From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mx0a-001b2d01.pphosted.com ([148.163.156.1]:56252 "EHLO mx0a-001b2d01.pphosted.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750795AbdLNIfk (ORCPT ); Thu, 14 Dec 2017 03:35:40 -0500 Received: from pps.filterd (m0098394.ppops.net [127.0.0.1]) by mx0a-001b2d01.pphosted.com (8.16.0.21/8.16.0.21) with SMTP id vBE8Yu0s033044 for ; Thu, 14 Dec 2017 03:35:40 -0500 Received: from e06smtp11.uk.ibm.com (e06smtp11.uk.ibm.com [195.75.94.107]) by mx0a-001b2d01.pphosted.com with ESMTP id 2eun0ya69r-1 (version=TLSv1.2 cipher=AES256-SHA bits=256 verify=NOT) for ; Thu, 14 Dec 2017 03:35:39 -0500 Received: from localhost by e06smtp11.uk.ibm.com with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted for from ; Thu, 14 Dec 2017 08:35:37 -0000 Date: Thu, 14 Dec 2017 09:35:32 +0100 From: Heiko Carstens To: gregkh@linuxfoundation.org Cc: schwidefsky@de.ibm.com, stable@vger.kernel.org Subject: Re: FAILED: patch "[PATCH] s390: always save and restore all registers on context switch" failed to apply to 4.4-stable tree References: <151302783222933@kroah.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <151302783222933@kroah.com> Message-Id: <20171214083532.GB4238@osiris> Sender: stable-owner@vger.kernel.org List-ID: On Mon, Dec 11, 2017 at 10:30:32PM +0100, gregkh@linuxfoundation.org wrote: > > The patch below does not apply to the 4.4-stable tree. > If someone wants it applied there, or to any other stable or longterm > tree, then please email the backport, including the original git commit > id to . Please find below the patch applies against 4.4-stable. >>From 7ad104f73a7eece168875e2bc19d7c3f8944a938 Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Mon, 20 Nov 2017 12:38:44 +0100 Subject: [PATCH] s390: always save and restore all registers on context switch commit fbbd7f1a51965b50dd12924841da0d478f3da71b upstream. The switch_to() macro has an optimization to avoid saving and restoring register contents that aren't needed for kernel threads. There is however the possibility that a kernel thread execve's a user space program. In such a case the execve'd process can partially see the contents of the previous process, which shouldn't be allowed. To avoid this, simply always save and restore register contents on context switch. Fixes: fdb6d070effba ("switch_to: dont restore/save access & fpu regs for kernel threads") Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/include/asm/switch_to.h | 19 +++++++++---------- 1 file changed, 9 insertions(+), 10 deletions(-) diff --git a/arch/s390/include/asm/switch_to.h b/arch/s390/include/asm/switch_to.h index dde6b52359c5..ff2fbdafe689 100644 --- a/arch/s390/include/asm/switch_to.h +++ b/arch/s390/include/asm/switch_to.h @@ -29,17 +29,16 @@ static inline void restore_access_regs(unsigned int *acrs) } #define switch_to(prev,next,last) do { \ - if (prev->mm) { \ - save_fpu_regs(); \ - save_access_regs(&prev->thread.acrs[0]); \ - save_ri_cb(prev->thread.ri_cb); \ - } \ + /* save_fpu_regs() sets the CIF_FPU flag, which enforces \ + * a restore of the floating point / vector registers as \ + * soon as the next task returns to user space \ + */ \ + save_fpu_regs(); \ + save_access_regs(&prev->thread.acrs[0]); \ + save_ri_cb(prev->thread.ri_cb); \ update_cr_regs(next); \ - if (next->mm) { \ - set_cpu_flag(CIF_FPU); \ - restore_access_regs(&next->thread.acrs[0]); \ - restore_ri_cb(next->thread.ri_cb, prev->thread.ri_cb); \ - } \ + restore_access_regs(&next->thread.acrs[0]); \ + restore_ri_cb(next->thread.ri_cb, prev->thread.ri_cb); \ prev = __switch_to(prev,next); \ } while (0) -- 2.13.5