The interrupter register set includes a register that says whether interrupts
are pending for each event ring (the IP bit).  Each MSI-X vector will get its
own interrupter set with separate IP bits.  The status register includes an
"Event Interrupt (EINT)" bit that is set when an IP bit is set in any of the
interrupters.

When PCI interrupts are used, the EINT bit exactly mirrors the IP bit in the
single interrupter set, and it is a waste of time to check both registers when
trying to figure out if the xHC interrupted or another device on the shared IRQ
line interrupted.  Only check the IP bit to reduce register reads.

The IP bit is automatically cleared by the xHC when MSI or MSI-X is enabled.  It
doesn't make sense to read that register to check for shared interrupts (since
MSI and MSI-X aren't shared).  It also doesn't make sense to write to that
register to clear the IP bit, since it is cleared by the hardware.

We can tell whether MSI or MSI-X is enabled by looking at the irq number in
hcd->irq.  If it's -1, we know MSI or MSI-X is enabled.
Signed-off-by: NSarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Now that the event handler functions no longer use xhci_set_hc_event_deq()
to update the event ring dequeue pointer, that function is not used by
anything in xhci-ring.c.  Move that function into xhci-mem.c and make it
static.
Signed-off-by: NSarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

The xHCI specification suggests that writing the hardware event ring dequeue
pointer register too often can be an expensive operation for the xHCI hardware
to manage.  It suggests minimizing the number of writes to that register.

Originally, the driver wrote the event ring dequeue pointer after each
event was processed.  Depending on how the event ring moderation register
is set up and how fast the transfers are completing, there may be several
events processed for each interrupt.  This patch makes the hardware event
ring dequeue pointer be written only once per interrupt.

Make the transfer event handler and port status event handler only write
the software event ring dequeue pointer.  Move the updating of the
hardware event ring dequeue pointer into the interrupt function.  Move the
contents of xhci_set_hc_event_deq() into the interrupt handler.  The
interrupt handler must clear the event handler busy flag, so it might as
well also write the dequeue pointer to the same register.  This eliminates
two 32-bit PCI reads and two 32-bit PCI writes.
Reported-by: NAndiry Xu <andiry.xu@amd.com>
Signed-off-by: NSarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

xhci_handle_event() is now only called from within xhci-ring.c, so make it
static.
Signed-off-by: NSarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Remove a duplicate register read of the interrupt pending register from
xhci_irq().  Also, remove waiting on the posted write of that register.
The host will see it eventually.  It will probably read the register
itself before deciding whether to interrupt the system again, forcing the
posted write to complete.
Signed-off-by: NSarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

When we move xhci_work() into xhci_irq(), we don't need to read the operational
register status field twice.
Signed-off-by: NSarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Most of the work for interrupt handling is done in xhci-ring.c, so it makes
sense to move the functions that are first called when an interrupt happens
(xhci_irq() or xhci_msi_irq()) into xhci-ring.c, so that the compiler can better
optimize them.

Shorten some lines to make it pass checkpatch.
Signed-off-by: NSarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

I've been using perf to measure the top symbols while transferring 1GB of data
on a USB 3.0 drive with dd.  This is using the raw disk with /dev/sdb, with a
block size of 1K.

During performance testing, the top symbol was xhci_triad_to_transfer_ring(), a
function that should return immediately if streams are not enabled for an
endpoint.  It turned out that the functions to find the endpoint ring was
defined in xhci-mem.c and used in xhci-ring.c and xhci-hcd.c.  I moved a copy of
xhci_triad_to_transfer_ring() and xhci_urb_to_transfer_ring() into xhci-ring.c
and declared them static.  I also made a static version of
xhci_urb_to_transfer_ring() in xhci.c.

This improved throughput on a 1GB read of the raw disk with dd from
186MB/s to 195MB/s, and perf reported sampling the xhci_triad_to_transfer_ring()
0.06% of the time, rather than 9.26% of the time.
Signed-off-by: NSarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Add support for the Quatech SSU-100 single port usb to serial device.
This driver is based on the ftdi_sio.c driver and the original
serqt_usb driver from Quatech.
Signed-off-by: NBill Pemberton <wfp5p@virginia.edu>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This patch (as1400) adds runtime-PM support to usb-storage.  It
utilizes the SCSI layer's runtime-PM implementation, so its scope is
limited.  Currently the only effect is that disk-like devices (such as
card readers or flash drives) will be autosuspended if they aren't
mounted and their device files aren't open.  This would apply, for
example, to card readers that don't contain a memory card.

Unfortunately this won't interact very well with the removable-media
polling normally carried out by hal or DeviceKit.  Maybe those
programs can be changed to use a longer polling interval, or maybe the
default autosuspend time for usb-storage should be set to something
below 1 second.
Signed-off-by: NAlan Stern <stern@rowland.harvard.edu>
Cc: James Bottomley <James.Bottomley@suse.de>
Cc: Matthew Dharm <mdharm-usb@one-eyed-alien.net>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

The patch below on gregkh tree only creates 'lpm' file under
ehci->debug_dir, but not removes it when unloading module,

	 USB: EHCI: EHCI 1.1 addendum: preparation

which can make loading of ehci-hcd module failed after unloading it.

This patch replaces debugfs_remove with debugfs_remove_recursive
to remove ehci debugfs dir and files. It does fix the bug above,
and may simplify the removing procedure.

Also, remove the debug_registers, debug_async and debug_periodic
field from ehci_hcd struct since they are useless now.
Signed-off-by: NMing Lei <tom.leiming@gmail.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This adds USB 2.0 support to ssb ohci driver.
This patch was used in OpenWRT for a long time now.

CC: Steve Brown <sbrown@cortland.com>
Signed-off-by: NHauke Mehrtens <hauke@hauke-m.de>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Signed-off-by: NAlessio Igor Bogani <abogani@texware.it>
Cc: stable <stable@kernel.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This patch is to add a US Interface, Inc. "Navigator" USB device.
Specifically, it's a HAM Radio USB sound modem that also
incorporates three pairs of unique FTDI serial ports.  The standard
Linux FTDI serial driver will only recognize the first two serial
ports of an unknown FDTI derived device and this patch adds in
recognition to these specific new IDs.
Signed-off-by: NDavid A. Ranch <dranch@trinnet.net>
Cc: stable <stable@kernel.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This patch (as1410) makes a slight change to the strategy used for
choosing a default configuration.  Currently we skip configs whose
first interface is RNDIS, if the kernel wasn't built with the
corresponding driver.  This risks losing access to the other
interfaces in those configs.  In addition, if there is only one config
then we will end up not configuring the device at all.

This changes the logic; now such configurations will be skipped only
if there is at least one other config.
Signed-off-by: NAlan Stern <stern@rowland.harvard.edu>
Tested-by: NAdam Kropelin <akropel1@rochester.rr.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This patch adds the product IDs of Huawei's K3765 and K4505 mobile
broadband usb modems to option.c. It also adds a quirk to the option
probe function so that binding to the device's network interface(class
0xff) is avoided. This is necessary to allow another driver to bind to
that, and to avoid programs like wvdial opening a nonfunctioning tty
during modem discovery.
Signed-off-by: NAndrew Bird <ajb@spheresystems.co.uk>
Cc: stable <stable@kernel.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Moved the serial parameter handling code out of "#ifdef
CONFIG_USB_FILE_STORAGE_TEST".

This modifies Yann Cantin's commit "USB: Add a serial number
parameter to g_file_storage" module as per Alan Stern's request.
Signed-off-by: NMichal Nazarewicz <m.nazarewicz@samsung.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>
Acked-by: NAlan Stern <stern@rowland.harvard.edu>
Tested-by: NAnand Gadiyar <gadiyar@ti.com>
Cc: David Brownell <david-b@pacbell.net>
Cc: Yann Cantin <yann.cantin@laposte.net>

imx21_hc_reset() uses schedule_timeout() without setting state to
STATE_(UN)INTERRUPTIBLE. As it is called in cycle without checking of
pending signals, use schedule_timeout_uninterruptible().
Signed-off-by: NKulikov Vasiliy <segooon@gmail.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

I have added the ProductID=0xe729 VendorID=FTDI_VID=0x0403 which will
enable support for the Segway Robotic Mobility Platform (RMP200) in the
ftdi_sio kernel module.  Currently, users of the Segway RMP200 must use
a RUN+="/sbin/modprobe -q ftdi-sio product=0xe729 vendor=0x0403 in a
udev rule to get the ftdi_sio module to handle the usb interface and
mount it on /dev/ttyXXX.  This is not a good solution because some users
will have multiple USB to Serial converters which will use the ftdi_sio
module.
Signed-off-by: NJohn Rogers <jgrogers@gmail.com>
Cc: stable <stable@kernel.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This patch adds support for clock gating of the HS/OTG block. On S5PV210
otg gating clock is initally disabled so the driver needs to get and
enable it before it can access its registers.
Signed-off-by: NMarek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

S5PV210 SoCs has 2 USB PHY interfaces, both enabled by writing zero to
S3C_PHYPWR register. HS/OTG driver uses only PHY0, so do not touch bits
related to PHY1.
Signed-off-by: NMarek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

drivers/usb/gadget/s3c-hsotg.c: In function ‘s3c_hsotg_otgreset’:
drivers/usb/gadget/s3c-hsotg.c:2816: error: ‘MHZ’ undeclared (first use in this function)
drivers/usb/gadget/s3c-hsotg.c:2816: error: (Each undeclared identifier is reported only once
drivers/usb/gadget/s3c-hsotg.c:2816: error: for each function it appears in.)
Signed-off-by: NMarek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

The PLL that drives the USB clock supports 3 input clocks: 12, 24 and 48Mhz.
This patch adds support to the USB driver for setting the correct register bit
according to the given clock.

This depends on the following patch:
[PATCH] ARM: S3C64XX: Add USB external clock definition
Signed-off-by: NMaurus Cuelenaere <mcuelenaere@gmail.com>
Signed-off-by: NBen Dooks <ben-linux@fluff.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

If there is more data in the request than we could fit into a single
hardware request, then check when the OutDone event is received if
we have more data, and if so, schedule the new data instead of trying
to complete the request (and in the case of EP0, sending a 0 packet
in the middle of a transfer).

Also, move the debug message about the current transfer state before
the warning about a bad transfer.
Signed-off-by: NBen Dooks <ben-linux@fluff.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

The EP0 out limit is the same as the IN limit, so make them the same.
Signed-off-by: NBen Dooks <ben-linux@fluff.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

The maximum length for any EP0 IN request on EP0 is 127 bytes, not 128
as the driver currently has it.
Signed-off-by: NBen Dooks <ben-linux@fluff.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Before trying a new setup transaction after getting an EP0 in complete
interrupt, check that the driver did not try and send more EP0 IN data
before enqueing a new setup transaction.

This fixes a bug where we cannot send all of the IN data in one go
so split the transfer, but then fail to send all the data as we start
waiting for a new OUT transaction
Signed-off-by: NBen Dooks <ben-linux@fluff.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Limit the IN FIFO write to a single packet per attempt at writing,
as per the specifications and ensure that we don't return fifo-full
so that we can continue writing packets if we have the space.
Signed-off-by: NBen Dooks <ben-linux@fluff.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Add support for the dedicated FIFO mode on newer SoCs such as the S5PV210
partly to improve support and to fix the bug where any non-EP0 IN endpoint
requires its own FIFO allocation.

To fix this, we ensure that any non-zero IN endpoint is given a TXFIFO
using the same allocation method as the periodic case (all our current
hardware has enough FIFOs and FIFO memory for a 1:1 mapping) and ensure
that the necessary transmission done interrupt is enabled.

The default settings from reset for the core point all EPs at FIFO0,
used for the control endpoint. However, the controller documentation
states that all IN endpoints _must_ have a unique FIFO to avoid any
contention during transmission.

Note, this leaves us with a large IN FIFO for EP0 (which re-uses the
old NPTXFIFO) for an endpoint which cannot shift more than a pair of
packets at a time... this is a waste, but it looks like we cannot
re-allocate space to the individual IN FIFOs as they are already
maxed out (to be confirmed).
Signed-off-by: NBen Dooks <ben-linux@fluff.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

The USB documentation suggest that the FIFOs should be reset when a
bus reset event happens. Use the s3c_hsotg_init_fifo() to ensure that
the FIFO layout is correct and that the FIFOs are flushed before
acknowledging the reset.
Signed-off-by: NBen Dooks <ben-linux@fluff.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

In shared fifo mode (used on older SoCs) the periodic in fifo beahves
much more like a packet buffer, discarding old data when writing new
data. Avoid this by ensuring that we do not load new transactions in
when there is data sitting already in the FIFO.

Note, this may not be an observed bug, we are fixing the case that this
may happen.
Signed-off-by: NBen Dooks <ben-linux@fluff.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Fix a problem where we have been underestimating the space available in
the IN PTX/NPTX FIFOs by assuming that they where simply word aligned
instead of in number-of-words. This means all length calculations need
to be multiplied-by-4.

Note, we do not change the information about fifo size or start addresses
available to userspace as we assume the user can multiply by four easily
and is already knows these values are in words.
Signed-off-by: NBen Dooks <ben-linux@fluff.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Up the FIFO size for the TX to 1024 entries, as this now seems to work
with all the cores. This fixes a problem when using large packets on
a core with MPS set to 512 can hang due to insufficient space for the
writes.

The hang arises due to getting the non-periodic FIFO empty IRQ but
not being able to satisfy any requests since there is never enough
space to write 512 bytes into the buffer. This means we end up with
a stream of interrupt requests.

It is easier to up the TX FIFO to fill the space we left for it
than to try and fix the positions in the code where we should have
limited the max-packet size to < TXFIFOSIZE, since the TXFIFOSIZE
depends on how the TX FIFOs have been setup.
Signed-off-by: NBen Dooks <ben-linux@fluff.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

If kzalloc() fails exit with -ENOMEM.
Signed-off-by: NKulikov Vasiliy <segooon@gmail.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

MS Windows mounts removable storage in "Removal optimized mode" by
default. All the writes to the media are synchronous which is achieved
by setting FUA (Force Unit Access) bit in SCSI WRITE(10,12) commands.
This prevents I/O requests aggregation in block layer dramatically
decreasing performance.

This patch brings an option to accept or ignore mentioned bit
 a) via specifying module parameter "nofua", or
 b) through sysfs entry
	/sys/devices/platform/_UDC_/gadget/gadget-lunX/nofua
	(_UDC_ is the name of the USB Device Controller driver)

Patch is based on the work that was done by Denis Karpov for Maemo 5
platform.
Signed-off-by: NAndy Shevchenko <ext-andriy.shevchenko@nokia.com>
Acked-by: NAlan Stern <stern@rowland.harvard.edu>
Cc: Denis Karpov <ext-denis.2.karpov@nokia.com>
Cc: Adrian Hunter <adrian.hunter@nokia.com>
Cc: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Bring a strict way to get the 'ro' parameter from the user.

The patch followed by this one adds another boolean parameter. To be consistent
Michał Nazarewicz proposed to use simple_strtol() in both cases (correspondend
discussion in LKML [1]). Due to simple_strtol() doesn't return error in a good
way and we have a boolean parameter the strict_strtoul() is used.

[1] http://lkml.org/lkml/2010/7/14/169Signed-off-by: NAndy Shevchenko <ext-andriy.shevchenko@nokia.com>
Acked-by: NAlan Stern <stern@rowland.harvard.edu>
Cc: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Enable isochronous urb enqueue.
Signed-off-by: NAndiry Xu <andiry.xu@amd.com>
Signed-off-by: NSarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Isochronous endpoint needs a bigger size of transfer ring. Isochronous URB
consists of multiple packets, each packet needs a isoc td to carry, and
there will be multiple trbs inserted to the ring at one time. One segment
is too small for isochronous endpoints, and it will result in
room_on_ring() check failure and the URB is failed to enqueue.

Allocate bigger ring for isochronous endpoint. 8 segments should be enough.
This will be replaced with dynamic ring expansion in the future.
Signed-off-by: NAndiry Xu <andiry.xu@amd.com>
Signed-off-by: NSarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This patch implements isochronous urb enqueue and interrupt handler part.

When an isochronous urb is passed to xHCI driver, first check the transfer
ring to guarantee there is enough room for the whole urb. Then update the
start_frame and interval field of the urb. Always assume URB_ISO_ASAP
is set, and never use urb->start_frame as input.

The number of isoc TDs is equal to urb->number_of_packets. One isoc TD is
consumed every Interval. Each isoc TD consists of an Isoch TRB chained to
zero or more Normal TRBs.

Call prepare_transfer for each TD to do initialization; then calculate the
number of TRBs needed for each TD. If the data required by an isoc TD is
physically contiguous (not crosses a page boundary), then only one isoc TRB
is needed; otherwise one or more additional normal TRB shall be chained to
the isoc TRB by the host.

Set TRB_IOC to the last TRB of each isoc TD. Do not ring endpoint doorbell
to start xHC procession until all the TDs are inserted to the endpoint
transer ring.

In irq handler, update urb status and actual_length, increase
urb_priv->td_cnt. When all the TDs are completed(td_cnt is equal to
urb_priv->length), giveback the urb to usbcore.
Signed-off-by: NAndiry Xu <andiry.xu@amd.com>
Signed-off-by: NSarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

Add urb_priv data structure to xHCI driver. This structure allows multiple
xhci TDs to be linked to one urb, which is essential for isochronous
transfer. For non-isochronous urb, only one TD is needed for one urb;
for isochronous urb, the TD number for the urb is equal to
urb->number_of_packets.

The length field of urb_priv indicates the number of TDs in the urb.
The td_cnt field indicates the number of TDs already processed by xHC.
When td_cnt matches length, the urb can be given back to usbcore.

When an urb is dequeued or cancelled, add all the unprocessed TDs to the
endpoint's cancelled_td_list. When process a cancelled TD, increase
td_cnt field. When td_cnt equals urb_priv->length, giveback the
cancelled urb.
Signed-off-by: NAndiry Xu <andiry.xu@amd.com>
Signed-off-by: NSarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>