On Thu, Feb 14, 2019 at 02:24:13AM -0600, Bruce Dubbs via lfs-dev wrote: > On 2/14/19 1:22 AM, Ken Moffat via lfs-dev wrote: > > > > I beg leave to doubt that this machine will ever be able to boot > > without secure boot. Certainly, the manufacturer (and others) seem > > to go out of their way to mess with the UEFI so that it is hard to > > boot linux, and I suspect the legacy bios boot path is poorly > > exercised because the machines all whip with W. Soem of the ACPI > > and related settings can best be described as "poor, but adequate > > for W". > > > > And although I've updated the "bios" by two versions, the updater > > comes in an (unsigned!) zip file which W said was "downloaded from > > the internet", and it uses a windows executable to do the update. > > > > Anyway, I'm not about to try removing W - the machine was bought > > for a purpose, I've already destroyed enough hardware in the past > > without breaking some more ;-) > > LOL. > > I must be lucky. I've never run into a situation that needs a firmware > update. My understanding, and I'm no expert here, is that a lot of updates > add CPU microcode updates. We can do those when we build the kernel. Maybe > W needs the firmware to do it. >
When ryzens first came out, there were memory issues (i.e. some brands worked, others didn't), fixed by "bios updates". My purchase last year was late enough to avoid that, but my impression is that the updates on desktop motherboards sometimes needed W. > I suppose that some updates fix hardware initialization problems and > possibly the kernel asks the firmware what hardware exists. > On the current laptop, a couple of errors in the ioapic, which lead to interrups not getting remapped (I think that means they all get routed to the boot CPU, but not sure) - that's the kernel bug in my last-week's .sig. On mine, it seems to perhaps work correctly on a few boots, so maybe a race somewhere. The bios update didn't fix that, but with Manjaro the boot parms seem to do the trick. > I will note that what I've read says the UEFI does not have a "bios" mode, > but sets up the boot via secure mode on/off and what is in the efi > partition. I just use the term 'firmware' as generic for the old bios > firmware and the newer uefi firmware. > True, but manufacturers still refer to bios updates. > It is probably too late, but did you see this: > > https://makeawebsitehub.com/best-laptops-linux/ > Yes, but "best" usually means expensive, and anyway I want to get away from intel (competition is good, and provides greater test-coverage such as amdgpu). ĸen -- The beauty of reading a page of de Selby is that it leads one inescapably to the conclusion that one is not, of all nincompoops, the greates. -- du Garbandier -- http://lists.linuxfromscratch.org/listinfo/lfs-dev FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
