[gccsdk] Dynamic libraries vs pthreads

Lee Noar leenoar at sky.com
Fri Aug 29 12:35:28 PDT 2008


Peter Naulls wrote:
> Peter Naulls wrote:
>> John Tytgat wrote:
>>> In message <48B58F7B.7030303 at sky.com>
>>>           Lee Noar <leenoar at sky.com> wrote:
>>>
>>>> Ah, right, thanks. You're right, I do have a revision K so that 
>>>> probably explains it. Perhaps I should commit this change to 
>>>> pthread_yield.
>>>
>>> Yes, that makes sense (with suitable comment so we don't forget why it
>>> like that please).
>>>
>>> Does this also mean that this could be a cause of crashes on some RiscPC
>>> machines with e.g. the Firefox port before ? Or this is something which
>>> can only happen with the current shared libraries code ?
>>
>> In particular, does the existing static build of Firefox 2.0.0.14
>> work on Lee's machine?
> 
> I meant of course, the 2.0.0.12-riscpc version, unless he wants to
> build 2.0.0.14 himself.

Yes, 2.0.0.12-riscpc does work on my machine, and I see that Firefox 
uses pthread_yield() quite extensively. Whatever triggers this bug must 
be fairly rare as it doesn't always fail. I don't see how it could be 
specific to shared library code as it really isn't any different to 
non-PIC code as far as the CPU is concerned.

Lee.




More information about the gcc mailing list