Updates for NekRS v25 #1257
Updates for NekRS v25 #1257
Conversation
|
Job Precheck, step Clang format on 193f563 wanted to post the following: Your code requires style changes. A patch was auto generated and copied here
Alternatively, with your repository up to date and in the top level of your repository:
|
|
|
||
| freePointer(nrs->usrwrk); | ||
| nrs->o_usrwrk.free(); | ||
| freePointer(usrwrk); |
There was a problem hiding this comment.
now that we're creating additional arrays to manage data NekRS used to, we'll also want to free them here/somewhere equivalent
| nrs_t * nrs = (nrs_t *)nrsPtr(); | ||
| mesh_t * mesh = temperatureMesh(); | ||
|
|
||
| const auto sid = nrs->scalar->nameToIndex.find("temperature")->second; |
There was a problem hiding this comment.
out of curiosity, would the temperature scalar ID ever be different from 0, if temperature exists in the problem? If the scalars now have arbitrary ordering this will be important to know since it'll impact a lot of other places in the code where we'd need the same kind of index usage as here
| auto vc = _nek_mesh->volumeCoupling(); | ||
| int id = vc.element[elem_id] * mesh->Np; | ||
|
|
||
| auto [x, y, z] = nekrs::host_xyz(); |
There was a problem hiding this comment.
do we need host_xyz(), or could we get away with just the host_x, host_y, and host_z functions?
| case field::velocity_y_squared: | ||
| case field::velocity_z_squared: | ||
| n_values = n * nrs->NVfields; | ||
| n_values = n * 3; //nVFields |
There was a problem hiding this comment.
did NVfields get deleted? not an issue, just asking for my understanding
|
Thanks @nandu90 , this is a great start. I think the overall design will be compatible with the rest of Cardinal - can you point me to where the host-side solution fields are being updated from the device-side solutions? I may have missed it or it hasn't been added yet to the draft PR (in which case we'll need to add it probably in |
3 participants
Work in progress