[Bug target/42159] app compiled with 4.4.2 SIGABRTs after a trivial nested throw/stack unwinding

["vlad at demoninsight dot com" <gcc-bugzilla at gcc dot gnu dot org>]

------- Comment #14 from vlad at demoninsight dot com  2009-12-01 05:50 -------
(In reply to comment #13)
> (In reply to comment #12)
> > Compile-time.
> 
> configure with --enable-checking=release to turn off checks that are enabled by
> default in pre-release builds, that will give a better comparison between the
> 4.4.2 release and 4.5 snapshot
> 

Thank you for a useful tip. I have re-built the same snapshot:

Using built-in specs.
COLLECT_GCC=gcc-4.5
COLLECT_LTO_WRAPPER=/sw/lib/gcc4.5.trunk/libexec/gcc/x86_64-apple-darwin10/4.5.0/lto-wrapper
Target: x86_64-apple-darwin10
Configured with: ../src/configure --enable-checking=release
--prefix=/sw/lib/gcc4.5.trunk --mandir=/sw/share/man --infodir=/sw/share/info
--enable-languages=c,c++,fortran --with-gmp=/sw --with-libiconv-prefix=/sw
--with-ppl=/sw --with-cloog=/sw --with-system-zlib
--x-includes=/usr/X11R6/include --x-libraries=/usr/X11R6/lib
--disable-libjava-multilib --build=x86_64-apple-darwin10
--host=x86_64-apple-darwin10 --target=x86_64-apple-darwin10
Thread model: posix
gcc version 4.5.0 20091128 (experimental) (GCC) 

It did make a noticeable difference. With 4.4.2 and 4.5 snapshot installed in
parallel dirs, I ran the same build in two shells and simply pointing the
builder from one version to another. The times are quite reproducible:

4.4.2: ~ 337 sec
4.5:   ~ 466 sec

(averages of two runs each). So, 4.5 snapshot is 40% slower at compile time.
This is quite a bit better than without --enable-checking=release (which was
close to 2x slower) but seems a tad heavy price to pay to just have exceptions
working ...


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42159