All of lore.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, Nikolay Borisov <nborisov@suse.com>,
	Mathieu Desnoyers <mathieu.desnoyers@efficios.com>,
	David Sterba <dsterba@suse.com>
Subject: [PATCH 4.15 37/52] btrfs: Fix memory barriers usage with device stats counters
Date: Mon, 19 Mar 2018 19:08:35 +0100	[thread overview]
Message-ID: <20180319180737.073560552@linuxfoundation.org> (raw)
In-Reply-To: <20180319180734.976730813@linuxfoundation.org>

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

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

From: Nikolay Borisov <nborisov@suse.com>

commit 9deae9689231964972a94bb56a79b669f9d47ac1 upstream.

Commit addc3fa74e5b ("Btrfs: Fix the problem that the dirty flag of dev
stats is cleared") reworked the way device stats changes are tracked. A
new atomic dev_stats_ccnt counter was introduced which is incremented
every time any of the device stats counters are changed. This serves as
a flag whether there are any pending stats changes. However, this patch
only partially implemented the correct memory barriers necessary:

- It only ordered the stores to the counters but not the reads e.g.
  btrfs_run_dev_stats
- It completely omitted any comments documenting the intended design and
  how the memory barriers pair with each-other

This patch provides the necessary comments as well as adds a missing
smp_rmb in btrfs_run_dev_stats. Furthermore since dev_stats_cnt is only
a snapshot at best there was no point in reading the counter twice -
once in btrfs_dev_stats_dirty and then again when assigning stats_cnt.
Just collapse both reads into 1.

Fixes: addc3fa74e5b ("Btrfs: Fix the problem that the dirty flag of dev stats is cleared")
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

---
 fs/btrfs/volumes.c |   18 ++++++++++++++++--
 fs/btrfs/volumes.h |   12 ++++++++++++
 2 files changed, 28 insertions(+), 2 deletions(-)

--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -7093,10 +7093,24 @@ int btrfs_run_dev_stats(struct btrfs_tra
 
 	mutex_lock(&fs_devices->device_list_mutex);
 	list_for_each_entry(device, &fs_devices->devices, dev_list) {
-		if (!device->dev_stats_valid || !btrfs_dev_stats_dirty(device))
+		stats_cnt = atomic_read(&device->dev_stats_ccnt);
+		if (!device->dev_stats_valid || stats_cnt == 0)
 			continue;
 
-		stats_cnt = atomic_read(&device->dev_stats_ccnt);
+
+		/*
+		 * There is a LOAD-LOAD control dependency between the value of
+		 * dev_stats_ccnt and updating the on-disk values which requires
+		 * reading the in-memory counters. Such control dependencies
+		 * require explicit read memory barriers.
+		 *
+		 * This memory barriers pairs with smp_mb__before_atomic in
+		 * btrfs_dev_stat_inc/btrfs_dev_stat_set and with the full
+		 * barrier implied by atomic_xchg in
+		 * btrfs_dev_stats_read_and_reset
+		 */
+		smp_rmb();
+
 		ret = update_dev_stat_item(trans, fs_info, device);
 		if (!ret)
 			atomic_sub(stats_cnt, &device->dev_stats_ccnt);
--- a/fs/btrfs/volumes.h
+++ b/fs/btrfs/volumes.h
@@ -498,6 +498,12 @@ static inline void btrfs_dev_stat_inc(st
 				      int index)
 {
 	atomic_inc(dev->dev_stat_values + index);
+	/*
+	 * This memory barrier orders stores updating statistics before stores
+	 * updating dev_stats_ccnt.
+	 *
+	 * It pairs with smp_rmb() in btrfs_run_dev_stats().
+	 */
 	smp_mb__before_atomic();
 	atomic_inc(&dev->dev_stats_ccnt);
 }
@@ -523,6 +529,12 @@ static inline void btrfs_dev_stat_set(st
 				      int index, unsigned long val)
 {
 	atomic_set(dev->dev_stat_values + index, val);
+	/*
+	 * This memory barrier orders stores updating statistics before stores
+	 * updating dev_stats_ccnt.
+	 *
+	 * It pairs with smp_rmb() in btrfs_run_dev_stats().
+	 */
 	smp_mb__before_atomic();
 	atomic_inc(&dev->dev_stats_ccnt);
 }

  parent reply	other threads:[~2018-03-19 18:32 UTC|newest]

Thread overview: 62+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-03-19 18:07 [PATCH 4.15 00/52] 4.15.12-stable review Greg Kroah-Hartman
2018-03-19 18:07 ` [PATCH 4.15 01/52] x86/cpufeatures: Add Intel Total Memory Encryption cpufeature Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 02/52] x86/cpufeatures: Add Intel PCONFIG cpufeature Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 03/52] selftests/x86/entry_from_vm86: Exit with 1 if we fail Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 04/52] selftests/x86/entry_from_vm86: Add test cases for POPF Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 05/52] x86/vm86/32: Fix POPF emulation Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 06/52] x86/speculation, objtool: Annotate indirect calls/jumps for objtool on 32-bit kernels Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 07/52] x86/speculation: Remove Skylake C2 from Speculation Control microcode blacklist Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 08/52] KVM: x86: Fix device passthrough when SME is active Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 09/52] x86/mm: Fix vmalloc_fault to use pXd_large Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 10/52] parisc: Handle case where flush_cache_range is called with no context Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 11/52] ALSA: pcm: Fix UAF in snd_pcm_oss_get_formats() Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 12/52] ALSA: hda - Revert power_save option default value Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 13/52] ALSA: seq: Fix possible UAF in snd_seq_check_queue() Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 14/52] ALSA: seq: Clear client entry before deleting else at closing Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 15/52] drm/nouveau/bl: Fix oops on driver unbind Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 16/52] drm/nouveau/mmu: ALIGN_DOWN correct variable Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 17/52] drm/amdgpu: fix prime teardown order Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 18/52] drm/radeon: " Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 19/52] drm/amdgpu/dce: Dont turn off DP sink when disconnected Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 20/52] fs: Teach path_connected to handle nfs filesystems with multiple roots Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 21/52] KVM: arm/arm64: Reduce verbosity of KVM init log Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 22/52] KVM: arm/arm64: Reset mapped IRQs on VM reset Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 23/52] kvm: arm/arm64: vgic-v3: Tighten synchronization for guests using v2 on v3 Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 24/52] KVM: arm/arm64: vgic: Dont populate multiple LRs with the same vintid Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 25/52] lock_parent() needs to recheck if dentry got __dentry_killed under it Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 26/52] fs/aio: Add explicit RCU grace period when freeing kioctx Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 27/52] fs/aio: Use RCU accessors for kioctx_table->table[] Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 28/52] RDMAVT: Fix synchronization around percpu_ref Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 29/52] irqchip/gic-v3-its: Ensure nr_ites >= nr_lpis Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 30/52] nvme: fix subsystem multiple controllers support check Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 31/52] xfs: preserve i_rdev when recycling a reclaimable inode Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 32/52] btrfs: Fix NULL pointer exception in find_bio_stripe Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 33/52] btrfs: add missing initialization in btrfs_check_shared Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 34/52] btrfs: alloc_chunk: fix DUP stripe size handling Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 35/52] btrfs: Fix use-after-free when cleaning up fs_devs with a single stale device Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 36/52] btrfs: remove spurious WARN_ON(ref->count < 0) in find_parent_nodes Greg Kroah-Hartman
2018-03-19 18:08 ` Greg Kroah-Hartman [this message]
2018-03-19 18:08 ` [PATCH 4.15 38/52] scsi: qla2xxx: Fix smatch warning in qla25xx_delete_{rsp|req}_que Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 39/52] scsi: qla2xxx: Fix NULL pointer access for fcport structure Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 40/52] scsi: qla2xxx: Fix logo flag for qlt_free_session_done() Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 41/52] scsi: qla2xxx: Fix crashes in qla2x00_probe_one on probe failure Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 42/52] usb: dwc2: fix STM32F7 USB OTG HS compatible Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 43/52] dt-bindings: usb: fix the STM32F7 DWC2 OTG HS core binding Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 44/52] USB: gadget: udc: Add missing platform_device_put() on error in bdc_pci_probe() Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 45/52] usb: dwc3: Fix GDBGFIFOSPACE_TYPE values Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 46/52] usb: dwc3: core: Power-off core/PHYs on system_suspend in host mode Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 47/52] usb: dwc3: of-simple: fix oops by unbalanced clk disable call Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 48/52] usb: gadget: udc: renesas_usb3: fix oops in renesas_usb3_remove() Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 49/52] phy: phy-brcm-usb: Fix two DT properties to match bindings doc Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 50/52] phy: phy-brcm-usb-init: Some Low Speed keyboards fail on 7271 Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 51/52] phy: phy-brcm-usb-init: DRD mode can cause crash on startup Greg Kroah-Hartman
2018-03-19 18:08 ` [PATCH 4.15 52/52] phy: phy-brcm-usb-init: Power down USB 3.0 PHY when XHCI disabled Greg Kroah-Hartman
2018-03-19 23:20 ` [PATCH 4.15 00/52] 4.15.12-stable review kernelci.org bot
2018-03-20  6:47 ` Naresh Kamboju
2018-03-20  7:44   ` Greg Kroah-Hartman
2018-03-20 12:24 ` Thadeu Lima de Souza Cascardo
2018-03-21 11:07   ` Greg Kroah-Hartman
2018-03-20 16:11 ` Guenter Roeck
2018-03-21 12:29   ` Greg Kroah-Hartman
2018-03-20 20:11 ` Shuah Khan
2018-03-21  9:42   ` 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=20180319180737.073560552@linuxfoundation.org \
    --to=gregkh@linuxfoundation.org \
    --cc=dsterba@suse.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mathieu.desnoyers@efficios.com \
    --cc=nborisov@suse.com \
    --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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.