On Fri, 08 May 2015 13:45:36 +0100, Pekka Paalanen <[email protected]> wrote:
+pixman_image_t *
+fence_image_create_bits (pixman_format_code_t format,
+ int min_width, int height,
+ pixman_bool_t stride_fence)
Since you picked me up on subtleties of coding style, it's only fair that
I point out that in function definitions, the arguments should either all
be on one line, or one per line (with identifier names justified).
+ pixels = fence_malloc (stride * (unsigned)height);
I wonder if it wouldn't be better to reduce the size here by one page if
stride_fence==false - otherwise you've got a gap of one page between the
end of the last row of pixel data and the end fence.
+pixman_image_t *
+fence_image_create_bits (pixman_format_code_t format,
+ int min_width, int height,
+ pixman_bool_t stride_fence)
+{
+ return pixman_image_create_bits (format, width, height, NULL, 0);
+}
Perhaps a comment here to say that the auto-malloced pixel array is also
auto-freed when the image reference count drops to zero, so there's no
need for a destroy function in this case?
Ben
_______________________________________________
Pixman mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/pixman
