From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Return-Path: Date: Sun, 8 May 2016 11:05:58 +0200 From: Christoph Hellwig To: Bjorn Helgaas Cc: Christoph Hellwig , agordeev@redhat.com, pjw@netapp.com, axboe@fb.com, keith.busch@intel.com, linux-pci@vger.kernel.org, linux-nvme@lists.infradead.org Subject: Re: [PATCH 1/2] PCI: Provide sensible irq vector alloc/free routines Message-ID: <20160508090558.GA11677@lst.de> References: <1462457096-19795-1-git-send-email-hch@lst.de> <1462457096-19795-2-git-send-email-hch@lst.de> <20160506160408.GB15666@localhost> <20160506162823.GB7999@lst.de> <20160506163510.GA20754@localhost> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: <20160506163510.GA20754@localhost> List-ID: On Fri, May 06, 2016 at 11:35:10AM -0500, Bjorn Helgaas wrote: > I don't know what sort of clues you'll need for IRQ spreading, but > that sounds like a good reason for the argument. So if you will need > the flags argument for something else, then I think it's OK as-is. The basic idea is that we want to spread the MSI(-X) vectors around properly right when allocating them. See the series around this patch when I last posted it. For for now various drivers use different ways of spreading them, so we'll need to make it opt-in in the beginning. In the long run we'll hopefully have everyone converted to the common algorithm and can get rid of the flag, but that might take a while.