On Tue, Jul 09, 2019 at 09:01:30PM +0100, Stuart Henderson wrote:
> On 2019/07/09 17:32, Chris Humphries wrote:
> > On Tue, Jul 09, 2019 at 08:57:43AM +0200, Raimo Niskanen wrote:
> > > On Mon, Jul 08, 2019 at 04:40:19PM +0100, Stuart Henderson wrote:
> > > > On 2019/07/08 09:50, Raimo Niskanen wrote:
> > > > > On Thu, Jul 04, 2019 at 09:13:22AM +0200, Raimo Niskanen wrote:
> > > > > > I just freed a Lenovo U350 IdeaPad from Windows 10 Home and 
> > > > > > installed
> > > > > > OpenBSD 6.5 on it (the installer gets better and better - now I 
> > > > > > installed
> > > > > > the sets via https and cdn.openbsd.org; really slick!), and when 
> > > > > > starting X
> > > > > > it panicked.
> > > > > > 
> > > > > > During the first installation I selected to start xenodm and at 
> > > > > > firstboot
> > > > > > it panicked right after firmware installation during kernel 
> > > > > > relinking.
> > > > > > Both firmware files and kernel appeared to be damaged after reboot 
> > > > > > and fsck.
> > > > > > 
> > > > > > So this is after a second installation where I selected to not start
> > > > > > xenodm, which allowed me to update firmware, syspatch reboot again
> > > > > > and wait for quiescent filesystem before starting xenodm, and then;
> > > > > > it panicked.
> > > > > > 
> > > > > > I hope it is simply not that machdep.allowaperture=0; if that was 
> > > > > > needed
> > > > > > Xorg should just refuse to start, right?
> > > > > 
> > > > > It was not.  With machdep.allowaperture=0 I get the same panic.
> > > > > 
> > > > > Suggestion, anyone?
> > > > 
> > > > Quickly summarizing the panic message etc from the screenshot, it's an
> > > > "index out of bounds" panic in aml_parse:4042, called from 
> > > > acpivout_set_param.
> > > > 
> > > > As ever, first try a -current snapshot. If it still fails, run
> > > > sendbug -P > somefile as root (to generate a report including acpi 
> > > > tables)
> > > > and copy to a system you can conveniently email from, include 
> > > > information
> > > > from this report (so that the email is self-contained) and preferably
> > > > re-type the more important parts from panic.jpg (so it's in plaintext 
> > > > and
> > > > developers can more easily figure out if it's something they might know
> > > > about).
> > > 
> > > Ok.  Thank you!  I will do that.
> > > 
> > > > 
> > > > Maybe also worth trying the newer BIOS versions that are available
> > > > (1CCN25WW/1CCN26WW).
> > > 
> > > Aah...  I saw that, but the only BIOS updates I can find are in the form 
> > > of
> > > Windows 7 .exe files, and I wiped the Windows installation.  So I see two
> > > possible ways - either to try to reinstall Windows from an old restore DVD
> > > that I seriously doubt to be readable, or to find a trustworthy download 
> > > of
> > > a Windows 7 Live CD (or maybe 8, but the places I have seen them feel
> > > a bit shady).
> > > 
> > > Or are there other black magic ways to find and update Lenovo BIOS:es for
> > > this machine generation?
> > 
> > 
> > I think on that same screen where the Windows BIOS firmware updater
> > exe files are is also a link to an ISO, which you can burn to a cd or
> > usb stick. It isn't obvious, search the web page for "iso".
> > 
> > I did the same thing yesterday with my Thinkpad T420, even though the
> > link to the ISO was broken on the page, a google search found the
> > direct link.
> 
> For most thinkpads you can fetch an iso and use "geteltorito" (in
> ports) to extract an image that can be dd'd to a USB stick if you don't
> have a CD drive. But not for the ideapads.
> 
> You can download a clean Windows 10 iso from Microsoft, and it will
> install without a product key, just disables a few things if it's not
> activated that are unlikely to matter for this ...
> 
> Or you might be able to do something with a DOS version of the Phoenix
> flash programmer phlash16.exe and the files you get from unzipping
> 1CCN26WW.exe.
> 
> Given the hassle levels involved I would have a good attempt at seeing
> if it will work without the update though :)

I did the BIOS update, but no improvement.  The panic, at least, is exactly
the same, expect for small differences in the arg0 addresses.

For the record, this is how I did it:

I have a Windows 10 computer and used it to follow:
  https://www.intowindows.com/4-tools-to-create-windows-to-go-usb-of-windows-10/
method 1: Rufus.

Rufus is now at version 3.5 and worked nicely to create a Windows To Go
installation of Windows 10 Pro En 64-bit on a USB stick from a Windows 10
October 2018 ISO file that I downloaded from:
  https://www.microsoft.com/en-us/software-download/windows10ISO

The BIOS was downloaded from: 
  https://download.lenovo.com/eol/index.html
for IdeaPad U350 Notebook:
  
https://download.lenovo.com/UserFiles/Driver/en/Downloads%20and%20Drivers/U350/BIOS/1CCN26WW64.exe

I selected an SD card whith card reader as USB disk since it was the only
32 GB I had within reach, so creation of the Windows To Go installation
with Rufus took maybe more than an hour.  Then I could write the BIOS
installer right into the Windows To Go installation.  Booting the thing
on the laptop took a solid half hour (what on earth is Windows actually
doing while it says things like "Preparing Stuff"???), before it started
to ask about region keyboard and things, and a few minutes later I could
launch 1CCN26WW64.exe that completed without a glitch and rebooted.
When the reboot went into starting BIOS again, I pulled the stick,
and voila!  The machine had a new BIOS.

Alas, still no X!

Thanks for all valuable tips, everyone!
-- 

/ Raimo Niskanen, Erlang/OTP, Ericsson AB

Reply via email to