search

llvm / llvm-project

The LLVM Project is a collection of modular and reusable compiler and toolchain technologies.
http://llvm.org
Other
28.75k stars 11.89k forks source link

JSONTest.Integers is flaky #45815

Open nico opened 4 years ago

nico commented 4 years ago
Bugzilla Link 46470
Version trunk
OS All
CC @sam-mccall

Extended Description

Just failed on a Windows bot once here: https://logs.chromium.org/logs/chromium/buildbucket/cr-buildbucket.appspot.com/8876405792703724976/+/steps/package_clang/0/stdout

nico commented 4 years ago

disabled the test in https://github.com/llvm/llvm-project/commit/fb1aa286c1400ad75cd066df6c0233332b3fd9b8

nico commented 4 years ago

For posterity, pasting the failure from comment 0:

FAIL: LLVM-Unit :: Support/./SupportTests.exe/JSONTest.Integers (3808 of 58774) **** TEST 'LLVM-Unit :: Support/./SupportTests.exe/JSONTest.Integers' FAILED **** Note: Google Test filter = JSONTest.Integers [==========] Running 1 test from 1 test case. [----------] Global test environment set-up. [----------] 1 test from JSONTest [ RUN ] JSONTest.Integers C:\b\s\w\ir\cache\builder\src\third_party\llvm\llvm\unittests\Support\JSONTest.cpp(344): error: Expected: T.Str Which is: "6.9175290276410819e+18" To be equal to: s(T.Val) Which is: "6.9175290276410818e+18" Dynamically exact integer. Stored as double, convertible. C:\b\s\w\ir\cache\builder\src\third_party\llvm\llvm\unittests\Support\JSONTest.cpp(350): error: Expected: T.Str Which is: "6.9175290276410819e+18" To be equal to: s(*Doc) Which is: "6.9175290276410818e+18" Dynamically exact integer. Stored as double, convertible. [ FAILED ] JSONTest.Integers (1 ms) [----------] 1 test from JSONTest (1 ms total)

[----------] Global test environment tear-down [==========] 1 test from 1 test case ran. (1 ms total) [ PASSED ] 0 tests. [ FAILED ] 1 test, listed below: [ FAILED ] JSONTest.Integers

1 FAILED TEST

sam-mccall commented 4 years ago

Thanks for flagging this. I agree we should ifdef out this test on the bad version of MSVC.

(I think the test is worth keeping even though this particular failure mode isn't really worth writing about)

If you don't mind doing this I'd appreciate it, as I'm not familiar with MSVC version detection and don't have a good way to verify whether it's working.

nico commented 4 years ago

https://bugs.chromium.org/p/chromium/issues/detail?id=1099867#c4 has a good analysis.

nico commented 4 years ago

Actually, it's not flaky, it fails consistently after we updated our MSVC version.