Re: Regression tests fail with musl libc because libpq.so can't be loaded - Mailing list pgsql-bugs

From Andres Freund
Subject Re: Regression tests fail with musl libc because libpq.so can't be loaded
Date
Msg-id 20240322001948.atp2eetusylgjyw2@awork3.anarazel.de
Whole thread Raw
In response to Re: Regression tests fail with musl libc because libpq.so can't be loaded  (Wolfgang Walther <walther@technowledgy.de>)
Responses Re: Regression tests fail with musl libc because libpq.so can't be loaded  (Andrew Dunstan <andrew@dunslane.net>)
Re: Regression tests fail with musl libc because libpq.so can't be loaded  (Wolfgang Walther <walther@technowledgy.de>)
List pgsql-bugs
Hi,

On 2024-03-21 21:16:46 +0100, Wolfgang Walther wrote:
> Right. The latter not only confuses musl, but also makes /proc/<pid>/environ
> return garbage. This is also mentioned at the bottom of main.c, which has a
> workaround for the specific case of UBSan depending on that. This is kind of
> funny: Because we are relying on undefined behavior regarding the
> modification of environ, we need a workaround for the
> "UndefinedBehaviorSanitizer" - I guess by failing without this workaround,
> it wanted to tell us something..

I don't think that's quite a fair description. Ubsan is basically doing
undefined things itself, so it's turtles all the way down.


> So summarizing:

FWIW, independent of which fix we go with, I think we need a buildfarm animal
using musl. Even better if one of the CI tasks can be made to use musl as
well.

Greetings,

Andres Freund



pgsql-bugs by date:

Previous
From: Bruce Momjian
Date:
Subject: Re: Regression tests fail with musl libc because libpq.so can't be loaded
Next
From: Tomas Vondra
Date:
Subject: Re: Index plan returns different results to sequential scan