Currently, when an underrun occurs, a reset request is sent to the
histogram hardware. This has been observed to cause a hang in the histogram
collection cycle. The fix for this hang is to instead rely on the existing
checks for histogram validity in the histgram read worker. When the
underrun occurs, the histogram is invalidated, which in turn causes the
histogram read worker to reset the histogram collection when it attempts to
read its contents.
Change-Id: I66d33ddedef76fdcb6c4a3d8095c0e8ca695e33d
Signed-off-by: Carl Vanderlip <carlv@codeaurora.org>
(cherry picked from commit 2dd5234c094157bf77781b8646ac9ad114f819b0)
Signed-off-by: Pravin Tamkhane <pravint@codeaurora.org>
Make histogram enable/disable functions also enable/disable the histogram
"interrupt enabled" bit on the MDP enabled interrupts register.
CRs-Fixed: 398556
Signed-off-by: Carl Vanderlip <carlv@codeaurora.org>
Change-Id: I56f811e88cf3fd44ae0a2df8a6e5de32ba9a0c1c
(cherry picked from commit 1026832c51cbadeea9a717fa20fb42b6a7c5661f)
Signed-off-by: Pravin Tamkhane <pravint@codeaurora.org>
Histogram expects the display to be on while running, thus moving its
disable call to before disabling the panel timing generator ensures this
assertion.
Change-Id: I308b3e0c24f0111fb8549d0a0ae025910d649b75
Signed-off-by: Carl Vanderlip <carlv@codeaurora.org>
Conflicts:
drivers/video/msm/mdp4_overlay_dsi_video.c
drivers/video/msm/mdp4_overlay_lcdc.c
Signed-off-by: Pravin Tamkhane <pravint@codeaurora.org>
Handle the SET_FRAMERATE_MODE subdev ioctl for the Venus subdevice. Not
handling this causes VIDIOC_S_PARM to error out.
Change-Id: I13ecc92957ee526f1c6be5b8f3adb3ab15163dc0
Signed-off-by: Deva Ramasubramanian <dramasub@codeaurora.org>
Signed-off-by: Neha Pandey <nehap@codeaurora.org>
Avoid freeing input buffer context when recon buffer allocation fails.
On other failures such as QUEUE_BUFFER fail, clean up and free the
allocated buffers.
Change-Id: I12f2b76c1f6f78e4db30b3ce9cb781a20fcbdc16
Signed-off-by: Deva Ramasubramanian <dramasub@codeaurora.org>
CRs-Fixed: 415787
Signed-off-by: Neha Pandey <nehap@codeaurora.org>
Poor video quality is observed as VUI fps information is not present in
the encoded bit stream. Add support for VUI timing info in video driver
and an additional V4L2 control to set from user space.
CRs-Fixed: 399017
Signed-off-by: Srinu Gorle <sgorle@codeaurora.org>
Conflicts:
include/linux/videodev2.h
Change-Id: I970d4c479009da2d08531a9e9b8489a84cc5d44c
Signed-off-by: Neha Pandey <nehap@codeaurora.org>
The V4L2 framework doesn't appear to be clearing out vendor specific
flags between each QBUF call. Hence clear out all flags that we set
explicitly in the driver.
Change-Id: Id6ca10aeffe3821dacfe4d08b8c00ba48badcbb2
Signed-off-by: Deva Ramasubramanian <dramasub@codeaurora.org>
Signed-off-by: Neha Pandey <nehap@codeaurora.org>
Use proper flags while flushing from vidc instead of using the
ports. Also clean up some logs.
Change-Id: Ia1a1b64d90c6f2d56e895897b068b4f7d4cbeeb3
Signed-off-by: Deva Ramasubramanian <dramasub@codeaurora.org>
Signed-off-by: Neha Pandey <nehap@codeaurora.org>
The mdp_updates variable is getting incremented when giving a buffer to
MDP subdev. It should be getting incremented when we get it back from
the subdev.
Change-Id: I467f512e5d33484fd5c82579c168d66d3c48dd5a
Signed-off-by: Deva Ramasubramanian <dramasub@codeaurora.org>
Signed-off-by: Neha Pandey <nehap@codeaurora.org>
The client needs a way to know if a buffer contains the sps/pps. Hence
mark the v4l2 buffer with V4L2_QCOM_BUF_FLAG_CODECCONFIG.
Change-Id: I3bba1390ba1a8a3f1d61f8d015aefce1fd79ac74
Signed-off-by: Deva Ramasubramanian <dramasub@codeaurora.org>
Signed-off-by: Neha Pandey <nehap@codeaurora.org>
Sometimes during suspend resume, the mdp task thread
fails to dequeue buffer and errors out and exits.
Later when kthread_stop is called on mdp task, the
device crashes in kthread_stop because the mdp task
thread is invalid.
Change-Id: I42aee61b82c97c378c8497ea70eb53b3eeb24191
CRs-fixed: 394139
Signed-off-by: Gopikrishnaiah Anandan <agopik@codeaurora.org>
Signed-off-by: Arun Menon <menon@codeaurora.org>
Signed-off-by: Neha Pandey <nehap@codeaurora.org>
Due to improper ION handle check, unmap is not happening.
Added proper ion handle check, to unmap buffers.
Change-Id: I0fcf82307deee9845cdc98c76ac6e87af6dbcd62
CRs-Fixed: 401244
Signed-off-by: Srinu Gorle <sgorle@codeaurora.org>
Signed-off-by: Neha Pandey <nehap@codeaurora.org>
Add new V4L2 control to set AU delimiter for AVC.
Change-Id: I8c72cb89f18c9b2b550b906d797eafb73ac4ebbe
CRs-Fixed: 389030
Signed-off-by: Srinu Gorle <sgorle@codeaurora.org>
Signed-off-by: Neha Pandey <nehap@codeaurora.org>
Define Device Tree bindings for wfd and update the driver to match the
specification.
Change-Id: Ia204cf32aa1ccd5fd51999844dd73d740e90f0e2
Signed-off-by: Deva Ramasubramanian <dramasub@codeaurora.org>
Signed-off-by: Neha Pandey <nehap@codeaurora.org>
Move the domain specific mappings into subdevices to abstract away iommu
from wfd-ioctl.c. This is especially needed as the encoder subdevices
for MSM8x60 and MSM8x74 differ in the iommu domain to which they map the
buffer to. This target specific handling needs to taken care of by the
respective subdevices rather than by wfd-ioctl which should be target
agnostic.
Signed-off-by: Deva Ramasubramanian <dramasub@codeaurora.org>
Conflicts:
drivers/media/video/msm_wfd/mdp-4-subdev.c
Change-Id: I83b2cbcff2b86b0f9cce9844b1e3b1607f5881ca
Signed-off-by: Neha Pandey <nehap@codeaurora.org>
Display commit will work as overlay commit when
MDP_DISPLAY_COMMIT_OVERLAY flag is set. It will be
non-blocking if wait_for_finish is false.
Signed-off-by: Ken Zhang <kenz@codeaurora.org>
Signed-off-by: Naseer Ahmed <naseer@codeaurora.org>
Change-Id: Ibd143dbfcd2fdd7a8fab3038c5af4e2fbc9ea198
Frame buffer maintains a reference count (f_count) which increased
by fget() and decreased by fput(). overlay play will apply
fget() to frame buffer if it was the soruce buffer and fput()
will be applied to it after pipe commit done. However there has
extra fput() apply to frame buffer at the time pipe is freed.
This patch fix the problem by clear MDP_MEMORY_ID_TYPE_FB from
pipe's flags to avoid extra fput() applying to frame buffer.
CRs-fixed: 411394
Change-Id: Ib2b72e6277480b5ace23cda554be485d9d477b3d
Signed-off-by: Kuogee Hsieh <khsieh@codeaurora.org>
Signed-off-by: Siddhartha Agrawal <agrawals@codeaurora.org>
cmdlist_commit perform dsi clock control which turns off
dsi clock unexpectly when dsi controller is configured as
command mode during booting up. This cause display panel
can not be lit up. This patch drop clock control mechanism
from cmdlist_commit and move dsi clock control mechanism to
caller.
Change-Id: I1882ea3ee1c90ac5d76958c986740582929a5e3b
Signed-off-by: Kuogee Hsieh <khsieh@codeaurora.org>
Signed-off-by: Siddhartha Agrawal <agrawals@codeaurora.org>
On timeout while waiting for the vsync, send the current
timestamp to the userspace. This resolves the infinite
wait seen during the bootup.
Change-Id: Ib1426c7c9a21c37758d7352740938627d4613fd6
Signed-off-by: Siddhartha Agrawal <agrawals@codeaurora.org>
vsync sysfs entries should be created in mdp_probe instead of
creating in panel on. This avoids waiting for the first event
control in framework
Change-Id: I18b05b8d2a65d489afc6e3274ec8dc7a5b9a5f9a
Signed-off-by: Mayank Chopra <makchopra@codeaurora.org>
Signed-off-by: Siddhartha Agrawal <agrawals@codeaurora.org>
vsync sysfs entries should be created in mdp_probe instead of
creating in panel on. This avoids waiting for the first event
control in framework
Change-Id: I983731481ea89bb7666be5f2a9f75a1e04fd963f
Signed-off-by: Mayank Chopra <makchopra@codeaurora.org>
Signed-off-by: Siddhartha Agrawal <agrawals@codeaurora.org>
Replace sending of vsync timestamp using uevents with sysfs entry
as uevent result in increase in power numbers due to broadcast in
nature.
Change-Id: I22cfbd68c8bcbe4a47e94074393f1cbf79920c58
Signed-off-by: Mayank Chopra <makchopra@codeaurora.org>
Signed-off-by: Siddhartha Agrawal <agrawals@codeaurora.org>
Conflicts:
drivers/video/msm/mdp_dma.c
Signed-off-by: Siddhartha Agrawal <agrawals@codeaurora.org>
Avoid eternal wait for vsync event when client tries to
read the corresponding sysfs entry. This will avoid hangs when
client tries to read in corner cases where the Timing
Generator or the clock or IRQ gets disabled.
CRs-fixed: 406752
Change-Id: I773687220b868823a9e4bfdae74361ee6487c5c6
Signed-off-by: Pradeep Jilagam <pjilagam@codeaurora.org>
Signed-off-by: Siddhartha Agrawal <agrawals@codeaurora.org>
Change to make the vsync IOCTL no wait during the disable sequence.
This is to increase the performance. Handle the synchronization of
vsync interrupt using state variables in drivers from now.
Change-Id: Ia5ec14493f233a95f21d01ea627200a6c278239f
Signed-off-by: Padmanabhan Komanduru <pkomandu@codeaurora.org>
Signed-off-by: Siddhartha Agrawal <agrawals@codeaurora.org>
Handle enabling/disabling the vsync interrupt and mdp clock
enabling/disabling in a better way on MDP 3.03 targets. This
will avoid target crashes due to unclocked register access.
CRs-fixed: 388751
Change-Id: I5c4a409772464ce7d06869374dcba5ad7e335955
Signed-off-by: Padmanabhan Komanduru <pkomandu@codeaurora.org>
Signed-off-by: Siddhartha Agrawal <agrawals@codeaurora.org>
When mdp commits a buffer to play, functions inside pan display, e.g.
mdp4_overlay_mdp_perf_upd are based on the current state of play, so
extending mutex to the scope of pan display to avoid race conditions.
Change-Id: I2e55567fd21de3738be066a2dee298d8122f12a3
Signed-off-by: Huaibin Yang <huaibiny@codeaurora.org>
Signed-off-by: Siddhartha Agrawal <agrawals@codeaurora.org>
Disable IRQ for DSI_CMD_TERM was already called in the isr.
Removing the extra call present in the cmds_rx API's.
Change-Id: I0000dd3166beb2a3b0c788327e1d7520e0c11a13
Signed-off-by: Siddhartha Agrawal <agrawals@codeaurora.org>
Add wait4videodone to commit dcs commands to dsi controller
while video engine is busy so that dcs commands will be
transmitted to panel at next beginning of BLLP.
Change-Id: I7ca40124813452f25a60f899437aeb9bce6feb14
Signed-off-by: Kuogee Hsieh <khsieh@codeaurora.org>
Signed-off-by: Siddhartha Agrawal <agrawals@codeaurora.org>
Pipe is not un-staged from mixer when it is unset.
It expects a pan_display() to un-stage it out of mixer.
Return fail at overlay_set() if pipe was still stage
at mixer.
CRs-fixed: 399193, 393979
Change-Id: I562f6e133bb239b30f4f53669dbf1cc5ddf80d02
Signed-off-by: Kuogee Hsieh <khsieh@codeaurora.org>
Signed-off-by: Siddhartha Agrawal <agrawals@codeaurora.org>
There has VG1 pipe commit (pan display) happen after system
suspended. This left VG1 still staged up at mdp mixer. Once timing
generator is enabled at resume, VG1 pipe start fetching contents from
address 0 since VG1 has not yet be configured. This cause page fault.
This patch has sanity check at system suspend to make sure no any pipe
stage up at mixer after suspend.
Change-Id: Idcf974ceb4afe2a3ec55b9603b700fa497f84045
Signed-off-by: Kuogee Hsieh <khsieh@codeaurora.org>
Signed-off-by: Siddhartha Agrawal <agrawals@codeaurora.org>
For the panels that have low v_back_porch, if h/w FIR scaling is used,
during vbp time, mdp clk should be fast enough to output a line from 4
input lines. MDP rate equation for very small scaling, e.g. 1 or 2
pixels' upscaling or downscaling cannot cover this case, where
underruns are seen on those panels.
Change-Id: I4f35f98c2c1bc4ef04dfee193427bd3ff674945d
Signed-off-by: Huaibin Yang <huaibiny@codeaurora.org>
Signed-off-by: Siddhartha Agrawal <agrawals@codeaurora.org>
MDP4.1 blend operation registers are not double buffered and
take effect as soon as they are configured. Configuring these
registers at any place other than between dma and vsync produce
flickers as new blend values would get applied to previous
mixer configurations. Configure blend operation registers just
after dma is over, before vsync.
Change-Id: I377c11222f7a0cd77e98e90edec6e2e1dc0c17ca
Signed-off-by: Pradeep Jilagam <pjilagam@codeaurora.org>
Signed-off-by: Siddhartha Agrawal <agrawals@codeaurora.org>
Update LPA platform driver to register restart callback.
The implementation of the callback is to issue a pending buffer
written by the application to restart the session when the dequeue
process is stopped due to underrun issue.
The callback is triggered from ALSA core when the framework
detects that the renderer stopped pulling data but the application
is still writing data. It runs in atomic context.
Fixes:
-A disturbance was heard when switching from one song
to the other, or when the song is paused for greater
than 3 seconds and then resumed
-This was due to the previous value of the track
volume being used in the lpa driver
-This was fixed by setting 0 as the initial volume
Change-Id: I035760937d032038fb3f3383dc87ecbdc5f4700d
Signed-off-by: Subhash Chandra Bose Naripeddy <snariped@codeaurora.org>
If actual_freq of gpu is not updated when a gpu floor
is set, TZ is unaware of the gpu frequency change and
computations are based on an incorrect value. By setting
actual_freq, TZ is notified of the new gpu frequency at
the next idle exit.
Change-Id: Ifc69996ef309ed4f8a5d46a2048757e57dd1a8ec
Signed-off-by: Steve Muckle <smuckle@codeaurora.org>
(cherry picked from commit 81599d50a0d122c52c966279640d6d8c64def656)
The active and leakage energy offsets for each frequency are
best left configurable from userspace to make tuning easier.
Change-Id: I435a30d46cc02861dd7e8371b21d0433c0aa6a28
Signed-off-by: Steve Muckle <smuckle@codeaurora.org>
(cherry picked from commit 8104d6dde814e8dbb4f91a3e37a19627e7512f50)
It is desirable for performance reasons to be able to run dcvs
with different algorithm parameters depending on the number of
CPUs that are online. The algorithm parameters are already replicated
per-CPU, but all CPUs operate in DCVS with the same parameters at
any given time.
These per-CPU parameters now correspond to the total number of CPUs
online. For example, if 2 CPUs are online, the algorithm parameters
for CPU2 are used.
Change-Id: If308a2fa949ac669823e759e0fd8e2cdbf6e4c03
Signed-off-by: Steve Muckle <smuckle@codeaurora.org>
(cherry picked from commit c48381e2336125410549d7df1425dc8009c0c495)
Performance is enhanced if the GPU frequency is floored at its
maximum nominal frequency when multiple CPUs are online.
Change-Id: Ib6bfd5007d660064844876b06f552f4d8c84865c
Signed-off-by: Steve Muckle <smuckle@codeaurora.org>
(cherry picked from commit 749f301c94312d2ec827c67cecb858e02383c8ab)
Once a gpu calls stop_sink and is no longer participating in dcvs, we
should not call set_floor_frequency callbacks for that gpu.
Change-Id: I3cde82dc9975bcb37f065ea7ab8eb88a8f7fc933
Signed-off-by: Steve Muckle <smuckle@codeaurora.org>
(cherry picked from commit 77c66a3c6c27729aa0594b2f1b140aaf8f556b67)
If there is a pending frequency update when a core stops
acting as a dcvs sink, it should just be dropped, rather
than trying to honor that last frequency request.
Change-Id: Iac30cc7e06a8a70eecc4da712fdf160b788a7c2c
Signed-off-by: Steve Muckle <smuckle@codeaurora.org>
(cherry picked from commit 016af00bf66bb267cab573131598edbf83204f3f)
Temperature updates should not be sent continuously just
because a core has registered with dcvs. Instead, start
and stop the updates when the core starts and stops as a
dcvs frequency sink.
Change-Id: If7721f523eb44026c421dd7ec4fa3975bc65cd09
Signed-off-by: Steve Muckle <smuckle@codeaurora.org>
(cherry picked from commit 4eb5d7cf22bbea963251ff70a7b4e28a2d883b77)
To avoid spending a lot of time at once in the cpu hotremove
path, do not perform multiple hotremoves back to back.
Change-Id: I78a643f73718e4e08786193c71bbe2237aa790d2
Signed-off-by: Steve Muckle <smuckle@codeaurora.org>
(cherry picked from commit 9c0aa035f7aeac1b1039125bf27d51d11861fa03)
System performance is enhanced if the gpu frequency is given a
minimum corresponding to various frequency levels of CPU 0.
Change-Id: Iba168d708524fc8ef164428bb5f4e0631a499342
Signed-off-by: Steve Muckle <smuckle@codeaurora.org>
(cherry picked from commit 682c7a01c1d86518cdc7bec25cb413498811137b)
Target reset happens when trying to do a ion_unmap_iommu
on the above fd, which was never mapped. Added a check for
non-zero virtual address.
Change-Id: I9a4d81eea6ee0cdbe3e8bbcb0c1d05c005680555
CRs-Fixed: 394779
Signed-off-by: Srinu Gorle <sgorle@codeaurora.org>
Signed-off-by: Neha Pandey <nehap@codeaurora.org>
DCVS only expects a core to be registered once and kgsl reregisters
the core every time the policy changes. Kgsl should only do the
sink start and stop every time and only do the register the first time.
Change-Id: Id54f078d6013586899bf85fc2462e6c473ffc6b5
Signed-off-by: Suman Tatiraju <sumant@codeaurora.org>
(cherry picked from commit 229d24c088699c5f7a3591e5335dcc1668e5e913)
CPU frequencies can differ between different revisions, so create
the table of cpu frequencies dynamically based on data from
acpuclock rather than using a fixed table.
Some of the data in the table comes from sync points that will be
implemented between CPU and GPU frequencies. The sync points are
defined here, but the sync implementation itself will come in a
subsequent commit.
Change-Id: I24cf17b199df1e5d064c1894b269e49118b39ac7
Signed-off-by: Steve Muckle <smuckle@codeaurora.org>
(cherry picked from commit f9a8749b17961be653e9cbb426b1b72cada7f61c)
It is not necessary to run these threads at a realtime
priority level.
Change-Id: I8c6492f22e803811db5757c51beeeb81f93f32b6
Signed-off-by: Steve Muckle <smuckle@codeaurora.org>
(cherry picked from commit 8862f41f1c632ba84f0a21e8045d70278fd67fd3)
It is useful to be able to toggle at runtime whether
particular DCVS points are used as transient levels. This
adds a sysfs node per core which when read returns a list
of the currently enabled DCVS transient levels. Writing
the frequency of a DCVS transient level toggles whether it
is enabled or not.
Change-Id: I69ebb6974e97efa832798047259b9acdfd08aa7c
Signed-off-by: Steve Muckle <smuckle@codeaurora.org>
(cherry picked from commit 2f6db34b3216695d4f624d2204d5e1356e62c3d5)
This patch ensures that camera is recovered by propogating the
error event when sensor is pulled off.
Change-Id: I4147e02342c2538db8337d3e04735cf97b2d7635
Signed-off-by: Azam Sadiq Pasha Kapatrala Syed <akapatra@codeaurora.org>
