From: Patrick Steinhardt <ps@pks.im>
Date: Fri, 02 Oct 2026 19:34:04 GMT
Subject: Re: [PATCH 2/2] packfile: fix corruption due to stale delta base cache entries
Message-ID: <asAHLJvsz0gpSB7j@pks.im>
In-Reply-To: <20261002133405.1284-1-guillaume.chauvel@gmail.com>

On Fri, Oct 02, 2026 at 03:34:05PM +0200, Guillaume Chauvel wrote:
> On Fri, Oct 02, 2026 at 09:34:07AM +0200, Patrick Steinhardt wrote:
> > Note that the added test reliably reproduces the above bug on my machine
> > that uses NixOS at c59305bab206 (cosmic-applets: add missing runtime
> > dependency (#566040), 2026-10-01) with glibc 2.44-25. But as we rely on
> > specific allocation behaviour of glibc it is very likely that the test
> > will not work on other platforms.
> 
> What about forcing the address reuse in the test for deterministic
> behavior ?
> A test helper can close a pack and move a second packed_git, whose
> delta base sits at the same offset, into its memory.
> 
> This was written with the help of AI tools.
> To apply on top of [PATCH 1/2].

You can do that, of course. But given the amount of code that we'd have
to add to test for a very specific edge case it doesn't quite feel
reasonable to me to add all this infrastructure.

Patrick