07 Jun, 2015
1 commit
-
DTO/DCRC errors were not being informed to the mmc core since
commit ae4bf788ee ("mmc: omap_hsmmc: consolidate error report handling of
iHSMMC IRQ"). This commit made sure 'end_trans' is never set on DTO/DCRC
errors. This is because after this commit 'host->data' is checked after
it has been cleared to NULL by omap_hsmmc_dma_cleanup().Because 'end_trans' is never set, omap_hsmmc_xfer_done() is never invoked
making core layer not to be aware of DTO/DCRC errors. Because of this
any command invoked after DTO/DCRC error leads to a hang.Fix this by checking for 'host->data' before it is actually cleared.
Fixes: ae4bf788ee ("mmc: omap_hsmmc: consolidate error report handling of
iHSMMC IRQ")Reported-by: Yan Liu
Tested-by: Yan Liu
Signed-off-by: Kishon Vijay Abraham I
Signed-off-by: Sekhar Nori
29 May, 2015
2 commits
-
We can't rely just on dr_mode to decide if we're in host or gadget
mode when we're configured as otg/dual-role. So while dr_mode is
OTG, we find out from the otg state machine if we're in host
or gadget mode and take the necessary actions during suspend/resume.Also make sure that we disable OTG irq and events during system suspend
so that we don't lockup the system during system suspend/resume.Known Issue:
when otg-host, usb device disconnects and reconnects after system suspend/resume.Signed-off-by: Roger Quadros
Signed-off-by: Sekhar Nori -
There is a race happening during dwc3_drd_init() that causes
otg events to get disabled. This is what happens.dwc3_otg_irq() happens immediately when PRTCAP is set to OTG,
even though OEVTEN is 0. This is because BIT 31 IRQ of
OEVT can't be disabled by OEVTEN.
We configure OEVTEN in dwc3_otg_init() but dwc3_otg_irq() has
already saved OEVTEN as 0 into dwc->oevten. So finally when
dwc3_irq_thread_irq() is called we save 0 into OEVTEN
thus disabling OTG irqs forever.We fix this by disabling IRQs when configuring OEVTEN in
dwc3_otg_init().eviewed-by: Felipe Balbi
Signed-off-by: Roger Quadros
Signed-off-by: Sekhar Nori
26 May, 2015
2 commits
-
commit 5fa63bd959 (usb: xhci: Fix suspend/resume when used
with OTG core) removes assigning xhci->main_hcd from xhci_gen_setup
and adds it in the probe of xhci-plat and xhci-pci.In the case of xhci-pci, xhci_mem_init is invoked before main_hcd is
initialized in the probe causing a null pointer deferencing error.Fix it by initializing xhci->main_hcd in xhci_gen_setup and removing
it from xhci_pci_probe().Fixes: 5fa63bd959cc ("usb: xhci: Fix suspend/resume when used with OTG core")
Acked-by: Roger Quadros
Signed-off-by: Kishon Vijay Abraham I
Signed-off-by: Sekhar Nori -
CONFIG_USB_OTG_WHITELIST is to be enabled for compliance to OTG
specification. However, on TI SoCs, we only support/test USB
dual-role device, not the full OTG specification.Moreover, with CONFIG_USB_OTG_WHITELIST enabled, on AM335x
beaglebone black, on connecting a USB webcam we get:[ 74.960459] musb-hdrc musb-hdrc.1.auto: otg: usb_otg_kick_fsm: invalid host/gadget device
[ 75.080228] usb 1-1: new high-speed USB device number 4 using musb-hdrc
[ 75.270151] usb 1-1: device v03f0 pa707 is not supported
[ 75.275670] hub 1-0:1.0: unable to enumerate USB device on port 1This is because kernel refuses to enumerate peripherals which
are not listed in kernel's OTG whitelist. This is needed for
full compliance with OTG specification.Disable CONFIG_USB_OTG_WHITELIST since we support only dual-role
device and do not need to reject non-whitelisted devices during
enumeration.Suggested-by: Roger Quadros
Signed-off-by: Sekhar Nori
18 May, 2015
3 commits
-
The VBUS line of USB2 is connected to VBUS detect logic on
the PMIC. Use the palmas-usb driver to report VBUS events
to the USB driver.As the palmas-usb driver supports GPIO based ID reporting
provide the GPIO for ID pin as well.Signed-off-by: Roger Quadros
Signed-off-by: Sekhar Nori -
Some palmas based chip variants do not have OTG based ID logic.
For these variants we rely on GPIO based USB ID detection.These chips do have VBUS comparator for VBUS detection so we
continue to use the old way of detecting VBUS.Signed-off-by: Roger Quadros
Signed-off-by: Sekhar Nori -
[ Upstream commit fa31409a82ee050e52caad9e4c483fe3edca163a ]
Using $(PWD) breaks builds when make was invoked from outside
of the kernel tree.Signed-off-by: Yegor Yefremov
Signed-off-by: Felipe Balbi
Signed-off-by: Jacob Stiffler
Signed-off-by: Sekhar Nori
14 May, 2015
2 commits
-
Multiple failures are seemed to be caused by 8250 DMA:
* board crash due to dma errors if it is suspended while crypto
operations are in progress
* edma unmanaged events errors caused nand write failures during
suspend
* board aborts suspend sequence due to late interrupt generated by
presumably uart dmaSo disable it to have a more stable system.
Signed-off-by: Carlos Hernandez
-
The extcon us gpio driver requests id and vbus gpios during
probe. This could fail with -EPROBE_DEFER in case the gpiochip
that has the specific gpio pin is not initialised. For eg: the
gpiochip could be an i2c client and is not probed at the time that
the extcon driver gets probed.Propagating the EPROBE_DEFER error lets the extcon driver probe
function to be called again until it succeeds or error out with
something other than EPROBE_DEFER.Acked-by: Roger Quadros
Signed-off-by: Subramaniam Chanderashekarapuram
[nsekhar@ti.com: move EPROBE_DEFER check inside IS_ERR() check]
Signed-off-by: Sekhar Nori
13 May, 2015
5 commits
-
When hardware and software timer are used simultaneously there
are instances where both will trigger timeout interrupt to MMC
core and also instances where any of them provide any timeout
to MMC core which results in MMC core kernel task to hung
infinitely.Instead of using software timeout only for larger delays requested
when using HS200/SDR104 cards which results in hardware and
software timer race conditions, so move all the timeout request
to use software timer when HS200/SDR104 card is connected and
use hardware timer when other type cards are connected.Also start the software timer after queueing to DMA to ensure
we are more likely to expire within correct limits. To be ever
more sure that we won't expire this soft timer too early, we're
adding a 1000000ns slack to the data timeout requested by the
upper layer.Fixes: d029ff6f ("mmc: host: omap_hsmmc: add software timer when timeout greater than hardware capablility")
Suggested-by: Felipe Balbi
Signed-off-by: Mugunthan V N
Signed-off-by: Sekhar Nori -
When SATA is used by u-boot and not powered down correctly *and*
if AHCI_PLATFORM is m, clock auto-disable will kick in
and disable sata_refclk thus causing Errata i783 and breaking SATA.Since we can't guarantee what's done in u-boot, we take a safe approach by
keeping the sata_refclk ON in pipe3 probe so that it doesn't get auto-disabled
by the kernel. We keep track of this with the sata_refclk_enabled flag
and do an additional disable in ti_pipe3_disable_refclk() when needed.Signed-off-by: Roger Quadros
Signed-off-by: Sekhar Nori -
The pcf compatible field for i2c client 1 at slave address 21 should
be nxp,pcf8575 and not ti,pcf8575. Fix the same.Signed-off-by: Subramaniam Chanderashekarapuram
Signed-off-by: Sekhar Nori -
We don't want leave any pending timers if we're
going into suspend or removing the MMC driver.Signed-off-by: Felipe Balbi
Signed-off-by: Mugunthan V N
Signed-off-by: Sekhar Nori -
that's a nicer way of using the timer API.
Signed-off-by: Felipe Balbi
Signed-off-by: Mugunthan V N
Signed-off-by: Sekhar Nori
12 May, 2015
3 commits
-
This print was left back by mistake. Get rid of it.
Fixes: 08f6d9796f02 ('phy: ti-pipe3: i783 workaround for SATA lockup after dpll unlock/relock')
Signed-off-by: Roger Quadros
Signed-off-by: Sekhar Nori -
gpio2_8 is connected to the PCIe_RESETn line and it has to be driven low
in order to reset the PCIe cards. So added gpios property to pcie
dt node and also set the flags to GPIO_ACTIVE_LOW.Signed-off-by: Kishon Vijay Abraham I
Signed-off-by: Sekhar Nori -
The PERST# line in am57x-evm is connected to a gpio line and PERST#
should be driven high to indicate the clocks are stable (As per
Figure 2-10: Power Up of the PCIe CEM spec 3.0).Add support in pci-dra7xx driver to make gpio drive PERST#
line here.Signed-off-by: Kishon Vijay Abraham I
Signed-off-by: Sekhar Nori
08 May, 2015
2 commits
-
SATA_PLL_SOFT_RESET bit of CTRL_CORE_SMA_SW_0 must be toggled
between a SATA DPLL unlock and re-lock to prevent SATA lockup.Introduce a new DT parameter 'syscon-pllreset' to provide the syscon
regmap access to this register which sits in the control module.If the register is not provided we fallback to the old behaviour
i.e. SATA DPLL refclk will not be disabled and we prevent SoC low
power states.NOTE: If SATA is enabled in u-boot will break if AHCI_PLATFORM is m
as kernel clock autodisable kicks in and u-boot hadn't disabled
SATA correctly. We can't do anything about this in the kernel
and u-boot must be fixed.Signed-off-by: Roger Quadros
Signed-off-by: Sekhar Nori -
This register is required to be passed to the SATA PHY
driver to workaround errata i783 - SATA Lockup After SATA DPLL Unlock/Relock.Signed-off-by: Roger Quadros
Signed-off-by: Sekhar Nori
05 May, 2015
20 commits
-
[ Upstream commit 90bdf403db4ca75ee60ca5a760dc342cef96b5e1 ]
The driver was recently adapted to a core API change, but the
change was incomplete, missing out the suspend helper and
leaving an extraneous local variable around:usb/phy/phy-mv-usb.c: In function 'mv_otg_update_state':
usb/phy/phy-mv-usb.c:341:18: warning: unused variable 'phy' [-Wunused-variable]usb/phy/phy-mv-usb.c: In function 'mv_otg_suspend':
usb/phy/phy-mv-usb.c:861:16: error: 'struct usb_phy' has no member named 'state'Signed-off-by: Arnd Bergmann
Fixes: e47d92545c297 ("usb: move the OTG state from the USB PHY to the OTG structure")
Signed-off-by: Felipe Balbi
Signed-off-by: Sekhar Nori -
The rx_running flag should show if DMA is currently active. However
there is a window between when the flag is set/cleared and when
the DMA is started/stopped. Because the flag is queried from both
hard and soft irq contexts, the driver can make incorrect
decisions and do things like start a DMA transfer using a buffer
that is already setup to be used for a DMA transfer.This patch adds a spinlock to synchronize the rx_running flag and
close the above mentioned window.Signed-off-by: John Ogness
Signed-off-by: Sekhar Nori -
[ Upstream commit d560fed6abe0f9975b509e4fb824e08ac19adc93 ]
Required for moving drivers to the nanosecond based interfaces.
Signed-off-by: Thomas Gleixner
Signed-off-by: John Stultz
Signed-off-by: Felipe Balbi
Signed-off-by: Sekhar Nori -
We support dual-role via CONFIG_USB_OTG so enable it.
Switch from EXTCON_GPIO to EXTCON_USB_GPIO as we've moved
to using the new driver for USB.Signed-off-by: Roger Quadros
Signed-off-by: Sekhar Nori -
commit 33b23a79d28c (arm: boot: dts: am437x-sk: add missing USB pins)
mistakenly added a wrong pinctrl address for usb2_phy2_pins.
This looks like a copy paste error as it is the address of usb2_phy1_pins.Signed-off-by: Roger Quadros
Signed-off-by: Sekhar Nori -
USB1 port is micro-AB type and can function as peripheral
as well as host. Enable dual-role mode for USB1.Signed-off-by: Roger Quadros
Signed-off-by: Sekhar Nori -
Gdget controller might not be always active during suspend/
resume when we are operating in dual-role/otg mode.
Check if we're active and only if we are then perform
necessary actions during suspend/resume.Signed-off-by: Roger Quadros
Signed-off-by: Sekhar Nori -
Without this we loose OTG controller register context and malfunction
after a system suspend-resume.Signed-off-by: Roger Quadros
Signed-off-by: Sekhar Nori -
If the ID pin event is not available over extcon
then we rely on the OTG controller to provide us ID and VBUS
information.We still don't support any OTG features but just
dual-role operation.Signed-off-by: Roger Quadros
Signed-off-by: Sekhar Nori -
Add OTG and GHWPARAMS6 register definitions
Signed-off-by: Roger Quadros
Signed-off-by: Sekhar Nori -
Now that we have dual-role support working at USB core,
enable dual-role support for usb1 controller.Signed-off-by: Roger Quadros
Signed-off-by: Sekhar Nori -
Register with the USB OTG core. Since we don't support
OTG yet we just work as a dual-role device even
if device tree says "otg".Use extcon framework to get VBUS/ID cable events and
kick the OTG state machine.Signed-off-by: Roger Quadros
Signed-off-by: Sekhar Nori -
This is the a_set_b_hnp_enable flag in the OTG state machine
diagram and must be set when the A-Host has successfully set
the b_hnp_enable feature of the OTG-B-Peripheral attached to it.When this bit changes we kick our OTG FSM to make note of the
change and act accordingly.Signed-off-by: Roger Quadros
Signed-off-by: Sekhar Nori -
DRD mode is a reduced functionality OTG mode. In this mode
we don't support SRP, HNP and dynamic role-swap.In DRD operation, the controller mode (Host or Peripheral)
is decided based on the ID pin status. Once a cable plug (Type-A
or Type-B) is attached the controller selects the state
and doesn't change till the cable in unplugged and a different
cable type is inserted.As we don't need most of the complex OTG states and OTG timers
we implement a lean DRD state machine in usb-otg.c.
The DRD state machine is only interested in 2 hardware inputs
'id' and 'vbus; that are still passed via the origintal struct otg_fsm.Most of the usb-otg.c functionality remains the same except
adding a new parameter to usb_otg_register() to indicate that
the OTG controller needs to operate in DRD mode.Signed-off-by: Roger Quadros
Signed-off-by: Sekhar Nori -
we need to unlock the usb_lock mutex before calling
usb_otg_register_gadget() else it will cause a
circular locking dependency.Signed-off-by: Roger Quadros
Signed-off-by: Sekhar Nori -
The OTG state machine needs a mechanism to start and
stop the gadget controller. Add usb_gadget_start()
and usb_gadget_stop().Register with OTG core when gadget function driver
is available and unregister when function driver is unbound.Ignore softconnect sysfs control when we're in OTG
mode as OTG FSM takes care of gadget softconnect using
the b_bus_req mechanism.Signed-off-by: Roger Quadros
[nsekhar@ti.com: sparse warning fix]
Signed-off-by: Sekhar Nori -
The existing usb_add/remove_hcd() functionality
remains unchanged for non-OTG devices. For OTG
devices they only register the HCD with the OTG core.Introduce usb_otg_add/remove_hcd() for use by OTG core.
These functions actually add/remove the HCD.Signed-off-by: Roger Quadros
Signed-off-by: Sekhar Nori -
The OTG core instantiates the OTG Finite State Machine
per OTG controller and manages starting/stopping the
host and gadget controllers based on the bus state.It provides APIs for the following tasks
- Registering an OTG capable controller
- Registering Host and Gadget controllers to OTG core
- Providing inputs to and kicking the OTG state machineTODO:
- sysfs interface to allow application inputs to OTG state machine
- otg class?Signed-off-by: Roger Quadros
[nsekhar@ti.com: remove unused if() block in usb_otg_work() and sparse fixes]
Signed-off-by: Sekhar Nori -
The OTG core will use struct otg_gadget_ops to
start/stop the gadget controller.The main purpose of this interface is to avoid directly
calling usb_gadget_start/stop() from the OTG core as they
wouldn't be defined in the built-in symbol table if
CONFIG_USB_GADGET is m.Signed-off-by: Roger Quadros
Signed-off-by: Sekhar Nori -
The OTG core will use struct otg_hcd_ops to
add/remove the HCD controller.The main purpose of this interface is to avoid directly
calling usb_add/remove_hcd() from the OTG core as they
wouldn't be defined in the built-in symbol table if
CONFIG_USB is m.Signed-off-by: Roger Quadros
Signed-off-by: Sekhar Nori
