# Test status — latest numbers Snapshot of the current xts5 (X Test Suite) and rendercheck (RENDER smoke) pass rates. This file is the headline only; run-by-run history and debugging notes live in [`xts-baseline.md`](xts-baseline.md) and `status.md`. ## xts5 — full run #6, yserver/KMS bare-metal (eiger/aarch64, 2026-06-25) **3993 / 5987 test purposes PASS (66.7%)** — +32 vs run #5 (3961), FAIL 818 → 813, UNRES 80 → 67. Movers: XIproto +8, Xlib4 +4, Xlib8 +4, Xlib9 +4, Xlib7 +3, Xproto +3, Xlib12 / Xlib13 / Xt13 +2 each. For context, Xorg itself only passes 77% of the test suite. | scenario | cases | tests | PASS | FAIL | UNRES | UNTST | UNSUP | NOTIU | Δ PASS | |-----------|------:|------:|-----:|-----:|------:|------:|------:|------:|-------:| | Xproto | 122 | 389 | 361 | 6 | 1 | 19 | 2 | 0 | +3 | | Xlib3 | 109 | 162 | 107 | 18 | 3 | 27 | 6 | 1 | 0 | | Xlib4 | 29 | 324 | 171 | 113 | 7 | 17 | 11 | 5 | +4 | | Xlib5 | 15 | 84 | 59 | 18 | 0 | 5 | 2 | 0 | 0 | | Xlib6 | 8 | 50 | 6 | 15 | 0 | 29 | 0 | 0 | 0 | | Xlib7 | 58 | 172 | 86 | 28 | 0 | 13 | 45 | 0 | +3 | | Xlib8 | 29 | 165 | 92 | 37 | 4 | 22 | 10 | 0 | +4 | | Xlib9 | 46 | 1472 | 833 | 376 | 0 | 36 | 23 | 201 | +4 | | Xlib10 | 23 | 95 | 25 | 29 | 5 | 35 | 1 | 0 | 0 | | Xlib11 | 33 | 195 | 74 | 49 | 3 | 4 | 22 | 43 | 0 | | Xlib12 | 27 | 138 | 96 | 12 | 1 | 15 | 2 | 12 | +2 | | Xlib13 | 32 | 269 | 207 | 23 | 23 | 10 | 3 | 3 | +2 | | Xlib14 | 45 | 58 | 46 | 7 | 0 | 5 | 0 | 0 | 0 | | Xlib15 | 45 | 159 | 125 | 1 | 0 | 33 | 0 | 0 | 0 | | Xlib16 | 30 | 105 | 82 | 0 | 0 | 22 | 1 | 0 | 0 | | Xlib17 | 55 | 131 | 102 | 8 | 0 | 21 | 0 | 0 | 0 | | Xopen | 8 | 127 | 122 | 3 | 0 | 0 | 2 | 0 | 0 | | Xt3 | 21 | 73 | 73 | 0 | 0 | 0 | 0 | 0 | 0 | | Xt4 | 33 | 192 | 94 | 0 | 0 | 98 | 0 | 0 | 0 | | Xt5 | 10 | 69 | 26 | 0 | 0 | 41 | 0 | 0 | 0 | | Xt6 | 7 | 71 | 67 | 4 | 0 | 0 | 0 | 0 | 0 | | Xt7 | 11 | 106 | 96 | 1 | 0 | 6 | 0 | 3 | 0 | | Xt8 | 7 | 43 | 35 | 4 | 0 | 4 | 0 | 0 | 0 | | Xt9 | 33 | 189 | 122 | 2 | 8 | 55 | 2 | 0 | 0 | | Xt10 | 8 | 17 | 16 | 0 | 0 | 1 | 0 | 0 | 0 | | Xt11 | 58 | 285 | 246 | 3 | 0 | 34 | 0 | 0 | −1 | | Xt12 | 22 | 67 | 55 | 0 | 1 | 11 | 0 | 0 | 0 | | Xt13 | 39 | 178 | 126 | 5 | 0 | 47 | 0 | 0 | +2 | | Xt14 | 2 | 18 | 18 | 0 | 0 | 0 | 0 | 0 | 0 | | Xt15 | 1 | 2 | 0 | 0 | 0 | 0 | 2 | 0 | 0 | | XtC | 29 | 147 | 88 | 1 | 1 | 56 | 1 | 0 | 0 | | XtE | 1 | 1 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | | ShapeExt | 11 | 11 | 11 | 0 | 0 | 0 | 0 | 0 | 0 | | XI | 36 | 316 | 222 | 49 | 10 | 28 | 2 | 5 | +1 | | XIproto | 35 | 107 | 103 | 1 | 0 | 3 | 0 | 0 | +8 | | **total** | **1078** | **5987** | **3993** | **813** | **67** | **697** | **137** | **273** | **+32** | ShapeExt, Xlib16 and Xt3/4/5/10/14 are fully clean (zero FAIL/UNRES). yserver survived the whole sweep with zero panics in the server log; 2 NORESULTs (`Xt5/XtUnmanageChild(ren)`, unchanged). Movement vs #5 came from regular bugfixing, no focused xts work: XIproto +8 (UNTST → PASS), Xlib4/8/9 +4 each, Xlib7/Xproto +3. The minor new FAILs (XtC +1, Xt11 −1 PASS fontset/resource, XIproto ChangeFeedbackControl) are unrelated to the e27 input fixes — those touched XI2 delivery, which xts5 doesn't exercise. Largest FAIL buckets / next targets: 1. **Xlib9 (376)** — remaining drawing/GetImage content semantics. 2. **Xlib4 (113)** — depth-mismatch BadMatch (CWBorderPixmap parser needed), colormap visual-type checks, bit-gravity pixel cluster, stacking-order pixel checks, BadAccess event-mask conflicts. 3. **Xlib11 (49)** — residual grab semantics. 4. **XI (49)** — XInput-1.x device functions + XTest-through-XI1 gaps. 5. **Xlib8 (37)** / **Xlib7 (28)** — events / colormap sections. Previous full runs: - #5 — 2026-06-07 22:03:01 (bee, HW): 3961/5987 PASS (66.2%) — `xts/results/2026-06-07-22:03:01/`. - #4 — 2026-06-07 17:14:17 (bee, HW): 3747/5987 PASS (62.6%) — `xts/results/2026-06-07-17:14:17/`. Last run before the Xlib4 BadX work and the desktop-input-fixes branch. - #3 — 2026-06-06 (air, M1): 3419/5987 PASS (57.1%) — `xts/results/2026-06-06-20:26:54/`. - #2 — 2026-06-05 (M2) + 2026-06-06 air XI row: 3370/5987 PASS (56.3%) — `xts/results/2026-06-05-13:20:07/` (+ `2026-06-06-00:58:03` for XI). - #1 — 2026-06-04 (first ever to complete): 2784/5987 PASS (46.5%) — `xts/results/2026-06-04-15:48:44/`. Aborted run between #3 and #4 (`xts/results/2026-06-07-14:01:34/`): 2999/5987 PASS, 1290 UNRES — the GetImage BadMatch cascade caused by an unguarded `XConfigureWindow` on the root window, fixed by `77f785b` before run #4. ## rendercheck — bare-metal 2026-06-04, rendercheck 1.6, 900 s/test | category | PASS | TOTAL | |-------------|------:|------:| | fill | 64 | 64 | | dcoords | 2 | 2 | | scoords | 1 | 1 | | mcoords | 1 | 1 | | tscoords | 2 | 2 | | tmcoords | 2 | 2 | | blend | 5 | 5 | | composite | 5 | 5 | | cacomposite | 5 | 5 | | gradients | 6081 | 6081 | | repeat | 380 | 380 | | triangles | 570 | 570 | | bug7366 | 1 | 1 | | **total** | **7119** | **7119** | **100% pass.** > Use rendercheck ≥ 1.6. Version 1.5 has a bug in > `gradients::render_to_gradient_test` that trips even against the > host X server.