On Thu, 2009-07-02 at 23:20 +0200, Maarten Maathuis wrote: > This smells like taping over a bigger issue imo. The only reason the > old code did this because it didn't have interrupts to signal clock > changes (and such). I have no other solution, nor personal experience, > because i've been away (and still am). I just want to give my 2 cents. I agree, this really should not be required. The display channel should stall on its own until the appropriate IRQs have been handled, if required.
It might be useful to know *which* of these "fixes" the issue to track down the real problem. Ben. > > Maarten. > _______________________________________________ > Nouveau mailing list > [email protected] > http://lists.freedesktop.org/mailman/listinfo/nouveau _______________________________________________ Nouveau mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/nouveau
