On Fri, Aug 08, 2014 at 11:34:07AM +0100, Siluvery, Arun wrote: > On 08/08/2014 10:57, Chris Wilson wrote: > >On Fri, Aug 08, 2014 at 10:52:57AM +0100, arun.siluvery@xxxxxxxxxxxxxxx wrote: > >>From: Arun Siluvery <arun.siluvery@xxxxxxxxxxxxxxx> > >> > >>Workarounds for bdw are currently applied in init_clock_gating() but they > >>are lost following a gpu reset. Some of the registers are part of register > >>state context and they are restored with every context switch so initializing > >>WAs in golden render state ensures that they are applied even when we start > >>with an uninitialized context or during hw initialization followed by a reset. > > > >Interesting, but let's try to keep the opaque blobs minimal. The > >comments for w/a are even more valuable than the code. > I agree, I will add comments to each workaround. > We are looking at augmenting workarounds to the null batch in render > state setup function itself. Do you have any comments with that > approach? Other than changing its name, that does actually seem like a useful juncture to do things since it is the first userspace operation on the RCS ring within that context. And if it has to be within the context, that is indeed more troublesome to do without adding to the renderstate. -Chris -- Chris Wilson, Intel Open Source Technology Centre _______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx http://lists.freedesktop.org/mailman/listinfo/intel-gfx
