Re: [PATCH 2/6] ipr: use pci_irq_allocate_vectors

2016-10-11 Thread Martin K. Petersen
> "Christoph" == Christoph Hellwig writes: Christoph> Switch the ipr driver to use pci_alloc_irq_vectors. We need Christoph> to two calls to pci_alloc_irq_vectors as ipr only supports Christoph> multiple MSI-X vectors, but not multiple MSI vectors. Christoph> Otherwise this

Re: [PATCH 2/6] ipr: use pci_irq_allocate_vectors

2016-09-29 Thread Christoph Hellwig
On Thu, Sep 29, 2016 at 09:01:44AM -0500, Brian King wrote: > Thanks Christoph. Very nice. As I was reviewing the patch, I noticed > the additional PCI_IRQ_AFFINITY flag, which is currently not being set > in this patch. Is the intention to set that globally by default, or > should I follow up

Re: [PATCH 2/6] ipr: use pci_irq_allocate_vectors

2016-09-29 Thread Brian King
Thanks Christoph. Very nice. As I was reviewing the patch, I noticed the additional PCI_IRQ_AFFINITY flag, which is currently not being set in this patch. Is the intention to set that globally by default, or should I follow up with a one liner to add that to the ipr driver in the next patch set I

Re: [PATCH 2/6] ipr: use pci_irq_allocate_vectors

2016-09-19 Thread Martin K. Petersen
> "Christoph" == Christoph Hellwig writes: Christoph> Switch the ipr driver to use pci_alloc_irq_vectors. We need Christoph> to two calls to pci_alloc_irq_vectors as ipr only supports Christoph> multiple MSI-X vectors, but not multiple MSI vectors. Christoph> Otherwise this

[PATCH 2/6] ipr: use pci_irq_allocate_vectors

2016-09-11 Thread Christoph Hellwig
Switch the ipr driver to use pci_alloc_irq_vectors. We need to two calls to pci_alloc_irq_vectors as ipr only supports multiple MSI-X vectors, but not multiple MSI vectors. Otherwise this cleans up a lot of cruft and allows to use a common request_irq loop for irq types, which happens to only