PaStiX issueshttps://gitlab.inria.fr/solverstack/pastix/-/issues2024-03-07T15:53:13+01:00https://gitlab.inria.fr/solverstack/pastix/-/issues/68Base and int64 at the same time2024-03-07T15:53:13+01:00Rafal brzegowyBase and int64 at the same timeQuestion: are there any contraindications to building and maintaining the base and int64 versions (mingw/msys2)? Can I expect any conflict (that I have regulard and int64 versions installed at the same time)?
Like openblas example int64...Question: are there any contraindications to building and maintaining the base and int64 versions (mingw/msys2)? Can I expect any conflict (that I have regulard and int64 versions installed at the same time)?
Like openblas example int64:
https://packages.msys2.org/package/mingw-w64-x86_64-openblas64?repo=mingw64
and base:
https://packages.msys2.org/package/mingw-w64-x86_64-openblas?repo=mingw64
I plan to build it like openblas:
https://github.com/msys2/MINGW-packages/blob/23895685131b3ec659324fc66b4541c4e11d42ac/mingw-w64-openblas/PKGBUILDhttps://gitlab.inria.fr/solverstack/pastix/-/issues/67Not issue rather a request for advice/help2024-01-15T15:59:55+01:00Rafal brzegowyNot issue rather a request for advice/helpMay I ask for advice/help on how to modify the CalculiX file: https://github.com/Dhondtguido/CalculiX/blob/master/src/pastix.c so that ccx it works with the main version of PaStiX (with the changes introduced: https://gitlab.inria.fr/pas...May I ask for advice/help on how to modify the CalculiX file: https://github.com/Dhondtguido/CalculiX/blob/master/src/pastix.c so that ccx it works with the main version of PaStiX (with the changes introduced: https://gitlab.inria.fr/pastix/pastix/-/merge_requests?scope=all&state=merged&label_name%5B%5D=CalculiX) instead the modified version (based on a much older version: https://github.com/Dhondtguido/PaStiX4CalculiX)PaStiX 6.4.0https://gitlab.inria.fr/solverstack/pastix/-/issues/66Error handling when factorization fail2023-12-11T11:55:33+01:00MARAIT GillesError handling when factorization failAt the moment when factorization fail, pastix_print_error prints an error message and calls `exit(1)`, stopping the program.
For the calling code it can be a problem, because no error handling can be done. In particular, when the Choles...At the moment when factorization fail, pastix_print_error prints an error message and calls `exit(1)`, stopping the program.
For the calling code it can be a problem, because no error handling can be done. In particular, when the Cholesky factorization fails, it's impossible to recover from it, when it could be convenient to be able to handle it and try an LDLT factorization instead (after all Cholesky factorization can be used as a way to check if a matrix is SPD).
https://gitlab.inria.fr/solverstack/pastix/-/blob/master/kernels/core_zpotrfsp.c?ref_type=heads#L88
Would it be possible to have the factorization function return an error instead ? Or is there a hidden way to handle this kind of error ?PaStiX 6.4.0https://gitlab.inria.fr/solverstack/pastix/-/issues/62Re-enable Julia in the CI2023-12-11T11:58:43+01:00Mathieu FavergeRe-enable Julia in the CIDue to image upgrade, the Julia implementation could not work anymore. The manifest files should be upgraded and the CI restored.Due to image upgrade, the Julia implementation could not work anymore. The manifest files should be upgraded and the CI restored.https://gitlab.inria.fr/solverstack/pastix/-/issues/58error: conflicting types for 'MPI_User_function'2023-12-11T20:30:54+01:00RAMET Pierreerror: conflicting types for 'MPI_User_function'My name is Hannes Panzer and I am a research associate at the iwb of the Technical University of Munich.
In my research team, we are performing numerical simulations based on the finite element simulation tool CalculiX. Recently, we wan...My name is Hannes Panzer and I am a research associate at the iwb of the Technical University of Munich.
In my research team, we are performing numerical simulations based on the finite element simulation tool CalculiX. Recently, we wanted to install the newest version (2.19) of the tool, along with PaStiX as the Solver, to increase the computational speed by performing calculations on the GPU. However, so far we were not able to install PaStiX without any error message, which is the following in our case:
error: conflicting types for 'MPI_User_function'
There are many similar error messages concerning MPI, even though the option MPI is turned off in the make file.
We wanted to ask you, if you have already encountered any error messages of this type when trying to compile PaStiX, and if you have an idea, what causes the problem and how it could be solved.RAMET PierreRAMET Pierrehttps://gitlab.inria.fr/solverstack/pastix/-/issues/56Solve with L / Mutiplication with L (Cholesky factorization)2021-05-27T16:00:13+02:00DURUFLE MarcSolve with L / Mutiplication with L (Cholesky factorization)Hi,
in the previous releases of pastix (5.xx), it was possible to solve independently
```
L x = b
or
L^T x = b
```
for the Cholesky factorisation. It was also possible to perform a multiplication by the factor L (or L^T) with the followi...Hi,
in the previous releases of pastix (5.xx), it was possible to solve independently
```
L x = b
or
L^T x = b
```
for the Cholesky factorisation. It was also possible to perform a multiplication by the factor L (or L^T) with the following parameters :
```c
iparm[IPARM_TRANSPOSE_SOLVE] = API_SOLVE_BACKWARD_ONLY;
iparm[IPARM_TRANSPOSE_SOLVE] = API_SOLVE_FORWARD_ONLY;
iparm[IPARM_TRANSPOSE_SOLVE] = API_SOLVE_UTRMV;
iparm[IPARM_TRANSPOSE_SOLVE] = API_SOLVE_LTRMV;
```
How is it possible to perform these tasks with the new version of pastix (6.xx) ?
Thanks for your help.https://gitlab.inria.fr/solverstack/pastix/-/issues/55Free csc user for the distributed case2021-05-26T11:02:52+02:00DURUFLE MarcFree csc user for the distributed caseHi,
I would like to free the csc matrix during the factorization, I write :
iparm[IPARM_FREE_CSCUSER] = 1;
It works fine for the sequential version. However, for the distributed version, it does not work
because pastix_task_solve needs...Hi,
I would like to free the csc matrix during the factorization, I write :
iparm[IPARM_FREE_CSCUSER] = 1;
It works fine for the sequential version. However, for the distributed version, it does not work
because pastix_task_solve needs the array loc2glob which has been freed by the previous tasks. Is there a way
to avoid freeing this array such that the solve can be completed correctly ?
In terms of memory storage, when I have a moderate fill-in (for instance around 2), this option is interesting ? I was wondering if the arrays are freed before allocating more space for L factor because of fill-in.https://gitlab.inria.fr/solverstack/pastix/-/issues/46Performance model and memory problems on GPU2019-11-25T14:11:25+01:00Ghost UserPerformance model and memory problems on GPUDear developers,
first of all, thank you very much developing such a great solver.
In my current project I am using PastiX' hybrid GPU mode through PaRSEC. It works great and the speedup is really showing. However, there are currently ...Dear developers,
first of all, thank you very much developing such a great solver.
In my current project I am using PastiX' hybrid GPU mode through PaRSEC. It works great and the speedup is really showing. However, there are currently two problems I am facing:
1. A default performance model and cost function is provided for CPU and GPU (common/model.c) but I wonder how to reasonably determine coefficients for my target hardware? I found the CMake-option "PASTIX_GENERATE_MODEL", but can't figure out whether i can turn that data into coefficients.
2. When factorizing some of my larger matrices and choosing a larger block size at the same time, PaRSEC returns illegal memory access errors during factorization. I assume this is happens because more memory than available is requested from the GPU. Do you possibly know a solution for this problem or are you aware of this problem?
Best regards,
Peter