bpo-41513: Second attempt to add accuracy tests for math.hypot() #22327
Conversation
|
After 9 hours, the Travis CI seems to be wedged. Perhaps closing and reopening will trigger a new build. |
|
If you want to test Gentoo, you can add the buildbot label. |
|
🤖 New build scheduled with the buildbot fleet by @rhettinger for commit 5edc606 🤖 If you want to schedule another build, you need to add the ":hammer: test-with-buildbots" label again. |
|
🤖 New build scheduled with the buildbot fleet by @rhettinger for commit 9255115 🤖 If you want to schedule another build, you need to add the ":hammer: test-with-buildbots" label again. |
|
🤖 New build scheduled with the buildbot fleet by @rhettinger for commit 6da426d 🤖 If you want to schedule another build, you need to add the ":hammer: test-with-buildbots" label again. |
|
🤖 New build scheduled with the buildbot fleet by @rhettinger for commit 1cd8b13 🤖 If you want to schedule another build, you need to add the ":hammer: test-with-buildbots" label again. |
Formerly, these tests succeeded on all buildbots except the Gentoo AMD Athlon build.
Am attempting to get it to pass by adding the same guards used in testFsum(). This requires IEEE-754 and will skip the test in double rounding environments.
If this still fails, will revert and abandon the effort to add accuracy tests.
https://bugs.python.org/issue41513