diff options
author | Szabolcs Nagy <nsz@port70.net> | 2016-10-04 03:58:56 +0200 |
---|---|---|
committer | Rich Felker <dalias@aerifal.cx> | 2016-10-20 01:32:27 -0400 |
commit | 688d3da0f1730daddbc954bbc2d27cc96ceee04c (patch) | |
tree | 3f769f22d7cb42a40eb618535bc31ae3c0fd5d90 /src/math/pow.c | |
parent | 87643335ac699b14527bbf807956cf8911b86bfa (diff) | |
download | musl-688d3da0f1730daddbc954bbc2d27cc96ceee04c.tar.gz musl-688d3da0f1730daddbc954bbc2d27cc96ceee04c.tar.bz2 musl-688d3da0f1730daddbc954bbc2d27cc96ceee04c.tar.xz musl-688d3da0f1730daddbc954bbc2d27cc96ceee04c.zip |
math: fix pow signed shift ub
j is int32_t and thus j<<31 is undefined if j==1, so j is changed to
uint32_t locally as a quick fix, the generated code is not affected.
(this is a strict conformance fix, future c standard may allow 1<<31,
see DR 463. the bug was inherited from freebsd fdlibm, the proper fix
is to use uint32_t for all bit hacks, but that requires more intrusive
changes.)
reported by Daniel Sabogal
Diffstat (limited to 'src/math/pow.c')
-rw-r--r-- | src/math/pow.c | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/src/math/pow.c b/src/math/pow.c index b66f632d..3ddc1b6f 100644 --- a/src/math/pow.c +++ b/src/math/pow.c @@ -125,11 +125,11 @@ double pow(double x, double y) else if (iy >= 0x3ff00000) { k = (iy>>20) - 0x3ff; /* exponent */ if (k > 20) { - j = ly>>(52-k); + uint32_t j = ly>>(52-k); if ((j<<(52-k)) == ly) yisint = 2 - (j&1); } else if (ly == 0) { - j = iy>>(20-k); + uint32_t j = iy>>(20-k); if ((j<<(20-k)) == iy) yisint = 2 - (j&1); } |