fix(ppc-manual): 543 dead links, from two generator bugs and wrong relative paths #48

Merged
fabi merged 1 commits from fix/ppc-manual-dead-links into main 2026-09-17 05:11:50 +00:00
Owner

The manual had 545 relative links to pages that don't exist. After this PR, 2 remain, both intentional: TEMPLATE.md's <group>/<FORM> placeholders.

links cause fix
350 generator: category pages linked <slug>.md relative to categories/ link ../<category>/<slug>.md
49 generator: form pages linked each member into its own category directory, so VMX128 siblings pointed at vmx128/ although their family page is under vmx/ link into the family's directory
109 hand-written Related sections linked other categories' pages as if they sat in the same directory retargeted through the page index (e.g. fpu/fdivx.md../control/mffsx.md)
29 a family page linking to itself, e.g. [vrefp128](vrefp128.md) on vrefp.md plain text
6 links to instructions the manual has no page for (vupkhsh128, vsubeuwm, vupkhub, …) plain text

Canary pin unchanged. Upstream canary_experimental moved (f21ebd49eaee0871dd7), so a plain regeneration would re-pin all 350 family pages. I regenerated at the existing pin (--canary-ref f21ebd49e…) to keep this PR to links. For a later re-pin: none of those upstream commits touch src/xenia/cpu/ppc or the x64 backend, so the #45 review would carry over as is.

Checked:

  • Generator at the pin: 0 family pages changed, and a second run changes nothing.
  • A link checker over every page finds only the two template placeholders.
  • Spot checks: categories/memory.md../memory/dcbf.md; forms/VX128_5.md../vmx/vsldoi.md; memory/stdcx.md../alu/sync.md.

🤖 Generated with Claude Code

The manual had **545 relative links to pages that don't exist**. After this PR, 2 remain, both intentional: `TEMPLATE.md`'s `<group>`/`<FORM>` placeholders. | links | cause | fix | |---:|---|---| | 350 | generator: category pages linked `<slug>.md` relative to `categories/` | link `../<category>/<slug>.md` | | 49 | generator: form pages linked each member into its *own* category directory, so VMX128 siblings pointed at `vmx128/` although their family page is under `vmx/` | link into the family's directory | | 109 | hand-written Related sections linked other categories' pages as if they sat in the same directory | retargeted through the page index (e.g. `fpu/fdivx.md` → `../control/mffsx.md`) | | 29 | a family page linking to itself, e.g. `[vrefp128](vrefp128.md)` on `vrefp.md` | plain text | | 6 | links to instructions the manual has no page for (`vupkhsh128`, `vsubeuwm`, `vupkhub`, …) | plain text | **Canary pin unchanged.** Upstream `canary_experimental` moved (`f21ebd49e` → `aee0871dd7`), so a plain regeneration would re-pin all 350 family pages. I regenerated at the existing pin (`--canary-ref f21ebd49e…`) to keep this PR to links. For a later re-pin: none of those upstream commits touch `src/xenia/cpu/ppc` or the x64 backend, so the #45 review would carry over as is. **Checked:** - Generator at the pin: 0 family pages changed, and a second run changes nothing. - A link checker over every page finds only the two template placeholders. - Spot checks: `categories/memory.md` → `../memory/dcbf.md`; `forms/VX128_5.md` → `../vmx/vsldoi.md`; `memory/stdcx.md` → `../alu/sync.md`. 🤖 Generated with [Claude Code](https://claude.com/claude-code)
fabi added 1 commit 2026-09-16 20:37:54 +00:00
fix(ppc-manual): 543 dead links, from two generator bugs and wrong relative paths
All checks were successful
CI / Native — linux (pull_request) Successful in 2h2m48s
CI / WASM — Web (pull_request) Successful in 29m51s
CI / Formatting (pull_request) Successful in 1m36s
21161154d1
- Category pages linked each family as `<slug>.md`, relative to categories/,
  where no family page lives. They now link `../<category>/<slug>.md`.
- Form pages linked a member into its *own* category directory, so every
  VMX128 sibling (`vsldoi128`) pointed at vmx128/ although its family page is
  under vmx/. They now link into the family's directory.
- Hand-written "Related" and sibling mentions linked other categories' pages
  as if they were in the same directory. 109 are retargeted through the page
  index; 29 that pointed a family page at itself (`vrefp128` on vrefp.md) and
  6 naming instructions the manual has no page for are plain text now.

Regenerated at the existing Canary pin (f21ebd49e): upstream has moved on, and
re-pinning belongs in its own change. The generator reports 0 family pages
changed and is idempotent; the only dead links left are TEMPLATE.md's
placeholders.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
fabi merged commit e7325577e4 into main 2026-09-17 05:11:50 +00:00
Sign in to join this conversation.