[ Thread Index |
Date Index
| More lists.liballeg.org/allegro-developers Archives
]
Speaking of fogotten patches, I get a strange result when doing svn diff
in my 4.2 checkout.
The first diff seems to fix a real warning, so I'll commit it.
The last one seems to make sense somehow, no idea where it is from
though. Guess if I can verify that it is correct (looks like it is,
right?), I'll commit it as well.
The windows one likely must be from a thread on allegro.cc long ago, I'd
be inclined to simply commit it as well. Either it does fix a deadlock
as it claims, or else it shouldn't do harm :)
Index: src/i386/icsprite.c
===================================================================
--- src/i386/icsprite.c (revision 7648)
+++ src/i386/icsprite.c (working copy)
@@ -18,6 +18,7 @@
#include <stdio.h>
#include <string.h>
+#include <errno.h> /* For perror() */
#include "allegro.h"
#include "allegro/internal/aintern.h"
Index: src/win/wsystem.c
===================================================================
--- src/win/wsystem.c (revision 7648)
+++ src/win/wsystem.c (working copy)
@@ -215,6 +215,9 @@
_free_win_digi_driver_list();
_free_win_midi_driver_list();
+ /* make sure we are not paused and deadlock */
+ _win_reset_switch_mode();
+
/* unhook or close window */
exit_directx_window();
Index: src/c/cdefs8.h
===================================================================
--- src/c/cdefs8.h (revision 7648)
+++ src/c/cdefs8.h (working copy)
@@ -37,7 +37,11 @@
/* Blender for putpixel (DRAW_MODE_TRANS). */
#define PP_BLENDER unsigned char*
+#ifdef DEBUGMODE
+#define MAKE_PP_BLENDER(c) (!color_map && (al_assert(__FILE__, __LINE__), 0), (color_map->data[(c) & 0xFF]))
+#else
#define MAKE_PP_BLENDER(c) (color_map->data[(c) & 0xFF])
+#endif
#define PP_BLEND(b,o,n) ((b)[(o) & 0xFF])
/* Blender for draw_trans_*_sprite. */
--
Elias Pschernig