replies; 7+ messages in thread
From: hjl.tools at gmail dot com @ 2010-12-15 20:15 UTC (permalink / raw)
  To: gcc-bugs

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46969

H.J. Lu <hjl.tools at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2010.12.15 20:14:58
                 CC|                            |irar at il dot ibm.com
   Target Milestone|---                         |4.6.0
            Summary|-fcompare-debug failure     |[4.6 Regression]
                   |with -O -ftree-vectorize    |-fcompare-debug failure
                   |-ftree-parallelize-loops=2  |with -O -ftree-vectorize
                   |                            |-ftree-parallelize-loops=2
     Ever Confirmed|0                           |1

--- Comment #2 from H.J. Lu <hjl.tools at gmail dot com> 2010-12-15 20:14:58 UTC ---
It is caused by revision 161797:

http://gcc.gnu.org/ml/gcc-cvs/2010-07/msg00151.html


^ permalink raw reply	[flat|nested] 7+ messages in thread

* [Bug tree-optimization/46969] [4.6 Regression] -fcompare-debug failure with -O -ftree-vectorize -ftree-parallelize-loops=2
  2010-12-15 19:27 [Bug tree-optimization/46969] New: -O -ftree-vectorize -ftree-parallelize-loops=2 -fcompare-debug zsojka at seznam dot cz
  2010-12-15 19:30 ` [Bug tree-optimization/46969] -fcompare-debug failure with -O -ftree-vectorize -ftree-parallelize-loops=2 zsojka at seznam dot cz
  2010-12-15 20:15 ` [Bug tree-optimization/46969] [4.6 Regression] " hjl.tools at gmail dot com
@ 2010-12-16 11:35 ` irar at il dot ibm.com
  2010-12-16 13:29 ` jakub at gcc dot gnu.org
                   ` (2 subsequent siblings)
  5 siblings, 0 replies; 7+ messages in thread
From: irar at il dot ibm.com @ 2010-12-16 11:35 UTC (permalink / raw)
  To: gcc-bugs

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46969

--- Comment #3 from Ira Rosen <irar at il dot ibm.com> 2010-12-16 11:35:19 UTC ---
This revision changed the realignment for the loop, we now use loop peeling to
make the accesses aligned. Maybe auto-par doesn't handle the peeled loop
correctly?

Also after a small change in the code I couldn't see the error anymore, even
after reverting the change.


^ permalink raw reply	[flat|nested] 7+ messages in thread

* [Bug tree-optimization/46969] [4.6 Regression] -fcompare-debug failure with -O -ftree-vectorize -ftree-parallelize-loops=2
  2010-12-15 19:27 [Bug tree-optimization/46969] New: -O -ftree-vectorize -ftree-parallelize-loops=2 -fcompare-debug zsojka at seznam dot cz
                   ` (2 preceding siblings ...)
  2010-12-16 11:35 ` irar at il dot ibm.com
@ 2010-12-16 13:29 ` jakub at gcc dot gnu.org
  2010-12-18 21:07 ` jakub at gcc dot gnu.org
  2010-12-18 21:08 ` jakub at gcc dot gnu.org
  5 siblings, 0 replies; 7+ messages in thread
From: jakub at gcc dot gnu.org @ 2010-12-16 13:29 UTC (permalink / raw)
  To: gcc-bugs

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46969

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |ASSIGNED
         AssignedTo|unassigned at gcc dot       |jakub at gcc dot gnu.org
                   |gnu.org                     |

--- Comment #4 from Jakub Jelinek <jakub at gcc dot gnu.org> 2010-12-16 13:18:15 UTC ---
Created attachment 22784
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=22784
gcc46-pr46959.patch

Untested fix.


^ permalink raw reply	[flat|nested] 7+ messages in thread

* [Bug tree-optimization/46969] [4.6 Regression] -fcompare-debug failure with -O -ftree-vectorize -ftree-parallelize-loops=2
  2010-12-15 19:27 [Bug tree-optimization/46969] New: -O -ftree-vectorize -ftree-parallelize-loops=2 -fcompare-debug zsojka at seznam dot cz
                   ` (3 preceding siblings ...)
  2010-12-16 13:29 ` jakub at gcc dot gnu.org
@ 2010-12-18 21:07 ` jakub at gcc dot gnu.org
  2010-12-18 21:08 ` jakub at gcc dot gnu.org
  5 siblings, 0 replies; 7+ messages in thread
From: jakub at gcc dot gnu.org @ 2010-12-18 21:07 UTC (permalink / raw)
  To: gcc-bugs

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46969

--- Comment #5 from Jakub Jelinek <jakub at gcc dot gnu.org> 2010-12-18 21:07:15 UTC ---
Author: jakub
Date: Sat Dec 18 21:07:12 2010
New Revision: 168034

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=168034
Log:
    PR tree-optimization/46969
    * tree-parloops.c (struct reduction_info): Add reduc_version.
    (reduction_info_hash): Return reduc_version field.
    (reduction_phi): Set reduc_version to gimple_uid (phi).
    (build_new_reduction): Set reduc_version to SSA_NAME_VERSION of
    phi result.
    (set_reduc_phi_uids): New function.
    (gather_scalar_reductions): Call it at the end through htab_traverse.

    * gcc.dg/autopar/pr46969.c: New test.

Added:
    trunk/gcc/testsuite/gcc.dg/autopar/pr46969.c
Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/testsuite/ChangeLog
    trunk/gcc/tree-parloops.c


^ permalink raw reply	[flat|nested] 7+ messages in thread

* [Bug tree-optimization/46969] [4.6 Regression] -fcompare-debug failure with -O -ftree-vectorize -ftree-parallelize-loops=2
  2010-12-15 19:27 [Bug tree-optimization/46969] New: -O -ftree-vectorize -ftree-parallelize-loops=2 -fcompare-debug zsojka at seznam dot cz
                   ` (4 preceding siblings ...)
  2010-12-18 21:07 ` jakub at gcc dot gnu.org
@ 2010-12-18 21:08 ` jakub at gcc dot gnu.org
  5 siblings, 0 replies; 7+ messages in thread
From: jakub at gcc dot gnu.org @ 2010-12-18 21:08 UTC (permalink / raw)
  To: gcc-bugs

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46969

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|                            |FIXED

--- Comment #6 from Jakub Jelinek <jakub at gcc dot gnu.org> 2010-12-18 21:08:14 UTC ---
Fixed.


^ permalink raw reply	[flat|nested] 7+ messages in thread

end of thread, other threads:[~2010-12-18 21:08 UTC | newest]

Thread overview: 7+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2010-12-15 19:27 [Bug tree-optimization/46969] New: -O -ftree-vectorize -ftree-parallelize-loops=2 -fcompare-debug zsojka at seznam dot cz
2010-12-15 19:30 ` [Bug tree-optimization/46969] -fcompare-debug failure with -O -ftree-vectorize -ftree-parallelize-loops=2 zsojka at seznam dot cz
2010-12-15 20:15 ` [Bug tree-optimization/46969] [4.6 Regression] " hjl.tools at gmail dot com
2010-12-16 11:35 ` irar at il dot ibm.com
2010-12-16 13:29 ` jakub at gcc dot gnu.org
2010-12-18 21:07 ` jakub at gcc dot gnu.org
2010-12-18 21:08 ` jakub at gcc dot gnu.org

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).