[Bug regression/55168] New: [4.8 Regression]: gcc.c-torture/compile/pr44119.c ICE, all but -O0

[Bug regression/55168] New: [4.8 Regression]: gcc.c-torture/compile/pr44119.c ICE, all but -O0

[hp at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55168

             Bug #: 55168
           Summary: [4.8 Regression]: gcc.c-torture/compile/pr44119.c ICE,
                    all but -O0
    Classification: Unclassified
           Product: gcc
           Version: 4.7.0
            Status: UNCONFIRMED
          Keywords: ice-on-valid-code
          Severity: normal
          Priority: P3
         Component: regression
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: hp@gcc.gnu.org
                CC: steven@gcc.gnu.org
              Host: x86_64-unknown-linux-gnu
            Target: cris-axis-elf


This test previously passed, now it fails.
A patch in the revision range (last_known_working:first_known_failing)
193042:193049
exposed or caused this regression.  Since then it fails as follows:

Running
/tmp/hpautotest-gcc0/gcc/gcc/testsuite/gcc.c-torture/compile/compile.exp ...
FAIL: gcc.c-torture/compile/pr44119.c  -O1  (internal compiler error)
FAIL: gcc.c-torture/compile/pr44119.c  -O1  (test for excess errors)
FAIL: gcc.c-torture/compile/pr44119.c  -O2  (internal compiler error)
FAIL: gcc.c-torture/compile/pr44119.c  -O2  (test for excess errors)
FAIL: gcc.c-torture/compile/pr44119.c  -O3 -fomit-frame-pointer  (internal
compiler error)
FAIL: gcc.c-torture/compile/pr44119.c  -O3 -fomit-frame-pointer  (test for
excess errors)
FAIL: gcc.c-torture/compile/pr44119.c  -O3 -fomit-frame-pointer -funroll-loops 
(internal compiler error)
FAIL: gcc.c-torture/compile/pr44119.c  -O3 -fomit-frame-pointer -funroll-loops 
(test for excess errors)
FAIL: gcc.c-torture/compile/pr44119.c  -O3 -fomit-frame-pointer
-funroll-all-loops -finline-functions  (internal compiler error)
FAIL: gcc.c-torture/compile/pr44119.c  -O3 -fomit-frame-pointer
-funroll-all-loops -finline-functions  (test for excess errors)
FAIL: gcc.c-torture/compile/pr44119.c  -O3 -g  (internal compiler error)
FAIL: gcc.c-torture/compile/pr44119.c  -O3 -g  (test for excess errors)
FAIL: gcc.c-torture/compile/pr44119.c  -Os  (internal compiler error)
FAIL: gcc.c-torture/compile/pr44119.c  -Os  (test for excess errors)
FAIL: gcc.c-torture/compile/pr44119.c  -O2 -flto -fno-use-linker-plugin
-flto-partition=none  (internal compiler error)
FAIL: gcc.c-torture/compile/pr44119.c  -O2 -flto -fno-use-linker-plugin
-flto-partition=none  (test for excess errors)
FAIL: gcc.c-torture/compile/pr44119.c  -O2 -flto -fuse-linker-plugin
-fno-fat-lto-objects  (internal compiler error)
FAIL: gcc.c-torture/compile/pr44119.c  -O2 -flto -fuse-linker-plugin
-fno-fat-lto-objects  (test for excess errors)

The messages in gcc.log are similar:

Executing on host: /tmp/hpautotest-gcc0/cris-elf/gccobj/gcc/xgcc
-B/tmp/hpautotest-gcc0/cris-elf/gccobj/gcc/  -fno-diagnostics-show-caret   -O1 
-w -c   -isystem
/tmp/hpautotest-gcc0/cris-elf/gccobj/cris-elf/./newlib/targ-include -isystem
/tmp/hpautotest-gcc0/gcc/newlib/libc/include  -o pr44119.o
/tmp/hpautotest-gcc0/gcc/gcc/testsuite/gcc.c-torture/compile/pr44119.c   
(timeout = 300)
/tmp/hpautotest-gcc0/gcc/gcc/testsuite/gcc.c-torture/compile/pr44119.c: In
function 'func_96':

/tmp/hpautotest-gcc0/gcc/gcc/testsuite/gcc.c-torture/compile/pr44119.c:45:1:
internal compiler error: Floating point exception

0x84ab35 crash_signal

    /tmp/hpautotest-gcc0/gcc/gcc/toplev.c:333

0x7d4bbf predict_loops

    /tmp/hpautotest-gcc0/gcc/gcc/predict.c:1524

0x7d67b4 tree_estimate_probability()

    /tmp/hpautotest-gcc0/gcc/gcc/predict.c:2259

0x7d75d4 tree_estimate_probability_driver

    /tmp/hpautotest-gcc0/gcc/gcc/predict.c:2296


Author of only suspect patch in revision range CC:ed.

[Bug regression/55168] [4.8 Regression]: gcc.c-torture/compile/pr44119.c ICE, all but -O0

[hp at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55168

Hans-Peter Nilsson <hp at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jakub at gcc dot gnu.org

--- Comment #1 from Hans-Peter Nilsson <hp at gcc dot gnu.org> 2012-11-01 20:26:05 UTC ---
Found the patch, added co-author to CC.

[Bug regression/55168] [4.8 Regression]: gcc.c-torture/compile/pr44119.c ICE, all but -O0

[hubicka at ucw dot cz]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55168

--- Comment #2 from Jan Hubicka <hubicka at ucw dot cz> 2012-11-01 20:49:18 UTC ---
This actually looks like a previously latent issue in predict.c For all but
estimate_num_iterations_int.  It uses the funny definition of number of
iterations (i.e. 0 means that loop will exit at first invocation of the exit
condition and that is what it will be predicted with when nitercst == 1.

Index: predict.c
===================================================================
--- predict.c   (revision 193062)
+++ predict.c   (working copy)
@@ -1433,8 +1437,8 @@ predict_loops (void)
             the loop, use it to predict this exit.  */
          else if (n_exits == 1)
            {
-             nitercst = estimated_stmt_executions_int (loop);
-             if (nitercst < 0)
+             nitercst = estimated_stmt_executions_int (loop) + 1;
+             if (nitercst <= 0)
                continue;
              if (nitercst > max)
                nitercst = max;

Re: [Bug regression/55168] New: [4.8 Regression]: gcc.c-torture/compile/pr44119.c ICE, all but -O0

[Jan Hubicka]

This actually looks like a previously latent issue in predict.c For all but
estimate_num_iterations_int.  It uses the funny definition of number of
iterations (i.e. 0 means that loop will exit at first invocation of the exit
condition and that is what it will be predicted with when nitercst == 1.

Index: predict.c
===================================================================
--- predict.c   (revision 193062)
+++ predict.c   (working copy)
@@ -1433,8 +1437,8 @@ predict_loops (void)
             the loop, use it to predict this exit.  */
          else if (n_exits == 1)
            {
-             nitercst = estimated_stmt_executions_int (loop);
-             if (nitercst < 0)
+             nitercst = estimated_stmt_executions_int (loop) + 1;
+             if (nitercst <= 0)
                continue;
              if (nitercst > max)
                nitercst = max;

[Bug regression/55168] [4.8 Regression]: gcc.c-torture/compile/pr44119.c ICE, all but -O0

[hp at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55168

Hans-Peter Nilsson <hp at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |hubicka at ucw dot cz

--- Comment #3 from Hans-Peter Nilsson <hp at gcc dot gnu.org> 2012-11-04 20:26:26 UTC ---
(In reply to comment #2)
> This actually looks like a previously latent issue in predict.c For all but
> estimate_num_iterations_int.  It uses the funny definition of number of
> iterations ...

Honza, the regression went away in (193100:193109].  I see you made a series of
predict.c improvements there.  Is fixing this bug reasonably an expected effect
of those patches, so this PR can be closed?  As opposed to "I thought of that
but that bug is expected to still be there", I mean.

[Bug regression/55168] [4.8 Regression]: gcc.c-torture/compile/pr44119.c ICE, all but -O0

[hubicka at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55168

Jan Hubicka <hubicka at gcc dot gnu.org> changed:

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

--- Comment #4 from Jan Hubicka <hubicka at gcc dot gnu.org> 2012-11-06 14:57:22 UTC ---
Should be fixed now.