From mboxrd@z Thu Jan 1 00:00:00 1970 From: Malcolm Crossley Subject: Re: VT-d interrup remapping errata workaround Date: Tue, 9 Jul 2013 10:34:13 +0100 Message-ID: <51DBD915.6060901@citrix.com> References: <51DACCD702000078000E342C@nat28.tlf.novell.com> <20130708133356.GB12204@hmsreliant.think-freely.org> <51DADDFF02000078000E3511@nat28.tlf.novell.com> <20130708171133.GF12204@hmsreliant.think-freely.org> <51DBCC7E02000078000E3770@nat28.tlf.novell.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii"; Format="flowed" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <51DBCC7E02000078000E3770@nat28.tlf.novell.com> List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xen.org Errors-To: xen-devel-bounces@lists.xen.org To: Jan Beulich Cc: Andrew Cooper , Joerg Roedel , Neil Horman , xen-devel List-Id: xen-devel@lists.xenproject.org On 09/07/13 07:40, Jan Beulich wrote: >>>> On 08.07.13 at 19:11, Neil Horman wrote: >> Thank you I'll see if I can find an online version to point others to, but >> regardless, errata 61 seems to be the kicker here. Looks like we need to add >> and new pci id to the quirk, and check for some additional revisions if that one >> is seen. I'll get to it shortly. Thanks for sending this out. Lets hope this >> problem doesn't exist in even more chipsets. > Erratum 61? That doesn't require remapping to be disabled as > workaround. 62 and 69 do. > > But you make a very good point regarding revisions: 62 also > applies to rev 0x12, and 69 applies to both 0x12 and 0x22. > Xen doesn't check for either of them right now. Malcolm - this > in turn means that keying off of device ID 0x342e indeed isn't > correct/complete. Are you going to be able to submit a Xen > side fix relatively soon, or shall I take care of this? I didn't notice that errata 69 is not listed as fixed on rev 0x22 for the X58 chipset, sorry about that. Have we had confirmation from Intel about this particular errata affecting the X58? It seems strange that they both have the same revision ID but different fixes for errata. If we have got confirmation then I'll create a patch straight away, I'll try our Intel contacts as well, it might speed things up a bit. Malcolm > Jan >