From 157eb059a2d65a2f0aafbccc6872c221710b86e1 Mon Sep 17 00:00:00 2001 From: Arthur Hanson Rasmusson Date: Mon, 12 Oct 2026 05:09:17 -0400 Subject: vgsp-sema: re-arm ring/push-buffer discovery latches on channel registration; scanner streams anonymous high-VA buffers Second-client wedge fix: a new GSP session re-registers the same channel handles/ranks, but the per-rank one-shot ring/push-buffer discovery latches stayed set from the previous session, so the new session's ring was never re-autopinned (ring_auto=0 pb=0/0, nomap growing) and the guest spun in its channel-creation wait. svc_chan_rearm_rank() clears the latches on every channel registration; all post-fix cycles launch the second client. Scanner: guest_pin_stream.py now streams the client's anonymous high-VA mappings (cudaMalloc buffers were skipped by the old len(parts)<6 check), so the host can learn their pages. Diagnostics: PT-window lines, per-window first words, post-OK qmdwin w48..59. Remaining: the fill-time svc_ipin_lookup miss for the second client's a/b/c pins (they exist in the pin log but are not seen at fill time) - next session. --- tools/guest-diag/guest_pin_stream.py | 15 ++++++++++++--- 1 file changed, 12 insertions(+), 3 deletions(-) (limited to 'tools') diff --git a/tools/guest-diag/guest_pin_stream.py b/tools/guest-diag/guest_pin_stream.py index 04aba73..739aeca 100644 --- a/tools/guest-diag/guest_pin_stream.py +++ b/tools/guest-diag/guest_pin_stream.py @@ -778,10 +778,19 @@ def scan_vmas(fd, pid, out): try: for line in maps: parts = line.split() - if len(parts) < 6: + if len(parts) < 5: continue - path = parts[5] - if not any(p in path for p in VMA_PATHS): + # Anonymous mappings (no path field) are the client's UVM sysmem + # buffers when the driver maps them without a /dev/zero backing: + # the second client's cudaMalloc'd data buffers show up exactly + # like that (e.g. 75d7b0000000-75d7b0021000 rw-p 00000000 00:00 0) + # and were skipped by the old len<6 check, so their pages never + # reached the host and the forwarded launch could not resolve + # them (all-zero results for the second client). Accept them at + # the high-VA range; the VMA_LO bound and the pagemap present/PFN + # filters keep the set to real GPU mappings. + path = parts[5] if len(parts) >= 6 else '' + if path and not any(p in path for p in VMA_PATHS): continue try: a, b = parts[0].split('-') -- cgit v1.2.3