<div dir="ltr"><div class="gmail_extra"><div class="gmail_quote">On 1 June 2018 at 22:51, Dylan Baker <span dir="ltr"><<a href="mailto:dylan@pnwbakers.com" target="_blank">dylan@pnwbakers.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">Quoting Samuel Pitoiset (2018-06-01 08:58:42)<br>
<div><div class="gmail-h5">> <br>
> <br>
> On 06/01/2018 05:48 PM, Dylan Baker wrote:<br>
> > Quoting Alex Smith (2018-06-01 07:56:38)<br>
> >> On 1 June 2018 at 15:48, Dylan Baker <<a href="mailto:dylan@pnwbakers.com">dylan@pnwbakers.com</a>> wrote:<br>
> >><br>
> >> Quoting Alex Smith (2018-05-31 08:44:18)<br>
> >> > With GFX9 merged shaders, active_stages would be set to the original<br>
> >> > stages specified if shaders were not cached, but to the stages still<br>
> >> > present after merging if they were.<br>
> >> ><br>
> >> > Be consistent and use the original stages.<br>
> >> ><br>
> >> > Signed-off-by: Alex Smith <<a href="mailto:asmith@feralinteractive.com">asmith@feralinteractive.com</a>><br>
> >> > Cc: "18.1" <<a href="mailto:mesa-stable@lists.freedesktop.org">mesa-stable@lists.<wbr>freedesktop.org</a>><br>
> >> > ---<br>
> >> > src/amd/vulkan/radv_pipeline.c | 7 ++-----<br>
> >> > 1 file changed, 2 insertions(+), 5 deletions(-)<br>
> >> ><br>
> >> > diff --git a/src/amd/vulkan/radv_<wbr>pipeline.c b/src/amd/vulkan/radv_<br>
> >> pipeline.c<br>
> >> > index 52734a308a..18dcc43ebe 100644<br>
> >> > --- a/src/amd/vulkan/radv_<wbr>pipeline.c<br>
> >> > +++ b/src/amd/vulkan/radv_<wbr>pipeline.c<br>
> >> > @@ -1964,6 +1964,8 @@ void radv_create_shaders(struct radv_pipeline<br>
> >> *pipeline,<br>
> >> > _mesa_sha1_compute(modules[i]<wbr>->nir-><br>
> >> <a href="http://info.name" rel="noreferrer" target="_blank">info.name</a>,<br>
> >> > strlen(modules[i]-><br>
> >> nir-><a href="http://info.name" rel="noreferrer" target="_blank">info.name</a>),<br>
> >> > modules[i]->sha1);<br>
> >> > +<br>
> >> > + pipeline->active_stages |=<br>
> >> mesa_to_vk_shader_stage(i);<br>
> >> > }<br>
> >> > }<br>
> >> ><br>
> >> > @@ -1979,10 +1981,6 @@ void radv_create_shaders(struct radv_pipeline<br>
> >> *pipeline,<br>
> >> ><br>
> >> > if (radv_create_shader_variants_<wbr>from_pipeline_cache(device,<br>
> >> cache, hash, pipeline->shaders) &&<br>
> >> > (!modules[MESA_SHADER_<wbr>GEOMETRY] || pipeline->gs_copy_shader))<br>
> >> {<br>
> >> > - for (unsigned i = 0; i < MESA_SHADER_STAGES; ++i) {<br>
> >> > - if (pipeline->shaders[i])<br>
> >> > - pipeline->active_stages |=<br>
> >> mesa_to_vk_shader_stage(i);<br>
> >> > - }<br>
> >> > return;<br>
> >> > }<br>
> >> ><br>
> >> > @@ -2015,7 +2013,6 @@ void radv_create_shaders(struct radv_pipeline<br>
> >> *pipeline,<br>
> >> > stage ? stage->pName<br>
> >> : "main", i,<br>
> >> > stage ? stage-><br>
> >> pSpecializationInfo : NULL,<br>
> >> > flags);<br>
> >> > - pipeline->active_stages |= mesa_to_vk_shader_stage(i);<br>
> >> ><br>
> >> > /* We don't want to alter meta shaders IR directly so<br>
> >> clone it<br>
> >> > * first.<br>
> >> > --<br>
> >> > 2.14.3<br>
> >> ><br>
> >> > ______________________________<wbr>_________________<br>
> >> > mesa-stable mailing list<br>
> >> > <a href="mailto:mesa-stable@lists.freedesktop.org">mesa-stable@lists.freedesktop.<wbr>org</a><br>
> >> > <a href="https://lists.freedesktop.org/mailman/listinfo/mesa-stable" rel="noreferrer" target="_blank">https://lists.freedesktop.org/<wbr>mailman/listinfo/mesa-stable</a><br>
> >><br>
> >> Hi Alex,<br>
> >><br>
> >> This doesn't apply cleanly to the 18.1 tree with the following collision:<br>
> >><br>
> >> ++<<<<<<< HEAD<br>
> >> <br>
> >> + stage ? stage-><br>
> >> pSpecializationInfo : NULL);<br>
> >> + pipeline->active_stages |= mesa_to_vk_shader_stage(i);<br>
> >> <br>
> >> ++=======<br>
> >> <br>
> >> + stage ? stage-><br>
> >> pSpecializationInfo : NULL,<br>
> >> + flags);<br>
> >> <br>
> >> ++>>>>>>> 0fa51bfdbe5... radv: Set active_stages the same whether or not<br>
> >> shaders were cached<br>
> >><br>
> >> I can remove the flags field (which doesn't exist in 18.1) before merging<br>
> >> if you<br>
> >> think that's the right thing to do. Does that seem reasonable?<br>
> >><br>
> >><br>
> >> Yes, that's correct. Should just be removing the pipeline->active_stages line<br>
> >> at that point.<br>
> >><br>
> >> Thanks,<br>
> >> Alex<br>
> > <br>
> > Thanks! This is in the tree for 18.1.2.<br>
> <br>
> Apparently, this series regresses some CTS on my side, Bas will <br>
> investigate, would be nice to not release 18.1.2 without this fixed. :)<br>
> <br>
<br>
</div></div>Let me know what you all want to do, 18.1.1 happened today, and 18.1.2 is<br>
planned for release on the 15th, so you have some time to figure it out.<br></blockquote><div><br></div><div>This commit fixes the issue: <a href="https://cgit.freedesktop.org/mesa/mesa/commit/?id=2835b6baf446d0ff3b3df6eefc57b248a505af36">https://cgit.freedesktop.org/mesa/mesa/commit/?id=2835b6baf446d0ff3b3df6eefc57b248a505af36</a></div><div><br></div><div>Thanks,</div><div>Alex</div><div> </div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
<span class="gmail-HOEnZb"><font color="#888888"><br>
Dylan<br>
</font></span></blockquote></div><br></div></div>