gunrock: CUDA 9: CUDA error 77: an illegal memory access was encountered

../../../gunrock_build/bin/bfs market /data/gunrock_dataset/large/indochina-2004/indochina-2004.mtx --src=0 --undirected --idempotence --queue-sizing=6.5 --in-sizing=4 --iteration-num=10 --direction-optimized --device=0 --traversal-mode=LB_CULL --do_a=100 --do_b=100 --jsondir=./eval/ > ./eval/indochina-2004.bfs.output.txt

[/home/mosama/gunrock/gunrock/app/enactor_loop.cuh, 781 @ gpu 1] FullQueue_Core failed. (CUDA error 77: an illegal memory access was encountered)
[/home/mosama/gunrock/tests/bfs/test_bfs.cu, 481 @ gpu 1] BFS Enact failed (CUDA error 77: an illegal memory access was encountered)

About this issue

Original URL
State: closed
Created 7 years ago
Comments: 18 (18 by maintainers)

Commits related to this issue

corrected *_sync() warper for cuda 9; issue #250 should be solved — committed to gunrock/gunrock by sgpyc 7 years ago
corrected *_sync() warper for cuda 9; issue #250 should be solved — committed to gunrock/gunrock by sgpyc 7 years ago
solved illegal memory access when using LB_CULL with V2V, issue #250 should be okey now — committed to gunrock/gunrock by sgpyc 7 years ago
solved illegal memory access when using LB_CULL with V2V, issue #250 should be okey now — committed to Laurawly/gunrock by sgpyc 7 years ago

Most upvoted comments

The sssp run output is here

sgpyc@mario:~/Projects/gunrock_master/tests/sssp$ ./bin/test_sssp_9.0_x86_64 market /data/gunrock_dataset/large/soc-LiveJournal1/soc-LiveJournal1.mtx --src=0 --undirected --iteration-num=10 --device=1 --traversal-mode=LB_CULL
Loading Matrix-market coordinate-formatted graph ...
  Reading directly from stored binary CSR arrays ...
Done reading (1s).

Degree Histogram (4847571 vertices, 85702474 edges):
    Degree   0: 962 (0.02%)
    Degree 2^0: 1057993 (21.83%)
    Degree 2^1: 919014 (18.96%)
    Degree 2^2: 833607 (17.20%)
    Degree 2^3: 742607 (15.32%)
    Degree 2^4: 616847 (12.72%)
    Degree 2^5: 404231 (8.34%)
    Degree 2^6: 184810 (3.81%)
    Degree 2^7: 62310 (1.29%)
    Degree 2^8: 19429 (0.40%)
    Degree 2^9: 4855 (0.10%)
    Degree 2^10: 704 (0.01%)
    Degree 2^11: 161 (0.00%)
    Degree 2^12: 29 (0.00%)
    Degree 2^13: 11 (0.00%)
    Degree 2^14: 1 (0.00%)

Source vertex: 0
Using 1 GPU(s): [ 1 ].
Using traversal mode LB_CULL
__________________________
--------------------------
iteration 0 elapsed: 420.994043 ms, src = 0, #iteration = 49
__________________________
--------------------------
iteration 1 elapsed: 398.401022 ms, src = 0, #iteration = 45
__________________________
--------------------------
iteration 2 elapsed: 395.196915 ms, src = 0, #iteration = 45
__________________________
--------------------------
iteration 3 elapsed: 408.931017 ms, src = 0, #iteration = 48
__________________________
--------------------------
iteration 4 elapsed: 407.134056 ms, src = 0, #iteration = 45
__________________________
--------------------------
iteration 5 elapsed: 428.952932 ms, src = 0, #iteration = 48
__________________________
--------------------------
iteration 6 elapsed: 420.816183 ms, src = 0, #iteration = 49
__________________________
--------------------------
iteration 7 elapsed: 407.245159 ms, src = 0, #iteration = 48
__________________________
--------------------------
iteration 8 elapsed: 412.925005 ms, src = 0, #iteration = 46
__________________________
--------------------------
iteration 9 elapsed: 435.800076 ms, src = 0, #iteration = 49
Computing reference value ...
CPU SSSP finished in 5711.313965 msec.


First 40 labels of the GPU result.
[0:0 1:1 2:1 3:2 4:3 5:3 6:2 7:2 8:7 9:3 10:2 11:2 12:2 13:1 14:3 15:7 16:5 17:2 18:2 19:1 20:2 21:2 22:3 23:5 24:6 25:31 26:1 27:7 28:7 29:3 30:7 31:4 32:1 33:3 34:2 35:4 36:3 37:6 38:2 39:2 ]
Label Validity: 
CORRECT
First 40 labels of the reference CPU result.
[0:0 1:1 2:1 3:2 4:3 5:3 6:2 7:2 8:7 9:3 10:2 11:2 12:2 13:1 14:3 15:7 16:5 17:2 18:2 19:1 20:2 21:2 22:3 23:5 24:6 25:31 26:1 27:7 28:7 29:3 30:7 31:4 32:1 33:3 34:2 35:4 36:3 37:6 38:2 39:2 ]

        Memory Usage(B)  #keys0,0        #keys0,1
GPU_1    834666496       5817087         5817087
         queue_sizing =  1.200000        1.200000

 [SSSP] finished.
 avg. elapsed: 413.6396 ms
 iterations: 49
 min. elapsed: 395.1969 ms
 max. elapsed: 435.8001 ms
 rate: 207.1643 MiEdges/s
 src: 0
 nodes_visited: 4843953
 edges_visited: 85691368
 nodes queued: 24580847
 edges queued: 457374007
 nodes redundance: 407.45%
 edges redundance: 433.75%
 load time: 1278.6779 ms
 preprocess time: 544.5180 ms
 postprocess time: 105.4420 ms
 total time: 22295.7089 ms

sgpyc on Nov 3, 2017