From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id F06ADC433F5 for ; Thu, 21 Oct 2021 17:09:41 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id C6A6A61A2E for ; Thu, 21 Oct 2021 17:09:41 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S231826AbhJURL4 (ORCPT ); Thu, 21 Oct 2021 13:11:56 -0400 Received: from mail.kernel.org ([198.145.29.99]:42682 "EHLO mail.kernel.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S230281AbhJURLr (ORCPT ); Thu, 21 Oct 2021 13:11:47 -0400 Received: by mail.kernel.org (Postfix) with ESMTPSA id EA4EE61A0A; Thu, 21 Oct 2021 17:09:28 +0000 (UTC) Date: Thu, 21 Oct 2021 18:09:26 +0100 From: Catalin Marinas To: Andreas Gruenbacher Cc: Linus Torvalds , Al Viro , Christoph Hellwig , "Darrick J. Wong" , Jan Kara , Matthew Wilcox , cluster-devel , linux-fsdevel , Linux Kernel Mailing List , "ocfs2-devel@oss.oracle.com" , Josef Bacik , Will Deacon Subject: Re: [RFC][arm64] possible infinite loop in btrfs search_ioctl() Message-ID: References: MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Thu, Oct 21, 2021 at 04:42:33PM +0200, Andreas Gruenbacher wrote: > On Thu, Oct 21, 2021 at 12:06 PM Catalin Marinas > wrote: > > On Thu, Oct 21, 2021 at 02:46:10AM +0200, Andreas Gruenbacher wrote: > > > When a page fault would occur, we > > > get back an error instead, and then we try to fault in the offending > > > pages. If a page is resident and we still get a fault trying to access > > > it, trying to fault in the same page again isn't going to help and we > > > have a true error. > > > > You can't be sure the second fault is a true error. The unlocked > > fault_in_*() may race with some LRU scheme making the pte not accessible > > or a write-back making it clean/read-only. copy_to_user() with > > pagefault_disabled() fails again but that's a benign fault. The > > filesystem should re-attempt the fault-in (gup would correct the pte), > > disable page faults and copy_to_user(), potentially in an infinite loop. > > If you bail out on the second/third uaccess following a fault_in_*() > > call, you may get some unexpected errors (though very rare). Maybe the > > filesystems avoid this problem somehow but I couldn't figure it out. > > Good point, we can indeed only bail out if both the user copy and the > fault-in fail. > > But probing the entire memory range in fault domain granularity in the > page fault-in functions still doesn't actually make sense. Those > functions really only need to guarantee that we'll be able to make > progress eventually. From that point of view, it should be enough to > probe the first byte of the requested memory range, so when one of > those functions reports that the next N bytes should be accessible, > this really means that the first byte surely isn't permanently > inaccessible and that the rest is likely accessible. Functions > fault_in_readable and fault_in_writeable already work that way, so > this only leaves function fault_in_safe_writeable to worry about. I agree, that's why generic_perform_write() works. It does a get_user() from the first byte in that range and the subsequent copy_from_user() will make progress of at least one byte if it was readable. Eventually it will hit the byte that faults. The gup-based fault_in_*() are a bit more problematic. Your series introduces fault_in_safe_writeable() and I think for MTE doing a _single_ get_user(uaddr) (in addition to the gup checks for write) would be sufficient as long as generic_file_read_iter() advances by at least one byte (eventually). This discussion started with the btrfs search_ioctl() where, even if some bytes were written in copy_to_sk(), it always restarts from an earlier position, reattempting to write the same bytes. Since copy_to_sk() doesn't guarantee forward progress even if some bytes are writable, Linus' suggestion was for fault_in_writable() to probe the whole range. I consider this overkill since btrfs is the only one that needs probing every 16 bytes. The other cases like the new fault_in_safe_writeable() can be fixed by probing the first byte only followed by gup. I think we need to better define the semantics of the fault_in + uaccess sequences. For uaccess, we document "a hard requirement that not storing anything at all (i.e. returning size) should happen only when nothing could be copied" (from linux/uaccess.h). I think we can add a requirement for the new size_t-based fault_in_* variants without mandating that the whole range is probed, something like: "returning leftover < size guarantees that a subsequent user access at uaddr copies at least one byte eventually". I said "eventually" but maybe we can come up with some clearer wording for a liveness property. Such requirement would ensure that infinite loops of fault_in_* + uaccess make progress as long as they don't reset the probed range. Code like btrfs search_ioctl() would need to be adjusted to avoid such range reset and guarantee forward progress. -- Catalin From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id C51D2C433EF for ; Thu, 21 Oct 2021 17:09:39 +0000 (UTC) Received: from mx0a-00069f02.pphosted.com (mx0a-00069f02.pphosted.com [205.220.165.32]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 6918B61A0C for ; Thu, 21 Oct 2021 17:09:39 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.4.1 mail.kernel.org 6918B61A0C Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=arm.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=oss.oracle.com Received: from pps.filterd (m0246627.ppops.net [127.0.0.1]) by mx0b-00069f02.pphosted.com (8.16.1.2/8.16.1.2) with SMTP id 19LGi3U3030729; Thu, 21 Oct 2021 17:09:38 GMT Received: from userp3020.oracle.com (userp3020.oracle.com [156.151.31.79]) by mx0b-00069f02.pphosted.com with ESMTP id 3btkxa0431-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Thu, 21 Oct 2021 17:09:38 +0000 Received: from pps.filterd (userp3020.oracle.com [127.0.0.1]) by userp3020.oracle.com (8.16.1.2/8.16.1.2) with SMTP id 19LH5dKl049740; Thu, 21 Oct 2021 17:09:37 GMT Received: from oss.oracle.com (oss-old-reserved.oracle.com [137.254.22.2]) by userp3020.oracle.com with ESMTP id 3br8gwd4bu-1 (version=TLSv1 cipher=AES256-SHA bits=256 verify=NO); Thu, 21 Oct 2021 17:09:37 +0000 Received: from localhost ([127.0.0.1] helo=lb-oss.oracle.com) by oss.oracle.com with esmtp (Exim 4.63) (envelope-from ) id 1mdbZQ-00060N-4G; Thu, 21 Oct 2021 10:09:36 -0700 Received: from aserp3020.oracle.com ([141.146.126.70]) by oss.oracle.com with esmtp (Exim 4.63) (envelope-from ) id 1mdbZN-0005zz-PO for ocfs2-devel@oss.oracle.com; Thu, 21 Oct 2021 10:09:33 -0700 Received: from pps.filterd (aserp3020.oracle.com [127.0.0.1]) by aserp3020.oracle.com (8.16.1.2/8.16.1.2) with SMTP id 19LH7FxU098929 for ; Thu, 21 Oct 2021 17:09:33 GMT Received: from mx0a-00069f01.pphosted.com (mx0a-00069f01.pphosted.com [205.220.165.26]) by aserp3020.oracle.com with ESMTP id 3bqpj95fy8-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK) for ; Thu, 21 Oct 2021 17:09:33 +0000 Received: from pps.filterd (m0246574.ppops.net [127.0.0.1]) by mx0b-00069f01.pphosted.com (8.16.1.2/8.16.1.2) with SMTP id 19LFbgvG011544 for ; Thu, 21 Oct 2021 17:09:32 GMT Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by mx0b-00069f01.pphosted.com with ESMTP id 3bub0q9744-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=NO) for ; Thu, 21 Oct 2021 17:09:32 +0000 Received: by mail.kernel.org (Postfix) with ESMTPSA id EA4EE61A0A; Thu, 21 Oct 2021 17:09:28 +0000 (UTC) Date: Thu, 21 Oct 2021 18:09:26 +0100 From: Catalin Marinas To: Andreas Gruenbacher Message-ID: References: MIME-Version: 1.0 Content-Disposition: inline In-Reply-To: X-Source-IP: 198.145.29.99 X-ServerName: mail.kernel.org X-Proofpoint-SPF-Result: pass X-Proofpoint-SPF-Record: v=spf1 mx include:_spf.kernel.org ~all X-Proofpoint-Virus-Version: vendor=nai engine=6300 definitions=10144 signatures=668683 X-Proofpoint-Spam-Reason: safe X-Spam: OrgSafeList X-SpamRule: orgsafelist Cc: cluster-devel , Jan Kara , Will Deacon , Linux Kernel Mailing List , Josef Bacik , Christoph Hellwig , Al Viro , linux-fsdevel , Linus Torvalds , "ocfs2-devel@oss.oracle.com" Subject: Re: [Ocfs2-devel] [RFC][arm64] possible infinite loop in btrfs search_ioctl() X-BeenThere: ocfs2-devel@oss.oracle.com X-Mailman-Version: 2.1.9 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Sender: ocfs2-devel-bounces@oss.oracle.com Errors-To: ocfs2-devel-bounces@oss.oracle.com X-Proofpoint-Virus-Version: vendor=nai engine=6300 definitions=10144 signatures=668683 X-Proofpoint-Spam-Details: rule=notspam policy=default score=0 mlxlogscore=999 mlxscore=0 adultscore=0 spamscore=0 phishscore=0 bulkscore=0 suspectscore=0 malwarescore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2109230001 definitions=main-2110210085 X-Proofpoint-ORIG-GUID: 14pgdmWXlb017ZF8THmr5J2VZF0kdyeN X-Proofpoint-GUID: 14pgdmWXlb017ZF8THmr5J2VZF0kdyeN On Thu, Oct 21, 2021 at 04:42:33PM +0200, Andreas Gruenbacher wrote: > On Thu, Oct 21, 2021 at 12:06 PM Catalin Marinas > wrote: > > On Thu, Oct 21, 2021 at 02:46:10AM +0200, Andreas Gruenbacher wrote: > > > When a page fault would occur, we > > > get back an error instead, and then we try to fault in the offending > > > pages. If a page is resident and we still get a fault trying to access > > > it, trying to fault in the same page again isn't going to help and we > > > have a true error. > > > > You can't be sure the second fault is a true error. The unlocked > > fault_in_*() may race with some LRU scheme making the pte not accessible > > or a write-back making it clean/read-only. copy_to_user() with > > pagefault_disabled() fails again but that's a benign fault. The > > filesystem should re-attempt the fault-in (gup would correct the pte), > > disable page faults and copy_to_user(), potentially in an infinite loop. > > If you bail out on the second/third uaccess following a fault_in_*() > > call, you may get some unexpected errors (though very rare). Maybe the > > filesystems avoid this problem somehow but I couldn't figure it out. > > Good point, we can indeed only bail out if both the user copy and the > fault-in fail. > > But probing the entire memory range in fault domain granularity in the > page fault-in functions still doesn't actually make sense. Those > functions really only need to guarantee that we'll be able to make > progress eventually. From that point of view, it should be enough to > probe the first byte of the requested memory range, so when one of > those functions reports that the next N bytes should be accessible, > this really means that the first byte surely isn't permanently > inaccessible and that the rest is likely accessible. Functions > fault_in_readable and fault_in_writeable already work that way, so > this only leaves function fault_in_safe_writeable to worry about. I agree, that's why generic_perform_write() works. It does a get_user() from the first byte in that range and the subsequent copy_from_user() will make progress of at least one byte if it was readable. Eventually it will hit the byte that faults. The gup-based fault_in_*() are a bit more problematic. Your series introduces fault_in_safe_writeable() and I think for MTE doing a _single_ get_user(uaddr) (in addition to the gup checks for write) would be sufficient as long as generic_file_read_iter() advances by at least one byte (eventually). This discussion started with the btrfs search_ioctl() where, even if some bytes were written in copy_to_sk(), it always restarts from an earlier position, reattempting to write the same bytes. Since copy_to_sk() doesn't guarantee forward progress even if some bytes are writable, Linus' suggestion was for fault_in_writable() to probe the whole range. I consider this overkill since btrfs is the only one that needs probing every 16 bytes. The other cases like the new fault_in_safe_writeable() can be fixed by probing the first byte only followed by gup. I think we need to better define the semantics of the fault_in + uaccess sequences. For uaccess, we document "a hard requirement that not storing anything at all (i.e. returning size) should happen only when nothing could be copied" (from linux/uaccess.h). I think we can add a requirement for the new size_t-based fault_in_* variants without mandating that the whole range is probed, something like: "returning leftover < size guarantees that a subsequent user access at uaddr copies at least one byte eventually". I said "eventually" but maybe we can come up with some clearer wording for a liveness property. Such requirement would ensure that infinite loops of fault_in_* + uaccess make progress as long as they don't reset the probed range. Code like btrfs search_ioctl() would need to be adjusted to avoid such range reset and guarantee forward progress. -- Catalin _______________________________________________ Ocfs2-devel mailing list Ocfs2-devel@oss.oracle.com https://oss.oracle.com/mailman/listinfo/ocfs2-devel From mboxrd@z Thu Jan 1 00:00:00 1970 From: Catalin Marinas Date: Thu, 21 Oct 2021 18:09:26 +0100 Subject: [Cluster-devel] [RFC][arm64] possible infinite loop in btrfs search_ioctl() In-Reply-To: References: Message-ID: List-Id: To: cluster-devel.redhat.com MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit On Thu, Oct 21, 2021 at 04:42:33PM +0200, Andreas Gruenbacher wrote: > On Thu, Oct 21, 2021 at 12:06 PM Catalin Marinas > wrote: > > On Thu, Oct 21, 2021 at 02:46:10AM +0200, Andreas Gruenbacher wrote: > > > When a page fault would occur, we > > > get back an error instead, and then we try to fault in the offending > > > pages. If a page is resident and we still get a fault trying to access > > > it, trying to fault in the same page again isn't going to help and we > > > have a true error. > > > > You can't be sure the second fault is a true error. The unlocked > > fault_in_*() may race with some LRU scheme making the pte not accessible > > or a write-back making it clean/read-only. copy_to_user() with > > pagefault_disabled() fails again but that's a benign fault. The > > filesystem should re-attempt the fault-in (gup would correct the pte), > > disable page faults and copy_to_user(), potentially in an infinite loop. > > If you bail out on the second/third uaccess following a fault_in_*() > > call, you may get some unexpected errors (though very rare). Maybe the > > filesystems avoid this problem somehow but I couldn't figure it out. > > Good point, we can indeed only bail out if both the user copy and the > fault-in fail. > > But probing the entire memory range in fault domain granularity in the > page fault-in functions still doesn't actually make sense. Those > functions really only need to guarantee that we'll be able to make > progress eventually. From that point of view, it should be enough to > probe the first byte of the requested memory range, so when one of > those functions reports that the next N bytes should be accessible, > this really means that the first byte surely isn't permanently > inaccessible and that the rest is likely accessible. Functions > fault_in_readable and fault_in_writeable already work that way, so > this only leaves function fault_in_safe_writeable to worry about. I agree, that's why generic_perform_write() works. It does a get_user() from the first byte in that range and the subsequent copy_from_user() will make progress of at least one byte if it was readable. Eventually it will hit the byte that faults. The gup-based fault_in_*() are a bit more problematic. Your series introduces fault_in_safe_writeable() and I think for MTE doing a _single_ get_user(uaddr) (in addition to the gup checks for write) would be sufficient as long as generic_file_read_iter() advances by at least one byte (eventually). This discussion started with the btrfs search_ioctl() where, even if some bytes were written in copy_to_sk(), it always restarts from an earlier position, reattempting to write the same bytes. Since copy_to_sk() doesn't guarantee forward progress even if some bytes are writable, Linus' suggestion was for fault_in_writable() to probe the whole range. I consider this overkill since btrfs is the only one that needs probing every 16 bytes. The other cases like the new fault_in_safe_writeable() can be fixed by probing the first byte only followed by gup. I think we need to better define the semantics of the fault_in + uaccess sequences. For uaccess, we document "a hard requirement that not storing anything at all (i.e. returning size) should happen only when nothing could be copied" (from linux/uaccess.h). I think we can add a requirement for the new size_t-based fault_in_* variants without mandating that the whole range is probed, something like: "returning leftover < size guarantees that a subsequent user access@uaddr copies at least one byte eventually". I said "eventually" but maybe we can come up with some clearer wording for a liveness property. Such requirement would ensure that infinite loops of fault_in_* + uaccess make progress as long as they don't reset the probed range. Code like btrfs search_ioctl() would need to be adjusted to avoid such range reset and guarantee forward progress. -- Catalin