Re: MSan failures in pack-bitmap

From: Taylor Blau <me@ttaylorr.com>
Date: Jun 9, 2024, 20:30 UTC
Message-ID: <ZmYQ4f+qmQIxf+5g@nand.local>
In-Reply-To: <xmqqh6e1c1bf.fsf@gitster.g>

On Sun, Jun 09, 2024 at 01:23:48PM -0700, Junio C Hamano wrote:

Show 10 quoted lines

> Taylor Blau <me@ttaylorr.com> writes:
>
> > I don't think the issue here is a corrupt on-disk bitmap or MIDX. The
> > regression that Kyle reported was a logic issue reading a
> > ordinary/non-corrupt MIDX file, not a result of some on-disk corruption.
> > ...
> > But I do not think that repository corruption is a likely outcome here,
>
> Yes, exactly.  What I was asking was how on-disk corruption that
> would result from the logic issue of reading a correct MIDX file.

OK, I think I misunderstood your original email.

> ... there is no reasonable work around here, is there?  We do not
> even know exactly what kind of breakage this will cause (which is
> expected, as we would be "reading" a garbage value and basing future
> actions on it)?

I think the end-result would be that packs sent between client and server during fetches or pushes would be broken, resulting in index-pack failing to index them, thus causing one side or the other to (correctly) throw away the received pack.

The only thing I can think of to avoid this would be to enable multi-pack reuse, which is only possible if your MIDX has a BTMP chunk. The other work around would be to disable pack-reuse altogether, which is very unfortunate.

Thanks, Taylor