This patch (as1536) fixes a bug in the USB serial core.  Unloading and
reloading a serial driver while a serial device is plugged in causes
errors because of the code in usb_serial_disconnect() that tries to
make sure the port_remove method is called.  With the new order of
driver registration introduced in the 3.4 kernel, this is definitely
not the right thing to do (if indeed it ever was).

The patch removes that whole section code, along with the mechanism
for keeping track of each port's registration state, which is no
longer needed.  The driver core can handle all that stuff for us.

Note: This has been tested only with one or two USB serial drivers.
In theory, other drivers might still run into trouble.  But if they
do, it will be the fault of the drivers, not of this patch -- that is,
the drivers will need to be fixed.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Reported-and-tested-by: Johan Hovold <jhovold@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Now that module_driver() can handle varargs, use it instead of rolling
our own version.

Cc: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Now that Alan Stern has cleaned up the usb serial driver registration,
we have the ability to create a module_usb_serial_driver macro to make
things a bit simpler, like the other *_driver macros created.

But, as we need two functions here, we can't reuse the existing
module_driver() macro, so we need to roll our own.

Here's a patch implementing module_usb_serial_driver() and it converts
the pl2303 driver to use it, showing a nice cleanup.

Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

No one uses them anymore, they should be using the safer
usb_serial_register_drivers() and usb_serial_deregister_drivers()
functions instead.

Thanks to Alan Stern for writing these functions and porting all
in-kernel users to them.

Cc: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

This patch (as1522) adds two new routines to the usb-serial core, for
registering and unregistering serial drivers.  Instead of registering
the usb_driver and usb_serial_drivers separately, with error checking
for each one, the drivers can all be registered and unregistered by a
single function call.  This reduces duplicated code.

More importantly, the new core routines change the order in which the
drivers are registered.  Currently the usb-serial drivers are all
registered first and the usb_driver is done last, which leaves a
window for problems.  A udev script may quickly add a new dynamic-ID
for a usb-serial driver, causing the corresponding usb_driver to be
probed.  If the usb_driver hasn't been registered yet then an oops
will occur.

The new routine prevents such problems by registering the usb_driver
first.  To insure that it gets probed properly for already-attached
serial devices, we call driver_attach() after all the usb-serial
drivers have been registered.

Along with adding the new routines, the patch modifies the "generic"
serial driver to use them.  Further patches will similarly modify all
the other in-tree USB serial drivers.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Add macro which prints an error message only once if port is used a
console.

Reporting errors in a write path when port is used as a console could
otherwise result in an infinite loop.

Signed-off-by: Johan Hovold <jhovold@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Add support for multiple read urbs to generic read implementation.

Use a static array of two read urbs for now which is enough to get a
50% throughput increase in one test setup.

Signed-off-by: Johan Hovold <jhovold@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

Remove no longer used write_urb_busy field from struct usb_serial_port.

Signed-off-by: Johan Hovold <jhovold@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

* 'tty-next' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/tty-2.6: (76 commits)
  pch_uart: reference clock on CM-iTC
  pch_phub: add new device ML7213
  n_gsm: fix UIH control byte : P bit should be 0
  n_gsm: add a documentation
  serial: msm_serial_hs: Add MSM high speed UART driver
  tty_audit: fix tty_audit_add_data live lock on audit disabled
  tty: move cd1865.h to drivers/staging/tty/
  Staging: tty: fix build with epca.c driver
  pcmcia: synclink_cs: fix prototype for mgslpc_ioctl()
  Staging: generic_serial: fix double locking bug
  nozomi: don't use flush_scheduled_work()
  tty/serial: Relax the device_type restriction from of_serial
  MAINTAINERS: Update HVC file patterns
  tty: phase out of ioctl file pointer for tty3270 as well
  tty: forgot to remove ipwireless from drivers/char/pcmcia/Makefile
  pch_uart: Fix DMA channel miss-setting issue.
  pch_uart: fix exclusive access issue
  pch_uart: fix auto flow control miss-setting issue
  pch_uart: fix uart clock setting issue
  pch_uart : Use dev_xxx not pr_xxx
  ...

Fix up trivial conflicts in drivers/misc/pch_phub.c (same patch applied
twice, then changes to the same area in one branch)

When a driver doesn't know how much data a device is going to send,
the buffer size should be at least as big as the endpoint's maxpacket
value.  The serial drivers don't follow this rule; many of them
request only 256-byte bulk-in buffers.  As a result, they suffer
overflow errors if a high-speed device wants to send a lot of data,
because high-speed bulk endpoints are required to have a maxpacket
size of 512.

This patch (as1450) fixes the problem by using the driver's
bulk_in_size value as a minimum, always allocating buffers no smaller
than the endpoint's maxpacket size.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Tested-by: Flynn Marquardt <flynn@flynnux.de>
CC: <stable@kernel.org> [after .39-rc1 is out]
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>