//
// Generated by NVIDIA NVVM Compiler
//
// Compiler Build ID: CL-19122697
// Cuda compilation tools, release 7.0, V7.0.17
// Based on LLVM 3.4svn
//

.version 4.2
.target sm_53
.address_size 32

	// .weak	cudaMalloc

.weak .func  (.param .b32 func_retval0) cudaMalloc(
	.param .b32 cudaMalloc_param_0,
	.param .b32 cudaMalloc_param_1
)
{
	.reg .s32 	%r<2>;


	mov.u32 	%r1, 30;
	st.param.b32	[func_retval0+0], %r1;
	ret;
}

	// .weak	cudaFuncGetAttributes
.weak .func  (.param .b32 func_retval0) cudaFuncGetAttributes(
	.param .b32 cudaFuncGetAttributes_param_0,
	.param .b32 cudaFuncGetAttributes_param_1
)
{
	.reg .s32 	%r<2>;


	mov.u32 	%r1, 30;
	st.param.b32	[func_retval0+0], %r1;
	ret;
}

	// .weak	cudaDeviceGetAttribute
.weak .func  (.param .b32 func_retval0) cudaDeviceGetAttribute(
	.param .b32 cudaDeviceGetAttribute_param_0,
	.param .b32 cudaDeviceGetAttribute_param_1,
	.param .b32 cudaDeviceGetAttribute_param_2
)
{
	.reg .s32 	%r<2>;


	mov.u32 	%r1, 30;
	st.param.b32	[func_retval0+0], %r1;
	ret;
}

	// .weak	cudaGetDevice
.weak .func  (.param .b32 func_retval0) cudaGetDevice(
	.param .b32 cudaGetDevice_param_0
)
{
	.reg .s32 	%r<2>;


	mov.u32 	%r1, 30;
	st.param.b32	[func_retval0+0], %r1;
	ret;
}

	// .weak	cudaOccupancyMaxActiveBlocksPerMultiprocessor
.weak .func  (.param .b32 func_retval0) cudaOccupancyMaxActiveBlocksPerMultiprocessor(
	.param .b32 cudaOccupancyMaxActiveBlocksPerMultiprocessor_param_0,
	.param .b32 cudaOccupancyMaxActiveBlocksPerMultiprocessor_param_1,
	.param .b32 cudaOccupancyMaxActiveBlocksPerMultiprocessor_param_2,
	.param .b32 cudaOccupancyMaxActiveBlocksPerMultiprocessor_param_3
)
{
	.reg .s32 	%r<2>;


	mov.u32 	%r1, 30;
	st.param.b32	[func_retval0+0], %r1;
	ret;
}

	// .weak	cudaOccupancyMaxActiveBlocksPerMultiprocessorWithFlags
.weak .func  (.param .b32 func_retval0) cudaOccupancyMaxActiveBlocksPerMultiprocessorWithFlags(
	.param .b32 cudaOccupancyMaxActiveBlocksPerMultiprocessorWithFlags_param_0,
	.param .b32 cudaOccupancyMaxActiveBlocksPerMultiprocessorWithFlags_param_1,
	.param .b32 cudaOccupancyMaxActiveBlocksPerMultiprocessorWithFlags_param_2,
	.param .b32 cudaOccupancyMaxActiveBlocksPerMultiprocessorWithFlags_param_3,
	.param .b32 cudaOccupancyMaxActiveBlocksPerMultiprocessorWithFlags_param_4
)
{
	.reg .s32 	%r<2>;


	mov.u32 	%r1, 30;
	st.param.b32	[func_retval0+0], %r1;
	ret;
}

	// .globl	KernelBandwidth
.visible .entry KernelBandwidth(
	.param .u32 KernelBandwidth_param_0,
	.param .u64 KernelBandwidth_param_1
)
{
	.reg .pred 	%p<2>;
	.reg .f32 	%f<2>;
	.reg .s32 	%r<18>;
	.reg .s64 	%rd<22>;


	ld.param.u32 	%r2, [KernelBandwidth_param_0];
	ld.param.u64 	%rd1, [KernelBandwidth_param_1];
	cvta.to.global.u32 	%r1, %r2;
	mov.u32 	%r3, %tid.x;
	cvt.u64.u32	%rd2, %r3;
	mov.u32 	%r4, %ntid.x;
	mov.u32 	%r5, %tid.y;
	mul.wide.u32 	%rd3, %r5, %r4;
	mov.u32 	%r6, %ntid.y;
	mul.wide.u32 	%rd4, %r6, %r4;
	mov.u32 	%r7, %tid.z;
	cvt.u64.u32	%rd5, %r7;
	mul.lo.s64 	%rd6, %rd5, %rd4;
	mov.u32 	%r8, %ntid.z;
	cvt.u64.u32	%rd7, %r8;
	mul.lo.s64 	%rd8, %rd7, %rd4;
	mov.u32 	%r9, %ctaid.x;
	cvt.u64.u32	%rd9, %r9;
	mul.lo.s64 	%rd10, %rd9, %rd8;
	mov.u32 	%r10, %nctaid.x;
	cvt.u64.u32	%rd11, %r10;
	mul.lo.s64 	%rd12, %rd11, %rd8;
	mov.u32 	%r11, %ctaid.y;
	cvt.u64.u32	%rd13, %r11;
	mov.u32 	%r12, %nctaid.y;
	mov.u32 	%r13, %ctaid.z;
	mul.wide.u32 	%rd14, %r13, %r12;
	add.s64 	%rd15, %rd14, %rd13;
	mul.lo.s64 	%rd16, %rd12, %rd15;
	add.s64 	%rd17, %rd3, %rd2;
	add.s64 	%rd18, %rd17, %rd6;
	add.s64 	%rd19, %rd18, %rd10;
	add.s64 	%rd20, %rd19, %rd16;
	rem.s64 	%rd21, %rd20, %rd1;
	cvt.u32.u64	%r14, %rd21;
	shl.b32 	%r15, %r14, 4;
	add.s32 	%r16, %r1, %r15;
	ld.global.f32 	%f1, [%r16];
	setp.neu.f32	%p1, %f1, 0f4640E400;
	@%p1 bra 	BB6_2;

	mov.u32 	%r17, 0;
	st.global.u32 	[%r1], %r17;

BB6_2:
	ret;
}