public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	stable@vger.kernel.org, Ilya Dryomov <idryomov@redhat.com>,
	Josh Durgin <jdurgin@redhat.com>, Alex Elder <elder@linaro.org>
Subject: [PATCH 3.18 21/57] rbd: fix rbd_dev_parent_get() when parent_overlap == 0
Date: Tue,  3 Feb 2015 15:14:12 -0800	[thread overview]
Message-ID: <20150203231214.989717704@linuxfoundation.org> (raw)
In-Reply-To: <20150203231211.486950145@linuxfoundation.org>

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

------------------

From: Ilya Dryomov <idryomov@redhat.com>

commit ae43e9d05eb4bd324155292f889fbd001c4faea8 upstream.

The comment for rbd_dev_parent_get() said

    * We must get the reference before checking for the overlap to
    * coordinate properly with zeroing the parent overlap in
    * rbd_dev_v2_parent_info() when an image gets flattened.  We
    * drop it again if there is no overlap.

but the "drop it again if there is no overlap" part was missing from
the implementation.  This lead to absurd parent_ref values for images
with parent_overlap == 0, as parent_ref was incremented for each
img_request and virtually never decremented.

Fix this by leveraging the fact that refresh path calls
rbd_dev_v2_parent_info() under header_rwsem and use it for read in
rbd_dev_parent_get(), instead of messing around with atomics.  Get rid
of barriers in rbd_dev_v2_parent_info() while at it - I don't see what
they'd pair with now and I suspect we are in a pretty miserable
situation as far as proper locking goes regardless.

Signed-off-by: Ilya Dryomov <idryomov@redhat.com>
Reviewed-by: Josh Durgin <jdurgin@redhat.com>
Reviewed-by: Alex Elder <elder@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

---
 drivers/block/rbd.c |   20 ++++++--------------
 1 file changed, 6 insertions(+), 14 deletions(-)

--- a/drivers/block/rbd.c
+++ b/drivers/block/rbd.c
@@ -2098,32 +2098,26 @@ static void rbd_dev_parent_put(struct rb
  * If an image has a non-zero parent overlap, get a reference to its
  * parent.
  *
- * We must get the reference before checking for the overlap to
- * coordinate properly with zeroing the parent overlap in
- * rbd_dev_v2_parent_info() when an image gets flattened.  We
- * drop it again if there is no overlap.
- *
  * Returns true if the rbd device has a parent with a non-zero
  * overlap and a reference for it was successfully taken, or
  * false otherwise.
  */
 static bool rbd_dev_parent_get(struct rbd_device *rbd_dev)
 {
-	int counter;
+	int counter = 0;
 
 	if (!rbd_dev->parent_spec)
 		return false;
 
-	counter = atomic_inc_return_safe(&rbd_dev->parent_ref);
-	if (counter > 0 && rbd_dev->parent_overlap)
-		return true;
-
-	/* Image was flattened, but parent is not yet torn down */
+	down_read(&rbd_dev->header_rwsem);
+	if (rbd_dev->parent_overlap)
+		counter = atomic_inc_return_safe(&rbd_dev->parent_ref);
+	up_read(&rbd_dev->header_rwsem);
 
 	if (counter < 0)
 		rbd_warn(rbd_dev, "parent reference overflow");
 
-	return false;
+	return counter > 0;
 }
 
 /*
@@ -4236,7 +4230,6 @@ static int rbd_dev_v2_parent_info(struct
 		 */
 		if (rbd_dev->parent_overlap) {
 			rbd_dev->parent_overlap = 0;
-			smp_mb();
 			rbd_dev_parent_put(rbd_dev);
 			pr_info("%s: clone image has been flattened\n",
 				rbd_dev->disk->disk_name);
@@ -4282,7 +4275,6 @@ static int rbd_dev_v2_parent_info(struct
 	 * treat it specially.
 	 */
 	rbd_dev->parent_overlap = overlap;
-	smp_mb();
 	if (!overlap) {
 
 		/* A null parent_spec indicates it's the initial probe */



  parent reply	other threads:[~2015-02-03 23:50 UTC|newest]

Thread overview: 66+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-02-03 23:13 [PATCH 3.18 00/57] 3.18.6-stable review Greg Kroah-Hartman
2015-02-03 23:13 ` [PATCH 3.18 01/57] x86, build: replace Perl script with Shell script Greg Kroah-Hartman
2015-02-03 23:13 ` [PATCH 3.18 02/57] spi: dw: Fix detecting FIFO depth Greg Kroah-Hartman
2015-02-03 23:13 ` [PATCH 3.18 03/57] spi: dw-mid: fix FIFO size Greg Kroah-Hartman
2015-02-03 23:13 ` [PATCH 3.18 04/57] vm: add VM_FAULT_SIGSEGV handling support Greg Kroah-Hartman
2015-02-10  8:22   ` Konstantin Khlebnikov
2015-02-11  3:43     ` Greg Kroah-Hartman
2015-02-11  3:49       ` Linus Torvalds
2015-02-11  4:16         ` Greg Kroah-Hartman
2015-02-11  5:34       ` Konstantin Khlebnikov
2015-02-16  9:50     ` Luis Henriques
2015-02-03 23:13 ` [PATCH 3.18 05/57] arc: mm: Fix build failure Greg Kroah-Hartman
2015-02-03 23:13 ` [PATCH 3.18 06/57] vm: make stack guard page errors return VM_FAULT_SIGSEGV rather than SIGBUS Greg Kroah-Hartman
2015-02-03 23:13 ` [PATCH 3.18 07/57] ASoC: wm8960: Fix capture sample rate from 11250 to 11025 Greg Kroah-Hartman
2015-02-03 23:13 ` [PATCH 3.18 08/57] ASoC: pcm512x: Fix DSP program selection Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 09/57] ASoC: fsl_esai: Fix incorrect xDC field width of xCCR registers Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 10/57] ASoC: soc-compress.c: fix NULL dereference Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 11/57] ASoC: simple-card: Fix crash in asoc_simple_card_unref() Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 12/57] ASoC: omap-mcbsp: Correct CBM_CFS dai format configuration Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 13/57] udf: Release preallocation on last writeable close Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 14/57] can: kvaser_usb: Do not sleep in atomic context Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 15/57] can: kvaser_usb: Send correct context to URB completion Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 16/57] can: kvaser_usb: Retry the first bulk transfer on -ETIMEDOUT Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 17/57] can: kvaser_usb: Fix state handling upon BUS_ERROR events Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 18/57] powerpc/xmon: Fix another endiannes issue in RTAS call from xmon Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 19/57] ALSA: seq-dummy: remove deadlock-causing events on close Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 20/57] rbd: drop parent_ref in rbd_dev_unprobe() unconditionally Greg Kroah-Hartman
2015-02-03 23:14 ` Greg Kroah-Hartman [this message]
2015-02-03 23:14 ` [PATCH 3.18 22/57] USB: Add OTG PET device to TPL Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 23/57] usb-storage/SCSI: blacklist FUA on JMicron 152d:2566 USB-SATA controller Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 24/57] uas: Add no-report-opcodes quirk for Simpletech devices with id 4971:8017 Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 25/57] i2c: s3c2410: fix ABBA deadlock by keeping clock prepared Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 26/57] Input: synaptics - adjust min/max for Lenovo ThinkPad X1 Carbon 2nd Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 27/57] Input: elantech - add more Fujtisu notebooks to force crc_enabled Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 28/57] Input: i8042 - add noloop quirk for Medion Akoya E7225 (MD98857) Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 29/57] nfs: fix dio deadlock when O_DIRECT flag is flipped Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 30/57] NFSv4.1: Fix an Oops in nfs41_walk_client_list Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 31/57] mac80211: properly set CCK flag in radiotap Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 32/57] mac80211: only roll back station states for WDS when suspending Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 33/57] nl80211: fix per-station group key get/del and memory leak Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 34/57] pinctrl: at91: allow to have disabled gpio bank Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 35/57] ARM: mvebu: dont set the PL310 in I/O coherency mode when I/O coherency is disabled Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 36/57] dm thin: dont allow messages to be sent to a pool target in READ_ONLY or FAIL mode Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 37/57] dm cache: fix missing ERR_PTR returns and handling Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 38/57] drm/vmwgfx: Replace the hw mutex with a hw spinlock Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 41/57] spi/pxa2xx: Clear cur_chip pointer before starting next message Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 42/57] drivers/rtc/rtc-s5m.c: terminate s5m_rtc_id array with empty element Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 43/57] regulator: core: fix race condition in regulator_put() Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 44/57] drivers: net: cpsw: discard dual emac default vlan configuration Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 45/57] drm: fix fb-helper vs MST dangling connector ptrs (v2) Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 46/57] drm/i915: Only fence tiled region of object Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 47/57] drm/i915: BDW Fix Halo PCI IDs marked as ULT Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 48/57] drm/i915: Init PPGTT before context enable Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 49/57] drm/i915: fix inconsistent brightness after resume Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 50/57] quota: Switch ->get_dqblk() and ->set_dqblk() to use bytes as space units Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 51/57] memcg: remove extra newlines from memcg oom kill log Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 52/57] perf/x86/intel: Add model number for Airmont Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 53/57] perf/rapl: Fix crash in rapl_scale() Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 54/57] HID: rmi: Check for additional ACM registers appended to F11 data report Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 55/57] can: c_can: end pending transmission on network stop (ifdown) Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 56/57] clocksource: arch_timer: Only use the virtual counter (CNTVCT) on arm64 Greg Kroah-Hartman
2015-02-03 23:14 ` [PATCH 3.18 57/57] xen/arm/arm64: introduce xen_arch_need_swiotlb Greg Kroah-Hartman
2015-02-04 14:03 ` [PATCH 3.18 00/57] 3.18.6-stable review Guenter Roeck
2015-02-04 19:22   ` Greg Kroah-Hartman
2015-02-04 17:30 ` Shuah Khan
2015-02-04 19:22   ` Greg Kroah-Hartman

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=20150203231214.989717704@linuxfoundation.org \
    --to=gregkh@linuxfoundation.org \
    --cc=elder@linaro.org \
    --cc=idryomov@redhat.com \
    --cc=jdurgin@redhat.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=stable@vger.kernel.org \
    /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