parsec
parsec copied to clipboard
ICLDisco
PaRSEC is a generic framework for architecture aware scheduling and management of micro-tasks on distributed, GPU accelerated, many-core heterogeneous architectures. PaRSEC assigns computation threads...
Be more explicit when we fail to allocate memory: display values that led to the failure to help debug.
Reenable detection of AVX capabilities even if compiler would not generate vectorized AVX code (because libraries like mkl can still use it). With the current status, the AVX availability on...
## Describe the bug parsec_ce is not initialized by default when we create the taskpool. It is only initialized lazily after the context is started. When enabling dynamic termination detection...
In some cases CPU SYRKs would execute when there is a valid GPU hook, because the CPU load (which is never updated hence is always having an eta of 0+time_estimate)...
Coverity identified 692 legitimate issues with the node. Luckily for us most of them are in the generated code and have relatively simple solutions. This PR is a placeholder to...
When writing in PTG like this: `READ D super.rank_of(&desc_f_data->super, 0, n); %})` It shows: `#error Expression return desc_f_data->super.rank_of(&desc_f_data->super, 0, n); has not been generated` @therault suggested using a task local...
**[Original report](https://bitbucket.org/icldistcomp/parsec/issue/242) by Qinglei Cao (Bitbucket: [Qinglei_Cao](https://bitbucket.org/Qinglei_Cao), ).** ---------------------------------------- If a taskpool runs on GPU, it seems that status on GPU \(some GPU data\_copy\_t, load between devices\) is not reset...
**[Original report](https://bitbucket.org/icldistcomp/parsec/issue/271) by Omri Mor (Bitbucket: [omrimor2](https://bitbucket.org/omrimor2), GitHub: [omor1](https://github.com/omor1)).** ---------------------------------------- ```c diag = /* 0 if includes diagonal, 1 otherwise */ if (uplo == matrix_Upper) { nmax = n-diag; for...