bugtraq - Intel CPU bug forcing page table switch during syscalls?

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>] [day] [month] [year] [list]

Date: Wed, 3 Jan 2018 22:31:29 +0100
From: Pavel Machek <pavel@....cz>
To: bugtraq@...urityfocus.com
Subject: Intel CPU bug forcing page table switch during syscalls?

Hi!

It looks like there's Intel CPU bug, allowing prefetch from kernel
memory. It seems to be reason KASLR patches are pushed so fast to Linux.

https://mobile.twitter.com/brainsmoke/status/948561799875502080/photo/1
https://forums.freebsd.org/threads/63955/page-2#post-371276

Hmm.

Does that mean we can do

   u16 *peek_addr = <somewhere into kernel>;
   char cacheline1[64];
   char cacheline2[64];

   wbinvd();

   if (*peek_addr == 0x1234)
      (volatile char *) cacheline1[0];
   else
      (volatile char *) cacheline2[0];

Thread will certainly die to SIGSEGV here, but from other thread we
should be able to tell if cacheline1 or cacheline2 is in cache... and
therefore read unreadable memory....?

									Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

Download attachment "signature.asc" of type "application/pgp-signature" (182 bytes)