On 31/07/16 21:59, Martin Storsjö wrote: > If the value is a 32 bit value (e.g. a constant like 0), the larger > shifts are undefined, and e.g. clang end up only writing 32 bits > for an AV_WN64(ptr, 0); > > This fixes the VP8 decoder in build configurations with clang, with > fast unaligned support and the gcc version of AV_WN macro disabled. > --- > libavutil/intreadwrite.h | 36 ++++++++++++++++++------------------ > 1 file changed, 18 insertions(+), 18 deletions(-)
Sounds Ok. _______________________________________________ libav-devel mailing list [email protected] https://lists.libav.org/mailman/listinfo/libav-devel
