GPU_PTX 3.SB+membar.gl+po+po-p0:p1:p2-xgygzs "Mgl-d-Wg-Rg Fr-w-g Po-d-Wg-Rs Fr-w-s Po-d-Ws-Rg Fr-w-g" {} P0 | P1 | P2 ; mov.s32 r0, 1 | mov.s32 r0, 1 | mov.s32 r0, 1 ; st.cg.s32 [r1], r0 | st.cg.s32 [r1], r0 | st.cg.s32 [r1], r0 ; membar.gl | ld.cg.s32 r2, [r3] | ld.cg.s32 r2, [r3] ; ld.cg.s32 r2, [r3] | | ; exists (0:r2=0 /\ 1:r2=0 /\ 2:r2=0) (device (kernel (cta (warp P0) (warp P1) (warp P2)))) x:global; y:global; z:local Generated assembler //START _litmus_P2 mov.s32 sr02, 1 ; st.cg.s32 [sr12], sr02 ; ld.cg.s32 sr22, [sr32] ; //START _litmus_P1 mov.s32 sr01, 1 ; st.cg.s32 [sr11], sr01 ; ld.cg.s32 sr21, [sr31] ; //START _litmus_P0 mov.s32 sr00, 1 ; st.cg.s32 [sr10], sr00 ; membar.gl ; ld.cg.s32 sr20, [sr30] ; Binary '/tmp/dir9df400.tmp/3.SB+membar.gl+po+po-p0:p1:p2-xgygzs-0.exe' successfully loaded Specification clusters: 3 Specification: [[(795, 2, 'R10'), (797, 3, 'R0')], [(870, 2, 'R10'), (869, 3, 'R0')], [(936, 2, 'R12'), (935, 3, 'R0')]] Cluster 0: OK Cluster 1: OK Cluster 2: OK !!SUCCESS!! Test 3.SB+membar.gl+po+po-p0:p1:p2-xgygzs Allowed Histogram (7 states) 8546 :>0:r2=1; 1:r2=0; 2:r2=0; 41 :>0:r2=0; 1:r2=1; 2:r2=0; 18708 :>0:r2=1; 1:r2=1; 2:r2=0; 16 :>0:r2=0; 1:r2=0; 2:r2=1; 14602 :>0:r2=1; 1:r2=0; 2:r2=1; 72 :>0:r2=0; 1:r2=1; 2:r2=1; 8015 :>0:r2=1; 1:r2=1; 2:r2=1; No Witnesses Positive: 0, Negative: 50000 Condition exists (0:r2=0 /\ 1:r2=0 /\ 2:r2=0) is NOT validated Observation 3.SB+membar.gl+po+po-p0:p1:p2-xgygzs Never 0 50000 Time 3.SB+membar.gl+po+po-p0:p1:p2-xgygzs 4.85 -- device info -- Name: GeForce GTX 750 Ti Major revision number: 5 Minor revision number: 0 Total global memory: 2147287040 Total shared memory per block: 49152 Number of multiprocessors: 5 Machine: formal Revision 1383, version 5.01 GPU_PTX experimental backend Command line: ./litmus -mach GTX750.cfg -a 8 -gpu-verf-comp true big_bucket/3.SB+membar.gl+po+po-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_50 */ /* 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: 5 */ /* gpu_min_ctas: 4 */ /* gpu_max_threads: 256 */ /* gpu_min_threads: 128 */