Xen devel - Sep 2008 - Megaraid SAS driver failing in Xen-3.3.0 but was working in Xen-3.2.2-rc3

On Xen-3.3.0, domain0 Megaraid SAS (SAS 1068 controller) driver is not loading
correctly if vtd support in Xen is enabled.
It fails at the point of initializing firmware.

I wasn''t seeing this error with Xen-3.2.2-rc3 (Unstable version),
though with vtd disabled in Xen-3.3.0, it is working.

Looks like a degrade problem.

Any clues?
Thx,
Venkat



_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Venkat,
 
Can you post the serial console output? 
 
Randy (Weidong)

________________________________

From: Kumar, Venkat [mailto:Venkat.Kumar@lsi.com] 
Sent: 2008年9月17日 20:38
To: xen-devel@lists.xensource.com
Cc: Han, Weidong; Cui, Dexuan
Subject: Megaraid SAS driver failing in Xen-3.3.0 but was working in
Xen-3.2.2-rc3



On Xen-3.3.0, domain0 Megaraid SAS (SAS 1068 controller) driver is not loading
correctly if vtd support in Xen is enabled.

It fails at the point of initializing firmware.

 

I wasn’t seeing this error with Xen-3.2.2-rc3 (Unstable version), though with
vtd disabled in Xen-3.3.0, it is working.

 

Looks like a degrade problem.

 

Any clues?

Thx,

Venkat

 



_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Hi Randy,

I am attaching the complete output in a file.

But for your reference, a part of it I am pasting in the body.
=================================================================================(XEN)
PCI add device 02:0e.0
(XEN) [VT-D]iommu.c:775: iommu_page_fault: iommu->reg = ffff828bfff55000
(XEN) [VT-D]iommu.c:744: iommu_fault_status: Fault Overflow
(XEN) [VT-D]iommu.c:729: iommu_fault:DMA Write: 2:e.0 addr cddf2000 REASON 2
iommu->reg = ffff828bfff55000
(XEN) print_vtd_entries: iommu = ffff8300cde0db00 bdf = 2:e:0 gmfn = cddf2
(XEN)     root_entry = ffff83012bfe5000
(XEN)     root_entry[2] = 12779e001
(XEN)     context = ffff83012779e000
(XEN)     context[70] = 0_0
(XEN)     ctxt_entry[70] not present
(XEN) PCI add device 00:1b.0
(XEN) PCI remove device 02:0e.0
=================================================================================
02:0e.0 is the pci id of Megaraid SAS controller.
For some reason you can see that the device is removed.
I guess this output would give you some idea.


Thx,
Venkat
________________________________
From: Han, Weidong [mailto:weidong.han@intel.com]
Sent: Thursday, September 18, 2008 8:01 AM
To: Kumar, Venkat; xen-devel@lists.xensource.com
Cc: Cui, Dexuan
Subject: RE: Megaraid SAS driver failing in Xen-3.3.0 but was working in
Xen-3.2.2-rc3

Venkat,

Can you post the serial console output?

Randy (Weidong)

________________________________
From: Kumar, Venkat [mailto:Venkat.Kumar@lsi.com]
Sent: 2008年9月17日 20:38
To: xen-devel@lists.xensource.com
Cc: Han, Weidong; Cui, Dexuan
Subject: Megaraid SAS driver failing in Xen-3.3.0 but was working in
Xen-3.2.2-rc3
On Xen-3.3.0, domain0 Megaraid SAS (SAS 1068 controller) driver is not loading
correctly if vtd support in Xen is enabled.
It fails at the point of initializing firmware.

I wasn’t seeing this error with Xen-3.2.2-rc3 (Unstable version), though with
vtd disabled in Xen-3.3.0, it is working.

Looks like a degrade problem.

Any clues?
Thx,
Venkat




_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

per this log, we can see the context mapping is not created for 02:0e.0.
 
I checked the code, I suspect context mapping() for PCI device is incorrect (see
case DEV_TYPE_PCI block code in domain_context_mapping()). Obviously the PCI
device is not mapped. Espen changed this code in c/s 17972. Espen, can you give
some explanation?
 
Venkat, pls try attached patch. and pls post your "lspci -v" output.
 
Randy (Weidong)


________________________________

From: Kumar, Venkat [mailto:Venkat.Kumar@lsi.com] 
Sent: 2008年9月18日 12:45
To: Han, Weidong
Cc: Cui, Dexuan; xen-devel@lists.xensource.com
Subject: RE: Megaraid SAS driver failing in Xen-3.3.0 but was working in
Xen-3.2.2-rc3



 

Hi Randy,

 

I am attaching the complete output in a file.

 

But for your reference, a part of it I am pasting in the body.

=================================================================================
(XEN) PCI add device 02:0e.0

(XEN) [VT-D]iommu.c:775: iommu_page_fault: iommu->reg = ffff828bfff55000

(XEN) [VT-D]iommu.c:744: iommu_fault_status: Fault Overflow

(XEN) [VT-D]iommu.c:729: iommu_fault:DMA Write: 2:e.0 addr cddf2000 REASON 2
iommu->reg = ffff828bfff55000

(XEN) print_vtd_entries: iommu = ffff8300cde0db00 bdf = 2:e:0 gmfn = cddf2

(XEN)     root_entry = ffff83012bfe5000

(XEN)     root_entry[2] = 12779e001

(XEN)     context = ffff83012779e000

(XEN)     context[70] = 0_0

(XEN)     ctxt_entry[70] not present

(XEN) PCI add device 00:1b.0

(XEN) PCI remove device 02:0e.0

=================================================================================
 

02:0e.0 is the pci id of Megaraid SAS controller.

For some reason you can see that the device is removed.

I guess this output would give you some idea.

 

 

Thx,

Venkat

________________________________

From: Han, Weidong [mailto:weidong.han@intel.com] 
Sent: Thursday, September 18, 2008 8:01 AM
To: Kumar, Venkat; xen-devel@lists.xensource.com
Cc: Cui, Dexuan
Subject: RE: Megaraid SAS driver failing in Xen-3.3.0 but was working in
Xen-3.2.2-rc3

 

Venkat,

 

Can you post the serial console output? 

 

Randy (Weidong)

 

________________________________

From: Kumar, Venkat [mailto:Venkat.Kumar@lsi.com] 
Sent: 2008年9月17日 20:38
To: xen-devel@lists.xensource.com
Cc: Han, Weidong; Cui, Dexuan
Subject: Megaraid SAS driver failing in Xen-3.3.0 but was working in
Xen-3.2.2-rc3

On Xen-3.3.0, domain0 Megaraid SAS (SAS 1068 controller) driver is not loading
correctly if vtd support in Xen is enabled.

It fails at the point of initializing firmware.

 

I wasn’t seeing this error with Xen-3.2.2-rc3 (Unstable version), though with
vtd disabled in Xen-3.3.0, it is working.

 

Looks like a degrade problem.

 

Any clues?

Thx,

Venkat

 




_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Hi the patch works and we are able to insert the module in the kernel.
Subsequently we’re dedicating the MgaRAID SAS card to an HVM guest as well but
once that is done we’re unable to see the volumes i.e. the disk is not showing
up through fdisk -l in the HVM guest. The error received on the dmesg of HVM is
that the firmware doesn’t initialize: -

megasas: 00.00.03.01 Sun May 14 22:49:52 PDT 2006
megasas: 0x1000:0x0411:0x1000:0x1008: bus 0:slot 6:func 0
PCI: Enabling device 0000:00:06.0 (0080 -> 0082)
ACPI: PCI Interrupt 0000:00:06.0[A] -> GSI 40 (level, low) -> IRQ 169
megasas: Waiting for FW to come to ready state
megasas: Failed to init firmware
ACPI: PCI interrupt for device 0000:00:06.0 disabled

This Megasas card however worked fine on an HVM on a dell dq35 machine. Does
anyone face the same issue?
Bhaskar.
________________________________
From: Han, Weidong [mailto:weidong.han@intel.com]
Sent: Thursday, September 18, 2008 12:02 PM
To: Kumar, Venkat
Cc: Cui, Dexuan; xen-devel@lists.xensource.com; espen.skoglund@netronome.com
Subject: RE: Megaraid SAS driver failing in Xen-3.3.0 but was working in
Xen-3.2.2-rc3

per this log, we can see the context mapping is not created for 02:0e.0.

I checked the code, I suspect context mapping() for PCI device is incorrect (see
case DEV_TYPE_PCI block code in domain_context_mapping()). Obviously the PCI
device is not mapped. Espen changed this code in c/s 17972. Espen, can you give
some explanation?

Venkat, pls try attached patch. and pls post your "lspci -v" output.

Randy (Weidong)

________________________________
From: Kumar, Venkat [mailto:Venkat.Kumar@lsi.com]
Sent: 2008年9月18日 12:45
To: Han, Weidong
Cc: Cui, Dexuan; xen-devel@lists.xensource.com
Subject: RE: Megaraid SAS driver failing in Xen-3.3.0 but was working in
Xen-3.2.2-rc3

Hi Randy,

I am attaching the complete output in a file.

But for your reference, a part of it I am pasting in the body.
=================================================================================(XEN)
PCI add device 02:0e.0
(XEN) [VT-D]iommu.c:775: iommu_page_fault: iommu->reg = ffff828bfff55000
(XEN) [VT-D]iommu.c:744: iommu_fault_status: Fault Overflow
(XEN) [VT-D]iommu.c:729: iommu_fault:DMA Write: 2:e.0 addr cddf2000 REASON 2
iommu->reg = ffff828bfff55000
(XEN) print_vtd_entries: iommu = ffff8300cde0db00 bdf = 2:e:0 gmfn = cddf2
(XEN)     root_entry = ffff83012bfe5000
(XEN)     root_entry[2] = 12779e001
(XEN)     context = ffff83012779e000
(XEN)     context[70] = 0_0
(XEN)     ctxt_entry[70] not present
(XEN) PCI add device 00:1b.0
(XEN) PCI remove device 02:0e.0
=================================================================================
02:0e.0 is the pci id of Megaraid SAS controller.
For some reason you can see that the device is removed.
I guess this output would give you some idea.


Thx,
Venkat
________________________________
From: Han, Weidong [mailto:weidong.han@intel.com]
Sent: Thursday, September 18, 2008 8:01 AM
To: Kumar, Venkat; xen-devel@lists.xensource.com
Cc: Cui, Dexuan
Subject: RE: Megaraid SAS driver failing in Xen-3.3.0 but was working in
Xen-3.2.2-rc3

Venkat,

Can you post the serial console output?

Randy (Weidong)

________________________________
From: Kumar, Venkat [mailto:Venkat.Kumar@lsi.com]
Sent: 2008年9月17日 20:38
To: xen-devel@lists.xensource.com
Cc: Han, Weidong; Cui, Dexuan
Subject: Megaraid SAS driver failing in Xen-3.3.0 but was working in
Xen-3.2.2-rc3
On Xen-3.3.0, domain0 Megaraid SAS (SAS 1068 controller) driver is not loading
correctly if vtd support in Xen is enabled.
It fails at the point of initializing firmware.

I wasn’t seeing this error with Xen-3.2.2-rc3 (Unstable version), though with
vtd disabled in Xen-3.3.0, it is working.

Looks like a degrade problem.

Any clues?
Thx,
Venkat



_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Hi, Bhanskar.

Could you try my patch?
My MegaRAID SAS card works well with this patch.

If you are using ioemu, please try c/s 18430.
http://xenbits.xensource.com/xen-unstable.hg?rev/1a785d213573

If you are using ioemu-romote, please try c/s 4761.
http://xenbits.xensource.com/ext/qemu-xen-unstable.hg?rev/5f415e92f4e5

Thanks,

--
Yuji Shimada

On Tue, 30 Sep 2008 20:54:04 +0800
"Jayaraman, Bhaskar" <Bhaskar.Jayaraman@lsi.com> wrote:
> Hi the patch works and we are able to insert the module in the kernel.
Subsequently we’re dedicating the MgaRAID SAS card to an HVM guest as well but
once that is done we’re unable to see the volumes i.e. the disk is not showing
up through fdisk -l in the HVM guest. The error received on the dmesg of HVM is
that the firmware doesn’t initialize: -
> 
> megasas: 00.00.03.01 Sun May 14 22:49:52 PDT 2006
> megasas: 0x1000:0x0411:0x1000:0x1008: bus 0:slot 6:func 0
> PCI: Enabling device 0000:00:06.0 (0080 -> 0082)
> ACPI: PCI Interrupt 0000:00:06.0[A] -> GSI 40 (level, low) -> IRQ 169
> megasas: Waiting for FW to come to ready state
> megasas: Failed to init firmware
> ACPI: PCI interrupt for device 0000:00:06.0 disabled
> 
> This Megasas card however worked fine on an HVM on a dell dq35 machine.
Does anyone face the same issue?
> Bhaskar.
> ________________________________
> From: Han, Weidong [mailto:weidong.han@intel.com]
> Sent: Thursday, September 18, 2008 12:02 PM
> To: Kumar, Venkat
> Cc: Cui, Dexuan; xen-devel@lists.xensource.com;
espen.skoglund@netronome.com
> Subject: RE: Megaraid SAS driver failing in Xen-3.3.0 but was working in
Xen-3.2.2-rc3
> 
> per this log, we can see the context mapping is not created for 02:0e.0.
> 
> I checked the code, I suspect context mapping() for PCI device is incorrect
(see case DEV_TYPE_PCI block code in domain_context_mapping()). Obviously the
PCI device is not mapped. Espen changed this code in c/s 17972. Espen, can you
give some explanation?
> 
> Venkat, pls try attached patch. and pls post your "lspci -v"
output.
> 
> Randy (Weidong)
> 
> ________________________________
> From: Kumar, Venkat [mailto:Venkat.Kumar@lsi.com]
> Sent: 2008年9月18日 12:45
> To: Han, Weidong
> Cc: Cui, Dexuan; xen-devel@lists.xensource.com
> Subject: RE: Megaraid SAS driver failing in Xen-3.3.0 but was working in
Xen-3.2.2-rc3
> 
> Hi Randy,
> 
> I am attaching the complete output in a file.
> 
> But for your reference, a part of it I am pasting in the body.
>
=================================================================================>
(XEN) PCI add device 02:0e.0
> (XEN) [VT-D]iommu.c:775: iommu_page_fault: iommu->reg = ffff828bfff55000
> (XEN) [VT-D]iommu.c:744: iommu_fault_status: Fault Overflow
> (XEN) [VT-D]iommu.c:729: iommu_fault:DMA Write: 2:e.0 addr cddf2000 REASON
2 iommu->reg = ffff828bfff55000
> (XEN) print_vtd_entries: iommu = ffff8300cde0db00 bdf = 2:e:0 gmfn = cddf2
> (XEN)     root_entry = ffff83012bfe5000
> (XEN)     root_entry[2] = 12779e001
> (XEN)     context = ffff83012779e000
> (XEN)     context[70] = 0_0
> (XEN)     ctxt_entry[70] not present
> (XEN) PCI add device 00:1b.0
> (XEN) PCI remove device 02:0e.0
>
=================================================================================>
> 02:0e.0 is the pci id of Megaraid SAS controller.
> For some reason you can see that the device is removed.
> I guess this output would give you some idea.
> 
> 
> Thx,
> Venkat
> ________________________________
> From: Han, Weidong [mailto:weidong.han@intel.com]
> Sent: Thursday, September 18, 2008 8:01 AM
> To: Kumar, Venkat; xen-devel@lists.xensource.com
> Cc: Cui, Dexuan
> Subject: RE: Megaraid SAS driver failing in Xen-3.3.0 but was working in
Xen-3.2.2-rc3
> 
> Venkat,
> 
> Can you post the serial console output?
> 
> Randy (Weidong)
> 
> ________________________________
> From: Kumar, Venkat [mailto:Venkat.Kumar@lsi.com]
> Sent: 2008年9月17日 20:38
> To: xen-devel@lists.xensource.com
> Cc: Han, Weidong; Cui, Dexuan
> Subject: Megaraid SAS driver failing in Xen-3.3.0 but was working in
Xen-3.2.2-rc3
> On Xen-3.3.0, domain0 Megaraid SAS (SAS 1068 controller) driver is not
loading correctly if vtd support in Xen is enabled.
> It fails at the point of initializing firmware.
> 
> I wasn’t seeing this error with Xen-3.2.2-rc3 (Unstable version), though
with vtd disabled in Xen-3.3.0, it is working.
> 
> Looks like a degrade problem.
> 
> Any clues?
> Thx,
> Venkat
> 
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Is this the patch that you mean (pls. see below)?   I added this patch onto a
recent Xen 3.4-unstable and with CentOS 5.2 for dom0, and the VT-d passthru is
working.  I am able to access SATA drives attached to the onboard 1068
controller on the Dell T7400.  Access works for HVMs running either CentOS and
Windows.

I haven’t tried the MR SAS card tho.  I do also have the onboard Broadcom NIC
working with the VT-d passthru.

HTH,
Ed

Here is the patch I used:

diff -r 694b7daa353c xen/drivers/passthrough/vtd/iommu.c
--- a/xen/drivers/passthrough/vtd/iommu.c       Wed Sep 17 14:16:02 2008 +0100
+++ b/xen/drivers/passthrough/vtd/iommu.c       Thu Sep 18 14:11:09 2008 +0800
@@ -1265,6 +1265,7 @@ static int domain_context_mapping(struct
                  bus, PCI_SLOT(devfn), PCI_FUNC(devfn));

         ob = bus; odf = devfn;
+        ret = domain_context_mapping_one(domain, drhd->iommu, bus, devfn);
         if ( !find_pcie_endpoint(&bus, &devfn, &secbus) )
         {
             gdprintk(XENLOG_WARNING VTDPREFIX,


-----Original Message-----
From: xen-devel-bounces@lists.xensource.com
[mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Yuji Shimada
Sent: Wednesday, October 01, 2008 12:08 AM
To: Jayaraman, Bhaskar; weidong.han@intel.com; espen.skoglund@netronome.com;
xen-devel@lists.xensource.com
Subject: Re: [Xen-devel] RE: Megaraid SAS driver failing in Xen-3.3.0 but was
working in Xen-3.2.2-rc3

Hi, Bhanskar.

Could you try my patch?
My MegaRAID SAS card works well with this patch.

If you are using ioemu, please try c/s 18430.
http://xenbits.xensource.com/xen-unstable.hg?rev/1a785d213573

If you are using ioemu-romote, please try c/s 4761.
http://xenbits.xensource.com/ext/qemu-xen-unstable.hg?rev/5f415e92f4e5

Thanks,

--
Yuji Shimada

On Tue, 30 Sep 2008 20:54:04 +0800
"Jayaraman, Bhaskar" <Bhaskar.Jayaraman@lsi.com> wrote:
> Hi the patch works and we are able to insert the module in the kernel.
Subsequently we’re dedicating the MgaRAID SAS card to an HVM guest as well but
once that is done we’re unable to see the volumes i.e. the disk is not showing
up through fdisk -l in the HVM guest. The error received on the dmesg of HVM is
that the firmware doesn’t initialize: -
>
> megasas: 00.00.03.01 Sun May 14 22:49:52 PDT 2006
> megasas: 0x1000:0x0411:0x1000:0x1008: bus 0:slot 6:func 0
> PCI: Enabling device 0000:00:06.0 (0080 -> 0082)
> ACPI: PCI Interrupt 0000:00:06.0[A] -> GSI 40 (level, low) -> IRQ 169
> megasas: Waiting for FW to come to ready state
> megasas: Failed to init firmware
> ACPI: PCI interrupt for device 0000:00:06.0 disabled
>
> This Megasas card however worked fine on an HVM on a dell dq35 machine.
Does anyone face the same issue?
> Bhaskar.
> ________________________________
> From: Han, Weidong [mailto:weidong.han@intel.com]
> Sent: Thursday, September 18, 2008 12:02 PM
> To: Kumar, Venkat
> Cc: Cui, Dexuan; xen-devel@lists.xensource.com;
espen.skoglund@netronome.com
> Subject: RE: Megaraid SAS driver failing in Xen-3.3.0 but was working in
Xen-3.2.2-rc3
>
> per this log, we can see the context mapping is not created for 02:0e.0.
>
> I checked the code, I suspect context mapping() for PCI device is incorrect
(see case DEV_TYPE_PCI block code in domain_context_mapping()). Obviously the
PCI device is not mapped. Espen changed this code in c/s 17972. Espen, can you
give some explanation?
>
> Venkat, pls try attached patch. and pls post your "lspci -v"
output.
>
> Randy (Weidong)
>
> ________________________________
> From: Kumar, Venkat [mailto:Venkat.Kumar@lsi.com]
> Sent: 2008年9月18日 12:45
> To: Han, Weidong
> Cc: Cui, Dexuan; xen-devel@lists.xensource.com
> Subject: RE: Megaraid SAS driver failing in Xen-3.3.0 but was working in
Xen-3.2.2-rc3
>
> Hi Randy,
>
> I am attaching the complete output in a file.
>
> But for your reference, a part of it I am pasting in the body.
>
=================================================================================>
(XEN) PCI add device 02:0e.0
> (XEN) [VT-D]iommu.c:775: iommu_page_fault: iommu->reg = ffff828bfff55000
> (XEN) [VT-D]iommu.c:744: iommu_fault_status: Fault Overflow
> (XEN) [VT-D]iommu.c:729: iommu_fault:DMA Write: 2:e.0 addr cddf2000 REASON
2 iommu->reg = ffff828bfff55000
> (XEN) print_vtd_entries: iommu = ffff8300cde0db00 bdf = 2:e:0 gmfn = cddf2
> (XEN)     root_entry = ffff83012bfe5000
> (XEN)     root_entry[2] = 12779e001
> (XEN)     context = ffff83012779e000
> (XEN)     context[70] = 0_0
> (XEN)     ctxt_entry[70] not present
> (XEN) PCI add device 00:1b.0
> (XEN) PCI remove device 02:0e.0
>
=================================================================================>
> 02:0e.0 is the pci id of Megaraid SAS controller.
> For some reason you can see that the device is removed.
> I guess this output would give you some idea.
>
>
> Thx,
> Venkat
> ________________________________
> From: Han, Weidong [mailto:weidong.han@intel.com]
> Sent: Thursday, September 18, 2008 8:01 AM
> To: Kumar, Venkat; xen-devel@lists.xensource.com
> Cc: Cui, Dexuan
> Subject: RE: Megaraid SAS driver failing in Xen-3.3.0 but was working in
Xen-3.2.2-rc3
>
> Venkat,
>
> Can you post the serial console output?
>
> Randy (Weidong)
>
> ________________________________
> From: Kumar, Venkat [mailto:Venkat.Kumar@lsi.com]
> Sent: 2008年9月17日 20:38
> To: xen-devel@lists.xensource.com
> Cc: Han, Weidong; Cui, Dexuan
> Subject: Megaraid SAS driver failing in Xen-3.3.0 but was working in
Xen-3.2.2-rc3
> On Xen-3.3.0, domain0 Megaraid SAS (SAS 1068 controller) driver is not
loading correctly if vtd support in Xen is enabled.
> It fails at the point of initializing firmware.
>
> I wasn’t seeing this error with Xen-3.2.2-rc3 (Unstable version), though
with vtd disabled in Xen-3.3.0, it is working.
>
> Looks like a degrade problem.
>
> Any clues?
> Thx,
> Venkat
>
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Hi, Nadolski,

If you use resent Xen 3.4-unstable, you don''t need applying the patch.
Because the patch has already been applied in Xen 3.4-unstable on Sep 05th.

If not, e.g.Xen 3.3-stable, we can apply the patch to assign MegaRaid SAS.

Please get the patch from following URL:
  ioemu       : c/s 18430: 1a785d213573
    http://xenbits.xensource.com/xen-unstable.hg?rev/1a785d213573
  ioemu-remote: c/s  4761: 5f415e92f4e5
    http://xenbits.xensource.com/ext/qemu-xen-unstable.hg?rev/5f415e92f4e5

Thanks,

--
Yuji Shimada

Here is my patch that I mention.

ioemu:
--- a/tools/ioemu/hw/pt-msi.c	Thu Sep 04 16:29:38 2008 +0100
+++ b/tools/ioemu/hw/pt-msi.c	Fri Sep 05 11:18:20 2008 +0100
@@ -313,7 +313,7 @@ int pt_msix_init(struct pt_dev *dev, int
 
     table_off = pci_read_long(pd, pos + PCI_MSIX_TABLE);
     bar_index = dev->msix->bar_index = table_off & PCI_MSIX_BIR;
-    table_off &= table_off & ~PCI_MSIX_BIR;
+    table_off = dev->msix->table_off = table_off & ~PCI_MSIX_BIR;
     dev->msix->table_base = dev->pci_dev->base_addr[bar_index];
     PT_LOG("get MSI-X table bar base %llx\n",
            (unsigned long long)dev->msix->table_base);


ioemu-remote:
--- a/hw/pt-msi.c	Wed Sep 03 11:26:06 2008 +0100
+++ b/hw/pt-msi.c	Fri Sep 05 10:32:49 2008 +0100
@@ -313,7 +313,7 @@ int pt_msix_init(struct pt_dev *dev, int
 
     table_off = pci_read_long(pd, pos + PCI_MSIX_TABLE);
     bar_index = dev->msix->bar_index = table_off & PCI_MSIX_BIR;
-    table_off &= table_off & ~PCI_MSIX_BIR;
+    table_off = dev->msix->table_off = table_off & ~PCI_MSIX_BIR;
     dev->msix->table_base = dev->pci_dev->base_addr[bar_index];
     PT_LOG("get MSI-X table bar base %llx\n",
            (unsigned long long)dev->msix->table_base);


On Wed, 1 Oct 2008 12:52:25 -0600
"Nadolski, Ed" <Ed.Nadolski@lsi.com> wrote:
> Is this the patch that you mean (pls. see below)?   I added this patch onto
a recent Xen 3.4-unstable and with CentOS 5.2 for dom0, and the VT-d passthru is
working.  I am able to access SATA drives attached to the onboard 1068
controller on the Dell T7400.  Access works for HVMs running either CentOS and
Windows.
> 
> I haven’t tried the MR SAS card tho.  I do also have the onboard Broadcom
NIC working with the VT-d passthru.
> 
> HTH,
> Ed
> 
> Here is the patch I used:
> 
> diff -r 694b7daa353c xen/drivers/passthrough/vtd/iommu.c
> --- a/xen/drivers/passthrough/vtd/iommu.c       Wed Sep 17 14:16:02 2008
+0100
> +++ b/xen/drivers/passthrough/vtd/iommu.c       Thu Sep 18 14:11:09 2008
+0800
> @@ -1265,6 +1265,7 @@ static int domain_context_mapping(struct
>                   bus, PCI_SLOT(devfn), PCI_FUNC(devfn));
> 
>          ob = bus; odf = devfn;
> +        ret = domain_context_mapping_one(domain, drhd->iommu, bus,
devfn);
>          if ( !find_pcie_endpoint(&bus, &devfn, &secbus) )
>          {
>              gdprintk(XENLOG_WARNING VTDPREFIX,
> 
> 
> -----Original Message-----
> From: xen-devel-bounces@lists.xensource.com
[mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Yuji Shimada
> Sent: Wednesday, October 01, 2008 12:08 AM
> To: Jayaraman, Bhaskar; weidong.han@intel.com;
espen.skoglund@netronome.com; xen-devel@lists.xensource.com
> Subject: Re: [Xen-devel] RE: Megaraid SAS driver failing in Xen-3.3.0 but
was working in Xen-3.2.2-rc3
> 
> Hi, Bhanskar.
> 
> Could you try my patch?
> My MegaRAID SAS card works well with this patch.
> 
> If you are using ioemu, please try c/s 18430.
> http://xenbits.xensource.com/xen-unstable.hg?rev/1a785d213573
> 
> If you are using ioemu-romote, please try c/s 4761.
> http://xenbits.xensource.com/ext/qemu-xen-unstable.hg?rev/5f415e92f4e5
> 
> Thanks,
> 
> --
> Yuji Shimada
> 
> On Tue, 30 Sep 2008 20:54:04 +0800
> "Jayaraman, Bhaskar" <Bhaskar.Jayaraman@lsi.com> wrote:
> 
> > Hi the patch works and we are able to insert the module in the kernel.
Subsequently we’re dedicating the MgaRAID SAS card to an HVM guest as well but
once that is done we’re unable to see the volumes i.e. the disk is not showing
up through fdisk -l in the HVM guest. The error received on the dmesg of HVM is
that the firmware doesn’t initialize: -
> >
> > megasas: 00.00.03.01 Sun May 14 22:49:52 PDT 2006
> > megasas: 0x1000:0x0411:0x1000:0x1008: bus 0:slot 6:func 0
> > PCI: Enabling device 0000:00:06.0 (0080 -> 0082)
> > ACPI: PCI Interrupt 0000:00:06.0[A] -> GSI 40 (level, low) ->
IRQ 169
> > megasas: Waiting for FW to come to ready state
> > megasas: Failed to init firmware
> > ACPI: PCI interrupt for device 0000:00:06.0 disabled
> >
> > This Megasas card however worked fine on an HVM on a dell dq35
machine. Does anyone face the same issue?
> > Bhaskar.
> > ________________________________
> > From: Han, Weidong [mailto:weidong.han@intel.com]
> > Sent: Thursday, September 18, 2008 12:02 PM
> > To: Kumar, Venkat
> > Cc: Cui, Dexuan; xen-devel@lists.xensource.com;
espen.skoglund@netronome.com
> > Subject: RE: Megaraid SAS driver failing in Xen-3.3.0 but was working
in Xen-3.2.2-rc3
> >
> > per this log, we can see the context mapping is not created for
02:0e.0.
> >
> > I checked the code, I suspect context mapping() for PCI device is
incorrect (see case DEV_TYPE_PCI block code in domain_context_mapping()).
Obviously the PCI device is not mapped. Espen changed this code in c/s 17972.
Espen, can you give some explanation?
> >
> > Venkat, pls try attached patch. and pls post your "lspci -v"
output.
> >
> > Randy (Weidong)
> >
> > ________________________________
> > From: Kumar, Venkat [mailto:Venkat.Kumar@lsi.com]
> > Sent: 2008年9月18日 12:45
> > To: Han, Weidong
> > Cc: Cui, Dexuan; xen-devel@lists.xensource.com
> > Subject: RE: Megaraid SAS driver failing in Xen-3.3.0 but was working
in Xen-3.2.2-rc3
> >
> > Hi Randy,
> >
> > I am attaching the complete output in a file.
> >
> > But for your reference, a part of it I am pasting in the body.
> >
=================================================================================>
> (XEN) PCI add device 02:0e.0
> > (XEN) [VT-D]iommu.c:775: iommu_page_fault: iommu->reg =
ffff828bfff55000
> > (XEN) [VT-D]iommu.c:744: iommu_fault_status: Fault Overflow
> > (XEN) [VT-D]iommu.c:729: iommu_fault:DMA Write: 2:e.0 addr cddf2000
REASON 2 iommu->reg = ffff828bfff55000
> > (XEN) print_vtd_entries: iommu = ffff8300cde0db00 bdf = 2:e:0 gmfn =
cddf2
> > (XEN)     root_entry = ffff83012bfe5000
> > (XEN)     root_entry[2] = 12779e001
> > (XEN)     context = ffff83012779e000
> > (XEN)     context[70] = 0_0
> > (XEN)     ctxt_entry[70] not present
> > (XEN) PCI add device 00:1b.0
> > (XEN) PCI remove device 02:0e.0
> >
=================================================================================>
>
> > 02:0e.0 is the pci id of Megaraid SAS controller.
> > For some reason you can see that the device is removed.
> > I guess this output would give you some idea.
> >
> >
> > Thx,
> > Venkat
> > ________________________________
> > From: Han, Weidong [mailto:weidong.han@intel.com]
> > Sent: Thursday, September 18, 2008 8:01 AM
> > To: Kumar, Venkat; xen-devel@lists.xensource.com
> > Cc: Cui, Dexuan
> > Subject: RE: Megaraid SAS driver failing in Xen-3.3.0 but was working
in Xen-3.2.2-rc3
> >
> > Venkat,
> >
> > Can you post the serial console output?
> >
> > Randy (Weidong)
> >
> > ________________________________
> > From: Kumar, Venkat [mailto:Venkat.Kumar@lsi.com]
> > Sent: 2008年9月17日 20:38
> > To: xen-devel@lists.xensource.com
> > Cc: Han, Weidong; Cui, Dexuan
> > Subject: Megaraid SAS driver failing in Xen-3.3.0 but was working in
Xen-3.2.2-rc3
> > On Xen-3.3.0, domain0 Megaraid SAS (SAS 1068 controller) driver is not
loading correctly if vtd support in Xen is enabled.
> > It fails at the point of initializing firmware.
> >
> > I wasn’t seeing this error with Xen-3.2.2-rc3 (Unstable version),
though with vtd disabled in Xen-3.3.0, it is working.
> >
> > Looks like a degrade problem.
> >
> > Any clues?
> > Thx,
> > Venkat
> >
> 
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xensource.com
> http://lists.xensource.com/xen-devel

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel