commit 502296030ec6b0329e00f9fb15018e170cc63037 upstream.
There appear to be a few different ways that Wacom devices can deal with
confidence:
1. If the device looses confidence in a touch, it will first clear
the tipswitch flag in one report, and then clear the confidence
flag in a second report. This behavior is used by e.g. DTH-2452.
2. If the device looses confidence in a touch, it will clear both
the tipswitch and confidence flags within the same report. This
behavior is used by some AES devices.
3. If the device looses confidence in a touch, it will clear *only*
the confidence bit. The tipswitch bit will remain set so long as
the touch is tracked. This behavior may be used in future devices.
The driver does not currently handle situation 3 properly. Touches that
loose confidence will remain "in prox" and essentially frozen in place
until the tipswitch bit is finally cleared. Not only does this result
in userspace seeing a stuck touch, but it also prevents pen arbitration
from working properly (the pen won't send events until all touches are
up, but we don't currently process events from non-confident touches).
This commit centralizes the checking of the confidence bit in the
wacom_wac_finger_slot() function and has 'prox' depend on it. In the
case where situation 3 is encountered, the treat the touch as though
it was removed, allowing both userspace and the pen arbitration to
act normally.
Signed-off-by: Tatsunosuke Tobita <tatsunosuke.tobita@wacom.com>
Signed-off-by: Ping Cheng <ping.cheng@wacom.com>
Signed-off-by: Jason Gerecke <jason.gerecke@wacom.com>
Fixes: 7fb0413baa ("HID: wacom: Use "Confidence" flag to prevent reporting invalid contacts")
Cc: stable@vger.kernel.org
Signed-off-by: Jiri Kosina <jkosina@suse.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit a25a7df518fc71b1ba981d691e9322e645d2689c upstream.
Previous version of ad7091r event handler received the ADC state pointer
and retrieved the iio device from driver data field with dev_get_drvdata().
However, no driver data have ever been set, which led to null pointer
dereference when running the event handler.
Pass the iio device to the event handler and retrieve the ADC state struct
from it so we avoid the null pointer dereference and save the driver from
filling the driver data field.
Fixes: ca69300173 ("iio: adc: Add support for AD7091R5 ADC")
Signed-off-by: Marcelo Schmitt <marcelo.schmitt@analog.com>
Link: https://lore.kernel.org/r/5024b764107463de9578d5b3b0a3d5678e307b1a.1702746240.git.marcelo.schmitt1@gmail.com
Cc: <Stable@vger.kernel.org>
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit ad362fe07fecf0aba839ff2cc59a3617bd42c33f upstream.
There is a potential UAF scenario in the case of an LPI translation
cache hit racing with an operation that invalidates the cache, such
as a DISCARD ITS command. The root of the problem is that
vgic_its_check_cache() does not elevate the refcount on the vgic_irq
before dropping the lock that serializes refcount changes.
Have vgic_its_check_cache() raise the refcount on the returned vgic_irq
and add the corresponding decrement after queueing the interrupt.
Cc: stable@vger.kernel.org
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20240104183233.3560639-1-oliver.upton@linux.dev
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 7b95382f965133ef61ce44aaabc518c16eb46909 upstream.
When the VMM writes to ISPENDR0 to set the state pending state of
an SGI, we fail to convey this to the HW if this SGI is already
backed by a GICv4.1 vSGI.
This is a bit of a corner case, as this would only occur if the
vgic state is changed on an already running VM, but this can
apparently happen across a guest reset driven by the VMM.
Fix this by always writing out the pending_latch value to the
HW, and reseting it to false.
Reported-by: Kunkun Jiang <jiangkunkun@huawei.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Reviewed-by: Zenghui Yu <yuzenghui@huawei.com>
Cc: stable@vger.kernel.org # 5.10+
Link: https://lore.kernel.org/r/7e7f2c0c-448b-10a9-8929-4b8f4f6e2a32@huawei.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 1c6d984f523f67ecfad1083bb04c55d91977bb15 upstream.
kvm_guest_cpu_offline() tries to disable kvmclock regardless if it is
present in the VM. It leads to write to a MSR that doesn't exist on some
configurations, namely in TDX guest:
unchecked MSR access error: WRMSR to 0x12 (tried to write 0x0000000000000000)
at rIP: 0xffffffff8110687c (kvmclock_disable+0x1c/0x30)
kvmclock enabling is gated by CLOCKSOURCE and CLOCKSOURCE2 KVM paravirt
features.
Do not disable kvmclock if it was not enabled.
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Fixes: c02027b574 ("x86/kvm: Disable kvmclock on all CPUs on shutdown")
Reviewed-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Wanpeng Li <wanpengli@tencent.com>
Cc: stable@vger.kernel.org
Message-Id: <20231205004510.27164-6-kirill.shutemov@linux.intel.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 4e11c29873a8a296a20f99b3e03095e65ebf897d upstream.
We found a failure when using the iperf tool during WiFi performance
testing, where some MSIs were received while clearing the interrupt
status, and these MSIs cannot be serviced.
The interrupt status can be cleared even if the MSI status remains pending.
As such, given the edge-triggered interrupt type, its status should be
cleared before being dispatched to the handler of the underling device.
[kwilczynski: commit log, code comment wording]
Link: https://lore.kernel.org/linux-pci/20231211094923.31967-1-jianjun.wang@mediatek.com
Fixes: 43e6409db6 ("PCI: mediatek: Add MSI support for MT2712 and MT7622")
Signed-off-by: qizhong cheng <qizhong.cheng@mediatek.com>
Signed-off-by: Jianjun Wang <jianjun.wang@mediatek.com>
Signed-off-by: Krzysztof Wilczyński <kwilczynski@kernel.org>
[bhelgaas: rewrap comment]
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 2217fffcd63f86776c985d42e76daa43a56abdf1 upstream.
Commit 6f5e193bfb ("PCI: dwc: Fix dw_pcie_ep_raise_msix_irq() to get
correct MSI-X table address") modified dw_pcie_ep_raise_msix_irq() to
support iATUs which require a specific alignment.
However, this support cannot have been properly tested.
The whole point is for the iATU to map an address that is aligned,
using dw_pcie_ep_map_addr(), and then let the writel() write to
ep->msi_mem + aligned_offset.
Thus, modify the address that is mapped such that it is aligned.
With this change, dw_pcie_ep_raise_msix_irq() matches the logic in
dw_pcie_ep_raise_msi_irq().
Link: https://lore.kernel.org/linux-pci/20231128132231.2221614-1-nks@flawful.org
Fixes: 6f5e193bfb ("PCI: dwc: Fix dw_pcie_ep_raise_msix_irq() to get correct MSI-X table address")
Signed-off-by: Niklas Cassel <niklas.cassel@wdc.com>
Signed-off-by: Krzysztof Wilczyński <kwilczynski@kernel.org>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Cc: stable@vger.kernel.org # 5.7
Cc: Kishon Vijay Abraham I <kishon@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit f0dd488e11e71ac095df7638d892209c629d9af2 upstream.
AP BSSID configuration is missing at AP start. Without this fix, FW returns
STA interface MAC address after first init. When hostapd restarts, it gets MAC
address from netdev before driver sets STA MAC to netdev again. Now MAC address
between hostapd and net interface are different causes STA cannot connect to
AP. After that MAC address of uap0 mlan0 become the same. And issue disappears
after following hostapd restart (another issue is AP/STA MAC address become the
same).
This patch fixes the issue cleanly.
Signed-off-by: David Lin <yu-hao.lin@nxp.com>
Fixes: 12190c5d80 ("mwifiex: add cfg80211 start_ap and stop_ap handlers")
Cc: stable@vger.kernel.org
Reviewed-by: Francesco Dolcini <francesco.dolcini@toradex.com>
Tested-by: Rafael Beims <rafael.beims@toradex.com> # Verdin iMX8MP/SD8997 SD
Acked-by: Brian Norris <briannorris@chromium.org>
Signed-off-by: Kalle Valo <kvalo@kernel.org>
Link: https://msgid.link/20231215005118.17031-1-yu-hao.lin@nxp.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 5894d0089cbc146063dcc0239a78ede0a8142efb upstream.
The rtlwifi driver comes with custom code to write into PCIe Link
Control register. RMW access for the Link Control register requires
locking that is already provided by the standard PCIe capability
accessors.
Convert the custom RMW code writing into LNKCTL register to standard
RMW capability accessors. The accesses are changed to cover the full
LNKCTL register instead of touching just a single byte of the register.
Fixes: 0c8173385e ("rtl8192ce: Add new driver")
Cc: stable@vger.kernel.org
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Kalle Valo <kvalo@kernel.org>
Link: https://lore.kernel.org/r/20231124084725.12738-3-ilpo.jarvinen@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit b3943b3c2971444364e03224cfc828c5789deada upstream.
Ever since introduction in the commit 0c8173385e ("rtl8192ce: Add new
driver") the rtlwifi code has, according to comments, attempted to
disable/enable ASPM of the upstream bridge by writing into its LNKCTL
register. However, the code has never been correct because it performs
the writes to the device instead of the upstream bridge.
Worse yet, the offset where the PCIe capabilities reside is derived
from the offset of the upstream bridge. As a result, the write will use
an offset on the device that does not relate to the LNKCTL register
making the ASPM disable/enable code outright dangerous.
Because of those problems, there is no indication that the driver needs
disable/enable ASPM on the upstream bridge. As the Capabilities offset
is not correctly calculated for the write to target device's LNKCTL
register, the code is not disabling/enabling device's ASPM either.
Therefore, just remove the upstream bridge related ASPM disable/enable
code entirely.
The upstream bridge related ASPM code was the only user of the struct
mp_adapter members num4bytes, pcibridge_pciehdr_offset, and
pcibridge_linkctrlreg so those are removed as well.
Note: This change does not remove the code related to changing the
device's ASPM on purpose (which is independent of this flawed code
related to upstream bridge's ASPM).
Suggested-by: Bjorn Helgaas <bhelgaas@kernel.org>
Fixes: 0c8173385e ("rtl8192ce: Add new driver")
Fixes: 886e14b65a ("rtlwifi: Eliminate raw reads and writes from PCIe portion")
Cc: stable@vger.kernel.org
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Kalle Valo <kvalo@kernel.org>
Link: https://lore.kernel.org/r/20231124084725.12738-2-ilpo.jarvinen@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit e874a79250b39447765ac13272b67ac36ccf2a75 upstream.
Commit 495184ac91 ("mt76: mt7915: add support for applying
pre-calibration data") was fundamentally broken and never worked.
The idea (before NVMEM support) was to expand the MTD function and pass
an additional offset. For normal EEPROM load the offset would always be
0. For the purpose of precal loading, an offset was passed that was
internally the size of EEPROM, since precal data is right after the
EEPROM.
Problem is that the offset value passed is never handled and is actually
overwrite by
offset = be32_to_cpup(list);
ret = mtd_read(mtd, offset, len, &retlen, eep);
resulting in the passed offset value always ingnored. (and even passing
garbage data as precal as the start of the EEPROM is getting read)
Fix this by adding to the current offset value, the offset from DT to
correctly read the piece of data at the requested location.
Cc: stable@vger.kernel.org
Fixes: 495184ac91 ("mt76: mt7915: add support for applying pre-calibration data")
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit afc95681c3068956fed1241a1ff1612c066c75ac upstream.
In some cases the firmware expects cbndx 1 to be assigned to the GMU,
so we also want the default domain for the GMU to be an identy domain.
This way it does not get a context bank assigned. Without this, both
of_dma_configure() and drm/msm's iommu_domain_attach() will trigger
allocating and configuring a context bank. So GMU ends up attached to
both cbndx 1 and later cbndx 2. This arrangement seemingly confounds
and surprises the firmware if the GPU later triggers a translation
fault, resulting (on sc8280xp / lenovo x13s, at least) in the SMMU
getting wedged and the GPU stuck without memory access.
Cc: stable@vger.kernel.org
Signed-off-by: Rob Clark <robdclark@chromium.org>
Tested-by: Johan Hovold <johan+linaro@kernel.org>
Reviewed-by: Robin Murphy <robin.murphy@arm.com>
Link: https://lore.kernel.org/r/20231210180655.75542-1-robdclark@gmail.com
Signed-off-by: Will Deacon <will@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 715d82ba636cb3629a6e18a33bb9dbe53f9936ee upstream.
The following case can cause a crash due to missing attach_btf:
1) load rawtp program
2) load fentry program with rawtp as target_fd
3) create tracing link for fentry program with target_fd = 0
4) repeat 3
In the end we have:
- prog->aux->dst_trampoline == NULL
- tgt_prog == NULL (because we did not provide target_fd to link_create)
- prog->aux->attach_btf == NULL (the program was loaded with attach_prog_fd=X)
- the program was loaded for tgt_prog but we have no way to find out which one
BUG: kernel NULL pointer dereference, address: 0000000000000058
Call Trace:
<TASK>
? __die+0x20/0x70
? page_fault_oops+0x15b/0x430
? fixup_exception+0x22/0x330
? exc_page_fault+0x6f/0x170
? asm_exc_page_fault+0x22/0x30
? bpf_tracing_prog_attach+0x279/0x560
? btf_obj_id+0x5/0x10
bpf_tracing_prog_attach+0x439/0x560
__sys_bpf+0x1cf4/0x2de0
__x64_sys_bpf+0x1c/0x30
do_syscall_64+0x41/0xf0
entry_SYSCALL_64_after_hwframe+0x6e/0x76
Return -EINVAL in this situation.
Fixes: f3a9507554 ("bpf: Allow trampoline re-attach for tracing and lsm programs")
Cc: stable@vger.kernel.org
Signed-off-by: Jiri Olsa <olsajiri@gmail.com>
Acked-by: Jiri Olsa <olsajiri@gmail.com>
Acked-by: Song Liu <song@kernel.org>
Signed-off-by: Dmitrii Dolgov <9erthalion6@gmail.com>
Link: https://lore.kernel.org/r/20240103190559.14750-4-9erthalion6@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit da9065caa594d19b26e1a030fd0cc27bd365d685 upstream.
In min_key_size_set():
if (val > hdev->le_max_key_size || val < SMP_MIN_ENC_KEY_SIZE)
return -EINVAL;
hci_dev_lock(hdev);
hdev->le_min_key_size = val;
hci_dev_unlock(hdev);
In max_key_size_set():
if (val > SMP_MAX_ENC_KEY_SIZE || val < hdev->le_min_key_size)
return -EINVAL;
hci_dev_lock(hdev);
hdev->le_max_key_size = val;
hci_dev_unlock(hdev);
The atomicity violation occurs due to concurrent execution of set_min and
set_max funcs.Consider a scenario where setmin writes a new, valid 'min'
value, and concurrently, setmax writes a value that is greater than the
old 'min' but smaller than the new 'min'. In this case, setmax might check
against the old 'min' value (before acquiring the lock) but write its
value after the 'min' has been updated by setmin. This leads to a
situation where the 'max' value ends up being smaller than the 'min'
value, which is an inconsistency.
This possible bug is found by an experimental static analysis tool
developed by our team, BassCheck[1]. This tool analyzes the locking APIs
to extract function pairs that can be concurrently executed, and then
analyzes the instructions in the paired functions to identify possible
concurrency bugs including data races and atomicity violations. The above
possible bug is reported when our tool analyzes the source code of
Linux 5.17.
To resolve this issue, it is suggested to encompass the validity checks
within the locked sections in both set_min and set_max funcs. The
modification ensures that the validation of 'val' against the
current min/max values is atomic, thus maintaining the integrity of the
settings. With this patch applied, our tool no longer reports the bug,
with the kernel configuration allyesconfig for x86_64. Due to the lack of
associated hardware, we cannot test the patch in runtime testing, and just
verify it according to the code logic.
[1] https://sites.google.com/view/basscheck/
Fixes: 18f81241b7 ("Bluetooth: Move {min,max}_key_size debugfs ...")
Cc: stable@vger.kernel.org
Signed-off-by: Gui-Dong Han <2045gemini@gmail.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 21528c69a0d8483f7c6345b1a0bc8d8975e9a172 upstream.
Documentation/filesystems/ramfs-rootfs-initramfs.rst states:
If CONFIG_TMPFS is enabled, rootfs will use tmpfs instead of ramfs by
default. To force ramfs, add "rootfstype=ramfs" to the kernel command
line.
This currently does not work when root= is provided since then
saved_root_name contains a string and rootfstype= is ignored. Therefore,
ramfs is currently always chosen when root= is provided.
The current behavior for rootfs's filesystem is:
root= | rootfstype= | chosen rootfs filesystem
------------+-------------+--------------------------
unspecified | unspecified | tmpfs
unspecified | tmpfs | tmpfs
unspecified | ramfs | ramfs
provided | ignored | ramfs
rootfstype= should be respected regardless whether root= is given,
as shown below:
root= | rootfstype= | chosen rootfs filesystem
------------+-------------+--------------------------
unspecified | unspecified | tmpfs (as before)
unspecified | tmpfs | tmpfs (as before)
unspecified | ramfs | ramfs (as before)
provided | unspecified | ramfs (compatibility with before)
provided | tmpfs | tmpfs (new)
provided | ramfs | ramfs (new)
This table represents the new behavior.
Fixes: 6e19eded36 ("initmpfs: use initramfs if rootfstype= or root= specified")
Cc: <stable@vger.kernel.org>
Signed-off-by: Rob Landley <rob@landley.net>
Link: https://lore.kernel.org/lkml/8244c75f-445e-b15b-9dbf-266e7ca666e2@landley.net/
Reviewed-and-Tested-by: Mimi Zohar <zohar@linux.ibm.com>
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Link: https://lore.kernel.org/r/20231120011248.396012-1-stefanb@linux.ibm.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 0a535eddbe0dc1de4386046ab849f08aeb2f8faf upstream.
If IOSQE_ASYNC is set and we fail importing an iovec for a readv or
writev request, then we leave ->bytes_done uninitialized and hence the
eventual failure CQE posted can potentially have a random res value
rather than the expected -EINVAL.
Setup ->bytes_done before potentially failing, so we have a consistent
value if we fail the request early.
Cc: stable@vger.kernel.org
Reported-by: xingwei lee <xrivendell7@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit b1db244ffd041a49ecc9618e8feb6b5c1afcdaa7 upstream.
When deactivating the catch-all set element, check the state in the next
generation that represents this transaction.
This bug uncovered after the recent removal of the element busy mark
a2dd0233cb ("netfilter: nf_tables: remove busy mark and gc batch API").
Fixes: aaa31047a6 ("netfilter: nftables: add catch-all set element support")
Cc: stable@vger.kernel.org
Reported-by: lonial con <kongln9170@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 6f64f866aa1ae6975c95d805ed51d7e9433a0016 upstream.
Before calling add partition or resize partition, there is no check
on whether the length is aligned with the logical block size.
If the logical block size of the disk is larger than 512 bytes,
then the partition size maybe not the multiple of the logical block size,
and when the last sector is read, bio_truncate() will adjust the bio size,
resulting in an IO error if the size of the read command is smaller than
the logical block size.If integrity data is supported, this will also
result in a null pointer dereference when calling bio_integrity_free.
Cc: <stable@vger.kernel.org>
Signed-off-by: Min Li <min15.li@samsung.com>
Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20230629142517.121241-1-min15.li@samsung.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 15e4c1f462279b4e128f27de48133e0debe9e0df upstream.
The driver's fsync() is supposed to flush any pending operation to
hardware. It is implemented in this driver by cancelling the queued
deferred IO first, then schedule it for "immediate execution" by calling
schedule_delayed_work() again with delay=0. However, setting delay=0
only means the work is scheduled immediately, it does not mean the work
is executed immediately. There is no guarantee that the work is finished
after schedule_delayed_work() returns. After this driver's fsync()
returns, there can still be pending work. Furthermore, if close() is
called by users immediately after fsync(), the pending work gets
cancelled and fsync() may do nothing.
To ensure that the deferred IO completes, use flush_delayed_work()
instead. Write operations to this driver either write to the device
directly, or invoke schedule_delayed_work(); so by flushing the
workqueue, it can be guaranteed that all previous writes make it to the
device.
Fixes: 5e841b88d2 ("fb: fsync() method for deferred I/O flush.")
Cc: stable@vger.kernel.org
Signed-off-by: Nam Cao <namcao@linutronix.de>
Reviewed-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Helge Deller <deller@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 7c45eaa813476bd195ac1227a64b52f9cf2e2030 upstream.
There are register accesses in the function imx_uart_rs485_config(). The
clock must be enabled for these accesses. This was ensured by calling it
via the function uart_rs485_config() in the probe() function within the
range where the clock is enabled. With the commit 7c7f9bc986 ("serial:
Deassert Transmit Enable on probe in driver-specific way") it was removed
from the probe() function and is now only called through the function
uart_add_one_port() which is located at the end of the probe() function.
But the clock is already switched off in this area. To ensure that the
clock is enabled during register access, move the disabling of the clock
to the very end of the probe() function. To avoid leaking enabled clocks
on error also add an error path for exiting with disabling the clock.
Fixes: 7c7f9bc986 ("serial: Deassert Transmit Enable on probe in driver-specific way")
Cc: stable <stable@kernel.org>
Signed-off-by: Christoph Niedermaier <cniedermaier@dh-electronics.com>
Reviewed-by: Lukas Wunner <lukas@wunner.de>
Link: https://lore.kernel.org/r/20231226113647.39376-1-cniedermaier@dh-electronics.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 2dd23cc4d0e6aa55cf9fb3b05f2f4165b01de81c upstream.
In mon_bin_vma_fault():
offset = vmf->pgoff << PAGE_SHIFT;
if (offset >= rp->b_size)
return VM_FAULT_SIGBUS;
chunk_idx = offset / CHUNK_SIZE;
pageptr = rp->b_vec[chunk_idx].pg;
The code is executed without holding any lock.
In mon_bin_vma_close():
spin_lock_irqsave(&rp->b_lock, flags);
rp->mmap_active--;
spin_unlock_irqrestore(&rp->b_lock, flags);
In mon_bin_ioctl():
spin_lock_irqsave(&rp->b_lock, flags);
if (rp->mmap_active) {
...
} else {
...
kfree(rp->b_vec);
rp->b_vec = vec;
rp->b_size = size;
...
}
spin_unlock_irqrestore(&rp->b_lock, flags);
Concurrent execution of mon_bin_vma_fault() with mon_bin_vma_close() and
mon_bin_ioctl() could lead to atomicity violations. mon_bin_vma_fault()
accesses rp->b_size and rp->b_vec without locking, risking array
out-of-bounds access or use-after-free bugs due to possible modifications
in mon_bin_ioctl().
This possible bug is found by an experimental static analysis tool
developed by our team, BassCheck[1]. This tool analyzes the locking APIs
to extract function pairs that can be concurrently executed, and then
analyzes the instructions in the paired functions to identify possible
concurrency bugs including data races and atomicity violations. The above
possible bug is reported when our tool analyzes the source code of
Linux 6.2.
To address this issue, it is proposed to add a spin lock pair in
mon_bin_vma_fault() to ensure atomicity. With this patch applied, our tool
never reports the possible bug, with the kernel configuration allyesconfig
for x86_64. Due to the lack of associated hardware, we cannot test the
patch in runtime testing, and just verify it according to the code logic.
[1] https://sites.google.com/view/basscheck/
Fixes: 19e6317d24 ("usb: mon: Fix a deadlock in usbmon between ...")
Cc: <stable@vger.kernel.org>
Signed-off-by: Gui-Dong Han <2045gemini@gmail.com>
Link: https://lore.kernel.org/r/20240105052412.9377-1-2045gemini@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 5962ded777d689cd8bf04454273e32228d7fb71f upstream.
When typec_altmode_put_partner is called by a plug altmode upon release,
the port altmode the plug belongs to will not remove its reference to the
plug. The check to see if the altmode being released is a plug evaluates
against the released altmode's partner instead of the calling altmode, so
change adev in typec_altmode_put_partner to properly refer to the altmode
being released.
Because typec_altmode_set_partner calls get_device() on the port altmode,
add partner_adev that points to the port altmode in typec_put_partner to
call put_device() on. typec_altmode_set_partner is not called for port
altmodes, so add a check in typec_altmode_release to prevent
typec_altmode_put_partner() calls on port altmode release.
Fixes: 8a37d87d72 ("usb: typec: Bus type for alternate modes")
Cc: <stable@vger.kernel.org>
Co-developed-by: Christian A. Ehrhardt <lk@c--e.de>
Signed-off-by: Christian A. Ehrhardt <lk@c--e.de>
Signed-off-by: RD Babiera <rdbabiera@google.com>
Tested-by: Christian A. Ehrhardt <lk@c--e.de>
Acked-by: Heikki Krogerus <heikki.krogerus@linux.intel.com>
Link: https://lore.kernel.org/r/20240103181754.2492492-2-rdbabiera@google.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 40c304109e866a7dc123661a5c8ca72f6b5e14e0 upstream.
Supposed DMA cross 4k bounder problem should be fixed at DEV_VER_V2, but
still met problem when do ISO transfer if sg enabled.
Data pattern likes below when sg enabled, package size is 1k and mult is 2
[UVC Header(8B) ] [data(3k - 8)] ...
The received data at offset 0xd000 will get 0xc000 data, len 0x70. Error
happen position as below pattern:
0xd000: wrong
0xe000: wrong
0xf000: correct
0x10000: wrong
0x11000: wrong
0x12000: correct
...
To avoid DMA cross 4k bounder at ISO transfer, reduce burst len according
to start DMA address's alignment.
Cc: <stable@vger.kernel.org>
Fixes: 7733f6c32e ("usb: cdns3: Add Cadence USB3 DRD Driver")
Signed-off-by: Frank Li <Frank.Li@nxp.com>
Link: https://lore.kernel.org/r/20231224153816.1664687-4-Frank.Li@nxp.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 92f02efa1d86d7dcaef7f38a5fe3396c4e88a93c upstream.
ISO basic transfer is
ITP(SOF) Package_0 Package_1 ... Package_n
CDNS3 DMA start dma transfer from memmory to internal FIFO when get SOF,
controller will transfer data to usb bus from internal FIFO when get IN
token.
According USB spec defination:
Maximum number of packets = (bMaxBurst + 1) * (Mult + 1)
Internal memory should be the same as (bMaxBurst + 1) * (Mult + 1). DMA
don't fetch data advance when ISO transfer, so only reserve
(bMaxBurst + 1) * (Mult + 1) internal memory for ISO transfer.
Need save Mult and bMaxBurst information and set it into EP_CFG register,
otherwise only 1 package is sent by controller, other package will be
lost.
Cc: <stable@vger.kernel.org>
Fixes: 7733f6c32e ("usb: cdns3: Add Cadence USB3 DRD Driver")
Signed-off-by: Frank Li <Frank.Li@nxp.com>
Link: https://lore.kernel.org/r/20231224153816.1664687-3-Frank.Li@nxp.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 1b8be5ecff26201bafb0a554c74e91571299fb94 upstream.
When IP version >= DEV_VER_V2, gadget:sg_supported is true. So uvc gadget
function driver will use sg to equeue data, first is 8bytes header, the
second is 1016bytes data.
cdns3_prepare_trb: ep2in: trb 0000000000ac755f, dma buf: 0xbf455000, size: 8, burst: 128 ctrl: 0x00000415 (C=1, T=0, ISP, CHAIN, Normal)
cdns3_prepare_trb: ep2in: trb 00000000a574e693, dma buf: 0xc0200fe0, size: 1016, burst: 128 ctrl: 0x00000405 (C=1, T=0, ISP, Normal)
But cdns3_ep_run_transfer() can't correctly handle this case, which only
support one TRB for ISO transfer.
The controller requires duplicate the TD for each SOF if priv_ep->interval
is not 1. DMA will read data from DDR to internal FIFO when get SOF. Send
data to bus when receive IN token. DMA always refill FIFO when get SOF
regardless host send IN token or not. If host send IN token later, some
frames data will be lost.
Fixed it by below major steps:
1. Calculate numembers of TRB base on sg_nums and priv_ep->interval.
2. Remove CHAIN flags for each end TRB of TD when duplicate TD.
3. The controller requires LINK TRB must be first TRB of TD. When check
there are not enough TRBs lefts, just fill LINK TRB for left TRBs.
.... CHAIN_TRB DATA_TRB, CHAIN_TRB DATA_TRB, LINK_TRB ... LINK_TRB
^End of TRB List
Cc: <stable@vger.kernel.org>
Fixes: 7733f6c32e ("usb: cdns3: Add Cadence USB3 DRD Driver")
Signed-off-by: Frank Li <Frank.Li@nxp.com>
Link: https://lore.kernel.org/r/20231224153816.1664687-2-Frank.Li@nxp.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 128d849074d05545becf86e713715ce7676fc074 upstream.
After the chipidea driver introduce extcon for id and vbus, it's able
to wakeup from another irq source, in case the system with extcon ID
cable, wakeup from usb ID cable and device removal, the usb device
disconnect irq may come firstly before the extcon notifier while system
resume, so we will get 2 "wakeup" irq, one for usb device disconnect;
and one for extcon ID cable change(real wakeup event), current driver
treat them as 2 successive wakeup irq so can't handle it correctly, then
finally the usb irq can't be enabled. This patch adds a check to bypass
further usb events before controller resume finished to fix it.
Fixes: 1f874edcb7 ("usb: chipidea: add runtime power management support")
cc: <stable@vger.kernel.org>
Acked-by: Peter Chen <peter.chen@kernel.org>
Signed-off-by: Xu Yang <xu.yang_2@nxp.com>
Signed-off-by: Li Jun <jun.li@nxp.com>
Link: https://lore.kernel.org/r/20231228110753.1755756-2-xu.yang_2@nxp.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit afe28cd686aeb77e8d9140d50fb1cf06a7ecb731 upstream.
This reverts commit e835c0a4e2.
Don't omit soft-reset. During initialization, the driver may need to
perform a soft reset to ensure the phy is ready when the controller
updates the GCTL.PRTCAPDIR or other settings by issuing phy soft-reset.
Many platforms often have access to DCTL register for soft-reset despite
being host-only. If there are actual reported issues from the platforms
that don't expose DCTL registers, then we will need to revisit (perhaps
to teach dwc3 to perform xhci's soft-reset USBCMD.HCRST).
Cc: <stable@vger.kernel.org>
Fixes: e835c0a4e2 ("usb: dwc3: don't reset device side if dwc3 was configured as host-only")
Signed-off-by: Thinh Nguyen <Thinh.Nguyen@synopsys.com>
Link: https://lore.kernel.org/r/7668ab11a48f260820825274976eb41fec7f54d1.1703282469.git.Thinh.Nguyen@synopsys.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit e9d40b215e38480fd94c66b06d79045717a59e9c upstream.
Current implementation blocks the running operations when Plug-out and
Plug-In is performed continuously, process gets stuck in
dwc3_thread_interrupt().
Code Flow:
CPU1
->Gadget_start
->dwc3_interrupt
->dwc3_thread_interrupt
->dwc3_process_event_buf
->dwc3_process_event_entry
->dwc3_endpoint_interrupt
->dwc3_ep0_interrupt
->dwc3_ep0_inspect_setup
->dwc3_ep0_stall_and_restart
By this time if pending_list is not empty, it will get the next request
on the given list and calls dwc3_gadget_giveback which will unmap request
and call its complete() callback to notify upper layers that it has
completed. Currently dwc3_gadget_giveback status is set to -ECONNRESET,
whereas it should be -ESHUTDOWN based on condition if not dwc->connected
is true.
Cc: <stable@vger.kernel.org>
Fixes: d742220b35 ("usb: dwc3: ep0: giveback requests on stall_and_restart")
Signed-off-by: Uttkarsh Aggarwal <quic_uaggarwa@quicinc.com>
Link: https://lore.kernel.org/r/20231222094704.20276-1-quic_uaggarwa@quicinc.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 71fee48fb772ac4f6cfa63dbebc5629de8b4cc09 upstream.
When offlining and onlining CPUs the overall reported idle and iowait
times as reported by /proc/stat jump backward and forward:
cpu 132 0 176 225249 47 6 6 21 0 0
cpu0 80 0 115 112575 33 3 4 18 0 0
cpu1 52 0 60 112673 13 3 1 2 0 0
cpu 133 0 177 226681 47 6 6 21 0 0
cpu0 80 0 116 113387 33 3 4 18 0 0
cpu 133 0 178 114431 33 6 6 21 0 0 <---- jump backward
cpu0 80 0 116 114247 33 3 4 18 0 0
cpu1 52 0 61 183 0 3 1 2 0 0 <---- idle + iowait start with 0
cpu 133 0 178 228956 47 6 6 21 0 0 <---- jump forward
cpu0 81 0 117 114929 33 3 4 18 0 0
Reason for this is that get_idle_time() in fs/proc/stat.c has different
sources for both values depending on if a CPU is online or offline:
- if a CPU is online the values may be taken from its per cpu
tick_cpu_sched structure
- if a CPU is offline the values are taken from its per cpu cpustat
structure
The problem is that the per cpu tick_cpu_sched structure is set to zero on
CPU offline. See tick_cancel_sched_timer() in kernel/time/tick-sched.c.
Therefore when a CPU is brought offline and online afterwards both its idle
and iowait sleeptime will be zero, causing a jump backward in total system
idle and iowait sleeptime. In a similar way if a CPU is then brought
offline again the total idle and iowait sleeptimes will jump forward.
It looks like this behavior was introduced with commit 4b0c0f294f
("tick: Cleanup NOHZ per cpu data on cpu down").
This was only noticed now on s390, since we switched to generic idle time
reporting with commit be76ea6144 ("s390/idle: remove arch_cpu_idle_time()
and corresponding code").
Fix this by preserving the values of idle_sleeptime and iowait_sleeptime
members of the per-cpu tick_sched structure on CPU hotplug.
Fixes: 4b0c0f294f ("tick: Cleanup NOHZ per cpu data on cpu down")
Reported-by: Gerald Schaefer <gerald.schaefer@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/r/20240115163555.1004144-1-hca@linux.ibm.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 9a9ab0d963621d9d12199df9817e66982582d5a5 upstream.
Task A calls binder_update_page_range() to allocate and insert pages on
a remote address space from Task B. For this, Task A pins the remote mm
via mmget_not_zero() first. This can race with Task B do_exit() and the
final mmput() refcount decrement will come from Task A.
Task A | Task B
------------------+------------------
mmget_not_zero() |
| do_exit()
| exit_mm()
| mmput()
mmput() |
exit_mmap() |
remove_vma() |
fput() |
In this case, the work of ____fput() from Task B is queued up in Task A
as TWA_RESUME. So in theory, Task A returns to userspace and the cleanup
work gets executed. However, Task A instead sleep, waiting for a reply
from Task B that never comes (it's dead).
This means the binder_deferred_release() is blocked until an unrelated
binder event forces Task A to go back to userspace. All the associated
death notifications will also be delayed until then.
In order to fix this use mmput_async() that will schedule the work in
the corresponding mm->async_put_work WQ instead of Task A.
Fixes: 457b9a6f09 ("Staging: android: add binder driver")
Reviewed-by: Alice Ryhl <aliceryhl@google.com>
Signed-off-by: Carlos Llamas <cmllamas@google.com>
Link: https://lore.kernel.org/r/20231201172212.1813387-4-cmllamas@google.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit c7ec4f2d684e17d69bbdd7c4324db0ef5daac26a upstream.
While frontends may submit zero-size requests (wasting a precious slot),
core networking code as of at least 3ece782693 ("sock: skb_copy_ubufs
support for compound pages") can't deal with SKBs when they have all
zero-size fragments. Respond to empty requests right when populating
fragments; all further processing is fragment based and hence won't
encounter these empty requests anymore.
In a way this should have been that way from the beginning: When no data
is to be transferred for a particular request, there's not even a point
in validating the respective grant ref. That's no different from e.g.
passing NULL into memcpy() when at the same time the size is 0.
This is XSA-448 / CVE-2023-46838.
Cc: stable@vger.kernel.org
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Juergen Gross <jgross@suse.com>
Reviewed-by: Paul Durrant <paul@xen.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 50d6281ce9 upstream.
If CONFIG_DMA_DECLARE_COHERENT is not set,
make ARCH=x86_64 CROSS_COMPILE=x86_64-linux-gnu- will be failed, like this:
drivers/remoteproc/remoteproc_core.c: In function ‘rproc_rvdev_release’:
./include/linux/dma-map-ops.h:182:42: error: statement with no effect [-Werror=unused-value]
#define dma_release_coherent_memory(dev) (0)
^
drivers/remoteproc/remoteproc_core.c:464:2: note: in expansion of macro ‘dma_release_coherent_memory’
dma_release_coherent_memory(dev);
^~~~~~~~~~~~~~~~~~~~~~~~~~~
cc1: all warnings being treated as errors
The return type of function dma_release_coherent_memory in CONFIG_DMA_DECLARE_COHERENT area is void, so in !CONFIG_DMA_DECLARE_COHERENT area it should neither return any value nor be defined as zero.
Reported-by: Hulk Robot <hulkci@huawei.com>
Fixes: e61c451476 ("dma-mapping: Add dma_release_coherent_memory to DMA API")
Signed-off-by: Ren Zhijie <renzhijie2@huawei.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20220630123528.251181-1-renzhijie2@huawei.com
Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit c6d05e0762ab276102246d24affd1e116a46aa0c upstream.
Each transaction is associated with a 'struct binder_buffer' that stores
the metadata about its buffer area. Since commit 74310e06be ("android:
binder: Move buffer out of area shared with user space") this struct is
no longer embedded within the buffer itself but is instead allocated on
the heap to prevent userspace access to this driver-exclusive info.
Unfortunately, the space of this struct is still being accounted for in
the total buffer size calculation, specifically for async transactions.
This results in an additional 104 bytes added to every async buffer
request, and this area is never used.
This wasted space can be substantial. If we consider the maximum mmap
buffer space of SZ_4M, the driver will reserve half of it for async
transactions, or 0x200000. This area should, in theory, accommodate up
to 262,144 buffers of the minimum 8-byte size. However, after adding
the extra 'sizeof(struct binder_buffer)', the total number of buffers
drops to only 18,724, which is a sad 7.14% of the actual capacity.
This patch fixes the buffer size calculation to enable the utilization
of the entire async buffer space. This is expected to reduce the number
of -ENOSPC errors that are seen on the field.
Fixes: 74310e06be ("android: binder: Move buffer out of area shared with user space")
Signed-off-by: Carlos Llamas <cmllamas@google.com>
Reviewed-by: Alice Ryhl <aliceryhl@google.com>
Link: https://lore.kernel.org/r/20231201172212.1813387-6-cmllamas@google.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit acc657692aed438e9931438f8c923b2b107aebf9 upstream.
Fix the size check added to dns_resolver_preparse() for the V1 server-list
header so that it doesn't give EINVAL if the size supplied is the same as
the size of the header struct (which should be valid).
This can be tested with:
echo -n -e '\0\0\01\xff\0\0' | keyctl padd dns_resolver desc @p
which will give "add_key: Invalid argument" without this fix.
Fixes: 1997b3cb4217 ("keys, dns: Fix missing size check of V1 server-list header")
Reported-by: Pengfei Xu <pengfei.xu@intel.com>
Link: https://lore.kernel.org/r/ZZ4fyY4r3rqgZL+4@xpf.sh.intel.com/
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Petr Vorel <pvorel@suse.cz>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 716089b417cf98d01f0dc1b39f9c47e1d7b4c965 ]
The expected result value for the call to of_count_phandle_with_args()
was updated from 7 to 8, but the accompanying error message was
forgotten.
Fixes: 4dde83569832f937 ("of: Fix double free in of_parse_phandle_with_args_map")
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/r/20240111085025.2073894-1-geert+renesas@glider.be
Signed-off-by: Rob Herring <robh@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 4dde83569832f9377362e50f7748463340c5db6b ]
In of_parse_phandle_with_args_map() the inner loop that
iterates through the map entries calls of_node_put(new)
to free the reference acquired by the previous iteration
of the inner loop. This assumes that the value of "new" is
NULL on the first iteration of the inner loop.
Make sure that this is true in all iterations of the outer
loop by setting "new" to NULL after its value is assigned to "cur".
Extend the unittest to detect the double free and add an additional
test case that actually triggers this path.
Fixes: bd6f2fd5a1 ("of: Support parsing phandle argument lists through a nexus node")
Cc: Stephen Boyd <stephen.boyd@linaro.org>
Signed-off-by: "Christian A. Ehrhardt" <lk@c--e.de>
Link: https://lore.kernel.org/r/20231229105411.1603434-1-lk@c--e.de
Signed-off-by: Rob Herring <robh@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 2f3be3ca779b11c332441b10e00443a2510f4d7b ]
The hwmgr->backend, (i.e. data) allocated by kzalloc is not freed in
the error-handling paths of smu7_get_evv_voltages and
smu7_update_edc_leakage_table. However, it did be freed in the
error-handling of phm_initializa_dynamic_state_adjustment_rule_settings,
by smu7_hwmgr_backend_fini. So the lack of free in smu7_get_evv_voltages
and smu7_update_edc_leakage_table is considered a memleak in this patch.
Fixes: 599a7e9fe1 ("drm/amd/powerplay: implement smu7 hwmgr to manager asics with smu ip version 7.")
Fixes: 8f0804c6b7 ("drm/amd/pm: add edc leakage controller setting")
Signed-off-by: Zhipeng Lu <alexious@zju.edu.cn>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 2f1888281e67205bd80d3e8f54dbd519a9653f26 ]
The iser_reg_resources structure has two pointers to MR but only one
mr_valid field. The implementation assumes that we use only *sig_mr when
pi_enable is true. Otherwise, we use only *mr. However, it is only
sometimes correct. Read commands without protection information occur even
when pi_enble is true. For example, the following SCSI commands have a
Data-In buffer but never have protection information: READ CAPACITY (16),
INQUIRY, MODE SENSE(6), MAINTENANCE IN. So, we use
*sig_mr for some SCSI commands and *mr for the other SCSI commands.
In most cases, it works fine because the remote invalidation is applied.
However, there are two cases when the remote invalidation is not
applicable.
1. Small write commands when all data is sent as an immediate.
2. The target does not support the remote invalidation feature.
The lazy invalidation is used if the remote invalidation is impossible.
Since, at the lazy invalidation, we always invalidate the MR we want to
use, the wrong MR may be invalidated.
To fix the issue, we need a field per MR that indicates the MR needs
invalidation. Since the ib_mr structure already has such a field, let's
use ib_mr.need_inval instead of iser_reg_resources.mr_valid.
Fixes: b76a439982 ("IB/iser: Use IB_WR_REG_MR_INTEGRITY for PI handover")
Link: https://lore.kernel.org/r/20231219072311.40989-1-sergeygo@nvidia.com
Acked-by: Max Gurtovoy <mgurtovoy@nvidia.com>
Signed-off-by: Sergey Gorenko <sergeygo@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
