In the current implementation functionfs generates a EFAULT for async read
operations if the read buffer size is larger than the URB data size. Since
a application does not necessarily know how much data the host side is
going to send it typically supplies a buffer larger than the actual data,
which will then result in a EFAULT error.

This behaviour was introduced while refactoring the code to use iov_iter
interface in commit c993c39b ("gadget/function/f_fs.c: use put iov_iter
into io_data"). The original code took the minimum over the URB size and
the user buffer size and then attempted to copy that many bytes using
copy_to_user(). If copy_to_user() could not copy all data a EFAULT error
was generated. Restore the original behaviour by only generating a EFAULT
error when the number of bytes copied is not the size of the URB and the
target buffer has not been fully filled.

Commit 342f39a6 ("usb: gadget: f_fs: fix check in read operation")
already fixed the same problem for the synchronous read path.

Fixes: c993c39b ("gadget/function/f_fs.c: use put iov_iter into io_data")
Acked-by: Michal Nazarewicz <mina86@mina86.com>
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

This function compiles to 298 bytes of machine code, has ~10 callsites.

This is a USB 2.0 device, USB 2.0 is limited to ~40 MB/s, so should be
almost never CPU bound. No need to optimize for speed this agressively.
Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com>
CC: Felipe Balbi <balbi@ti.com>
CC: linux-usb@vger.kernel.org
CC: linux-kernel@vger.kernel.org
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

The previous code could use the first USB-DMAC with IPMMU if iommus
property was set into this device node. However, in this case, it
could not control the second USB-DMAC with IPMMU because a parameter
of IPMMU (micro-TLB id) is different with each USB-DMAC.

So, this patch uses the usb_gadget_{un}map_request_by_dev() APIs for
IPMMU. (Then, iommus property should be set into USB-DMAC node(s).)
Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

Since usbhsg_dma_map_ctrl() needs DMA device structure in the near future,
this patch changes arguments of dma_map_ctrl() to give such data.
(This patch is only change the argument.)
Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

Since usbhsf_dma_{un}map() will use the "fifo" data in the near future,
this patch changes function call orfer in usbhsf_dma_prepare_push().
Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

If the following environment, the first argument of DMA API should
be set to a DMAC's device structure, not a udc controller's one.
 - A udc controller needs an external DMAC device (like a DMA Engine).
 - The external DMAC enables IOMMU.

So, this patch add usb_gadget_{un}map_request_by_dev() API to set
a DMAC's device structure by a udc controller driver.
Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

The argument of dev_err() in usb_gadget_map_request() should be dev
instead of &gadget->dev.

Fixes: 7ace8fc8 ("usb: gadget: udc: core: Fix argument of dma_map_single for IOMMU")
Cc: <stable@vger.kernel.org> # v4.3+
Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>

Debugfs init failure is not so important. We can continue our job on
this failure. Also no break need for debugfs_create_file call failure.
Signed-off-by: Du, Changbin <changbin.du@intel.com>

[felipe.balbi@linux.intel.com :
	- remove out-of-memory message, we get that from OOM.
	- switch dev_err() to dev_dbg() ]
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

dma_status bit flag is set but never really used
so get rid of it.
Reported-by: Felipe Balbi <balbi@kernel.org>
Signed-off-by: Roger Quadros <rogerq@ti.com>
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

It makes no sense to interrupt in the middle of
chained transfer. This patch just makes sure we
don't do that.
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

trace already adds a newline character for us, we
don't need to do it ourselves.
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

instead of limiting link TRB only to Isoc endpoints,
let's use it for all endpoint types, this way we are
more likely to transfer more data before a
XferComplete event.
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

By moving our % DWC3_NUM_TRB operation to the
increment helpers, the rest of the driver can be
simplified.

It's also a good practice to make sure we will have
a single place dealing with details about how to
increment our enqueue and dequeue pointers.
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

Add three little helpers which will aid in making
the code slightly easier to read. One helper
increments enqueue pointer, another increments
dequeue pointer and the last one tests if we're
dealing with the last TRB.
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

instead of using a bitwise and, let's rely on the %
operator since that's a lot more clear. Also, GCC
will optimize % 256 to nothing anyway.
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

We *know* that we have 1 PAGE (4096 bytes) for our
TRB poll. We also know the size of each TRB and know
that we can fit 256 of them in one PAGE. By using a
u8 type we can make sure that:

	enqueue++ % 256;

gets optimized to an increment only.
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

No functional changes. Merely adding useful
documentation for future readers.
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

This makes it clear that we're dealing with a queue
of TRBs. No functional changes. While at that, also
rename start_slot to first_trb_index for similar
reasons.
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

The DWC3 OMAP driver supports DT-boot only, as result dma_mask will be
always configured properly from DT -
of_platform_device_create_pdata()->of_dma_configure(). More over,
dwc3-omap.c can be built as module and in this case it's unsafe to
assign local variable as dma_mask.

Hence, remove dma_mask configuration code.
Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com>
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

Some freescale QorIQ platforms require to disable receiver detection
in P3 for correct detection of USB devices. If GUSB3PIPECTL(DISRXDETINP3)
is set, Core will change PHY power state to P2 and then perform receiver
detection. After receiver detection, Core will change PHY power state to
P3. Same quirk would be added in dts file in future patches.
Signed-off-by: Sriram Dash <sriram.dash@nxp.com>
Signed-off-by: Rajesh Bhagat <rajesh.bhagat@nxp.com>
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

Add snps,dis_rxdet_inp3_quirk property which disables receiver detection
in PHY P3 power state.
Signed-off-by: Sriram Dash <sriram.dash@nxp.com>
Signed-off-by: Rajesh Bhagat <rajesh.bhagat@nxp.com>
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

Synopsys Databook says we should move link to U0
before issuing a Start Transfer command. We could
require the gadget driver to call
usb_gadget_wakeup() however I feel that changing all
gadget drivers to keep track of Link State and
conditionally call usb_gadget_wakeup() would be far
too much work. For now we will handle this at the
UDC level, but at some point composite.c should be
one handling this.
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

we will need this from StartTransfer to make sure
link is in U0 before starting a transfer.
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

Synopsys Databook 2.60a has a note that if we're
sending an endpoint command we _must_ make sure that
DWC3_GUSB2PHY(n).SUSPHY bit is cleared.

This patch implements that particular detail.
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

dwc3_send_gadget_ep_cmd() had three return
points. That becomes a pain to track when we need to
debug something or if we need to add more code
before returning.

Let's combine all three return points into a single
one just by introducing a local 'ret' variable.
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

Sort IDs in groups to be easily found when needed.

There is no functional change.
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

It seems there are leftovers of some assignments which are not used
anymore.  Compiler even warns us about:

drivers/usb/gadget/udc/pch_udc.c:2022:22: warning: variable ‘dev’ set \
but not used [-Wunused-but-set-variable]

drivers/usb/gadget/udc/pch_udc.c:2639:9: warning: variable ‘ret’ set \
but not used [-Wunused-but-set-variable]

Remove them and shut compiler about.
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

Try to enable MSI in case hardware supports it. At least Intel Quark is
known SoC which indeed does.
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

devres API allows to make error paths cleaner and less error
prone. Convert the driver to use it.
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

There is no need to repeat the work that is already done in the PCI
driver core. The patch removes excerpts from suspend and resume
callbacks.

Note that there is no more calls performed to enable or disable a PCI
device during suspend-resume cycle. Nowadays they seems to be
superfluous. Someone can read more in [1].

While here, convert PM ops to use modern API.

[1] https://www.kernel.org/doc/ols/2009/ols2009-pages-319-330.pdfSigned-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>

[felipe.balbi@linux.intel.com: fixed build break and checkpatch error ]
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

The phy-am335x driver selects 'USB_COMMON', but all other drivers
use 'depends on' for that symbol, and it depends on USB || USB_GADGET
itself, which causes a Kconfig warning:

warning: (AM335X_PHY_USB) selects USB_COMMON which has unmet direct dependencies (USB_SUPPORT && (USB || USB_GADGET))

As suggested by Felipe Balbi, this turns the logic around, and makes
'USB_COMMON' selected by everything else that needs it, so we can
remove the dependencies.

Fixes: 59f042f6 ("usb: phy: phy-am335x: bypass first VBUS sensing for host-only mode")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Felipe Balbi <balbi@kernel.org>
Reviewed-by: Peter Chen <peter.chen@nxp.com>
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

we don't plan on using multiple event buffers, but
if we find a good use case for it, this little trick
will help us avoid a loop in hardirq handler looping
for each and every event buffer.
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

we will be using a single event buffer and that
renders ev_buffs array unnecessary. Let's remove it
in favor of a single pointer to a single event
buffer.
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

We never, ever route any of the other event buffers
so we might as well drop support for them.

Until someone has a real, proper benefit for
multiple event buffers, we will rely on a single
one. This also helps reduce memory footprint of
dwc3.ko which won't allocate memory for the extra
event buffers.
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

coccicheck found this pattern which could be
converted to PTR_ERR_OR_ZERO(). No functional
changes.
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

coccicheck found this pattern which could be
converted to PTR_ERR_OR_ZERO(). No functional
changes.
Acked-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

request_list and req_queued were, well, weird naming
choices.

Let's give those better names and call them,
respectively, pending_list and started_list. These
new names better reflect what these lists are
supposed to do.

While at that also rename req->queued to req->started.
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

previously we were using a maximum of 32 TRBs per
endpoint. With each TRB being 16 bytes long, we were
using 512 bytes of memory for each endpoint.

However, SLAB/SLUB will always allocate PAGE_SIZE
chunks. In order to better utilize the memory we
allocate and to allow deeper queues for gadgets
which would benefit from it (g_ether comes to mind),
let's increase the maximum to 256 TRBs which rounds
up to 4096 bytes for each endpoint.
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

CSP bit of TRB Control is useful for protocols such
CDC EEM/ECM/NCM where we're transferring in blocks
of MTU-sized requests (usually MTU is 1500 bytes).

We know we will always have a short packet after two
(for HS) wMaxPacketSize packets and, usually, we
will have a long(-ish) queue of requests (for our
g_ether gadget, we have at least 10
requests).

Instead of always stopping the queue processing to
interrupt, giveback and restart, let's tell dwc3 to
interrupt but continue processing following request
if we have anything already pending in the queue.

This gave me a considerable improvement of 40% on my
test setup.
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>

That FIFO resizing logic was added to support OMAP5
ES1.0 which had a bogus default FIFO size. I can't
remember the exact size of default FIFO, but it was
less than one bulk superspeed packet (<1024) which
would prevent USB3 from ever working on OMAP5 ES1.0.

However, OMAP5 ES1.0 support has been dropped by
commit aa2f4b16 ("ARM: OMAP5: id: Remove ES1.0
support") which renders FIFO resizing unnecessary.
Tested-by: Kishon Vijay Abraham I <kishon@ti.com>
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>