Date: Wed, 9 Dec 2020 23:45:06 +0100 From: Alban Hertroys <haramrae@gmail.com> To: John Kennedy <warlock@phouka.net> Cc: freebsd-current@freebsd.org Subject: Re: KLD zfs.ko: depends on kernel - not available or version mismatch Message-ID: <3027159B-1851-4005-A569-059A1CAACFCC@gmail.com> In-Reply-To: <X9A7GHdxrVlf4Jj7@phouka1.phouka.net> References: <42AC7323-5AD6-401D-9A7D-F1D962EE5717@gmail.com> <X8%2BebiETBdNpbRXt@phouka1.phouka.net> <2B044A92-500F-4121-85DB-D486865C75B5@gmail.com> <X9A7GHdxrVlf4Jj7@phouka1.phouka.net>
next in thread | previous in thread | raw e-mail | index | archive | help
> On 9 Dec 2020, at 3:48, John Kennedy <warlock@phouka.net> wrote: >=20 >> I had to copy over several files from /etc and /usr/local/etc and = re-installed the most important packages. This was admittedly a bit = messy, it is possible that I forgot to copy something over. >> (Originally my intention was to dd the contents of the spinning disk = over, but apparently that disk has a few wonky sectors, dd failed after = a few device timeouts) >=20 > ... so, no guarantee that things are totally sane. >=20 > The "sane" we're looking for is how you can presumably be booting a = kernel > located at /boot/kernel/kernel and not have it match the kernel = modules > found under /boot/kernel. (...) > What I have built in my source tree is the kernel/zfs module I'd = expect: >=20 > # md5 -r /usr/obj/usr/src/amd64.amd64/sys/GENERIC/kernel = /usr/obj/usr/src/amd64.amd64/sys/GENERIC/modules/usr/src/sys/modules/zfs/z= fs.ko /boot/kernel/kernel /boot/kernel/zfs.ko | sort > 941ab52d075e444da6eea7fb56213e10 /boot/kernel/kernel > 941ab52d075e444da6eea7fb56213e10 = /usr/obj/usr/src/amd64.amd64/sys/GENERIC/kernel > 97d4e0c8ffed1f75e924bf8768a95ff1 /boot/kernel/zfs.ko > 97d4e0c8ffed1f75e924bf8768a95ff1 = /usr/obj/usr/src/amd64.amd64/sys/GENERIC/modules/usr/src/sys/modules/zfs/z= fs.ko >=20 > What are you seeing after your installkernel equivalent? It turns out that I was being fooled by the BIOS. Even though I selected = the device that this kernel and modules were on as the device to boot = from, the actual kernel was still coming from the old spinning disk! It would probably have taken me significantly longer to figure that out = without your hints, as I was trying to solve the wrong problem. So = thanks a lot for that. Having things to verify was a tremendous help. I used the opportunity to switch to EFI booting, which took me the = better part of the evening, but that's working now and booting the = correct kernel. It=E2=80=99s even booting into a 1280x720 resolution = with the help of the drm-devel-kmod. The next challenge is getting Xorg to run on this Navi-10 GPU; so far I = get stuck with "[KMS] drm report modesetting isn't supported=E2=80=9D. > Your hashes won't match mine due to non-reproducible build. >=20 > I'd make sure you don't have anything in /boot/modules or otherwise = load any > extra modules until sanity is restored (just to reduce random = variables). Ah yes, I wasn=E2=80=99t aware of /boot/modules. Last time I used = CURRENT, modules were still in the kernel directory. Hence, that was = also where I pointed kldload to to test my modules, which explains part = of the confusion (and there=E2=80=99s no modules.old=E2=80=A6). Alban Hertroys -- There is always an exception to always.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3027159B-1851-4005-A569-059A1CAACFCC>