Skip to content

Conversation

@daniel-noland
Copy link
Collaborator

All of the commits save wip are ready for review. Will finish breaking up the last commit tomorrow morning

@qmonnet
Copy link
Member

qmonnet commented Oct 30, 2025

All of the commits save wip are ready for review.

... but wip is the only one actually related to DPDK 😆. You should have done a separate PR with the first 7 commits, I'd have merged them already 🙂

@mvachhar
Copy link
Contributor

Does this fix #481?

@daniel-noland daniel-noland force-pushed the pr/daniel-noland/split-init branch 3 times, most recently from 7edf8d1 to 2000a0f Compare November 1, 2025 23:56
This commit introduces most of the new dependencies needed for this patch set.

Signed-off-by: Daniel Noland <[email protected]>
The derives used here are not quite correct and mess with the display impl derivation.
This breaks compiles downstream when the serde feature is enabled.

Signed-off-by: Daniel Noland <[email protected]>
This are just random, trivial nitpicks I found while reading the mgmt
code.

Signed-off-by: Daniel Noland <[email protected]>
This module nesting isn't actually doing anything other than makingi
test names long.

Signed-off-by: Daniel Noland <[email protected]>
This should make our stack traces better by enriching them with the
color-eyere / tracing data.

Signed-off-by: Daniel Noland <[email protected]>
Signed-off-by: Daniel Noland <[email protected]>
These never should have been public.

Signed-off-by: Daniel Noland <[email protected]>
This is needed to ensure safe deserialization of the generated rkyv.

Signed-off-by: Daniel Noland <[email protected]>
This test was scheduled for removal and is now being removed 🤷

Signed-off-by: Daniel Noland <[email protected]>
Signed-off-by: Daniel Noland <[email protected]>
Avoiding runtime code here lets us build support tables at compile
time.

Signed-off-by: Daniel Noland <[email protected]>
This module scans the running system and produces a hardware report
about it.

Signed-off-by: Daniel Noland <[email protected]>
This module defines our supported devices.

Signed-off-by: Daniel Noland <[email protected]>
Signed-off-by: Daniel Noland <[email protected]>
Missing this before.

Signed-off-by: Daniel Noland <[email protected]>
This is needed for intel virtual functions (which are common).

Signed-off-by: Daniel Noland <[email protected]>
Signed-off-by: Daniel Noland <[email protected]>
This is needed to rkyv generated configurations from init to dataplane.

Signed-off-by: Daniel Noland <[email protected]>
Signed-off-by: Daniel Noland <[email protected]>
No need for cell here when lazylock keeps everything nice and safe :)

Signed-off-by: Daniel Noland <[email protected]>
Signed-off-by: Daniel Noland <[email protected]>
These are old and don't seem to reflect current goals.

Signed-off-by: Daniel Noland <[email protected]>
@daniel-noland daniel-noland force-pushed the pr/daniel-noland/split-init branch from 33185cb to eb961e4 Compare November 2, 2025 05:19
@daniel-noland
Copy link
Collaborator Author

I am breaking this into 3 prs. Two of them are up and available now as #963 and #964

@qmonnet
Copy link
Member

qmonnet commented Nov 3, 2025

Should we close this one, in that case?

@daniel-noland
Copy link
Collaborator Author

closing as this is showing up in other prs

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.

4 participants