From b8004cdcce75f9cbe0070f9f52b49fa914bdebf3 Mon Sep 17 00:00:00 2001 From: Arthur Hanson Rasmusson Date: Sun, 15 Nov 2026 14:11:01 -0500 Subject: docs(wiki): 2026-11-15 status refresh - DS stage image authored+live (st0=st1=020a0481; DS 0x84000420, two boots); draw-issue mac=0x22 and stores remain; STATE-CURRENT snapshot --- wiki/Execution-Frontier.md | 48 ++++++++++++++++++++++++++-------------------- 1 file changed, 27 insertions(+), 21 deletions(-) (limited to 'wiki/Execution-Frontier.md') diff --git a/wiki/Execution-Frontier.md b/wiki/Execution-Frontier.md index 5f0acfc..1425aaf 100644 --- a/wiki/Execution-Frontier.md +++ b/wiki/Execution-Frontier.md @@ -14,8 +14,7 @@ > [`../docs/04-implementation/03-execution-frontier.md`](../docs/04-implementation/03-execution-frontier.md); > the docs tree wins on disagreement. -Status: 2026-11-15, tree `master @ 444ae2c1` plus in-flight graphics changes in -the working tree. +Status: 2026-11-15, tree `master @ 692a7eb0`. Tags: `[live]` observed; `[wip]` partial; `[wall]` current blocker; theories are labelled as such. @@ -120,16 +119,20 @@ keep-windows land, the part-launch poison is eliminated (the client's own real fault; they are redirected to a mapped scratch slot), the 2799-word pack completes (2799/2799), the first `MME-PACK-REARM` fires, the fixed-CB region is aliased, and `DRAW_IAB` (`mac=0x22`) runs with the SPH-BIND prelude intact -(84 words) and the DS stage VAs populated. What is missing is -**window/screen pixels through the NVIDIA path**: the DS stage-0/1 image is -the last named gate (the synthesized binds read `family+0x900`, which is zero -in the live family page, while the client's VTG SPH image is at -`family+0x600`), so the draw never stores (`wrote=0`, `px0=0`); the -window/screen present path is unproven, glxgears' window is black, and a -synchronous window draw wedges X on some boots. The per-boot binding and X -first-generation start lottery are now the dominant throughput limiter. The -guest Vulkan pipeline's inline-to-memory upload arm is its gap. The full -picture lives on [`Graphics-Execution.md`](Graphics-Execution.md). +(84 words) and the DS stage VAs populated. The **DS stage-0/1 image is now +authored and live**: a VTG SPH at `family+0x900` (`SPH-AUTHOR`, first word +`020a0481`) is what both vertex slots fetch (`DS-IMAGE st0=st1=020a0481 +code=ff007321`), the window's own later slot-1 bind is rewritten to it, and +the DS value fell `0x86040626 -> 0x84000420` (reproduced on two boots). What +is missing is **window/screen pixels through the NVIDIA path**: the remaining +DS bits are stage type/context (SPH5/SPH10/SPH18), the 728-B draw window's +released macros (`0x33/0x39/0x1/0x0`) do not include `DRAW_IAB` (`mac=0x22`), +so the draw still does not store (`wrote=0`, `px0=0`); the window/screen +present path is unproven, glxgears' window is black, and a synchronous window +draw wedges X on some boots. The per-boot binding and X first-generation +start lottery are now the dominant throughput limiter. The guest Vulkan +pipeline's inline-to-memory upload arm is its gap. The full picture lives on +[`Graphics-Execution.md`](Graphics-Execution.md). No breadth beyond these is claimed: Windows guests, migration and performance numbers remain planned. @@ -198,10 +201,12 @@ captures and reaches the engine. The MME chain now completes and executes the draw macro: multi-frame keep-windows land, the 2799-word pack completes, the first `MME-PACK-REARM` fires, the fixed-CB region is aliased, and `DRAW_IAB` (`mac=0x22`) runs with the SPH-BIND prelude intact and the DS stage VAs -populated. The remaining gap is the **DS stage-0/1 image** (the synthesized -binds read `family+0x900`, which is zero in the live family page, while the -client's VTG SPH image is at `family+0x600`; the `+0x600` bind is in flight) -so the draw stores. glxgears' window is black, a window draw wedges X on some +populated. The **DS stage-0/1 image is now authored and live** (a VTG SPH at +`family+0x900`, fetched by both vertex slots as `st0=st1=020a0481`; DS fell +`0x86040626 -> 0x84000420`, reproduced on two boots); the remaining bits are +stage type/context (SPH5/SPH10/SPH18), and the 728-B window's released macros +(`0x33/0x39/0x1/0x0`) do not include the `DRAW_IAB` call, so the draws still +do not store. glxgears' window is black, a window draw wedges X on some boots, and the per-boot binding plus X first-generation start lottery is the dominant throughput limiter ([`Graphics-Execution.md`](Graphics-Execution.md)). @@ -218,11 +223,12 @@ dominant throughput limiter 4. **Hold the acceptance battery** — vector-add, matrix-multiply and the transformer kernel; each entry is accepted only on its own verification output (`bad=0` + `STATUS OK`). -5. **The graphics pixel frontier** — finish the DS stage-0/1 image bind - (`family+0x600`) so the executed `DRAW_IAB` macro stores (`wrote>0`) at the - bound render target, then make the window/screen present land, and reduce - the per-boot binding/X start lottery so the proven off-screen frames become - window pixels; the lanes, the walls and the milestone ladder are on +5. **The graphics pixel frontier** — the DS stage image is authored and live + for both vertex slots; next get the draw-issue `DRAW_IAB` (`mac=0x22`) call + into the window so it stores (`wrote>0`) at the bound render target, then + make the window/screen present land, and reduce the per-boot binding/X + start lottery so the proven off-screen frames become window pixels; the + lanes, the walls and the milestone ladder are on [`Graphics-Execution.md`](Graphics-Execution.md). ## 7. Theories, resolved -- cgit v1.2.3