linux-stable.git/drivers/input/rmi4, branch linux-5.4.y

Input: synaptics-rmi - fix crash with unsupported versions of F34

2025-06-27T10:02:49+00:00

[ Upstream commit ca39500f6af9cfe6823dc5aa8fbaed788d6e35b2 ]

Sysfs interface for updating firmware for RMI devices is available even
when F34 probe fails. The code checks for presence of F34 "container"
pointer and then tries to use the function data attached to the
sub-device. F34 assigns the function data early, before it knows if
probe will succeed, leaving behind a stale pointer.

Fix this by expanding checks to not only test for presence of F34
"container" but also check if there is driver data assigned to the
sub-device, and call dev_set_drvdata() only after we are certain that
probe is successful.

This is not a complete fix, since F34 will be freed during firmware
update, so there is still a race when fetching and accessing this
pointer. This race will be addressed in follow-up changes.

Reported-by: Hanno Böck 
Fixes: 29fd0ec2bdbe ("Input: synaptics-rmi4 - add support for F34 device reflash")
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/aBlAl6sGulam-Qcx@google.com
Signed-off-by: Dmitry Torokhov 
Signed-off-by: Sasha Levin

Input: synaptics-rmi4 - convert to use sysfs_emit() APIs

2025-06-27T10:02:49+00:00

[ Upstream commit 9dedc915937c33302df7fcab01c45e7936d6195a ]

Follow the advice of the Documentation/filesystems/sysfs.rst and show()
should only use sysfs_emit() or sysfs_emit_at() when formatting the value
to be returned to user space.

Reported-by: Zeal Robot 
Signed-off-by: zhang songyi 
Link: https://lore.kernel.org/r/20220927070936.258300-1-zhang.songyi@zte.com.cn
Signed-off-by: Dmitry Torokhov 
Stable-dep-of: ca39500f6af9 ("Input: synaptics-rmi - fix crash with unsupported versions of F34")
Signed-off-by: Sasha Levin

Input: synaptics-rmi4 - fix UAF of IRQ domain on driver removal

2024-11-08T15:20:43+00:00

commit fbf8d71742557abaf558d8efb96742d442720cc2 upstream.

Calling irq_domain_remove() will lead to freeing the IRQ domain
prematurely. The domain is still referenced and will be attempted to get
used via rmi_free_function_list() -> rmi_unregister_function() ->
irq_dispose_mapping() -> irq_get_irq_data()'s ->domain pointer.

With PaX's MEMORY_SANITIZE this will lead to an access fault when
attempting to dereference embedded pointers, as in Torsten's report that
was faulting on the 'domain->ops->unmap' test.

Fix this by releasing the IRQ domain only after all related IRQs have
been deactivated.

Fixes: 24d28e4f1271 ("Input: synaptics-rmi4 - convert irq distribution to irq_domain")
Reported-by: Torsten Hilbrich 
Signed-off-by: Mathias Krause 
Link: https://lore.kernel.org/r/20240222142654.856566-1-minipli@grsecurity.net
Signed-off-by: Dmitry Torokhov 
Signed-off-by: Tzung-Bi Shih 
Signed-off-by: Sasha Levin

Input: synaptics-rmi4 - fail probing if memory allocation for "phys" fails

2024-04-13T10:51:38+00:00

[ Upstream commit bc4996184d56cfaf56d3811ac2680c8a0e2af56e ]

While input core can work with input->phys set to NULL userspace might
depend on it, so better fail probing if allocation fails. The system must
be in a pretty bad shape for it to happen anyway.

Signed-off-by: Kunwu Chan 
Link: https://lore.kernel.org/r/20240117073124.143636-1-chentao@kylinos.cn
Signed-off-by: Dmitry Torokhov 
Signed-off-by: Sasha Levin

Input: synaptics-rmi4 - fix use after free in rmi_unregister_function()

2023-11-20T09:30:15+00:00

[ Upstream commit eb988e46da2e4eae89f5337e047ce372fe33d5b1 ]

The put_device() calls rmi_release_function() which frees "fn" so the
dereference on the next line "fn->num_of_irqs" is a use after free.
Move the put_device() to the end to fix this.

Fixes: 24d28e4f1271 ("Input: synaptics-rmi4 - convert irq distribution to irq_domain")
Signed-off-by: Dan Carpenter 
Link: https://lore.kernel.org/r/706efd36-7561-42f3-adfa-dd1d0bd4f5a1@moroto.mountain
Signed-off-by: Dmitry Torokhov 
Signed-off-by: Sasha Levin

Input: synaptics-rmi4 - handle reset delay when using SMBus trsnsport

2023-11-08T10:23:40+00:00

[ Upstream commit 5030b2fe6aab37fe42d14f31842ea38be7c55c57 ]

Touch controllers need some time after receiving reset command for the
firmware to finish re-initializing and be ready to respond to commands
from the host. The driver already had handling for the post-reset delay
for I2C and SPI transports, this change adds the handling to
SMBus-connected devices.

SMBus devices are peculiar because they implement legacy PS/2
compatibility mode, so reset is actually issued by psmouse driver on the
associated serio port, after which the control is passed to the RMI4
driver with SMBus companion device.

Note that originally the delay was added to psmouse driver in
92e24e0e57f7 ("Input: psmouse - add delay when deactivating for SMBus
mode"), but that resulted in an unwanted delay in "fast" reconnect
handler for the serio port, so it was decided to revert the patch and
have the delay being handled in the RMI4 driver, similar to the other
transports.

Tested-by: Jeffery Miller 
Link: https://lore.kernel.org/r/ZR1yUFJ8a9Zt606N@penguin
Signed-off-by: Dmitry Torokhov 
Signed-off-by: Sasha Levin

Input: synaptics-rmi4 - fix error return code in rmi_driver_probe()

2020-06-03T06:21:17+00:00

[ Upstream commit 5caab2da63207d6d631007f592f5219459e3454d ]

Fix to return a negative error code from the input_register_device()
error handling case instead of 0, as done elsewhere in this function.

Signed-off-by: Wei Yongjun 
Link: https://lore.kernel.org/r/20200428134948.78343-1-weiyongjun1@huawei.com
Cc: stable@vger.kernel.org
Signed-off-by: Dmitry Torokhov 
Signed-off-by: Sasha Levin

Input: synaptics-rmi4 - really fix attn_data use-after-free

2020-06-03T06:21:17+00:00

[ Upstream commit d5a5e5b5fa7b86c05bf073acc0ba98fa280174ec ]

Fix a use-after-free noticed by running with KASAN enabled. If
rmi_irq_fn() is run twice in a row, then rmi_f11_attention() (among
others) will end up reading from drvdata->attn_data.data, which was
freed and left dangling in rmi_irq_fn().

Commit 55edde9fff1a ("Input: synaptics-rmi4 - prevent UAF reported by
KASAN") correctly identified and analyzed this bug. However the attempted
fix only NULLed out a local variable, missing the fact that
drvdata->attn_data is a struct, not a pointer.

NULL out the correct pointer in the driver data to prevent the attention
functions from copying from it.

Fixes: 55edde9fff1a ("Input: synaptics-rmi4 - prevent UAF reported by KASAN")
Fixes: b908d3cd812a ("Input: synaptics-rmi4 - allow to add attention data")
Signed-off-by: Evan Green 
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20200427145537.1.Ic8f898e0147beeee2c005ee7b20f1aebdef1e7eb@changeid
Signed-off-by: Dmitry Torokhov 
Signed-off-by: Sasha Levin

Revert "Input: synaptics-rmi4 - don't increment rmiaddr for SMBus transfers"

2020-01-29T15:45:27+00:00

commit 8ff771f8c8d55d95f102cf88a970e541a8bd6bcf upstream.

This reverts commit a284e11c371e446371675668d8c8120a27227339.

This causes problems (drifting cursor) with at least the F11 function that
reads more than 32 bytes.

The real issue is in the F54 driver, and so this should be fixed there, and
not in rmi_smbus.c.

So first revert this bad commit, then fix the real problem in F54 in another
patch.

Signed-off-by: Hans Verkuil 
Reported-by: Timo Kaufmann 
Fixes: a284e11c371e ("Input: synaptics-rmi4 - don't increment rmiaddr for SMBus transfers")
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20200115124819.3191024-2-hverkuil-cisco@xs4all.nl
Signed-off-by: Dmitry Torokhov 
Signed-off-by: Greg Kroah-Hartman

Input: synaptics-rmi4 - don't increment rmiaddr for SMBus transfers

2019-12-13T07:42:42+00:00

commit a284e11c371e446371675668d8c8120a27227339 upstream.

This increment of rmi_smbus in rmi_smb_read/write_block() causes
garbage to be read/written.

The first read of SMB_MAX_COUNT bytes is fine, but after that
it is nonsense. Trial-and-error showed that by dropping the
increment of rmiaddr everything is fine and the F54 function
properly works.

I tried a hack with rmi_smb_write_block() as well (writing to the
same F54 touchpad data area, then reading it back), and that
suggests that there too the rmiaddr increment has to be dropped.
It makes sense that if it has to be dropped for read, then it has
to be dropped for write as well.

It looks like the initial work with F54 was done using i2c, not smbus,
and it seems nobody ever tested F54 with smbus. The other functions
all read/write less than SMB_MAX_COUNT as far as I can tell, so this
issue was never noticed with non-F54 functions.

With this change I can read out the touchpad data correctly on my
Lenovo X1 Carbon 6th Gen laptop.

Signed-off-by: Hans Verkuil 
Link: https://lore.kernel.org/r/8dd22e21-4933-8e9c-a696-d281872c8de7@xs4all.nl
Cc: stable@vger.kernel.org
Signed-off-by: Dmitry Torokhov 
Signed-off-by: Greg Kroah-Hartman