On Wed, Oct 30, 2013 at 1:02 AM, Sebastian Huber <sebastian.hu...@embedded-brains.de> wrote: > > In the file "libgcc/config/arm/t-elf", there is this block: > > # Currently there is a bug somewhere in GCC's alias analysis > # or scheduling code that is breaking _fpmul_parts in fp-bit.c. > # Disabling function inlining is a workaround for this problem. > HOST_LIBGCC2_CFLAGS += -fno-inline > > This -fno-inline seems to be there since at least 2011. Is this bug still > present in GCC?
I traced the use of this option back to revision 33028, committed in April 2000: http://gcc.gnu.org/ml/gcc-patches/2000-04/msg00305.html . I didn't try to track down where the patch was added on the branch. The bug, whatever it was, is almost certainly long gone. In any case, the comment provides enough information to check: the bug showed up when doing floating-point multiplications using software floating point. Ian
