Jason Garrett-Glaser <[email protected]> writes:

>> -        if (i > 63) {
>> -            av_log(ctx->avctx, AV_LOG_ERROR, "ac tex damaged %d, %d\n", n, 
>> i);
>> -            break;
>> -        }
>> +        k = ++i & 63;
>
> Why is this faster?  At best it seems it should be the same; can the
> other restructuring not be done without it for some reason?

I've sometimes noticed gcc disliking loops with more than one entry or
exit point.  This particular change seems to make little difference in
this case though.

-- 
Måns Rullgård
[email protected]
_______________________________________________
libav-devel mailing list
[email protected]
https://lists.libav.org/mailman/listinfo/libav-devel

Reply via email to