[Mesa-dev] [PATCH] i965 : optimized bucket index calculation

[Ian Romanick]

On 09/08/2017 08:38 AM, Jason Ekstrand wrote:
> In general, I'm very concerned about how this handles rounding
> behavior.  Almost everywhere, you round down when what you want to do is
> round up.  Also, as I said on IRC, I'd like to see some asserts in
> add_bucket so that we are sure this calculation is correct.  In
> particular, I'd like to see
> 
> assert(bucket_for_size(size) == &bufmgr->cache_bucket[i]);
> assert(bucket_for_size(size - 2048) == &bufmgr->cache_bucket[i]);
> assert(bucket_for_size(size + 1) != &bufmgr->cache_bucket[i]);
> 
> We need to check on both sides of size to be 100% sure we're doing our
> rounding correctly.
> 
> On Fri, Sep 8, 2017 at 1:11 AM, <aravindan.muthukumar at intel.com
> <mailto:aravindan.muthukumar at intel.com>> wrote:
> 
>     From: Aravindan Muthukumar <aravindan.muthukumar at intel.com
>     <mailto:aravindan.muthukumar at intel.com>>
> 
>     Avoiding the loop which was running with O(n) complexity.
>     Now the complexity has been reduced to O(1)
> 
>     Tested with piglit.
>     Slight performance improvement (~1%) in 3d mark.
> 
> 
> Which 3dmark test?  Also, what's the error in that 1%?

And, is that Atom or Core?