mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Greg KH <gregkh@suse.de>
To: linux-kernel@vger.kernel.org, stable@kernel.org
Cc: stable-review@kernel.org, torvalds@linux-foundation.org,
	akpm@linux-foundation.org, alan@lxorguk.ukuu.org.uk,
	NeilBrown <neilb@suse.de>
Subject: [25/73] md: Fix - again - partition detection when array becomes active
Date: Fri, 04 Mar 2011 16:55:30 -0800	[thread overview]
Message-ID: <20110305005637.029391792@clark.kroah.org> (raw)
In-Reply-To: <20110305005720.GA16162@kroah.com>

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

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

From: NeilBrown <neilb@suse.de>

commit f0b4f7e2f29af678bd9af43422c537dcb6008603 upstream.

Revert
    b821eaa572fd737faaf6928ba046e571526c36c6
and
    f3b99be19ded511a1bf05a148276239d9f13eefa

When I wrote the first of these I had a wrong idea about the
lifetime of 'struct block_device'.  It can disappear at any time that
the block device is not open if it falls out of the inode cache.

So relying on the 'size' recorded with it to detect when the
device size has changed and so we need to revalidate, is wrong.

Rather, we really do need the 'changed' attribute stored directly in
the mddev and set/tested as appropriate.

Without this patch, a sequence of:
   mknod / open / close / unlink

(which can cause a block_device to be created and then destroyed)
will result in a rescan of the partition table and consequence removal
and addition of partitions.
Several of these in a row can get udev racing to create and unlink and
other code can get confused.

With the patch, the rescan is only performed when needed and so there
are no races.

This is suitable for any stable kernel from 2.6.35.

Reported-by: "Wojcik, Krzysztof" <krzysztof.wojcik@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

---
 drivers/md/md.c |   22 +++++++++++++++++++++-
 drivers/md/md.h |    2 ++
 2 files changed, 23 insertions(+), 1 deletion(-)

--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -4614,6 +4614,7 @@ static int do_md_run(mddev_t *mddev)
 	}
 	set_capacity(mddev->gendisk, mddev->array_sectors);
 	revalidate_disk(mddev->gendisk);
+	mddev->changed = 1;
 	kobject_uevent(&disk_to_dev(mddev->gendisk)->kobj, KOBJ_CHANGE);
 out:
 	return err;
@@ -4702,6 +4703,7 @@ static void md_clean(mddev_t *mddev)
 	mddev->sync_speed_min = mddev->sync_speed_max = 0;
 	mddev->recovery = 0;
 	mddev->in_sync = 0;
+	mddev->changed = 0;
 	mddev->degraded = 0;
 	mddev->safemode = 0;
 	mddev->bitmap_info.offset = 0;
@@ -4811,6 +4813,7 @@ static int do_md_stop(mddev_t * mddev, i
 
 		set_capacity(disk, 0);
 		mutex_unlock(&mddev->open_mutex);
+		mddev->changed = 1;
 		revalidate_disk(disk);
 
 		if (mddev->ro)
@@ -5994,7 +5997,7 @@ static int md_open(struct block_device *
 	atomic_inc(&mddev->openers);
 	mutex_unlock(&mddev->open_mutex);
 
-	check_disk_size_change(mddev->gendisk, bdev);
+	check_disk_change(bdev);
  out:
 	return err;
 }
@@ -6009,6 +6012,21 @@ static int md_release(struct gendisk *di
 
 	return 0;
 }
+
+static int md_media_changed(struct gendisk *disk)
+{
+	mddev_t *mddev = disk->private_data;
+
+	return mddev->changed;
+}
+
+static int md_revalidate(struct gendisk *disk)
+{
+	mddev_t *mddev = disk->private_data;
+
+	mddev->changed = 0;
+	return 0;
+}
 static const struct block_device_operations md_fops =
 {
 	.owner		= THIS_MODULE,
@@ -6019,6 +6037,8 @@ static const struct block_device_operati
 	.compat_ioctl	= md_compat_ioctl,
 #endif
 	.getgeo		= md_getgeo,
+	.media_changed  = md_media_changed,
+	.revalidate_disk= md_revalidate,
 };
 
 static int md_thread(void * arg)
--- a/drivers/md/md.h
+++ b/drivers/md/md.h
@@ -270,6 +270,8 @@ struct mddev_s
 	atomic_t			active;		/* general refcount */
 	atomic_t			openers;	/* number of active opens */
 
+	int				changed;	/* True if we might need to
+							 * reread partition info */
 	int				degraded;	/* whether md should consider
 							 * adding a spare
 							 */



  parent reply	other threads:[~2011-03-05  1:00 UTC|newest]

Thread overview: 74+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-03-05  0:57 [00/73] 2.6.37.3-stable review Greg KH
2011-03-05  0:55 ` [01/73] usb: musb: omap2430: fix kernel panic on reboot Greg KH
2011-03-05  0:55 ` [02/73] USB: add quirks entry for Keytouch QWERTY Panel Greg KH
2011-03-05  0:55 ` [03/73] USB: Add Samsung SGH-I500/Android modem ID switch to visor driver Greg KH
2011-03-05  0:55 ` [04/73] USB: Add quirk for Samsung Android phone modem Greg KH
2011-03-05  0:55 ` [05/73] USB: serial/usb_wwan, fix tty NULL dereference Greg KH
2011-03-05  0:55 ` [06/73] sierra: add new ID for Airprime/Sierra USB IP modem Greg KH
2011-03-05  0:55 ` [07/73] Revert "Bluetooth: Enable USB autosuspend by default on btusb" Greg KH
2011-03-05  0:55 ` [08/73] p54pci: update receive dma buffers before and after processing Greg KH
2011-03-05  0:55 ` [09/73] Revert "drm/radeon/kms: switch back to min->max pll post divider iteration" Greg KH
2011-03-05  0:55 ` [10/73] tcp: fix inet_twsk_deschedule() Greg KH
2011-03-05  0:55 ` [11/73] mm: prevent concurrent unmap_mapping_range() on the same inode Greg KH
2011-03-05  0:55 ` [12/73] staging: brcm80211: remove assert to avoid panic since 2.6.37 kernel Greg KH
2011-03-05  0:55 ` [13/73] staging: brcm80211: bugfix for softmac crash on multi cpu configurations Greg KH
2011-03-05  0:55 ` [14/73] staging: usbip: vhci: update reference count for usb_device Greg KH
2011-03-05  0:55 ` [15/73] staging: usbip: vhci: give back URBs from in-flight unlink requests Greg KH
2011-03-05  0:55 ` [16/73] staging: usbip: vhci: refuse to enqueue for dead connections Greg KH
2011-03-05  0:55 ` [17/73] staging: usbip: vhci: use urb->dev->portnum to find port Greg KH
2011-03-05  0:55 ` [18/73] epoll: prevent creating circular epoll structures Greg KH
2011-03-05  0:55 ` [19/73] swiotlb: fix wrong panic Greg KH
2011-03-05  0:55 ` [20/73] ldm: corrupted partition table can cause kernel oops Greg KH
2011-03-05  0:55 ` [21/73] drivers/rtc/rtc-ds3232.c: fix time range difference between linux and RTC chip Greg KH
2011-03-05  0:55 ` [22/73] mm: fix dubious code in __count_immobile_pages() Greg KH
2011-03-05  0:55 ` [23/73] md: correctly handle probe of an mdp device Greg KH
2011-03-05  0:55 ` [24/73] md: avoid spinlock problem in blk_throtl_exit Greg KH
2011-03-05  0:55 ` Greg KH [this message]
2011-03-05  0:55 ` [26/73] Fix over-zealous flush_disk when changing device size Greg KH
2011-03-05  0:55 ` [27/73] PM: Make ACPI wakeup from S5 work again when CONFIG_PM_SLEEP is unset Greg KH
2011-03-05  0:55 ` [28/73] x86 quirk: Fix polarity for IRQ0 pin2 override on SB800 systems Greg KH
2011-03-05  0:55 ` [29/73] xhci: Avoid BUG() in interrupt context Greg KH
2011-03-05  0:55 ` [30/73] xhci: Clarify some expressions in the TRB math Greg KH
2011-03-05  0:55 ` [31/73] xhci: Fix errors in the running total calculations " Greg KH
2011-03-05  0:55 ` [32/73] xhci: Fix an error in count_sg_trbs_needed() Greg KH
2011-03-05  0:55 ` [33/73] USB: Reset USB 3.0 devices on (re)discovery Greg KH
2011-03-05  0:55 ` [34/73] USB: prevent buggy hubs from crashing the USB stack Greg KH
2011-03-05  0:55 ` [35/73] usb: musb: core: set has_tt flag Greg KH
2011-03-05  0:55 ` [36/73] ALSA: HDA: Add a new Conexant codec 506e (20590) Greg KH
2011-03-05  0:55 ` [37/73] ALSA: usb-audio: fix oops due to cleanup race when disconnecting Greg KH
2011-03-05  0:55 ` [38/73] ALSA: HDA: Fix mic initialization in VIA auto parser Greg KH
2011-03-05  0:55 ` [39/73] ALSA: HDA: Add ideapad quirk for two Dell machines Greg KH
2011-03-05  0:55 ` [40/73] ocfs2: Check heartbeat mode for kernel stacks only Greg KH
2011-03-05  0:55 ` [41/73] Ocfs2/refcounttree: Fix a bug for refcounttree to writeback clusters in a right number Greg KH
2011-03-05  0:55 ` [42/73] drm: fix unsigned vs signed comparison issue in modeset ctl ioctl Greg KH
2011-03-05  0:55 ` [43/73] ACPI / debugfs: Fix buffer overflows, double free Greg KH
2011-03-05  0:55 ` [44/73] mfd: Avoid tps6586x burst writes Greg KH
2011-03-05  0:55 ` [45/73] mfd: Fix NULL pointer due to non-initialized ucb1x00-ts absinfo Greg KH
2011-03-05  0:55 ` [46/73] x86: Use u32 instead of long to set reset vector back to 0 Greg KH
2011-03-05  0:55 ` [47/73] Bluetooth: add Atheros BT AR9285 fw supported Greg KH
2011-03-05  0:55 ` [48/73] Bluetooth: fix crash with quirky dongles doing sound Greg KH
2011-03-05  0:55 ` [49/73] Bluetooth: Add Atheros BT AR5BBU12 fw supported Greg KH
2011-03-05  0:55 ` [50/73] eukrea-tlv320: fix platform_name Greg KH
2011-03-05  0:55 ` [51/73] ASoC: correct pxa AC97 DAI names Greg KH
2011-03-05  0:55 ` [52/73] fuse: fix hang of single threaded fuseblk filesystem Greg KH
2011-03-05  0:55 ` [53/73] clockevents: Prevent oneshot mode when broadcast device is periodic Greg KH
2011-03-05  0:55 ` [54/73] ext2: Fix link count corruption under heavy link+rename load Greg KH
2011-03-05  0:56 ` [55/73] mm: vmstat: use a single setter function and callback for adjusting percpu thresholds Greg KH
2011-03-05  0:56 ` [56/73] e1000e: 82579 PHY incorrectly identified during init Greg KH
2011-03-05  0:56 ` [57/73] Staging: comedi: Add MODULE_LICENSE and similar to NI modules Greg KH
2011-03-05  0:56 ` [58/73] fix cfg80211_wext_siwfreq lock ordering Greg KH
2011-03-05  0:56 ` [59/73] tg3: Restrict phy ioctl access Greg KH
2011-03-05  0:56 ` [60/73] drm/i915: fix memory corruption with GM965 and >4GB RAM Greg KH
2011-03-05  0:56 ` [61/73] blk-throttle: Do not use kblockd workqueue for throtl work Greg KH
2011-03-05  0:56 ` [62/73] block: add @force_kblockd to __blk_run_queue() Greg KH
2011-03-05  0:56 ` [63/73] block: blk-flush shouldnt call directly into q->request_fn() __blk_run_queue() Greg KH
2011-03-05  0:56 ` [64/73] block: kill loop_mutex Greg KH
2011-03-05  0:56 ` [65/73] ath9k_htc: Fix an endian issue Greg KH
2011-03-05  0:56 ` [66/73] p54usb: add Senao NUB-350 usbid Greg KH
2011-03-05  0:56 ` [67/73] nilfs2: fix regression that i-flag is not set on changeless checkpoints Greg KH
2011-03-05  0:56 ` [68/73] carl9170: add Airlive X.USB a/b/g/n USBID Greg KH
2011-03-05  0:56 ` [69/73] r8169: disable ASPM Greg KH
2011-03-05  0:56 ` [70/73] dccp: fix oops on Reset after close Greg KH
2011-03-05  0:56 ` [71/73] e1000e: disable broken PHY wakeup for ICH10 LOMs, use MAC wakeup instead Greg KH
2011-03-05  0:56 ` [72/73] DNS: Fix a NULL pointer deref when trying to read an error key [CVE-2011-1076] Greg KH
2011-03-05  0:56 ` [73/73] arp_notify: unconditionally send gratuitous ARP for NETDEV_NOTIFY_PEERS Greg KH

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=20110305005637.029391792@clark.kroah.org \
    --to=gregkh@suse.de \
    --cc=akpm@linux-foundation.org \
    --cc=alan@lxorguk.ukuu.org.uk \
    --cc=linux-kernel@vger.kernel.org \
    --cc=neilb@suse.de \
    --cc=stable-review@kernel.org \
    --cc=stable@kernel.org \
    --cc=torvalds@linux-foundation.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

all inboxes | Powered by JetHome®