Re: [PATCH 6/6] pack-objects: reuse on-disk deltas for thin "have" objects

Jeff King [off-list ref] writes:

Ah, yeah, I think you're right. We call find_patch_start(), which thinks
the "---" line is the end of the commit message. That makes sense when
parsing trailers out of "format-patch" output, but not when we know we
have just the commit message.

Yes, but that does not explain what we are seeing.  If the code
mistakenly thinks that the log message ends before that table, then
it should have inserted the S-o-b: _before_ that table, but that is
not happening.

So there are three issues; (1) find-patch-start uses too weak a
logic to find the beginning of a patch section (2) even if it found
the right place, its caller does not tell "commit --amend -s" where
the log message ends correctly and (3) some callchains that get
there know they only have a log message but there is no way to take
advantage of that information and skip the call to find-patch-start.