Re: [bitbake-devel] [oe-core][PATCH 1/2] devtools: Initial recipe for repo 2.17.3

["Richard Purdie" <richard.purdie@linuxfoundation.org>]

On Tue, 2021-11-09 at 11:26 +0000, Jasper Orschulko wrote:
> 
> 
> > e) fetcher output is deterministic
> >  (i.e. if you fetch configuration XXX now it will match in future
> > exactly in 
> >  a clean build with a new DL_DIR)
> 
> check. When a fixed refspec is set within the recipe, the fetcher will
> check, if all repositories in the repo manifest are set to a fixed
> refspec as well and otherwise throw an error.

When you say "fixed refspec", will that be a definitive sha revision or a tag?
We always force resolution of tags as they tend to cause problems and can change
even if it is bad form.

> > f) network access is expected to work with the standard linux proxy
> > variables
> >  environment but only in the do_fetch tasks)
> 
> this should work, as we keep the GIT_PROXY_COMMAND environment and run
> repo via runfetchcmd(). Not further tested though.

If you're using runfetchcmd that should work.

> 
> > g) access during parsing has to be minimal, a "git ls-remote" for an
> > AUTOREV 
> >  git recipe might be ok but you can't expect to checkout a git tree
> 
> unfortunately, do to the nature of repo, we need to clone the manifest
> repo, so that we can run "git ls-remote" on the referenced git repos
> and therefore ensure that 

This is potentially a big issue. Cloning operations during parsing is pretty
horrible. We'd not expect any thing being written out like that during a parse.
It would probably work "ok" for one recipe but if you start getting the hundreds
of git recipes we have in some layers, it wouldn't scale if we allowed that :(.

Not sure what to recommend here but it is definitely problematic.

> 
Cheers,

Richard