Currently, the UVC function is activated when open on the corresponding
v4l2 device is called. On another open the activation of the function
fails since the deactivation counter in `usb_function_activate` equals
0. However the error is not returned to userspace since the open of the
v4l2 device is successful.
On a close the function is deactivated (since deactivation counter still
equals 0) and the video is disabled in `uvc_v4l2_release`, although the
UVC application potentially is streaming.
Move activation of UVC function to subscription on UVC_EVENT_SETUP
because there we can guarantee for a userspace application utilizing
UVC. Block subscription on UVC_EVENT_SETUP while another application
already is subscribed to it, indicated by `bool func_connected` in
`struct uvc_device`. Extend the `struct uvc_file_handle` with member
`bool is_uvc_app_handle` to tag it as the handle used by the userspace
UVC application.
With this a process is able to check capabilities of the v4l2 device
without deactivating the function for the actual UVC application.
Reviewed-By: Michael Tretter <m.tretter@pengutronix.de>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Thomas Haemmerle <thomas.haemmerle@wolfvision.net>
Signed-off-by: Michael Tretter <m.tretter@pengutronix.de>
Signed-off-by: Michael Grzeschik <m.grzeschik@pengutronix.de>
Acked-by: Felipe Balbi <balbi@kernel.org>
Link: https://lore.kernel.org/r/20211003201355.24081-1-m.grzeschik@pengutronix.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Now that Kees Cook has added a definition for HCS_N_PORTS_MAX in
commit 72dd184323 ("USB: EHCI: Add register array bounds to HCS
ports"), the code in ehci_hub_control() which sanitizes port index
values can be improved a little.
The idea behind this change is that it prevents a possible
out-of-bounds pointer computation, which the compiler might be able to
detect since the port_status[] array now has a fixed length rather
than a variable length.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Link: https://lore.kernel.org/r/20211002190217.GA537967@rowland.harvard.edu
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The commit 217504a055 ("usb: typec: ucsi: Work around PPM
losing change information") had solved this issue
previously, but in a really complex manner. The core issue
is that on some platforms the EC firmware does not interrupt
the driver on unplug event in some cases, mainly when the
cable is unplugged immediately after the plug-in.
From now on handling that problem by simply re-checking new
connections.
Signed-off-by: Heikki Krogerus <heikki.krogerus@linux.intel.com>
Link: https://lore.kernel.org/r/20210920142419.54493-8-heikki.krogerus@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
UCSI does not tell the driver explicitly when the firmware
(PPM in UCSI lingo) has actually detected the partner
alternate modes, there is no specific change event for that.
That's why they have to be checked with any notification
that informs that PD contract with that partner has been
achieved.
Previously the alternate modes were checked always when the
firmware (PPM) informed that something with the partner had
changed, but on some platforms the EC firmware does not
generate separate events for generic partner changes at all.
On those platforms the EC firmware notifies the driver only
about connections, or separately about the PD contract if it
was not achieved soon enough after the initial connection
event.
Signed-off-by: Heikki Krogerus <heikki.krogerus@linux.intel.com>
Link: https://lore.kernel.org/r/20210920142419.54493-6-heikki.krogerus@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The "poll worker" that is introduced here is first used for
checking partner alternate modes, but it can later be used
for any partner task that requires a separate job to be
scheduled to the connector specific workqueues.
The mechanism allows the partner device specific tasks to be
polling tasks and also delayed tasks if necessary.
By polling the partner alternate modes with this mechanism
the long command completion timeout value can be reduced
back to normal. The long command completion timeout was only
used to work around a problem on some platforms where the EC
firmware (PPM) didn't return BUSY even when it should with
the alt mode commands.
Signed-off-by: Heikki Krogerus <heikki.krogerus@linux.intel.com>
Link: https://lore.kernel.org/r/20210920142419.54493-4-heikki.krogerus@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The USB2.0 spec chapter 11.24.2.13 says that the USB port which is going
under test needs to be put in suspend state before sending the test
command. Many hubs, don't enforce this precondition and they work fine
without this step. But there are some "special" hubs, which requires to
disable the port power before sending the test command.
Because the USB spec mention that the port should be suspended, also
do this step before sending the test command. This could rise the
problem with other hubs which are not compliant with the spec and the
test command will not work if the port is suspend. If such hubs are
found, a similar workaround like the disable part could be implemented
to skip the suspend port command.
Signed-off-by: Razvan Heghedus <heghedus.razvan@gmail.com>
Link: https://lore.kernel.org/r/20210915121615.3790-1-heghedus.razvan@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Johan writes:
USB-serial fixes for 5.15-rc3
Here's a fix for a regression affecting some CP2102 devices and a host
of new device ids.
Included are also a couple of cleanups of duplicate device ids, which
are also tagged for stable to keep the tables in sync, and a trivial
patch to help debugging cp210x issues.
All have been in linux-next with no reported issues. Note however that
the last last two device-id commits were rebased to fix up a lore link
in a commit message (as the patch itself never made it to the list).
* tag 'usb-serial-5.15-rc3' of https://git.kernel.org/pub/scm/linux/kernel/git/johan/usb-serial:
USB: serial: option: add device id for Foxconn T99W265
USB: serial: cp210x: add ID for GW Instek GDM-834x Digital Multimeter
USB: serial: cp210x: add part-number debug printk
USB: serial: cp210x: fix dropped characters with CP2102
USB: serial: option: remove duplicate USB device ID
USB: serial: mos7840: remove duplicated 0xac24 device ID
USB: serial: option: add Telit LN920 compositions
Some CP2102 do not support event-insertion mode but return no error when
attempting to enable it.
This means that any event escape characters in the input stream will not
be escaped by the device and consequently regular data may be
interpreted as escape sequences and be removed from the stream by the
driver.
The reporter's device has batch number DCL00X etched into it and as
discovered by the SHA2017 Badge team, counterfeit devices with that
marking can be detected by sending malformed vendor requests. [1][2]
Tests confirm that the possibly counterfeit CP2102 returns a single byte
in response to a malformed two-byte part-number request, while an
original CP2102 returns two bytes. Assume that every CP2102 that behaves
this way also does not support event-insertion mode (e.g. cannot report
parity errors).
[1] https://mobile.twitter.com/sha2017badge/status/1167902087289532418
[2] https://hackaday.com/2017/08/14/hands-on-with-the-shacamp-2017-badge/#comment-3903376
Reported-by: Malte Di Donato <malte@neo-soft.org>
Tested-by: Malte Di Donato <malte@neo-soft.org>
Fixes: a7207e9835 ("USB: serial: cp210x: add support for line-status events")
Cc: stable@vger.kernel.org # 5.9
Link: https://lore.kernel.org/r/20210922113100.20888-1-johan@kernel.org
Signed-off-by: Johan Hovold <johan@kernel.org>
Drop the line-status conversion helper and do the conversion in place
instead.
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Johan Hovold <johan@kernel.org>
Now that the driver is using usb_control_msg_recv(), the line status
handling can be simplified further by reading directly into the status
variable and doing the endian conversion in place.
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Johan Hovold <johan@kernel.org>
Clean up the line-status handling by dropping redundant initialisations
and returning early on errors.
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Johan Hovold <johan@kernel.org>
ScanLogic SL11R-IDE with firmware older than 2.6c (the latest one) has
broken tag handling, preventing the device from working at all:
usb 1-1: new full-speed USB device number 2 using uhci_hcd
usb 1-1: New USB device found, idVendor=04ce, idProduct=0002, bcdDevice= 2.60
usb 1-1: New USB device strings: Mfr=1, Product=1, SerialNumber=0
usb 1-1: Product: USB Device
usb 1-1: Manufacturer: USB Device
usb-storage 1-1:1.0: USB Mass Storage device detected
scsi host2: usb-storage 1-1:1.0
usbcore: registered new interface driver usb-storage
usb 1-1: reset full-speed USB device number 2 using uhci_hcd
usb 1-1: reset full-speed USB device number 2 using uhci_hcd
usb 1-1: reset full-speed USB device number 2 using uhci_hcd
usb 1-1: reset full-speed USB device number 2 using uhci_hcd
Add US_FL_BULK_IGNORE_TAG to fix it. Also update my e-mail address.
2.6c is the only firmware that claims Linux compatibility.
The firmware can be upgraded using ezotgdbg utility:
https://github.com/asciilifeform/ezotgdbg
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Ondrej Zary <linux@zary.sk>
Cc: stable <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/20210913210106.12717-1-linux@zary.sk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
There is a race present where the DWC3 runtime resume runs in parallel
to the UDC unbind sequence. This will eventually lead to a possible
scenario where we are enabling the run/stop bit, without a valid
composition defined.
Thread#1 (handling UDC unbind):
usb_gadget_remove_driver()
-->usb_gadget_disconnect()
-->dwc3_gadget_pullup(0)
--> continue UDC unbind sequence
-->Thread#2 is running in parallel here
Thread#2 (handing next cable connect)
__dwc3_set_mode()
-->pm_runtime_get_sync()
-->dwc3_gadget_resume()
-->dwc->gadget_driver is NOT NULL yet
-->dwc3_gadget_run_stop(1)
--> _dwc3gadget_start()
...
Fix this by tracking the pullup disable routine, and avoiding resuming
of the DWC3 gadget. Once the UDC is re-binded, that will trigger the
pullup enable routine, which would handle enabling the DWC3 gadget.
Acked-by: Felipe Balbi <balbi@kernel.org>
Signed-off-by: Wesley Cheng <wcheng@codeaurora.org>
Link: https://lore.kernel.org/r/20210917021852.2037-1-wcheng@codeaurora.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
