[amdgpu] Add `-enable-ocl-mangling-mismatch-workaround`.

[hliao]

Summary

• Add -enable-ocl-mangling-mismatch-workaround to work around the mismatch on OCL name mangling so far.

[hliao]

This's the short-term solution until all issues exposed after the frontend change are fixed. This change will be reverted once the frontend change is landed.

[rampitec]

This comment was removed by rampitec.

[rampitec]

This comment was removed by rampitec.

[rampitec]

LGTM as a short term w/a.

[rampitec]

Oops, too soon. Please add "amdgpu" to the switch.

[yaxunl]

LGTM. Thanks!

[hliao]

Add 'amdgpu-' prefix in that option.

[rampitec]

Thanks!

llvm/lib/Target/AMDGPU/AMDGPULibFunc.cpp

 #include "AMDGPU.h"
 #include "llvm/ADT/StringExtras.h"
 #include "llvm/ADT/StringMap.h"
 #include "llvm/ADT/StringSwitch.h"
 #include "llvm/IR/DerivedTypes.h"
 #include "llvm/IR/Function.h"
 #include "llvm/IR/Module.h"
 #include "llvm/IR/ValueSymbolTable.h"
+#include "llvm/Support/CommandLine.h"
 #include "llvm/Support/raw_ostream.h"
 
 using namespace llvm;
 
+static cl::opt<bool> EnableOCLManglingMismatchWA(
+    "enable-ocl-mangling-mismatch-workaround", cl::init(true), cl::ReallyHidden,
+    cl::desc("Enable the workaround for OCL name mangling mismatch."));
+
 namespace {
 
 enum EManglingParam {
     E_NONE,
     EX_EVENT,
     EX_FLOAT4,
     EX_INTV4,
     EX_RESERVEDID,
void operator()(Stream& os, AMDGPULibFunc::Param p) {
     if (p.PtrKind) {
       if (trySubst(os, p)) return;
       os << 'P';
       if (p.PtrKind & AMDGPULibFunc::CONST) os << 'K';
       if (p.PtrKind & AMDGPULibFunc::VOLATILE) os << 'V';
       unsigned AS = UseAddrSpace
                         ? AMDGPULibFuncBase::getAddrSpaceFromEPtrKind(p.PtrKind)
                         : 0;
-      if (AS != 0) os << "U3AS" << AS;
+      if (EnableOCLManglingMismatchWA || AS != 0)
+        os << "U3AS" << AS;
       Ptr = p;
       p.PtrKind = 0;
     }
 
     if (p.VectorSize > 1) {
       if (trySubst(os, p)) goto exit;
       Str.push_back(p);
       os << "Dv" << static_cast<unsigned>(p.VectorSize) << '_';

llvm/test/CodeGen/AMDGPU/simplify-libcalls.ll

 ; RUN: opt -S -O1 -mtriple=amdgcn-- -amdgpu-simplify-libcall < %s | FileCheck -enable-var-scope -check-prefix=GCN -check-prefix=GCN-POSTLINK %s
-; RUN: opt -S -O1 -mtriple=amdgcn-- -amdgpu-simplify-libcall -amdgpu-prelink  <%s | FileCheck -enable-var-scope -check-prefix=GCN -check-prefix=GCN-PRELINK %s
+; RUN: opt -S -O1 -mtriple=amdgcn-- -amdgpu-simplify-libcall -amdgpu-prelink -enable-ocl-mangling-mismatch-workaround=0 <%s | FileCheck -enable-var-scope -check-prefix=GCN -check-prefix=GCN-PRELINK %s
 ; RUN: opt -S -O1 -mtriple=amdgcn-- -amdgpu-use-native -amdgpu-prelink < %s | FileCheck -enable-var-scope -check-prefix=GCN -check-prefix=GCN-NATIVE %s
 ; RUN: opt -S -passes='default<O1>' -mtriple=amdgcn-- -amdgpu-simplify-libcall < %s | FileCheck -enable-var-scope -check-prefix=GCN -check-prefix=GCN-POSTLINK %s
-; RUN: opt -S -passes='default<O1>' -mtriple=amdgcn-- -amdgpu-simplify-libcall -amdgpu-prelink  <%s | FileCheck -enable-var-scope -check-prefix=GCN -check-prefix=GCN-PRELINK %s
+; RUN: opt -S -passes='default<O1>' -mtriple=amdgcn-- -amdgpu-simplify-libcall -amdgpu-prelink -enable-ocl-mangling-mismatch-workaround=0 <%s | FileCheck -enable-var-scope -check-prefix=GCN -check-prefix=GCN-PRELINK %s
 ; RUN: opt -S -passes='default<O1>' -mtriple=amdgcn-- -amdgpu-use-native -amdgpu-prelink < %s | FileCheck -enable-var-scope -check-prefix=GCN -check-prefix=GCN-NATIVE %s
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_sincos
 ; GCN-POSTLINK: call fast float @_Z3sinf(
 ; GCN-POSTLINK: call fast float @_Z3cosf(
 ; GCN-PRELINK: call fast float @_Z6sincosfPf(
 ; GCN-NATIVE: call fast float @_Z10native_sinf(
 ; GCN-NATIVE: call fast float @_Z10native_cosf(