kernel/drivers/usb/serial/mos7840.c, branch linux-6.9.y

USB: serial: mos7840: fix crash on resume

2024-07-18T11:22:46Z

commit c15a688e49987385baa8804bf65d570e362f8576 upstream. Since commit c49cfa917025 ("USB: serial: use generic method if no alternative is provided in usb serial layer"), USB serial core calls the generic resume implementation when the driver has not provided one. This can trigger a crash on resume with mos7840 since support for multiple read URBs was added back in 2011. Specifically, both port read URBs are now submitted on resume for open ports, but the context pointer of the second URB is left set to the core rather than mos7840 port structure. Fix this by implementing dedicated suspend and resume functions for mos7840. Tested with Delock 87414 USB 2.0 to 4x serial adapter. Signed-off-by: Dmitry Smirnov [ johan: analyse crash and rewrite commit message; set busy flag on resume; drop bulk-in check; drop unnecessary usb_kill_urb() ] Fixes: d83b405383c9 ("USB: serial: add support for multiple read urbs") Cc: stable@vger.kernel.org # 3.3 Signed-off-by: Johan Hovold Signed-off-by: Greg Kroah-Hartman

USB: serial: return errors from break handling

2023-06-07T15:00:23Z

Start propagating errors to user space when setting the break state fails. This will be used by follow-on changes to also report when a driver or device does not support break control. Tested-by: Corey Minyard Signed-off-by: Johan Hovold

treewide: Convert del_timer() to timer_shutdown()

2022-12-25T21:38:09Z

Due to several bugs caused by timers being re-armed after they are shutdown and just before they are freed, a new state of timers was added called "shutdown". After a timer is set to this state, then it can no longer be re-armed. The following script was run to find all the trivial locations where del_timer() or del_timer_sync() is called in the same function that the object holding the timer is freed. It also ignores any locations where the timer->function is modified between the del_timer*() and the free(), as that is not considered a "trivial" case. This was created by using a coccinelle script and the following commands: $ cat timer.cocci @@ expression ptr, slab; identifier timer, rfield; @@ ( - del_timer(&ptr->timer); + timer_shutdown(&ptr->timer); | - del_timer_sync(&ptr->timer); + timer_shutdown_sync(&ptr->timer); ) ... when strict when != ptr->timer ( kfree_rcu(ptr, rfield); | kmem_cache_free(slab, ptr); | kfree(ptr); ) $ spatch timer.cocci . > /tmp/t.patch $ patch -p1 < /tmp/t.patch Link: https://lore.kernel.org/lkml/20221123201306.823305113@linutronix.de/ Signed-off-by: Steven Rostedt (Google) Acked-by: Pavel Machek [ LED ] Acked-by: Kalle Valo [ wireless ] Acked-by: Paolo Abeni [ networking ] Signed-off-by: Linus Torvalds

usb: serial: Make ->set_termios() old ktermios const

2022-08-30T12:22:35Z

There should be no reason to adjust old ktermios which is going to get discarded anyway. Reviewed-by: Andy Shevchenko Signed-off-by: Ilpo Järvinen Link: https://lore.kernel.org/r/20220816115739.10928-8-ilpo.jarvinen@linux.intel.com Signed-off-by: Greg Kroah-Hartman

USB: serial: mos7840: remove duplicated 0xac24 device ID

2021-09-20T09:19:55Z

0xac24 device ID is already defined and used via BANDB_DEVICE_ID_USO9ML2_4. Remove the duplicate from the list. Fixes: 27f1281d5f72 ("USB: serial: Extra device/vendor ID for mos7840 driver") Signed-off-by: Krzysztof Kozlowski Cc: stable@vger.kernel.org Signed-off-by: Johan Hovold

USB: serial: mos7840: drop buffer-callback return-value comments

2021-05-21T13:46:20Z

The driver write_room and chars_in_buffer callbacks used to incorrectly return a negative errno in case they were called with a NULL port driver-data pointer or if some other always-true sanity checks failed. The bogus sanity checks were later removed by commit ce039bd4b21f ("USB: serial: mos7840: drop paranoid port checks") and 7b2faede671a ("USB: serial: mos7840: drop port driver data accessors") but the function-header comments were never updated to match. Drop the outdated return-value comments. Reviewed-by: Greg Kroah-Hartman Signed-off-by: Johan Hovold

USB: serial: make usb_serial_driver::chars_in_buffer return uint

2021-05-19T08:56:41Z

tty_operations::chars_in_buffer is being switched to return uint. Do the same for usb_serial_driver's chars_in_buffer. Signed-off-by: Jiri Slaby [ johan: amend commit message ] Signed-off-by: Johan Hovold

USB: serial: make usb_serial_driver::write_room return uint

2021-05-19T08:56:40Z

Line disciplines expect a positive value or zero returned from tty->ops->write_room (invoked by tty_write_room). Both of them are being updated to return an unsigned int. Switch also usb_serial_driver::write_room and all its users. Signed-off-by: Jiri Slaby [ johan: amend commit message, drop unrelated comment change ] Signed-off-by: Johan Hovold

USB: serial: stop reporting legacy UART types

2021-04-08T07:46:02Z

The TIOCGSERIAL ioctl can be used to set and retrieve the UART type for legacy UARTs, but some USB serial drivers have been reporting back random types in order to "make user-space happy". Some applications have historically expected TIOCGSERIAL to be implemented, but judging from the Debian sources, the port type not being PORT_UNKNOWN is only used to check for the existence of legacy serial ports (ttySn). Drivers like ftdi_sio have been using PORT_UNKNOWN for twenty years (and option for 10 years) without anyone complaining so let's stop reporting back anything else. In the unlikely event that this do cause problems, this should be fixed tree-wide anyway (e.g. for all USB serial drivers and also CDC-ACM). Reviewed-by: Greg Kroah-Hartman Signed-off-by: Johan Hovold

USB: serial: add generic support for TIOCSSERIAL

2021-04-08T07:46:02Z

TIOCSSERIAL is a horrid, underspecified, legacy interface which for most serial devices is only useful for setting the close_delay and closing_wait parameters. The closing_wait parameter determines how long to wait for the transfer buffers to drain during close and the default timeout of 30 seconds may not be sufficient at low line speeds. In other cases, when for example flow is stopped, the default timeout may instead be too long. Add generic support for TIOCSSERIAL and TIOCGSERIAL with handling of the three common parameters close_delay, closing_wait and line for the benefit of all USB serial drivers while still allowing drivers to implement further functionality through the existing callbacks. This currently includes a few drivers that report their base baud clock rate even if that is really only of interest when setting custom divisors through the deprecated ASYNC_SPD_CUST interface; an interface which only the FTDI driver actually implements. Some drivers have also been reporting back a fake UART type, something which should no longer be needed and will be dropped by a follow-on patch. Reviewed-by: Greg Kroah-Hartman Signed-off-by: Johan Hovold

kernel/drivers/usb/serial/mos7840.c, branch linux-6.9.y

USB: serial: mos7840: fix crash on resume

USB: serial: return errors from break handling

treewide: Convert del_timer*() to timer_shutdown*()

usb: serial: Make ->set_termios() old ktermios const

USB: serial: mos7840: remove duplicated 0xac24 device ID

USB: serial: mos7840: drop buffer-callback return-value comments

USB: serial: make usb_serial_driver::chars_in_buffer return uint

USB: serial: make usb_serial_driver::write_room return uint

USB: serial: stop reporting legacy UART types

USB: serial: add generic support for TIOCSSERIAL

treewide: Convert del_timer() to timer_shutdown()