2008-12-20 06:41:08

by Junio C Hamano

[permalink] [raw]
Subject: [ANNOUNCE] GIT 1.6.0.6

The latest maintenance release GIT 1.6.0.6 is available at the
usual places:

http://www.kernel.org/pub/software/scm/git/

git-1.6.0.6.tar.{gz,bz2} (source tarball)
git-htmldocs-1.6.0.6.tar.{gz,bz2} (preformatted docs)
git-manpages-1.6.0.6.tar.{gz,bz2} (preformatted docs)

The RPM binary packages for a few architectures are also provided
as courtesy.

RPMS/$arch/git-*-1.6.0.6-1.fc9.$arch.rpm (RPM)

Among miscellaneous fixes, this contains a local gitweb security fix.
Maintenance releases for older versions (v1.5.4.7, v1.5.5.6 and v1.5.6.6)
are also available at the same place.

----------------------------------------------------------------

Changes since v1.6.0.5 are as follows:

David Aguilar (1):
git-mergetool: properly handle "git mergetool -- filename"

Deskin Miller (1):
git-svn: Make following parents atomic

Jim Meyering (1):
git-config.txt: fix a typo

Johannes Schindelin (2):
fast-import: close pack before unlinking it
fast-export: deal with tag objects that do not have a tagger

Junio C Hamano (6):
work around Python warnings from AsciiDoc
git-show: do not segfault when showing a bad tag
pager: do not dup2 stderr if it is already redirected
gitweb: do not run "git diff" that is Porcelain
GIT 1.5.4.7
fast-import: make tagger information optional

Linus Torvalds (1):
fsck: reduce stack footprint

Markus Heidelberg (1):
Documentation: fix typos, grammar, asciidoc syntax

Miklos Vajna (1):
SubmittingPatches: mention the usage of real name in Signed-off-by: lines

Nicolas Pitre (1):
make sure packs to be replaced are closed beforehand

Wu Fengguang (1):
git-send-email: handle email address with quoted comma


2008-12-20 10:54:40

by Jakub Narebski

[permalink] [raw]
Subject: Re: [Security] gitweb local privilege escalation (fix)

Junio C Hamano <[email protected]> writes:

> Current gitweb has a possible local privilege escalation bug that allows a
> malicious repository owner to run a command of his choice by specifying
> diff.external configuration variable in his repository and running a
> crafted gitweb query.
>
> Recent (post 1.4.3) gitweb itself never generates a link that would result
> in such a query, and the safest and cleanest fix to this issue is to
> simply drop the support for it. Maintenance release v1.6.0.6, v1.5.6.6,
> v1.5.5.6 and v1.5.4.7 are already available at k.org (see the announcement
> for v1.6.0.6 I sent out a few minutes ago), and the master branch and
> others pushed out tonight have the same fix.

>From what I have found diff.external works only since v1.5.4 (see
commit cbe02100), so when gitweb started using git-diff for old
legacy links to not use $tmpdir and /usr/bin/diff -u it wasn't an
issue...

--
Jakub Narebski
Poland
ShadeHawk on #git

2008-12-20 11:56:50

by Nanako Shiraishi

[permalink] [raw]
Subject: Re: [Security] gitweb local privilege escalation (fix)

Quoting Jakub Narebski <[email protected]>:

> Junio C Hamano <[email protected]> writes:
>
>> Current gitweb has a possible local privilege escalation bug that allows a
>> malicious repository owner to run a command of his choice by specifying
>> diff.external configuration variable in his repository and running a
>> crafted gitweb query.
>>
>> Recent (post 1.4.3) gitweb itself never generates a link that would result
>> in such a query, and the safest and cleanest fix to this issue is to
>> simply drop the support for it. Maintenance release v1.6.0.6, v1.5.6.6,
>> v1.5.5.6 and v1.5.4.7 are already available at k.org (see the announcement
>> for v1.6.0.6 I sent out a few minutes ago), and the master branch and
>> others pushed out tonight have the same fix.
>
> From what I have found diff.external works only since v1.5.4 (see
> commit cbe02100), so when gitweb started using git-diff for old
> legacy links to not use $tmpdir and /usr/bin/diff -u it wasn't an
> issue...

Sorry, but I do not understand what you are trying to say.

The maintenance releases Junio listed above go only as far back as v1.5.4.X series, so it covers the necessary range of versions, and no further back than that. Do you see a problem with that?

--
Nanako Shiraishi
http://ivory.ap.teacup.com/nanako3/