summaryrefslogtreecommitdiffstats
path: root/wiki/open-iov.org/index.php?title=User_talk:Arthur&action=edit&redlink=1.html
diff options
context:
space:
mode:
authorArthur Hanson Rasmusson <arthur@vgpu.io>2026-10-19 20:03:33 -0400
committerArthur Hanson Rasmusson <arthur@vgpu.io>2026-10-19 20:03:39 -0400
commitae7848021ec10dd25c48f3244c4bca7adc79f2d5 (patch)
tree1a34ab8b4f7d0027179189f4f8159cf4ee530d49 /wiki/open-iov.org/index.php?title=User_talk:Arthur&action=edit&redlink=1.html
parent79952c695ab929982a03996226ed180c2bbc58c7 (diff)
downloadopen-iov-ae7848021ec10dd25c48f3244c4bca7adc79f2d5.tar.gz
open-iov-ae7848021ec10dd25c48f3244c4bca7adc79f2d5.zip
vgsp: fail4 - route pending-pass releases through opdef withholding (cold-run gap)
The fwd-defer arm took priority over svc_opdef_release, so a pass whose op was still pending recorded its fence release in svc_fwd_rel (applied at the hold break / fallback) instead of the opdef slot - the fence fired while the copy was still owed, the driver refilled the ping-pong staging, and the late capture/read got refilled bytes (cold-run vector_add gap: one to four chunks per cold boot). svc_fwd_defer_arm() is now skipped when svc.op_tr_pending, so those releases go through svc_opdef_release and apply only when the retried pass has no pending op (the op has run by then). Measured cold-run vector_add bad across this iteration: 193536 -> 100352 (r323) -> 479232 (r324 wrong queue-pending experiment, reverted) -> 66560 (r325, ~1 chunk). All other acceptance items green on r325 (mail OK, smi OK, matmul OK, transformer OK).
Diffstat (limited to 'wiki/open-iov.org/index.php?title=User_talk:Arthur&action=edit&redlink=1.html')
0 files changed, 0 insertions, 0 deletions