[cairo] Re: about glitz yv12 support

[ma liang]

Hello, David:
    Follow the last email I sent. My video frame displayed to my cairo
surface divided into 3 part, one biggest in the bottom, and the other
two in the top left and top right respectively(maybe the Y, V, and U
data displayed separate).
    I am using xine-lib's video output, it seems that it's in the
TOP_DOWN format, but, I haven't flip it yet while I am using
GLITZ_PIXEL_SCANLINE_ORDER_BOTTOM_UP pixel_format(the frame seems
displayed in the normal pixel trun), I am using cario surface to
display them, how could I flip it in cairo surface if I need?

BTW:
1. When I use glitz_set_pixels() with a buffer
glitz_buffer_create_for_data() returned, must the  bytes_per_line's
value be "(width +3) & -4" NOT "(((width * pf.masks.bpp) / 8) +3) &
-4"? This seems not nature on the point of a glitz user.

2. on the glitz_pixel.c's line 1226 of current glitz cvs code, May be
the yoffset parameter of tex_sub_image_2d() should change from
"texture->box.y2 - box.y2" to "texture->height - box.y2"?