Re: [Xen-devel] Regression PCI passthrough from 4.5.5 to 4.6.0-rc1

On Thu, Nov 02, 2017 at 05:11:17PM +, Roger Pau Monné wrote: > On Thu, Nov 02, 2017 at 07:02:49PM +0200, Pasi Kärkkäinen wrote: > > On Thu, Aug 24, 2017 at 09:23:16AM +0100, Roger Pau Monné wrote: > > > On Wed, Aug 23, 2017 at 07:13:00PM +0200, Andreas Kinzler wrote: > > > > > > > From a brief

Re: [Xen-devel] Regression PCI passthrough from 4.5.5 to 4.6.0-rc1

On Thu, Nov 02, 2017 at 07:02:49PM +0200, Pasi Kärkkäinen wrote: > On Thu, Aug 24, 2017 at 09:23:16AM +0100, Roger Pau Monné wrote: > > On Wed, Aug 23, 2017 at 07:13:00PM +0200, Andreas Kinzler wrote: > > > > > > From a brief look it looks like this would be doable, but the way > > > > > these

Re: [Xen-devel] Regression PCI passthrough from 4.5.5 to 4.6.0-rc1

On Thu, Aug 24, 2017 at 09:23:16AM +0100, Roger Pau Monné wrote: > On Wed, Aug 23, 2017 at 07:13:00PM +0200, Andreas Kinzler wrote: > > > > > From a brief look it looks like this would be doable, but the way > > > > these flags are being communicated is rather ugly (the values used > > > > here >

Re: [Xen-devel] Regression PCI passthrough from 4.5.5 to 4.6.0-rc1

On Wed, Aug 23, 2017 at 07:13:00PM +0200, Andreas Kinzler wrote: > > > > From a brief look it looks like this would be doable, but the way > > > these flags are being communicated is rather ugly (the values used > > > here > > > > aren't part of the public interface, and hence it wasn't

Re: [Xen-devel] Regression PCI passthrough from 4.5.5 to 4.6.0-rc1

> From a brief look it looks like this would be doable, but the way these flags are being communicated is rather ugly (the values used here > aren't part of the public interface, and hence it wasn't immediately > clear whether using one of the unused bits would be an option, but > it looks like

Re: [Xen-devel] Regression PCI passthrough from 4.5.5 to 4.6.0-rc1

On Mon, Aug 21, 2017 at 04:18:56PM +0100, Roger Pau Monné wrote: > On Mon, Aug 21, 2017 at 09:14:45AM -0600, Jan Beulich wrote: > > >>> On 21.08.17 at 16:49, wrote: > > > Another option is to (ab)use the msi.gflags field to add another flag > > > in order to signal Xen

Re: [Xen-devel] Regression PCI passthrough from 4.5.5 to 4.6.0-rc1

On Mon, Aug 21, 2017 at 09:14:45AM -0600, Jan Beulich wrote: > >>> On 21.08.17 at 16:49, wrote: > > Another option is to (ab)use the msi.gflags field to add another flag > > in order to signal Xen whether the interrupt should be unmasked. This > > is in line with what you

Re: [Xen-devel] Regression PCI passthrough from 4.5.5 to 4.6.0-rc1

>>> On 21.08.17 at 16:49, wrote: > Another option is to (ab)use the msi.gflags field to add another flag > in order to signal Xen whether the interrupt should be unmasked. This > is in line with what you suggest below. From a brief look it looks like this would be doable,

Re: [Xen-devel] Regression PCI passthrough from 4.5.5 to 4.6.0-rc1

On Mon, Aug 21, 2017 at 06:22:05AM -0600, Jan Beulich wrote: > >>> On 21.08.17 at 11:46, wrote: > > Xen never detects the MSIX table entry unmask because it happens > > before the MSIX is bound to the guest, so the Xen msixtbl handlers are > > not aware of this memory

Re: [Xen-devel] Regression PCI passthrough from 4.5.5 to 4.6.0-rc1

>>> On 21.08.17 at 14:32, wrote: >> > - Make QEMU setup the vectors when the table entries are unmasked, >>>even if MSIX is not enabled. >>> - Provide an hypercall so QEMU can unmask MSIX vectors on behalf of >>>the guest. This would be used to unmask the entries if MSIX

Re: [Xen-devel] Regression PCI passthrough from 4.5.5 to 4.6.0-rc1

- Make QEMU setup the vectors when the table entries are unmasked, even if MSIX is not enabled. - Provide an hypercall so QEMU can unmask MSIX vectors on behalf of the guest. This would be used to unmask the entries if MSIX is enabled with table entries already unmasked. Neither

Re: [Xen-devel] Regression PCI passthrough from 4.5.5 to 4.6.0-rc1

>>> On 21.08.17 at 11:46, wrote: > Xen never detects the MSIX table entry unmask because it happens > before the MSIX is bound to the guest, so the Xen msixtbl handlers are > not aware of this memory region. > > The two possible solutions I see to this are: > > - Make

Re: [Xen-devel] Regression PCI passthrough from 4.5.5 to 4.6.0-rc1

On Mon, Aug 21, 2017 at 11:09:01AM +0200, Andreas Kinzler wrote: > Hello Roger & Jan, > > > > > Could you please try the patch below and paste the output you get on > > > > the Xen console? > > > Output is in attached file. Does it help? > > Yes, thanks. I'm quite sure the problem is that MSIX

Re: [Xen-devel] Regression PCI passthrough from 4.5.5 to 4.6.0-rc1

On Thu, Aug 17, 2017 at 07:36:20PM +0200, Andreas Kinzler wrote: > On Tue, 15 Aug 2017 11:55:10 +0200, Roger Pau Monné > wrote: > > Could you please try the patch below and paste the output you get on > > the Xen console? > > Output is in attached file. Does it help? Yes,

Re: [Xen-devel] Regression PCI passthrough from 4.5.5 to 4.6.0-rc1

>>> On 17.08.17 at 19:36, wrote: > On Tue, 15 Aug 2017 11:55:10 +0200, Roger Pau Monné > wrote: >> Could you please try the patch below and paste the output you get on >> the Xen console? > > Output is in attached file. Does it help? For the moment I'll

Re: [Xen-devel] Regression PCI passthrough from 4.5.5 to 4.6.0-rc1

On Tue, 15 Aug 2017 11:55:10 +0200, Roger Pau Monné wrote: Could you please try the patch below and paste the output you get on the Xen console? Output is in attached file. Does it help? Regards Andreas(XEN) MSIX ctrl write. Enabled: 0 Maskall: 0. Configured entries:

Re: [Xen-devel] Regression PCI passthrough from 4.5.5 to 4.6.0-rc1

On Tue, Aug 15, 2017 at 06:12:32AM -0600, Jan Beulich wrote: > >>> On 15.08.17 at 13:00, wrote: > > On Tue, Aug 15, 2017 at 10:55:10AM +0100, Roger Pau Monné wrote: > >> On Mon, Aug 14, 2017 at 02:08:56PM +0200, Andreas Kinzler wrote: > >> > On Mon, 14 Aug 2017 13:56:58

Re: [Xen-devel] Regression PCI passthrough from 4.5.5 to 4.6.0-rc1

>>> On 15.08.17 at 13:00, wrote: > On Tue, Aug 15, 2017 at 10:55:10AM +0100, Roger Pau Monné wrote: >> On Mon, Aug 14, 2017 at 02:08:56PM +0200, Andreas Kinzler wrote: >> > On Mon, 14 Aug 2017 13:56:58 +0200, Roger Pau Monné >> > wrote: >> > > > > I

Re: [Xen-devel] Regression PCI passthrough from 4.5.5 to 4.6.0-rc1

On Tue, Aug 15, 2017 at 10:55:10AM +0100, Roger Pau Monné wrote: > On Mon, Aug 14, 2017 at 02:08:56PM +0200, Andreas Kinzler wrote: > > On Mon, 14 Aug 2017 13:56:58 +0200, Roger Pau Monné > > wrote: > > > > > I defined XEN_PT_LOGGING_ENABLED in xen_pt.h as requested without

Re: [Xen-devel] Regression PCI passthrough from 4.5.5 to 4.6.0-rc1

On Mon, Aug 14, 2017 at 02:08:56PM +0200, Andreas Kinzler wrote: > On Mon, 14 Aug 2017 13:56:58 +0200, Roger Pau Monné > wrote: > > > > I defined XEN_PT_LOGGING_ENABLED in xen_pt.h as requested without the > > > > "hack" patch. Log is attached. Does it help? > > > It tells

Re: [Xen-devel] Regression PCI passthrough from 4.5.5 to 4.6.0-rc1

On Mon, 14 Aug 2017 13:56:58 +0200, Roger Pau Monné wrote: > I defined XEN_PT_LOGGING_ENABLED in xen_pt.h as requested without the > "hack" patch. Log is attached. Does it help? It tells me that there's nothing unexpected on that side. As I think I had indicated before,

Re: [Xen-devel] Regression PCI passthrough from 4.5.5 to 4.6.0-rc1

On Mon, Aug 14, 2017 at 05:44:00AM -0600, Jan Beulich wrote: > >>> On 14.08.17 at 13:31, wrote: > > On Mon, 31 Jul 2017 12:12:45 +0200, Jan Beulich wrote: > > "Andreas Kinzler" 07/17/17 6:32 PM >>> > > Jan, I still have access to the

Re: [Xen-devel] Regression PCI passthrough from 4.5.5 to 4.6.0-rc1

>>> On 14.08.17 at 13:31, wrote: > On Mon, 31 Jul 2017 12:12:45 +0200, Jan Beulich wrote: > "Andreas Kinzler" 07/17/17 6:32 PM >>> > Jan, I still have access to the hardware so perhaps we can finally > solve > this problem.

Re: [Xen-devel] Regression PCI passthrough from 4.5.5 to 4.6.0-rc1

On Mon, 31 Jul 2017 12:12:45 +0200, Jan Beulich wrote: "Andreas Kinzler" 07/17/17 6:32 PM >>> Jan, I still have access to the hardware so perhaps we can finally solve this problem. Feel free to go ahead; I'll be on vacation for the next three weeks.

Re: [Xen-devel] Regression PCI passthrough from 4.5.5 to 4.6.0-rc1

>>> "Andreas Kinzler" 07/17/17 6:32 PM >>> >>> Jan, I still have access to the hardware so perhaps we can finally solve >>> this problem. >> Feel free to go ahead; I'll be on vacation for the next three weeks. > >Perhaps we can shortcut debugging a bit because I looked through the

Re: [Xen-devel] Regression PCI passthrough from 4.5.5 to 4.6.0-rc1

Hi, On Mon, Jul 17, 2017 at 06:32:42PM +0200, Andreas Kinzler wrote: > Hello Jan, Pasi, all > > >>Jan, I still have access to the hardware so perhaps we can finally solve > >>this problem. > >Feel free to go ahead; I'll be on vacation for the next three weeks. > > Perhaps we can shortcut

Re: [Xen-devel] Regression PCI passthrough from 4.5.5 to 4.6.0-rc1

Hello Jan, Pasi, all Jan, I still have access to the hardware so perhaps we can finally solve this problem. Feel free to go ahead; I'll be on vacation for the next three weeks. Perhaps we can shortcut debugging a bit because I looked through the patches of XenServer 7.2 and found the

Re: [Xen-devel] Regression PCI passthrough from 4.5.5 to 4.6.0-rc1

>>> On 10.07.17 at 18:56, wrote: > Hello Jan, Pasi, all > >>> I noticed that PCI passthrough for an LSI SAS HBA 9211 did not longer work > (at least under Windows) when using Xen 4.8.1. >>> I then bisected through various released versions and finally I narrowed it > down to

Re: [Xen-devel] Regression PCI passthrough from 4.5.5 to 4.6.0-rc1

On Fri, Jul 07, 2017 at 06:39:03PM +0200, Andreas Kinzler wrote: > Hello, > > I noticed that PCI passthrough for an LSI SAS HBA 9211 did not longer work > (at least under Windows) when using Xen 4.8.1. > I then bisected through various released versions and finally I narrowed it > down to > >

[Xen-devel] Regression PCI passthrough from 4.5.5 to 4.6.0-rc1

Hello, I noticed that PCI passthrough for an LSI SAS HBA 9211 did not longer work (at least under Windows) when using Xen 4.8.1. I then bisected through various released versions and finally I narrowed it down to 4.5.5 (with qemu from Xen 4.6.5) -> working 4.6.0-rc1 (with qemu from Xen 4.6.5)