nvme: Replace potentially long DELAY()'s with pause()'s
ClosedPublic
Actions

Authored by mav on Mar 17 2021, 2:45 AM.

Details

Reviewers

Commits

rGd00a723a2a5c: nvme: Replace potentially long DELAY() with pause().
rGed407c92e43f: nvme: Replace potentially long DELAY() with pause().
rG4fbbe523653b: nvme: Replace potentially long DELAY() with pause().

Summary

In some cases nvme(4) may wait minutes for controller response before timeout, for example in case of broken hardware. Doing so in a tight spin loop makes whole system unresponsive.

Test Plan

Without the patch hot-plug of broken NVMe SSD makes system console stuck for two minutes. With the patch system remains responsive and just prints "controller ready did not become 1 within 120500 ms" two minutes later.

Diff Detail

Repository

rG FreeBSD src repository

Lint

Lint Not Applicable

Unit

Tests Not Applicable

Event Timeline

mav created this revision.Mar 17 2021, 2:45 AM

Herald added a subscriber: dab. · View Herald TranscriptMar 17 2021, 2:45 AM

mav requested review of this revision.Mar 17 2021, 2:45 AM

Seems reasonable... we used to run this before interrupts were enabled, and now we don't so the DELAYs are a bit OBE now anyway.

sys/dev/nvme/nvme_ctrlr.c
273	Do we need to worry about this being 0 when hz < 1000?
1557	Ditto...

mav added inline comments.Mar 17 2021, 5:06 AM

sys/dev/nvme/nvme_ctrlr.c
273	No for the last 10 years since 9e3ae31c7a935255092fff24544c3d120bbd15bc: "Given that the typical usage of pause() is pause("zzz", hz / N), where N can be greater than hz in some cases, simply ignore a timeout value of zero." The zero there is handled as one.