GPU_PTX WRW+2W+membar.gl+membar.cta-p0:p1:p2-xsys "Rf-w-s Mgl-d-Rs-Ws Ws-w-s Mcta-d-Ws-Ws Ws-w-s" {} P0 | P1 | P2 ; mov.s32 r0, 2 | ld.cg.s32 r0, [r1] | mov.s32 r0, 2 ; st.cg.s32 [r1], r0 | membar.gl | st.cg.s32 [r1], r0 ; | mov.s32 r2, 1 | membar.cta ; | st.cg.s32 [r3], r2 | mov.s32 r2, 1 ; | | st.cg.s32 [r3], r2 ; exists (0::x=2 /\ 1::y=2 /\ 1:r0=2) (device (kernel (cta (warp P0) (warp P1) (warp P2)))) x:local; y:local Generated assembler //START _litmus_P2 mov.s32 sr02, 2 ; st.cg.s32 [sr12], sr02 ; membar.cta ; mov.s32 sr22, 1 ; st.cg.s32 [sr32], sr22 ; //START _litmus_P1 ld.cg.s32 sr01, [sr11] ; membar.gl ; mov.s32 sr21, 1 ; st.cg.s32 [sr31], sr21 ; //START _litmus_P0 mov.s32 sr00, 2 ; st.cg.s32 [sr10], sr00 ; Binary '/tmp/dira45aaf.tmp/WRW+2W+membar.gl+membar.cta-p0:p1:p2-xsys-0.exe' successfully loaded Specification clusters: 3 Specification: [[(553, 2, 'R10'), (555, 2, 'R12')], [(617, 3, 'R4'), (615, 2, 'R12')], [(667, 2, 'R10')]] Cluster 0: OK Cluster 1: OK Cluster 2: OK !!SUCCESS!! Test WRW+2W+membar.gl+membar.cta-p0:p1:p2-xsys Allowed Histogram (9 states) 18266 :>1:r0=0; 0::x=1; 1::y=1; 3230 :>1:r0=1; 0::x=1; 1::y=1; 10945 :>1:r0=2; 0::x=1; 1::y=1; 6480 :>1:r0=0; 0::x=2; 1::y=1; 2043 :>1:r0=1; 0::x=2; 1::y=1; 1394 :>1:r0=2; 0::x=2; 1::y=1; 6482 :>1:r0=0; 0::x=1; 1::y=2; 1089 :>1:r0=2; 0::x=1; 1::y=2; 71 :>1:r0=0; 0::x=2; 1::y=2; No Witnesses Positive: 0, Negative: 50000 Condition exists (0::x=2 /\ 1::y=2 /\ 1:r0=2) is NOT validated Observation WRW+2W+membar.gl+membar.cta-p0:p1:p2-xsys Never 0 50000 Time WRW+2W+membar.gl+membar.cta-p0:p1:p2-xsys 12.02 -- device info -- Name: Tesla C2075 Major revision number: 2 Minor revision number: 0 Total global memory: 5636554752 Total shared memory per block: 49152 Number of multiprocessors: 14 Machine: fractus Revision 1365M, version 5.01 GPU_PTX experimental backend Command line: ./litmus -mach TeslaC2075.cfg -a 8 -gpu-verf-comp true big_bucket2/WRW+2W+membar.gl+membar.cta-p0:p1:p2-xsys-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_20 */ /* 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: 12 */ /* gpu_min_ctas: 9 */ /* gpu_max_threads: 512 */ /* gpu_min_threads: 128 */