[libc-commits] [PATCH] D82591: [libc] Add the remaining long double flavors of nearest integer functions.

Thu Jun 25 22:53:51 PDT 2020

sivachandra added a comment.

In D82591#2115239 <https://reviews.llvm.org/D82591#2115239>, @tschuett wrote:

> Did you miss the aarch64 entry points?

Yes and no :)
The most convenient way for me to verify the patch on aarch64 is to upload the patch and then use `arc patch` to apply it on the aarch64 machine. So, I did not add them to the aarch64 config in the first round as I had not tested them on aarch64. But, added now as I have now tested on aarch64 as well.

================
Comment at: libc/test/src/math/ceill_test.cpp:23
+TEST(CeillTest, SpecialNumbers) {
+  ASSERT_TRUE(FPBits::zero() == __llvm_libc::ceill(FPBits::zero()));
+  ASSERT_TRUE(FPBits::negZero() == __llvm_libc::ceill(FPBits::negZero()));
----------------
asteinhauser wrote:
> ASSERT_EQ(a, b) here and in the all following ASSERT_TRUE(a == b).
Our unittest framework supports only integer types to me checked with `ASSERT_EQ` and `ASSERT_NE`. But, I took your suggestion and converted wherever I could.

Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D82591/new/

https://reviews.llvm.org/D82591