On 9/18/2015 6:35 PM, Ganesh Ajjanagadde wrote: > On Fri, Sep 18, 2015 at 5:30 PM, James Almer <jamr...@gmail.com> wrote: >> On 9/17/2015 9:46 AM, Ganesh Ajjanagadde wrote: >>> Commit 7404f3bdb switched bitrate to 64 bits. >>> This triggers -Wabsolute-value on clang, e.g >>> http://fate.ffmpeg.org/log.cgi?time=20150917122742&log=compile&slot=x86_64-darwin-clang-3.7-O3. >>> Therefore, usage of abs is changed to llabs, which is available on all of >>> the platforms. >>> Have not tested whether LLONG_MAX (C99 feature) is available on Microsoft >>> or not. >> >> Use int64_t and INT64_MAX instead. > > I can definitely do that, but am curious about this. The codebase > already uses long long (see e.g ffprobe), and morally llabs returns > long long; hence I use it. As for the INT64_MAX, I do not have a > current example to know whether it is safe on all platforms, but > again, the "clean" solution is LLONG_MAX due to the type of llabs.
grepping INT64_MAX shows a lot of hits in the codebase, so it's safe to use. Also, int64_t is used in hundreds of files whereas long long is used in about seven. C guarantees that long long is at least 64bits so int64_t should be ok for this patch, but in any case wait for someone else to comment before sending a new patch. _______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
