Re: [PATCH v2 -next] genirq: Separate warning messages of enable of suspended irq

[chenlifu]

在 2022/9/20 14:42, chenlifu 写道:
> 在 2022/7/28 18:29, Chen Lifu 写道:
> > Both unbalanced enable for irq and enable of suspended irq
> > print same warning messages, such as:
> > ......
> > [20220314100142]------------[ cut here ]------------
> > [20220314100142]WARNING: CPU: 1 PID: 10332 at kernel/irq/manage.c:723 
> > __enable_irq+0x58/0xa4
> > [20220314100142]Unbalanced enable for IRQ 38
> > ......
> >
> > It's confusing. Separate the warning messages to make it clear.
> > No functional changes.
> >
> > Signed-off-by: Chen Lifu <chenlifu@xxxxxxxxxx>
> > ---
> >   kernel/irq/manage.c | 8 +++++---
> >   1 file changed, 5 insertions(+), 3 deletions(-)
> >
> > diff --git a/kernel/irq/manage.c b/kernel/irq/manage.c
> > index 40fe7806cc8c..bf4f5f63b2d3 100644
> > --- a/kernel/irq/manage.c
> > +++ b/kernel/irq/manage.c
> > @@ -775,17 +775,19 @@ void disable_nmi_nosync(unsigned int irq)
> >   void __enable_irq(struct irq_desc *desc)
> >   {
> >       switch (desc->depth) {
> >       case 0:
> > - err_out:
> >           WARN(1, KERN_WARNING "Unbalanced enable for IRQ %d\n",
> >                irq_desc_get_irq(desc));
> >           break;
> >       case 1: {
> > -        if (desc->istate & IRQS_SUSPENDED)
> > -            goto err_out;
> > +        if (desc->istate & IRQS_SUSPENDED) {
> > +            WARN(1, KERN_WARNING "Enable of suspended IRQ %d\n",
> > +                 irq_desc_get_irq(desc));
> > +            break;
> > +        }
> >           /* Prevent probing on this irq: */
> >           irq_settings_set_noprobe(desc);
> >           /*
> >            * Call irq_startup() not irq_enable() here because the
> >            * interrupt might be marked NOAUTOEN. So irq_startup()
>
> Friendly ping ...
>
> .

Friendly ping ...