Re: [PATCH 2/2] ptrace: fix set_task_blockstep()->update_debugctlmsr() logic

[Oleg Nesterov <oleg@redhat.com>]

On 08/03, Sebastian Andrzej Siewior wrote:
>
> On 08/03/2012 06:29 PM, Oleg Nesterov wrote:
>> diff --git a/arch/x86/kernel/step.c b/arch/x86/kernel/step.c
>> index afa60db..636402e 100644
>> --- a/arch/x86/kernel/step.c
>> +++ b/arch/x86/kernel/step.c
>> @@ -166,12 +166,18 @@ static void set_task_blockstep(struct task_struct *task, bool on)
>>   	else
>>   		clear_tsk_thread_flag(task, TIF_BLOCKSTEP);
>>
>> +	if (task != current)
>> +		return;
>> +
>> +	/* ensure irq/preemption can't change debugctl in between */
>> +	local_irq_disable();
>>   	debugctl = get_debugctlmsr();
>>   	if (on)
>>   		debugctl |= DEBUGCTLMSR_BTF;
>>   	else
>>   		debugctl&= ~DEBUGCTLMSR_BTF;
>>   	update_debugctlmsr(debugctl);
>> +	local_irq_enable();
>
> wouldn't preempt_disable() be enough?

preempt_disable() can't help if interrupt handler changes
other bits in between?

Oleg.