New issue
Advanced search Search tips
Note: Color blocks (like or ) mean that a user may not be available. Tooltip shows the reason.

Issue 844018 link

Starred by 1 user

Issue metadata

Status: Fixed
Owner:
Last visit > 30 days ago
Closed: May 2018
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: Linux
Pri: 3
Type: Bug

Blocking:
issue 843511
issue 844392



Sign in to add a comment

blink_platform_unittests 's WebGLImageConversionTest.ConvertRGBA8toRGBA8 failing on ToTLinuxMSan

Project Member Reported by thakis@chromium.org, May 17 2018

Issue description

Started here:

https://ci.chromium.org/buildbot/chromium.clang/ToTLinuxMSan/2543

Because I added blink_platform_unittests to the bot's tests in https://chromium-review.googlesource.com/#/c/1060293/

However, it already runs on the chromium.memory main waterfall, and it passes there: https://ci.chromium.org/buildbot/chromium.memory/Linux%20MSan%20Tests/9859

The stack says "  ORIGIN: invalid (0). Might be a bug in MemorySanitizer origin tracking.", so maybe it's a regression on tot? Did anything in msan change since the last roll (which was just 31h ago, to r332335)?

kcc: ^




Full stack:

[ RUN      ] WebGLImageConversionTest.ConvertRGBA8toRGBA8
==5864==WARNING: MemorySanitizer: use-of-uninitialized-value
    #0 0x6a08457 in PackOneRowOfRGBA8LittleToRGBA8 third_party/blink/renderer/platform/graphics/cpu/x86/webgl_image_conversion_sse.h:182:30
    #1 0x6a08457 in Pack<0, 2, unsigned char, unsigned char> third_party/blink/renderer/platform/graphics/gpu/webgl_image_conversion.cc:892:0
    #2 0x6a08457 in blink::WebGLImageConversion::PackPixels(unsigned char const*, blink::WebGLImageConversion::DataFormat, unsigned int, unsigned char*) third_party/blink/renderer/platform/graphics/gpu/webgl_image_conversion.cc:3215:0
    #3 0x1453d82 in PackPixels third_party/blink/renderer/platform/graphics/gpu/webgl_image_conversion_test.cc:25:5
    #4 0x1453d82 in blink::WebGLImageConversionTest_ConvertRGBA8toRGBA8_Test::TestBody() third_party/blink/renderer/platform/graphics/gpu/webgl_image_conversion_test.cc:122:0
    #5 0x5a3a064 in testing::Test::Run() third_party/googletest/src/googletest/src/gtest.cc:0:0
    #6 0x5a3d953 in testing::TestInfo::Run() third_party/googletest/src/googletest/src/gtest.cc:2667:11
    #7 0x5a3f399 in testing::TestCase::Run() third_party/googletest/src/googletest/src/gtest.cc:2785:28
    #8 0x5a765d4 in testing::internal::UnitTestImpl::RunAllTests() third_party/googletest/src/googletest/src/gtest.cc:5047:43
    #9 0x5a74e87 in testing::UnitTest::Run() third_party/googletest/src/googletest/src/gtest.cc:0:0
    #10 0x8d937d0 in RUN_ALL_TESTS third_party/googletest/src/googletest/include/gtest/gtest.h:2329:46
    #11 0x8d937d0 in base::TestSuite::Run() base/test/test_suite.cc:275:0
    #12 0x2301806 in (anonymous namespace)::runTestSuite(base::TestSuite*) third_party/blink/renderer/platform/testing/run_all_tests.cc:44:27
    #13 0x8db6821 in Run base/callback.h:96:12
    #14 0x8db6821 in base::(anonymous namespace)::LaunchUnitTestsInternal(base::OnceCallback<int ()>, unsigned long, int, bool, base::OnceCallback<void ()>) base/test/launcher/unit_test_launcher.cc:225:0
    #15 0x8db5fce in base::LaunchUnitTests(int, char**, base::OnceCallback<int ()>) base/test/launcher/unit_test_launcher.cc:576:10
    #16 0x23015ce in main third_party/blink/renderer/platform/testing/run_all_tests.cc:62:14
    #17 0x7f4ad9520f44 in __libc_start_main /build/eglibc-ripdx6/eglibc-2.19/csu/libc-start.c:287:0
    #18 0x879029 in _start ??:0:0
SUMMARY: MemorySanitizer: use-of-uninitialized-value (/b/s/w/ir/out/Release/blink_platform_unittests+0x6a08457)
  ORIGIN: invalid (0). Might be a bug in MemorySanitizer origin tracking.
    This could still be a bug in your code, too!
Exiting
 

Comment 2 by r...@chromium.org, May 17 2018

Owner: euge...@chromium.org
https://reviews.llvm.org/D46785 / r332402 "[msan] Instrument masked.store, masked.load intrinsics." seems like a possible culprit.
Something did change:
http://llvm.org/viewvc/llvm-project?rev=332402&view=rev

It looks relevant. I'll take a look today or tomorrow.

Comment 4 by thakis@chromium.org, May 18 2018

Blocking: 844392
Status: Started (was: Untriaged)
It's not my change. I'm not sure which change it is exactly, but it did trigger a bug in MSan: https://bugs.llvm.org/show_bug.cgi?id=37523

It is pretty easy to fix.



Status: Fixed (was: Started)
This should do it:
http://llvm.org/viewvc/llvm-project?view=revision&revision=332761

Comment 7 by thakis@chromium.org, May 22 2018

Blocking: 843511

Sign in to add a comment