Prep patches for modularity and pinned repo packages #2785
Conversation
The only entrypoint so far doesn't use it (`countme`), but a future new entrypoint will. Also mention that the commands should add themselves to the array in libmain.cxx if applicable so it shows up in `--help`.
This dates from the cbindgen era and isn't needed anymore.
This will be necessary for Rust-based client commands to be able to do the regular "post-deployment package diff".
rust/src/treefile.rs
Outdated
| @@ -271,8 +271,10 @@ fn merge_basic_field<T>(dest: &mut Option<T>, src: &mut Option<T>) { | |||
| /// the `packages` key. | |||
| fn merge_vec_field<T>(dest: &mut Option<Vec<T>>, src: &mut Option<Vec<T>>) { | |||
| if let Some(mut srcv) = src.take() { | |||
| if let Some(ref mut destv) = dest { | |||
| srcv.append(destv); | |||
| if let Some(mut destv) = dest.take() { | |||
There was a problem hiding this comment.
The next patches will effectively add a unit test for this right? If so fine as is but if not it'd be good to add one.
There was a problem hiding this comment.
Yes, #2787 has tests which rely on this.
There was a problem hiding this comment.
Ohhh wait, yes I think the postprocess scripts are relying on the current ordering.
|
The vmcheck failure there is a seen-before race around zincati/prom metrics. (Not sure if we filed it anywhere). Not sure what was up with the compose failure. Restarted CI. |
|
EDIT: See above, this is the Vec ordering.
|
This made me hesitate for a bit so add a comment about it to make it clear. The RPMOSTREE_ASSEMBLE_TYPE_SERVER_BASE path was used by the `ex container`, which was subsequently ripped out.
jlebon
force-pushed
the
pr/modularity-prep-patches
branch
from
89ad445
to
6ee97b5
Compare
|
Ahhh heh, alrighty then. Indeed postprocess scripts is a good reason not to change those semantics. I dropped that commit and will just adapt #2787. |
See individual commit messages. Split out of #2760.