[PATCH 1/2] ftrace: disable preemption on the testing of recursion

王贇 yun.wang at linux.alibaba.com
Wed Oct 13 13:38:28 AEDT 2021



On 2021/10/13 上午10:30, Steven Rostedt wrote:
> On Wed, 13 Oct 2021 10:04:52 +0800
> 王贇 <yun.wang at linux.alibaba.com> wrote:
> 
>> I see, while the user can still check smp_processor_id() after trylock
>> return bit 0...
> 
> But preemption would have already been disabled. That's because a bit 0
> means that a recursion check has already been made by a previous
> caller and this one is nested, thus preemption is already disabled.
> If bit is 0, then preemption had better be disabled as well.

Thanks for the explain, now I get your point :-)

Let's make bit 0 an exemption then.

Regards,
Michael Wang

> 
> -- Steve
> 


More information about the Linuxppc-dev mailing list