GeminiGraph
GeminiGraph copied to clipboard
thu-pacman
`numa_set_interleave_mask` problem
Hi,xiaowei, I'm trying to run pagerank on scale16_s.bin, which from here from issue #6 , but I get a segment fault, it looks like following. I think the numa_set_interleave_mask() function which was called in your init() function in file core/graph.hpp:174 causes this problem, because when I comment out this line code, the program run correctly , get no error and print the result.
Here is detail information: OS:Ubuntu 18.04 gcc version: 7.5.0 opemmp version : Open MPI 2.1.1 (Language: C++) slurm version: 17.11.2 cpu infomation:
Architecture: x86_64
CPU op-mode(s): 32-bit, 64-bit
Byte Order: Little Endian
CPU(s): 24
On-line CPU(s) list: 0-23
Thread(s) per core: 2
Core(s) per socket: 6
Socket(s): 2
NUMA node(s): 2
Vendor ID: GenuineIntel
CPU family: 6
Model: 62
Model name: Intel(R) Xeon(R) CPU E5-2620 v2 @ 2.10GHz
Stepping: 4
CPU MHz: 1200.597
CPU max MHz: 2600.0000
CPU min MHz: 1200.0000
BogoMIPS: 4200.13
Virtualization: VT-x
L1d cache: 32K
L1i cache: 32K
L2 cache: 256K
L3 cache: 15360K
NUMA node0 CPU(s): 0-5,12-17
NUMA node1 CPU(s): 6-11,18-23
Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc cpuid aperfmperf pni pclmulqdq dtes64 monitor ds_cpl vmx smx est tm2 ssse3 cx16 xtpr pdcm pcid dca sse4_1 sse4_2 x2apic popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm cpuid_fault epb pti ssbd ibrs ibpb stibp tpr_shadow vnmi flexpriority ept vpid fsgsbase smep erms xsaveopt dtherm ida arat pln pts md_clear flush_l1d
I use this command,
srun -N 2 ./toolkits/pagerank ./data/scale16_s.bin 65536 20
I have 1 node named ca33 as control machine and 2 nodes named ca97 and ca100 for computing, I get this error.
WARNING: There is at least non-excluded one OpenFabrics device found, but there are no active ports detected (or Open MPI was unable to use them). This is most certainly not what you wanted. Check your cables, subnet manager configuration, etc. The openib BTL will be ignored for this job. Local host: ca97
[ca100:15129] *** Process received signal *** [ca100:15129] Signal: Segmentation fault (11) [ca100:15129] Signal code: Address not mapped (1) [ca100:15129] Failing at address: (nil) [ca100:15129] [ 0] /lib/x86_64-linux-gnu/libpthread.so.0(+0x12980)[0x7f8e3c1c9980] [ca100:15129] [ 1] /usr/lib/x86_64-linux-gnu/libnuma.so.1(numa_bitmask_equal+0xd)[0x7f8e3d0b592d] [ca100:15129] [ 2] /usr/lib/x86_64-linux-gnu/libnuma.so.1(numa_set_interleave_mask+0x13)[0x7f8e3d0b64e3] [ca100:15129] [ 3] /home/yinwei/dev/GeminiGraph/./toolkits/pagerank(+0xd2f9)[0x5582c2c632f9] [ca100:15129] [ 4] /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xe7)[0x7f8e3bde7bf7] [ca100:15129] [ 5] /home/yinwei/dev/GeminiGraph/./toolkits/pagerank(+0xd8ba)[0x5582c2c638ba] [ca100:15129] *** End of error message *** [ca97:17078] *** Process received signal *** [ca97:17078] Signal: Segmentation fault (11) [ca97:17078] Signal code: Address not mapped (1) [ca97:17078] Failing at address: (nil) [ca97:17078] [ 0] /lib/x86_64-linux-gnu/libpthread.so.0(+0x12980)[0x7f0c137b7980] [ca97:17078] [ 1] /usr/lib/x86_64-linux-gnu/libnuma.so.1(numa_bitmask_equal+0xd)[0x7f0c146a392d] [ca97:17078] [ 2] /usr/lib/x86_64-linux-gnu/libnuma.so.1(numa_set_interleave_mask+0x13)[0x7f0c146a44e3] [ca97:17078] [ 3] /home/yinwei/dev/GeminiGraph/./toolkits/pagerank(+0xd2f9)[0x55739a4a72f9] [ca97:17078] [ 4] /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xe7)[0x7f0c133d5bf7] [ca97:17078] [ 5] /home/yinwei/dev/GeminiGraph/./toolkits/pagerank(+0xd8ba)[0x55739a4a78ba] [ca97:17078] *** End of error message *** srun: error: ca100: task 1: Segmentation fault srun: error: ca97: task 0: Segmentation fault (core dumped)
when I comment out this line core/graph.hpp:174, make and run again, it will get no error. i don't know why.
WARNING: There is at least non-excluded one OpenFabrics device found, but there are no active ports detected (or Open MPI was unable to use them). This is most certainly not what you wanted. Check your cables, subnet manager configuration, etc. The openib BTL will be ignored for this job. Local host: ca97
edge_data_size:0 unit_size:4 edge_unit_size:8 vertices size is 65536 edges size is 2728272 iterations is 20 delta(0)=1.000000 delta(1)=1.201973 delta(2)=0.299845 delta(3)=0.121878 delta(4)=0.053445 delta(5)=0.023630 delta(6)=0.010217 delta(7)=0.004646 delta(8)=0.001908 delta(9)=0.000956 delta(10)=0.000324 delta(11)=0.000227 delta(12)=0.000150 delta(13)=0.000127 delta(14)=0.000108 delta(15)=0.000092 delta(16)=0.000078 delta(17)=0.000066 delta(18)=0.000056 delta(19)=0.000048 exec_time=0.105598(s) pr_sum=47355.429527 pr[4]=1348.359586 delta(0)=1.000000 delta(1)=1.201973 delta(2)=0.299845 delta(3)=0.121878 delta(4)=0.053445 delta(5)=0.023630 delta(6)=0.010217 delta(7)=0.004646 delta(8)=0.001908 delta(9)=0.000956 delta(10)=0.000324 delta(11)=0.000227 delta(12)=0.000150 delta(13)=0.000127 delta(14)=0.000108 delta(15)=0.000092 delta(16)=0.000078 delta(17)=0.000066 delta(18)=0.000056 delta(19)=0.000048 exec_time=0.102557(s) pr_sum=47355.429527 pr[4]=1348.359586 delta(0)=1.000000 delta(1)=1.201973 delta(2)=0.299845 delta(3)=0.121878 delta(4)=0.053445 delta(5)=0.023630 delta(6)=0.010217 delta(7)=0.004646 delta(8)=0.001908 delta(9)=0.000956 delta(10)=0.000324 delta(11)=0.000227 delta(12)=0.000150 delta(13)=0.000127 delta(14)=0.000108 delta(15)=0.000092 delta(16)=0.000078 delta(17)=0.000066 delta(18)=0.000056 delta(19)=0.000048 exec_time=0.107026(s) pr_sum=47355.429527 pr[4]=1348.359586 delta(0)=1.000000 delta(1)=1.201973 delta(2)=0.299845 delta(3)=0.121878 delta(4)=0.053445 delta(5)=0.023630 delta(6)=0.010217 delta(7)=0.004646 delta(8)=0.001908 delta(9)=0.000956 delta(10)=0.000324 delta(11)=0.000227 delta(12)=0.000150 delta(13)=0.000127 delta(14)=0.000108 delta(15)=0.000092 delta(16)=0.000078 delta(17)=0.000066 delta(18)=0.000056 delta(19)=0.000048 exec_time=0.091182(s) pr_sum=47355.429527 pr[4]=1348.359586 delta(0)=1.000000 delta(1)=1.201973 delta(2)=0.299845 delta(3)=0.121878 delta(4)=0.053445 delta(5)=0.023630 delta(6)=0.010217 delta(7)=0.004646 delta(8)=0.001908 delta(9)=0.000956 delta(10)=0.000324 delta(11)=0.000227 delta(12)=0.000150 delta(13)=0.000127 delta(14)=0.000108 delta(15)=0.000092 delta(16)=0.000078 delta(17)=0.000066 delta(18)=0.000056 delta(19)=0.000048 exec_time=0.083132(s) pr_sum=47355.429527 pr[4]=1348.359586 delta(0)=1.000000 delta(1)=1.201973 delta(2)=0.299845 delta(3)=0.121878 delta(4)=0.053445 delta(5)=0.023630 delta(6)=0.010217 delta(7)=0.004646 delta(8)=0.001908 delta(9)=0.000956 delta(10)=0.000324 delta(11)=0.000227 delta(12)=0.000150 delta(13)=0.000127 delta(14)=0.000108 delta(15)=0.000092 delta(16)=0.000078 delta(17)=0.000066 delta(18)=0.000056 delta(19)=0.000048 exec_time=0.077558(s) pr_sum=47355.429527 pr[4]=1348.359586 edge_data_size:0 unit_size:4 edge_unit_size:8 vertices size is 65536 edges size is 2728272 iterations is 20 delta(0)=1.000000 delta(1)=1.201973 delta(2)=0.299845 delta(3)=0.121878 delta(4)=0.053445 delta(5)=0.023630 delta(6)=0.010217 delta(7)=0.004646 delta(8)=0.001908 delta(9)=0.000956 delta(10)=0.000324 delta(11)=0.000227 delta(12)=0.000150 delta(13)=0.000127 delta(14)=0.000108 delta(15)=0.000092 delta(16)=0.000078 delta(17)=0.000066 delta(18)=0.000056 delta(19)=0.000048 exec_time=0.098069(s) pr_sum=47355.429527 pr[4]=1348.359586 delta(0)=1.000000 delta(1)=1.201973 delta(2)=0.299845 delta(3)=0.121878 delta(4)=0.053445 delta(5)=0.023630 delta(6)=0.010217 delta(7)=0.004646 delta(8)=0.001908 delta(9)=0.000956 delta(10)=0.000324 delta(11)=0.000227 delta(12)=0.000150 delta(13)=0.000127 delta(14)=0.000108 delta(15)=0.000092 delta(16)=0.000078 delta(17)=0.000066 delta(18)=0.000056 delta(19)=0.000048 exec_time=0.113920(s) pr_sum=47355.429527 pr[4]=1348.359586 delta(0)=1.000000 delta(1)=1.201973 delta(2)=0.299845 delta(3)=0.121878 delta(4)=0.053445 delta(5)=0.023630 delta(6)=0.010217 delta(7)=0.004646 delta(8)=0.001908 delta(9)=0.000956 delta(10)=0.000324 delta(11)=0.000227 delta(12)=0.000150 delta(13)=0.000127 delta(14)=0.000108 delta(15)=0.000092 delta(16)=0.000078 delta(17)=0.000066 delta(18)=0.000056 delta(19)=0.000048 exec_time=0.087034(s) pr_sum=47355.429527 pr[4]=1348.359586 delta(0)=1.000000 delta(1)=1.201973 delta(2)=0.299845 delta(3)=0.121878 delta(4)=0.053445 delta(5)=0.023630 delta(6)=0.010217 delta(7)=0.004646 delta(8)=0.001908 delta(9)=0.000956 delta(10)=0.000324 delta(11)=0.000227 delta(12)=0.000150 delta(13)=0.000127 delta(14)=0.000108 delta(15)=0.000092 delta(16)=0.000078 delta(17)=0.000066 delta(18)=0.000056 delta(19)=0.000048 exec_time=0.091063(s) pr_sum=47355.429527 pr[4]=1348.359586 delta(0)=1.000000 delta(1)=1.201973 delta(2)=0.299845 delta(3)=0.121878 delta(4)=0.053445 delta(5)=0.023630 delta(6)=0.010217 delta(7)=0.004646 delta(8)=0.001908 delta(9)=0.000956 delta(10)=0.000324 delta(11)=0.000227 delta(12)=0.000150 delta(13)=0.000127 delta(14)=0.000108 delta(15)=0.000092 delta(16)=0.000078 delta(17)=0.000066 delta(18)=0.000056 delta(19)=0.000048 exec_time=0.093679(s) pr_sum=47355.429527 pr[4]=1348.359586 delta(0)=1.000000 delta(1)=1.201973 delta(2)=0.299845 delta(3)=0.121878 delta(4)=0.053445 delta(5)=0.023630 delta(6)=0.010217 delta(7)=0.004646 delta(8)=0.001908 delta(9)=0.000956 delta(10)=0.000324 delta(11)=0.000227 delta(12)=0.000150 delta(13)=0.000127 delta(14)=0.000108 delta(15)=0.000092 delta(16)=0.000078 delta(17)=0.000066 delta(18)=0.000056 delta(19)=0.000048 exec_time=0.079862(s) pr_sum=47355.429527 pr[4]=1348.359586
Hi @iosmers , have you tried running the program via mpirun with "--bind-to none"?
