just one of five disks detected at arcmsr(4) since upgrade to NetBSD 10

[Bert Kiers <kiers%original.xs4all.nl@localhost>]

Hi,

This morning I upgraded a box from 9.3_STABLE to NetBSD 10.0. With 
NetBSD 9.3 it probed five disks at the Areca ARC-1222 RAID controller, 
now just one. A powercycle and "scsictl scsibus0 scan all all" did not help.


NetBSD 9.3:

Apr 18 12:22:20 janeway /netbsd: [   1.0262686] arcmsr0 at pci1 dev 0 
function 0
Apr 18 12:22:20 janeway /netbsd: [   1.0262686] arcmsr0: interrupting at 
ioapic0 pin 16
Apr 18 12:22:20 janeway /netbsd: [   1.0262686] arcmsr0: Areca ARC-1222 
Host Adapter RAID controller
Apr 18 12:22:20 janeway /netbsd: [   1.0262686] arcmsr0: 8 ports, 256MB 
SDRAM, firmware <V1.51 2012-07-04>
Apr 18 12:22:20 janeway /netbsd: [   1.0262686] scsibus0 at arcmsr0: 16 
targets, 8 luns per target
...
Apr 18 12:22:20 janeway /netbsd: [   1.2718484] sd0 at scsibus0 target 0 
lun 0: <HGST, HDN724040ALE640, R001> disk fixed
Apr 18 12:22:20 janeway /netbsd: [   1.2718484] sd0: 3726 GB, 64600 cyl, 
252 head, 480 sec, 512 bytes/sect x 7814037168 sectors
Apr 18 12:22:20 janeway /netbsd: [   1.2718484] sd1 at scsibus0 target 0 
lun 1: <HGST, HDN724040ALE640, R001> disk fixed
Apr 18 12:22:20 janeway /netbsd: [   1.2718484] sd1: 3726 GB, 64600 cyl, 
252 head, 480 sec, 512 bytes/sect x 7814037168 sectors
Apr 18 12:22:20 janeway /netbsd: [   1.2718484] sd2 at scsibus0 target 0 
lun 2: <HGST, HDN724040ALE640, R001> disk fixed
Apr 18 12:22:20 janeway /netbsd: [   1.2718484] sd2: 3726 GB, 64600 cyl, 
252 head, 480 sec, 512 bytes/sect x 7814037168 sectors
Apr 18 12:22:20 janeway /netbsd: [   1.2718484] sd3 at scsibus0 target 0 
lun 3: <HGST, HDN724040ALE640, R001> disk fixed
Apr 18 12:22:20 janeway /netbsd: [   1.2718484] sd3: 3726 GB, 64600 cyl, 
252 head, 480 sec, 512 bytes/sect x 7814037168 sectors
Apr 18 12:22:20 janeway /netbsd: [   1.2718484] sd4 at scsibus0 target 0 
lun 4: <HGST, HDN724040ALE640, R001> disk fixed
Apr 18 12:22:20 janeway /netbsd: [   1.2718484] sd4: 3726 GB, 64600 cyl, 
252 head, 480 sec, 512 bytes/sect x 7814037168 sectors
Apr 18 12:22:20 janeway /netbsd: [   1.2811828] autoconfiguration error: 
arcmsr0: unable to query firmware for sensor info


NetBSD 10.0:

Apr 18 14:08:20 janeway /netbsd: [   1.0299139] arcmsr0 at pci1 dev 0 
function 0
Apr 18 14:08:20 janeway /netbsd: [   1.0299139] arcmsr0: interrupting at 
ioapic0 pin 16
Apr 18 14:08:20 janeway /netbsd: [   1.0299139] arcmsr0: Areca ARC-1222 
Host Adapter RAID controller
Apr 18 14:08:20 janeway /netbsd: [   1.0299139] arcmsr0: 8 ports, 256MB 
SDRAM, firmware <V1.51 2012-07-04>
Apr 18 14:08:20 janeway /netbsd: [   1.0299139] scsibus0 at arcmsr0: 16 
targets, 8 luns per target
...
Apr 18 14:08:20 janeway /netbsd: [   1.0700815] sd0 at scsibus0 target 0 
lun 0: <HGST, HDN724040ALE640, R001> disk fixed
Apr 18 14:08:20 janeway /netbsd: [   1.0700815] sd0: 3726 GB, 64600 cyl, 
252 head, 480 sec, 512 bytes/sect x 7814037168 sectors
...
Apr 18 14:08:20 janeway /netbsd: [   1.0765269] autoconfiguration error: 
arcmsr0: unable to query firmware for sensor info


root@janeway:/home/kiers# sysctl hw.disknames
hw.disknames = sd0 wd0 wd1


The card has ethernet with a webinterface and everything is fine there.
The five disks are a zpool, but that seems irrelevant.


Any ideas before I send-pr?

Thanks,
Bert