[Mesa-dev] [PATCH 0/6] Constants through builtins evaluation rewrite

[Olivier Galibert]

On Wed, May 02, 2012 at 11:11:36PM +0200, Olivier Galibert wrote:
> What's missing is a piglit run, which I currently have technical
> problems with.

Thanks to very pertinent advice by ajax and anholt I've been able to
do a piglit run.  With three additional bugfix patches I've sent
independently there are no regressions.

The three patches:
- one to piglit, which assumes that atan is more precise than it is in mesa
- one to mesa to ensure that acos(1) is exactly zero
- another one to mesa that fixes a number of bugs w.r.t relational operators and vectors

They are all on the appropriate mailing-lists.

Best,

  OG.