[PATCH] D30551: [AMDGPU] Add builtin functions readlane ds_permute mov_dpp

Matt Arsenault via Phabricator via cfe-commits cfe-commits at lists.llvm.org
Thu Mar 2 14:54:17 PST 2017


arsenm added inline comments.


================
Comment at: lib/CodeGen/CGBuiltin.cpp:8391-8398
+  case AMDGPU::BI__builtin_amdgcn_ds_permute:
+    return emitBinaryBuiltin(*this, E, Intrinsic::amdgcn_ds_permute);
+  case AMDGPU::BI__builtin_amdgcn_ds_bpermute:
+    return emitBinaryBuiltin(*this, E, Intrinsic::amdgcn_ds_bpermute);
+  case AMDGPU::BI__builtin_amdgcn_readfirstlane:
+    return emitUnaryBuiltin(*this, E, Intrinsic::amdgcn_readfirstlane);
+  case AMDGPU::BI__builtin_amdgcn_readlane:
----------------
Since these don't have mangling, you could add GCCBuiltin to the intrinsic definition and then you wouldn't need these to be handled here


https://reviews.llvm.org/D30551





More information about the cfe-commits mailing list