Hello,
I am trying to spawn multiple threads to get few equation solver running at the same time. For example I have 50 set of matrix to be solved. I spawn 6 threads. Each thread is going to run pardiso in different phases (analysis, factorize and solve) but each thread is not necessarily going to finish analysis, factorize and solve before the next thread starts analysis. The code keep crashing inside pardiso call. Intel inspector with "Theading Error Analysis/Locate Dead Locks and Data Races" points to the problem of pardiso. The sequence of analysis, I am doing could be....
thread#1, analysis, thread#2, analysis, thread#1 factorize, thread#3 analysis, thread#2 analysis, thread#1 solve .... you get my point.
If I can not solve this with my executable at current situation, can you please point in the direction of a workaround so that I can run multiple equation solver from my executable?
Sukla
链接已复制
Hi Sukla,
Thanks for posting in Intel communities.
Could you please provide us with the following details so that we can look into your issue.
- Sample reproducer of the code
- Steps to reproduce (if any)
- Environment details (Operating System, architecture, etc.)
Thanks and Regards,
Praneeth Achanta