Skip to content
This repository has been archived by the owner on Mar 20, 2023. It is now read-only.

New allocators with sensible names and more #872

Draft
wants to merge 6 commits into
base: master
Choose a base branch
from

Conversation

ohm314
Copy link
Contributor

@ohm314 ohm314 commented Oct 24, 2022

Description

(WIP)

Fixes #806

How to test this?

check out nrn, change to this branch in coreneuron submodule, build and test as per nrn repo instructions

Test System

  • OS: macos (for now)
  • Compiler: lang 13
  • Backend: CPU (for now)

Use certain branches in CI pipelines.

CI_BRANCHES:NEURON_BRANCH=master,NMODL_BRANCH=master,SPACK_BRANCH=develop

Omar Awile added 4 commits October 12, 2022 16:40
work in progress - some tests fail
Reintroduce MemoryManaged class such that things can be moved slowly to
new allocators
@bbpbuildbot
Copy link
Collaborator

Logfiles from GitLab pipeline #80445 (:no_entry:) have been uploaded here!

Status and direct links:

Omar Awile added 2 commits October 24, 2022 14:47
UnifiedMemManaged and HostMemManaged now have overridden new and delete
operators so that these operations happen correctly behind the scenes.

NrnThread now inherits from UnifiedMemManaged.

Also fixed a bunch of minor issues
@bbpbuildbot
Copy link
Collaborator

Logfiles from GitLab pipeline #80508 (:no_entry:) have been uploaded here!

Status and direct links:

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

CoreNEURON memory allocation routines are confusing
2 participants