Hi folks, I've been trying to play with this report and was able to reproduce on v6.8-rc2, in a simple qemu VM. But the thing is: after looking similar reports in MLs, this seems quite the same report as [0], so a dup. And we even have a candidate fix for it, in the form of Thomas's patch (https://lore.kernel.org/all/87r0jwquhv.ffs@tglx/). I've tested this patch and it works, preventing the crash. So... Jann: could you help me confirm the reproducer here is the same of the other report, in which you nailed it to accessing the VSYSCALL region? For me it's quite similar, but I'm not experienced in reading this kind of BPF program... Thomas: could you maybe re-submit/merge this patch, if you still agree this is the proper fix? There's a Tested-by from Hou Tao in that thread, and feel free to add mine as well! Thanks in advance and let me know if I can test more stuff / provide more data, etc - I'm glad to help here. Cheers, Guilherme [0] https://lore.kernel.org/all/000000000000c84343060a850bd0@xxxxxxxxxx/ ("[syzbot] [mm?] BUG: unable to handle kernel paging request in copy_from_kernel_nofault")
