[PATCH] PCI:MSI Return -ENOSPC when requested vectors is not enough

[Christoph Hellwig]

On Wed, Jan 11, 2017 at 12:18:53PM -0600, Bjorn Helgaas wrote:
> Unless Christoph objects, I'll apply this, but I don't understand the
> situation with 17a51f12cfbd.  That commit doesn't check for EINVAL or
> ENOSPC so I don't know what the connection with this patch is.

I don't think that commit is the culprit.

> I know Christoph said he changed something in ahci to treat all errors
> the same, but I don't know where that is, either.

"ahci: always fall back to single-MSI mode"