This patch fixes a suspend/resume issue where the driver is blindly
calling ehci_suspend/resume functions when the ehci hasn't been setup.
This results in a crash during suspend/resume operations.
Signed-off-by: NAndy Gross <andy.gross@linaro.org>
Tested-by: NPramod Gurav <pramod.gurav@linaro.org>
Acked-by: NAlan Stern <stern@rowland.harvard.edu>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Submitters of device tree binding documentation may forget to CC
the subsystem maintainer if this is missing.
Signed-off-by: NGeert Uytterhoeven <geert@linux-m68k.org>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: linux-usb@vger.kernel.org
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Starting with commit 0b52297f ("reset: Add support for shared reset
controls") there is a reference count for reset control assertions. The
goal is to allow resets to be shared by multiple devices and an assert
will take effect only when all instances have asserted the reset.

In order to preserve backwards-compatibility, all reset controls become
exclusive by default. This is to ensure that reset_control_assert() can
immediately assert in hardware.

However, this new behaviour triggers the following warning in the EHCI
driver for Tegra:

[    3.365019] ------------[ cut here ]------------
[    3.369639] WARNING: CPU: 0 PID: 1 at drivers/reset/core.c:187 __of_reset_control_get+0x16c/0x23c
[    3.382151] Modules linked in:
[    3.385214] CPU: 0 PID: 1 Comm: swapper/0 Not tainted 4.6.0-rc6-next-20160503 #140
[    3.392769] Hardware name: NVIDIA Tegra SoC (Flattened Device Tree)
[    3.399046] [<c010fa50>] (unwind_backtrace) from [<c010b120>] (show_stack+0x10/0x14)
[    3.406787] [<c010b120>] (show_stack) from [<c0347dcc>] (dump_stack+0x90/0xa4)
[    3.414007] [<c0347dcc>] (dump_stack) from [<c011f4fc>] (__warn+0xe8/0x100)
[    3.420964] [<c011f4fc>] (__warn) from [<c011f5c4>] (warn_slowpath_null+0x20/0x28)
[    3.428525] [<c011f5c4>] (warn_slowpath_null) from [<c03cc8cc>] (__of_reset_control_get+0x16c/0x23c)
[    3.437648] [<c03cc8cc>] (__of_reset_control_get) from [<c0526858>] (tegra_ehci_probe+0x394/0x518)
[    3.446600] [<c0526858>] (tegra_ehci_probe) from [<c04516d8>] (platform_drv_probe+0x4c/0xb0)
[    3.455029] [<c04516d8>] (platform_drv_probe) from [<c044fe78>] (driver_probe_device+0x1ec/0x330)
[    3.463892] [<c044fe78>] (driver_probe_device) from [<c0450074>] (__driver_attach+0xb8/0xbc)
[    3.472320] [<c0450074>] (__driver_attach) from [<c044e1ec>] (bus_for_each_dev+0x68/0x9c)
[    3.480489] [<c044e1ec>] (bus_for_each_dev) from [<c044f338>] (bus_add_driver+0x1a0/0x218)
[    3.488743] [<c044f338>] (bus_add_driver) from [<c0450768>] (driver_register+0x78/0xf8)
[    3.496738] [<c0450768>] (driver_register) from [<c010178c>] (do_one_initcall+0x40/0x170)
[    3.504909] [<c010178c>] (do_one_initcall) from [<c0c00ddc>] (kernel_init_freeable+0x158/0x1f8)
[    3.513600] [<c0c00ddc>] (kernel_init_freeable) from [<c0810784>] (kernel_init+0x8/0x114)
[    3.521770] [<c0810784>] (kernel_init) from [<c0107778>] (ret_from_fork+0x14/0x3c)
[    3.529361] ---[ end trace 4bda87dbe4ecef8a ]---

The reason is that Tegra SoCs have three EHCI controllers, each with a
separate reset line. However the first controller contains UTMI pads
configuration registers that are shared with its siblings and that are
reset as part of the first controller's reset. There is special code in
the driver to assert and deassert this shared reset at probe time, and
it does so irrespective of which controller is probed first to ensure
that these shared registers are reset before any of the controllers are
initialized. Unfortunately this means that if the first controller gets
probed first, it will request its own reset line and will subsequently
request the same reset line again (temporarily) to perform the reset.
This used to work fine before the above-mentioned commit, but now
triggers the new WARN.

Work around this by making sure we reuse the controller's reset if the
controller happens to be the first controller.

Cc: Philipp Zabel <p.zabel@pengutronix.de>
Cc: Hans de Goede <hdegoede@redhat.com>
Reviewed-by: NPhilipp Zabel <p.zabel@pengutronix.de>
Signed-off-by: NThierry Reding <treding@nvidia.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

There are three EHCI controllers on Tegra SoCs, each with its own reset
line. However, the first controller contains a set of UTMI configuration
registers that are shared with its siblings. These registers will only
be reset as part of the first controller's reset. For proper operation
it must be ensured that the UTMI configuration registers are reset
before any of the EHCI controllers are enabled, irrespective of the
probe order.

Commit a47cc24c ("USB: EHCI: tegra: Fix probe order issue leading to
broken USB") introduced code that ensures the first controller is always
reset before setting up any of the controllers, and is never again reset
afterwards.

This code, however, grabs the wrong reset. Each EHCI controller has two
reset controls attached: 1) the USB controller reset and 2) the UTMI
pads reset (really the first controller's reset). In order to reset the
UTMI pads registers the code must grab the second reset, but instead it
grabbing the first.

Fixes: a47cc24c ("USB: EHCI: tegra: Fix probe order issue leading to broken USB")
Acked-by: NJon Hunter <jonathanh@nvidia.com>
Cc: stable@vger.kernel.org
Signed-off-by: NThierry Reding <treding@nvidia.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

parport subsystem has introduced parport_del_port() to delete a port
when it is going away. Without parport_del_port() the registered port
will not be unregistered.
To reproduce and verify the error:
Command to be used is : ls /sys/bus/parport/devices
1) without the device attached there is no output as there is no
registered parport.
2) Attach the device, and the command will show "parport0".
3) Remove the device and the command still shows "parport0".
4) Attach the device again and we get "parport1".

With the patch applied:
1) without the device attached there is no output as there is no
registered parport.
2) Attach the device, and the command will show "parport0".
3) Remove the device and there is no output as "parport0" is now
removed.
4) Attach device again to get "parport0" again.

Cc: <stable@vger.kernel.org> # 4.2+
Signed-off-by: NSudip Mukherjee <sudip.mukherjee@codethink.co.uk>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Since ed_schedule begins with marking the ED as "operational",
the ED may be left in such state even if scheduling actually
fails.

This allows future submission attempts to smuggle this ED to the
hardware behind the scheduler's back and without linking it to
the ohci->eds_in_use list.

The former causes bandwidth saturation and data loss on isoc
endpoints, the latter crashes the kernel when attempt is made
to unlink such ED from this list.

Fix ed_schedule to update ED state only on successful return.
Signed-off-by: NMichal Pecio <michal.pecio@gmail.com>
Acked-by: NAlan Stern <stern@rowland.harvard.edu>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Ensure that the endpoint is stopped by clearing REQPKT before
clearing DATAERR_NAKTIMEOUT before rotating the queue on the
dedicated bulk endpoint.
This addresses an issue where a race could result in the endpoint
receiving data before it was reprogrammed resulting in a warning
about such data from musb_rx_reinit before it was thrown away.
The data thrown away was a valid packet that had been correctly
ACKed which meant the host and device got out of sync.
Signed-off-by: NAndrew Goodbody <andrew.goodbody@cambrionix.com>
Cc: stable@vger.kernel.org
Signed-off-by: NBin Liu <b-liu@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

shared_fifo endpoints would only get a previous tx state cleared
out, the rx state was only cleared for non shared_fifo endpoints
Change this so that the rx state is cleared for all endpoints.
This addresses an issue that resulted in rx packets being dropped
silently.
Signed-off-by: NAndrew Goodbody <andrew.goodbody@cambrionix.com>
Cc: stable@vger.kernel.org
Signed-off-by: NBin Liu <b-liu@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Incorrect cppi dma channel is referenced in musb_rx_dma_iso_cppi41(),
which causes kernel NULL pointer reference oops later when calling
cppi41_dma_channel_program().

Fixes: 069a3fd1 (usb: musb: Remove ifdefs for musb_host_rx in musb_host.c
part1)

Cc: <stable@vger.kernel.org> # v4.2+
Reported-by: NMatwey V. Kornilov <matwey@sai.msu.ru>
Acked-by: NTony Lindgren <tony@atomide.com>
Signed-off-by: NBin Liu <b-liu@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

If the session bit was not set in the backup of devctl register,
restoring devctl would clear the session bit. Therefor, only restore
devctl register when the session bit was set in the backup.

This solves the device enumeration failure in otg mode exposed by commit
56f487c7 (PM / Runtime: Update last_busy in rpm_resume).

Cc: <stable@vger.kernel.org> # v4.2+
Signed-off-by: NBin Liu <b-liu@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

We need to check the state for the PHY with delayed_work
as otherwise MUSB will get confused and idles immediately.
Signed-off-by: NTony Lindgren <tony@atomide.com>
Signed-off-by: NBin Liu <b-liu@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

There's no longer any need for custom initcall level for
musb.
Signed-off-by: NTony Lindgren <tony@atomide.com>
Signed-off-by: NBin Liu <b-liu@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

With the pull up being handled with delayed work, we can
now finally remove pm_runtime_set_irq_safe that blocks the
MUSB glue from idling.
Signed-off-by: NTony Lindgren <tony@atomide.com>
Signed-off-by: NBin Liu <b-liu@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

With PM runtime behaving, these are all now unnecessary.
Doing pm_runtime_get(musb->controller) will keep the parent
glue layer also active.
Signed-off-by: NTony Lindgren <tony@atomide.com>
Signed-off-by: NBin Liu <b-liu@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

At least on n900 we have phy-twl4030-usb only generating cable
interrupts, and then have a separate USB PHY.

In order for musb to know the real cable status, we need to
clear any cached state until musb is ready. Otherwise the cable
status interrupts will get just ignored if the status does
not change from the initial state.

To do this, let's add a return value to musb_mailbox(), and
reset cached linkstat to MUSB_UNKNOWN on error. Sorry to cause
a bit of churn here, I should have added that already last time
patching musb_mailbox().
Signed-off-by: NTony Lindgren <tony@atomide.com>
Signed-off-by: NBin Liu <b-liu@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

At least 2430 glue layer pulls d+ high on start up even if there are
no gadgets configured. This is bad at least for anything using a separate
battery charger chip as it can confuse the charger detection.

Let's fix the issue by removing the bogus glue layer code tinkering
with the SESSION bit. As pointed out Bin Liu <b-liu@ti.com> and
Sergei Shtylyov <sergei.shtylyov@cogentembedded.com>, the SESSION
bit just starts host mode if ID pin is grounded, and starts the
srp is ID pin is floating. So without the ID pin changing, it's
unable to force musb mode to anything. And just for starting a
host mode, things work fine without this code.
Signed-off-by: NTony Lindgren <tony@atomide.com>
Signed-off-by: NBin Liu <b-liu@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

This is no longer needed with PM runtime at least for 2430 glue.
We can now rely only on PM runtime and cable detection.

The other glue layers can probably remove try_idle too, but that
needs to be tested for each platform before doing it.
Signed-off-by: NTony Lindgren <tony@atomide.com>
Signed-off-by: NBin Liu <b-liu@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

This simplifies things and allows idling both MUSB and PHY
when nothing is configured. Let's just return early from PM
runtime if musb is not yet initialized.

Let's also warn if PHY is not configured.
Signed-off-by: NTony Lindgren <tony@atomide.com>
Signed-off-by: NBin Liu <b-liu@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

We may have drivers loaded but no configured gadgets and MUSB may be in
host mode. If gadgets are configured during host mode, PM runtime will
get confused.

Disable PM runtime from gadget state, and do it based on the cable
and last state.

Note that we may get multiple cable events, so we need to keep track
of the power state.
Signed-off-by: NTony Lindgren <tony@atomide.com>
Signed-off-by: NBin Liu <b-liu@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

We have MUSB setting pm_runtime_irq_safe with the following
commits:

30a70b02 ("usb: musb: fix obex in g_nokia.ko causing kernel panic")
3e43a072 ("usb: musb: core: add pm_runtime_irq_safe()")

Let's fix things to use delayed work so we can remove the
pm_runtime_irq_safe.

Note that we may want to set this up in a generic way in the
gadget framework eventually.
Signed-off-by: NTony Lindgren <tony@atomide.com>
Signed-off-by: NBin Liu <b-liu@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

The conditional use of PM runtime does not work properly
for musb gadget. On cable disconnect we may not get any
USB_EVENT_NONE leaving the PM runtime call unpaired.

Let's fix the issue by making sure the PM runtime calls are
paired within the functions. The glue layer will take care
of the rest.
Signed-off-by: NTony Lindgren <tony@atomide.com>
Signed-off-by: NBin Liu <b-liu@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Let's not tinker with the PM runtime of musb core from the omap2430
wrapper. This allows us to initialize PM runtime for musb core later
on instead of doing it in stages. And omap2430 wrapper has no need
to for accessing musb core at this point.

Note that this does not remove all the PM runtime calls from the
glue layer, those will get removed in a later patch.
Signed-off-by: NTony Lindgren <tony@atomide.com>
Signed-off-by: NBin Liu <b-liu@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Let's make the PM runtime use the standard autosuspend calls.
Commit 5de85b9d ("PM / runtime: Re-init runtime PM states at probe
error and driver unbind") means we must pair use_autosuspend with
dont_use_autosuspend and then use put_sync to properly idle the
device.

Note that we'll be removing the PM runtime calls from the glue
layer to the MUSB core in the next patch. And we can also remove
the pointless FIXME comment now.
Signed-off-by: NTony Lindgren <tony@atomide.com>
Signed-off-by: NBin Liu <b-liu@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

We have remove() already calling shutdown(), so let's drop it
and move the code to remove(). No code changes, we'll drop the
the FIXME in the following patch with more clean-up.
Signed-off-by: NTony Lindgren <tony@atomide.com>
Signed-off-by: NBin Liu <b-liu@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Looks like at least 2430 glue won't idle reliably with the 200 ms
autosuspend delay. This causes deeper idle states being blocked for
the whole SoC when disconnecting OTG A cable.

Increasing the delay to 500 ms seems to idle both MUSB and the PHY
reliably. This is probably because of time needed by the hardware
based negotiation between MUSB and the PHY.
Signed-off-by: NTony Lindgren <tony@atomide.com>
Signed-off-by: NBin Liu <b-liu@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

When the board is powering attached usb devices via the otg port
sometimes / on some devices it takes slightly too long for the Vbus
detection code in phy-sun4i-usb.c to signal that Vbus is high after
enabling Vbus and the musb hardware signals a MUSB_INTR_VBUSERROR
interrupt.

This commit sets the otg state to A_WAIT_VRISE upon enabling Vbus
making musb_stage0_irq() ignore the first VBUSERR_RETRY_COUNT
VBUSERROR interrupts, fixing connection issues in these cases.
Signed-off-by: NHans de Goede <hdegoede@redhat.com>
[b-liu@ti.com: revise subject prefix]
Signed-off-by: NBin Liu <b-liu@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Move the mode handling to the platform_set_mode callback.
Signed-off-by: NHans de Goede <hdegoede@redhat.com>
[b-liu@ti.com: revise subject prefix]
Signed-off-by: NBin Liu <b-liu@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Now that the DMA engine check was moved to musb_tx_dma_porgram(), both
musb_tx_dma_set_mode_cppi_tusb() and musb_tx_dma_set_mode_mentor() always
return 0, so we can  make both these functions *void*.
Signed-off-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
[b-liu@ti.com: revise subject prefix]
Signed-off-by: NBin Liu <b-liu@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Commit 754fe4a9 ("usb: musb: Remove ifdefs for TX DMA for musb_host.c")
looks incomplete: the DMA engine checks are  done outside the Mentor/UX500
handler  but inside the CPPI/TUSB handler. Move the checks out of the CPPI/
TUSB handler into its caller, musb_tx_dma_program().
Signed-off-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
[b-liu@ti.com: revise subject prefix]
Signed-off-by: NBin Liu <b-liu@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

urb->status is set when endpoint csr RXSTALL, H_ERROR, DATAERROR or
INCOMPRX bit is set. Those bits mean a broken pipe, so don't start next
urb when any of these bits is set by checking urb->status.

To minimize the risk of regression, only do so for RX, until we have a
test case to understand the behavior of TX.

The patch fixes system freeze issue caused by repeatedly invoking RX ISR
while removing a usb uart device connected to a hub, in which case the
hub has no chance to report the disconnect event due to the kernel is
busy in processing the RX interrupt flooding.

Fix checkpatch complaint (qh != NULL) as while.
Reported-by: NMax Uvarov <muvarov@gmail.com>
Tested-by: NYegor Yefremov <yegorslists@googlemail.com>
Signed-off-by: NBin Liu <b-liu@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

The MUSB Programming Guide states that the driver should clear RXCSR
bit2 when the controller sets the bit.
Signed-off-by: NBin Liu <b-liu@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

I got one of these cards for testing uas with, it seems that with streams
it dma-s all over the place, corrupting memory. On my first tests it
managed to dma over the BIOS of the motherboard somehow and completely
bricked it.

Tests on another motherboard show that it does work with streams disabled.

Cc: stable@vger.kernel.org
Signed-off-by: NHans de Goede <hdegoede@redhat.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Several people have reported that UBSAN doesn't like the pointer
arithmetic in ehci_hub_control():

	u32 __iomem	*status_reg = &ehci->regs->port_status[
				(wIndex & 0xff) - 1];
	u32 __iomem	*hostpc_reg = &ehci->regs->hostpc[(wIndex & 0xff) - 1];

If wIndex is 0 (and it often is), these calculations underflow and
UBSAN complains.

According to the C standard, pointer computations leading to locations
outside the bounds of an array object (other than 1 position past the
end) are undefined.  In this case, the compiler would be justified in
concluding the wIndex can never be 0 and then optimizing away the
tests for !wIndex that occur later in the subroutine.  (Although,
since ehci->regs->port_status and ehci->regs->hostpc are both 0-length
arrays and are thus GCC extensions to the C standard, it's not clear
what the compiler is really allowed to do.)

At any rate, we can avoid all these difficulties, at the cost of
making the code slightly longer, by not decrementing the index when it
is equal to 0.  The runtime effect is minimal, and anyway
ehci_hub_control() is not on a hot path.
Signed-off-by: NAlan Stern <stern@rowland.harvard.edu>
Reported-by: NValdis Kletnieks <Valdis.Kletnieks@vt.edu>
Reported-by: NMeelis Roos <mroos@linux.ee>
Reported-by: NMartin_MOKREJÅ <mmokrejs@gmail.com>
Reported-by: N"Navin P.S" <navinp1912@gmail.com>
CC: Andrey Ryabinin <ryabinin.a.a@gmail.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Commit 198de51d ("USB: uas: Limit qdepth at the scsi-host level")
removed the scsi_change_queue_depth() call from uas_slave_configure()
assuming that the slave would inherit the host's queue_depth, which
that commit sets to the same value.

This is incorrect, without the scsi_change_queue_depth() call the slave's
queue_depth defaults to 1, introducing a performance regression.

This commit restores the call, fixing the performance regression.

Cc: stable@vger.kernel.org
Fixes: 198de51d ("USB: uas: Limit qdepth at the scsi-host level")
Reported-by: NTom Yan <tom.ty89@gmail.com>
Signed-off-by: NHans de Goede <hdegoede@redhat.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Commit 198de51d ("USB: uas: Limit qdepth at the scsi-host level") made
qdepth limit set in host template (`.can_queue = MAX_CMNDS`) redundant.
Removing it to avoid confusion.
Signed-off-by: NTom Yan <tom.ty89@gmail.com>
Acked-by: NHans de Goede <hdegoede@redhat.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

The Acer C120 LED Projector is a USB-3 connected pico projector which
takes both its power and video data from USB-3.

In combination with some hubs this device does not play well with
lpm, so disable lpm for it.
Signed-off-by: NHans de Goede <hdegoede@redhat.com>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Properly sort all the entries by vendor id.
Signed-off-by: NHans de Goede <hdegoede@redhat.com>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

commit b1c127ae ("usb: host: xhci: plat: make use of new methods in
xhci_plat_priv") sets xhci->quirks before calling xhci_gen_setup(), which
will overwrite them.

Don't overwite the quirks, just add the new ones

Fixes: b1c127ae ("usb: host: xhci: plat: make use of new methods in xhci_plat_priv")
Reported-by: NYoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Cc: Felipe Balbi <felipe.balbi@linux.intel.com>
Signed-off-by: NMathias Nyman <mathias.nyman@linux.intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

On some platforms, the clocks might be registered by a platform
driver. When this is the case, the clock platform driver may very well
be probed after xhci-plat, in which case the first probe() invocation
of xhci-plat will receive -EPROBE_DEFER as the return value of
devm_clk_get().

The current code handles that as a normal error, and simply assumes
that this means that the system doesn't have a clock for the XHCI
controller, and continues probing without calling
clk_prepare_enable(). Unfortunately, this doesn't work on systems
where the XHCI controller does have a clock, but that clock is
provided by another platform driver. In order to fix this situation,
we handle the -EPROBE_DEFER error condition specially, and abort the
XHCI controller probe(). It will be retried later automatically, the
clock will be available, devm_clk_get() will succeed, and the probe()
will continue with the clock prepared and enabled as expected.

In practice, such issue is seen on the ARM64 Marvell 7K/8K platform,
where the clocks are registered by a platform driver.

Cc: <stable@vger.kernel.org>
Signed-off-by: NThomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: NMathias Nyman <mathias.nyman@linux.intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

If commands timeout we mark them for abortion, then stop the command
ring, and turn the commands to no-ops and finally restart the command
ring.

If the host is working properly the no-op commands will finish and
pending completions are called.
If we notice the host is failing, driver clears the command ring and
completes, deletes and frees all pending commands.

There are two separate cases reported where host is believed to work
properly but is not. In the first case we successfully stop the ring
but no abort or stop command ring event is ever sent and host locks up.

The second case is if a host is removed, command times out and driver
believes the ring is stopped, and assumes it will be restarted, but
actually ends up timing out on the same command forever.
If one of the pending commands has the xhci->mutex held it will block
xhci_stop() in the remove codepath which otherwise would cleanup pending
commands.

Add a check that clears all pending commands in case host is removed,
or we are stuck timing out on the same command. Also restart the
command timeout timer when stopping the command ring to ensure we
recive an ring stop/abort event.

Cc: stable <stable@vger.kernel.org>
Tested-by: NJoe Lawrence <joe.lawrence@stratus.com>
Signed-off-by: NMathias Nyman <mathias.nyman@linux.intel.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>