[PATCH 6/6] RADEONPrepareAccess_CS: fallback to DFS when pixmap is in VRAM

[Karl Tomlinson]

Alex Deucher writes:

> 2010/9/1  <xmail at karlt.net>:
>> Michel Dänzer writes:

>>> , still it would be good to make some additional measurements
>>> to make sure there aren't any bad regressions.
>>
>> I'll look into it.  Let me know if there is a subset of x11perf
>> tests that is interesting here.  I just saw "This may take a
>> while" and I don't have "a while" today, but I guess running all
>> the tests is better anyway.
>
> Probably gtkperf would be more useful since lots of x11perf tests
> measure stuff that's never used in most desktops.  Still, would useful
> to see how things are affected overall even if many of them aren't
> used much.

I did some gtkperf runs against xserver 1.8.2 (with unpatched exa).

With the compositing window manager running, I couldn't see any
differences in results either in the total number or individual
tests.

With compositing disabled, some runs with the unpatched driver
were 10-15 % slower, but the particular set of tests that were
slower in each of the slower runs was not consistent, so I
wouldn't read too much into this.

The results with the patched driver were very similar to the
fastest runs with the unpatched driver.