DRM and latency..

2004-05-19 Thread Dave Airlie
It has been pointed out to me by Andrew and Fernando Pablo Lopez-Lezcano that we do a lot of sleeping in code that probably should reschedule rather than sleep,radeon_do_wait_for_idle being a prime example, this has a DRM_UDELAY(1), this messes up audio latencys, Now I'm not sure that wholesale

Re: DRM and latency..

2004-05-20 Thread Keith Whitwell
Dave Airlie wrote: It has been pointed out to me by Andrew and Fernando Pablo Lopez-Lezcano that we do a lot of sleeping in code that probably should reschedule rather than sleep,radeon_do_wait_for_idle being a prime example, this has a DRM_UDELAY(1), this messes up audio latencys, Now I'm not sur

Re: DRM and latency..

2004-05-20 Thread Dieter Nützel
Am Donnerstag, 20. Mai 2004 06:52 schrieb Dave Airlie: > It has been pointed out to me by Andrew and Fernando Pablo Lopez-Lezcano > that we do a lot of sleeping in code that probably should reschedule > rather than sleep,radeon_do_wait_for_idle being a prime example, this has > a DRM_UDELAY(1), th

Re: DRM and latency..

2004-05-20 Thread Joe O
On Thu, 20 May 2004, Keith Whitwell wrote: > Dave Airlie wrote: > > It has been pointed out to me by Andrew and Fernando Pablo Lopez-Lezcano > > that we do a lot of sleeping in code that probably should reschedule > > rather than sleep,radeon_do_wait_for_idle being a prime example, this has > >