https://gcc.gnu.org/bugzilla/show_bug.cgi?id=123132
--- Comment #6 from ro at CeBiTec dot Uni-Bielefeld.DE <ro at CeBiTec dot Uni-Bielefeld.DE> --- > --- Comment #5 from Jose E. Marchesi <jemarch at gcc dot gnu.org> --- > So endianness it was 8-) > > However, there are still some tests failing with -j32. This is due to a race > condition. Clearly the infrastructure we have to compile the modules used by > a > particular test doesn't play well with parallel make check when there is a > large number of cores... That's a bit weird though: I don't see this neither on i386-pc-solaris2.11 with -j56 nor on sparc-sun-solaris2.11 with -j96. The only failures I still see (either sparc or x86) are the widening-bits-[23].a68 ones (PR testsuite/123008)
