Re: [PATCH 14/42] PCI: aardvark: Don't mask irq when mapping

[Pali Rohár]

On Friday 07 May 2021 10:20:39 Marc Zyngier wrote:
> On Thu, 06 May 2021 16:31:25 +0100,
> Pali Rohár <pali@xxxxxxxxxx> wrote:
> > 
> > By default, all Legacy INTx interrupts are masked, so there is no need to
> > mask this interrupt during irq_map callback.
> 
> What guarantees that they are actually masked? I would actually assume
> that the HW is in an unknown state at boot time.

Function advk_pcie_setup_hw() during driver probing mask all INTx interrupts:

    advk_writel(pcie, PCIE_ISR1_ALL_MASK, PCIE_ISR1_MASK_REG);

Individual INTx interrupts can be unmasked by PCIE_ISR1_INTX_ASSERT(val).
Macro PCIE_ISR1_ALL_MASK contains all bits from PCIE_ISR1_INTX_ASSERT(val).

> Thanks,
> 
> 	M.
> 
> > 
> > Signed-off-by: Pali Rohár <pali@xxxxxxxxxx>
> > Reviewed-by: Marek Behún <kabel@xxxxxxxxxx>
> > ---
> >  drivers/pci/controller/pci-aardvark.c | 1 -
> >  1 file changed, 1 deletion(-)
> > 
> > diff --git a/drivers/pci/controller/pci-aardvark.c b/drivers/pci/controller/pci-aardvark.c
> > index 2aced8c9ae9f..08f1157e1c5e 100644
> > --- a/drivers/pci/controller/pci-aardvark.c
> > +++ b/drivers/pci/controller/pci-aardvark.c
> > @@ -940,7 +940,6 @@ static int advk_pcie_irq_map(struct irq_domain *h,
> >  {
> >  	struct advk_pcie *pcie = h->host_data;
> >  
> > -	advk_pcie_irq_mask(irq_get_irq_data(virq));
> >  	irq_set_status_flags(virq, IRQ_LEVEL);
> >  	irq_set_chip_and_handler(virq, &pcie->irq_chip,
> >  				 handle_level_irq);
> > -- 
> > 2.20.1
> > 
> > 
> 
> -- 
> Without deviation from the norm, progress is not possible.