On Mon, Oct 19, 2015 at 11:49 AM, Petri Gynther <pgynther@xxxxxxxxxx> wrote:
>
> We need to enter NMI context when NMI interrupt fires.
>
> Signed-off-by: Petri Gynther <pgynther@xxxxxxxxxx>
> ---
> arch/mips/kernel/traps.c | 2 ++
> 1 file changed, 2 insertions(+)
>
> diff --git a/arch/mips/kernel/traps.c b/arch/mips/kernel/traps.c
> index fdb392b..efcedd4 100644
> --- a/arch/mips/kernel/traps.c
> +++ b/arch/mips/kernel/traps.c
> @@ -1856,12 +1856,14 @@ void __noreturn nmi_exception_handler(struct pt_regs *regs)
> {
> char str[100];
>
> + nmi_enter();
> raw_notifier_call_chain(&nmi_chain, 0, regs);
> bust_spinlocks(1);
> snprintf(str, 100, "CPU%d NMI taken, CP0_EPC=%lx\n",
> smp_processor_id(), regs->cp0_epc);
> regs->cp0_epc = read_c0_errorepc();
> die(str, regs);
> + nmi_exit();
> }
>
> #define VECTORSPACING 0x100 /* for EI/VI mode */
> --
> 2.6.0.rc2.230.g3dd15c0
>
Any comments/concerns about this patch?
On our systems, we have seen stack traces like this:
<4>[158549.586000] : [<800103c0>] show_stack+0x78/0x90
<4>[158549.586000] : [<807340a0>] dump_stack+0xd0/0x144
<4>[158549.586000] : [<8006475c>] dequeue_task_idle+0x38/0x4c
<4>[158549.586000] : [<80735c8c>] __schedule+0x49c/0xa68
<4>[158549.586000] : [<80736294>] schedule+0x3c/0x9c
<4>[158549.586000] : [<80739930>] schedule_timeout+0x144/0x254
<4>[158549.586000] : [<800964f8>] msleep+0x40/0x54
<4>[158549.586000] : [<80010604>] die+0x124/0x130
<4>[158549.586000] : [<80012abc>] set_vi_handler+0x0/0x24
<4>[158549.586000] :
<3>[158549.586000] bad: scheduling from the idle thread!
80012a24 <nmi_exception_handler>:
...
80012ab4: 0c004138 jal 800104e0 <die>
80012ab8: 02002821 move a1,s0
80012abc <set_vi_handler>:
...
