This has now also pulled in #215. So this will also not have that vulnerability,

Feedback: I could build it on my build machine (ubuntu 22.10 server as VM). Had to clean up a little after the first run that failed on dkms, but was just build debris still lying around, git clean -df took care of it. It is running happily on my vintage T500, will test it on a Ryzen box with NVidia card and 2 screens, too.

@jglathe awesome! Thank you!

Install of 6.1.6 on vega just done and works nicely. Multiple displays on NVidia, sound, everything there.

System:
  Host: vega Kernel: 6.1.6-76060106-generic x86_64 bits: 64
    Desktop: GNOME 42.3.1 Distro: Pop!_OS 22.04 LTS
Machine:
  Type: Desktop Mobo: ASRock model: X570M Pro4 serial: M80-D6017501251
    UEFI: American Megatrends v: P3.90 date: 10/19/2022
CPU:
  Info: 16-core model: AMD Ryzen 9 3950X bits: 64 type: MT MCP cache:
    L2: 8 MiB
  Speed (MHz): avg: 2362 min/max: 2200/4761 cores: 1: 2200 2: 2200 3: 2200
    4: 2200 5: 2200 6: 3500 7: 3500 8: 2200 9: 2200 10: 2200 11: 2200 12: 2200
    13: 3500 14: 2200 15: 2200 16: 2200 17: 2200 18: 2200 19: 2200 20: 2200
    21: 2200 22: 2200 23: 2200 24: 2200 25: 2200 26: 2200 27: 2200 28: 2200
    29: 3500 30: 2200 31: 2200 32: 2200
Graphics:
  Device-1: NVIDIA TU116 [GeForce GTX 1660 SUPER] driver: nvidia v: 525.78.01
  Display: server: X.Org v: 1.21.1.3 driver: X: loaded: nvidia
    unloaded: fbdev,modesetting,nouveau,vesa gpu: nvidia resolution:
    1: 2560x1440~60Hz 2: 2560x1440~60Hz
  OpenGL: renderer: NVIDIA GeForce GTX 1660 SUPER/PCIe/SSE2
    v: 4.6.0 NVIDIA 525.78.01
Audio:
  Device-1: NVIDIA TU116 High Definition Audio driver: snd_hda_intel
  Device-2: AMD Starship/Matisse HD Audio driver: snd_hda_intel
  Sound Server-1: ALSA v: k6.1.6-76060106-generic running: yes
  Sound Server-2: PipeWire v: 0.3.63 running: yes
Network:
  Device-1: Intel I211 Gigabit Network driver: igb
  IF: enp4s0 state: up speed: 1000 Mbps duplex: full mac: a8:a1:59:38:b6:d7
Drives:
  Local Storage: total: 1.4 TiB used: 52.03 GiB (3.6%)
  ID-1: /dev/nvme0n1 vendor: Silicon Power model: SPCC M.2 PCIe SSD
    size: 476.94 GiB
  ID-2: /dev/nvme1n1 vendor: Silicon Power model: SPCC M.2 PCIe SSD
    size: 953.87 GiB
Partition:
  ID-1: / size: 460.03 GiB used: 48.76 GiB (10.6%) fs: ext4
    dev: /dev/nvme0n1p3
  ID-2: /boot/efi size: 497 MiB used: 384.2 MiB (77.3%) fs: vfat
    dev: /dev/nvme0n1p1
Swap:
  ID-1: swap-1 type: partition size: 4 GiB used: 0 KiB (0.0%) dev: /dev/dm-0
  ID-2: swap-2 type: zram size: 16 GiB used: 0 KiB (0.0%) dev: /dev/zram0
Sensors:
  System Temperatures: cpu: N/A mobo: N/A gpu: nvidia temp: 33 C
  Fan Speeds (RPM): N/A gpu: nvidia fan: 44%
Info:
  Processes: 483 Uptime: 5m Memory: 62.72 GiB used: 5.98 GiB (9.5%)
  Shell: Sudo inxi: 3.3.13

sudo inxi -Gay
Graphics:
  Device-1: NVIDIA TU116 [GeForce GTX 1660 SUPER] vendor: Micro-Star MSI
    driver: nvidia v: 525.78.01 alternate: nvidiafb,nouveau,nvidia_drm pcie:
    gen: 1 speed: 2.5 GT/s lanes: 16 link-max: gen: 3 speed: 8 GT/s ports:
    active: none off: DP-1,HDMI-A-1 empty: DP-2,DP-3 bus-ID: 09:00.0
    chip-ID: 10de:21c4 class-ID: 0300
  Display: server: X.Org v: 1.21.1.3 compositor: gnome-shell v: 42.3.1
    driver: X: loaded: nvidia unloaded: fbdev,modesetting,nouveau,vesa
    gpu: nvidia display-ID: :1 screens: 1
  Screen-1: 0 s-res: 5120x1440 s-dpi: 108 s-size: 1204x342mm (47.4x13.5")
    s-diag: 1252mm (49.3")
  Monitor-1: DP-0 pos: right res: 2560x1440 hz: 60 dpi: 109
    size: 597x336mm (23.5x13.2") diag: 685mm (27")
  Monitor-2: HDMI-0 pos: primary,left res: 2560x1440 hz: 60 dpi: 109
    size: 597x336mm (23.5x13.2") diag: 685mm (27")
  OpenGL: renderer: NVIDIA GeForce GTX 1660 SUPER/PCIe/SSE2
    v: 4.6.0 NVIDIA 525.78.01 direct render: Yes

To get the whole package of ./rebuild.sh installed without errors you need to upgrade pciutils and libpci to 3.9.0:

libpci3_3.9.0-3_amd64.deb libpci-dev_3.9.0-3_amd64.deb pciutils_3.9.0-3_amd64.deb

Thanks! Cherry-picked that fix, so that shouldn't be an issue anymore

Unfortunately that didn't fix suspend on pang10

5ee6a55 should fix the pang10/11 suspend issue

Yep, looks like that fixed pang11! 🎉

Looks like bcmwl-kernel-source is throwing some DKMS errors when I try to install it. Here is the log:
make.log

Guessing https://github.com/pop-os/bcmwl/ just needs an update as well?

https://github.com/pop-os/bcmwl/tree/bcmwl-rebase fixes that for me

Yep, that fixes it for me as well

I have installed the staging kernel on my Maingear Vector Pro 2 which is running AMD 6900HX and the laptop keyboard is still not working.

@13r0ck can you elaborate on the Mira/Nvidia/Plymouth issue?

On the first few attempts, I struggled to see a Mira/Nvidia system progress beyond entering the decryption password on boot with this kernel, which is a worry.

I've attached a log (ripped from journalctl) from the last failed attempt to boot with this Kernel PR present.
Failure.txt

I'm in the process of seeing whether I can find some way to induce a boot failure now I've had the system boot successfully.

Previous log was after a failure to boot Old_Kern after adding this PR. This is after a failure to boot 6.1.6:
Failure2.txt

Two back to back boot attempts, one good and one bad. They diverge at line 762 (I believe they are literally identical for the first 761 line, time excepted). Not sure the implications there of. This appears to be timing related; looks like this is the point at which different aspects are processed in parallel.

good.txt
bad.txt

nvidia driver balking

At least on DevOne, things are looking good.

DevOneLinux6-1-6.txt

But, Mira is still misbehaving. The fact that it doesn't fail to move from Plymouth decrypt to login every time is odd to me, and guidance on desired debugging would be helpful.

Alright, I've got a better idea of what's going on, and I've eliminated Mesa 22.3.4 from the equation. With this kernel installed, MST breaks after suspend and resume on Dev One. So from a fresh boot, everything is working fine.

After suspending and resuming with this 6.1.6 on a fresh and updated install, MST stops working when I plug the USB-C cable in. Display settings shows both monitors I have connected, but the upstream one says "No signal" and goes back to sleep, while the downstream one never even wakes up.

When booting back into oldkern (6.0.12), both monitors work fine after suspending and resuming, then connecting the USB-C cable. I've replicated this on my own Dev One as well.

Could you enable MGLRU by default?

The kernel config option is CONFIG_LRU_GEN_ENABLED

https://www.phoronix.com/news/MGLRU-Reaches-mm-stable

@lolcosmo good find!. Currently too much of our QA has been done already on this kernel release. But I will enable it on our 6.2 series kernel and see if it passes QA.

Until the next release you could build your own kernel with that enabled

git clone https://github.com/pop-os/linux
cd linux
vim debian.master/config/config.common.ubuntu # <- this is where to enable MGLRU
sudo apt build-dep linux-system76 -y
dpkg-buildpackage -us -uc -b
cd ..
sudo dpkg -i *.deb
reboot

@leviport I cherry-picked eeda912 from 6.1.9, that should fix the MST bug on DevOne. It did for my machine

Sorry, but I don't think this fixes it. If anything, it seems worse now. Instead of just not activating either of the displays, now it sometimes makes the built-in display go black with only the mouse cursor shown, and act locked up (can't move the mouse cursor) until I unplug the USB-C cable. I tried to make these things happen while booted into oldkern (6.0.12) on the same install, but it just kept working as expected between reboots and suspend/resume cycles.

I'll do a reinstall and try again soon, just in case it was some kind of fluke.

Closed in favor of #222