linux-stable.git/drivers/rpmsg, branch linux-5.4.y

rpmsg: qcom_smd: Fix uninitialized return variable in __qcom_smd_send()

2025-06-27T10:02:48+00:00

[ Upstream commit 5de775df3362090a6e90046d1f2d83fe62489aa0 ]

The "ret" variable isn't initialized if we don't enter the loop.  For
example,  if "channel->state" is not SMD_CHANNEL_OPENED.

Fixes: 33e3820dda88 ("rpmsg: smd: Use spinlock in tx path")
Signed-off-by: Dan Carpenter 
Link: https://lore.kernel.org/r/aAkhvV0nSbrsef1P@stanley.mountain
Signed-off-by: Bjorn Andersson 
Signed-off-by: Sasha Levin

rpmsg: glink: Propagate TX failures in intentless mode as well

2024-12-14T18:44:39+00:00

commit 7a68f9fa97357a0f2073c9c31ed4101da4fce93e upstream.

As support for splitting transmission over several messages using
TX_DATA_CONT was introduced it does not immediately return the return
value of qcom_glink_tx().

The result is that in the intentless case (i.e. intent == NULL), the
code will continue to send all additional chunks. This is wasteful, and
it's possible that the send operation could incorrectly indicate
success, if the last chunk fits in the TX fifo.

Fix the condition.

Fixes: 8956927faed3 ("rpmsg: glink: Add TX_DATA_CONT command while sending")
Reviewed-by: Chris Lew 
Signed-off-by: Bjorn Andersson 
Signed-off-by: Bjorn Andersson 
Link: https://lore.kernel.org/r/20230418163018.785524-2-quic_bjorande@quicinc.com
Signed-off-by: Greg Kroah-Hartman

rpmsg: glink: use only lower 16-bits of param2 for CMD_OPEN name length

2024-12-14T18:44:32+00:00

[ Upstream commit 06c59d97f63c1b8af521fa5aef8a716fb988b285 ]

The name len field of the CMD_OPEN packet is only 16-bits and the upper
16-bits of "param2" are a different "prio" field, which can be nonzero in
certain situations, and CMD_OPEN packets can be unexpectedly dropped
because of this.

Fix this by masking out the upper 16 bits of param2.

Fixes: b4f8e52b89f6 ("rpmsg: Introduce Qualcomm RPM glink driver")
Signed-off-by: Jonathan Marek 
Reviewed-by: Dmitry Baryshkov 
Link: https://lore.kernel.org/r/20241007235935.6216-1-jonathan@marek.ca
Signed-off-by: Bjorn Andersson 
Signed-off-by: Sasha Levin

rpmsg: glink: Fix GLINK command prefix

2024-12-14T18:44:32+00:00

[ Upstream commit 4e816d0318fdfe8932da80dbf04ba318b13e4b3a ]

The upstream GLINK driver was first introduced to communicate with the
RPM on MSM8996, presumably as an artifact from that era the command
defines was prefixed RPM_CMD, while they actually are GLINK_CMDs.

Let's rename these, to keep things tidy. No functional change.

Signed-off-by: Bjorn Andersson 
Reviewed-by: Chris Lew 
Signed-off-by: Bjorn Andersson 
Link: https://lore.kernel.org/r/20230214225933.2025595-1-quic_bjorande@quicinc.com
Stable-dep-of: 06c59d97f63c ("rpmsg: glink: use only lower 16-bits of param2 for CMD_OPEN name length")
Signed-off-by: Sasha Levin

rpmsg: glink: Send READ_NOTIFY command in FIFO full case

2024-12-14T18:44:32+00:00

[ Upstream commit b16a37e1846c9573a847a56fa2f31ba833dae45a ]

The current design sleeps unconditionally in TX FIFO full case and
wakeup only after sleep timer expires which adds random delays in
clients TX path.

Avoid sleep and use READ_NOTIFY command so that writer can be woken up
when remote notifies about read completion by sending IRQ.

Signed-off-by: Deepak Kumar Singh 
Signed-off-by: Arun Kumar Neelakantam 
Signed-off-by: Bjorn Andersson 
Link: https://lore.kernel.org/r/1596086296-28529-7-git-send-email-deesin@codeaurora.org
Stable-dep-of: 06c59d97f63c ("rpmsg: glink: use only lower 16-bits of param2 for CMD_OPEN name length")
Signed-off-by: Sasha Levin

rpmsg: glink: Add TX_DATA_CONT command while sending

2024-12-14T18:44:32+00:00

[ Upstream commit 8956927faed366b60b0355f4a4317a10e281ced7 ]

With current design the transport can send packets of size upto
FIFO_SIZE which is 16k and return failure for all packets above 16k.

Add TX_DATA_CONT command to send packets greater than 16k by splitting
into 8K chunks.

Signed-off-by: Arun Kumar Neelakantam 
Signed-off-by: Deepak Kumar Singh 
Signed-off-by: Bjorn Andersson 
Link: https://lore.kernel.org/r/1596086296-28529-4-git-send-email-deesin@codeaurora.org
Stable-dep-of: 06c59d97f63c ("rpmsg: glink: use only lower 16-bits of param2 for CMD_OPEN name length")
Signed-off-by: Sasha Levin

rpmsg: virtio: Free driver_override when rpmsg_remove()

2024-02-23T07:24:48+00:00

commit d5362c37e1f8a40096452fc201c30e705750e687 upstream.

Free driver_override when rpmsg_remove(), otherwise
the following memory leak will occur:

unreferenced object 0xffff0000d55d7080 (size 128):
  comm "kworker/u8:2", pid 56, jiffies 4294893188 (age 214.272s)
  hex dump (first 32 bytes):
    72 70 6d 73 67 5f 6e 73 00 00 00 00 00 00 00 00  rpmsg_ns........
    00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
  backtrace:
    [<000000009c94c9c1>] __kmem_cache_alloc_node+0x1f8/0x320
    [<000000002300d89b>] __kmalloc_node_track_caller+0x44/0x70
    [<00000000228a60c3>] kstrndup+0x4c/0x90
    [<0000000077158695>] driver_set_override+0xd0/0x164
    [<000000003e9c4ea5>] rpmsg_register_device_override+0x98/0x170
    [<000000001c0c89a8>] rpmsg_ns_register_device+0x24/0x30
    [<000000008bbf8fa2>] rpmsg_probe+0x2e0/0x3ec
    [<00000000e65a68df>] virtio_dev_probe+0x1c0/0x280
    [<00000000443331cc>] really_probe+0xbc/0x2dc
    [<00000000391064b1>] __driver_probe_device+0x78/0xe0
    [<00000000a41c9a5b>] driver_probe_device+0xd8/0x160
    [<000000009c3bd5df>] __device_attach_driver+0xb8/0x140
    [<0000000043cd7614>] bus_for_each_drv+0x7c/0xd4
    [<000000003b929a36>] __device_attach+0x9c/0x19c
    [<00000000a94e0ba8>] device_initial_probe+0x14/0x20
    [<000000003c999637>] bus_probe_device+0xa0/0xac

Signed-off-by: Xiaolei Wang 
Fixes: b0b03b811963 ("rpmsg: Release rpmsg devices in backends")
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20231215020049.78750-1-xiaolei.wang@windriver.com
Signed-off-by: Mathieu Poirier 
Signed-off-by: Greg Kroah-Hartman

rpmsg: Fix possible refcount leak in rpmsg_register_device_override()

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

commit d7bd416d35121c95fe47330e09a5c04adbc5f928 upstream.

rpmsg_register_device_override need to call put_device to free vch when
driver_set_override fails.

Fix this by adding a put_device() to the error path.

Fixes: bb17d110cbf2 ("rpmsg: Fix calling device_lock() on non-initialized device")
Reviewed-by: Krzysztof Kozlowski 
Signed-off-by: Hangyu Hua 
Link: https://lore.kernel.org/r/20220624024120.11576-1-hbh25y@gmail.com
Signed-off-by: Mathieu Poirier 
Signed-off-by: Lee Jones 
Signed-off-by: Greg Kroah-Hartman

rpmsg: glink: Release driver_override

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

commit fb80ef67e8ff6a00d3faad4cb348dafdb8eccfd8 upstream.

Upon termination of the rpmsg_device, driver_override needs to be freed
to avoid leaking the potentially assigned string.

Fixes: 42cd402b8fd4 ("rpmsg: Fix kfree() of static memory on setting driver_override")
Fixes: 39e47767ec9b ("rpmsg: Add driver_override device attribute for rpmsg_device")
Reviewed-by: Chris Lew 
Signed-off-by: Bjorn Andersson 
Signed-off-by: Bjorn Andersson 
Link: https://lore.kernel.org/r/20230109223931.1706429-1-quic_bjorande@quicinc.com
Signed-off-by: Lee Jones 
Signed-off-by: Greg Kroah-Hartman

rpmsg: Fix calling device_lock() on non-initialized device

2023-11-08T10:23:39+00:00

commit bb17d110cbf270d5247a6e261c5ad50e362d1675 upstream.

driver_set_override() helper uses device_lock() so it should not be
called before rpmsg_register_device() (which calls device_register()).
Effect can be seen with CONFIG_DEBUG_MUTEXES:

  DEBUG_LOCKS_WARN_ON(lock->magic != lock)
  WARNING: CPU: 3 PID: 57 at kernel/locking/mutex.c:582 __mutex_lock+0x1ec/0x430
  ...
  Call trace:
   __mutex_lock+0x1ec/0x430
   mutex_lock_nested+0x44/0x50
   driver_set_override+0x124/0x150
   qcom_glink_native_probe+0x30c/0x3b0
   glink_rpm_probe+0x274/0x350
   platform_probe+0x6c/0xe0
   really_probe+0x17c/0x3d0
   __driver_probe_device+0x114/0x190
   driver_probe_device+0x3c/0xf0
   ...

Refactor the rpmsg_register_device() function to use two-step device
registering (initialization + add) and call driver_set_override() in
proper moment.

This moves the code around, so while at it also NULL-ify the
rpdev->driver_override in error path to be sure it won't be kfree()
second time.

Fixes: 42cd402b8fd4 ("rpmsg: Fix kfree() of static memory on setting driver_override")
Reported-by: Marek Szyprowski 
Signed-off-by: Krzysztof Kozlowski 
Tested-by: Marek Szyprowski 
Link: https://lore.kernel.org/r/20220429195946.1061725-2-krzysztof.kozlowski@linaro.org
Signed-off-by: Greg Kroah-Hartman 
Signed-off-by: Lee Jones 
Signed-off-by: Greg Kroah-Hartman