On 2023-09-04 04:52, Andrew Gregory wrote:
On 08/25/23 at 07:36pm, Dragan Simic wrote:
On 2023-08-21 18:32, Dragan Simic wrote:
> On 2023-08-21 18:19, Andrew Gregory wrote:
...
I am trying to reproduce this to wrap my head around what's actually
happening and verify that this actually fixes it, but I am unable
to get pacman to execute a hook with fd 0 closed with or without
the ARM patches. Can you provide a reproducible case?
Ah, I know very well how it feels, I spent at least a couple of days
hunting the issue down and fixing it.
One of the issues with reproducing the error is that you must have an
old systemd version installed, which still contains assert(fd) checks
in src/tmpfiles/tmpfiles.c that actually trigger the error. IIRC,
systemd removed those checks about a couple of years ago, presumably
to fix similar complaints.
You may also want to see my original report of the issue, which also
contains the fix:
https://gitlab.manjaro.org/manjaro-arm/packages/core/pacman/-/issues/1
Just checking, have you managed to reproduce the issue? If not,
please let
me know, and I'll put together some kind of instructions for creating
an
appropriate environment and a test case that reproduces the error.
I still haven't been able to reproduce pacman running a hook with fd 0
closed.
I've put together a hook/script to test the hook execution environment:
https://github.com/andrewgregory/pachooks/commit/b91e5de6f203b9be9fbf2f4fec54d9301a40ec1f?diff=unified
I apologize for my delayed response. I've had some health issues,
unfortunately, and I still haven't recovered fully.
Thank you very much for putting together this test hook for pacman,
which is really neat. I'll get back to you with the further steps to
reproduce the issue as soon as possible.
