<html>
<head>
<base href="https://bugs.freedesktop.org/">
</head>
<body><table border="1" cellspacing="0" cellpadding="8">
<tr>
<th>Bug ID</th>
<td><a class="bz_bug_link
bz_status_NEW "
title="NEW - NIR verification crashes on piglit tests"
href="https://bugs.freedesktop.org/show_bug.cgi?id=107366">107366</a>
</td>
</tr>
<tr>
<th>Summary</th>
<td>NIR verification crashes on piglit tests
</td>
</tr>
<tr>
<th>Product</th>
<td>Mesa
</td>
</tr>
<tr>
<th>Version</th>
<td>git
</td>
</tr>
<tr>
<th>Hardware</th>
<td>Other
</td>
</tr>
<tr>
<th>OS</th>
<td>All
</td>
</tr>
<tr>
<th>Status</th>
<td>NEW
</td>
</tr>
<tr>
<th>Keywords</th>
<td>bisected, regression
</td>
</tr>
<tr>
<th>Severity</th>
<td>normal
</td>
</tr>
<tr>
<th>Priority</th>
<td>medium
</td>
</tr>
<tr>
<th>Component</th>
<td>Drivers/DRI/i965
</td>
</tr>
<tr>
<th>Assignee</th>
<td>danylo.piliaiev@gmail.com
</td>
</tr>
<tr>
<th>Reporter</th>
<td>mark.a.janes@intel.com
</td>
</tr>
<tr>
<th>QA Contact</th>
<td>intel-3d-bugs@lists.freedesktop.org
</td>
</tr>
<tr>
<th>CC</th>
<td>baker.dylan.c@gmail.com, jason@jlekstrand.net
</td>
</tr></table>
<p>
<div>
<pre>Running piglit tests with NIR_TEST_CLONE=true NIR_TEST_SERIALIZE=true,
thousands of piglit tests crash:
NIR_TEST_CLONE=true NIR_TEST_SERIALIZE=true
/tmp/build_root/m64/lib/piglit/bin/shader_runner
/tmp/build_root/m64/lib/piglit/generated_tests/spec/arb_tessellation_shader/execution/built-in-functions/tcs-op-mult-float-mat4.shader_test
-auto -fbo
piglit: debug: Requested an OpenGL 3.2 Core Context, and received a matching
4.5 context
Segmentation fault
Bisected to:
d219521379626ebf2bff63e4a9c2f92725b3926a is the first bad commit
commit d219521379626ebf2bff63e4a9c2f92725b3926a
Author: Danylo Piliaiev <<a href="mailto:danylo.piliaiev@gmail.com">danylo.piliaiev@gmail.com</a>>
Date: Wed Jul 11 15:29:00 2018 +0300
i965: Sweep NIR after linking phase to free held memory
After optimization passes and many trasfromations most of memory
NIR holds is a garbage which was being freed only after shader deletion.
Freeing it at the end of linking will save memory which would be useful
in case there are a lot of complex shaders being compiled.
The common case for this issue is 32bit game running under Wine.
The cost of the optimization is around ~3-5% of compilation speed
with complex shaders.
V2: by Jason Ekstrand
- Move nir_sweep up, right after the last change of NIR
Bugzilla: <a class="bz_bug_link
bz_status_RESOLVED bz_closed"
title="RESOLVED FIXED - BRW allocates too much heap memory"
href="show_bug.cgi?id=103274">https://bugs.freedesktop.org/show_bug.cgi?id=103274</a>
Signed-off-by: Danylo Piliaiev <<a href="mailto:danylo.piliaiev@globallogic.com">danylo.piliaiev@globallogic.com</a>>
Reviewed-by: Jason Ekstrand <<a href="mailto:jason@jlekstrand.net">jason@jlekstrand.net</a>>
Cc: <a href="mailto:mesa-stable@lists.freedesktop.org">mesa-stable@lists.freedesktop.org</a></pre>
</div>
</p>
<hr>
<span>You are receiving this mail because:</span>
<ul>
<li>You are the QA Contact for the bug.</li>
</ul>
</body>
</html>