Monday Patch Queue Review update (2025-02-10)

[Carlos O'Donell]

Most recent meeting status is always here:
https://sourceware.org/glibc/wiki/PatchworkReviewMeetings#Update

Meeting: 2025-02-10 @ 0900h EST5EDT

Video/Audio: https://bbb.linuxfoundation.org/room/adm-alk-1uu-7fu

IRC: #glibc on OFTC.

Review new patches and restart review at the top.

​ * State NEW delegate NOBODY at 539 patches.
 * Started discussion around executable stack and if we would do anything different?
  * codonell: Do we have patches to review to change this?
  * adhemerval: There is some discussion on the bugzilla. Should we fix this? What is the proper way to fix this.
  * Note: https://github.com/conda-forge/pytorch-cpu-feedstock/issues/350#issue-2840876825
  * adhemerval: If the DSO and executable doesn't require executable stack, then it can be worked around, otherwise you have to use GCC14's alternate support.
  * Note: https://sourceware.org/bugzilla/show_bug.cgi?id=32653
  * fweimer: Avoid bringing back the old code.
  * codonell: I've reopened bug 32653 to review the options.
  * adhemerval: I'll post an option using the personality flag.
 * codonell: Anything remaining for 2.41?
 * andreas: Manual update on the website.
 * codonell: I've started an update of the 2.41 documentation.
 * Starting at 106250
 * Add tests for Guarded Control Stack (Yury)
  * yury: Adds tests. Adds compiler tests.
  * Adhemerval to review given review for v1-v3.
 * mig_strncpy: ensure destination string is null terminated (Flavio)
  * codonell: This needs Hurd maintainer review. Skipping for now.
 * htl: move pthread_barrierattr* and pthread_barrier* into libc (gfleury)
  * Already applied.
 * [2/2] linux: use __getrandom_nocancel in getentropy (Christian)
  * codonell: Note, only 2/2 was posted.
  * Note: https://pubs.opengroup.org/onlinepubs/9799919799/functions/getentropy.html
  * Reviewer Pool: Arjun Shankar
   * [v2,1/4] Makeconfig: Add libgcc directory to rtld-prefix search path (Andreas)
    * Andreas: First patch is changed. Initial review.
    * Reviewer Pool: Adhemerval
 * help debugging flaky posix/tst-waitid.c test (Aurelien)
  * codonell: I'll comment.
 * linux: check mremap w/ MREMAP_DONTUNMAP correctly (Liu)
  * Reviewer Pool: DJ
 * Stopping at 106133.
 * fweimer: Link map hardening patch review would be good.
  * https://patchwork.sourceware.org/project/glibc/list/?series=43633
  * adhemerval: We cannot use memory protections on most architectures that support it?
   * fweimer: Memory protection keys is a performance enhancement only.
   * fweimer: The semantics are incorrect on aarch64, and x86_64.
 * cupertino: I would like to discuss what I'm doing with malloc and get feedback.
  * https://patchwork.sourceware.org/project/glibc/list/?series=42499 (v3)
  * We know the leak is not within the application.
  * Try to address things that could benefit from enhancements.
  * Review of the strategy would be valuable.
  * Reviewer Pool: Florian Weimer
 * yury: Is there a next version of the memory sealing patch?
  * adhemerval: Discussions ongoing with criu developers. I don't have an answer yet.
  * We needed to give FF a way to disable this if they were applying a loader hack.
  * We can make this per-binary e.g. binary sets it, and sets it for all dependents.
  * fweimer: We would do it more like execstack handling.
  * adhemerval: That's how it's done.

-- 
Cheers,
Carlos.