linux-stable.git/drivers, branch v7.2.1

ptp: vmclock: prevent read-only mappings from becoming writable

2026-08-27T12:35:27+00:00

commit a5edadbae57e2298a56cf7a4e774a027905a331f upstream.

vmclock_miscdev_mmap() rejects writable mappings of the shared vmclock
ABI page with -EROFS, but leaves VM_MAYWRITE set.  Userspace can map the
page read-only and then upgrade it to writable with mprotect(), after
which the guest can corrupt the host-written timekeeping data (sequence
counter, UTC time, TSC offset) that the vmclock ABI defines as read-only.

Clear VM_MAYWRITE on the read-only path so the mapping cannot be
upgraded, as i915 does for its read-only objects and as fixed in drm/vc4
(CVE-2026-68445) and drm/panthor (CVE-2024-53071).

Cc: stable@vger.kernel.org
Fixes: 205032724226 ("ptp: Add support for the AMZNC10C 'vmclock' device")
Signed-off-by: Abdifatah Suruur 
Link: https://patch.msgid.link/20260813174707.14809-1-suruurism@gmail.com
Signed-off-by: Jakub Kicinski 
Signed-off-by: Greg Kroah-Hartman

Bluetooth: hci_aml: validate firmware segment lengths

2026-08-27T12:35:27+00:00

commit 2bf6b9baca9372ea51b6d0f2820dc9bf29a83ef4 upstream.

aml_download_firmware() reads two lengths from the firmware header and
uses them to build pointers before checking that the header and segment
data are present. A truncated or inconsistent firmware image can make
the driver read past firmware->data while constructing TCI commands.

Reject images shorter than the header and ensure that the ICCM and DCCM
ranges fit within the loaded firmware before downloading either segment.

Fixes: 37bac77e4649 ("Bluetooth: hci_uart: Add support for Amlogic HCI UART")
Cc: stable@vger.kernel.org
Signed-off-by: Laxman Acharya Padhya 
Signed-off-by: Luiz Augusto von Dentz 
Signed-off-by: Greg Kroah-Hartman

HID: input: read battery capacity from its actual report offset

2026-08-27T12:35:26+00:00

commit d07644524b6511b622ee7b0e2e68c9ee43d522a4 upstream.

hidinput_query_battery_capacity() assumes the state-of-charge value is
the first byte following the report ID (buf[1]) and ignores where the
battery field actually sits within the report.

An Apple Magic Trackpad 2 precedes the AbsoluteStateOfCharge byte with a
byte of status flags in its battery reports, so this query returns the
flags byte instead of the charge level.

The device happens to make that easy to observe, because it exposes the
same cell twice: its report descriptor declares AbsoluteStateOfCharge in
two reports (0x90 and 0x9b), so hidinput_setup_battery() registers two
power supplies. Only the first one is refreshed by hid-magicmouse -- it
uses hid_get_battery(), which returns the first battery of the list --
and that refresh goes through the report event path, which parses the
field correctly. Nothing ever reports the second one, so every read of
its capacity takes the query path above. On a USB-C Magic Trackpad over
USB, on an unpatched 7.1.5:

  hid--battery-144 = 100%  (Charging)      <- report event path
  hid--battery-155 =   3%  (Discharging)   <- query path

Both are the same physical battery. A raw HIDIOCGINPUT of the two
reports at that same moment:

  report 0x90 -> [90 03 64]
  report 0x9b -> [9b 03 64 64 00 00 10 00 00 00 00 00 00 00]
                     ^flags ^SoC = 0x64 = 100%

The device answers correctly in both cases; only the offset the kernel
reads the capacity from is wrong. 0x03 is the flags byte (present,
charging), reported as "3%".

Bluetooth takes the same query path for its capacity, where the trackpad
reported a bogus near-constant ~4% -- 0b100, the FullyCharged flag --
regardless of the real charge.

Store the battery field's offset within the report at setup time and use
it when querying, so the capacity is read from its real position. The
report event path already parses the field correctly through the HID
core; only the explicit GET_REPORT query was wrong.

Devices whose capacity field is the first field in the report have a
report_offset of 0 and are unaffected (buf[1 + 0] == buf[1]).

Fixes: 581c4484769e ("HID: input: map digitizer battery usage")
Cc: stable@vger.kernel.org
Signed-off-by: Jose Villaseñor Montfort 
Reviewed-by: Alec Hall 
Signed-off-by: Jiri Kosina 
Signed-off-by: Greg Kroah-Hartman

HID: hyperv: validate initial device info bounds

2026-08-27T12:35:26+00:00

commit 934b7778aa7b7c8f6bb073d2a73ba3674885bae0 upstream.

The Hyper-V synthetic HID host supplies SYNTH_HID_INITIAL_DEVICE_INFO
messages that contain a HID descriptor followed by the report descriptor
bytes. mousevsc_on_receive_device_info() trusts bLength and
wDescriptorLength without checking that the received packet contains both
byte ranges.

A malformed host or backend message can therefore make the guest read
past the received VMBus packet while copying the report descriptor. Pass
the received initial-device-info size into the parser and reject
descriptor lengths that exceed the packet.

Impact: A malicious Hyper-V host or backend can crash a guest by sending
a short initial device-info message with an oversized HID report
descriptor length.

Fixes: b95f5bcb811e ("HID: Move the hid-hyperv driver out of staging")
Cc: stable@vger.kernel.org
Assisted-by: Codex:gpt-5-5-xhigh
Signed-off-by: Michael Bommarito 
Signed-off-by: Jiri Kosina 
Signed-off-by: Greg Kroah-Hartman

HID: uclogic: fix use-after-free of inrange_timer on remove

2026-08-27T12:35:26+00:00

commit 506fd50a9027340f0e9dcc587d10ccb03312dba6 upstream.

uclogic_remove() cancels the pen in-range timer and then stops the
device:

	timer_delete_sync(&drvdata->inrange_timer);
	hid_hw_stop(hdev);

timer_delete_sync() only guarantees the timer is idle at that instant.
uclogic_raw_event_pen() keeps delivering pen reports until hid_hw_stop()
stops the transport several lines later, and every report with
pen->inrange == UCLOGIC_PARAMS_PEN_INRANGE_NONE re-arms the timer:

	mod_timer(&drvdata->inrange_timer, jiffies + msecs_to_jiffies(100));

A report landing between the timer_delete_sync() call and the transport
teardown in hid_hw_stop() re-arms inrange_timer after it was cancelled.
uclogic_remove() then returns and the devm drvdata is freed, while
hid_hw_stop() has already freed the input device drvdata->pen_input
points at, so when the timer fires ~100 ms later
uclogic_inrange_timeout() dereferences freed memory -- a use-after-free
in timer-softirq context.

Swapping the two calls is not a fix: stopping the device first frees
drvdata->pen_input via hidinput_disconnect() while the timer may still
be pending, so a timer already armed before removal fires on the freed
input device in the window before timer_delete_sync() runs.

Use timer_shutdown_sync() before hid_hw_stop() instead. It cancels the
timer, waits for a running callback while pen_input is still valid, and
prevents any further re-arming -- a later mod_timer() from an in-flight
report is silently ignored -- so the timer is provably dead before
hid_hw_stop() frees the inputs. This is the ordering the timer core
documents for this "timer re-armed from another path" teardown case.

Fixes: 01309e29eb95 ("HID: uclogic: Support in-range reporting emulation")
Cc: stable@vger.kernel.org
Signed-off-by: Ibrahim Hashimov 
Assisted-by: AuditCode-AI:2026.07
Signed-off-by: Jiri Kosina 
Signed-off-by: Greg Kroah-Hartman

HID: sensor: custom: Fix use-after-free in enable_sensor

2026-08-27T12:35:26+00:00

commit ad8fb82b04422f49530d2aa2753cc81d1c60102c upstream.

enable_sensor_store() can call set_power_report_state(), which
dereferences sensor_inst->power_state and sensor_inst->report_state.
These pointers refer to entries in sensor_inst->fields.

Create the field attributes before exposing the enable_sensor sysfs
attribute, so enable_sensor cannot be accessed before the state it
depends on has been initialized.

On remove, delete enable_sensor before freeing the field attributes,
so a concurrent sysfs write cannot dereference freed memory through
power_state or report_state.

Reported-by: Sashiko AI Review 
Link: https://sashiko.dev/#/patchset/20260623021950.1736413-1-haoxiang_li2024@163.com?part=1
Fixes: 4a7de0519df5 ("HID: sensor: Custom and Generic sensor support")
Cc: stable@vger.kernel.org
Signed-off-by: Haoxiang Li 
Acked-by: Srinivas Pandruvada 
Signed-off-by: Jiri Kosina 
Signed-off-by: Greg Kroah-Hartman

HID: ft260: fix stack-use-after-return write in I2C read race

2026-08-27T12:35:26+00:00

commit bf3e39df3a397fd82967a31d17c4e02c7feab221 upstream.

ft260_i2c_read() points dev->read_buf at a caller-supplied buffer
(often an on-stack variable), arms a completion and waits up to five
seconds for the device to return the data. The HID input callback
ft260_raw_event() runs in the input/IRQ path, independent of the
dev->lock mutex held by the read path, and copies the device-supplied
payload into dev->read_buf after a plain NULL check.

These two paths share read_buf, read_idx and read_len with no
serialization. If the device delays its response until the read
times out, ft260_i2c_read() resets the controller, clears read_buf
and returns, unwinding the stack frame the buffer lived in. A
response that arrives at that moment lets ft260_raw_event() pass the
NULL check and then memcpy() the device-controlled payload into the
now-freed stack location, a bounded but attacker-influenced
stack-use-after-return write triggerable by malicious or
malfunctioning hardware.

Add a dedicated spinlock that serializes every access to read_buf,
read_idx and read_len. ft260_raw_event() now holds it across the
NULL check, the memcpy and the index update, while the read path
takes it when arming and when clearing the buffer, so the teardown
can no longer slip between the check and the copy.

Fixes: 6a82582d9fa4 ("HID: ft260: add usb hid to i2c host bridge driver")
Cc: stable@vger.kernel.org
Signed-off-by: Raman Varabets 
Reviewed-by: Michael Zaidman 
Signed-off-by: Jiri Kosina 
Signed-off-by: Greg Kroah-Hartman

HID: core: fix number/pointer type confusion on long items

2026-08-27T12:35:26+00:00

commit 28abce951343fcec26e397610868efa4e1395c3f upstream.

When fetch_item() is called by hid_scan_report() on an item with
HID_ITEM_TAG_LONG, it stores a pointer to the item data in
item->data.longdata instead of storing a value directly in
item->data.{u8/u16/u32}.

When item_udata() or item_sdata() encounters such an item, it incorrectly
assumes that the item is in short format, and therefore returns the lower
part of a kernel pointer reinterpreted as a number.

When a HID device is connected whose descriptor contains a
HID_GLOBAL_ITEM_TAG_REPORT_SIZE encoded in long format with size=4, this
causes the lower half of a kernel pointer to be printed into dmesg as a
number, like this:

    hid (null): invalid report_size 107953555

To fix it, let item_udata() and item_sdata() verify that the item is in
short format.

Note that this bug only affects hid_scan_report(), while the main parsing
pass hid_parse_collections() will always bail out when encountering a long
item.

Sidenote: There are currently no users of data.longdata; maybe we should
just remove any parsing of long-format descriptors as a follow-up.

Fixes: 3dc8fc083dbf ("HID: Use hid_parser for pre-scanning the report descriptors")
Cc: stable@vger.kernel.org
Signed-off-by: Jann Horn 
Signed-off-by: Jiri Kosina 
Signed-off-by: Greg Kroah-Hartman

HID: rapoo: fix missing hid_is_usb() check

2026-08-27T12:35:26+00:00

commit b57af2448268c30c25509a832b6ff6dc27176b28 upstream.

to_usb_interface() can only be used on a hid_device whose parent is really
USB; uhid can create devices that identify as being on BUS_USB, but don't
actually have a USB parent.
Fix the use of to_usb_interface() without a hid_is_usb() check.

Add a dependency on USB_HID for hid_is_usb(), as other HID drivers do; the
alternative would be to provide a simple stub implementation on !USB_HID
builds.

I have verified that it is currently possible to trigger a kernel splat due
to this bug in an ASAN build, and that this commit fixes the issue.

Fixes: b3b1c68fb726 ("HID: rapoo: Add support for side buttons on RAPOO 0x2015 mouse")
Cc: stable@vger.kernel.org
Signed-off-by: Jann Horn 
Signed-off-by: Jiri Kosina 
Signed-off-by: Greg Kroah-Hartman

HID: nintendo: stop device IO before hid_hw_stop on probe failure

2026-08-27T12:35:26+00:00

commit 1f74d3bff6fe04a64e02ab3661d2e0d554565aa6 upstream.

nintendo_hid_probe() calls hid_device_io_start() before joycon_init()
and joycon_leds_create().  If either fails, the error path jumps to
err_close which calls hid_hw_close()/hid_hw_stop() without first calling
hid_device_io_stop().

hid_hw_stop() does not stop device IO, so hid_input_report() may still
run and access driver data that is being torn down, resulting in a
use-after-free.

Add an err_io_stop label that calls hid_device_io_stop() before
hid_hw_close(), and point the two post-io_start error paths at it.

Fixes: 2af16c1f846b ("HID: nintendo: add nintendo switch controller driver")
Cc: stable@vger.kernel.org
Signed-off-by: Jiangshan Yi 
Signed-off-by: Jiri Kosina 
Signed-off-by: Greg Kroah-Hartman