summaryrefslogtreecommitdiff
path: root/vp9/common/vp9_reconinter.c
AgeCommit message (Collapse)Author
2014-11-24Fix a tautological assert.Alex Converse
Change-Id: I90ad08823e1d038384536fa9f458caadc2c87f38
2014-10-09Rename highbitdepth functions to use highbd prefixDeb Mukherjee
Uses highbd_ prefix convention consistently. Change-Id: I58f7f799a7ff8e32701bcd71c955bcf1cdd4581e
2014-09-29Miscellaneous decoder changes for high bitdepthDeb Mukherjee
Also includes yv12 config changes. Change-Id: Iacf40d8bf486815b54c32a127ce3cd4516b7e44f
2014-09-25Avoid calling vp9_is_scaled two times in a function.hkuang
Use a local variable to hold the result of vp9_is_scaled. Change-Id: I5e203909805923e20eefef596bc84424da47dbe2
2014-09-25Fix a couple of commentsYaowu Xu
The first comment is obselete given the way is now normative in VP9 bitstream. The second comment line was too long. Change-Id: I6546585babf60d466485ddcf2daa6d2fa79e999a
2014-09-24Correct the condition for border extensionYaowu Xu
As reported in issue #850, the condition for border extension was not complete. This commit added the case when the scaling is enabled. This fixes issue #850. Change-Id: I67768b23f0dcc4ac9a9aa0a0825b0fe8cb85a72e
2014-09-24Merge "Fix invalid memory access on 2x downscale."Yaowu Xu
2014-09-19Remove mi_grid_* structures.hkuang
mi_grid_* are arrays of pointer to pointer. They save the pointers that point to the MIs in cm->mi. But they are unnecessary and complicated. The original goal was to remove MODE_INFO_t copy. But with an extra MODE_INFO_t pointer inside MODE_INFO_t, same goal could be achieved. This commit totally removes the mi_grid_* structures. But there are still many dummy MODE_INFO_t inside cm->mi which are a waste of memory. Next commit will do on-demand MODE_INFO_t allocation in order to save these memories. Change-Id: I3a05cf1610679fed26e0b2eadd315a9ae91afdd6
2014-09-18Adds high bitdepth convolve, interpred & scalingDeb Mukherjee
Change-Id: Ie51c352a6b250547207cbc1ebba833a01ed053e3
2014-09-15Fix invalid memory access on 2x downscale.Alexander Voronov
The issue was discovered on bitstream with 2x vertical downscale. For zero MVs, y_pad is set to 1 only when vertical convolution is required. The original code assumes that for y_step_q4 == 32 we don't perform vertical convolution. But vp9_setup_scale_factors_for_frame() sets convolve functions so that when x_step and y_step are both not equal to 16, convolve in both directions is performed. And convolve() unconditionally subtracts one stride from source pointer when calls convolve_horiz(). This leads to invalid memory access. Change-Id: I882dfa6081a58e172b5ffa55842bfcd6727f10bf
2014-07-11Remove an unused parameterYaowu Xu
Change-Id: I6ad6fd75dc3c9e6218d88148cf49e205398e2af5
2014-07-10Fix a bug in boundary checking.hkuang
Change-Id: Ifc741da9da6f61c8d3c1f675ec6b8a96570f877d
2014-06-18BITSTREAM: Handle transform size and motion vectors more logically for non-420.Alex Converse
This breaks the profile 1 bitstream. Don't force non420 uv transform size to 1/4 y size. In the 4:2:0 case the chroma corresponding to a luma block is 1/4 its size. In the 4:4:4 case chroma and luma planes are the same size. Disallowing larger transforms can result in a loss of compression efficiency and is inconsistent. For sub-8x8 blocks only average corresponding motion vectors. 4:2:0 and profile 0 behavior remains unchanged. Change-Id: I560ae07183012c6734dd1860ea54ed6f62f3cae8
2014-05-16Removing MACROBLOCKD dependency from loop filter.Dmitry Kovalev
Change-Id: I9ef40f3d95ab8f94f69e92ea25678a40956bc1ce
2014-04-03Removing interp_kernel from MACROBLOCKD.Dmitry Kovalev
Now interp_kernel is obtained when it is really required (based on mbmi->interp_filter value). Change-Id: I4c7a93c179d1045eba16e7526c293d02c9b8b47e
2014-04-01Renaming two members in MACROBLOCKD struct.Dmitry Kovalev
Renames: mi_8x8 -> mi mode_info_stride -> mi_stride Change-Id: I66f3e5fd1e7b7f46f108af5bb711c5fd9493c1be
2014-03-24Change back the scaling calculation.hkuang
Let the calculation to be compatible with Google's HW implementation. Change-Id: I22e179888cdb0419e230351c0a47661b37051fef
2014-03-17Fix the md5 mismatch for some scale cases.hkuang
Fixes issue #731 Change-Id: Id313e84b8fb4ff20f6a4e1ed11cb601927888318
2014-03-12Revert "With on demand border extension, clamping the MV"Yaowu Xu
This reverts commit b0fec6ab4a61ded1ab2ade188987631f53c4e9c1. Change-Id: I9acd8ee0423f22d92138f11579611ff959331013
2014-03-12Revert "Remove dec_build_inter_predictors() parameters"Yaowu Xu
This reverts commit 9650b9d72aa236e76c54b4f0acebd6bf1d6bbe48. Change-Id: I841c4a4734170fda63469e32adc10703aa4bf0fa
2014-03-03vp9_reconinter.h static functions in header converted to globalJim Bankoski
Change-Id: I916944950deb22f4c2301d83a803b732bf3ecd77
2014-03-01Remove dec_build_inter_predictors() parametersYaowu Xu
There were two parameters not in use, this commit removed them. Change-Id: Ia03a73b9a2521400bed539df45574e34214ed93a
2014-02-14Cleanup some comments.Adrian Grange
Change-Id: I568861ba1d43620865ad9a98a97eef37a51fd856
2014-02-04Merge "Moving x1 & y1 calculation under if condition."Dmitry Kovalev
2014-02-04With on demand border extension, clamping the MVhkuang
is not longer needed. Change-Id: I40c37ef18c67ab27fc336694dfca3c43a87c47ca
2014-02-03Moving x1 & y1 calculation under if condition.Dmitry Kovalev
Change-Id: Iae787d491f7cfe24855ef8f2d04e2c6c19350378
2014-02-03vp9_filter.h: rename interp_kernel typeJames Zern
-> InterpKernel avoids conflicts in variable names, fixing the build with various toolchains. broken since: 8691565 Removing subpix_fn_table struct. Change-Id: Ib5f6fdbcb494a97b62c75b99d4d826ff25d4c981
2014-01-24Removing subpix_fn_table struct.Dmitry Kovalev
We don't use different filter kernels for x and y, it is always one kernel for both directions. Change-Id: Iefcbb02ec74bf46ea20d9dca672a3efd5d631517
2014-01-17Use a temp buffer for reconstruction whenhkuang
reference buffer is out of boarder. Change-Id: Ic7ad136e54a4d68abe0fd4345146a86b0ba824e1
2014-01-03Adding RefBuffer struct.Dmitry Kovalev
Adding RefBuffer to simplify reference buffer management. The struct has a pointer to image data and scale factors relative to the current frame. Change-Id: If38eb1491ff687cc11428aee339f3e052e2c5d9e
2013-12-19Using single struct to represent scale factors.Dmitry Kovalev
Moving back to scale_factors struct. We don't need anymore x_offset_q4 and y_offset_q4 because both values are calculated locally inside vp9_scale_mv function. Change-Id: I78a2122ba253c428a14558bda0e78ece738d2b5b
2013-12-19Call set_scaled_offsets() just before scale_mv() call.Dmitry Kovalev
Before mv scaling it is required to calculate x_offset_q4/y_offset_q4 by calling set_scaled_offsets(). Now offset configuration can not be missed because it happens just before scale_mv(). Change-Id: I7dd1a85b85811a6cc67c46c9b01e6ccbbb06ce3a
2013-12-15Merge "Add support to pass in external frame buffers."Frank Galligan
2013-12-15Add support to pass in external frame buffers.Frank Galligan
VP9 decoder can now use frame buffers passed in by the application. Change-Id: I599527ec85c577f3f5552831d79a693884fafb73
2013-12-11Merge "Fix valgrind error."hkuang
2013-12-11Fix valgrind error.hkuang
Temporarily change memcpy to memmove. Change-Id: I700a197bc1ce496be1ddad7118429c5da465b0ca
2013-12-10Refactor inter_predictor function.hkuang
Change-Id: Ic429b2f16462e926f30efb3af4da3080026359d8
2013-12-09Implenment on demand border extension. In place extendhkuang
the border now. Next commit will totally remove the border. Change-Id: Ic1e1ca9cc34f81c688715b3948689b47df63a151
2013-12-05Renaming constants.Dmitry Kovalev
NUM_YV12_BUFFERS => FRAME_BUFFERS ALLOWED_REFS_PER_FRAME => REFS_PER_FRAME NUM_REF_FRAMES_LOG2 => REF_FRAMES_LOG2 NUM_REF_FRAMES => REF_FRAMES NUM_FRAME_CONTEXTS_LOG2 => FRAME_CONTEXTS_LOG2 NUM_FRAME_CONTEXTS => FRAME_CONTEXTS Change-Id: I4e1ada08f25d8fa30fdf03aebe1b1c9df0f87e63
2013-12-04Cleaning up vp9_dec_build_inter_predictors_sb function.Dmitry Kovalev
Using get_plane_block_size() instead of manipulation with subsampling values, calculating all required values only once without redundant calls to b_width_log2(). Change-Id: I00303f2a0926f9c4cb17f34591adda60615f8919
2013-11-25Use separate inter predictors for enc/decJingning Han
The decoder will construct inter predictor using lazy border extension, while the encoder, going with multiple runs of motion search in the rate- distortion optimization loop for each block, does border extension at frame level. This commit makes separate the inter predictors for encoder and decoder, respectively. Change-Id: Ieca2fecba3a7201a6d64ef9f219e5d91e50559c3
2013-11-25Merge "Separate setup_scale_factor/extend_frame_borders"Jingning Han
2013-11-22Separate setup_scale_factor/extend_frame_bordersJingning Han
This commit takes out vp9_extend_frame_borders from vp9_setup_scale_factors. The refactoring is for the preparation of the use of lazy border extension at decoder. This makes it necessary to handle border extension separately at encoder/decoder. The use of vp9_extend_frame_borders will be removed, when lazy border extension is ready. Change-Id: Ia3baba3d179d5f11eee1634f19b3b319d2a59186
2013-11-20Removing plane_block_{width, height} functions.Dmitry Kovalev
Change-Id: I29c0dfcf41a1253d5e2a0d2ff740c0c38ebaa5a2
2013-11-19Move vp9_setup_interp_filter() to encoderYaowu Xu
As it is used in encoder only. Change-Id: I5f2a8abbe72bb18cbf6ce36a3dc7e132aeae8ec2
2013-11-11Removing function pointers from inter prediction.Dmitry Kovalev
Removing foreach_predicted_block_visitor and calling build_inter_predictors directly. Change-Id: I11bb3c872b99b47c2680b01b0dbcc01c558c4a2b
2013-10-29Fix x_offset_q4/y_offset_q4 calculationYunqing Wang
"<< SUBPEL_BITS" needs to be added in the calculation. Call set_scaled_offsets() to calculate x_offset_q4 and y_offset_q4. Change-Id: Ied130ea771510e918f51cd1dc3abe57f4c0962b5
2013-10-24Merge changes I1868fb75,I9ff504c6Dmitry Kovalev
* changes: Renaming INTERPOLATIONFILTERTYPE to INTERPOLATION_TYPE. Adding VP9_FRAME_MARKER constant.
2013-10-23Renaming INTERPOLATIONFILTERTYPE to INTERPOLATION_TYPE.Dmitry Kovalev
Change-Id: I1868fb75ed88bfa65c1c2ca24677d65f2894d713
2013-10-22Improve scale_factors structYunqing Wang
The ref's scale_factors are set at frame level, and then copied for each partition block. Since the struct members are mostly constant, this patch separated the constant and non-constant members, and reduced struct copying. This gave 0.5% ~ 1.4% decoder speed gain. Change-Id: I94043bf5a6995c8042da52e5c661818dfa6f6d4c
generated by cgit v1.2.3 (git 2.25.1) at 2024-10-10 21:28:43 +0000