1. Add jpege_api_v2 module.
2. Move all control process to jpege_api_v2
3. Add hal_jpege_vepu1_v2 and hal_jpege_vepu2_v2 modules.
3. Add default rc for jpege.
Change-Id: I08c8adeb2d6613cbc074ee4978b9533739c21bb1
Signed-off-by: Herman Chen <herman.chen@rock-chips.com>
1. Set default node to 800 and info to 80.
2. Fix mpp_trie_test compile warning.
Signed-off-by: Herman Chen <herman.chen@rock-chips.com>
Change-Id: Ie63ce123e18fefc3108d3be7cbdc50e08c21f846
On FIXQP mode all qp will be set to init_quality.
Change-Id: Ic614c583e5dba0e5e4e129f46437adc87cd13fa0
Signed-off-by: Johnson Ding <johnson.ding@rock-chips.com>
For better binary comaptibility users should use this method to
configure encoders.
NOTE: api only init once.
Change-Id: I5ced15b504e4933560895e6e11f62613ae89e350
Signed-off-by: Herman Chen <herman.chen@rock-chips.com>
MppTrie will use the location of the input stream pointer as its
indexing pointer.
Change-Id: I2cb0b7d9cc2452db5698a971d08a24b98e5bc587
Signed-off-by: Herman Chen <herman.chen@rock-chips.com>
We are going to use new configure method for all encoders.
Change-Id: Id478d2567714ee9c6ddee3d410a4a855825f1c42
Signed-off-by: Herman Chen <herman.chen@rock-chips.com>
H.264 qp_sum is calculated by summon up 16x16 macroblock qp.
H.265 qp_sum is calcultaed by summon up 8x8 block qp.
Change-Id: I08d483d9cb997b25b9b6258259cb3b2290da3e4a
Signed-off-by: Herman Chen <herman.chen@rock-chips.com>
1. Fix h.264 encoder roi buffer leak error.
2. Set all qparea to [1, 51] range.
Change-Id: I8434fc5fc60071ec533aba68fc2fe3be5a087bcd
Signed-off-by: Herman Chen <herman.chen@rock-chips.com>
When gop / fps / rc_mode in RcCfg is changed the header should be
resend.
Change-Id: Id055cfe2f752dc741dc05012f954198c00077caa
Signed-off-by: Herman Chen <herman.chen@rock-chips.com>
NOTE: encoder may have more than 5 address to add offset patch.
Increase max offset patch number from 5 to 16.
Change-Id: I0fe3c5f7a24596f87d8248fd37b4c4aa5a059dcc
Signed-off-by: Herman Chen <herman.chen@rock-chips.com>
Due to complex input image content mjpeg encoder required larger buffer
to avoid overflow and hardware mmu fault.
Signed-off-by: Herman Chen <herman.chen@rock-chips.com>
Change-Id: I4417ae5e62f16673a2ea06a5954463ce34b6bb8a
On vepu541 the H.264 width and height should be aligned to 16 first
which is different to H.265.
Change-Id: I6f20960916aedf47cd778c22d7dc1ad860b03bc9
Signed-off-by: Herman Chen <herman.chen@rock-chips.com>
1. according frame type set max min quality qp
2. fix reenc check ins_bsp calc issue
Change-Id: Iededb7e64c40cc04432272ad3dd4b7d55c2aebd5
Signed-off-by: sayon.chen <sayon.chen@rock-chips.com>
On kernel 4.19 the send registers and recv registers can be different.
But due to compatibility kernel 4.4 the register copy is needed.
Change-Id: I1af7bbf8c27671627bda1679283693ec7eb4b0b5
Signed-off-by: Herman Chen <herman.chen@rock-chips.com>
New vepu541 encoder support relative qp delta config in ROI region.
Signed-off-by: Herman Chen <herman.chen@rock-chips.com>
Change-Id: Iac57469c30ea09ca03368cc3e02372a326c525ec
use "-r numerator:denominator:flex"
for set the input to the same fps as the output, such as 50:1:1
or "-r numerator:denominator/flex-numerator:denominator:flex"
for set input and output separately, such as 40:1:1/30:1:0.
the default parameter is 30:1:0.
Signed-off-by: Xinhuang Li <buluess.li@rock-chips.com>
Change-Id: Ief2935354b82aba9c192b4af92636519d3e75423
1. Use mpp_device_patch_add to setup bitstream buffer.
2. Add more log for hw_status.
3. Add madi and madp update.
Signed-off-by: Herman Chen <herman.chen@rock-chips.com>
Change-Id: Ie7e049e7e1af34688cbf8db8525515b508e69be4
Active sps parameters does not need to be updated in parsing
SEI information, which may cause incorrect updates.
Change-Id: I346ca8cdc6d19cba7e4a991fa2881e26567d1f00
Signed-off-by: Rimon Xu <rimon.xu@rock-chips.com>
When rc mode changed from CQP to CBR, bitrate configed from outside of
mpp maybe keeped the same, but bitrate value at mpp_rc should be
changed.
Change-Id: I845c05977ed53f664a70d1310684ba8ebfe5ee2b
Signed-off-by: Johnson Ding <johnson.ding@rock-chips.com>
The FBC data should be in this form:
+---------------------+
| header (4K aligned) |
+---------------------+
| |
| payload data |
| |
+---------------------+
The payload data size is equal to the orignal format size.
Signed-off-by: Herman Chen <herman.chen@rock-chips.com>
Change-Id: I2855c89db43c3e2cc302ac40e96e36ef6147f041
