Monday Patch Queue Review update (2023-10-30)

Monday Patch Queue Review update (2023-10-30)

[Carlos O'Donell]

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

Meeting: 2023-10-30 @ 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 454 patches.
 * Carlos's SLI at 212.53813559322035 days average patch age in queue and 100318 accumulated patch days.
 * Starting at 78663.
 * Started by reviewing the oldest open patches in the queue.
 * Carlos: I could create a new project here https://gitlab.com/gnutools and we could add the scripts there.
 * Siddhesh: Works for me.
 * Carlos: I have the AI to create a patchwork scripts repo called "gnutools-bots"
 * Maxim: A bot that enforces policy should have a way for people to contribute.
 * Siddesh: Should I pull the trigger on sending emails for commits that don't match?
 * Carlos: Yes, start by send email to us, and we'll review the results, and then turn it live.
 * Carlos: Once we have a repo we can ask for help to automate re-applying the patches.
 * Carlos : AI split DJ's apply-patch bot into two, one for head and one for applied at the time of sent/base-rev.
 * libio: Remove comment in <stdio.h> (Johnny) - Changes requested.
 * Add a tunable to decorate anonymous memory maps (Adhemerval)
  * Ian Rogers from Google provided the initial RFC.
  * Adhemerval: I expanded the idea the annotate the pthread stack.
  * Performance implications: To set the name requires setting the mm lock.
  * It can show on specific workloads, so it's behind a tunable, off by default for now.
 * test: Run the tst-tls-allocation-failure-static-patched with test-wrapper. (Yanzhang)
  * Carlos: I saw the test failing in December for x86_64 cross to AArch64.
  * Carlos: Can we pre-commit CI test cross-test-ssh.sh?
  * Maxim: We do not use cross-test-ssh.sh currently. From the technical point of view you want a container on the host to drive testing, and a second container on the target that will run the tests.
  * Carlos: Yes, the OpSec for this case is very difficult for pre-commit CI.
  * Maxim: I raised that I would like the applypatch bot to run for all GNU Toolchain projects.
  * Carlos: I'll talk to DJ about doing that.
 * Stopped at 78463.

-- 
Cheers,
Carlos.

Re: Monday Patch Queue Review update (2023-10-30)

[DJ Delorie]

"Carlos O'Donell" <carlos@redhat.com> writes:
>  * Carlos : AI split DJ's apply-patch bot into two, one for head and
>  one for applied at the time of sent/base-rev.

I was thinking of hacking up something to rescan the queue every weekend
and maintain a "still applies" flag on each, just in time for the review
meeting; is this what you were thinking of?