Re: BUG: Bad page state in process systemd-udevd (was: [PATCH v9 bpf-next 1/9] x86/Kconfig: select HAVE_ARCH_HUGE_VMALLOC with HAVE_ARCH_HUGE_VMAP)
From: Edgecombe, Rick P
Date: Mon Mar 28 2022 - 15:22:14 EST
On Mon, 2022-03-28 at 06:37 +0000, Song Liu wrote:
> + Rick, who highlighted some potential issues with this. (also
> attached
> the stack trace).
Yea, this looks like something else. Just a wild guess, but could some
vmalloc() caller be mucking with the struct page of the backing
allocation in way that trips up compound pages?