ci: make the matrix retry budget fit inside the job cap (#245)

The retry budget was larger than the job cap on both engines, so the last
attempt was always killed partway and the cell reported `cancelled`.

Measured on master run 32643567648 (post-#243), the 5 JRuby cells:

  jruby-10.0 rails81  suite 713s, no hang, 1 attempt   -> 12m02  pass
  jruby-10.0 rails71  suite 545s + 6m hang -> 15m t/o;
                      attempt 2 clean at 543s          -> 24m20  pass
  jruby-10.0 rails80  attempt 1 hung -> 15m t/o;
                      attempt 2 killed at the 25m cap  -> cancelled
  jruby-10.0 rails72  same                             -> cancelled
  jruby-head rails81  same                             -> cancelled

3 of 5 JRuby cells gate nothing. The cause is arithmetic, not the driver:
max_attempts 3 x timeout_minutes 15 = 45 min against timeout-minutes 25.
Attempt 3 could never start, and attempt 2 had only 25 - 15 - 0.7 = 9.3 min
to finish a run that measures 9-12 min -- so whether a cell survived a hang
came down to which gemfile it drew. MRI has the same shape, smaller: 3x3 = 9
against a cap of 8.

  - JRuby job cap 25 -> 31, so 1 + 15 + 15 fits.
  - max_attempts 3 -> 2 on both engines, since 3 was never reachable.

A doubly-hung JRuby cell now costs 31 min instead of 25, but today's 25 min
buys no verdict at all. Coverage, drivers and the full-ci/cron opt-in model
are unchanged.

Also adds the concurrency group Lint never had, so superseded PR pushes stop
running the linter to completion.

Author: pftg

.github/workflows/lint.yml

@@ -13,6 +13,12 @@ on:
 env:
   RUBY_YJIT_ENABLE: 1
 
+# Test has had this since it was written; Lint never did, so every push to a
+# PR ran a fresh linter to completion alongside the superseded ones.
+concurrency:
+  group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}
+  cancel-in-progress: true
+
 jobs:
   lint:
     name: Ruby & YAML

.github/workflows/test.yml

@@ -108,7 +108,10 @@ jobs:
       contains(github.event.pull_request.labels.*.name, 'full-ci')
     needs: [ functional-test ]
     runs-on: ubuntu-latest
-    timeout-minutes: ${{ contains(matrix.ruby-version, 'jruby') && 25 || 8 }}
+    # Must fit `max_attempts * timeout_minutes` below, plus ~1 min of setup,
+    # or the last attempt gets killed mid-run and the cell reports `cancelled`
+    # -- a dead gate. JRuby: 1 + 15 + 15 = 31. MRI: 1 + 3 + 3 = 7, under 8.
+    timeout-minutes: ${{ contains(matrix.ruby-version, 'jruby') && 31 || 8 }}
     continue-on-error: ${{ matrix.experimental }}
     strategy:
       matrix:
@@ -151,17 +154,23 @@ jobs:
           ruby-cache-version: ${{ matrix.ruby-version }}-${{ matrix.gemfile }}-1
           cache-apt-packages: true
 
-      - name: Run tests (with 2 retries)
+      - name: Run tests (with 1 retry)
         uses: nick-fields/retry@v4
         with:
-          # 7 minutes stopped being enough for JRuby once the suite grew past
-          # ~600 tests: every JRuby cell on master now burns all three attempts
-          # on `Timeout of 420000ms hit` and dies at the job cap, so the cell
-          # has stopped gating anything. Locally the full suite is 390s on
-          # JRuby with vips and 434s with chunky_png -- the budget is the
-          # problem, not the driver.
+          # Measured on master run 32643567648: a clean JRuby attempt is
+          # 545-713s depending on the gemfile, so 15 min is ~26% headroom over
+          # the slowest. MRI is 128s against 3 min.
           timeout_minutes: ${{ contains(matrix.ruby-version, 'jruby') && 15 || 3 }}
-          max_attempts: 3
+          # Two, not three. A third attempt never fit inside the job cap on
+          # either engine (JRuby 3x15=45 > 25; MRI 3x3=9 > 8), so it only ever
+          # got killed partway and reported the cell as `cancelled`. Keeping
+          # the retry budget inside the cap is what makes the cell a real gate.
+          #
+          # The retry exists solely for the intermittent JRuby teardown hang
+          # (#244), which strands the process for minutes *after* the suite
+          # prints `Finished in ...`. Once #244 is fixed, drop this to a single
+          # attempt and the JRuby cap to 16 -- that halves the JRuby bill.
+          max_attempts: 2
           command: bin/rake test
 
   matrix-screenshot-driver: