Re: loader_4th.efi "IPv4 PXE" net booting no longer works
- In reply to: Bjoern A. Zeeb: "Re: loader_4th.efi "IPv4 PXE" net booting no longer works"
- Go to: [ bottom of page ] [ top of archives ] [ this month ]
Date: Sat, 03 Oct 2026 23:50:50 UTC
On Sat, 3 Oct 2026, Bjoern A. Zeeb wrote: > On Sat, 26 Sep 2026, Warner Losh wrote: > >> On Sat, Sep 26, 2026 at 7:36 AM Bjoern A. Zeeb < >> bzeeb-lists@lists.zabbadoz.net> wrote: >> >>> On Sat, 26 Sep 2026, Warner Losh wrote: >>> >>>> On Sat, Sep 26, 2026 at 6:17 AM Bjoern A. Zeeb < >>>> bzeeb-lists@lists.zabbadoz.net> wrote: >>>> >>>>> Hi, >>>>> >>>>> sorry I cannot gather more parts of the console before this easily; this >>>>> is a remote machine booting via IPv4 (just in case you wonder with me). >>>>> >>>>> It's a EDK II firmware, boot options is IPv4 PXE. Boot file is >>> loader.efi, >>>>> which is loader_4th.efi in that setup. >>>>> >>>>> I updated to a kernel/world from the last 48 hours and the boot hangs: >>> .. >>>>> >>>>> Do I need to start a bisect? >>>>> >>>> >>>> If you have multiple NICs, it may be that we assume unit 0 in the >>>> committed code. I got a report yesterday thta I saw just before >>>> I was heading to bed, so I worked up >>>> >>>> http://people.freebsd.org/~imp/net-unit.patch >>>> >>>> to see if it fixed the problem for the reporter and sent it off. I then >>>> told claude to add a two-nic booting from the second nic scenario >>>> to boot-test.sh and went to bed. Now I wake up to this :) I'll see >>>> if that tests things, but in the mean time, it wouldn't hurt to try >>>> the above. >>> >>> I'll give it a shot as soon as I can reboot again (may be a few hours >>> still) >>> but the machine in question here has a single NIC only. >>> >>> If it does not help I'll try to narrow things down and report back. >>> >> >> If my patch doesn't work, try commenting out the call >> to maybe_download_initmd() in main.c > > I could not yet do that. The patch works and seems to make things boot again. Thank you so much! >> Finally, how long did you wait at the end? There's a 5 minute timeout that >> might tell us something if it times out vs other behavior. > > I got another test machine I could at least boot once but that indeed does > have multiple interfaces. > > It says: > > Setting currdev to md0: > MD not present > Setting currdev to net0: > Portmapper returned 0. TCP-only NFS server? > Portmapper returned 0. TCP-only NFS server? > > /bz > > -- For there shall arise false Christs, and false prophets, and shall shew great signs and wonders; insomuch that, if it were possible, they shall deceive the very elect. Behold, I have told you before. [Mat. 24:24-25]