[FFmpeg-devel] [PATCH 2/5] avcodec/allcodecs: Dont play with NULLs

Andreas Rheinhardt andreas.rheinhardt at outlook.com
Mon Mar 21 22:55:22 EET 2022


James Almer:
> 
> 
> On 3/21/2022 5:19 PM, Michael Niedermayer wrote:
>> Fixes: member access within null pointer of type 'const FFCodec' (aka
>> 'const struct FFCodec')
>> Fixes:
>> 45726/clusterfuzz-testcase-minimized-ffmpeg_DEMUXER_fuzzer-6554445419249664
>>
>>
>> Found-by: continuous fuzzing process
>> https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
>> Signed-off-by: Michael Niedermayer <michael at niedermayer.cc>
>> ---
>>   libavcodec/allcodecs.c | 7 ++++---
>>   1 file changed, 4 insertions(+), 3 deletions(-)
>>
>> diff --git a/libavcodec/allcodecs.c b/libavcodec/allcodecs.c
>> index b1aa7e266b..22d56760ec 100644
>> --- a/libavcodec/allcodecs.c
>> +++ b/libavcodec/allcodecs.c
>> @@ -882,10 +882,11 @@ const AVCodec *av_codec_iterate(void **opaque)
>>         ff_thread_once(&av_codec_static_init, av_codec_init_static);
>>   -    if (c)
>> +    if (c) {
>>           *opaque = (void*)(i + 1);
>> -
>> -    return &c->p;
>> +        return &c->p;
>> +    }
>> +    return NULL;
> 
> Can't you just do
> 
> return (const AVCodec *)c;
> 
> Or is that aliasing a problem?
> 

There is no aliasing problem: It is perfectly legal to cast a pointer to
a struct to a pointer to its first member. But it unnecessarily
circumvents the type system, so this patch here is better.

- Andreas


More information about the ffmpeg-devel mailing list