idx_mask computation inside the precision loop

[jianle973]

Move ref_mv_idx_to_search() from the pre-loop into the real BAWP rd search loop. The 33-frame test results are (on top of commit 63e3858, sep-30) as follows, average results of A1-5 and B1:

     psnr-y psnr-u psnr-v psnr-yuv enc-time dec-time

speed 1: 0.00% -0.00% +0.01% +0.00% 99.52% 99.78%
speed 2: -0.00% +0.00% -0.00% -0.00% 99.16% 98.89%

[yeqing-wu]

Move ref_mv_idx_to_search() from the pre-loop into the real BAWP rd search loop. The 33-frame $A test results are (on top of commit 63e3858, sep-30) as follows, average results of A1-5 and B1:

     psnr-y psnr-u psnr-v psnr-yuv enc-time dec-time

speed 1: 0.00% -0.00% +0.01% +0.00% 99.52% 99.78% speed 2: -0.00% +0.00% -0.00% -0.00% 99.16% 98.89%

I think we have some format issues for the commit body message. what is $A?

[jianle973]

Move ref_mv_idx_to_search() from the pre-loop into the real BAWP rd search loop. The 33-frame $A test results are (on top of commit 63e3858, sep-30) as follows, average results of A1-5 and B1:

     psnr-y psnr-u psnr-v psnr-yuv enc-time dec-time

speed 1: 0.00% -0.00% +0.01% +0.00% 99.52% 99.78% speed 2: -0.00% +0.00% -0.00% -0.00% 99.16% 98.89%

I think we have some format issues for the commit body message. what is $A?

It's a typo. Fixed.

[yeqing-wu]

LGTM, some minor comments. Thanks

[yeqing-wu]

LGTM, thanks