GPU_PTX Z6.0+po+po+membar.cta-p0:p1:p2-xgyszs "Po-d-Wg-Ws Rf-w-s Po-d-Rs-Ws Ws-w-s Mcta-d-Ws-Rg Fr-w-g" {} P0 | P1 | P2 ; mov.s32 r0, 1 | ld.cg.s32 r0, [r1] | mov.s32 r0, 2 ; st.cg.s32 [r1], r0 | mov.s32 r2, 1 | st.cg.s32 [r1], r0 ; mov.s32 r2, 1 | st.cg.s32 [r3], r2 | membar.cta ; st.cg.s32 [r3], r2 | | ld.cg.s32 r2, [r3] ; exists (1::z=2 /\ 1:r0=1 /\ 2:r2=0) (device (kernel (cta (warp P0) (warp P1) (warp P2)))) x:global; y:local; z:local Generated assembler //START _litmus_P2 mov.s32 sr02, 2 ; st.cg.s32 [sr12], sr02 ; membar.cta ; ld.cg.s32 sr22, [sr32] ; //START _litmus_P1 ld.cg.s32 sr01, [sr11] ; mov.s32 sr21, 1 ; st.cg.s32 [sr31], sr21 ; //START _litmus_P0 mov.s32 sr00, 1 ; st.cg.s32 [sr10], sr00 ; mov.s32 sr20, 1 ; st.cg.s32 [sr30], sr20 ; Binary '/tmp/dire85d71.tmp/Z6.0+po+po+membar.cta-p0:p1:p2-xgyszs-0.exe' successfully loaded Specification clusters: 3 Specification: [[(822, 2, 'R10'), (824, 3, 'R0')], [(909, 3, 'R0'), (907, 2, 'R12')], [(984, 2, 'R10'), (986, 2, 'R12')]] Cluster 0: OK Cluster 1: OK Cluster 2: OK !!SUCCESS!! Test Z6.0+po+po+membar.cta-p0:p1:p2-xgyszs Allowed Histogram (8 states) 5073 :>1:r0=0; 2:r2=0; 1::z=1; 1531 :>1:r0=1; 2:r2=0; 1::z=1; 9914 :>1:r0=0; 2:r2=1; 1::z=1; 12390 :>1:r0=1; 2:r2=1; 1::z=1; 1780 :>1:r0=0; 2:r2=0; 1::z=2; 8 *>1:r0=1; 2:r2=0; 1::z=2; 14295 :>1:r0=0; 2:r2=1; 1::z=2; 5009 :>1:r0=1; 2:r2=1; 1::z=2; Ok Witnesses Positive: 8, Negative: 49992 Condition exists (1::z=2 /\ 1:r0=1 /\ 2:r2=0) is validated Observation Z6.0+po+po+membar.cta-p0:p1:p2-xgyszs Sometimes 8 49992 Time Z6.0+po+po+membar.cta-p0:p1:p2-xgyszs 4.02 -- 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/Z6.0+po+po+membar.cta-p0:p1:p2-xgyszs-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 */