Use sdx4df to do 1-step refinement
Reset src buffer only once in vp9_int_pro_motion_estimation
Reduce the number of full block SAD calls
Prevent integer overflow in choose_partitioning
Remove unnecessary speed feature checking
Apply fast motion search to golden reference frame
Refactor to remove GLOBAL_MOTION
Merge "Separate rd_thresh adaption by ref_frame"
Separate rd_thresh adaption by ref_frame
Merge "Modify update golden reference update under aq-mode=3 mode."
Modify update golden reference update under aq-mode=3 mode.
Merge "Accumulate tx_totals counters in multi-threaded encoder"
Merge "Enable using Golden reference in choose_partition()"
Merge "Allow q adjustment for VPX_CQ and VPX_CBR."
Accumulate tx_totals counters in multi-threaded encoder
Fix vp9_compute_qdelta_by_rate loop behavior
Enable using Golden reference in choose_partition()
Allow q adjustment for VPX_CQ and VPX_CBR.
Refactor read_intra_frame_mode_info
Merge "Don't waste time partitioning skip superblocks."
Merge "Skip golden ref frame check when it is same as last ref frame"
Skip golden ref frame check when it is same as last ref frame
Don't waste time partitioning skip superblocks.
Merge "vp9_ethread: fix me consts initialization to support aq_mode=3 encoding"
Merge "Move pred_mv assign outside integral projection motion search"
Merge "Reduce default overshoot/undershoot params to 25"
Merge "Modify the setting of transform skip flags in non-rd mode"
vp9_ethread: fix me consts initialization to support aq_mode=3 encoding
Merge "Cherry pick libyuv VS fix"
Remove some unnecessary code in thread context copy.
Merge "Only wait for previous frame's motion vector if needed."
Merge "Set the threads/tile_columns in vpx_temporal_svc_encoder."
Set the threads/tile_columns in vpx_temporal_svc_encoder.
Modify the setting of transform skip flags in non-rd mode
Merge changes I1b972c94,I9c897d32
Only wait for previous frame's motion vector if needed.
Merge "Fix a tsan error bug in frame parallel decode."
Merge changes I219c287b,I6adee670
Merge "Fix misleading indentation."
Call encoder control before running ethread test.
Reduce default overshoot/undershoot params to 25
Merge "Use 8 threads by default for VP9 encodes"
Merge "VP9: turn on tile-columns and frame-parallel-mode by default"
Merge "Don't inline cost_coeffs."
Merge "Fix valgrind memcpy memory overlaps warning"
Fix misleading indentation.
Don't copy thread data for the main thread.
Move pred_mv assign outside integral projection motion search
Merge "Move integral projection motion search to vp9_mcomp.c"
Declare function used by 'once' with 'void' parameters
Merge "Fix variable shadowing."
Merge "Small rationalization of code in vp9_first_pass"
Merge "Remove unnecessary loopfilter init."
Merge "Make encoder buffer allocation dynamic"
Small rationalization of code in vp9_first_pass
Merge "Use SAD value to set chroma cost flag"
Move integral projection motion search to vp9_mcomp.c
Use SAD value to set chroma cost flag
Merge "Properly handle the boundary blocks for integral projection search"
Make encoder buffer allocation dynamic
Merge "dc quantizer fix for 32x32 transforms"
Merge "Scale the normalization factor depending on the block size"
Use 8 threads by default for VP9 encodes
Properly handle the boundary blocks for integral projection search
dc quantizer fix for 32x32 transforms
Merge "fix a race condition caused by intra function pointer initialization"
Merge "Adapt color sensitiviy threshold to luma signal energy"
Fix valgrind memcpy memory overlaps warning
Fix a tsan error bug in frame parallel decode.
Scale the normalization factor depending on the block size
VP9: turn on tile-columns and frame-parallel-mode by default
Adapt color sensitiviy threshold to luma signal energy
Remove unnecessary loopfilter init.
fix a race condition caused by intra function pointer initialization
Merge "libyuv: update to r1305"
ResizeCspTest: request PSNR packets
Use variance metric for integral projection vector match
Merge "Fix source frame border extension"
Fix source frame border extension
Merge "Fix calc_highbd_psnr"
Merge "Make SVC compatible with external resize."
Merge "Fix arithmetic overflow warnings."
Merge "use LC_ALL=C to sort libvpx_*srcs.txt"
Merge "Refactor integral projection based motion estimation"
Merge "Fix high bit-depth loop-filter sse2 compiling issue - part 4"
Merge "Fix high bit-depth loop-filter sse2 compiling issue - part 3"
Merge "Fix high bit-depth loop-filter sse2 compiling issue - part 2"
Merge "Fix high bit-depth loop-filter sse2 compiling issue - part 1"
test-data.mk: fix perf test data dependency
use LC_ALL=C to sort libvpx_*srcs.txt
tools_common.sh: use $$ in VPX_TEST_OUTPUT_DIR
Merge "tools_common.sh: add directory name to error output"