On Wed, Nov 17, 2010 at 03:03:20PM +1100, matthew green wrote: > > > - hppa seems to have a mutex implemented without cas. is it broken? > > FWIW, the sparc implementation is only broken on SMP.
My recollection is that there were problems with the assembly that we were using, we didn't get to the bottom of it. On top of that I think there is a race condition with the interlock trick that we were going to use for sparc but I can't remember what it was. Possibly related to preemption or interrupts (i.e. mutex_enter inteerupted with the interlock part held but mutex owner not set). If it comes to me I'll post back.
