[v4,19/19] fs: handle inode->i_version more efficiently
diff mbox

Message ID 1515418164.3486.18.camel@kernel.org
State New
Headers show

Commit Message

Jeff Layton Jan. 8, 2018, 1:29 p.m. UTC
On Mon, 2018-01-08 at 14:21 +0100, Krzysztof Kozlowski wrote:
> On Mon, Jan 8, 2018 at 1:56 PM, Jeff Layton <jlayton@kernel.org> wrote:
> > On Sun, 2018-01-07 at 13:44 +0100, Krzysztof Kozlowski wrote:
> > > On Fri, Dec 22, 2017 at 1:05 PM, Jeff Layton <jlayton@kernel.org> wrote:
> > > > From: Jeff Layton <jlayton@redhat.com>
> > > > 
> > > > Since i_version is mostly treated as an opaque value, we can exploit that
> > > > fact to avoid incrementing it when no one is watching. With that change,
> > > > we can avoid incrementing the counter on writes, unless someone has
> > > > queried for it since it was last incremented. If the a/c/mtime don't
> > > > change, and the i_version hasn't changed, then there's no need to dirty
> > > > the inode metadata on a write.
> > > > 
> > > > Convert the i_version counter to an atomic64_t, and use the lowest order
> > > > bit to hold a flag that will tell whether anyone has queried the value
> > > > since it was last incremented.
> > > > 
> > > > When we go to maybe increment it, we fetch the value and check the flag
> > > > bit.  If it's clear then we don't need to do anything if the update
> > > > isn't being forced.
> > > > 
> > > > If we do need to update, then we increment the counter by 2, and clear
> > > > the flag bit, and then use a CAS op to swap it into place. If that
> > > > works, we return true. If it doesn't then do it again with the value
> > > > that we fetch from the CAS operation.
> > > > 
> > > > On the query side, if the flag is already set, then we just shift the
> > > > value down by 1 bit and return it. Otherwise, we set the flag in our
> > > > on-stack value and again use cmpxchg to swap it into place if it hasn't
> > > > changed. If it has, then we use the value from the cmpxchg as the new
> > > > "old" value and try again.
> > > > 
> > > > This method allows us to avoid incrementing the counter on writes (and
> > > > dirtying the metadata) under typical workloads. We only need to increment
> > > > if it has been queried since it was last changed.
> > > > 
> > > > Signed-off-by: Jeff Layton <jlayton@redhat.com>
> > > > ---
> > > >  include/linux/fs.h       |   2 +-
> > > >  include/linux/iversion.h | 208 ++++++++++++++++++++++++++++++++++-------------
> > > >  2 files changed, 154 insertions(+), 56 deletions(-)
> > > > 
> > > 
> > > Hi,
> > > 
> > > On recent linux-next my ARM/Exynos boards fail to boot over nfsroot.
> > > This commit popped up through bisect (log at the end). Systemd
> > > timeouts on some device-specific services, including mounting ext4
> > > /home:
> > > 
> > > [ *** ] (1 of 4) A start job is running for…ress polling (1min 41s / no limit)
> > > [ TIME ] Timed out waiting for device dev-ttySAC2.device.
> > > Jan 07 13:29:38 [DEPEND] Dependency failed for Serial Getty on ttySAC2.
> > > Jan 07 13:29:38 [ TIME ] Timed out waiting for device
> > > dev-disk-by\x2dlabel-home.device.
> > > Jan 07 13:29:38 [DEPEND] Dependency failed for /home.
> > > Jan 07 13:29:38 [DEPEND] Dependency failed for Local File Systems.
> > > Jan 07 13:29:38 [DEPEND] Dependency failed for File System Check on
> > > /dev/disk/by-label/home.
> > > Jan 07 13:30:02 [ *** ] (1 of 2) A start job is running for…ress
> > > polling (1min 53s / no limit)
> > > 
> > > Kernel command line:
> > > console=tty1 console=ttySAC2,115200n8
> > > ip=192.168.1.11:192.168.1.10:192.168.1.1:255.255.255.0::eth0:none
> > > nfsrootdebug root=/dev/nfs
> > > nfsroot=192.168.1.10:/srv/nfs/odroidxu3,vers=4,nolock rootwait rw
> > > smsc95xx.macaddr=00:1e:06:61:7a:93 no_console_suspend
> > > 
> > > /home is /dev/mmcblk1p2:
> > > kozik@odroidxu3:~$ tune2fs -l /dev/mmcblk1p2
> > > tune2fs 1.43.7 (16-Oct-2017)
> > > Filesystem volume name:   home
> > > Last mounted on:          /home
> > > Filesystem UUID:          3f9dbeba-2738-45d3-807e-c1b2e21128ed
> > > Filesystem magic number:  0xEF53
> > > Filesystem revision #:    1 (dynamic)
> > > Filesystem features:      has_journal ext_attr resize_inode dir_index
> > > filetype needs_recovery extent flex_bg sparse_super large_file
> > > uninit_bg dir_nlink extra_isize
> > > Filesystem flags:         signed_directory_hash
> > > Default mount options:    user_xattr acl
> > > Filesystem state:         clean
> > > Errors behavior:          Continue
> > > Filesystem OS type:       Linux
> > > Inode count:              1430800
> > > Block count:              5717760
> > > Reserved block count:     285888
> > > Free blocks:              5467576
> > > Free inodes:              1428301
> > > First block:              0
> > > Block size:               4096
> > > Fragment size:            4096
> > > Reserved GDT blocks:      1022
> > > Blocks per group:         32768
> > > Fragments per group:      32768
> > > Inodes per group:         8176
> > > Inode blocks per group:   511
> > > Flex block group size:    16
> > > Filesystem created:       Thu May 21 12:17:05 2015
> > > Last mount time:          Thu Dec 21 13:31:26 2017
> > > Last write time:          Thu Dec 21 13:31:26 2017
> > > Mount count:              1
> > > Maximum mount count:      -1
> > > Last checked:             Thu Dec 21 13:31:25 2017
> > > Check interval:           0 (<none>)
> > > Lifetime writes:          126 GB
> > > Reserved blocks uid:      0 (user root)
> > > Reserved blocks gid:      0 (group root)
> > > First inode:              11
> > > Inode size:           256
> > > Required extra isize:     28
> > > Desired extra isize:      28
> > > Journal inode:            8
> > > Default directory hash:   half_md4
> > > Directory Hash Seed:      42e17e23-86b2-4356-ad63-78aa51651d03
> > > Journal backup:           inode blocks
> > > 
> > > 
> > > Full dmesg log:
> > > http://www.krzk.eu/#/builders/1/builds/1258/steps/10/logs/serial0
> > > 
> > > The regular boot from rootfs on SD card also fails - but without any
> > > serial console logs (just "Starting kernel...") so the real cause is
> > > unknown.
> > > 
> > > Any hints?
> > > 
> > > Best regards,
> > > Krzysztof
> > > 
> > > 
> > > bisect log:
> > > # bad: [73005e1a35fd67c644b0645c9e4c1efabd0fe62c] Add linux-next
> > > specific files for 20180103
> > > # good: [30a7acd573899fd8b8ac39236eff6468b195ac7d] Linux 4.15-rc6
> > > git bisect start 'next/master' 'next/stable'
> > > # bad: [c1d290f9ce8daa2b0a79d2fe48c1b7c3c5370f5a] Merge
> > > remote-tracking branch 'crypto/master'
> > > git bisect bad c1d290f9ce8daa2b0a79d2fe48c1b7c3c5370f5a
> > > # bad: [55695d94f0915121d106cd2d1ab94983a32f3e9a] Merge
> > > remote-tracking branch 'hid/for-next'
> > > git bisect bad 55695d94f0915121d106cd2d1ab94983a32f3e9a
> > > # good: [cffae1eead0dd91be1a3069a8348127bb00158f3] Merge
> > > remote-tracking branch 'realtek/for-next'
> > > git bisect good cffae1eead0dd91be1a3069a8348127bb00158f3
> > > # good: [5f889f1176dc99636c6bf8af7c286decc888c007] Merge
> > > remote-tracking branch 'btrfs/next'
> > > git bisect good 5f889f1176dc99636c6bf8af7c286decc888c007
> > > # good: [984c35877f36bee305e43a1c58176169854d85cf] Merge
> > > remote-tracking branch 'xfs/for-next'
> > > git bisect good 984c35877f36bee305e43a1c58176169854d85cf
> > > # bad: [f9fec502daea2a869232b6dff33ba3de79dd0d61] Merge
> > > remote-tracking branch 'printk/for-next'
> > > git bisect bad f9fec502daea2a869232b6dff33ba3de79dd0d61
> > > # good: [c71d227fc4133f949dae620ed5e3a250b43f2415] make kernel-side
> > > POLL... arch-independent
> > > git bisect good c71d227fc4133f949dae620ed5e3a250b43f2415
> > > # good: [416d20e8c31107f5dfd45d1d80d1e6c8e4871180] Merge branches
> > > 'work.get_user_pages_fast', 'work.wmci', 'work.sock_recvmsg',
> > > 'misc.netdrv', 'misc.poll', 'work.mqueue', 'work.whack-a-mole' and
> > > 'work.misc' into for-next
> > > git bisect good 416d20e8c31107f5dfd45d1d80d1e6c8e4871180
> > > # good: [325a1de4a691512a48c1426b943a7b0b9f8d6744] xfs: convert to new
> > > i_version API
> > > git bisect good 325a1de4a691512a48c1426b943a7b0b9f8d6744
> > > # good: [a94fe10fb114c169e7ddaecd8251521886409121] checkpatch: add
> > > pF/pf deprecation warning
> > > git bisect good a94fe10fb114c169e7ddaecd8251521886409121
> > > # good: [6b3911dffd1184fdcd63299a5fee59ac000f2067] btrfs: only dirty
> > > the inode in btrfs_update_time if something was changed
> > > git bisect good 6b3911dffd1184fdcd63299a5fee59ac000f2067
> > > # bad: [448f8c749a7a0ae03505823910ec45a112678048] Merge
> > > remote-tracking branch 'iversion/iversion-next'
> > > git bisect bad 448f8c749a7a0ae03505823910ec45a112678048
> > > # bad: [8618bff776758ebff5b55211e7b5a60a0fc119a5] fs: handle
> > > inode->i_version more efficiently
> > > git bisect bad 8618bff776758ebff5b55211e7b5a60a0fc119a5
> > > # first bad commit: [8618bff776758ebff5b55211e7b5a60a0fc119a5] fs:
> > > handle inode->i_version more efficiently
> > 
> > That's really strange. I'm afraid I have no idea what could be going on.
> > 
> > With NFS, we really just treat i_version as an opaque value, so I'm not
> > sure how this patch in particular would affect anything there. We _do_
> > increment it if you have a write delegation in some cases, but not many
> > servers hand those out.
> 
> About the NFS server, Arch Linux on Raspberry Pi (so 32-bit, ARMv6):
> Linux pi 4.9.70-1-ARCH #1 SMP Mon Dec 18 19:38:00 UTC 2017 armv6l GNU/Linux
> 
> The /etc/nfs.conf is default except:
> [nfsd]
> vers2=n
> vers3=n
> 
> The client logs for nfsroot mounts are:
> Jan 08 14:07:25 :: running hook [net_nfs4]
> Jan 08 14:07:25 IP-Config: eth0 hardware address ba:17:70:7e:87:d1 mtu 1500
> Jan 08 14:07:25 IP-Config: eth0 guessed broadcast address 192.168.1.255
> Jan 08 14:07:25 IP-Config: eth0 complete (from 192.168.1.10):
> Jan 08 14:07:25 address: 192.168.1.11 broadcast: 192.168.1.255
> netmask: 255.255.255.0
> Jan 08 14:07:25 gateway: 192.168.1.1 dns0 : 0.0.0.0 dns1 : 0.0.0.0
> Jan 08 14:07:25 rootserver: 192.168.1.10 rootpath:
> Jan 08 14:07:25 filename :
> Jan 08 14:07:25 NFS-Mount: 192.168.1.10:/srv/nfs/odroidxu3
> Jan 08 14:07:25 Waiting 10 seconds for device /dev/nfs ...
> Jan 08 14:07:36 ERROR: device '/dev/nfs' not found. Skipping fsck.
> Jan 08 14:07:36 Mount cmd:
> Jan 08 14:07:36 /opt/tools/buildbot/arch-arm-bin/mount.nfs4 -o
> vers=4,nolock 192.168.1.10:/srv/nfs/odroidxu3 /new_root
> 
> Only root (/) is froom NFS. The /home comes from sdcard (/dev/mmcblk1).
> 
> > ext4 will only touch the i_version field if you mount it with '-o
> > iversion'. Are you doing that here?
> 
> The /home is mounted by systemd from /etc/fstab:
>       LABEL=home /home ext4 defaults 0 2
> 
> > 
> > Have you run the bisect more than once? Is this maybe an intermittent
> > problem, and the bisect has landed on the wrong commit?
> 
> I just run tests on commits around again (but not full bisect) on
> current next (next-20180108):
> fbf97ece47d66d22 - works
> 3da7bcdd695bae43 ("fs: handle inode->i_version more efficiently") -
> fails: http://www.krzk.eu/#/builders/1/builds/1267
> 
> 
> Best regards,
> Krzysztof

Ok, thanks. If you're seeing hangs then that might imply that we have
some sort of excessive looping going on in the cmpxchg loops.

Could you apply the patch below and let me know if it causes either of
the warnings to pop? That might at least point us in the right
direction:

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

[PATCH] DEBUG: throw warning if we are looping excessively in new
 iversion code

Signed-off-by: Jeff Layton <jlayton@redhat.com>
---
 include/linux/iversion.h | 6 ++++++
 1 file changed, 6 insertions(+)

Comments

Krzysztof Kozlowski Jan. 8, 2018, 5:29 p.m. UTC | #1
On Mon, Jan 08, 2018 at 08:29:24AM -0500, Jeff Layton wrote:
> On Mon, 2018-01-08 at 14:21 +0100, Krzysztof Kozlowski wrote:
> > On Mon, Jan 8, 2018 at 1:56 PM, Jeff Layton <jlayton@kernel.org> wrote:
> > > On Sun, 2018-01-07 at 13:44 +0100, Krzysztof Kozlowski wrote:
> > > > On Fri, Dec 22, 2017 at 1:05 PM, Jeff Layton <jlayton@kernel.org> wrote:
> > > > > From: Jeff Layton <jlayton@redhat.com>
> > > > > 
> > > > > Since i_version is mostly treated as an opaque value, we can exploit that
> > > > > fact to avoid incrementing it when no one is watching. With that change,
> > > > > we can avoid incrementing the counter on writes, unless someone has
> > > > > queried for it since it was last incremented. If the a/c/mtime don't
> > > > > change, and the i_version hasn't changed, then there's no need to dirty
> > > > > the inode metadata on a write.
> > > > > 
> > > > > Convert the i_version counter to an atomic64_t, and use the lowest order
> > > > > bit to hold a flag that will tell whether anyone has queried the value
> > > > > since it was last incremented.
> > > > > 
> > > > > When we go to maybe increment it, we fetch the value and check the flag
> > > > > bit.  If it's clear then we don't need to do anything if the update
> > > > > isn't being forced.
> > > > > 
> > > > > If we do need to update, then we increment the counter by 2, and clear
> > > > > the flag bit, and then use a CAS op to swap it into place. If that
> > > > > works, we return true. If it doesn't then do it again with the value
> > > > > that we fetch from the CAS operation.
> > > > > 
> > > > > On the query side, if the flag is already set, then we just shift the
> > > > > value down by 1 bit and return it. Otherwise, we set the flag in our
> > > > > on-stack value and again use cmpxchg to swap it into place if it hasn't
> > > > > changed. If it has, then we use the value from the cmpxchg as the new
> > > > > "old" value and try again.
> > > > > 
> > > > > This method allows us to avoid incrementing the counter on writes (and
> > > > > dirtying the metadata) under typical workloads. We only need to increment
> > > > > if it has been queried since it was last changed.
> > > > > 
> > > > > Signed-off-by: Jeff Layton <jlayton@redhat.com>
> > > > > ---
> > > > >  include/linux/fs.h       |   2 +-
> > > > >  include/linux/iversion.h | 208 ++++++++++++++++++++++++++++++++++-------------
> > > > >  2 files changed, 154 insertions(+), 56 deletions(-)
> > > > > 
> > > > 
> > > > Hi,
> > > > 
> > > > On recent linux-next my ARM/Exynos boards fail to boot over nfsroot.
> > > > This commit popped up through bisect (log at the end). Systemd
> > > > timeouts on some device-specific services, including mounting ext4
> > > > /home:
> > > > 
> > > > [ *** ] (1 of 4) A start job is running for…ress polling (1min 41s / no limit)
> > > > [ TIME ] Timed out waiting for device dev-ttySAC2.device.
> > > > Jan 07 13:29:38 [DEPEND] Dependency failed for Serial Getty on ttySAC2.
> > > > Jan 07 13:29:38 [ TIME ] Timed out waiting for device
> > > > dev-disk-by\x2dlabel-home.device.
> > > > Jan 07 13:29:38 [DEPEND] Dependency failed for /home.
> > > > Jan 07 13:29:38 [DEPEND] Dependency failed for Local File Systems.
> > > > Jan 07 13:29:38 [DEPEND] Dependency failed for File System Check on
> > > > /dev/disk/by-label/home.
> > > > Jan 07 13:30:02 [ *** ] (1 of 2) A start job is running for…ress
> > > > polling (1min 53s / no limit)
> > > > 
> > > > Kernel command line:
> > > > console=tty1 console=ttySAC2,115200n8
> > > > ip=192.168.1.11:192.168.1.10:192.168.1.1:255.255.255.0::eth0:none
> > > > nfsrootdebug root=/dev/nfs
> > > > nfsroot=192.168.1.10:/srv/nfs/odroidxu3,vers=4,nolock rootwait rw
> > > > smsc95xx.macaddr=00:1e:06:61:7a:93 no_console_suspend
> > > > 
> > > > /home is /dev/mmcblk1p2:
> > > > kozik@odroidxu3:~$ tune2fs -l /dev/mmcblk1p2
> > > > tune2fs 1.43.7 (16-Oct-2017)
> > > > Filesystem volume name:   home
> > > > Last mounted on:          /home
> > > > Filesystem UUID:          3f9dbeba-2738-45d3-807e-c1b2e21128ed
> > > > Filesystem magic number:  0xEF53
> > > > Filesystem revision #:    1 (dynamic)
> > > > Filesystem features:      has_journal ext_attr resize_inode dir_index
> > > > filetype needs_recovery extent flex_bg sparse_super large_file
> > > > uninit_bg dir_nlink extra_isize
> > > > Filesystem flags:         signed_directory_hash
> > > > Default mount options:    user_xattr acl
> > > > Filesystem state:         clean
> > > > Errors behavior:          Continue
> > > > Filesystem OS type:       Linux
> > > > Inode count:              1430800
> > > > Block count:              5717760
> > > > Reserved block count:     285888
> > > > Free blocks:              5467576
> > > > Free inodes:              1428301
> > > > First block:              0
> > > > Block size:               4096
> > > > Fragment size:            4096
> > > > Reserved GDT blocks:      1022
> > > > Blocks per group:         32768
> > > > Fragments per group:      32768
> > > > Inodes per group:         8176
> > > > Inode blocks per group:   511
> > > > Flex block group size:    16
> > > > Filesystem created:       Thu May 21 12:17:05 2015
> > > > Last mount time:          Thu Dec 21 13:31:26 2017
> > > > Last write time:          Thu Dec 21 13:31:26 2017
> > > > Mount count:              1
> > > > Maximum mount count:      -1
> > > > Last checked:             Thu Dec 21 13:31:25 2017
> > > > Check interval:           0 (<none>)
> > > > Lifetime writes:          126 GB
> > > > Reserved blocks uid:      0 (user root)
> > > > Reserved blocks gid:      0 (group root)
> > > > First inode:              11
> > > > Inode size:           256
> > > > Required extra isize:     28
> > > > Desired extra isize:      28
> > > > Journal inode:            8
> > > > Default directory hash:   half_md4
> > > > Directory Hash Seed:      42e17e23-86b2-4356-ad63-78aa51651d03
> > > > Journal backup:           inode blocks
> > > > 
> > > > 
> > > > Full dmesg log:
> > > > http://www.krzk.eu/#/builders/1/builds/1258/steps/10/logs/serial0
> > > > 
> > > > The regular boot from rootfs on SD card also fails - but without any
> > > > serial console logs (just "Starting kernel...") so the real cause is
> > > > unknown.
> > > > 
> > > > Any hints?
> > > > 
> > > > Best regards,
> > > > Krzysztof
> > > > 
> > > > 
> > > > bisect log:
> > > > # bad: [73005e1a35fd67c644b0645c9e4c1efabd0fe62c] Add linux-next
> > > > specific files for 20180103
> > > > # good: [30a7acd573899fd8b8ac39236eff6468b195ac7d] Linux 4.15-rc6
> > > > git bisect start 'next/master' 'next/stable'
> > > > # bad: [c1d290f9ce8daa2b0a79d2fe48c1b7c3c5370f5a] Merge
> > > > remote-tracking branch 'crypto/master'
> > > > git bisect bad c1d290f9ce8daa2b0a79d2fe48c1b7c3c5370f5a
> > > > # bad: [55695d94f0915121d106cd2d1ab94983a32f3e9a] Merge
> > > > remote-tracking branch 'hid/for-next'
> > > > git bisect bad 55695d94f0915121d106cd2d1ab94983a32f3e9a
> > > > # good: [cffae1eead0dd91be1a3069a8348127bb00158f3] Merge
> > > > remote-tracking branch 'realtek/for-next'
> > > > git bisect good cffae1eead0dd91be1a3069a8348127bb00158f3
> > > > # good: [5f889f1176dc99636c6bf8af7c286decc888c007] Merge
> > > > remote-tracking branch 'btrfs/next'
> > > > git bisect good 5f889f1176dc99636c6bf8af7c286decc888c007
> > > > # good: [984c35877f36bee305e43a1c58176169854d85cf] Merge
> > > > remote-tracking branch 'xfs/for-next'
> > > > git bisect good 984c35877f36bee305e43a1c58176169854d85cf
> > > > # bad: [f9fec502daea2a869232b6dff33ba3de79dd0d61] Merge
> > > > remote-tracking branch 'printk/for-next'
> > > > git bisect bad f9fec502daea2a869232b6dff33ba3de79dd0d61
> > > > # good: [c71d227fc4133f949dae620ed5e3a250b43f2415] make kernel-side
> > > > POLL... arch-independent
> > > > git bisect good c71d227fc4133f949dae620ed5e3a250b43f2415
> > > > # good: [416d20e8c31107f5dfd45d1d80d1e6c8e4871180] Merge branches
> > > > 'work.get_user_pages_fast', 'work.wmci', 'work.sock_recvmsg',
> > > > 'misc.netdrv', 'misc.poll', 'work.mqueue', 'work.whack-a-mole' and
> > > > 'work.misc' into for-next
> > > > git bisect good 416d20e8c31107f5dfd45d1d80d1e6c8e4871180
> > > > # good: [325a1de4a691512a48c1426b943a7b0b9f8d6744] xfs: convert to new
> > > > i_version API
> > > > git bisect good 325a1de4a691512a48c1426b943a7b0b9f8d6744
> > > > # good: [a94fe10fb114c169e7ddaecd8251521886409121] checkpatch: add
> > > > pF/pf deprecation warning
> > > > git bisect good a94fe10fb114c169e7ddaecd8251521886409121
> > > > # good: [6b3911dffd1184fdcd63299a5fee59ac000f2067] btrfs: only dirty
> > > > the inode in btrfs_update_time if something was changed
> > > > git bisect good 6b3911dffd1184fdcd63299a5fee59ac000f2067
> > > > # bad: [448f8c749a7a0ae03505823910ec45a112678048] Merge
> > > > remote-tracking branch 'iversion/iversion-next'
> > > > git bisect bad 448f8c749a7a0ae03505823910ec45a112678048
> > > > # bad: [8618bff776758ebff5b55211e7b5a60a0fc119a5] fs: handle
> > > > inode->i_version more efficiently
> > > > git bisect bad 8618bff776758ebff5b55211e7b5a60a0fc119a5
> > > > # first bad commit: [8618bff776758ebff5b55211e7b5a60a0fc119a5] fs:
> > > > handle inode->i_version more efficiently
> > > 
> > > That's really strange. I'm afraid I have no idea what could be going on.
> > > 
> > > With NFS, we really just treat i_version as an opaque value, so I'm not
> > > sure how this patch in particular would affect anything there. We _do_
> > > increment it if you have a write delegation in some cases, but not many
> > > servers hand those out.
> > 
> > About the NFS server, Arch Linux on Raspberry Pi (so 32-bit, ARMv6):
> > Linux pi 4.9.70-1-ARCH #1 SMP Mon Dec 18 19:38:00 UTC 2017 armv6l GNU/Linux
> > 
> > The /etc/nfs.conf is default except:
> > [nfsd]
> > vers2=n
> > vers3=n
> > 
> > The client logs for nfsroot mounts are:
> > Jan 08 14:07:25 :: running hook [net_nfs4]
> > Jan 08 14:07:25 IP-Config: eth0 hardware address ba:17:70:7e:87:d1 mtu 1500
> > Jan 08 14:07:25 IP-Config: eth0 guessed broadcast address 192.168.1.255
> > Jan 08 14:07:25 IP-Config: eth0 complete (from 192.168.1.10):
> > Jan 08 14:07:25 address: 192.168.1.11 broadcast: 192.168.1.255
> > netmask: 255.255.255.0
> > Jan 08 14:07:25 gateway: 192.168.1.1 dns0 : 0.0.0.0 dns1 : 0.0.0.0
> > Jan 08 14:07:25 rootserver: 192.168.1.10 rootpath:
> > Jan 08 14:07:25 filename :
> > Jan 08 14:07:25 NFS-Mount: 192.168.1.10:/srv/nfs/odroidxu3
> > Jan 08 14:07:25 Waiting 10 seconds for device /dev/nfs ...
> > Jan 08 14:07:36 ERROR: device '/dev/nfs' not found. Skipping fsck.
> > Jan 08 14:07:36 Mount cmd:
> > Jan 08 14:07:36 /opt/tools/buildbot/arch-arm-bin/mount.nfs4 -o
> > vers=4,nolock 192.168.1.10:/srv/nfs/odroidxu3 /new_root
> > 
> > Only root (/) is froom NFS. The /home comes from sdcard (/dev/mmcblk1).
> > 
> > > ext4 will only touch the i_version field if you mount it with '-o
> > > iversion'. Are you doing that here?
> > 
> > The /home is mounted by systemd from /etc/fstab:
> >       LABEL=home /home ext4 defaults 0 2
> > 
> > > 
> > > Have you run the bisect more than once? Is this maybe an intermittent
> > > problem, and the bisect has landed on the wrong commit?
> > 
> > I just run tests on commits around again (but not full bisect) on
> > current next (next-20180108):
> > fbf97ece47d66d22 - works
> > 3da7bcdd695bae43 ("fs: handle inode->i_version more efficiently") -
> > fails: http://www.krzk.eu/#/builders/1/builds/1267
> > 
> > 
> > Best regards,
> > Krzysztof
> 
> Ok, thanks. If you're seeing hangs then that might imply that we have
> some sort of excessive looping going on in the cmpxchg loops.
> 
> Could you apply the patch below and let me know if it causes either of
> the warnings to pop? That might at least point us in the right
> direction:

No new warnings with attached patch (except existing already lockdep:
"INFO: trying to register non-static key.").

Systemd timeouts on mounting /home but after entering rescue shell there
is no problem running mount /home:
	Give root password for maintenance
	(or press Control-D to continue): 
	root@odroidxu3:~# mount /home
	[  220.659331] EXT4-fs (mmcblk1p2): mounted filesystem with ordered data mode. Opts: (null)

Best regards,
Krzysztof

--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Jeff Layton Jan. 8, 2018, 6 p.m. UTC | #2
On Mon, 2018-01-08 at 18:29 +0100, Krzysztof Kozlowski wrote:
> On Mon, Jan 08, 2018 at 08:29:24AM -0500, Jeff Layton wrote:
> > On Mon, 2018-01-08 at 14:21 +0100, Krzysztof Kozlowski wrote:
> > > On Mon, Jan 8, 2018 at 1:56 PM, Jeff Layton <jlayton@kernel.org> wrote:
> > > > On Sun, 2018-01-07 at 13:44 +0100, Krzysztof Kozlowski wrote:
> > > > > On Fri, Dec 22, 2017 at 1:05 PM, Jeff Layton <jlayton@kernel.org> wrote:
> > > > > > From: Jeff Layton <jlayton@redhat.com>
> > > > > > 
> > > > > > Since i_version is mostly treated as an opaque value, we can exploit that
> > > > > > fact to avoid incrementing it when no one is watching. With that change,
> > > > > > we can avoid incrementing the counter on writes, unless someone has
> > > > > > queried for it since it was last incremented. If the a/c/mtime don't
> > > > > > change, and the i_version hasn't changed, then there's no need to dirty
> > > > > > the inode metadata on a write.
> > > > > > 
> > > > > > Convert the i_version counter to an atomic64_t, and use the lowest order
> > > > > > bit to hold a flag that will tell whether anyone has queried the value
> > > > > > since it was last incremented.
> > > > > > 
> > > > > > When we go to maybe increment it, we fetch the value and check the flag
> > > > > > bit.  If it's clear then we don't need to do anything if the update
> > > > > > isn't being forced.
> > > > > > 
> > > > > > If we do need to update, then we increment the counter by 2, and clear
> > > > > > the flag bit, and then use a CAS op to swap it into place. If that
> > > > > > works, we return true. If it doesn't then do it again with the value
> > > > > > that we fetch from the CAS operation.
> > > > > > 
> > > > > > On the query side, if the flag is already set, then we just shift the
> > > > > > value down by 1 bit and return it. Otherwise, we set the flag in our
> > > > > > on-stack value and again use cmpxchg to swap it into place if it hasn't
> > > > > > changed. If it has, then we use the value from the cmpxchg as the new
> > > > > > "old" value and try again.
> > > > > > 
> > > > > > This method allows us to avoid incrementing the counter on writes (and
> > > > > > dirtying the metadata) under typical workloads. We only need to increment
> > > > > > if it has been queried since it was last changed.
> > > > > > 
> > > > > > Signed-off-by: Jeff Layton <jlayton@redhat.com>
> > > > > > ---
> > > > > >  include/linux/fs.h       |   2 +-
> > > > > >  include/linux/iversion.h | 208 ++++++++++++++++++++++++++++++++++-------------
> > > > > >  2 files changed, 154 insertions(+), 56 deletions(-)
> > > > > > 
> > > > > 
> > > > > Hi,
> > > > > 
> > > > > On recent linux-next my ARM/Exynos boards fail to boot over nfsroot.
> > > > > This commit popped up through bisect (log at the end). Systemd
> > > > > timeouts on some device-specific services, including mounting ext4
> > > > > /home:
> > > > > 
> > > > > [ *** ] (1 of 4) A start job is running for…ress polling (1min 41s / no limit)
> > > > > [ TIME ] Timed out waiting for device dev-ttySAC2.device.
> > > > > Jan 07 13:29:38 [DEPEND] Dependency failed for Serial Getty on ttySAC2.
> > > > > Jan 07 13:29:38 [ TIME ] Timed out waiting for device
> > > > > dev-disk-by\x2dlabel-home.device.
> > > > > Jan 07 13:29:38 [DEPEND] Dependency failed for /home.
> > > > > Jan 07 13:29:38 [DEPEND] Dependency failed for Local File Systems.
> > > > > Jan 07 13:29:38 [DEPEND] Dependency failed for File System Check on
> > > > > /dev/disk/by-label/home.
> > > > > Jan 07 13:30:02 [ *** ] (1 of 2) A start job is running for…ress
> > > > > polling (1min 53s / no limit)
> > > > > 
> > > > > Kernel command line:
> > > > > console=tty1 console=ttySAC2,115200n8
> > > > > ip=192.168.1.11:192.168.1.10:192.168.1.1:255.255.255.0::eth0:none
> > > > > nfsrootdebug root=/dev/nfs
> > > > > nfsroot=192.168.1.10:/srv/nfs/odroidxu3,vers=4,nolock rootwait rw
> > > > > smsc95xx.macaddr=00:1e:06:61:7a:93 no_console_suspend
> > > > > 
> > > > > /home is /dev/mmcblk1p2:
> > > > > kozik@odroidxu3:~$ tune2fs -l /dev/mmcblk1p2
> > > > > tune2fs 1.43.7 (16-Oct-2017)
> > > > > Filesystem volume name:   home
> > > > > Last mounted on:          /home
> > > > > Filesystem UUID:          3f9dbeba-2738-45d3-807e-c1b2e21128ed
> > > > > Filesystem magic number:  0xEF53
> > > > > Filesystem revision #:    1 (dynamic)
> > > > > Filesystem features:      has_journal ext_attr resize_inode dir_index
> > > > > filetype needs_recovery extent flex_bg sparse_super large_file
> > > > > uninit_bg dir_nlink extra_isize
> > > > > Filesystem flags:         signed_directory_hash
> > > > > Default mount options:    user_xattr acl
> > > > > Filesystem state:         clean
> > > > > Errors behavior:          Continue
> > > > > Filesystem OS type:       Linux
> > > > > Inode count:              1430800
> > > > > Block count:              5717760
> > > > > Reserved block count:     285888
> > > > > Free blocks:              5467576
> > > > > Free inodes:              1428301
> > > > > First block:              0
> > > > > Block size:               4096
> > > > > Fragment size:            4096
> > > > > Reserved GDT blocks:      1022
> > > > > Blocks per group:         32768
> > > > > Fragments per group:      32768
> > > > > Inodes per group:         8176
> > > > > Inode blocks per group:   511
> > > > > Flex block group size:    16
> > > > > Filesystem created:       Thu May 21 12:17:05 2015
> > > > > Last mount time:          Thu Dec 21 13:31:26 2017
> > > > > Last write time:          Thu Dec 21 13:31:26 2017
> > > > > Mount count:              1
> > > > > Maximum mount count:      -1
> > > > > Last checked:             Thu Dec 21 13:31:25 2017
> > > > > Check interval:           0 (<none>)
> > > > > Lifetime writes:          126 GB
> > > > > Reserved blocks uid:      0 (user root)
> > > > > Reserved blocks gid:      0 (group root)
> > > > > First inode:              11
> > > > > Inode size:           256
> > > > > Required extra isize:     28
> > > > > Desired extra isize:      28
> > > > > Journal inode:            8
> > > > > Default directory hash:   half_md4
> > > > > Directory Hash Seed:      42e17e23-86b2-4356-ad63-78aa51651d03
> > > > > Journal backup:           inode blocks
> > > > > 
> > > > > 
> > > > > Full dmesg log:
> > > > > http://www.krzk.eu/#/builders/1/builds/1258/steps/10/logs/serial0
> > > > > 
> > > > > The regular boot from rootfs on SD card also fails - but without any
> > > > > serial console logs (just "Starting kernel...") so the real cause is
> > > > > unknown.
> > > > > 
> > > > > Any hints?
> > > > > 
> > > > > Best regards,
> > > > > Krzysztof
> > > > > 
> > > > > 
> > > > > bisect log:
> > > > > # bad: [73005e1a35fd67c644b0645c9e4c1efabd0fe62c] Add linux-next
> > > > > specific files for 20180103
> > > > > # good: [30a7acd573899fd8b8ac39236eff6468b195ac7d] Linux 4.15-rc6
> > > > > git bisect start 'next/master' 'next/stable'
> > > > > # bad: [c1d290f9ce8daa2b0a79d2fe48c1b7c3c5370f5a] Merge
> > > > > remote-tracking branch 'crypto/master'
> > > > > git bisect bad c1d290f9ce8daa2b0a79d2fe48c1b7c3c5370f5a
> > > > > # bad: [55695d94f0915121d106cd2d1ab94983a32f3e9a] Merge
> > > > > remote-tracking branch 'hid/for-next'
> > > > > git bisect bad 55695d94f0915121d106cd2d1ab94983a32f3e9a
> > > > > # good: [cffae1eead0dd91be1a3069a8348127bb00158f3] Merge
> > > > > remote-tracking branch 'realtek/for-next'
> > > > > git bisect good cffae1eead0dd91be1a3069a8348127bb00158f3
> > > > > # good: [5f889f1176dc99636c6bf8af7c286decc888c007] Merge
> > > > > remote-tracking branch 'btrfs/next'
> > > > > git bisect good 5f889f1176dc99636c6bf8af7c286decc888c007
> > > > > # good: [984c35877f36bee305e43a1c58176169854d85cf] Merge
> > > > > remote-tracking branch 'xfs/for-next'
> > > > > git bisect good 984c35877f36bee305e43a1c58176169854d85cf
> > > > > # bad: [f9fec502daea2a869232b6dff33ba3de79dd0d61] Merge
> > > > > remote-tracking branch 'printk/for-next'
> > > > > git bisect bad f9fec502daea2a869232b6dff33ba3de79dd0d61
> > > > > # good: [c71d227fc4133f949dae620ed5e3a250b43f2415] make kernel-side
> > > > > POLL... arch-independent
> > > > > git bisect good c71d227fc4133f949dae620ed5e3a250b43f2415
> > > > > # good: [416d20e8c31107f5dfd45d1d80d1e6c8e4871180] Merge branches
> > > > > 'work.get_user_pages_fast', 'work.wmci', 'work.sock_recvmsg',
> > > > > 'misc.netdrv', 'misc.poll', 'work.mqueue', 'work.whack-a-mole' and
> > > > > 'work.misc' into for-next
> > > > > git bisect good 416d20e8c31107f5dfd45d1d80d1e6c8e4871180
> > > > > # good: [325a1de4a691512a48c1426b943a7b0b9f8d6744] xfs: convert to new
> > > > > i_version API
> > > > > git bisect good 325a1de4a691512a48c1426b943a7b0b9f8d6744
> > > > > # good: [a94fe10fb114c169e7ddaecd8251521886409121] checkpatch: add
> > > > > pF/pf deprecation warning
> > > > > git bisect good a94fe10fb114c169e7ddaecd8251521886409121
> > > > > # good: [6b3911dffd1184fdcd63299a5fee59ac000f2067] btrfs: only dirty
> > > > > the inode in btrfs_update_time if something was changed
> > > > > git bisect good 6b3911dffd1184fdcd63299a5fee59ac000f2067
> > > > > # bad: [448f8c749a7a0ae03505823910ec45a112678048] Merge
> > > > > remote-tracking branch 'iversion/iversion-next'
> > > > > git bisect bad 448f8c749a7a0ae03505823910ec45a112678048
> > > > > # bad: [8618bff776758ebff5b55211e7b5a60a0fc119a5] fs: handle
> > > > > inode->i_version more efficiently
> > > > > git bisect bad 8618bff776758ebff5b55211e7b5a60a0fc119a5
> > > > > # first bad commit: [8618bff776758ebff5b55211e7b5a60a0fc119a5] fs:
> > > > > handle inode->i_version more efficiently
> > > > 
> > > > That's really strange. I'm afraid I have no idea what could be going on.
> > > > 
> > > > With NFS, we really just treat i_version as an opaque value, so I'm not
> > > > sure how this patch in particular would affect anything there. We _do_
> > > > increment it if you have a write delegation in some cases, but not many
> > > > servers hand those out.
> > > 
> > > About the NFS server, Arch Linux on Raspberry Pi (so 32-bit, ARMv6):
> > > Linux pi 4.9.70-1-ARCH #1 SMP Mon Dec 18 19:38:00 UTC 2017 armv6l GNU/Linux
> > > 
> > > The /etc/nfs.conf is default except:
> > > [nfsd]
> > > vers2=n
> > > vers3=n
> > > 
> > > The client logs for nfsroot mounts are:
> > > Jan 08 14:07:25 :: running hook [net_nfs4]
> > > Jan 08 14:07:25 IP-Config: eth0 hardware address ba:17:70:7e:87:d1 mtu 1500
> > > Jan 08 14:07:25 IP-Config: eth0 guessed broadcast address 192.168.1.255
> > > Jan 08 14:07:25 IP-Config: eth0 complete (from 192.168.1.10):
> > > Jan 08 14:07:25 address: 192.168.1.11 broadcast: 192.168.1.255
> > > netmask: 255.255.255.0
> > > Jan 08 14:07:25 gateway: 192.168.1.1 dns0 : 0.0.0.0 dns1 : 0.0.0.0
> > > Jan 08 14:07:25 rootserver: 192.168.1.10 rootpath:
> > > Jan 08 14:07:25 filename :
> > > Jan 08 14:07:25 NFS-Mount: 192.168.1.10:/srv/nfs/odroidxu3
> > > Jan 08 14:07:25 Waiting 10 seconds for device /dev/nfs ...
> > > Jan 08 14:07:36 ERROR: device '/dev/nfs' not found. Skipping fsck.
> > > Jan 08 14:07:36 Mount cmd:
> > > Jan 08 14:07:36 /opt/tools/buildbot/arch-arm-bin/mount.nfs4 -o
> > > vers=4,nolock 192.168.1.10:/srv/nfs/odroidxu3 /new_root
> > > 
> > > Only root (/) is froom NFS. The /home comes from sdcard (/dev/mmcblk1).
> > > 
> > > > ext4 will only touch the i_version field if you mount it with '-o
> > > > iversion'. Are you doing that here?
> > > 
> > > The /home is mounted by systemd from /etc/fstab:
> > >       LABEL=home /home ext4 defaults 0 2
> > > 
> > > > 
> > > > Have you run the bisect more than once? Is this maybe an intermittent
> > > > problem, and the bisect has landed on the wrong commit?
> > > 
> > > I just run tests on commits around again (but not full bisect) on
> > > current next (next-20180108):
> > > fbf97ece47d66d22 - works
> > > 3da7bcdd695bae43 ("fs: handle inode->i_version more efficiently") -
> > > fails: http://www.krzk.eu/#/builders/1/builds/1267
> > > 
> > > 
> > > Best regards,
> > > Krzysztof
> > 
> > Ok, thanks. If you're seeing hangs then that might imply that we have
> > some sort of excessive looping going on in the cmpxchg loops.
> > 
> > Could you apply the patch below and let me know if it causes either of
> > the warnings to pop? That might at least point us in the right
> > direction:
> 
> No new warnings with attached patch (except existing already lockdep:
> "INFO: trying to register non-static key.").
> 

Yeah, I saw that in the original logs and it looks unrelated (and
harmless).

> Systemd timeouts on mounting /home but after entering rescue shell there
> is no problem running mount /home:
> 	Give root password for maintenance
> 	(or press Control-D to continue): 
> 	root@odroidxu3:~# mount /home
> 	[  220.659331] EXT4-fs (mmcblk1p2): mounted filesystem with ordered data mode. Opts: (null)
> 

Ok, thanks for testing it. So I guess we can probably rule out excessive
looping in those functions as the issue.

To make sure I understand the problem: When systemd tries to do the
initial mount of /home (which is an ext4 filesystem), it hangs. But once
it drops to the shell, it works, if you do the mount by hand.

Is that correct?

If so, then is it possible to trigger sysrq commands during the hanging
mount attempt? Maybe you could use e.g. sysrq-l, sysrq-w, etc. to
determine what it's blocking on?

Thanks,
Krzysztof Kozlowski Jan. 8, 2018, 6:33 p.m. UTC | #3
On Mon, Jan 08, 2018 at 01:00:19PM -0500, Jeff Layton wrote:
> On Mon, 2018-01-08 at 18:29 +0100, Krzysztof Kozlowski wrote:

(...)

> > > Ok, thanks. If you're seeing hangs then that might imply that we have
> > > some sort of excessive looping going on in the cmpxchg loops.
> > > 
> > > Could you apply the patch below and let me know if it causes either of
> > > the warnings to pop? That might at least point us in the right
> > > direction:
> > 
> > No new warnings with attached patch (except existing already lockdep:
> > "INFO: trying to register non-static key.").
> > 
> 
> Yeah, I saw that in the original logs and it looks unrelated (and
> harmless).
> 
> > Systemd timeouts on mounting /home but after entering rescue shell there
> > is no problem running mount /home:
> > 	Give root password for maintenance
> > 	(or press Control-D to continue): 
> > 	root@odroidxu3:~# mount /home
> > 	[  220.659331] EXT4-fs (mmcblk1p2): mounted filesystem with ordered data mode. Opts: (null)
> > 
> 
> Ok, thanks for testing it. So I guess we can probably rule out excessive
> looping in those functions as the issue.
> 
> To make sure I understand the problem: When systemd tries to do the
> initial mount of /home (which is an ext4 filesystem), it hangs. But once
> it drops to the shell, it works, if you do the mount by hand.
> 
> Is that correct?

Yes, although it also timeouts on setting up /dev/ttySAC2 (serial
console).

> If so, then is it possible to trigger sysrq commands during the hanging
> mount attempt? Maybe you could use e.g. sysrq-l, sysrq-w, etc. to
> determine what it's blocking on?

Yes, I have sysrq. Blocked state (w):
##################
[ ***  ] (2 of 4) A start job is running for…v-ttySAC2.device (1min / 1min 30s)
*** break sent ***
[***   ] (2 of 4) A start job is running for…v-ttySAC2.device (1min / 1min 30s)[  110.962917] sysrq: SysRq : Show Blocked State
[  110.965931]   task                PC stack   pid father
[  110.971360] lvm             D    0   219      1 0x00000000
[  110.976649] [<c087682c>] (__schedule) from [<c0876ca0>] (schedule+0x4c/0xac)
[  110.983643] [<c0876ca0>] (schedule) from [<c07ae19c>] (rpc_wait_bit_killable+0x2c/0xf0)
[  110.991621] [<c07ae19c>] (rpc_wait_bit_killable) from [<c0877328>] (__wait_on_bit+0x80/0xb4)
[  111.000025] [<c0877328>] (__wait_on_bit) from [<c08773e8>] (out_of_line_wait_on_bit+0x8c/0x94)
[  111.008603] [<c08773e8>] (out_of_line_wait_on_bit) from [<c07aec78>] (__rpc_execute+0x19c/0x274)
[  111.017370] [<c07aec78>] (__rpc_execute) from [<c07a5eb4>] (rpc_run_task+0x134/0x14c)
[  111.025164] [<c07a5eb4>] (rpc_run_task) from [<c0315b20>] (nfs4_call_sync_sequence+0x58/0x78)
[  111.033648] [<c0315b20>] (nfs4_call_sync_sequence) from [<c031c490>] (nfs4_proc_access+0xf4/0x12c)
[  111.042586] [<c031c490>] (nfs4_proc_access) from [<c02fe408>] (nfs_do_access+0x230/0x3e8)
[  111.050716] [<c02fe408>] (nfs_do_access) from [<c02fe88c>] (nfs_permission+0x2a0/0x2c0)
[  111.058694] [<c02fe88c>] (nfs_permission) from [<c0224aa4>] (link_path_walk+0x6c/0x508)
[  111.066658] [<c0224aa4>] (link_path_walk) from [<c0225030>] (path_lookupat+0x8c/0x1ec)
[  111.074541] [<c0225030>] (path_lookupat) from [<c02270fc>] (path_openat+0x770/0x964)
[  111.082253] [<c02270fc>] (path_openat) from [<c0227e78>] (do_filp_open+0x60/0xc4)
[  111.089711] [<c0227e78>] (do_filp_open) from [<c0215a14>] (do_sys_open+0x114/0x1c4)
[  111.097339] [<c0215a14>] (do_sys_open) from [<c0108880>] (ret_fast_syscall+0x0/0x28)
[  111.105036] udevadm         D    0   259      1 0x00000000
[  111.110482] [<c087682c>] (__schedule) from [<c0876ca0>] (schedule+0x4c/0xac)
[  111.117508] [<c0876ca0>] (schedule) from [<c07ae19c>] (rpc_wait_bit_killable+0x2c/0xf0)
[  111.125494] [<c07ae19c>] (rpc_wait_bit_killable) from [<c0877328>] (__wait_on_bit+0x80/0xb4)
[  111.133903] [<c0877328>] (__wait_on_bit) from [<c08773e8>] (out_of_line_wait_on_bit+0x8c/0x94)
[  111.142482] [<c08773e8>] (out_of_line_wait_on_bit) from [<c07aec78>] (__rpc_execute+0x19c/0x274)
[  111.151236] [<c07aec78>] (__rpc_execute) from [<c07a5eb4>] (rpc_run_task+0x134/0x14c)
[  111.159028] [<c07a5eb4>] (rpc_run_task) from [<c0315b20>] (nfs4_call_sync_sequence+0x58/0x78)
[  111.167523] [<c0315b20>] (nfs4_call_sync_sequence) from [<c031c68c>] (nfs4_proc_getattr+0xbc/0xe4)
[  111.176452] [<c031c68c>] (nfs4_proc_getattr) from [<c0303c90>] (__nfs_revalidate_inode+0x8c/0x130)
[  111.185378] [<c0303c90>] (__nfs_revalidate_inode) from [<c02fe758>] (nfs_permission+0x16c/0x2c0)
[  111.194132] [<c02fe758>] (nfs_permission) from [<c0224aa4>] (link_path_walk+0x6c/0x508)
[  111.202096] [<c0224aa4>] (link_path_walk) from [<c0225030>] (path_lookupat+0x8c/0x1ec)
[  111.209980] [<c0225030>] (path_lookupat) from [<c0227714>] (filename_lookup+0x8c/0xe8)
[  111.217865] [<c0227714>] (filename_lookup) from [<c0215014>] (SyS_faccessat+0x9c/0x208)
[  111.225839] [<c0215014>] (SyS_faccessat) from [<c0108880>] (ret_fast_syscall+0x0/0x28)
[***   ] (3 of 4) A start job is running for…l-home.device (1min 2s / 1min 30s)
##################

Another blocked (after few seconds):
[  125.364311] sysrq: SysRq : Show Blocked State
[  125.367292]   task                PC stack   pid father
[  125.372688] udevadm         D    0   259      1 0x00000000
[  125.377998] [<c087682c>] (__schedule) from [<c0876ca0>] (schedule+0x4c/0xac)
[  125.384990] [<c0876ca0>] (schedule) from [<c07ae19c>] (rpc_wait_bit_killable+0x2c/0xf0)
[  125.392961] [<c07ae19c>] (rpc_wait_bit_killable) from [<c0877328>] (__wait_on_bit+0x80/0xb4)
[  125.401365] [<c0877328>] (__wait_on_bit) from [<c08773e8>] (out_of_line_wait_on_bit+0x8c/0x94)
[  125.409944] [<c08773e8>] (out_of_line_wait_on_bit) from [<c07aec78>] (__rpc_execute+0x19c/0x274)
[  125.418711] [<c07aec78>] (__rpc_execute) from [<c07a5eb4>] (rpc_run_task+0x134/0x14c)
[  125.426505] [<c07a5eb4>] (rpc_run_task) from [<c0315b20>] (nfs4_call_sync_sequence+0x58/0x78)
[  125.434989] [<c0315b20>] (nfs4_call_sync_sequence) from [<c031eab0>] (nfs4_proc_lookup_common+0x100/0x354)
[  125.444608] [<c031eab0>] (nfs4_proc_lookup_common) from [<c031edac>] (nfs4_proc_lookup+0x3c/0x88)
[  125.453456] [<c031edac>] (nfs4_proc_lookup) from [<c02ff46c>] (nfs_lookup_revalidate+0x224/0x35c)
[  125.462294] [<c02ff46c>] (nfs_lookup_revalidate) from [<c02233c4>] (lookup_fast+0x294/0x478)
[  125.470689] [<c02233c4>] (lookup_fast) from [<c02235d4>] (walk_component+0x2c/0x2e0)
[  125.478395] [<c02235d4>] (walk_component) from [<c0224bd4>] (link_path_walk+0x19c/0x508)
[  125.486455] [<c0224bd4>] (link_path_walk) from [<c0225030>] (path_lookupat+0x8c/0x1ec)
[  125.494342] [<c0225030>] (path_lookupat) from [<c0227714>] (filename_lookup+0x8c/0xe8)
[  125.502226] [<c0227714>] (filename_lookup) from [<c021d074>] (SyS_readlinkat+0x44/0xec)
[  125.510208] [<c021d074>] (SyS_readlinkat) from [<c0108880>] (ret_fast_syscall+0x0/0x28)
[    **] (3 of 4) A start job is running for…-home.device (1min 20s / 1min 30s)

##################

Another blocked (after few seconds):
[    **] (3 of 4) A start job is running for…-home.device (1min 26s / 1min 30s)[  136.281417] sysrq: SysRq : Show Blocked State
[  136.284403]   task                PC stack   pid father
[  136.289818] udevadm         D    0   259      1 0x00000000
[  136.295116] [<c087682c>] (__schedule) from [<c0876ca0>] (schedule+0x4c/0xac)
[  136.302105] [<c0876ca0>] (schedule) from [<c07ae19c>] (rpc_wait_bit_killable+0x2c/0xf0)
[  136.310079] [<c07ae19c>] (rpc_wait_bit_killable) from [<c0877328>] (__wait_on_bit+0x80/0xb4)
[  136.318484] [<c0877328>] (__wait_on_bit) from [<c08773e8>] (out_of_line_wait_on_bit+0x8c/0x94)
[  136.327064] [<c08773e8>] (out_of_line_wait_on_bit) from [<c07aec78>] (__rpc_execute+0x19c/0x274)
[  136.335830] [<c07aec78>] (__rpc_execute) from [<c07a5eb4>] (rpc_run_task+0x134/0x14c)
[  136.343625] [<c07a5eb4>] (rpc_run_task) from [<c0315b20>] (nfs4_call_sync_sequence+0x58/0x78)
[  136.352110] [<c0315b20>] (nfs4_call_sync_sequence) from [<c031eab0>] (nfs4_proc_lookup_common+0x100/0x354)
[  136.361729] [<c031eab0>] (nfs4_proc_lookup_common) from [<c031edac>] (nfs4_proc_lookup+0x3c/0x88)
[  136.370576] [<c031edac>] (nfs4_proc_lookup) from [<c02ff46c>] (nfs_lookup_revalidate+0x224/0x35c)
[  136.379416] [<c02ff46c>] (nfs_lookup_revalidate) from [<c02233c4>] (lookup_fast+0x294/0x478)
[  136.387809] [<c02233c4>] (lookup_fast) from [<c02235d4>] (walk_component+0x2c/0x2e0)
[  136.395514] [<c02235d4>] (walk_component) from [<c0224bd4>] (link_path_walk+0x19c/0x508)
[  136.403575] [<c0224bd4>] (link_path_walk) from [<c02269f4>] (path_openat+0x68/0x964)
[  136.411285] [<c02269f4>] (path_openat) from [<c0227e78>] (do_filp_open+0x60/0xc4)
[  136.418743] [<c0227e78>] (do_filp_open) from [<c0215a14>] (do_sys_open+0x114/0x1c4)
[  136.426372] [<c0215a14>] (do_sys_open) from [<c0108880>] (ret_fast_syscall+0x0/0x28)

##################


And list of all CPU backtraces (l):

*** break sent ***
[   87.197309] sysrq: SysRq : Show backtrace of all active CPUs
[   87.201610] NMI backtrace for cpu 0
[   87.205044] CPU: 0 PID: 0 Comm: swapper/0 Not tainted 4.15.0-rc3-00022-g3da7bcdd695b #1101
[   87.213297] Hardware name: SAMSUNG EXYNOS (Flattened Device Tree)
[   87.219399] [<c01103a4>] (unwind_backtrace) from [<c010cac0>] (show_stack+0x10/0x14)
[   87.227098] [<c010cac0>] (show_stack) from [<c08620ac>] (dump_stack+0x98/0xc4)
[   87.234280] [<c08620ac>] (dump_stack) from [<c0868078>] (nmi_cpu_backtrace+0x10c/0x110)
[   87.242251] [<c0868078>] (nmi_cpu_backtrace) from [<c0868184>] (nmi_trigger_cpumask_backtrace+0x108/0x150)
[   87.251884] [<c0868184>] (nmi_trigger_cpumask_backtrace) from [<c041f384>] (__handle_sysrq+0x170/0x264)
[   87.261237] [<c041f384>] (__handle_sysrq) from [<c0439888>] (s3c24xx_serial_rx_drain_fifo+0x1ec/0x214)
[   87.270506] [<c0439888>] (s3c24xx_serial_rx_drain_fifo) from [<c043a6b8>] (s3c24xx_serial_rx_chars+0x6c/0x1a0)
[   87.280473] [<c043a6b8>] (s3c24xx_serial_rx_chars) from [<c043a834>] (s3c64xx_serial_handle_irq+0x48/0x60)
[   87.290104] [<c043a834>] (s3c64xx_serial_handle_irq) from [<c017ceb0>] (__handle_irq_event_percpu+0x9c/0x128)
[   87.299975] [<c017ceb0>] (__handle_irq_event_percpu) from [<c017cf58>] (handle_irq_event_percpu+0x1c/0x58)
[   87.309589] [<c017cf58>] (handle_irq_event_percpu) from [<c017cfcc>] (handle_irq_event+0x38/0x5c)
[   87.318425] [<c017cfcc>] (handle_irq_event) from [<c0180740>] (handle_fasteoi_irq+0xb8/0x174)
[   87.326914] [<c0180740>] (handle_fasteoi_irq) from [<c017c15c>] (generic_handle_irq+0x24/0x34)
[   87.335492] [<c017c15c>] (generic_handle_irq) from [<c017c714>] (__handle_domain_irq+0x7c/0xec)
[   87.344159] [<c017c714>] (__handle_domain_irq) from [<c0101544>] (gic_handle_irq+0x58/0x9c)
[   87.352473] [<c0101544>] (gic_handle_irq) from [<c010d7b0>] (__irq_svc+0x70/0xb0)
[   87.359912] Exception stack(0xc0d01ee8 to 0xc0d01f30)
[   87.364927] 1ee0:                   c087cc70 c0d0bf80 00000000 00000000 eef6d3c0 00000000
[   87.373094] 1f00: ed4e8000 00000102 cfe02400 c0d08db8 ed491880 c0d01f6c 00000001 c0d01f38
[   87.381234] 1f20: c087cc70 c087cc74 60070013 ffffffff
[   87.386254] [<c010d7b0>] (__irq_svc) from [<c087cc74>] (_raw_spin_unlock_irq+0x28/0x5c)
[   87.394252] [<c087cc74>] (_raw_spin_unlock_irq) from [<c0144b94>] (finish_task_switch+0xa0/0x1b0)
[   87.403097] [<c0144b94>] (finish_task_switch) from [<c0876830>] (__schedule+0x238/0x65c)
[   87.411140] [<c0876830>] (__schedule) from [<c0877158>] (schedule_idle+0x38/0x78)
[   87.418591] [<c0877158>] (schedule_idle) from [<c0162030>] (cpu_startup_entry+0x18/0x1c)
[   87.426658] [<c0162030>] (cpu_startup_entry) from [<c0c00c88>] (start_kernel+0x39c/0x3a8)
[   87.434792] Sending NMI from CPU 0 to CPUs 1-7:
[   87.439278] NMI backtrace for cpu 3
[   87.439294] CPU: 3 PID: 0 Comm: swapper/3 Not tainted 4.15.0-rc3-00022-g3da7bcdd695b #1101
[   87.439300] Hardware name: SAMSUNG EXYNOS (Flattened Device Tree)
[   87.439322] PC is at arch_cpu_idle+0x24/0x3c
[   87.439333] LR is at arch_cpu_idle+0x20/0x3c
[   87.439341] pc : [<c01093d8>]    lr : [<c01093d4>]    psr: 60000013
[   87.439348] sp : ee8d5fb8  ip : 00000001  fp : c0afe104
[   87.439354] r10: 00000000  r9 : 00000000  r8 : c0c65730
[   87.439363] r7 : 00000008  r6 : c0d08c8c  r5 : c0d08c20  r4 : ee8d4000
[   87.439370] r3 : 00000000  r2 : 00000000  r1 : 00000000  r0 : c01093d4
[   87.439381] Flags: nZCv  IRQs on  FIQs on  Mode SVC_32  ISA ARM  Segment none
[   87.439389] Control: 10c5387d  Table: 6cd4806a  DAC: 00000051
[   87.439399] CPU: 3 PID: 0 Comm: swapper/3 Not tainted 4.15.0-rc3-00022-g3da7bcdd695b #1101
[   87.439405] Hardware name: SAMSUNG EXYNOS (Flattened Device Tree)
[   87.439425] [<c01103a4>] (unwind_backtrace) from [<c010cac0>] (show_stack+0x10/0x14)
[   87.439442] [<c010cac0>] (show_stack) from [<c08620ac>] (dump_stack+0x98/0xc4)
[   87.439460] [<c08620ac>] (dump_stack) from [<c086801c>] (nmi_cpu_backtrace+0xb0/0x110)
[   87.439476] [<c086801c>] (nmi_cpu_backtrace) from [<c010f2ac>] (handle_IPI+0xd8/0x1b0)
[   87.439492] [<c010f2ac>] (handle_IPI) from [<c0101584>] (gic_handle_irq+0x98/0x9c)
[   87.439506] [<c0101584>] (gic_handle_irq) from [<c010d7b0>] (__irq_svc+0x70/0xb0)
[   87.439514] Exception stack(0xee8d5f68 to 0xee8d5fb0)
[   87.439527] 5f60:                   c01093d4 00000000 00000000 00000000 ee8d4000 c0d08c20
[   87.439542] 5f80: c0d08c8c 00000008 c0c65730 00000000 00000000 c0afe104 00000001 ee8d5fb8
[   87.439552] 5fa0: c01093d4 c01093d8 60000013 ffffffff
[   87.439569] [<c010d7b0>] (__irq_svc) from [<c01093d8>] (arch_cpu_idle+0x24/0x3c)
[   87.439586] [<c01093d8>] (arch_cpu_idle) from [<c0161c20>] (do_idle+0x190/0x230)
[   87.439600] [<c0161c20>] (do_idle) from [<c0162030>] (cpu_startup_entry+0x18/0x1c)
[   87.439615] [<c0162030>] (cpu_startup_entry) from [<4010192c>] (0x4010192c)
[   87.439626] NMI backtrace for cpu 1
[   87.439641] CPU: 1 PID: 0 Comm: swapper/1 Not tainted 4.15.0-rc3-00022-g3da7bcdd695b #1101
[   87.439647] Hardware name: SAMSUNG EXYNOS (Flattened Device Tree)
[   87.439659] PC is at arch_cpu_idle+0x24/0x3c
[   87.439669] LR is at arch_cpu_idle+0x20/0x3c
[   87.439677] pc : [<c01093d8>]    lr : [<c01093d4>]    psr: 600d0013
[   87.439684] sp : ee8d1fb8  ip : 00000001  fp : c0afe104
[   87.439691] r10: 00000000  r9 : 00000000  r8 : c0c65730
[   87.439699] r7 : 00000002  r6 : c0d08c8c  r5 : c0d08c20  r4 : ee8d0000
[   87.439707] r3 : 00000000  r2 : 00000000  r1 : 00000000  r0 : c01093d4
[   87.439715] Flags: nZCv  IRQs on  FIQs on  Mode SVC_32  ISA ARM  Segment none
[   87.439723] Control: 10c5387d  Table: 6cd4806a  DAC: 00000051
[   87.439733] CPU: 1 PID: 0 Comm: swapper/1 Not tainted 4.15.0-rc3-00022-g3da7bcdd695b #1101
[   87.439738] Hardware name: SAMSUNG EXYNOS (Flattened Device Tree)
[   87.439758] [<c01103a4>] (unwind_backtrace) from [<c010cac0>] (show_stack+0x10/0x14)
[   87.439775] [<c010cac0>] (show_stack) from [<c08620ac>] (dump_stack+0x98/0xc4)
[   87.439793] [<c08620ac>] (dump_stack) from [<c086801c>] (nmi_cpu_backtrace+0xb0/0x110)
[   87.439808] [<c086801c>] (nmi_cpu_backtrace) from [<c010f2ac>] (handle_IPI+0xd8/0x1b0)
[   87.439824] [<c010f2ac>] (handle_IPI) from [<c0101584>] (gic_handle_irq+0x98/0x9c)
[   87.439838] [<c0101584>] (gic_handle_irq) from [<c010d7b0>] (__irq_svc+0x70/0xb0)
[   87.439845] Exception stack(0xee8d1f68 to 0xee8d1fb0)
[   87.439859] 1f60:                   c01093d4 00000000 00000000 00000000 ee8d0000 c0d08c20
[   87.439873] 1f80: c0d08c8c 00000002 c0c65730 00000000 00000000 c0afe104 00000001 ee8d1fb8
[   87.439884] 1fa0: c01093d4 c01093d8 600d0013 ffffffff
[   87.439901] [<c010d7b0>] (__irq_svc) from [<c01093d8>] (arch_cpu_idle+0x24/0x3c)
[   87.439917] [<c01093d8>] (arch_cpu_idle) from [<c0161c20>] (do_idle+0x190/0x230)
[   87.439931] [<c0161c20>] (do_idle) from [<c0162030>] (cpu_startup_entry+0x18/0x1c)
[   87.439944] [<c0162030>] (cpu_startup_entry) from [<4010192c>] (0x4010192c)
[   87.439955] NMI backtrace for cpu 2
[   87.439968] CPU: 2 PID: 0 Comm: swapper/2 Not tainted 4.15.0-rc3-00022-g3da7bcdd695b #1101
[   87.439979] Hardware name: SAMSUNG EXYNOS (Flattened Device Tree)
[   87.439993] PC is at arch_cpu_idle+0x24/0x3c
[   87.440003] LR is at arch_cpu_idle+0x20/0x3c
[   87.440011] pc : [<c01093d8>]    lr : [<c01093d4>]    psr: 60070013
[   87.440019] sp : ee8d3fb8  ip : 00000001  fp : c0afe104
[   87.440027] r10: 00000000  r9 : 00000000  r8 : c0c65730
[   87.440036] r7 : 00000004  r6 : c0d08c8c  r5 : c0d08c20  r4 : ee8d2000
[   87.440044] r3 : 00000000  r2 : 00000000  r1 : 00000000  r0 : c01093d4
[   87.440054] Flags: nZCv  IRQs on  FIQs on  Mode SVC_32  ISA ARM  Segment none
[   87.440064] Control: 10c5387d  Table: 6cd3006a  DAC: 00000051
[   87.440076] CPU: 2 PID: 0 Comm: swapper/2 Not tainted 4.15.0-rc3-00022-g3da7bcdd695b #1101
[   87.440082] Hardware name: SAMSUNG EXYNOS (Flattened Device Tree)
[   87.440102] [<c01103a4>] (unwind_backtrace) from [<c010cac0>] (show_stack+0x10/0x14)
[   87.440118] [<c010cac0>] (show_stack) from [<c08620ac>] (dump_stack+0x98/0xc4)
[   87.440135] [<c08620ac>] (dump_stack) from [<c086801c>] (nmi_cpu_backtrace+0xb0/0x110)
[   87.440153] [<c086801c>] (nmi_cpu_backtrace) from [<c010f2ac>] (handle_IPI+0xd8/0x1b0)
[   87.440169] [<c010f2ac>] (handle_IPI) from [<c0101584>] (gic_handle_irq+0x98/0x9c)
[   87.440184] [<c0101584>] (gic_handle_irq) from [<c010d7b0>] (__irq_svc+0x70/0xb0)
[   87.440192] Exception stack(0xee8d3f68 to 0xee8d3fb0)
[   87.440206] 3f60:                   c01093d4 00000000 00000000 00000000 ee8d2000 c0d08c20
[   87.440221] 3f80: c0d08c8c 00000004 c0c65730 00000000 00000000 c0afe104 00000001 ee8d3fb8
[   87.440233] 3fa0: c01093d4 c01093d8 60070013 ffffffff
[   87.440251] [<c010d7b0>] (__irq_svc) from [<c01093d8>] (arch_cpu_idle+0x24/0x3c)
[   87.440269] [<c01093d8>] (arch_cpu_idle) from [<c0161c20>] (do_idle+0x190/0x230)
[   87.440285] [<c0161c20>] (do_idle) from [<c0162030>] (cpu_startup_entry+0x18/0x1c)
[   87.440299] [<c0162030>] (cpu_startup_entry) from [<4010192c>] (0x4010192c)
[   87.440323] NMI backtrace for cpu 4
[   87.440359] CPU: 4 PID: 0 Comm: swapper/4 Not tainted 4.15.0-rc3-00022-g3da7bcdd695b #1101
[   87.440375] Hardware name: SAMSUNG EXYNOS (Flattened Device Tree)
[   87.440422] PC is at arch_cpu_idle+0x24/0x3c
[   87.440450] LR is at arch_cpu_idle+0x20/0x3c
[   87.440469] pc : [<c01093d8>]    lr : [<c01093d4>]    psr: 600d0013
[   87.440487] sp : ee8d7fb8  ip : 00000001  fp : c0afe104
[   87.440504] r10: 00000000  r9 : 00000000  r8 : c0c65730
[   87.440523] r7 : 00000010  r6 : c0d08c8c  r5 : c0d08c20  r4 : ee8d6000
[   87.440541] r3 : 00000000  r2 : 00000000  r1 : 00000000  r0 : c01093d4
[   87.440565] Flags: nZCv  IRQs on  FIQs on  Mode SVC_32  ISA ARM  Segment none
[   87.440586] Control: 10c5387d  Table: 6cd5006a  DAC: 00000051
[   87.440615] CPU: 4 PID: 0 Comm: swapper/4 Not tainted 4.15.0-rc3-00022-g3da7bcdd695b #1101
[   87.440629] Hardware name: SAMSUNG EXYNOS (Flattened Device Tree)
[   87.440690] [<c01103a4>] (unwind_backtrace) from [<c010cac0>] (show_stack+0x10/0x14)
[   87.440736] [<c010cac0>] (show_stack) from [<c08620ac>] (dump_stack+0x98/0xc4)
[   87.440782] [<c08620ac>] (dump_stack) from [<c086801c>] (nmi_cpu_backtrace+0xb0/0x110)
[   87.440822] [<c086801c>] (nmi_cpu_backtrace) from [<c010f2ac>] (handle_IPI+0xd8/0x1b0)
[   87.440859] [<c010f2ac>] (handle_IPI) from [<c0101584>] (gic_handle_irq+0x98/0x9c)
[   87.440891] [<c0101584>] (gic_handle_irq) from [<c010d7b0>] (__irq_svc+0x70/0xb0)
[   87.440908] Exception stack(0xee8d7f68 to 0xee8d7fb0)
[   87.440944] 7f60:                   c01093d4 00000000 00000000 00000000 ee8d6000 c0d08c20
[   87.440979] 7f80: c0d08c8c 00000010 c0c65730 00000000 00000000 c0afe104 00000001 ee8d7fb8
[   87.441003] 7fa0: c01093d4 c01093d8 600d0013 ffffffff
[   87.441048] [<c010d7b0>] (__irq_svc) from [<c01093d8>] (arch_cpu_idle+0x24/0x3c)
[   87.441089] [<c01093d8>] (arch_cpu_idle) from [<c0161c20>] (do_idle+0x190/0x230)
[   87.441124] [<c0161c20>] (do_idle) from [<c0162030>] (cpu_startup_entry+0x18/0x1c)
[   87.441155] [<c0162030>] (cpu_startup_entry) from [<4010192c>] (0x4010192c)
[   87.441181] NMI backtrace for cpu 5
[   87.441216] CPU: 5 PID: 0 Comm: swapper/5 Not tainted 4.15.0-rc3-00022-g3da7bcdd695b #1101
[   87.441232] Hardware name: SAMSUNG EXYNOS (Flattened Device Tree)
[   87.441269] PC is at arch_cpu_idle+0x24/0x3c
[   87.441298] LR is at arch_cpu_idle+0x20/0x3c
[   87.441318] pc : [<c01093d8>]    lr : [<c01093d4>]    psr: 60010113
[   87.441336] sp : ee8e1fb8  ip : 00000001  fp : c0afe104
[   87.441352] r10: 00000000  r9 : 00000000  r8 : c0c65730
[   87.441372] r7 : 00000020  r6 : c0d08c8c  r5 : c0d08c20  r4 : ee8e0000
[   87.441389] r3 : 00000000  r2 : 00000000  r1 : 00000000  r0 : c01093d4
[   87.441412] Flags: nZCv  IRQs on  FIQs on  Mode SVC_32  ISA ARM  Segment none
[   87.441433] Control: 10c5387d  Table: 6ccd406a  DAC: 00000051
[   87.441461] CPU: 5 PID: 0 Comm: swapper/5 Not tainted 4.15.0-rc3-00022-g3da7bcdd695b #1101
[   87.441476] Hardware name: SAMSUNG EXYNOS (Flattened Device Tree)
[   87.441528] [<c01103a4>] (unwind_backtrace) from [<c010cac0>] (show_stack+0x10/0x14)
[   87.441574] [<c010cac0>] (show_stack) from [<c08620ac>] (dump_stack+0x98/0xc4)
[   87.441619] [<c08620ac>] (dump_stack) from [<c086801c>] (nmi_cpu_backtrace+0xb0/0x110)
[   87.441659] [<c086801c>] (nmi_cpu_backtrace) from [<c010f2ac>] (handle_IPI+0xd8/0x1b0)
[   87.441694] [<c010f2ac>] (handle_IPI) from [<c0101584>] (gic_handle_irq+0x98/0x9c)
[   87.441727] [<c0101584>] (gic_handle_irq) from [<c010d7b0>] (__irq_svc+0x70/0xb0)
[   87.441745] Exception stack(0xee8e1f68 to 0xee8e1fb0)
[   87.441780] 1f60:                   c01093d4 00000000 00000000 00000000 ee8e0000 c0d08c20
[   87.441815] 1f80: c0d08c8c 00000020 c0c65730 00000000 00000000 c0afe104 00000001 ee8e1fb8
[   87.441839] 1fa0: c01093d4 c01093d8 60010113 ffffffff
[   87.441884] [<c010d7b0>] (__irq_svc) from [<c01093d8>] (arch_cpu_idle+0x24/0x3c)
[   87.441925] [<c01093d8>] (arch_cpu_idle) from [<c0161c20>] (do_idle+0x190/0x230)
[   87.441959] [<c0161c20>] (do_idle) from [<c0162030>] (cpu_startup_entry+0x18/0x1c)
[   87.441992] [<c0162030>] (cpu_startup_entry) from [<4010192c>] (0x4010192c)
[   87.442017] NMI backtrace for cpu 6
[   87.442051] CPU: 6 PID: 0 Comm: swapper/6 Not tainted 4.15.0-rc3-00022-g3da7bcdd695b #1101
[   87.442067] Hardware name: SAMSUNG EXYNOS (Flattened Device Tree)
[   87.442103] PC is at arch_cpu_idle+0x24/0x3c
[   87.442132] LR is at arch_cpu_idle+0x20/0x3c
[   87.442152] pc : [<c01093d8>]    lr : [<c01093d4>]    psr: 60010013
[   87.442169] sp : ee8e3fb8  ip : 00000001  fp : c0afe104
[   87.442185] r10: 00000000  r9 : 00000000  r8 : c0c65730
[   87.442204] r7 : 00000040  r6 : c0d08c8c  r5 : c0d08c20  r4 : ee8e2000
[   87.442222] r3 : 00000000  r2 : 00000000  r1 : 00000000  r0 : c01093d4
[   87.442244] Flags: nZCv  IRQs on  FIQs on  Mode SVC_32  ISA ARM  Segment none
[   87.442264] Control: 10c5387d  Table: 6ccd006a  DAC: 00000051
[   87.442292] CPU: 6 PID: 0 Comm: swapper/6 Not tainted 4.15.0-rc3-00022-g3da7bcdd695b #1101
[   87.442306] Hardware name: SAMSUNG EXYNOS (Flattened Device Tree)
[   87.442359] [<c01103a4>] (unwind_backtrace) from [<c010cac0>] (show_stack+0x10/0x14)
[   87.442404] [<c010cac0>] (show_stack) from [<c08620ac>] (dump_stack+0x98/0xc4)
[   87.442448] [<c08620ac>] (dump_stack) from [<c086801c>] (nmi_cpu_backtrace+0xb0/0x110)
[   87.442486] [<c086801c>] (nmi_cpu_backtrace) from [<c010f2ac>] (handle_IPI+0xd8/0x1b0)
[   87.442522] [<c010f2ac>] (handle_IPI) from [<c0101584>] (gic_handle_irq+0x98/0x9c)
[   87.442555] [<c0101584>] (gic_handle_irq) from [<c010d7b0>] (__irq_svc+0x70/0xb0)
[   87.442572] Exception stack(0xee8e3f68 to 0xee8e3fb0)
[   87.442605] 3f60:                   c01093d4 00000000 00000000 00000000 ee8e2000 c0d08c20
[   87.442642] 3f80: c0d08c8c 00000040 c0c65730 00000000 00000000 c0afe104 00000001 ee8e3fb8
[   87.442666] 3fa0: c01093d4 c01093d8 60010013 ffffffff
[   87.442708] [<c010d7b0>] (__irq_svc) from [<c01093d8>] (arch_cpu_idle+0x24/0x3c)
[   87.442747] [<c01093d8>] (arch_cpu_idle) from [<c0161c20>] (do_idle+0x190/0x230)
[   87.442783] [<c0161c20>] (do_idle) from [<c0162030>] (cpu_startup_entry+0x18/0x1c)
[   87.442814] [<c0162030>] (cpu_startup_entry) from [<4010192c>] (0x4010192c)
[   87.442839] NMI backtrace for cpu 7
[   87.442873] CPU: 7 PID: 0 Comm: swapper/7 Not tainted 4.15.0-rc3-00022-g3da7bcdd695b #1101
[   87.442888] Hardware name: SAMSUNG EXYNOS (Flattened Device Tree)
[   87.442925] PC is at arch_cpu_idle+0x24/0x3c
[   87.442954] LR is at arch_cpu_idle+0x20/0x3c
[   87.442973] pc : [<c01093d8>]    lr : [<c01093d4>]    psr: 60000013
[   87.442990] sp : ee8e5fb8  ip : 00000001  fp : c0afe104
[   87.443007] r10: 00000000  r9 : 00000000  r8 : c0c65730
[   87.443025] r7 : 00000080  r6 : c0d08c8c  r5 : c0d08c20  r4 : ee8e4000
[   87.443043] r3 : 00000000  r2 : 00000000  r1 : 00000000  r0 : c01093d4
[   87.443065] Flags: nZCv  IRQs on  FIQs on  Mode SVC_32  ISA ARM  Segment none
[   87.443086] Control: 10c5387d  Table: 6ccb006a  DAC: 00000051
[   87.443113] CPU: 7 PID: 0 Comm: swapper/7 Not tainted 4.15.0-rc3-00022-g3da7bcdd695b #1101
[   87.443127] Hardware name: SAMSUNG EXYNOS (Flattened Device Tree)
[   87.443179] [<c01103a4>] (unwind_backtrace) from [<c010cac0>] (show_stack+0x10/0x14)
[   87.443223] [<c010cac0>] (show_stack) from [<c08620ac>] (dump_stack+0x98/0xc4)
[   87.443265] [<c08620ac>] (dump_stack) from [<c086801c>] (nmi_cpu_backtrace+0xb0/0x110)
[   87.443303] [<c086801c>] (nmi_cpu_backtrace) from [<c010f2ac>] (handle_IPI+0xd8/0x1b0)
[   87.443340] [<c010f2ac>] (handle_IPI) from [<c0101584>] (gic_handle_irq+0x98/0x9c)
[   87.443374] [<c0101584>] (gic_handle_irq) from [<c010d7b0>] (__irq_svc+0x70/0xb0)
[   87.443391] Exception stack(0xee8e5f68 to 0xee8e5fb0)
[   87.443424] 5f60:                   c01093d4 00000000 00000000 00000000 ee8e4000 c0d08c20
[   87.443459] 5f80: c0d08c8c 00000080 c0c65730 00000000 00000000 c0afe104 00000001 ee8e5fb8
[   87.443482] 5fa0: c01093d4 c01093d8 60000013 ffffffff
[   87.443527] [<c010d7b0>] (__irq_svc) from [<c01093d8>] (arch_cpu_idle+0x24/0x3c)
[   87.443568] [<c01093d8>] (arch_cpu_idle) from [<c0161c20>] (do_idle+0x190/0x230)
[   87.443604] [<c0161c20>] (do_idle) from [<c0162030>] (cpu_startup_entry+0x18/0x1c)
[   87.443634] [<c0162030>] (cpu_startup_entry) from [<4010192c>] (0x4010192c)
[   89.892591] systemd-journald[225]: Received request to flush runtime journal from PID 1
[   90.289569] systemd-journald[225]: File /var/log/journal/7a4e9999283c4067874e93fe7b5d73eb/system.journal corrupted or uncleanly shut down




Best regards,
Krzysztof

--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Patch
diff mbox

diff --git a/include/linux/iversion.h b/include/linux/iversion.h
index 36ca56005c36..fe92bcc29f4c 100644
--- a/include/linux/iversion.h
+++ b/include/linux/iversion.h
@@ -168,6 +168,7 @@  static inline bool
 inode_maybe_inc_iversion(struct inode *inode, bool force)
 {
 	u64 cur, old, new;
+	int i = 0;
 
 	/*
 	 * The i_version field is not strictly ordered with any other inode
@@ -193,6 +194,8 @@  inode_maybe_inc_iversion(struct inode *inode, bool force)
 		if (likely(old == cur))
 			break;
 		cur = old;
+		if (++i > 1000)
+			WARN_ONCE(1, "Too much looping!");
 	}
 	return true;
 }
@@ -258,6 +261,7 @@  static inline u64
 inode_query_iversion(struct inode *inode)
 {
 	u64 cur, old, new;
+	int i = 0;
 
 	cur = inode_peek_iversion_raw(inode);
 	for (;;) {
@@ -277,6 +281,8 @@  inode_query_iversion(struct inode *inode)
 		if (likely(old == cur))
 			break;
 		cur = old;
+		if (++i > 1000)
+			WARN_ONCE(1, "Too much looping!");
 	}
 	return cur >> I_VERSION_QUERIED_SHIFT;
 }