From mboxrd@z Thu Jan 1 00:00:00 1970 From: Stephen Hemminger Subject: Re: What to do about UIO breakage in 2.0 Date: Tue, 28 Apr 2015 09:33:24 -0700 Message-ID: <20150428093324.441d0015@urahara> References: <20150427150624.68ef88bc@urahara> <533710CFB86FA344BFBF2D6802E602860466909F@SHSMSX101.ccr.corp.intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Cc: "dev-VfR2kkLFssw@public.gmane.org" To: "Qiu, Michael" Return-path: In-Reply-To: <533710CFB86FA344BFBF2D6802E602860466909F-0J0gbvR4kThpB2pF5aRoyrfspsVTdybXVpNB7YpNyf8@public.gmane.org> List-Id: patches and discussions about DPDK List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dev-bounces-VfR2kkLFssw@public.gmane.org Sender: "dev" On Tue, 28 Apr 2015 08:07:18 +0000 "Qiu, Michael" wrote: > On 4/28/2015 6:06 AM, Stephen Hemminger wrote: > > I raised the issue, but people seem to be ignoring that fact that igb_uio > > was broken by the introduction of UIO PCI generic in 2.0. > > What do you mean about igb_uio broken? > > Thanks, > Michael > > > > There are three options: > > 1. Remove IGB_UIO only use UIO PCI generic. > > Downside there is no MSI-X support for UIO PCI generic. > > 2. Revert UIO PCI generic support > > 3. Replace both of the above with something better. > > > > I am working on #3 but it will not be ready for 2.0.1 and there > > is no solution for users of 2.0 and any future stable code. > > > If you use link state interrupt, then the base code in eal_interrupts is now broken when using igb_uio. The code was changed to work with uio_pci_generic. The new code does read/write of PCI INTX config and will not work when using igb_uio which uses MSI-X. Some possible solutions are: 1. Remove igb_uio completely 2. Revert/remove uio_pci_generic and use old read/write to enable irq 3. Make code look at driver type (sending patches for that now). 4. Implement a better UIO method for MSI-X which is what I will submit when Rx IRQ support is ready.