GPU_PTX 3.LB+membar.gl+data+addr-p0:p1:p2-xsyszg "Mgl-d-Rs-Ws Rf-w-s Data-d-s-Wg Rf-w-g Addr-d-g-Ws Rf-w-s" {} P0 | P1 | P2 ; ld.cg.s32 r0, [r1] | ld.cg.s32 r0, [r1] | ld.cg.s32 r0, [r1] ; membar.gl | and.b32 r2, r0 2147483648 | and.b32 r2, r0 2147483648 ; mov.s32 r2, 1 | add.s32 r2, r2 1 | cvt.u64.u32 r3, r2 ; st.cg.s32 [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) (warp P1) (warp P2)))) x:local; y:local; z:global 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] ; and.b32 sr21, sr01, 2147483648 ; add.s32 sr21, sr21, 1 ; st.cg.s32 [sr31], sr21 ; //START _litmus_P0 ld.cg.s32 sr00, [sr10] ; membar.gl ; mov.s32 sr20, 1 ; st.cg.s32 [sr30], sr20 ; Binary '/tmp/dir6209cc.tmp/3.LB+membar.gl+data+addr-p0:p1:p2-xsyszg-0.exe' successfully loaded Specification clusters: 3 Specification: [[(826, 3, 'R0'), (824, 2, 'R4')], [(900, 3, 'R0'), (898, 2, 'R12')], [(985, 3, 'R0'), (983, 2, 'R12')]] Cluster 0: OK Cluster 1: OK Cluster 2: OK !!SUCCESS!! Test 3.LB+membar.gl+data+addr-p0:p1:p2-xsyszg Allowed Histogram (6 states) 26298 :>0:r0=0; 1:r0=0; 2:r0=0; 3608 :>0:r0=1; 1:r0=0; 2:r0=0; 4164 :>0:r0=0; 1:r0=1; 2:r0=0; 2 :>0:r0=1; 1:r0=1; 2:r0=0; 15349 :>0:r0=0; 1:r0=0; 2:r0=1; 579 :>0:r0=1; 1:r0=0; 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.gl+data+addr-p0:p1:p2-xsyszg Never 0 50000 Time 3.LB+membar.gl+data+addr-p0:p1:p2-xsyszg 4.26 -- 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.LB+membar.gl+data+addr-p0:p1:p2-xsyszg-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 */