GPU_PTX W+RWC+membar.cta+ctrl+po-p0:p1:p2-xgyszg "Mcta-d-Wg-Ws Rf-w-s Ctrl-d-s-Rg Fr-w-g Po-d-Wg-Rg Fr-w-g" {} P0 | P1 | P2 ; mov.s32 r0, 1 | ld.cg.s32 r0, [r1] | mov.s32 r0, 1 ; st.cg.s32 [r1], r0 | and.b32 r2, r0 2147483648 | st.cg.s32 [r1], r0 ; membar.cta | setp.eq.b32 p2, r2, 0 | ld.cg.s32 r2, [r3] ; mov.s32 r2, 1 | @p2 bra LC00 | ; st.cg.s32 [r3], r2 | mov.s64 r4, 3494 | ; | LC00: | ; | ld.cg.s32 r3, [r4] | ; exists (1:r0=1 /\ 1:r3=0 /\ 2:r2=0) (device (kernel (cta (warp P0) (warp P1) (warp P2)))) x:global; y:local; z:global Generated assembler //START _litmus_P2 mov.s32 sr02, 1 ; st.cg.s32 [sr12], sr02 ; ld.cg.s32 sr22, [sr32] ; //START _litmus_P1 ld.cg.s32 sr01, [sr11] ; and.b32 sr21, sr01, 2147483648 ; setp.eq.b32 sp21, sr21, 0 ; @sp21 bra LC00 ; mov.s64 sr41, 3494 ; LC00: ld.cg.s32 sr31, [sr41] ; //START _litmus_P0 mov.s32 sr00, 1 ; st.cg.s32 [sr10], sr00 ; membar.cta ; mov.s32 sr20, 1 ; st.cg.s32 [sr30], sr20 ; Binary '/tmp/dir7f23b4.tmp/W+RWC+membar.cta+ctrl+po-p0:p1:p2-xgyszg-0.exe' successfully loaded Specification clusters: 3 Specification: [[(618, 2, 'R10'), (621, 3, 'R4')], [(679, 3, 'R4'), (680, 3, 'R5')], [(740, 2, 'R10'), (743, 2, 'R12')]] Cluster 0: OK Cluster 1: OK Cluster 2: OK !!SUCCESS!! Test W+RWC+membar.cta+ctrl+po-p0:p1:p2-xgyszg Allowed Histogram (7 states) 1669 :>1:r0=0; 1:r3=0; 2:r2=0; 12922 :>1:r0=0; 1:r3=1; 2:r2=0; 957 :>1:r0=1; 1:r3=1; 2:r2=0; 14490 :>1:r0=0; 1:r3=0; 2:r2=1; 114 :>1:r0=1; 1:r3=0; 2:r2=1; 16367 :>1:r0=0; 1:r3=1; 2:r2=1; 3481 :>1:r0=1; 1:r3=1; 2:r2=1; No Witnesses Positive: 0, Negative: 50000 Condition exists (1:r0=1 /\ 1:r3=0 /\ 2:r2=0) is NOT validated Observation W+RWC+membar.cta+ctrl+po-p0:p1:p2-xgyszg Never 0 50000 Time W+RWC+membar.cta+ctrl+po-p0:p1:p2-xgyszg 8.47 -- device info -- Name: GeForce GT 540M Major revision number: 2 Minor revision number: 1 Total global memory: 1072889856 Total shared memory per block: 49152 Number of multiprocessors: 2 Machine: DeepThought3 Revision 1371, version 5.01 GPU_PTX experimental backend Command line: ./litmus -mach GTX540m.cfg -a 8 -gpu-verf-comp true big_bucket/W+RWC+membar.cta+ctrl+po-p0:p1:p2-xgyszg-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_21 */ /* 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: 4 */ /* gpu_min_ctas: 4 */ /* gpu_max_threads: 128 */ /* gpu_min_threads: 128 */