[Mesa-dev] [PATCH 3/3] i965: Apply a homebrew workaround for GPU hang in OGLC api-texcoord.

[Chris Wilson]

On Tue, 19 Jul 2011 15:44:12 -0700, Eric Anholt <eric at anholt.net> wrote:
> The behavior of flushes in the hardware is a maze of twisty passages,
> and strangely the VS constants appear to be loaded during a pipeline
> flush instead of at the time of the packet emit according to the
> simulator.  On moving the STATE_BASE_ADDRESS packet to where it really
> needed to live (in order for data loads by other packets to be
> correct), we sometimes no longer got a flush between those packets
> where we apparently needed it.  This replicates the flushes implied by
> a STATE_BASE_ADDRESS update, fixing the GPU hangs in OGLC and the
> "engine" demo.

Whilst you are deep in flushing territory, do you know of any similar
rules regarding 3DSTATE_CC_STATE_POINTERS? rendercheck demands a NP or
flush before the blend mode is changed.
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre