From patchwork Fri Dec 27 19:48:29 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Theodore Ts'o X-Patchwork-Id: 11311307 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 3BDCC138C for ; Fri, 27 Dec 2019 19:48:38 +0000 (UTC) Received: from kanga.kvack.org (kanga.kvack.org [205.233.56.17]) by mail.kernel.org (Postfix) with ESMTP id CCF48208C4 for ; Fri, 27 Dec 2019 19:48:37 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org CCF48208C4 Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=mit.edu Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=owner-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix) id 2D84E8E0005; Fri, 27 Dec 2019 14:48:36 -0500 (EST) Delivered-To: linux-mm-outgoing@kvack.org Received: by kanga.kvack.org (Postfix, from userid 40) id 288808E0001; Fri, 27 Dec 2019 14:48:36 -0500 (EST) X-Original-To: int-list-linux-mm@kvack.org X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id 176748E0005; Fri, 27 Dec 2019 14:48:36 -0500 (EST) X-Original-To: linux-mm@kvack.org X-Delivered-To: linux-mm@kvack.org Received: from forelay.hostedemail.com (smtprelay0216.hostedemail.com [216.40.44.216]) by kanga.kvack.org (Postfix) with ESMTP id F30AA8E0001 for ; Fri, 27 Dec 2019 14:48:35 -0500 (EST) Received: from smtpin23.hostedemail.com (10.5.19.251.rfc1918.com [10.5.19.251]) by forelay03.hostedemail.com (Postfix) with SMTP id 9DD448249980 for ; Fri, 27 Dec 2019 19:48:35 +0000 (UTC) X-FDA: 76311958590.23.flesh25_10b4be9ee0c0e X-Spam-Summary: 75,0,0,96ad35d4443c12cf,d41d8cd98f00b204,tytso@mit.edu,:linux-fsdevel@vger.kernel.org::tytso@mit.edu,RULES_HIT:1:2:41:69:355:379:421:541:800:960:966:973:988:989:1260:1311:1314:1345:1437:1515:1605:1730:1747:1777:1792:2196:2199:2393:2553:2559:2562:2693:3138:3139:3140:3141:3142:3622:3865:3866:3867:3868:3870:3871:3872:3874:4052:4250:4321:4385:4605:5007:6119:6238:6261:7619:7903:9707:10004:11026:11473:11657:11658:11914:12043:12296:12297:12438:12523:12555:12679:12683:12698:12737:12895:12986:13161:13229:13870:13894:13972:14096:14394:21080:21450:21451:21627:21740:21987:21990:30004:30012:30029:30045:30054:30056:30080:30090,0,RBL:18.9.28.11-irl.urbl.hostedemail.com-127.0.0.175,CacheIP:none,Bayesian:0.5,0.5,0.5,Netcheck:none,DomainCache:0,MSF:not bulk,SPF:ft,MSBL:0,DNSBL:neutral,Custom_rules:0:0:0,LFtime:24,LUA_SUMMARY:none X-HE-Tag: flesh25_10b4be9ee0c0e X-Filterd-Recvd-Size: 10782 Received: from outgoing.mit.edu (outgoing-auth-1.mit.edu [18.9.28.11]) by imf19.hostedemail.com (Postfix) with ESMTP for ; Fri, 27 Dec 2019 19:48:34 +0000 (UTC) Received: from callcc.thunk.org (96-72-84-49-static.hfc.comcastbusiness.net [96.72.84.49] (may be forged)) (authenticated bits=0) (User authenticated as tytso@ATHENA.MIT.EDU) by outgoing.mit.edu (8.14.7/8.12.4) with ESMTP id xBRJmWrs000588 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-GCM-SHA384 bits=256 verify=NOT); Fri, 27 Dec 2019 14:48:33 -0500 Received: by callcc.thunk.org (Postfix, from userid 15806) id 4972F420485; Fri, 27 Dec 2019 14:48:32 -0500 (EST) From: "Theodore Ts'o" To: Linux Filesystem Development List Cc: linux-mm@kvack.org, "Theodore Ts'o" Subject: [PATCH] memcg: fix a crash in wb_workfn when a device disappears Date: Fri, 27 Dec 2019 14:48:29 -0500 Message-Id: <20191227194829.150110-1-tytso@mit.edu> X-Mailer: git-send-email 2.24.1 MIME-Version: 1.0 X-Bogosity: Ham, tests=bogofilter, spamicity=0.000000, version=1.2.4 Sender: owner-linux-mm@kvack.org Precedence: bulk X-Loop: owner-majordomo@kvack.org List-ID: Without memcg, there is a one-to-one mapping between the bdi and bdi_writeback structures. In this world, things are fairly straightforward; the first thing bdi_unregister() does is to shutdown the bdi_writeback structure (or wb), and part of that writeback ensures that no other work queued against the wb, and that the wb is fully drained. With memcg, however, there is a one-to-many relationship between the bdi and bdi_writeback structures; that is, there are multiple wb objects which can all point to a single bdi. There is a refcount which prevents the bdi object from being released (and hence, unregistered). So in theory, the bdi_unregister() *should* only get called once its refcount goes to zero (bdi_put will drop the refcount, and when it is zero, release_bdi gets called, which calls bdi_unregister. Unfortunately, del_gendisk() in block/gen_hd.c never got the memo about the Brave New memcg World, and calls bdi_unregister directly. It does this without informing the file system, or the memcg code, or anything else. This causes the root wb associated with the bdi to be unregistered, but none of the memcg-specific wb's are shutdown. So when one of these wb's are woken up to do delayed work, they try to dereference their wb->bdi->dev to fetch the device name, but unfortunately bdi->dev is now NULL, thanks to the bdi_unregister() called by del_gendisk(). As a result, *boom*. Fortunately, it looks like the rest of the writeback path is perfectly happy with bdi->dev and bdi->owner being NULL, so the simplest fix is to create a bdi_dev_name() function which can handle bdi->dev being NULL. This also allows us to bulletproof the writeback tracepoints to prevent them from dereferencing a NULL pointer and crashing the kernel if one is tracing with memcg's enabled, and an iSCSI device dies or a USB storage stick is pulled. Google-Bug-Id: 145475544 Tested: fs smoke test Signed-off-by: Theodore Ts'o Reported-by: kbuild test robot Reported-by: kbuild test robot --- fs/fs-writeback.c | 2 +- include/linux/backing-dev.h | 9 ++++++++ include/trace/events/writeback.h | 37 +++++++++++++++----------------- mm/backing-dev.c | 1 + 4 files changed, 28 insertions(+), 21 deletions(-) diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c index 335607b8c5c0..76ac9c7d32ec 100644 --- a/fs/fs-writeback.c +++ b/fs/fs-writeback.c @@ -2063,7 +2063,7 @@ void wb_workfn(struct work_struct *work) struct bdi_writeback, dwork); long pages_written; - set_worker_desc("flush-%s", dev_name(wb->bdi->dev)); + set_worker_desc("flush-%s", bdi_dev_name(wb->bdi)); current->flags |= PF_SWAPWRITE; if (likely(!current_is_workqueue_rescuer() || diff --git a/include/linux/backing-dev.h b/include/linux/backing-dev.h index 97967ce06de3..d968a50a0be5 100644 --- a/include/linux/backing-dev.h +++ b/include/linux/backing-dev.h @@ -504,4 +504,13 @@ static inline int bdi_rw_congested(struct backing_dev_info *bdi) (1 << WB_async_congested)); } +extern const char *bdi_unknown_name; + +static inline const char *bdi_dev_name(struct backing_dev_info *bdi) +{ + if (!bdi || !bdi->dev) + return bdi_unknown_name; + return dev_name(bdi->dev); +} + #endif /* _LINUX_BACKING_DEV_H */ diff --git a/include/trace/events/writeback.h b/include/trace/events/writeback.h index ef50be4e5e6c..d94def25e4dc 100644 --- a/include/trace/events/writeback.h +++ b/include/trace/events/writeback.h @@ -67,8 +67,8 @@ DECLARE_EVENT_CLASS(writeback_page_template, TP_fast_assign( strscpy_pad(__entry->name, - mapping ? dev_name(inode_to_bdi(mapping->host)->dev) : "(unknown)", - 32); + bdi_dev_name(mapping ? inode_to_bdi(mapping->host) : + NULL), 32); __entry->ino = mapping ? mapping->host->i_ino : 0; __entry->index = page->index; ), @@ -111,8 +111,7 @@ DECLARE_EVENT_CLASS(writeback_dirty_inode_template, struct backing_dev_info *bdi = inode_to_bdi(inode); /* may be called for files on pseudo FSes w/ unregistered bdi */ - strscpy_pad(__entry->name, - bdi->dev ? dev_name(bdi->dev) : "(unknown)", 32); + strscpy_pad(__entry->name, bdi_dev_name(bdi), 32); __entry->ino = inode->i_ino; __entry->state = inode->i_state; __entry->flags = flags; @@ -193,7 +192,7 @@ TRACE_EVENT(inode_foreign_history, ), TP_fast_assign( - strncpy(__entry->name, dev_name(inode_to_bdi(inode)->dev), 32); + strncpy(__entry->name, bdi_dev_name(inode_to_bdi(inode)), 32); __entry->ino = inode->i_ino; __entry->cgroup_ino = __trace_wbc_assign_cgroup(wbc); __entry->history = history; @@ -222,7 +221,7 @@ TRACE_EVENT(inode_switch_wbs, ), TP_fast_assign( - strncpy(__entry->name, dev_name(old_wb->bdi->dev), 32); + strncpy(__entry->name, bdi_dev_name(old_wb->bdi), 32); __entry->ino = inode->i_ino; __entry->old_cgroup_ino = __trace_wb_assign_cgroup(old_wb); __entry->new_cgroup_ino = __trace_wb_assign_cgroup(new_wb); @@ -255,7 +254,7 @@ TRACE_EVENT(track_foreign_dirty, struct address_space *mapping = page_mapping(page); struct inode *inode = mapping ? mapping->host : NULL; - strncpy(__entry->name, dev_name(wb->bdi->dev), 32); + strncpy(__entry->name, bdi_dev_name(wb->bdi), 32); __entry->bdi_id = wb->bdi->id; __entry->ino = inode ? inode->i_ino : 0; __entry->memcg_id = wb->memcg_css->id; @@ -288,7 +287,7 @@ TRACE_EVENT(flush_foreign, ), TP_fast_assign( - strncpy(__entry->name, dev_name(wb->bdi->dev), 32); + strncpy(__entry->name, bdi_dev_name(wb->bdi), 32); __entry->cgroup_ino = __trace_wb_assign_cgroup(wb); __entry->frn_bdi_id = frn_bdi_id; __entry->frn_memcg_id = frn_memcg_id; @@ -318,7 +317,7 @@ DECLARE_EVENT_CLASS(writeback_write_inode_template, TP_fast_assign( strscpy_pad(__entry->name, - dev_name(inode_to_bdi(inode)->dev), 32); + bdi_dev_name(inode_to_bdi(inode)), 32); __entry->ino = inode->i_ino; __entry->sync_mode = wbc->sync_mode; __entry->cgroup_ino = __trace_wbc_assign_cgroup(wbc); @@ -361,9 +360,7 @@ DECLARE_EVENT_CLASS(writeback_work_class, __field(ino_t, cgroup_ino) ), TP_fast_assign( - strscpy_pad(__entry->name, - wb->bdi->dev ? dev_name(wb->bdi->dev) : - "(unknown)", 32); + strscpy_pad(__entry->name, bdi_dev_name(wb->bdi), 32); __entry->nr_pages = work->nr_pages; __entry->sb_dev = work->sb ? work->sb->s_dev : 0; __entry->sync_mode = work->sync_mode; @@ -416,7 +413,7 @@ DECLARE_EVENT_CLASS(writeback_class, __field(ino_t, cgroup_ino) ), TP_fast_assign( - strscpy_pad(__entry->name, dev_name(wb->bdi->dev), 32); + strscpy_pad(__entry->name, bdi_dev_name(wb->bdi), 32); __entry->cgroup_ino = __trace_wb_assign_cgroup(wb); ), TP_printk("bdi %s: cgroup_ino=%lu", @@ -438,7 +435,7 @@ TRACE_EVENT(writeback_bdi_register, __array(char, name, 32) ), TP_fast_assign( - strscpy_pad(__entry->name, dev_name(bdi->dev), 32); + strscpy_pad(__entry->name, bdi_dev_name(bdi), 32); ), TP_printk("bdi %s", __entry->name @@ -463,7 +460,7 @@ DECLARE_EVENT_CLASS(wbc_class, ), TP_fast_assign( - strscpy_pad(__entry->name, dev_name(bdi->dev), 32); + strscpy_pad(__entry->name, bdi_dev_name(bdi), 32); __entry->nr_to_write = wbc->nr_to_write; __entry->pages_skipped = wbc->pages_skipped; __entry->sync_mode = wbc->sync_mode; @@ -514,7 +511,7 @@ TRACE_EVENT(writeback_queue_io, ), TP_fast_assign( unsigned long *older_than_this = work->older_than_this; - strscpy_pad(__entry->name, dev_name(wb->bdi->dev), 32); + strscpy_pad(__entry->name, bdi_dev_name(wb->bdi), 32); __entry->older = older_than_this ? *older_than_this : 0; __entry->age = older_than_this ? (jiffies - *older_than_this) * 1000 / HZ : -1; @@ -600,7 +597,7 @@ TRACE_EVENT(bdi_dirty_ratelimit, ), TP_fast_assign( - strscpy_pad(__entry->bdi, dev_name(wb->bdi->dev), 32); + strscpy_pad(__entry->bdi, bdi_dev_name(wb->bdi), 32); __entry->write_bw = KBps(wb->write_bandwidth); __entry->avg_write_bw = KBps(wb->avg_write_bandwidth); __entry->dirty_rate = KBps(dirty_rate); @@ -665,7 +662,7 @@ TRACE_EVENT(balance_dirty_pages, TP_fast_assign( unsigned long freerun = (thresh + bg_thresh) / 2; - strscpy_pad(__entry->bdi, dev_name(wb->bdi->dev), 32); + strscpy_pad(__entry->bdi, bdi_dev_name(wb->bdi), 32); __entry->limit = global_wb_domain.dirty_limit; __entry->setpoint = (global_wb_domain.dirty_limit + @@ -726,7 +723,7 @@ TRACE_EVENT(writeback_sb_inodes_requeue, TP_fast_assign( strscpy_pad(__entry->name, - dev_name(inode_to_bdi(inode)->dev), 32); + bdi_dev_name(inode_to_bdi(inode)), 32); __entry->ino = inode->i_ino; __entry->state = inode->i_state; __entry->dirtied_when = inode->dirtied_when; @@ -800,7 +797,7 @@ DECLARE_EVENT_CLASS(writeback_single_inode_template, TP_fast_assign( strscpy_pad(__entry->name, - dev_name(inode_to_bdi(inode)->dev), 32); + bdi_dev_name(inode_to_bdi(inode)), 32); __entry->ino = inode->i_ino; __entry->state = inode->i_state; __entry->dirtied_when = inode->dirtied_when; diff --git a/mm/backing-dev.c b/mm/backing-dev.c index c360f6a6c844..62f05f605fb5 100644 --- a/mm/backing-dev.c +++ b/mm/backing-dev.c @@ -21,6 +21,7 @@ struct backing_dev_info noop_backing_dev_info = { EXPORT_SYMBOL_GPL(noop_backing_dev_info); static struct class *bdi_class; +const char *bdi_unknown_name = "(unknown)"; /* * bdi_lock protects bdi_tree and updates to bdi_list. bdi_list has RCU