Hey all,
I recently ran into interesting behavior with git-config (which I
originally thought was a bug).
I was converting repos from the deprecated core.fsyncObjectFiles to the new
core.fsync option suite; I wrote some automation to do that, using
"git config -l" to detect previously-converted repos.
But then some weekly repo maintenance automation complained about the
deprecated fsync option being present in some of the repo configs, even
though I thought those repos had been converted to the new settings.
I did some digging, and it turns out that "git config -l" was reporting
nothing (no output) in Git 2.37.4. I did some testing, and found that Git
2.35.1 correctly reporting the repo's config settings.
Interestingly, the maintenance automation runs fsck and some other things,
and reports the presence of the deprecated fsync setting (which is what
made me notice); so that code path does read the config and run (and
complain about the presence of deprecated settings).
I did a git bisect between 2.35.1 and 2.37.4, and it looks like the
following commit changes the behavior:
8959555cee7ec045958f9b6dd62e541affb7e7d9 is the first bad commit
commit 8959555cee7ec045958f9b6dd62e541affb7e7d9
Author: Johannes Schindelin <johannes.schindelin@xxxxxx>
Date: Wed Mar 2 12:23:04 2022 +0100
setup_git_directory(): add an owner check for the top-level directory
It poses a security risk to search for a git directory outside of the
directories owned by the current user.
[full commit message clipped]
So... my maintenance automation runs as root, and the repo directories are
uid/gid'd someone else (though, the config file inside the [bare] repo
happens to be owned by root)... so I suppose what I'm observing is expected
behavior?
I guess this leaves me with two questions:
1. Why does "git config" refuse to run due to this security check, but
other git commands ("git fsck," at least) run?
2. I think it might be useful to warn the user that the behavior they're
expecting isn't happening due to this security check, instead of just
outputting objectively wrong information (i.e. that no config options
exist when they actually do exist); I'd be curious what others think.
best,
preed
--
J. Paul Reed
https://jpaulreed.com
PGP: 0x41AA0EF1
