Adds a `docker-args-process-build` trigger to the resource plugin so
limits set via `dokku resource:limit --process-type build APP` are
applied during the build phase. Only `build.limit.*` properties are
read - defaults do not inherit, since builds typically need more memory
than runtime and a leaked tiny default would cause confusing OOM
failures. Reservations are never applied at build time. Allowed flags
are filtered per builder: herokuish gets cpu, memory, memory-swap, and
nvidia-gpu; dockerfile gets memory and memory-swap; pack, nixpacks,
railpack, lambda, and null emit nothing because their underlying CLIs
do not accept docker run resource flags. The dockerfile builder
whitelists the new memory flags and corrects pre-existing typos where
`--ssh` mapped to `--platform` and `--ulimit` mapped to `--tag`.
As there can be custom schedulers, including scheduler-specific information on the resource management page can be a disingenuous - the values don't map to kubernetes, for example - and thus we should call out to the scheduler docs (and underlying docker docs) to make the mapping more clear.
This was removed in c9bbddcfcb as it was more complicated for operators to figure out _exactly_ which combination of limits/reservations would apply to their applications. Unfortunately, the documentation was never updated.
We may re-add this functionality in the future, but for now, cleaning up the docs is the best option.
[ci skip]
Global limits make it harder to reason about the inheritance in the system. Rather than support global limits, Dokku should provide primitives to automatically set them on an app at provision time, or explicitly set them for an app. Both of these are likely best suited for a higher-level provisioner, or a community plugin.
