[Bug c++/11746] New: __attribute__((__noreturn__)) causing compiler SEGV

["eljay at adobe dot com" <gcc-bugzilla at gcc dot gnu dot org>]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11746

           Summary: __attribute__((__noreturn__)) causing compiler SEGV
           Product: gcc
           Version: 3.2
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: c++
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: eljay at adobe dot com
                CC: gcc-bugs at gcc dot gnu dot org
 GCC build triplet: i686-pc-linux-gnu
  GCC host triplet: i686-pc-linux-gnu
GCC target triplet: i686-pc-linux-gnu

--- foo.cpp ---
extern int okay();
extern void good(int i = (okay(),0));
extern void boom() __attribute__((__noreturn__));
extern void crash(int i = (1?(void)0:boom(),0)); // <-- crash
extern void crash_this_too(int i = (boom(),0)); // <-- alternative crash
--- end-of-file ---

I would not expect a crash to occur at that line.

I realize that the above looks somewhat unusual, but it serves as a mechanism to
do Eiffel-like constraints.  The above minimal code is distilled from the
real-world example, and doesn't demonstrate the utility value of the original
situation.  The original situation spans many lines of code, and involves
__assert_fail from cassert / assert.h in particular.

No crash with GCC 2.95 or RH-GCC 2.96.

> gcc -v
Reading specs from
/user/unicore/i80386linux/compiler/gcc3.2/linux2.4/lib/gcc-lib/i686-pc-linux-gnu/3.2/specs
Configured with: ../configure
--prefix=/user/unicore/i80386linux/compiler/gcc3.2/linux2.4
--enable-languages=c,c++ --enable-threads
Thread model: posix
gcc version 3.2

> uname -a
Linux angelo 2.4.7-10smp #1 SMP Thu Sep 6 17:09:31 EDT 2001 i686 unknown

Same crash on...
NT 5.1 / Cygwin GCC 3.2 prerelease-3
Sun 2.8 GCC 3.2

I have not confirmed that the crash still occurs with GCC 3.2.3 or GCC 3.3.1 or
GCC 3.4.