public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
* [Bug lto/47943] New: PRE fails to move a load before a loop with LTO
@ 2011-03-01 13:10 jamborm at gcc dot gnu.org
  2011-03-01 13:11 ` [Bug lto/47943] " jamborm at gcc dot gnu.org
                   ` (2 more replies)
  0 siblings, 3 replies; 4+ messages in thread
From: jamborm at gcc dot gnu.org @ 2011-03-01 13:10 UTC (permalink / raw)
  To: gcc-bugs

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=47943

           Summary: PRE fails to move a load before a loop with LTO
           Product: gcc
           Version: 4.6.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: lto
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: jamborm@gcc.gnu.org


When compiling the provided testcase with LTO, PRE fails to hoist the
load of reg->node from the loop.  Though this slows down the testcase
only marginally, it has a substantial performance impact for
462.libquantum from SPEC 2006.

When compiling the test case without LTO, PRE does move the load
before the loop.

Observed at least with trunk revision 170357 (2011-02-21), compiling
at -Ofast with and without -flto.


^ permalink raw reply	[flat|nested] 4+ messages in thread

* [Bug lto/47943] PRE fails to move a load before a loop with LTO
  2011-03-01 13:10 [Bug lto/47943] New: PRE fails to move a load before a loop with LTO jamborm at gcc dot gnu.org
@ 2011-03-01 13:11 ` jamborm at gcc dot gnu.org
  2011-03-01 13:34 ` rguenth at gcc dot gnu.org
  2011-03-02  0:45 ` jamborm at gcc dot gnu.org
  2 siblings, 0 replies; 4+ messages in thread
From: jamborm at gcc dot gnu.org @ 2011-03-01 13:11 UTC (permalink / raw)
  To: gcc-bugs

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=47943

--- Comment #1 from Martin Jambor <jamborm at gcc dot gnu.org> 2011-03-01 13:11:23 UTC ---
Created attachment 23499
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=23499
Testcase

The testcase.


^ permalink raw reply	[flat|nested] 4+ messages in thread

* [Bug lto/47943] PRE fails to move a load before a loop with LTO
  2011-03-01 13:10 [Bug lto/47943] New: PRE fails to move a load before a loop with LTO jamborm at gcc dot gnu.org
  2011-03-01 13:11 ` [Bug lto/47943] " jamborm at gcc dot gnu.org
@ 2011-03-01 13:34 ` rguenth at gcc dot gnu.org
  2011-03-02  0:45 ` jamborm at gcc dot gnu.org
  2 siblings, 0 replies; 4+ messages in thread
From: rguenth at gcc dot gnu.org @ 2011-03-01 13:34 UTC (permalink / raw)
  To: gcc-bugs

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=47943

Richard Guenther <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |WAITING
   Last reconfirmed|                            |2011.03.01 13:34:01
     Ever Confirmed|0                           |1

--- Comment #2 from Richard Guenther <rguenth at gcc dot gnu.org> 2011-03-01 13:34:01 UTC ---
I think this is a dup of PR47924 as I now get

<bb 7>:
  pretmp.8_39 = reg_3(D)->node;
  pretmp.10_19 = 1 << control2_20(D);
  pretmp.10_42 = 1 << control1_10(D);
  pretmp.10_43 = pretmp.10_19 | pretmp.10_42;

<bb 3>:
  # i_17 = PHI <i_37(10), 0(7)>


^ permalink raw reply	[flat|nested] 4+ messages in thread

* [Bug lto/47943] PRE fails to move a load before a loop with LTO
  2011-03-01 13:10 [Bug lto/47943] New: PRE fails to move a load before a loop with LTO jamborm at gcc dot gnu.org
  2011-03-01 13:11 ` [Bug lto/47943] " jamborm at gcc dot gnu.org
  2011-03-01 13:34 ` rguenth at gcc dot gnu.org
@ 2011-03-02  0:45 ` jamborm at gcc dot gnu.org
  2 siblings, 0 replies; 4+ messages in thread
From: jamborm at gcc dot gnu.org @ 2011-03-02  0:45 UTC (permalink / raw)
  To: gcc-bugs

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=47943

Martin Jambor <jamborm at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|WAITING                     |RESOLVED
         Resolution|                            |DUPLICATE

--- Comment #3 from Martin Jambor <jamborm at gcc dot gnu.org> 2011-03-02 00:45:15 UTC ---
Indeed. And libquantum is now as fast as the non-LTO one (again).  Thanks.

*** This bug has been marked as a duplicate of bug 47924 ***


^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2011-03-02  0:45 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2011-03-01 13:10 [Bug lto/47943] New: PRE fails to move a load before a loop with LTO jamborm at gcc dot gnu.org
2011-03-01 13:11 ` [Bug lto/47943] " jamborm at gcc dot gnu.org
2011-03-01 13:34 ` rguenth at gcc dot gnu.org
2011-03-02  0:45 ` jamborm 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).