On Fri, Aug 10, 2012 at 12:51:53PM -0600, Chris Friesen wrote:
> On 08/03/2012 02:33 AM, Lukas Hejtmanek wrote:
> >I also tried OFED package from Mellanox which seems to have better SR-IOV
> >support (at least mlx4_ib does not complain that SR-IOV is not supported).
> >However, it does not work
On 08/03/2012 02:33 AM, Lukas Hejtmanek wrote:
I also tried OFED package from Mellanox which seems to have better SR-IOV
support (at least mlx4_ib does not complain that SR-IOV is not supported).
However, it does not work when SR-IOV enabled:
Last I heard they were not officially providing
On 08/03/2012 02:33 AM, Lukas Hejtmanek wrote:
I also tried OFED package from Mellanox which seems to have better SR-IOV
support (at least mlx4_ib does not complain that SR-IOV is not supported).
However, it does not work when SR-IOV enabled:
Last I heard they were not officially providing
On Fri, Aug 10, 2012 at 12:51:53PM -0600, Chris Friesen wrote:
On 08/03/2012 02:33 AM, Lukas Hejtmanek wrote:
I also tried OFED package from Mellanox which seems to have better SR-IOV
support (at least mlx4_ib does not complain that SR-IOV is not supported).
However, it does not work when
On Mon, Aug 06, 2012 at 10:07:06AM -0400, Konrad Rzeszutek Wilk wrote:
> > good catch. I forgot to pass swiotl=force for DomU in Xen. So now, it seems
> > that mlx4_core works, mlx4_en (ethernet part) works as well. Unfortunately,
> > the IB part does not. IB layer complains that SR-IOV is
On Sun, Aug 05, 2012 at 10:05:00AM +0200, Lukas Hejtmanek wrote:
> Hi,
>
> On Fri, Aug 03, 2012 at 06:49:59AM -0700, Konrad Wilk wrote:
> > This looks like you are using PV PCI passthrough? If so, did you
> > remember to use 'iommu=soft' to enable the Xen-SWIOTLB in your guest?
> > And are you
On Sun, Aug 05, 2012 at 10:05:00AM +0200, Lukas Hejtmanek wrote:
Hi,
On Fri, Aug 03, 2012 at 06:49:59AM -0700, Konrad Wilk wrote:
This looks like you are using PV PCI passthrough? If so, did you
remember to use 'iommu=soft' to enable the Xen-SWIOTLB in your guest?
And are you booting
On Mon, Aug 06, 2012 at 10:07:06AM -0400, Konrad Rzeszutek Wilk wrote:
good catch. I forgot to pass swiotl=force for DomU in Xen. So now, it seems
that mlx4_core works, mlx4_en (ethernet part) works as well. Unfortunately,
the IB part does not. IB layer complains that SR-IOV is currently
Hi,
On Fri, Aug 03, 2012 at 06:49:59AM -0700, Konrad Wilk wrote:
> This looks like you are using PV PCI passthrough? If so, did you
> remember to use 'iommu=soft' to enable the Xen-SWIOTLB in your guest?
> And are you booting with more than 4GB? Or is less than 3GB (so that you have
> a nice gap
Hi,
On Fri, Aug 03, 2012 at 06:49:59AM -0700, Konrad Wilk wrote:
This looks like you are using PV PCI passthrough? If so, did you
remember to use 'iommu=soft' to enable the Xen-SWIOTLB in your guest?
And are you booting with more than 4GB? Or is less than 3GB (so that you have
a nice gap in
On Fri, Aug 3, 2012 at 1:33 AM, Lukas Hejtmanek wrote:
> On Wed, Aug 01, 2012 at 04:36:14PM -0700, Yinghai Lu wrote:
>> > so it seems, that pic=nocsr is a must now.
>>
>> yes. Or you have bios provide SRIOV support or 64 bit resource in _CRS.
>
> Well, I can use PCI passthrough in Xen now,
).
- Original Message -
From: xhejt...@ics.muni.cz
To: ying...@kernel.org
Cc: rol...@kernel.org, linux-kernel@vger.kernel.org, linux-...@vger.kernel.org
Sent: Friday, August 3, 2012 4:34:03 AM GMT -05:00 US/Canada Eastern
Subject: Re: mellanox mlx4_core and SR-IOV
On Wed, Aug 01, 2012 at 04:36:14PM -0700
On Wed, Aug 01, 2012 at 04:36:14PM -0700, Yinghai Lu wrote:
> > so it seems, that pic=nocsr is a must now.
>
> yes. Or you have bios provide SRIOV support or 64 bit resource in _CRS.
Well, I can use PCI passthrough in Xen now, however, it seems SR-IOV does not
work in case of Mellanox mlx4
On Wed, Aug 01, 2012 at 04:36:14PM -0700, Yinghai Lu wrote:
so it seems, that pic=nocsr is a must now.
yes. Or you have bios provide SRIOV support or 64 bit resource in _CRS.
Well, I can use PCI passthrough in Xen now, however, it seems SR-IOV does not
work in case of Mellanox mlx4 driver.
).
- Original Message -
From: xhejt...@ics.muni.cz
To: ying...@kernel.org
Cc: rol...@kernel.org, linux-kernel@vger.kernel.org, linux-...@vger.kernel.org
Sent: Friday, August 3, 2012 4:34:03 AM GMT -05:00 US/Canada Eastern
Subject: Re: mellanox mlx4_core and SR-IOV
On Wed, Aug 01, 2012 at 04:36:14PM -0700
On Fri, Aug 3, 2012 at 1:33 AM, Lukas Hejtmanek xhejt...@ics.muni.cz wrote:
On Wed, Aug 01, 2012 at 04:36:14PM -0700, Yinghai Lu wrote:
so it seems, that pic=nocsr is a must now.
yes. Or you have bios provide SRIOV support or 64 bit resource in _CRS.
Well, I can use PCI passthrough in Xen
On Wed, Aug 1, 2012 at 3:08 PM, Lukas Hejtmanek wrote:
> On Wed, Aug 01, 2012 at 02:32:17PM -0700, Yinghai Lu wrote:
>> yes, i knew that.
>>
>> one patch in my for-pci-next should address that.
>>
>>
On Wed, Aug 01, 2012 at 02:32:17PM -0700, Yinghai Lu wrote:
> yes, i knew that.
>
> one patch in my for-pci-next should address that.
>
> http://git.kernel.org/?p=linux/kernel/git/yinghai/linux-yinghai.git;a=patch;h=fcce563f868e296f46a2eeaa88d6959bcee26a2d
this is probably only half-way. well
On Wed, Aug 01, 2012 at 02:27:34PM -0700, Yinghai Lu wrote:
> you may try to boot with pci=nocrs
ok, pci=nocrs i got:
02:00.0 InfiniBand: Mellanox Technologies MT26428 [ConnectX VPI PCIe 2.0 5GT/s
- IB QDR / 10GigE] (rev b0)
02:00.1 InfiniBand: Mellanox Technologies MT25400 Family [ConnectX-2
On Wed, Aug 1, 2012 at 1:56 PM, Lukas Hejtmanek wrote:
> On Wed, Aug 01, 2012 at 11:29:02AM -0700, Yinghai Lu wrote:
>> iov bar is not assigned by BIOS, and kernel can not find range for it too.
>>
>> Lukas, can you post whole boot log with PCI_DEBUG enabled? That will
>> tell exact why kernel
On Wed, Aug 1, 2012 at 1:28 PM, Lukas Hejtmanek wrote:
> On Wed, Aug 01, 2012 at 11:29:02AM -0700, Yinghai Lu wrote:
>> On Wed, Aug 1, 2012 at 10:37 AM, Roland Dreier wrote:
>> > On Wed, Aug 1, 2012 at 6:38 AM, Lukas Hejtmanek
>> > wrote:
>> >> [3.558296] mlx4_core :02:00.0: not enough
On Wed, Aug 01, 2012 at 11:29:02AM -0700, Yinghai Lu wrote:
> iov bar is not assigned by BIOS, and kernel can not find range for it too.
>
> Lukas, can you post whole boot log with PCI_DEBUG enabled? That will
> tell exact why kernel does not assign them.
>
> Recent kernel from 3.4... should
On Wed, Aug 01, 2012 at 11:29:02AM -0700, Yinghai Lu wrote:
> On Wed, Aug 1, 2012 at 10:37 AM, Roland Dreier wrote:
> > On Wed, Aug 1, 2012 at 6:38 AM, Lukas Hejtmanek
> > wrote:
> >> [3.558296] mlx4_core :02:00.0: not enough MMIO resources for
> >> SR-IOV (nres: 0, iov->nres: 1)
> >
>
On Wed, Aug 1, 2012 at 10:37 AM, Roland Dreier wrote:
> On Wed, Aug 1, 2012 at 6:38 AM, Lukas Hejtmanek wrote:
>> [3.558296] mlx4_core :02:00.0: not enough MMIO resources for SR-IOV
>> (nres: 0, iov->nres: 1)
>
> This comes from the core sriov_enable() function, not anything in mlx4.
>
On Wed, Aug 1, 2012 at 6:38 AM, Lukas Hejtmanek wrote:
> [3.558296] mlx4_core :02:00.0: not enough MMIO resources for SR-IOV
> (nres: 0, iov->nres: 1)
This comes from the core sriov_enable() function, not anything in mlx4.
(although my kernel doesn't have the print of nres in that
Hello,
I tried to use SR-IOV virtualizaton for Mellanox ConnectX2 card with
mlx4_core driver with kernel 3.5.0. I built firware for the IB card with
sriov_en = true, lspci shows:
02:00.0 InfiniBand: Mellanox Technologies MT26428 [ConnectX VPI PCIe 2.0 5GT/s
- IB QDR / 10GigE] (rev b0)
Hello,
I tried to use SR-IOV virtualizaton for Mellanox ConnectX2 card with
mlx4_core driver with kernel 3.5.0. I built firware for the IB card with
sriov_en = true, lspci shows:
02:00.0 InfiniBand: Mellanox Technologies MT26428 [ConnectX VPI PCIe 2.0 5GT/s
- IB QDR / 10GigE] (rev b0)
On Wed, Aug 1, 2012 at 6:38 AM, Lukas Hejtmanek xhejt...@ics.muni.cz wrote:
[3.558296] mlx4_core :02:00.0: not enough MMIO resources for SR-IOV
(nres: 0, iov-nres: 1)
This comes from the core sriov_enable() function, not anything in mlx4.
(although my kernel doesn't have the print of
On Wed, Aug 1, 2012 at 10:37 AM, Roland Dreier rol...@kernel.org wrote:
On Wed, Aug 1, 2012 at 6:38 AM, Lukas Hejtmanek xhejt...@ics.muni.cz wrote:
[3.558296] mlx4_core :02:00.0: not enough MMIO resources for SR-IOV
(nres: 0, iov-nres: 1)
This comes from the core sriov_enable()
On Wed, Aug 01, 2012 at 11:29:02AM -0700, Yinghai Lu wrote:
On Wed, Aug 1, 2012 at 10:37 AM, Roland Dreier rol...@kernel.org wrote:
On Wed, Aug 1, 2012 at 6:38 AM, Lukas Hejtmanek xhejt...@ics.muni.cz
wrote:
[3.558296] mlx4_core :02:00.0: not enough MMIO resources for
SR-IOV
On Wed, Aug 01, 2012 at 11:29:02AM -0700, Yinghai Lu wrote:
iov bar is not assigned by BIOS, and kernel can not find range for it too.
Lukas, can you post whole boot log with PCI_DEBUG enabled? That will
tell exact why kernel does not assign them.
Recent kernel from 3.4... should enable
On Wed, Aug 1, 2012 at 1:28 PM, Lukas Hejtmanek xhejt...@ics.muni.cz wrote:
On Wed, Aug 01, 2012 at 11:29:02AM -0700, Yinghai Lu wrote:
On Wed, Aug 1, 2012 at 10:37 AM, Roland Dreier rol...@kernel.org wrote:
On Wed, Aug 1, 2012 at 6:38 AM, Lukas Hejtmanek xhejt...@ics.muni.cz
wrote:
[
On Wed, Aug 1, 2012 at 1:56 PM, Lukas Hejtmanek xhejt...@ics.muni.cz wrote:
On Wed, Aug 01, 2012 at 11:29:02AM -0700, Yinghai Lu wrote:
iov bar is not assigned by BIOS, and kernel can not find range for it too.
Lukas, can you post whole boot log with PCI_DEBUG enabled? That will
tell exact
On Wed, Aug 01, 2012 at 02:27:34PM -0700, Yinghai Lu wrote:
you may try to boot with pci=nocrs
ok, pci=nocrs i got:
02:00.0 InfiniBand: Mellanox Technologies MT26428 [ConnectX VPI PCIe 2.0 5GT/s
- IB QDR / 10GigE] (rev b0)
02:00.1 InfiniBand: Mellanox Technologies MT25400 Family [ConnectX-2
On Wed, Aug 01, 2012 at 02:32:17PM -0700, Yinghai Lu wrote:
yes, i knew that.
one patch in my for-pci-next should address that.
http://git.kernel.org/?p=linux/kernel/git/yinghai/linux-yinghai.git;a=patch;h=fcce563f868e296f46a2eeaa88d6959bcee26a2d
this is probably only half-way. well mlx
On Wed, Aug 1, 2012 at 3:08 PM, Lukas Hejtmanek xhejt...@ics.muni.cz wrote:
On Wed, Aug 01, 2012 at 02:32:17PM -0700, Yinghai Lu wrote:
yes, i knew that.
one patch in my for-pci-next should address that.
36 matches
Mail list logo
