-
Notifications
You must be signed in to change notification settings - Fork 41
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
cargo test segfaulting on i386 #155
Comments
Thanks for the report. I’ll see if I can reproduce. |
@plugwash Can you provide more information about the environments you're running this on? I'm not sure I have easy access to an i386 or armhf environment. Was there a particular container or VM image that you ran these tests on? |
Could you try a build again from the
Or if you have already cloned then run
It would be interesting if |
I suspect this has something to do with either our bindings to On a 32 bit Ubuntu VM I'm seeing the following when trying to compile:
Doesn't sound good. |
I should point out I am able to repro this crash.
Running under GDB it seems the issue is somewhere in the Oniguruma "string table" code
From the stack trace it looks like the
This also appears to still be an issue on the |
Provisional fix for this in #156. If you could check that out and report back it would be much appreciated. |
Testing with a clone of https://github.com/rust-onig and fetching the submodule failed as expected with the master branch and passed with the 32-bit-segfaults branch. I've backported the commit to the Debian package and am now running tests in that environment. The Debian autopkgtest seems to take a lot longer than plain cargo test and produce crazy ammounts of output, so i'm still waiting for it to finish, it's got way past the point where it failed before though. |
Nice. If it all looks good I'll pull together a new point release of |
I have uploaded the patched rust-onig-sys to debian and tests for rust-onig 5.0.0 are now passing on amd64, i386 and armhf. ppc64el is still waiting for test results (but if that does fail it will presumably be a totally separate bug) . arm64 does not seem to be getting tested for some reason. I probably won't update rust-onig and it's reverse dependencies rust-sytect and rust-bat to the new major versions myself, i'll leave the decision on if and when to do that up to others. |
Debianhas recently upgraded rust-bindgen to 0.55.1 in unstable. To keep dependencies satisfiable the update of bindgen required updating rust-onig-sys as well and it was upgraded to 69.5.1.
However, after the update of bindgen onig-sys it was discovered that the tests for onig started failing on i386 and armhf and this is currently blocking the testing migration of the new rust-bindgen package. On i386 the tests fail with a segfault and on armhf they faile with a bus error. I performed further testing on i386.
First I tried updating to the latest version of the onig crate and the error persisted.
I then tried doing a test outside of the "debcargo" environment, I first tried using
an upstream git checkout but that failed to find it's embedded copy of libonig
so I decided to use cargo clone instead.
cargo install cargo clone
cargo clone onig
cd onig
cargo test
And again that failed in the same way
The text was updated successfully, but these errors were encountered: