pcm/FloatConvert: fix compile-time integer overflow for S32

The compile-time calculation for `factor` overflows because `1<<31` cannot be represented by `int`. By casting to `uintmax_t` first, we can avoid this overflow. Closes #380

pcm/FloatConvert: fix compile-time integer overflow for S32
cc5fab28 · Max Kellermann · a3f7127e · cc5fab28 · cc5fab28
Commit cc5fab28 authored Oct 29, 2018 by Max Kellermann
Show whitespace changes
Inline Side-by-side

Showing with 2 additions and 1 deletion

NEWS NEWS +1 -0

FloatConvert.hxx src/pcm/FloatConvert.hxx +1 -1

No files found.
--- a/NEWS
+++ b/NEWS
 ver 0.20.23 (not yet released)
 * protocol
  - emit "player" idle event when restarting the current song
+* fix broken float to s32 conversion
 * new clang crash bug workaround

 ver 0.20.22 (2018/10/23)

--- a/src/pcm/FloatConvert.hxx
+++ b/src/pcm/FloatConvert.hxx
@@ -34,7 +34,7 @@ struct FloatToIntegerSampleConvert {
 	typedef typename SrcTraits::long_type SL;
 	typedef typename DstTraits::value_type DV;

-	static constexpr SV factor = 1 << (DstTraits::BITS - 1);
+	static constexpr SV factor = uintmax_t(1) << (DstTraits::BITS - 1);
 	static_assert(factor > 0, "Wrong factor");

 	gcc_const