On Mon, Mar 03, 2025 at 04:02:46PM +0100, Mark Kettenis wrote: > > Date: Sun, 2 Mar 2025 21:34:06 -0500 > > From: Josh Grosse <[email protected]> > > Hi Josh, > > > On Mon, Feb 24, 2025 at 04:05:48PM +0100, Hrvoje Popovski wrote: > > > On 24.2.2025. 14:48, Mark Kettenis wrote: > > > >> Date: Mon, 24 Feb 2025 10:07:51 +0100 > > > >> From: Hrvoje Popovski <[email protected]> > > > > Hi Hrvoje, > > > > > > > >> Hi all, > > > >> > > > >> today I've sysupgrade my box and sysupgrade won't finish. It stops > > > >> after > > > >> > > > >> > > > >> Location of sets? (cd0 disk http nfs or 'done') [done] done > > > >> Making all device nodes... done > > > >> installboot: EFIIOC_VAR_SET: Boot000A: Invalid argument > > > > Interesting. That is not expected. Can you install the efivar > > > > package on that machine and send me the output of > > > > > > > > # efivar -l > > > > > > > > for that machine? > > > > > > Hi, > > > > > > here's output from command above > > > > > > snifi# efivar -l > > > 8be4df61-93ca-11d2-aa0d-00e098032b8c-SetupMode > > > > I've just upgraded to a new snapshot, and installboot failed during > > the upgrade because of the -c option, with: > > > > installboot: EFIIOC_VAR_NEXT: Function not implemented > > This may be related. > > > > A dmesg is attached. > > > > OpenBSD 7.7-beta (GENERIC) #569: Sun Mar 2 11:59:21 MST 2025 > > [email protected]:/usr/src/sys/arch/amd64/compile/GENERIC > > real mem = 17034862592 (16245MB) > > avail mem = 16492183552 (15728MB) > > random: good seed from bootblocks > > mpath0 at root > > scsibus0 at mpath0: 256 targets > > mainbus0 at root > > bios0 at mainbus0: SMBIOS rev. 2.7 @ 0xbcd2d000 (66 entries) > > bios0: vendor LENOVO version "GMET91WW (2.39 )" date 06/03/2021 > > bios0: LENOVO 20BE00BTUS > > efi0 at bios0: UEFI 2.3.1 > > efi0: Lenovo rev 0x2390 > > acpi0 at bios0: ACPI 5.0 > > landry@ ran into this on a very similar Lenovo machine (same EFI and > ACPI versions). I just committed a change that turned the error in a > warning which should make the installer continue. > > The error itself is unexpected, but I believe it is the result of a > buggy EFI implementation. Google tells me this is a T540p, so this > hardware is a decade old by now and UEFI 2.3.1 is quite old as well. > So I'm not surprised about those bugs and I'm just going to accept > that configuring the EFI boot manager won't work on these machines.
Yup, T540p, with the latest and greatest BIOS from 2021. Thanks for looking into it, and thank you for making the change to installboot to circumvent Lenovo's bug.
