[Bug 109081] [bisected] [HSW] Regression in clipping.user_defined.clip_* vulkancts tests

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=109081

--- Comment #7 from Timothy Arceri <t_arceri at yahoo.com.au> ---
(In reply to Ian Romanick from comment #6)
> (In reply to Timothy Arceri from comment #5)
> > (In reply to Ian Romanick from comment #3)
> > > I've submitted an MR for a possible fix for problem #2.
> > > 
> > > https://gitlab.freedesktop.org/mesa/mesa/merge_requests/26
> > > 
> > > I'm have an idea for problem #1, but I don't know that I'll get to it before
> > > the new year.  Maybe Tim will beat me to it. ;)
> > 
> > The problem with peephole select is that we loose opportunities to apply the
> > growing list of opts in nir_opt_if(). In this case we no longer recognise
> > the pattern in opt_peel_loop_initial_if()
> 
> I suspected it was something like that.  I was planning to go through and
> beef up the places that handle if-statements to also handle bcsel.  That
> doesn't (conceptually) seem to hard.  What do you think?

It should be fine for opt_peel_loop_initial_if() but for things like
opt_if_evaluate_condition_use() and
https://patchwork.freedesktop.org/patch/263442/ it's not really possible to do.

-- 
You are receiving this mail because:
You are the QA Contact for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/intel-3d-bugs/attachments/20181218/8143c48d/attachment.html>