Re: mips64eb seems to be mostly 32-bit
Simon Burge <[email protected]>
| Newsgroups | gmane.os.netbsd.ports.mips.devel |
|---|---|
| Message-ID | <[email protected]> |
Hi Denis, Denis Ovsienko wrote: > On Sun, 4 Jul 2021 17:01:14 +0200 > Martin Husemann <[email protected]> wrote: > > > On Sun, Jul 04, 2021 at 12:59:02PM +0100, Denis Ovsienko wrote: > > > [ 3.1606850] boot device: sd0 > > > [ 3.1708506] root on sd0a dumps on sd0b > > > > This is clearly wrong. Do you have any bootargs set in u-boot? > > NetBSD kernel receives no arguments, as far as I can tell: > > U-Boot 1.1.1 (UBNT Build Version: e102_003_eace7) (May 27 2019 - > 06:35:40) > > fatload usb 0 $(loadaddr) netbsd > bootoctlinux $(loadaddr) numcores=$(numcores) endbootargs The problem here is that "fatload usb 0 $(loadaddr) netbsd" doesn't leave any info at all on how or where that kernel was loaded from (that I'm aware of, certainly not in a uboot variable). If you look at how Linux boots on an ER-4 (I don't have a USB in my Erlite at the moment), you'll see bootoctlinux ... root=/dev/mmcblk0p2 We need to do something similar for NetBSD. For the DD images that have root on dk1 and a wedge called "octeon-root", either of the following work fatload usb 0 $(loadaddr) netbsd bootoctlinux $(loadaddr) root=dk1 or fatload usb 0 $(loadaddr) netbsd bootoctlinux $(loadaddr) root=wedge:octeon-root Both of these will boot like this: [ 2.7199912] dk0 at sd0: "octeon-boot", 163840 blocks at 32768, type: msdos [ 2.7199912] dk1 at sd0: "octeon-root", 15458304 blocks at 196608, type: ffs ... [ 2.7410579] boot device: sd0 [ 2.7410579] root on dk1 [ 2.7499912] root file system type: ffs Cheers, Simon.