From mboxrd@z Thu Jan 1 00:00:00 1970 From: agordeev@redhat.com (Alexander Gordeev) Date: Tue, 8 Oct 2013 11:07:16 +0200 Subject: [PATCH RFC 00/77] Re-design MSI/MSI-X interrupts enablement pattern In-Reply-To: <20131007182117.GC27396@htj.dyndns.org> References: <20131007182117.GC27396@htj.dyndns.org> Message-ID: <20131008090716.GA10561@dhcp-26-207.brq.redhat.com> On Mon, Oct 07, 2013@02:21:17PM -0400, Tejun Heo wrote: > Whee.... that's a lot more than I expected. I was just scanning > multiple msi users. Maybe we can stage the work in more manageable > steps so that you don't have to go through massive conversion only to > do it all over again afterwards and likewise people don't get > bombarded on each iteration? Maybe we can first update pci / msi code > proper, msi and then msix? Multipe MSIs is just a handful of drivers, really. MSI-X impact still will be huge. But if we opt a different name for the new pci_enable_msix() then we could first update pci/msi, then drivers (in few stages possibly) and finally remove the old implementation. > tejun -- Regards, Alexander Gordeev agordeev at redhat.com