Age | Commit message (Collapse) | Author |
|
Bug: 240346564
Test: Build successfully, check sysfs for decon symlinks
Change-Id: Ie5e696b040fa5950d0a60960969aa1b3daa8b981
Signed-off-by: Taylor Nelms <tknelms@google.com>
Signed-off-by: Will McVicker <willmcvicker@google.com>
|
|
aosp/android14-gs-pixel-6.1
* aosp/android-gs-raviole-5.10-android14-qpr2: (145 commits)
display: nt37290: delay DDIC DSC commands for RRS
panel: s6e3hc3: update idle vrefresh in backlight state
panel: s6e3fc3-p10: add log when 8-10 bit cmds are batched together
drm: samsung: add callback to panel when cmds are queued
panel: s6e3fc3-p10: send te2 command only for sleep in/out
panel: s6e3fc3-p10: avoid batching 8-10 bit commands together
drm: samsung: don't sleep in __dpp_disable()
panel: allow to update panel idle mode later
drm: samsung: add crtc fps in log
drm: samsung: Update parallel dsim_write_data for Dual DSI
drm: samsung: support to suspend a decon without hibernation
display: nt37290: update refresh rate when enabling panel
panel: add the progress type of mode switch & keep higher BTS while booting for RRS
display: s6e3hc4: delay DDIC DSC commands for RRS
drm: samsung: decon can suspend as long as crtc is enabled
drm: samsung: reset timer while increasing block count of hibernation
panel: correct TE refresh rate type information
drm:samsung: remove histogram hibernation WARN() message
panel: s6e3hc4: disable panel features before adjusting Vreg
panel: s6e3hc4: disable panel features before adjusting Vreg
drm: samsung: rewrite the disable_outputs
Revert "Revert "drm:samsung: add histogram_always_on support""
Revert "drm:samsung: add histogram_always_on support"
drm:samsung: add histogram_always_on support
panel: s6e3hc4: use fixed TE in AOD mode
drm: samsung: add trace for frame transfer
drm: samsung: trigger emergency reboot if fail to set PLL_EN
display: s6e3hc4: show idle refresh rate in AOD mode
display: nt37290: show idle refresh rate in AOD mode
Revert "Revert "panel: predict nearest te timestamp for mipi sync timing if possible""
drm: samsung: improve debug log for sysmmu fault issue
Revert "panel: s6e3hc4: update AOD brightness"
drm: samsung: correct some DPU events information
drm/samsung: decon: support TE width measurement
panel: s6e3hc4: increase source amplifier power in AOD
panel: s6e3hc4: update init and AOD command sequence
samsung/panel: add the functions of waiting for vsync
drm: samsung: set dma mask while decon probe
drm: samsung: support sbwc encoded writeback
panel: s6e3hc4: update AOD brightness
panel: s6e3hc4: increase source amplifier power in AOD
Revert "panel: s6e3hc4: update AOD power voltage"
panel: s6e3hc4: update AOD power voltage
panel: s6e3hc4: update init and AOD command sequence
samsung/panel: add the functions of waiting for vsync
drm: samsung: dsim: handle mode change on TUI exit
drm: samsung: move commit work to crtc state
drm: samsung: change tui sequence for video mode
Revert "drm: samsung: tui: skip blankmode transition for video mode panels"
drm: samsung: improve debug messages for secure contents
drm: samsung: call local_clock to log timestamp
samsung: cal_9845: check PLL_EN right after config it
drm: samsung: add more log to debug stability issues
drm: samsung: dsim: cleanup dsim command packet prepare
drm: samsung: dsim: check for fifo status before wait
drm: samsung: dsim: cleanup packet-go even after failure
panel: s6e3hc4: update command sequence to mitigate artifacts
samsung: panel: don't duplicate setting brightness=0
samsung: panel: don't duplicate setting brightness=0
panel: s6e3hc4: update command sequence to mitigate artifacts
Revert "drm: samsung: restore immutable flag for lp_mode"
drm: samsung: recovery: isolate recovery only to a single pipeline
drm: samsung: add MIPI_CMD_SYNC_NONE to mipi_sync_mode
drm/samsung: don't allow enable vblank if decon is not ready
panel: s6e3hc4: Update panel voltages settings
panel: s6e3hc4: Update panel voltages settings
drm: samsung: restore immutable flag for lp_mode
drm: samsung: turn off cwb after framestart interrupt
Revert "drm: samsung: rewrite the disable_outputs"
panel: check if get_te_usec exists before call it
panel: remove time_window_wait_crtc2
panel: add refresh_rate sysfs node
panel: add the function to get current TE pulse time
Revert "panel: predict nearest te timestamp for mipi sync timing if possible"
drm/samsung: don't sleep in decon_disable_irqs()
drm: samsung: tui: skip blankmode transition for video mode panels
drm: samsung: add event log to check decon config
drm: samsung: skip decon suspend/resume if decon is disabled
drm: samsung: tui: skip blankmode transition for video mode panels
samsung: drm: rename decon sysfs counter node
samsung: support sysfs counter node for decon
drm/samsung: fix the race between toggling decon and interrupt handler
drm/samsung: ensure handover dpp could get disabled unconditionally
drm: samsung: limit max_disp_freq for concurrent writeback
drm: samsung: rewrite the disable_outputs
drm: samsung: corrrect ppc when calculating DPU ACLK
drm: samsung: add rcd_id property
panel: provide way to customize the parse_regulators.
drm: samsung: add rcd_id property
panel: nt37290: force to disable LHBM while disabling panel
drm: samsung: keep rpm active before dumping dpu
drm: samsung: add more dpu event log
samsung: drm: rename decon sysfs counter node
panel: predict nearest te timestamp for mipi sync timing if possible
drm: samsung: add panel command tracing
drm: samsung: dump dpp/rcd/cgc in decon On and correct debug message
drm: samsung: protect to init and deinit DSIM/DPHY hardware.
Revert "samsung/cal_9845: avoid concurrent dphy reset of multiple dsim channels"
samsung: support sysfs counter node for decon
drm: samsung: dsim: handle cont splash handover
...
Change-Id: If425cb72f063e57ce12853714295e7fd293b20af
Signed-off-by: Will McVicker <willmcvicker@google.com>
|
|
Delay DDIC DSC commands to the same VSYNC with framestart
Bug: 263139741
Test: Boot on successfully
Test: Switch resolution in Setting and check the timing in trace
Change-Id: Ibc63bc52046ec8af7c2d749d9a30868354d10140
Signed-off-by: Ken Lin <lyenting@google.com>
|
|
Simply update panel idle vrefresh follow by self refresh for
IDLE_MODE_ON_INACTIVITY mode to provide correct information in backlight
state.
Bug: 283374166
Test: check backlight state in normal and AOD mode
Change-Id: I4d530b4433130afb21753ca79ebb24ecb0b95f61
Signed-off-by: Midas Chien <midaschieh@google.com>
|
|
- add warning when 8 bit or 10 bit change command is not flushed
- add warning with call stack when 8/10 bit cmds are batched together
Bug: 290186193
Test: Display works properly
Change-Id: I4c9225dad77384328c21d949f5651339bbdeb2b3
Signed-off-by: Safayat Ullah <safayat@google.com>
|
|
Bug: 290186193
Test: build success
Change-Id: I2880af374f4cfe28451b48202cf239c73e3c5ca1
Signed-off-by: Safayat Ullah <safayat@google.com>
|
|
- Only send the TE2 command during init
- As lp settings use different register,
send the lp settings as well always.
- No need to send it during 60/90Hz or AOD on/off
Bug: 290186193
Test: hw verify
Change-Id: Iab84db36e0e30e77428193036c3ae865a7b446bf
Signed-off-by: Safayat Ullah <safayat@google.com>
|
|
- Control the batching from panel driver side
- Always force flush when 8 bit or 10 bit change happens
- Add log whenever we use the 8-10 bit change
Bug: 290186193
Test: Refresh rate, lhbm, AOD, display on/off
Change-Id: I225c83a329061aed5d9be589e66b9799c6561d1a
Signed-off-by: Safayat Ullah <safayat@google.com>
|
|
The caller may grab spin lock, so use disable_irq_nosync() instead of
disable_irq() to avoid sleep.
Bug: 287560601
Test: display suspend/resume
Change-Id: I4306aaf036c4a6918add677819fe0ff09111c09d
Signed-off-by: Ken Huang <kenbshuang@google.com>
(cherry picked from commit d2aab83add76cbe327b8fe1e93d9a0973315e616)
|
|
Updating the panel idle mode in the composer thread may be blocked by
the display driver when a frame is under transfer, resulting in frame
drop. To avoid this, allow the panel idle mode to be updated at the
pre-commit of the next frame, which ensures that command can be sent
without delay.
Bug: 263362191
Test: AoD, check refresh rate indicator
Test: update panel_idle, min_vrefresh and idle_delay_ms randomly
Change-Id: Ifcd2679c823c0e6aae3a8661d9f8bd209beada45
Signed-off-by: Midas Chien <midaschieh@google.com>
|
|
Currently, the display driver prints a refresh rate log based on
BTS FPS, which may not align with the actual crtc FPS.
Bug: 304763911
Test: check log
Change-Id: I72b600ee64186c0dd11197f1a5af651e8035bd9c
Signed-off-by: Ken Huang <kenbshuang@google.com>
(cherry picked from commit bf6d563a148fe1e313a6b82b177166c6033c768b)
|
|
Bug: 298975477
Bug: 268588097
Test: Check the perfetto trace
Change-Id: Ibd922747e5b5c32e21357f51d69f5143d8c8217c
Signed-off-by: Yuti Amonkar <yuti.amonkar@samsung.com>
Signed-off-by: Leo Chen <yinchiuan@google.com>
(Cherry picked from commit f62323821c68ae1bac743f8fb640b37e4d087f66)
|
|
Bug: 300854197
SBMerger: 558810260
Change-Id: Idbea22b483d0688d52fdde190e88dc23ad845b7a
Signed-off-by: SecurityBot <android-nexus-securitybot@system.gserviceaccount.com>
|
|
android13-gs-pixel-5.10-udc-qpr1" into android13-gs-pixel-5.10-udc-qpr1
|
|
Conflict files:
samsung/panel/panel-samsung-s6e3hc4.c
Bug: 292447561
SBMerger: 558810260
Change-Id: I8139867a3f5769154fc46180d22e54f2b2c4e2e4
|
|
If a decon doesn't support hibernation, decon_suspend() will return
directly and sometimes may cause power domain fails to suspend dpu.
Bug: 296744776
Test: echo mem > /sys/power/state
Test: echo 1 > /d/dri/0/crtc-0/recovery
Change-Id: Ib0d9008bd335d1200581e17979b16a6bb877d93a
Signed-off-by: Ken Huang <kenbshuang@google.com>
(cherry picked from commit ce371a081aacdcf8a988bafbf19f75ebbff633e6)
|
|
When mode setting is required before panel initialization, the mode is
set when enabling the panel. In addition to changing the resolution,
we also need to update the refresh rate.
Bug: 297581026
Test: boot, on/off, AoD in/out, refresh rate switch
Change-Id: Ic5d7e76bca286324c703075ed8db79ec76c2872d
Signed-off-by: Midas Chien <midaschieh@google.com>
|
|
for RRS
Add the variable to indicate the mode switch is in progress with
refresh rate or/and resolution change in the same atomic commit.
Keep higher BTS in case we switch resolution and refresh rate
from high to low at the same time to avoid noises.
Bug: 283929569
Test: Boot on successfully
Test: Verified by PTE
Change-Id: I97685717527b70353eec2a65dc1317d4c70cf67d
Signed-off-by: Ken Lin <lyenting@google.com>
|
|
Delay DDIC DSC commands to the same VSYNC with framestart
Bug: 283929569
Test: Boot on successfully
Test: Verified by PTE
Change-Id: I0971f56200286e408211cce83de9c9126fc3ee9b
Signed-off-by: Ken Lin <lyenting@google.com>
|
|
There is a race condition during decon suspend if decon state has
been updated earlier than decon suspend while crtc is still enabled.
In that case, it makes pm suspend misunderstand dpu had suspended.
Bug: 282372962
Test: unplug usb to test AoD and suspend/resume
Change-Id: I133d14c049ad9c11cb02818d479513fe18b83aca
Signed-off-by: Ken Huang <kenbshuang@google.com>
(cherry picked from commit 9f1c59dd946fff4e1c7357fa81303065efb1987e)
|
|
Ensuring each call to increase block count can reset hibernation
timer. Besides, add more trace to debug hibernation issues.
Bug: 287054924
Test: check trace
Change-Id: I9a3ff1ae41e989b78bbac66c8173d2eb583482fb
Signed-off-by: Ken Huang <kenbshuang@google.com>
(cherry picked from commit 591a68c82f279f3f9d7a63c867fad09fe913dd3c)
|
|
If new rr has take effect, use current mode to get TE refresh rate type
information for wait TE low.
Bug: 290199919
Test: AOD, UDFPS
Change-Id: If68d35edf35637f498779ac7ce2a2e7949e2a7eb
Signed-off-by: Midas Chien <midaschieh@google.com>
(cherry picked from commit 3b8e8d398c39413a0c540952df82d6612a89b225)
|
|
Remove histogram hibernation WARN() message to avoid confusion.
We have more cases of pending frames at the time of hibernation
than initially anticipated.
Bug: 287900304
Change-Id: I3b8270eaf0f58ff22e120d7299ac8cba7e08b510
Signed-off-by: Oleg Matcovschi <omatcovschi@google.com>
|
|
Bug: 289967462
Test: PTE display test
Change-Id: Iaeffc837051ef1f3f7f256e4a41ae709cb8d27e8
Signed-off-by: Shiyong Li <shiyongli@google.com>
|
|
Bug: 289967462
Test: PTE display test
Change-Id: Iaeffc837051ef1f3f7f256e4a41ae709cb8d27e8
Signed-off-by: Shiyong Li <shiyongli@google.com>
|
|
This patch rewrite the disable outputs for touch. Now adopting
the following sequence. Touch should be disabled before decon.
1. Disable all bridges (include touch)
2. Disable crtcs (decon)
3. Disable encoders (dsim)
4. post disable on all bridges
Bug: 255678834
Test: Single tap to wake up
Change-Id: I53d5b8d1906e80c98cbec1c09100b9a040700048
Signed-off-by: Leo Chen <yinchiuan@google.com>
|
|
SBMerger: 526756187
Change-Id: Ief95a8b72785b5261970866f23bda8fa5339c471
Signed-off-by: SecurityBot <android-nexus-securitybot@system.gserviceaccount.com>
|
|
This reverts commit 10257b6c36f6aa404385dbd72ea69b865e9fb5f4.
Reason for revert: flaky test as mentioned in b/285587335#comment22
Bug: 261376040
Change-Id: I277fcc417968b9fa3ae755dd2ca33fc3d723a108
Signed-off-by: Leo Chen <yinchiuan@google.com>
|
|
This reverts commit f3deb4be40c0d55b5c676c47acb65d8605111a5f.
Reason for revert: b/285587335
Bug: 285587335
Signed-off-by: Oleg Matcovschi <omatcovschi@google.com>
Change-Id: I959256cc689a75fe1ed96cedead254a6a6da7218
|
|
SBMerger: 526756187
Change-Id: I2be715cb3c9b76f3f7bb6e8a3ff7cfbca38e737b
Signed-off-by: SecurityBot <android-nexus-securitybot@system.gserviceaccount.com>
|
|
Changes:
* track histogram engine status via simple state machine.
* preserve histogram_bins data when required.
* immediate return cached histogram data in context of
histogram_request ioctl whenever possible.
Note, histogram is enabled as long as DRM histogram properties are
valid.
Bug: 261376040
Change-Id: Iadc71d6af32651b760194156ce902e0b47fc2d8b
Signed-off-by: Oleg Matcovschi <omatcovschi@google.com>
|
|
Bug: 277560840
Change-Id: I0a457a0a13302d9b1f77c7a8e73a23f671bf3397
Signed-off-by: Wilson Sung <wilsonsung@google.com>
|
|
Set fixed TE to have a consistent TE frequency (30Hz) in AOD mode.
Bug: 281917185
Test: capture trace, signal measurement, AOD in/out
Change-Id: I7850d1bf9941f00dcb08fbc176784030a5c4bfde
Signed-off-by: Chris Lu <luchris@google.com>
|
|
Add the trace which will be changed according to frame start and
frame done IRQs.
Bug: 283419973
Test: capture the trace and check "frame_transfer"
Change-Id: I90417482fcba520f068b0cbb681528939b3dcf5c
Signed-off-by: Chris Lu <luchris@google.com>
(cherry picked from commit a32af6fcb0950c573fc2bfd3d6de53cf4bde02df)
|
|
Scandump doesn't record the right timing after receiving itmon error.
Trigger emergency reboot to ensure the issue timing.
Bug: 279094593
Test: trigger ramdump by hard code
Change-Id: I3da1f5febd335de8ffaf472b7fb27ff5c76ba61e
Signed-off-by: Ken Huang <kenbshuang@google.com>
|
|
The refresh rate indicator reads the sysfs node of display state,
and the state shows the actual refresh rate according to whether
panel idle refresh rate is set. This change sets idle refresh rate
to 10 in AOD mode so that the indicator can show correct refresh
rate.
Bug: 273577459
Test: cat sys/class/backlight/panel0-backlight/state while AOD
Change-Id: I5b023fd5a4f177c99bf2ea2cd36930d4b020ae19
Signed-off-by: Leo Chen <yinchiuan@google.com>
(Cherry picked from commit c18b655dde3e58c7c5274fc5420753557795c3c5)
|
|
The refresh rate indicator reads the sysfs node of display state,
and the state shows the actual refresh rate according to whether
panel idle refresh rate is set. This change sets idle refresh rate
to 10 in AOD mode so that the indicator can show correct refresh
rate.
Bug: 273577459
Test: cat sys/class/backlight/panel0-backlight/state while AOD
Change-Id: If8fb450caf5b12913d43ae2281bdeaa04ff77394
Signed-off-by: Leo Chen <yinchiuan@google.com>
(Cherry picked from commit c18b655dde3e58c7c5274fc5420753557795c3c5)
|
|
possible""
This reverts commit e302f182ae12d5d5ce8cec93e5aefe8dfb4c866d.
Bug: 280727794
Test: refresh rate switch, LHBM
Change-Id: I1372996beaade519e49a6256b91c273c43bde4d2
Signed-off-by: Midas Chien <midaschieh@google.com>
|
|
Bug: 278834131
Test: trigger sysmmu fault and check last kmsg
Change-Id: I809763e39b0441c545f65bbbb8c3bed87a92317c
Signed-off-by: Ken Huang <kenbshuang@google.com>
|
|
This reverts commit e65e7c4f651fd8a11c7eeba894e99bae8a3eb3a5.
Reason for revert: revert AOD changes since it's not needed
Bug: 280519993
Test: enter/exit AOD, change AOD brightness
Change-Id: I040de7e329b145c2876e153f918f56ffa7bc21a2
Signed-off-by: Chris Lu <luchris@google.com>
|
|
Use "CH" for id of dpp device. Add zpos to ATOMIC_COMMIT event. Correct
resource occupancy information in DECON_RSC_OCCUPANCY event. Add marco
to map id of plane to dpp device.
Bug: 247790341
Test: check dpu event
Signed-off-by: Midas Chien <midaschieh@google.com>
Change-Id: I92e308ed22331eb1b72cc7523891eed5448d5435
|
|
Report both TE rising and falling edge event to measure TE width.
Bug: 266484665
Test: echo 1 > /d/dri/0/crtc-0/force_te_on, check perfetto trace
Change-Id: I1f8c2a69739fee825b7de371efe0812e0fe554cc
Signed-off-by: Shiyong Li <shiyongli@google.com>
(cherry picked from commit 2d0a3e6aa273582a97cca581cd99a9c3dd4daa90)
|
|
Bug: 255246572
Change-Id: Idd002e6ca61c6f950d2e6e0f059fffe997dbab76
Signed-off-by: Robin Peng <robinpeng@google.com>
|
|
Bug: 274187749
Test: check AOD flicker
Change-Id: I95ee814a8e6a0f80f83c967fd76f8e69beb99783
Signed-off-by: Shiyong Li <shiyongli@google.com>
|
|
- set Vreg = 4.5V instead of 6.4V at display-off and 6.8V at
display-on.
- disable idle before setting Vreg=4.5V
- refactor AOD enter/exit command sequence, remove lp_off mode
Bug: 274187749
Test: check perfetto trace, enter and exit AOD
Signed-off-by: Shiyong Li <shiyongli@google.com>
Change-Id: If124d4ef0a0de56189cb5311598ecd98d402b1f3
|
|
Bug: 274187749
Change-Id: Iedd736ff7753da1ada8669e8f30c7603f87838ba
Signed-off-by: Shiyong Li <shiyongli@google.com>
|
|
Bug: 261700544
Test: add debug log to check coherent_dma_mask in plane update
Change-Id: I7bdb312d46de32913af17ddabb10d65845ffa3f2
Signed-off-by: Ken Huang <kenbshuang@google.com>
|
|
SBMerger: 516612970
Change-Id: Ic325879d956ee882d8e978604aa6f05627ae0660
Signed-off-by: SecurityBot <android-nexus-securitybot@system.gserviceaccount.com>
|
|
The feature to support sbwc encoded writeback was added. And,
DRM_FORMAT_NV12 to writeback also was added.
Bug: 187000170
Test: ./build_cloudripper.sh
Signed-off-by: Kwanghyun Chung <khyun.chung@samsung.com>
Change-Id: Id38229165b799799fecd653e2386a4d87bdfe368
|
|
support 10, 20 and 30 nits.
Bug: 274187749
Test: turn on/off screen, check AOD
Change-Id: Ieb764fbfd6a27f5b6671b52bbaff114b4556956d
Signed-off-by: Shiyong Li <shiyongli@google.com>
|