[Bug rtl-optimization/54739] New: [4.8 regression] FAIL: gcc.dg/lower-subreg-1.c scan-rtl-dump subreg1 "Splitting reg"

[Bug rtl-optimization/54739] New: [4.8 regression] FAIL: gcc.dg/lower-subreg-1.c scan-rtl-dump subreg1 "Splitting reg"

[schwab@linux-m68k.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54739

             Bug #: 54739
           Summary: [4.8 regression] FAIL: gcc.dg/lower-subreg-1.c
                    scan-rtl-dump subreg1 "Splitting reg"
    Classification: Unclassified
           Product: gcc
           Version: 4.8.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: rtl-optimization
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: schwab@linux-m68k.org
                CC: uweigand@gcc.gnu.org
            Target: m68k-*-*


gcc.dg/lower-subreg-1.c fails on m68k since r191805.

[Bug rtl-optimization/54739] [4.8 regression] FAIL: gcc.dg/lower-subreg-1.c scan-rtl-dump subreg1 "Splitting reg"

[danglin at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54739

John David Anglin <danglin at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2012-09-29
                 CC|                            |danglin at gcc dot gnu.org
     Ever Confirmed|0                           |1

--- Comment #1 from John David Anglin <danglin at gcc dot gnu.org> 2012-09-29 22:26:54 UTC ---
Also seen on hppa.

[Bug rtl-optimization/54739] [4.8 regression] FAIL: gcc.dg/lower-subreg-1.c scan-rtl-dump subreg1 "Splitting reg"

[ebotcazou at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54739

Eric Botcazou <ebotcazou at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Target|m68k-*-*                    |
                 CC|                            |ebotcazou at gcc dot
                   |                            |gnu.org

--- Comment #2 from Eric Botcazou <ebotcazou at gcc dot gnu.org> 2012-09-30 09:52:23 UTC ---
And on the SPARC.

[Bug rtl-optimization/54739] [4.8 regression] FAIL: gcc.dg/lower-subreg-1.c scan-rtl-dump subreg1 "Splitting reg"

[uweigand at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54739

--- Comment #3 from Ulrich Weigand <uweigand at gcc dot gnu.org> 2012-10-01 12:16:53 UTC ---
It seems all three of those targets have an "iordi3" pattern that triggers even
for 32-bit compiles.  In this case, the lower-subreg pass now no longer splits
the register, so that the DImode pattern is actually used.  (Prior to my patch,
the register would have been split anyway.)

The test case is intended to run on 32-bit targets where an ior:DI operation is
supposed to be split; it will now fail on targets with an iordi3 pattern.

For those targets, I guess it's up the target maintainers to decide whether:

- you want the iordi3 pattern to trigger since it gives better code than having
lower-subreg split the operation: in this case, just disable the test case for
your target (this is what I did for ARM)

or

- you'd really prefer to have lower-subreg split the operation, in which case
you should remove the iordi3 pattern

[Bug rtl-optimization/54739] [4.8 regression] FAIL: gcc.dg/lower-subreg-1.c scan-rtl-dump subreg1 "Splitting reg"

[dave.anglin at bell dot net]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54739

--- Comment #4 from dave.anglin at bell dot net 2012-10-01 23:51:23 UTC ---
Thanks Ulrich for your analysis.  On parisc, I believe it would be  
best to remove
iordi3, anddi3, etc, for 32-bit targets if they are successfully  
split.  Testing a fix.

--
John David Anglin    dave.anglin@bell.net

[Bug rtl-optimization/54739] [4.8 regression] FAIL: gcc.dg/lower-subreg-1.c scan-rtl-dump subreg1 "Splitting reg"

[ebotcazou at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54739

--- Comment #5 from Eric Botcazou <ebotcazou at gcc dot gnu.org> 2012-10-04 07:58:31 UTC ---
Thanks.  The code generated for the testcase is certainly better on the SPARC
if lower-subreg doesn't split the DImode objects (which probably means that the
patch was also beneficial for this architecture) so I'll go the ARM way.

[Bug rtl-optimization/54739] [4.8 regression] FAIL: gcc.dg/lower-subreg-1.c scan-rtl-dump subreg1 "Splitting reg"

[ebotcazou at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54739

--- Comment #6 from Eric Botcazou <ebotcazou at gcc dot gnu.org> 2012-10-04 08:15:37 UTC ---
Author: ebotcazou
Date: Thu Oct  4 08:15:11 2012
New Revision: 192066

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=192066
Log:
    PR rtl-optimization/54739
    * gcc.dg/lower-subreg-1.c: Also skip on SPARC.

Modified:
    trunk/gcc/testsuite/ChangeLog
    trunk/gcc/testsuite/gcc.dg/lower-subreg-1.c

[Bug rtl-optimization/54739] [4.8 regression] FAIL: gcc.dg/lower-subreg-1.c scan-rtl-dump subreg1 "Splitting reg"

[danglin at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54739

--- Comment #7 from John David Anglin <danglin at gcc dot gnu.org> 2012-10-06 00:33:56 UTC ---
Author: danglin
Date: Sat Oct  6 00:33:52 2012
New Revision: 192152

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=192152
Log:
    PR rtl-optimization/54739
    * config/pa/pa.md: Remove DImode and, not and, ior and xor patterns
    for 32-bit targets.  Adjust expanders.


Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/config/pa/pa.md

[Bug rtl-optimization/54739] [4.8 regression] FAIL: gcc.dg/lower-subreg-1.c scan-rtl-dump subreg1 "Splitting reg"

[danglin at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54739

--- Comment #8 from John David Anglin <danglin at gcc dot gnu.org> 2012-10-06 00:46:25 UTC ---
Applied patch fixes testcase on parisc.

[Bug rtl-optimization/54739] [4.8 regression] FAIL: gcc.dg/lower-subreg-1.c scan-rtl-dump subreg1 "Splitting reg"

[schwab at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54739

--- Comment #9 from Andreas Schwab <schwab at gcc dot gnu.org> 2012-10-06 11:42:18 UTC ---
Author: schwab
Date: Sat Oct  6 11:42:13 2012
New Revision: 192156

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=192156
Log:
PR rtl-optimization/54739
* config/m68k/m68k.md (anddi3, iordi3, xordi3, one_cmpldi2):
Remove.

Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/config/m68k/m68k.md

[Bug rtl-optimization/54739] [4.8 regression] FAIL: gcc.dg/lower-subreg-1.c scan-rtl-dump subreg1 "Splitting reg"

[schwab@linux-m68k.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54739

Andreas Schwab <schwab@linux-m68k.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |eager at eagercon dot com,
                   |                            |nickc at redhat dot com

--- Comment #10 from Andreas Schwab <schwab@linux-m68k.org> 2012-10-06 11:51:44 UTC ---
mcore and microblaze are the only targets not yet covered.

[Bug rtl-optimization/54739] [4.8 regression] FAIL: gcc.dg/lower-subreg-1.c scan-rtl-dump subreg1 "Splitting reg"

[nickc at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54739

--- Comment #11 from Nick Clifton <nickc at gcc dot gnu.org> 2012-10-09 09:22:00 UTC ---
Author: nickc
Date: Tue Oct  9 09:21:47 2012
New Revision: 192244

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=192244
Log:
    PR rtl-optimization/54739
    * config/mcore/mcore.md: (anddi3, iordi3, xordi3): Delete
    patterns.

Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/config/mcore/mcore.md

[Bug rtl-optimization/54739] [4.8 regression] FAIL: gcc.dg/lower-subreg-1.c scan-rtl-dump subreg1 "Splitting reg"

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54739

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |4.8.0

[Bug rtl-optimization/54739] [4.8 regression] FAIL: gcc.dg/lower-subreg-1.c scan-rtl-dump subreg1 "Splitting reg"

[eager at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54739

--- Comment #12 from eager at gcc dot gnu.org 2012-12-05 17:27:10 UTC ---
Author: eager
Date: Wed Dec  5 17:27:05 2012
New Revision: 194226

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=194226
Log:
PR rtl-optimization/54739
    * config/microblaze/microblaze.md: (anddi3, iordi3, xordi3): Delete
patterns.

Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/config/microblaze/microblaze.md

[Bug rtl-optimization/54739] [4.8 regression] FAIL: gcc.dg/lower-subreg-1.c scan-rtl-dump subreg1 "Splitting reg"

[steven at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54739

Steven Bosscher <steven at gcc dot gnu.org> changed:

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

--- Comment #13 from Steven Bosscher <steven at gcc dot gnu.org> 2012-12-05 18:48:19 UTC ---
per comments #10 - #12