GAPI FLUID: SIMD for SubRC kernel#21231
Conversation
ae2b4df to
f46b0b1
Compare
|
@anna-khakimova any review here? |
| return saturate<DST>(x - y, roundf); | ||
| } | ||
|
|
||
| // there is no difference between sub and subr (due to documentation) |
There was a problem hiding this comment.
Could you please give me a link to this documents?
There was a problem hiding this comment.
There was a problem hiding this comment.
I haven't seen such statement in the mentioned documentation.
There was a problem hiding this comment.
I mean the order of arguments. It is the same as it in sub function
There was a problem hiding this comment.
The difference is obvious. SubC: SRC - C; SubRC: C - SRC.
It's necessary either use sub() instead of subr() and then remove subr() at all, or roll back these changes and leave everything as it was.
There was a problem hiding this comment.
I would prefer the second variant.
|
@alexgiving Could you please resolve the conflicts? |
|
@alexgiving could you please change the performance report. It's necessary to leave 2 numbers after the decimal point. |
f46b0b1 to
5b3c528
Compare
anna-khakimova
left a comment
There was a problem hiding this comment.
Looks good. Thank you!
|
@alalek could you merge pr please? |
…luid GAPI FLUID: SIMD for SubRC kernel * SIMD for SubRC * Reverse subrc
Performance report
Pull Request Readiness Checklist
See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request
Patch to opencv_extra has the same branch name.