attached example program produces wrong results when compiled as 64 bit binary. it works ok when compiled with -Os optimization OR into 32 bit binary (forced with -m32 switch). it grabs two timestamps (format: YYYYMMDDHH:MM:SS), converts them with strptime into tm struct and displays unix epoch time generated with mktime() - when compiled as 64 bit binary, one timestamp has wrong value (the difference between original timestamps is 180s, whereas it shows it's 3420 seconds after using mktime()). it shows proper 180 seconds difference when compiled with -m32 or -Os. my gcc -v: Reading specs from /usr/lib/gcc/x86_64-pc-linux-gnu/3.4.1/specs Configured with: /usr/local/portage/portage/gcc-3.4.1/work/gcc-3.4.1/configure --prefix=/usr --bindir=/usr/x86_64-pc-linux-gnu/gcc-bin/3.4 --includedir=/usr/lib/gcc/x86_64-pc-linux-gnu/3.4.1/include --datadir=/usr/share/gcc-data/x86_64-pc-linux-gnu/3.4 --mandir=/usr/share/gcc-data/x86_64-pc-linux-gnu/3.4/man --infodir=/usr/share/gcc-data/x86_64-pc-linux-gnu/3.4/info --enable-shared --host=x86_64-pc-linux-gnu --target=x86_64-pc-linux-gnu --with-system-zlib --enable-languages=c,c++,objc --enable-threads=posix --enable-long-long --disable-checking --disable-libunwind-exceptions --enable-cstdio=stdio --enable-version-specific-runtime-libs --with-gxx-include-dir=/usr/lib/gcc/x86_64-pc-linux-gnu/3.4.1/include/g++-v3 --with-local-prefix=/usr/local --disable-werror --enable-shared --enable-nls --without-included-gettext --enable-multilib --enable-__cxa_atexit --enable-clocale=gnu Thread model: posix gcc version 3.4.1 (Gentoo Linux 3.4.1, ssp-3.4-2, pie-8.7.6.3)
Created attachment 6906 [details] wrong timestamps test case
So this fails at -O0 right? Also does this fail with an unmodified compiler as the compiler you are using has been modified by Gentoo.
yes, it behaves ugly with -O0 (only with -Wall in fact). I'll try to test how does it work with vanilla gcc 3.4.1 soon.
the bug is absent with plain gcc from gnu.org. moreover, the original bug seems to be gone after glibc upgrade, so I guess this successfully classifies this bug as missed...
So closing as worksforme.
unfortunately - the "bug" is still there until I have setlocale(LC_TIME, "C") before calling sesstime(). in fact it generates wrong results whatever LC_TIME value I use (and 64 bit code is generated). setlocale() has no effect when compiled with -m32 (it works ok then). this makes me wonder if it may be rather some glibc issue rather than gcc problem. I'd appreciate any hints on that.