This is the mail archive of the gcc-help@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

RE: WELCOME to gcc-help@gcc.gnu.org


Hello,
 
I'm working on a project for which a MN103 compiler is needed. We
currently use Matsushita's standard compiler, which in turn is gcc 3.2.2
patched and corrected by RedHat. This one works well, but the
corrections seem to have never been released to the open source
community.
 
So, in order to get rid of this compiler (which is furthermore very
expensive) I am looking for a free GCC compiler for MN103 (i.e. one
directly provided by GNU) and I have tested GCC 3.3.2 and GCC 3.4.5 for
MN103 (cross compiler).
 
Unfortunately, those compilers seem not to work correctly when used with
optimization enabled (-O2). I cannot give a lot of information on this
bug, except that our test suites crash when -O2 optimizations are
activated (but not with Matsushita's GCC compiler). The bug is not so
frequent and seem to happen especially when a lot of context switches
occur (in multithreaded stress tests).
 
So, my question is : do you have any information about a potential bug
related to optimization in GCC's MN103 port ? Can you give me some
advice ?
 
Also, I think RedHat should have contributed back to the open source
community by giving their corrections and patches to the FSF. Am I right
?
 
Best regards,


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]