Page size for transaction descriptors for CI13XXX has nothing
common with page size from MM. Using platform and configuration
specific PAGE_SIZE is wrong.
Signed-off-by: NArtem Leonenko <tikkeri@gmail.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

CI13xxx UDC driver doesn't call complete() callback for requests
with flag no_interrupt set. Thus gadget drivers (like g_ether) are
never notifed about successfully (or not) transmitted requests. As
a result in case of g_ether and queued request with no_interrupt=1
fields g_ether is never notifed about sent packets and TX stalls.

Solution: treat no_interrupt flag like all other UDC drivers do and
call complete() callback for all requests.
Signed-off-by: NArtem Leonenko <tikkeri@gmail.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

There is one nasty scenario when CI13xxx driver fails:
       a) two or more rx requests are queued (g_ether does that)
       b) rx request completed, interrupt fires and ci13xxx dequeues rq
       c) request complete() callback gets called and in turn it calls ep_queue()
               c1) in ep_queue() request gets added to the TAIL of the rx queue list
       d) ep gets primed with rq from (b)
       e) interrupt fires
       f) request gets popped from queue head for hw dequeue
       G) requets from queue head wasn't enqueued
               g1)  isr_tr_complete_low() doesn't
               enqueue more requests and it doesn't prime EP,
               rx traffic stalls

Solution:
       a) enque queued requests ASAP, i.e. before calling complete() callback.
       b) don't HW enqueue and prime endpoint with recently added request and
       use the oldest request in the queue.

Fixed issues:
       a) ep_queue() may return an error code despite request was successfully
       added to the queue (if _hardware_enqueue() fails)
       b) Added requests are always processed in LIFO order, even if they are
       added in complete() callback
       c) Finally more than two and more queued requests are processed consistently,
       even if they were added in complete() callback

The fix was successfully tested on MIPS based SoC with 4KEc CPU core and
CI13612 USB core. Board successfully boots with NFS root using g_ether
on ci13xxx udc.
Signed-off-by: NArtem Leonenko <tikkeri@gmail.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Built-in gadget drivers have NULL-ifed unbind() function. Checking
whether unbind() is NULL will never let any compiled into kernel
driver attach.
Signed-off-by: NArtem Leonenko <tikkeri@gmail.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This patches makes possible to use composite framework and f_ncm
NCM function driver to build a standalone NCM gadget device.
Signed-off-by: NYauheni Kaliuta <yauheni.kaliuta@nokia.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Initial submittion of NCM link function driver.

The driver's logic is based on f_ecm driver and does not
use most of the NCM advantages like frame grouping and alignment.
Signed-off-by: NYauheni Kaliuta <yauheni.kaliuta@nokia.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

NCM is a Network Control Model, subclass of USB CDC class,
specification is available at http://www.usb.org/developers/devclass_docs

This patch makes possible for u_ether to use multiply of wMaxPacketSize
predefined size transfers without ZLP (Zero Length Packet), required
by NCM spec.
Signed-off-by: NYauheni Kaliuta <yauheni.kaliuta@nokia.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This patch fixes an issue where the driver does not handle out data in
setup transactions.

The per endpoint cached status register is cleared in the
pch_udc_svc_control_out function. When there is out data available the
function pch_udc_svc_data_out is called which tries to pick it up the
status, which now is cleared to 0. When the status is 0, the function
doesn't start reading the data from the FIFO.

There is a second bug in all this, pch_udc_svc_data_out takes the
endpoint number (0 for EP0), while pch_udc_svc_control_out passes the
endpoint index (1 for EP0). Effectively pch_udc_svc_data_out picks up
the wrong internal ep structure.

This patch makes sure to put back the cached status and pass the
endpoint number rather than index when calling pch_udc_svc_data_out.
Signed-off-by: NRichard Röjfors <richard.rojfors@pelagicore.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Building pch_udc in linux-next fails, this patch fixes the a compile
error:

drivers/usb/gadget/pch_udc.c: In function ‘usb_gadget_register_driver’:
drivers/usb/gadget/pch_udc.c:2645: error: ‘struct usb_gadget_driver’ has no member named ‘bind’
drivers/usb/gadget/pch_udc.c:2664: error: ‘struct usb_gadget_driver’ has no member named ‘bind’

And a couple of compiler warnings and checkpatch warnings.
Signed-off-by: NRichard Röjfors <richard.rojfors@pelagicore.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

OTG driver takes care of putting hardware in low power mode.  Hence
not registered for any runtime PM callbacks.
Signed-off-by: NPavankumar Kondeti <pkondeti@codeaurora.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

The actual suspend/resume work is delegated to bus glue driver, which
is responsible for putting hardware in low power mode.
Signed-off-by: NPavankumar Kondeti <pkondeti@codeaurora.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

MSM SoC has chipidea USB controller. So use ci13xxx_udc core.
This driver depends on transceiver driver for clock control,
PHY initialization, VBUS detection.  Register for notify_event
callback to perform MSM specific quirks after controller is reset
and stopped.
Signed-off-by: NPavankumar Kondeti <pkondeti@codeaurora.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Introduces ci13xxx_udc_driver struct for bus glue drivers to hint
ci13xxx_udc core about their special requirements.  The flags include
avoiding hardware register access when controller is not in peripheral
mode, enabling pull-up upon VBUS, disabling streaming mode and dependency
on transceiver driver.

Initialize gadget_ops in udc_probe so that transceiver can notify VBUS
presence even when no gadget driver is bounded.

A notify_event callback is embedded in the same struct. This patch implements
two events called CONTROLLER_RESET_EVENT and CONTROLLER_STOPPED_EVENT to
notify the bus glue driver after resetting and stopping the controller for
performing SoC specific quirks.
Signed-off-by: NPavankumar Kondeti <pkondeti@codeaurora.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

dma_alloc_coherent() which is internally called by dma_pool_alloc()
flags a warning if device's coherent DMA mask.  Hence initialize
gadget device's coherent DMA mask to it's parent mask.
Signed-off-by: NPavankumar Kondeti <pkondeti@codeaurora.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

dma_pool_alloc() require sleeping context when called with GFP_KERNEL
argument.  Hence release the spin lock before calling dma_pool_alloc().

usb_ep_alloc_request can also be called with non-atomic GFP flags.  Hence
get rid off spin lock while allocation request memory.

Use GFP_ATOMIC flag for allocating request for ep0 in interrupt handler.
Signed-off-by: NPavankumar Kondeti <pkondeti@codeaurora.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Move PCI bus code from ci13xxx_udc to a new file ci13xxx_pci.  SoC's
which has MIPS USB core can include the ci13xxx_udc and keep bus glue
code in their respective gadget controller drivers.
Signed-off-by: NPavankumar Kondeti <pkondeti@codeaurora.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

IS_ERR() already implies unlikely(), so it can be omitted here.
Signed-off-by: NTobias Klauser <tklauser@distanz.ch>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Commit 06c3859f (usb: gadget/imx-udc: remove usage of deprecated symbol
USBD_INT0) was a bit precipitant because the name used instead didn't
match the usual naming scheme for irqs on arm/imx.  I renamed the irq to
the right name in e0830004 (ARM: imx: dynamically allocate imx_udc
device) when 06c3859f didn't hit Linus' tree, so I missed to add a
compat #define.

This patch allows compiling imx_udc.c with and without e0830004.
Signed-off-by: NUwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

change all ocurrences of musb_hdrc to musb-hdrc.

We will call glue layer drivers musb-<glue layer>,
so in order to keep things somewhat standard, let's
change the underscore into a dash.
Signed-off-by: NFelipe Balbi <balbi@ti.com>

If g_audio fails to open the sound control device, it crashes at removal:

Insertion:
[ 4143.836536] g_audio gadget: unable to open sound control device file: /dev/snd/controlC0
[ 4143.836543] g_audio gadget: we need at least one control device
[ 4143.836551] g_audio gadget: Linux USB Audio Gadget, version: Dec 18, 2008
[ 4143.836558] g_audio gadget: g_audio ready

Removal:
[ 4146.802643] BUG: unable to handle kernel paging request at 00023018
[ 4146.802655] IP: [<c10af9f5>] filp_close+0xa/0x5b
[ 4146.802674] *pdpt = 0000000015426001 *pde = 0000000000000000
[ 4146.802684] Oops: 0000 [#1] PREEMPT SMP
[ 4146.802692] last sysfs file: /sys/power/state
[ 4146.802701] Modules linked in: g_audio(-) ioh_udc fuse asix usbnet [last unloaded: g_audio]
[ 4146.802719]
[ 4146.802728] Pid: 1394, comm: rmmod Not tainted 2.6.33.5-26.1-ivi #1 To be filled by O.E.M./To be filled by O.E.M.
[ 4146.802738] EIP: 0060:[<c10af9f5>] EFLAGS: 00010206 CPU: 0
[ 4146.802746] EIP is at filp_close+0xa/0x5b
[ 4146.802753] EAX: 00023000 EBX: 00023000 ECX: 00000046 EDX: df842680
[ 4146.802760] ESI: e071cd4c EDI: df842680 EBP: ddbbbef0 ESP: ddbbbee4
[ 4146.802768]  DS: 007b ES: 007b FS: 00d8 GS: 0033 SS: 0068
[ 4146.802776] Process rmmod (pid: 1394, ti=ddbba000 task=dd95a4f0 task.ti=ddbba000)
[ 4146.802782] Stack:
[ 4146.802787]  d540c280 e071cd4c df2bc000 ddbbbefc e071b82c df11e440 ddbbbf04 e071c622
[ 4146.802804] <0> ddbbbf28 e071c47f 00000008 e071cd74 df11e464 df2bc01c df2bc000 e071ce68
[ 4146.802822] <0> 00000880 ddbbbf38 e07fd1b8 e071cef0 00000000 ddbbbf40 e071b9f4 ddbbbf48
[ 4146.802842] Call Trace:
[ 4146.802857]  [<e071b82c>] ? gaudio_cleanup+0x87/0xe0 [g_audio]
[ 4146.802869]  [<e071c622>] ? audio_unbind+0x8/0xc [g_audio]
[ 4146.802881]  [<e071c47f>] ? composite_unbind+0x8d/0xcb [g_audio]
[ 4146.802895]  [<e07fd1b8>] ? usb_gadget_unregister_driver+0x7b/0xc0 [ioh_udc]
[ 4146.802908]  [<e071b9f4>] ? usb_composite_unregister+0x15/0x17 [g_audio]
[ 4146.802920]  [<e071c633>] ? cleanup+0xd/0xf [g_audio]
[ 4146.802932]  [<c105a938>] ? sys_delete_module+0x185/0x1dd
[ 4146.802944]  [<c101c3ea>] ? do_page_fault+0x248/0x276
[ 4146.802956]  [<c10027d0>] ? sysenter_do_call+0x12/0x26
[ 4146.802962] Code: 12 5f 3a 00 8b 43 04 8b 40 0c 0f b3 30 3b 73 44 73 03 89 73 44 89 f8 e8 f1 61 3a 00 5b 5e 5f 5d c3 55 89 e5 57 89 d7 56 53 89 c3 <8b> 40 18 85 c0 75 0f 68 32 15 5e c1 31 f6 e8 52 39 3a 00 5a eb
[ 4146.803058] EIP: [<c10af9f5>] filp_close+0xa/0x5b SS:ESP 0068:ddbbbee4
[ 4146.803071] CR2: 0000000000023018
[ 4146.803112] ---[ end trace 0989a7e023da0434 ]---

This patch makes sure not to assign the_card if gaudio_open_snd_dev fails,
since the parent function will deallocate the card.

Also make sure all filp's in gaudio_open_snd_dev is assigned NULL upon error
and gaudio_close_snd_dev only cleanups when the filp's are non-NULL.
Signed-off-by: NRichard Röjfors <richard.rojfors@pelagicore.com>
Cc: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This patch add USB client support Marvell PXA9xx/PXA168 chips. The USB
controller in PXA9xx/PXA168 is a High-Speed OTG controller. The available
endpoints is different between PXA9xx and PXA168.

NOTE:
It is the first version of Marvell PXA9xx/PXA168 USB controller driver.
The support for OTG mode will be added in later patch.
PXA9xx and PXA168 has integrated UTMI PHY in the chips. The initialization
for the PHY is a little different between PXA9xx and PXA168.
Signed-off-by: NChao Xie <chao.xie@marvell.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This patch adds test mode support for Langwell gadget driver.
Signed-off-by: NHenry Yuan <hang.yuan@intel.com>
Signed-off-by: NAndy Luo <yifei.luo@intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This patch modifies the composite gadget to set vbus_draw current limitation
during suspend state. This current limitation in suspend state shouldn't be
more than 2.5mA
Signed-off-by: NHao Wu <hao.wu@intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This commit fixes warning in f_fs.c introduced by "usb:
gadget: f_fs: remove custom printk() wrappers":

In file included from drivers/usb/gadget/g_ffs.c:64:
drivers/usb/gadget/f_fs.c:30:1: warning: "pr_fmt" redefined
Signed-off-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

The big kernel lock has been removed from all these files at some point,
leaving only the #include.

Remove this too as a cleanup.
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

This patch adds the USB device driver of EG20T(Topcliff) PCH.

EG20T PCH is the platform controller hub that is going to be used in
Intel's upcoming general embedded platform. All IO peripherals in
EG20T PCH are actually devices sitting on AMBA bus.
EG20T PCH has USB device I/F. Using this I/F, it is able to access system
devices connected to USB device.
Signed-off-by: NToshiharu Okada <toshiharu-linux@dsn.okisemi.com>
Acked-by: NMichał Nazarewicz <m.nazarewicz@samsung.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This commit removes custom printk() wrappers from the f_fs.c
file.  They served little purpose above what pr_*() family of
macros provides.  Only FVDBG() has been left but renamed to
pr_vdebug() to match other uses.
Signed-off-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This commit changes FunctionFS as to make it more compliant
with coding style as well as fixes several typos.
Signed-off-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Signed-off-by: NJoe Perches <joe@perches.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

compile fix for bug introduced by 969affff)
Signed-off-by: NJosh Wu <josh.wu@atmel.com>
Cc: Jiri Kosina <jkosina@suse.cz>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

OK, the USB gadget serial driver actually has a couple of problems.  On
gs_open(), it always allocates and queues an additional QUEUE_SIZE (16)
worth of requests, so with a loop like this:

    i=1 ; while echo $i > /dev/ttyGS0 ; do let i++ ; done

eventually we run into OOM (Out of Memory).

Technically, it is not a leak as everything gets freed up when the USB
connection is broken, but not on gs_close().

With a USB device/gadget controller driver that has limited resources
(e.g., Marvell has a this MAX_XDS_FOR_TR_CALLS of 64 for transmit and
receive), so even after 4

    stty -F /dev/ttyGS0

we cannot transmit anymore.  We can still receive (not necessarily
reliably) as now we have 16 * 4 = 64 descriptors/buffers ready, but the
device is otherwise not usable.
Signed-off-by: NJim Sung <jsung@syncadence.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

The commit below cleaned up error handling, in part by introducing a
registered flag bit.  This however was not added to the device
structure leding to build failures:

  commit 319feaab
  Author: Dan Carpenter <error27@gmail.com>
  Date:   Tue Oct 5 18:55:34 2010 +0200

    usb: gadget: goku_udc: Fix error path

Add the missing registered flag bit.
Signed-off-by: NAndy Whitcroft <apw@canonical.com>
Acked-by: NDan Carpenter <error27@gmail.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Commit 126512e3 added support for FSL's USB
controller on powerpc. In this commit the Open Firmware code was selected
and compiled unconditionally.

This breaks on ARM systems from FSL which use the same driver (.i.e. the i.MX
series), because ARM don't have OF support (yet). This patch fixes the problem
by only selecting the OF code on systems with Open Firmware support.
Reported-by: NUwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: NMarc Kleine-Budde <mkl@pengutronix.de>
Compile-Tested-by: NUwe Kleine-König <u.kleine-koenig@pengutronix.de>
Acked-by: NGrant Likely <grant.likely@secretlab.ca>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Take handling of the control requests out from dummy_timer to a different
function.
Signed-off-by: NTatyana Brokhman <tlinder@codeaurora.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This commit is purely style clean ups.
Signed-off-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This commit removes call to the complete() function done in
fsg_unbind() which was never needed there but was a leftover
form file_storage.c.
Signed-off-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This commit drops START_TRANSFER_OR() and START_TRANSFER()
macros with a pair of nice inline functions which are actually
more readable and easier to use.
Signed-off-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This commit removes an "OR" macro defined in Mass Storage
Function in favour of a two argument version of "?:" operator
(which is a GCC extension).
Signed-off-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This commit fixes an issue with error recovery after
device_register() fails in Mass Storage Function.  The device
needs to be put to avoid resource leakage.
Signed-off-by: NRahul Ruikar <rahul.ruikar@gmail.com>
[mina86@mina86.com: updated commit message]
Signed-off-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This commit fixes some issues with File-backed Storage Gadget
error recovery when registering LUN's devices.

First of all, when device_register() fails the device still
needs to be put.  However, because lun_release() decreases
fsg->ref reference counter the counter must be incremented
beforehand.

Second of all, after any of the device_create_file()s fails,
device_unregister() is called which in turn (indirectly) calls
lun_release() which decrements fsg->ref.  So, again, the
reference counter must be incremented beforehand.

Lastly, if the first or the second device_create_file()
succeeds, the files are never removed.  To fix it,
device_remove_file() needs to be called.  This is done by
simply marking LUN as registered prior to creating files so
that fsg_unbind() can handle removing files.
Signed-off-by: NMichal Nazarewicz <mina86@mina86.com>
Reported-by: NRahul Ruikar <rahul.ruikar@gmail.com>
Acked-by: NAlan Stern <stern@rowland.harvard.edu>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>