linux-kernel - Re: [PATCH 2/1] apply: reject input that touches outside $cwd

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20150130182456.GA29477@peff.net>
Date:	Fri, 30 Jan 2015 13:24:56 -0500
From:	Jeff King <peff@...f.net>
To:	Junio C Hamano <gitster@...ox.com>
Cc:	Git Mailing List <git@...r.kernel.org>,
	Josh Boyer <jwboyer@...oraproject.org>,
	"Linux-Kernel@...r. Kernel. Org" <linux-kernel@...r.kernel.org>,
	twaugh@...hat.com, Linus Torvalds <torvalds@...ux-foundation.org>
Subject: Re: [PATCH 2/1] apply: reject input that touches outside $cwd

On Thu, Jan 29, 2015 at 03:48:14PM -0800, Junio C Hamano wrote:

> By default, a patch that affects outside the working area is
> rejected as a mistake; Git itself never creates such a patch
> unless the user bends backwards and specifies nonstandard
> prefix to "git diff" and friends.
> 
> When `git apply` is used without either `--index` or `--cached`
> option as a "better GNU patch", the user can pass `--allow-uplevel`
> option to override this safety check.  This cannot be used to escape
> outside the working tree when using `--index` or `--cached` to apply
> the patch to the index.

It looks like your new --allow-uplevel goes to verify_path(). So this
isn't just about "..", but it will also protect against applying a patch
inside ".git". Which seems like a good thing to me, but I wonder if the
option name is a little misleading. It is really about applying the same
checks we do for index paths to the non-index mode of "git apply".

>  * Meant to apply on top of the previous one, but these two are
>    about separate and orthogonal issues.

I agree they are orthogonal in concept, though I doubt the symlink tests
here would pass without the previous one (since verify_path does not
know or care about crossing symlink boundaries).

-Peff
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/