Messages in this thread | | | Date | Sun, 15 Nov 2015 16:59:57 -0800 | Subject | Re: BUG: unable to handle kernel paging request at ffffe8ff7fc00001 | From | Linus Torvalds <> |
| |
On Sun, Nov 15, 2015 at 2:28 PM, Kyle Sanderson <kyle.leet@gmail.com> wrote: > [] BUG: unable to handle kernel paging request at ffffe8ff7fc00001 > [] IP: [<ffffffff810a174f>] kstat_irqs+0x4f/0x90 > [] CPU: 2 PID: 1078 Comm: usage.pl Not tainted 4.1.7-hardened-r1 #1 > [] Hardware name: Supermicro Super Server/X10SRi-F, BIOS 1.0b 04/21/2015 RSI: 000060f700000001 > [] Call Trace: > [] [<>] kstat_irqs_usr+0x1e/0x40 > [] [<>] show_stat+0x5ca/0x690 > [] [<>] seq_read+0xcd/0x3b0 > [] [<>] proc_reg_read+0x43/0x70 > [] [<>] __vfs_read+0x23/0xd0 > [] [<>] vfs_read+0xc0/0x1d0 > [] [<>] SyS_read+0x41/0xb0 > [] [<>] system_call_fastpath+0x16/0x72 > [] Code: 83 78 48 00 74 56 4c 8b ...
The code ends up being
mov 0x48(%r13),%rsi mov __per_cpu_offset(,%rcx,8),%rcx add (%rsi,%rcx,1),%ebx <-- trapping instruction
which is just the
sum += *per_cpu_ptr(desc->kstat_irqs, cpu);
part of kstat_irqs().
Your registers being
RSI: 000060f700000001 RCX: ffff88087fc00000
and it's RSI that makes no sense - RCX looks like a real kernel pointer. So it looks like it's the "desc->kstat_irqs" thing that is for some reason garbage.
I don't see any sane possible reason this would happen, though. Thomas, does this look like anything you've seen before?
Linus
| |