CI: Improve Pages job

Code coverage and RTLMeter complete (concluding 'skipped') on branch
pushes where they do no work, and each completion fires the Pages
workflow_run trigger. Rebuilding and redeploying master's pages for
these no-op events is wasteful.

Gate the build job so a workflow_run whose upstream conclusion is
'skipped' does nothing; deploy and notify need build, so they
cascade-skip.

Route such no-op runs to their own throwaway concurrency group. In the
shared 'pages' group they would, as the newest queued run, cancel a
pending real run and then deploy nothing (concurrency keeps only the
in-progress and newest pending run). Real runs keep sharing 'pages' and
serialize as before.

Add retry to deploy step to reduce flakyness due to the Pages back end.

Fix trigger on workflow file changes.

.github/workflows/pages.yml

@@ -7,7 +7,7 @@ name: Pages
 on:
   push:
     branches: [master]
-    paths: ["ci/**", ".github/workflows"]
+    paths: ["ci/**", ".github/workflows/**"]
   workflow_dispatch:
   workflow_run:
     workflows: ["Code coverage", "RTLMeter"]
@@ -22,8 +22,11 @@ permissions:
 # Allow only one concurrent deployment, skipping runs queued between the run
 # in-progress and latest queued. However, do NOT cancel in-progress runs as we
 # want to allow these deployments to complete.
+# A skipped-upstream run does no work (see the build job's if:), so put it in its
+# own throwaway group: otherwise, as the newest queued run, it would cancel a
+# pending real run and then deploy nothing.
 concurrency:
-  group: "pages"
+  group: ${{ (github.event_name == 'workflow_run' && github.event.workflow_run.conclusion == 'skipped') && format('pages-skip-{0}', github.run_id) || 'pages' }}
   cancel-in-progress: false
 
 defaults:
@@ -34,6 +37,10 @@ jobs:
   build:
     name: Build content
     runs-on: ubuntu-24.04
+    # A skipped upstream run (e.g. Code coverage / RTLMeter on a branch push)
+    # still fires workflow_run; don't rebuild and redeploy for it. deploy and
+    # notify need this job, so they cascade-skip too.
+    if: ${{ github.event_name != 'workflow_run' || github.event.workflow_run.conclusion != 'skipped' }}
     outputs:
       pr-run-ids: ${{ steps.build.outputs.pr-run-ids }}
     steps:
@@ -58,9 +65,16 @@ jobs:
     runs-on: ubuntu-24.04
     environment:
       name: github-pages
-      url: ${{ steps.deployment.outputs.page_url }}
+      url: ${{ steps.deploy-2.outputs.page_url || steps.deploy-1.outputs.page_url }}
     steps:
+      # GitHub's Pages backend intermittently fails a deployment mid-sync, retry
       - name: Deploy to GitHub Pages
+        id: deploy-1
+        continue-on-error: true
+        uses: actions/deploy-pages@cd2ce8fcbc39b97be8ca5fce6e763baed58fa128  # v5
+      - name: Deploy to GitHub Pages (retry)
+        id: deploy-2
+        if: steps.deploy-1.outcome == 'failure'
         uses: actions/deploy-pages@cd2ce8fcbc39b97be8ca5fce6e763baed58fa128  # v5
 
   notify: