[Bug ipa/102081] New: [12 regression] ICE building compiler starting with r12-3159

[Bug ipa/102081] New: [12 regression] ICE building compiler starting with r12-3159

[seurer at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102081

            Bug ID: 102081
           Summary: [12 regression] ICE building compiler starting with
                    r12-3159
           Product: gcc
           Version: 9.4.1
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: ipa
          Assignee: unassigned at gcc dot gnu.org
          Reporter: seurer at gcc dot gnu.org
                CC: marxin at gcc dot gnu.org
  Target Milestone: ---

g:f075b8c5adcf9cb6336563c472c8d624c54184db, r12-3159 

/home/seurer/gcc/git/build/gcc-test/./gcc/xgcc
-B/home/seurer/gcc/git/build/gcc-test/./gcc/
-B/home/seurer/gcc/git/install/gcc-test/powerpc64le-unknown-linux-gnu/bin/
-B/home/seurer/gcc/git/install/gcc-test/powerpc64le-unknown-linux-gnu/lib/
-isystem
/home/seurer/gcc/git/install/gcc-test/powerpc64le-unknown-linux-gnu/include
-isystem
/home/seurer/gcc/git/install/gcc-test/powerpc64le-unknown-linux-gnu/sys-include
   -g -O2 -O2  -g -O2 -DIN_GCC    -W -Wall -Wno-narrowing -Wwrite-strings
-Wcast-qual -Wstrict-prototypes -Wmissing-prototypes -Wold-style-definition 
-isystem ./include   -fPIC -mlong-double-128 -mno-minimal-toc -g -DIN_LIBGCC2
-fbuilding-libgcc -fno-stack-protector   -fPIC -mlong-double-128
-mno-minimal-toc -I. -I. -I../.././gcc -I/home/seurer/gcc/git/gcc-test/libgcc
-I/home/seurer/gcc/git/gcc-test/libgcc/.
-I/home/seurer/gcc/git/gcc-test/libgcc/../gcc
-I/home/seurer/gcc/git/gcc-test/libgcc/../include
-I/home/seurer/gcc/git/gcc-test/libgcc/../libdecnumber/dpd
-I/home/seurer/gcc/git/gcc-test/libgcc/../libdecnumber -DHAVE_CC_TLS  -o
decQuad.o -MT decQuad.o -MD -MP -MF decQuad.dep -c
/home/seurer/gcc/git/gcc-test/libgcc/../libdecnumber/decQuad.c
during IPA pass: modref
/home/seurer/gcc/git/gcc-test/libgcc/../libdecnumber/decQuad.c: In function
'__decQuadReduce':
/home/seurer/gcc/git/gcc-test/libgcc/../libdecnumber/decQuad.c:141: internal
compiler error: in verify, at ipa-modref-tree.h:335
  141 | 
      | 
0x1088092b modref_ref_node<int>::verify()
        /home/seurer/gcc/git/gcc-test/gcc/ipa-modref-tree.h:335
0x1088092b modref_ref_node<int>::insert_access(modref_access_node, unsigned
long, bool)
        /home/seurer/gcc/git/gcc-test/gcc/ipa-modref-tree.h:355
0x10880c5f modref_tree<int>::insert(int, int, modref_access_node, bool)
        /home/seurer/gcc/git/gcc-test/gcc/ipa-modref-tree.h:616
0x10871ca3 record_access
        /home/seurer/gcc/git/gcc-test/gcc/ipa-modref.c:713
0x10873077 analyze_load
        /home/seurer/gcc/git/gcc-test/gcc/ipa-modref.c:1217
0x10765a1b walk_stmt_load_store_addr_ops(gimple*, void*, bool (*)(gimple*,
tree_node*, tree_node*, void*), bool (*)(gimple*, tree_node*, tree_node*,
void*), bool (*)(gimple*, tree_node*, tree_node*, void*))
        /home/seurer/gcc/git/gcc-test/gcc/gimple-walk.c:800
0x10879437 analyze_stmt
        /home/seurer/gcc/git/gcc-test/gcc/ipa-modref.c:1268
0x10879437 analyze_function
        /home/seurer/gcc/git/gcc-test/gcc/ipa-modref.c:2130
0x1087b3a3 modref_generate
        /home/seurer/gcc/git/gcc-test/gcc/ipa-modref.c:2208
0x10a5ca1b execute_ipa_summary_passes(ipa_opt_pass_d*)
        /home/seurer/gcc/git/gcc-test/gcc/passes.c:2248
0x104e2fdb ipa_passes
        /home/seurer/gcc/git/gcc-test/gcc/cgraphunit.c:2179
0x104e2fdb symbol_table::compile()
        /home/seurer/gcc/git/gcc-test/gcc/cgraphunit.c:2289
0x104e6eff symbol_table::compile()
        /home/seurer/gcc/git/gcc-test/gcc/cgraphunit.c:2269
0x104e6eff symbol_table::finalize_compilation_unit()
        /home/seurer/gcc/git/gcc-test/gcc/cgraphunit.c:2537



commit f075b8c5adcf9cb6336563c472c8d624c54184db (HEAD)
Author: Jan Hubicka <hubicka@ucw.cz>
Date:   Thu Aug 26 15:33:56 2021 +0200

    Fix off-by-one error in try_merge_with

[Bug ipa/102081] [12 regression] ICE building compiler starting with r12-3159

[seurer at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102081

seurer at gcc dot gnu.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Target|                            |powerpc64le-linux-gnu
              Build|                            |powerpc64le-linux-gnu
                 CC|                            |hubicka at gcc dot gnu.org
               Host|                            |powerpc64le-linux-gnu

--- Comment #1 from seurer at gcc dot gnu.org ---
This is during stage 1.

[Bug ipa/102081] [12 regression] ICE building compiler starting with r12-3159

[dcb314 at hotmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102081

David Binderman <dcb314 at hotmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |dcb314 at hotmail dot com

--- Comment #2 from David Binderman <dcb314 at hotmail dot com> ---
Might be related:

during GIMPLE pass: modref
bug749.c: In function ‘CopyMatrix3D’:
bug749.c:4:6: internal compiler error: in verify, at ipa-modref-tree.h:335

For this C source code and compiler flag -O3:

typedef double Matrix3D[][3];
Matrix3D *CopyMatrix3D_dest;
int CopyMatrix3D_x;
void CopyMatrix3D(Matrix3D *src) {
  int y;
  CopyMatrix3D_x = 0;
  for (; CopyMatrix3D_x < 3; ++CopyMatrix3D_x) {
    y = 0;
    for (; y < 3; ++y)
      (*CopyMatrix3D_dest)[CopyMatrix3D_x][y] = (*src)[CopyMatrix3D_x][y];
  }
}

[Bug ipa/102081] [12 regression] ICE building compiler starting with r12-3159

[hjl.tools at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102081

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
     Ever confirmed|0                           |1
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2021-08-26

--- Comment #3 from H.J. Lu <hjl.tools at gmail dot com> ---
On Linux/x86_64,

f075b8c5adcf9cb6336563c472c8d624c54184db is the first bad commit
commit f075b8c5adcf9cb6336563c472c8d624c54184db
Author: Jan Hubicka <hubicka@ucw.cz>
Date:   Thu Aug 26 15:33:56 2021 +0200

    Fix off-by-one error in try_merge_with

caused

FAIL: gcc.c-torture/compile/960514-1.c   -O3 -fomit-frame-pointer
-funroll-loops -fpeel-loops -ftracer -finline-functions  (internal compiler
error)
FAIL: gcc.c-torture/compile/960514-1.c   -O3 -fomit-frame-pointer
-funroll-loops -fpeel-loops -ftracer -finline-functions  (test for excess
errors)
FAIL: gcc.c-torture/compile/960514-1.c   -O3 -g  (internal compiler error)
FAIL: gcc.c-torture/compile/960514-1.c   -O3 -g  (test for excess errors)
FAIL: gcc.dg/vect/O3-pr45971.c (internal compiler error)
FAIL: gcc.dg/vect/O3-pr45971.c (test for excess errors)
FAIL: gfortran.dg/pr68251.f90   -O  (internal compiler error)
FAIL: gfortran.dg/pr68251.f90   -O  (test for excess errors)

with GCC configured with



To reproduce:

$ cd {build_dir}/gcc && make check
RUNTESTFLAGS="compile.exp=gcc.c-torture/compile/960514-1.c
--target_board='unix{-m32}'"
$ cd {build_dir}/gcc && make check
RUNTESTFLAGS="compile.exp=gcc.c-torture/compile/960514-1.c
--target_board='unix{-m32\ -march=cascadelake}'"
$ cd {build_dir}/gcc && make check
RUNTESTFLAGS="compile.exp=gcc.c-torture/compile/960514-1.c
--target_board='unix{-m64\ -march=cascadelake}'"
$ cd {build_dir}/gcc && make check
RUNTESTFLAGS="vect.exp=gcc.dg/vect/O3-pr45971.c --target_board='unix{-m32}'"
$ cd {build_dir}/gcc && make check
RUNTESTFLAGS="vect.exp=gcc.dg/vect/O3-pr45971.c --target_board='unix{-m32\
-march=cascadelake}'"
$ cd {build_dir}/gcc && make check
RUNTESTFLAGS="vect.exp=gcc.dg/vect/O3-pr45971.c --target_board='unix{-m64\
-march=cascadelake}'"
$ cd {build_dir}/gcc && make check RUNTESTFLAGS="dg.exp=gfortran.dg/pr68251.f90
--target_board='unix{-m32\ -march=cascadelake}'"
$ cd {build_dir}/gcc && make check RUNTESTFLAGS="dg.exp=gfortran.dg/pr68251.f90
--target_board='unix{-m64\ -march=cascadelake}'"

[Bug ipa/102081] [12 regression] ICE building compiler starting with r12-3159

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102081

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Version|9.4.1                       |12.0
   Target Milestone|---                         |12.0

--- Comment #4 from Richard Biener <rguenth at gcc dot gnu.org> ---
part of this might be already fixed

[Bug ipa/102081] [12 regression] ICE building compiler starting with r12-3159

[hubicka at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102081

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |WAITING

--- Comment #5 from Jan Hubicka <hubicka at gcc dot gnu.org> ---
I believe that this should be fixed.  Is it still reproducing for you? (the
testcase in comment #2 compiles for me)

[Bug ipa/102081] [12 regression] ICE building compiler starting with r12-3159

[dcb314 at hotmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102081

--- Comment #6 from David Binderman <dcb314 at hotmail dot com> ---
(In reply to Jan Hubicka from comment #5)
> I believe that this should be fixed.  Is it still reproducing for you? (the
> testcase in comment #2 compiles for me)

Code in comment 2 compiles fine for me now.

[Bug ipa/102081] [12 regression] ICE building compiler starting with r12-3159

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102081

Martin Liška <marxin at gcc dot gnu.org> changed:

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

--- Comment #7 from Martin Liška <marxin at gcc dot gnu.org> ---
Then let's close it as fixed.