libyuv: Updates to gyp file convert and general

libyuv: Updates to gyp file convert and general

[fbarchard]

good changes... just style stuff

http://webrtc-codereview.appspot.com/239017/diff/1/libyuv.gyp
File libyuv.gyp (right):

http://webrtc-codereview.appspot.com/239017/diff/1/libyuv.gyp#newcode21
libyuv.gyp:21: '.',
.?  should that be source/ ?

http://webrtc-codereview.appspot.com/239017/diff/1/source/convert.cc
File source/convert.cc (left):

http://webrtc-codereview.appspot.com/239017/diff/1/source/convert.cc#oldcode45
source/convert.cc:45: uint8* out = dst_frame + dst_stride * src_height * 3 -
dst_stride * 3;
yikes!

http://webrtc-codereview.appspot.com/239017/diff/1/source/convert.cc
File source/convert.cc (right):

http://webrtc-codereview.appspot.com/239017/diff/1/source/convert.cc#newcode33
source/convert.cc:33: int
could you fix the style here?   int on same line as I420ToRGB24

http://webrtc-codereview.appspot.com/239017/diff/1/source/convert.cc#newcode118
source/convert.cc:118: int tmpR, tmpG, tmpB;
should be lower case or tmp_r

http://webrtc-codereview.appspot.com/239017/diff/1/source/convert.cc#newcode181
source/convert.cc:181: dst_frame == NULL)
should always have {}

http://webrtc-codereview.appspot.com/239017/diff/1/source/general.cc
File source/general.cc (right):

http://webrtc-codereview.appspot.com/239017/diff/1/source/general.cc#newcode19
source/general.cc:19: int
move int down one line

http://webrtc-codereview.appspot.com/239017/diff/1/source/general.cc#newcode32
source/general.cc:32: if (height < 0 || width < 0 ||  dst_ystride < 0 ||
dst_ustride < 0 ||
will want inversion support, but can wait

http://webrtc-codereview.appspot.com/239017/diff/1/source/general.cc#newcode55
source/general.cc:55: const int halfuvWidth = (width + 1) >> 1;
lowercase and/or _

http://webrtc-codereview.appspot.com/239017/diff/1/source/general.cc#newcode57
source/general.cc:57: for (wind = 0; wind < halfuvWidth; wind++) {
prefer ++wind

http://webrtc-codereview.appspot.com/239017/diff/1/source/general.cc#newcode280
source/general.cc:280: memset(dst_frame, 127, pad_width >> 2);
note that planar as I420Rect that is about 2x faster than memset

[mikhals]

Uploaded new patch following review. 

-Mikhal

http://webrtc-codereview.appspot.com/239017/diff/1/libyuv.gyp
File libyuv.gyp (right):

http://webrtc-codereview.appspot.com/239017/diff/1/libyuv.gyp#newcode21
libyuv.gyp:21: '.',
Actually on purpose, means this folder - for projects that link to it. 
On 2011/10/27 23:24:26, fbarchard wrote:
> .?  should that be source/ ?

http://webrtc-codereview.appspot.com/239017/diff/1/source/convert.cc
File source/convert.cc (right):

http://webrtc-codereview.appspot.com/239017/diff/1/source/convert.cc#newcode33
source/convert.cc:33: int
On 2011/10/27 23:24:26, fbarchard wrote:
> could you fix the style here?   int on same line as I420ToRGB24

Done.

http://webrtc-codereview.appspot.com/239017/diff/1/source/convert.cc#newcode118
source/convert.cc:118: int tmpR, tmpG, tmpB;
On 2011/10/27 23:24:26, fbarchard wrote:
> should be lower case or tmp_r

Done.

http://webrtc-codereview.appspot.com/239017/diff/1/source/convert.cc#newcode181
source/convert.cc:181: dst_frame == NULL)
Does not have to for one liners
On 2011/10/27 23:24:26, fbarchard wrote:
> should always have {}

http://webrtc-codereview.appspot.com/239017/diff/1/source/general.cc
File source/general.cc (right):

http://webrtc-codereview.appspot.com/239017/diff/1/source/general.cc#newcode32
source/general.cc:32: if (height < 0 || width < 0 ||  dst_ystride < 0 ||
dst_ustride < 0 ||
Later on. For now added to be clear about capabilities. 
On 2011/10/27 23:24:26, fbarchard wrote:
> will want inversion support, but can wait

http://webrtc-codereview.appspot.com/239017/diff/1/source/general.cc#newcode55
source/general.cc:55: const int halfuvWidth = (width + 1) >> 1;
On 2011/10/27 23:24:26, fbarchard wrote:
> lowercase and/or _

Done.

http://webrtc-codereview.appspot.com/239017/diff/1/source/general.cc#newcode57
source/general.cc:57: for (wind = 0; wind < halfuvWidth; wind++) {
On 2011/10/27 23:24:26, fbarchard wrote:
> prefer ++wind

Done.

http://webrtc-codereview.appspot.com/239017/diff/1/source/general.cc#newcode280
source/general.cc:280: memset(dst_frame, 127, pad_width >> 2);
I'm still hoping to get rid of this one, so let's see. 
On 2011/10/27 23:24:26, fbarchard wrote:
> note that planar as I420Rect that is about 2x faster than memset

[fbarchard]

LGTM