cascardo/linux.git
7 years agogreybus: hd/es2: add cport_clear callback and ARPC
Johan Hovold [Wed, 10 Aug 2016 10:58:44 +0000 (12:58 +0200)]
greybus: hd/es2: add cport_clear callback and ARPC

Add a host-device cport_clear callback, which will be called as part
of the new connection tear-down sequence to reset the CPort state.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Acked-by: Sandeep Patil <sspatil@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: hd/es2: add cport_quiesce callback and ARPC
Johan Hovold [Wed, 10 Aug 2016 10:58:43 +0000 (12:58 +0200)]
greybus: hd/es2: add cport_quiesce callback and ARPC

Add a host-device cport_quiesce callback, which will be called as part
of the new connection tear-down sequence to disable flow control after
first making sure that enough peer buffer space is available for the
next messages about to be sent.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Acked-by: Sandeep Patil <sspatil@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: hd/es2: add cport_connected callback and ARPC
Johan Hovold [Wed, 10 Aug 2016 10:58:42 +0000 (12:58 +0200)]
greybus: hd/es2: add cport_connected callback and ARPC

Add a host-device cport_connected callback, which will be called after a
connection has been created and that can be used by the host-device
driver to make sure its internal state is updated to match the CPort
attributes set by the SVC.

This callback will eventually replace the cport_features_enable
callback.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Acked-by: Sandeep Patil <sspatil@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: fix up copyright dates on arpc.h
Greg Kroah-Hartman [Wed, 10 Aug 2016 15:02:05 +0000 (17:02 +0200)]
greybus: fix up copyright dates on arpc.h

I got them wrong, Johan was right, my fault.

Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: lights: enable multi color LED support
Kris Huang [Tue, 9 Aug 2016 03:28:37 +0000 (11:28 +0800)]
greybus: lights: enable multi color LED support

A backport (commit 79c4de08c0e5a26b04a4ac9e6543dad6379f0b40) was
applied in kernel which adding attribute-group support in led-class.
Remove the LED_HAVE_GROUPS flag check entirely that allow led
drivers to create custom attributes like color/fade_in/fade_out.

Testing Done: Compiled and verified on EVT2 and gpbridge-test module
              with device class daughter board.

Signed-off-by: Kris Huang <huang_kris@projectara.com>
Reviewed-by: Rui Miguel Silva <rui.silva@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: greybus-driver: Add intf_oops operation
Georgi Dobrev [Tue, 9 Aug 2016 21:37:32 +0000 (14:37 -0700)]
greybus: greybus-driver: Add intf_oops operation

Add intf_oops operation to SVC Protocol. This
operation will notify the AP about a fatal error
in a module. The request has two arguments:
    -u8 intf - the interface in question
    -u8 reason - reason of the error
The response has no payload.
Upon receiving the Request, the driver disables
the Interface.

Signed-off-by: Georgi Dobrev <gdobrev@mm-sol.com>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Ashwin Chaugule <ashwinch@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: camera: Remove reference to ara subdevice
Jacopo Mondi [Wed, 10 Aug 2016 07:08:19 +0000 (09:08 +0200)]
greybus: camera: Remove reference to ara subdevice

Remove last occurrence of "ara" term in camera driver.
Replace reference to "ara subdevice" with "gmp subdevice"

Signed-off-by: Jacopo Mondi <jacopo.mondi@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: arpc: move arpc definitions to their own header
Johan Hovold [Wed, 10 Aug 2016 10:58:41 +0000 (12:58 +0200)]
greybus: arpc: move arpc definitions to their own header

Move the ARPC definitions to their own header.

ARPC is not part of greybus, but is rather an implementation-specific
means of communicating with a certain class of host-device hardware.

Note that the same is true for the APBA USB vendor requests, but we keep
them in the greybus header for the time being.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: es2: fix USB vendor-request prefixes
Johan Hovold [Wed, 10 Aug 2016 10:58:40 +0000 (12:58 +0200)]
greybus: es2: fix USB vendor-request prefixes

Make sure to use the common GB_APB prefix for all APBA USB vendor
requests.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: greybus_protocols: fix apba vendor-request comment
Johan Hovold [Wed, 10 Aug 2016 10:58:39 +0000 (12:58 +0200)]
greybus: greybus_protocols: fix apba vendor-request comment

Fix copy-paste error in an APBA USB vendor-request comment.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: sdio: increase maximum segment size
Rui Miguel Silva [Tue, 9 Aug 2016 22:35:35 +0000 (23:35 +0100)]
greybus: sdio: increase maximum segment size

Adjust maximum segment size of the sg list to meet the maximum request
size, this will allow to have less segments which increase the
performance of data movement during transfer operations.

Test Done: using mmc_test with best-case read/write performance and see
that only one segment is used and that the results are better (288KiB vs
300KiB in read operation).

Signed-off-by: Rui Miguel Silva <rui.silva@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: greybus_protocol: fix order of sdio get caps response
Rui Miguel Silva [Tue, 9 Aug 2016 22:35:34 +0000 (23:35 +0100)]
greybus: greybus_protocol: fix order of sdio get caps response

Order of the field of the sdio get caps operation response were wrong,
that influence later the frequencies used by core, during normal
operation.

Tested: verified that the values inserted by the fw are the correct ones
for the field.

Suggested-by: Jackson Chang <jacksonc@bsquare.com>
Signed-off-by: Rui Miguel Silva <rui.silva@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: interface: fix timesync registration sequencing
David Lin [Fri, 5 Aug 2016 22:08:40 +0000 (15:08 -0700)]
greybus: interface: fix timesync registration sequencing

When enabling an interface, control device should be registered after
having successfully added the timesync. Similarly for the interface
disable path, control device should be removed first before removing
timesync.

Testing Done:
 - Enable and disable the interface

Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Johan Hovold <johan@hovoldconsulting.com>
Suggested-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: David Lin <dtwlin@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: audio: Report jack removal along with module removal
Vaibhav Agarwal [Fri, 5 Aug 2016 12:46:30 +0000 (18:16 +0530)]
greybus: audio: Report jack removal along with module removal

For GB module with jack slot supported, headset/headphone can still
be inserted at the time of module removal. In this case, above layer is
unaware about jack removal event which happened due to module removal.
This may lead to inconsistent state in above HAL layer.  Fix this by
reporting jack removal event explicitly.

Signed-off-by: Vaibhav Agarwal <vaibhav.agarwal@linaro.org>
Reviewed-by: Mark Greer <mgreer@animalcreek.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: lights: Add runtime pm support
Kris Huang [Fri, 5 Aug 2016 04:59:08 +0000 (12:59 +0800)]
greybus: lights: Add runtime pm support

Modify Lights greybus driver to support runtime PM framework.
The suspend and resume function have been tested with gpbridge-test
image by sysfs. Lights functions work well on suspend/resume.

Testing Done: Compiled and verified on EVT2 and gpbridge-test module
              with device class daughter board.

Signed-off-by: Kris Huang <huang_kris@projectara.com>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: SDIO: Add runtime pm support
Jackson Chang [Fri, 5 Aug 2016 02:22:02 +0000 (10:22 +0800)]
greybus: SDIO: Add runtime pm support

Modify SDIO greybus driver to support runtime PM framework.
To enable SDIO runtime PM, it needs to remove MMC_CAP_NEEDS_POLL
and add MMC_CAP2_CORE_RUNTIME_PM in set_host_caps().
The suspend function and resume function have been tested
with micron-sdio image by sysfs. SDIO functions work well
on suspend/resume.

Testing Done: Compiled and verified on EVT2.0 + Micron ARA
              SD module with USB connector

Signed-off-by: Jackson Chang <chang_jackson@projectara.com>
Reviewed-by: Rui Miguel Silva <rui.silva@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: audio: Maintain module stream state for each data connection
Vaibhav Agarwal [Thu, 4 Aug 2016 09:44:39 +0000 (15:14 +0530)]
greybus: audio: Maintain module stream state for each data connection

For SPK module, each data connection corresponds to codec DAI. Now
stream state is maintained for each DAI. So, need to maintain stream
state for each DAI/data connection for individual module as well.

Signed-off-by: Vaibhav Agarwal <vaibhav.agarwal@linaro.org>
Reviewed-by: Mark Greer <mgreer@animalcreek.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: audio: Maintain runtime stream params for each DAI
Vaibhav Agarwal [Thu, 4 Aug 2016 09:44:38 +0000 (15:14 +0530)]
greybus: audio: Maintain runtime stream params for each DAI

Runtime streams are required while configuring GB module plugged-in
during active stream. Currently, it is maintained for single stream.
However, this should be maintained for a stream corresponding to each
DAI. Fix this!

Signed-off-by: Vaibhav Agarwal <vaibhav.agarwal@linaro.org>
Reviewed-by: Mark Greer <mgreer@animalcreek.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: audio: Update pm runtime support in dai_ops callback
Vaibhav Agarwal [Thu, 4 Aug 2016 09:44:37 +0000 (15:14 +0530)]
greybus: audio: Update pm runtime support in dai_ops callback

Ensure pm runtime get_sync/put protection in codec_dai ops callback
functions before accessing apbridge.

Signed-off-by: Vaibhav Agarwal <vaibhav.agarwal@linaro.org>
Reviewed-by: Mark Greer <mgreer@animalcreek.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: Remove unused field from data_connection
Vaibhav Agarwal [Thu, 4 Aug 2016 09:44:36 +0000 (15:14 +0530)]
greybus: Remove unused field from data_connection

Audio codec driver internally maintains a struct containing info about
module's data connection. Remove unused field from this struct.

Signed-off-by: Vaibhav Agarwal <vaibhav.agarwal@linaro.org>
Reviewed-by: Mark Greer <mgreer@animalcreek.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: audio: Find data connection based on id
Vaibhav Agarwal [Thu, 4 Aug 2016 09:44:35 +0000 (15:14 +0530)]
greybus: audio: Find data connection based on id

Currently we are using dai->name to identify data connection from list
for a module. Now since we are enabling data path based on widget,
dai->name might be invalid by the time driver receives disable request
for a widget. So, use id fetched from AIF widget->sname to identify data
connection for a module.

Signed-off-by: Vaibhav Agarwal <vaibhav.agarwal@linaro.org>
Reviewed-by: Mark Greer <mgreer@animalcreek.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: audio: Add id to identify data connection
Vaibhav Agarwal [Thu, 4 Aug 2016 09:44:34 +0000 (15:14 +0530)]
greybus: audio: Add id to identify data connection

Added id field to data connection struct. This is used to identify which
data connection to use while enabling interface between module & APB.

Signed-off-by: Vaibhav Agarwal <vaibhav.agarwal@linaro.org>
Reviewed-by: Mark Greer <mgreer@animalcreek.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: audio: Use AIF widget to enable path between module & APB
Vaibhav Agarwal [Thu, 4 Aug 2016 09:44:33 +0000 (15:14 +0530)]
greybus: audio: Use AIF widget to enable path between module & APB

Currently, SPK Amp switch is used to identify when to enable data path
between module and APB. With headset, other switch controls added, it is
not possible to use this switch to control data path. Instead path
should be established based on AIF widget.

Signed-off-by: Vaibhav Agarwal <vaibhav.agarwal@linaro.org>
Reviewed-by: Mark Greer <mgreer@animalcreek.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: audio: Update parameters for gbaudio_module_update API
Vaibhav Agarwal [Thu, 4 Aug 2016 09:44:32 +0000 (15:14 +0530)]
greybus: audio: Update parameters for gbaudio_module_update API

Earlier, module path was enabled based on module's control switch e.g.
'SPK Amp switch'.  Thus widget's name was sufficient to parse and
identify the direction. Now individual modules' path will be enabled
based on AIF widget status. So, it is required to get complete widget
details, say w->type is used to identify direction (playback/capture)
and w->sname is used to identify module's DATA connection used for
communication via greybus.

Signed-off-by: Vaibhav Agarwal <vaibhav.agarwal@linaro.org>
Reviewed-by: Mark Greer <mgreer@animalcreek.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: audio: Update dai_driver table with appropriate fields
Vaibhav Agarwal [Thu, 4 Aug 2016 09:44:31 +0000 (15:14 +0530)]
greybus: audio: Update dai_driver table with appropriate fields

Currently, the stream name for the DAI driver is generically set to "GB
Audio Playback" and "GB Audio Capture". This is OK since we use a single
interface on APB1 but that could change in the future.  Update the DAI
driver table entries to properly reflect the interface used.

Signed-off-by: Vaibhav Agarwal <vaibhav.agarwal@linaro.org>
Reviewed-by: Mark Greer <mgreer@animalcreek.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: audio: Split helper APIs based on stream direction
Vaibhav Agarwal [Thu, 4 Aug 2016 09:44:30 +0000 (15:14 +0530)]
greybus: audio: Split helper APIs based on stream direction

Now, module is enabled/disabled based on widget event only and not
during startup/shutdown callbacks. Thus, we needn't compare codec_state
in enable/disable module helper APIs. Also, these can be further
simplified based on stream direction.

Signed-off-by: Vaibhav Agarwal <vaibhav.agarwal@linaro.org>
Reviewed-by: Mark Greer <mgreer@animalcreek.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: audio: Enable audio path based on control switch state only
Vaibhav Agarwal [Thu, 4 Aug 2016 09:44:29 +0000 (15:14 +0530)]
greybus: audio: Enable audio path based on control switch state only

As per current implementation, audio data is played from each individual
SPK module connected to endo frame. This is not a valid requirement in
case of capture/headset path. So, provide a mechanism to enable
individual module path based on it's control switch state.

Signed-off-by: Vaibhav Agarwal <vaibhav.agarwal@linaro.org>
Reviewed-by: Mark Greer <mgreer@animalcreek.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: audio: Remove un-necessary goto statement
Vaibhav Agarwal [Thu, 4 Aug 2016 09:44:28 +0000 (15:14 +0530)]
greybus: audio: Remove un-necessary goto statement

For most of the helper functions, goto statement is widely used.  It was
originally used with an intent of single exit point for the function
with some cleanup required. This is no more the case. So, simplify code
by avoiding un-necessary gotos.

Signed-off-by: Vaibhav Agarwal <vaibhav.agarwal@linaro.org>
Reviewed-by: Mark Greer <mgreer@animalcreek.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: arche-platform: Add support for SPI bus sharing for Mihi
Vaibhav Hiremath [Thu, 28 Jul 2016 08:17:37 +0000 (13:47 +0530)]
greybus: arche-platform: Add support for SPI bus sharing for Mihi

In case of Mihi, SPI bus is shared between APB1 and APB2
SPI ROMs, so their FW flashing must be sequential and
arche-platform driver should make sure that they are mutual
exclusive in nature.

So this patch adds certain restrictions to the user of the
arche-platform driver,

 - User can no longer flash APB1 and APB2 SPI ROM in parallel
 - SPI bus becomes an resource, so user must claim it by moving
   respective APB device into FW_FLASHING mode and release it
   by exiting FW_FLASHING mode. User can exit FW_FLASHING mode by
   switching to any other modes (ACTIVE, OFF, STANDBY).
 - If APB1 is in FW_FLASHING mode, APB2 can no longer enter into
   FW_FLASHING mode. User will get -EBUSY.

Having said that, while APB1 is into FW_FLASHING mode,
APB2 can independently boot from its own SPI ROM.

Testing Done: Tested by simulating usecase on EVT2.
 - Made sure that APB1 and APB2 FW_FLASHING mode is mutual exclusive
   in nature. Confirmed that an attempt on second device return -EBUSY.
 - Added simulating code, where printed state of dummy gpio for
   spi-en and verified that it shows right pin status for both APBs

Signed-off-by: Vaibhav Hiremath <vaibhav.hiremath@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: arche-platform: Reset SVC & APB only if turned off by SUSPEND_PREPARE
Vaibhav Hiremath [Wed, 3 Aug 2016 11:45:38 +0000 (17:15 +0530)]
greybus: arche-platform: Reset SVC & APB only if turned off by SUSPEND_PREPARE

There is possible race condition in arche platform driver for
maintaining of ->state field, when shared with timesync driver.

Assume device booted fine, all connected modules have been enumerated
correctly. As part of suspend-resume operation we have pm_notifier
callback, where we turn off SVC & APB and coldboot on resume.
In the process of resume, all modules gets enumerated again, and
timesync driver does come into picture everytime.
So when timesync driver requests arche-platform to change the state
to TIMESYNC and in the middle of sync operation, if suspend gets
triggered, then execution lands into arche-platform->pm_notifier callback,
leading to race condition in the driver, where it checks for (state != ACTIVE)
in PM_SUSPEND_PREPARE and returns, but in PM_POST_SUSPEND it just simply
coldboots SVC & APB, which would inbalance all resources (including IRQ).

So we need a add check in PM_POST_SUSPEND, to make sure that, we only
coldboot devices if they are in to off state,
i.e. if (state != OFF) then return.

Testing Done: Done regressive suspend/resume testing on EVT2 platform.
Note that, I some time hit issue.

Signed-off-by: Vaibhav Hiremath <vaibhav.hiremath@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: es2: rename USB vendor-request timeout define
Johan Hovold [Wed, 3 Aug 2016 12:09:35 +0000 (14:09 +0200)]
greybus: es2: rename USB vendor-request timeout define

Give the USB vendor-request timeout define a more descriptive name.

Also drop the since-long obsolete comments about allowing "the SVC to do
something" and "SVC messages go down our control pipe".

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: es2: add define for ARPC CPort requests
Johan Hovold [Wed, 3 Aug 2016 12:09:34 +0000 (14:09 +0200)]
greybus: es2: add define for ARPC CPort requests

Add dedicated define for ARPC CPort requests instead of using the
default timeout for USB vendor requests.

We still allow responses to take 500 ms to arrive, but note that this
adds on top of the 500ms already allowed for a requests to be
acknowledged. This should probably be tightened up at some point.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: es2: allow ARPC result to be ignored
Johan Hovold [Wed, 3 Aug 2016 12:09:33 +0000 (14:09 +0200)]
greybus: es2: allow ARPC result to be ignored

Allow user of ARPC to pass a NULL pointer for the ARPC result.

For simple requests there may only be one error code for remote errors,
or the caller may simply not care to differentiate them.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: es2: always set result value
Johan Hovold [Wed, 3 Aug 2016 12:09:32 +0000 (14:09 +0200)]
greybus: es2: always set result value

Make sure to always set the result value for ARPC instead of forcing
every caller to do it in order to avoid compiler warnings.

The ARPC result should still be ignored unless arpc_sync returns
-EREMOTEIO.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: es2: clean up ARPC symbol names
Johan Hovold [Wed, 3 Aug 2016 12:09:31 +0000 (14:09 +0200)]
greybus: es2: clean up ARPC symbol names

Add a _req suffix to request message structures, and a _TYPE_ infix to
request type defines.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: connection: fix offloaded-connection ping cport id
Johan Hovold [Wed, 3 Aug 2016 12:09:30 +0000 (14:09 +0200)]
greybus: connection: fix offloaded-connection ping cport id

The host-device cport_ping callback expects the AP side cport id, but
was incorrectly passed the interface cport id.

Note that no host-device driver currently implements this callback, and
that it is soon even to be removed.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: control: remove some braces
Johan Hovold [Wed, 3 Aug 2016 12:09:29 +0000 (14:09 +0200)]
greybus: control: remove some braces

Remove some no-longer-needed braces around an error path.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: remove unused protocol-version messages
Johan Hovold [Wed, 3 Aug 2016 12:09:28 +0000 (14:09 +0200)]
greybus: remove unused protocol-version messages

Remove the unused protocol-version messages.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: audio: Remove GET_TX/RX_DELAY message types
Mark Greer [Wed, 3 Aug 2016 03:30:28 +0000 (20:30 -0700)]
greybus: audio: Remove GET_TX/RX_DELAY message types

The 'GB_AUDIO_TYPE_GET_TX_DELAY' and 'GB_AUDIO_TYPE_GET_RX_DELAY'
are no longer a part of the Greybus Audio Device Class Protocol
so remove support for them.  The message numbers are not coalesced
to prevent compatibility issues between the AP and the module.

Testing Done: Played music using a speaker module

Signed-off-by: Mark Greer <mgreer@animalcreek.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: audio: apbridgea: Remove GET_TX/RX_DELAY message types
Mark Greer [Wed, 3 Aug 2016 03:36:07 +0000 (20:36 -0700)]
greybus: audio: apbridgea: Remove GET_TX/RX_DELAY message types

The 'AUDIO_APBRIDGEA_TYPE_GET_TX_DELAY' and
'AUDIO_APBRIDGEA_TYPE_GET_RX_DELAY' message types have been
removed from the AP <-> APBrigdeA Audio Protocol so remove
them from the code.  Do not coalesce the message type numbers
to prevent compatibility issues between the AP and APBridgeA.

Testing Done: Played music using a speaker module

Signed-off-by: Mark Greer <mgreer@animalcreek.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: timesync: Printout strobe count on sync failure
Bryan O'Donoghue [Tue, 2 Aug 2016 12:18:29 +0000 (13:18 +0100)]
greybus: timesync: Printout strobe count on sync failure

If we failed to synchronize the FrameTime it would be useful to know how
many of the expected strobes arrived, for example a value of 0/5 would
indicate the SVC was completely dead but a value of 4/5 would indicate one
GPIO pulse got lost i.e. the AP was too slow reacting to an interrupt and
completely missed one of the strobe events. In either case the actual
number of strobes is a useful thing to print out.

Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Reviewed-by: Alex Elder <elder@linaro.org>
Reviewed-by: Vaibhav Hiremath <vaibhav.hiremath@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: timesync: Ensure parallel synchronous calls succeed
Bryan O'Donoghue [Tue, 2 Aug 2016 12:18:28 +0000 (13:18 +0100)]
greybus: timesync: Ensure parallel synchronous calls succeed

The guard for initiating a new synchronization operation should allow for
that resync to happen in every single state except for INVALID. This patch
fixes by ensuring the guard does just that. With local testing it was
possible to break a sync to a Module. This hasn't been observed in a buglog
but should be fixed anyway.

Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Acked-by: David Lin <dtwlin@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: timesync: Implement a retry mechanism
Bryan O'Donoghue [Tue, 2 Aug 2016 12:18:30 +0000 (13:18 +0100)]
greybus: timesync: Implement a retry mechanism

It's possible the AP could miss an incoming SVC timesync pulse i.e. the AP
could have interrupts switched off for long enough that one SVC GPIO strobe
ends up over-lapping another one. TimeSync should be able to deal with this
type of transitory failure by retrying a failed synchronous TimeSync resync
up to 'n' number of times. For this patch 'n' has been set to five, which
is a hand-wavy choice that 'feels' right.

Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: svc_watchdog: use schedule_delayed_work helper
David Lin [Tue, 2 Aug 2016 03:51:38 +0000 (20:51 -0700)]
greybus: svc_watchdog: use schedule_delayed_work helper

Instead of using the queue_delayed_work call for delayed work on
system_wq, use the schedule_delayed_work helper to be more consistent on
the style.

Testing Done:
 - Check SVC watchdog is pining after the change.

Signed-off-by: David Lin <dtwlin@google.com>
Reviewed-by: Vaibhav Hiremath <vaibhav.hiremath@linaro.org>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: interface: delete control device upon enable failure
David Lin [Tue, 2 Aug 2016 22:34:46 +0000 (15:34 -0700)]
greybus: interface: delete control device upon enable failure

There is an issue that when an interface failed to be enabled due to
timesync failure, a previously added control device is not deleted as
part of the error clean-up. This causes a leak in the sysfs file when
the interface is disabled. This would eventually cause this particular
interface to be unable to register future control device even after
unipro_reset. See failure logs below:

[  906.495261] greybus 1-3.3: failed to add to timesync: -19
[  906.516497] greybus 1-3.3: failed to re-enable interface: -19
[  907.016016] greybus 1-3.3: Interface removed
...
[ 1623.677343] ------------[ cut here ]------------
[ 1623.681116] WARNING: at kernel/arche/fs/sysfs/dir.c:530 sysfs_add_one+0x98/0xb4()
[ 1623.681128] sysfs: cannot create duplicate filename '/bus/greybus/devices/1-3.3.ctrl'
[ 1623.681252] Call trace:
[ 1623.681265] [<ffffffc000207b40>] dump_backtrace+0x0/0x268
[ 1623.681272] [<ffffffc000207db8>] show_stack+0x10/0x1c
[ 1623.681284] [<ffffffc000ccb890>] dump_stack+0x1c/0x28
[ 1623.681295] [<ffffffc00021f9dc>] warn_slowpath_common+0x74/0x9c
[ 1623.681301] [<ffffffc00021fa60>] warn_slowpath_fmt+0x5c/0x80
[ 1623.681307] [<ffffffc000366624>] sysfs_add_one+0x94/0xb4
[ 1623.681315] [<ffffffc0003670b4>] sysfs_do_create_link_sd+0x100/0x1c8
[ 1623.681320] [<ffffffc0003671a8>] sysfs_create_link+0x2c/0x38
[ 1623.681332] [<ffffffc0005d5890>] bus_add_device+0xd8/0x190
[ 1623.681338] [<ffffffc0005d39ec>] device_add+0x2b4/0x604
[ 1623.681349] [<ffffffbffc006dfc>] gb_control_add+0x10/0x40 [greybus]
[ 1623.681362] [<ffffffbffc003dac>] gb_interface_enable+0x20c/0x3b8 [greybus]
[ 1623.681373] [<ffffffbffc002a30>] gb_module_add+0x124/0x174 [greybus]
[ 1623.681385] [<ffffffbffc0082cc>] gb_svc_intf_set_power_mode+0xdd4/0xfe8 [greybus]
[ 1623.681394] [<ffffffc00023888c>] process_one_work+0x268/0x3c8
[ 1623.681400] [<ffffffc000239a64>] worker_thread+0x204/0x358
[ 1623.681410] [<ffffffc00023f43c>] kthread+0xb8/0xc4
[ 1623.681414] ---[ end trace 44489577dd9220db ]---
[ 1623.681818] greybus 1-3.3.ctrl: failed to register control device: -17

Testing Done:
 - Continuous unipro_reset stress test

Signed-off-by: David Lin <dtwlin@google.com>
Reviewed-by: Johan Hovold <johan@hovoldconsulting.com>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Jeffrey Carlyle <jcarlyle@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: es2: fix arpc active-list corruption
Johan Hovold [Wed, 27 Jul 2016 14:37:22 +0000 (16:37 +0200)]
greybus: es2: fix arpc active-list corruption

Fix ARPC active-list corruption due to incorrect list-interface usage.

The corruption manifested itself as ARPC timeouts due to expected
responses not being recognised whenever more than one ARPC was active.

This could be seen for example when two interfaces were being runtime
suspended in parallel:

[  165.739896] usb 1-1.1: invalid arpc response id received: 13
[  165.794743] greybus 1-5.5: gb_interface_refclk_set - 0
[  166.241202] usb 1-1.1: failed to execute ARPC: -110

Fortunately the impact of this bug has so far been limited to such
timeouts and error messages due to ARPC currently only being used for
CPort reset in the connection tear-down path.

Reported-by: Mark Greer <mgreer@animalcreek.com>
Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Reviewed-by: Alex Elder <elder@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: es2: fix arpc error message
Johan Hovold [Wed, 27 Jul 2016 14:37:21 +0000 (16:37 +0200)]
greybus: es2: fix arpc error message

Add missing le16_to_cpu() to an ARPC error message in the receive path,
and also use %u to print the unsigned id.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Reviewed-by: Alex Elder <elder@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: es2: fix arpc response-allocation error handling
Johan Hovold [Wed, 27 Jul 2016 14:37:20 +0000 (16:37 +0200)]
greybus: es2: fix arpc response-allocation error handling

The wrong pointer was checked for allocation failures when allocating
the ARPC response buffer, something which would lead to allocation
failures going undetected.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Reviewed-by: Alex Elder <elder@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: replace Ara references
Johan Hovold [Thu, 28 Jul 2016 09:40:53 +0000 (11:40 +0200)]
greybus: replace Ara references

Replace all occurrences of the term "Ara" with "GMP" in core.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: Documentation/sysfs: replace Ara references
Johan Hovold [Thu, 28 Jul 2016 09:40:52 +0000 (11:40 +0200)]
greybus: Documentation/sysfs: replace Ara references

Refer to the MDK as GMP MDK.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: arche-apb-ctrl: Rename ara,init-disable => arche,init-disable
Vaibhav Hiremath [Thu, 28 Jul 2016 08:26:44 +0000 (13:56 +0530)]
greybus: arche-apb-ctrl: Rename ara,init-disable => arche,init-disable

Inline with other properties used for arche-platform driver,
rename ara,init-disable => arche,init-disable.

Testing Done: Boot tested on EVT2 platform

Signed-off-by: Vaibhav Hiremath <vaibhav.hiremath@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: camera: Rename debug and metadata mbus formats
Gjorgji Rosikopulos [Wed, 27 Jul 2016 18:47:01 +0000 (11:47 -0700)]
greybus: camera: Rename debug and metadata mbus formats

Change ARA prefix to GB_CAM for greyus camera specific
formats.

Change-Id: I1a0552516e8ea727c48085c59fc49f2409a89486
Signed-off-by: Gjorgji Rosikopulos <grosikopulos@mm-sol.com>
Signed-off-by: Blagovest Borisov Kolenichev <kolenichev_blagovest@projectara.com>
Acked-by: Laurent Pinchart <laurent.pinchart@linaro.org>
7 years agogreybus: audio: add runtime pm to enumerated control and DAPM widget
David Lin [Mon, 25 Jul 2016 23:29:20 +0000 (16:29 -0700)]
greybus: audio: add runtime pm to enumerated control and DAPM widget

There's an issue that the userspace is not able to control both the
enumerated control and DAPM widget when audio bundle is in the SUSPEND
state. This patch fixes the issue by adding pm_runtime_get/put() calls
for the both controls.

Testing Done:
 - Use tinymix to get and put both enumerated control and DAPM widget as
   the followings, and observe audio bundle is able to wake up from
   suspend.

$ tinymix "GB 3 PB source" 1
$ tinymix "GB 3 PB source"
GB 3 PB source: AIF1 >AIF2

$ tinymix "GB 3 AIF1_RX MUX" 2
$ tinymix "GB 3 AIF1_RX MUX"
GB 3 AIF1_RX MUX: Stereo Left >Right

Reported-by: Vaibhav Agarwal <vaibhav.agarwal@linaro.org>
Signed-off-by: David Lin <dtwlin@google.com>
Reviewed-by: Vaibhav Agarwal <vaibhav.agarwal@linaro.org>
Reviewed-by: Mark Greer <mgreer@animalcreek.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: svc_watchdog: Add sysfs file to change the behavior of bite
David Lin [Tue, 26 Jul 2016 23:27:28 +0000 (16:27 -0700)]
greybus: svc_watchdog: Add sysfs file to change the behavior of bite

Currently, AP performs unipro_reset if SVC fails to response to its
ping. While this error recovery is best suited for the end-user
experience, errors in the UniPro network could potentially go unnoticed
by the QA and fishfooders in the development phase of the project. This
patch adds an option to trigger a kernel panic so logs can be collected
for analysis.

Testing Done:
 - Reproduce issue and observe kernel panic when
   watchdob_control is changed to 'panic'

Signed-off-by: David Lin <dtwlin@google.com>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: firmware: Update Documentation and sample application
Viresh Kumar [Mon, 25 Jul 2016 21:38:09 +0000 (14:38 -0700)]
greybus: firmware: Update Documentation and sample application

Update documentation and sample application to capture the 'status' byte
in backend version operation and new error types in backend firmware
update operation.

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: firmware: Add 'status' byte to backend fw version response
Viresh Kumar [Mon, 25 Jul 2016 21:38:08 +0000 (14:38 -0700)]
greybus: firmware: Add 'status' byte to backend fw version response

The backend processor may not be ready to return the version of firmware
it is running by the time AP requests for it. The greybus specification
is updated to return 1-byte 'status' to return the error type, RETRY is
one of them.

This patch implements that in greybus now.

Note that the version structure was common across interface and backend
version requests earlier, but that is changing as well.

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: firmware: Add new status types for backend updated request
Viresh Kumar [Mon, 25 Jul 2016 21:38:07 +0000 (14:38 -0700)]
greybus: firmware: Add new status types for backend updated request

The specification got updated with two more status values, add their
support in greybus.

As retry isn't really an error, skip printing error messages for it as
well.

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: interface: Add quirk for no PM for S2 Loader
Viresh Kumar [Tue, 26 Jul 2016 20:41:04 +0000 (13:41 -0700)]
greybus: interface: Add quirk for no PM for S2 Loader

S2 Loader doesn't support runtime PM operations currently and we will
fail to suspend/resume the bundle for firmware management protocols.

Once that happens, the bundle and its connections will be pretty much
useless as we would have tried to disable/enable all connections during
such an operation and the S2 loader doesn't expect the connections to go
away during normal operation (except in the case of mode-switch).

This patch defines a new quirk GB_INTERFACE_QUIRK_NO_PM and uses a new
interface init status value (GB_INIT_S2_LOADER_INITIALIZED) which will
be advertised by S2 Loader now in the init status.

After detecting the currently running stage as S2 Loader, the kernel
wouldn't attempt suspending or resuming the bundle.

Reviewed-by: David Lin <dtwlin@google.com>
Reviewed-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: interface: Rename *_NO_PM as *_NO_BUNDLE_ACTIVATE
Viresh Kumar [Tue, 26 Jul 2016 20:41:03 +0000 (13:41 -0700)]
greybus: interface: Rename *_NO_PM as *_NO_BUNDLE_ACTIVATE

Its a special quirk just for the bootrom as it doesn't have any PM
operations implemented. As the greybus bootrom bundle driver doesn't try
to do any PM stuff, this quirk is used only to skip bundle activate
operation currently.

Rename the GB_INTERFACE_QUIRK_NO_PM quirk to
GB_INTERFACE_QUIRK_NO_BUNDLE_ACTIVATE to suit its purpose better as the
GB_INTERFACE_QUIRK_NO_PM will be used for other quirk now.

Reviewed-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: firmware: Add runtime PM support
Viresh Kumar [Tue, 26 Jul 2016 20:41:02 +0000 (13:41 -0700)]
greybus: firmware: Add runtime PM support

This patch implements runtime PM support for firmware management bundle.
This guarantees that the bundle will be active, while the AP or the
Module is trying to exchange any operations over any of the CPorts.

- Firmware Management CPort:

  Runtime PM get/put calls are placed around the ioctl calls, which are
  all implemented as blocking ioctls.

- Component Authentication CPort:

  Runtime PM get/put calls are placed around the ioctl calls, which are
  all implemented as blocking ioctls.

- SPI:

  Uses the interface provided by spilib.c and runtime PM get/put are
  called around connection usage.

- firmware-download:

  This is the most tricky one. All operations on this CPort are
  initiated from the Module and not from the AP. And the AP needs to do
  runtime_pm_get() before any request is received over this CPort.

  The module doesn't send any request over this connection, unless the
  AP has requested the module over firmware management CPort to download
  a firmware package over firmware download CPort.

  And so the runtime PM get/put calls around the ioctls in
  fw-management.c are sufficient to handle the firmware management CPort
  as well.

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: operation: clean up request handler
Johan Hovold [Tue, 26 Jul 2016 15:11:30 +0000 (17:11 +0200)]
greybus: operation: clean up request handler

Clean up the incoming request handler somewhat by passing a properly
aligned header and dropping the now redundant id and type parameters.

Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Laurent Pinchart <laurent.pinchart@linaro.org>
Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Reviewed-by: Alex Elder <elder@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: operation: fix broken response tracepoint
Johan Hovold [Tue, 26 Jul 2016 15:11:29 +0000 (17:11 +0200)]
greybus: operation: fix broken response tracepoint

The response-received tracepoint is currently broken. Instead of parsing
the received message header it was tracing a bunch of zeroed fields from
an uninitialised response buffer.

Fix this by moving the tracepoint after were the response buffer is
initialised.

Fixes: 7cb496e6890e ("greybus: tracing: fix message traces")
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Reviewed-by: Alex Elder <elder@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: operation: fix broken response error messages
Johan Hovold [Tue, 26 Jul 2016 15:11:28 +0000 (17:11 +0200)]
greybus: operation: fix broken response error messages

The operation type included in the error message printed for malformed
responses has never been correct. An uninitialised buffer was used to
retrieve the type, resulting in the type always being reported as 0.

Fix this by passing a properly aligned header to the response handler,
and drop the now redundant id and result parameters.

Fixes: cb0ef0c019ab ("operation: print message type on errors")
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Laurent Pinchart <laurent.pinchart@linaro.org>
Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Reviewed-by: Alex Elder <elder@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: HID: Add runtime pm support
Philip Yang [Fri, 22 Jul 2016 21:48:14 +0000 (05:48 +0800)]
greybus: HID: Add runtime pm support

Modify HID greybus driver to support runtime PM framework.
The suspend and resume fucntion have been test with tutorial-hid,
e-ink-display and gpbridge-test image by sysfs resume, all HID
buttons work well on suspend/resume.

Testing Done: Compiled and verified on EVT2, DB3.5, GPB test module
              with daughter board, Red module.

Signed-off-by: Philip Yang <yang_philip@projectara.com>
Reviewed-by: David Lin <dtwlin@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: bootrom: Rename download_firmware() as find_firmware()
Viresh Kumar [Fri, 22 Jul 2016 21:13:44 +0000 (14:13 -0700)]
greybus: bootrom: Rename download_firmware() as find_firmware()

The download_firmware() function isn't downloading the firmware but just
finding if one is available or not. The same applies to the error
message printed by it.

Replace 'download' with 'find' to make it sound better.

Reported-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: bootrom: Create gb_bootrom_cancel_timeout()
Viresh Kumar [Fri, 22 Jul 2016 21:13:43 +0000 (14:13 -0700)]
greybus: bootrom: Create gb_bootrom_cancel_timeout()

We set timeouts using gb_bootrom_set_timeout(), which hides the internal
implementation, i.e. workqueues.

While canceling timeouts, we do cancel_delayed_work_sync(), which
exposes the internal implementation and doesn't look that clean.

Create gb_bootrom_cancel_timeout() to hide the internal implementation
here as well.

Reviewed-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: bootrom: Set timeout before sending AP_READY
Viresh Kumar [Fri, 22 Jul 2016 21:13:42 +0000 (14:13 -0700)]
greybus: bootrom: Set timeout before sending AP_READY

A special race is observed in the case where the phone boots with
modules attached to it.

In that case, the AP processes a later FIRMWARE_SIZE request from the
bootrom, before processing the response to the earlier AP_READY request.
And because of that, we set the timeout from gb_bootrom_probe() after
trying to cancel it from gb_bootrom_firmware_size_request().

And with that, if the firmware package isn't available for the kernel to
directly read from, then we get a timeout print message like below:

[   23.669764] bootrom 1-3.3.1: Firmware file 'ara_00000126_00001001_fffe0001_ffe5001a_s2l.tftf' requested
[   24.680528] bootrom 1-3.3.1: Timed out waiting for Firmware Size Request from the Module

Note that this doesn't happen if the module is inserted into the frame,
after the phone is already booted.

This behavior is the result of an earlier commit cdd1cb3f4943
("operation: add completion work queue").

This perhaps happens only at boot time because the UNBOUND wq
(connection->wq) gets a chance to run on other CPUs, while the BOUND wq
(gb_operation_completion_wq) doesn't at that stage (24 seconds since
boot).

Setting the timeout before sending the AP_READY request fixes it for
now.

Reported-by: Johan Hovold <johan@hovoldconsulting.com>
Reviewed-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: bootrom: Don't print duplicate error messages
Viresh Kumar [Fri, 22 Jul 2016 21:13:41 +0000 (14:13 -0700)]
greybus: bootrom: Don't print duplicate error messages

On failing to find a firmware image, we get these today:

bootrom 1-3.3.1: Firmware request for ara_00000126_00001001_fffe0001_ffe5001a_s2l.tftf has failed : -12
bootrom 1-3.3.1: gb_bootrom_firmware_size_request: failed to download firmware (-12)

Which are more or less duplicate, as they print error for the same root
cause.

With this patch this is all we get now:

bootrom 1-3.3.1: failed to download ara_00000126_00001001_fffe0001_ffe5001a_s2l.tftf firmware (-12)

Reported-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: bootrom: Skip setting timeout in failure path of size request
Viresh Kumar [Fri, 22 Jul 2016 21:13:40 +0000 (14:13 -0700)]
greybus: bootrom: Skip setting timeout in failure path of size request

The currently set value of next_request_type in the error path of
gb_bootrom_firmware_size_request() is not correct, as it should have
been NEXT_REQ_FIRMWARE_SIZE for the failure case (as we should be
waiting for another similar request).

But, if an error occurs in gb_bootrom_firmware_size_request(), then the
ES3 bootrom will never be able to recover from it and send another
request. And so there is no point waiting for another request and timing
out.

Skip doing that in error path.

Reported-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: vibrator: integrate runtime pm
Ann Chen [Fri, 22 Jul 2016 07:33:55 +0000 (15:33 +0800)]
greybus: vibrator: integrate runtime pm

Integrate greybus drivers with the Linux Kernel RuntimePM framework
for vibrator driver.

Testing Done: AP side (kernel) can control the vibrator driver with
suspend and resume.

Signed-off-by: Ann Chen <chen_ann@projectara.com>
Reviewed-by: David Lin <dtwlin@google.com>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: control: Print bundle-id in print messages
Viresh Kumar [Fri, 22 Jul 2016 05:41:05 +0000 (22:41 -0700)]
greybus: control: Print bundle-id in print messages

The new power management specific operations added to the control
protocol do not print the bundle-id in the error messages and it is not
possible to identify which bundle-id the operation failed for.

Fix that and do minor rewriting of the print messages to make them more
readable.

Tested on EVT 2.0 with gpbridge-test module.

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Sandeep Patil <sspatil@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: operation: print id when synchronous operation timeout
David Lin [Fri, 22 Jul 2016 20:46:25 +0000 (13:46 -0700)]
greybus: operation: print id when synchronous operation timeout

In case of a synchronous operation timeout error, it's helpful for
purpose of debugging to print the operation id in the error message, so
that we know if the response is received at a later time after operation
time out.

Testing Done:
 - Observe the error message below when response comes later after
operation timeout:

[  792.973978] greybus greybus1: 0/0:0: synchronous operation id 0x0005
of type 0x21 failed: -110
[  800.646694] greybus greybus1: 0/0:0: unexpected response id 0x0005
received

Signed-off-by: David Lin <dtwlin@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: audio: Avoid using ARA keyword
Vaibhav Agarwal [Fri, 22 Jul 2016 04:11:30 +0000 (09:41 +0530)]
greybus: audio: Avoid using ARA keyword

It is suggested to avoid using ARA keyword externally. So we need to
update GB codec driver. Also, codec name is currently set to 'gb-codec'.
However, it makes more sense to name it as apb-dummy-codec, since it is
used to control various audio modules connected to APB via greybus.

Signed-off-by: Vaibhav Agarwal <vaibhav.agarwal@linaro.org>
Reviewed-by: Mark Greer <mgreer@animalcreek.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: spilib: make spilib independent of gbphy
Viresh Kumar [Thu, 21 Jul 2016 21:45:06 +0000 (14:45 -0700)]
greybus: spilib: make spilib independent of gbphy

spilib is used by multiple users currently (spi.c and fw-core.c) but
commit aa52b62a0556 broke that hierarchy and introduced gbphy dependent
code in spilib.

This may have unreliable consequences as we are doing following
operation unconditionally now:

gbphy_dev = to_gbphy_dev(spi->parent);
gbphy_runtime_get_sync(gbphy_dev);

which may not go well when the parent is of type &bundle->dev
(fw-core.c).

This patch introduces spilib_ops and lets the users of the core register
them. This shall have no functional change for the spi.c usecase and
shall fix the unreliable results for the fw-core.c usecase.

Tested by writing to mtd0 dev and verifying (with print messages) that
the below routines are getting called for a gpbridge-test module.

Fixes: aa52b62a0556 ("spi: Add runtime_pm support")
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: firmware: Improve test application
Viresh Kumar [Wed, 20 Jul 2016 23:39:00 +0000 (16:39 -0700)]
greybus: firmware: Improve test application

It can take arguments not and we can perform all the operations with a
single binary, sorry for missing that initially.

Usage: ./firmware <gb-fw-mgmt-X> <type: interface/backend> <firmware-tag> <timeout>

And all of them have default values, etc.

Tested with a semco 13 MP module.

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: timesync: do not print frametime by default
David Lin [Thu, 21 Jul 2016 00:18:39 +0000 (17:18 -0700)]
greybus: timesync: do not print frametime by default

Currently frametime are being printed whenever an Interface is
runtime-resumed. This is unnecessarily chatty. This patch moves the
frametime print from using pr_info() to dev_dbg().

Testing Done:
 - $ echo "module greybus +p" > /d/dynamic_debug/control

[   97.699395] greybus 1-svc: greybus frametime: ap=2042778707
1-svc=2042775185 greybus1=2042778397 1-6.6=2042778390

Signed-off-by: David Lin <dtwlin@google.com>
Reviewed-by: Jeffrey Carlyle <jcarlyle@google.com>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: pm: add error handling to bundle activation
Bartosz Golaszewski [Thu, 21 Jul 2016 16:09:34 +0000 (18:09 +0200)]
greybus: pm: add error handling to bundle activation

The firmware now keeps the underlying hardware disabled until
receiving the first Bundle Activate request. Additionally: requesting
transition to a state the bundle is already in is no longer an error.

We can now add proper error handling to the bundle activate call.

Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: control: suppress bundle_activate() for bootrom
Johan Hovold [Thu, 21 Jul 2016 12:24:11 +0000 (14:24 +0200)]
greybus: control: suppress bundle_activate() for bootrom

We always knew backward compatibility with the ES3 bootrom, which was
finalised about a year ago, would be a pain. Here we go again.

The bootrom does not support control requests added after it was burnt
into ROM for obvious reasons. This means that we need to suppress
sending the new bundle_activate() operation to any interface executing
the legacy bootrom.

Do so by adding a new NO_PM interface quirk (we can use the
control-protocol version for this later once we bump it).

Note that the interface-disable path (e.g. for power down) is already
handled by the FORCED_DISABLE quirk, and that the suspend/resume
paths are currently avoided by making sure that the bootrom bundle
never suspends.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Acked-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
Tested-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: firmware.h: rename TFTF firmware filenames
Joel Porquet [Thu, 21 Jul 2016 22:57:38 +0000 (15:57 -0700)]
greybus: firmware.h: rename TFTF firmware filenames

Replace prefix "ara_" with prefix "gmp_" for TFTF firmware filenames.

Testing Done: compiled and flashed EVT2 (bootimage) and tested that
              "gmp_" prefixed tftf files are expected. Pushed one from
      FDK and checked that it's properly loaded.

Signed-off-by: Joel Porquet <porquet_joel@projectara.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Acked-by: Sandeep Patil <sspatil@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: interface: fix interface_id attribute group
Johan Hovold [Thu, 21 Jul 2016 10:48:57 +0000 (12:48 +0200)]
greybus: interface: fix interface_id attribute group

All interfaces, regardless of type, should have an interface_id
attribute reflecting its position on the frame.

This has been reported to cause an assertion failure in libmoduleutil
for dummy modules.

Testing done: Verified that the attribute is present for registered
dummy interfaces.

Reported-by: Gjorgji Rosikopulos <grosikopulos@mm-sol.com>
Tested-by: Gjorgji Rosikopulos <grosikopulos@mm-sol.com>
Reviewed-by: Patrick Titiano <ptitiano@baylibre.com>
Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: timesync: probe shouldn't complete until FrameTime sync does
Bryan O'Donoghue [Wed, 20 Jul 2016 18:31:18 +0000 (19:31 +0100)]
greybus: timesync: probe shouldn't complete until FrameTime sync does

Currently the probe() function contains the asynchronous() variant of
FrameTime synchronization. This patch converts to the synchronous() version
of synchronization. This is required for two reasons first a probe() cannot
reasonably be considered to be complete without successfully completing a
time synchronization for Interfaces that care about that sync. Secondly
scheduling the operation asynchronously means its possible the PM-runtime
suspend() path can execute before the async timesync operation completes.
For both reasons we want to run synchronization - synchronously.

Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: loopback_test: Add -f argument to cancel all tests
Ryan Lim [Wed, 20 Jul 2016 15:14:04 +0000 (08:14 -0700)]
greybus: loopback_test: Add -f argument to cancel all tests

To make sure tests on all devices, enabled or not, are cancelled,
specify -f when starting new loopback test.

Signed-off-by: Ryan Lim <limryan@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: loopback_test: Fix -z argument bug
Ryan Lim [Wed, 20 Jul 2016 15:14:03 +0000 (08:14 -0700)]
greybus: loopback_test: Fix -z argument bug

The -z argument was missing the break statement needed in the switch
block and was not in the getopt list.

Signed-off-by: Ryan Lim <limryan@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: loopback_test: Cancel only the tests running on selected devices
Ryan Lim [Wed, 20 Jul 2016 15:14:02 +0000 (08:14 -0700)]
greybus: loopback_test: Cancel only the tests running on selected devices

When starting a loopback test, it cancels all currently running tests on
all loopback devices. When -m argument is given, which runs the test on
specific loopback devices, only the tests running on the selected
devices need to be cancelled before starting new tests.

Signed-off-by: Ryan Lim <limryan@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: timesync: reduce initial startup time
Bryan O'Donoghue [Tue, 19 Jul 2016 14:04:37 +0000 (15:04 +0100)]
greybus: timesync: reduce initial startup time

Currently we have a long initial delay when initiating the first operation
to synchronize the FrameTime. That made sense during development of the
feature since we did synchronization asynchronously and wanted to give some
grace time for new Interfaces to be added before initiating a new
synchronization. On the PM runtime resume path though we do a synchronous
restoration of FrameTime and in this case waiting for 1 second before
initiating the first operation is clearly wrong.

This patch reduces the initial operation time bringing the synchronous call
time down by 990 milliseconds.

Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: move all firmware prefix strings to one place
Greg Kroah-Hartman [Wed, 20 Jul 2016 17:11:37 +0000 (10:11 -0700)]
greybus: move all firmware prefix strings to one place

The firmware prefix is hard-coded in a few different places.  Put it all
in one handy #define, for when/if we ever decide to change it in the
future...

Testing: 'strings gb-firmware.ko gb-bootrom.ko | grep ara_' produced the
 same output before and after this patch.

Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Acked-by: Sandeep Patil <sspatil@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: interface: hibernate UniPro link in activation error path
Johan Hovold [Wed, 20 Jul 2016 14:40:27 +0000 (16:40 +0200)]
greybus: interface: hibernate UniPro link in activation error path

To be well-behaved, we should hibernate the link before disabling UniPro
in case the link has already been established (i.e. when the interface
type has been detected as UniPro or Greybus).

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Reviewed-by: Alex Elder <elder@linaro.org>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Sandeep Patil <sspatil@google.com>
Reviewed-by: Patrick Titiano <ptitiano@baylibre.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: interface: fix power_state documentation
Johan Hovold [Wed, 20 Jul 2016 14:40:26 +0000 (16:40 +0200)]
greybus: interface: fix power_state documentation

Reword the power_state attribute documentation to make it clear that an
attempt to boot and possibly enumerate an interface is made when writing
"on" to the file, and that on errors the interface will again be powered
down.

Drop the incorrect claim that writing the currently read value has no
effect, since no such guarantees can be made (e.g. several writers may
be sleeping on the interface mutex).

Also fix some minor language issues.

Reviewed-by: Sandeep Patil <sspatil@google.com>
Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Reviewed-by: Alex Elder <elder@linaro.org>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Patrick Titiano <ptitiano@baylibre.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: interface: add power_state attribute
Viresh Kumar [Wed, 20 Jul 2016 14:40:25 +0000 (16:40 +0200)]
greybus: interface: add power_state attribute

User space needs the capability of powering ON or OFF an Interface for
multiple use cases. For example, userspace may want an Interface
currently in its S3 boot stage, to boot into its S2 Loader stage to
update the bridge SPI flash. Or the Interface is running its S2 Loader
stage and updated the SPI flash with the new S2 Loader firmware and
wants to boot into the new S2 Loader firmware.

Another use case can be, Android wants to disable (not eject) a
misbehaving module.

Add a 'power_state' sysfs file within the interface directory. It can be
read to know the current power state of the Interface and can be written
to power ON or power OFF an Interface. Possible values that can be
written or read from it are: "on" and "off".

Testing Done: Tested by enabling/disabling camera module on EVT 2.0.

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
CC: David Lin <dtwlin@google.com>
Reviewed-by: Johan Hovold <johan@hovoldconsulting.com>
[johan: drop es3-quirk duplication, add to power attribute group, fix
return value, drop tags ]
Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Reviewed-by: Alex Elder <elder@linaro.org>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Sandeep Patil <sspatil@google.com>
Reviewed-by: Patrick Titiano <ptitiano@baylibre.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: interface: prevent reactivation during removal
Johan Hovold [Wed, 20 Jul 2016 14:40:24 +0000 (16:40 +0200)]
greybus: interface: prevent reactivation during removal

Make sure to prevent an interface that is going away from being
reactivated.

This is needed to preemptively close a race between the upcoming feature
to reactivate a powered-down interface and physical removal (i.e.
module_removed event processing) as well as logical removal (e.g. the
current system-suspend hack).

Reviewed-by: Sandeep Patil <sspatil@google.com>
Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Reviewed-by: Alex Elder <elder@linaro.org>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Patrick Titiano <ptitiano@baylibre.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: interface: make sure type is invariant during reactivation
Johan Hovold [Wed, 20 Jul 2016 14:40:23 +0000 (16:40 +0200)]
greybus: interface: make sure type is invariant during reactivation

An interface is not expected to change its type after a power down and
reactivation so make sure to treat that as a fatal error.

This is complicated by the current Toshiba ES3 hack which requires us
to retry activation as Greybus interfaces are sometimes misdetected as
UniPro interfaces. Handle that by only retrying activation the first
time an interface is activated, and for interfaces already detected as
having Greybus type.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Reviewed-by: Alex Elder <elder@linaro.org>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Sandeep Patil <sspatil@google.com>
Reviewed-by: Patrick Titiano <ptitiano@baylibre.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: interface: clean up ES3 activation-retry hack
Johan Hovold [Wed, 20 Jul 2016 14:40:22 +0000 (16:40 +0200)]
greybus: interface: clean up ES3 activation-retry hack

Clean up the ES3 activation retry-hack and isolate it in the interface
code.

This way the retry hack can be reused when we soon start allowing
interfaces to be reactivated after having been powered down.

Reviewed-by: Sandeep Patil <sspatil@google.com>
Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Reviewed-by: Alex Elder <elder@linaro.org>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Patrick Titiano <ptitiano@baylibre.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: interface: do not read DME during reactivation
Johan Hovold [Wed, 20 Jul 2016 14:40:21 +0000 (16:40 +0200)]
greybus: interface: do not read DME during reactivation

It will soon be possible to reactivate an interface that has been
powered down after already having been activated. In that case there's
no need to re-read the DME attributes as part of activation as the
values are already cached.

Reviewed-by: Sandeep Patil <sspatil@google.com>
Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Reviewed-by: Alex Elder <elder@linaro.org>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Patrick Titiano <ptitiano@baylibre.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: interface: use an enum for interface type
Johan Hovold [Wed, 20 Jul 2016 14:40:20 +0000 (16:40 +0200)]
greybus: interface: use an enum for interface type

Use an enum for the interface type instead of using the SVC protocol
values directly.

Suggested-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Reviewed-by: Alex Elder <elder@linaro.org>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Sandeep Patil <sspatil@google.com>
Reviewed-by: Patrick Titiano <ptitiano@baylibre.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: module: suppress activation error message for dummy interfaces
Johan Hovold [Tue, 19 Jul 2016 13:24:50 +0000 (15:24 +0200)]
greybus: module: suppress activation error message for dummy interfaces

We currently handle dummy interfaces by deactivating them using
activation error paths, but we don't want the corresponding
module_inserted error message to be printed.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: interface: amend interface registration message
Johan Hovold [Tue, 19 Jul 2016 13:24:49 +0000 (15:24 +0200)]
greybus: interface: amend interface registration message

Amend the interface registration message with the detected interface
type, and only print the Ara VID/PID and DDBL1 attributes for the types
for which they exist.

Also drop the now redundant message about a detected dummy interface
from the activate operation helper.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: interface: make attributes type dependent
Johan Hovold [Tue, 19 Jul 2016 13:24:48 +0000 (15:24 +0200)]
greybus: interface: make attributes type dependent

Make most interface attributes type dependent (e.g only UniPro and
Greybus interfaces should have a DDBL1 Manufacturer ID attribute).

Note that the power attributes (e.g. current_now) will only be visible
for UniPro- and Greybus-type interfaces (i.e. interfaces that can draw
power).

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: interface: partition attribute group
Johan Hovold [Tue, 19 Jul 2016 13:24:47 +0000 (15:24 +0200)]
greybus: interface: partition attribute group

Partition the current attribute group into four groups for UniPro,
Greybus, power and common attributes.

This is a step in refining the interface-type handling as attributes are
type dependent.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: interface: add interface-type attribute
Johan Hovold [Tue, 19 Jul 2016 13:24:46 +0000 (15:24 +0200)]
greybus: interface: add interface-type attribute

Add an interface-type string attribute that represents the detected
interface type as either "dummy", "unipro", "greybus", or "unknown".

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Sandeep Patil <sspatil@google.com>
Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
7 years agogreybus: arche-platform: Add support for init-off feature
Vaibhav Hiremath [Tue, 19 Jul 2016 08:01:46 +0000 (13:31 +0530)]
greybus: arche-platform: Add support for init-off feature

Disable wake_detect interrupt after request

As part of SW-4344/SW-7061, now we are enabling FW flashing
to all builds. That means check for need of FW upgrade is
going to be present in all builds, and moving to FW_FlASHING
mode from active is heavy operation; so the idea here is
simplify this process and save the boot time due to switching
back-n-forth between ACTIVE<=>FW_FLASHING modes.

So we decided to put unipro into OFF state by default on boot,
which can be changed through DT property. If arche-platform
device node has "arche,init-off" property set, then unipro
will be in OFF state on boot. User can bring it back by

 # echo active > /sys/devices/arche_platform*/state

And to simply the exit code of probe() fn the
arche_platform_coldboot_seq() has been shifted to the
bottom of the _probe() fn.

Testing Done: Tested on EVT2 platform, with and without
"arche,init-off" property, multiple times.
Note: I am seeing SW-7128, which is not related to these changes.

Signed-off-by: Vaibhav Hiremath <vaibhav.hiremath@linaro.org>
Tested-by: Michael Scott <michael.scott@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>