From d9645c7d200ddd806148c219bc3335ac708cce65 Mon Sep 17 00:00:00 2001 From: Arthur Hanson Rasmusson Date: Mon, 16 Nov 2026 14:40:34 -0500 Subject: vgsp: eng89j - replay the whole admitted window (40 chunks, 48-slot ring) eng89i admitted the 14837-word window and chunked it into 6 chunks, but 6*384 words covers only 18% of the window - the DRAW_IAB call at the end was never replayed. Raise SVC_GL_STREAM_CHUNK_MAX 6 -> 40 (39 chunks for 14837 words at the proven 384-word chunk) and the candidate ring 32 -> 48 so the chunks are not evicted before the pump drains them. Context: c952-c957 submitted chunk 0 and got a FECS priv-error (badf1201, mme_pc=513 mac=0xf1) before the release; c956 reached MME-CARRY-ARM (off=1632/2799) and CALL-HOLD window=121, then CANCEL reason= chan-ctx-reset. The chunk-0 poison and the carry fire remain the next gates after full-window coverage. --- kernel/gsp_sema_service.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/kernel/gsp_sema_service.c b/kernel/gsp_sema_service.c index 4b3c24c..3ef6e75 100644 --- a/kernel/gsp_sema_service.c +++ b/kernel/gsp_sema_service.c @@ -9250,7 +9250,7 @@ struct svc_gr_cand_rec { * evict its own later chunks (or the X channel's pending candidate) before * the pump drained them. */ -#define SVC_GR_CAND_RING 32 +#define SVC_GR_CAND_RING 48 static struct svc_gr_cand_rec svc_gr_cand_ring[SVC_GR_CAND_RING]; static u32 svc_gr_cand_pos; #define svc_gr_cand (svc_gr_cand_ring[svc_gr_cand_pos]) @@ -11761,7 +11761,7 @@ MODULE_PARM_DESC(sema_screen_blit, * section's base candidate; only stream_off/stream_n differ, so every * chunk launch carries the same release identity and program images. */ -#define SVC_GL_STREAM_CHUNK_MAX 6 +#define SVC_GL_STREAM_CHUNK_MAX 40 /* * Words per chunk: 384 is the largest window the physical engine is * PROVEN to retire (r997: `3D stream marked h=0x1000013 -- cgit v1.2.3