From 90ab9baa309ae3c54d371546872c918c4c801920 Mon Sep 17 00:00:00 2001 From: Arthur Hanson Rasmusson Date: Tue, 10 Nov 2026 05:25:15 -0500 Subject: wiki: 2026-11-10 status refresh - guest GL draws execute on the physical GPU Refresh STATE-CURRENT, the Main Page and the status/frontier pages: the GSP event-IRQ fix completes swaps; the 2D path is green end-to-end; the guest's real GL draws now forward and execute on the physical GPU (work_gr backend_ok=9); the draw writes 0 pages because its constant-buffer (MVP) page is unresolved and forced to the seed page; the MME front-end defer, GR stale-wedge and the X DDX get-word/notifier completion remain. Published to open-iov.org. --- wiki/Graphics-Execution.md | 96 ++++++++++++++++++++++++---------------------- 1 file changed, 50 insertions(+), 46 deletions(-) (limited to 'wiki/Graphics-Execution.md') diff --git a/wiki/Graphics-Execution.md b/wiki/Graphics-Execution.md index ffbcfaf..7771d6a 100644 --- a/wiki/Graphics-Execution.md +++ b/wiki/Graphics-Execution.md @@ -12,7 +12,7 @@ > [`../docs/04-implementation/04-real-render-walls.md`](../docs/04-implementation/04-real-render-walls.md) > (the measured walls); the docs tree wins on disagreement. -Status: 2026-11-09, tree `master @ 21e0d611`. Tags: `[live]` observed end to +Status: 2026-11-09, tree `master @ cb8e822b`. Tags: `[live]` observed end to end; `[wip]` partially built; `[wall]` current blocker; `[target]` defined success state; `[planned]` designed, not built. @@ -89,10 +89,12 @@ fidelity: kick channel (`0xbfef0002`) is decoded from its ring head with its class-2D notification authored. The X DDX waits are fixed (class-aware display arm, `0x8000 -> 0` notification completion, decoded GL release - applied), and with GLX forced (`SDL_VIDEO_X11_FORCE_EGL=0`) the windowed - GL path selects `libnvidia-glcore`. What remains is the kernel-visible GL - present/kick completion, the DDX get-word/notifier recycle, and physical - dispatch of the client's draws (§6). + applied), forcing GLX selects `libnvidia-glcore`, and the event-IRQ fix + (`a478f0b4`) lets the guest service the completion, so glxgears completes + swaps and the guest's real GL draws now forward and execute on the + physical GPU (`work_gr backend_ok=9`). What remains is the constant-buffer + (MVP) content, the MME front-end defer and GR stale-wedge, and the DDX + get-word/notifier completion (§6). ## 4. The channel and object picture @@ -139,10 +141,12 @@ until it posts, so the X server's mode set finishes. For the X DDX's own 2D push buffers, completion also requires the GPU-written engine get word to advance and the notifier to be recycled to `0xffff` after a kick; both are host-side levers now, and a diagnostic write of the get word released the -DDX's reserve loop. For a windowed GL client the present is a kick whose -completion the guest kernel observes by polling its per-slot -`/dev/nvidia-modeset` fd; the host must complete the exact notifier page the -kernel watches and signal that event, not only write the notifier memory. +DDX's reserve loop. For a windowed GL client the present is a kick; the GSP +event-IRQ fix now sets the RISC-V SWGEN0 IRQSTAT/IRQMASK/IRQDEST and latches +the CPU-tree leaf, so the guest services the event and glxgears completes +swaps. The next completion step is the constant buffer: the draw's CB page +must be resolved and copied into the seed page so the engine produces +fragments. During bring-up the module also carries a **diagnostics-only mirror** for apps that spin on a shared progress word; it exists so a run can proceed while the @@ -151,32 +155,35 @@ requires it switched off, along with any external poke. ## 6. Where the frontier is (ranked walls) -1. **Kernel-visible GL present completion** `[wall]`. With GLX forced the - windowed client uses `libnvidia-glcore`; the minimal `glxgears` then hangs - in the glcore kick wait: the client polls its per-slot - `/dev/nvidia-modeset` fd (`poll(POLLIN, 1 s)`). The host's sweep completes - the beef notifier memory, but the guest kernel never signals the event the - poll waits on. The exact notifier page the kernel watches must be - completed and the guest event signalled (the kick-notify executor), not - just the memory written. -2. **The X DDX push-buffer completion** `[wall]`. The DDX's 2D reserve loop +1. **The constant buffer (MVP) content** `[wall]`. The guest's GL draws now + execute on the physical GPU, but the engine writes 0 pages. The draw's + constant-buffer source is forced to the seed page because + `cb-copy src=0x0cd70000 ... first8=00000000 ... src=synthetic-id` reads all + zeros; a guest-RAM scan found no MVP matrix and the `0x3f800000` vidmem + hits are shader microcode immediates. With a zero MVP every vertex + collapses to the origin, so there are no fragments. Decode the CB-bind + method on the beef channel, resolve that VA through the channel VAS, and + copy the real page into the seed page. +2. **Execution stability** `[wall]`. The MME front-end is busy/deferred + (`MME-FE-DEFER` then forced loads) and the GR channel stale-wedges + (`GR-STALL-stale-wedge`, recovered), so the pipeline restarts mid-frame. + Fixing the CB page may unstick the MME; the wedge recovery needs a durable + fix. +3. **The X DDX push-buffer completion** `[wall]`. The DDX's 2D reserve loop exits only when the GPU-written engine get word advances (`get <= put`) and - the kick requires the notifier to be recycled to `0xffff` after - completion. A diagnostic host write released the loop; the kernel levers - to publish/advance the get word and recycle the notifier are next. -3. **The X DDX footprint** `[wall]`. The DDX allocates 4 GB of system pages at + the kick requires the notifier to be recycled to `0xffff` after completion. + `sema_vas_cand_ms=1` makes X green on attempt 1, but X can still die after a + few forward frames into the RM-degraded state. +4. **The X DDX footprint** `[wall]`. The DDX allocates 4 GB of system pages at startup; an 8 GB guest has no headroom once GZDoom's own footprint lands and both are OOM-killed. Current runs use a 12 GB guest; the allocation is a suspected emulation gap (a 512 MiB emulated VRAM falling back to a 4 GB sysmem allocation) to fix. -4. **Physical dispatch of the GL draws** `[wall]`. The direct-GL channels - submit and get host-completed releases, but the forwarding path reports no - dispatch to the physical GR engine (`forwarded=0`); the draws are decoded - and host-completed, not executed on hardware. 5. **Non-black frames and sustained rate** `[wall]`/`[target]`. Once the - kernel-visible present completion and the DDX get-word/notifier completion - land, the capture must show motion, then 60 fps in one VM with no forcing, - then two VMs on two mediated devices with non-black frames in both. + constant buffer, the MME/wedge stability and the DDX get-word/notifier + completion land, the capture must show motion, then 60 fps in one VM with + no forcing, then two VMs on two mediated devices with non-black frames in + both. ## 7. What is proven today @@ -186,21 +193,19 @@ requires it switched off, along with any external poke. | Display engine serves the modeset push buffers; notifier plus `UPDATE -> FINISHED` | `[live]` (campaign configuration) | | nvidia-modeset/NvKms registers (`/dev/dri/card0`, `renderD128`, `fb0`); `nvidia-smi -L` enumerates | `[live]` | | `glxinfo` reports the emulated GPU with the NVIDIA renderer (OpenGL 4.6.0, direct rendering) | `[live]` | -| GZDoom obtains a real GL context; windowed reaches renderer startup | `[live]` (pixels frontier) | -| GZDoom reaches `MAP01` on Mesa llvmpipe (not the emulated vGPU path) | `[wip]` | -| Forced GLX (`SDL_VIDEO_X11_FORCE_EGL=0`) selects `libnvidia-glcore` | `[live]` | -| Windowed GL present/kick completion (minimal `glxgears`) | `[wall]` | -| X server survives the GZDoom run on a 12 GB guest (the DDX's 4 GB sysmem allocation forces the size) | `[live]` | -| GL section capture: the X server's `0xC597` stream decoded | `[live]` | +| X server green with `sema_vas_cand_ms=1`; glxinfo 4.6.0 direct | `[live]` | +| GSP event-IRQ fix wakes the guest poll; glxgears completes swaps | `[live]` | +| 2D path end-to-end: accelerated X fills read back via `xwd` | `[live]` | +| Guest's real GL draws forward and execute on the physical GPU (`work_gr backend_ok=9`) | `[live]` | +| Beef GR channel `0xbeef0100` captures real GL draw sections | `[live]` | +| GL section capture: the X channel's `0xC597` stream decoded | `[live]` | | Verbatim 3D method-stream replay retired by the engine (no MMU fault) | `[live]` | | Synthesized launch executes real `sm_75` SASS; Turing `V3` program header accepted | `[live]` | -| 2D screen fills and copy-engine present ops execute | `[live]` | -| Direct GL client: device-init gate handled, channels allocate, readback path returns clean | `[live]` | -| Direct GL client channels submit and retire (GR `0xbeef0100`, CE `0xbeef0101`/`0xbeef0102`); DDX kick channel decoded | `[live]` | -| Guest kernel observes the kick completion and signals the modeset event | `[wall]` | -| NVIDIA Vulkan ICD creates a device | `[wall]` | +| X server survives the run on a 12 GB guest (the DDX's 4 GB sysmem allocation forces the size) | `[live]` | +| Constant-buffer (MVP) page resolves and the draw writes render-target pixels | `[wall]` | +| MME front-end defer / GR stale-wedge stability | `[wall]` | | DDX push-buffer get-word publish and notifier recycle | `[wall]` | -| Physical dispatch of the direct-GL draws (`forwarded=0`) | `[wall]` | +| NVIDIA Vulkan ICD creates a device | `[wall]` | | Non-black pixels from the emulated vGPU's 3D pipeline | `[wall]` | | Windows/DirectX guests | `[planned]` | @@ -208,11 +213,10 @@ requires it switched off, along with any external poke. * No non-black pixel has been produced by the emulated vGPU's 3D pipeline yet. * Display/KMS bring-up (NvKms, `/dev/dri`, X, a real GL context) is not a - rendered-frame claim; GZDoom reaching `MAP01` on Mesa llvmpipe is not the - emulated-vGPU path, and the later forced-GLX run selects `libnvidia-glcore` - but hangs in the present/kick wait. The kernel-visible present completion, - the DDX get-word/notifier completion and the physical dispatch of the GL - draws are the current walls. + rendered-frame claim; the guest's GL draws now execute on the physical GPU + and glxgears completes swaps, but the frames are black. The constant-buffer + (MVP) content, the MME/wedge stability and the DDX get-word/notifier + completion are the current walls. * No frame-rate or latency numbers: the scheduled measurement is defined (60 fps over 60 s, no forcing) but not yet met. * The `virgl`-based demo videos are a different, labelled path — not this one. -- cgit v1.2.3