Define AVX blend intrinsics

This defines the `_mm256_blendv_pd` and `_mm256_blendv_ps` intrinsics.
The `_mm256_blend_pd` and `_mm256_blend_ps` intrinsics are not available
as LLVM intrinsics. In Clang they are implemented using the
shufflevector builtin.

Intel reference: https://software.intel.com/en-us/node/524070.

Author: ruuda

src/etc/platform-intrinsics/x86/avx.json

@@ -8,6 +8,13 @@
             "ret": "f(32-64)",
             "args": ["0", "0"]
         },
+        {
+            "intrinsic": "256_blendv_{0.data_type}",
+            "width": [256],
+            "llvm": "blendv.{0.data_type}.256",
+            "ret": "f(32-64)",
+            "args": ["0", "0", "0"]
+        },
         {
             "intrinsic": "256_broadcast_{0.data_type}",
             "width": [256],