https://gcc.gnu.org/bugzilla/show_bug.cgi?id=117527

            Bug ID: 117527
           Summary: [15 regression] 511.povary_r with
                    march_native_ofast_lto build fail since r15-5021
           Product: gcc
           Version: 15.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: tree-optimization
          Assignee: unassigned at gcc dot gnu.org
          Reporter: haochen.jiang at intel dot com
  Target Milestone: ---

I have got a build fail on 511.povary_y with march_native_ofast_lto.

Known to fail: x86 since Cascade Lake and Zen3, aarch64 for Amphere (and
probably Graviton3).

The case will also happen w/o -flto.

After eliminating -flto, it seems that when building lighting.cpp, I got:

$
/export/users/haochenj/CI_Framework/task/20241111102138/install/applications.compilers.source.gnu-toolchain.gcc/bin/g++
-m64 -c -o lighting.o -DSPEC -DNDEBUG -Ifrontend -Ibase -I. -Ispec_qsort
-DSPEC_AUTO_SUPPRESS_OPENMP  -Ofast -march=native -mfpmath=sse              
-fno-finite-math-only     -DSPEC_LP64  lighting.cpp
lighting.cpp: In function ‘void pov::do_light_area_light(LIGHT_SOURCE*,
double*, RAY*, RAY*, double*)’:
lighting.cpp:1359:18: warning: ISO C++17 does not allow ‘register’ storage
class specifier [-Wregister]
 1359 |     register int i;
      |                  ^
lighting.cpp: In function ‘void pov::block_area_light(LIGHT_SOURCE*, double*,
RAY*, RAY*, double*, float*, int, int, int, int, int)’:
lighting.cpp:1596:22: warning: ISO C++17 does not allow ‘register’ storage
class specifier [-Wregister]
 1596 |         register int i;
      |                      ^
lighting.cpp:1723:26: warning: ISO C++17 does not allow ‘register’ storage
class specifier [-Wregister]
 1723 |             register int i;
      |                          ^
g++: internal compiler error: Segmentation fault signal terminated program
cc1plus
Please submit a full bug report, with preprocessed source (by using
-freport-bug).
See <https://gcc.gnu.org/bugs/> for instructions.

The strangest thing is that it doesn't show backtrace for the ICE.

I will try to reduce that first.
  • [Bug tree-optimization/117527]... haochen.jiang at intel dot com via Gcc-bugs

Reply via email to