Homec4science
Diffusion lammps (efficient_neuronet) Browse

lammps/lib/gpu/lj_class2_long.ptx
6b1774dc2052efficient_neuronet

Last Change
Raw File

lj_class2_long.ptx
View Options

.version 2.3
.target sm_20
.address_size 64
// compiled with /usr/local/cuda/open64/lib//be
// nvopencc 4.0 built on 2011-05-12
//-----------------------------------------------------------
// Compiling /tmp/tmpxft_00009d9c_00000000-9_lal_lj_class2_long.cpp3.i (/home/sjplimp/ccBI#.JI7tD2)
//-----------------------------------------------------------
//-----------------------------------------------------------
// Options:
//-----------------------------------------------------------
// Target:ptx, ISA:sm_20, Endian:little, Pointer Size:64
// -O3 (Optimization level)
// -g0 (Debug level)
// -m2 (Report advisories)
//-----------------------------------------------------------
.file 1 "<command-line>"
.file 2 "/tmp/tmpxft_00009d9c_00000000-8_lal_lj_class2_long.cudafe2.gpu"
.file 3 "/usr/lib/gcc/x86_64-redhat-linux/4.4.5/include/stddef.h"
.file 4 "/usr/local/cuda/include/crt/device_runtime.h"
.file 5 "/usr/local/cuda/include/host_defines.h"
.file 6 "/usr/local/cuda/include/builtin_types.h"
.file 7 "/usr/local/cuda/include/device_types.h"
.file 8 "/usr/local/cuda/include/driver_types.h"
.file 9 "/usr/local/cuda/include/surface_types.h"
.file 10 "/usr/local/cuda/include/texture_types.h"
.file 11 "/usr/local/cuda/include/vector_types.h"
.file 12 "/usr/local/cuda/include/device_launch_parameters.h"
.file 13 "/usr/local/cuda/include/crt/storage_class.h"
.file 14 "/usr/include/bits/types.h"
.file 15 "/usr/include/time.h"
.file 16 "lal_lj_class2_long.cu"
.file 17 "/usr/local/cuda/include/common_functions.h"
.file 18 "/usr/local/cuda/include/math_functions.h"
.file 19 "/usr/local/cuda/include/math_constants.h"
.file 20 "/usr/local/cuda/include/device_functions.h"
.file 21 "/usr/local/cuda/include/sm_11_atomic_functions.h"
.file 22 "/usr/local/cuda/include/sm_12_atomic_functions.h"
.file 23 "/usr/local/cuda/include/sm_13_double_functions.h"
.file 24 "/usr/local/cuda/include/sm_20_atomic_functions.h"
.file 25 "/usr/local/cuda/include/sm_20_intrinsics.h"
.file 26 "/usr/local/cuda/include/surface_functions.h"
.file 27 "/usr/local/cuda/include/texture_fetch_functions.h"
.file 28 "/usr/local/cuda/include/math_functions_dbl_ptx3.h"
.global .texref pos_tex;
.global .texref q_tex;
.entry kernel_pair (
.param .u64 __cudaparm_kernel_pair_x_,
.param .u64 __cudaparm_kernel_pair_lj1,
.param .u64 __cudaparm_kernel_pair_lj3,
.param .s32 __cudaparm_kernel_pair_lj_types,
.param .u64 __cudaparm_kernel_pair_sp_lj_in,
.param .u64 __cudaparm_kernel_pair_dev_nbor,
.param .u64 __cudaparm_kernel_pair_dev_packed,
.param .u64 __cudaparm_kernel_pair_ans,
.param .u64 __cudaparm_kernel_pair___val_paramengv,
.param .s32 __cudaparm_kernel_pair_eflag,
.param .s32 __cudaparm_kernel_pair_vflag,
.param .s32 __cudaparm_kernel_pair_inum,
.param .s32 __cudaparm_kernel_pair_nbor_pitch,
.param .u64 __cudaparm_kernel_pair_q_,
.param .f32 __cudaparm_kernel_pair_cut_coulsq,
.param .f32 __cudaparm_kernel_pair_qqrd2e,
.param .f32 __cudaparm_kernel_pair_g_ewald,
.param .s32 __cudaparm_kernel_pair_t_per_atom)
{
.reg .u32 %r<86>;
.reg .u64 %rd<64>;
.reg .f32 %f<167>;
.reg .pred %p<21>;
.shared .align 16 .b8 __cuda___cuda_local_var_32541_33_non_const_sp_lj112[32];
.shared .align 4 .b8 __cuda___cuda_local_var_32635_55_non_const_red_acc144[3072];
// __cuda_local_var_32553_10_non_const_f = 64
// __cuda_local_var_32555_9_non_const_virial = 16
// __cuda_local_var_32589_43_non_const_r6inv = 44
// __cuda_local_var_32589_50_non_const_r3inv = 40
// __cuda_local_var_32589_57_non_const_prefactor = 52
// __cuda_local_var_32589_68_non_const__erfc = 48
.loc 16 36 0
$LDWbegin_kernel_pair:
.loc 16 41 0
ld.param.u64 %rd1, [__cudaparm_kernel_pair_sp_lj_in];
ldu.global.f32 %f1, [%rd1+0];
.loc 16 42 0
ld.global.f32 %f2, [%rd1+4];
.loc 16 43 0
ld.global.f32 %f3, [%rd1+8];
.loc 16 44 0
ld.global.f32 %f4, [%rd1+12];
st.shared.v4.f32 [__cuda___cuda_local_var_32541_33_non_const_sp_lj112+0], {%f1,%f2,%f3,%f4};
.loc 16 45 0
ld.global.f32 %f5, [%rd1+16];
.loc 16 46 0
ld.global.f32 %f6, [%rd1+20];
.loc 16 47 0
ld.global.f32 %f7, [%rd1+24];
.loc 16 48 0
ld.global.f32 %f8, [%rd1+28];
st.shared.v4.f32 [__cuda___cuda_local_var_32541_33_non_const_sp_lj112+16], {%f5,%f6,%f7,%f8};
.loc 16 56 0
mov.f32 %f9, 0f00000000; // 0
mov.f32 %f10, %f9;
mov.f32 %f11, 0f00000000; // 0
mov.f32 %f12, %f11;
mov.f32 %f13, 0f00000000; // 0
mov.f32 %f14, %f13;
mov.f32 %f15, 0f00000000; // 0
mov.f32 %f16, %f15;
mov.f32 %f17, 0f00000000; // 0
mov.f32 %f18, %f17;
mov.f32 %f19, 0f00000000; // 0
mov.f32 %f20, %f19;
ld.param.s32 %r1, [__cudaparm_kernel_pair_t_per_atom];
cvt.s32.u32 %r2, %tid.x;
div.s32 %r3, %r2, %r1;
cvt.s32.u32 %r4, %ntid.x;
div.s32 %r5, %r4, %r1;
cvt.s32.u32 %r6, %ctaid.x;
mul.lo.s32 %r7, %r6, %r5;
add.s32 %r8, %r3, %r7;
ld.param.s32 %r9, [__cudaparm_kernel_pair_inum];
setp.ge.s32 %p1, %r8, %r9;
@%p1 bra $Lt_0_30978;
.loc 16 61 0
cvt.s64.s32 %rd2, %r8;
mul.wide.s32 %rd3, %r8, 4;
ld.param.u64 %rd4, [__cudaparm_kernel_pair_dev_nbor];
add.u64 %rd5, %rd3, %rd4;
ld.global.s32 %r10, [%rd5+0];
ld.param.s32 %r11, [__cudaparm_kernel_pair_nbor_pitch];
cvt.s64.s32 %rd6, %r11;
mul.wide.s32 %rd7, %r11, 4;
add.u64 %rd8, %rd7, %rd5;
ld.global.s32 %r12, [%rd8+0];
sub.s32 %r13, %r1, 1;
and.b32 %r14, %r13, %r2;
cvt.s64.s32 %rd9, %r14;
mul.wide.s32 %rd10, %r14, 4;
ld.param.u64 %rd11, [__cudaparm_kernel_pair_dev_packed];
setp.ne.u64 %p2, %rd11, %rd4;
@%p2 bra $Lt_0_22530;
cvt.s32.s64 %r15, %rd6;
mul.lo.s32 %r16, %r15, %r1;
mov.s32 %r17, %r16;
mul.lo.s32 %r18, %r13, %r8;
add.s32 %r19, %r15, %r18;
cvt.s64.s32 %rd12, %r19;
mul.wide.s32 %rd13, %r19, 4;
add.u64 %rd14, %rd8, %rd13;
and.b32 %r20, %r13, %r12;
cvt.s64.s32 %rd15, %r20;
div.s32 %r21, %r12, %r1;
mul.lo.s32 %r22, %r16, %r21;
cvt.s64.s32 %rd16, %r22;
add.u64 %rd17, %rd15, %rd16;
mul.lo.u64 %rd18, %rd17, 4;
add.u64 %rd19, %rd14, %rd18;
add.u64 %rd20, %rd10, %rd14;
bra.uni $Lt_0_22274;
$Lt_0_22530:
add.u64 %rd21, %rd7, %rd8;
ld.global.s32 %r23, [%rd21+0];
cvt.s64.s32 %rd22, %r23;
mul.wide.s32 %rd23, %r23, 4;
add.u64 %rd24, %rd11, %rd23;
cvt.s64.s32 %rd25, %r12;
mul.wide.s32 %rd26, %r12, 4;
add.u64 %rd19, %rd24, %rd26;
mov.s32 %r17, %r1;
add.u64 %rd20, %rd10, %rd24;
$Lt_0_22274:
.loc 16 64 0
mov.u32 %r24, %r10;
mov.s32 %r25, 0;
mov.u32 %r26, %r25;
mov.s32 %r27, 0;
mov.u32 %r28, %r27;
mov.s32 %r29, 0;
mov.u32 %r30, %r29;
tex.1d.v4.f32.s32 {%f21,%f22,%f23,%f24},[pos_tex,{%r24,%r26,%r28,%r30}];
mov.f32 %f25, %f21;
mov.f32 %f26, %f22;
mov.f32 %f27, %f23;
mov.f32 %f28, %f24;
.loc 16 65 0
mov.u32 %r31, %r10;
mov.s32 %r32, 0;
mov.u32 %r33, %r32;
mov.s32 %r34, 0;
mov.u32 %r35, %r34;
mov.s32 %r36, 0;
mov.u32 %r37, %r36;
tex.1d.v4.f32.s32 {%f29,%f30,%f31,%f32},[q_tex,{%r31,%r33,%r35,%r37}];
mov.f32 %f33, %f29;
setp.ge.u64 %p3, %rd20, %rd19;
@%p3 bra $Lt_0_32514;
cvt.rzi.ftz.s32.f32 %r38, %f28;
cvt.s64.s32 %rd27, %r17;
ld.param.s32 %r39, [__cudaparm_kernel_pair_lj_types];
mul.lo.s32 %r40, %r39, %r38;
ld.param.u64 %rd28, [__cudaparm_kernel_pair_lj1];
mov.f32 %f34, 0f00000000; // 0
mov.f32 %f35, 0f00000000; // 0
mov.f32 %f36, 0f00000000; // 0
mov.f32 %f37, 0f00000000; // 0
mov.f32 %f38, 0f00000000; // 0
mov.u64 %rd29, __cuda___cuda_local_var_32541_33_non_const_sp_lj112;
$Lt_0_23298:
//<loop> Loop body line 65, nesting depth: 1, estimated iterations: unknown
.loc 16 69 0
ld.global.s32 %r41, [%rd20+0];
.loc 16 72 0
shr.s32 %r42, %r41, 30;
and.b32 %r43, %r42, 3;
cvt.s64.s32 %rd30, %r43;
mul.wide.s32 %rd31, %r43, 4;
add.u64 %rd32, %rd29, %rd31;
ld.shared.f32 %f39, [%rd32+0];
.loc 16 73 0
mov.f32 %f40, 0f3f800000; // 1
ld.shared.f32 %f41, [%rd32+16];
sub.ftz.f32 %f42, %f40, %f41;
.loc 16 76 0
and.b32 %r44, %r41, 1073741823;
mov.u32 %r45, %r44;
mov.s32 %r46, 0;
mov.u32 %r47, %r46;
mov.s32 %r48, 0;
mov.u32 %r49, %r48;
mov.s32 %r50, 0;
mov.u32 %r51, %r50;
tex.1d.v4.f32.s32 {%f43,%f44,%f45,%f46},[pos_tex,{%r45,%r47,%r49,%r51}];
mov.f32 %f47, %f43;
mov.f32 %f48, %f44;
mov.f32 %f49, %f45;
mov.f32 %f50, %f46;
cvt.rzi.ftz.s32.f32 %r52, %f50;
sub.ftz.f32 %f51, %f26, %f48;
sub.ftz.f32 %f52, %f25, %f47;
sub.ftz.f32 %f53, %f27, %f49;
mul.ftz.f32 %f54, %f51, %f51;
fma.rn.ftz.f32 %f55, %f52, %f52, %f54;
fma.rn.ftz.f32 %f56, %f53, %f53, %f55;
add.s32 %r53, %r52, %r40;
cvt.s64.s32 %rd33, %r53;
mul.wide.s32 %rd34, %r53, 16;
add.u64 %rd35, %rd34, %rd28;
ld.global.f32 %f57, [%rd35+8];
setp.gt.ftz.f32 %p4, %f57, %f56;
@!%p4 bra $Lt_0_26114;
rcp.approx.ftz.f32 %f58, %f56;
ld.global.f32 %f59, [%rd35+12];
setp.lt.ftz.f32 %p5, %f56, %f59;
@!%p5 bra $Lt_0_24322;
.loc 16 92 0
rsqrt.approx.ftz.f32 %f60, %f56;
mul.ftz.f32 %f61, %f58, %f60;
mov.f32 %f62, %f61;
.loc 16 93 0
mul.ftz.f32 %f63, %f61, %f61;
mov.f32 %f64, %f63;
.loc 16 94 0
mul.ftz.f32 %f65, %f63, %f39;
ld.global.v2.f32 {%f66,%f67}, [%rd35+0];
mul.ftz.f32 %f68, %f66, %f61;
sub.ftz.f32 %f69, %f68, %f67;
mul.ftz.f32 %f70, %f65, %f69;
bra.uni $Lt_0_24066;
$Lt_0_24322:
.loc 16 96 0
mov.f32 %f70, 0f00000000; // 0
$Lt_0_24066:
ld.param.f32 %f71, [__cudaparm_kernel_pair_cut_coulsq];
setp.gt.ftz.f32 %p6, %f71, %f56;
@!%p6 bra $Lt_0_24834;
.loc 20 518 0
rsqrt.approx.ftz.f32 %f72, %f58;
ld.param.f32 %f73, [__cudaparm_kernel_pair_g_ewald];
mul.ftz.f32 %f74, %f73, %f72;
mul.ftz.f32 %f75, %f74, %f74;
neg.ftz.f32 %f76, %f75;
mov.f32 %f77, 0f3fb8aa3b; // 1.4427
mul.ftz.f32 %f78, %f76, %f77;
ex2.approx.ftz.f32 %f79, %f78;
.loc 16 103 0
mov.f32 %f80, 0f3f800000; // 1
mov.f32 %f81, 0f3ea7ba05; // 0.327591
fma.rn.ftz.f32 %f82, %f81, %f74, %f80;
rcp.approx.ftz.f32 %f83, %f82;
mov.f32 %f84, 0f3e827906; // 0.25483
mov.f32 %f85, 0fbe91a98e; // -0.284497
mov.f32 %f86, 0f3fb5f0e3; // 1.42141
mov.f32 %f87, 0fbfba00e3; // -1.45315
mov.f32 %f88, 0f3f87dc22; // 1.06141
fma.rn.ftz.f32 %f89, %f88, %f83, %f87;
fma.rn.ftz.f32 %f90, %f83, %f89, %f86;
fma.rn.ftz.f32 %f91, %f83, %f90, %f85;
fma.rn.ftz.f32 %f92, %f83, %f91, %f84;
mul.ftz.f32 %f93, %f83, %f92;
mul.ftz.f32 %f94, %f79, %f93;
mov.f32 %f95, %f94;
.loc 16 104 0
mov.u32 %r54, %r44;
mov.s32 %r55, 0;
mov.u32 %r56, %r55;
mov.s32 %r57, 0;
mov.u32 %r58, %r57;
mov.s32 %r59, 0;
mov.u32 %r60, %r59;
tex.1d.v4.f32.s32 {%f96,%f97,%f98,%f99},[q_tex,{%r54,%r56,%r58,%r60}];
mov.f32 %f100, %f96;
ld.param.f32 %f101, [__cudaparm_kernel_pair_qqrd2e];
mul.ftz.f32 %f102, %f101, %f33;
mul.ftz.f32 %f103, %f102, %f100;
div.approx.ftz.f32 %f104, %f103, %f72;
mov.f32 %f105, %f104;
.loc 16 105 0
mov.f32 %f106, 0f3f906ebb; // 1.12838
mul.ftz.f32 %f107, %f74, %f106;
fma.rn.ftz.f32 %f108, %f79, %f107, %f94;
sub.ftz.f32 %f109, %f108, %f42;
mul.ftz.f32 %f110, %f104, %f109;
bra.uni $Lt_0_24578;
$Lt_0_24834:
.loc 16 107 0
mov.f32 %f110, 0f00000000; // 0
$Lt_0_24578:
.loc 16 111 0
add.ftz.f32 %f111, %f110, %f70;
mul.ftz.f32 %f112, %f111, %f58;
fma.rn.ftz.f32 %f36, %f52, %f112, %f36;
.loc 16 112 0
fma.rn.ftz.f32 %f35, %f51, %f112, %f35;
.loc 16 113 0
fma.rn.ftz.f32 %f34, %f53, %f112, %f34;
ld.param.s32 %r61, [__cudaparm_kernel_pair_eflag];
mov.u32 %r62, 0;
setp.le.s32 %p7, %r61, %r62;
@%p7 bra $Lt_0_25602;
.loc 16 116 0
mov.f32 %f113, %f105;
mov.f32 %f114, %f95;
sub.ftz.f32 %f115, %f114, %f42;
fma.rn.ftz.f32 %f116, %f113, %f115, %f37;
selp.f32 %f37, %f116, %f37, %p6;
@!%p5 bra $Lt_0_25602;
.loc 16 120 0
ld.param.u64 %rd36, [__cudaparm_kernel_pair_lj3];
add.u64 %rd37, %rd36, %rd34;
ld.global.v4.f32 {%f117,%f118,%f119,_}, [%rd37+0];
mov.f32 %f120, %f64;
mov.f32 %f121, %f62;
mul.ftz.f32 %f122, %f117, %f121;
sub.ftz.f32 %f123, %f122, %f118;
mul.ftz.f32 %f124, %f120, %f123;
sub.ftz.f32 %f125, %f124, %f119;
fma.rn.ftz.f32 %f38, %f39, %f125, %f38;
$Lt_0_25602:
$Lt_0_25090:
ld.param.s32 %r63, [__cudaparm_kernel_pair_vflag];
mov.u32 %r64, 0;
setp.le.s32 %p8, %r63, %r64;
@%p8 bra $Lt_0_26114;
.loc 16 124 0
mov.f32 %f126, %f10;
mul.ftz.f32 %f127, %f52, %f52;
fma.rn.ftz.f32 %f128, %f112, %f127, %f126;
mov.f32 %f10, %f128;
.loc 16 125 0
mov.f32 %f129, %f12;
fma.rn.ftz.f32 %f130, %f112, %f54, %f129;
mov.f32 %f12, %f130;
.loc 16 126 0
mov.f32 %f131, %f14;
mul.ftz.f32 %f132, %f53, %f53;
fma.rn.ftz.f32 %f133, %f112, %f132, %f131;
mov.f32 %f14, %f133;
.loc 16 127 0
mov.f32 %f134, %f16;
mul.ftz.f32 %f135, %f51, %f52;
fma.rn.ftz.f32 %f136, %f112, %f135, %f134;
mov.f32 %f16, %f136;
.loc 16 128 0
mov.f32 %f137, %f18;
mul.ftz.f32 %f138, %f52, %f53;
fma.rn.ftz.f32 %f139, %f112, %f138, %f137;
mov.f32 %f18, %f139;
.loc 16 129 0
mul.ftz.f32 %f140, %f51, %f53;
fma.rn.ftz.f32 %f19, %f112, %f140, %f19;
mov.f32 %f20, %f19;
$Lt_0_26114:
$Lt_0_23554:
.loc 16 68 0
mul.lo.u64 %rd38, %rd27, 4;
add.u64 %rd20, %rd20, %rd38;
setp.lt.u64 %p9, %rd20, %rd19;
@%p9 bra $Lt_0_23298;
bra.uni $Lt_0_22786;
$Lt_0_32514:
mov.f32 %f34, 0f00000000; // 0
mov.f32 %f35, 0f00000000; // 0
mov.f32 %f36, 0f00000000; // 0
mov.f32 %f37, 0f00000000; // 0
mov.f32 %f38, 0f00000000; // 0
$Lt_0_22786:
mov.u32 %r65, 1;
setp.le.s32 %p10, %r1, %r65;
@%p10 bra $Lt_0_28930;
.loc 16 134 0
mov.u64 %rd39, __cuda___cuda_local_var_32635_55_non_const_red_acc144;
cvt.s64.s32 %rd40, %r2;
mul.wide.s32 %rd41, %r2, 4;
add.u64 %rd42, %rd39, %rd41;
mov.f32 %f141, %f36;
st.shared.f32 [%rd42+0], %f141;
mov.f32 %f142, %f35;
st.shared.f32 [%rd42+512], %f142;
mov.f32 %f143, %f34;
st.shared.f32 [%rd42+1024], %f143;
mov.f32 %f144, %f38;
st.shared.f32 [%rd42+1536], %f144;
mov.f32 %f145, %f37;
st.shared.f32 [%rd42+2048], %f145;
shr.s32 %r66, %r1, 31;
mov.s32 %r67, 1;
and.b32 %r68, %r66, %r67;
add.s32 %r69, %r68, %r1;
shr.s32 %r70, %r69, 1;
mov.s32 %r71, %r70;
mov.u32 %r72, 0;
setp.ne.u32 %p11, %r70, %r72;
@!%p11 bra $Lt_0_27394;
$Lt_0_27906:
setp.ge.u32 %p12, %r14, %r71;
@%p12 bra $Lt_0_28162;
add.u32 %r73, %r2, %r71;
cvt.u64.u32 %rd43, %r73;
mul.wide.u32 %rd44, %r73, 4;
add.u64 %rd45, %rd39, %rd44;
ld.shared.f32 %f146, [%rd45+0];
add.ftz.f32 %f141, %f146, %f141;
st.shared.f32 [%rd42+0], %f141;
ld.shared.f32 %f147, [%rd45+512];
add.ftz.f32 %f142, %f147, %f142;
st.shared.f32 [%rd42+512], %f142;
ld.shared.f32 %f148, [%rd45+1024];
add.ftz.f32 %f143, %f148, %f143;
st.shared.f32 [%rd42+1024], %f143;
ld.shared.f32 %f149, [%rd45+1536];
add.ftz.f32 %f144, %f149, %f144;
st.shared.f32 [%rd42+1536], %f144;
ld.shared.f32 %f150, [%rd45+2048];
add.ftz.f32 %f145, %f150, %f145;
st.shared.f32 [%rd42+2048], %f145;
$Lt_0_28162:
shr.u32 %r71, %r71, 1;
mov.u32 %r74, 0;
setp.ne.u32 %p13, %r71, %r74;
@%p13 bra $Lt_0_27906;
$Lt_0_27394:
mov.f32 %f36, %f141;
mov.f32 %f35, %f142;
mov.f32 %f34, %f143;
mov.f32 %f38, %f144;
mov.f32 %f37, %f145;
ld.param.s32 %r75, [__cudaparm_kernel_pair_vflag];
mov.u32 %r76, 0;
setp.le.s32 %p14, %r75, %r76;
@%p14 bra $Lt_0_28930;
mov.f32 %f141, %f10;
st.shared.f32 [%rd42+0], %f141;
mov.f32 %f142, %f12;
st.shared.f32 [%rd42+512], %f142;
mov.f32 %f143, %f14;
st.shared.f32 [%rd42+1024], %f143;
mov.f32 %f144, %f16;
st.shared.f32 [%rd42+1536], %f144;
mov.f32 %f145, %f18;
st.shared.f32 [%rd42+2048], %f145;
mov.f32 %f151, %f19;
st.shared.f32 [%rd42+2560], %f151;
mov.s32 %r77, %r70;
@!%p11 bra $Lt_0_29442;
$Lt_0_29954:
setp.ge.u32 %p15, %r14, %r77;
@%p15 bra $Lt_0_30210;
add.u32 %r78, %r2, %r77;
cvt.u64.u32 %rd46, %r78;
mul.wide.u32 %rd47, %r78, 4;
add.u64 %rd48, %rd39, %rd47;
ld.shared.f32 %f152, [%rd48+0];
add.ftz.f32 %f141, %f152, %f141;
st.shared.f32 [%rd42+0], %f141;
ld.shared.f32 %f153, [%rd48+512];
add.ftz.f32 %f142, %f153, %f142;
st.shared.f32 [%rd42+512], %f142;
ld.shared.f32 %f154, [%rd48+1024];
add.ftz.f32 %f143, %f154, %f143;
st.shared.f32 [%rd42+1024], %f143;
ld.shared.f32 %f155, [%rd48+1536];
add.ftz.f32 %f144, %f155, %f144;
st.shared.f32 [%rd42+1536], %f144;
ld.shared.f32 %f156, [%rd48+2048];
add.ftz.f32 %f145, %f156, %f145;
st.shared.f32 [%rd42+2048], %f145;
ld.shared.f32 %f157, [%rd48+2560];
add.ftz.f32 %f151, %f157, %f151;
st.shared.f32 [%rd42+2560], %f151;
$Lt_0_30210:
shr.u32 %r77, %r77, 1;
mov.u32 %r79, 0;
setp.ne.u32 %p16, %r77, %r79;
@%p16 bra $Lt_0_29954;
$Lt_0_29442:
mov.f32 %f10, %f141;
mov.f32 %f12, %f142;
mov.f32 %f14, %f143;
mov.f32 %f16, %f144;
mov.f32 %f18, %f145;
mov.f32 %f20, %f151;
$Lt_0_28930:
$Lt_0_26882:
mov.u32 %r80, 0;
setp.ne.s32 %p17, %r14, %r80;
@%p17 bra $Lt_0_30978;
ld.param.u64 %rd49, [__cudaparm_kernel_pair___val_paramengv];
add.u64 %rd50, %rd49, %rd3;
ld.param.s32 %r81, [__cudaparm_kernel_pair_eflag];
mov.u32 %r82, 0;
setp.le.s32 %p18, %r81, %r82;
@%p18 bra $Lt_0_31490;
st.global.f32 [%rd50+0], %f38;
cvt.s64.s32 %rd51, %r9;
mul.wide.s32 %rd52, %r9, 4;
add.u64 %rd53, %rd52, %rd50;
st.global.f32 [%rd53+0], %f37;
add.u64 %rd50, %rd52, %rd53;
$Lt_0_31490:
ld.param.s32 %r83, [__cudaparm_kernel_pair_vflag];
mov.u32 %r84, 0;
setp.le.s32 %p19, %r83, %r84;
@%p19 bra $Lt_0_32002;
mov.f32 %f158, %f10;
st.global.f32 [%rd50+0], %f158;
cvt.s64.s32 %rd54, %r9;
mul.wide.s32 %rd55, %r9, 4;
add.u64 %rd56, %rd55, %rd50;
mov.f32 %f159, %f12;
st.global.f32 [%rd56+0], %f159;
add.u64 %rd57, %rd55, %rd56;
mov.f32 %f160, %f14;
st.global.f32 [%rd57+0], %f160;
add.u64 %rd58, %rd55, %rd57;
mov.f32 %f161, %f16;
st.global.f32 [%rd58+0], %f161;
add.u64 %rd50, %rd55, %rd58;
mov.f32 %f162, %f18;
st.global.f32 [%rd50+0], %f162;
mov.f32 %f163, %f20;
add.u64 %rd59, %rd55, %rd50;
st.global.f32 [%rd59+0], %f163;
$Lt_0_32002:
ld.param.u64 %rd60, [__cudaparm_kernel_pair_ans];
mul.lo.u64 %rd61, %rd2, 16;
add.u64 %rd62, %rd60, %rd61;
mov.f32 %f164, %f165;
st.global.v4.f32 [%rd62+0], {%f36,%f35,%f34,%f164};
$Lt_0_30978:
$Lt_0_21762:
.loc 16 137 0
exit;
$LDWend_kernel_pair:
} // kernel_pair
.entry kernel_pair_fast (
.param .u64 __cudaparm_kernel_pair_fast_x_,
.param .u64 __cudaparm_kernel_pair_fast_lj1_in,
.param .u64 __cudaparm_kernel_pair_fast_lj3_in,
.param .u64 __cudaparm_kernel_pair_fast_sp_lj_in,
.param .u64 __cudaparm_kernel_pair_fast_dev_nbor,
.param .u64 __cudaparm_kernel_pair_fast_dev_packed,
.param .u64 __cudaparm_kernel_pair_fast_ans,
.param .u64 __cudaparm_kernel_pair_fast___val_paramengv,
.param .s32 __cudaparm_kernel_pair_fast_eflag,
.param .s32 __cudaparm_kernel_pair_fast_vflag,
.param .s32 __cudaparm_kernel_pair_fast_inum,
.param .s32 __cudaparm_kernel_pair_fast_nbor_pitch,
.param .u64 __cudaparm_kernel_pair_fast_q_,
.param .f32 __cudaparm_kernel_pair_fast_cut_coulsq,
.param .f32 __cudaparm_kernel_pair_fast_qqrd2e,
.param .f32 __cudaparm_kernel_pair_fast_g_ewald,
.param .s32 __cudaparm_kernel_pair_fast_t_per_atom)
{
.reg .u32 %r<88>;
.reg .u64 %rd<76>;
.reg .f32 %f<170>;
.reg .pred %p<24>;
.shared .align 4 .b8 __cuda___cuda_local_var_32654_33_non_const_sp_lj3320[32];
.shared .align 16 .b8 __cuda___cuda_local_var_32652_34_non_const_lj13360[1936];
.shared .align 16 .b8 __cuda___cuda_local_var_32653_34_non_const_lj35296[1936];
.shared .align 4 .b8 __cuda___cuda_local_var_32749_55_non_const_red_acc7232[3072];
// __cuda_local_var_32665_10_non_const_f = 64
// __cuda_local_var_32667_9_non_const_virial = 16
// __cuda_local_var_32703_43_non_const_r6inv = 44
// __cuda_local_var_32703_50_non_const_r3inv = 40
// __cuda_local_var_32703_57_non_const_prefactor = 52
// __cuda_local_var_32703_68_non_const__erfc = 48
.loc 16 147 0
$LDWbegin_kernel_pair_fast:
cvt.s32.u32 %r1, %tid.x;
mov.u32 %r2, 7;
setp.gt.s32 %p1, %r1, %r2;
@%p1 bra $Lt_1_24066;
.loc 16 155 0
mov.u64 %rd1, __cuda___cuda_local_var_32654_33_non_const_sp_lj3320;
cvt.s64.s32 %rd2, %r1;
mul.wide.s32 %rd3, %r1, 4;
ld.param.u64 %rd4, [__cudaparm_kernel_pair_fast_sp_lj_in];
add.u64 %rd5, %rd4, %rd3;
ld.global.f32 %f1, [%rd5+0];
add.u64 %rd6, %rd3, %rd1;
st.shared.f32 [%rd6+0], %f1;
$Lt_1_24066:
mov.u64 %rd1, __cuda___cuda_local_var_32654_33_non_const_sp_lj3320;
mov.u32 %r3, 120;
setp.gt.s32 %p2, %r1, %r3;
@%p2 bra $Lt_1_24578;
.loc 16 157 0
mov.u64 %rd7, __cuda___cuda_local_var_32652_34_non_const_lj13360;
cvt.s64.s32 %rd8, %r1;
mul.wide.s32 %rd9, %r1, 16;
ld.param.u64 %rd10, [__cudaparm_kernel_pair_fast_lj1_in];
add.u64 %rd11, %rd10, %rd9;
add.u64 %rd12, %rd9, %rd7;
ld.global.v4.f32 {%f2,%f3,%f4,%f5}, [%rd11+0];
st.shared.v4.f32 [%rd12+0], {%f2,%f3,%f4,%f5};
ld.param.s32 %r4, [__cudaparm_kernel_pair_fast_eflag];
mov.u32 %r5, 0;
setp.le.s32 %p3, %r4, %r5;
@%p3 bra $Lt_1_25090;
.loc 16 159 0
mov.u64 %rd13, __cuda___cuda_local_var_32653_34_non_const_lj35296;
ld.param.u64 %rd14, [__cudaparm_kernel_pair_fast_lj3_in];
add.u64 %rd15, %rd14, %rd9;
add.u64 %rd16, %rd9, %rd13;
ld.global.v4.f32 {%f6,%f7,%f8,%f9}, [%rd15+0];
st.shared.v4.f32 [%rd16+0], {%f6,%f7,%f8,%f9};
$Lt_1_25090:
mov.u64 %rd13, __cuda___cuda_local_var_32653_34_non_const_lj35296;
$Lt_1_24578:
mov.u64 %rd13, __cuda___cuda_local_var_32653_34_non_const_lj35296;
mov.u64 %rd7, __cuda___cuda_local_var_32652_34_non_const_lj13360;
.loc 16 168 0
mov.f32 %f10, 0f00000000; // 0
mov.f32 %f11, %f10;
mov.f32 %f12, 0f00000000; // 0
mov.f32 %f13, %f12;
mov.f32 %f14, 0f00000000; // 0
mov.f32 %f15, %f14;
mov.f32 %f16, 0f00000000; // 0
mov.f32 %f17, %f16;
mov.f32 %f18, 0f00000000; // 0
mov.f32 %f19, %f18;
mov.f32 %f20, 0f00000000; // 0
mov.f32 %f21, %f20;
.loc 16 170 0
bar.sync 0;
ld.param.s32 %r6, [__cudaparm_kernel_pair_fast_t_per_atom];
div.s32 %r7, %r1, %r6;
cvt.s32.u32 %r8, %ntid.x;
div.s32 %r9, %r8, %r6;
cvt.s32.u32 %r10, %ctaid.x;
mul.lo.s32 %r11, %r10, %r9;
add.s32 %r12, %r7, %r11;
ld.param.s32 %r13, [__cudaparm_kernel_pair_fast_inum];
setp.ge.s32 %p4, %r12, %r13;
@%p4 bra $Lt_1_34818;
.loc 16 175 0
cvt.s64.s32 %rd17, %r12;
mul.wide.s32 %rd18, %r12, 4;
ld.param.u64 %rd19, [__cudaparm_kernel_pair_fast_dev_nbor];
add.u64 %rd20, %rd18, %rd19;
ld.global.s32 %r14, [%rd20+0];
ld.param.s32 %r15, [__cudaparm_kernel_pair_fast_nbor_pitch];
cvt.s64.s32 %rd21, %r15;
mul.wide.s32 %rd22, %r15, 4;
add.u64 %rd23, %rd22, %rd20;
ld.global.s32 %r16, [%rd23+0];
sub.s32 %r17, %r6, 1;
and.b32 %r18, %r17, %r1;
cvt.s64.s32 %rd24, %r18;
mul.wide.s32 %rd25, %r18, 4;
ld.param.u64 %rd26, [__cudaparm_kernel_pair_fast_dev_packed];
setp.ne.u64 %p5, %rd26, %rd19;
@%p5 bra $Lt_1_26370;
cvt.s32.s64 %r19, %rd21;
mul.lo.s32 %r20, %r19, %r6;
mov.s32 %r21, %r20;
mul.lo.s32 %r22, %r17, %r12;
add.s32 %r23, %r19, %r22;
cvt.s64.s32 %rd27, %r23;
mul.wide.s32 %rd28, %r23, 4;
add.u64 %rd29, %rd23, %rd28;
and.b32 %r24, %r17, %r16;
cvt.s64.s32 %rd30, %r24;
div.s32 %r25, %r16, %r6;
mul.lo.s32 %r26, %r20, %r25;
cvt.s64.s32 %rd31, %r26;
add.u64 %rd32, %rd30, %rd31;
mul.lo.u64 %rd33, %rd32, 4;
add.u64 %rd34, %rd29, %rd33;
add.u64 %rd35, %rd25, %rd29;
bra.uni $Lt_1_26114;
$Lt_1_26370:
add.u64 %rd36, %rd22, %rd23;
ld.global.s32 %r27, [%rd36+0];
cvt.s64.s32 %rd37, %r27;
mul.wide.s32 %rd38, %r27, 4;
add.u64 %rd39, %rd26, %rd38;
cvt.s64.s32 %rd40, %r16;
mul.wide.s32 %rd41, %r16, 4;
add.u64 %rd34, %rd39, %rd41;
mov.s32 %r21, %r6;
add.u64 %rd35, %rd25, %rd39;
$Lt_1_26114:
.loc 16 178 0
mov.u32 %r28, %r14;
mov.s32 %r29, 0;
mov.u32 %r30, %r29;
mov.s32 %r31, 0;
mov.u32 %r32, %r31;
mov.s32 %r33, 0;
mov.u32 %r34, %r33;
tex.1d.v4.f32.s32 {%f22,%f23,%f24,%f25},[pos_tex,{%r28,%r30,%r32,%r34}];
mov.f32 %f26, %f22;
mov.f32 %f27, %f23;
mov.f32 %f28, %f24;
mov.f32 %f29, %f25;
.loc 16 179 0
mov.u32 %r35, %r14;
mov.s32 %r36, 0;
mov.u32 %r37, %r36;
mov.s32 %r38, 0;
mov.u32 %r39, %r38;
mov.s32 %r40, 0;
mov.u32 %r41, %r40;
tex.1d.v4.f32.s32 {%f30,%f31,%f32,%f33},[q_tex,{%r35,%r37,%r39,%r41}];
mov.f32 %f34, %f30;
setp.ge.u64 %p6, %rd35, %rd34;
@%p6 bra $Lt_1_36354;
cvt.rzi.ftz.s32.f32 %r42, %f29;
cvt.s64.s32 %rd42, %r21;
mul.lo.s32 %r43, %r42, 11;
cvt.rn.f32.s32 %f35, %r43;
mov.f32 %f36, 0f00000000; // 0
mov.f32 %f37, 0f00000000; // 0
mov.f32 %f38, 0f00000000; // 0
mov.f32 %f39, 0f00000000; // 0
mov.f32 %f40, 0f00000000; // 0
$Lt_1_27138:
//<loop> Loop body line 179, nesting depth: 1, estimated iterations: unknown
.loc 16 184 0
ld.global.s32 %r44, [%rd35+0];
.loc 16 187 0
shr.s32 %r45, %r44, 30;
and.b32 %r46, %r45, 3;
cvt.s64.s32 %rd43, %r46;
mul.wide.s32 %rd44, %r46, 4;
add.u64 %rd45, %rd1, %rd44;
ld.shared.f32 %f41, [%rd45+0];
.loc 16 188 0
mov.f32 %f42, 0f3f800000; // 1
ld.shared.f32 %f43, [%rd45+16];
sub.ftz.f32 %f44, %f42, %f43;
.loc 16 191 0
and.b32 %r47, %r44, 1073741823;
mov.u32 %r48, %r47;
mov.s32 %r49, 0;
mov.u32 %r50, %r49;
mov.s32 %r51, 0;
mov.u32 %r52, %r51;
mov.s32 %r53, 0;
mov.u32 %r54, %r53;
tex.1d.v4.f32.s32 {%f45,%f46,%f47,%f48},[pos_tex,{%r48,%r50,%r52,%r54}];
mov.f32 %f49, %f45;
mov.f32 %f50, %f46;
mov.f32 %f51, %f47;
mov.f32 %f52, %f48;
sub.ftz.f32 %f53, %f27, %f50;
sub.ftz.f32 %f54, %f26, %f49;
sub.ftz.f32 %f55, %f28, %f51;
mul.ftz.f32 %f56, %f53, %f53;
fma.rn.ftz.f32 %f57, %f54, %f54, %f56;
fma.rn.ftz.f32 %f58, %f55, %f55, %f57;
add.ftz.f32 %f59, %f35, %f52;
cvt.rzi.ftz.s32.f32 %r55, %f59;
cvt.s64.s32 %rd46, %r55;
mul.wide.s32 %rd47, %r55, 16;
add.u64 %rd48, %rd47, %rd7;
ld.shared.f32 %f60, [%rd48+8];
setp.gt.ftz.f32 %p7, %f60, %f58;
@!%p7 bra $Lt_1_29954;
rcp.approx.ftz.f32 %f61, %f58;
ld.shared.f32 %f62, [%rd48+12];
setp.lt.ftz.f32 %p8, %f58, %f62;
@!%p8 bra $Lt_1_28162;
.loc 16 206 0
rsqrt.approx.ftz.f32 %f63, %f58;
mul.ftz.f32 %f64, %f61, %f63;
mov.f32 %f65, %f64;
.loc 16 207 0
mul.ftz.f32 %f66, %f64, %f64;
mov.f32 %f67, %f66;
.loc 16 208 0
mul.ftz.f32 %f68, %f66, %f41;
ld.shared.v2.f32 {%f69,%f70}, [%rd48+0];
mul.ftz.f32 %f71, %f69, %f64;
sub.ftz.f32 %f72, %f71, %f70;
mul.ftz.f32 %f73, %f68, %f72;
bra.uni $Lt_1_27906;
$Lt_1_28162:
.loc 16 210 0
mov.f32 %f73, 0f00000000; // 0
$Lt_1_27906:
ld.param.f32 %f74, [__cudaparm_kernel_pair_fast_cut_coulsq];
setp.gt.ftz.f32 %p9, %f74, %f58;
@!%p9 bra $Lt_1_28674;
.loc 20 518 0
rsqrt.approx.ftz.f32 %f75, %f61;
ld.param.f32 %f76, [__cudaparm_kernel_pair_fast_g_ewald];
mul.ftz.f32 %f77, %f76, %f75;
mul.ftz.f32 %f78, %f77, %f77;
neg.ftz.f32 %f79, %f78;
mov.f32 %f80, 0f3fb8aa3b; // 1.4427
mul.ftz.f32 %f81, %f79, %f80;
ex2.approx.ftz.f32 %f82, %f81;
.loc 16 217 0
mov.f32 %f83, 0f3f800000; // 1
mov.f32 %f84, 0f3ea7ba05; // 0.327591
fma.rn.ftz.f32 %f85, %f84, %f77, %f83;
rcp.approx.ftz.f32 %f86, %f85;
mov.f32 %f87, 0f3e827906; // 0.25483
mov.f32 %f88, 0fbe91a98e; // -0.284497
mov.f32 %f89, 0f3fb5f0e3; // 1.42141
mov.f32 %f90, 0fbfba00e3; // -1.45315
mov.f32 %f91, 0f3f87dc22; // 1.06141
fma.rn.ftz.f32 %f92, %f91, %f86, %f90;
fma.rn.ftz.f32 %f93, %f86, %f92, %f89;
fma.rn.ftz.f32 %f94, %f86, %f93, %f88;
fma.rn.ftz.f32 %f95, %f86, %f94, %f87;
mul.ftz.f32 %f96, %f86, %f95;
mul.ftz.f32 %f97, %f82, %f96;
mov.f32 %f98, %f97;
.loc 16 218 0
mov.u32 %r56, %r47;
mov.s32 %r57, 0;
mov.u32 %r58, %r57;
mov.s32 %r59, 0;
mov.u32 %r60, %r59;
mov.s32 %r61, 0;
mov.u32 %r62, %r61;
tex.1d.v4.f32.s32 {%f99,%f100,%f101,%f102},[q_tex,{%r56,%r58,%r60,%r62}];
mov.f32 %f103, %f99;
ld.param.f32 %f104, [__cudaparm_kernel_pair_fast_qqrd2e];
mul.ftz.f32 %f105, %f104, %f34;
mul.ftz.f32 %f106, %f105, %f103;
div.approx.ftz.f32 %f107, %f106, %f75;
mov.f32 %f108, %f107;
.loc 16 219 0
mov.f32 %f109, 0f3f906ebb; // 1.12838
mul.ftz.f32 %f110, %f77, %f109;
fma.rn.ftz.f32 %f111, %f82, %f110, %f97;
sub.ftz.f32 %f112, %f111, %f44;
mul.ftz.f32 %f113, %f107, %f112;
bra.uni $Lt_1_28418;
$Lt_1_28674:
.loc 16 221 0
mov.f32 %f113, 0f00000000; // 0
$Lt_1_28418:
.loc 16 225 0
add.ftz.f32 %f114, %f113, %f73;
mul.ftz.f32 %f115, %f114, %f61;
fma.rn.ftz.f32 %f38, %f54, %f115, %f38;
.loc 16 226 0
fma.rn.ftz.f32 %f37, %f53, %f115, %f37;
.loc 16 227 0
fma.rn.ftz.f32 %f36, %f55, %f115, %f36;
ld.param.s32 %r63, [__cudaparm_kernel_pair_fast_eflag];
mov.u32 %r64, 0;
setp.le.s32 %p10, %r63, %r64;
@%p10 bra $Lt_1_29442;
.loc 16 230 0
mov.f32 %f116, %f108;
mov.f32 %f117, %f98;
sub.ftz.f32 %f118, %f117, %f44;
fma.rn.ftz.f32 %f119, %f116, %f118, %f39;
selp.f32 %f39, %f119, %f39, %p9;
@!%p8 bra $Lt_1_29442;
.loc 16 233 0
add.u64 %rd49, %rd47, %rd13;
mov.f32 %f120, %f67;
ld.shared.v4.f32 {%f121,%f122,%f123,_}, [%rd49+0];
mov.f32 %f124, %f65;
mul.ftz.f32 %f125, %f121, %f124;
sub.ftz.f32 %f126, %f125, %f122;
mul.ftz.f32 %f127, %f120, %f126;
.loc 16 234 0
sub.ftz.f32 %f128, %f127, %f123;
fma.rn.ftz.f32 %f40, %f41, %f128, %f40;
$Lt_1_29442:
$Lt_1_28930:
ld.param.s32 %r65, [__cudaparm_kernel_pair_fast_vflag];
mov.u32 %r66, 0;
setp.le.s32 %p11, %r65, %r66;
@%p11 bra $Lt_1_29954;
.loc 16 238 0
mov.f32 %f129, %f11;
mul.ftz.f32 %f130, %f54, %f54;
fma.rn.ftz.f32 %f131, %f115, %f130, %f129;
mov.f32 %f11, %f131;
.loc 16 239 0
mov.f32 %f132, %f13;
fma.rn.ftz.f32 %f133, %f115, %f56, %f132;
mov.f32 %f13, %f133;
.loc 16 240 0
mov.f32 %f134, %f15;
mul.ftz.f32 %f135, %f55, %f55;
fma.rn.ftz.f32 %f136, %f115, %f135, %f134;
mov.f32 %f15, %f136;
.loc 16 241 0
mov.f32 %f137, %f17;
mul.ftz.f32 %f138, %f53, %f54;
fma.rn.ftz.f32 %f139, %f115, %f138, %f137;
mov.f32 %f17, %f139;
.loc 16 242 0
mov.f32 %f140, %f19;
mul.ftz.f32 %f141, %f54, %f55;
fma.rn.ftz.f32 %f142, %f115, %f141, %f140;
mov.f32 %f19, %f142;
.loc 16 243 0
mul.ftz.f32 %f143, %f53, %f55;
fma.rn.ftz.f32 %f20, %f115, %f143, %f20;
mov.f32 %f21, %f20;
$Lt_1_29954:
$Lt_1_27394:
.loc 16 183 0
mul.lo.u64 %rd50, %rd42, 4;
add.u64 %rd35, %rd35, %rd50;
setp.lt.u64 %p12, %rd35, %rd34;
@%p12 bra $Lt_1_27138;
bra.uni $Lt_1_26626;
$Lt_1_36354:
mov.f32 %f36, 0f00000000; // 0
mov.f32 %f37, 0f00000000; // 0
mov.f32 %f38, 0f00000000; // 0
mov.f32 %f39, 0f00000000; // 0
mov.f32 %f40, 0f00000000; // 0
$Lt_1_26626:
mov.u32 %r67, 1;
setp.le.s32 %p13, %r6, %r67;
@%p13 bra $Lt_1_32770;
.loc 16 248 0
mov.u64 %rd51, __cuda___cuda_local_var_32749_55_non_const_red_acc7232;
cvt.s64.s32 %rd52, %r1;
mul.wide.s32 %rd53, %r1, 4;
add.u64 %rd54, %rd51, %rd53;
mov.f32 %f144, %f38;
st.shared.f32 [%rd54+0], %f144;
mov.f32 %f145, %f37;
st.shared.f32 [%rd54+512], %f145;
mov.f32 %f146, %f36;
st.shared.f32 [%rd54+1024], %f146;
mov.f32 %f147, %f40;
st.shared.f32 [%rd54+1536], %f147;
mov.f32 %f148, %f39;
st.shared.f32 [%rd54+2048], %f148;
shr.s32 %r68, %r6, 31;
mov.s32 %r69, 1;
and.b32 %r70, %r68, %r69;
add.s32 %r71, %r70, %r6;
shr.s32 %r72, %r71, 1;
mov.s32 %r73, %r72;
mov.u32 %r74, 0;
setp.ne.u32 %p14, %r72, %r74;
@!%p14 bra $Lt_1_31234;
$Lt_1_31746:
setp.ge.u32 %p15, %r18, %r73;
@%p15 bra $Lt_1_32002;
add.u32 %r75, %r1, %r73;
cvt.u64.u32 %rd55, %r75;
mul.wide.u32 %rd56, %r75, 4;
add.u64 %rd57, %rd51, %rd56;
ld.shared.f32 %f149, [%rd57+0];
add.ftz.f32 %f144, %f149, %f144;
st.shared.f32 [%rd54+0], %f144;
ld.shared.f32 %f150, [%rd57+512];
add.ftz.f32 %f145, %f150, %f145;
st.shared.f32 [%rd54+512], %f145;
ld.shared.f32 %f151, [%rd57+1024];
add.ftz.f32 %f146, %f151, %f146;
st.shared.f32 [%rd54+1024], %f146;
ld.shared.f32 %f152, [%rd57+1536];
add.ftz.f32 %f147, %f152, %f147;
st.shared.f32 [%rd54+1536], %f147;
ld.shared.f32 %f153, [%rd57+2048];
add.ftz.f32 %f148, %f153, %f148;
st.shared.f32 [%rd54+2048], %f148;
$Lt_1_32002:
shr.u32 %r73, %r73, 1;
mov.u32 %r76, 0;
setp.ne.u32 %p16, %r73, %r76;
@%p16 bra $Lt_1_31746;
$Lt_1_31234:
mov.f32 %f38, %f144;
mov.f32 %f37, %f145;
mov.f32 %f36, %f146;
mov.f32 %f40, %f147;
mov.f32 %f39, %f148;
ld.param.s32 %r77, [__cudaparm_kernel_pair_fast_vflag];
mov.u32 %r78, 0;
setp.le.s32 %p17, %r77, %r78;
@%p17 bra $Lt_1_32770;
mov.f32 %f144, %f11;
st.shared.f32 [%rd54+0], %f144;
mov.f32 %f145, %f13;
st.shared.f32 [%rd54+512], %f145;
mov.f32 %f146, %f15;
st.shared.f32 [%rd54+1024], %f146;
mov.f32 %f147, %f17;
st.shared.f32 [%rd54+1536], %f147;
mov.f32 %f148, %f19;
st.shared.f32 [%rd54+2048], %f148;
mov.f32 %f154, %f20;
st.shared.f32 [%rd54+2560], %f154;
mov.s32 %r79, %r72;
@!%p14 bra $Lt_1_33282;
$Lt_1_33794:
setp.ge.u32 %p18, %r18, %r79;
@%p18 bra $Lt_1_34050;
add.u32 %r80, %r1, %r79;
cvt.u64.u32 %rd58, %r80;
mul.wide.u32 %rd59, %r80, 4;
add.u64 %rd60, %rd51, %rd59;
ld.shared.f32 %f155, [%rd60+0];
add.ftz.f32 %f144, %f155, %f144;
st.shared.f32 [%rd54+0], %f144;
ld.shared.f32 %f156, [%rd60+512];
add.ftz.f32 %f145, %f156, %f145;
st.shared.f32 [%rd54+512], %f145;
ld.shared.f32 %f157, [%rd60+1024];
add.ftz.f32 %f146, %f157, %f146;
st.shared.f32 [%rd54+1024], %f146;
ld.shared.f32 %f158, [%rd60+1536];
add.ftz.f32 %f147, %f158, %f147;
st.shared.f32 [%rd54+1536], %f147;
ld.shared.f32 %f159, [%rd60+2048];
add.ftz.f32 %f148, %f159, %f148;
st.shared.f32 [%rd54+2048], %f148;
ld.shared.f32 %f160, [%rd60+2560];
add.ftz.f32 %f154, %f160, %f154;
st.shared.f32 [%rd54+2560], %f154;
$Lt_1_34050:
shr.u32 %r79, %r79, 1;
mov.u32 %r81, 0;
setp.ne.u32 %p19, %r79, %r81;
@%p19 bra $Lt_1_33794;
$Lt_1_33282:
mov.f32 %f11, %f144;
mov.f32 %f13, %f145;
mov.f32 %f15, %f146;
mov.f32 %f17, %f147;
mov.f32 %f19, %f148;
mov.f32 %f21, %f154;
$Lt_1_32770:
$Lt_1_30722:
mov.u32 %r82, 0;
setp.ne.s32 %p20, %r18, %r82;
@%p20 bra $Lt_1_34818;
ld.param.u64 %rd61, [__cudaparm_kernel_pair_fast___val_paramengv];
add.u64 %rd62, %rd61, %rd18;
ld.param.s32 %r83, [__cudaparm_kernel_pair_fast_eflag];
mov.u32 %r84, 0;
setp.le.s32 %p21, %r83, %r84;
@%p21 bra $Lt_1_35330;
st.global.f32 [%rd62+0], %f40;
cvt.s64.s32 %rd63, %r13;
mul.wide.s32 %rd64, %r13, 4;
add.u64 %rd65, %rd64, %rd62;
st.global.f32 [%rd65+0], %f39;
add.u64 %rd62, %rd64, %rd65;
$Lt_1_35330:
ld.param.s32 %r85, [__cudaparm_kernel_pair_fast_vflag];
mov.u32 %r86, 0;
setp.le.s32 %p22, %r85, %r86;
@%p22 bra $Lt_1_35842;
mov.f32 %f161, %f11;
st.global.f32 [%rd62+0], %f161;
cvt.s64.s32 %rd66, %r13;
mul.wide.s32 %rd67, %r13, 4;
add.u64 %rd68, %rd67, %rd62;
mov.f32 %f162, %f13;
st.global.f32 [%rd68+0], %f162;
add.u64 %rd69, %rd67, %rd68;
mov.f32 %f163, %f15;
st.global.f32 [%rd69+0], %f163;
add.u64 %rd70, %rd67, %rd69;
mov.f32 %f164, %f17;
st.global.f32 [%rd70+0], %f164;
add.u64 %rd62, %rd67, %rd70;
mov.f32 %f165, %f19;
st.global.f32 [%rd62+0], %f165;
mov.f32 %f166, %f21;
add.u64 %rd71, %rd67, %rd62;
st.global.f32 [%rd71+0], %f166;
$Lt_1_35842:
ld.param.u64 %rd72, [__cudaparm_kernel_pair_fast_ans];
mul.lo.u64 %rd73, %rd17, 16;
add.u64 %rd74, %rd72, %rd73;
mov.f32 %f167, %f168;
st.global.v4.f32 [%rd74+0], {%f38,%f37,%f36,%f167};
$Lt_1_34818:
$Lt_1_25602:
.loc 16 251 0
exit;
$LDWend_kernel_pair_fast:
} // kernel_pair_fast
c4science · Help