This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: ARM THUMB: fundamental bug in handling of far jumps?
On Tue, 2002-01-29 at 22:46, Richard Henderson wrote:
> Not-taken branches are normally better for performance. Might not
> apply to the thumb though.
afaict at least on arm7tdmi branches are indifferent to whether they are
taken or not. But this is probably different on all bigger ARM cpus. So
having this architecture dependent would in some cases only pay off if
it would be specific to the architecture, not only the target - so a
simple compile time flag wouldn't be enough.
> I could tell you how to turn it off (modifying FIXED_REGS), but
> I don't know enough about the arm or thumb to say if that would
> get you into trouble because of something else.
Thanks. Don't know if I come around to trying it, though (changing
project assignment here at the company...)
greets from Zürich
-- vbi