Hi Andy, On Thu, 8 Aug 2024 15:32:07 +0300 Andy Shevchenko <andy.shevchenko@xxxxxxxxx> wrote: > On Wed, Aug 7, 2024 at 1:10 PM Herve Codina <herve.codina@xxxxxxxxxxx> wrote: > > On Mon, 5 Aug 2024 22:13:38 +0200 > > Andy Shevchenko <andy.shevchenko@xxxxxxxxx> wrote: > > > On Mon, Aug 5, 2024 at 12:19 PM Herve Codina <herve.codina@xxxxxxxxxxx> wrote: > > ... > > > > > + if (!pdev->irq) > > > > + return ERR_PTR(-EOPNOTSUPP); > > > > > > Before even trying to get it via APIs? (see below as well) > > > Also, when is it possible to have 0 here? > > > > pdev->irq can be 0 if the PCI device did not request any IRQ > > (i.e. PCI_INTERRUPT_PIN in PCI config header is 0). > > > I use that to check whether or not INTx is supported. > > But why do you need that? What happens if you get a new device that > supports let's say MSI? > > > Even if this code is present in the LAN966x PCI driver, it can be use as a > > starting point for other drivers and may be moved to a common part in the > > future. > > > > Do you think I should remove it ? > > I think pci_alloc_vectors() should be enough. Make it to be the first > call, if you think it's better. > Thanks for your answer. I will remove the pdev->irq check an rely pci_alloc_vectors(). Not sure that I will move the call to the first call. Best regards. Hervé
