[libc-commits] [libc] [llvm] [libc] Modify Exponent decomposition functions to work with emulated Float128 (PR #221540)

[via libc-commits]

================
@@ -610,11 +610,18 @@ TEST(LlvmLibcSharedMathTest, AllEmuFloat128) {
                                   Float128(0.0), Float128(0.0)));
   EXPECT_FP_EQ(Float128(0.0), LIBC_NAMESPACE::shared::fminimum_numf128(
                                   Float128(0.0), Float128(0.0)));
+  int frexpf128_exp = 0;
+  EXPECT_FP_EQ(Float128(0.0), LIBC_NAMESPACE::shared::frexpf128(
+                                  Float128(0.0), &frexpf128_exp));
+  EXPECT_EQ(0, frexpf128_exp);
----------------
Hari-Shankar-Karthik wrote:

No, no problems. I removed the test for it below (inside AllFloat128) and added it here in the same style as was present for `frexpf`. Should I instead copy over the old test here?

https://github.com/llvm/llvm-project/pull/221540