[PATCH v2] ARM: Fix "BUG: scheduling while atomic: swapper/0/0x00000002

[Philby John]

Hello Santosh,

On 04/17/2010 01:45 PM, Santosh Shilimkar wrote:
> This patch fixes the preempt leak in the cpuidle path invoked from
> cpu-hotplug. The fix is suggested by Russell King and is based
> on x86 idea of calling init_idle() on the idle task when it's
> re-used which also resets the preempt count amongst other things
>
> dump:
> BUG: scheduling while atomic: swapper/0/0x00000002
> Modules linked in:
> Backtrace:
> [<c0024f90>] (dump_backtrace+0x0/0x110) from [<c0173bc4>] (dump_stack+0x18/0x1c)
>   r7:c02149e4 r6:c033df00 r5:c7836000 r4:00000000
> [<c0173bac>] (dump_stack+0x0/0x1c) from [<c003b4f0>] (__schedule_bug+0x60/0x70)
> [<c003b490>] (__schedule_bug+0x0/0x70) from [<c0174214>] (schedule+0x98/0x7b8)
>   r5:c7836000 r4:c7836000
> [<c017417c>] (schedule+0x0/0x7b8) from [<c00228c4>] (cpu_idle+0xb4/0xd4)
> # [<c0022810>] (cpu_idle+0x0/0xd4) from [<c0171dd8>] (secondary_start_kernel+0xe0/0xf0)
>   r5:c7836000 r4:c0205f40
> [<c0171cf8>] (secondary_start_kernel+0x0/0xf0) from [<c002d57c>] (prm_rmw_mod_reg_bits+0x88/0xa4)
>   r7:c02149e4 r6:00000001 r5:00000001 r4:c7836000
> Backtrace aborted due to bad frame pointer<c7837fbc>
>
> Signed-off-by: Santosh Shilimkar<santosh.shilimkar at ti.com>
> Cc: Russell King<rmk+kernel at arm.linux.org.uk>
> Cc: Catalin Marinas<catalin.marinas at arm.com>
> Cc: Sundar R IYER<sundar.iyer at stericsson.com>
> ---
>   arch/arm/kernel/smp.c |    8 +++++++-
>   1 files changed, 7 insertions(+), 1 deletions(-)
>
> diff --git a/arch/arm/kernel/smp.c b/arch/arm/kernel/smp.c
> index 57162af..3e14629 100644
> --- a/arch/arm/kernel/smp.c
> +++ b/arch/arm/kernel/smp.c
> @@ -86,8 +86,14 @@ int __cpuinit __cpu_up(unsigned int cpu)
>   			return PTR_ERR(idle);
>   		}
>   		ci->idle = idle;
> +	} else {
> +		/*
> +		 * Since idle thread getting re-used, call
> +		 * init_idle() to set up the idle thread for given
> +		 * CPU. Additionally it also resets the preempt count
> +		 */
> +		init_idle(idle, cpu);

Since Sundar was the first to suggest this fix, shouldn't his SOB be 
included as well instead of CC ? Or did I miss a mail thread ?


Regards,
Philby