Hi Dejin and Andy,

[...]

quoted

quoted

Question: wouldn't you need to call pci_free_irq_vectors() somewhere,
possibly to pcim_release() callback?  Although, I am not sure where the
right place would be.

I am asking, as the documentation (see [4]) suggests that one would have
to release allocated IRQ vectors (relevant exceprt):

It's done in pcim_release() but not explicitly.

        if (dev->msi_enabled)
                pci_disable_msi(dev);
        if (dev->msix_enabled)
                pci_disable_msix(dev);

Maybe above can be replaced by pci_free_irq_vectors() to be sure that any
future change to PCI IRQ allocation APIs.

Yes, I have checked and currently the above code is equivalent to
pci_free_irq_vectors().

Dejin, please update your patch accordingly.

Hi Andy and Krzysztof,

I have modified it and sent patch v2. thank you very much!

Thank you Andy for double-checking!  Much appreciated.

Moving to pci_free_irq_vectors() directly looks great as we are also
removing the surplus checks that pcim_release() is doing - since both
the pci_disable_msi() and the pci_disable_msix() are doing internal
validation to see whether MSI/MSI-X is currently enabled.

Thank you again, Dejin and Andy. :)

Krzysztof