When writing an empty string to either 'qw_sign' or 'landingPage' sysfs attributes, the store functions attempt to access page[l - 1] before validating that the length 'l' is greater than zero. This patch fixes the vulnerability by adding a check at the beginning of os_desc_qw_sign_store() and webusb_landingPage_store() to handle the zero-length input case gracefully by returning immediately. Signed-off-by: Xinyu Liu <katieeliu@tencent.com> Cc: stable <stable@kernel.org> Link: https://lore.kernel.org/r/tencent_B1C9481688D0E95E7362AB2E999DE8048207@qq.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
A race condition occurs when gs_start_io() calls either gs_start_rx() or
gs_start_tx(), as those functions briefly drop the port_lock for
usb_ep_queue(). This allows gs_close() and gserial_disconnect() to clear
port.tty and port_usb, respectively.
Use the null-safe TTY Port helper function to wake up TTY.
Example
CPU1: CPU2:
gserial_connect() // lock
gs_close() // await lock
gs_start_rx() // unlock
usb_ep_queue()
gs_close() // lock, reset port.tty and unlock
gs_start_rx() // lock
tty_wakeup() // NPE
Fixes: 35f95fd7f234 ("TTY: usb/u_serial, use tty from tty_port")
Cc: stable <stable@kernel.org>
Signed-off-by: Kuen-Han Tsai <khtsai@google.com>
Reviewed-by: Prashanth K <prashanth.k@oss.qualcomm.com>
Link: https://lore.kernel.org/linux-usb/20240116141801.396398-1-khtsai@google.com/
Link: https://lore.kernel.org/r/20250617050844.1848232-2-khtsai@google.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This reverts commit ffd603f214237e250271162a5b325c6199a65382.
Commit ffd603f21423 ("usb: gadget: u_serial: Add null pointer check in
gs_start_io") adds null pointer checks at the beginning of the
gs_start_io() function to prevent a null pointer dereference. However,
these checks are redundant because the function's comment already
requires callers to hold the port_lock and ensure port.tty and port_usb
are not null. All existing callers already follow these rules.
The true cause of the null pointer dereference is a race condition. When
gs_start_io() calls either gs_start_rx() or gs_start_tx(), the port_lock
is temporarily released for usb_ep_queue(). This allows port.tty and
port_usb to be cleared.
Fixes: ffd603f21423 ("usb: gadget: u_serial: Add null pointer check in gs_start_io")
Cc: stable <stable@kernel.org>
Signed-off-by: Kuen-Han Tsai <khtsai@google.com>
Reviewed-by: Prashanth K <prashanth.k@oss.qualcomm.com>
Link: https://lore.kernel.org/r/20250617050844.1848232-1-khtsai@google.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Move this API to the canonical timer_*() namespace. [ tglx: Redone against pre rc1 ] Signed-off-by: Ingo Molnar <mingo@kernel.org> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Link: https://lore.kernel.org/all/aB2X0jCKQO56WdMt@gmail.com
Pull USB / Thunderbolt updates from Greg KH:
"Here is the big set of USB and Thunderbolt changes for 6.16-rc1.
Included in here are the following:
- USB offload support for audio devices.
I think this takes the record for the most number of patch series
(30+) over the longest period of time (2+ years) to get merged
properly.
Many props go to Wesley Cheng for seeing this effort through, they
took a major out-of-tree hacked-up-monstrosity that was created by
multiple vendors for their specific devices, got it all merged into
a semi-coherent set of changes, and got all of the different major
subsystems to agree on how this should be implemented both with
changes to their code as well as userspace apis, AND wrangled the
hardware companies into agreeing to go forward with this, despite
making them all redo work they had already done in their private
device trees.
This feature offers major power savings on embedded devices where a
USB audio stream can continue to flow while the rest of the system
is sleeping, something that devices running on battery power really
care about. There are still some more small tweaks left to be done
here, and those patches are still out for review and arguing among
the different hardware companies, but this is a major step forward
and a great example of how to do upstream development well.
- small number of thunderbolt fixes and updates, things seem to be
slowing down here (famous last words...)
- xhci refactors and reworking to try to handle some rough corner
cases in some hardware implementations where things don't always
work properly
- typec driver updates
- USB3 power management reworking and updates
- Removal of some old and orphaned UDC gadget drivers that had not
been used in a very long time, dropping over 11 thousand lines from
the tree, always a nice thing, making up for the 12k lines added
for the USB offload feature.
- lots of little updates and fixes in different drivers
All of these have been in linux-next for over 2 weeks, the USB offload
logic has been in there for 8 weeks now, with no reported issues"
* tag 'usb-6.16-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb: (172 commits)
ALSA: usb-audio: qcom: fix USB_XHCI dependency
ASoC: qdsp6: fix compile-testing without CONFIG_OF
usb: misc: onboard_usb_dev: fix build warning for CONFIG_USB_ONBOARD_DEV_USB5744=n
usb: typec: tipd: fix typo in TPS_STATUS_HIGH_VOLAGE_WARNING macro
USB: typec: fix const issue in typec_match()
USB: gadget: udc: fix const issue in gadget_match_driver()
USB: gadget: fix up const issue with struct usb_function_instance
USB: serial: pl2303: add new chip PL2303GC-Q20 and PL2303GT-2AB
USB: serial: bus: fix const issue in usb_serial_device_match()
usb: usbtmc: Fix timeout value in get_stb
usb: usbtmc: Fix read_stb function and get_stb ioctl
ALSA: qc_audio_offload: try to reduce address space confusion
ALSA: qc_audio_offload: avoid leaking xfer_buf allocation
ALSA: qc_audio_offload: rename dma/iova/va/cpu/phys variables
ALSA: usb-audio: qcom: Fix an error handling path in qc_usb_audio_probe()
usb: misc: onboard_usb_dev: Fix usb5744 initialization sequence
dt-bindings: usb: ti,usb8041: Add binding for TI USB8044 hub controller
usb: misc: onboard_usb_dev: Add support for TI TUSB8044 hub
usb: gadget: lpc32xx_udc: Use USB API functions rather than constants
usb: gadget: epautoconf: Use USB API functions rather than constants
...
Pull HID updates from Jiri Kosina: - support for Apple Magic Mouse 2 USB-C (Aditya Garg) - power management improvement for multitouch devices (Werner Sembach) - fix for ACPI initialization in intel-thc driver (Wentao Guan) - adaptation of HID drivers to use new gpio_chip's line setter callbacks (Bartosz Golaszewski) - fix potential OOB in usbhid_parse() (Terry Junge) - make it possible to set hid_mouse_ignore_list dynamically (the same way we handle other quirks) (Aditya Garg) - other small assorted fixes and device ID additions * tag 'hid-for-linus-2025060301' of git://git.kernel.org/pub/scm/linux/kernel/git/hid/hid: HID: multitouch: Disable touchpad on firmware level while not in use HID: core: Add functions for HID drivers to react on first open and last close call HID: HID_APPLETB_BL should depend on X86 HID: HID_APPLETB_KBD should depend on X86 HID: appletb-kbd: Use secs_to_jiffies() instead of msecs_to_jiffies() HID: intel-thc-hid: intel-thc: make read-only arrays static const HID: magicmouse: Apple Magic Mouse 2 USB-C support HID: mcp2221: use new line value setter callbacks HID: mcp2200: use new line value setter callbacks HID: cp2112: use new line value setter callbacks HID: cp2112: use lock guards HID: cp2112: hold the lock for the entire direction_output() call HID: cp2112: destroy mutex on driver detach HID: intel-thc-hid: intel-quicki2c: pass correct arguments to acpi_evaluate_object HID: corsair-void: Use to_delayed_work() HID: hid-logitech: use sysfs_emit_at() instead of scnprintf() HID: quirks: Add HID_QUIRK_IGNORE_MOUSE quirk HID: usbhid: Eliminate recurrent out-of-bounds bug in usbhid_parse() HID: Kysona: Add periodic online check
gadget_match_driver() takes a const pointer, and then decides to cast it
away into a non-const one, which is not a good thing to do overall. Fix
this up by properly setting the pointers to be const to preserve that
attribute.
Fixes: d69d80484598 ("driver core: have match() callback in struct bus_type take a const *")
Link: https://lore.kernel.org/r/2025052139-rash-unsaddle-7c5e@gregkh
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In struct usb_function, the struct usb_function_instance pointer variable "fi" is listed as const, but it is written to in numerous places, making the const marking of it a total lie. Fix this up by just removing the const pointer attribute as this is modified in numerous places. Link: https://lore.kernel.org/r/2025052145-undress-puma-f7cf@gregkh Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Use the function usb_endpoint_type() rather than constants. The Coccinelle semantic patch is as follows: @@ struct usb_endpoint_descriptor *epd; @@ - (epd->bmAttributes & \(USB_ENDPOINT_XFERTYPE_MASK\|3\)) + usb_endpoint_type(epd) Signed-off-by: Chen Ni <nichen@iscas.ac.cn> Link: https://lore.kernel.org/r/20250519063120.724793-1-nichen@iscas.ac.cn Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Use the function usb_endpoint_type() rather than constants. The Coccinelle semantic patch is as follows: @@ struct usb_endpoint_descriptor *epd; @@ - (epd->bmAttributes & \(USB_ENDPOINT_XFERTYPE_MASK\|3\)) + usb_endpoint_type(epd) Signed-off-by: Chen Ni <nichen@iscas.ac.cn> Link: https://lore.kernel.org/r/20250519062545.724727-1-nichen@iscas.ac.cn Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>