https://gcc.gnu.org/bugzilla/show_bug.cgi?id=87552

--- Comment #10 from Frank Scheiner <frank.scheiner at web dot de> ---
Looking into this, the first test results that show this issue seem to be from
2018-10-10 ([1], 9.0.0 20181010 (experimental) [trunk revision 265003]):

```
FAIL: gcc.c-torture/compile/20010102-1.c   -O3 -fomit-frame-pointer
-funroll-loops -fpeel-loops -ftracer -finline-functions  (internal compiler
error)
FAIL: gcc.c-torture/compile/20010102-1.c   -O3 -fomit-frame-pointer
-funroll-loops -fpeel-loops -ftracer -finline-functions  (test for excess
errors)
[...]
configure flags: --prefix=/usr --build=ia64-suse-linux --enable-shared
--with-system-zlib CFLAGS='-O2 -g' CXXFLAGS='-O2 -g' --enable-checking=release
```

...and these FAILs are already gone again on 2018-10-12 ([2], 9.0.0 20181012
(experimental) [trunk revision 265073]) which used the exact same configure
flags.

[1]: https://gcc.gnu.org/pipermail/gcc-testresults/2018-October/503423.html

[2]: https://gcc.gnu.org/pipermail/gcc-testresults/2018-October/503935.html

So looking into gcc-old one or some of the 38 changes between [3] and [4] must
have fixed or masked that problem somehow.

[3]:
https://gcc.gnu.org/git/?p=gcc-old.git;a=commit;h=f92b7cd1ecc0033282b7b41b36657aa4318d9671

[4]:
https://gcc.gnu.org/git/?p=gcc-old.git;a=commit;h=f8c147a5f5ad459fba646c377783bace5d7547d6

The most recent test results for ia64 ([5]) seem to confirm that this problem
is gone:

```
gcc/testsuite/gcc/gcc.log:17805:PASS: gcc.c-torture/compile/20010102-1.c   -O3
-fomit-frame-pointer -funroll-loops -fpeel-loops -ftracer -finline-functions 
(test for excess errors)
gcc/testsuite/gcc/gcc.log:17808:PASS: gcc.c-torture/compile/20010102-1.c   -O3
-g  (test for excess errors)
```

[5]: https://gcc.gnu.org/pipermail/gcc-testresults/2026-March/870522.html

Also two quick compile tests in Ski seeem to confirm that (though with GCC
15.2.0 in EPIC Slack):

```
root@darkstar:~/pr87552# lscpu
Architecture:           ia64
  CPU op-mode(s):       64-bit
  Byte Order:           Little Endian
CPU(s):                 1
  On-line CPU(s) list:  0
Vendor ID:              HP Ski Simulator
  Model name:           MonteSkito
    CPU family:         31
    Model:              0
    Thread(s) per core: 1
    Core(s) per socket: 1
    Socket(s):          1
    BogoMIPS:           165.15
    Flags:              branchlong, 16-byte atomic ops

root@darkstar:~/pr87552# gcc --version
gcc (GCC) 15.2.0
Copyright (C) 2025 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

root@darkstar:~/pr87552# time gcc -c -O3 -fomit-frame-pointer -funroll-loops
-fpeel-loops -ftracer -finline-functions ./20010102-1.c
./20010102-1.c: In function ‘_obstack_newchunk’:
./20010102-1.c:61:1: warning: old-style function definition
[-Wold-style-definition]
   61 | _obstack_newchunk (h, length)
      | ^~~~~~~~~~~~~~~~~

real    0m8.565s
user    0m8.250s
sys     0m0.312s
root@darkstar:~/pr87552# echo $?
0

root@darkstar:~/pr87552# time gcc -c -O3 -fomit-frame-pointer -funroll-loops
-fpeel-loops -ftracer -finline-functions -fselective-scheduling
-fselective-scheduling2 -fsel-sched-pipelining
-fsel-sched-pipelining-outer-loops ./20010102-1.c
./20010102-1.c: In function ‘_obstack_newchunk’:
./20010102-1.c:61:1: warning: old-style function definition
[-Wold-style-definition]
   61 | _obstack_newchunk (h, length)
      | ^~~~~~~~~~~~~~~~~

real    0m9.727s
user    0m9.593s
sys     0m0.125s
root@darkstar:~/pr87552# echo $?
0
```

Reply via email to