| CVE |
Vendors |
Products |
Updated |
CVSS v3.1 |
| In the Linux kernel, the following vulnerability has been resolved:
memcg: bypass the reclaim and oom killer for dying tasks once oom_reaper is done
At Meta, we are seeing instances where an OOM killed job is stuck in the
exit path for several hours. In one particular case, the job was stuck
for more than 8 hours and I had to manually remove the memory.max limits
to allow the process to exit.
The job was a single process job and had ~55 GiB memory.max and zswap
enabled. It had almost 0 anon in memory and ~111 GiB in zswap compressed
to ~51 GiB zswap pool (i.e. almost all of memory.current was zswap).
Nothing was left on the LRUs to reclaim.
On further inspection, I observed ~20k threads of that process stuck with
the following stack:
[<0>] mem_cgroup_out_of_memory+0x4e/0xa0
[<0>] charge_memcg+0x8bf/0x990
[<0>] mem_cgroup_swapin_charge_folio+0x4e/0x80
[<0>] __read_swap_cache_async+0x10c/0x260
[<0>] swapin_readahead+0x116/0x3f0
[<0>] do_swap_page+0x13c/0x1ce0
[<0>] handle_mm_fault+0x61d/0x11f0
[<0>] do_user_addr_fault+0x3e7/0x6d0
[<0>] exc_page_fault+0x8f/0x110
[<0>] asm_exc_page_fault+0x22/0x30
[<0>] __get_user_8+0x14/0x20
[<0>] futex_cleanup+0x27/0x1c0
[<0>] futex_exit_release+0x47/0x60
[<0>] do_exit+0x107/0x940
[<0>] do_group_exit+0x81/0xa0
[<0>] get_signal+0x2b1/0x6e0
[<0>] arch_do_signal_or_restart+0x1a/0x1c0
[<0>] exit_to_user_mode_loop+0xa8/0x1c0
[<0>] do_syscall_64+0x152/0x250
[<0>] entry_SYSCALL_64_after_hwframe+0x4b/0x53
In addition the dmesg was filled with "Out of memory and no killable
processes..." messages.
I have no idea why oom reaper was not able to reap/unmap the process. My
guess is that since oom reaper tries to acquire mmap_lock in read mode
limited number of times and then gives up, there might be a thread of that
process which had mmap_lock in write mode at that time.
My initial suspicion was the futex_cleanup and kernel page fault causing
infinite fault and charge retries but that was put to rest in previous
discussions happened on similar problem [1].
My current theory is that it is just a simple slow serialization behind
the oom_lock. Unlike page allocator, memcg charge code takes the oom_lock
without the "try". Though memcg oom code uses mutex_lock_killable(), note
that in the call stack get_signal() consumes SIGKILL (or
sigdelset(SIGKILL)) before calling do_group_exit(). So this
mutex_lock_killable() is just a mutex_lock() here. Therefore 10s of
thousands of threads are waiting on oom_lock and one by one they get
-EFAULT from get_user() in the futex cleanup code and bails out.
Discussion from [1] led to commit a75ffa26122b ("memcg, oom: do not bypass
oom killer for dying tasks") which routes dying tasks into the OOM path
precisely so the oom_reaper can reap their mm and free the memory
asynchronously. But the reaper is best-effort and one-shot: if it cannot
take mmap_lock for read (e.g. a sibling thread holds it for write) it
sets MMF_OOM_SKIP and never retries, leaving only the glacial
oom_lock-serialized synchronous drain.
Once MMF_OOM_SKIP is set there is no more asynchronous reclaim coming for
the mm, so a dying task charging against it has nothing left to wait for:
it frees its memory only once it finishes exiting. Running reclaim and
the (no-victim) OOM killer for it is then pointless, and doing it for 10s
of thousands of exiting threads is what serializes them behind oom_lock.
So before reclaim, if current is an OOM victim whose reaper is done, fail
the charge.
Reproduced with 20k threads, each parking a robust futex head on its own
zswapped page, OOM-group-killed while a sibling holds mmap_lock for write
so the reaper gives up and sets MMF_OOM_SKIP. Tested on next-20260728 and
baseline show ~90 seconds exit time while with the patch the exit time
reduced to ~3 seconds. |
| In the Linux kernel, the following vulnerability has been resolved:
scsi: qla2xxx: Fix response queue over-consumption in __qla_consume_iocb()
qla24xx_process_response_queue() advances ring_ptr past the head IOCB
before dispatching, so by the time __qla_consume_iocb() runs, ring_ptr
already points at the first continuation IOCB. The function however
looped purex->entry_count times starting at ring_ptr. As entry_count
includes the head, this consumed one entry too many: it stamped
RESPONSE_PROCESSED on the next, unrelated IOCB and advanced the ring
past it, silently dropping a legitimate firmware response. The head
IOCB's signature was also never marked.
Mark the head processed and account for it, then consume only the
entry_count - 1 continuation IOCBs, matching __qla_copy_purex_to_buffer(). |
| In the Linux kernel, the following vulnerability has been resolved:
mm/secretmem: properly account locked pages
secretmem accounts folios by treating memory as if it were mlock()'d and
thus limited by the RLIMIT_MEMLOCK limit.
However the folios are unevictable and remain so until the inode is
evicted, eliminating usual mlock() semantics - mapping folios then
unmapping them does not clear their unevictable state, since it depends on
AS_UNEVICTABLE, not PG_mlocked.
A user can therefore easily work around the RLIMIT_MEMLOCK limit - simply
map then unmap and VmLck no longer counts the secretmem range. Worse,
folios are not accounted in the process's RSS, meaning the OOM killer
won't know to kill the process.
Repeatedly mapping/unmapping (or forking) can then result in the
consumption of all available system memory with unevictable folios and
cause system instability.
A secretmem fd can be passed between processes and over fork so a
per-process limit simply does not make sense, so follow the precedent set
by io_uring, perf, skbuff, iommufd and xdp by tracking the number of
locked pages in user_struct->locked_vm.
Since the scope tracked is actually inode lifetime, the RLIMIT_MEMLOCK
applies per-user not per-process, so it doesn't make sense to bypass for
users with CAP_IPC_LOCK, therefore remove this bypass.
There is simply no reason to carry on marking the mapping as mlock()'d
since it's misleading and the lifecycle is now correctly handled, so
remove this too.
Note that secretmem does not support any form of truncation (including
hole punching) and the folios are unreclaimable, so the folios need only
be accounted on fault and unaccounted on inode destruction.
__secretmem_account_pages() is more or less a duplicate of the code that
io_uring etc. use, but since this is a bug fix that needs backporting,
defer any de-duplication efforts to a follow-up.
test_mlock_limit() asserts mlock_future_ok() on mmap(), however this has
been removed, so remove the test altogether for the fix. A new test will
be sent separately for upstream. |
| In the Linux kernel, the following vulnerability has been resolved:
drm/sysfb: simpledrm: Improve stride validation
Validate the computed stride against the maximum value INT_MAX. |
| In the Linux kernel, the following vulnerability has been resolved:
ACPI: bus: Introduce acpi_bus_get_primary_device()
The function used for obtaining the first "physical" device for which
the given ACPI one is the ACPI companion, acpi_get_first_physical_node(),
may return a stale device pointer (mostly in theory) because
acpi_unbind_one() may run as a whole after dropping the ACPI device's
physical_node_lock in acpi_get_first_physical_node() and before it
returns. The last reference to the "physical" device may be dropped
then before the pointer to it is returned to the caller.
If that happens and the acpi_get_first_physical_node() caller invokes
get_device() on the pointer obtained from it, which is done by the
majority of its callers, a use-after-free will occur.
To prepare for addressing this problem, introduce a new function for
getting the first "physical" device associated with the given ACPI one
(the "primary physical device") that will also reference count the
device in question before returning a pointer to it.
Make that new function and acpi_get_first_physical_node() share the
physical node list lookup code.
No intentional functional impact. |
| In the Linux kernel, the following vulnerability has been resolved:
arm64: hibernate: mask DAIF before restoring hibernated kernel
The arm64 hibernate code manages the exception masking in an unsound
way, leading to potential crashes and/or warnings during resume.
When a hibernation image is saved in `swsusp_arch_suspend()`, all DAIF
exceptions are masked (by virtue of `local_daif_save()`), and the
suspended image is saved assuming that all DAIF exceptions will remain
masked when the image is restored.
When a hibernation image is resumed by `swsusp_arch_resume()`, only
interrupts are masked (by virtue of `local_irq_disable()` in
`resume_target_kernel()`). When pseudo-NMI is enabled the DAIF.IF bits
will be clear, and regardless of pseudo-NMI the DAIF.DA bits will be
clear.
This means that there are two problems:
(1) It is possible to take Debug, SError, or pseudo-NMI exceptions
during the resume process. This is unsafe, as during the resume
process both the old ane new kernels will tranisently be in an
inconsistent state, and swsusp_arch_suspend_exit() won't retain
an executable mapping of any exception vectors.
Any exception taken here will be fatal and silent.
(2) When re-entering the resumed kernel, some DAIF bits will be clear
unexpectedly. This permits Debug, SError, or pseudo-NMI exceptions
to be taken for a short period while the resumed kernel is not yet
in a consistent state.
This is detected by CONFIG_ARM64_DEBUG_PRIORITY_MASKING.
Avoid these issues by masking all DAIF exceptions during resume. |
| In the Linux kernel, the following vulnerability has been resolved:
clk: eswin: Zero-initialize stack-allocated clk_init_data
eswin_clk_register_pll() and eswin_register_clkdiv() declare a struct
clk_init_data on the stack and only initialize some of its fields
(parent_data respectively parent_hws). clk_core_populate_parent_map()
checks parent_names first and parent_data second before falling back
to parent_hws, so leftover stack garbage in the uninitialized fields
hijacks parent resolution and the clk core dereferences a bogus
pointer:
Unable to handle kernel NULL pointer dereference at virtual address 000000000000000c
Oops [#1]
epc : __clk_register+0x31a/0x7f0
[<ffffffff805dc774>] __clk_register+0x31a/0x7f0
[<ffffffff805dcd76>] devm_clk_hw_register+0x2a/0x94
[<ffffffff805e319a>] eswin_register_clkdiv+0x80/0xd0
[<ffffffff805e34a0>] eswin_clk_register_clks+0x162/0x1a0
[<ffffffff805e3736>] eic7700_clk_probe+0x146/0x180
[<ffffffff8065d23c>] platform_probe+0x3c/0x7a
Observed on EIC7700 hardware (with the driver backported to a 6.17
tree); whether the bug triggers depends entirely on what the stack
happens to contain when the registration helpers run.
Zero-initialize both structures. |
| In the Linux kernel, the following vulnerability has been resolved:
wifi: ath11k: cap out-of-range rx MCS instead of leaving bogus rate
ath11k can receive HT/VHT/HE frames whose reported MCS is above the
maximum that can be expressed in the corresponding mac80211 rate space
(e.g. an HE frame reported with MCS 12, while HE tops out at MCS 11).
The frame itself is valid and decodes correctly, but for such a frame
ath11k_dp_rx_h_rate() leaves rx_status->rate_idx set to the out-of-range
value and never assigns rx_status->encoding, so it stays RX_ENC_LEGACY
from the ath11k_dp_rx_h_ppdu() initialization. Once that frame reaches
mac80211 it trips the rate sanity check and the frame is dropped with a
splat:
ath11k_pci 0000:03:00.0: Received with invalid mcs in HE mode 12
WARNING: CPU: 0 PID: 0 at net/mac80211/rx.c:5433 ieee80211_rx_list+0xb0a/0xe90 [mac80211]
Dropping the frame would discard otherwise valid data, so instead cap the
reported MCS to the maximum the rate space can express and deliver the
frame. Set rx_status->encoding before the range check and assign rate_idx
from the capped value, so a frame with an out-of-range MCS no longer
leaves partial or bogus rate metadata behind. Also downgrade the logging
level since they are not treated as invalid frames now. The only loss is
that such a frame is reported as the capped MCS in the rx rate statistics.
Tested-on: WCN6855 hw2.1 PCI WLAN.HSP.1.1-03125-QCAHSPSWPL_V1_V2_SILICONZ_LITE-3.6510.41 |
| In the Linux kernel, the following vulnerability has been resolved:
HID: hidpp: fix potential UAF in hidpp_connect_event()
If input_register_device() fails, we call input_free_device(), but keep
stale pointer to the old device in hidpp->input, which could potentially
lead to UAF. Fix that by resetting it to NULL before returning from
hidpp_connect_event(). |
| In the Linux kernel, the following vulnerability has been resolved:
net: stmmac: xgmac2: disable RBUE in default RX interrupt mask
Enabling the RX Buffer Unavailable (RBUE) interrupt is counterproductive
and can trigger a MAC interrupt storm under heavy RX pressure. When the
DMA runs out of RX descriptors it fires RBUE continuously until software
refills the ring.
However, RBUE is redundant: the normal RX completion interrupt (RIE)
already triggers NAPI, which processes completed descriptors and refills
the ring, causing the DMA to resume. The RBUE handler itself only sets
handle_rx - the same outcome as RIE.
On Agilex5 under heavy RX pressure, the MAC interrupt (which includes
RBUE) was observed firing 1,821,811,555 times against only 2,618,627
actual RX completions - a ~695x ratio - confirming the severity of the
storm.
RBUE does not provide OOM recovery. If page_pool is exhausted,
stmmac_rx_refill() cannot advance the DMA tail pointer, the DMA stays
suspended, and RBUE fires again on the next NAPI completion - a storm
with no forward progress. This patch trades that storm for a clean
stall with the same RX outcome. Proper OOM recovery is a pre-existing
gap outside the scope of this fix.
Note: as a consequence of disabling RBUE, the rx_buf_unav_irq ethtool
counter will always read 0 on XGMAC2 devices. This behaviour is already
inconsistent across DWMAC core versions.
Remove RBUE from XGMAC_DMA_INT_DEFAULT_EN and XGMAC_DMA_INT_DEFAULT_RX
to prevent the interrupt storm while keeping normal RX handling intact. |
| In the Linux kernel, the following vulnerability has been resolved:
clocksource/drivers/timer-sun4i: Advertise a real minimum delta
sun4i_clkevt_next_event() compensates for the timer stop/start
synchronization delay by programming evt - TIMER_SYNC_TICKS into the
hardware interval register. The clockevent device currently advertises
TIMER_SYNC_TICKS as min_delta_ticks, so the clockevents core is allowed
to call set_next_event() with evt == TIMER_SYNC_TICKS.
That programs a zero-tick interval. With oneshot/highres/nohz timer
operation this can leave the next event stuck, which was observed as a
boot hang on Allwinner D1 after the clockevents core started reusing
forced minimum-delta events.
Advertise one extra tick instead, so the smallest event accepted by the
core still programs at least one hardware tick after the synchronization
compensation. |
| In the Linux kernel, the following vulnerability has been resolved:
nfsd: convert nfsd_net boolean flags to unsigned long flags word
nfsd_net contains several boolean fields that are accessed from
concurrent contexts without serialization. In particular,
nfsd4_end_grace() guards its drain path with a plain bool:
if (nn->grace_ended)
return;
nn->grace_ended = true;
The read and the write are independent, and nothing in struct
nfsd_net serializes them. At least two contexts can reach this
code with no lock held:
laundromat path
laundry_wq kworker
nfs4_laundromat()
nfsd4_end_grace()
RECLAIM_COMPLETE path
nfsd compound kthread
nfsd4_reclaim_complete()
inc_reclaim_complete()
nfsd4_end_grace()
Both callers can observe grace_ended == false on different CPUs,
both store true, and both proceed into nfsd4_record_grace_done(),
which invokes the active client_tracking_ops->grace_done callback.
For tracking ops that drain reclaim_str_hashtbl (legacy_tracking_ops
via nfsd4_recdir_purge_old, and the cld v1+ ops via
nfsd4_cld_grace_done), grace_done calls nfs4_release_reclaim(),
which walks every bucket of reclaim_str_hashtbl with no lock and
calls nfs4_remove_reclaim_record() (list_del + kfree) on each
entry. Two concurrent walkers corrupt the list and double-free
every nfs4_client_reclaim. A concurrent nfsd4_find_reclaim_client()
iterating the same bucket reads through freed memory.
A third call site exists in nfs4_state_start_net() on the
skip_grace startup path, but it runs under nfsd_mutex before any
client has connected and before the laundromat's first delayed
work fires, so it cannot race with the two callers above.
Replace the scattered boolean fields in nfsd_net with a single
unsigned long flags word and an enum nfsd_net_flag for the bit
positions. The grace_ended race is fixed by using
test_and_set_bit(), which is atomic on all architectures. The
remaining flags (grace_end_forced, in_grace, somebody_reclaimed,
track_reclaim_completes, nfsd_net_up, lockd_up) are converted to
use test_bit/set_bit/clear_bit for consistency. This avoids
sub-word cmpxchg issues on architectures like Hexagon that only
support word-sized atomic operations. |
| In the Linux kernel, the following vulnerability has been resolved:
staging: media: tegra-video: fix of_node_put() on VIP parse errors
tegra_vip_channel_of_parse() initializes np from dev->of_node without
taking a reference, but its error paths drop one through the
err_node_put label. This underflows the refcount of the VIP device's
OF node when endpoint parsing fails on a malformed device tree.
The only reference the function takes on np is the success-path
of_node_get() stored in vip->chan.of_node, and that one is already
released by the tegra_vip_init() error path and by tegra_vip_exit().
Return errors directly instead of jumping to the bogus cleanup label. |
| In the Linux kernel, the following vulnerability has been resolved:
ipv6: use RCU iterator to dump route exceptions
rt6_nh_dump_exceptions() uses hlist_for_each_entry() to iterate over
RCU-protected exception lists. The caller holds rcu_read_lock(), but does
not hold rt6_exception_lock, so rt6_insert_exception() can concurrently
add an entry with hlist_add_head_rcu().
KCSAN reports this race (irrelevant details omitted):
==================================================================
BUG: KCSAN: data-race in rt6_insert_exception / rt6_nh_dump_exceptions
write (marked) to 0xffff8a7c44c59620 of 8 bytes by interrupt on cpu 5:
rt6_insert_exception+0x3bb/0x760
__ip6_rt_update_pmtu+0x4fe/0x750
ip6_sk_update_pmtu+0x19a/0x3b0
udpv6_err+0x3ff/0x800
icmpv6_notify+0x1e1/0x440
icmpv6_rcv+0x8c0/0xab0
ip6_protocol_deliver_rcu+0x616/0x840
ip6_input_finish+0xb9/0x160
...
entry_SYSCALL_64_after_hwframe+0x77/0x7f
read to 0xffff8a7c44c59620 of 8 bytes by task 549 on cpu 14:
rt6_nh_dump_exceptions+0xb3/0x260
rt6_dump_route+0x53e/0x5f0
fib6_dump_node+0x6d/0xf0
fib6_walk_continue+0x290/0x2d0
fib6_dump_table+0x28d/0x360
inet6_dump_fib+0x37d/0x620
rtnl_dumpit+0x7b/0xd0
netlink_dump+0x3ae/0x7e0
...
entry_SYSCALL_64_after_hwframe+0x77/0x7f
4 locks held by dumper/549:
...
#1: (rcu_read_lock){....}-{1:3}, at: inet6_dump_fib+0x88/0x620
#2: (&tb->tb6_lock){+.-.}-{3:3}, at: fib6_dump_table+0x1e9/0x360
#3: (rcu_read_lock){....}-{1:3}, at: rt6_dump_route+0x483/0x5f0
value changed: 0xffff8a7c44e05700 -> 0xffff8a7c45d60100
Reported by Kernel Concurrency Sanitizer on:
CPU: 14 UID: 0 PID: 549 Comm: dumper Not tainted
7.2.0-rc7-virtme #38 PREEMPT(lazy)
...
Use hlist_for_each_entry_rcu() to safely iterate over the exception list. |
| In the Linux kernel, the following vulnerability has been resolved:
svcrdma: Reject Write/Reply chunks with segcount 0
A peer can send a Write or Reply chunk whose segcount field is zero.
xdr_check_write_chunk() only rejects segcount > rc_maxpages, so zero
passes the range check, and xdr_inline_decode(stream, 0) returns the
current (non-NULL) cursor without advancing. The function returns
true and pcl_alloc_write() then links a struct svc_rdma_chunk with
ch_segcount == 0 onto rc_write_pcl or rc_reply_pcl.
An earlier patch in this series made pcl_for_each_segment() safe for
ch_segcount == 0, so this no longer drives the memory walk it used
to. Rejecting the malformed frame at the decode boundary is still
worthwhile as defense in depth: it keeps degenerate zero-segment
chunks off the parsed chunk lists entirely, so any future consumer
that walks ch_segments directly cannot observe one, and it makes the
zero-floor easy to backport to trees where the macro change is more
intrusive. RFC 8166 has no meaning for a Write/Reply chunk that
describes no remote buffer, so no legitimate client is affected.
xdr_check_reply_chunk() funnels Reply chunks through
xdr_check_write_chunk() and inherits the same rejection.
pcl_alloc_write() also links each chunk onto the parsed chunk list
before filling its segment array. If a future change weakens the
segcount-0 rejection, an incomplete chunk is visible to consumers
during the fill loop. Reorder so that list_add_tail() follows the
segment fill loop, ensuring only fully-populated chunks appear on
the list. |
| In the Linux kernel, the following vulnerability has been resolved:
drm/gud: validate TV mode names before creating enum property
The GUD protocol returns TV mode names as fixed-size
GUD_CONNECTOR_TV_MODE_NAME_LEN entries and requires each name to be
NUL-terminated.
gud_connector_add_tv_mode() currently passes each fixed-size entry
directly to drm_mode_create_tv_properties_legacy(), which eventually
reaches drm_property_add_enum() and strlen(). If a device returns an
entry without a terminating NUL byte, strlen() reads past the end of
the slot and can run beyond the allocated buffer, triggering an
out-of-bounds read.
Validate that each returned TV mode name contains a NUL terminator
within its fixed-size slot before passing it to the DRM property code.
If a malformed entry is found, reject the device response with -EIO.
This fixes the out-of-bounds read without changing the handling of
valid devices, and avoids silently truncating malformed protocol data. |
| Rojo's "rojo serve" HTTP API (default port 34872) has no Host/Origin header validation, making it vulnerable to DNS rebinding. A malicious webpage can read all project source, write malicious code to files on disk, and launch local programs via opener::open() with no user interaction beyond visiting the page. |
| In the Linux kernel, the following vulnerability has been resolved:
s390/vfio-ap: fix potential use of uninitialized apm_filtered bitmap
The DECLARE_BITMAP(apm_filtered, AP_DEVICES) macro allocates the bitmap
on the stack without zero-initializing it.
In vfio_ap_mdev_hot_plug_cfg(), the vfio_ap_mdev_filter_matrix() function
is only called to initialize and populate apm_filtered if either
filter_adapters or filter_domains is true. If the hot plug configuration
change only adds control domains (meaning filter_cdoms is true, but
filter_adapters and filter_domains are both false),
vfio_ap_mdev_filter_matrix() is bypassed.
Consequently, apm_filtered is passed to reset_queues_for_apids() with
uninitialized stack garbage. This can cause reset_queues_for_apids() to
interpret arbitrary stack garbage bits as valid APIDs to reset, potentially
performing unintended guest hardware queue resets.
Fix this by zero-initializing the apm_filtered bitmap at the beginning of
vfio_ap_mdev_hot_plug_cfg() using bitmap_zero(). |
| In the Linux kernel, the following vulnerability has been resolved:
memcg: make the v1 soft limit knob inert
The v1 soft limit has been deprecated since v6.12 and nobody has reported
depending on it. Start the removal by decoupling the interface from the
implementation: keep memory.soft_limit_in_bytes, but ignore writes to it
and always report the maximum value on read similar to what
memory.kmem.limit_in_bytes already does.
Writes are still parsed, so malformed input keeps returning -EINVAL. The
knob now also behaves the same everywhere: it used to return -EOPNOTSUPP
on PREEMPT_RT, where soft limit reclaim has always been disabled.
This also fixes the syzbot report linked below. Soft limit reclaim is the
only caller that runs shrink_lruvec() from kswapd against a specific
memcg, so it is the only way to reach lru_gen_shrink_lruvec() and in turn
set_mm_walk(), which warns when called from kswapd. |
| In the Linux kernel, the following vulnerability has been resolved:
ext4: fix circular lock dependency in ext4_ext_migrate
Move iput(tmp_inode) after ext4_writepages_up_write() to avoid a
circular lock dependency between s_writepages_rwsem and sb_internal
(freeze protection).
The deadlock scenario:
CPU0 (EXT4_IOC_MIGRATE) CPU1 (orphan cleanup during mount)
---- ----
ext4_ext_migrate()
ext4_writepages_down_write()
s_writepages_rwsem (write)
ext4_evict_inode()
sb_start_intwrite() [sb_internal]
...
ext4_writepages()
s_writepages_rwsem (read) [BLOCKED]
iput(tmp_inode)
ext4_evict_inode()
sb_start_intwrite() [BLOCKED]
The tmp_inode is a temporary inode with nlink=0 created solely for
building the extent tree. Its eviction does not require
s_writepages_rwsem protection, so deferring iput() until after
releasing the rwsem is safe. |
