[dantdmdarealguy]

What

Since 56c2a53 (fix(relinker): redraw console progress and format runtime layout as a tree, 2026-10-13), the relinker no longer prints a Guest module: <path> line for each converted bundled module. It prints Converted modules: N and then lists each .guest.prx under app0/ in its Arrange the files like this tree (core/relinker/main.cpp L438-452, core/relinker/cli/include/RuntimeLayout.hpp). It writes each converted module into app0/ beside the output executable, keeping the module's own folder (GuestModuleBuilder.cpp L292).

Two guides still tell the reader to use the files "printed as Guest module:", a line that no longer appears in the output:

• docs/dev/GAME_TESTING.md L15 (Before a run)
• docs/user/USAGE.md L175 (Runtime layout)

Both sentences now point at the .guest.prx files that the relinker writes into app0/ and lists in the tree it prints. The rest of each paragraph is unchanged. That includes keeping the original .prx and the error 193 note.

Tested

Windows 11 Pro, Intel Core i9-14900K, 128 GB RAM, NVIDIA GeForce RTX 5080 (driver 32.0.16.1742), WinLibs GCC 15.2.0. Release build of main 5a79f2645 plus --target libs relinker, rebased onto that commit.

• ScreenTester5 (PPSA69420), unwrapped with tools/unwrap_self.py, was relinked with relinker --windows --registry source/eboot.elf run/app.exe. The run ends with:

  Conversion complete
  Output file: ...\run\app.exe
  External prx references: 208
  Converted modules: 1
  
  Arrange the files like this:
  folder/
    |-- app.exe
    |-- libs/
          |-- *.prx (system libraries)
    |-- app0/
          |-- <another app files>
          |-- sce_module/
                |-- libc.prx.guest.prx
  

  The output contains no Guest module line (grep -c = 0), and the file was written to run/app0/sce_module/libc.prx.guest.prx. Laid out that way, with the original libc.prx kept beside it, ScreenTester5 starts (Selected GPU name=NVIDIA GeForce RTX 5080), opens its window and exits with 0 when the window is closed after 25 s.
• git grep -n '"Guest module: "' -- core finds nothing on 5a79f2645. 56c2a53 removed the print (git log -S). git grep "Guest module:" -- docs finds 2 lines on main (GAME_TESTING.md:15, USAGE.md:175) and none with this change.
• python tools/check_conventions.py --base origin/main: All conventions checks passed.
• ctest -R "^(patched_libraries|conventions_doc_links|test_registration)$": 3/3 passed (367 patched libraries loaded, 0 failed).
• python tools/pr_overlap.py --branch on the rebased branch: no overlap with the PRs it could compare. It could not compare 9 PRs that conflict with main (feat(aot): consolidate native ARM64 compiler, runtime and consumers #5021, perf(agc): build the pipelines of a submission ahead on a compile pool #5925, fix(tools): recognize guest module identities in import audit #6171, feat(agc): select Vulkan devices by driver name #6788, perf(recompiler): keep runtime-specialized shader modules in the shader disk cache #6924, feat(agc): look up table images at run time through a device image table on the GPU heap (APS5_GPU_PATH) #7642, fix(agc): reclaim idle images safely during cache creation #7800, fix(relinker): rewrite XSAVEOPT to XSAVE under --to-rosetta #7918, feat(relinker): trace function imports on macOS #9408), all on docs/user/USAGE.md. I fetched each head (pull/<n>/head) and diffed it against its merge base with main for docs/user/USAGE.md and docs/dev/GAME_TESTING.md. None of these diffs contains Guest module (or guest.prx), and none touches GAME_TESTING.md.

Checklist

• Based on current main; no other open PR implements the same functions
• One topic per PR; follow-ups go in a new PR
• No comments in code except technical debt
• Unimplemented paths throw (NotImplemented_nid_no_patch); silent stubs are listed under Silent stubs in their library's technical debt file
• No notes, investigation .md files or images added to the repository (attach them to this PR)
• New third-party code is a submodule built from source
• Generic behaviour, not specific to one title
• Depends on: none
• Fixes: none
• Unblocks: none
• GPUs tested: n/a
• AI-assisted: yes, some AI assistance was used

[MarcZX10]

Reviewed the two documentation updates against the current relinker output and runtime layout implementation. The CLI now prints a converted-module count and a layout tree under app0/, and no longer emits the old per-module “Guest module:” line. The replacements accurately describe where the generated .guest.prx files are written and preserve the requirement to keep the original module alongside them. The PR changes only those two stale instructions, follows the template, and reports focused documentation and Windows runtime checks. The conventions, intake, eligibility and label checks pass; executable build jobs were skipped, which is appropriate for this documentation-only diff.

[github-actions]

Some comparisons could not be completed because a pull request conflicts with the base branch. Review these shared files manually before concluding that there is no overlap.

Pull request	Shared files not compared
PR 5021	docs/user/USAGE.md
PR 5925	docs/user/USAGE.md
PR 6171	docs/user/USAGE.md
PR 6924	docs/user/USAGE.md
PR 7642	docs/user/USAGE.md
PR 7918	docs/user/USAGE.md
PR 7942	docs/user/USAGE.md
PR 9408	docs/user/USAGE.md
PR 9508	docs/user/USAGE.md
PR 10193	docs/user/USAGE.md

[MarcZX10]

Rechecked the two changed instructions against the relinker output described in the PR and the reported Windows runtime-layout reproduction. Both now point readers to the generated .guest.prx files in app0/ and the printed layout tree; GAME_TESTING.md also preserves the instruction to keep the original .prx. The reported focused documentation checks pass. Hosted eligibility, intake, conventions and labels pass; build/relinker jobs are skipped for this documentation-only diff. I did not run local tests.

[MarcZX10]

The two documentation replacements match the current relinker output described in the PR, and the focused documentation/runtime tests plus policy checks are reported passing. The branch is still based on a6dcea9, while upstream main is now ab347d1; although these two documentation files have no intervening upstream changes and GitHub reports the PR mergeable, please update the branch/checklist to current main before merging. I could not independently retrieve the path-filtered diffs for the nine uncomparable PRs listed in the overlap warning in this pass; the PR’s exact-string manual check is useful, so please refresh the overlap check on the updated base and retain the documented finding that those diffs do not mention the removed Guest module: output. I did not run local tests.

[dantdmdarealguy]

Rebased onto current main (5a79f2645). The commit is unchanged (same patch-id), and both docs files took the upstream edits from #8135/#8807/#8973 without conflicts. Because #8135 added a line above it, the GAME_TESTING sentence is now L15; I've updated the What section. On the new base, git grep "Guest module:" -- docs finds the 2 lines on main and none with the change. A ScreenTester5 relink built from 5a79f2645 still prints Converted modules: 1 and the Arrange the files like this tree with sce_module/libc.prx.guest.prx, with no Guest module line. The title starts with that layout. check_conventions.py passes and patched_libraries/conventions_doc_links/test_registration pass 3/3. I refreshed pr_overlap.py on the updated base. It can't compare 9 PRs, all on USAGE.md: #5021, #5925, #6171, #6788, #6924, #7642, #7800, #7918 and #9408 (#9408 is new; #6947 has since been closed). I diffed each of their heads against its merge base for USAGE.md and GAME_TESTING.md: none mentions Guest module, and none touches GAME_TESTING.md.

---

[dantdmdarealguy]

Rebased onto main (5e6ae6e): the USAGE.md paragraph gained a sentence upstream about modules that were not relinked (--module-dir). I kept it and only replaced the Guest module: sentence, as before. check_conventions.py passes.