On 05/02/2014 11:01 PM, Bruce Dubbs wrote:
Frans de Boer wrote:
By the way, don't you think that it is strange that additional errors
surface - at different tests - when code optimization is explicitly
switched off?
From the compiler writer's point of view, getting optimization right is
really hard. It doesn't surprise me when things like this show up when
a new version of the compiler comes out. The issues may be in the
compiler or may be poor code. Both need to be examined. Generally
after a while, the issues get worked out.
-- Bruce
I know it is hard, especially with modern day complex instructions
microprocessors. Many years ago I made an optimizing C-compiler for the
Z80 (yes, I am from that era), but when I tried it to apply to the then
state of the art i8086, I gave up because for a fun project it cost me
to much time. And that was a relatively simple 16-bit processor.
Okay, what I said before, maybe just wait until gcc-4.9.1 comes along
and accept the current failures as being glitches without severe
repercussions (I hope).
Frans.
--
http://lists.linuxfromscratch.org/listinfo/lfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page
