io_uring: restore RCU read section in io_req_local_work_add()
The task-work refactor that moved io_req_local_work_add() out of
io_uring.c into the new io_uring/tw.c dropped the whole-body
guard(rcu)() that used to cover the function body.
For DEFER_TASKRUN rings the ring teardown still relies on that RCU read
section pairing with its grace period:
/* pairs with RCU read section in io_req_local_work_add() */
if (ctx->flags & IORING_SETUP_DEFER_TASKRUN)
synchronize_rcu();
io_ring_ctx_free(ctx);
io_req_local_work_add() keeps dereferencing ctx after mpscq_push() has
published the request to the work list (ctx->cq_wait_nr, and
ctx->submitter_task in the final wake_up_state()), without holding a ctx
reference across that window. The RCU read section was the only thing
guaranteeing an in-flight adder had finished touching ctx before
io_ring_ctx_free() ran; synchronize_rcu() only waits for readers that
are actually inside an RCU read-side critical section. With the guard
gone the grace period no longer pairs with anything on the add side, so
ctx can be freed and reused while io_req_local_work_add() is still using
it.
smb: client: mask server-provided mode to 07777 in modefromsid
When modefromsid is active, parse_dacl() applies the server-provided
sub_auth[2] value from the NFS mode SID to cf_mode without masking to
07777. Apply the correct masking, same as in the read path.
Fixes: e2f8fbfb8d09c ("cifs: get mode bits from special sid on stat") Signed-off-by: Norbert Manthey <nmanthey@amazon.de> Assisted-by: Kiro:claude-opus-4.6 Cc: stable@vger.kernel.org Signed-off-by: Steve French <stfrench@microsoft.com>
Matt Bobrowski [Thu, 9 Jul 2026 02:53:16 +0000 (02:53 +0000)]
bpf: Fix UAF in sock clone early bailouts
Similar to recent commit 9b51a6155d14 ("bpf,fork: wipe ->bpf_storage
before bailouts that access it"), sk_clone() performs an initial
shallow copy of the socket field ->sk_bpf_storage via sock_copy()
for the cloned socket newsk.
If sk_clone() bails out early (e.g. if sk_filter_charge() fails) prior
to calling bpf_sk_storage_clone(), newsk->sk_bpf_storage still points
to the parent socket's BPF local storage. When newsk is subsequently
freed via sk_free(), the deallocation path (__sk_destruct() ->
bpf_sk_storage_free()) destroys the parent socket's BPF local storage,
leading to a use-after-free (UAF) on the parent socket.
Fix this by resetting newsk->sk_bpf_storage to NULL immediately after
sock_copy() in sk_clone(), and remove the now redundant initialization
from bpf_sk_storage_clone().
Fixes: 6ac99e8f23d4 ("bpf: Introduce bpf sk local storage") Fixes: f12dd75959b0 ("bpf: net: Set sk_bpf_storage back to NULL for cloned sk") Signed-off-by: Matt Bobrowski <mattbobrowski@google.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Reviewed-by: Kuniyuki Iwashima <kuniyu@google.com> Acked-by: Daniel Borkmann <daniel@iogearbox.net> Link: https://lore.kernel.org/bpf/20260709025316.999913-1-mattbobrowski@google.com
When the HDA source was reorganized some of the cirrus* files were moved
into a new 'side-codecs' subdirectory. But MAINTAINERS wasn't updated to
add a cirrus* file reference to cover these moved files.
ALSA: hda/cirrus_scodec: Make Kconfig visible if KUNIT
Make the Kconfig item for cirrus_scodec visible if CONFIG_KUNIT is
enabled. This is so that its KUnit test can be enabled by KUnit
scripts without requiring a large amount of irrelevant additional
components.
The general rule for KUNIT_ALL_TESTS is that it should only enable
tests for components that are already selected. However, the UML
environment does not support ACPI, which means the HDA codec drivers
that use cirrus_scodec cannot be selected. But cirrus_scodec does not
need ACPI.
By making the Kconfig option visible if CONFIG_KUNIT, the KUnit test
can be enabled with only the minimal set of functionality that is
required for cirrus_scodec.
This is still compliant with the KUNIT_ALL_TESTS rule "only tests
for enabled modules" because by default cirrus_scodec will only be
enabled if the drivers that use it are enabled. It must be
intentionally enabled to force it to be included for testing.
Merge tag 'net-7.2-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net
Pull networking fixes from Paolo Abeni:
"Including fixes from netfilter, Bluetooth and batman-adv.
Current release - regressions:
- bluetooth: fix using chan->conn as indication to no remote netdev
Current release - new code bugs:
- netfilter: cap to maximum number of expectation per master on
updates
Previous releases - regressions:
- bluetooth:
- fix UAF of hci_conn_params in add_device_complete
- fix null ptr deref in hci_abort_conn()
- igmp: remove multicast group from hash table on device destruction
- batman-adv: prevent TVLV OOB check overflow
- eth: mlx5/mlx5e:
- fix off-by-one in single-FDB error rollback
- skip peer flow cleanup when LAG seq is unavailable
- fix crashes in dynamic per-channel stats and HV VHCA agent
- eth: mana: Sync page pool RX frags for CPU
Previous releases - always broken:
- netfilter:
- mark malformed IPv6 extension headers for hotdrop
- terminate table name before find_table_lock()
- ipvs: use parsed transport offset in TCP state lookup
* tag 'net-7.2-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (94 commits)
macsec: don't read an unset MAC header in macsec_encrypt()
dibs: loopback: validate offset and size in move_data()
octeontx2-af: fix VF bringup affecting PF promiscuous state
ethtool: rss: Fix hfunc and input_xfrm parsing on big endian
net/mlx5: Fix L3 tunnel entropy refcount leak
net: macb: drop in-flight Tx SKBs on close
net: mana: Sync page pool RX frags for CPU
net: mana: Validate the packet length reported by the NIC
selftests/net: fix EVP_MD_CTX leak in tcp_mmap
ipvs: ensure inner headers in ICMP errors are in headroom
ipvs: use parsed transport offset in SCTP state lookup
ipvs: use parsed transport offset in TCP state lookup
ipvs: pass parsed transport offset to state handlers
netfilter: handle unreadable frags
netfilter: flowtable: support IPIP tunnel with direct xmit
netfilter: flowtable: IPIP tunnel hardware offload is not yet support
netfilter: flowtable: use dst in this direction when pushing IPIP header
netfilter: ipset: allocate the proper memory for the generic hash structure
netfilter: ipset: cleanup the add/del backlog when resize failed
netfilter: ipset: exclude gc when resize is in progress
...
Eric Biggers [Thu, 9 Jul 2026 02:29:54 +0000 (22:29 -0400)]
crypto: aes - Fix conditions for selecting MAC dependencies
Starting in commit 7137cbf2b5c9 ("crypto: aes - Add cmac, xcbc, and
cbcmac algorithms using library"), the aes module (CRYPTO_AES) supports
CBC based MACs using the corresponding library functions.
To avoid including unneeded functionality, that support honors the
existing CRYPTO_CMAC, CRYPTO_XCBC, and CRYPTO_CCM kconfig options. The
dependencies are selected if at least one of those is enabled.
However, the select statements don't correctly handle the case where
CRYPTO_AES=y and (for example) CRYPTO_CMAC=m. In that case the
dependencies get selected at level 'm', due to how the kconfig language
works. That causes a linker error.
Fix this by changing the selection conditions to use '!= n'.
A similar issue also exists for CRYPTO_LIB_AES's conditional selection
of CRYPTO_LIB_UTILS. The same '!= n' would work, but instead just make
CRYPTO_LIB_AES always select CRYPTO_LIB_UTILS. CRYPTO_LIB_UTILS is
lightweight, and it's needed by most AES modes and many other things.
Fixes: 7137cbf2b5c9 ("crypto: aes - Add cmac, xcbc, and cbcmac algorithms using library") Fixes: 309a7e514da7 ("lib/crypto: aes: Add support for CBC-based MACs") Cc: stable@vger.kernel.org Link: https://patch.msgid.link/20260709022954.45113-1-ebiggers@kernel.org Signed-off-by: Eric Biggers <ebiggers@kernel.org>
Eric Biggers [Thu, 9 Jul 2026 02:27:47 +0000 (22:27 -0400)]
lib/crypto: docs: Improve introduction sentence
Make it clear that lib/crypto/ is a kernel-internal library. It's easy
for people to come across this page, especially the HTML version online,
without that context.
Eric Biggers [Thu, 9 Jul 2026 02:26:51 +0000 (22:26 -0400)]
lib/crypto: docs: Fix some sentence fragments
Currently, the section about the library API for each algorithm begins
with a noun phrase that was intended to serve as an elaboration on the
title. It's better to use complete sentences.
wifi: ath12k: Flush the posted write after writing to PCIE_SOC_GLOBAL_RESET
ath12k_pci_soc_global_reset() tries to reset the device by writing to the
PCIE_SOC_GLOBAL_RESET register. But it doesn't do a read-back to ensure
that the write gets flushed to the device before the delay.
This may lead to the delay on the host to be insufficient, if the posted
write doesn't reach the device before the delay.
So add a read-back after writing to the PCIE_SOC_GLOBAL_RESET register and
before the delay.
wifi: ath11k: Flush the posted write after writing to PCIE_SOC_GLOBAL_RESET
ath11k_pci_soc_global_reset() tries to reset the device by writing to the
PCIE_SOC_GLOBAL_RESET register. But it doesn't do a read-back to ensure
that the write gets flushed to the device before the delay.
This may lead to the delay on the host to be insufficient, if the posted
write doesn't reach the device before the delay.
So add a read-back after writing to the PCIE_SOC_GLOBAL_RESET register and
before the delay.
Compile tested only.
Fixes: f3c603d412b3 ("ath11k: reset MHI during power down and power up") Reported-by: Alex Williamson <alex@shazbot.org> Closes: https://lore.kernel.org/linux-pci/20260622160822.09350246@shazbot.org Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@oss.qualcomm.com> Reviewed-by: Baochen Qiang <baochen.qiang@oss.qualcomm.com> Reviewed-by: Raj Kumar Bhagat <raj.bhagat@oss.qualcomm.com> Link: https://patch.msgid.link/20260623141649.41087-1-manivannan.sadhasivam@oss.qualcomm.com Signed-off-by: Jeff Johnson <jeff.johnson@oss.qualcomm.com>
Dmitry Morgun [Sat, 30 May 2026 11:42:52 +0000 (11:42 +0000)]
wifi: ath11k: fix potential buffer underflow in ath11k_hal_rx_msdu_list_get()
When the first entry in msdu_details has a zero buffer address,
the code accesses msdu_details[i - 1] with i == 0, causing a
buffer underflow.
Fix similarly to ath12k_wifi7_hal_rx_msdu_list_get() by adding
a separate check for i == 0 before the main condition to prevent
the out-of-bounds access.
Found by Linux Verification Center (linuxtesting.org) with SVACE.
Fixes: d5c65159f289 ("ath11k: driver for Qualcomm IEEE 802.11ax devices") Signed-off-by: Dmitry Morgun <d.morgun@ispras.ru> Reviewed-by: Rameshkumar Sundaram <rameshkumar.sundaram@oss.qualcomm.com> Reviewed-by: Baochen Qiang <baochen.qiang@oss.qualcomm.com> Link: https://patch.msgid.link/20260530114252.42615-1-d.morgun@ispras.ru Signed-off-by: Jeff Johnson <jeff.johnson@oss.qualcomm.com>
wifi: ath10k: fix skb leak on incomplete msdu during rx pop
When ath10k_htt_rx_pop_paddr32_list() or
ath10k_htt_rx_pop_paddr64_list() encounters an incomplete frame
(RX_ATTENTION_FLAGS_MSDU_DONE not set), it returns -EIO without
purging the skb list built up so far, leaking any skbs already
queued in the list.
Other early-exit paths within these same functions already call
__skb_queue_purge() before returning an error. Add it before the
-EIO return as well to be consistent and prevent the leak.
Wentao Liang [Tue, 9 Jun 2026 09:25:28 +0000 (09:25 +0000)]
wifi: ath11k: fix refcount leak in ath11k_ahb_fw_resources_init()
of_get_child_by_name() returns a node pointer with refcount
incremented, but the error path when ath11k_ahb_setup_msa_resources()
fails does not release it. Add the missing of_node_put() to avoid
leaking the reference.
Matthew Rosato [Thu, 9 Jul 2026 13:54:04 +0000 (09:54 -0400)]
KVM: s390: pci: Fix handling of AIF enable without AISB
When a guest seeks to register IRQs without a summary bit specified,
ensure that the associated GAITE then stores 0 for the guest AISB
location instead of virt_to_phys(page_address(NULL)).
Fixes: 3c5a1b6f0a18 ("KVM: s390: pci: provide routines for enabling/disabling interrupt forwarding") Cc: stable@vger.kernel.org Reviewed-by: Farhan Ali <alifm@linux.ibm.com> Signed-off-by: Matthew Rosato <mjrosato@linux.ibm.com> Signed-off-by: Christian Borntraeger <borntraeger@linux.ibm.com>
drm/imagination: fix error checking of pvr_vm_context_lookup()
Since pvr_vm_context_lookup() returns either NULL or a pointer, then stop
using IS_ERR() for checking the return value.
Using IS_ERR() leads to the kernel oops reported below. It can be
reproduced by passing an invalid VM context handle from userspace to the
DRM_IOCTL_PVR_CREATE_CONTEXT ioctl.
firmware: arm_scmi: Rate-limit queue-full warnings in IRQ context
The scmi_notify() function is called from interrupt context to queue
received notification events onto a per-protocol kfifo. When the kfifo
is full, it logs a warning via dev_warn() for every dropped event.
Under conditions where the platform sends a burst of SCMI notifications
faster than the deferred worker can drain the queue, this results in a
flood of dev_warn() calls from IRQ context. Each call acquires the
console lock and may execute blocking console writes, causing the CPU
to be held in interrupt context for an extended period and leading to
observable system stalls.
Fix this by switching to dev_warn_ratelimited() to limit the frequency
of log messages when the notification queue is full. This reduces
console overhead in interrupt context and prevents CPU stalls caused by
excessive logging, while still preserving diagnostic visibility.
Steve Dunnagan [Wed, 1 Jul 2026 19:59:20 +0000 (15:59 -0400)]
firmware: arm_scmi: Use 64-bit division for clock rate rounding
SCMI clock range descriptors report rates as 64-bit values. When handling
a range clock, scmi_clock_determine_rate() rounds the requested rate up to
the next supported step using the SCMI RATE_STEP value.
The current code uses div64_ul() for this calculation. Since div64_ul()
takes an unsigned long divisor, the 64-bit RATE_STEP value can be truncated
on 32-bit builds. In the worst case, a non-zero 64-bit step can be narrowed
to zero before the division.
Store RATE_STEP in a u64, reject a malformed zero step, and use
DIV64_U64_ROUND_UP() so the divisor is handled as a 64-bit value.
This does not change behavior for valid firmware reporting a non-zero step
that fits in unsigned long.
Tested on Xunlong Orange Pi 5 Plus / RK3588 with SCMI over SMC. SCMI
clocks probed successfully before and after the change. SCMI-backed CPU
clocks were exercised through cpufreq-dt by switching each CPU policy
between its lowest and highest available OPP.
gpu/buddy: bail out of try_harder when alignment cannot be honoured
The try_harder contiguous fallback could return a range whose start
offset did not match the caller's min_block_size. When a candidate's
start is misaligned, realign it: free the misaligned run and reallocate
exactly @size at the next lower min_block_size boundary. This keeps the
returned size unchanged with no surplus to trim, and rejects the request
only when no aligned candidate fits.
v2: align misaligned candidates down to min_block_size instead of
bailing out, for both the RHS and LHS paths (Matthew).
Fixes: 0a1844bf0b53 ("drm/buddy: Improve contiguous memory allocation") Suggested-by: Christian König <christian.koenig@amd.com> Cc: Matthew Auld <matthew.auld@intel.com> Cc: Christian König <christian.koenig@amd.com> Cc: Timur Kristóf <timur.kristof@gmail.com> Cc: stable@vger.kernel.org Reviewed-by: Matthew Auld <matthew.auld@intel.com> Tested-by: John Olender <john.olender@gmail.com> Signed-off-by: Arunpravin Paneer Selvam <Arunpravin.PaneerSelvam@amd.com> Link: https://patch.msgid.link/20260709131050.1022759-1-Arunpravin.PaneerSelvam@amd.com
cifs: Show reason why autodisabling serverino support
Extend cifs_autodisable_serverino() function to print also text message why
the function was called.
The text message is printed just once for mount then autodisabling
serverino support. Once the serverino support is disabled for mount it will
not be re-enabled. So those text messages do not cause flooding logs.
This change allows to debug issues why cifs.ko decide to turn off server
inode number support and hence disable support for detection of hardlinks.
Signed-off-by: Pali Rohár <pali@kernel.org> Signed-off-by: Steve French <stfrench@microsoft.com>
smb/client: fix incorrect nlink returned by fstat()
Reproducer:
1. mount -t cifs //${server_ip}/export /mnt
2. touch /mnt/file1; ln /mnt/file1 /mnt/file2; ln /mnt/file1 /mnt/file3
3. C program: int fd = open("/mnt/file1", O_RDONLY);
4. C program: struct stat stbuf; fstat(fd, &stbuf);
stbuf.st_nlink is always 1, should be 3
Setting `unknown_nlink` to true in `SMB2_open()` triggers the
`CIFS_FATTR_UNKNOWN_NLINK` flag in `cifs_open_info_to_fattr()`,
which safely preserves the existing i_nlink in
`cifs_nlink_fattr_to_inode()`.
smb/client: pass cifs_open_info_data to SMB2_open()
Let SMB2_open() fill the smb2_file_all_info embedded in cifs_open_info_data
directly. This removes the temporary smb2_file_all_info copy in
smb2_open_file().
Signed-off-by: ChenXiaoSong <chenxiaosong@kylinos.cn> Signed-off-by: Steve French <stfrench@microsoft.com>
Guangshuo Li [Wed, 8 Jul 2026 12:35:20 +0000 (20:35 +0800)]
smb: client: fix overflow in passthrough ioctl bounds check
smb2_ioctl_query_info() validates the PASSTHRU_FSCTL response payload
before copying it to userspace.
The payload offset and length both come from 32-bit fields. The bounds
check currently adds OutputOffset and qi.input_buffer_length directly, so
the addition can wrap in 32-bit arithmetic before the result is compared
against the response buffer length.
A malicious server can use a large OutputOffset and a small OutputCount
to make the wrapped sum pass the bounds check. The later copy_to_user()
then reads from io_rsp + OutputOffset, outside the response buffer.
Use size_add() for the offset plus length check so overflow is treated as
out of bounds.
Fixes: 2b1116bbe898 ("CIFS: Use common error handling code in smb2_ioctl_query_info()") Signed-off-by: Guangshuo Li <lgs201920130244@gmail.com> Signed-off-by: Steve French <stfrench@microsoft.com>
Daehyeon Ko [Fri, 3 Jul 2026 08:36:33 +0000 (17:36 +0900)]
macsec: don't read an unset MAC header in macsec_encrypt()
macsec_encrypt() reads the Ethernet header via eth_hdr(skb)
(skb->head + skb->mac_header) to memmove() the 12 source/destination MAC
bytes forward and make room for the SecTAG.
On the AF_PACKET SOCK_RAW + PACKET_QDISC_BYPASS transmit path the skb
reaches the macsec ndo_start_xmit() with the MAC header unset, so
eth_hdr(skb) resolves to skb->head + (u16)~0 and the read is out of
bounds: a 12-byte heap over-read that is also emitted on the wire as the
frame's outer source/destination MAC. KASAN reports a slab-out-of-bounds
read in macsec_start_xmit() on 6.0; on current mainline a CONFIG_DEBUG_NET
build flags it as an unset mac header in skb_mac_header().
On the TX path the L2 header is at skb->data, so use skb_eth_hdr(), added
by commit 96cc4b69581d ("macvlan: do not assume mac_header is set in
macvlan_broadcast()") for exactly this purpose.
Dust Li [Tue, 7 Jul 2026 07:43:18 +0000 (15:43 +0800)]
dibs: loopback: validate offset and size in move_data()
The loopback move_data() performs a memcpy into the registered DMB
without checking whether offset + size exceeds the DMB length. Unlike
real ISM hardware, which enforces memory region bounds natively, the
software loopback has no such protection.
A peer-supplied out-of-bounds offset or oversized write would result in
an OOB write past the allocated kernel buffer. Add an explicit bounds
check before the memcpy to reject such requests with -EINVAL.
Fixes: f7a22071dbf3 ("net/smc: implement DMB-related operations of loopback-ism") Cc: stable@vger.kernel.org Reported-by: Federico Kirschbaum <federico.kirschbaum@xbow.com> Signed-off-by: Dust Li <dust.li@linux.alibaba.com> Reported-by: Baul Lee <baul.lee@xbow.com> Link: https://patch.msgid.link/20260707074318.1448662-1-dust.li@linux.alibaba.com Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Shuicheng Lin [Tue, 30 Jun 2026 19:22:21 +0000 (19:22 +0000)]
drm/xe/userptr: Stub notifier_lock helpers when DRM_GPUSVM=n
When CONFIG_DRM_GPUSVM=n (e.g. um-allyesconfig), the only caller of
xe_pt_svm_userptr_notifier_lock() is compiled out, triggering:
drivers/gpu/drm/xe/xe_pt.c:1418:13: warning:
'xe_pt_svm_userptr_notifier_lock' defined but not used
[-Wunused-function]
The helpers cannot simply be removed in this case: the matching
xe_pt_svm_userptr_notifier_unlock() is also referenced from
xe_pt_update_ops_run(), which lives outside any DRM_GPUSVM ifdef and is
gated only at runtime by pt_update_ops->needs_svm_lock. The symbol must
exist in all builds.
Provide empty static inline stubs for !DRM_GPUSVM, matching the pattern
used by xe_svm_notifier_lock()/_unlock() in xe_svm.h.
Fixes: dca6e08c923a ("drm/xe/userptr: Hold notifier_lock for write on inject test path") Reported-by: kernel test robot <lkp@intel.com> Closes: https://lore.kernel.org/oe-kbuild-all/202606302210.QqcLbOEN-lkp@intel.com/ Reviewed-by: Matthew Brost <matthew.brost@intel.com> Link: https://patch.msgid.link/20260630192221.2998168-1-shuicheng.lin@intel.com Signed-off-by: Shuicheng Lin <shuicheng.lin@intel.com>
(cherry picked from commit 3359422bf0a1140e96d783a19a397686e580a3ca) Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Guangshuo Li [Wed, 8 Jul 2026 07:34:22 +0000 (15:34 +0800)]
drm/xe: free madvise VMA array on L2 flush failure
xe_vm_madvise_ioctl() allocates madvise_range.vmas in get_vmas().
After get_vmas() succeeds with at least one VMA, error paths must go
through free_vmas so the array is released before the madvise details are
destroyed.
The L2 flush validation path added for PAT madvise rejects some
SVM/userptr ranges after get_vmas() has succeeded, but jumps directly to
madv_fini. This skips kfree(madvise_range.vmas), leaking the VMA array on
each failed ioctl.
Jump to free_vmas instead, matching the other validation failure paths
after get_vmas() has succeeded.
Fixes: 4f39a194d41e ("drm/xe/xe3p_lpg: Restrict UAPI to enable L2 flush optimization") Signed-off-by: Guangshuo Li <lgs201920130244@gmail.com> Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patch.msgid.link/20260708073422.725186-1-lgs201920130244@gmail.com Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
(cherry picked from commit c3a1c3579b1250060da73507a4acef712974c78a) Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Anas Khan [Thu, 2 Jul 2026 11:28:20 +0000 (16:58 +0530)]
drm/xe: remove duplicate <kunit/test-bug.h> include
xe_pci.c includes <kunit/test-bug.h> twice, separated only by the
<kunit/test.h> include. Drop the redundant second include; this is a
non-functional cleanup flagged by scripts/checkincludes.pl.
Matthew Brost [Thu, 2 Jul 2026 21:58:05 +0000 (14:58 -0700)]
drm/xe: Wait on external BO kernel fences in exec IOCTL
Before arming a user job, xe_exec_ioctl() only added the VM's
dma-resv KERNEL slot as a dependency. That slot covers rebinds and
the kernel operations of the VM's private BOs, but not external BOs
(bo->vm == NULL), which carry their kernel operations (evictions,
moves, ...) in their own dma-resv KERNEL slot.
The DMA_RESV_USAGE_KERNEL slot is the cross-driver contract for
memory management operations that must complete before the BO or its
backing store may be used: any accessor is required to wait on the
KERNEL fences before touching the resv. By skipping the external BOs'
KERNEL slots, the exec path violated that contract and could schedule
a user job while a kernel operation on an external BO mapped by the VM
was still in flight, racing against it and potentially reading or
writing memory that was being moved.
Replace the VM-only dependency with an iteration over every object
locked by the exec, adding each object's KERNEL slot as a job
dependency. This covers the VM resv (rebinds and private BOs) as well
as every external BO, mirroring the drm_gpuvm_resv_add_fence() call
that later publishes the job fence to the same set of objects.
Long-running mode continues to skip this, as before.
Fixes: dd08ebf6c352 ("drm/xe: Introduce a new DRM driver for Intel GPUs") Cc: stable@vger.kernel.org Assisted-by: GitHub_Copilot:claude-opus-4.8 Signed-off-by: Matthew Brost <matthew.brost@intel.com> Reviewed-by: Matthew Auld <matthew.auld@intel.com> Link: https://patch.msgid.link/20260702215805.4011228-1-matthew.brost@intel.com
(cherry picked from commit a6b842acf3ddd1efc53a56de9260cfa718fb35e7) Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Matthew Brost [Thu, 2 Jul 2026 01:24:34 +0000 (18:24 -0700)]
drm/xe: Fix PTE index in xe_vm_populate_pgtable() for chunked binds
xe_vm_populate_pgtable() indexed the source PTE array (update->pt_entries)
by the per-call loop counter, assuming each call starts at the first entry
of the update. That holds for the CPU bind path
(xe_migrate_update_pgtables_cpu), which populates a whole update in a single
call, but not for the GPU bind path: write_pgtable() splits an update into
MAX_PTE_PER_SDI (510) sized MI_STORE_DATA_IMM chunks, invoking the populate
callback once per chunk with an advancing qword_ofs but a fresh command-
buffer destination pointer.
As a result, every chunk after the first re-read pt_entries from index 0
instead of from its true offset, so PTEs beyond the first 510 entries of a
single update were programmed with the wrong physical pages, shifting the
mapping by exactly MAX_PTE_PER_SDI pages.
This stayed latent because a single update only exceeds 510 qwords when a
large (e.g. 2M) region is bound as individual 4K PTEs rather than a single
huge-page entry, which happens when the backing store is sufficiently
fragmented. It was surfaced by the BO defrag path, which deliberately
rebinds such fragmented ranges via the GPU bind path, producing
deterministic data corruption offset by 510 pages.
Index pt_entries by the chunk's absolute offset relative to update->ofs so
both the CPU and GPU paths pick the correct entries.
Fixes: dd08ebf6c352 ("drm/xe: Introduce a new DRM driver for Intel GPUs") Cc: stable@vger.kernel.org Assisted-by: GitHub_Copilot:claude-opus-4.8 Signed-off-by: Matthew Brost <matthew.brost@intel.com> Reviewed-by: Matthew Auld <matthew.auld@intel.com> Link: https://patch.msgid.link/20260702012434.3861171-1-matthew.brost@intel.com
(cherry picked from commit e6f2d0b757c4fb577a513c577140109d1d292a9a) Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
octeontx2-af: fix VF bringup affecting PF promiscuous state
Mbox handling of nix_set_rx_mode for a VF with promiscuous and
all_multi flags set to false causes deletion of the PF's promiscuous
and allmulti MCAM rules. This occurs because the APIs that
enable/disable these rules operate only on the PF, even when the
mbox request is made via a VF interface.
Guard both rvu_npc_enable_allmulti_entry() and
rvu_npc_enable_promisc_entry() disable paths with an is_vf() check so
that a VF bringing up or tearing down its interface cannot inadvertently
clear the PF's MCAM rules.
Fixes: 967db3529eca ("octeontx2-af: add support for multicast/promisc packet replication feature") Signed-off-by: Harman Kalra <hkalra@marvell.com> Signed-off-by: Nitin Shetty J <nshettyj@marvell.com> Link: https://patch.msgid.link/20260702045616.3002773-2-nshettyj@marvell.com Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Paolo Abeni [Thu, 9 Jul 2026 09:42:56 +0000 (11:42 +0200)]
Merge tag 'nf-26-07-08' of https://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf
Florian Westphal says:
====================
netfilter: updates for net
The following patchset contains Netfilter fixes for *net*.
Most of these are LLM fixes for old issues flagged by sashiko/LLMs.
Many of these trigger drive-by-findings in sashiko. In particular:
- many load/store tearing and missing memory barriers, races
etc. in ipset, esp. with GC and resizing.
Keeping the proposed patches spinning for yet-another-iteration
keeps legit fixes back, so I prefer to add these now and follow
up with other reports later.
- flowtable work queue still has possible races with teardown,
but same rationale as with ipset: drive-by findings, not
problems coming with the flowtable IPIP changeset in this PR.
- ever since unreadable frag skb support was added in 6.12, we can no
longer do: BUG_ON(skb_copy_bits( ...): it will fire with such skbs.
Mina Almasry is looking at similar patterns elsewhere in the stack.
1) Guard skb->mac_header adjustment after IPv6 defragmentation in
nf_conntrack_reasm. From Xiang Mei.
2) NUL-terminate ebtables table names before calling find_table_lock() to
prevent stack-out-of-bounds reads. Also from Xiang Mei.
3) Zero the ebtables chainstack array, else error unwind may free bogus
pointer when CPU mask is sparse. All three issues date from 2.6 days.
4) Ensure ebtables module names are c-strings, same bug pattern as 2).
Bug added in 4.6.
5) Fix catchall element handling for inverted lookups in nft_lookup. Fold the
catchall lookup into ext before computing the match status. Was like
this ever since catchall elements got introduced in 5.13.
From Tamaki Yanagawa.
6-9) ipset updates from Jozsef Kadlecsik:
- mark rcu protected areas correctly
- address gc and resize clash in the comment extension
- add/del backlog cleanup in the error path
- allocate right size for the generic hash structure
10-12): IPIP flowtable updates from Pablo Neira Ayuso:
- Use the current direction's route when pushing IPIP headers
Fix incorrect headroom and fragmentation offset calculations.
- Avoid hardware offload for IPIP tunnels due to lack of driver support.
- Support IPIP tunnels with direct xmit in netfilter flowtable.
dst_cache and dst_cookie are moved outside the union to share route
state across flows. This is a followup to work done in 6.19 cycle.
13) Don't BUG() on skb_copy_bits error. Handle unreadable fragments by
either returning an error or restricting the copy operations to linear area,
This became an issue when unreable frag support was merged in 6.12.
14-16): IPVS updates from Yizhou Zhao:
- Pass parsed transport offset to IPVS state handlers.
update callback signatures.
- use correct transport header offset on state lookp in TCP.
As-is it was possible for ipv6 extension header data to be
treated as L4 header.
- same for SCTP. This was also broken since 2.6 days.
17) Ensure inner IP headers in ICMP errors are in the skb headroom after
stripping outer headers. Add more checks for the length of inner headers.
This was broken since 3.7 days.
From Julian Anastasov.
netfilter pull request nf-26-07-08
* tag 'nf-26-07-08' of https://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf:
ipvs: ensure inner headers in ICMP errors are in headroom
ipvs: use parsed transport offset in SCTP state lookup
ipvs: use parsed transport offset in TCP state lookup
ipvs: pass parsed transport offset to state handlers
netfilter: handle unreadable frags
netfilter: flowtable: support IPIP tunnel with direct xmit
netfilter: flowtable: IPIP tunnel hardware offload is not yet support
netfilter: flowtable: use dst in this direction when pushing IPIP header
netfilter: ipset: allocate the proper memory for the generic hash structure
netfilter: ipset: cleanup the add/del backlog when resize failed
netfilter: ipset: exclude gc when resize is in progress
netfilter: ipset: mark the rcu locked areas properly
netfilter: nft_lookup: fix catchall element handling with inverted lookups
netfilter: ebtables: module names must be null-terminated
netfilter: ebtables: zero chainstack array
netfilter: ebtables: terminate table name before find_table_lock()
netfilter: nf_conntrack_reasm: guard mac_header adjustment after IPv6 defrag
====================
Gal Pressman [Mon, 6 Jul 2026 05:50:17 +0000 (08:50 +0300)]
ethtool: rss: Fix hfunc and input_xfrm parsing on big endian
ETHTOOL_A_RSS_HFUNC and ETHTOOL_A_RSS_INPUT_XFRM are NLA_U32 attributes,
but ethnl_rss_set() and ethnl_rss_create_doit() parse them with
ethnl_update_u8(), which reads a single byte.
On little endian this happens to read the least significant byte and
works as long as the value fits in a byte. On big endian it reads the
most significant byte, so the requested value is parsed incorrectly.
The destination fields in struct ethtool_rxfh_param are u8, so the
attribute can't be read directly with ethnl_update_u32().
Cap the hfunc policy at U8_MAX so an out of range value is rejected
instead of being silently truncated into the u8 field, and add
ethnl_update_u8_u32() to read the full u32 and narrow it into the u8
destination.
Fixes: 82ae67cbc423 ("ethtool: rss: support setting hfunc via Netlink") Fixes: d3e2c7bab124 ("ethtool: rss: support setting input-xfrm via Netlink") Fixes: a166ab7816c5 ("ethtool: rss: support creating contexts via Netlink") Reviewed-by: Dragos Tatulea <dtatulea@nvidia.com> Reviewed-by: Nimrod Oren <noren@nvidia.com> Signed-off-by: Gal Pressman <gal@nvidia.com> Link: https://patch.msgid.link/20260706055017.3355806-1-gal@nvidia.com Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Li RongQing [Fri, 3 Jul 2026 14:14:23 +0000 (22:14 +0800)]
net/mlx5: Fix L3 tunnel entropy refcount leak
mlx5_tun_entropy_refcount_inc() counts both VXLAN and L2-to-L3
tunnel reformat entries as entropy-enabling users. The matching
decrement path only handled VXLAN, leaving L2-to-L3 tunnel entries
counted after release.
Handle MLX5_REFORMAT_TYPE_L2_TO_L3_TUNNEL in
mlx5_tun_entropy_refcount_dec() as well so the enabling entry
refcount remains balanced.
Fixes: f828ca6a2fb6 ("net/mlx5e: Add support for hw encapsulation of MPLS over UDP") Signed-off-by: Li RongQing <lirongqing@baidu.com> Reviewed-by: Simon Horman <horms@kernel.org> Reviewed-by: Tariq Toukan <tariqt@nvidia.com> Link: https://patch.msgid.link/20260703141423.1723-1-lirongqing@baidu.com Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Kailang Yang [Thu, 9 Jul 2026 06:57:06 +0000 (14:57 +0800)]
ALSA: hda/realtek - Fixed Headphone noise issue for Dell QCM1255
This platform booted with Ubuntu 24.04 with Pipewire audio server. So,
it has pop noise with headphone. But it's normal with Pulseaudio server.
This patch was the workaround. Connect the headphones to DAC 0x2.
The popping sound will disappear.
The MACB driver has since forever leaked the outgoing SKBs that
have not yet been marked as completed. They live in queue->tx_skb
which gets freed without remorse nor checking.
macb_free_consistent() gets called in a few codepaths, but only close will
trigger the added expressions. In macb_open() and macb_alloc_consistent()
failure cases, queues' tx_skb just got allocated and are empty.
Paolo Abeni [Thu, 9 Jul 2026 08:36:14 +0000 (10:36 +0200)]
Merge branch 'fix-mana-rx-with-bounce-buffering'
Dexuan Cui says:
====================
Fix MANA RX with bounce buffering
With swiotlb=force, the MANA NIC fails to work properly due to commit 730ff06d3f5c ("net: mana: Use page pool fragments for RX buffers instead
of full pages to improve memory efficiency.").
This happens because, with the standard MTU=1500, the aforementioned
commit uses page pool frags with PP_FLAG_DMA_MAP, but fails to call
page_pool_dma_sync_for_cpu() to sync the received packet for CPU acces
before handing the RX buffer to the stack.
Here patch #2 adds the required page_pool_dma_sync_for_cpu().
Patch #1 validates the packet length reported by the NIC. With patch #2,
page_pool_dma_sync_for_cpu() uses the packet length, so we don't want
to blindly trust the packet length, just in case.
There is no change between v2 and v3.
v3 just swaps the order of the 2 patches in v2, as suggested by Simon [3].
Dexuan Cui [Thu, 2 Jul 2026 04:12:37 +0000 (21:12 -0700)]
net: mana: Sync page pool RX frags for CPU
MANA allocates RX buffers from page pool fragments when frag_count is
greater than 1. In that case the buffers remain DMA mapped by page pool
and the RX completion path does not call dma_unmap_single(). As a result,
the implicit sync-for-CPU normally performed by dma_unmap_single() is
missing before the packet data is passed to the networking stack.
This breaks RX on configurations which require explicit DMA syncing, for
example when booted with swiotlb=force.
Fix this by recording the page pool page and DMA sync offset when the RX
buffer is allocated, and syncing the received packet range for CPU access
before handing the RX buffer to the stack.
Fixes: 730ff06d3f5c ("net: mana: Use page pool fragments for RX buffers instead of full pages to improve memory efficiency.") Cc: stable@vger.kernel.org Reviewed-by: Haiyang Zhang <haiyangz@microsoft.com> Signed-off-by: Dexuan Cui <decui@microsoft.com> Link: https://patch.msgid.link/20260702041237.617719-3-decui@microsoft.com Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Dexuan Cui [Thu, 2 Jul 2026 04:12:36 +0000 (21:12 -0700)]
net: mana: Validate the packet length reported by the NIC
Validate the packet length reported in the RX CQE before passing it
to skb processing. The CQE is supplied by the NIC device and should
not be blindly trusted.
Cc: stable@vger.kernel.org Reviewed-by: Haiyang Zhang <haiyangz@microsoft.com> Signed-off-by: Dexuan Cui <decui@microsoft.com> Fixes: ca9c54d2d6a5 ("net: mana: Add a driver for Microsoft Azure Network Adapter (MANA)") Link: https://patch.msgid.link/20260702041237.617719-2-decui@microsoft.com Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Wang Yan [Thu, 2 Jul 2026 02:59:49 +0000 (10:59 +0800)]
selftests/net: fix EVP_MD_CTX leak in tcp_mmap
In tcp_mmap.c, both child_thread() and main() allocate an EVP_MD_CTX
via EVP_MD_CTX_new() when integrity checking is enabled, but neither
function releases the context. child_thread() misses the free in its
common cleanup block, and main() returns without freeing the context.
This results in a SHA256 context leak on every run that uses the
‑i (integrity) option. Add the missing EVP_MD_CTX_free() calls to
the appropriate cleanup paths to fix the leak.
KVM: s390: Fix dat_crste_walk_range() early return
If a walk entry handler for a lower level returns a value,
dat_crste_walk_range() will not return immediately, but instead loop
again and move to the next entry.
This means that some entries are potentially skipped, and early return
is ignored. Skipped entries might lead to all kinds of issues, given
that the caller expects them to not be skipped. Early return is often
used to interrupt a walk when a rescheduling is needed; if it is
ignored it can lead to stalls.
Fix by breaking from the loop immediately if the walk to a lower level
returned non-zero.
KVM: s390: vsie: Avoid potential deadlock with real spaces
The natural lock ordering is mmu_lock -> children_lock, but in
gmap_create_shadow() the reverse order is used when handling shadowing
of real address spaces.
Convert the inner locking of kvm->mmu_lock to a trylock; return -EAGAIN
if the lock is busy, and let the caller try again.
This path is not expected to happen in real-life scenarios, so its
performance is not important.
Fixes: a2c17f9270cc ("KVM: s390: New gmap code") Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com> Reviewed-by: Christian Borntraeger <borntraeger@linux.ibm.com> Signed-off-by: Christian Borntraeger <borntraeger@linux.ibm.com>
kvm_s390_gisc_register() registers the guest ISC before pinning
the guest interrupt forwarding pages and allocating the AISB bit.
If any of the later setup steps fails, the function unwinds the
pinned pages and other local state, but does not unregister the
GISC reference. Add the missing kvm_s390_gisc_unregister() to the
error unwind path.
Fixes: 3c5a1b6f0a18 ("KVM: s390: pci: provide routines for enabling/disabling interrupt forwarding") Cc: stable@vger.kernel.org Signed-off-by: Haoxiang Li <haoxiang_li2024@163.com> Reviewed-by: Matthew Rosato <mjrosato@linux.ibm.com> Tested-by: Matthew Rosato <mjrosato@linux.ibm.com> Acked-by: Claudio Imbrenda <imbrenda@linux.ibm.com> Reviewed-by: Christian Borntraeger <borntraeger@linux.ibm.com> Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Message-ID: <20260624061910.2794734-1-haoxiang_li2024@163.com> Signed-off-by: Christian Borntraeger <borntraeger@linux.ibm.com>
drm/fb-helper: Only consider active CRTCs for vblank sync
Only synchronize fbdev output to the vblank of an active CRTC. Go over
the list of CRTCs and pick the first that matches. Fixes warnings as
the one shown below
[ 77.201354] WARNING: drivers/gpu/drm/drm_vblank.c:1320 at drm_crtc_wait_one_vblank+0x194/0x1cc [drm], CPU#1: kworker/1:7/1867
[ 77.201354] omapdrm omapdrm.0: [drm] vblank wait timed out on crtc 0
This currently happens if the fbdev output is not on CRTC 0.
Atomic and non-atomic drivers require distinct code paths. As for other
fbdev operations, implement both and select the correct one at runtime.
Not finding an active CRTC is not a bug. Do not wait in this case, but
flush the display update as before.
v4:
- avoid possible deadlocks with locking context (Sashiko)
v3:
- drop excessive state validation (Jani)
- acquire plane and CRTC mutices (Sashiko)
v2:
- move look-up code into separate helper
- support drivers with legacy modesetting
v1:
- see https://lore.kernel.org/dri-devel/1c9e0e24-9c4a-4259-8700-cf9e5fd60ca3@suse.de/
Co-authored-by: H. Nikolaus Schaller <hns@goldelico.com> Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Fixes: d8c4bddcd8bcb ("drm/fb-helper: Synchronize dirty worker with vblank") Tested-by: Icenowy Zheng <zhengxingda@iscas.ac.cn> Tested-by: H. Nikolaus Schaller <hns@goldelico.com> Closes: https://bugs.debian.org/1138033 Acked-by: Maxime Ripard <mripard@kernel.org> Link: https://patch.msgid.link/20260702145021.226932-1-tzimmermann@suse.de
Merge tag 'iio-fixes-for-7.2a' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/jic23/iio into char-misc-linus
Jonathan writes:
IIO: 1st set of fixes for the 7.2 cycle
Usual mixed bag of recently introduced issues and much older ones.
core
- Ensure kfifo is reset before fd is allocated avoiding concurrent use of
fifo with reset.
multiple drivers
- Fix up missing Kconfig dependencies.
hid-sensors
- Add support for multibyte read as necessary precursor to...
- Fix stale or zero output when reading raw values for quaternions.
adi,adis
- Add IRQF_NO_THREAD to ensure interrupt is not pushed to the software
interrupt chip used for trigger demux in the IIO core from a thread.
bosch,bmc150
- Hardening against device returning a reserved out of range value for
how many entries are in the FIFO.
bosch,bmi160
- Add IRQF_NO_THREAD to ensure interrupt is not pushed to the software
interrupt chip used for trigger demux in the IIO core from a thread.
dynaimage,al3010
- Fix wrong scale for highest gain_range due to too many digits in the
micro part (val2).
freescale,mpl3115
- Fix unbalanced runtime pm on error in read_raw().
invensens,icm42600
- Avoid wrong divisor for fifo timestamps when using the watermark
interrupt.
- Fix timestamp accuracy loss due to excessive divisor for calculations.
kionix,kxsd9
- Fix unbalanced runtime pm on an error in write_raw().
microchip,mcp37feb02
- Fix an uninitialized reference voltage value for particular DT config.
melix,mlx90635
- Build on basis of right Kconfig symbol.
nxp,lpc32xx
- Ensure completion initialized before requesting irq. Hardening against
spurious IRQ.
nxp,saradc
- Fix a delay calculation.
sharp,gp2ap0002
- Fix unbalanced runtime pm on error in read_raw().
st,lsm6dsx
- Fix an issue seen in wild where an unplanned CPU reset can leave the
device on the wrong register page, thus leaving the driver wedged.
st,st_sensors library
- Make sure to handle a device that provides data as big endian correctly.
st,spear
- Ensure completion initialized before requesting irq. Hardening against
spurious IRQ.
taos,tsl2591
- Don't eat return from devm_request_threaded_irq() as that breaks
deferred probing.
ti,ads1119
- Fix a pm reference count leak in an error path.
ti,ads124s08
- Handle gpio look up errors correctly.
* tag 'iio-fixes-for-7.2a' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/jic23/iio: (28 commits)
iio: event: Fix event FIFO reset race
iio: imu: inv_icm42600: fix timestamp clock period by using lower value
iio: light: al3010: fix incorrect scale for the highest gain range
iio: adc: nxp-sar-adc: Fix the delay calculation in nxp_sar_adc_wait_for()
iio: light: tsl2591: return actual error from probe IRQ failure
iio: imu: inv_icm42600: fix timestamping by limiting FIFO reading
iio: imu: st_lsm6dsx: deselect shub page before reading whoami
iio: adc: ad7779: add missing 'select IIO_TRIGGERED_BUFFER' to Kconfig
iio: adc: ad4130: add missing `select IIO_TRIGGERED_BUFFER` to Kconfig
iio: adc: ti-ads124s08: Return reset GPIO lookup errors
iio: temperature: Build mlx90635 with CONFIG_MLX90635
iio: light: al3320a: add missing REGMAP_I2C to Kconfig
iio: light: al3010: add missing REGMAP_I2C to Kconfig
iio: light: al3000a: add missing REGMAP_I2C to Kconfig
iio: common: st_sensors: honour channel endianness in read_axis_data
iio: imu: bmi160: add IRQF_NO_THREAD to data-ready trigger IRQ
iio: imu: adis: add IRQF_NO_THREAD to non-FIFO trigger IRQ
iio: hid-sensor-rotation: Fix stale or zero output when reading raw values
HID: sensor-hub: Add sensor_hub_input_attr_read_values() for multi-byte reads
iio: adc: spear: Initialize completion before requesting IRQ
...
scsi: lpfc: Fix memory leak in lpfc_sli4_driver_resource_setup()
The memory allocated for mboxq using mempool_alloc() is not freed in
some of the early exit error paths. Fix that by moving the
mempool_free() call to an earlier point after last use.
Fixes: d79c9e9d4b3d ("scsi: lpfc: Support dynamic unbounded SGL lists on G7 hardware.") Cc: stable@vger.kernel.org Signed-off-by: Abdun Nihaal <nihaal@cse.iitm.ac.in> Reviewed-by: Justin Tee <justin.tee@broadcom.com> Link: https://patch.msgid.link/20260707065304.949135-1-nihaal@cse.iitm.ac.in Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
scsi: sg: Report request-table problems when any status is set
SG_GET_REQUEST_TABLE reports per-request diagnostic state through
sg_req_info::problem. The field is meant to indicate whether there is an
error to report for a completed request.
sg_fill_request_table() currently combines masked_status, host_status
and driver_status with bitwise AND. This only reports a problem when all
three status fields are non-zero at the same time. A normal target check
condition, for example, has masked_status set while host_status and
driver_status may both be zero, so the request is incorrectly reported
as clean.
Use the same condition as sg_new_read(), which sets SG_INFO_CHECK when
any of the three status fields is non-zero.
Steven Rostedt [Tue, 30 Jun 2026 22:54:12 +0000 (18:54 -0400)]
scsi: ufs: core: tracing: Do not dereference pointers in TP_printk()
The trace events in drivers/ufs/core/ufs_trace.h were converted to take
a pointer to the hba structure as an argument for the tracepoint and
then in TP_printk() the printing of the dev_name from the ring buffer
was converted to using the dev dereferenced pointer from the hba saved
pointer.
This is not allowed as the TP_printk() is executed at the time the trace
event is read from /sys/kernel/tracing/trace file. That can happen
literally, seconds, minutes, hours, weeks, days, or even months later!
There is no guarantee that the hba pointer will still exist by the time
it is dereferenced when the "trace" file is read.
Instead, save the device name from the hba pointer at the time the
tracepoint is called and place it into the ring buffer event. Then the
TP_printk() can read the name directly from the ring buffer and remove
the possibility that it will read a freed pointer and crash the kernel.
This was detected when testing the trace event code that looks for
TP_printk() parameters doing illegal derferences[1]
Cc: stable@vger.kernel.org Fixes: 583e518e7100 ("scsi: ufs: core: Add hba parameter to trace events") Signed-off-by: Steven Rostedt <rostedt@goodmis.org> Reviewed-by: Peter Wang <peter.wang@mediatek.com> Reviewed-by: Bart Van Assche <bvanassche@acm.org> Link: https://patch.msgid.link/20260630185412.283c26c5@gandalf.local.home Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Timur Tabi [Wed, 8 Jul 2026 23:57:22 +0000 (18:57 -0500)]
regulator: core: regulator_lock_two() should test for EDEADLK not EDEADLOCK
Compare against -EDEADLK, which is what ww_mutex_lock() actually
returns and what every other deadlock check in this file already uses.
Function regulator_lock_two() acquires two regulators via
regulator_lock_nested() -> ww_mutex_lock(). On contention,
ww_mutex_lock() returns -EDEADLK, which is the caller's signal to drop
the lock it holds and retry the acquisition in the canonical order.
However, regulator_lock_two() tests the return value against -EDEADLOCK
rather than -EDEADLK. On most architectures, EDEADLK and EDEADLOCK are
the same value, so the comparison happens to be correct and the bug is
invisible. But on MIPS, SPARC, and PowerPC, those two errors have
different values. The test is wrong: a genuine -EDEADLK backoff no
longer matches -EDEADLOCK, so instead of unlocking and retrying, the
code falls into WARN_ON(ret) and returns with only one of the two
regulators locked.
In practice, this is a bug only on MIPS, because the regulator core is
not built or used on the other two platforms.
In general, EDEADLK is preferred over EDEADLOCK for new code.
Zizhi Wo [Sat, 4 Jul 2026 01:53:20 +0000 (09:53 +0800)]
smb: client: fix busy dentry warning on unmount after DIO
Commit c68337442f03 ("cifs: Fix busy dentry used after unmounting") fixed
the issue in cifs where deferred close of a file led to a dentry reference
count not being released in umount, by flushing deferredclose_wq in
cifs_kill_sb() to solve it.
However, the cifs DIO path suffers from the same busy-dentry problem caused
by a delayed dentry reference-count release:
If the umount path is triggered before [5], it results warning:
BUG: Dentry 00000000eab1f070{i=9a917b66ae404fec,n=test} still in use (1)
[unmount of cifs cifs]
The existing per-inode ictx->io_count wait in cifs_evict_inode() does not
help: it lives in the inode eviction path, which runs after
shrink_dcache_for_umount() has already warned about the busy dentries.
Fix it by adding a per-superblock outstanding-rreq counter that is
incremented in cifs_init_request() and decremented in cifs_free_request().
In cifs_kill_sb(), before kill_anon_super(), wait for this counter to reach
0 - which guarantees that all cleanup_work for this sb have run and thus
all relevant cfile puts are queued on fileinfo_put_wq or serverclose_wq.
Then drain the workqueue so the dentry refs are dropped.
This is a targeted wait, not a flush of the system-wide system_dfl_wq.
Fixes: 340cea84f691c ("cifs: open files should not hold ref on superblock") Signed-off-by: Zizhi Wo <wozizhi@huawei.com> Signed-off-by: Steve French <stfrench@microsoft.com>
SFU fifos are natively supported (created and recognized) at least by:
- Microsoft POSIX subsystem
- OpenNT/Interix subsystem
- Microsoft SFU (Windows Services for UNIX)
- Microsoft SUA (Subsystem for UNIX-based Applications)
- Windows NFS server (up to the Windows Server 2008 R2)
Windows NFS server since Windows Server 2012 uses new reparse point format
for storing new fifos, but still can recognize this old format (also in the
latest Windows Server 2022 version).
SFU-style fifo is empty regular file which has system attribute set.
These SFU-style fifos are already recognized by Linux SMB client.
But Linux SMB client is currently creating new SFU fifos in different
format which is not compatible with all those SFU-style consumers. Fix this
by creating new fifos in correct SFU format which would be recognized by
all those applications and also by existing Linux SMB clients.
This change affects only creating new fifos when mount option -o sfu is used.
Signed-off-by: Pali Rohár <pali@kernel.org> Acked-by: Paulo Alcantara (Red Hat) <pc@manguebit.org> Signed-off-by: Steve French <stfrench@microsoft.com>
SFU sockets are natively supported by Interix 3.0 subsystem and also by
later versions. It is part of Microsoft SFU (Windows Services for UNIX) and
Microsoft SUA (Subsystem for UNIX-based Applications). They can be created
and existing (stored on local disk or remote SMB share) can be recognized.
SFU sockets are recognized also by NFS server included in Windows Server.
Windows NFS server versions since Windows Server 2012 uses new reparse
point format for storing new sockets, but still can recognize this old
format (also in the latest Windows Server 2022 version).
SFU-style socket is a regular file which has system attribute set and
content of the file is one zero byte.
These SFU-style sockets are already recognized by Linux SMB client.
But Linux SMB client is currently creating new SFU socket in different
format which is not compatible with all those SFU applications. Fix this by
creating new sockets in correct SFU format which would be recognized by all
SFU, SUA, NFS and existing Linux SMB clients.
This change affects only creating new sockets when mount option -o sfu is used.
Signed-off-by: Pali Rohár <pali@kernel.org> Acked-by: Paulo Alcantara (Red Hat) <pc@manguebit.org> Signed-off-by: Steve French <stfrench@microsoft.com>
smb: client: fix atime clamp check in read completion
cifs_rreq_done() updates the inode atime to current_time(inode) after a
netfs read. It then preserves the CIFS rule that atime should not be
older than mtime, because some applications break if atime is less than
mtime. That rule only requires clamping when atime < mtime.
The current check uses the raw non-zero result of timespec64_compare().
It therefore takes the clamp path for both atime < mtime and
atime > mtime. The latter is the normal case when reading an older file:
the newly recorded atime is newer than the file mtime. The completion
handler then immediately moves atime back to mtime, losing the access
time that was just recorded. Userspace tools that rely on atime, such as
stat, find -atime, backup tools or cold-data classifiers, can therefore
see a recently read CIFS file as not recently accessed.
This is easy to miss because the bug is silent: read I/O still succeeds,
no error is reported, and many systems either do not check atime after
reads or mount with policies such as relatime/noatime. It becomes
visible when a CIFS file has an mtime older than the current time, the
file is read, and the local inode atime is inspected before a later
revalidation replaces the cached timestamps.
Clamp only when atime is actually older than mtime. This matches the
same atime/mtime rule used when applying CIFS inode attributes.
Fixes: 69c3c023af25 ("cifs: Implement netfslib hooks") Cc: stable@vger.kernel.org Signed-off-by: Xu Rao <raoxu@uniontech.com> Signed-off-by: Steve French <stfrench@microsoft.com>
ASoC: tas2562: fix deprecated 'shut-down' GPIO always cleared after lookup
In tas2562_parse_dt(), the fallback lookup for the deprecated
"shut-down" GPIO property is broken due to a missing pair of braces.
The code intends to reset sdz_gpio to NULL only when the lookup
returns an error that is not -EPROBE_DEFER (so the driver gracefully
continues without a GPIO). However, without braces the statement:
tas2562->sdz_gpio = NULL;
falls outside the IS_ERR() check and is executed unconditionally
for every path through the if block, including a successful GPIO
lookup.
This means any device using the deprecated 'shut-down' DT property
will always have sdz_gpio == NULL after probe, making the GPIO
completely non-functional.
Fix this by adding the missing braces to scope the NULL assignment
inside the IS_ERR() branch, matching the pattern already used for
the primary 'shutdown' GPIO lookup above.
David Francis [Mon, 6 Jul 2026 14:19:04 +0000 (10:19 -0400)]
drm/amdkfd: Check bounds on CRIU restore queue type and mqd size
We weren't checking whether the values provided in the private
data in kfd CRIU restore were within bounds.
For queue type, add a KFD_QUEUE_TYPE_MAX and ensure the provided
type is less than it.
For mqd_size, add new function mqd_size_from_queue_type and confirm
that the provided mqd_size matches expectations.
Reviewed-by: David Yat Sin <david.yatsin@amd.com> Signed-off-by: David Francis <David.Francis@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit f19d8086f6644083c913d70bfdeee20e1b6f46a5) Cc: stable@vger.kernel.org
Yang Wang [Mon, 6 Jul 2026 01:29:24 +0000 (09:29 +0800)]
drm/amd/pm: fix smu14 power limit range calculation
SMU14 derives the default PPT limit from SocketPowerLimitAc/Dc, but
MsgLimits.Power may expose a different firmware limit for the same PPT0
throttler. Using those values independently as fixed min/max bases can
report an incorrect configurable power range.
Keep the socket power limit as the default value and as the fallback for
current-limit queries. Calculate the reported range from both firmware
values instead, using the lower value as the minimum base and the higher
value as the maximum base before applying OD percentages.
Signed-off-by: Yang Wang <kevinyang.wang@amd.com> Reviewed-by: Kenneth Feng <kenneth.feng@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit c936b8126b444401318fcbeb1828488cc5312dee) Cc: stable@vger.kernel.org
David Francis [Thu, 21 May 2026 13:18:59 +0000 (09:18 -0400)]
drm/amdkfd: Check bounds in allocate_event_notification_slot
The valid event ids go from 0 to KFD_SIGNAL_EVENT_LIMIT
allocate_event_notification_slot has an option to specify
an event id to allocate at, used by CRIU. We weren't checking
the bounds on that value.
Check them.
v2: Lower bounds check is unecessary because of idr_alloc
already rejecting negative numbers. Upper bounds check should
be KFD_SIGNAL_EVENT_LIMIT since the signal mode mappings might
not yet exist
Signed-off-by: David Francis <David.Francis@amd.com> Reviewed-by: David Yat Sin <david.yatsin@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit 6853f1f6cbbeb3f53ebbbd7286536aeb2c5d5f50) Cc: stable@vger.kernel.org
Function kfd_process_free_id() should skip over
the primary kfd process because its context id
is fixed assigned, not allocated through the ida table.
This function should only work on secondary contexts.
Fixes: fac682a1d1af ("amdkfd: identify a secondary kfd process by its id") Signed-off-by: Zhu Lingshan <lingshan.zhu@amd.com> Reviewed-by: Felix Kuehling <felix.kuehling@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit 8799ba6fb6a48438aea20c82e74c2f2a3d2b2e7a) Cc: stable@vger.kernel.org
David Francis [Tue, 30 Jun 2026 13:58:18 +0000 (09:58 -0400)]
drm/amdkfd: Don't acquire buffers during CRIU queue restore.
kfd_criu_restore_queue's call of kfd_queue_acquire_buffers was
failing for multiple reasons
- The ctl_stack_size set by the CRIU plugin doesn't match
what is expected by acquire_buffers
- The svm buffer cannot be acquired at this point because
CRIU may not have restored it, or may have restored it
to a different address.
The only reason acquire_buffers was necessary here was to
avoid a null ptr dereference in init_user_queue.
Just put in a check for that dereference; it doesn't appear to
come up in real use cases right now. That is, there is no
usage of CRIU with shared MES.
This is a partial revert of
commit 20a5e7ffdfec ("drm/amdkfd: Properly acquire queue buffers in CRIU restore")
Fixes: 20a5e7ffdfec ("drm/amdkfd: Properly acquire queue buffers in CRIU restore") Reviewed-by: David Yat Sin <david.yatsin@amd.com> Signed-off-by: David Francis <David.Francis@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit 1cafa8b29e029eac3ddf64604f891b35dbf6262b) Cc: stable@vger.kernel.org
David Francis [Tue, 12 May 2026 18:49:41 +0000 (14:49 -0400)]
drm/amdkfd: Check bounds on CRIU restore event id
The valid amdkfd event ids go from 0 to KFD_SIGNAL_EVENT_LIMIT - 1.
During CRIU restore, ensure that the provided event ids are
in that range.
v2: No need for lower bound check since idr_alloc rejects negative
inputs
v3: Also change error message to reflect new error condition
Reviewed-by: David Yat Sin <david.yatsin@amd.com> Signed-off-by: David Francis <David.Francis@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit 5c6c247992d4d9200e073b83f4ec6c703c096845)
Cao Guanghui [Wed, 17 Jun 2026 06:00:53 +0000 (14:00 +0800)]
dm era: fix error code propagation in era_ctr()
era_ctr() replaces the actual error codes returned by dm_get_device()
and dm_set_target_max_io_len() with hardcoded -EINVAL, discarding
the real reason for the failure (e.g. -ENODEV, -ENOMEM). This makes
it harder for users to diagnose problems and is inconsistent with
other dm targets (dm-thin, dm-verity, dm-flakey, dm-ebs) which
propagate the original error.
Fix all three sites to return 'r' instead of -EINVAL.
Signed-off-by: Cao Guanghui <caoguanghui@kylinos.cn> Reviewed-by: Su Yue <glass.su@suse.com> Reviewed-by: Ming-Hung Tsai <mtsai@redhat.com> Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Cao Guanghui [Wed, 17 Jun 2026 06:00:52 +0000 (14:00 +0800)]
dm era: fix NULL pointer dereference in metadata_open()
metadata_open() returns NULL when kzalloc_obj() fails, but the
caller era_ctr() only checks IS_ERR(md). Since IS_ERR(NULL)
returns false, the NULL pointer is treated as a valid result
and later assigned to era->md, leading to a NULL pointer
dereference when the metadata is accessed.
Fix this by returning ERR_PTR(-ENOMEM) on allocation failure,
consistent with dm-cache-metadata.c, dm-thin-metadata.c, and
dm-clone-metadata.c which all use ERR_PTR(-ENOMEM) for the
same pattern.
Fixes: eec40579d848 ("dm: add era target") Signed-off-by: Cao Guanghui <caoguanghui@kylinos.cn> Reviewed-by: Su Yue <glass.su@suse.com> Reviewed-by: Ming-Hung Tsai <mtsai@redhat.com> Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
dm: avoid leaking the caller's thread keyring via the table device file
The refactoring in commit a28d893eb327 ("md: port block device access to file")
accidentally causes the caller's thread keyring to be kept alive long
beyond the caller's lifetime.
As a result, "cryptsetup luksSuspend" silently fails to wipe the
LUKS volume key from memory.
In detail: "cryptsetup luksOpen" uses its supposedly ephemeral thread
keyring to pass the volume key to the kernel. dm-crypt's
crypt_set_keyring_key() copies the key material into its own
crypt_config structure and then drops its own reference to the key in
the keyring with key_put().
With this fix, restoring pre-v6.9 behavior, the copy in the thread
keyring is then promptly garbage collected, such that exactly one copy
of the volume key remains. This single copy is correctly wiped from
memory on "cryptsetup luksSuspend".
Without this fix, the thread keyring and the volume key in it remains.
This second copy is only freed on "luksClose". "luksSuspend" neither
knows about this copy nor has any way to remove it, so the key remains
recoverable from RAM after a suspend that is documented to have wiped it.
This fix should not introduce new security problems, as the code is
anyway gated by CAP_SYS_ADMIN. The device-mapper core, not the calling
task, is the legitimate owner of this long-lived file.
Fixes: a28d893eb327 ("md: port block device access to file") Closes: https://gitlab.com/cryptsetup/cryptsetup/-/work_items/993 Link: https://www.speicherleck.de/iblech/cryptsetup-luksSuspend-issue-reproduction/ Signed-off-by: Ingo Blechschmidt <iblech@speicherleck.de> Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Cc: stable@vger.kernel.org Tested-by: Ondrej Kozina <okozina@redhat.com>
Samuel Moelius [Mon, 29 Jun 2026 15:47:40 +0000 (15:47 +0000)]
dm-pcache: reject option groups without values
The pcache target parses optional arguments as name/value pairs. A
table that advertises one optional argument and supplies only a
recognized option name, for example "cache_mode", reaches
parse_cache_opts() with argc == 1. The parser consumes the name,
decrements argc to zero, then calls dm_shift_arg() again for the value.
dm_shift_arg() returns NULL when no arguments remain, and the following
strcmp() dereferences that NULL pointer.
Check that each recognized option has a value before consuming it. This
keeps valid "cache_mode writeback" and "data_crc true/false" tables
unchanged while making malformed tables fail during target construction
with a precise missing-value error.
Ming-Hung Tsai [Tue, 30 Jun 2026 12:17:44 +0000 (20:17 +0800)]
dm thin metadata: fix metadata snapshot consistency on commit failure
__reserve_metadata_snap() and __release_metadata_snap() modify the
superblock's held_root directly in the block_manager's buffer. If the
subsequent metadata commit fails, the held_root gets flushed to disk
through the abort_transaction path, resulting in inconsistent metadata.
Reproducer 1: __reserve_metadata_snap()
1. Create a 2 MiB metadata device and make the region after the 14th
block inaccessible, to trigger metadata commit failure in the
subsequent reserve_metadata_snap operation. The 14th block will be
the shadow destination for the index block.
3. Take a metadata snapshot to trigger metadata commit failure and
transaction abort. However, the held_root is written to disk,
breaking metadata consistency.
dmsetup message tpool 0 "reserve_metadata_snap"
thin_check v1.2.2 result:
Bad reference count for metadata block 6. Expected 2, but space map contains 1.
Bad reference count for metadata block 7. Expected 2, but space map contains 1.
Bad reference count for metadata block 13. Expected 1, but space map contains 0.
Reproducer 2: __release_metadata_snap()
1. Create a 2 MiB metadata device and make the region after the 16th
block inaccessible, to trigger metadata commit failure in the
subsequent release_metadata_snap operation. The 16th block will be
the shadow destination for the index block.
3. Reserve then release the metadata snapshot, to trigger metadata
commit failure and transaction abort. The held_root gets removed
from the on-disk superblock, causing inconsistent metadata.
Bad reference count for metadata block 6. Expected 1, but space map contains 2.
Bad reference count for metadata block 7. Expected 1, but space map contains 2.
1 metadata blocks have leaked.
Fix by deferring the held_root update to commit time.
Additionally, move the existing-snapshot check in __reserve_metadata_snap
before the shadow operation to avoid unnecessary work. In
__release_metadata_snap, clear pmd->held_root before btree deletion so
partial failure leaks blocks rather than leaving a stale reference, and
unlock the snapshot block before decrementing its refcount.
if (neras && *neras <= v->fec->roots)
fio->erasures[(*neras)++] = i;
This allows *neras to reach roots + 1 (the post-increment pushes it past
roots). This value is then passed as no_eras to decode_rs8(). Inside the
RS decoder (lib/reed_solomon/decode_rs.c:113-121), the erasure locator
polynomial loop writes lambda[j] where j can reach nroots + 1 — one
element past the end of lambda[] (which is sized nroots + 1, valid
indices 0..nroots). The out-of-bounds write lands on syn[0], corrupting
the syndrome buffer.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Assisted-by: Claude:claude-opus-4-6 Cc: stable@vger.kernel.org Fixes: a739ff3f543a ("dm verity: add support for forward error correction") Reviewed-by: Sami Tolvanen <samitolvanen@google.com> Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Samuel Moelius [Thu, 2 Jul 2026 00:27:35 +0000 (00:27 +0000)]
dm era: fix out-of-bounds memory access for non-zero start sector
dm-era tracks writes in target-relative blocks, but era_map() calculates
the writeset block before applying the target offset. Tables with a
non-zero start sector can therefore pass an absolute mapped-device block
to metadata_current_marked().
If the absolute block is beyond the current writeset size,
writeset_marked() tests past the end of the in-core bitset. KASAN reports
this as a vmalloc-out-of-bounds access.
Apply the target offset before calculating the era block so writeset
lookups use the target-relative block number.
dm-log: fix a bitset_size overflow on 32bit machines
Commit c20e36b7631d ("dm log: fix out-of-bounds write due to
region_count overflow") made sure that region_count could fit in an
unsigned int. But the bitmap memory isn't allocated based on
region_count. It uses bitset_size (a size_t variable). The first step of
calculating bitset_size is to set it to region_count, rounded up to a
multiple of BITS_PER_LONG. If region_size is less than BITS_PER_LONG
smaller than UINT_MAX, it will get rounded up to 2^32. On a 32bit
architecture, this will make bitset_size wrap around to 0 and fail,
despite region_count being valid.
Since bitset_size gets divided by 8, it can hold any valid region_count.
It just needs a special case to handle the rollover. If it is 0, the
value rolled over, and bitset size should be set to the number of bytes
needed to hold 2^32 bits.
Signed-off-by: Benjamin Marzinski <bmarzins@redhat.com> Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Fixes: c20e36b7631d ("dm log: fix out-of-bounds write due to region_count overflow") Cc: stable@vger.kernel.org
drm/amdgpu: fix lifetime issue of amdgpu_vm_get_task_info_pasid()
The vm pointer returned from amdgpu_vm_get_vm_from_pasid() is only
valid while the lock is still being held. Once xa_unlock_irqrestore is
called and returned, the pointer is no longer under lock and is subject
to modification. Since, the caller still dereferences vm->task_info in
amdgpu_vm_get_task_info_vm() after the lock is removed, this causes a
use after unlock problem.
Remove the lifetime issue present in amdgpu_vm_get_task_info_pasid()
through removing the amdgpu_vm_get_vm_from_pasid() function from
amdgpu_vm.c and making the relevant code inline to hold the lock while
it is still in use.
Signed-off-by: Shahyan Soltani <shahyan.soltani@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit 9d01579f3f868b333acc901815972685989092c7) Cc: stable@vger.kernel.org
drm/amdgpu: trigger GPU recovery when userq destroy fails to unmap a hung queue
Destroying a hung user queue issues a MES REMOVE_QUEUE that times out,
The destroy path only logged the error and freed the queue, so the
next userq submission failed and forced a GPU reset attributed to an innocent workload.
Kick the userq reset work when unmap fails so the GPU is recovered at
destroy time.
Acked-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Jesse Zhang <Jesse.Zhang@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit 8396b9de4198a54ec4760a94a179347540a9764d) Cc: stable@vger.kernel.org
NFS: Charge unstable writes by request size, not folio size
nfs_folio_mark_unstable() and nfs_folio_clear_commit() charge and
uncharge NR_WRITEBACK/WB_WRITEBACK by folio_nr_pages(folio) once per
*request* added to or removed from a commit list. This is correct only
when a folio has a single associated request. When pg_test splits a
folio into N sub-folio requests (e.g. pNFS flexfiles striping with a
stripe unit smaller than the folio size, or plain wsize-limited
splitting), each of the N requests independently charges the whole
folio's page count, inflating the accounting by a factor of N per
folio. With large folios and small stripe units this reaches multiple
orders of magnitude: a 2 MiB folio split into 512 4 KiB requests can
charge up to 512x its real size, pushing global dirty+writeback
accounting past the system's dirty threshold and forcing every
buffered writer on the host into the hard-throttle path, including
unrelated in-kernel NFS server threads sharing the box.
Charge each request only for the pages it actually covers.
Fixes: 0c493b5cf16e ("NFS: Convert buffered writes to use folios") Cc: stable@vger.kernel.org Signed-off-by: Benjamin Coddington <bcodding@hammerspace.com> Assisted-By: Claude Sonnet 5 <noreply@anthropic.com> Signed-off-by: Anna Schumaker <anna.schumaker@hammerspace.com>
NFSv4: include MAY_WRITE in open permission mask for O_TRUNC
POSIX requires write permission to truncate a file, so an open() that
specifies O_TRUNC must be authorized for write access regardless of the
O_ACCMODE access mode.
nfs_open_permission_mask() builds the access mask passed to
nfs_may_open(), which is the local authorization gate for OPENs the
client serves itself from a cached write delegation via the
can_open_delegated() path in nfs4_try_open_cached(). The mask is
derived from O_ACCMODE alone, so an open(O_RDONLY | O_TRUNC) against a
file the caller cannot write requests only MAY_READ and passes the
local check. The OPEN is then satisfied locally and the truncation is
issued to the server as a SETATTR(size=0) over the delegation stateid,
which the server accepts under standard write-delegation semantics.
POSIX requires that this open fail with EACCES.
Include MAY_WRITE in the mask whenever O_TRUNC is set so the local
check matches the access the server would have enforced.
Suggested-by: Trond Myklebust <trondmy@kernel.org> Fixes: af22f94ae02a ("NFSv4: Simplify _nfs4_do_access()") Cc: stable@vger.kernel.org Signed-off-by: Benjamin Coddington <bcodding@hammerspace.com> Signed-off-by: Anna Schumaker <anna.schumaker@hammerspace.com>
The driver has an OF match table wired to .of_match_table, but does
not export the table with MODULE_DEVICE_TABLE().
Add the missing MODULE_DEVICE_TABLE(of, ...) entry so module alias
information is generated for OF based module autoloading.
This is a source-level fix. It does not claim dynamic hardware
reproduction; the evidence is the driver-owned match table, its use by
the platform driver, and the missing module alias publication.
The driver has an OF match table wired to .of_match_table, but does
not export the table with MODULE_DEVICE_TABLE().
Add the missing MODULE_DEVICE_TABLE(of, ...) entry so module alias
information is generated for OF based module autoloading.
This is a source-level fix. It does not claim dynamic hardware
reproduction; the evidence is the driver-owned match table, its use by
the platform driver, and the missing module alias publication.
Finite-slice EXT tasks need the periodic scheduler tick to expire their
slices even when nohz_full is enabled.
Add a regression test that selects a nohz_full CPU and exercises both
infinite-to-finite and finite-to-finite slice transitions across an idle
interval. For each finite task, verify that its ops.tick() callback is
invoked.
Skip the test when an allowed nohz_full CPU and a separate housekeeping
CPU are not available.
Signed-off-by: Andrea Righi <arighi@nvidia.com> Signed-off-by: Tejun Heo <tj@kernel.org>
Andrea Righi [Wed, 8 Jul 2026 07:46:48 +0000 (09:46 +0200)]
sched_ext: Enable tick for finite slices on nohz_full
set_next_task_scx() updates the tick dependency before __schedule()
updates rq->curr. When switching from a non-EXT task, such as idle, to
an EXT task with a finite slice, sched_update_tick_dependency() checks
the outgoing task and can allow the tick to remain stopped.
The dependency can also be lost without a slice-type transition. After a
finite-slice task leaves the CPU idle, the enqueue path can clear the
dependency against the idle rq->curr. SCX_RQ_CAN_STOP_TICK still records
a finite slice, so another finite task skips the transition block and
can run without the ticks needed to expire its slice.
The reverse mismatch can also happen when the last finite-slice EXT task
is dequeued: sub_nr_running() updates the dependency before rq->curr
changes, so the outgoing task state can keep the dependency set after
the CPU goes idle.
Fix this by unconditionally enabling the scheduler tick whenever a
finite-slice EXT task is selected on a nohz_full CPU. Moreover, when the
last runnable EXT task leaves, ignore the outgoing EXT slice state so
the generic scheduler can correctly re-evaluate and clear the tick
dependency.
audit: fix potential integer overflow in audit_log_n_hex()
The function calculates new_len as len << 1 for hex encoding. This
has two overflow risks: the shift itself can overflow when len is
large, and the result can be truncated when assigned to new_len
(declared as int) from the size_t calculation.
Fix by using check_shl_overflow() to catch shift overflow and
changing new_len and loop counter i to size_t to prevent truncation.
Cc: stable@vger.kernel.org Fixes: 168b7173959f ("AUDIT: Clean up logging of untrusted strings") Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Ricardo Robaina <rrobaina@redhat.com>
[PM: remove vertical whitspace noise] Signed-off-by: Paul Moore <paul@paul-moore.com>
Andrea Righi [Wed, 8 Jul 2026 08:43:27 +0000 (10:43 +0200)]
sched_ext: Preserve rq tracking across local DSQ dispatch
dispatch_to_local_dsq() can run from scx_bpf_dsq_move_to_local() while
ops.dispatch() has recorded the current rq. Moving a task to a local DSQ
may switch to the source or destination rq before synchronously invoking
ops.dequeue() through the following path:
The nested callback saves the recorded rq and restores it on return. If
the rq tracking does not follow the lock switch, update_locked_rq() can
trigger the following lockdep assertion while restoring an rq which is
no longer held:
Introduce switch_rq_lock() to update the tracking state together with
each rq lock handoff. Use it in dispatch_to_local_dsq(),
move_remote_task_to_local_dsq() and the in-balance paths of
scx_dsq_move(), ensuring that scx_locked_rq() consistently refers to the
rq whose lock is actually held throughout the lock dance.
Fixes: 7fb39e4eb4c3 ("sched_ext: Save and restore scx_locked_rq across SCX_CALL_OP") Cc: stable@vger.kernel.org # 7.1+ Signed-off-by: Andrea Righi <arighi@nvidia.com> Signed-off-by: Tejun Heo <tj@kernel.org>
Chuck Lever [Mon, 4 May 2026 10:28:19 +0000 (06:28 -0400)]
SUNRPC: pin upper rpc_clnt across the TLS connect_worker
The TLS connect path has a use-after-free: nothing pins the
upper rpc_clnt across the delayed connect_worker. xs_connect()
stores task->tk_client in sock_xprt::clnt as a raw pointer
and queues the worker; for TLS-secured transports that worker
is xs_tcp_tls_setup_socket(), which reads several fields out
of the saved pointer (cl_timeout, cl_program, cl_prog,
cl_vers, cl_cred, cl_stats) to construct the args for the
inner handshake rpc_clnt.
The xprt does not reference the rpc_clnt; the rpc_clnt
references the xprt. xs_destroy() does cancel the
connect_worker, but it runs only when the xprt's refcount
drops to zero, which cannot happen until the rpc_clnt
releases its cl_xprt reference in rpc_free_client_work().
When a TLS handshake fails fatally (for example, an mTLS
mount whose client cert does not match the server), the
connecting task is woken with -EACCES and exits, the mount
caller invokes rpc_shutdown_client(), and the upper rpc_clnt
is freed before the queued connect_worker fires.
xs_tcp_tls_setup_socket() then dereferences the freed clnt,
producing the refcount_t underflow Michael Nemanov reported.
Take a reference on the upper rpc_clnt in xs_connect() for
TLS transports via a new rpc_hold_client() helper, and drop
it in the connect_worker's exit path with rpc_release_client().
The xprt_lock_connect() / xprt_unlock_connect() pairing
already serialises xs_connect() with xs_tcp_tls_setup_socket(),
so the take and release are balanced one-for-one.
The non-TLS connect worker (xs_tcp_setup_socket) never reads
sock_xprt::clnt, so leave that path alone and avoid the
clnt-holds-xprt-holds-clnt cycle that would otherwise prevent
xprt destruction.
Reported-by: Michael Nemanov <michael.nemanov@vastdata.com> Closes: https://lore.kernel.org/linux-nfs/40e3d522-dfcf-4fc1-9c55-b5e81f1536d5@vastdata.com/ Fixes: 75eb6af7acdf ("SUNRPC: Add a TCP-with-TLS RPC transport class") Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Tested-by: Michael Nemanov <michael.nemanov@vastdata.com> Reviewed-by: Michael Nemanov <michael.nemanov@vastdata.com> Signed-off-by: Anna Schumaker <anna.schumaker@hammerspace.com>
Chuck Lever [Mon, 4 May 2026 10:28:18 +0000 (06:28 -0400)]
SUNRPC: release lower rpc_clnt if killed waiting for XPRT_LOCKED
xs_tcp_tls_setup_socket() creates a temporary "lower" rpc_clnt with
rpc_create() to drive the inner TLS handshake, then waits for
XPRT_LOCKED on its xprt with TASK_KILLABLE so a stuck handshake can
be aborted by signal. When the wait is interrupted, the function
jumps to out_unlock without releasing lower_clnt. The success path
and the out_close error path both call
rpc_shutdown_client(lower_clnt); only the killed-wait path skips it,
leaking the clnt and its underlying xprt.
Call rpc_shutdown_client() on this path before joining out_unlock.
xprt_release_write() is not needed here because XPRT_LOCKED was
never acquired.
Fixes: 26e8bfa30dac ("SUNRPC/TLS: Lock the lower_xprt during the tls handshake") Signed-off-by: Chuck Lever <chuck.lever@oracle.com> Tested-by: Michael Nemanov <michael.nemanov@vastdata.com> Reviewed-by: Michael Nemanov <michael.nemanov@vastdata.com> Signed-off-by: Anna Schumaker <anna.schumaker@hammerspace.com>
KVM: nVMX: Don't use vmcs01.GUEST_CR3 to snapshot L1's CR3 when EPT is disabled
Add a dedicated field in "struct nested_vmx" to track L1's pre-VM-Enter CR3
instead of using vmcs01.GUEST_CR3, which isn't anywhere near as safe as the
comment purports it to be. E.g. in addition to the warn_on_missed_cc bug
(that was fixed by relocating the consistency check), if getting vmcs12
pages (during actual nested VM-Entry) fails and EPT is disabled (in KVM),
KVM will return control to userspace with vmcs01.GUEST_CR3 holding a guest-
controlled value.
Alternatively, KVM could force a reload of vmcs01.GUEST_CR3 by resetting
the MMU context in the error path, but as above, the safety of the vmcs01
approach is extremely questionable, e.g. it took all of ~4 months for the
code to break.
KVM: nVMX: Move vTPR vs. TPR Threshold consistency check into "normal" checks
Move the off-by-default consistency check for vmcs12.tpr_threshold vs.
the virtual APIC vTPR into the "normal" controls checks, as waiting until
KVM has loaded some amount of state is unnecessary and actively dangerous.
Specifically, failure to unwind vmcs01.GUEST_CR3 to KVM's value when EPT
is disabled results in KVM running L1 with an L1-controlled CR3, not with
KVM's CR3!
Alternatively, KVM could simply reset the MMU to force a reload of
vmcs01.GUEST_CR3, but the _only_ reason the check was shoved into a "late"
flow was to wait until the vmcs12 pages were retrieved. Rather than build
up more crusty code, simply access vTPR using a regular guest memory access
(performance isn't a concern). To circumvent the restrictions that led to
KVM deferring nested_get_vmcs12_pages(), (a) use a VM-scoped API to read
guest memory so that it always hits non-SMM memslots (for RSM), and (b)
skip the check (since its off-by-default anyways) when the vCPU doesn't
want to run, i.e. when userspace is restoring/stuffing state.
If reading guest memory fails, simply skip the consistency check, as KVM's
de facto ABI is that VMX instruction accesses to non-existent memory get
PCI Bus Error semantics, where reads return 0xFFs. And if vTPR=0xFF, then
the vTPR is guaranteed to be greater than or equal to TPR_THRESHOLD.
KVM: x86: Ignore pending PV EOI if the vCPU has since disabled PV EOIs
Ignore KVM's internal "service pending PV EOI" request if the vCPU has
disabled PV EOIs since the request was made. Asserting that PV EOIs are
enabled can fail if reading guest memory in pv_eoi_get_user() fails, i.e.
if pv_eoi_test_and_clr_pending() bails early, *and* the vCPU also disables
PV EOIs.
Hui Wang [Tue, 7 Jul 2026 08:23:48 +0000 (16:23 +0800)]
selftests/rseq: Fix a building error for riscv arch
RISC-V rseq selftests include asm/fence.h from tools/arch/riscv,
but the rseq Makefile only adds tools/include in the CFLAGS, this
results in the building failure both for native and cross build:
In file included from rseq.h:131,
from rseq.c:37:
rseq-riscv.h:11:10: fatal error: asm/fence.h: No such file or directory
To fix it, add the matching tools/arch/$(ARCH)/include path in the
CFLAGS and derive ARCH from SUBARCH for standalone native builds where
ARCH is not set.
Fixes: c92786e179e0 ("KVM: riscv: selftests: Use the existing RISCV_FENCE macro in `rseq-riscv.h`") Cc: stable@vger.kernel.org Signed-off-by: Hui Wang <hui.wang@canonical.com> Link: https://patch.msgid.link/20260707082348.36896-1-hui.wang@canonical.com Signed-off-by: Paul Walmsley <pjw@kernel.org>
leixiang [Mon, 22 Jun 2026 07:51:01 +0000 (15:51 +0800)]
KVM: x86: Nullify irqfd->producer if updating IRTE for bypass fails
Nullify irqfd->producer if updating the IRTE for bypass fails, as leaving a
dangling pointer will result in a use-after-free if the irqfd is reachable
through KVM's routing, but the producer is freed separately. E.g. for VFIO
PCI, the producer is embedded in struct "vfio_pci_irq_ctx" and freed when
the vector is disabled, which can happen independent of routing updates.
Fixes: 77e1b8332d1d ("KVM: x86: Decouple device assignment from IRQ bypass") Cc: stable@vger.kernel.org Signed-off-by: leixiang <leixiang@kylinos.cn> Link: https://patch.msgid.link/1782119051448443.14545.seg@mailgw.kylinos.cn
[sean: drop PPC change, massage changelog] Signed-off-by: Sean Christopherson <seanjc@google.com>
projects / thirdparty / linux.git / log
