[FFmpeg-devel] [PATCH] avutil/rational: Document what is to be expected from av_d2q() of doubles representing rational numbers

[Michael Niedermayer]

Signed-off-by: Michael Niedermayer <michael at niedermayer.cc>
---
 libavutil/rational.h | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/libavutil/rational.h b/libavutil/rational.h
index 8cbfc8e0669..c6ac9fb1d91 100644
--- a/libavutil/rational.h
+++ b/libavutil/rational.h
@@ -168,6 +168,10 @@ static av_always_inline AVRational av_inv_q(AVRational q)
  * In case of infinity, the returned value is expressed as `{1, 0}` or
  * `{-1, 0}` depending on the sign.
  *
+ * In general rational numbers with |num| <= 1<<26 && |den| <= 1<<26
+ * can be recovered exactly from their double representation.
+ * (no exceptions where found within 100M random ones)
+ *
  * @param d   `double` to convert
  * @param max Maximum allowed numerator and denominator
  * @return `d` in AVRational form
-- 
2.17.1