[Ffmpeg-devel] [PATCH] from DivX, Part 13: NULL callback calls

[Steve Lhomme]

Michael Niedermayer wrote:
> Hi
> 
> On Fri, Dec 16, 2005 at 04:03:58PM -1000, Steve Lhomme wrote:
> 
>>Some tests to avoid calling NULL callbacks.
> 
> 
>>Index: libavcodec/imgconvert.c
>>===================================================================
>>RCS file: /cvsroot/ffmpeg/ffmpeg/libavcodec/imgconvert.c,v
>>retrieving revision 1.58
>>diff -u -r1.58 imgconvert.c
>>--- libavcodec/imgconvert.c	11 Jul 2005 00:43:21 -0000	1.58
>>+++ libavcodec/imgconvert.c	17 Dec 2005 01:51:48 -0000
>>@@ -256,7 +256,7 @@
>>     int i; 
>>     
>>     for (i=0; i < PIX_FMT_NB; i++)
>>-         if (!strcmp(pix_fmt_info[i].name, name))
>>+         if (pix_fmt_info[i].name && !strcmp(pix_fmt_info[i].name, name))
> 
> 
> a pixel format with no name is a bug

A crash is a bigger bug IMO. You can't wait for a software to crash to 
solve possible problems. Maybe I just added the test just in case, I 
don't remember that one.

>> 	     break;
>>     return i;
>> }
>>Index: libavcodec/utils.c
>>===================================================================
>>RCS file: /cvsroot/ffmpeg/ffmpeg/libavcodec/utils.c,v
>>retrieving revision 1.162
>>diff -u -r1.162 utils.c
>>--- libavcodec/utils.c	2 Dec 2005 00:12:37 -0000	1.162
>>+++ libavcodec/utils.c	17 Dec 2005 01:58:36 -0000
>>@@ -962,7 +962,7 @@
>>         return -1;
>>     }
>> 
>>-    if (avctx->codec->close)
>>+    if (avctx->codec && avctx->codec->close)
>>         avctx->codec->close(avctx);
> 
> 
> how did this become NULL here?

IIRC that was for some raw DV samples. But I could be wrong.