I bought a new laptop. YAY! I tried a couple distros but ultimately got my beloved Manjaro KDE running when installing from the Architect ISO. It is great … until it freezes. Unfortunately it freezes a lot and I have to hold down power to shut it off.
Since I can leave an Ubuntu 17.10.1 (the only distro that didn’t cry for help on a live stick) running all day long regardless of the PCIe Bus Error, I think it makes sense to also look elsewhere. So, full sitrep:
The Setup
- Lenovo 720s-13ARR
- Ryzen 2500U with Vega 8
- 8GB RAM (not upgradable)
- Samsung 960 EVO 1TB (upgraded)
- Intel 8265 wifi / bluetooth (upgraded)
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
nvme0n1 259:0 0 931,5G 0 disk
├─nvme0n1p1 259:1 0 300M 0 part /boot/efi
├─nvme0n1p2 259:2 0 131,2G 0 part /
├─nvme0n1p3 259:3 0 50G 0 part
└─nvme0n1p4 259:4 0 750G 0 part
The 50GB and 750GB partitions are currently empty.
They were meant for a secondary OS (not sure if I’ll go that route though) and just bulk storage.
I am not using swap at all.
I am running Manjaro KDE 17.1.5 on kernel 4.15.4-2 at the time of this post.
It has been installed to an external drive using the architect ISO.
It was clonezilla’d over to the internal 960 EVO.
Bootup
From hitting enter in grub it goes through
Loading Linux 4.15.4-2-MANJARO x64 ...
Loading initial ramdisk ...
… blanks the screen (nothing unusual up to here) and then goes on with
[ *timecodewhatever*] [Firmware Bug]: AMD-Vi: IOAPIC[4] not in IVRS table
[ *timecodewhatever*] [Firmware Bug]: AMD-Vi: IOAPIC[5] not in IVRS table
[ *timecodewhatever*] [Firmware Bug]: AMD-Vi: No southbridge IOAPIC found
[ *timecodewhatever*] AMD-Vi: Disabling interrupt remapping
[ *timecodewhatever*] AMD-Vi: Unable to write to IOMMU perf counter.
starting version 237
/dev/nvme0n1p2: recovering journal
/dev/nvme0n1p2: clean, numbers/biggernumbers files, numbers/biggernumbers blocks
sp5100_tco: I/O adress 0x0cd6 already in use
At this point a hangup can happen and I have to force shutdown by holding down power.
If it doesn’t, it moves on with the PCIe Bus Error messages looping
[ *timecodewhatever*] pcieport 0000:00:01.2: PCIe Bus Error: serverity=Uncorrected (Non-Fatal), type=Transaction Layer, id=000a (Requester ID)
[ *timecodewhatever*] pcieport 0000:00:01.2: device [1022:15d3] error status/mask=00100000/04400000
[ *timecodewhatever*] pcieport 0000:00:01.2: [20] Unsupported Request (First)
[ *timecodewhatever*] pcieport 0000:00:01.2: TLP Header:34000000 02000010 00000000 09f409f4
… for about 2x the screen until I get a bunch of [OK] lines and get into the GUI.
I’m not sure if these hangups have gotten worse over time. It feels like it has gotten more and more but it might just be getting more and more annoying.
Running
As mentioned the system freezes a lot while in use and I have to hard reset it by holding down power to shut it off. After a fresh start I have left the system running the same Manjaro install for multiple hours and it was still fine when I came back. I have managed to crash it running a game in minutes. I don’t believe it is temp related since I am playing (well, hardly) Hyper Light Drifter, which should not be taxing on either CPU or GPU.
The USB ports are sometimes just … uhm, turned off. And it isn’t just with drives not being mounted, even input devices aren’t recognized and won’t even get power.
Done so far
In grub I have blacklisted shpchp without any observable (for a noob) effect and sp5100_tco which at least got rid of that one error seemingly without any negative outcome otherwise.
Smart as well as fsck had nothing bad to say about my drive or any partition on it.
I have set pci=noaer to get rid of the wall of PCIe Bus Error messages.
What is next
I will try to get temp readings from the system while pushing it.
I really don’t want to nuke this install but I will try that if I can’t figure out the freezing issue.
And that is it. Sooo …
