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>
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>
This change handles panel updates for WFD panel via
overlay_commit interface, along with conglomeration of
some common code across pan_update and overlay_commit ioctls
Signed-off-by: Kalyan Thota <kalyant@codeaurora.org>
Conflicts:
drivers/video/msm/mdp4_overlay.c
drivers/video/msm/msm_fb.c
include/linux/msm_mdp.h
Change-Id: I9e332856782c59ab598bb388a2ec482076746ad2
Signed-off-by: Sudhir Sharma <sudsha@codeaurora.org>
To improve performance, hardware vsync event at mdp driver
is passed to framework so that both composition and frame
update are aligned at vsync event. This patch will upgrade
wfd to incorporate vsync driven frame push in consistence
with other interfaces.
Change-Id: If088826ec387446301f1f6acc311d909edb04163
Signed-off-by: Kuogee Hsieh <khsieh@codeaurora.org>
(cherry picked from commit d74db5fcf24a223570137087aa5632593b315fcc)
Conflicts:
drivers/video/msm/mdp4_overlay.c
drivers/video/msm/mdp4_overlay_writeback.c
Signed-off-by: Sudhir Sharma <sudsha@codeaurora.org>
The mdp4 version reflects mdp's capability and is referred to at
many places. Currently, it is acquired by reading mdp version
register. Since the MDP version is defined in the board file,
this patch uses that version at probe time.
Change-Id: Iae2db1fb8609cb80f3705b5e47d3fefedc15fec5
Signed-off-by: Kuogee Hsieh <khsieh@codeaurora.org>
Signed-off-by: Naseer Ahmed <naseer@codeaurora.org>
Currently, alpha drop detection is done per frame which
means multiple blending stages share same alpha drop status
which does not reflect the alpha effect truly on each blending
stage. This patch enforces that the alpha drop detection
is done per blending stage so that alpha effect is applied
properly at each blending stage.
Change-Id: I2f456623c696192cf3388a635584a4a49d4720a0
Signed-off-by: Kuogee Hsieh <khsieh@codeaurora.org>
Signed-off-by: Naseer Ahmed <naseer@codeaurora.org>
When the blended background pipe has dropped alpha and it is going to
be blended with foreground pipe on the next stage,
MDP4_BLEND_BG_ALPHA_FG_CONST is taken. Foreground alpha constant is
0xFF, so we need to use this alpha value to show background, instead
of its inverted value.
Change-Id: I9412141b28a46ed3b58079a56969e79dea4ed57c
Signed-off-by: Huaibin Yang <huaibiny@codeaurora.org>
Signed-off-by: Naseer Ahmed <naseer@codeaurora.org>
Overlay commit and display commit were two separate
ioctls doing similar things. This change consolidates
the two into one display commit ioctl. With this, we
can use the RGB0 pipe with an IOMMU framebuffer.
Change-Id: I2b51e16a6678e6cec1bd3e56a00a476addb1296f
Signed-off-by: Kuogee Hsieh <khsieh@codeaurora.org>
Signed-off-by: Naseer Ahmed <naseer@codeaurora.org>
Since iommu buffers will be unmapped after two newer commits
had passed, there is no need to request iommu driver to further
delay unmapping an iommu buffer.
Change-Id: I8cf8d406cf509908035d12c9a04f0add12f3c7c7
Signed-off-by: Kuogee Hsieh <khsieh@codeaurora.org>
Signed-off-by: Naseer Ahmed <naseer@codeaurora.org>
Frame buffer's (fb0) ref_count's need to be balanced. overlay_play()
increases fb0's ref_count (fget_light). This patch will decrease fb0's
ref_count by calling fput_light after commit.
Change-Id: I352da1fce99b344c059f615a6fe88cc6e7387d89
Signed-off-by: Kuogee Hsieh <khsieh@codeaurora.org>
Signed-off-by: Naseer Ahmed <naseer@codeaurora.org>
Signed-off-by: Iliyan Malchev <malchev@google.com>
To enable left-right flip x-scaling bit should be set for
MDP 4.2 and above versions. This fixes the issue when RGB layer
pipes are not properly rotated.
Change-Id: I0afae4aecde39ce3ac4dcbefc0bda93510e7d18a
Signed-off-by: Siddhartha Agrawal <agrawals@codeaurora.org>
Signed-off-by: Naseer Ahmed <naseer@codeaurora.org>
For HDMI as primary case, we need to blackout the pipes
when they are unset so that any residue is not visible
on the DTV during suspend.
Change-Id: I2bf36f1002dd185c26fe6c4291977737daa73ae1
Signed-off-by: Ajay Singh Parmar <aparmar@codeaurora.org>
Signed-off-by: Aravind Venkateswaran <aravindh@codeaurora.org>
Both vsync enabled and disabled are controlled by framework
through ioctl. Driver reports vsync event to framework via
sysfs. Meanwhile, mdp related clocks are enabled by frame
work at vsync enabled request and disabled by driver if there
is no any display update within specifid vsync period.
Change-Id: Ice5be3a6db5930b95bdd2f718d9256bf73936e23
Signed-off-by: Kuogee Hsieh <khsieh@codeaurora.org>
(cherry picked from commit 3c406e4eec79b710b7f5b5b67c23db7af871be40)
Wait for fence signaled before commit, signal previous
fence after commit done. Same as pan display
Signed-off-by: Ken Zhang <kenz@codeaurora.org>
Signed-off-by: Naseer Ahmed <naseer@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. Also, for smart panel, wait4vsync is removed.
Change-Id: I5219e85b837b8bab9d00a9f40c1048d2a2355194
Signed-off-by: Mayank Chopra <makchopra@codeaurora.org>
Signed-off-by: Kuogee Hsieh <khsieh@codeaurora.org>
Conflicts:
drivers/video/msm/mdp4_overlay_dsi_video.c
Signed-off-by: Naseer Ahmed <naseer@codeaurora.org>
MSMFB_OVERLAY_COMMIT ioctl is used to display on external.
This is used instead of the default PAN mechanism to avoid delays in that path.
This ioctl for external also waits for DMAE thus fixing tearing.
Signed-off-by: Kuogee Hsieh <khsieh@codeaurora.org>
Signed-off-by: Arun Kumar K.R <akumarkr@codeaurora.org>
Clear original pipe attributes in the pipe free call.
Change-Id: I776869e7997e5b70e4c2b102e1a394034247bbcd
Signed-off-by: Kalyan Thota <kalyant@codeaurora.org>
Signed-off-by: Naseer Ahmed <naseer@codeaurora.org>
Adds support to enable or disable mixer commit during pipe staging operations.
Depending on the current activity, this selection can be made.
Change-Id: Ia2f5dbab58f73ef4de90abf7f278332d8d73df46
Signed-off-by: Pradeep Jilagam <pjilagam@codeaurora.org>
Signed-off-by: Ajay Dudani <adudani@codeaurora.org>
Signed-off-by: Iliyan Malchev <malchev@google.com>
An iommu page fault may heppen When a pipe_commit is executed
in between of frame dropped and pipe_queue during overlay play.
An mutex is necessary to mutual exclusive between pipe_commit
and overlay_play.
Change-Id: I736e821c6c648a6ebd435303a1906a40e9b75791
Signed-off-by: Kuogee Hsieh <khsieh@codeaurora.org>
Signed-off-by: Ajay Dudani <adudani@codeaurora.org>
To avoid double freeing writeback at end of blt mode, let
both alloc and free writeback buffer done at do_blt().
Change-Id: Ida5d24589c19be8f6f655eca6add00badca356b2
Signed-off-by: Kuogee Hsieh <khsieh@codeaurora.org>
Signed-off-by: Siddhartha Agrawal <agrawals@codeaurora.org>
Signed-off-by: Ajay Dudani <adudani@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: Ajay Dudani <adudani@codeaurora.org>
Fix the problem crashed while playing video clips and VT camera test.
ion_share_dma_buf is called in pr_debug.
So, it is called depends on build option and makes crash problem.
When in resume or boot time, baselayer pipes need to be committed for
mdp to play, otherwize underruns and blue screens are seen. The
previous pipe stage up func has this commit which is removed now, so
the commit calls have to be added for the cases where base pipe is
initialized.
Change-Id: I63a7d0ac3fb078b3d7f604eb646a0430326a478c
Signed-off-by: Huaibin Yang <huaibiny@codeaurora.org>
Signed-off-by: Siddhartha Agrawal <agrawals@codeaurora.org>
Currently mdp clk and bw requests are predefined in a 4-entry table
passed from the board file, and the logic to decide which level to use
is mainly based on source resolution. This patch is intended to
address several issues with this approach.
One major issue is clk and bandwidth depends on seperate things, and
need to be considered seperately. e.g. with mdp composition of
multiple pipes, bw request may be high but clk requirement may still
be low. The current approach that binds these two things together
causes inefficiency of power.
Another major issue is that there is no logic to calculate mdp clk
requirement of a single pipe based upon panel clk and downscale
parameters. Further the worst case of mdp clk should be determined by
all pipe usage. Without proper logic, many underrun have been
experienced, and blt mode may not be enalbed properly.
Also mdp_clk or blt mode update must be on right timing especially
between these two pathes: overlay play and pan display. In the
situations of performance from high to low or from low to high clk and
blt must be handled properly to avoid underruns.
In a summary, to support many different panels and targets and
complicated mdp pipe usage(mdp composition), mdp driver related to clk
and bw needs to be implemented differently. This patch is to seperate
clk and bw, and maily to deal with mdp clk and blt. Later increasing
granuity of bw will be added.
Change-Id: I678fbf86d6997ed7b602ce81cf2e0fff6164d129
Signed-off-by: Siddhartha Agrawal <agrawals@codeaurora.org>
Since stage commit had been splited out of stage up fucnction,
it needed to be called to stage pipes into mixer befroe
kickoff mixer. Add this function into overlay writeback for WFD.
Change-Id: I11e4486d92b89a15e58531033d474b3137b5935a
Signed-off-by: Kuogee Hsieh <khsieh@codeaurora.org>
Using RGB888 as source format and rgb565 as destination pixel format,
will cause image corruption on the display. Fix this issue by
setting the destination format as RGB888 in MDP and DSI controller
pixel format as RGB565 for DSI specific panels.
CRs-Fixed: 377484
Change-Id: If58976ee9cd4825efb39437170e54796fa2213e0
Signed-off-by: Chandan Uddaraju <chandanu@codeaurora.org>
At run time, mdp BLT (writeback) mode is enabled and disabled
base on performance level to avoid mdp underrun. Move BLT
enabled/disabled logic to pan display so that enabled/disable
is done per commit instead of per queue. Since all panels
are shared same logic, dsi video mode and lcdc panel BLT logic
have been changed accordingly.
Change-Id: I6e04bf0f7bf1c13634e19b18d8ac4368affbc4d8
Signed-off-by: Kuogee Hsieh <khsieh@codeaurora.org>
Base layer pipe is cloned and cached at satge array.
Retrive the real base layer pipe before used it.
Change-Id: I6fc606559684156348a57b500083ed4089fc71b5
Signed-off-by: Kuogee Hsieh <khsieh@codeaurora.org>
A few places have unnecessary mutex unlock calls, and iommu pipe free
depends on pipe parameters so move it before pipe is zeroed.
Change-Id: I7ef39c66be611907b1b0c38675b15f8bc919d0c0
Signed-off-by: Huaibin Yang <huaibiny@codeaurora.org>
During pipe free, some content of pipe is backed up and then after
pipe is zeroed out, the content get restored, in between there should
be no actions to change the content since any change will be wiped out
by the restore.
Change-Id: Ia6da0162d1f9b614682e1fc36edab1db003b42ab
Signed-off-by: Huaibin Yang <huaibiny@codeaurora.org>
Fixed iommu map buffer dropped after overlay unset by saving
origianl iommu list before zero out overlay pipe structure.
Change-Id: I498dd4cf62e93f4571883fe20a0298e9fad4ccab
Signed-off-by: Kuogee Hsieh <khsieh@codeaurora.org>
When HDMI is primary, dtv is associated with node 0. In this case,
DTV vsync and ctrl need to be called for node 0.
CRs-Fixed: 383877
Change-Id: I425cddf453dd159259883e71f92c32918ca8e6c7
Signed-off-by: Ajay Singh Parmar <aparmar@codeaurora.org>
When a mdp pipe is in solid fill mode, the pipe outputs constant color
and pipe itself does not fetch data from memory, so pipe src size
should not matter. However, when src size is greater than dst size in
solid mode which means downscaling, underrun happens. In this
situation, setting src size to be same as dst size can remove underrun.
CRs-fixed: 368478
Change-Id: Ie3ffc34486129cddcc872fe6e5317873db985b97
Signed-off-by: Huaibin Yang <huaibiny@codeaurora.org>
MDP4 blend logic differs from GPU blend implementation as MDP4
does not differentiate between pre-multiplied alpha layers and
non pre-multiplied alpha layers. Check layers for pre-multiplied
alpha and change blend logic accordingly.
CRs-Fixed: 355755
CRs-Fixed: 361729
Change-Id: Id8253d3888f1ebbadba1b6bbd87e7d98bac22da7
Signed-off-by: Mayank Chopra <makchopra@codeaurora.org>
When mdp pipe is freed at end of video play back, Writeback
buffer is freed while mdp still fetchs data from it. This cause
system crashed and blue screen shows on screen. This patch
will free writeback buffer at correct time.
Change-Id: I46a5031c13d86f943d6604713e1841b482dffe96
Signed-off-by: Kuogee Hsieh <khsieh@codeaurora.org>
This patch will queue multiple surfaces and commit those
surfaces into mdp at same instance so that surfaces will
be blended and displayed at same time. Hardware vsync event
is delivered to the user space frame work via uevent. Both
queue and commit are controlled by frame work and synchonized
with vsync event. Therefore frame rate will match with vsync rate.
Change-Id: If630a6d94fd38483ee313f575b1a71ed8bd65a52
Signed-off-by: Kuogee Hsieh <khsieh@codeaurora.org>
Add support for CSC and QSEED configuration through MSMFB_OVERLAY_SET and
MSMFB_OVERLAY_PLAY ioctls. Provide better integration than previous attempt
with the existing Overlay and PostProcessing APIs.
CRs-Fixed: 380750
Change-Id: If895391074ef1e7d7a6824d289e1ed9da6a52180
Signed-off-by: Carl Vanderlip <carlv@codeaurora.org>
Remove old implementation of HSIC and QSEED smoothing/sharpening overlay
integration. Need to replace with new implementation that better integrates
the existing functions of the two APIs.
CRs-Fixed: 380750
Change-Id: Id3469fad16764ed88e74a8da75bb873f726ba366
Signed-off-by: Carl Vanderlip <carlv@codeaurora.org>
Signed-off-by: Pravin Tamkhane <pravint@codeaurora.org>
MDP pipe or mixer size registers have 12-bit number, so boundary
checks are needed for those parameters. If not, input req sizes could
be very large 32 bit unsigned numbers, e.g. 0xffffffff, unsigned
operations may have overflow to pass all later checks, causing MDP
underruns or hang issues.
CRs-fixed: 380469
Change-Id: I096d7022ede21d54266b46a140782ab0348dd54b
Signed-off-by: Huaibin Yang <huaibiny@codeaurora.org>
Tearing was observed while playing 1080p H264 videos in
potrait mode. Enabled blt mode check for MIPI command
mode panel as well.
CRs-Fixed: 369365
Signed-off-by: Siddhartha Agrawal <agrawals@codeaurora.org>
Change-Id: Iac980050e5966db4731240c11e89e1f68118f2b7
For handling split IOMMU domain cases, BLT mode requires DMA to map
for reads and Overlay for writes. DMA and Overlay blt addresses are
added to configure the respective registers to avoid page faults and
normal operation of supported panel interfaces.
Change-Id: Id007cf96588f817bcfe8559177abb073a9b00755
Signed-off-by: Ravishangar Kalyanam <rkalya@codeaurora.org>
After writeback mode is switched for DSI Video interface, wait for
overlay done for prefill frame to avoid unnecessary fetches from previous
buffers resulting in page faults.
CRs-Fixed: 372389
Signed-off-by: Ravishangar Kalyanam <rkalya@codeaurora.org>
Change-Id: Iacd3775a718fe8e08404b4d119d447a88b628617
