diff options
| author | Benjamin Kramer <benny.kra@googlemail.com> | 2011-03-30 17:02:54 +0000 |
|---|---|---|
| committer | Benjamin Kramer <benny.kra@googlemail.com> | 2011-03-30 17:02:54 +0000 |
| commit | 8398512f89623144c8bb45ce9deb4f74f76480fc (patch) | |
| tree | f36a0e103a6f63fa6537a9eda91358964e17d0d7 /unittests | |
| parent | 546739656ec9469499d3866d87dca6fdbcf2eee0 (diff) | |
| download | llvm-8398512f89623144c8bb45ce9deb4f74f76480fc.tar.gz llvm-8398512f89623144c8bb45ce9deb4f74f76480fc.tar.bz2 llvm-8398512f89623144c8bb45ce9deb4f74f76480fc.tar.xz | |
Avoid turning a floating point division with a constant power of two into a denormal multiplication.
Some platforms may treat denormals as zero, on other platforms multiplication
with a subnormal is slower than dividing by a normal.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@128555 91177308-0d34-0410-b5e6-96231b3b80d8
Diffstat (limited to 'unittests')
| -rw-r--r-- | unittests/ADT/APFloatTest.cpp | 6 |
1 files changed, 2 insertions, 4 deletions
diff --git a/unittests/ADT/APFloatTest.cpp b/unittests/ADT/APFloatTest.cpp index dea4a65b88..5f05b86169 100644 --- a/unittests/ADT/APFloatTest.cpp +++ b/unittests/ADT/APFloatTest.cpp @@ -589,10 +589,8 @@ TEST(APFloatTest, exactInverse) { EXPECT_TRUE(APFloat(1.17549435e-38f).getExactInverse(&inv)); EXPECT_TRUE(inv.bitwiseIsEqual(APFloat(8.5070592e+37f))); - // Large float - EXPECT_TRUE(APFloat(1.7014118e38f).getExactInverse(&inv)); - EXPECT_TRUE(inv.bitwiseIsEqual(APFloat(5.8774718e-39f))); - + // Large float, inverse is a denormal. + EXPECT_FALSE(APFloat(1.7014118e38f).getExactInverse(0)); // Zero EXPECT_FALSE(APFloat(0.0).getExactInverse(0)); // Denormalized float |