From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755430AbYHRSkv (ORCPT ); Mon, 18 Aug 2008 14:40:51 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1753448AbYHRSkn (ORCPT ); Mon, 18 Aug 2008 14:40:43 -0400 Received: from mga11.intel.com ([192.55.52.93]:48170 "EHLO mga11.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753387AbYHRSkm (ORCPT ); Mon, 18 Aug 2008 14:40:42 -0400 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="4.32,229,1217833200"; d="scan'208";a="607689256" Message-Id: <20080818183605.860693000@linux-os.sc.intel.com> References: <20080818183605.719267000@linux-os.sc.intel.com> User-Agent: quilt/0.46-1 Date: Mon, 18 Aug 2008 11:35:47 -0700 From: Suresh Siddha To: greg@kroah.com, mingo@elte.hu Cc: stable@kernel.org, linux-kernel@vger.kernel.org, Suresh Siddha , Linus Torvalds Subject: [stable 2/2] x64, fpu: fix possible FPU leakage in error conditions Content-Disposition: inline; filename=fix_fpu_leakage.patch Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org [Upstream commit: 6ffac1e90a17ea0aded5c581204397421eec91b6] On Thu, Jul 24, 2008 at 03:43:44PM -0700, Linus Torvalds wrote: > So how about this patch as a starting point? This is the RightThing(tm) to > do regardless, and if it then makes it easier to do some other cleanups, > we should do it first. What do you think? restore_fpu_checking() calls init_fpu() in error conditions. While this is wrong(as our main intention is to clear the fpu state of the thread), this was benign before commit 92d140e21f1 ("x86: fix taking DNA during 64bit sigreturn"). Post commit 92d140e21f1, live FPU registers may not belong to this process at this error scenario. In the error condition for restore_fpu_checking() (especially during the 64bit signal return), we are doing init_fpu(), which saves the live FPU register state (possibly belonging to some other process context) into the thread struct (through unlazy_fpu() in init_fpu()). This is wrong and can leak the FPU data. For the signal handler restore error condition in restore_i387(), clear the fpu state present in the thread struct(before ultimately sending a SIGSEGV for badframe). For the paranoid error condition check in math_state_restore(), send a SIGSEGV, if we fail to restore the state. Signed-off-by: Suresh Siddha Cc: Cc: Linus Torvalds Signed-off-by: Ingo Molnar --- Index: linux-2.6.26.2/arch/x86/kernel/signal_64.c =================================================================== --- linux-2.6.26.2.orig/arch/x86/kernel/signal_64.c 2008-08-18 11:16:43.000000000 -0700 +++ linux-2.6.26.2/arch/x86/kernel/signal_64.c 2008-08-18 11:16:50.000000000 -0700 @@ -104,7 +104,16 @@ clts(); task_thread_info(current)->status |= TS_USEDFPU; } - return restore_fpu_checking((__force struct i387_fxsave_struct *)buf); + err = restore_fpu_checking((__force struct i387_fxsave_struct *)buf); + if (unlikely(err)) { + /* + * Encountered an error while doing the restore from the + * user buffer, clear the fpu state. + */ + clear_fpu(tsk); + clear_used_math(); + } + return err; } /* Index: linux-2.6.26.2/arch/x86/kernel/traps_64.c =================================================================== --- linux-2.6.26.2.orig/arch/x86/kernel/traps_64.c 2008-08-06 09:19:01.000000000 -0700 +++ linux-2.6.26.2/arch/x86/kernel/traps_64.c 2008-08-18 11:17:43.000000000 -0700 @@ -1141,7 +1141,14 @@ } clts(); /* Allow maths ops (or we recurse) */ - restore_fpu_checking(&me->thread.xstate->fxsave); + /* + * Paranoid restore. send a SIGSEGV if we fail to restore the state. + */ + if (unlikely(restore_fpu_checking(&me->thread.xstate->fxsave))) { + stts(); + force_sig(SIGSEGV, me); + return; + } task_thread_info(me)->status |= TS_USEDFPU; me->fpu_counter++; } Index: linux-2.6.26.2/include/asm-x86/i387.h =================================================================== --- linux-2.6.26.2.orig/include/asm-x86/i387.h 2008-08-18 11:16:43.000000000 -0700 +++ linux-2.6.26.2/include/asm-x86/i387.h 2008-08-18 11:16:50.000000000 -0700 @@ -62,8 +62,6 @@ #else : [fx] "cdaSDb" (fx), "m" (*fx), "0" (0)); #endif - if (unlikely(err)) - init_fpu(current); return err; } --