On Oct 18 15:01, Julia wrote:
> Actually, it seems that trace_pci_nvme_irq_pin is emitted even if the
> IRQ is not asserted due to a setting of the interrupt masks. Which is
> weird because there's no corresponding one for deasserting. Possibly
> this should be reworded for 'interrupt is high (but
Actually, it seems that trace_pci_nvme_irq_pin is emitted even if the IRQ is
not asserted due to a setting of the interrupt masks. Which is weird because
there's no corresponding one for deasserting. Possibly this should be reworded
for 'interrupt is high (but might be masked?)', or just leave
The NVMe subsystem logs "pulsing IRQ pin" when it is asserting the PCI(e)
IRQ. This is confusing as it implies a short pulse, not the level-triggered
interrupts PCI(e) uses.
Also remove the pci_irq_pulse() function marked with FIXME as it is no
longer used by any calls.
Signed-off-by: julia
---
