Trying to build a custom Proxmox kernel from https://github.com/proxmox/pve-kernel currently fails as it tries to fetch the ubuntu-kernel submodule from https://github.com/proxmox/mirror_ubuntu-kernels which 404s (presumably its a private repo)...
kernelnewbies generally has a good list (ie https://kernelnewbies.org/Linux_6.17 https://kernelnewbies.org/Linux_6.16 and https://kernelnewbies.org/Linux_6.15), no need to use an LLM here.
Apparently I spoke too soon on the -ice-fix builds. The buffer_head slab allocations do rise for the first few hours, but eventually seem to flatten out. I see something like 300MB of used buffer_head slab allocations (per slabtop) which of...
As I posted last Thursday, I tested that as well with no change. That kernel does *not* contain the "ice: fix Rx page leak on multi-buffer frames" fix (which is the -ice-fix builds), though of course those also don't fix the issue.
Just to confirm what others are seeing, the 6.14.11-2-ice-fix-1-pve kernel shows a similar rate of buffer_head slab leaking as other 6.14 kernels (with 9k MTU ceph on an ice card).
After upgrading to 6.14.11-3-pve I still see `buffer_head` slab increasing slowly over time. Its obvious, but might be worth noting that the ice memory leak fix previously identified isn't in the above list.
After upgrading to proxmox 9 my hosts that use the ice driver (ie Intel 25G NICs) for the ceph network (with jumbo frames) show a pretty quick memory leak (something like 10GB/day). I assume it was fixed in upstream kernel commit...