It seems this bug or something very similar may have surfaced again. When a package postinst replaces an apparmor profile,
Jan 6 15:52:43 hapatus kernel: [103287.157730] apparmor_parser: page allocation failure. order:9, mode:0x40d0 Jan 6 15:52:43 hapatus kernel: [103287.178618] apparmor_parser: page allocation failure. order:7, mode:0x40d0
happens, and after that point, it seems a kernel process is taking all available CPU time to do something in an infinite loop, and all available RAM seems to be continuously allocated by the kernel.
It seems this bug or something very similar may have surfaced again. When a package postinst replaces an apparmor profile,
Jan 6 15:52:43 hapatus kernel: [103287.157730] apparmor_parser: page allocation failure. order:9, mode:0x40d0
Jan 6 15:52:43 hapatus kernel: [103287.178618] apparmor_parser: page allocation failure. order:7, mode:0x40d0
happens, and after that point, it seems a kernel process is taking all available CPU time to do something in an infinite loop, and all available RAM seems to be continuously allocated by the kernel.