Skip to content
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

Fix #574 - Revert "Improve compatibility of binfmt-bypass with really old Docker images" #662

Open
wants to merge 1 commit into
base: master
Choose a base branch
from

Conversation

trinitronx
Copy link

@trinitronx trinitronx commented Sep 15, 2024

This reverts commit c810e70.

Notes:

…with really old Docker images"

This reverts commit c810e70.

Notes:

- [CentOS 7 reached End of Life][1] on June 30, 2024
- [RedHat moved CentOS Stream 8 & 9 upstream of RHEL][2]
- Arch Linux & Manjaro unstable channel have GlibC with symbol version `2.14`:

    U memcpy@GLIBC_2.14

[1]: https://www.redhat.com/en/topics/linux/centos-linux-eol
[2]: https://www.theregister.com/2023/06/23/red_hat_centos_move/
@TheAssassin
Copy link
Owner

Looks okay to me. Is there an open issue which this PR solves?

@TheAssassin
Copy link
Owner

Regarding new builds, I'm still working on the CI bit, making some progress there. But I suppose since this fixes some Arch issue, AIL would be built by AUR anyway.

@TheAssassin TheAssassin mentioned this pull request Sep 19, 2024
2 tasks
@trinitronx
Copy link
Author

trinitronx commented Oct 11, 2024

Looks okay to me. Is there an open issue which this PR solves?

Yes, this should fix #574 which was a failed to build from source (FTBFS) issue on Arch Linux. (Downstream issue mentioned on AUR here)

It looks like the problem was caused by the patchelf script here and differing glibc symbol versions. I gave a bit more explanation here based on what I found.

The gist of it was that the problem was introduced in c810e70, which was to support some old CentOS 7 versions in docker (according to comments in the code & commit message). CentOS 7 is now EoL, so this hack can probably be removed, as it's causing these build issues on distros with newer GlibC.

The other thing that appears to have been fixed is #558 1. After rebuilding the AUR package with this PR's changes, appimagelauncherd no longer segfaults and creates lots of coredumps with SystemD restarting it (and using lots of disk space for coredump files). Meanwhile, the current version of extra/appimagelauncher 2.2.0-10 does still segfault in this way, causing the coredump file buildup.

Footnotes

  1. I have been running my packaged version built from this PR since around September 15th, and haven't seen the segfault issue since then. I suspect that the patchelf hack had opened up the door to some undefined behavior due to potential for GlibC linkage / ABI issues.

@TheAssassin
Copy link
Owner

Since packaging is under rework (making progress, actually), I'd like to postpone this PR for now and reconsider it when the new build system has been merged.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants