Commit 6e740f9a8 (ARM: OMAP: Move omap-pm-noop.c local to mach-omap2)
moved omap-pm-noop to be local to mach-omap2. However, the makefile
entry got placed within ifeq ($(CONFIG_PM),y) which was not the
case earlier.
Fix the issue by moving it out of the ifeq ($(CONFIG_PM),y) in
the makefile as these stubs are needed also when PM is not set.
Reported-by: Russell King <rmk+kernel@arm.linux.org.uk>
Signed-off-by: Tony Lindgren <tony@atomide.com>
cpu_hotplug_pm_callback should have higher priority than
bsp_pm_callback which depends on cpu_hotplug_pm_callback to disable cpu hotplug
to avoid race during bsp online checking.
This is to hightlight the priorities between the two callbacks in case people
may overlook the order.
Ideally the priorities should be defined in macro/enum instead of fixed values.
To do that, a seperate patchset may be pushed which will touch serveral other
generic files and is out of scope of this patchset.
Signed-off-by: Fenghua Yu <fenghua.yu@intel.com>
Link: http://lkml.kernel.org/r/1352835171-3958-7-git-send-email-fenghua.yu@intel.com
Reviewed-by: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Because x86 BIOS requires CPU0 to resume from sleep, suspend or hibernate can't
be executed if CPU0 is detected offline. To make suspend or hibernate and
further resume succeed, CPU0 must be online.
Signed-off-by: Fenghua Yu <fenghua.yu@intel.com>
Link: http://lkml.kernel.org/r/1352835171-3958-6-git-send-email-fenghua.yu@intel.com
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Add smp_store_boot_cpu_info() to store cpu info for BSP during boot time.
Now smp_store_cpu_info() stores cpu info for bringing up BSP or AP after
it's offline.
Continue to online CPU0 in native_cpu_up().
Continue to offline CPU0 in native_cpu_disable().
Signed-off-by: Fenghua Yu <fenghua.yu@intel.com>
Link: http://lkml.kernel.org/r/1352835171-3958-5-git-send-email-fenghua.yu@intel.com
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
If CONFIG_BOOTPARAM_HOTPLUG_CPU is turned on, CPU0 hotplug feature is enabled
by default.
If CONFIG_BOOTPARAM_HOTPLUG_CPU is not turned on, CPU0 hotplug feature is not
enabled by default. The kernel parameter cpu0_hotplug can enable CPU0 hotplug
feature at boot.
Currently the feature is supported on Intel platforms only.
Signed-off-by: Fenghua Yu <fenghua.yu@intel.com>
Link: http://lkml.kernel.org/r/1352835171-3958-4-git-send-email-fenghua.yu@intel.com
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
New config switch CONFIG_BOOTPARAM_HOTPLUG_CPU0 sets default state of whether
the CPU0 hotplug is on or off.
If the switch is off, CPU0 is not hotpluggable by default. But the CPU0 hotplug
feature can still be turned on by kernel parameter cpu0_hotplug at boot.
If the switch is on, CPU0 is always hotpluggable.
The default value of the switch is off.
Signed-off-by: Fenghua Yu <fenghua.yu@intel.com>
Link: http://lkml.kernel.org/r/1352835171-3958-3-git-send-email-fenghua.yu@intel.com
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
If CONFIG_BOOTPARAM_HOTPLUG_CPU0 is turned on, CPU0 is hotpluggable. Otherwise,
by default CPU0 is not hotpluggable and kernel parameter cpu0_hotplug enables
CPU0 online/offline feature.
The documentations point out two known CPU0 dependencies. First, resume from
hibernate or suspend always starts from CPU0. So hibernate and suspend are
prevented if CPU0 is offline. Another dependency is PIC interrupts always go
to CPU0.
It's said that some machines may depend on CPU0 to poweroff/reboot. But I
haven't seen such dependency on a few tested machines.
Please let me know if you see any CPU0 dependencies on your machine.
Signed-off-by: Fenghua Yu <fenghua.yu@intel.com>
Link: http://lkml.kernel.org/r/1352835171-3958-2-git-send-email-fenghua.yu@intel.com
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Flush the cache so that the instructions written to the XOL area are
visible.
Signed-off-by: Rabin Vincent <rabin@rab.in>
Acked-by: Ananth N Mavinakayanahalli <ananth@in.ibm.com>
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Header length should be validated for all ACPI tables before accessing
any non-header field.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Link: http://lkml.kernel.org/r/509A9E6002000078000A7079@nat28.tlf.novell.com
Acked-by: Matt Fleming <matt.fleming@intel.com>
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Currently after the reset the sysc is written with hardcoded values.
The patch reads the sysc register and writes back the same value
after reset.
- Some unnecessary rev checks can be optimised.
- Also due to whatever reason the hwmod flags are changed
we will not reset the values.
- In some of the cases the minor values of the 2430 register
is different(0x37) in that case the autoidle setting may be missed.
Signed-off-by: Shubhrajyoti D <shubhrajyoti@ti.com>
Signed-off-by: Wolfram Sang <w.sang@pengutronix.de>
Currently the restore is done based on the flag
OMAP_I2C_FLAG_RESET_REGS_POSTIDLE.
This helps the following
- The driver is always capable of restoring regardless
of the off mode support being there or not.
- While testing omap2430 it is found that in case of certain
error paths (timeout) a reset is done. However the restore
never happens as it is dependent on the POSTIDLE flag.
The other option would be to call a restore in the reset
case. As there are only a few registers to be restored
the penalty in the idle case should not be much.
Reviewed-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Shubhrajyoti D <shubhrajyoti@ti.com>
Signed-off-by: Wolfram Sang <w.sang@pengutronix.de>
Implement reset as a separate function.
This will enable us to make sure that we don't do the
calculation again on every transfer.
Also at probe the reset is not added as the hwmod is doing that
for us.
Reviewed-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Shubhrajyoti D <shubhrajyoti@ti.com>
Signed-off-by: Wolfram Sang <w.sang@pengutronix.de>
re-factor omap_i2c_init() so that we can re-use it for resume.
While at it also remove the bufstate variable as we write it
in omap_i2c_resize_fifo for every transfer.
Reviewed-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Shubhrajyoti D <shubhrajyoti@ti.com>
Signed-off-by: Wolfram Sang <w.sang@pengutronix.de>
The commit [i2c: omap: use revision check for OMAP_I2C_FLAG_APPLY_ERRATA_I207]
uses the revision id instead of the flag. So the flag can be safely removed.
Reviewed-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Shubhrajyoti D <shubhrajyoti@ti.com>
Signed-off-by: Wolfram Sang <w.sang@pengutronix.de>
The dtrev is used only for the comments. Remove the same and use
the scheme instead to know if it is version2.
Signed-off-by: Shubhrajyoti D <shubhrajyoti@ti.com>
Reviewed-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Wolfram Sang <w.sang@pengutronix.de>
The errata i207 is enabled for 2430 and 3xxx. Use the revision check
to enable the erratum instead.
Reviewed-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Shubhrajyoti D <shubhrajyoti@ti.com>
Signed-off-by: Wolfram Sang <w.sang@pengutronix.de>
The revision register on OMAP4 is a 16-bit lo and a 16-bit
hi. Currently the driver reads only the lower 8-bits.
Fix the same by preventing the truncating of the rev register
for OMAP4.
Also use the scheme bit ie bit-14 of the hi register to know if it
is OMAP_I2C_IP_VERSION_2.
On platforms previous to OMAP4 the offset 0x04 is IE register whose
bit-14 reset value is 0, the code uses the same to its advantage.
Also since the omap_i2c_read_reg uses reg_map_ip_* a raw_readw is done
to fetch the revision register.
The dev->regs is populated after reading the rev_hi. A NULL check
has been added in the resume handler to prevent the access before
the setting of the regs.
Signed-off-by: Shubhrajyoti D <shubhrajyoti@ti.com>
Reviewed-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Wolfram Sang <w.sang@pengutronix.de>
Some GPIO expanders need some early pin control muxing. Due to
legacy boards sometimes the driver uses subsys_initcall instead of
module_init. This patch takes advantage of defer probe feature
and pin control in order to wait until pin control probing before
GPIO driver probing. It has been tested on OMAP5 board with TCA6424
driver.
Signed-off-by: Sebastien Guiriec <s-guiriec@ti.com>
Acked-by: Shubhrajyoti D <shubhrajyoti@ti.com>
Reviewed-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Wolfram Sang <w.sang@pengutronix.de>
if we allow compiler reorder our writes, we could
fall into a situation where dev->buf_len is reset
for no apparent reason.
This bug was found with a simple script which would
transfer data to an i2c client from 1 to 1024 bytes
(a simple for loop), when we got to transfer sizes
bigger than the fifo size, dev->buf_len was reset
to zero before we had an oportunity to handle XDR
Interrupt. Because dev->buf_len was zero, we entered
omap_i2c_transmit_data() to transfer zero bytes,
which would mean we would just silently exit
omap_i2c_transmit_data() without actually writing
anything to DATA register. That would cause XDR
IRQ to trigger forever and we would never transfer
the remaining bytes.
After adding the memory barrier, we also drop resetting
dev->buf_len to zero in omap_i2c_xfer_msg() because
both omap_i2c_transmit_data() and omap_i2c_receive_data()
will act until dev->buf_len reaches zero, rendering the
other write in omap_i2c_xfer_msg() redundant.
This patch has been tested with pandaboard for a few
iterations of the script mentioned above.
Signed-off-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Wolfram Sang <w.sang@pengutronix.de>
if we allow compiler reorder our writes, we could
fall into a situation where dev->buf_len is reset
for no apparent reason.
This bug was found with a simple script which would
transfer data to an i2c client from 1 to 1024 bytes
(a simple for loop), when we got to transfer sizes
bigger than the fifo size, dev->buf_len was reset
to zero before we had an oportunity to handle XDR
Interrupt. Because dev->buf_len was zero, we entered
omap_i2c_transmit_data() to transfer zero bytes,
which would mean we would just silently exit
omap_i2c_transmit_data() without actually writing
anything to DATA register. That would cause XDR
IRQ to trigger forever and we would never transfer
the remaining bytes.
After adding the memory barrier, we also drop resetting
dev->buf_len to zero in omap_i2c_xfer_msg() because
both omap_i2c_transmit_data() and omap_i2c_receive_data()
will act until dev->buf_len reaches zero, rendering the
other write in omap_i2c_xfer_msg() redundant.
This patch has been tested with pandaboard for a few
iterations of the script mentioned above.
Signed-off-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Wolfram Sang <w.sang@pengutronix.de>
Work claiming wants to be SMP-safe.
And by the time we try to claim a work, if it is already executing
concurrently on another CPU, we want to succeed the claiming and queue
the work again because the other CPU may have missed the data we wanted
to handle in our work if it's about to complete there.
This scenario is summarized below:
CPU 1 CPU 2
----- -----
(flags = 0)
cmpxchg(flags, 0, IRQ_WORK_FLAGS)
(flags = 3)
[...]
xchg(flags, IRQ_WORK_BUSY)
(flags = 2)
func()
if (flags & IRQ_WORK_PENDING)
(not true)
cmpxchg(flags, flags, IRQ_WORK_FLAGS)
(flags = 3)
[...]
cmpxchg(flags, IRQ_WORK_BUSY, 0);
(fail, pending on CPU 2)
This state machine is synchronized using [cmp]xchg() on the flags.
As such, the early IRQ_WORK_PENDING check in CPU 2 above is racy.
By the time we check it, we may be dealing with a stale value because
we aren't using an atomic accessor. As a result, CPU 2 may "see"
that the work is still pending on another CPU while it may be
actually completing the work function exection already, leaving
our data unprocessed.
To fix this, we start by speculating about the value we wish to be
in the work->flags but we only make any conclusion after the value
returned by the cmpxchg() call that either claims the work or let
the current owner handle the pending work for us.
Changelog-heavily-inspired-by: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Acked-by: Steven Rostedt <rostedt@goodmis.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Paul Gortmaker <paul.gortmaker@windriver.com>
Cc: Anish Kumar <anish198519851985@gmail.com>
The IRQ_WORK_BUSY flag is set right before we execute the
work. Once this flag value is set, the work enters a
claimable state again.
So if we have specific data to compute in our work, we ensure it's
either handled by another CPU or locally by enqueuing the work again.
This state machine is guanranteed by atomic operations on the flags.
So when we set IRQ_WORK_BUSY without using an xchg-like operation,
we break this guarantee as in the following summarized scenario:
CPU 1 CPU 2
----- -----
(flags = 0)
old_flags = flags;
(flags = 0)
cmpxchg(flags, old_flags,
old_flags | IRQ_WORK_FLAGS)
(flags = 3)
[...]
flags = IRQ_WORK_BUSY
(flags = 2)
func()
(sees flags = 3)
cmpxchg(flags, old_flags,
old_flags | IRQ_WORK_FLAGS)
(give up)
cmpxchg(flags, 2, 0);
(flags = 0)
CPU 1 claims a work and executes it, so it sets IRQ_WORK_BUSY and
the work is again in a claimable state. Now CPU 2 has new data to process
and try to claim that work but it may see a stale value of the flags
and think the work is still pending somewhere that will handle our data.
This is because CPU 1 doesn't set IRQ_WORK_BUSY atomically.
As a result, the data expected to be handle by CPU 2 won't get handled.
To fix this, use xchg() to set IRQ_WORK_BUSY, this way we ensure the CPU 2
will see the correct value with cmpxchg() using the expected ordering.
Changelog-heavily-inspired-by: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Acked-by: Steven Rostedt <rostedt@goodmis.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Paul Gortmaker <paul.gortmaker@windriver.com>
Cc: Anish Kumar <anish198519851985@gmail.com>
In device tree enabled setups requiring preallocated memory for storing keymap
is quite often awkward, so let's provide an option of allocating it directly
in matrix_keypad_build_keymap().
Reviewed-by: Alban Bedel <alban.bedel@avionic-design.de>
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Usage of TSTAMP_* macros has gone in 2010 in 730947bc (V4L/DVB: vivi:
clean up and a major overhaul) but the macros remain. Say goodbye to
them.
Signed-off-by: Kirill Smelkov <kirr@mns.spb.ru>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Use a proper list iterator instead of an open-coded loop and remove a
superfluous list head initialisation.
Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
This patch makes a loop look cleaner by replacing a "break" and a "continue"
in its body by a single "if".
Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
Instead of listing explicitly all boards in the .dt_compat field of
the DT_MACHINE_START structure for Armada 370/XP, use instead a
compatible string that is common to all boards using the Armada
370/XP.
This allows to add new boards by just using a different Device Tree,
without having to modify the source code of the kernel.
Note that the name of the array containing the compatible string is
also renamed, to reflect the fact that it no longer contains the list
of all boards.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Acked-by: Gregory CLEMENT <gregory.clement@free-electrons.com>
Tested-by: Gregory CLEMENT <gregory.clement@free-electrons.com>
All the Device Tree files for Armada 370 and XP SoCs and boards use
the "armada-370-xp" common compatible string, except
armada-370-xp.dtsi which was specifying "armada_370_xp".
Fix this inconsistency by making armada-370-xp.dtsi declare a
compatible string of "armada-370-xp" like everyone else.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Acked-by: Gregory CLEMENT <gregory.clement@free-electrons.com>
Tested-by: Gregory CLEMENT <gregory.clement@free-electrons.com>
This platform, available from Globalscale has an Armada 370. For now,
only the serial port is supported. Support for network, USB and other
peripherals will be added as drivers for them become available for
Armada 370 in mainline.
Signed-off-by: Gregory CLEMENT <gregory.clement@free-electrons.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
---
This is 3.8 material.
Changes from original version posted by Gregory:
* Renamed .dts file to armada-370-mirabox.dts
* Change compatible string to 'globalscale,mirabox'
* Remove compatible string from armada-370-xp.c
* Removed references to MBX0001
This platform, available in Japan from PlatHome, has a dual-core
Armada XP, the MV78260. For now, only the two serial ports and the
three front LEDs are supported. Support for SMP, network, SATA, USB
and other peripherals will be added as drivers for them become
available for Armada XP in mainline.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Acked-by: Gregory CLEMENT <gregory.clement@free-electrons.com>
Tested-by: Gregory CLEMENT <gregory.clement@free-electrons.com>
---
This is 3.8 material.
Changes since v2:
* Renamed the .dts file to armada-xp-openblocks-ax3-4.dts
* Removed the compatible string from armada-370-xp.c (which now only
lists the common SoC compatible string)
Changes since v1:
* Renamed the board to OpenBlocks AX3-4, since there is a variant
called AX3-2 which has less RAM, and no mini PCIe port. Requested
by Andrew Lunn.
* Fix the amount of memory to 3 GB. In fact, the board has 1 GB
soldered, and 2 GB in a SODIMM slot (which is therefore
removable). But as the board is delivered as is, we'll assume it
has 3 GB of memory by default.
Add support for retrieving TTC configuration from device tree. This
includes the ability to pull information about the driving clocks from
the of_clk bindings.
Signed-off-by: Josh Cartwright <josh.cartwright@ni.com>
Acked-by: Michal Simek <michal.simek@xilinx.com>
Make the Zynq platform use the newly created zynq clk bindings.
Signed-off-by: Josh Cartwright <josh.cartwright@ni.com>
Acked-by: Michal Simek <michal.simek@xilinx.com>
Provide simplified models for the necessary clocks on the zynq-7000
platform. Currently, the PLLs, the CPU clock network, and the basic
peripheral clock networks (for SDIO, SMC, SPI, QSPI, UART) are modelled.
OF bindings are also provided and documented.
Signed-off-by: Josh Cartwright <josh.cartwright@ni.com>
Signed-off-by: Soren Brinkmann <soren.brinkmann@xilinx.com>
Acked-by: Mike Turquette <mturquette@linaro.org>
Acked-by: Michal Simek <michal.simek@xilinx.com>
The recent change for USB-audio disconnection race fixes introduced a
mutex deadlock again. There is a circular dependency between
chip->shutdown_rwsem and pcm->open_mutex, depicted like below, when a
device is opened during the disconnection operation:
A. snd_usb_audio_disconnect() ->
card.c::register_mutex ->
chip->shutdown_rwsem (write) ->
snd_card_disconnect() ->
pcm.c::register_mutex ->
pcm->open_mutex
B. snd_pcm_open() ->
pcm->open_mutex ->
snd_usb_pcm_open() ->
chip->shutdown_rwsem (read)
Since the chip->shutdown_rwsem protection in the case A is required
only for turning on the chip->shutdown flag and it doesn't have to be
taken for the whole operation, we can reduce its window in
snd_usb_audio_disconnect().
Reported-by: Jiri Slaby <jslaby@suse.cz>
Cc: <stable@vger.kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
A selection of voltage or current values (AKA states) should always
be specified when using a GPIO regulator. If there are no switchable
states then the fixed regulators should be used instead.
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
If consumers wish to set voltages based on a tolerance it stands to reason
that they will also want to query for support in the same manner.
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: Liam Girdwood <lrg@ti.com>