[Bug target/55433] New: [4.8 Regression] ICE on excessive reloads

[Bug target/55433] New: [4.8 Regression] ICE on excessive reloads

[PHHargrove at lbl dot gov]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55433

             Bug #: 55433
           Summary: [4.8 Regression] ICE on excessive reloads
    Classification: Unclassified
           Product: gcc
           Version: 4.8.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: target
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: PHHargrove@lbl.gov
                CC: gary@intrepid.com, nenad@intrepid.com


The 9-line stand-alone test case at the end of this report fails as show below
on MacOS (at least versions 10.6 and 10.8) on x86-64 machines when targeting
the ILP32 ABI.  It notably does NOT fail on Linux/x86-64.

This was seen in SVN r193679 (on MacOSX 10.6) and r193709 (on MacOSX 10.8).

--cut--
$ gcc -m32 -O1 -fpreprocessed -c testcase.c
testcase.c: In function 'test':
testcase.c:10:1: internal compiler error: Max. number of generated reload insns
per insn is achieved (90)

 }
 ^

testcase.c:9:1: internal compiler error: Abort trap
gcc: internal compiler error: Abort trap (program cc1)
Abort trap
--cut--

Note that while the error message is the same as bug 55247, I have confirmed
that the test case attached to that bug report has NOT regressed.  Therefore, I
believe this is a different bug (though possibly related).

Some additional data:
+  -m32 is required (no bug seen w/ -m64)
+  Linux host is fine (w/ -m32 and -m64)
+  -O1 is sufficient, but -O2 and -O3 fail in the same manner
+  removing "unsigned" from the typedef eliminates the problem
+  adding -mno-sse to the command line eliminates the problem


The testcase:

typedef unsigned long long tick_t; // NOTE: no bug if "unsigned" is removed
extern int foo(void);
extern tick_t tick(void);
double test(void) {
  struct { tick_t ticks; } st;
  st.ticks = tick();
  foo(); // NOTE: no bug if this call is removed
  return (double)st.ticks;
}

[Bug target/55433] [4.8 Regression] ICE on excessive reloads

[PHHargrove at lbl dot gov]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55433

--- Comment #1 from Paul H. Hargrove <PHHargrove at lbl dot gov> 2012-11-21 22:48:05 UTC ---
Minor correction:

The line or error output in the Description that reads
   testcase.c:10:1: internal compiler error: [...]
should instead read
   testcase.c:9:1: internal compiler error: [...]

[line number off-by-one due to "pilot error" when entering the report]

[Bug target/55433] [4.8 Regression] ICE on excessive reloads

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55433

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

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

[Bug target/55433] [4.8 Regression][LRA] ICE on excessive reloads

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55433

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Target|                            |i?86-darwin
                 CC|                            |vmakarov at gcc dot gnu.org
               Host|                            |x86_64-darwin
            Summary|[4.8 Regression] ICE on     |[4.8 Regression][LRA] ICE
                   |excessive reloads           |on excessive reloads

--- Comment #2 from Richard Biener <rguenth at gcc dot gnu.org> 2012-12-07 10:19:34 UTC ---
Cannot reproduce on x86_64-linux even with -msseregparm -mfpmath=sse.

[Bug target/55433] [4.8 Regression][LRA] ICE on excessive reloads

[PHHargrove at lbl dot gov]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55433

--- Comment #3 from Paul H. Hargrove <PHHargrove at lbl dot gov> 2013-01-15 09:15:44 UTC ---
This bug is still present and "biting me" as of r195176.
Is there anything I can do to help make some progress on this?

-Paul

[Bug target/55433] [4.8 Regression][LRA] ICE on excessive reloads

[vmakarov at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55433

--- Comment #4 from Vladimir Makarov <vmakarov at gcc dot gnu.org> 2013-01-16 18:44:14 UTC ---
(In reply to comment #3)
> This bug is still present and "biting me" as of r195176.
> Is there anything I can do to help make some progress on this?
>
I managed to reproduce it on Linux and started to work on it.  I hope it will
be fixed at the end of this week.

[Bug target/55433] [4.8 Regression][LRA] ICE on excessive reloads

[vmakarov at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55433

--- Comment #5 from Vladimir Makarov <vmakarov at gcc dot gnu.org> 2013-01-18 18:15:02 UTC ---
Author: vmakarov
Date: Fri Jan 18 18:14:52 2013
New Revision: 195302

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=195302
Log:
2013-01-18  Vladimir Makarov  <vmakarov@redhat.com>

    PR target/55433
    * lra-constraints.c (curr_insn_transform): Reuse original insn for
    secondary memory move.
    (inherit_reload_reg): Use rclass instead of cl for
    check_secondary_memory_needed_p.

2013-01-18  Vladimir Makarov  <vmakarov@redhat.com>

    PR target/55433
    * gcc.target/i386/pr55433.c: New.


Added:
    trunk/gcc/testsuite/gcc.target/i386/pr55433.c
Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/lra-constraints.c
    trunk/gcc/testsuite/ChangeLog

[Bug target/55433] [4.8 Regression][LRA] ICE on excessive reloads

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55433

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
                 CC|                            |jakub at gcc dot gnu.org
         Resolution|                            |FIXED

--- Comment #6 from Jakub Jelinek <jakub at gcc dot gnu.org> 2013-01-18 20:30:39 UTC ---
Fixed.

[Bug target/55433] [4.8 Regression][LRA] ICE on excessive reloads

[vmakarov at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55433

--- Comment #7 from Vladimir Makarov <vmakarov at gcc dot gnu.org> 2013-01-20 19:10:03 UTC ---
Author: vmakarov
Date: Sun Jan 20 19:09:58 2013
New Revision: 195322

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=195322
Log:
2013-01-20  Vladimir Makarov  <vmakarov@redhat.com>

    PR target/55433
    * lra-constraints.c (curr_insn_transform): Don't reuse original
    insn for secondary memory move when memory mode should be
    different.


Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/lra-constraints.c