On 10/21/17 6:41 AM, Stefan Sperling wrote: > On Sat, Oct 21, 2017 at 05:48:30AM -0400, RD Thrush wrote: >> On 10/20/17 1:07 PM, Stefan Sperling wrote: >>> On Fri, Oct 20, 2017 at 11:08:07AM -0400, [email protected] wrote: >>>>> Synopsis: bsd.rd hangs "iwm0: could not read firmware iwm-7265-16 (error >>>>> 2)" >>>>> Category: kernel >>>>> Environment: >>>> System : OpenBSD 6.2 >>>> Details : OpenBSD 6.2-current (RAMDISK_CD) #159: Fri Oct 20 >>>> 00:02:05 MDT 2017 >>>> >>>> [email protected]:/usr/src/sys/arch/amd64/compile/RAMDISK_CD >>>> >>>> Architecture: OpenBSD.amd64 >>>> Machine : amd64 >>>>> Description: >>>> I'm unable to finish booting bsd.rd. It hangs w/ the following text >>>> on the screen (transcribed from camera image): >>>> root on rd0a swap on rd0b dump on rd0b >>>> iwm0: could not read firmware iwm-7265-16 (error 2) >>>> iwm0: failed to load init firmware >>>> >>>> The hang has happened with the last few snaps. (Since I have not been >>>> able to test since the appended Sep 9 dmesg, the problem arose between >>>> Sep 9 and Oct 18). >>>> >>>>> How-To-Repeat: >>>> boot /bsd.rd >>> >>> Works fine here with an iwm 7265. >>> >>> Are you sure iwm is implicated? What happens if you disable the iwm >>> driver with config(8) or with 'boot -c'? >> >> As of "OpenBSD 6.2-current (RAMDISK_CD) #161: Fri Oct 20 20:32:01 MDT 2017", >> the message no longer appears and the install proceeds as expected. I've >> appended the associated sendbug info in case its useful. > > Which message no longer appears? > > Perhaps I should clarify that these messages are to be expected in bsd.rd: > > iwm0: could not read firmware iwm-7265-16 (error 2) > iwm0: failed to load init firmware > > When the root filesystem is mounted, the driver attemts to load firmware > files which are required to initialize the hardware. > And, in particular, to find the card's MAC address. > > The Intel firmware files aren't part of the bsd.rd root filesystem, which > means this attempt will fail and print errors. > > But the files will be found during upgrades as soon as the root filesystem > of the installed system is mounted at /mnt and network interfaces are > started by the upgrade script > > So these messages are expected. What is not expected is a system hang.
Thanks for the clarification. I'm probably mistaken with my comment that the iwm0: message no longer displays. I'm used to reviewing the serial interface logs on other OpenBSD boxes. These intel nucs don't have a serial interface and I must rely on my faulty memory about bsd.rd's dmesg. In any case, the hang did not occur w/ the forementioned snapshot.
