CHROMIUM: ALSA: Add new DSP loader callback routines

BUG=chrome-os-partner:12327

Change-Id: I1dff2d4d672b0b03eb486fb97295367db4b07354
Signed-off-by: Chih-Chung Chang <chihchung@chromium.org>
Signed-off-by: Ian Minett <ian_minett@creativelabs.com>
Reviewed-on: https://gerrit.chromium.org/gerrit/32126
Reviewed-by: Dylan Reid <dgreid@chromium.org>

CHROMIUM: ALSA: Make snd_sgbuf_get_{ptr|addr}() available for non-SG cases.

BUG=chrome-os-partner:12327

Change-Id: Iae062d860bc4044db79284290ff28b29c58b29ce
Signed-off-by: Chih-Chung Chang <chihchung@chromium.org>
Signed-off-by: Ian Minett <ian_minett@creativelabs.com>
Reviewed-on: https://gerrit.chromium.org/gerrit/32134

CHROMIUM: ramoops: Copy the entire buffer to deal with bit rot

If unclean shutdown is done by yanking power rails, the RAM may be in a
state where it is not getting refreshed for a short period of time. This
can cause bit rotting in RAM and throw off the wraparound and buffer
size calculation logic. To get around this, copy the entire console
buffer. Note that this, of course, won't do anything to fix the bit
rotting but the intent here is that some logs are better than none.

BUG=chrome-os-partner:13249
TEST=on a snow or link system do a warm reboot by pressing
alt-volume_up-r keys together. The system should reboot. Upon bootup,
check contents of "/dev/pstore/console-ramoops" and ensure they are from
dmesg of the previous boot.

Change-Id: I68b2b10a6b3fc29329979cc0c659b440436b771f
Signed-off-by: Sameer Nanda <snanda@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/32180
Reviewed-by: Doug Anderson <dianders@chromium.org>

exynos:gsc:modify number of output/capture buffers

gscaler src buffer count as well as destination buffer
count increased to 32.
This is required to interface to MFC for csc, as MFC
demands 32 capture buffers for some H264 streams

BUG=streams with more than 16 dpb buffers fail
TEST=tested through browser for streams with 20 dpbs

Change-Id: I817d65c122eb8974b6c658f65fe48e1e1131c83e
Signed-off-by: Kiran AVND <avnd.kiran@samsung.com>
Reviewed-on: https://gerrit.chromium.org/gerrit/32133
Tested-by: kiran <kiran@chromium.org>
Reviewed-by: Shaik Ameer Basha <shaik.ameer@samsung.com>
Reviewed-by: Pawel Osciak <posciak@google.com>
Commit-Ready: Arun Mankuzhi <arun@chromium.org>
Reviewed-by: Arun Mankuzhi <arun@chromium.org>

CHROMIUM: dm: fix newlines in printouts

Turns out that we print double newlines:

[    3.799802] device-mapper: init: attempting early device configuration.
[    3.799807]

So, don't add the newline at the end of DM*() macros.

BUG=none
TEST=look at dmesg after booting

Change-Id: I039a3451c5a757641f1c0211845fc0bdc218b2bc
Signed-off-by: Olof Johansson <olofj@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/32154
Reviewed-by: Mandeep Singh Baines <msb@chromium.org>

arm: exynos: Gracefully exit on suspend failure

As per Pg 641 Pt 7 of Exynos_5250_SCP_UM_REV1.00:

When there are pending interrupt events, WFI/WFE instruction are
ignored. To cancel the power-down sequence follow these steps:
1) Disable system power-down through CENTRAL_SEQ_CONFIGURATION
and CENTRAL_SEQ_CONFIGURATION_DMC registers
2) Clear WAKEUP_STAT and WAKEUP_MASK_DMC registers
3) Enable interrupt service routine for CPU

Code for early wakeup already exists; remove the panic on suspend
failure, clear the wakeup state register and return error from
cpu_suspend to indicate a failed suspend.

BUG=chrome-os-partner:12807
TEST=I still have not been able to reproduce this bug. I am on ToT
kernel, u-boot and ec running Jon's suspend_test.sh script. On
commenting out cpu_do_idle in the suspend code, the system resumes
successfully taking the early_wakeup path.
cat /sys/kernel/debug/suspend_stats shows
(success: 0
 fail: 1
 last_failed_errno: -16)

Change-Id: I082664ae77db6116bccff5555a8d43412c780277
Signed-off-by: Abhilash Kesavan <a.kesavan@samsung.com>
Reviewed-on: https://gerrit.chromium.org/gerrit/30992
Reviewed-by: Jon Kliegman <kliegs@chromium.org>
Commit-Ready: Olof Johansson <olofj@chromium.org>
Reviewed-by: Olof Johansson <olofj@chromium.org>

CHROMIUM: ALSA: hda/ca0132 - make initialization faster.

Remove unnecessary dspio_send() from dspio_write_wait().
Check dsp status before sleep and sleep for shorter periods.

BUG=chrome-os-partner:12307
TEST=build and install, check dmesg.

Signed-off-by: Chih-Chung Chang <chihchung@chromium.org>
Change-Id: I0e1d5b0d29c446f29beea704c1b2cca22524c142
Reviewed-on: https://gerrit.chromium.org/gerrit/32023
Reviewed-by: Dylan Reid <dgreid@chromium.org>
Tested-by: Chih-Chung Chang <chihchung@chromium.org>
Commit-Ready: Chih-Chung Chang <chihchung@chromium.org>

CHROMIUM: config: exynos5: enable xfrm_user module

xfrm_user kernel module is needed for supporting LT2P/IPsec VPN via
strongSwan.

BUG=chromium-os:21495
TEST=Build chromeos-kernel and test L2TP/IPsec VPN.

Change-Id: Ic7635a1057642d8eaee0a15f6d1db210804a2912
Signed-off-by: Ben Chan <benchan@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/32084
Reviewed-by: Darin Petkov <petkov@chromium.org>

HROMIUM: ALSA: hda/ca0132 - latency patch from Creative.

This change improves the latency and report the correct estimate
delay when dsp loaded or not. DSP parameter set incorrect issue has
also been fixed.

BUG=chrome-os-partner:12926
TEST=build kernel, record some sound by arecord command. Google hangout
should not hear echo.

Change-Id: Ibade3a0cfbd956379ff1690cd1e0f3f31eedd6f9
Signed-off-by: Hsinyu Chao <hychao@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31827
Reviewed-by: Puneet Kumar <puneetster@chromium.org>

UPSTREAM: [media] uvcvideo: Reset the bytesused field when recycling an erroneous buffer

Buffers marked as erroneous are recycled immediately by the driver if
the nodrop module parameter isn't set. The buffer payload size is reset
to 0, but the buffer bytesused field isn't. This results in the buffer
being immediately considered as complete, leading to an infinite loop in
interrupt context.

Fix the problem by resetting the bytesused field when recycling the
buffer.

Cc: <stable@vger.kernel.org>
Signed-off-by: Jayakrishnan Memana <jayakrishnan.memana@maxim-ic.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
(cherry picked from commit 8a3f0ede2b3f5477122060af1a816c6bbf09fcd2)

BUG=chromium-os:33591
TEST=compile

Change-Id: Ibbb507bff1e57b7a92e84c36457de04d2115b3b8
Signed-off-by: Mandeep Singh Baines <msb@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/32071
Reviewed-by: Vincent Palatin <vpalatin@chromium.org>
Reviewed-by: Pawel Osciak <posciak@google.com>

CHROMIUM: gobi: call usbnet_disconnect after qc_deregister

usbnet_disconnect should be called after qc_deregister and at the end of
qcnet_disconnect as qc_deregister may still need to access the
underlying usb_device during client_free -> cid_free -> write_sync.

BUG=chrome-os-partner:12812
TEST=Tested the following on a system with a Gobi modem:
1. Run suspend/resume stress test.
2. Manually connect to a cellular network, suspend and then resume the
   system. Observe from syslog how qcusbnet is released and
   re-initialized.

Change-Id: I0e5f2fd12896992e0da5b99e8b92274f87e1640c
Signed-off-by: Ben Chan <benchan@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/32063
Reviewed-by: Gary Morain <gmorain@chromium.org>

media:gscalar: fix csc RGB out/in format settings

macros used to set RGB output and input csc aren't correct.
This causes RGB(16,16,16) to appear where we expect RGB(0,0,0).

BUG:chrome-os-partner:13370
TEST: http://code.entropywave.com/test-media/gamut/gamut-theora-bt470.ogv
      Transcode using ffmpeg: to gamut-theora-bt470.mp4(h264 content)
      Play on snow; with HW decode enabled
      checkboard pattern not observed, playback similar to software decoding.

Change-Id: Ia12390e39f68bde6cb4a5a6202324b96761dd26e
Signed-off-by: Kiran AVND <avnd.kiran@samsung.com>
Reviewed-on: https://gerrit.chromium.org/gerrit/32020
Reviewed-by: Shaik Ameer Basha <shaik.ameer@samsung.com>
Tested-by: Shaik Ameer Basha <shaik.ameer@samsung.com>
Reviewed-by: Sunil Mazhavanchery <sunilm@samsung.com>
Reviewed-by: Pawel Osciak <posciak@google.com>
Commit-Ready: Kiran Avnd <kiran@chromium.org>

Revert "CHROMIUM: ramoops: Copy the entire buffer to deal with bit rot"

This reverts commit 4aa82798833de5023d60a81cd1abd118da03a0f1

Change-Id: Ie3272c25d9c0f2aced79588faac172395d07afb4
Reviewed-on: https://gerrit.chromium.org/gerrit/32061
Commit-Ready: Andrew de los Reyes <adlr@chromium.org>
Reviewed-by: Andrew de los Reyes <adlr@chromium.org>
Tested-by: Andrew de los Reyes <adlr@chromium.org>

gpu: vithar: Fix compilation error when disabling T6XX_DVFS

This patch fixes the build errors that occur when T6XX_DVFS is disabled.

Signed-off-by: Shariq Hasnain <shariq.786@samsung.com>
BUG=None
TEST=Build tested with T6XX_DVFS disabled.

Change-Id: I999489ac6e55a92fc68bac5ff67fb91cf926865c
Reviewed-on: https://gerrit.chromium.org/gerrit/29584
Commit-Ready: shariq hasnain <shariq.786@samsung.com>
Reviewed-by: shariq hasnain <shariq.786@samsung.com>
Tested-by: shariq hasnain <shariq.786@samsung.com>
Commit-Ready: Sam Leffler <sleffler@chromium.org>
Reviewed-by: Sam Leffler <sleffler@chromium.org>

gpu: vithar: GPU ASV code added

This patch select the appropiate supply voltage for GPU.

Signed-off-by: Shariq Hasnain <shariq.786@samsung.com>
BUG=chrome-os-partner:11795
TEST=Build and boot snow. A group is chosen from ASV voltage
table and webGL applications are running without any issues.

Change-Id: I96faafc4e5d3a6654be54e55d60d887b01643614
Reviewed-on: https://gerrit.chromium.org/gerrit/31694
Commit-Ready: shariq hasnain <shariq.786@samsung.com>
Reviewed-by: shariq hasnain <shariq.786@samsung.com>
Tested-by: shariq hasnain <shariq.786@samsung.com>
Commit-Ready: Sam Leffler <sleffler@chromium.org>
Reviewed-by: Sam Leffler <sleffler@chromium.org>

UPSTREAM: dmaengine: pl330: dont complete descriptor for cyclic dma

Commit eab215855803 ("dmaengine: pl330: dont complete descriptor for
cyclic dma") wrongly completes descriptor for cyclic dma, hence following
BUG_ON is still hit with cyclic DMA operations.

kernel BUG at drivers/dma/dmaengine.h:53!

BUG=chrome-os-partner:12565
TEST=built for emerge-daisy chromeos-kernel and booted on one board.

Signed-off-by: Tushar Behera <tushar.behera@linaro.org>
Acked-by: Jassi Brar <jaswinder.singh@linaro.org>
Signed-off-by: Vinod Koul <vinod.koul@linux.intel.com>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Alim Akhtar <alim.akhtar@gmail.com>
(cherry picked from commit 30c1dc0ff30b5552e8af555265dbeac5637cbb48)

Change-Id: Ia705fcbb3b19b0d3a51295bf273d039fb76a1a72
Reviewed-on: https://gerrit.chromium.org/gerrit/31830
Commit-Ready: naveen krishna chatradhi <naveen@chromium.org>
Tested-by: naveen krishna chatradhi <naveen@chromium.org>
Reviewed-by: Hatim Ali R.V <hatim.rv@samsung.com>
Reviewed-by: naveen krishna chatradhi <naveen@chromium.org>

UPSTREAM: DMA: PL330: Fix racy mutex unlock

pl330_update() stores a pointer to the thrd->req that finished, which
contains a pointer to the corresponding pl330_req.  This is done with
the pl330_lock held.  Then, it iterates through the req_done list,
calling the callback for each of the requests that are done.  The
problem is that the driver releases the lock before calling the
callback for each of the callbacks.  pl330_submit_req() running in
another processor can then acquire the lock and insert another request
in one of the thrd->req that hasn't been processed yet, replacing the
pointer to pl330_req there.  When the callback returns in
pl330_update() and the next rqdone is popped from the list, it
dereferences the pl330_req pointer to the just scheduled pl330_req,
instead of the one that has finished, calling pl330 with the wrong r.

This patch fixes this by storing the pointer to pl330_req directly in
the list.

BUG=chrome-os-partner:12565
TEST=built for emerge-daisy chromeos-kernel and booted on one board.

Signed-off-by: Javi Merino <javi.merino@arm.com>
Cc: Jassi Brar <jaswinder.singh@linaro.org>
Acked-by: Jassi Brar <jaswinder.singh@linaro.org>
Signed-off-by: Vinod Koul <vinod.koul@linux.intel.com>
Signed-off-by: Alim Akhtar <alim.akhtar@gmail.com>
(cherry picked from commit fdec53d5203ee3d44be7c09d1524e3a6287d46a7)

Change-Id: If0f7f4dd93dc12acf38047c8d90406bd172eb875
Reviewed-on: https://gerrit.chromium.org/gerrit/31829
Commit-Ready: naveen krishna chatradhi <naveen@chromium.org>
Tested-by: naveen krishna chatradhi <naveen@chromium.org>
Reviewed-by: Hatim Ali R.V <hatim.rv@samsung.com>
Reviewed-by: naveen krishna chatradhi <naveen@chromium.org>

CHROMIUM: ASoC: Samsung: Don't reset playing streams.

b796fed30720 "Asoc: Samsung: Add I2S S/W RST in startup function"
introduced a bug where whenever a new stream was started, all i2s
streams would be reset.  This caused all full-duplex use cases to fail,
no hangouts, no GTalk.

Instead of always resetting, only reset if nothing is active.  The reset
will still happen when it is needed, but won't if something is already
playing.

BUG=chrome-os-partner:13282
TEST=aplay and arecord at the same time should work.

Change-Id: I1bc7c87b8deff710f408565b2b834c25d4d0053d
Signed-off-by: Dylan Reid <dgreid@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31681
Reviewed-by: Doug Anderson <dianders@chromium.org>

CHROMIUM: ramoops: Copy the entire buffer to deal with bit rot

If unclean shutdown is done by yanking power rails, the RAM may be in a
state where it is not getting refreshed for a short period of time. This
can cause bit rotting in RAM and throw off the wraparound and buffer
size calculation logic. To get around this, copy the entire console
buffer. Note that this, of course, won't do anything to fix the bit
rotting but the intent here is that some logs are better than none.

Also squeezed in a minor cleanup of renaming headp to tailp since it
really tracks the tail, not head.

BUG=chrome-os-partner:13249
TEST=on a snow system do a warm reboot by either issuing the
"warm_reboot" command from servo ec console or by pressing
alt-volume_up-r keys together. The system should reboot. Upon bootup,
check contents of "/dev/pstore/console-ramoops" and ensure they are from
dmesg of the previous boot.

Change-Id: I999e95bf854fff26265404a680255f25cc2f7fe6
Signed-off-by: Sameer Nanda <snanda@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31909
Reviewed-by: Mandeep Singh Baines <msb@chromium.org>

CHROMIUM: Correct i2c wait free time calculation

We should be using usecs_to_jiffies() for microsecond values. Correct
this.

BUG=chrome-os-partner:12420
BRANCH=snow
TEST=manual
On kernel:

while true; do i2cdump  -f -y 4 0x48; done >/dev/null

On EC:

pmu 1000

See that there are no errors.

Change-Id: I589721e200ad9ad7d872ee952aece84ca2bec923
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31924
Reviewed-by: Doug Anderson <dianders@chromium.org>

CHROMIUM: Increase i2c arbitration wait time

We currently intend to wait 50ms for the EC to release the i2c bus
(although due to a bug soon to be fixed, we wait longer).

The longest transaction we expect the EC to perform is 4 bytes, which
should be about 0.5ms, but if there is a bug in the EC driver the time
could wind out significantly.

Change the wait time to 500ms.

BUG=chrome-os-partner:12420
BRANCH=snow
TEST=manual
On kernel:

while true; do i2cdump  -f -y 4 0x48; done >/dev/null

On EC:

pmu 1000

See that there are no errors.

Change-Id: I2ab5c6f0ac1182485ac1a4f60e4e76ba83fb3987
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31923
Reviewed-by: Doug Anderson <dianders@chromium.org>

CHROMIUM: dts: Set no-ack-irq for tps65090

Set no-ack-irq to prevent the tps from acknowledging interrupts
on the tps chip.

BUG=chrome-os-partner:13109
TEST=Tested with and without devicetree flag set.
  Validated message in probe matches devicetree setting
  Unplugged and replugged power cable, confirmed battery icon
    updated correctly and debug messages (removed in CL) outputted
    the correct logic path.

Change-Id: I90df39c9fa4f7fa34df9701dc3e50522361cddea
Signed-off-by: Jonathan Kliegman <kliegs@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31896
Reviewed-by: Doug Anderson <dianders@chromium.org>

drivers: mfd: tps65090: Add devicetree flag to not ack tps interrupts

On some systems the tps65090 chip can be connected to multiple masters.
Add a device tree option to disable checking interrupts if a configuration
is present to prevent race conditions with only one device seeing the
interrupt.

BUG=chrome-os-partner:13109
TEST=Tested with and without devicetree flag set.
  Validated message in probe matches devicetree setting
  Unplugged and replugged power cable, confirmed battery icon
    updated correctly and debug messages (removed in CL) outputted
    the correct logic path.

Change-Id: I1d196373d65b36def67eeeb053f60c2471729d9c
Signed-off-by: Jonathan Kliegman <kliegs@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31895
Reviewed-by: Doug Anderson <dianders@chromium.org>

CHROMIUM: Input: cyapa - skip cyapa_dump_data if no DEBUG

ftrace shows this debugging function is still consuming several ~10 usec
even with DEBUG disabled.

Signed-off-by: Daniel Kurtz <djkurtz@chromium.org>
BUG=chromium-os:33910
TEST=cd /sys/kernel/debug/tracing/
TEST=echo 'cyapa*' > set_ftrace_filter
TEST=echo 'cyapa_irq' > set_graph_function
TEST=echo function_graph > current_tracer
TEST=move finger on touch surface
TEST=less trace
  => cyapa_dump_data() only present when DEBUG defined in cyapa.c

Change-Id: Ia8c789dfb01dda52c31167174a8a4731c965b603
Reviewed-on: https://gerrit.chromium.org/gerrit/31464
Commit-Ready: Daniel Kurtz <djkurtz@chromium.org>
Reviewed-by: Daniel Kurtz <djkurtz@chromium.org>
Tested-by: Daniel Kurtz <djkurtz@chromium.org>

UPSTREAM: ARM: Remove current_mm per-cpu variable

The current_mm variable was used to store the new mm between the
switch_mm() and switch_to() calls where an IPI to reset the context
could have set the wrong mm. Since the interrupts are disabled during
context switch, there is no need for this variable, current->active_mm
already points to the current mm when interrupts are re-enabled.

Reviewed-by: Will Deacon <will.deacon@arm.com>
Tested-by: Will Deacon <will.deacon@arm.com>
Reviewed-by: Frank Rowand <frank.rowand@am.sony.com>
Tested-by: Marc Zyngier <Marc.Zyngier@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
(cherry picked from commit e323969ccda2d69f02e047c08b03faa09215c72a)

BUG=chromium-os:34007
TEST=generic system testing

Change-Id: I3ba9043520e275d593526f97aa2ecdd877effa4a
Signed-off-by: Olof Johansson <olofj@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31852
Reviewed-by: Mandeep Singh Baines <msb@chromium.org>

CHROMIUM: mfd: tps65090 uses rising edge interrupt

This is to avoid a conflict with the EC, which uses the preceding
falling edge for interrupts.

BUG=chrome-os-partner:13109
TEST=udevadm monitor, plug and unplug AC.  Should see power_supply
events.

Change-Id: If4a8740a6ca55821928410f693c2ad9b128e66b5
Signed-off-by: Simon Que <sque@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31735
Reviewed-by: Simon Glass <sjg@chromium.org>

CHROMIUM: dts: Configure BUCK2,3,4 to standby opmode.

BUCK2,3,4 drive vdd_arm, vdd_int and vdd_g3d respectively.  These
exynos regulators can safely be turned off during standby.  They were
previously off during S3 but recent op_mode additions to the device
tree have set them to always on.

CL disables these regulators in standby.

BUG=chrome-os-partner:11256
TEST=manual,
enter standby ... measure power on arm,int,g3d to be zero.  Resume
from suspend without issues.

Signed-off-by: Todd Broch <tbroch@chromium.org>
Change-Id: Ib714e1fe166afbd2607574eaabc4819a492b3fa1
Reviewed-on: https://gerrit.chromium.org/gerrit/31660
Reviewed-by: Abhilash Kesavan <a.kesavan@samsung.com>
Reviewed-by: Bryan Freed <bfreed@chromium.org>

CHROMIUM: leds: Fix chromeos keyboard LED suspend behavior

Should use the built-in suspend/resume functions in led-class.c instead.

Also renamed the set function to correctly reflect the device type.

BUG=chrome-os-partner:13273
TEST=set the keyboard backlight to various levels (low, medium, full)
and run powerd_suspend.  The backlight should turn off.  Resume.  The
backlight should be restored to its previous value (maybe with a delay).

Change-Id: Ifd56debd6f157b71729bf563b4242c58709549f7
Signed-off-by: Simon Que <sque@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31722
Reviewed-by: Duncan Laurie <dlaurie@chromium.org>
Reviewed-by: Vincent Palatin <vpalatin@chromium.org>

CHROMIUM: Input: atmel_mxt_ts - disable input_dev when updating FW & Config

When the FW update is in progress, the linux input layer may
call open/close on the driver, which translates to writing
certain values into the device T9 control register, which does
not exist at all when the system is in bootloader mode and the
writing to register will make the firmware update corrupted.

So unregister the input device from input subsystem when doing
FW update and re-create it when update finishes. Also doing the
same for config update.

Signed-off-by: Yufeng Shen <miletus@chromium.org>
BUG=chromium-os:33981
TEST=Boot system and let the device auto-update the FW and make
     sure the update is successful.

Change-Id: If8f7ea2953f1a825cbcad4a548cf563345518c7d
Reviewed-on: https://gerrit.chromium.org/gerrit/31638
Reviewed-by: Daniel Kurtz <djkurtz@chromium.org>
Reviewed-by: Benson Leung <bleung@chromium.org>
Commit-Ready: Yufeng Shen <miletus@chromium.org>
Tested-by: Yufeng Shen <miletus@chromium.org>

arm: exynos: tweak busfreq sysfs support

Silence console msgs generated when reading the cur_freq data due
to null pointers.

Signed-off-by: Sam Leffler <sleffler@chromium.org>
BUG=none
TEST=cat busfreq/cur_freq and note no console messages

Change-Id: I14867d8d29aee92f3dd2a5daf0eef0bda2fffcf3
Reviewed-on: https://gerrit.chromium.org/gerrit/31618
Reviewed-by: Abhilash Kesavan <a.kesavan@samsung.com>
Reviewed-by: Bryan Freed <bfreed@chromium.org>
Commit-Ready: Sam Leffler <sleffler@chromium.org>
Tested-by: Sam Leffler <sleffler@chromium.org>

UPSTREAM: ARM: 7445/1: mm: update CONTEXTIDR register to contain PID of current process

This patch introduces a new Kconfig option which, when enabled, causes
the kernel to write the PID of the current task into the PROCID field
of the CONTEXTIDR on context switch. This is useful when analysing
hardware trace, since writes to this register can be configured to emit
an event into the trace stream.

The thread notifier for writing the PID is deliberately kept separate
from the ASID-writing code so that we can support newer processors using
LPAE, where the ASID is stored in TTBR0. As such, the switch_mm code is
updated to perform a read-modify-write sequence to ensure that we don't
clobber the PID on CPUs using the classic 2-level page tables.

Signed-off-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
(cherry picked from commit 575320d625d5b5eb115575a1f5e17af456e69577)

BUG=chromium-os:34007
TEST=general system testing

Change-Id: I4a4ba8b8911efac06914ed0c2974d16b502d78f2
Signed-off-by: Olof Johansson <olofj@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31715
Reviewed-by: Mandeep Singh Baines <msb@chromium.org>

UPSTREAM: ARM: Remove __ARCH_WANT_INTERRUPTS_ON_CTXSW on ASID-capable CPUs

Since the ASIDs must be unique to an mm across all the CPUs in a system,
the __new_context() function needs to broadcast a context reset event to
all the CPUs during ASID allocation if a roll-over occurred. Such IPIs
cannot be issued with interrupts disabled and ARM had to define
__ARCH_WANT_INTERRUPTS_ON_CTXSW.

This patch changes the check_context() function to
check_and_switch_context() called from switch_mm(). In case of
ASID-capable CPUs (ARMv6 onwards), if a new ASID is needed and the
interrupts are disabled, it defers the __new_context() and
cpu_switch_mm() calls to the post-lock switch hook where the interrupts
are enabled. Setting the reserved TTBR0 was also moved to
check_and_switch_context() from cpu_v7_switch_mm().

Reviewed-by: Will Deacon <will.deacon@arm.com>
Tested-by: Will Deacon <will.deacon@arm.com>
Reviewed-by: Frank Rowand <frank.rowand@am.sony.com>
Tested-by: Marc Zyngier <Marc.Zyngier@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
(cherry picked from commit 7fec1b57b8a925d83c194f995f83d9f8442fd48e)

BUG=chromium-os:34007
TEST=general system testing

Change-Id: I0628afd963eb7af1dfa0c3230f08f670feba9e9d
Signed-off-by: Olof Johansson <olofj@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31714
Reviewed-by: Mandeep Singh Baines <msb@chromium.org>

UPSTREAM: ARM: Use TTBR1 instead of reserved context ID

On ARMv7 CPUs that cache first level page table entries (like the
Cortex-A15), using a reserved ASID while changing the TTBR or flushing
the TLB is unsafe.

This is because the CPU may cache the first level entry as the result of
a speculative memory access while the reserved ASID is assigned. After
the process owning the page tables dies, the memory will be reallocated
and may be written with junk values which can be interpreted as global,
valid PTEs by the processor. This will result in the TLB being populated
with bogus global entries.

This patch avoids the use of a reserved context ID in the v7 switch_mm
and ASID rollover code by temporarily using the swapper_pg_dir pointed
at by TTBR1, which contains only global entries that are not tagged
with ASIDs.

Reviewed-by: Frank Rowand <frank.rowand@am.sony.com>
Tested-by: Marc Zyngier <Marc.Zyngier@arm.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
[catalin.marinas@arm.com: add LPAE support]
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
(cherry picked from commit 3c5f7e7b4a0346de670b08f595bd15e7eec91f97)

BUG=chromium-os:34007
TEST=general system testing

Change-Id: I853a937d0a045c0487e97d74783064329d42eaf0
Signed-off-by: Olof Johansson <olofj@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31713
Reviewed-by: Mandeep Singh Baines <msb@chromium.org>

Add implementation of poll() to dma_buf using KDS

BUG=chrome-os-partner:11949
TEST=local build, run on ARM

Change-Id: I955e12be4d52e771eccc267befff8b42f316c150
Signed-off-by: Ray Smith <Raymond.Smith@arm.com>
Reviewed-on: https://gerrit.chromium.org/gerrit/27056
Commit-Ready: John Sheu <sheu@chromium.org>
Reviewed-by: John Sheu <sheu@chromium.org>
Tested-by: John Sheu <sheu@chromium.org>

KDS: Fix incorrect timeout

kds_waitall uses wait_event_interruptible() and
wait_event_interruptible_timeout() incorrectly.

BUG=chrome-os-partner:11949
TEST=build, run on ARM

Change-Id: Ide44ccff6e902c6d8b5483d67335e36e094da5d2
Signed-off-by: David Garbett <David.Garbett@arm.com>
Reviewed-on: https://gerrit.chromium.org/gerrit/29878
Commit-Ready: John Sheu <sheu@chromium.org>
Reviewed-by: John Sheu <sheu@chromium.org>
Tested-by: John Sheu <sheu@chromium.org>

CHROMIUM: ALSA: hda/ca0132 - cache firmware on first load.

So we won't load it from at resume time.
Also remove the size check for the speakereq file.

BUG=chrome-os-partner:12927
TEST=build and install, audio still works.

Change-Id: Ic1eb133f5d5a33fbf0e823736a62ff391648d939
Signed-off-by: Chih-Chung Chang <chihchung@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31093
Reviewed-by: Dylan Reid <dgreid@chromium.org>

drm/exynos: Implement fimd_display_power_on

Also fix fimd_power_on to turn the panel off in the off case.

BUG=chromium-os-partner:12832
TEST=by hand, screen goes completely black after the timeout.

Change-Id: Ia79f5e53009110715395a5984e3e3f96809c7aed
Signed-off-by: Stéphane Marchesin <marcheu@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31634
Reviewed-by: John Sheu <sheu@chromium.org>

drm/exynos: Fix doxygen comment for the event.

BUG=none
TEST=compiles

Change-Id: I505539c5cb25ae38f672d21dac3bda837a6c66ec
Signed-off-by: Stéphane Marchesin <marcheu@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31390

leds: chromeos keyboard name should contain "kbd_backlight"

The driver sysfs directory should be "chromeos::kbd_backlight"
This is to match how other keyboard backlight drivers do it.

BUG=chromium-os:32959
TEST=See /sys/class/leds, there should be "chromeos::kbd_backlight"

Change-Id: Icaed85f110211a1e51f15ebcbb90b9e9b5e63291
Signed-off-by: Simon Que <sque@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31626
Reviewed-by: Duncan Laurie <dlaurie@chromium.org>

CHROMIUM: chromeos_laptop: Use LEDs for keyboard backlight

Previously using regular backlight driver, now using LEDs driver.

BUG=chromium-os:32959
TEST=Go to /sys/class/leds/keyboard_leds, make sure you can change
keyboard backlight by writing to "brightness"

Change-Id: I780629a014030639dae1497712b7a4d7ed599e26
Signed-off-by: Simon Que <sque@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31543
Reviewed-by: Duncan Laurie <dlaurie@chromium.org>

CHROMIUM: config: Enable keyboard backlight LEDs

BUG=chromium-os:32959
TEST=None (see later patch(es))

Change-Id: I487688054743eae0743201d923ba8f2c1ebc3ff5
Signed-off-by: Simon Que <sque@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31542
Reviewed-by: Duncan Laurie <dlaurie@chromium.org>

CHROMIUM: leds: Add Chrome OS keyboard LEDs driver

It is ACPI-based keyboard backlight LEDs.

BUG=chromium-os:32959
TEST=None (see next patch(es) for build/test info)

Change-Id: I0871851b1c67af33a9b9bd83c1dc93b1f8253ac5
Signed-off-by: Simon Que <sque@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31541
Reviewed-by: Duncan Laurie <dlaurie@chromium.org>

regulator: max77686: Add support for various operating modes

MAX77686 supports a low power mode as well as a standby mode for
its LDOs and BUCKs. Initially add LDO8 standby mode as it leaks
~160mW through a suspend/resume cycle. The idea would be to set
these modes for the rest of the LDOs and BUCKs as well.
Note: The device tree bindings have been taken from the upstream
MFD driver with the new optional bindings for operating mode and
buck ramp delay added.

BUG=chrome-os-partner:12522
TEST=Verify 1350_ldoin power during suspend along with follow-on
patches.

Change-Id: I1144254ebf6cf096bc5b11217c3045aff41d8db0
Signed-off-by: Abhilash Kesavan <a.kesavan@samsung.com>
Reviewed-on: https://gerrit.chromium.org/gerrit/30257
Reviewed-by: Todd Broch <tbroch@chromium.org>

CHROMIUM: dts: Add regulator operating mode bindings

Some of the LDOs and BUCKs on the MAX77686 PMIC can be put
into a low power or standby state. Add bindings to control
the operating mode (primarily enabling the standby mode for
LDO8).

BUG=chrome-os-partner:12660
TEST=Verify that the standby mode bit is being set using:
i2cget -y -f 0 0x9 0x47 (get a value of 0x48).
Suspend the system (rtcwake -m mem -s 25) and measure 1350_ldoin_mw
power. Prior to the patches 1350_ldoin_mw showed ~95mW and after it
showed ~2mW. Verified DP across 20 suspend/resume cycles.
HDMI S2R support is on-going; with or without these patches the
behaviour of HDMI is identical (corrupted images on cable insertion
post-resume).

Change-Id: I58f7975e9cb14a051139988b6237d3f5599589c2
Signed-off-by: Abhilash Kesavan <a.kesavan@samsung.com>
Reviewed-on: https://gerrit.chromium.org/gerrit/31596
Reviewed-by: Todd Broch <tbroch@chromium.org>

exynos: Store mwifiex_ds_misc_subsc_evt in mwifiex_private

Since mwifiex_ds_misc_subsc_evt is used in an asynchronous case, store
the structure in the long-lived mwifiex_private instead of on the
calling stack.
Ideally, we would use this storage for the synchronous case, as well,
and explicitly clear the buffer on the "end" of the operation such that
we could prove only one message was in flight and that we were never
corrupting another.  That would require a more invasive change to create
such an "end".
This fixes a problem where the response of the asynchronous operation
would corrupt a stack frame potentially in use by a different thread.

BUG=chrome-os-partner:12367
TEST=Walked back and forth between Wifi access points on the same
network while poster circle was actively animating on the screen (this
constant flushing of KDS resources creates an easy-to-recognize panic
when this corruption occurs), watching the wifi reset without panic.

Change-Id: I06b9c522d1f50e0de1c5754399d154cfd93830a4
Signed-off-by: Jeff Disher <disher@google.com>
Reviewed-on: https://gerrit.chromium.org/gerrit/31473
Reviewed-by: Sam Leffler <sleffler@chromium.org>
Commit-Ready: Jeff Disher <disher@chromium.org>
Tested-by: Jeff Disher <disher@chromium.org>

DRM: HDMI: Initialized AUI and AVI Header info for HDMI IP

Audio and Video Header information has to be initialized to HDMI IP,

Fix: For compliance test suite TestID 7-19
Test: AVI Info Frame missing
AVI InfoFrame is transmitted on every two video field.
Test: AUI InfoFrame missing
Audio InfoFrame is transmitted on every two video fields.

Added code for getting VIC.

BUG=chrome-os-partner:13173
TEST=Verified that all the compliance tests pass
for 480p,720p and 1080p in
Agilent Technologies,N5998A
protocol analyser.

Change-Id: I2b1aeedeb33818f839f395d5f9097f3d50f76cc6
Signed-off-by: Fahad Kunnathadi <fahad.k@samsung.com>
Signed-off-by: Shirish S <s.shirish@samsung.com>
Reviewed-on: https://gerrit.chromium.org/gerrit/31429
Reviewed-by: Sean Paul <seanpaul@chromium.org>
Tested-by: Shirish S <shirish@chromium.org>
Commit-Ready: Shirish S <shirish@chromium.org>

drm/exynos: enable dvi mode for dvi monitor

Hdmi monitor and dvi monitor can be distinguished with edid.
This patch enables dvi mode if dvi monitor is connected and does
not enable audio feature for dvi mode because dvi has no audio
feature.

This patch is inspired from  commit id:
872d20d66c0b7de0787675dce8569a61e4d9bc0 in drm-next branch.

BUG=chrome-os-partner:13173
TEST=Verified that all the dvi related compliance tests pass
Agilent Technologies,N5998A
protocol analyser.

Change-Id: I94c7ba951512bb9f1926880e9b8b41d8668495a3
Signed-off-by: Shirish S <s.shirish@samsung.com>
Reviewed-on: https://gerrit.chromium.org/gerrit/31428
Reviewed-by: Sean Paul <seanpaul@chromium.org>
Tested-by: Shirish S <shirish@chromium.org>
Commit-Ready: Shirish S <shirish@chromium.org>

CHROMIUM: Input: synaptics - add retry mechanism for reconnect

On the Samsung Series 5 Chromebook, the Synaptics ClickPad will sometimes not
send an "0xFA" PS/2 ACK in response to a command byte during the reconnect
sequence following a resume from suspend. From our experiments, the failure can
happen during any byte of any of the commands in the reconnect batch.

This failure results in a timeout in the psmouse driver. In addition, the
ClickPad will often also respond to the next PS/2 command byte with an "0xFE"
PS/2 RESEND response.

This patch makes the synaptics reconnect-at-resume more tolerant to such
failures.

Signed-off-by: Chung-yih Wang <cywang@chromium.org>
BUG=chromium-os:31522
TEST=on device; run platform_TouchpadResume test

Change-Id: Ic13d67fc8f3cad72fb5e6f6146f81060bfee5474
Reviewed-on: https://gerrit.chromium.org/gerrit/31099
Reviewed-by: Daniel Kurtz <djkurtz@chromium.org>
Commit-Ready: Chung-yih Wang <cywang@chromium.org>
Reviewed-by: Chung-yih Wang <cywang@chromium.org>
Tested-by: Chung-yih Wang <cywang@chromium.org>

CHROMIUM: config: exynos5: Enable busfreq support

Enable support for Samsung's busfreq driver which controls VDD_INT.

BUG=chrome-os-partner:11792
TEST=Measurements for variation in VDD_INT on Daisy show a reduction
of ~70-80mW at home screen.

Signed-off-by: Abhilash Kesavan <a.kesavan@samsung.com>
Change-Id: I875e5d2ea9c80e777a1ac1d2b57c88acb70cb59b
Reviewed-on: https://gerrit.chromium.org/gerrit/28944
Reviewed-by: Bryan Freed <bfreed@chromium.org>
Tested-by: Arjun.K.V <arjun.kv@samsung.com>

arm: exynos: Add busfreq support for smdk5250

The busfreq driver controls the MIF and INT frequencies and
voltages.
Note: MIF cannot be varied for DDR3 boards and hence has been
disabled.

BUG=chrome-os-partner:11792
TEST=Measurements for variation in VDD_INT on Daisy show a reduction
of ~70-80mW at home screen.

Change-Id: I02e6cae2c1aec2b88c17b5bcccd3ebfc9d439f98
Signed-off-by: Abhilash Kesavan <a.kesavan@samsung.com>
Reviewed-on: https://gerrit.chromium.org/gerrit/28943
Reviewed-by: Sam Leffler <sleffler@chromium.org>
Reviewed-by: Bryan Freed <bfreed@chromium.org>
Tested-by: Arjun.K.V <arjun.kv@samsung.com>

arm: exynos: Add ppmu support for Exynos5

Platform Performance Monitoring Unit (PPMU) is used to monitor and
gather performance information for custom IPs such such as MFC. This
information is used by the busfreq driver to vary/control VDD_INT.

BUG=chrome-os-partner:11792
TEST=Build and boot Snow

Change-Id: I0c94bfb23c2715afa0d26ca8595b59c7780f3a45
Signed-off-by: Abhilash Kesavan <a.kesavan@samsung.com>
Reviewed-on: https://gerrit.chromium.org/gerrit/28942
Reviewed-by: Bryan Freed <bfreed@chromium.org>
Tested-by: Arjun.K.V <arjun.kv@samsung.com>

cpufreq: exynos: Updated the DVFS table with ASV for EXYNOS5250

Choose from the newly added DVFS/ASV table for EXYNOS5250 based on
4/30 released guide. Different lot samples are supported.

BUG=chrome-os-partner:11794
TEST=Check that the ASV group is chosen based on lot

Change-Id: I392eae626cbee0f6ed7962cfecea5614296fd92b
Signed-off-by: Abhilash Kesavan <a.kesavan@samsung.com>
Reviewed-on: https://gerrit.chromium.org/gerrit/28393
Reviewed-by: Prashanth Godrehal <prashanth.g@samsung.com>
Reviewed-by: Bryan Freed <bfreed@chromium.org>
Tested-by: Arjun.K.V <arjun.kv@samsung.com>
Commit-Ready: Arjun.K.V <arjun.kv@samsung.com>

arm: exynos: Add Adaptive Supply Voltage for Exynos5250

Add Adaptive Supply Voltage support for Exynos5250. ASV lists out
various voltage settings per frequency appropriate for the silicon.
This table will be used by cpufreq and busfreq drivers to set the
core and INT voltages.

BUG=chrome-os-partner:11794
TEST=Build and boot to Login prompt 10 times on Snow and vary the
cpu frequency via sysfs. This is not good enough but I don't have
a cpufreq stress test.

Change-Id: I97cc1b5877aa56d237fa6b458693eccd98046122
Signed-off-by: Abhilash Kesavan <a.kesavan@samsung.com>
Reviewed-on: https://gerrit.chromium.org/gerrit/28392
Reviewed-by: Prashanth Godrehal <prashanth.g@samsung.com>
Reviewed-by: Bryan Freed <bfreed@chromium.org>
Tested-by: Arjun.K.V <arjun.kv@samsung.com>
Commit-Ready: Arjun.K.V <arjun.kv@samsung.com>

UPSTREAM: drm/exynos: fixed wrong pageflip finish event for interlace mode

Pageflip finish event for interlace mode has bug on checking top
field vsync because of comparing between dma address converted
by start coordinates and non-converted dma address.

Add layer update change specifice to exynos5.

BUG=chrome-os-partner:13173
TEST=Verified the same patch with HDMI analser,
     for chrome os version 22.0.1229

Signed-off-by: Seung-Woo Kim <sw0312.kim@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
(cherry picked from commit 8379e4823d04e2552e1395c2410fc16733e28c6e)

Change-Id: I75b9b3fbef7bd1acd809eb97709b7b503938f414
Signed-off-by: Fahad Kunnathadi <fahad.k@samsung.com>
Reviewed-on: https://gerrit.chromium.org/gerrit/31430
Reviewed-by: Sean Paul <seanpaul@chromium.org>
Tested-by: Shirish S <shirish@chromium.org>
Commit-Ready: Shirish S <shirish@chromium.org>

drm/exynos: hdmi: Correct the timing values for HDMI

The current method of calculation uses DRM_MODE_FLAG_PVSYNC
and DRM_MODE_FLAG_PHSYNC, which makes it non-usable for "Nx480/576"
etc resolutions whose flags use a higher value for flags variable.

Hence now,DRM_MODE_FLAG_NVSYNC and DRM_MODE_FLAG_NHSYNC are used for
calculating HSYNC and VSYNC Polarity.

Also as per the Exynos 5 IP design, the hsync start and hsync end
has a offset of 2 pixels.This patch corrects this.

The above modifications are required for passing the HDMI compliance
tests.

BUG=chrome-os-partner:13173
TEST=Verified that all the compliance tests pass
for 480p,720p and 1080p in
Agilent Technologies,N5998A protocol analyser.

Change-Id: I25d1bb2a77dcd109ec0669555b3d87f1a277299a
Signed-off-by: Shirish S <s.shirish@samsung.com>
Reviewed-on: https://gerrit.chromium.org/gerrit/31447
Reviewed-by: Sean Paul <seanpaul@chromium.org>
Tested-by: Shirish S <shirish@chromium.org>
Commit-Ready: Shirish S <shirish@chromium.org>

drm/exynos: mixer: Set mixer pipe

Set the mixer pipe correctly so we don't call finish pageflip on the
wrong crtc.

BUG=chrome-os-partner:12889
TEST=Tested on snow with hdmi

Change-Id: I78144c690be3872d9fe06923e91e1f427269f9c7
Signed-off-by: Sean Paul <seanpaul@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31555
Reviewed-by: Stéphane Marchesin <marcheu@chromium.org>

gpu: vithar: Change the GPU DVFS algorithm

Changed the GPU DVFS algorithm as follows:
1) Added a new OPP (350 MHz) to mali_dvfs_infotbl.
2) Removed the configurable option for setting the number of levels.
Now, there are 7 levels by default.
3) Modified the scaling logic. Now voltage is upscaled after utilization
exceed the max threshold certain times. The max and min threshold values
are also changed.
4) Modified the GPU DVFS table with updated voltage values.
5) GPU start with 266 Mhz when loaded after being idle for sometime.

BUG=chrome-os-partner:11795
TEST=Less power consumed whle running webGL apps.
For 'Aquarium' the power consumed by vithar is reduce to 739.63mW from
999.37mW.

Change-Id: I3ff5092c30e45c950d69323c16ab7241b0bceaef
Signed-off-by: Shariq Hasnain <shariq.786@samsung.com>
Reviewed-on: https://gerrit.chromium.org/gerrit/29582
Reviewed-by: Sam Leffler <sleffler@chromium.org>
Reviewed-by: Abhilash Kesavan <a.kesavan@samsung.com>
Tested-by: Arjun.K.V <arjun.kv@samsung.com>
Commit-Ready: Abhilash Kesavan <a.kesavan@samsung.com>

gpu: vithar: Enable GPU frequency/voltage scaling

Enable support for GPU DVFS; it supports the following frequencies:
533, 450, 400, 266, 160, 100 MHz.

BUG=chrome-os-partner:11795
TEST=Build and boot Snow. Check voltage variation when the
GPU DVFS is functioning.
Build with MALI_DVFS_DEBUG define enabled, boot and check
GPU DVFS debug prints

Change-Id: Ifd4ea158e12ff15285529808e590196ec88fe776
Signed-off-by: Abhilash Kesavan <a.kesavan@samsung.com>
Signed-off-by: Doug Anderson <dianders@chromium.org>
[dianders removed unneeded CSTD_UNUSED]
Reviewed-on: https://gerrit.chromium.org/gerrit/27976
Reviewed-by: Prashanth Godrehal <prashanth.g@samsung.com>
Tested-by: Arjun.K.V <arjun.kv@samsung.com>
Reviewed-by: Sam Leffler <sleffler@chromium.org>

CHROMIUM: ALSA: hda - Cancel power work before suspending.

If snd_hda_call_codec_suspend is called from the system suspend path
wait for powersave work to finish.  Without the call to _sync, the
power_work function can hang the workqueue if it has progressed past the
check of the power_on variable before this function sets it to 0.  In
that case the power_work function will call snd_hda_power_up (via it's
own call of hda_call_codec_suspend).  snd_hda_power_up will then call
cancel_delayed_work_sync for the work it is currently executing on.

BUG=chrome-os-partner:13008
TEST=suspend_test.sh from bug.

Change-Id: Ic127eeecdfb8995be5e2c630028f4e3cf5477a2e
Signed-off-by: Dylan Reid <dgreid@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31285
Reviewed-by: Sonny Rao <sonnyrao@chromium.org>
Commit-Ready: Sonny Rao <sonnyrao@chromium.org>
Tested-by: Sonny Rao <sonnyrao@chromium.org>

drm/i915: Send the AVI Infoframe mode code.

This is needed for HDMI conformance.

BUG=none
TEST=compiles

Change-Id: I483c81c76ddb39b47c84df3a0a2614a4aca12b3d
Signed-off-by: Stéphane Marchesin <marcheu@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/30286
Reviewed-by: Sean Paul <seanpaul@chromium.org>

drm/exynos: Only map GEM buffers into the IOMMU when needed.

We used to map all GEM buffers into the IOMMU, which would result in
eventually running out of space. Instead, map/unmap them as they get
used as a FB.

BUG=none
TEST=by hand, compiled and ran daisy, tried webGL and video mode changes.

Change-Id: Ic04b9618f51bf5e0e35de4c5498127e2cbdb4f97
Signed-off-by: Stéphane Marchesin <marcheu@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31391
Reviewed-by: Sean Paul <seanpaul@chromium.org>
Reviewed-by: Pawel Osciak <posciak@google.com>

Asoc: Samsung: Add I2S S/W RST in startup function

I2S module need to be reset after S2R. Keeping the S/W rst
control part in resume didn't help in playing audio after resume.
So this patch adds S/W RST control part in startup function which
gets triggered for every new audio stream playback.

BUG=chrome-os-partner:12579
TEST=tested by playing audio before suspend and after resume
using aplay command.

Change-Id: I27aa97da2c7c6cb42e24d19bf41f294732cfbd6e
Signed-off-by: Padmavathi Venna <padma.v@samsung.com>
Reviewed-on: https://gerrit.chromium.org/gerrit/31335
Reviewed-by: Dylan Reid <dgreid@chromium.org>
Reviewed-by: Alim Akhtar <alim.akhtar@samsung.com>
Commit-Ready: Alim Akhtar <alim.akhtar@samsung.com>

Asoc: Samsung: Add support for I2S runtime S2R

This patch adds runtime suspend to resume support for I2S

BUG=chrome-os-partner:12579
TEST=Tested by playing audio before suspend and after resume
using aplay command. I2S get suspended after probe and resume
upon audio playback.

Change-Id: Icd507188bbe2bb39e3bb9240ef6793b4ba937901
Signed-off-by: Padmavathi Venna <padma.v@samsung.com>
Reviewed-on: https://gerrit.chromium.org/gerrit/29991
Reviewed-by: Jon Kliegman <kliegs@chromium.org>
Reviewed-by: Dylan Reid <dgreid@chromium.org>
Commit-Ready: Alim Akhtar <alim.akhtar@samsung.com>
Reviewed-by: Alim Akhtar <alim.akhtar@samsung.com>

CHROMIUM: scsi: make sure we avoid resume race

Added the same check in suspend to shutdown to prevent the race where we are
not resuming the disk yet, but we start to shutdown.

BUG=chrome-os-partner:11780
TEST=power_Resume, powerd_suspend

Change-Id: Ifb0b07854857e9baaa7647b8e655b553842f3326
Signed-off-by: Derek Basehore <dbasehore@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31427
Reviewed-by: Sonny Rao <sonnyrao@chromium.org>

CHROMIUM: ata: make sure we avoid resume race

Added code to make sure shutdown, freeze, etc. do not race with resume. This is
done by checking if there is work pending on the resume_work work_struct and
waiting for it to complete if there is.

BUG=chrome-os-partner:11780
TEST=power_Resume, powerd_suspend

Change-Id: I5d010ee341116b0498440fb3093085ab9163eb24
Signed-off-by: Derek Basehore <dbasehore@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31426
Reviewed-by: Sonny Rao <sonnyrao@chromium.org>

drm/exynos: fix double invocation of drm_prime_(init/delete)_file_priv

The double invocations are incorrect but seem to be safe so I don't
think this will fix any bugs.

BUG=none
TEST=Added printks to the init and delete methods.

Before:

[    7.639366] Mandeep: drm_prime_init_file ee3675d0
[    7.639377] Mandeep: drm_prime_init_file ee3675d0
[    7.639507] Mandeep: drm_prime_destroy_file ee3675d0
[    7.639518] Mandeep: drm_prime_destroy_file ee3675d0
[    7.639802] Mandeep: drm_prime_init_file ee372390
[    7.639810] Mandeep: drm_prime_init_file ee372390
[    8.473316] Mandeep: drm_prime_init_file ee356390
[    8.473331] Mandeep: drm_prime_init_file ee356390

After:

[    6.363842] Mandeep: drm_prime_init_file edc2e5d0
[    6.363994] Mandeep: drm_prime_destroy_file edc2e5d0
[    6.364260] Mandeep: drm_prime_init_file edc2e750
[    8.004837] Mandeep: drm_prime_init_file ee36ded0

Change-Id: I3350913fe158557d6ddf67c786a3abb578f6ccf4
Signed-off-by: Mandeep Singh Baines <msb@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31381
Reviewed-by: Stéphane Marchesin <marcheu@chromium.org>
Reviewed-by: Pawel Osciak <posciak@google.com>

drm/exynos: Cleanup warning

BUG=none
TEST=compiles

Change-Id: Ie0ec41c7d3721f85e1dce4f07241ecfbd80cbed4
Signed-off-by: Stéphane Marchesin <marcheu@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31388
Reviewed-by: Mandeep Singh Baines <msb@chromium.org>

Revert "video/exynos: dp: Add reporter to log dp errors"

We no longer need this hack, DP is not to blame for the flicker. This reverts commit 4d23813361c0a31fe314d6ffa523fa6dbe6942b9

Change-Id: I2675d8c39cfc809ac891c5458465217162757432
Reviewed-on: https://gerrit.chromium.org/gerrit/31399
Reviewed-by: Doug Anderson <dianders@chromium.org>
Commit-Ready: Sean Paul <seanpaul@chromium.org>
Tested-by: Sean Paul <seanpaul@chromium.org>

CHROMIUM: config: turn off DYNAMIC_DEBUG on ARM

This was on during development to catch some of the wifi problems. It
should be disabled now for release.

BUG=chromium-os:32752
TEST=usual wifi tests

Change-Id: Ib006f06f7c29ba617326c62d87d5974609c6689b
Signed-off-by: Olof Johansson <olofj@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31392
Reviewed-by: Gary Morain <gmorain@chromium.org>

drm/exynos: Factor all the finish_pageflips implementations.

Factor all finish_pageflips implementations in exynos_drm_crtc.c so we can
modify it without changing three places.

BUG=none
TEST=by hand, ran the device, browsed, ran webgl demos.

Change-Id: I2abafbe352bf684f31d1fc737e045832b521aebb
Reviewed-on: https://gerrit.chromium.org/gerrit/31300
Commit-Ready: Stéphane Marchesin <marcheu@chromium.org>
Reviewed-by: Stéphane Marchesin <marcheu@chromium.org>
Tested-by: Stéphane Marchesin <marcheu@chromium.org>
Reviewed-by: Mandeep Singh Baines <msb@chromium.org>

drivers: mfd: tps65090: disable wake capabilities for tps65090

Disable the ability for the device to wake the system from the driver.
This can be turned back on via the wakeup-source devicetree setting
to configure the i2c interface to turn on wakeup for this device.

BUG=chrome-os-partner:13020
TEST=suspended and saw no immediate wake
  ran 'udevadm monitor' and confirmed battery events still present
  Add wakeup-sources; to dts to test that wakeup works if configured
    Saw device wake system up.
    Wrote disabled to
     /sys/class/i2c-dev/i2c-4/device/4-0048/power/wakeup
    Saw device no longer woke system

Change-Id: Ia88c326cf568239edf18cb5524353d89f6e494c0
Signed-off-by: Jonathan Kliegman <kliegs@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31247
Reviewed-by: Doug Anderson <dianders@chromium.org>

arm: exynos: Clear ENABLE_WAKEUP_SW bit when entering suspend

Setting this bit to 0 causes the system to wait until suspended
to use the wakeup masks.  With it being set high previously,
masked interrupts were being received and processed before the
EINT_WAKEUP_MASK was configured.

BUG=chrome-os-partner:13020
TEST=On a system that generated spurious interrupts on EINT[22]:
  Suspend, notice resume waits for input or rtc
  Resume via: trackpad, keyboard, power button, lid, rtc0 and rtc1
Signed-off-by: Jonathan Kliegman <kliegs@chromium.org>
Change-Id: I9fc5148ed95a81bca1ac4ba1526eb6f93585ce36
Reviewed-on: https://gerrit.chromium.org/gerrit/31341
Tested-by: Jon Kliegman <kliegs@chromium.org>
Reviewed-by: Doug Anderson <dianders@chromium.org>
Commit-Ready: Jon Kliegman <kliegs@chromium.org>

drm: exynos: Enable MIE block to handle 6bit dithering

This patch enables MIE block on Gaia and programs MIE
to handle 8bit to 6 bit Dithering. Previously, FIMD output
was bypassing MIE since MIE was not being used.
Note that, even though MIE is regarded as different IP
and has its own SFR space, it is programed inside FIMD driver.

BUG=chrome-os-partner:11072
TEST=Checked the browser borders on Snow. Banding is not visible.

Change-Id: If4da91ba3af039232ba9582a079346099ac31a4d
Signed-off-by: Ajay Kumar <ajaykumar.rs@samsung.com>
Reviewed-on: https://gerrit.chromium.org/gerrit/31114
Reviewed-by: Shirish S <shirish@chromium.org>
Reviewed-by: Stéphane Marchesin <marcheu@chromium.org>
Commit-Ready: Olof Johansson <olofj@chromium.org>
Tested-by: Olof Johansson <olofj@chromium.org>

arm: exynos: Remove hardcode wakeup unmask for EINT_0

For legacy reasons EINT_0 was being forced on for all
exynos systems as a wake interrupt.  For boards that need
EINT_0 they should probably enable it with enable_irq_wake

BUG=None
TEST=Observed normal suspend/resume cycle.  Confirmed
  trackpad, rtc, keyboard, lid and power button wakes.
Signed-off-by: Jonathan Kliegman <kliegs@chromium.org>
Change-Id: Icb6687a7f6b01cb6207a799948d957781fd4a7da
Reviewed-on: https://gerrit.chromium.org/gerrit/31337
Tested-by: Jon Kliegman <kliegs@chromium.org>
Commit-Ready: Thomas Abraham <thomas.ab@samsung.com>
Reviewed-by: Thomas Abraham <thomas.ab@samsung.com>
Reviewed-by: Doug Anderson <dianders@chromium.org>
Commit-Ready: Jon Kliegman <kliegs@chromium.org>

drm/exynos: remove unneeded drm_vblank_off

The drm_vblank_off calls in finish_page_flip are not needed. drm_vblank_put
will schedule an off 5 s in the future.

BUG=none
TEST=Suspend/resume. Multipe VT Switches. Multiple LCD off/on. Some browsing.

Change-Id: I08ac4a154995fcccc817c60c51a8a486972be1c5
Signed-off-by: Mandeep Singh Baines <msb@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31281

Add two new quirks:
SDHCI_QUIRK2_BROKEN_UHS:
    Disables all UHS modes.

SDHCI_QUIRK2_BROADCOM_REGISTERS:
    Bit twiddles some Broadcom-specific registers and supresses an error
    about the 64k bar0.

Add PCI fixes structs:
One for the chip itself with the required quirks in general (lightly tested)
and one that is Parrot specific which adds the new SDHCI_QUIRK2_BROKEN_UHS
quirk, disables ADMA mode, and adds a delay after power.

BUG=chrome-os-partner:12366
TEST=Excercise IO on SD/SDHC/MMC cards on a Parrot device.

Signed-off-by: Stephen Hurd <shurd@broadcom.com>
Change-Id: I278a19084b2dd9ddf3aedf2aaf14a9962080610d
Reviewed-on: https://gerrit.chromium.org/gerrit/31240
Commit-Ready: Stephen Hurd <shurd@broadcom.com>
Reviewed-by: Stephen Hurd <shurd@broadcom.com>
Tested-by: Stephen Hurd <shurd@broadcom.com>
Commit-Ready: Grant Grundler <grundler@chromium.org>
Reviewed-by: Grant Grundler <grundler@chromium.org>
Reviewed-by: Dave Parker <dparker@chromium.org>

CHROMIUM: Input: atmel_mxt_ts - save and restore t9_ctrl on wakeup disabled

Signed-off-by: Benson Leung <bleung@chromium.org>
BUG=chrome-os-partner:12318
TEST=Close and open the lid to suspend and resume the system.
Make sure that the touchscreen does not generate a lot of
stray events on resume.

Change-Id: I57b34aab52149cc78cd55a79640b496239bbfa5e
Reviewed-on: https://gerrit.chromium.org/gerrit/31270
Reviewed-by: Yufeng Shen <miletus@chromium.org>
Commit-Ready: Benson Leung <bleung@chromium.org>
Tested-by: Benson Leung <bleung@chromium.org>

CHROMIUM: Input: atmel_mxt_ts - Suppress handle messages used for resume

The position information at resume time while the system is in a
low power state may be garbage, so do not actually report this
information up to the input layer. Just perform the reads to clear
out the status on the device side.

BUG=chrome-os-partner:12318
TEST=Suspend/resume using lid close. Check that on resume,
the touch device doesn't show a bunch of garbage data
from when the lid was just being opened.

Change-Id: I4e60ab672d6191b9141e08e9b0bfdbf42649a95e
Signed-off-by: Benson Leung <bleung@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31194
Reviewed-by: Yufeng Shen <miletus@chromium.org>

CHROMIUM: Input: atmel_mxt_ts - disable reporting on stop

Leave the touch device scanning and enabled, but disable reporting
on mxt_stop. This will prevent a needless recalibration due to the T9_ctrl
register being set to 0 (disabling the object) and then set back to 0x83.

BUG=chrome-os-partner:12042,chrome-os-partner:9717
TEST=Log in to guest mode. Test the touch screen.
Log out. Verify that in the next session (after X has restarted)
touch performance is still good.

Change-Id: If5556325d69390a329d1ed406b5bba9f0d494896
Signed-off-by: Benson Leung <bleung@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/29951

CHROMIUM: Input: atmel_mxt_ts - Add defines for T9 Touch Control

Fix the use of magic numbers (such as 0x83) to write to
T9 Touch Control.

Signed-off-by: Benson Leung <bleung@chromium.org>
BUG=chromium-os:33368
TEST=Builds clean.

Change-Id: I28d650becaa103490541cd88b600fdb535b4196c
Reviewed-on: https://gerrit.chromium.org/gerrit/29558
Commit-Ready: Benson Leung <bleung@chromium.org>
Tested-by: Benson Leung <bleung@chromium.org>
Reviewed-by: Benson Leung <bleung@chromium.org>

CHROMIUM: i2c-s3c2410: do not special case HDMIPHY stuck bus detection

Commit "i2c-s3c2410: Add HDMIPHY quirk for S3C2440" added support for
HDMIPHY with some special handling in s3c24xx_i2c_set_master:

"due to unknown reason (probably HW bug in HDMIPHY and/or the controller)
a transfer fails to finish. The controller hangs after sending the last
byte, the workaround for this bug is resetting the controller after each
transfer"

The "unknown reason" was that the proper sequence for generating a STOP
condition wasn't being followed as per the datasheet. Since this is now
fixed, remove the special handling.

Signed-off-by: Daniel Kurtz <djkurtz@chromium.org>
BUG=chrome-os-partner:10089
TEST=No "s3c-i2c 12ce0000.i2c" error/warning messages in dmesg
TEST=HDMI output works
  1) EDID displayed in /var/log/Xorg.0.log
  2) Video shown on external monitor

Change-Id: I463f29c401d9cea37d081a0586c7ec2e71605631
Reviewed-on: https://gerrit.chromium.org/gerrit/29879
Commit-Ready: Daniel Kurtz <djkurtz@chromium.org>
Reviewed-by: Daniel Kurtz <djkurtz@chromium.org>
Tested-by: Daniel Kurtz <djkurtz@chromium.org>

drm/i915: Only apply the adaptive backlight modulation when it is enabled.

Otherwise things go very, very dark.

BUG=chromium-os:33854
TEST=by hand

Change-Id: I3122447a255652ba5c98863664c0f60e1859c490
Reviewed-on: https://gerrit.chromium.org/gerrit/31293
Tested-by: Stéphane Marchesin <marcheu@chromium.org>
Reviewed-by: Sean Paul <seanpaul@chromium.org>
Commit-Ready: Stéphane Marchesin <marcheu@chromium.org>

drm/exynos: track vblank events on a per crtc basis

The goal of the change is to make sure we send the vblank event on the
current vblank. My hope is to fix any races that might be causing flicker.
After this change I only see a flicker in the transition plymouth and
X11.

Simplified the code by tracking vblank events on a per-crtc basis. This
allowed me to remove all error paths from the callback. It also allowed
me to remove the vblank wait from the callback.

BUG=chrome-os-partner:12586
TEST=Suspend/resume. Multipe VT Switches. Multiple LCD off/on. Some browsing.
TEST=Power down and restart multiple times.

Change-Id: Ia71863161c6cb7f2dca20d14a68acd5e19b0db53
Signed-off-by: Mandeep Singh Baines <msb@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/30781
Reviewed-by: Stéphane Marchesin <marcheu@chromium.org>

CHROMIUM: drm/i915: Adaptive backlight support

This patch implements i915 adaptive backlight support for SNB and IVB.
The intended use for the adaptive backlight is to generate interrupts
whenever the luminance of the screen changes by some thresholds. The
main caveat with that implementation is that those additional
interrupts will wake up the CPU and consume more power. Instead, we
hook into the vblank handler and handle it from there. This makes the
implementation a little less intuitive but a lot more efficient.
We also need to compute the gamma correction from the interrupt
handler so we do this with a (new) fixed point module.

BUG=chrome-os-partner:11002
TEST=set the adaptive backlight property with:
TEST=xrandr --output eDP1 --set "Adaptive backlight" on
TEST=then watch a full screen movie and see the backlight level change
TEST=to read the backlight percentage:
TEST=echo $((`printf "%d\n" \`intel_reg_read 0x48254 | cut -f 2 -d :\``*100/4882))

Change-Id: I9b9631cacc7d90e2801a542a3789118521bc25f0
Signed-off-by: Stéphane Marchesin <marcheu@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/26748

CHROMIUM: config: remove unused mousedev and polldev

All modern input drivers use evdev, not legacy mousedev, and none of the
input drivers used in chromium OS use polldev.

Signed-off-by: Daniel Kurtz <djkurtz@chromium.org>
BUG=none
TEST=builds clean; notice 29 less functions in
     /sys/kernel/debug/tracing/available_filter_functions

Change-Id: I0fb2a2580faadb88710c3af11b1459dc5ab91363
Reviewed-on: https://gerrit.chromium.org/gerrit/30832
Commit-Ready: Daniel Kurtz <djkurtz@chromium.org>
Reviewed-by: Daniel Kurtz <djkurtz@chromium.org>
Tested-by: Daniel Kurtz <djkurtz@chromium.org>

ARM: exynos: Recover WiFi functionality on daisy.

A previous WiFi revamp for snow broke daisy.
This change gets it working again.

BUG=chrome-os-partner:11366
TEST=manual:boot daisy and verify wifi works.

Change-Id: I1dde8e61a2f31a7bb6e1e9582c2d5ac36bd56337
Signed-off-by: Bryan Freed <bfreed@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31147
Reviewed-by: Sam Leffler <sleffler@chromium.org>
Reviewed-by: Doug Anderson <dianders@chromium.org>

regulator: tps65090: Poll timeout bit on enable

Instead of just reading power good bit once, poll the register up to 5
times. This fixes an issue where the regulator is not setup and we read
the status too early. We will now wait until either the TOFET bit is
de-asserted, or ~5 ms elapse before checking for power good.

BUG=chrome-os-partner:12762
TEST=Tested on snow, no regressions found

Change-Id: Ied6dc840dcc4a94970cecac659efbffcbb55d651
Signed-off-by: Sean Paul <seanpaul@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31072
Reviewed-by: Doug Anderson <dianders@chromium.org>
Reviewed-by: Olof Johansson <olofj@chromium.org>

CHROMIUM: ath9k_hw: enable MAC hang detection for AR9300 chips

Enable MAC/Baseband hang detection for all AR9300 chips.
This routine helps to recover the chip from know stuck
states.

Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com>
Signed-off-by: Paul Stewart <pstew@chromium.org>
BUG=chrome-os-partner:12518
TEST=Recompile

Change-Id: I6dc4fb1f5d4ed17b884f556bd5d67fe1d892d7df
Reviewed-on: https://gerrit.chromium.org/gerrit/31089
Reviewed-by: Gary Morain <gmorain@chromium.org>
Commit-Ready: Paul Stewart <pstew@chromium.org>
Tested-by: Paul Stewart <pstew@chromium.org>

drm/exynos: add a flip_pending flag to track page flip complete

exynos does not have flip completed interrupt so we'll use the vblank
interrupt to detect when a flip is completed.

BUG=chrome-os-partner:12586
TEST=Suspend/resume. VT Switch. LCD off. Some browsing.

Change-Id: I3574c54dda47cc00a378b8e18f2d081d9ccc5fc0
Signed-off-by: Mandeep Singh Baines <msb@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/30780
Reviewed-by: Stéphane Marchesin <marcheu@chromium.org>
Reviewed-by: Pawel Osciak <posciak@google.com>

CHROMIUM: config: turn on HID_CHERRY=m

Turn on support for the Cherry cymotion keyboards, since some
people might want to use them with Chromeboxes.

BUG=chromium-os:31607
TEST=plug in a cherry cymotion linux keyboard. Sorry, don't have one (external bugreport)

Change-Id: I08839e5ef08d21d7540a609cbb7e2332b069a357
Signed-off-by: Olof Johansson <olofj@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31119
Reviewed-by: Benson Leung <bleung@chromium.org>

Revert "CHROMIUM: drivers: device tree support for gpio-charger"

This reverts commit 2e42da7e808a2c4f6713fb2618fca0529b4c053c.

BUG=chrome-os-partner:11739
TEST=build kernel succesfully

Signed-off-by: Simon Que <sque@chromium.org>
Change-Id: I35cb59b03e2f6bb1a87fa99f0fceb27d5e4137ad
Reviewed-on: https://gerrit.chromium.org/gerrit/31048
Reviewed-by: Olof Johansson <olofj@chromium.org>

CHROMIUM: input: cyapa: recover from watchdog triggered reset

Newer versions of Cypress Trackpad firmware use a watchdog timer
to reset the device after an internal failure. The reset operation
will put the device into bootloader mode, so the kernel driver
must use cyapa_detect to bring the trackpad into operational mode
on a failed i2c transaction.

BUG=chrome-os-partner:12838
TEST=test with trackpad v1.2 firmware on CYTRA_103002-00
Trigger a watchdog reset using an electrostatic discharge.
Ensure that the cyapa_detect case is triggered, and the pad
is functional after the event.

Change-Id: Ib2270eb48d80643756fdc06e944a1c9afac4e64f
Signed-off-by: Du, Dudley <dudl@cypress.com>
Reviewed-on: https://gerrit.chromium.org/gerrit/27496
Reviewed-by: Benson Leung <bleung@chromium.org>
Commit-Ready: Benson Leung <bleung@chromium.org>

Revert "CHROMIUM: HACK: mmc: core: Do not rescan after resume"

This reverts commit adabba54966cbfd3b0dc9b006bb30f9fa0a1a48e.
This is not needed as we have a fix for this hack.

BUG=chrome-os-partner:12802
TEST= after reverting this CL and with other 2 CL, sd card detection
works fine after s2r cycle.

Change-Id: Idccaf531bc272533e3918b703a55f56f15a8cb48
Signed-off-by: Alim Akhtar <alim.akhtar@samsung.com>
Reviewed-on: https://gerrit.chromium.org/gerrit/30692
Reviewed-by: Doug Anderson <dianders@chromium.org>
Reviewed-by: Olof Johansson <olofj@chromium.org>
Reviewed-by: Jon Kliegman <kliegs@chromium.org>

CHROMIUM: exynos: snow: Disable mmc on the external SD card slot

MMC is not supported on this slot on this system.

BUG=chrome-os-partner:11811

TEST=With patch series, plug in an MMC card and saw that it was
detected before this patch but isn't detected after.

Change-Id: Ie9d1cd13628d27e60b47572837c162b00563bf72
Signed-off-by: Doug Anderson <dianders@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31023
Reviewed-by: Olof Johansson <olofj@chromium.org>

CHROMIUM: mwifiex: Disable WOW by default.

Disable WOW for two reasons:
1. Appears to have some hand in WIFI device waking during Standby and
entering high energy state.
2. Is not required feature and often burns additional power.

BUG=chrome-os-partner:12164
TEST=manual,  associate wifi and measure power in suspend to be
reasonable.

Change-Id: I039650eb8c986a447022fd87bf87880cc7b47635
Signed-off-by: Todd Broch <tbroch@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31020
Reviewed-by: Gary Morain <gmorain@chromium.org>
Reviewed-by: Sam Leffler <sleffler@chromium.org>
Reviewed-by: Bing Zhao <bzhao@marvell.com>

CHROMIUM: scsi: don't wait for disk to wakeup on resume

This is done in the same way as the ata ports. Instead of waiting for the disk
to resume in sd_resume, we put a function into the workqueue and return. We also
make sure that a race does not happen between resume and suspend. This is done
by making the suspend code wait until the work_struct in the ata_port finishes.

There does not seem to be any differences between what this change does and what
happened before when it breaks. It seems that the disk will be woken up in
another place if the function to wake it up in resume fails.

Note: I have not decided exactly what to do for error paths and therefore there
are currently none when ata ports or disks fail to resume

BUG=chrome-os-partner:11780
TEST=power_Resume, powerd_suspend, opening and closing laptop, commenting out
the line that wakes up the disk and seeing the system behaviour (same before and
after change)

Change-Id: I05ba125ef55c2cc9e6849d9fec693c574f81221a
Signed-off-by: Derek Basehore <dbasehore@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/29673

CHROMIUM: ata: don't wait on ata ports for resume

This change causes dpm_resume to not block on resuming ata ports. This is done
by putting the function calls to wake up the ata ports in a workqueue. I make
sure that there is not a race between suspend and resume by making the suspend
code for ata_port wait until the work for resume is done.

I tested what would happen when the ata ports do not resume (by commenting out
the line that resumes and returning -1). We do wake up, but after ~2 minutes,
the machine reboots. The machine blocks on any io that requires disk. Mouse
clicks do not seem to register.

Without the change, pretty much the same thing occurs (from a user experience),
except that the mouse arrow is not visible.

BUG=chrome-os-partner:11780
TEST=power_Resume, powerd_suspend

Change-Id: Ica746d313349a3d8c5f6ba6912f6a3b18e5c2b3f
Signed-off-by: Derek Basehore <dbasehore@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/29672

mmc: dw_mmc: Add a DISABLE_MMC quirk that sets the core mmc cap

This quirk can be set from the device tree to disable mmc on a given
slot.

BUG=chrome-os-partner:11811
TEST=With patch series, plug in an MMC card and saw that it was
detected before this patch but isn't detected after.

Change-Id: I169b070578cff16b74f1556934e09ea3ea542b84
Signed-off-by: Doug Anderson <dianders@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/31022

mmc: core: Add a capability for disabling mmc cards

On some systems we need a way to disable MMC card support in a MMC/SD
card slot.  Add support in the core SD/MMC code to support this.

BUG=chrome-os-partner:11811
TEST=With patch series, plug in an MMC card and saw that it was
detected before this patch but isn't detected after.

Change-Id: I5b94a1d7deba85e0eb4709a419589cc3e9058eca
Signed-off-by: Doug Anderson <dianders@chromium.org>
Signed-off-by: Alim Akhtar <alim.akhtar@samsung.com>
Reviewed-on: https://gerrit.chromium.org/gerrit/31021

CHROMIUM: ARM: config: cros5250: set HDMI I2C speed to ~400 kHz

The HDMI I2C is hooked up internally to the HDMI core which supports
Fast Mode (400 kbit/s) I2C operation.

Note: actual rate is reported as 378 kHz.

Signed-off-by: Daniel Kurtz <djkurtz@chromium.org>
BUG=chrome-os-partner:8665
TEST=No i2c-8 (12ce0000.i2c) errors when plugging/unplugging hdmi cable

Change-Id: I79f259ce7d02786ba57bca2cf79ee8650617591e
Reviewed-on: https://gerrit.chromium.org/gerrit/29877
Commit-Ready: Daniel Kurtz <djkurtz@chromium.org>
Reviewed-by: Daniel Kurtz <djkurtz@chromium.org>
Tested-by: Daniel Kurtz <djkurtz@chromium.org>
Reviewed-by: Doug Anderson <dianders@chromium.org>