Tejun Heo
45a140e587
kernfs: restructure removal path to fix possible premature return ...
The recursive nature of kernfs_remove() means that, even if
kernfs_remove() is not allowed to be called multiple times on the same
node, there may be race conditions between removal of parent and its
descendants. While we can claim that kernfs_remove() shouldn't be
called on one of the descendants while the removal of an ancestor is
in progress, such rule is unnecessarily restrictive and very difficult
to enforce. It's better to simply allow invoking kernfs_remove() as
the caller sees fit as long as the caller ensures that the node is
accessible.
The current behavior in such situations is broken. Whoever enters
removal path first takes the node off the hierarchy and then
deactivates. Following removers either return as soon as it notices
that it's not the first one or can't even find the target node as it
has already been removed from the hierarchy. In both cases, the
following removers may finish prematurely while the nodes which should
be removed and drained are still being processed by the first one.
This patch restructures so that multiple removers, whether through
recursion or direction invocation, always follow the following rules.
* When there are multiple concurrent removers, only one puts the base
ref.
* Regardless of which one puts the base ref, all removers are blocked
until the target node is fully deactivated and removed.
To achieve the above, removal path now first deactivates the subtree,
drains it and then unlinks one-by-one. __kernfs_deactivate() is
called directly from __kernfs_removal() and drops and regrabs
kernfs_mutex for each descendant to drain active refs. As this means
that multiple removers can enter __kernfs_deactivate() for the same
node, the function is updated so that it can handle multiple
deactivators of the same node - only one actually deactivates but all
wait till drain completion.
The restructured removal path guarantees that a removed node gets
unlinked only after the node is deactivated and drained. Combined
with proper multiple deactivator handling, this guarantees that any
invocation of kernfs_remove() returns only after the node itself and
all its descendants are deactivated, drained and removed.
v2: Draining separated into a separate loop (used to be in the same
loop as unlink) and done from __kernfs_deactivate(). This is to
allow exposing deactivation as a separate interface later.
Root node removal was broken in v1 patch. Fixed.
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2014-01-10 13:48:08 -08:00
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jikos/trivial
2013-11-15 16:47:22 -08:00
fsl /bestcomm
gpiolib: add missing declarations
2013-12-03 13:10:47 +01:00
Merge branch 'i2c/for-next' of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux
2013-11-18 15:50:07 -08:00
mfd/rtc: s5m: fix register updating by adding regmap for RTC
2013-12-12 18:19:26 -08:00
Merge branch 'next' of git://git.infradead.org/users/vkoul/slave-dma
2013-11-20 13:20:24 -08:00
usb: wusbcore: fix deadlock in wusbhc_gtk_rekey
2013-12-02 15:21:04 -08:00
KEYS: Fix multiple key add into associative array
2013-12-02 11:24:18 +00:00
blk-mq: ensure that we set REQ_IO_STAT so diskstats work
2013-11-19 09:25:07 -07:00
cmdline-parser: fix build
2013-11-15 09:32:23 +09:00
x86, build, icc: Remove uninitialized_var() from compiler-intel.h
2013-12-10 15:01:01 -08:00
sched: remove INIT_COMPLETION
2013-11-15 09:32:21 +09:00
Revert "cpufreq: suspend governors on system suspend/hibernate"
2013-12-08 01:04:17 +01:00
dcache: allow word-at-a-time name hashing with big-endian CPUs
2013-12-12 10:39:01 -08:00
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jikos/trivial
2013-11-15 16:47:22 -08:00
Merge commit 'dmaengine-3.13-v2' of git://git.kernel.org/pub/scm/linux/kernel/git/djbw/dmaengine
2013-11-16 12:02:36 +05:30
efivars, efi-pstore: Hold off deletion of sysfs entry until the scan is completed
2013-11-28 20:16:55 +00:00
firmware: Introduce request_firmware_direct()
2013-12-08 18:22:32 -08:00
consolidate simple ->d_delete() instances
2013-11-15 22:04:17 -05:00
ftrace, perf: Avoid infinite event generation loop
2013-11-19 16:57:40 +01:00
genetlink: make multicast groups const, prevent abuse
2013-11-19 16:39:06 -05:00
HID: hid-sensor-hub: Add logical min and max
2013-12-02 21:05:30 +00:00
iio: hid-sensors: Fix power and report state
2013-12-02 21:05:32 +00:00
include/linux/hugetlb.h: make isolate_huge_page() an inline
2013-12-12 18:19:25 -08:00
macvlan: introduce macvlan_dev_real_dev() helper function
2013-11-15 17:55:48 -05:00
revert "softirq: Add support for triggering softirq work on softirqs"
2013-11-15 09:32:22 +09:00
xen-netback: fix fragment detection in checksum setup
2013-12-05 20:31:40 -05:00
genirq: Correct fuzzy and fragile IRQ_RETVAL() definition
2013-11-19 19:06:41 +01:00
include/linux/kernel.h: make might_fault() a nop for !MMU
2013-12-12 18:19:26 -08:00
kernfs: restructure removal path to fix possible premature return
2014-01-10 13:48:08 -08:00
PCI: Disable Bus Master only on kexec reboot
2013-12-07 14:20:28 -07:00
kfifo API type safety
2013-11-15 09:32:23 +09:00
kernfs: s/sysfs_dirent/kernfs_node/ and rename its friends accordingly
2013-12-11 15:28:36 -08:00
llists: move llist_reverse_order from raid5 to llist.c
2013-11-15 09:32:22 +09:00
mm: do not allocate page->ptl dynamically, if spinlock_t fits to long
2013-12-20 12:25:45 -08:00
math64: Add mul_u64_u32_shr()
2013-12-11 15:52:34 +01:00
drivers: base: Add prototype declaration to the header file
2013-12-20 12:20:26 -08:00
micrel: add support for KSZ8041RNLI
2013-12-11 13:54:13 -05:00
aio/migratepages: make aio migrate pages sane
2013-12-21 17:56:08 -05:00
mm: do not allocate page->ptl dynamically, if spinlock_t fits to long
2013-12-20 12:25:45 -08:00
mm: do not allocate page->ptl dynamically, if spinlock_t fits to long
2013-12-20 12:25:45 -08:00
net: unix: allow set_peek_off to fail
2013-12-10 21:45:15 -05:00
net: make neigh_priv_len in struct net_device 16bit instead of 8bit
2013-12-12 15:35:09 -05:00
NFS client bugfixes
2013-12-05 13:05:48 -08:00
NFS: Enabling v4.2 should not recompile nfsd and lockd
2013-11-19 16:20:40 -05:00
Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net
2013-12-15 11:56:47 -08:00
net/phy: Add the autocross feature for forced links on VSC82x4
2013-11-20 22:09:19 -05:00
pstore: Don't allow high traffic options on fragile devices
2013-12-20 13:12:01 -08:00
kexec: migrate to reboot cpu
2013-12-18 19:04:50 -08:00
sched/fair: Rework sched_fair time accounting
2013-12-11 15:52:35 +01:00
seq_file: introduce seq_setwidth() and seq_pad()
2013-11-15 09:32:20 +09:00
take read_seqbegin_or_lock() and friends to seqlock.h
2013-11-15 22:04:17 -05:00
security: shmem: implement kernel private shmem inodes
2013-12-02 11:24:19 +00:00
Revert "net: Handle CHECKSUM_COMPLETE more adequately in pskb_trim_rcsum()."
2013-12-02 17:26:05 -05:00
slab.h: remove duplicate kmalloc declaration and fix kernel-doc warnings
2013-11-24 11:01:16 -08:00
Merge branch 'slab/next' of git://git.kernel.org/pub/scm/linux/kernel/git/penberg/linux
2013-11-22 08:10:34 -08:00
Merge branch 'slab/next' of git://git.kernel.org/pub/scm/linux/kernel/git/penberg/linux
2013-11-22 08:10:34 -08:00
kernel-provide-a-__smp_call_function_single-stub-for-config_smp-fix
2013-11-15 09:32:22 +09:00
mm, hugetlb: convert hugetlbfs to use split pmd lock
2013-11-15 09:32:14 +09:00
This batch of changes is mostly clean ups and small bug fixes.
2013-11-16 12:23:18 -08:00
kernfs: s/sysfs_dirent/kernfs_node/ and rename its friends accordingly
2013-12-11 15:28:36 -08:00
ARM: tegra: Provide dummy powergate implementation
2013-11-25 13:34:51 -08:00
ftrace, perf: Avoid infinite event generation loop
2013-11-19 16:57:40 +01:00
usb: xhci: Link TRB must not occur within a USB payload burst
2013-12-02 11:57:10 -08:00
