Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754674AbcLaLKn (ORCPT ); Sat, 31 Dec 2016 06:10:43 -0500 Received: from mail-wj0-f195.google.com ([209.85.210.195]:36598 "EHLO mail-wj0-f195.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754647AbcLaLKl (ORCPT ); Sat, 31 Dec 2016 06:10:41 -0500 MIME-Version: 1.0 In-Reply-To: References: <20161230155634.8692-1-dsafonov@virtuozzo.com> <20161230155634.8692-5-dsafonov@virtuozzo.com> From: Dmitry Safonov <0x7f454c46@gmail.com> Date: Sat, 31 Dec 2016 14:10:19 +0300 Message-ID: Subject: Re: [RFC 4/4] x86/arch_prctl: add ARCH_{GET,SET}_TASK_SIZE To: Andy Lutomirski Cc: Dmitry Safonov , "linux-kernel@vger.kernel.org" , Thomas Gleixner , Ingo Molnar , "H. Peter Anvin" , Andy Lutomirski , "Kirill A. Shutemov" , X86 ML Content-Type: text/plain; charset=UTF-8 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1445 Lines: 30 Hi Andy, thanks for reviewing! 2016-12-31 5:02 GMT+03:00 Andy Lutomirski : > On Fri, Dec 30, 2016 at 7:56 AM, Dmitry Safonov wrote: >> Add arch_prctl getters/setters for size of virtual address space of task. >> This adds ability to change task's virtual address space limit. >> I need this for correctly restore virtual address space limits in CRIU. >> Currently, on x86 there are three task sizes: 3GB for some old 32 bit java >> apps, 4Gb for ordinary 32-bit compatible apps and 47-bits for native >> x86_64 processes. >> 32-bit applications are restored by CRIU with the help of 64-bit clone()-d >> child, and on restore we need to place correct address space limitations >> back - otherwise 32-bit restored application may mmap() address over >> 4Gb space and as this address will not fit into 4-byte pointer, it >> will silently reuse/corrupt the pointer that has the same lower 4-bytes. >> > > I agree we need something like this, but this particular justification > is a bit bogus. If 32-bit mmap() returns an address above 2^32, then > I think it's a straight-up bug. The address space limit shouldn't > have anything to do with it -- the kernel *knows* that it's a > "compat" syscall. Yep, I guess, I didn't realize that the real wrong thing is that compat syscall returns address above 4Gb and not the address space limits here. Thanks, will look into that. -- Dmitry