A sufficiently-large number of USB serial devices causes a reference leak
when /proc/tty/drivers/usbserial is read.
Signed-Off-By: NMatthias Urlichs <smurf@smurf.noris.de>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

USB serial outside of the kernel tree can not build properly due to
usb-serial.h being buried down in the source tree.  This patch moves the
location of the file to include/linux/usb and fixes up all of the usb
serial drivers to handle the move properly.

Cc: Sergei Organov <osv@javad.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This fix addresses two issues:
- Unattached port structures were not freed
- My initial fix for crash when eventd runs a work in a freed port
  did not go far enough
Signed-off-by: NPete Zaitcev <zaitcev@redhat.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Signed-off-by: NJörn Engel <joern@wohnheim.fh-wedel.de>
Signed-off-by: NAdrian Bunk <bunk@stusta.de>

I've always found this flag confusing.  Now that devfs is no longer around, it
has been renamed, and the documentation for when this flag should be used has
been updated.

Also fixes all drivers that use this flag.
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Also fixes all drivers that set this field.
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Some usbserial functions returns -EINVAL if the port doesn't exist or if
it's not opened. However, the right error code for such situations is
-ENODEV.
Signed-off-by: NLuiz Fernando N. Capitulino <lcapitulino@mandriva.com.br>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

If either the driver's open() method or try_module_get() fails, we need to
set 'tty->driver_data' and 'port->tty' to NULL in serial_open(), otherwise
we'll get an OOPS in usb_device_disconnect() when the device is disconnected.
Signed-off-by: NFrank Gevaerts <frank.gevaerts@fks.be>
Acked-by: NLuiz Fernando N. Capitulino <lcapitulino@mandriva.com.br>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Wait for the scheduled work to finish before freeing memory, prevent oops.
http://bugzilla.kernel.org/show_bug.cgi?id=6596Signed-off-by: NPete Zaitcev <zaitcev@redhat.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

I'm going to throw schedule_work away, it's retarded. But for starters,
let's have it encapsulated.

Also, generic and whiteheat were both calling usb_serial_port_softint
and scheduled work. Only one was necessary.
Signed-off-by: NPete Zaitcev <zaitcev@redhat.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Prevent sending further output to a USB-serial console after the dongle is
disconnected, take care not to leak kref.
Signed-off-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Prevent ENODEV on a /dev/ttyUSBx, used as a USB-serial console.

From: Paul Fulghum <paulkf@microgate.com>
Signed-off-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

If serial_open() fails at the port assignment or mutex_lock_interruptible()
is interrupted, the 'serial' object will never be freed.

We should call kref_put() when those errors happens.
Signed-off-by: NLuiz Fernando N. Capitulino <lcapitulino@mandriva.com.br>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

If the device is disconnected while serial_open() is executing and
either try_module_get() or the device specific open function fails, the
kref_put() call in the 'bailout_kref_put' label will free the memory
pointed out by 'port'.

The subsequent dereferences in the 'bailout_kref_put' label will be
invalid.

The fix is just to assure kref_put() is called after any 'port' usage.
Signed-off-by: NLuiz Fernando N. Capitulino <lcapitulino@mandriva.com.br>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

The usbserial's port semaphore used to synchronize serial_open()
and serial_close() are strict mutexes, convert them to the mutex
implementation.
Signed-off-by: NLuiz Capitulino <lcapitulino@mandriva.com.br>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Signed-off-by: NEric Sesterhenn <snakebyte@gmx.de>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This fixes a use-after-free bug in the usb-serial core.  It is simple to
trigger this (open a usb-serial port, then yank the device out before
closing the port.)  Thanks to Stefan Seyfried <seife@suse.de> for
reporting this, and to the slab debugging code which enabled it to be
tracked down.
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Additionally, one comment was reformulated by Joe Perches <joe@perches.com>.
Signed-off-by: NAdrian Bunk <bunk@stusta.de>

There is a race-condition in usb-serial driver that can be triggered if
a processes does 'port->tty->driver_data = NULL' in serial_close() while
other processes is in kernel-space about to call serial_ioctl() on the
same port.

This happens because a process can open the device while there is
another one closing it.

The patch below fixes that by adding a semaphore to ensure that no
process will open the device while another process is closing it.

Note that we can't use spinlocks here, since serial_open() and
serial_close() can sleep.
Signed-off-by: NLuiz Capitulino <lcapitulino@mandriva.com.br>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Checks if 'port' is NULL before using it in all tty operations, this
can avoid NULL pointer dereferences.
Signed-off-by: NLuiz Capitulino <lcapitulino@mandriva.com.br>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

It is no longer needed, so let's remove it, saving a bit of memory.
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This lets drivers, like the usb-serial ones, disable the ability to add
ids from sysfs.

The usb-serial drivers are "odd" in that they are really usb-serial bus
drivers, not usb bus drivers, so the dynamic id logic will have to go
into the usb-serial bus core for those drivers to get that ability.
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

These numbers are pointless, as they have not been changed in _years_,
so we should just remove them to stop pretending there is an actual
"version number" for these drivers.

This should also help reduce confusion when people try to ask for
support of a specific driver version, as there has been no way to tell
what they are talking about.
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This fixes up a lot of problems in sysfs with some of the usb serial
drivers, they had incorrect driver names.  Also saves a tiny ammount
of memory.
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Create a new file just for these things, as they just get in the
way in the source files.
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Don't duplicate something that's already in struct driver.
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

I'm tired of trying to explain why a "device_type" is really a driver.
This better describes exactly what this structure is.
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

The following patch removes unneeded casts for the following (void *) pointers:
	- tty_struct->driver_data
	- void *private argument of usb_serial_port_softint()
Signed-off-by: NTobias Klauser <tklauser@nuerscht.ch>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

[PATCH] USB: add ability for usb-serial drivers to determine if their write urb is currently being used.

This removes a lot of racy and buggy code by trying to check the status of the urb.
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

At module load time, if a generic device is found, the tty information
for the device is not set up properly (as the tty structures aren't initialized
yet.)  This can cause big problems for things like udev.  This patch fixes this.

Thanks to Kay Sievers for the original patch for this problem.
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Initial git repository build. I'm not bothering with the full history,
even though we have it. We can create a separate "historical" git
archive of that later if we want to, and in the meantime it's about
3.2GB when imported into git - space that would just make the early
git days unnecessarily complicated, when we don't have a lot of good
infrastructure for it.

Let it rip!