summaryrefslogtreecommitdiff
path: root/Build/source/libs/mpfr/mpfr-src/src/get_flt.c
diff options
context:
space:
mode:
Diffstat (limited to 'Build/source/libs/mpfr/mpfr-src/src/get_flt.c')
-rw-r--r--Build/source/libs/mpfr/mpfr-src/src/get_flt.c5
1 files changed, 4 insertions, 1 deletions
diff --git a/Build/source/libs/mpfr/mpfr-src/src/get_flt.c b/Build/source/libs/mpfr/mpfr-src/src/get_flt.c
index 168c58702db..c1ba68e4baf 100644
--- a/Build/source/libs/mpfr/mpfr-src/src/get_flt.c
+++ b/Build/source/libs/mpfr/mpfr-src/src/get_flt.c
@@ -50,7 +50,10 @@ mpfr_get_flt (mpfr_srcptr src, mpfr_rnd_t rnd_mode)
rnd_mode = negative ? MPFR_RNDD : MPFR_RNDU;
/* FIXME: The code below assumes the IEEE-754 binary32 format
- with subnormal support. */
+ with subnormal support. Fix it by converting to double, then
+ to float, and in case of binary radix (for which we want
+ correct rounding), handle double-rounding issues somewhere
+ in the code? */
/* the smallest positive normal float number is 2^(-126) = 0.5*2^(-125),
and the smallest positive subnormal number is 2^(-149) = 0.5*2^(-148) */