Skip to content

move front-io related sequencing logic into the server#2502

Open
Aaron-Hartwig wants to merge 4 commits into
masterfrom
aaron/front-io-server-seq
Open

move front-io related sequencing logic into the server#2502
Aaron-Hartwig wants to merge 4 commits into
masterfrom
aaron/front-io-server-seq

Conversation

@Aaron-Hartwig
Copy link
Copy Markdown
Contributor

Act III in the effort to decouple front-io related logic from board specific sequencers involves moving all the actual sequencing logic (e.g. FPGA bitstream loading) into the front-io-server. This diff is very noisy as a result of that, but the core logic in play here is not touched. The last piece to do here will be move all the transceivers and LED stuff over, but that warrants its own commit I think. I am bringing Medusa along for the ride for now even though it is untested. After the decoupling is complete I will have a follow on PR which was used to actually bring up that board.

@Aaron-Hartwig Aaron-Hartwig requested a review from mkeeter May 7, 2026 18:53
@Aaron-Hartwig Aaron-Hartwig self-assigned this May 7, 2026
Comment thread drv/front-io-server/build.rs Outdated
Comment thread drv/sidecar-seq-server/src/main.rs
Comment thread drv/front-io-server/src/main.rs Outdated
Comment thread drv/front-io-server/src/main.rs
Comment thread drv/front-io-server/src/main.rs
Comment thread drv/front-io-server/src/main.rs Outdated
Comment thread drv/front-io-server/src/main.rs
Comment thread drv/front-io-server/src/main.rs
Comment thread drv/front-io-server/src/main.rs
@mkeeter
Copy link
Copy Markdown
Collaborator

mkeeter commented May 11, 2026

This all looks pretty good! Have you tested on Sidecar hardware? Dunno if you've thought about how to test the bad oscillator case – I'd like to confirm that the PR still handles it correctly, but it's hard to artificially trigger...

@Aaron-Hartwig
Copy link
Copy Markdown
Contributor Author

Yeah, this was tested on my Sidecar/Cosmo setup and things worked. My unit does not have a chronically bad oscillator, but sidecar-b on niles does and is able to correctly do monorail's PHY reset dance (something I'd like to move into the front-io-server at some point...). There were some ringbuf improvement opportunities to give you a clear picture of what was happening, so I'm toying with those now.

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