[adriadam10]

What

Follow-up of #5106 for the two other test executables that compiled driver sources again: agc_driver_graphics_tests (29 Graphics/src and 2 Execution/src sources, the generated shader headers, plus BufferPool, TextureDetilerDescriptors, TextureCache and GpuColorTransfer from PostCompile.cmake) and agc_driver_bda_device_tests (8 driver sources plus BufferPool and GpuColorTransfer). Every one of them is already in libSceAgcDriver. Both now build only their tests/ files and link libSceAgcDriver; the recompiler include directories come from shader_recompiler_agc_driver's interface property instead of linking a second copy of the recompiler. About 45 compiles fewer per configuration.

Tested

Linux, GCC, Release: both executables build and link. The 5 tests that run them (agc_driver_graphics, agc_driver_bda_device, agc_sparse_buffer_holes, agc_descriptor_pool_reset_failure, agc_compute_descriptor_layout_flags) pass on RADV and on CI's lavapipe (25.2.8, GALLIUM_OVERRIDE_CPU_CAPS=avx, ANYPS5_REQUIRE_VULKAN=1).

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
• Unblocks: none
• AI-assisted: yes