feat: Overhaul Github Actions

[ajnavarro]

Github Actions refactoring

This PR aims to heavily simplify our CI and the way we are running our tests. To remove as much cognitive load as possible, Go default test commands are used for testing (a.k.a go test ./...) instead of intricated makefiles.

This PR is the first of several PRs cleaning and simplifying the test suite.

On this PR:

• We are simplifying all the CI workflows using reusable workflows on all existing Go modules living on the monorepo.
• Improvements made on contribs on chore: make contribs/ ci and makefiles dynamic #2013 are applied on this PR (CC @moul) (closes chore: make contribs/ ci and makefiles dynamic #2013)
• Codecov tasks are also simplified, creating a tag per module, making codecov happier when uploading the reports and the code coverage report more understandable and easy to follow. Old tags must be deleted after merging this PR.
• Now all tests are executed. The longest test suite is gnovm (6 mins). I don't think it is worth it to add the -short tag taking into account the extra work that it causes when tests are failing only in master because we are not running them on all the PRs. Apart from that, tests that should have been executed on CI are executed now.
• Docker images are only built on master branch.

TODO:

These are some of the missing improvements on this PR that will be addressed in following ones:

• Remove custom docker builds and use goreleaser
• Check examples workflow and see how we can better test them maybe using go tests.
• Check makefiles and simplify them
• Should we move portal loop stuff outside the repo?
• Test with several go versions when needed
• Add cache where it makes sense
• Fix race conditions and run tests with -race flag
• Rebump benchmarks if possible
• Build binaries and serve them as artifacts on every PR
• TestPackages should only test the standard libraries. Rename it to TestStdlibs
• Move failing challenges to issues

Extra checks:

• Added new tests, or not needed, or not feasible
• Provided an example (e.g. screenshot) to aid review or the PR is self-explanatory
• Updated the official documentation or not needed
• No breaking changes were made, or a BREAKING CHANGE: xxx message was included in the description
• Added references to related issues and PRs
• Provided any useful hints for running manual tests
• Added new benchmarks to generated graphs, if any. More info here.

[moul]

Related with #2013

[thehowl]

Overall, you seem to be also making a work to ensure that the "standard scenario" for testing is one where doing "go test ./..." just works. I can agree with that; I'd like to see the exceptions we have on the Makefiles (and copied into the CI) reduced as well.

However, consider that aside from the -run flags in gnovm/Makefile exclude some tests (like challenges, and the packages), consider also that we always pass the -short flag when testing the tests/files directory because some files (the ones with _long.gno) take a huge amount of time.

They are currently not covered; if you want to test them by adding a daily CI (rather than one at each commit) testing for them then go for it. But the CI should still run these with -short.

[codecov]

Codecov Report

Attention: Patch coverage is 0% with 10 lines in your changes are missing coverage. Please review.

Project coverage is 21.79%. Comparing base (977a3f4) to head (29bd559).

Files	Patch %	Lines
misc/autocounterd/cmd/cmd_start.go	0.00%	6 Missing ⚠️
misc/loop/cmd/snapshotter.go	0.00%	3 Missing ⚠️
misc/loop/cmd/cmd_serve.go	0.00%	1 Missing ⚠️

Additional details and impacted files

@@             Coverage Diff             @@
##           master    #2040       +/-   ##
===========================================
- Coverage   55.01%   21.79%   -33.22%     
===========================================
  Files         481       24      -457     
  Lines       67432     2092    -65340     
===========================================
- Hits        37097      456    -36641     
+ Misses      27318     1619    -25699     
+ Partials     3017       17     -3000     

Flag	Coverage Δ
contribs/gnofaucet	14.46% <ø> (?)
contribs/gnomd	0.00% <ø> (?)
gno.land-_test.gnokey	?
gno.land-_test.gnoland	?
gno.land-_test.pkgs	?
gnovm-_test.cmd	?
gnovm-_test.gnolang.native	?
gnovm-_test.gnolang.other	?
gnovm-_test.gnolang.pkg0	?
gnovm-_test.gnolang.pkg1	?
gnovm-_test.gnolang.pkg2	?
gnovm-_test.gnolang.realm	?
gnovm-_test.gnolang.stdlibs	?
gnovm-_test.pkg	?
go-1.21.x	?
go-1.22.x	?
misc	?
misc-_test.genstd	?
misc/autocounterd	0.00% <0.00%> (?)
misc/genproto	0.00% <ø> (?)
misc/genstd	73.90% <ø> (?)
misc/goscan	0.00% <ø> (?)
misc/logos	17.38% <ø> (?)
misc/loop	0.00% <0.00%> (?)
tm2-_test.flappy	?
tm2-_test.pkg.amino	?
tm2-_test.pkg.bft	?
tm2-_test.pkg.others	?

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[zivkovicmilos]

I might be crazy, but the CI on this PR is significantly faster and snappier 👀

Let's go 🚀

[zivkovicmilos]

@thehowl did you get a chance to recheck this? 👀

[thehowl]

overall LGTM. a few comments; on the ones from the previous review, please act on them and we should be good to go