From: Jeff King <peff@peff.net>
Date: Wed, 30 Sep 2026 20:53:11 GMT
Subject: Re: [PATCH 4/4] repack: retain cruft packs in MIDXs containing kept packs
Message-ID: <20260930205311.GC747209@coredump.intra.peff.net>
In-Reply-To: <e942c256334e4de31ec0a1cb2d5f8c7465d8696f.1790731662.git.me@ttaylorr.com>

On Tue, Sep 29, 2026 at 08:28:58PM -0500, Taylor Blau wrote:

> When performing a geometric repack with 'repack.midxMustContainCruft'
> set to "false", Git uses '--stdin-packs=follow' to copy (once-cruft)
> objects needed for reachability closure out of cruft packs. .keep packs
> do not need to participate in that walk, though they *are* included in
> the resulting MIDX.
> 
> A .keep pack can contain a commit that reaches an object whose only copy
> is in a cruft pack. When there is no previous MIDX and the repack writes
> a new pack, neither `midx_has_unknown_packs()` nor the `!names.nr`
> fallback require that cruft pack to be included. If the kept commit (or
> a descendant of it) is selected for bitmap coverage, the bitmap writer
> fails because the MIDX does not contain all of its reachable objects.
> 
> Include cruft packs whenever the MIDX contains kept packs. This also
> retains cruft when the kept packs happen to have full closure, or when
> '--pack-kept-objects' lets the repack walk them. It avoids having to
> establish their closure before deciding which packs the MIDX needs.

OK. This makes sense to me, but two questions:

  1. Is this going to kick in racily because of the .keep that we
     temporarily install during pushes? That could cause unexpected
     performance changes in a big repo when the midx sometimes has to
     randomly include cruft packs.

  2. I'd have thought that the solution would be to treat .keep packs
     like other included follow-packs: traverse them in the usual way.
     But maybe there are good reasons we didn't do that in the first
     place.

-Peff