[FFmpeg-devel] [FFmpeg-cvslog] Fix compilation on ARM with android gcc 4.7

[Christophe Gisquet]

Hi,

(no longer registered on -cvslog, so I'm sending this here)

I noticed some time earlier an issue with this code. Indeed, it uses
int*, which could be fine. But it seems to me it should be int32_t: it
is used as input to int32_to_float_fmul_scalar at some point.

The corresponding branch is deep away in my todo list, but here is
what I did regarding this:
https://github.com/kurosu/libav/commit/5a57bfa1a3024193503be0788a615e5abbe5bca8
(not yet ported to ffmpeg)

I don't understand why it didn't cause any issue when sizeof(int)==8.

FYI,
--
Christophe