platform/kernel/linux-rpi.git
9 years agogreybus: es2: sync up with recent es1.c changes
Greg Kroah-Hartman [Fri, 27 Mar 2015 10:38:07 +0000 (11:38 +0100)]
greybus: es2: sync up with recent es1.c changes

This brings the debug log functionality of es1.c to es2.c, along with
some other minor cleanups.

Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
Reviewed-by: Alex Elder <elder@linaro.org>
9 years agogreybus: debugfs: we shouldn't care if debugfs is working or not
Greg Kroah-Hartman [Fri, 27 Mar 2015 10:38:06 +0000 (11:38 +0100)]
greybus: debugfs: we shouldn't care if debugfs is working or not

This removes the error checking for debugfs initialization as we really
don't care if it failed or not.

Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
Reviewed-by: Alex Elder <elder@linaro.org>
9 years agogreybus: get rid of {conceal,reveal}_urb()
Alex Elder [Fri, 27 Mar 2015 02:25:06 +0000 (21:25 -0500)]
greybus: get rid of {conceal,reveal}_urb()

These clever macros were fine for early development, but they're
more of a distraction now.

Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: clean up some small messes
Alex Elder [Fri, 27 Mar 2015 02:25:01 +0000 (21:25 -0500)]
greybus: clean up some small messes

This is an old patch that I neglected to send out.  It's cleaning
up a couple things that got committed before I had a chance to
comment on them.

In operation.c there is a "FIXME" comment that is easily proven
wrong by inspection.

In gb_protocol_put(), there is another wrong "FIXME" comment as
well.  We can also use our cached copies of the protocol major
and minor version number in another spot.  And balance that
out by using a cached copy of the protocol id.

Signed-off-by: Alex Elder <elder@linaro.org>
Reviewed-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: Merge branch 'es1-fixes' into master
Greg Kroah-Hartman [Wed, 25 Mar 2015 08:26:13 +0000 (09:26 +0100)]
greybus: Merge branch 'es1-fixes' into master

Merge the es1 log file fixes into master.

Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: interface: put module->dev on failures
Viresh Kumar [Tue, 24 Mar 2015 14:44:29 +0000 (20:14 +0530)]
greybus: interface: put module->dev on failures

In order to decrement the reference count of module on failures, we must call
put_device(module->dev). This was missing for one of the error cases, fix it.

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: Greybus: Place module_init/exit() right after respective routines
Viresh Kumar [Tue, 24 Mar 2015 14:44:28 +0000 (20:14 +0530)]
greybus: Greybus: Place module_init/exit() right after respective routines

As mentioned in kernel coding guidelines.

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: manifest: don't need to check for desc_size == 0 anymore
Viresh Kumar [Tue, 24 Mar 2015 11:38:14 +0000 (17:08 +0530)]
greybus: manifest: don't need to check for desc_size == 0 anymore

identify_descriptor() doesn't return 0 anymore and so we don't need to check the
returned value against 0.

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: manifest: descriptor size should be >= header size
Viresh Kumar [Tue, 24 Mar 2015 11:38:13 +0000 (17:08 +0530)]
greybus: manifest: descriptor size should be >= header size

We are calculating descriptors expected size differently based on the type of
descriptor, that's fine but at few places we aren't taking size of the header
into account. And that looks wrong.

Lets make sure it is atleast as big as descriptor's header.

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: es1: allow the debug log to be stopped
Greg Kroah-Hartman [Tue, 24 Mar 2015 19:47:24 +0000 (20:47 +0100)]
greybus: es1: allow the debug log to be stopped

If you write 0 to the debugfs file, the log will stop being updated.

Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: es1: separate usb_log enable/disable logic into different functions
Greg Kroah-Hartman [Tue, 24 Mar 2015 19:45:31 +0000 (20:45 +0100)]
greybus: es1: separate usb_log enable/disable logic into different functions

One function shouldn't do two different things depending on a parameter
passed to it, so split usb_log_enable() into usb_log_disable()

Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: es1: fix checkpatch warning about blank lines needed
Greg Kroah-Hartman [Tue, 24 Mar 2015 19:34:02 +0000 (20:34 +0100)]
greybus: es1: fix checkpatch warning about blank lines needed

Add a blank line in apb1_log_enable_read() to make checkpatch happy.

Oh, and it makes the code more readable too...

Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: es1: move debugfs function to use kstrotoint_from_user()
Greg Kroah-Hartman [Tue, 24 Mar 2015 19:32:40 +0000 (20:32 +0100)]
greybus: es1: move debugfs function to use kstrotoint_from_user()

No need to duplicate built-in functions that the kernel has, so have the
core kernel parse the userspace string.  Saves us an allocation and
makes the logic simpler.

Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: es1: decimal modes are not what are wanted for debugfs
Greg Kroah-Hartman [Tue, 24 Mar 2015 19:10:58 +0000 (20:10 +0100)]
greybus: es1: decimal modes are not what are wanted for debugfs

decimal is not octal, so use the proper mode settings for the debugfs
files.

Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: es1: struct file_operations needs to be const
Greg Kroah-Hartman [Tue, 24 Mar 2015 19:08:12 +0000 (20:08 +0100)]
greybus: es1: struct file_operations needs to be const

We aren't changing these pointers, so mark them read-only as that is the
preferred way.

Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: es1: no need to check for NULL on debugfs_remove()
Greg Kroah-Hartman [Tue, 24 Mar 2015 19:06:41 +0000 (20:06 +0100)]
greybus: es1: no need to check for NULL on debugfs_remove()

The function can, and even expects NULL, so don't check.

Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: es1: use <linux/uaccess.h> and not <asm/uaccess.h>
Greg Kroah-Hartman [Tue, 24 Mar 2015 19:04:49 +0000 (20:04 +0100)]
greybus: es1: use <linux/uaccess.h> and not <asm/uaccess.h>

Asm is only for when you are doing arch-specific stuff, which we aren't
doing here.

Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: es1: fix tiny whitespace issues
Greg Kroah-Hartman [Tue, 24 Mar 2015 19:03:39 +0000 (20:03 +0100)]
greybus: es1: fix tiny whitespace issues

No trailing spaces or spaces before tabs are allowed.

Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: es1: fix build warning for apb1_log_enable_write
Greg Kroah-Hartman [Tue, 24 Mar 2015 18:55:14 +0000 (19:55 +0100)]
greybus: es1: fix build warning for apb1_log_enable_write

It's "const char __user *buf", not "char __user *buf".

'make check' is your friend.

Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: Dump log from APB1
Alexandre Bailon [Mon, 23 Mar 2015 16:52:37 +0000 (17:52 +0100)]
greybus: Dump log from APB1

On AP module (form factor), we don't have access to APBridge JTAG or UART.
But sometime, we still need to get log from APBridge. Add a new request in control endpoint
to get APBridge logs.
Logs can be accessed using debugfs (greybus/apb1_log).

Signed-off-by: Alexandre Bailon <abailon@baylibre.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
9 years agogreybus: core: Don't initialize greybus if it is disabled
Viresh Kumar [Fri, 20 Mar 2015 14:59:13 +0000 (20:29 +0530)]
greybus: core: Don't initialize greybus if it is disabled

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
9 years agogreybus: gpio: add error messages to callbacks not propagating errors
Johan Hovold [Thu, 19 Mar 2015 15:55:23 +0000 (16:55 +0100)]
greybus: gpio: add error messages to callbacks not propagating errors

Add error messages on failures to deactivate, set and get operation
handlers as any errors would not be detected by the upper layers (either
because the callbacks are declared void or expected to return a boolean
value).

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: gpio: refuse to set value of input pins
Johan Hovold [Thu, 19 Mar 2015 15:55:22 +0000 (16:55 +0100)]
greybus: gpio: refuse to set value of input pins

Add warning and refuse to set output value for pin configured as input,
as the result of such an operation is undefined.

Remove incorrect todo-comment suggesting that the driver could
implicitly switch direction as part of the call.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: gpio: clean up line-state allocation
Johan Hovold [Thu, 19 Mar 2015 15:51:17 +0000 (16:51 +0100)]
greybus: gpio: clean up line-state allocation

Clean up allocation of line-state array.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: gpio: use dev_err and dev_warn
Johan Hovold [Thu, 19 Mar 2015 15:51:16 +0000 (16:51 +0100)]
greybus: gpio: use dev_err and dev_warn

Use the more informative dev_err and dev_warn for messages, and make the
messages more uniform.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: gpio: fix error handling
Johan Hovold [Thu, 19 Mar 2015 15:51:15 +0000 (16:51 +0100)]
greybus: gpio: fix error handling

Make sure to propagate any errors detected up the call chain.

This specifically means that we will detect failed connection init,
something which is now handled more gracefully by greybus core.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: gpio: make gb_gpio_controller pointer naming consistent
Johan Hovold [Thu, 19 Mar 2015 15:51:14 +0000 (16:51 +0100)]
greybus: gpio: make gb_gpio_controller pointer naming consistent

Rename all struct gb_gpio_controller-pointer variables "ggc" for
consistency and readability reason.

This also fixes a bunch of lines that exceeded the 80 columns limit.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: gpio: remove unnecessary explicit cast
Johan Hovold [Thu, 19 Mar 2015 15:51:13 +0000 (16:51 +0100)]
greybus: gpio: remove unnecessary explicit cast

Remove unnecessary explicit cast of line value.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: gpio: remove overly defensive argument verification
Johan Hovold [Thu, 19 Mar 2015 15:51:12 +0000 (16:51 +0100)]
greybus: gpio: remove overly defensive argument verification

Remove overly defensive argument verification in gpio-chip callbacks. We
should trust gpiolib to get this right (or we would not even get any
callback) just like the other gpio drivers do.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: gpio: remove redundant argument verification
Johan Hovold [Thu, 19 Mar 2015 15:51:11 +0000 (16:51 +0100)]
greybus: gpio: remove redundant argument verification

Remove redundant verification of gpio numbers (which have already been
verified in the gpio-chip callbacks) from greybus-operation helpers.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: gpio: remove incorrect todo comments
Johan Hovold [Thu, 19 Mar 2015 15:51:10 +0000 (16:51 +0100)]
greybus: gpio: remove incorrect todo comments

The module reference count is incremented by gpiolib when a gpio is
requested, and the driver callbacks certainly do sleep.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: gpio: fix truncated debounce times
Johan Hovold [Thu, 19 Mar 2015 15:51:09 +0000 (16:51 +0100)]
greybus: gpio: fix truncated debounce times

Fix set_debounce, which silently truncated the time argument to 255us
even though we support 16-bit values.

Also remove the unnecessary explicit cast when verifying the argument.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: operation: remove unnecessary cast
Johan Hovold [Thu, 19 Mar 2015 15:46:19 +0000 (16:46 +0100)]
greybus: operation: remove unnecessary cast

Remove unnecessary cast of the message size in gb_connection_recv.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: operation: use dev_err in gb_operation_sync
Johan Hovold [Thu, 19 Mar 2015 15:46:18 +0000 (16:46 +0100)]
greybus: operation: use dev_err in gb_operation_sync

Use the more informative dev_err in gb_operation_sync, which includes
the connection device name in the error message (which in turn encodes
the module, interface, bundle and cport ids).

Add missing braces to conditional-construct branches while at it.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: connection: replace custom error function with dev_err
Johan Hovold [Thu, 19 Mar 2015 15:46:17 +0000 (16:46 +0100)]
greybus: connection: replace custom error function with dev_err

Remove custom connection error function and replace it with dev_err.

The standard error function provides more information in the message
prefix (e.g. includes the interface id), has a well-known semantics
(e.g. does does not add newlines to messages), and is even somewhat
shorter to type.

Note that some uses of the custom function were already adding double
newlines due to the non-standard semantics.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: ap: clean up svc link management error path
Johan Hovold [Thu, 19 Mar 2015 15:46:16 +0000 (16:46 +0100)]
greybus: ap: clean up svc link management error path

Return immediately on bundle-init failure when processing SVC link up.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: ap: fix typo in comment
Johan Hovold [Thu, 19 Mar 2015 15:46:15 +0000 (16:46 +0100)]
greybus: ap: fix typo in comment

Fix typo in svc_hotplug comment.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: ap: fix svc handshake protocol check
Johan Hovold [Thu, 19 Mar 2015 15:46:14 +0000 (16:46 +0100)]
greybus: ap: fix svc handshake protocol check

Fix incorrect SVC handshake protocol check, which would only bail out if
both major and minor protocol versions supported by the SVC differed.

Since we currently only support one version of the protocol, upgrade the
debug message to warning and bail unless the protocol versions match
perfectly for now.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: usb: silence compiler warning
Johan Hovold [Thu, 19 Mar 2015 15:46:13 +0000 (16:46 +0100)]
greybus: usb: silence compiler warning

This driver is being rewritten, but let's silence a pointer-to-int-cast
compiler warning meanwhile.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: hid: don't support OUTPUT report over interrupt channel
Viresh Kumar [Wed, 18 Mar 2015 06:11:39 +0000 (11:41 +0530)]
greybus: hid: don't support OUTPUT report over interrupt channel

There is no interrupt channel as such and so no need to support
->output_report().

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: core: place module_{init|exit}() right below the routines
Viresh Kumar [Thu, 19 Mar 2015 11:32:49 +0000 (17:02 +0530)]
greybus: core: place module_{init|exit}() right below the routines

To follow coding guidelines a bit :)

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: interface: remove double underscore from fn name
Viresh Kumar [Thu, 19 Mar 2015 11:32:48 +0000 (17:02 +0530)]
greybus: interface: remove double underscore from fn name

Also bring * closer to gb_interface_get_drvdata :)

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: operation: Fix comment mistake
Viresh Kumar [Thu, 19 Mar 2015 11:32:46 +0000 (17:02 +0530)]
greybus: operation: Fix comment mistake

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: operation: s/status/result to match field name
Viresh Kumar [Thu, 19 Mar 2015 11:32:45 +0000 (17:02 +0530)]
greybus: operation: s/status/result to match field name

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: Export greybus debugfs folder
Alexandre Bailon [Wed, 18 Mar 2015 14:42:51 +0000 (15:42 +0100)]
greybus: Export greybus debugfs folder

Add gb_debugfs_get method to access to gb_debug_root dentry,
in order to use it from other greybus modules.

Signed-off-by: Alexandre Bailon <abailon@baylibre.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: uart: remove packed-attribute from line-coding struct
Johan Hovold [Tue, 17 Mar 2015 17:24:30 +0000 (18:24 +0100)]
greybus: uart: remove packed-attribute from line-coding struct

Remove packed-attribute from line-coding struct, whose members are all
naturally aligned.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: pwm: fix config-request alignment
Johan Hovold [Tue, 17 Mar 2015 17:24:29 +0000 (18:24 +0100)]
greybus: pwm: fix config-request alignment

Fix alignment of the duty and period-fields in the config request,
which should follow the which-field without any inserted padding.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: gpio: fix set-debounce request alignment
Johan Hovold [Tue, 17 Mar 2015 17:24:28 +0000 (18:24 +0100)]
greybus: gpio: fix set-debounce request alignment

Fix alignment of the usec-field in the set-debounce request, which
should follow the which-field without any inserted padding.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: connection: fix oops after failed init
Johan Hovold [Tue, 17 Mar 2015 09:55:52 +0000 (10:55 +0100)]
greybus: connection: fix oops after failed init

Make sure not to call connection_exit for connections that have never
been initialised (e.g. due to failure to init).

This fixes oopses due to null-dereferences and use-after-free in
connection_exit callbacks (e.g. trying to remove a gpio-chip that has
never been added) when the bundle and interface are ultimately
destroyed.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: vibrator: fix memory leak in error path
Johan Hovold [Tue, 17 Mar 2015 09:55:51 +0000 (10:55 +0100)]
greybus: vibrator: fix memory leak in error path

Fix memory leak in connection_init error path.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: pwm: fix memory leak in error path
Johan Hovold [Tue, 17 Mar 2015 09:55:50 +0000 (10:55 +0100)]
greybus: pwm: fix memory leak in error path

Fix memory leak in connection_init error path.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: hid: add HID class driver
Viresh Kumar [Mon, 16 Mar 2015 11:19:37 +0000 (16:49 +0530)]
greybus: hid: add HID class driver

This adds HID transport layer driver for Greybus. Most of the stuff is
implemented, but is untested.

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: gpb: Fix print mistakes
Viresh Kumar [Tue, 10 Mar 2015 09:11:12 +0000 (14:41 +0530)]
greybus: gpb: Fix print mistakes

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Alex Elder <elder@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: connection: fix locking in gb_hd_connection_find
Johan Hovold [Mon, 2 Mar 2015 08:55:26 +0000 (09:55 +0100)]
greybus: connection: fix locking in gb_hd_connection_find

Fix unconditional re-enabling of interrupts in gb_hd_connection_find,
which can be called with local interrupts disabled from the USB
completion handler.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Reviewed-by: Alex Elder <elder@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: gpio.c: fix a bad irq number
Alexandre Bailon [Mon, 2 Mar 2015 16:32:43 +0000 (17:32 +0100)]
greybus: gpio.c: fix a bad irq number

When it receive an interrupt, the function gb_gpio_request_recv doesn't
use the good gpio number to get the irq number. Then, the expected irq is never fired.

Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: operation: fix locking issues
Johan Hovold [Mon, 2 Mar 2015 11:34:40 +0000 (12:34 +0100)]
greybus: operation: fix locking issues

Fix unconditional re-enabling of interrupts in various operation
functions that can all be called with local interrupts disabled from USB
completion handlers.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: es2: fix USB id to not be the same as ES1
Greg Kroah-Hartman [Mon, 2 Mar 2015 16:52:07 +0000 (08:52 -0800)]
greybus: es2: fix USB id to not be the same as ES1

We don't want to bind to the ES1 device, that would be bad.

Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: gpio: fix memory leaks at init and exit
Johan Hovold [Fri, 13 Feb 2015 06:58:04 +0000 (14:58 +0800)]
greybus: gpio: fix memory leaks at init and exit

Fix three related memory leaks in the init an exit callbacks, where the
gpio-lines array was never freed at all and the controller data wasn't
freed in the init error path.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: gpio: add interrupt handling support
Matt Porter [Tue, 17 Feb 2015 15:48:23 +0000 (10:48 -0500)]
greybus: gpio: add interrupt handling support

Adds gpio interrupt handling support using an irqchip/irqdomain
instantiation inside the GB GPIO driver. This implementation works
on older kernels such as 3.10 that do not have the gpiolib irqchip
helpers. Any line on a Greybus gpiochip may be configured as an
interrupt. Once configured, IRQ event messages received from a
module fire off the registered interrupt handler.

Signed-off-by: Matt Porter <mporter@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: bundle: fix sleep-while-atomic in gb_bundle_destroy
Johan Hovold [Fri, 13 Feb 2015 03:28:09 +0000 (11:28 +0800)]
greybus: bundle: fix sleep-while-atomic in gb_bundle_destroy

Make sure to release the spin lock protecting the interface bundle lists
before tearing down the connections and removing the bundle device,
which are operations that may sleep.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Reviewed-by: Alex Elder <elder@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: connection: fix non-atomic allocations under spin lock
Johan Hovold [Thu, 12 Feb 2015 03:22:47 +0000 (11:22 +0800)]
greybus: connection: fix non-atomic allocations under spin lock

Use GFP_ATOMIC for IDA memory allocations under spin lock, which must
not sleep.

Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Reviewed-by: Alex Elder <elder@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: operation: add missing gb_operation_response_send() export
Matt Porter [Tue, 3 Feb 2015 18:17:39 +0000 (13:17 -0500)]
greybus: operation: add missing gb_operation_response_send() export

Export gb_operation_response_send() for other modules

Signed-off-by: Matt Porter <mporter@linaro.org>
Reviewed-by: Alex Elder <elder@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: spi:fix sparse warnings
Viresh Kumar [Tue, 27 Jan 2015 03:38:04 +0000 (09:08 +0530)]
greybus: spi:fix sparse warnings

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: build: android: Fix script which locates .ko files and moves them into ramdisk
Michael Scott [Tue, 27 Jan 2015 20:14:09 +0000 (12:14 -0800)]
greybus: build: android: Fix script which locates .ko files and moves them into ramdisk

- Fixed incorrect use of $$GREYBUS_SRC_PATH variable
- Added quotes around find pattern to stop shell expansion of "*"

Signed-off-by: Michael Scott <michael.scott@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: spi: add bridged-PHY spi protocol driver
Viresh Kumar [Fri, 23 Jan 2015 07:37:45 +0000 (13:07 +0530)]
greybus: spi: add bridged-PHY spi protocol driver

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: es1.c: wait until the last possible minute to start the svc messages
Greg Kroah-Hartman [Thu, 22 Jan 2015 13:19:25 +0000 (21:19 +0800)]
greybus: es1.c: wait until the last possible minute to start the svc messages

When initializing the USB device, we were starting up the svc message
queue before the cport urbs were allocated.  This might not be an issue
for "slower" machines, but not having any allocated urbs for a cport
might be an issue if we were to handle svc messages.

So wait until everything is properly initialized and allocated before
starting the svc urb.

Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: ap.c: our workqueue should be ordered
Greg Kroah-Hartman [Thu, 22 Jan 2015 12:33:41 +0000 (20:33 +0800)]
greybus: ap.c: our workqueue should be ordered

SVC messages come in in an "order", so don't mess them up by processing
them out of order.  Fix this by making our work queue ordered, which
should keep everything in line.

Reported-by: Perry Hung <perry@leaflabs.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: build: android: add all greybus modules to the build
Michael Scott [Fri, 23 Jan 2015 00:22:47 +0000 (16:22 -0800)]
greybus: build: android: add all greybus modules to the build

We now have several modules generated by the greybus build.
Let's add any *.ko files we find to the buid.

Signed-off-by: Michael Scott <michael.scott@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: build: android: replace hard-coded build destination with variable
Michael Scott [Fri, 23 Jan 2015 00:22:20 +0000 (16:22 -0800)]
greybus: build: android: replace hard-coded build destination with variable

Make sure destination for greybus modules is consistent and easier to maintain

Signed-off-by: Michael Scott <michael.scott@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: Revert "protocol: dedup protocol find code"
Greg Kroah-Hartman [Fri, 23 Jan 2015 02:06:24 +0000 (10:06 +0800)]
greybus: Revert "protocol: dedup protocol find code"

This reverts commit 241b5fefc54eae95239b0f7dc4e2b0db49457729 as it's
wrong, we want to insert into the correct place in the list.

Reported-by: Alex Elder <elder@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: Revert "protocol.c: fix a kernel panic caused by __gb_protocol_register"
Greg Kroah-Hartman [Fri, 23 Jan 2015 02:05:58 +0000 (10:05 +0800)]
greybus: Revert "protocol.c: fix a kernel panic caused by __gb_protocol_register"

This reverts commit 57131bf309d34568dd3b8f8e9da7a7ba25e9495e, it isn't
going to be needed as the patch this fixes will be reverted.

Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: i2c: fix name conflict between function and struct: gb_i2c_transfer_request
Viresh Kumar [Thu, 22 Jan 2015 06:40:38 +0000 (12:10 +0530)]
greybus: i2c: fix name conflict between function and struct: gb_i2c_transfer_request

'gb_i2c_transfer_request' is the name given to a function and a struct. Though
we don't get any compilation errors/warnings about it, but the names should be
unique.

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: i2c: fix name conflict between function and struct: gb_i2c_transfer_response
Viresh Kumar [Thu, 22 Jan 2015 07:12:39 +0000 (12:42 +0530)]
greybus: i2c: fix name conflict between function and struct: gb_i2c_transfer_response

'gb_i2c_transfer_response' is the name given to a function and a struct. Though
we don't get any compilation errors/warnings about it, but the names should be
unique.

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: protocol.c: fix a kernel panic caused by __gb_protocol_register
Alexandre Bailon [Thu, 22 Jan 2015 07:23:37 +0000 (15:23 +0800)]
greybus: protocol.c: fix a kernel panic caused by __gb_protocol_register

__gb_protocol_register check if the protocol is not already registred,
and then register it. It register in existing->lists but at this point,
existing is always NULL (we exist just before if not).
Use gb_protocols instead.

Signed-off-by: Alexandre Bailon <abailon@baylibre.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: remove unused version-response structs
Viresh Kumar [Wed, 21 Jan 2015 12:42:37 +0000 (18:12 +0530)]
greybus: remove unused version-response structs

These aren't used anymore and so can be removed.

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Alex Elder <elder@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: create get_version() routines with the help of a macro
Viresh Kumar [Wed, 21 Jan 2015 12:42:36 +0000 (18:12 +0530)]
greybus: create get_version() routines with the help of a macro

This gets rid of lots of duplication of code.

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Alex Elder <elder@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: uart: s/REQ/TYPE
Viresh Kumar [Wed, 21 Jan 2015 12:42:35 +0000 (18:12 +0530)]
greybus: uart: s/REQ/TYPE

Request type for all other protocols is defined like: GB_<protocol>_TYPE_<operation>,
but for UART is like: GB_<protocol>_REQ_<operation>.

Replace REQ with TYPE to make it consistent. It will also be useful in a later
patch that creates get_version() routines with the help of a macro.

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Alex Elder <elder@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: Remove "gb-" prefix from .c files
Viresh Kumar [Wed, 21 Jan 2015 10:40:41 +0000 (16:10 +0530)]
greybus: Remove "gb-" prefix from .c files

Some files are still prefixed with "gb-" with the reasoning that the modules
would be named so, i.e.  gb-*.ko. But this can be done by playing a bit in
Makefile instead and keep uniform naming of .c files.

Lets try it.

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Alex Elder <elder@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: Remove "-gb" suffix from .c files
Viresh Kumar [Wed, 21 Jan 2015 10:40:40 +0000 (16:10 +0530)]
greybus: Remove "-gb" suffix from .c files

Some files are prefixed with "gb-" and some are suffixed with "-gb". The
rationale behind the first one is that the modules would be named so, i.e.
gb-*.ko. But there is no reason to keep the "-gb" suffix in the second case.

Remove the unnecessary suffix.

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Alex Elder <elder@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: es2: add ES2 Greybus Host driver
Greg Kroah-Hartman [Wed, 21 Jan 2015 02:24:15 +0000 (10:24 +0800)]
greybus: es2: add ES2 Greybus Host driver

This is just a copy of ES1 for now, things will start to diverge soon.
Any common functionality will be factored out over time.

Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: es1: no need to assign a variable we return directly
Greg Kroah-Hartman [Wed, 21 Jan 2015 01:47:21 +0000 (09:47 +0800)]
greybus: es1: no need to assign a variable we return directly

In next_free_urb(), just return usb_alloc_urb(), don't waste the time
assigning it to a local variable that we then return.

Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: protocol: dedup protocol find code
Rui Miguel Silva [Tue, 20 Jan 2015 17:26:18 +0000 (17:26 +0000)]
greybus: protocol: dedup protocol find code

in protocol register replace the protocol find code with the call to the already
existing function.

Signed-off-by: Rui Miguel Silva <rmfrfs@gmail.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: es1: release urb on error path
Rui Miguel Silva [Tue, 20 Jan 2015 16:38:44 +0000 (16:38 +0000)]
greybus: es1: release urb on error path

if error is return when submiting the urb, we need to make sure to release the
urb from the pool, or from the dinamicly allocated. As in it, factor out the free
code and create the free_urb function.

Signed-off-by: Rui Miguel Silva <rmfrfs@gmail.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: es1: remove useless statement
Rui Miguel Silva [Tue, 20 Jan 2015 16:31:45 +0000 (16:31 +0000)]
greybus: es1: remove useless statement

just return the result of usb_alloc_urb up, no need to rededunt check for NULL

Signed-off-by: Rui Miguel Silva <rmfrfs@gmail.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: gb_operation: drop operation refcount on cancel
Perry Hung [Wed, 14 Jan 2015 21:19:27 +0000 (16:19 -0500)]
greybus: gb_operation: drop operation refcount on cancel

An extra reference is taken out on an operation in
gb_operation_request_send(). If the response never arrives, we need to
put back the reference.

Signed-off-by: Perry Hung <perry@leaflabs.com>
Tested-by: Mitchell Tasman <tasman@leaflabs.com>
Reviewed-by: Alex Elder <elder@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: gb_operation: replace timeout workqueue
Perry Hung [Wed, 14 Jan 2015 21:19:26 +0000 (16:19 -0500)]
greybus: gb_operation: replace timeout workqueue

If an operation is issued and the response never comes back,
gb_operation_timeout() cancels the operation but never wakes up the
waiter in gb_operation_request_send().

This patch removes the timeout workqueue and changes the request wait to
wait_for_completion_interruptible_timeout(), with timeout set to
OPERATION_TIMEOUT_DEFAULT.

Signed-off-by: Perry Hung <perry@leaflabs.com>
Reviewed-by: Alex Elder <elder@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: gpio-gb: remove checks for negative offset variable
Bill Pemberton [Fri, 16 Jan 2015 18:57:32 +0000 (13:57 -0500)]
greybus: gpio-gb: remove checks for negative offset variable

offset is defined as unsigned so there is no point checking for
negative values of offset.

Signed-off-by: Bill Pemberton <wfp5p@worldbroken.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: Add FIXME warnings for possible NULL dereferences
Bill Pemberton [Fri, 16 Jan 2015 02:18:27 +0000 (21:18 -0500)]
greybus: Add FIXME warnings for possible NULL dereferences

Signed-off-by: Bill Pemberton <wfp5p@worldbroken.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: gb-vibrator: remove useless if in timeout_store()
Bill Pemberton [Fri, 16 Jan 2015 02:18:26 +0000 (21:18 -0500)]
greybus: gb-vibrator: remove useless if in timeout_store()

val is an unsigned long so there is no point in checking if it is less
than zero.

Signed-off-by: Bill Pemberton <wfp5p@worldbroken.com>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: Move the es1_ap_desc.c file to Documentation directory
Greg Kroah-Hartman [Thu, 15 Jan 2015 01:08:03 +0000 (17:08 -0800)]
greybus: Move the es1_ap_desc.c file to Documentation directory

This .c file isn't needed by the kernel driver, it's there for firmware
developers only, so just move it into the Documentation directory to
reduce confusion.

Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
Reviewed-by: Alex Elder <elder@linaro.org>
9 years agogreybus: rename es1-ap-usb.ko to gb-es1.ko
Greg Kroah-Hartman [Thu, 15 Jan 2015 01:08:02 +0000 (17:08 -0800)]
greybus: rename es1-ap-usb.ko to gb-es1.ko

Use the "gb-" prefix for the ES1 Host controller driver.

Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
Reviewed-by: Alex Elder <elder@linaro.org>
9 years agogreybus: rename battery-gb.c to gb-battery.c
Greg Kroah-Hartman [Thu, 15 Jan 2015 01:08:01 +0000 (17:08 -0800)]
greybus: rename battery-gb.c to gb-battery.c

Use the "gb" prefix for module names, not a suffix.

Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
Reviewed-by: Alex Elder <elder@linaro.org>
9 years agogreybus: rename vibrator-gb.ko to gb-vibrator.ko
Greg Kroah-Hartman [Thu, 15 Jan 2015 01:08:00 +0000 (17:08 -0800)]
greybus: rename vibrator-gb.ko to gb-vibrator.ko

Use the "gb" prefix for module names, not a suffix.

Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
Reviewed-by: Alex Elder <elder@linaro.org>
9 years agogreybus: rename gpbridge.ko to gb-phy.ko
Greg Kroah-Hartman [Thu, 15 Jan 2015 01:07:59 +0000 (17:07 -0800)]
greybus: rename gpbridge.ko to gb-phy.ko

This module provides the Bridged PHY protocols, so name the thing
properly.

Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
Reviewed-by: Alex Elder <elder@linaro.org>
9 years agogreybus: Makefile: provide install target
Greg Kroah-Hartman [Wed, 14 Jan 2015 22:55:24 +0000 (14:55 -0800)]
greybus: Makefile: provide install target

Provide an install Makefile target for those that want to install the
kernel modules.

Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
--
v3: resend to list, somehow this thread got taken private and v2 never
    made it there.

v2: add -a option to depmod, thanks to Mitchell

9 years agogreybus: i2c-gb: fix bad message size in gb_i2c
Alexandre Bailon [Wed, 14 Jan 2015 09:23:49 +0000 (10:23 +0100)]
greybus: i2c-gb: fix bad message size in gb_i2c

The data_in_size variable was set to 1 for the status byte.
But now, the status byte has move to header. Then, the status byte
is "allocated" twice and cause bad message size error.

Signed-off-by: Alexandre Bailon <abailon@baylibre.com>
Reviewed-by: Alex Elder <elder@linaro.org>
Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: sysfs: put a \n at the end of all sysfs files
Greg Kroah-Hartman [Wed, 24 Dec 2014 00:09:26 +0000 (16:09 -0800)]
greybus: sysfs: put a \n at the end of all sysfs files

Right now some sysfs attributes have \n and some do not, so fix that and
put \n at the end of all of them to make it easier to parse things
properly in userspace.

Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
9 years agogreybus: protocol: fix oops when no protocol is assigned
Greg Kroah-Hartman [Wed, 24 Dec 2014 21:12:10 +0000 (13:12 -0800)]
greybus: protocol: fix oops when no protocol is assigned

When removing a connection with no protocol assigned to it, the kernel
oopses as we always thought protocols were always there.

Fix that problem, oopses are bad.

Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
Reviewed-by: Alex Elder <elder@linaro.org>
9 years agogreybus: protocol: name protocols.
Greg Kroah-Hartman [Wed, 24 Dec 2014 21:01:45 +0000 (13:01 -0800)]
greybus: protocol: name protocols.

We want to be able to "blame" a protocol for things at times, so give
them a name we can refer to them by.  Announce when they are added or
removed from the system so we have a chance to know what is going on
in the kernel logs.

Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
Reviewed-by: Alex Elder <elder@linaro.org>
9 years agogreybus: module: get rid of global list of modules
Greg Kroah-Hartman [Wed, 24 Dec 2014 21:01:46 +0000 (13:01 -0800)]
greybus: module: get rid of global list of modules

Use the list that the driver core keeps of our structure, no need to
duplicate it with a local list as well.  This gets rid of a static lock
too, always a nice thing to do.

Signed-off-by: Greg Kroah-Hartman <greg@kroah.com>
Reviewed-by: Alex Elder <elder@linaro.org>