Closed juncaipeng closed 4 years ago
Maybe timeout is the same problem in https://github.com/PaddlePaddle/Paddle/pull/20162#issuecomment-554262103
The cpu information of the test machine obtained by lscpu:
flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush
mmx fxsr sse sse2 ss syscall nx pdpe1gb rdtscp lm constant_tsc arch_perfmon rep_good
nopl eagerfpu pni pclmulqdq ssse3 fma cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt
tsc_deadline_timer aes xsave avx f16c rdrand hypervisor lahf_lm abm 3dnowprefetch arat
fsgsbase tsc_adjust bmi1 hle avx2 smep bmi2 erms invpcid rtm mpx avx512f rdseed adx
smap avx512cd xsaveopt xsavec xgetbv1
@grygielski It seems that this machine doesn't have the right AVX512 flages needed for mkldnn 1.0 int8. could you help confirm this?
It definitely doesn't have VNNI flags so best implementations won't be chosen. However, there are avx512 flags so it shouldn't timeout so easily. It can be a problem similar to earlier ones with not propagating these flags. Is it virtual machine? Verbose logs would help here. @juncaipeng Could you run one of these tests with MKLDNN_VERBOSE=2 env. variable? If you are using bash just type this before test command. For example: MKLDNN_VERBOSE=2 ctest -R ...
Linux_Python27
http://ci.paddlepaddle.org/viewLog.html?buildId=243540&buildTypeId=Paddle_PrCiNight
http://ci.paddlepaddle.org/viewLog.html?buildId=243543&buildTypeId=Paddle_PrCiNight&tab=buildLog
http://ci.paddlepaddle.org/viewLog.html?buildId=243542&buildTypeId=Paddle_PrCiNight
http://ci.paddlepaddle.org/viewLog.html?buildId=243542&buildTypeId=Paddle_PrCiNight
PR_CI_Manylinux_Coverage_CPU
http://ci.paddlepaddle.org/viewLog.html?buildId=243523&buildTypeId=Paddle_PaddleManylinux_PrCiManylinuxCoverageCpu