https://github.com/sxx1314/pve-kernel

Thanks for the incredibly useful project — it's greatly appreciated!

Fetched in submodule path 'submodules/zfsonlinux/upstream', but it did not contain 6b18d7df3772cffa6469b00866d7eae585b912ae. Direct fetching of that commit failed.

After first trying a week ago, I'm continuing to see this error when fetching submodules. What's the recommended workaround?

ZFS has pushed another new release which provides fixes for some critical bugs, including one that can cause application crashes and system lockup (openzfs/zfs#13234), that crept into 2.1.3.

After update PVE kernel to 5.19 by "apt install pve-kernel-5.19-edge", I got errors in "Syslog" as snapshot below.

It happened on 5.18 as well but 5.17 work good.

Information of my NVMe as below.

root@pve:~# smartctl -a /dev/nvme0n1
smartctl 7.2 2020-12-30 r5155 [x86_64-linux-5.15.39-3-pve] (local build)
Copyright (C) 2002-20, Bruce Allen, Christian Franke, www.smartmontools.org

=== START OF INFORMATION SECTION ===
Model Number: Lexar 512GB SSD
Serial Number: MFM8182000030
Firmware Version: V1.19.B3
PCI Vendor/Subsystem ID: 0x1d97
IEEE OUI Identifier: 0xcaf25b
Total NVM Capacity: 512,110,190,592 [512 GB]
Unallocated NVM Capacity: 0
Controller ID: 0
NVMe Version: 1.4
Number of Namespaces: 1
Namespace 1 Size/Capacity: 512,110,190,592 [512 GB]
Namespace 1 Formatted LBA Size: 512
Namespace 1 IEEE EUI-64: caf25b 020000a827
Local Time is: Fri Aug 5 13:43:13 2022 CST
Firmware Updates (0x02): 1 Slot
Optional Admin Commands (0x0017): Security Format Frmw_DL Self_Test
Optional NVM Commands (0x004c): DS_Mngmt Wr_Zero Timestmp
Log Page Attributes (0x0e): Cmd_Eff_Lg Ext_Get_Lg Telmtry_Lg
Maximum Data Transfer Size: 32 Pages
Warning Comp. Temp. Threshold: 81 Celsius
Critical Comp. Temp. Threshold: 85 Celsius

Supported Power States
St Op Max Active Idle RL RT WL WT Ent_Lat Ex_Lat
0 + 5.00W - - 0 0 0 0 5 700

Supported LBA Sizes (NSID 0x1)
Id Fmt Data Metadt Rel_Perf
0 + 512 0 3

=== START OF SMART DATA SECTION ===
SMART overall-health self-assessment test result: PASSED

SMART/Health Information (NVMe Log 0x02)
Critical Warning: 0x00
Temperature: 43 Celsius
Available Spare: 100%
Available Spare Threshold: 10%
Percentage Used: 0%
Data Units Read: 47,430 [24.2 GB]
Data Units Written: 226,106 [115 GB]
Host Read Commands: 435,435
Host Write Commands: 2,705,916
Controller Busy Time: 6
Power Cycles: 19
Power On Hours: 72
Unsafe Shutdowns: 6
Media and Data Integrity Errors: 0
Error Information Log Entries: 0
Warning Comp. Temperature Time: 0
Critical Comp. Temperature Time: 0

Read 16 entries from Error Information Log failed: NVMe Status 0x13

I can't seem to figure out how to select the kernel as the boot kernal after install:

root@pve1000:~# ls
linux-tools-5.10_5.10.20-1_amd64.deb  pve-edge-kernel-5.10.20-1_5.10.20-1+zen21_amd64.deb
root@pve1000:~# dpkg -i *
Selecting previously unselected package linux-tools-5.10.
(Reading database ... 58239 files and directories currently installed.)
Preparing to unpack linux-tools-5.10_5.10.20-1_amd64.deb ...
Unpacking linux-tools-5.10 (5.10.20-1) ...
Preparing to unpack pve-edge-kernel-5.10.20-1_5.10.20-1+zen21_amd64.deb ...
Unpacking pve-edge-kernel-5.10.20-1 (5.10.20-1+zen21) over (5.10.20-1+zen21) ...
Setting up linux-tools-5.10 (5.10.20-1) ...
Setting up pve-edge-kernel-5.10.20-1 (5.10.20-1+zen21) ...
Examining /etc/kernel/postinst.d.
run-parts: executing /etc/kernel/postinst.d/apt-auto-removal 5.10.20-1 /boot/vmlinuz-5.10.20-1
run-parts: executing /etc/kernel/postinst.d/initramfs-tools 5.10.20-1 /boot/vmlinuz-5.10.20-1
update-initramfs: Generating /boot/initrd.img-5.10.20-1
run-parts: executing /etc/kernel/postinst.d/pve-auto-removal 5.10.20-1 /boot/vmlinuz-5.10.20-1
run-parts: executing /etc/kernel/postinst.d/zz-pve-efiboot 5.10.20-1 /boot/vmlinuz-5.10.20-1
Re-executing '/etc/kernel/postinst.d/zz-pve-efiboot' in new private mount namespace..
Copying and configuring kernels on /dev/disk/by-uuid/1967-F3F9
	Copying kernel and creating boot-entry for 5.4.101-1-pve
	Copying kernel and creating boot-entry for 5.4.73-1-pve
run-parts: executing /etc/kernel/postinst.d/zz-update-grub 5.10.20-1 /boot/vmlinuz-5.10.20-1
Generating grub configuration file ...
Found linux image: /boot/vmlinuz-5.10.20-1
Found initrd image: /boot/initrd.img-5.10.20-1
Found linux image: /boot/vmlinuz-5.4.101-1-pve
Found initrd image: /boot/initrd.img-5.4.101-1-pve
Found linux image: /boot/vmlinuz-5.4.73-1-pve
Found initrd image: /boot/initrd.img-5.4.73-1-pve
Found memtest86+ image: /ROOT/pve-1@/boot/memtest86+.bin
Found memtest86+ multiboot image: /ROOT/pve-1@/boot/memtest86+_multiboot.bin
Adding boot menu entry for EFI firmware configuration
done
Processing triggers for man-db (2.8.5-2) ...

Hi,

I'm trying to compile aqc111u driver and it looks like Kconfig is missing from v5.11.16-1

:~/aqc111u/Linux# make
make -C /lib/modules/5.11.16-1-edge/build SUBDIRS=/root/aqc111u/Linux modules
make[1]: Entering directory '/usr/src/linux-headers-5.11.16-1-edge'
SYNC include/config/auto.conf.cmd
can't find file Kconfig
make[3]: *** [scripts/kconfig/Makefile:71: syncconfig] Error 1
make[2]: *** [Makefile:602: syncconfig] Error 2
make[1]: *** [Makefile:710: include/config/auto.conf.cmd] Error 2
make[1]: Leaving directory '/usr/src/linux-headers-5.11.16-1-edge'
make: *** [Makefile:26: default] Error 2

5.15 has a regression in its config where UBSAN gets enabled again, leading to boot issues

Hi,

I was interested to know what specific fixes you have in for your zen2 kernel build? I am whether I see an improvement on my node using the latest 5.8.7 zen2 build vs the standard build.

Thanks.

Hi! Thanks for providing these edge kernels!

I've recently upgraded my PVE from stock 5.15 kernel to latest 5.19.4 and found my kern.log filled with i2c i2c-6: sendbytes: NAK bailout. I've done some searching and found out that this device is part of my Intel IGD and this error won't be thrown when there is an external monitor connected via HDMI, and it will again start showing up the second i remove the cable. However, this never happened before upgrading. Is there anything that I should tweak to tackle this error?

The following is the entry for i2c-6 by i2cdetect:
i2c-6 i2c i915 gmbus dpb I2C adapter

I'm using an Intel N5105 nuc with Intel JasperLake [UHD Graphics].

In Proxmox since the last kernel updates passthrough of Intel integrated graphics devices fail.
I googled and found that implementing some patches in Kernel 5.8+ would fix the issues.
Can you review and implement them please?
https://github.com/jam3st/vfio-hd4600

I can't say for sure, but I have crashes with the kernel 5.16 and 5.17 about every 1 - 3 days. I always keep the versions up to date - I'm currently on 5.17.0-edge. Sometimes the whole thing goes so far that the CPU fan turns very high, but nothing reacts anymore (no ping, like dead). The only thing that helps is to de-energise the computer. I am currently using a Minisforum Tiny (HX90) as my node, which has a built-in APU AMD Ryzen 9 5900HX.
Unfortunately I don't get any kernel panic logs that I could attach here.

Can I also use pve-edge-kernel for the Proxmox VE 7.0?

Thanks for your answer in advance.

Hi,

I came across your repo and I was wondering how hard it would be for me to compile kernel 5.17. I know it is really bleeding edge, but I would like to use it because I have a Ryzen 5600G Proxmox server running, and the new kernel contains optimalizations for my CPU.

LXC containers doesn't work with kernel 5.16.x and 5.15.19-2.

As mentionned, apparmor parser file was rolling back to old config.
When start any container with "pct start XXX -debug=1", i obtain this error message.

`
problem with monitor socket, but continuing anyway: got timeout

main: 256 Container is already running
in:256 - Container is already running
`

I have tested with apparmor enabled or disabled.

I use GPU passthrougth on one container and for this my /etc/subgid was modified. Is this the cause of my problem ?

We should add a job to our Github Actions workflow for automatically testing the built kernels using for example QEMU.

Hi @fabianishere

I'm trying to upgrade from 5.17 to 5.18, but on installing pve-kernel-5.18-edge, the following error is thrown:

dkms: running auto installation service for kernel 5.18.10-edge:Error! Your kernel headers for kernel 5.18.10-edge cannot be found. Please install the linux-headers-5.18.10-edge package

I can't seem to find linux-headers-5.18.10-edge either :)

Please help; thanks again!

Hello,

Thanks for updating the readme, after finding your repo I tested out 5.7.2 among others, but got stuck with the LXC Apparmor issue.

I followed your instructions to update /etc/apparmor.d/parser.conf, but it did not work successfully. In the end I had to update /etc/apparmor/parser.conf on my Proxmox host to get the apparmor service up an running again. Leaving the originally created /etc/apparmor.d/parser.conf also prevented the apparmor service from starting, so I needed to remove it.

My first github issue, so thanks!

Hi, I'm having trouble building DKMS modules on 5.10+ with this.

On 5.9.16 and the PVE 5.4 kernel they build totally fine.

Looking on a web search, people had this issue with 5.10rc4 but it was fixed in 5.10rc7 so I'm not sure it's related. https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=975571

On 5.10.16 and 5.11.0, both modules I tested have the following error:

make[3]: *** No rule to make target 'scripts/module.lds', needed by '/var/lib/dkms/zenpower/0.1.12/build/zenpower.ko'. Stop.

Full DKMS log for another module:

root@main-server ~/linux> dkms install vendor-reset/0.1.1 -k 5.10.16-1/x86_64 --force --verbose

Kernel preparation unnecessary for this kernel.  Skipping...

Building module:
make KDIR=/lib/modules/5.10.16-1/build clean
make -C /lib/modules/5.10.16-1/build M=/var/lib/dkms/vendor-reset/0.1.1/build clean
make[1]: Entering directory '/usr/src/linux-headers-5.10.16-1'
make[1]: Leaving directory '/usr/src/linux-headers-5.10.16-1'

{ make -j32 KERNELRELEASE=5.10.16-1 KDIR=/lib/modules/5.10.16-1/build; } >> /var/lib/dkms/vendor-reset/0.1.1/build/make.log 2>&1
(bad exit status: 2)
Error! Bad return status for module build on kernel: 5.10.16-1 (x86_64)
Consult /var/lib/dkms/vendor-reset/0.1.1/build/make.log for more information.

root@main-server ~/linux> cat /var/lib/dkms/vendor-reset/0.1.1/build/make.log
DKMS make.log for vendor-reset-0.1.1 for kernel 5.10.16-1 (x86_64)
Tue 16 Feb 2021 03:06:04 PM CST
make -C /lib/modules/5.10.16-1/build M=/var/lib/dkms/vendor-reset/0.1.1/build modules
make[1]: Entering directory '/usr/src/linux-headers-5.10.16-1'
  CC [M]  /var/lib/dkms/vendor-reset/0.1.1/build/src/module.o
  CC [M]  /var/lib/dkms/vendor-reset/0.1.1/build/src/vendor-reset-dev.o
  CC [M]  /var/lib/dkms/vendor-reset/0.1.1/build/src/ioctl.o
  CC [M]  /var/lib/dkms/vendor-reset/0.1.1/build/src/ftrace.o
  CC [M]  /var/lib/dkms/vendor-reset/0.1.1/build/src/hook.o
  CC [M]  /var/lib/dkms/vendor-reset/0.1.1/build/src/amd/common.o
  CC [M]  /var/lib/dkms/vendor-reset/0.1.1/build/src/amd/compat.o
  CC [M]  /var/lib/dkms/vendor-reset/0.1.1/build/src/amd/firmware.o
  CC [M]  /var/lib/dkms/vendor-reset/0.1.1/build/src/amd/navi10.o
  CC [M]  /var/lib/dkms/vendor-reset/0.1.1/build/src/amd/polaris10.o
  CC [M]  /var/lib/dkms/vendor-reset/0.1.1/build/src/amd/vega10.o
  CC [M]  /var/lib/dkms/vendor-reset/0.1.1/build/src/amd/vega20.o
  CC [M]  /var/lib/dkms/vendor-reset/0.1.1/build/src/amd/amdgpu/amdgpu_device.o
  CC [M]  /var/lib/dkms/vendor-reset/0.1.1/build/src/amd/amdgpu/amdgpu_discovery.o
  CC [M]  /var/lib/dkms/vendor-reset/0.1.1/build/src/amd/amdgpu/amdgpu_bios.o
  CC [M]  /var/lib/dkms/vendor-reset/0.1.1/build/src/amd/amdgpu/amdgpu_atomfirmware.o
  CC [M]  /var/lib/dkms/vendor-reset/0.1.1/build/src/amd/amdgpu/atom.o
  CC [M]  /var/lib/dkms/vendor-reset/0.1.1/build/src/amd/amdgpu/common_baco.o
  CC [M]  /var/lib/dkms/vendor-reset/0.1.1/build/src/amd/amdgpu/navi10_reg_init.o
  CC [M]  /var/lib/dkms/vendor-reset/0.1.1/build/src/amd/amdgpu/navi12_reg_init.o
  CC [M]  /var/lib/dkms/vendor-reset/0.1.1/build/src/amd/amdgpu/navi14_reg_init.o
  CC [M]  /var/lib/dkms/vendor-reset/0.1.1/build/src/amd/amdgpu/polaris_baco.o
  CC [M]  /var/lib/dkms/vendor-reset/0.1.1/build/src/amd/amdgpu/smu7_baco.o
  CC [M]  /var/lib/dkms/vendor-reset/0.1.1/build/src/amd/amdgpu/vega10_reg_init.o
  CC [M]  /var/lib/dkms/vendor-reset/0.1.1/build/src/amd/amdgpu/vega20_reg_init.o
  LD [M]  /var/lib/dkms/vendor-reset/0.1.1/build/vendor-reset.o
  MODPOST /var/lib/dkms/vendor-reset/0.1.1/build/Module.symvers
make[3]: *** No rule to make target 'scripts/module.lds', needed by '/var/lib/dkms/vendor-reset/0.1.1/build/vendor-reset.ko'.  Stop.
make[3]: *** Waiting for unfinished jobs....
  CC [M]  /var/lib/dkms/vendor-reset/0.1.1/build/vendor-reset.mod.o
make[2]: *** [scripts/Makefile.modpost:117: __modpost] Error 2
make[1]: *** [Makefile:1711: modules] Error 2
make[1]: Leaving directory '/usr/src/linux-headers-5.10.16-1'
make: *** [Makefile:8: build] Error 2

Thanks in advance.

Hello,
some network Cards are Problems with the current Kernel. is it possible to fix it with?
Look at this Treads also pls
Workaround with Kernel Parameters.

GRUB_CMDLINE_LINUX_DEFAULT="quiet intel_idle.max_cstate=1"

https://forum.proxmox.com/threads/langsame-verbindung-ins-wan-im-lan-alles-gut.100500/
https://forum.proxmox.com/threads/slow-internet-speed-on-pve-but-not-on-local-network.99514/
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1927925

Hi Fabian, I have run into this issue Missing br_netfilter module | Proxmox Support Forum when I tried to run K3S within LXC container.

It seems the fix of this issue is too slow to wait. I want to try you custom kernel to continue LXC K3S plan. Do you think it's gonna work?

Trying these kernels with a Coffee Lake ER Xeon processor (Xeon E-2276G) and getting a bunch of random kernel panics related to gvt. Should I be using the generic kernel or should I try the Cascade Lake kernel?

Hello,

After installing & using these latest kernels, I appear unable to update the ixgbe module (updating from default 5.1.0-k to 5.8.1):

root@host:/tmp/ixgbe-5.8.1/src# make install common.mk:82: *** Kernel header files not in any of the expected locations. common.mk:83: *** Install the appropriate kernel development package, e.g. common.mk:84: *** kernel-devel, for building kernel modules and try again. Stop.

Normally this seems to be a case of the headers not being installed correctly, but in the case of using these edge kernels, is there something missing or is there an additional package needed? I have installed:

linux-tools-5.7_5.7.8-1_amd64.deb
pve-edge-headers-5.7.8-1_5.7.8-1_amd64.deb
pve-edge-kernel-5.7.8-1-zen2_5.7.8-1_amd64.deb

root@host:/usr/src# ls linux-headers-5.3.13-3-pve linux-headers-5.3.18-3-pve linux-headers-5.4.41-1-pve linux-headers-5.4.44-2-pve linux-headers-5.6.8-1 linux-headers-5.7.8-1 linux-headers-5.3.18-2-pve linux-headers-5.4.34-1-pve linux-headers-5.4.44-1-pve linux-headers-5.5.19-1 linux-headers-5.7.2-1

Thanks for any assistance!

The change made in 5.16 branch to fix zfs submodule breaks unauthenticated access:

52ce8c9

Potential option here is to switch to https

... but I want to say: many thanks.
Please do not discontinue this excelent work.
I had 2 lenovo ThinkSystem SR630 with a lot of dmesg error.
After installed this pve-edge-kernel everything works as a charm.
Again many thanks.

Just wondering when to expect 5.8.x kernel releases.

I'm having the same issue experienced by this person on the proxmox forums
https://forum.proxmox.com/threads/no-zvol-device-link-found-after-300-sec-found-issue.94514/

tl;dr VMs fail to start with error 'no zvols found after 300 secs'

Have been using your builds for a while now, and it's been great!

I have had one issue though, detailed in this proxmox forum thread. It appears to have been resolved in the latest 5.15 kernel via some patches, and I wondered if they could be applied here as well please? The original commit was here I think.

I'm running the very latest 5.19 and still seeing the issue reproduced, so it seems to not have been fixed in the kernel itself.

Many thanks!

5.13.14 and 5.14 kernels won't boot on systems with NVDIMM due to UBSAN and NFIT module issues.

The issue is likely not specific to NVDIMM or NFIT and may appear in variety of different configs.

The solution in my case is to either remove Intel Optane P100 memory from the servers or disable UBSAN.

More info:

https://github.com/torvalds/linux/blob/master/lib/Kconfig.ubsan#L23

and my report to Ubuntu's kernel team:

https://lists.ubuntu.com/archives/kernel-team/2021-September/123828.html

thank you.

It looks like 5.11 is now EOL. What are you thoughts on jumping to 5.12?

I built it for my systems with the latest ZFS 2.1.0-rc5.

So far, it appears to be functional:

root@pve-bfs-1:~# uname -a
Linux pve-bfs-1 5.12.6-1-edge #1 SMP 5.12.6-1-edge (Wed, 19 May 2021 12:10:07 +0000) x86_64 GNU/Linux

dmesg | grep -i zfs

[ 6.037061] ZFS: Loaded module v2.1.0-rc5, ZFS pool version 5000, ZFS filesystem version 5

root@pve-bfs-1:~# modinfo zfs | grep -i version
version: 2.1.0-rc5
srcversion: 18176EA5C58E37536B4B6C2
vermagic: 5.12.6-1-edge SMP mod_unload modversions

zfs list

NAME USED AVAIL REFER MOUNTPOINT
zfs-test 420K 860G 96K /zfs-test

Other things tested: PCI passthrough, pmem passthough. ceph, corosync clustering, Mellanox NICs.

Hi

Due to this issue , I need this and this patches in the kernel.

Does the last of your kernels include those patches? If not, how hard would it be to include them? Thanks.

Looks like using the 5.16 pve edge reintroduces this issue or a similar one when using nginx inside an LXC
https://github.com/lxc/lxd/issues/8383
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=f8ad8187c3b536ee2b10502a8340c014204a1af0

I do not know exactly when this was introduced or if it affects earlier versions, I just noticed it on a low traffic website.

I'm not sure if the issue is with LXC or upstream with the kernel, I also am using 4K aligned SSDs which may be a problem however setting nginx's direct alignment to use 4k blocks didn't fix it.

As a workaround, you can set sendfile off in nginx conf.

Hi,

with https://github.com/fabianishere/pve-edge-kernel/releases/tag/v5.17.14-1 I get frequent reboots - about every hour - and services (lxc containers) are not starting (hanging).
I'm still looking for more and better error logs.

Anyone else having this problem?
(I know and apologize for the sparse information - I'm off site and since the machines are headless investigation into the causes or even the specific errors are slow. I'm currently working to roll back to a previous release.)

i knew there are some paid solution out there, but i hope there is some other solution that free for home user like me. thanks.

GPU pass-through runs stably for 3 months, except lxc can’t be used，kernel 5.8.

12700K GPU or unrecognized，5.16 Not supported？

Got this error when I compile my drivers:
make[2]: *** /lib/modules/5.9.16-1/build: No such file or directory. Stop.
Installed kernel with apt install ./pve-edge-kernel-5.9.16-1_5.9.16-1+zen21_amd64.deb command.
And cannot find linux headers package, the apt install linux-headers-`uname -r` doesn't work.

Hi,

any timeline on updating to Linux 5.10?

Thanks

With some amd video cards, they need to be reset via a nonstandard method in order to work with PCI pass through. A generic image would be wonderful.

gnif/linux@e3bb0b3

Forum post: https://forum.level1techs.com/t/navi-reset-kernel-patch/147547/47

https://del.dog/Navipatch.txt

i install the 5.16 after that, now container is starting.

There is a niche group of network engineers that use Proxmox and Eve-NG in an lxc container. For this to work properly, we need to patch the kernel with a specific bridge patch. I will attach the patch and some links with additional information. This patch can be dropped into the pve patches directory and should work fine.

https://lists.linuxfoundation.org/pipermail/bridge/2015-January/009291.html
https://interestingtraffic.nl/2017/11/21/an-oddly-specific-post-about-group_fwd_mask/
br_fwd_grp_mask.patch.zip

Hi Fabian!

Thanks for this project which has been very helpful to me!

Your Zen 2 kernel seems to work fine with Proxmox on my Zen 3 5950X, but I wonder if you are planning to add Zen 3.

Thanks again so much!

Best wishes and kindest regards,

Tom

I want to build the kernel with PPP(point-to-point protocol) support. But I found that the make menuconfig cannot be used.

Hello I created simple update script for unattended updates from cron. I am aware that automatic updates of custom build kernel aren't the best idea for hypervisor but I hope someone will find it useful.

You can see it here https://gist.github.com/pkejval/3c5a2dce94080185a05de6c08508ba04

Hello,

Is it possible at all to build a recent kernel with the latest rc1 of OpenZFS 2.0? I see there is a pre release up & would be interested in testing it.

Thanks as always!

Hello, I am having difficulties following the manual build process. I've tried this now in the following scenarios: Directly on my Proxmox host (Debian 11), a Debian 10 CT, an Ubuntu 21 CT, and a Debian 11 CT.

The result is always the same, the process ends here

  ...
  HDRINST usr/include/asm/fcntl.h
  HDRINST usr/include/asm/termios.h
  HDRINST usr/include/asm/ipcbuf.h
  HDRINST usr/include/asm/socket.h
  HDRINST usr/include/asm/termbits.h
  HDRINST usr/include/asm/unistd_32.h
  INSTALL /home/xaekai/src/pve-edge-kernel/debian/pve-kernel-libc-dev/usr/include
make[2]: Leaving directory '/home/xaekai/src/pve-edge-kernel/linux'
rm -rf debian/pve-kernel-libc-dev/usr/include/drm debian/pve-kernel-libc-dev/usr/include/scsi
find debian/pve-kernel-libc-dev/usr/include \( -name .install -o -name ..install.cmd \) -execdir rm {} +
# Move include/asm to arch-specific directory
mkdir -p debian/pve-kernel-libc-dev/usr/include/x86_64-linux-gnu
mv debian/pve-kernel-libc-dev/usr/include/asm debian/pve-kernel-libc-dev/usr/include/x86_64-linux-gnu/
test ! -d debian/pve-kernel-libc-dev/usr/include/arch || \
        mv debian/pve-kernel-libc-dev/usr/include/arch debian/pve-kernel-libc-dev/usr/include/x86_64-linux-gnu/
touch .usr_headers_install_mark
make[1]: Leaving directory '/home/xaekai/src/pve-edge-kernel'
make: *** [debian/rules:16: binary] Error 2
dpkg-buildpackage: error: fakeroot debian/rules binary subprocess returned exit status 2
debuild: fatal error at line 1182:
dpkg-buildpackage -us -uc -ui --jobs=auto -b failed

That "Line 1182" Of what? Some virtually compiled script that only exists during execution? Not helpful. It's been a few years since I was intimately familiar with deb packaging, so I'm out of my depth here.

The steps I took to get here this last time:
1.) Create fresh Debian 11 CT
2.) Login root through Proxmox console, confirm network, install sudoers and openssh, setup regular user with sudo
3.) Shell in from regular terminal
4.) Run these commands

sudo apt update
sudo apt upgrade
sudo apt install devscripts debhelper equivs git
mkdir src
cd src
git clone https://github.com/fabianishere/pve-edge-kernel
cd pve-edge-kernel
git checkout v5.13.5-1 
git submodule update --init --depth=1 --recursive linux
git submodule update --init --recursive
debian/rules debian/control
sudo mk-build-deps -i
export PVE_KERNEL_CFLAGS="-march=znver2"
debuild -ePVE* --jobs=auto -b -uc -us

What am I doing wrong?

i run openwrt in lxc, and use pppoe ,but not work,because the host pve kernel not found slhc.ko and ppp_generice.ko，can you compile its into pve kernel。i try to compile ,but fail 。thank you。

Hello Fabian,

are included in the kernel AMD P-state. When I try to load the module, I get the message that no driver is available.

Following the instructions on readme (on Proxmox 7.2 / 5.15), we get a warning:

Warning: apt-key is deprecated. Manage keyring files in trusted.gpg.d instead (see apt-key(8)).

Should the instructions be updated to use that instead?