On Tue, Dec 16, 2008 at 01:57:23PM +0100, Stefan Assmann wrote:
>Bernhard Reutner-Fischer wrote:
>> On Tue, Dec 16, 2008 at 11:04:49AM +0100, Stefan Assmann wrote:

>>> The current implementation relies on the fact that INTERNAL_SYSCALL is 
>>> defined
>>> for syscall sched_setaffinity. Otherwise the syscall will be treated as
>>> non-existant even if it is implemented. This patch adds a check whether the
>>> syscall is available if INTERNAL_SYSCALL is not defined. This was necessary 
>>> to
>>> get sched_setaffinity working on ia64.
>> 
>> All arches should define INTERNAL_SYSCALL, could you please add this
>> for ia64 instead?
>
>Yes, can do. Any objections why this patch shouldn't be applied anyway? What's
>the reason for not using the existing syscall?

We want to switch to use INTERNAL_SYSCALL everywhere (so we don't have to do
workarounds in libpthread, IIRC). Applying that patch would just paper over
the fact that ia64 is apparently not converted yet (as a couple of other
arches that apparently nobody uses or cares for) and as such doesn't fix the
real issue.

TIA,
_______________________________________________
uClibc mailing list
[email protected]
http://busybox.net/cgi-bin/mailman/listinfo/uclibc

Reply via email to