This adds multi-frame support to the NCM NTB's for
the gadget driver. This allows multiple network
packets to be put inside a single USB NTB with a
maximum size of 16kB.

It has a time out of 300ms to ensure that smaller
number of packets still maintain a normal latency.

Also the .fp_index and .next_fp_index have been
changed to .ndp_index and .next_ndp_index to
match the latest CDC-NCM specification and
help with maintenance.

Results transmitting from gadget to host.

Before the change:

TCP_STREAM Throughput (10^6bits/sec): 22.72
UDP_STREAM Throughput (10^6bits/sec): 25.94

Latency:
netperf -H 192.168.1.101 -v2 -l 50 -t TCP_RR -- -r 16384,16384
Trans.   RoundTrip  Throughput
Rate     Latency    10^6bits/s
per sec  usec/Tran  Outbound

100.83   9918.116   13.215

After the change:

TCP_STREAM Throughput (10^6bits/sec): 124.26
UDP_STREAM Throughput (10^6bits/sec): 227.48

Latency:
netperf -H 192.168.1.101 -v2 -l 50 -t TCP_RR -- -r 16384,16384
Trans.   RoundTrip  Throughput
Rate     Latency    10^6bits/s
per sec  usec/Tran  Outbound

156.80   6377.730   20.552
Signed-off-by: NJim Baxter <jim_baxter@mentor.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

The NDP was ignoring the wNextNdpIndex in the NDP which
means that NTBs containing multiple NDPs would have missed
frames.
Signed-off-by: NJim Baxter <jim_baxter@mentor.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

Even if the skb is cloned, we still need a ZLP or USB will stall.
Signed-off-by: NNathan Sullivan <nathan.sullivan@ni.com>
Acked-by: NBrad Mouring <brad.mouring@ni.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

Remove 'start_new' variable from dwc3_endpoint_transfer_complete(),
since this variable has not been used.
Signed-off-by: NJingoo Han <jg1.han@samsung.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

Use SIMPLE_DEV_PM_OPS macro and remove DEV_PM_OPS macro, in order
to make the code simpler.
Acked-by: NRoger Quadros <rogerq@ti.com>
Signed-off-by: NJingoo Han <jg1.han@samsung.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

Use the module_usb_composite_driver() macro where applicable to
eliminate the module_init/module_exit boilerplate in USB gadget composite
drivers.
Signed-off-by: NTobias Klauser <tklauser@distanz.ch>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

Introduce the module_usb_composite_driver macro as a convenience macro
for USB gadget composite driver modules, similar to module_usb_driver.
It is intended to be used by drivers which init/exit section does
nothing but calling usb_composite_probe/usb_composite_unrregister. By
using this macro it is possible to eliminate a few lines of boilerplate
code per USB gadget composite driver.

Based on f3a6a4b6 ("USB: Add helper macro for usb_driver boilerplate")
which introduced the according macro for USB drivers.
Signed-off-by: NTobias Klauser <tklauser@distanz.ch>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

Add support for OS descriptors. The new format of descriptors is used,
because the "flags" field is required for extensions. os_count gives
the number of OSDesc[] elements.
The format of descriptors is given in include/uapi/linux/usb/functionfs.h.

For extended properties descriptor the usb_ext_prop_desc structure covers
only a part of a descriptor, because the wPropertyNameLength is unknown
up front.
Signed-off-by: NAndrzej Pietrasiewicz <andrzej.p@samsung.com>
Acked-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

Provide helper functions to get pointers to particular locations within
a buffer holding an extended properties descriptor.
Signed-off-by: NAndrzej Pietrasiewicz <andrzej.p@samsung.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

ffs_do_desc() handles one descriptor, while ffs_do_descs() handles a number
of descriptors. The tho names are so similar that it causes confusion.
Rename to reflect their purpose better.
Signed-off-by: NAndrzej Pietrasiewicz <andrzej.p@samsung.com>
Acked-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

It is no longer needed and keeping it will break 64-bit ARM builds.
Signed-off-by: NThierry Reding <treding@nvidia.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

Following the name change of the NOP transceiver driver in commit
4525beeb (usb: phy: rename usb_nop_xceiv to usb_phy_generic), the
transceiver driver was no longer operable under its old name.

Register the transceiver driver before calling usb_get_phy() to make
sure we are noticed by an error message if it is not available.
Signed-off-by: NApelete Seketeli <apelete@seketeli.net>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

Commit e56e69cc ("usb: gadget: net2280: Use pr_* function")
includes a editing mistake on one of the #ifdef.

This patch fixes it.
Reported-by: NPaul Bolle <pebolle@tiscali.nl>
Signed-off-by: NRicardo Ribalda Delgado <ricardo.ribalda@gmail.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

commit c63d2225 ("usb: gadget: pxa25x_udc: use devm_ functions")
introduced the use of devm_gpio_request in this driver, but did not
correctly include the header file declaring this function, which
causes a build failure.

This changes pxa25x_udc to include linux/gpio.h instead of asm/gpio.h
to fix this.
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Cc: Himangi Saraogi <himangi774@gmail.com>
Cc: Julia Lawall <julia.lawall@lip6.fr>
Cc: Felipe Balbi <balbi@ti.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

This patch fixes an issue that the renesas_usbhs driver in gadget mode
cannot work correctly even if I disabled DMAC of the driver when I used
the g_zero driver and the testusb tool.

When a usb cable is re-connected, the renesas_usbhs driver calls the
usbhsp_flags_init() (via usbhs_hotplug() --> usbhs_mod_call(start) -->
usbhsg_try_start() --> usbhs_pipe_init()). However, the driver doesn't
call the usbhsp_flags_init() when usbhsg_ep_disable() is called.
So, if a gadget driver calls usb_ep_enable() and usb_ep_disable() again
and again, the renesas_usbhs driver will output the following log:

  renesas_usbhs renesas_usbhs: can't get pipe (BULK)
  renesas_usbhs renesas_usbhs: wrong recip request
Acked-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NYoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

This patch fixes an issue that the driver cannot push a new data when
a pipe is re-enabled after the pipe is queued.
Acked-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NYoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

Use clk_prepare_enable/clk_disable_unprepare to make the driver
work properly with common clock framework.
Signed-off-by: NVasily Khoruzhick <anarsoul@gmail.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

The musb/cppi41 code installs a hrtimer to work around DMA completion
interrupts that have fired too early on AM335x hardware. This timer
is currently programmed to first fire 140 microseconds after the DMA
completion callback. According to the commit which introduced it
(a655f481, "usb: musb: musb_cppi41: handle pre-mature TX complete
interrupt"), that value is is considered a 'rule of thumb' that worked
well with the test case described in the commit log.

Test show, however, that for USB audio devices and much smaller packet
sizes, the timer has to fire earlier in order to correctly handle the audio
stream. The original test case had output transfer sizes of 1514 bytes, and
a delay of 140 microseconds. For audio devices with 24 bytes channel size, 3
microseconds seem to work well.

Hence, let's assume that the time it takes to clear the bit correlates with
the number of bytes transferred. The referenced commit log mentions such a
suspicion as well. Let the timer fire in cppi41_channel->total_len/10
microseconds to correctly handle both cases.

Also, shorten the interval in which the timer fires again in case of
a non-empty early_tx list.

With these changes in place, both FS and HS audio devices appear to work
well on AM335x hardware.
Signed-off-by: NDaniel Mack <zonque@gmail.com>
Reported-by: NSebastian Reimers <sebastian.reimers@googlemail.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

This reverts commit 1af54b7a.

The commit tried to address cases in which isochronous transfers are 'not
reliable', most probably in the tests conducted, polling for the
MUSB_TXCSR_TXPKTRDY bit in MUSB_TXCSR is done too late.

Hence, it installs a work struct which basically busy-polls for the bit in a
rather agressive way by rescheduling the work if the FIFO is not empty. With
USB audio devices, tests have shown that it takes approximately 100
iterations of the asynchronous worker until the FIFO signals completion,
which leads to 100% CPU loads when streaming audio.

The issue the patch tried to address can be handled differently, which is
what the next patch does.
Signed-off-by: NDaniel Mack <zonque@gmail.com>
Reported-by: NSebastian Reimers <sebastian.reimers@googlemail.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

Just a cosmetic cleanup with no functional change.
Signed-off-by: NDaniel Mack <zonque@gmail.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

Before accessing any of an endpoint's CSR registers, make sure the
correct endpoint is selected. Otherwise, data read from or written to
the registers is likely to affect the wrong endpoint as long as the
connected device has more than one endpoint.

This, of course, leads to all sorts of strange effects such as stream
starvation and driver internal state machine confusion due to spurious
interrupts.
Signed-off-by: NDaniel Mack <zonque@gmail.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

The musb/cppi41 glue layer is capable of handling transactions that span
over more than one USB packet by reloading the DMA descriptors
partially. An urb is considered completed when either its transfer
buffer has been filled entirely (actual_length ==
transfer_buffer_length) or if a packet in the stream has less bytes than
the endpoint's wMaxPacketSize.

Once one of the above conditions is met, musb_dma_completion() is called
from cppi41_trans_done(). However, the final decision whether or not to
return the urb to its owner is made by the core and its determination of
the variable 'done' in musb_host_rx(). This code has currently no way of
knowing what the size of the last packet was, and whether or not to
give back the urb due to a short read.

Fix this by introducing a new boolean flag in 'struct dma_channel', and
set it from musb_cppi41.c. If set, it will make the core do what the
DMA layer decided and complete the urb.
Signed-off-by: NDaniel Mack <zonque@gmail.com>
Acked-by: NGeorge Cherian <george.cherian@ti.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

The datasheet says that MUSB_TXCSR_FLUSHFIFO is only valid when
MUSB_TXCSR_TXPKTRDY is set as well.

With this patch applied, the warning in this function does no longer
kick in when an USB soundcard is unplugged while the stream is active.
Signed-off-by: NDaniel Mack <zonque@gmail.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

On AM33xx platforms, unplugging a device in the middle of an active
transfer leads to a drop of MUSB_DEVCTL_HM in MUSB_DEVCTL before the
system is informed about a disconnect. This consequently makes the musb
core call the gadget code to handle the interrupt request, which then
crashes the kernel because the relevant pointers haven't been set up
for gadget mode.

To fix this, use is_host_active() rather than (devctl & MUSB_DEVCTL_HM)
in musb_interrupt() and musb_dma_completion() to detect whether the
controller is in host or peripheral mode. This information is provided
by the driver logic and does not rely on register contents.
Signed-off-by: NDaniel Mack <zonque@gmail.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

Just a little cleanup that removes unnecessary casts.
Signed-off-by: NDaniel Mack <zonque@gmail.com>
Acked-by: NGeorge Cherian <george.cherian@ti.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

Syscall mount returns -ENODEV error if requested FS type
has not been found. Returning the same error from FFS mount
callback makes value returned to userspace misleading.

Other file systems returns -ENOENT if requested device
has not been found. Adjust FFS to this convention to make
error codes meaningfull.
Acked-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NKrzysztof Opasiak <k.opasiak@samsung.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

It has already been covered by udc core
Signed-off-by: NPeter Chen <peter.chen@freescale.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

It has already been covered by udc core, besides, we do not
need unbind at .udc_start
Signed-off-by: NPeter Chen <peter.chen@freescale.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

It has already been covered by udc core
Signed-off-by: NPeter Chen <peter.chen@freescale.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

It has already been covered by udc core
Signed-off-by: NPeter Chen <peter.chen@freescale.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

It has already been covered by udc core
Signed-off-by: NPeter Chen <peter.chen@freescale.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

This patch makes the phy reset clk and reset line optional as this clk
is not available on boards like IFC6410 with APQ8064.
Signed-off-by: NSrinivas Kandagatla <srinivas.kandagatla@linaro.org>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

Add clock prepare and unprepare as required by clock framework.
Signed-off-by: NRobert Jarzmik <robert.jarzmik@free.fr>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

Add missing information about license. Some people will probably want to
reuse this code in their projects released under variety of licenses. For this
reason this example is under Public Domain license to avoid GPL limitations.
Acked-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NRobert Baldyga <r.baldyga@samsung.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

Since commit [ac8dde11: “Add flags to descriptors block”] functionfs
supports a new descriptor format, so we update example application
to make it using recomended version of descriptors.
Acked-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NRobert Baldyga <r.baldyga@samsung.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

We wrap numeric values of fs_count and hs_count fields in htole32,
because they should be in little-endian format.
Acked-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NRobert Baldyga <r.baldyga@samsung.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

Use of quirks improve readability and will be easier to add new devices
to this driver.
Suggested-by: NAlan Stern <stern@rowland.harvard.edu>
Signed-off-by: NRicardo Ribalda Delgado <ricardo.ribalda@gmail.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

Driver was using custom functions WARNING, ERROR, DEBUG, instead of
 pr_err, pr_dgb...

New ep_* macros have been created that use standard pr_* functions.
Signed-off-by: NRicardo Ribalda Delgado <ricardo.ribalda@gmail.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

- Move logical continuations to end of line
- Improve spacing
Signed-off-by: NRicardo Ribalda Delgado <ricardo.ribalda@gmail.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

Fix Code Style using checkpatch.pl criteria
Signed-off-by: NRicardo Ribalda Delgado <ricardo.ribalda@gmail.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>