Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1371023

[PATCH 3.13.y-ckt 39/97] jbd2: fix FS corruption possibility in jbd2_journal_destroy() on umount path

From Kamal Mostafa <kamal@canonical.com>
Newsgroups linux.kernel
Subject [PATCH 3.13.y-ckt 39/97] jbd2: fix FS corruption possibility in jbd2_journal_destroy() on umount path
Date 2016-04-05 01:50 +0200
Message-ID <rkmpR-2lv-47@gated-at.bofh.it> (permalink)
References <rkm6u-2bK-3@gated-at.bofh.it>
Organization linux.* mail to news gateway

Show all headers | View raw

3.13.11-ckt38 -stable review patch.  If anyone has any objections, please let me know.

---8<------------------------------------------------------------

From: OGAWA Hirofumi <hirofumi@mail.parknet.co.jp>

commit c0a2ad9b50dd80eeccd73d9ff962234590d5ec93 upstream.

On umount path, jbd2_journal_destroy() writes latest transaction ID
(->j_tail_sequence) to be used at next mount.

The bug is that ->j_tail_sequence is not holding latest transaction ID
in some cases. So, at next mount, there is chance to conflict with
remaining (not overwritten yet) transactions.

	mount (id=10)
	write transaction (id=11)
	write transaction (id=12)
	umount (id=10) <= the bug doesn't write latest ID

	mount (id=10)
	write transaction (id=11)
	crash

	mount
	[recovery process]
		transaction (id=11)
		transaction (id=12) <= valid transaction ID, but old commit
                                       must not replay

Like above, this bug become the cause of recovery failure, or FS
corruption.

So why ->j_tail_sequence doesn't point latest ID?

Because if checkpoint transactions was reclaimed by memory pressure
(i.e. bdev_try_to_free_page()), then ->j_tail_sequence is not updated.
(And another case is, __jbd2_journal_clean_checkpoint_list() is called
with empty transaction.)

So in above cases, ->j_tail_sequence is not pointing latest
transaction ID at umount path. Plus, REQ_FLUSH for checkpoint is not
done too.

So, to fix this problem with minimum changes, this patch updates
->j_tail_sequence, and issue REQ_FLUSH.  (With more complex changes,
some optimizations would be possible to avoid unnecessary REQ_FLUSH
for example though.)

BTW,

	journal->j_tail_sequence =
		++journal->j_transaction_sequence;

Increment of ->j_transaction_sequence seems to be unnecessary, but
ext3 does this.

Signed-off-by: OGAWA Hirofumi <hirofumi@mail.parknet.co.jp>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Signed-off-by: Kamal Mostafa <kamal@canonical.com>
---
 fs/jbd2/journal.c | 17 ++++++++++++-----
 1 file changed, 12 insertions(+), 5 deletions(-)

diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c
index ef5f39a..84db6c7 100644
--- a/fs/jbd2/journal.c
+++ b/fs/jbd2/journal.c
@@ -1423,11 +1423,12 @@ out:
 /**
  * jbd2_mark_journal_empty() - Mark on disk journal as empty.
  * @journal: The journal to update.
+ * @write_op: With which operation should we write the journal sb
  *
  * Update a journal's dynamic superblock fields to show that journal is empty.
  * Write updated superblock to disk waiting for IO to complete.
  */
-static void jbd2_mark_journal_empty(journal_t *journal)
+static void jbd2_mark_journal_empty(journal_t *journal, int write_op)
 {
 	journal_superblock_t *sb = journal->j_superblock;
 
@@ -1445,7 +1446,7 @@ static void jbd2_mark_journal_empty(journal_t *journal)
 	sb->s_start    = cpu_to_be32(0);
 	read_unlock(&journal->j_state_lock);
 
-	jbd2_write_superblock(journal, WRITE_FUA);
+	jbd2_write_superblock(journal, write_op);
 
 	/* Log is no longer empty */
 	write_lock(&journal->j_state_lock);
@@ -1730,7 +1731,13 @@ int jbd2_journal_destroy(journal_t *journal)
 	if (journal->j_sb_buffer) {
 		if (!is_journal_aborted(journal)) {
 			mutex_lock(&journal->j_checkpoint_mutex);
-			jbd2_mark_journal_empty(journal);
+
+			write_lock(&journal->j_state_lock);
+			journal->j_tail_sequence =
+				++journal->j_transaction_sequence;
+			write_unlock(&journal->j_state_lock);
+
+			jbd2_mark_journal_empty(journal, WRITE_FLUSH_FUA);
 			mutex_unlock(&journal->j_checkpoint_mutex);
 		} else
 			err = -EIO;
@@ -1991,7 +1998,7 @@ int jbd2_journal_flush(journal_t *journal)
 	 * the magic code for a fully-recovered superblock.  Any future
 	 * commits of data to the journal will restore the current
 	 * s_start value. */
-	jbd2_mark_journal_empty(journal);
+	jbd2_mark_journal_empty(journal, WRITE_FUA);
 	mutex_unlock(&journal->j_checkpoint_mutex);
 	write_lock(&journal->j_state_lock);
 	J_ASSERT(!journal->j_running_transaction);
@@ -2037,7 +2044,7 @@ int jbd2_journal_wipe(journal_t *journal, int write)
 	if (write) {
 		/* Lock to make assertions happy... */
 		mutex_lock(&journal->j_checkpoint_mutex);
-		jbd2_mark_journal_empty(journal);
+		jbd2_mark_journal_empty(journal, WRITE_FUA);
 		mutex_unlock(&journal->j_checkpoint_mutex);
 	}
 
-- 
2.7.4

Back to linux.kernel | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread

Thread

[3.13.y-ckt stable] Linux 3.13.11-ckt38 stable review Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:30 +0200
  [PATCH 3.13.y-ckt 26/97] quota: Fix possible GPF due to uninitialised pointers Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:30 +0200
  [PATCH 3.13.y-ckt 68/97] tracing: Fix trace_printk() to print when not using bprintk() Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:30 +0200
  [PATCH 3.13.y-ckt 21/97] Bluetooth: btusb: Add a new AR3012 ID 04ca:3014 Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:30 +0200
  [PATCH 3.13.y-ckt 93/97] paride: make 'verbose' parameter an 'int' again Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:30 +0200
  [PATCH 3.13.y-ckt 08/97] usb: retry reset if a device times out Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:30 +0200
  [PATCH 3.13.y-ckt 09/97] HID: core: do not scan reports if the group is already set Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:30 +0200
  [PATCH 3.13.y-ckt 02/97] EDAC, amd64_edac: Shift wrapping issue in f1x_get_norm_dct_addr() Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:30 +0200
  [PATCH 3.13.y-ckt 41/97] iser-target: Separate flows for np listeners and connections cma events Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:30 +0200
  [PATCH 3.13.y-ckt 86/97] rtc: vr41xx: Wire up alarm_irq_enable Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:30 +0200
  [PATCH 3.13.y-ckt 01/97] [stable-only] pipe: Fix buffer offset after partially failed read Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:30 +0200
  [PATCH 3.13.y-ckt 70/97] rapidio/rionet: fix deadlock on SMP Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:40 +0200
  [PATCH 3.13.y-ckt 50/97] ALSA: hda - Fix unconditional GPIO toggle via automute Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:40 +0200
  [PATCH 3.13.y-ckt 84/97] mtd: map: fix .set_vpp() documentation Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:40 +0200
  [PATCH 3.13.y-ckt 81/97] net/mlx5: Make command timeout way shorter Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:40 +0200
  [PATCH 3.13.y-ckt 72/97] MAINTAINERS: Update mailing list and web page for hwmon subsystem Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:40 +0200
  [PATCH 3.13.y-ckt 79/97] ethernet: micrel: fix some error codes Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:40 +0200
  [PATCH 3.13.y-ckt 69/97] fs/coredump: prevent fsuid=0 dumps into user-controlled directories Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:40 +0200
  [PATCH 3.13.y-ckt 75/97] clk: xgene: Add missing parenthesis when clearing divider value Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:40 +0200
  [PATCH 3.13.y-ckt 96/97] drivers/misc/ad525x_dpot: AD5274 fix RDAC read back errors Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:40 +0200
  [PATCH 3.13.y-ckt 54/97] x86/iopl: Fix iopl capability check on Xen PV Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:40 +0200
  [PATCH 3.13.y-ckt 90/97] ASoC: s3c24xx: use const snd_soc_component_driver pointer Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:40 +0200
  [PATCH 3.13.y-ckt 89/97] lpfc: fix misleading indentation Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:40 +0200
  [PATCH 3.13.y-ckt 94/97] ppp: ensure file->private_data can't be overridden Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:40 +0200
  [PATCH 3.13.y-ckt 83/97] ath9k: fix buffer overrun for ar9287 Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:40 +0200
  [PATCH 3.13.y-ckt 63/97] tracing: Fix crash from reading trace_pipe with sendfile Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:40 +0200
  [PATCH 3.13.y-ckt 65/97] target: Fix target_release_cmd_kref shutdown comp leak Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:40 +0200
  [PATCH 3.13.y-ckt 77/97] mdio-sun4i: oops in error handling in probe Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:40 +0200
  [PATCH 3.13.y-ckt 92/97] nbd: ratelimit error msgs after socket close Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:40 +0200
  [PATCH 3.13.y-ckt 66/97] KVM: VMX: avoid guest hang on invalid invept instruction Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:40 +0200
  [PATCH 3.13.y-ckt 95/97] clk: versatile: sp810: support reentrance Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:40 +0200
  [PATCH 3.13.y-ckt 62/97] tracing: Have preempt(irqs)off trace preempt disabled functions Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:40 +0200
  [PATCH 3.13.y-ckt 82/97] ipvs: correct initial offset of Call-ID header search in SIP persistence engine Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:40 +0200
  [PATCH 3.13.y-ckt 76/97] ppp: take reference on channels netns Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:40 +0200
  [PATCH 3.13.y-ckt 85/97] ARM: OMAP3: Add cpuidle parameters table for omap3430 Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:40 +0200
  [PATCH 3.13.y-ckt 78/97] net: Fix use after free in the recvmmsg exit path Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:40 +0200
  [PATCH 3.13.y-ckt 91/97] kbuild/mkspec: fix grub2 installkernel issue Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:40 +0200
  [PATCH 3.13.y-ckt 88/97] ipv4: fix broadcast packets reception Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:40 +0200
  [PATCH 3.13.y-ckt 97/97] perf stat: Document --detailed option Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:40 +0200
  [PATCH 3.13.y-ckt 61/97] USB: uas: Reduce can_queue to MAX_CMNDS Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:40 +0200
  [PATCH 3.13.y-ckt 74/97] ocfs2/dlm: fix BUG in dlm_move_lockres_to_recovery_list Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:40 +0200
  [PATCH 3.13.y-ckt 64/97] splice: handle zero nr_pages in splice_to_pipe() Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:40 +0200
  [PATCH 3.13.y-ckt 44/97] bus: imx-weim: Take the 'status' property value into account Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:50 +0200
  [PATCH 3.13.y-ckt 40/97] Bluetooth: btusb: Add a new AR3012 ID 13d3:3472 Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:50 +0200
  [PATCH 3.13.y-ckt 24/97] watchdog: rc32434_wdt: fix ioctl error handling Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:50 +0200
  [PATCH 3.13.y-ckt 37/97] drm/radeon: Don't drop DP 2.7 Ghz link setup on some cards. Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:50 +0200
  [PATCH 3.13.y-ckt 43/97] xtensa: clear all DBREAKC registers on start Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:50 +0200
  [PATCH 3.13.y-ckt 33/97] perf/x86/intel: Use PAGE_SIZE for PEBS buffer size on Core2 Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:50 +0200
  [PATCH 3.13.y-ckt 51/97] ALSA: usb-audio: Fix NULL dereference in create_fixed_stream_quirk() Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:50 +0200
  [PATCH 3.13.y-ckt 59/97] USB: usb_driver_claim_interface: add sanity checking Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:50 +0200
  [PATCH 3.13.y-ckt 53/97] nfsd: fix deadlock secinfo+readdir compound Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:50 +0200
  [PATCH 3.13.y-ckt 35/97] x86/PCI: Mark Broadwell-EP Home Agent & PCU as having non-compliant BARs Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:50 +0200
  [PATCH 3.13.y-ckt 23/97] [media] bttv: Width must be a multiple of 16 when capturing planar formats Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:50 +0200
  [PATCH 3.13.y-ckt 52/97] ALSA: usb-audio: Add sanity checks for endpoint accesses Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:50 +0200
  [PATCH 3.13.y-ckt 46/97] s390/pci: enforce fmb page boundary rule Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:50 +0200
  [PATCH 3.13.y-ckt 27/97] mtip32xx: Print exact time when an internal command is interrupted Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:50 +0200
  [PATCH 3.13.y-ckt 57/97] x86/apic: Fix suspicious RCU usage in smp_trace_call_function_interrupt() Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:50 +0200
  [PATCH 3.13.y-ckt 49/97] HID: i2c-hid: fix OOB write in i2c_hid_set_or_send_report() Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:50 +0200
  [PATCH 3.13.y-ckt 34/97] bcache: fix cache_set_flush() NULL pointer dereference on OOM Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:50 +0200
  [PATCH 3.13.y-ckt 45/97] ALSA: intel8x0: Add clock quirk entry for AD1981B on IBM ThinkPad X41. Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:50 +0200
  [PATCH 3.13.y-ckt 39/97] jbd2: fix FS corruption possibility in jbd2_journal_destroy() on umount path Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:50 +0200
  [PATCH 3.13.y-ckt 31/97] pinctrl-bcm2835: Fix cut-and-paste error in "pull" parsing Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:50 +0200
  [PATCH 3.13.y-ckt 56/97] Input: synaptics - handle spurious release of trackstick buttons, again Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:50 +0200
  [PATCH 3.13.y-ckt 55/97] Input: ims-pcu - sanity check against missing interfaces Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:50 +0200
  [PATCH 3.13.y-ckt 28/97] KVM: i8254: change PIT discard tick policy Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:50 +0200
  [PATCH 3.13.y-ckt 25/97] xfs: fix two memory leaks in xfs_attr_list.c error paths Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:50 +0200
  [PATCH 3.13.y-ckt 58/97] USB: iowarrior: fix oops with malicious USB descriptors Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:50 +0200
  [PATCH 3.13.y-ckt 60/97] USB: cdc-acm: more sanity checking Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:50 +0200
  [PATCH 3.13.y-ckt 47/97] Input: powermate - fix oops with malicious USB descriptors Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:50 +0200
  [PATCH 3.13.y-ckt 48/97] net: mvneta: enable change MAC address when interface is up Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:50 +0200
  [PATCH 3.13.y-ckt 38/97] sg: fix dxferp in from_to case Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:50 +0200
  [PATCH 3.13.y-ckt 42/97] xtensa: ISS: don't hang if stdin EOF is reached Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:50 +0200
  [PATCH 3.13.y-ckt 30/97] rt2x00: add new rt2800usb device Buffalo WLI-UC-G450 Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:50 +0200
  [PATCH 3.13.y-ckt 36/97] be2iscsi: set the boot_kset pointer to NULL in case of failure Kamal Mostafa <kamal@canonical.com> - 2016-04-05 01:50 +0200
  [PATCH 3.13.y-ckt 22/97] IB/srpt: Simplify srpt_handle_tsk_mgmt() Kamal Mostafa <kamal@canonical.com> - 2016-04-05 02:00 +0200
  [PATCH 3.13.y-ckt 06/97] 8250: use callbacks to access UART_DLL/UART_DLM Kamal Mostafa <kamal@canonical.com> - 2016-04-05 02:00 +0200
  [PATCH 3.13.y-ckt 05/97] [media] pwc: Add USB id for Philips Spc880nc webcam Kamal Mostafa <kamal@canonical.com> - 2016-04-05 02:00 +0200
  [PATCH 3.13.y-ckt 11/97] scripts/coccinelle: modernize & Kamal Mostafa <kamal@canonical.com> - 2016-04-05 02:00 +0200
  [PATCH 3.13.y-ckt 12/97] [media] adv7511: TX_EDID_PRESENT is still 1 after a disconnect Kamal Mostafa <kamal@canonical.com> - 2016-04-05 02:00 +0200
  [PATCH 3.13.y-ckt 13/97] [media] saa7134: Fix bytesperline not being set correctly for planar formats Kamal Mostafa <kamal@canonical.com> - 2016-04-05 02:00 +0200
  [PATCH 3.13.y-ckt 03/97] tty: Fix GPF in flush_to_ldisc(), part 2 Kamal Mostafa <kamal@canonical.com> - 2016-04-05 02:00 +0200
  [PATCH 3.13.y-ckt 14/97] perf tools: Dont stop PMU parsing on alias parse error Kamal Mostafa <kamal@canonical.com> - 2016-04-05 02:00 +0200
  [PATCH 3.13.y-ckt 20/97] md/raid5: Compare apples to apples (or sectors to sectors) Kamal Mostafa <kamal@canonical.com> - 2016-04-05 02:00 +0200
  [PATCH 3.13.y-ckt 04/97] [media] media: v4l2-compat-ioctl32: fix missing length copy in put_v4l2_buffer32 Kamal Mostafa <kamal@canonical.com> - 2016-04-05 02:00 +0200
  [PATCH 3.13.y-ckt 07/97] net: irda: Fix use-after-free in irtty_open() Kamal Mostafa <kamal@canonical.com> - 2016-04-05 02:00 +0200
  [PATCH 3.13.y-ckt 15/97] Bluetooth: btusb: Add new AR3012 ID 13d3:3395 Kamal Mostafa <kamal@canonical.com> - 2016-04-05 02:00 +0200
  [PATCH 3.13.y-ckt 17/97] aacraid: Fix memory leak in aac_fib_map_free Kamal Mostafa <kamal@canonical.com> - 2016-04-05 02:00 +0200
  [PATCH 3.13.y-ckt 18/97] mtd: onenand: fix deadlock in onenand_block_markbad Kamal Mostafa <kamal@canonical.com> - 2016-04-05 02:00 +0200
  [PATCH 3.13.y-ckt 16/97] Bluetooth: Add new AR3012 ID 0489:e095 Kamal Mostafa <kamal@canonical.com> - 2016-04-05 02:00 +0200
  [PATCH 3.13.y-ckt 19/97] PCI: Disable IO/MEM decoding for devices with non-compliant BARs Kamal Mostafa <kamal@canonical.com> - 2016-04-05 02:00 +0200
  [PATCH 3.13.y-ckt 10/97] HID: fix hid_ignore_special_drivers module parameter Kamal Mostafa <kamal@canonical.com> - 2016-04-05 02:00 +0200

csiph-web