On Sun, 3 Mar 2002, Christoph Hellwig wrote:

> On Fri, Mar 01, 2002 at 08:05:20PM -0500, Rob Radez wrote:
> > Hi,
> > I was just wondering what the point of the wrappers around the semaphore
> > locking was, so I ginned up this patch against the linux25 cvs tree.  I
> > believe it should be functionally equivalent and get rid of the ugly
> > wrappers.  Now I just need to get 2.5.6-pre2 compiling instead of
> > erroring out :-).  Hopefully you won't hate it, and hopefully I didn't
> > screw things up too badly.
>
> Umm, the internals of rw_semaphore are supposed to be opaque to
> architecture-independand code.  E.g include/asm-i386/rwsem.h in
> 2.5.6-pre2 doesn't even have an 'activity' member.
Oops, looking at sparc code too long ;-).  I was kind of hoping not to
have to include a patch to core code, so I guess my question is, is the
ugliness of including a private copy of the trylock code less ugly than
having a wrapper around a rw_semaphore?

Regards,
Rob Radez

_______________________________________________
Jfs-discussion mailing list
[EMAIL PROTECTED]
http://www-124.ibm.com/developerworks/oss/mailman/listinfo/jfs-discussion

Reply via email to