Convert all USB gadget sysfs attributes to use the _RO or _RW variants,
to make them easier to audit and ensure that the permissions are
correct.

Note, two are left using the DEVICE_ATTR() macro, as there is no
DEVICE_ATTR_WO() in Linus's tree, that will happen after 3.12-rc1 is
out, a follow-on patch will be sent then.
Reviewed-by: NFelipe Balbi <balbi@ti.com>
Acked-by: NFelipe Balbi <balbi@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

--
 drivers/usb/gadget/composite.c      |    8 +++-----
 drivers/usb/gadget/dummy_hcd.c      |    8 ++++----
 drivers/usb/gadget/f_mass_storage.c |   14 ++++++--------
 drivers/usb/gadget/net2272.c        |    4 ++--
 drivers/usb/gadget/net2280.c        |   18 +++++++++---------
 drivers/usb/gadget/storage_common.c |   25 ++++++++++++-------------
 drivers/usb/gadget/udc-core.c       |   14 +++++++-------
 7 files changed, 43 insertions(+), 48 deletions(-)

The name of udc state attribute file under sysfs is registered as
"state", while usb_gadget_set_state take it as "status" when it's
going to update. This patch fixes the typo.
Signed-off-by: NRong Wang <Rong.Wang@csr.com>
Signed-off-by: NBarry Song <Baohua.Song@csr.com>
Cc: <stable@kernel.org>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

usb_gadget_set_state() will call sysfs_notify()
which might sleep. Some users might want to call
usb_gadget_set_state() from the very IRQ handler
which actually changes the gadget state.

Instead of having every UDC driver add their own
workqueue for such a simple notification, we're
adding it generically to our struct usb_gadget,
so the details are hidden from all UDC drivers.
Acked-by: NAlan Stern <stern@rowland.harvard.edu>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

When CONFIG_HAS_DMA isn't enabled, the UDC core gets build errors:

drivers/built-in.o: In function `dma_set_coherent_mask':
include/linux/dma-mapping.h:93: undefined reference to `dma_supported'
include/linux/dma-mapping.h:93: undefined reference to `dma_supported'
drivers/built-in.o: In function `usb_gadget_unmap_request':
drivers/usb/gadget/udc-core.c:91: undefined reference to `dma_unmap_sg'
drivers/usb/gadget/udc-core.c:96: undefined reference to `dma_unmap_single'
drivers/built-in.o: In function `usb_gadget_map_request':
drivers/usb/gadget/udc-core.c:62: undefined reference to `dma_map_sg'
drivers/usb/gadget/udc-core.c:71: undefined reference to `dma_map_single'
drivers/usb/gadget/udc-core.c:74: undefined reference to `dma_mapping_error'

Prevent this by protecting the DMA API routines with preprocessor tests.
Signed-off-by: NAlan Stern <stern@rowland.harvard.edu>
CC: Geert Uytterhoeven <geert@linux-m68k.org>
Acked-by: NAlexander Shishkin <alexander.shishkin@linux.intel.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

udc-core would leak 'udc' memory in some error
cases. Fix it by kfree()ing udc on error path.
Reported-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

This patch (as1666) fixes a regression in the UDC core.  The core
takes care of unbinding gadget drivers, and it does the unbinding
before telling the UDC driver to turn off the controller hardware.
When the call to the udc_stop callback is made, the gadget no longer
has a driver.  The callback routine should not be invoked with a
pointer to the old driver; doing so can cause problems (such as
use-after-free accesses in net2280).

This patch should be applied, with appropriate context changes, to all
the stable kernels going back to 3.1.
Signed-off-by: NAlan Stern <stern@rowland.harvard.edu>
CC: <stable@vger.kernel.org>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

current code only allows the file owner (usually
root) to read current_speed and maximum_speed
sysfs files. Let anyone read those.
Signed-off-by: NFelipe Balbi <balbi@ti.com>

now that no UDC driver touches gadget.dev.release
we can assign our release function to it without
being afraid of breaking anything.
Signed-off-by: NFelipe Balbi <balbi@ti.com>

not all UDC drivers need a proper release function,
for those which don't need it, we udc-core will provide
a no-op release method so we can remove "redefinition"
of such methods in almost every UDC driver.
Signed-off-by: NFelipe Balbi <balbi@ti.com>

if we initialize gadget->dev.driver ourselves,
UDC drivers won't have to do the same, so we
can remove some duplicated code.
Signed-off-by: NFelipe Balbi <balbi@ti.com>

if udc-core always does it, we can delete
some extra lines from all UDC drivers. Besides,
it avoids mistakes from happening and propagating.
Signed-off-by: NFelipe Balbi <balbi@ti.com>

gadget's device pointer now is guaranteed to
have valid dma_mask, dma_parms and coherent_dma_mask
fields since we're always copying from our
parent device.
Signed-off-by: NFelipe Balbi <balbi@ti.com>

that's useful information to expose to userland.
Signed-off-by: NFelipe Balbi <balbi@ti.com>

We don't need the ->register_my_device flag
anymore because all UDC drivers have been
properly converted.

Let's remove every history of it.
Signed-off-by: NFelipe Balbi <balbi@ti.com>

Currently all UDC drivers are calling
device_register() before calling
usb_add_gadget_udc(). In order to avoid
code duplication, we can allow udc-core.c
register that device.

However that would become a really large patch,
so to cope with the meanwhile and allow us
to write bite-sized patches, we're adding
a flag which will be set by UDC driver once
it removes the code for registering the
gadget device.

Once all are converted, the new flag will
be removed.
Reviewed-by: NTomasz Figa <tomasz.figa@gmail.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

Those have been deprecated for a long time and
previous patches just converted all remaining
users of those.

Since there are no in-tree users and we don't
want any new users for them, let's obliterate
every piece of code related to those calls.
Signed-off-by: NFelipe Balbi <balbi@ti.com>

This patch adds udc_attach_driver() which allows to bind an UDC which is
specified by name to a driver. The name of available UDCs can be
obtained from /sys/class/udc. This interface is intended for configfs
interface.
Signed-off-by: NSebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

This commit removes the /sys/devices/platform/<UDC>/udc/<UDC>/is_dualspeed
file and is_dualspeeed line from /sys/devices/platform/ci13xxx_*/udc/device
file.

The is_dualspeed file is superseded by maximum_speed in the same directory
and is_dualspeed line in device file is superseded by max_speed line in
the same file.

The maximum_speed/max_speed specifies maximum speed supported by UDC.
To check if dualspeeed is supported, check if the value is >= 3.
Signed-off-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

This patch removes the global variable composite in composite.c.
The private data which was saved there is now passed via an additional
argument to the bind() function in struct usb_gadget_driver.

Only the "old-style" UDC drivers have to be touched here, new style are
doing it right because this change is made in udc-core.
Acked-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NSebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

This partly reverts 07a18bd7 ("usb gadget: don't save bind callback in
struct usb_composite_driver") and fixes new drivers. The section missmatch
problems was solved by whitelisting bind callback in modpost.
Acked-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NSebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

usb_gadget_remove_driver() runs through a four-step sequence to shut down
the gadget driver.  For the case of a composite gadget + at91 UDC, this
would look like:

    udc->driver->disconnect(udc->gadget);          // composite_disconnect()
    usb_gadget_disconnect(udc->gadget);            // at91_pullup(gadget, 0)
    udc->driver->unbind(udc->gadget);              // composite_unbind()
    usb_gadget_udc_stop(udc->gadget, udc->driver); // at91_stop()

The UDC driver can receive SETUP packets from the host up until the
point when usb_gadget_disconnect() returns.  On rare occasions, the
gadget driver may see this sequence:

    udc->driver->disconnect(udc->gadget);          // composite_disconnect()
    udc->driver->setup(udc->gadget, &ctrl);        // composite_setup()
    udc->driver->unbind(udc->gadget);              // composite_unbind()

Some gadget drivers, such as composite, assume this will never happen
and crash as a result.

The fix is to quiesce the UDC hardware (via usb_gadget_disconnect)
before running the gadget driver through the disconnect/unbind sequence.
Reviewed-by: NPeter Chen <peter.chen@freescale.com>
Signed-off-by: NKevin Cernekee <cernekee@gmail.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

This patch (as1548) fixes a recently-introduced incompatibility
between the UDC core and the dummy-hcd driver.  Commit
8ae8090c (usb: gadget: udc-core: fix
asymmetric calls in remove_driver) moved the usb_gadget_udc_stop()
call in usb_gadget_remove_driver() below the usb_gadget_disconnect()
call.

As a result, usb_gadget_disconnect() gets called at a time when the
gadget driver believes it has been unbound but dummy-hcd believes
it has not.  A nasty error ensues when dummy-hcd calls the gadget
driver's disconnect method a second time.

To fix the problem, this patch moves the gadget driver's unbind
notification after the usb_gadget_disconnect() call.  Now nothing
happens between the two unbind notifications, so nothing goes wrong.
Signed-off-by: NAlan Stern <stern@rowland.harvard.edu>
Tested-by: NAlexander Shishkin <alexander.shishkin@linux.intel.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

commit 6d258a4c (usb: gadget: udc-core: stop UDC on device-initiated
disconnect) introduced another case of asymmetric calls when issuing
a device-initiated disconnect. Fix it.
Reported-by: NBen Hutchings <ben@decadent.org.uk>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

During modprobe of gadget driver, pullup is called after
udc_start. In order to make the exit path symmetric when
removing a gadget driver, call pullup before ->udc_stop.

This is needed to avoid issues with PM where udc_stop
disables the module completely (put IP in reset state,
cut functional and interface clocks, and so on), which
prevents us from accessing the IP's address space,
thus creating the possibility of an abort exception
when we try to access IP's address space after clocks
are off.

Cc: stable@vger.kernel.org
Signed-off-by: NPartha Basak <p-basak2@ti.com>
Signed-off-by: NKishon Vijay Abraham I <kishon@ti.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

When we want to do device-initiated disconnect,
let's make sure we stop the UDC in order to
e.g. allow lower power states to be achieved by
turning off unnecessary clocks and/or stoping
PHYs.

When reconnecting, call ->udc_start() again to
make sure UDC is reinitialized.

Cc: stable@vger.kernel.org
Signed-off-by: NFelipe Balbi <balbi@ti.com>

such utilities are currently duplicated on all UDC
drivers basically with the same structure. Let's group
all implementations into one generic implementation
and get rid of that duplication.
Signed-off-by: NFelipe Balbi <balbi@ti.com>

This commit replaces usb_gadget's is_dualspeed field with
a max_speed field.

[ balbi@ti.com : Fixed DWC3 driver ]
Signed-off-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

Those are simply giving information about the current
state of the UDC, nothing really fancy. We can let
everybody read those.
Signed-off-by: NFelipe Balbi <balbi@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

usb_gadget_disconnect() is responsible of removing
data pullups. Before doing that we must, first, tell
gadget driver we're disconnecting (by calling disconnect
method on gadget driver structure), unbind the gadget
driver and stop the controller.
Signed-off-by: NFelipe Balbi <balbi@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

We should not be using dev_get_drvdata() because we
never call dev_set_drvdata(). Let's use container_of()
as all other sysfs attributes.
Signed-off-by: NFelipe Balbi <balbi@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

We should not be using dev_get_drvdata() because we
never call dev_set_drvdata(). Let's use container_of()
as all other sysfs attributes.
Signed-off-by: NFelipe Balbi <balbi@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

In a few places in the kernel, the code prints
a human-readable USB device speed (eg. "high speed").
This involves a switch statement sometimes wrapped
around in ({ ... }) block leading to code repetition.

To mitigate this issue, this commit introduces
usb_speed_string() function, which returns
a human-readable name of provided speed.

It also changes a few places switch was used to use
this new function.  This changes a bit the way the
speed is printed in few instances at the same time
standardising it.
Signed-off-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

This was somehow forgotten.
Signed-off-by: NSebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: NFelipe Balbi <balbi@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

udc_start() should only trigger the internal state machine and make
minimal house keeping. Before that call udc-core calls the bind()
callback and after the callback the pullup().

udc_stop() is simillar, udc-core calls pullup(), unbind() and finally
udc_stop().
Signed-off-by: NSebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: NFelipe Balbi <balbi@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

this class will be used to abstract away several of the duplicated
operations scattered among the USB gadget controller drivers.

Later, we can add an atomic notifier to tell interested drivers about
what's happening with the controller. Notifications such as suspend,
resume, enumerated, etc. will be useful, at a minimum, for implementing
usb charger detection.

As part of the converting process usb_gadget_probe_driver() is no longer
part of each udc but pushed into the ->stap() callback. The same for his
couterpart.

The core is currently set explicit to 'n'. It will be changed to 'y' once
all users are converted since it provides functions which clash with
other drivers.
Signed-off-by: NFelipe Balbi <balbi@ti.com>
Signed-off-by: NSebastian Andrzej Siewior <bigeasy@linutronix.de>
Acked-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>