Monday Patch Queue Review update (2024-02-12)

Monday Patch Queue Review update (2024-02-12)

[Carlos O'Donell]

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

Meeting: 2024-02-12 @ 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 is at 497 patches.
 * Carlos's SLI at 258 days average patch age in queue and 129620 accumulated patch days.
 * Patch review volunteers:
     https://inbox.sourceware.org/libc-alpha/F0A66BDC-9644-445F-9E30-2C1B90C2277C@linaro.org/
 * Carlos: Still working on my post the community. I saw with it for the week and I have concluded on a first come first serve design where we try to schedule out blocking work as feature delivery into the release.
 * Starting at 85595.
 * Patch reviewer pool: Adhemerval (azanella), Arjun (arjun), Carlos (codonell) , DJ(dj), Maxim (maximk), Simon.
 * 85594: [v3] x86-64: Update _dl_tlsdesc_dynamic to preserve vector registers - fweimer to review.
 * x86-64: Save APX registers in ld.so trampoline - adhemerval to review.
 * localedata: Set width of DEFAULT_IGNORABLE_CODE_POINTs to 0 [BZ #31370] - arju to review.
 * Enhanced x86 CPU diagnostics - hj to review.
 * Improve fortify support with clang - codonell to review.
 * [v2] Add public function syscall_no_errno
  * Initial version was CC'd to Rich Felker for musl comments.
  * Adhemerval to review since he started previous review.
 * 85461: x86: Improve ERMS usage on Zen3+ -- HJ to ping Noah on the review.
 * Stopped at 85344.
 * Florian: I will post to the list to start a discussion around eager allocation of TLS space.
 * Carlos: dlopen and TLS descriptors?
 * Florian: Yes.
 * HJ: ABI is clear we need to preserve all registers.
 * HJ: It is an implementation detail if we get rid of the slow path or keep it and preserve the registers.
 * Carlos: Why did we make this decision initially?
 * Adhemerval: Large TLS per thread, and you install multiple threads, then this can consume lots of memory.
 * Andreas: The clone3 fix for mips? Can we backport that? Only relevant for 2.38.
 * Adhemerval: I'll check the backport if I haven't done it already.
 * Carlos: What do we want to do for a patch review pool process? Do we stop once we hit the end of the pool or do we try to assign more to the pool?
 * [v2] tests: gracefully handle AppArmor userns containment - assigned to Maxim.
 * x86/cet: fix shadow stack test scripts - HJ to push
 * Carlos: Please remember to add Reviewed-by: to track review.
 * [v5] sysdeps/x86_64/multiarch/memmem-avx2.c: add memmem-avx2.c - Fails to apply. Fails CI.
 * Linux: Switch back to assembly syscall wrapper for prctl (bug 29770) - Asking Simon to review. Assigned to Maxim and Maxim will follow up with Simon to get a patchwork account.
 * Stopping at 85081.

-- 
Cheers,
Carlos.