-
Notifications
You must be signed in to change notification settings - Fork 280
Closed
Description
-- xsimd v11.0.0
-- Tests build type is Release
-- Configuring done
-- Generating done
-- Build files have been written to: /usr/ports/devel/xsimd/work/.build
ninja: no work to do.
[ 0% 1/1] cd /usr/ports/devel/xsimd/work/.build && /usr/local/bin/ctest --force-new-ctest-process
Test project /usr/ports/devel/xsimd/work/.build
Start 1: test_xsimd
1/1 Test #1: test_xsimd .......................***Failed 0.52 sec
[doctest] doctest version is "2.4.11"
[doctest] run with "--help" for options
===============================================================================
/usr/ports/devel/xsimd/work/xsimd-11.0.0/test/test_xsimd_api.cpp:695:
TEST CASE: [xsimd api | float types functions]<float>
exp10
/usr/ports/devel/xsimd/work/xsimd-11.0.0/test/test_xsimd_api.cpp:466: ERROR: CHECK_EQ( extract(xsimd::exp10(T(val))), std::pow(value_type(10), val) ) is NOT correct!
values: CHECK_EQ( 100, 100 )
===============================================================================
/usr/ports/devel/xsimd/work/xsimd-11.0.0/test/test_xsimd_api.cpp:695:
TEST CASE: [xsimd api | float types functions]<double>
exp10
/usr/ports/devel/xsimd/work/xsimd-11.0.0/test/test_xsimd_api.cpp:466: ERROR: CHECK_EQ( extract(xsimd::exp10(T(val))), std::pow(value_type(10), val) ) is NOT correct!
values: CHECK_EQ( 100, 100 )
===============================================================================
[doctest] test cases: 303 | 301 passed | 2 failed | 0 skipped
[doctest] assertions: 7929 | 7927 passed | 2 failed |
[doctest] Status: FAILURE!
0% tests passed, 1 tests failed out of 1
Total Test time (real) = 0.52 sec
The following tests FAILED:
1 - test_xsimd (Failed)
Errors while running CTest
FAILED: CMakeFiles/test.util
Verison: 11.0.0
clang-14
FreeBSD 13.1
Metadata
Metadata
Assignees
Labels
No labels
Type
Projects
Milestone
Relationships
Development
Select code repository
Activity
serge-sans-paille commentedon Apr 11, 2023
Thanks for reporting. Can you tell us if #916 fixed the issue?
serge-sans-paille commentedon Apr 25, 2023
@yurivict do you still encounter the issue with latest master branch?
doronbehar commentedon Jul 31, 2023
Here on NixOS, we try to build 11.1.0 on MacOS, which includes #918, and it didn't fix the issue for us. See progress at NixOS/nixpkgs#246200
yurivict commentedon Nov 9, 2023
11.2.0 still has this problem.
yurivict commentedon Feb 16, 2024
12.1.1 still has the problem:
yurivict commentedon May 3, 2024
Tests are still broken in 13.0.0
serge-sans-paille commentedon May 3, 2024
@yurivict which architecture?
yurivict commentedon May 3, 2024
amd64
serge-sans-paille commentedon May 4, 2024
and which os / compiler / libc ? Because we (obviously!) test on these arch...
doronbehar commentedon May 5, 2024
Here is a list of the tests and patches we apply to each OS / Architecture etc on NixOS:
https://github.com/NixOS/nixpkgs/blob/67939e1d931ecee79802d9038e7ebef096d96f53/pkgs/development/libraries/xsimd/default.nix#L17-L37
Loosen exp10 tests accuracy constraints
serge-sans-paille commentedon May 9, 2024
@yurivict does #917 do the the trick?
serge-sans-paille commentedon May 9, 2024
and thanks @doronbehar , that did help
doronbehar commentedon May 9, 2024
You meant #1021 I suppose :).
serge-sans-paille commentedon May 9, 2024
I do :-)
serge-sans-paille commentedon May 17, 2024
@doronbehar so, does it fix the bug?
doronbehar commentedon May 18, 2024
Only now I had time to delve in that. :) If I understand correctly, #1021 is supposed to maybe help us only with two patches we currently apply:
I also noticed many projects in the
@xtensor-stack
had version updates, so I currently test the updates with our CI in NixOS/nixpkgs#312745 . It will take time as this our CI is especially needed for Darwin tests, and for those platforms it is rather slow.doronbehar commentedon May 21, 2024
Ok so putting asside now #1024 (which (I hope) fixes #804), we didn't experience the failures reported here on any platform - including aarch64-darwin, for which it took our CI a few days to reach.
serge-sans-paille commentedon May 30, 2024
So I think we can close this issue, feel free to reopen if CI proves me wrong.