Skip to content

Provide template for an efficient private RAM #379

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Draft
wants to merge 1 commit into
base: main
Choose a base branch
from

Conversation

mandolaerik
Copy link
Contributor

@mandolaerik mandolaerik commented Jul 3, 2025

I did this mostly as an experiment, needs more testing and discussions before it can be added for real.

The idea is to allow reasonably easy access to a private RAM with good performance.

@syssimics syssimics requested a review from lwaern-intel July 3, 2025 20:51
@mandolaerik mandolaerik marked this pull request as draft July 3, 2025 20:54
Comment on lines +1417 to +1425
template ram {
param size;
connect ram is (init_as_subobj, destroy) {
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

A better design is probably to create a template for a private RAM connect, without DMI support, and then create a separate template which adds DMI magic to all private RAMs within a bank/subdev/port

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Oh, this adds some pain: the use of get_user_data requires that you have one separate direct_memory_update ports for each RAM; if we want a shared impl, then we instead need to keep a private hashtable from handles to pages.

@syssimics
Copy link
Contributor

PR Verification: ❌ failure

@mandolaerik mandolaerik force-pushed the pr/provide-template-for-an-efficient-private-ram branch from 00f4e70 to ddba890 Compare July 4, 2025 09:06
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants