CNA-VGAP-011: No test exercises PlatformSelection.cmake's reserved-identifier and unknown-name refusal paths

CNA snapshot 009d40f5  ·  Known Issues › Verification gaps  ·  source links pinned to 009d40f5

PlatformSelection.cmake fails configure for SDL12, EMSCRIPTEN, WIN32 off Windows, TERMINAL on Windows and unknown names, but the only platform-selection configure test, CnaWaylandPlatformSelection, covers Wayland alone.

Identifier

CNA-VGAP-011

Category

Verification gap

Subsystem

Build & CI

Status

Open

Verified against

CNA 009d40f5 (009d40f5dd085c4e674d3479675fac84b12b3e0a)

Evidence basis

Source-verified: read at TARGET, not executed

Tests touching this area

None

Affected contract

cmake/PlatformSelection.cmake configure-time refusals for reserved and unknown CNA_PLATFORM values

Expected behaviour

PlatformSelection.cmake makes these refusals a deliberate contract: it says the reserved-name failure is “a hard error on purpose”, because “falling back to SDL3 would build something other than what you asked for.” A contract that regresses silently -- a reserved name that starts to fall through to SDL3 -- deserves a test, the way Wayland's selection does.

Actual behaviour at TARGET

The one selection test registered in cmake/UnitTests.cmake is CnaWaylandPlatformSelection (WaylandPlatformSelection.cmake), which checks only that CNA_PLATFORM=WAYLAND is offered where buildable, refused with a package name where not, and changes no default. No test configures CNA_PLATFORM=SDL12, EMSCRIPTEN, WIN32 on a non-Windows host, TERMINAL on Windows, or an unknown name and asserts the FATAL_ERROR. The Sdl2OnlyRendererGate, RendererDefaultCase and RendererRetiredIdentityCase tests cover renderer selection, not the four platform refusal branches.

Source locations

• cmake/PlatformSelection.cmake — the four FATAL_ERROR refusal branches (reserved list, host-conditional, unknown name)
• cmake/UnitTests.cmake — CnaWaylandPlatformSelection is the only platform-selection configure test
• cmake/Tests/WaylandPlatformSelection.cmake — covers Wayland selection only

Evidence

Checked by reading PlatformSelection.cmake and every add_test in cmake/UnitTests.cmake and each fixture under cmake/Tests/ at 009d40f5. The Wayland selection fixture is the only one that drives platform selection; nothing drives the reserved or unknown branches. Not executed. The Development platforms overview page records the same: “No CMake test at this snapshot exercises the reserved-name path.”

Focused reproduction

No focused reproduction is known. Nothing has been invented here; the evidence above is what exists.

Current tests

CnaWaylandPlatformSelection (Wayland only), and the renderer-selection fixtures. None configures a reserved or unknown CNA_PLATFORM.

Regression test

A fixture like WaylandPlatformSelection.cmake that runs cmake against a tiny project including PlatformSelection.cmake with CNA_PLATFORM set to each reserved value, an unknown value, and a host-wrong value, asserting a non-zero exit and the message that names the reserved/available sets.

Blast radius

The guarantee that a reserved or misspelled CNA_PLATFORM fails loudly instead of falling back to SDL3 is unverified; a regression there would ship silently. The offered platforms themselves are exercised by the platform CI cells and the Wayland selection test.

Workaround

No workaround is known.