On Fri, May 26, 2023, Fabio Coatti wrote: > Hi all, > I'm using vanilla kernels on a gentoo-based laptop and since 6.3.2 What was the last kernel you used that didn't trigger this WARN? > I'm getting the kernel log below when using kvm VM on my box. Are you doing anything "interesting" when the WARN fires, or are you just running the VM and it random fires? Either way, can you provide your QEMU command line? > I know, kernel is tainted but avoiding to load nvidia driver could make > things complicated on my side; if needed for debug I can try to avoid it. Nah, don't worry about that at this point. > Not sure which other infos can be relevant in this context; if you > need more details just let me know, happy to provide them. > > [Fri May 26 09:16:35 2023] ------------[ cut here ]------------ > [Fri May 26 09:16:35 2023] WARNING: CPU: 5 PID: 4684 at > kvm_nx_huge_page_recovery_worker+0x38c/0x3d0 [kvm] Do you have the actual line number for the WARN? There are a handful of sanity checks in kvm_recover_nx_huge_pages(), it would be helpful to pinpoint which one is firing. My builds generate quite different code, and the code stream doesn't appear to be useful for reverse engineering the location.
