Description: Work around a gcc-4.5 behaviour change/bug(?). Also, this change is actually correct, as the arithmethics can easily overflow a 16 bit value.
Author: Martin Pitt <martin.pitt@ubuntu.com>
Bug: https://bugs.freedesktop.org//show_bug.cgi?id=33129
Bug-Ubuntu: https://launchpad.net/bugs/685352

Index: b/src/libply-splash-core/ply-pixel-buffer.c
===================================================================
--- a/src/libply-splash-core/ply-pixel-buffer.c
+++ b/src/libply-splash-core/ply-pixel-buffer.c
@@ -65,7 +65,7 @@
     {
       uint8_t alpha_1, red_1, green_1, blue_1;
       uint8_t red_2, green_2, blue_2;
-      uint_least16_t red, green, blue;
+      uint32_t red, green, blue;
 
       alpha_1 = (uint8_t) (pixel_value_1 >> 24);
       red_1 = (uint8_t) (pixel_value_1 >> 16);