PD#SWPL-182261
Problem:
output log continuously while the video is playing
Solution:
adjust print level
Verify:
s7d
Test:
DRM-OSD-44
Change-Id: I9c90096be174c026285a727f9e2dce327fc3b225
Signed-off-by: mingyang.he <mingyang.he@amlogic.com>
PD#SWPL-173046
Problem:
when switching from nonloopback to loopback, the connector
corresponding to crtc changes, but HWC still sets vrr_enable
to true. The previous field's brr_mode、brr、prev_vrefresh same
as the new filed, causing enable lcd failed
Solution:
not enable brr variables in duplicate_state,the brr is cal
in encoder_check
Verify:
s5 t5m
Test:
DRM-OSD-45
DRM-OSD-47
Change-Id: I7e7b2eeadc7714c37596f15ee2b34bc8f9dd7e06
Signed-off-by: congyang.huang <congyang.huang@amlogic.com>
PD#SWPL-180405
Problem:
didn't decrement fence reference count after dma_resv_get_excl_unlocked
Solution:
put video fence once more
Verify:
s6
Test:
DRM-OSD-28
Change-Id: I28ad558abf19980e6d7c631720c44496d3f5d611
Signed-off-by: mingyang.he <mingyang.he@amlogic.com>
PD#SWPL-181209
Problem:
dummy alpha have dummy bit in bit10~9, so the read value will be
shift and it is not equal to real value. rdma will fill the write
value according to read value when rdma_write_reg_bits be used for
write register.
Solution:
move global alpha setting to osdblend and write the register
at one time.
Verify:
s7d
Test:
DRM-OSD-77
Change-Id: I272633fd940baabaee81ec5b6a3c0f25c1354850
Signed-off-by: wenlong.zhang <wenlong.zhang@amlogic.com>
PD#SWPL-177148
Problem:
for save power consumption, suspend will turn off vpu power,
register will br reset.
Solution:
we need to do block register init again
Verify:
S7
Test:
DRM-OSD-75
Change-Id: Ia2e0a8e6404f624cc41f3aacffbce558179fcb70
Signed-off-by: wenlong.zhang <wenlong.zhang@amlogic.com>
PD#SWPL-176465
Problem:
10s for waiting dependencies in commit is too long.
Solution:
Use 100ms for waiting.
Verify:
s7d
Test:
DRM-TX-38
Change-Id: Ibc7c1e92eea0168421763a2b0acb90e5d4dce400
Signed-off-by: yujun.zhang <yujun.zhang@amlogic.com>
PD#SWPL-162724
Problem:
need nodes to resize video src and dst
Solution:
add nodes in sysfs
Verify:
sc2
Test:
DRM-OSD-73
Change-Id: I0fd226180f22b4222b425ce1e54f5da44dcaf90c
Signed-off-by: mingyang.he <mingyang.he@amlogic.com>
PD#SWPL-121840
Problem:
uboot enable vpp osd1 src control register when plugout hdmitx cable.
drm driver did not do logo_init.
when change bootanimation to bootvideo, then the screen will show osd
Solution:
disable pipeline's blocks when no plane was committed
Verify:
ah212, S5, T7c
Test:
DRM-OSD-27
Change-Id: I366ff01fcfbae6229e593d2a835c97cab64269af
Signed-off-by: yujun.zhang <yujun.zhang@amlogic.com>
PD#SWPL-179053
Problem:
osdblend ctrl enable premult of dinx for reva
Solution:
osdblend ctrl should set unpremult for dinx
and premult for core0&core1
Verify:
s7d
Test:
DRM-OSD-74
Change-Id: I41527fe80291e43daf4b48f4431498a51d6df75c
Signed-off-by: wenlong.zhang <wenlong.zhang@amlogic.com>
PD#SWPL-170739
Problem:
when change mode from ALLM to QMS and mode is same with brr, no need
to do mode change. when change mode from QMS to ALLM and QMS's mode is brr,
it also no need to do mode change.
Solution:
optimize seamless switch between ALLM and QMS
Verify:
s7d
Test:
DRM-OSD-5
Change-Id: Ic5fd113b0d8157ee02e9e33f00e5e85ba7ffcc6c
Signed-off-by: Ao Xu <ao.xu@amlogic.com>
PD#SWPL-168062
Problem:
compile error without dv
Solution:
fix error
Verify:
s5
Test:
S5
Change-Id: I591f88834b511375aafa875cd97e894ee3cfe478
Signed-off-by: yao liu <yao.liu@amlogic.com>
PD#SWPL-178017
Problem:
sg table used kmalloc, but vfree was used to release it
Solution:
change to kfree api
Verify:
S7D
Test:
DRM-OSD-71
Change-Id: I69ecaceeb6a1077550db520bb0cd22daac35cd6e
Signed-off-by: wenlong.zhang <wenlong.zhang@amlogic.com>
PD#SWPL-101847
Problem:
drm write registers to rdma table one by one, due to rdma table update
and rdma manager flush to real register are out of sync, so we want to
creat a fake table, the fake table will copy to rdma table after
pipeline register write done, and then config rdma vsync
Solution:
add rdma table for drm osd registers, and it can dynamic switching by
this node /sys/class/drm/card0/crtc0/rdma_table_switch.
echo 1 > /sys/class/drm/card0/crtc0/rdma_table_switch switch off
echo 0 > /sys/class/drm/card0/crtc0/rdma_table_switch switch on
Verify:
T3x s7 s5
Test:
DRM-OSD-33
Change-Id: I053daa7daaab672f8b1e9fc2c38a2d60552d6288
Signed-off-by: wenlong.zhang <wenlong.zhang@amlogic.com>
PD#SWPL-168643
Problem:
osd_cfg_sync_en bit of mif reg VIU_OSD2_CTRL_STAT gets
reset to be 1 after resuming.
Solution:
set osd_cfg_sync_en bit to be 0 in each commit if 1.
Verify:
s7
Test:
DRM-OSD-70
Change-Id: I7d3885eab088a597c25b41e0ce4b4377482315ba
Signed-off-by: yujun.zhang <yujun.zhang@amlogic.com>
PD#SWPL-168620
Problem:
when osd3 on/off, core2a and core2c reset, but lut
updated fail; drm enable osd3 is one vsync earlier
then dv core2c
Solution:
1.drm set core2c lut when osd3 off->on
2.add force_toggle_once debug
Verify:
s5
Test:
s5
Change-Id: I5fe2a7abd1f4b887da5f32a655a24418fee86b64
Signed-off-by: yao liu <yao.liu@amlogic.com>
PD#SWPL-166103
Problem:
on TCL55V6E TV, when STB change from 4k60hz y420,10bit
to 4K24hz y444,10bit by NFR, the TV may not decrypt
signal well and trigger a HPD reset to STB to handshake
again with low probability, this will cause Netflix
exit play and show tvq-pb-101(5.2.12). The possible
reasons may be as below
1.if the interval between previous mode disable and
new mode enable is short, TV may not detect input
signal change clearly
2.if hdcp encryption changes from enable to disable,
while signal not disabled, this may confuse TV, even
later hdmitx signal is disabled, there's probability TV
detection state machine go wrong to HPD reset logic
to request source device re-handshake.
Solution:
when disable hdmitx output, disable phy asap along with
hdcp disable, so that TV can detect signal change better
Verify:
ohm
Test:
DRM-TX-37
Change-Id: I8b897bc1f909989a459b23cc408e8b32ab8cdc84
Signed-off-by: hang cheng <hang.cheng@amlogic.com>
PD#SWPL-174924
Problem:
currently apk get hdcp status from hdmitx driver directly,
and need to use unify interface of drm to get hdcp status.
Solution:
add the property of getting hdcp topo info
Verify:
S7D
Test:
DRM-OSD-69
Change-Id: I49943fc84533c621bfdc422d2164264c5da020a8
Signed-off-by: linfang.zhao <linfang.zhao@amlogic.com>
PD#SWPL-174338
Problem:
display black when sent 4k afbc osd1&osd2
Solution:
reset osd1 and osd2 out addr
Verify:
s6
Test:
DRM-OSD-6
Change-Id: I2da0256e36dde43e705669b0f6e0b59650145111
Signed-off-by: mingyang.he <mingyang.he@amlogic.com>
PD#SWPL-166558
Problem:
global alpha function not effect in gfcd
Solution:
use osdblend dummy alpha to set global alpha
Verify:
S7D & s6
Test:
DRM-OSD-15
Change-Id: Ife2b66cb4a300b979483dbbc2528147efd594dcb
Signed-off-by: wenlong.zhang <wenlong.zhang@amlogic.com>
PD#SWPL-170918
PD#SWPL-161717
Problem:
Sink may support only 4k120, but its QMS range may be from
24 to 120.
Or sink may support upto 4k60, but 1080p120hz.
Solution:
Add the qms/game vrr range and vic lists to DRM group
Don't clear the EDID when forced_edid is true
Skip BRR 120 check when QMS tfr_max is false
Verify:
s7d/ross
Test:
DRM-TX-75
Change-Id: Ia65fa38bf18ae7399c9c4178665f8a3f18a03a74
Signed-off-by: zongdong.jiao <zongdong.jiao@amlogic.com>
PD#SWPL-170495
Problem:
when drm and fbdev post display in the same vsync, there is mismatched
between canvas index with osd hw that caused display abnormal.
Solution:
make sure canvas update according to real register
Verify:
S4
Test:
DRM-OSD-67
Change-Id: Ia2f209fa2ddf4d9270bae74f4609e81c6758153c
Signed-off-by: wenlong.zhang <wenlong.zhang@amlogic.com>
PD#TV-94929
Problem:
support video rotate function
Solution:
support video rotate function
Verify:
t5w
Test:
DRM-OSD-68
Change-Id: Ibdbafd97ac8874f3df7bab01e52fe2d639e4f567
Signed-off-by: wenlong.zhang <wenlong.zhang@amlogic.com>
PD#SWPL-171726
Problem:
WARN api print so many messages, it will lead to system
high latency occasionally
Solution:
remove wait vblank warning messages print
Verify:
s7d
Test:
DRM-OSD-24, DRM-OSD-54
Change-Id: I03b22d692edf78dc116cd50f74f5a7d52ad398e0
Signed-off-by: Ao Xu <ao.xu@amlogic.com>
PD#SWPL-174421
Problem:
afrc just support scanline layout, it must notify to
gpu explicitly.
Solution:
afrc just support scanline layout, it must notify to
gpu explicitly.
Verify:
s7d
Test:
DRM-OSD-66
Change-Id: I414b7658e97aba0cde67516b042e4948e8b89e85
Signed-off-by: wenlong.zhang <wenlong.zhang@amlogic.com>
PD#SWPL-168462
Problem:
Play AVC format video source, flash green screen when exiting playback
Solution:
disable video plane is not using work queue
Verify:
S7D
Test:
DRM-OSD-27
DRM-OSD-33
Change-Id: I1be41f2502d04b2f0701e54358a315c6d3c708bb
Signed-off-by: linfang.zhao <linfang.zhao@amlogic.com>
PD#SWPL-170861
Problem:
osd mif color related api need get cpu version.
osd mif color format has duplicated definition
Solution:
reorganize osd mif color realted interface
Verify:
s7d
Test:
DRM-OSD-2
Change-Id: I6fecf32c8295ac380a777b5e751d58fcce640338
Signed-off-by: Ao Xu <ao.xu@amlogic.com>
PD#SWPL-172323
Problem:
hdmi block screen while entering recovery
Solution:
use appropriate sequence_id to prevent the failure of
drm set mode during recovery mode
Verify:
s7
Test:
DRM-OSD-2
Change-Id: Ic2e50120abc4edbd25d8de34126f2d0b716e7584
Signed-off-by: congyang.huang <congyang.huang@amlogic.com>
PD#SWPL-170497
Problem:
handle8(vpp1) and handle9(vpp2) write same reg with different val
Solution:
set VPP_INTF_OSD3_CTRL in handle9(vpp2) only
Verify:
t3x
Test:
DRM-OSD-13
Change-Id: I0443e4fa05674f205c81c1a3ca70b843f715189c
Signed-off-by: mingyang.he <mingyang.he@amlogic.com>
PD#SWPL-168999
Problem:
small surface scale up 1.5, it show vague
Solution:
change scaler coef from COEFS_4POINT_TRIANGLE to COEFS_BICUBIC
Verify:
txhd2
Test:
DRM-OSD-8
Change-Id: I7f0ebdc01bb1f6329e9421bd9eafa8395fa14656
Signed-off-by: Ao Xu <ao.xu@amlogic.com>
congyang.huang