public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Chris Wright <chrisw@sous-sol.org>
To: linux-kernel@vger.kernel.org, stable@kernel.org
Cc: Justin Forbes <jmforbes@linuxtx.org>,
	Zwane Mwaikambo <zwane@arm.linux.org.uk>,
	"Theodore Ts'o" <tytso@mit.edu>,
	Randy Dunlap <rdunlap@xenotime.net>,
	Dave Jones <davej@redhat.com>,
	Chuck Wolber <chuckw@quantumlinux.com>,
	Chris Wedgwood <reviews@ml.cw.f00f.org>,
	Michael Krufky <mkrufky@linuxtv.org>,
	torvalds@osdl.org, akpm@osdl.org, alan@lxorguk.ukuu.org.uk,
	Miklos Szeredi <miklos@szeredi.hu>,
	Greg Kroah-Hartman <gregkh@suse.de>
Subject: [PATCH 34/61] fuse: fix hang on SMP
Date: Tue, 31 Oct 2006 21:34:14 -0800	[thread overview]
Message-ID: <20061101054119.573512000@sous-sol.org> (raw)
In-Reply-To: 20061101053340.305569000@sous-sol.org

[-- Attachment #1: fuse-fix-hang-on-smp.patch --]
[-- Type: text/plain, Size: 4810 bytes --]

-stable review patch.  If anyone has any objections, please let us know.
------------------

From: Miklos Szeredi <miklos@szeredi.hu>

Fuse didn't always call i_size_write() with i_mutex held which caused
rare hangs on SMP/32bit.  This bug has been present since fuse-2.2,
well before being merged into mainline.

The simplest solution is to protect i_size_write() with the
per-connection spinlock.  Using i_mutex for this purpose would require
some restructuring of the code and I'm not even sure it's always safe
to acquire i_mutex in all places i_size needs to be set.

Since most of vmtruncate is already duplicated for other reasons,
duplicate the remaining part as well, making all i_size_write() calls
internal to fuse.

Using i_size_write() was unnecessary in fuse_init_inode(), since this
function is only called on a newly created locked inode.

Reported by a few people over the years, but special thanks to Dana
Henriksen who was persistent enough in helping me debug it.

Signed-off-by: Miklos Szeredi <miklos@szeredi.hu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: Chris Wright <chrisw@sous-sol.org>

---
 fs/fuse/dir.c   |   30 +++++++++++++++++++++---------
 fs/fuse/file.c  |   12 +++++++++---
 fs/fuse/inode.c |    5 ++++-
 3 files changed, 34 insertions(+), 13 deletions(-)

--- linux-2.6.18.1.orig/fs/fuse/dir.c
+++ linux-2.6.18.1/fs/fuse/dir.c
@@ -935,14 +935,30 @@ static void iattr_to_fattr(struct iattr 
 	}
 }
 
+static void fuse_vmtruncate(struct inode *inode, loff_t offset)
+{
+	struct fuse_conn *fc = get_fuse_conn(inode);
+	int need_trunc;
+
+	spin_lock(&fc->lock);
+	need_trunc = inode->i_size > offset;
+	i_size_write(inode, offset);
+	spin_unlock(&fc->lock);
+
+	if (need_trunc) {
+		struct address_space *mapping = inode->i_mapping;
+		unmap_mapping_range(mapping, offset + PAGE_SIZE - 1, 0, 1);
+		truncate_inode_pages(mapping, offset);
+	}
+}
+
 /*
  * Set attributes, and at the same time refresh them.
  *
  * Truncation is slightly complicated, because the 'truncate' request
  * may fail, in which case we don't want to touch the mapping.
- * vmtruncate() doesn't allow for this case.  So do the rlimit
- * checking by hand and call vmtruncate() only after the file has
- * actually been truncated.
+ * vmtruncate() doesn't allow for this case, so do the rlimit checking
+ * and the actual truncation by hand.
  */
 static int fuse_setattr(struct dentry *entry, struct iattr *attr)
 {
@@ -993,12 +1009,8 @@ static int fuse_setattr(struct dentry *e
 			make_bad_inode(inode);
 			err = -EIO;
 		} else {
-			if (is_truncate) {
-				loff_t origsize = i_size_read(inode);
-				i_size_write(inode, outarg.attr.size);
-				if (origsize > outarg.attr.size)
-					vmtruncate(inode, outarg.attr.size);
-			}
+			if (is_truncate)
+				fuse_vmtruncate(inode, outarg.attr.size);
 			fuse_change_attributes(inode, &outarg.attr);
 			fi->i_time = time_to_jiffies(outarg.attr_valid,
 						     outarg.attr_valid_nsec);
--- linux-2.6.18.1.orig/fs/fuse/file.c
+++ linux-2.6.18.1/fs/fuse/file.c
@@ -481,8 +481,10 @@ static int fuse_commit_write(struct file
 		err = -EIO;
 	if (!err) {
 		pos += count;
-		if (pos > i_size_read(inode))
+		spin_lock(&fc->lock);
+		if (pos > inode->i_size)
 			i_size_write(inode, pos);
+		spin_unlock(&fc->lock);
 
 		if (offset == 0 && to == PAGE_CACHE_SIZE) {
 			clear_page_dirty(page);
@@ -586,8 +588,12 @@ static ssize_t fuse_direct_io(struct fil
 	}
 	fuse_put_request(fc, req);
 	if (res > 0) {
-		if (write && pos > i_size_read(inode))
-			i_size_write(inode, pos);
+		if (write) {
+			spin_lock(&fc->lock);
+			if (pos > inode->i_size)
+				i_size_write(inode, pos);
+			spin_unlock(&fc->lock);
+		}
 		*ppos = pos;
 	}
 	fuse_invalidate_attr(inode);
--- linux-2.6.18.1.orig/fs/fuse/inode.c
+++ linux-2.6.18.1/fs/fuse/inode.c
@@ -109,6 +109,7 @@ static int fuse_remount_fs(struct super_
 
 void fuse_change_attributes(struct inode *inode, struct fuse_attr *attr)
 {
+	struct fuse_conn *fc = get_fuse_conn(inode);
 	if (S_ISREG(inode->i_mode) && i_size_read(inode) != attr->size)
 		invalidate_inode_pages(inode->i_mapping);
 
@@ -117,7 +118,9 @@ void fuse_change_attributes(struct inode
 	inode->i_nlink   = attr->nlink;
 	inode->i_uid     = attr->uid;
 	inode->i_gid     = attr->gid;
+	spin_lock(&fc->lock);
 	i_size_write(inode, attr->size);
+	spin_unlock(&fc->lock);
 	inode->i_blksize = PAGE_CACHE_SIZE;
 	inode->i_blocks  = attr->blocks;
 	inode->i_atime.tv_sec   = attr->atime;
@@ -131,7 +134,7 @@ void fuse_change_attributes(struct inode
 static void fuse_init_inode(struct inode *inode, struct fuse_attr *attr)
 {
 	inode->i_mode = attr->mode & S_IFMT;
-	i_size_write(inode, attr->size);
+	inode->i_size = attr->size;
 	if (S_ISREG(inode->i_mode)) {
 		fuse_init_common(inode);
 		fuse_init_file_inode(inode);

--

  parent reply	other threads:[~2006-11-01  5:42 UTC|newest]

Thread overview: 91+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2006-11-01  5:33 [PATCH 00/61] 2.6.18-stable review Chris Wright
2006-11-01  5:33 ` [PATCH 01/61] [DECNET]: Fix sfuzz hanging on 2.6.18 Chris Wright
2006-11-01  5:33 ` [PATCH 02/61] splice: fix pipe_to_file() ->prepare_write() error path Chris Wright
2006-11-01  5:33 ` [PATCH 03/61] [S390] __div64_32 for 31 bit Chris Wright
2006-11-01  5:33 ` [PATCH 04/61] mm: fix a race condition under SMC + COW Chris Wright
2006-11-01  5:33 ` [PATCH 05/61] sky2: MSI test race and message Chris Wright
2006-11-01  5:33 ` [PATCH 06/61] sky2: pause parameter adjustment Chris Wright
2006-11-01  5:33 ` [PATCH 07/61] sky2: turn off PHY IRQ on shutdown Chris Wright
2006-11-01  5:33 ` [PATCH 08/61] ALSA: emu10k1: Fix outl() in snd_emu10k1_resume_regs() Chris Wright
2006-11-01  5:33 ` [PATCH 09/61] ALSA: powermac - Fix Oops when conflicting with aoa driver Chris Wright
2006-11-01  5:33 ` [PATCH 10/61] sound/pci/au88x0/au88x0.c: ioremap balanced with iounmap Chris Wright
2006-11-01  5:33 ` [PATCH 11/61] ALSA: Dereference after free in snd_hwdep_release() Chris Wright
2006-11-01  5:33 ` [PATCH 12/61] ALSA: Fix bug in snd-usb-usx2ys usX2Y_pcms_lock_check() Chris Wright
2006-11-01  5:33 ` [PATCH 13/61] ALSA: Repair snd-usb-usx2y for usb 2.6.18 Chris Wright
2006-11-01  5:33 ` [PATCH 14/61] sky2: accept multicast pause frames Chris Wright
2006-11-01  5:33 ` [PATCH 15/61] sky2: GMAC pause frame Chris Wright
2006-11-01  5:33 ` [PATCH 16/61] uml: fix processor selection to exclude unsupported processors and features Chris Wright
2006-11-01  5:33 ` [PATCH 17/61] SCSI: DAC960: PCI id table fixup Chris Wright
2006-11-01  5:33 ` [PATCH 18/61] Fix uninitialised spinlock in via-pmu-backlight code Chris Wright
2006-11-01  5:33 ` [PATCH 19/61] SERIAL: Fix resume handling bug Chris Wright
2006-11-01  5:34 ` [PATCH 20/61] SERIAL: Fix oops when removing suspended serial port Chris Wright
2006-11-01  5:34 ` [PATCH 21/61] Bluetooth: Check if DLC is still attached to the TTY Chris Wright
2006-11-01  5:34 ` [PATCH 22/61] JFS: pageno needs to be long Chris Wright
2006-11-01  5:34 ` [PATCH 23/61] SPARC64: Fix central/FHC bus handling on Ex000 systems Chris Wright
2006-11-01  5:34 ` [PATCH 24/61] SPARC64: Fix memory corruption in pci_4u_free_consistent() Chris Wright
2006-11-01  5:34 ` [PATCH 25/61] bcm43xx: fix watchdog timeouts Chris Wright
2006-11-01 15:13   ` John W. Linville
2006-11-01  5:34 ` [PATCH 26/61] DVB: fix dvb_pll_attach for mt352/zl10353 in cx88-dvb, and nxt200x Chris Wright
2006-11-01  5:34 ` [PATCH 27/61] ALSA: Fix re-use of va_list Chris Wright
2006-11-01  5:34 ` [PATCH 28/61] md: Fix bug where spares dont always get rebuilt properly when they become live Chris Wright
2006-11-01  5:34 ` [PATCH 29/61] md: Fix calculation of ->degraded for multipath and raid10 Chris Wright
2006-11-01  5:34 ` [PATCH 30/61] knfsd: Fix race that can disable NFS server Chris Wright
2006-11-01  7:11   ` Willy Tarreau
2006-11-04 21:06     ` Willy Tarreau
2006-11-05 23:55       ` Neil Brown
2006-11-06  4:11         ` Willy Tarreau
2006-11-01  5:34 ` [PATCH 31/61] SCSI: aic7xxx: avoid checking SBLKCTL register for certain cards Chris Wright
2006-11-01  5:34 ` [PATCH 32/61] IPoIB: Rejoin all multicast groups after a port event Chris Wright
2006-11-01  5:34 ` [PATCH 33/61] IB/mthca: Use mmiowb after doorbell ring Chris Wright
2006-11-01  5:34 ` Chris Wright [this message]
2006-11-01  5:34 ` [PATCH 35/61] Fix potential interrupts during alternative patching Chris Wright
2006-11-01  5:34 ` [PATCH 36/61] sky2: 88E803X transmit lockup (2.6.18) Chris Wright
2006-11-01  5:34 ` [PATCH 37/61] SCSI: aic7xxx: pause sequencer before touching SBLKCTL Chris Wright
2006-11-01  5:34 ` [PATCH 38/61] Audit: fix missing ifdefs in syscall classes hookup for generic targets Chris Wright
2006-11-01  5:34 ` [PATCH 39/61] NET: Fix skb_segment() handling of fully linear SKBs Chris Wright
2006-11-01  5:34 ` [PATCH 40/61] SCTP: Always linearise packet on input Chris Wright
2006-11-01  7:17   ` Willy Tarreau
2006-11-01  6:23     ` David Miller
2006-11-01  5:34 ` [PATCH 41/61] x86-64: Fix C3 timer test Chris Wright
2006-11-01  6:19   ` Len Brown
2006-11-01  5:34 ` [PATCH 42/61] uml: make Uml compile on FC6 kernel headers Chris Wright
2006-11-01  5:34 ` [PATCH 43/61] uml: remove warnings added by previous -stable patch Chris Wright
2006-11-01  5:34 ` [PATCH 44/61] ALSA: snd_rtctimer: handle RTC interrupts with a tasklet Chris Wright
2006-11-01  5:34 ` [PATCH 45/61] Watchdog: sc1200wdt - fix missing pnp_unregister_driver() Chris Wright
2006-11-01  7:45   ` Willy Tarreau
2006-11-01 13:14     ` Wim Van Sebroeck
2006-11-01  5:34 ` [PATCH 46/61] fix Intel RNG detection Chris Wright
2006-11-20 23:45   ` Dave Jones
2006-11-21  2:21     ` [stable] " Chris Wright
2006-11-21  9:32       ` Jan Beulich
2006-11-22  1:50         ` Chris Wright
2006-11-22  7:53           ` Jan Beulich
2006-11-24 20:27             ` Dave Jones
2006-11-27  8:30               ` Jan Beulich
2006-11-29  8:46               ` Jan Beulich
2006-12-13 19:50                 ` dean gaudet
2006-12-13 20:33                   ` Chris Wright
2006-12-13 23:00                     ` dean gaudet
2006-12-14  7:54                       ` Jan Beulich
2006-12-14  8:40                         ` dean gaudet
2006-12-14 10:12                           ` Jan Beulich
2006-11-21  9:05     ` Michael Buesch
2006-11-01  5:34 ` [PATCH 47/61] posix-cpu-timers: prevent signal delivery starvation Chris Wright
2006-11-01  5:34 ` [PATCH 48/61] rtc-max6902: month conversion fix Chris Wright
2006-11-01  5:34 ` [PATCH 49/61] ISDN: fix drivers, by handling errors thrown by ->readstat() Chris Wright
2006-11-01  6:02   ` Jeff Garzik
2006-11-01  7:49   ` Willy Tarreau
2006-11-01  9:18     ` Karsten Keil
2006-11-01  5:34 ` [PATCH 50/61] SPARC64: Fix PCI memory space root resource on Hummingbird Chris Wright
2006-11-01  5:34 ` [PATCH 51/61] PCI: Remove quirk_via_abnormal_poweroff Chris Wright
2006-11-01  6:20   ` Len Brown
2006-11-01  5:34 ` [PATCH 52/61] Reintroduce NODES_SPAN_OTHER_NODES for powerpc Chris Wright
2006-11-01  5:34 ` [PATCH 53/61] NFS: nfs_lookup - dont hash dentry when optimising away the lookup Chris Wright
2006-11-01  5:34 ` [PATCH 54/61] vmscan: Fix temp_priority race Chris Wright
2006-11-01  5:34 ` [PATCH 55/61] Use min of two prio settings in calculating distress for reclaim Chris Wright
2006-11-01  5:34 ` [PATCH 56/61] fill_tgid: fix task_struct leak and possible oops Chris Wright
2006-11-01  5:34 ` [PATCH 57/61] JMB 368 PATA detection Chris Wright
2006-11-01  5:34 ` [PATCH 58/61] tcp: cubic scaling error Chris Wright
2006-11-01  5:34 ` [PATCH 59/61] IPV6: fix lockup via /proc/net/ip6_flowlabel [CVE-2006-5619] Chris Wright
2006-11-01  5:34 ` [PATCH 60/61] md: check bio address after mapping through partitions Chris Wright
2006-11-01  5:34 ` [PATCH 61/61] usbfs: private mutex for open, release, and remove Chris Wright

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20061101054119.573512000@sous-sol.org \
    --to=chrisw@sous-sol.org \
    --cc=akpm@osdl.org \
    --cc=alan@lxorguk.ukuu.org.uk \
    --cc=chuckw@quantumlinux.com \
    --cc=davej@redhat.com \
    --cc=gregkh@suse.de \
    --cc=jmforbes@linuxtx.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=miklos@szeredi.hu \
    --cc=mkrufky@linuxtv.org \
    --cc=rdunlap@xenotime.net \
    --cc=reviews@ml.cw.f00f.org \
    --cc=stable@kernel.org \
    --cc=torvalds@osdl.org \
    --cc=tytso@mit.edu \
    --cc=zwane@arm.linux.org.uk \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox