git.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Jeff King <peff@peff.net>
To: Jonathan Nieder <jrnieder@gmail.com>
Cc: git@vger.kernel.org,
	"Linus Torvalds" <torvalds@linux-foundation.org>,
	"Ævar Arnfjörð Bjarmason" <avarab@gmail.com>
Subject: Re: [PATCH/RFC] fsck: complain when .gitignore and .gitattributes are symlinks
Date: Thu, 17 Jan 2019 12:00:06 -0500	[thread overview]
Message-ID: <20190117170005.GA27667@sigill.intra.peff.net> (raw)
In-Reply-To: <20190114230902.GG162110@google.com>

On Mon, Jan 14, 2019 at 03:09:02PM -0800, Jonathan Nieder wrote:

> From: Jeff King <peff@peff.net>
> Date: Sun, 13 May 2018 14:14:34 -0400
> 
> This case is already forbidden by verify_path(), so let's
> check it in fsck. It's easier to handle than .gitmodules,
> because we don't care about checking the blob content. This
> is really just about whether the name and mode for the tree
> entry are valid.

Hmm. I think this commit message isn't quite right, because we also
skipped the patches to touch gitignore/gitattributes in verify_path().

Are you thinking we should resurrect that behavior[1], too, or just
protect at the fsck level?

> It was omitted from that series because it does not address any known
> exploit, but to me it seems worthwhile anyway:
> 
> - if a client enables transfer.fsckObjects, this helps them protect
>   themselves against weird input that does *not* have a known exploit
>   attached, to
> 
> - it generally feels more simple and robust.  Git-related tools can
>   benefit from this kind of check as an indication of input they can
>   bail out on instead of trying to support.

I think I may just be restating your two points above, but what I'd
argue is:

  - even though there's no known-interesting exploit, this can cause Git
    to unexpectedly read arbitrary files outside of the repository
    directory. That in itself isn't necessarily evil, but it's weird.

  - there are potentially non-malicious bugs here, where we try to read
    .gitattributes out of the index, but obviously don't follow symlinks
    there

-Peff

[1] This wasn't a separate patch, but just an early iteration of the
    "ban symlinks in .gitmodules" patch. I think the incremental is
    just:

diff --git a/read-cache.c b/read-cache.c
index bfff271a3d..121c0bec69 100644
--- a/read-cache.c
+++ b/read-cache.c
@@ -937,7 +937,9 @@ static int verify_dotfile(const char *rest, unsigned mode)
 			return 0;
 		if (S_ISLNK(mode)) {
 			rest += 3;
-			if (skip_iprefix(rest, "modules", &rest) &&
+			if ((skip_iprefix(rest, "modules", &rest) ||
+			     skip_iprefix(rest, "ignore", &rest) ||
+			     skip_iprefix(rest, "attributes", &rest)) &&
 			    (*rest == '\0' || is_dir_sep(*rest)))
 				return 0;
 		}
@@ -966,7 +968,9 @@ int verify_path(const char *path, unsigned mode)
 				if (is_hfs_dotgit(path))
 					return 0;
 				if (S_ISLNK(mode)) {
-					if (is_hfs_dotgitmodules(path))
+					if (is_hfs_dotgitmodules(path) ||
+					    is_hfs_dotgitignore(path) ||
+					    is_hfs_dotgitattributes(path))
 						return 0;
 				}
 			}
@@ -974,7 +978,9 @@ int verify_path(const char *path, unsigned mode)
 				if (is_ntfs_dotgit(path))
 					return 0;
 				if (S_ISLNK(mode)) {
-					if (is_ntfs_dotgitmodules(path))
+					if (is_ntfs_dotgitmodules(path) ||
+					    is_ntfs_dotgitignore(path) ||
+					    is_ntfs_dotgitattributes(path))
 						return 0;
 				}
 			}

  reply	other threads:[~2019-01-17 17:00 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-01-14 23:09 [PATCH/RFC] fsck: complain when .gitignore and .gitattributes are symlinks Jonathan Nieder
2019-01-17 17:00 ` Jeff King [this message]
2019-01-17 20:13   ` Junio C Hamano
2019-01-17 21:24     ` Jeff King
2019-01-18  1:41       ` Ramsay Jones
2019-01-22  7:23         ` Jeff King
2019-01-22 18:19           ` Ramsay Jones

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20190117170005.GA27667@sigill.intra.peff.net \
    --to=peff@peff.net \
    --cc=avarab@gmail.com \
    --cc=git@vger.kernel.org \
    --cc=jrnieder@gmail.com \
    --cc=torvalds@linux-foundation.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).