================
@@ -1883,6 +1883,12 @@ let Log2LMUL = [-3, -2, -1, 0, 1, 2] in {
def vfncvt_rtz_x_f_w : RVVConvToNarrowingSignedBuiltin<"vfncvt_rtz_x">;
def vfncvt_rod_f_f_w : RVVConvBuiltin<"v", "vw", "xf", "vfncvt_rod_f">;
}
+
+// Zvfbfmin - Vector convert BF16 to FP32
+let Log2LMUL = [-2, -1, 0, 1, 2],
----------------
topperc wrote:
We can probably do something like what is done for Zvfh in
SemaRISCVVectorLookup.cpp. In practice, it probably doesn't matter match since
you can't declare a bfloat16 vector with Zvfhbfmin so you can't use the
intrinsic even if it was declared.
```
if (BaseType == BasicType::Float16) {
if ((Record.RequiredExtensions & RVV_REQ_Zvfhmin) == RVV_REQ_Zvfhmin) {
if (!TI.hasFeature("zvfhmin"))
continue;
} else if (!TI.hasFeature("zvfh")) {
continue;
}
}
```
https://github.com/llvm/llvm-project/pull/79618
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
