Re: ZFS deadlock in 14

Dag-Erling_Smørgrav <des_at_FreeBSD.org> wrote on
Date: Fri, 18 Aug 2023 14:16:12 UTC :

> Dag-Erling Smørgrav <des@FreeBSD.org> writes:
> > A kernel built from c47116e909 plus these two patches has so far built
> > 2,285 packages without a hitch, whereas normally it would have
> > deadlocked after well before reaching 500 packages. I'll do another run
> > without the patches tomorrow just to be sure.
> 
> Plot twist: c47116e909 _without_ the patches also appears to be working
> fine. The last kernel I know for sure deadlocks is b36f469a15, so I'm
> going to test cd25b0f740 and 28d2e3b5de.

I'm confused:

Tue, 18 Jul 2023
. . .
    • git: b36f469a15ec - main - zfs: set autotrim default to 'off' Yuri Pankov

But you had written:

Dag-Erling_Smørgrav <des_at_FreeBSD.org>
Date: Tue, 15 Aug 2023 17:37:45 UTC 
The attached script successfully deadlocks 9228ac3a69c4

where:

Thu, 10 Aug 2023
. . .
    • git: 9228ac3a69c4 - main - ixgbe: Add support for 82599 LS Kevin Bowling

(You had previously reported a poudriere deadlock there as well.)

Later on 2023-Aug-10 there is (about 11 or so commits later):

    • git: cd25b0f740f8 - main - zfs: cherry-pick fix from openzfs Martin Matuska 
    • git: 28d2e3b5dedf - main - zfs: cherry-pick fix from openzfs Martin Matuska


I'll note there was also this:

Dag-Erling_Smørgrav <des_at_FreeBSD.org>
Date: Sat, 12 Aug 2023 14:11:10 UTC 
. . .
Trying to narrow this range down, I did not get a deadlock with
4e8d558c9d1c (10 June) but I did with b7198dcfc039 (16 June), albeit
after building ~1800 packages.
. . .

So the count seems to sometimes go well above 500 or so.

===
Mark Millard
marklmi at yahoo.com