On Jan 14, 2013, at 17:49, Nicolas George wrote:
>> But Microsoft compiler only recognizes
>> _alloca().
>
> There you are.
#ifdef _MSC_VER
# define alloca(n) _alloca(n)
#endif
>
> There is another problem with alloca(): the scope of the allocation is
> per function, while the scope of a compound literal, as used by the macro
> currently, is per block.
Are you sure? I was under the impression that
n = 10;
{ char *foo = (char*) alloca(n); // char foo[n] in gcc :)
// fill foo
printf( "%s\n", foo );
}
is equivalent to
{ char foo[10];
// fill foo
printf( "%s\n", foo );
}
meaning that the stack-deallocation of *foo happens upon exit from the block,
like the releasing of foo[10] ...
> with the current macro, works as expected, whereas if the macro were based
> on alloca(), the error strings would stay allocated until the loop is
> completed. That is a memory leak. Probably not terrible, but still a leak.
In my book, a memory leak is memory that's allocated but never deallocated.
That's not the case here... it's more like a caching/garbage-collection scheme.
And as long as there are few errors being reported, there won't be much memory
to free.
R.
_______________________________________________
Libav-user mailing list
[email protected]
http://ffmpeg.org/mailman/listinfo/libav-user
