-
Notifications
You must be signed in to change notification settings - Fork 299
Broken CI runner with upgrade to 0.6.5 #132
Comments
(I am referring to #130 ) |
Hi @junderw. Thanks for reporting this. Do you have a link to your ci failure? If not, can you provide more info about the environment it's running on and what the test is doing? Thanks |
Sorry... it is a private repository, so no link. Additional info we found: it was only crashing on So perhaps the code used in 0.6.5 is relying on glibc? (googling the issue, it seems like random segfaults on alpine are usually caused by glibc not being there) |
Yes, that's also my suspicion. This was present in Can you confirm that you were using the latest versions of these libraries? Maybe you had an older For more info: |
@junderw any update on this? |
I'll go ahead and close this... it was probably a lockfile keeping a dependency from updating... but the funny thing was that it was working fine up until I updated a dependency that depended on this library... Perhaps another dependency had one of those deps and the highest common version requirement between this library and the other library was a version with the issue not fixed yet... Either way, we decided to switch to debian based image instead of alpine. |
We recently started updating an internal library that depends on
^0.6.3
of this library, and our CI was throwing segfaults.I tried changing it to
0.6.4
and it worked, then0.6.5
threw the same segfault at the same part. (It's our internal test for a signing helper function for transactions)I have temporarily fixed this issue by changing from
^0.6.3
to a specific version, but just thought I'd let you know about this issue. Reverting the change and publishing a 0.6.6 version would be appreciated. (This change should be put off until a major version bump)Thank you.
The text was updated successfully, but these errors were encountered: