[PATCH] D61319: [PR41674] [OpenCL] Fix initialisation of this via pointer

Anastasia Stulova via Phabricator via cfe-commits cfe-commits at lists.llvm.org
Thu May 2 03:27:14 PDT 2019


Anastasia added inline comments.


================
Comment at: test/CodeGenOpenCLCXX/addrspace-of-this.cl:94
 // Test the address space of 'this' when invoking a method that is declared in the file contex.
-// COMMON: %call1 = call i32 @_ZNU3AS41C7outsideEv(%class.C addrspace(4)* addrspacecast (%class.C addrspace(1)* @c to %class.C addrspace(4)*))
+// COMMON: %call2 = call i32 @_ZNU3AS41C7outsideEv(%class.C addrspace(4)* addrspacecast (%class.C addrspace(1)* @c to %class.C addrspace(4)*))
 
----------------
Ok I think we should regex name of call registers or may be even prevent checking it if it's not required. The same applies to other places btw.


================
Comment at: test/CodeGenOpenCLCXX/addrspace-of-this.cl:136
 // COMMON: [[CALL:%call[0-9]+]] = call spir_func dereferenceable(4) %class.C addrspace(4)* @_Z3foov()
-// EXPL: call void @_ZNU3AS41CC1EOU3AS4S_(%class.C addrspace(4)* [[C5GEN:%[0-9]+]], %class.C addrspace(4)* dereferenceable(4) %call4)
+// EXPL: call void @_ZNU3AS41CC1EOU3AS4S_(%class.C addrspace(4)* [[C5GEN:%[0-9]+]], %class.C addrspace(4)* dereferenceable(4) %call5)
 // IMPL: [[C5VOID:%[0-9]+]] = bitcast %class.C* %c5 to i8*
----------------
Shouldn't `%call5` just be `[[CALL]]`? If not I see little value in checking the name...


Repository:
  rC Clang

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D61319/new/

https://reviews.llvm.org/D61319





More information about the cfe-commits mailing list