linux-kernel - Re: [PATCH 0/5] i387: stable kernel backport

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: <CA+55aFyxB4kmmSOKDSRkAkaRx=G7o7JnuUGBu3_WVr9KbpWWYQ@mail.gmail.com>
Date:	Thu, 23 Feb 2012 14:11:48 -0800
From:	Linus Torvalds <torvalds@...ux-foundation.org>
To:	Willy Tarreau <w@....eu>
Cc:	Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
	"H. Peter Anvin" <hpa@...or.com>, stable@...r.kernel.org,
	Raphael Prevost <raphael@...o.asia>,
	Suresh Siddha <suresh.b.siddha@...el.com>,
	Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH 0/5] i387: stable kernel backport

On Thu, Feb 23, 2012 at 1:52 PM, Willy Tarreau <w@....eu> wrote:
>
> I would test this too, but apart from ensuring my kernel still boots,
> how do I ensure the patches do really fix what the ought to fix ? I
> must admit I didn't catch the initial issue they were supposed to fix
> unfortunately :-/

Almost nobody did.

This only happens on modern CPU's that support the new AES-NI
instructions, and only with a 32-bit kernel (although the very
unlikely preemption issues can happen on x86-64 too). And you need to
have the AES instructions called from interrupts, which probably only
happens with the mac80211 wireless networking stack.

And even then you need WPA2 to trigger it (I guess AES is sometimes
used with "extended WPA1" too, but I dunno).

So it's not impossible to trigger, but you do need to have a fairly
recent CPU that happily runs in 64-bit mode, and install a 32-bit
system on it. And it needs to use the right wireless setup.

It's possible that the right solution for really older kernels is just
to say that AES_NI depends on X86_64.

                      Linus
--
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/