GPU_PTX WRW+2W+addr+membar.cta-p0::p1:p2-xgyg "Rf-c-g Addr-d-g-Wg Ws-w-g Mcta-d-Wg-Wg Ws-c-g" {} P0 | P1 | P2 ; mov.s32 r0, 2 | ld.cg.s32 r0, [r1] | mov.s32 r0, 2 ; st.cg.s32 [r1], r0 | and.b32 r2, r0 2147483648 | st.cg.s32 [r1], r0 ; | cvt.u64.u32 r3, r2 | membar.cta ; | mov.s32 r4, 1 | mov.s32 r2, 1 ; | add.u64 r5, r5 r3 | st.cg.s32 [r3], r2 ; | st.cg.s32 [r5], r4 | ; exists (0::x=2 /\ 1::y=2 /\ 1:r0=2) (device (kernel (cta (warp P0)) (cta (warp P1) (warp P2)))) x:global; y:global 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] ; and.b32 sr21, sr01, 2147483648 ; cvt.u64.u32 sr31, sr21 ; mov.s32 sr41, 1 ; add.u64 sr51, sr51, sr31 ; st.cg.s32 [sr51], sr41 ; //START _litmus_P0 mov.s32 sr00, 2 ; st.cg.s32 [sr10], sr00 ; Binary '/tmp/dire3ddb6.tmp/WRW+2W+addr+membar.cta-p0::p1:p2-xgyg-0.exe' successfully loaded Specification clusters: 3 Specification: [[(552, 2, 'R8'), (555, 2, 'R12')], [(605, 3, 'R4'), (603, 2, 'R8')], [(652, 2, 'R8')]] Cluster 0: OK Cluster 1: OK Cluster 2: OK !!SUCCESS!! Test WRW+2W+addr+membar.cta-p0::p1:p2-xgyg Allowed Histogram (9 states) 19021 :>1:r0=0; 0::x=1; 1::y=1; 3075 :>1:r0=1; 0::x=1; 1::y=1; 13510 :>1:r0=2; 0::x=1; 1::y=1; 7417 :>1:r0=0; 0::x=2; 1::y=1; 1625 :>1:r0=1; 0::x=2; 1::y=1; 1232 :>1:r0=2; 0::x=2; 1::y=1; 3293 :>1:r0=0; 0::x=1; 1::y=2; 643 :>1:r0=2; 0::x=1; 1::y=2; 184 :>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+addr+membar.cta-p0::p1:p2-xgyg Never 0 50000 Time WRW+2W+addr+membar.cta-p0::p1:p2-xgyg 17.85 -- 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/WRW+2W+addr+membar.cta-p0::p1:p2-xgyg-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 */