Bug 926786 (CVE-2024-28757)

Summary:	<dev-libs/expat-2.6.2: vulnerable to billion laughs attacks with isolated use of external parsers
Product:	Gentoo Security	Reporter:	Sebastian Pipping <sping>
Component:	Vulnerabilities	Assignee:	Gentoo Security <security>
Status:	IN_PROGRESS ---
Severity:	normal	CC:	ajak, sping
Priority:	Normal
Version:	unspecified
Hardware:	All
OS:	Linux
URL:	https://github.com/libexpat/libexpat/commit/1d50b80cf31de87750103656f6eb693746854aa8
See Also:	https://github.com/libexpat/libexpat/issues/838 https://github.com/libexpat/libexpat/issues/839 https://github.com/libexpat/libexpat/pull/842
Whiteboard:	A3 [stable]
Package list:		Runtime testing required:	---
Bug Depends on:	924601, 939074, 930032
Bug Blocks:	923951

Description Sebastian Pipping gentoo-dev

2024-03-11 23:40:40 UTC

Quick heads up, upstream release 2.6.2 with the fix coming up shortly.

Comment 1 Larry the Git Cow gentoo-dev

2024-03-13 17:19:07 UTC

The bug has been referenced in the following commit(s):

https://gitweb.gentoo.org/repo/gentoo.git/commit/?id=b5ae6bfa8a1411b05e9fbd76792144b7824e9379

commit b5ae6bfa8a1411b05e9fbd76792144b7824e9379
Author:     Sebastian Pipping <sping@gentoo.org>
AuthorDate: 2024-03-13 17:18:05 +0000
Commit:     Sebastian Pipping <sping@gentoo.org>
CommitDate: 2024-03-13 17:18:32 +0000

    dev-libs/expat: 2.6.2 with fix for CVE-2024-28757
    
    Bug: https://bugs.gentoo.org/926786
    Signed-off-by: Sebastian Pipping <sping@gentoo.org>

 dev-libs/expat/Manifest           |   1 +
 dev-libs/expat/expat-2.6.2.ebuild | 101 ++++++++++++++++++++++++++++++++++++++
 2 files changed, 102 insertions(+)

Comment 2 John Helmert III archtester

2024-03-17 04:03:50 UTC

Please stable when ready.

Comment 3 Sebastian Pipping gentoo-dev

2024-03-17 12:39:04 UTC

(In reply to John Helmert III from comment #2)
> Please stable when ready.

I'm biased about it (being Expat's upstream maintainer), but someone will have to make a decision whether bug #924601 is a blocker to stabilization or not, I'm adding it as a potential blocker for now until someone  else removes it as a conscious decision.  ClusterShell tests will break through stabilization, not sure if the application itself.  We could in theory try to get an answer on that from Clustershell upstream (just asked at https://github.com/cea-hpc/clustershell/pull/556#issuecomment-2002439361), before making a move, with a timeout of a few days maybe.  What do you think?

I'm also making bug #923951 depend on bug #926786 (right here), please adjust as you see fits.