[LLVMbugs] [Bug 21365] New: Mesa llvmpipe lp_test_format regression with llvm-3.6.0svn
bugzilla-daemon at llvm.org
bugzilla-daemon at llvm.org
Thu Oct 23 13:54:33 PDT 2014
http://llvm.org/bugs/show_bug.cgi?id=21365
Bug ID: 21365
Summary: Mesa llvmpipe lp_test_format regression with
llvm-3.6.0svn
Product: new-bugs
Version: trunk
Hardware: PC
URL: https://bugs.freedesktop.org/show_bug.cgi?id=85377
OS: All
Status: NEW
Keywords: regression
Severity: normal
Priority: P
Component: new bugs
Assignee: unassignedbugs at nondot.org
Reporter: vlee at freedesktop.org
CC: brian.e.paul at gmail.com, chandlerc at gmail.com,
jfonseca at vmware.com, llvmbugs at cs.uiuc.edu,
sroland at vmware.com
Classification: Unclassified
llvm-3.6.0svn r220138 introduced a regression with Mesa llvmpipe lp_test_format
unit test.
$ lp_test_format
Testing PIPE_FORMAT_B8G8R8A8_UNORM (float) ...
Testing PIPE_FORMAT_B8G8R8A8_UNORM (unorm8) ...
Testing PIPE_FORMAT_B8G8R8X8_UNORM (float) ...
Testing PIPE_FORMAT_B8G8R8X8_UNORM (unorm8) ...
Testing PIPE_FORMAT_A8R8G8B8_UNORM (float) ...
Testing PIPE_FORMAT_A8R8G8B8_UNORM (unorm8) ...
Testing PIPE_FORMAT_X8R8G8B8_UNORM (float) ...
Testing PIPE_FORMAT_X8R8G8B8_UNORM (unorm8) ...
Testing PIPE_FORMAT_B5G5R5A1_UNORM (float) ...
Testing PIPE_FORMAT_B5G5R5A1_UNORM (unorm8) ...
Testing PIPE_FORMAT_B4G4R4A4_UNORM (float) ...
Testing PIPE_FORMAT_B4G4R4A4_UNORM (unorm8) ...
Testing PIPE_FORMAT_B5G6R5_UNORM (float) ...
Testing PIPE_FORMAT_B5G6R5_UNORM (unorm8) ...
Testing PIPE_FORMAT_R10G10B10A2_UNORM (float) ...
Testing PIPE_FORMAT_R10G10B10A2_UNORM (unorm8) ...
Testing PIPE_FORMAT_L8_UNORM (float) ...
Testing PIPE_FORMAT_L8_UNORM (unorm8) ...
Testing PIPE_FORMAT_A8_UNORM (float) ...
Testing PIPE_FORMAT_A8_UNORM (unorm8) ...
Testing PIPE_FORMAT_I8_UNORM (float) ...
Testing PIPE_FORMAT_I8_UNORM (unorm8) ...
Testing PIPE_FORMAT_L8A8_UNORM (float) ...
Testing PIPE_FORMAT_L8A8_UNORM (unorm8) ...
Testing PIPE_FORMAT_L16_UNORM (float) ...
Testing PIPE_FORMAT_L16_UNORM (unorm8) ...
Testing PIPE_FORMAT_UYVY (float) ...
Testing PIPE_FORMAT_UYVY (unorm8) ...
Testing PIPE_FORMAT_YUYV (float) ...
Testing PIPE_FORMAT_YUYV (unorm8) ...
Testing PIPE_FORMAT_R32_FLOAT (float) ...
Testing PIPE_FORMAT_R32_FLOAT (unorm8) ...
Testing PIPE_FORMAT_R32G32_FLOAT (float) ...
Testing PIPE_FORMAT_R32G32_FLOAT (unorm8) ...
Testing PIPE_FORMAT_R32G32B32_FLOAT (float) ...
Testing PIPE_FORMAT_R32G32B32_FLOAT (unorm8) ...
Testing PIPE_FORMAT_R32G32B32A32_FLOAT (float) ...
Segmentation fault (core dumped)
9b2d091a9c2509a766a233a64453462faa2bdffc is the first bad commit
commit 9b2d091a9c2509a766a233a64453462faa2bdffc
Author: Chandler Carruth <chandlerc at gmail.com>
Date: Sat Oct 18 06:36:22 2014 +0000
[InstCombine] Do an about-face on how LLVM canonicalizes (cast (load
...)) and (load (cast ...)): canonicalize toward the former.
Historically, we've tried to load using the type of the *pointer*, and
tried to match that type as closely as possible removing as many pointer
casts as we could and trading them for bitcasts of the loaded value.
This is deeply and fundamentally wrong.
Repeat after me: memory does not have a type! This was a hard lesson for
me to learn working on SROA.
There is only one thing that should actually drive the type used for
a pointer, and that is the type which we need to use to load from that
pointer. Matching up pointer types to the loaded value types is very
useful because it minimizes the physical size of the IR required for
no-op casts. Similarly, the only thing that should drive the type used
for a loaded value is *how that value is used*! Again, this minimizes
casts. And in fact, the *only* thing motivating types in any part of
LLVM's IR are the types used by the operations in the IR. We should
match them as closely as possible.
I've ended up removing some tests here as they were testing bugs or
behavior that is no longer present. Mostly though, this is just cleanup
to let the tests continue to function as intended.
The only fallout I've found so far from this change was SROA and I have
fixed it to not be impeded by the different type of load. If you find
more places where this change causes optimizations not to fire, those
too are likely bugs where we are assuming that the type of pointers is
"significant" for optimization purposes.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220138
91177308-0d34-0410-b5e6-96231b3b80d8
:040000 040000 982ecfc2706221c9731e5e8f54346370d2fe6283
667a68ea436b53840beda4b034cceee0e87dc6eb M lib
:040000 040000 a6f61cd09e99c468c8fdb8dd6bda10060b4f751f
cad7ad5d09cf0f43b3f47e36050c833b2debc9fc M test
bisect run success
--
You are receiving this mail because:
You are on the CC list for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-bugs/attachments/20141023/a5a63890/attachment.html>
More information about the llvm-bugs
mailing list