mirror of
https://gitee.com/bianbu-linux/linux-6.6
synced 2025-07-18 01:23:36 -04:00
commit a4b46d450c49f32e9d4247b421e58083fde304ce upstream.
Testing with KASAN and syzkaller revealed a bug in port.c:disable_store():
usb_hub_to_struct_hub() can return NULL if the hub that the port belongs to
is concurrently removed, but the function does not check for this
possibility before dereferencing the returned value.
It turns out that the first dereference is unnecessary, since hub->intfdev
is the parent of the port device, so it can be changed easily. Adding a
check for hub == NULL prevents further problems.
The same bug exists in the disable_show() routine, and it can be fixed the
same way.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Reported-and-tested-by: Yue Sun <samsun1006219@gmail.com>
Reported-by: xingwei lee <xrivendell7@gmail.com>
Link: https://lore.kernel.org/linux-usb/CAEkJfYON+ry7xPx=AiLR9jzUNT+i_Va68ACajOC3HoacOfL1ig@mail.gmail.com/
Fixes:
|
||
---|---|---|
.. | ||
buffer.c | ||
config.c | ||
devices.c | ||
devio.c | ||
driver.c | ||
endpoint.c | ||
file.c | ||
generic.c | ||
hcd-pci.c | ||
hcd.c | ||
hub.c | ||
hub.h | ||
Kconfig | ||
ledtrig-usbport.c | ||
Makefile | ||
message.c | ||
notify.c | ||
of.c | ||
otg_productlist.h | ||
phy.c | ||
phy.h | ||
port.c | ||
quirks.c | ||
sysfs.c | ||
urb.c | ||
usb-acpi.c | ||
usb.c | ||
usb.h |