nvme: complete requests when payload DMA mapping fails
ClosedPublic
Actions

Authored by seuros on Mon, Aug 24, 6:19 PM.

Details

Reviewers

adrian
imp
ngie

Commits

rGb90c2b13e975: nvme: complete requests when payload DMA mapping fails

Summary

bus_dmamap_load_mem() reports most mapping failures, including EFBIG,
only through its callback and then returns zero. nvme_payload_map()
logged the error without telling the submission path, so the tracker
stayed on the outstanding list with no command submitted and no
timeout armed, stalling all later I/O on the queue behind it.

Approved by: ngie (co-mentor)
MFC after: 1 week

Diff Detail

Repository

rG FreeBSD src repository

Lint

Lint Passed

Unit

No Test Coverage

Build Status

Buildable 76056
Build 72939: arc lint + arc unit

Event Timeline

seuros created this revision.Mon, Aug 24, 6:19 PM

Herald added subscribers: dab, imp. · View Herald TranscriptMon, Aug 24, 6:19 PM

seuros requested review of this revision.Mon, Aug 24, 6:19 PM

Harbormaster completed remote builds in B76056: Diff 184919.Mon, Aug 24, 6:19 PM

seuros added reviewers: adrian, imp.Mon, Aug 24, 6:20 PM

seuros added a child revision: D59152: nvme: limit namespace character-device I/O size.

rebase

Harbormaster completed remote builds in B76110: Diff 184987.Tue, Aug 25, 6:34 AM

imp added inline comments.Tue, Aug 25, 7:03 PM

sys/dev/nvme/nvme_qpair.c
1181	Should we set this to EINPROGRESS here? Are there paths out which don't do the DMA that aren't a mapping failure?
1200	I'd think that we don't want to wait here. If it's just to detect this error, can't we detect it in the callback instead?

ngie added a subscriber: ngie.Sat, Aug 29, 2:24 AM

ngie added inline comments.

sys/dev/nvme/nvme_qpair.c
1200	+1. Doing this in the callback (if possible) seems much cleaner so you don't have to deal with the truth table of err x tr->map_err conditional combinations and the associated complexity in all of the downstream callers.