Merge "Removing unused calc_partial_ssl_err() function."
Merge "As you go mbmi->skip_coeff"
Merge "Removing intra_mode_stats."
Merge "Removing pass number check from ivf_write_file_header()."
Merge "Save ~30% off tokenize_b"
Merge "Cleaning up vp9_rdopt.c."
Fixing invalid block width/height calculation.
Merge "Making motion search code more compact."
Merge "Cleaning up md5 example."
Merge "Removing unnecessary casts + cleanup."
Merge changes I8eda5762,Ia2ffca07
Merge "Removing unused error_bins[] field from VP9_COMP."
Merge "Reusing get_frame_new_buffer() function."
Merge "Cleaning up vp9_encodeframe.c."
Merge "Adding get_vp9_ref_frame_buffer() function."
Merge "Converting int_mv to MV."
Merge "Minor fix on an assert"
Merge "Change the strategy for deciding the display size"
Merge "fix a div by zero issue"
Removing unused error_bins[] field from VP9_COMP.
Removing unused switchable_interp_count[] field from VP9_COMP.
Using clamp() function instead of the same raw code.
Making motion search code more compact.
Removing pass number check from ivf_write_file_header().
Reusing get_frame_new_buffer() function.
Removing unnecessary casts + cleanup.
Merge "Adding mv_has_subpel() function."
Change the strategy for deciding the display size
Merge "Adding raw_read_frame() function to vpxdec."
Removing intra_mode_stats.
Removing unused calc_partial_ssl_err() function.
Merge "test_libvpx: fix negative CPU filter"
Merge "Cleaning up and fixing psnr calculation code."
Adding mv_has_subpel() function.
Adding raw_read_frame() function to vpxdec.
Merge "Enable reference frame masking for rt mode"
Enable reference frame masking for rt mode
No arf right before real scene cut.
Further rate control tweaks and fixes.
Merge "mips dsp-ase r2 vp9 decoder fixed bug in extend frame module (seg fault)"
mips dsp-ase r2 vp9 decoder fixed bug in extend frame module (seg fault)
Merge "cosmetics: vp9_reconinter.h: make some variables const"
Merge "datarate_test: fix some type conversions"
test_libvpx: fix negative CPU filter
Merge "Apply neon flags to intrinsic files"
Merge "Cleaning up vp9_rc_postencode_update() function."
Cleaning up and fixing psnr calculation code.
Merge "Cleaning up vp9_dx_iface.c."
Merge "Declare setup_buffer_inter in vp9_rdopt.h"
Merge "Enable skipping reference frame check in rd loop"
Merge "Removing mi_height_log2_lookup table."
Declare setup_buffer_inter in vp9_rdopt.h
Cleaning up vp9_encodeframe.c.
Removing mi_height_log2_lookup table.
Apply neon flags to intrinsic files
Adding get_vp9_ref_frame_buffer() function.
Merge "Don't use gf_update by default for 1-pass CBR."
Cleaning up vp9_dx_iface.c.
Merge "Fix Issue #679: vp9 C loop filter produces valgrind warning."
Cleaning up vp9_rc_postencode_update() function.
Don't use gf_update by default for 1-pass CBR.
Merge "Adding {get, set}_rate_correction_factor() functions."
Fix Issue #679: vp9 C loop filter produces valgrind warning.
Merge "Keep buffer clipped to maximum in change_config."
Revert "SSSE3 convolution optimization"
Enable skipping reference frame check in rd loop
Merge "Optimze inv 16x16 DCT with 10 non-zero coeffs - P2"
Merge "Optimze inv 16x16 DCT with 10 non-zero coeffs - P1"
Merge "Cleanups on refresh flags"
Merge "Use the correct member for initialization"
Merge "Simplify set_rt_speed_feature()"
Keep buffer clipped to maximum in change_config.
Merge "Renaming 'Sharpness' to 'sharpness'."
Simplify set_rt_speed_feature()
Optimze inv 16x16 DCT with 10 non-zero coeffs - P2
Merge "SSSE3 convolution optimization"
SSSE3 convolution optimization
Merge "Removing examples code generation and making them static."
Merge "Using VP9_COMMON instead of VP9_COMP."
Merge "VP8 for ARMv8 by using NEON intrinsics 01"
VP8 for ARMv8 by using NEON intrinsics 01
Merge "Fix rate allocation bug."
Use the correct member for initialization
Using VP9_COMMON instead of VP9_COMP.
Merge "Removing direct references to {lst_fb, gld_fb, alt_fb}_idx fields."
Merge "Install test sources for MSVS"
Merge "Cleanups around cpi->common."
Adding {get, set}_rate_correction_factor() functions.
Merge "Renaming 'Mode' to 'mode'."
Optimze inv 16x16 DCT with 10 non-zero coeffs - P1