On 15 September 2015 at 14:09, Toby Slight <tobysli...@gmail.com> wrote:
> The plot thickens... > > So I finally managed a successful UEFI install (I won't tell you how many > chickens I had to sacrifice...) with the latest snapshots (15/09/2015), and > following http://blog.jasper.la/openbsd-uefi-bootloader-howto/. > > This time I opted to keep it simple and not attempt my usual encrypted > install, so I'm not sure whether there was a fix in the latest snapshot or > the encryption of the previous attempt was to praise/blame respectively ... > I did intend to test an encrypted install with the latest sets, however the > randomly booting installer bug is still very much present (see chicken > sacrificing...), and after 30+ attempts to boot the usb again, my patience > gave out and I'm writing this email instead! > > Anyone have any ideas why loading the installer kernel is so flaky? It > doesn't happen when booting the installed kernels (both the previous, > broken, encrypted install, and the current, unencrypted, successful install > have no trouble getting past this stage). Anyway I have another picture, as > on one of my 50 odd attempts, it threw me a new error message that I hope > might be of use to someone... > > http://i.imgur.com/XOg0AM1.jpg > > Suddenly realised I could just attempt the encrypted install from the ramdisk kernel of the current install - doh! Unfortunately - it wasn't sucessful. I followed the exact same steps as yesterday: fdisk -ib 960 sd0 disklabel -E sd0 as one big RAID bioctl -c C -l /dev/sd0a softraid0 do stock install (wondered if I should have used the new fdisk -b flag to make the sd4 volume created by the bioctl command also and efi bootable disk?) After install formatted sd0i as msdos, mounted and copied over BOOTX64.EFI It did seem to get a bit further through the boot process this time before hitting ddb: http://i.imgur.com/4LjKoB6.jpg http://i.imgur.com/95dLnP7.jpg Strangely the bsd.sp kernel wouldn't boot at all, unlike with the 12th's snapshot, which got further than the mp kernel: http://i.imgur.com/z5RCzfw.jpg On a second attempt with the default mp kernel, it shit the bed earlier for some reason: http://i.imgur.com/yL33y8k.jpg http://i.imgur.com/VhjD6vz.jpg I also wondered how much these issues might have to do with a relatively recent BIOS update? I'm on Lenovo's latest - "G1ETA8WW (2.68 )" from 03/06/2015? Finally, is this kind of testing and information at all useful to the devs, or am I just creating unnecessary noise on the list, and should I just wait until it's a little further down the line? I've got a fair bit of free time in the next week or so, before the new term starts, so am happy to test. -- 0x2b || !0x2b
