From e77efad422961e5266ef30ef5ce9f49ae1fd76c1 Mon Sep 17 00:00:00 2001 From: Arthur Hanson Rasmusson Date: Wed, 21 Oct 2026 11:39:25 -0400 Subject: gr: r410 wedge quiet window - bounded idle hold before retrying a lane whose previous launch also no-wrote (r416: 58-launch sticky wedge clears after ~50 s quiet); wedge counters in state; wedge_recover v2 unmask The r416 boot reproduced the sticky no-execute wedge inside 60 s: 58 consecutive launches accepted zero writes while the lane stayed busy (wedge_recover v1's deterministic verify masked every one of them), and the first launches were green again only after the lane sat idle ~52 s. Both r409 (~37 s idle heal) and r416 therefore show the wedge clears on idleness and persists under back-to-back retries (r154/r409), so the retry loop now holds a quiet window (gsp_phys_gr_wedge_quiet_ms, default 20000, doubling on the second quiet retry, 0 disables) before the normal fresh-bank + resched retry when the previous launch also no-wrote. The whole-launch stall budget is relaxed only on this bounded recovery path. NOWRITE lines now carry the consecutive streak and quiet/heal counters, and the state file shows nw_streak/wedge_quiet/wedge_heals. wedge_recover v2 mutates the inputs every iteration so a stale (never executed) launch can no longer match the expected values, and classifies the readback STALE-PREV(NOWRITE) vs OTHER/PARTIAL. --- kernel/gsp_phys_gpu.h | 10 ++++++++++ 1 file changed, 10 insertions(+) (limited to 'kernel/gsp_phys_gpu.h') diff --git a/kernel/gsp_phys_gpu.h b/kernel/gsp_phys_gpu.h index c664e12..e02d6d8 100644 --- a/kernel/gsp_phys_gpu.h +++ b/kernel/gsp_phys_gpu.h @@ -399,6 +399,16 @@ int gsp_phys_gr_nowrite_retries_value(void); * and the sema service authors the QMD completion release in software. */ unsigned int gsp_phys_gr_wait_max_ms_value(void); +/* r410 wedge quiet window: when a launch's zero-write outcome follows a + * previous accepted zero-write launch (the sticky per-lane no-execute wedge), + * hold the retry for this many ms with NO submissions to the engine before + * re-publishing. The wedge is proven to clear when the lane stays quiet for + * tens of seconds (r409 ~37 s, r416 ~52 s); immediate retries keep it alive. + * Base value for the first quiet retry, doubled per further attempt, clamped + * [0,120000]; 0 disables the quiet path (retries run back-to-back as before). + * Default 20000. */ +unsigned int gsp_phys_gr_wedge_quiet_ms_value(void); + /* m73: use host-visible sysmem chunks for V7 instance/RAMFC (m62 shape). */ int gsp_phys_rmcall_gr_inst_sys_value(void); -- cgit v1.2.3