11-20-2018 01:24 PM
Against the sporadic freezes the kernel parameter "idle=nomwait" seems to help (see https://bugs.freedesktop.org/show_bug.cgi?id=106658).
11-21-2018 07:06 AM
It seems this may be a more widespread problem than just our Thinkpads. See this thread, it's happening with other Raven Ridge laptops as well. Lots of discussion here:
Looks like this might be a multi-part problem that requires fixes on both the kernel and bios
11-25-2018 11:12 AM
Indeed. However, I would order them as AMD > Linux > Lenovo because amd should definitely be first to provide integrators with proper specs and tools to implement their CPU features.
Linux should be able to fall back to normal enumeration just like Windows is doing at the moment and finally Lenovo should at least provide an option to disable the iommu or just disable it altogether until amd gets it working.
12-01-2018 01:53 PM - edited 12-01-2018 01:53 PM
There are notebooks with the same Ryzen CPU (Ryzen 3/5/7) and the same chipset that are working completly perfectly fine with Linux. So the problem here lies clearly at Lenovo, which has to fix the BIOS bug. And even if the bug is not in the BIOS, they have to report it to AMD and the Linux Kernel Team. Since there is still not any mention of the E585 or E485 in the kernel bug report, it seems Lenovo employees still did not have taken any action in the case. No Lenovo employee has made inquiries to the Linux Kernel Team, because the not-working Lenovo notebooks are not even mentioned there. So I only make Lenovo responsible for there faineance and the guilt lies by them.
12-01-2018 02:27 PM
Now I wrote a comment in the Linux bug report, so that the Linux Kernel Team is at least aware of that critical bug.
Only if we make pressure on all sides something will change.
12-01-2018 02:40 PM
I'm honestly doubting we're going to get any help with this. I contacted Lenovo support on Twitter about bluetooth not working in Linux, and they promptly directed me to a page stating that linux is not supported on this laptop. The same issue was already fixed in a bios update on the A485, but they didn't apply the fix to the E-series. I asked them to please forward my email on to the bios developers anyway, because they might be able to essentially just copy/paste the fix.. he said he would, but who knows.
The short url links to here:
Fixed in bios 1.06 on A485:
<1.06> - [Important] Fixed an issue that the BT status can't be got in Linux.
It's frustrating. The reason I chose the E-series over the A-series is that having a number pad is important to me, if I wanted a Ryzen thinkpad, it was my only option.
12-01-2018 02:53 PM - edited 12-01-2018 02:53 PM
This seems for me just a lame excuse. I mean we still speak about a ThinkPad. The ThinkPad is Lenovos high-quality product class with a high reputation in businesses. Even if it's just a E585 or E485, its still part of Lenovos highest-quality notebook product class. If this goes to the press that Linux is still not working after a half a year and Lenovo does not care any about it, that would not make a good impression to Lenovo's standing in the public.
I now also wrote a support ticket to Lenovo, maybe it helps a little. Everyone who has this problem, should also write a support ticket to Lenovo.
12-01-2018 05:57 PM
Honestly very disappointed with Lenovo after all the good things I've heard about them. This behaviour from them is quite disgusting in response to valid issues with their products. Similar thing happened to us a few months ago with the laggy keyboard issue. Took them AGES to finally release a patch after substantial pressure from the community and even worse was that they didn't even address the issue in the BIOS changelog... Which is why I went Googling before applying the latest BIOS update. For those having suspend issues, there's another forum post here outlining a fix. Still haven't updated BIOS because of it. Paid for a professional laptop to do work on, not to beta test for Lenovo.
12-02-2018 10:49 AM
It is now two months since this thread started in the forum with not a single official statement or fix from Lenovo. Also the suspend-bug was not fixed for almost 5 months. We have just userspace-workarounds for both of them. Does Lenovo still plan to react on their broken BIOS-IVRS-Table and the broken suspend which seems to be introduced with their recent bios? "No support for Linux" is in my eyes not equal with "we build bugs inside our products and let consumers fix them with individual workarounds". Wouldn't it be far easier to publish a fixed UEFI and have those bugs fixed for all consumers at once? Also in terms of publicity?
There is no need to re-invent the wheel. I assume a high probability that some (slightly adapted) patch for A485 can be applied here.