GPU_PTX 3.LB+membar.cta+po+addr-p0::p1:p2-xgygzs "Mcta-d-Rg-Wg Rf-c-g Po-d-Rg-Ws Rf-w-s Addr-d-s-Wg Rf-c-g" {} P0 | P1 | P2 ; ld.cg.s32 r0, [r1] | ld.cg.s32 r0, [r1] | ld.cg.s32 r0, [r1] ; membar.cta | mov.s32 r2, 1 | and.b32 r2, r0 2147483648 ; mov.s32 r2, 1 | st.cg.s32 [r3], r2 | cvt.u64.u32 r3, r2 ; st.cg.s32 [r3], r2 | | mov.s32 r4, 1 ; | | add.u64 r5, r5 r3 ; | | st.cg.s32 [r5], r4 ; exists (0:r0=1 /\ 1:r0=1 /\ 2:r0=1) (device (kernel (cta (warp P0)) (cta (warp P1) (warp P2)))) x:global; y:global; z:local Generated assembler //START _litmus_P2 ld.cg.s32 sr02, [sr12] ; and.b32 sr22, sr02, 2147483648 ; cvt.u64.u32 sr32, sr22 ; mov.s32 sr42, 1 ; add.u64 sr52, sr52, sr32 ; st.cg.s32 [sr52], sr42 ; //START _litmus_P1 ld.cg.s32 sr01, [sr11] ; mov.s32 sr21, 1 ; st.cg.s32 [sr31], sr21 ; //START _litmus_P0 ld.cg.s32 sr00, [sr10] ; membar.cta ; mov.s32 sr20, 1 ; st.cg.s32 [sr30], sr20 ; Binary '/tmp/dir1450e5.tmp/3.LB+membar.cta+po+addr-p0::p1:p2-xgygzs-0.exe' successfully loaded Specification clusters: 3 Specification: [[(681, 3, 'R0'), (679, 2, 'R4')], [(736, 3, 'R0'), (734, 2, 'R12')], [(790, 3, 'R0'), (787, 2, 'R14')]] Cluster 0: OK Cluster 1: OK Cluster 2: OK !!SUCCESS!! Test 3.LB+membar.cta+po+addr-p0::p1:p2-xgygzs Allowed Histogram (7 states) 21719 :>0:r0=0; 1:r0=0; 2:r0=0; 10954 :>0:r0=1; 1:r0=0; 2:r0=0; 8129 :>0:r0=0; 1:r0=1; 2:r0=0; 644 :>0:r0=1; 1:r0=1; 2:r0=0; 6499 :>0:r0=0; 1:r0=0; 2:r0=1; 1500 :>0:r0=1; 1:r0=0; 2:r0=1; 555 :>0:r0=0; 1:r0=1; 2:r0=1; No Witnesses Positive: 0, Negative: 50000 Condition exists (0:r0=1 /\ 1:r0=1 /\ 2:r0=1) is NOT validated Observation 3.LB+membar.cta+po+addr-p0::p1:p2-xgygzs Never 0 50000 Time 3.LB+membar.cta+po+addr-p0::p1:p2-xgygzs 16.02 -- device info -- Name: GeForce GTX TITAN Major revision number: 3 Minor revision number: 5 Total global memory: 6442123264 Total shared memory per block: 49152 Number of multiprocessors: 14 Machine: formal Revision 1383, version 5.01 GPU_PTX experimental backend Command line: ./litmus -mach GTXTitan.cfg -a 8 -gpu-verf-comp true big_bucket/3.LB+membar.cta+po+addr-p0::p1:p2-xgygzs-0.litmus -s 50000 Parameters #ifndef SIZE_OF_TEST #define SIZE_OF_TEST 50000 #endif #ifndef NUMBER_OF_RUN #define NUMBER_OF_RUN 1 #endif /* cc options: -arch=sm_35 */ /* barrier: user */ /* speedcheck: no */ /* gpu_general_bc: true */ /* gpu_mem_stress: true */ /* gpu-rand-threads: true */ /* gpu_cuda: true */ /* gpu_gen_addr: true */ /* gpu_cvt_global_to_gen: true */ /* gpu_cvt_shared_to_gen: true */ /* gpu_max_ctas: 14 */ /* gpu_min_ctas: 10 */ /* gpu_max_threads: 512 */ /* gpu_min_threads: 128 */