This is a re-write of the Motorola mdm6600_modem driver.
The usb_serial_generic_* handlers actually cover the USB bulk to serial data
path. So that code disappears.
The USB interrupt data to TIOCMGET ioctl path does not appear to be used by
pppd in userspace, using a temporary BUG_ON() to confirm.
So we end up with very little but a hook to register generic usb-serial
handlers for the MDM6600 vendor id.
I have not spent time testing behavior in low-power-modes, it is likely I will
need to over-ride some of the generic handlers at that point.
Change-Id: I65fdd06c2764416f365995761aa653878c0e393b
Signed-off-by: Nick Pelly <npelly@google.com>
moto_flashmdm_attach should also override bulk_out_buffers as
usb-serial expects they have the same size as bulk_out_buffer
Change-Id: I97487fa833ed3dee40a15c485ba8565944d0f9e4
Signed-off-by: Benoit Goby <benoit@android.com>
This allow's Motorola's mdm6600_modem driver to work with suspend.
Patch provided by Motorola.
mdm6600_modem will shortly be replaced by a re-write, but its useful to commit
this update for easier A/B testing.
Change-Id: I0489f139e59e42ba1b6fb1979cf8b3ea8196a126
Signed-off-by: Nick Pelly <npelly@google.com>
This driver receives notifications from cpcap-whisper and emulates the same
behavior as when the VBUS and ID USB pins are connected to the app processor.
Change-Id: I671e64c58559a2a3036273ef1fe8acb6af516fac
Signed-off-by: Benoit Goby <benoit@android.com>
The OTG driver disables the gadget device when the cable is
removed, so there is no need to check if the cable is plugged
before touching registers.
Change-Id: I0b1a3a8b07560d3eca2e2e25574b5219e3373808
Signed-off-by: Benoit Goby <benoit@android.com>
On suspend, dr_controller_stop disable interrupts and on resume, interrupts
are disabled until dr_controller_run is called, so it is safe to call
fsl_udc_clk_suspend/resume with interrupts and the spinlock unlocked.
Change-Id: I33618295ea096a4bfd796d1a07dfc9722e7786b0
Signed-off-by: Benoit Goby <benoit@android.com>
If the device connected to a port has out-of-band wakeup
signaling, the phy and controller may be powered off on bus suspend.
Change-Id: Ia206f05d01160411b97aefa83045cd759d35b66d
Signed-off-by: Benoit Goby <benoit@android.com>
If we use an OTG driver, the driver will detect VBUS changes and notify
the gadget driver through vbus_session. Enable/disable the gadget driver
in vbus session so that there is no need to check the OTG state on every
interrupt.
Change-Id: I617ad5742be2632b2257b71314db8f330be463d5
Signed-off-by: Benoit Goby <benoit@android.com>
There is no need anymore to check the OTG state on every interrupts and
use a work thread.
Moved the suspend code from usb_phy.c as this is ehci specific.
Change-Id: I523baab1476323a35360b1d802088370e42d0fd7
Signed-off-by: Benoit Goby <benoit@android.com>
Tegra quirk: The PORT_RESET bit in PORTSC1 does not need to be cleared
and there is no need to wait for it to clear. The bit will automatically
change to 0 when the bus-reset sequence is done and an interrupt will be
generated.
Change-Id: I645417013af46785a249096ebc06a1f688228d94
Signed-off-by: Benoit Goby <benoit@android.com>
only reset the controller when doing so won't also reset the phy (Tegra quirk)
Change-Id: I549a18977d0d5ebfa12c32016aa9e6bffaa8643c
Signed-off-by: Gary King <gking@nvidia.com>
On suspend, use phy_suspend save the phy registers so that there is
no need to reset the controller and re-enumerate devices on resume.
Change-Id: I00fe5b87a1b319044724494b8e635b540088a38b
Signed-off-by: Benoit Goby <benoit@android.com>
Reset the controller when the cable is unplugged to leave it in the idle
state. The OTG driver will reconfigure it on vbus/id pin detection.
Change-Id: I87903ec86f3c35af64a141f27a34cc0720a61b08
Signed-off-by: Benoit Goby <benoit@android.com>
The OTG driver will power it up on ID pin detection. This avoids
race conditions when the device is powered on with the otg cable
attached.
Change-Id: I6bd5f1d73284f9b8534ef3dfb936b81a9400fa5b
Signed-off-by: Benoit Goby <benoit@android.com>
Check the transceiver state before checking udc->stopped
Enable/disable the PHY and the clock on cable events
Change-Id: Id5a8a1b94f83da8060786f31181014dd1d546fc7
Signed-off-by: Benoit Goby <benoit@android.com>
Since these get allocated dmacoherent instead of noncacheable in armv7, we need
to do wmb before handing them to hardware.
Change-Id: I413eeb6da3bfeb754c4b475b19fe5823d83d3d04
Signed-off-by: Colin Cross <ccross@android.com>
Cc: Erik Gilling <konkers@android.com>
Based on work by Gary King.
Further abstraction of the chipidea core support needs to be done.
Signed-off-by: Colin Cross <ccross@android.com>
Cc: Erik Gilling <konkers@android.com>
Cc: Gary King <GKing@nvidia.com>
Remove USB_CONFIG_ATT_WAKEUP from bmAttributes as we don't
support remote wakeup. This fixes an issue with the USB 2.0
compliance test tool.
Change-Id: Ic15e44710069f3f05ace6a82bbfa2b85a3649027
Signed-off-by: Benoit Goby <benoit@android.com>
For android builds we disable the check for curlun->prevent_medium_removal.
Instead we let the framework manage unmounting policy, as we sometimes need
to unmount after the media has been removed.
This also helps support hosts that do not inform the device when the media
has been unmounted.
Signed-off-by: Mike Lockwood <lockwood@android.com>
Use this rather than calling usb_gadget_disconnect and usb_gadget_connect
directly to avoid sending USB disconnect events to userspace when resetting
the bus.
Signed-off-by: Mike Lockwood <lockwood@android.com>
Add switch to notify current USB configuration. This can be used to detect
USB connect and disconnect events.
Broadcast a change via the usb_composite class when a USB function is
enabled or disabled.
Rename usb_function.hidden to usb_function.disabled.
Signed-off-by: Mike Lockwood <lockwood@android.com>
Integrate support for android composite driver and platform data
within CONFIG_USB_ANDROID_MASS_STORAGE.
Signed-off-by: Mike Lockwood <lockwood@android.com>
As part of formating the mass storage device, Host sends the INQUIRY
scsi command. As per the standard, the command length for this
command should be 6 bytes, whereas the Vista host sends 12 bytes.
When the command length of the command is not equal to the standard
length, the device sends a phase error as part of the status phase.
When the host receives a phase error, it re-enumerates, hence the
error.
If the command is INQUIRY, and the command length is 12 bytes,
treating this as a good command and not sending the phase error
to the host fixes this issue.
Signed-off-by: Velempati Chiranjeevi <c_cvelem@qualcomm.com>
Signed-off-by: Mike Lockwood <lockwood@android.com>
Pulling in some code from file_storage.c, we now handle interface changes
in do_set_config(), which is now not called until any pending requests have
been successfully completed or cancelled.
This fixes a race condition that resulted in usb_ep_free_request() being called
while the request is still busy.
Signed-off-by: Mike Lockwood <lockwood@android.com>
The old code did not allways disable interrupts when called from thread
context, but tried to lock the same spinlock from interrupt context.
This was merged from a change to drivers/usb/function/mass_storage.c
in the android-msm-2.6.29 branch.
Signed-off-by: Arve Hjønnevåg <arve@android.com>
Signed-off-by: Mike Lockwood <lockwood@android.com>
Nick Pelly