Re: Default enable RCU list lockdep debugging with PROVE_RCU

[Paul E. McKenney]

On Mon, May 18, 2020 at 01:54:13PM +0800, Rong Chen wrote:
> 
> 
> On 5/18/20 5:47 AM, Paul E. McKenney wrote:
> > On Fri, May 15, 2020 at 02:36:26PM -0400, Qian Cai wrote:
> > > 
> > > > On May 14, 2020, at 2:13 PM, Paul E. McKenney <paulmck@xxxxxxxxxx> wrote:
> > > > 
> > > > Fair enough!  And yes, the Linux kernel is quite large, so I certainly am
> > > > not asking you to test the whole thing yourself.
> > > Ok, I saw 0day bot also started to report those which is good. For example,
> > > 
> > > lkml.org/lkml/2020/5/12/1358
> > > 
> > > which so far is nit blocking 0day on linux-next since it does not use panic_on_warn yet (while syzbot does).
> > > 
> > > Thus, I am more convinced that we should not revert the commit just for syzbot until someone could also convince 0day to select RCU_EXPERT and then DEBUG_RCU_LIST?
> > Let's ask the 0day people, now CCed, if they would be willing to
> > build with CONFIG_RCU_EXPERT=y and CONFIG_DEBUG_RCU_LIST=y on some
> > fraction of their testing.  ;-)
> > 
> > 							Thanx, Paul
> 
> Hi,
> 
> Thanks for your advice, we'll support it in the near future.

Thank you very much!

							Thanx, Paul