Re: 4.19.106-rt44 -- boot problems with irqwork: push most work into softirq context

[Pavel Machek]

Hi!

On Fri 2020-03-20 16:05:45, Steven Rostedt wrote:
> On Fri, 20 Mar 2020 20:54:32 +0100
> Pavel Machek <pavel@xxxxxxx> wrote:
> 
> > > Does this patch help?  
> > 
> > I don't think so. It also failed, and the failure seems to be
> > identical to me.
> > 
> > https://gitlab.com/cip-project/cip-kernel/linux-cip/tree/ci/pavel/linux-cip
> > https://lava.ciplatform.org/scheduler/job/13110
> > 
> 
> Can you send me a patch that shows the difference between the revert that
> you say works, and the upstream v4.19-rt tree (let me know which version
> of v4.19-rt you are basing it on).

I was using -rt44, and yes, I can probably generate better diffs.

But I guess I found it with code review: how does this look to you? I
applied it on top of your fix, and am testing. 2 successes so far.

Signed-off-by: Pavel Machek <pavel@xxxxxxx>

Best regards,
								Pavel

commit aa034c3060dbab96a7b6d6bf827504b394bed15b
Author: Pavel Machek <pavel@xxxxxx>
Date:   Sat Mar 21 22:58:43 2020 +0100

    It sounds like irq_work_queue() was queueing on wrong list?

diff --git a/kernel/irq_work.c b/kernel/irq_work.c
index 0ca75c77536b..dd654865c219 100644
--- a/kernel/irq_work.c
+++ b/kernel/irq_work.c
@@ -81,7 +81,8 @@ bool irq_work_queue(struct irq_work *work)
 
 	/* Queue the entry and raise the IPI if needed. */
 	preempt_disable();
-	if (IS_ENABLED(CONFIG_PREEMPT_RT_FULL) && !(work->flags & IRQ_WORK_HARD_IRQ))
+	if ((IS_ENABLED(CONFIG_PREEMPT_RT_FULL) && !(work->flags & IRQ_WORK_HARD_IRQ))
+	    || (work->flags & IRQ_WORK_LAZY))
 		list = this_cpu_ptr(&lazy_list);
 	else
 		list = this_cpu_ptr(&raised_list);



      	    	       	      	  	   	    	   Pavel
-- 
DENX Software Engineering GmbH,      Managing Director: Wolfgang Denk
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany

Attachment: signature.asc
Description: PGP signature