-
-
Notifications
You must be signed in to change notification settings - Fork 2.8k
MacOS test failures on some systems #467
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Comments
fixed with 6e5edc7 there is still an issue if we run all the tests at once, but I can't duplicate it on my macbook, and so I'm calling that a separate issue that 0.1.0 does not depend on. |
@snocl I noticed you are using MacOS. Do all tests pass for you when run like this?
After N child processes execute on Travis, it crashes (which is why I used the above workaround), but I can't reproduce the issue on my mac. |
No, after a bunch of successes I get:
|
I don’t know if this is related, but trying to run the |
I don't expect you to do this, but if you wanted to help troubleshoot this further: You can add Next we need to figure out how to run that command with lldb or gdb or some other way to get a backtrace and figure out what exactly happened. I have a mac running (I think - I'll check when I get home) 10.12 but I'm not an expert on how it works, is there any more version number or system information that we can compare to see what's different about your environment and mine? |
Running it again with
Running
and exits with error code 255. I can’t think of any other version numbers that might be relevant of the top of my head... |
Interestingly, running |
Interesting. I think this may be a separate issue. I think we can run that last command printed with a debugger and see what happened. That "malformed macho file" message is surprising. I don't know how the binary was able to run at all if it was malformed. |
I suspect this may be an LLD issue. Eventually we're going to have to improve the macho LLD code since the llvm team has been neglecting it |
Seems like it only fails if the build file is given as an absolute path:
EDIT: I think the |
That's really strange. The same test passes on travis as well as my mac. |
I will look into it later, but have to leave right now. |
I appreciate your help. For what it's worth, I think this is a bit of an edge case, and in my mind a relatively lower priority issue. If you wanted to play around with zig I wouldn't worry about this particular issue right now. |
Confirmed Heisenbug: launching it from |
I believe this is fixed now. Travis CI all tests passing. |
Uh oh!
There was an error while loading. Please reload this page.
The first commit that broke it was 9fb4d1f
The next few commits fixed it on my MacBook Air but did not fix it for Travis CI.
The text was updated successfully, but these errors were encountered: