Option --parallel-jobs is only supported by amdclang++

Upstream bug: https://github.com/GPUOpen-LibrariesAndSDKs/HIPRT/issues/21
--- a/contrib/Orochi/scripts/kernelCompile.py
+++ b/contrib/Orochi/scripts/kernelCompile.py
@@ -17,7 +17,7 @@ def compile( index ):
 	if index == 0 :
 		command = [
 			"hipcc",
-			"-x", "hip", "..\ParallelPrimitives\RadixSortKernels.h", "-O3", "-std=c++17", "-ffast-math", "--cuda-device-only", "--genco", "-I../", "-include", "hip/hip_runtime.h", "-parallel-jobs=15"]
+			"-x", "hip", "..\ParallelPrimitives\RadixSortKernels.h", "-O3", "-std=c++17", "-ffast-math", "--cuda-device-only", "--genco", "-I../", "-include", "hip/hip_runtime.h"]
 		#command.append( "--offload-arch=gfx1100" )
 		for i in enumArch( "gfx900" ):
 			command.append( "--offload-arch=" + i )
--- a/scripts/bitcodes/compile.py
+++ b/scripts/bitcodes/compile.py
@@ -117,15 +117,15 @@ def compileAmd():
     parallel_jobs = 15
 
     dst = 'hiprt' + hiprt_ver + '_' + hip_version + '_amd_lib' + postfix
-    cmd = hipccpath + ' -x hip ../../hiprt/impl/hiprt_kernels_bitcode.h -O3 -std=c++17 ' + targets + ' -fgpu-rdc -c --gpu-bundle-output -c -emit-llvm -I../../contrib/Orochi/ -I../../ -DHIPRT_BITCODE_LINKING -ffast-math -parallel-jobs=' + str(parallel_jobs) + ' -o ' + dst
+    cmd = hipccpath + ' -x hip ../../hiprt/impl/hiprt_kernels_bitcode.h -O3 -std=c++17 ' + targets + ' -fgpu-rdc -c --gpu-bundle-output -c -emit-llvm -I../../contrib/Orochi/ -I../../ -DHIPRT_BITCODE_LINKING -ffast-math -o ' + dst
     compileScript(cmd, dst)
 
     dst = 'hiprt' + hiprt_ver + '_' + hip_version + '_amd.hipfb'
-    cmd = hipccpath + ' -x hip ../../hiprt/impl/hiprt_kernels.h -O3 -std=c++17 ' + targets + ' -mllvm -amdgpu-early-inline-all=false -mllvm -amdgpu-function-calls=true --genco -I../../ -DHIPRT_BITCODE_LINKING -ffast-math -parallel-jobs=' + str(parallel_jobs) + ' -o ' + dst
+    cmd = hipccpath + ' -x hip ../../hiprt/impl/hiprt_kernels.h -O3 -std=c++17 ' + targets + ' -mllvm -amdgpu-early-inline-all=false -mllvm -amdgpu-function-calls=true --genco -I../../ -DHIPRT_BITCODE_LINKING -ffast-math -o ' + dst
     compileScript(cmd, dst)
 
     dst = 'oro_compiled_kernels.hipfb'
-    cmd = hipccpath + ' -x hip ../../contrib/Orochi/ParallelPrimitives/RadixSortKernels.h -O3 -std=c++17 ' + targets + ' --genco -I../../contrib/Orochi/ -include hip/hip_runtime.h -DHIPRT_BITCODE_LINKING -ffast-math -parallel-jobs=' + str(parallel_jobs) + ' -o ' + dst
+    cmd = hipccpath + ' -x hip ../../contrib/Orochi/ParallelPrimitives/RadixSortKernels.h -O3 -std=c++17 ' + targets + ' --genco -I../../contrib/Orochi/ -include hip/hip_runtime.h -DHIPRT_BITCODE_LINKING -ffast-math -o ' + dst
     compileScript(cmd, dst)
 
 
--- a/scripts/bitcodes/precompile_bitcode.py
+++ b/scripts/bitcodes/precompile_bitcode.py
@@ -127,12 +127,12 @@ def compileAmd():
 
     # compile custom function table
     hiprt_custom_func = 'hiprt' + hiprt_ver + '_' + hip_version + '_custom_func_table.bc'
-    cmd = hipccpath + ' -O3 -std=c++17 ' + targets + ' -fgpu-rdc -c --gpu-bundle-output -c -emit-llvm -I../../ -ffast-math ../../test/bitcodes/custom_func_table.cpp -parallel-jobs=15 -o ' + hiprt_custom_func
+    cmd = hipccpath + ' -O3 -std=c++17 ' + targets + ' -fgpu-rdc -c --gpu-bundle-output -c -emit-llvm -I../../ -ffast-math ../../test/bitcodes/custom_func_table.cpp -o ' + hiprt_custom_func
     compileScript('compiling ', cmd, hiprt_custom_func)
 
     # compiling unit test
     hiprt_unit_test = 'hiprt' + hiprt_ver + '_' + hip_version + '_unit_test'+ postfix
-    cmd = hipccpath + ' -O3 -std=c++17 ' + targets + ' -fgpu-rdc -c --gpu-bundle-output -c -emit-llvm -I../../ -ffast-math -D BLOCK_SIZE=64 -D SHARED_STACK_SIZE=16 ../../test/bitcodes/unit_test.cpp -parallel-jobs=15 -o ' + hiprt_unit_test
+    cmd = hipccpath + ' -O3 -std=c++17 ' + targets + ' -fgpu-rdc -c --gpu-bundle-output -c -emit-llvm -I../../ -ffast-math -D BLOCK_SIZE=64 -D SHARED_STACK_SIZE=16 ../../test/bitcodes/unit_test.cpp -o ' + hiprt_unit_test
     compileScript('compiling ', cmd, hiprt_unit_test)
 
     # linking