[TheAggie]

What

CB_COLOR_CONTROL.MODE 0 (disable) turns off color writes. It was accepted only when no color target was written. With a written target the draw was skipped (only normal color rendering with copy ROP is supported (CB_COLOR_CONTROL 0x00000000, mode disable)), even though the hardware writes no color in that mode.

• DecodeState: when the mode is disable, no export counts as written, so the draw has no color attachment and renders like the existing attachment-free path (depth only, or the screen scissor extent).
• colorControlSupported: mode disable is accepted with any ROP3 and other bits, which do not apply while color writes are off. The draw rejection check accepts the same.
• agc_driver_graphics (DisabledColorTests): a written color target with CB_COLOR_CONTROL 0x0 and 0xcc0000 decodes without a color attachment and keeps the scissor extent. It fails without the change.

Related to #1346, which drops COLOR_INVALID targets from the write mask. That PR handles disabled formats, this one the disabled mode; both touch the same check.

Tested

Windows 11 Pro (26200), MinGW-w64 GCC 15.2.0 (WinLibs r7), Release, NVIDIA GeForce RTX 5070 Ti.

• All agc_* and shader_* tests pass. They were run from the generated CTestTestfile.cmake commands, because antivirus blocks ctest.exe on this machine.
• PS5 OpenGL Showcase (PPSA99005, blackbearreloaded/ps5-opengl) issues a depth pre-pass with this state for its instanced scene. Its draws were skipped. With this change and the other pending fixes on a local branch (fix(recompiler): give vertex-path shaders a full NGG merged wave info #1412 and others), the scene renders. This change alone does not make it render.

Not tested on AMD, Intel or Linux.

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 in TechnicalDebt
• 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
• AI-assisted: yes

🤖 Generated with Claude Code

[TheAggie]

Showcase before/after on the combined local branch: without this change the instanced scene's draws are skipped; with it (together with #1415) the scene renders.

[adriadam10]

Checked build + full ctest on current main + this PR (457/457, agc_driver_graphics passes), and the new DisabledColorTests case fails without the State.cpp change. Mode 0 = CB_DISABLE matches Mesa (V_028808_CB_DISABLE). The branch conflicted with main in State.cpp/tests/Graphics.cpp (#1419 colorControl &= ~1u, COLOR_INVALID test block); rebased keeping both. Merging once CI is green on the rebased head.