-
-
Notifications
You must be signed in to change notification settings - Fork 5.5k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
unknown machine / CPUID related OpenBLAS issue #18692
Comments
Is this a virtual machine? From the error, it appears that the CPUID may be lying about what instructions it supports and angering OpenBLAS |
This is not a virtual machine. it's a compute cluster running Centos 5.5, with Rocks cluster management. So, I was under the impression this is a memory issue? |
What CPU / what's |
this is issue #18701 . |
Did you delete your response? From the |
yes i did delete it! sorry. julia> versioninfo()
Julia Version 0.5.0
Commit ee8dca9* (2016-09-24 19:29 UTC)
Platform Info:
System: Linux (x86_64-pc-linux-gnu)
CPU: AMD Opteron (tm) Processor 875
WORD_SIZE: 64
BLAS: libopenblas (USE64BITINT NO_AFFINITY OPTERON)
LAPACK: libopenblas64_
LIBM: libopenlibm
LLVM: libLLVM-3.7.1 (ORCJIT, k8-sse3) |
I'm still wrestling with julia running up against a virtual memory limit imposed by a SGE scheduler. I built julia-0.5 with the fix suggested in this issue comment. I requested 16GB of virtual and 16GB of total memory from the scheduler. From looking at
top
previously, I know that this job consumes 2GB of RES and 6482MB of VIRT memory on this machine, so should fit comfortably within those limits?What does this mean?
#define REGION_PG_COUNT 8*4096
insrc/gc-pages.c
too restrictive now?I now ran the unit tests and get the same behaviour.
The text was updated successfully, but these errors were encountered: