Re: Test with VT-d patches

Re: Test with VT-d patches

[gg]

Hello,

I try to explain the current state on my pc.

1) The patch KVM 1/2 was applied with the following change.
   The files "drivers/pci/iova.h" and "drivers/pci/intel-iommu.h" already exists.
   And I get both files rejected.
   So I take the headers from the KVM 1/2 Patchfile and use these a "iova.h" and "intel-iommu.h".
   Kernel compiles without errors.

2) What known bug is in the latest userspace patchfile? (Told by Yang, Sheng)


>Hello,
>
>* On Wednesday 03 September 2008 14:07:36 gg@seh.de wrote:
>> Hi,
>>
>> i make some more tests with
>> 1) an old APCI1500/PCI card no linux driver support -> so i don't need to
>> unload the module. 2) my second network-card Realtek 10/100 MBit.
>>
>> Both don't work at all. Only my first gigabit-onboard-network-card starts.
>>
>> Here the output from userspace/dmesg:
>>
>>
>> 1) // Applied Micro Circuits Corp. APCI1500 Signal processing controller
>>
>> Warning: No DNS servers found
>> Registered host PCI device 03:00.0 ("03:00.0") as guest device 00:03.0
>> assigned_dev_update_irq: Input/output error
>> assigned_dev_update_irq: Input/output error
>
>This means the devices shares the irq with some other device in the system. 
>See the "lspci -v" output for details.
>
>> 2) // 03:02.0 Ethernet controller: Realtek Semiconductor Co., Ltd.
>> RTL-8139/8139C/8139C+ (rev 10) Registered host PCI device 03:02.0
>> ("03:02.0") as guest device 00:03.0 assigned_dev_update_irq: Input/output
>> error
>>
>> [  369.195971] pci 0000:03:02.0: PCI INT A -> GSI 18 (level, low) -> IRQ 18
>> [  369.326026] kvm_vm_ioctl_assign_irq: couldn't allocate irq for pv device
>
>Same error.
>
>> Sould i update the BIOS if possible/available?
>
>That's not necessary. We don't support assigning devices that share the irq on 
>
>the host with some other device. You can try inserting one device at a time 
>and try different PCI slots.
>
>> Could it be a problem with the already assigned patches? (See the first
>> E-Mail :The VTD [PATCH1/2] seems already be applied. )
>
>No patches have already been applied to any tree. You'll definitely have to 
>apply the 1/2 patch as well.

Ok. On both PCI-Cards (1) APCI1500 Signal processing controller and (2) Realtek NIC
the IRQ is shared with my USB-UHCI-Controller.
On this PC there is no PS2-Connector (USB-Mouse/USB-Keyboard). 
When I unload the usb module on the host, I can not handle the system anymore.

I can start qemu/kvm remote over a serial-console. But then I cannot use the guest system without keyboard/mouse. (login and so)

"Han, Weidong" told to use IRQF_SHARED in reqeust_irq(). In which code?

Both tests are only made to get an idea how it works with other devices.

My primary efforts are to use my onboard Gigabit with VT-d direct on the Windows guest-system. This device don't share the IRQ with an other device.
But I have the problems with:
1) slow down the complete system and generate this Unbalanced IRQ 21 messages.
2) Packet statistic under windows shows on this device no package traffic in live-network. (no ping etc. possible)

Thanks for your hints.

Gregor

RE: Test with VT-d patches

[Han, Weidong]

gg@seh.de wrote:
> Hello,
> 
> I try to explain the current state on my pc.
> 
> 1) The patch KVM 1/2 was applied with the following change.
>    The files "drivers/pci/iova.h" and "drivers/pci/intel-iommu.h"
>    already exists. And I get both files rejected.
>    So I take the headers from the KVM 1/2 Patchfile and use these a
>    "iova.h" and "intel-iommu.h". Kernel compiles without errors.
> 
> 2) What known bug is in the latest userspace patchfile? (Told by
> Yang, Sheng) 
> 

in add_assigned_device(), you can simply comment out some lines between
following #ifdef and #endif as follows.

+#ifdef KVM_CAP_IOMMU
//+	r = kvm_check_extension(kvm_context, KVM_CAP_IOMMU);
//+	if (r)
+		assigned_devices[nr_assigned_devices].dma |=
+			KVM_DEV_ASSIGN_ENABLE_IOMMU;
+
//+	r = get_param_value(dma, sizeof dma, "dma", arg);
//+	if (r && !strncmp(dma, "none", 4)) {
//+		assigned_devices[nr_assigned_devices].dma &=
//+			~KVM_DEV_ASSIGN_ENABLE_IOMMU;
+	}
+#endif

Randy (Weidong)

> 
>> Hello,
>> 
>> * On Wednesday 03 September 2008 14:07:36 gg@seh.de wrote:
>>> Hi,
>>> 
>>> i make some more tests with
>>> 1) an old APCI1500/PCI card no linux driver support -> so i don't
>>> need to unload the module. 2) my second network-card Realtek 10/100
>>> MBit. 
>>> 
>>> Both don't work at all. Only my first gigabit-onboard-network-card
>>> starts. 
>>> 
>>> Here the output from userspace/dmesg:
>>> 
>>> 
>>> 1) // Applied Micro Circuits Corp. APCI1500 Signal processing
>>> controller 
>>> 
>>> Warning: No DNS servers found
>>> Registered host PCI device 03:00.0 ("03:00.0") as guest device
>>> 00:03.0 assigned_dev_update_irq: Input/output error
>>> assigned_dev_update_irq: Input/output error
>> 
>> This means the devices shares the irq with some other device in the
>> system. See the "lspci -v" output for details.
>> 
>>> 2) // 03:02.0 Ethernet controller: Realtek Semiconductor Co., Ltd.
>>> RTL-8139/8139C/8139C+ (rev 10) Registered host PCI device 03:02.0
>>> ("03:02.0") as guest device 00:03.0 assigned_dev_update_irq:
>>> Input/output error 
>>> 
>>> [  369.195971] pci 0000:03:02.0: PCI INT A -> GSI 18 (level, low)
>>> -> IRQ 18 [  369.326026] kvm_vm_ioctl_assign_irq: couldn't allocate
>>> irq for pv device 
>> 
>> Same error.
>> 
>>> Sould i update the BIOS if possible/available?
>> 
>> That's not necessary. We don't support assigning devices that share
>> the irq on 
>> 
>> the host with some other device. You can try inserting one device at
>> a time and try different PCI slots. 
>> 
>>> Could it be a problem with the already assigned patches? (See the
>>> first E-Mail :The VTD [PATCH1/2] seems already be applied. )
>> 
>> No patches have already been applied to any tree. You'll definitely
>> have to apply the 1/2 patch as well.
> 
> Ok. On both PCI-Cards (1) APCI1500 Signal processing controller and
> (2) Realtek NIC 
> the IRQ is shared with my USB-UHCI-Controller.
> On this PC there is no PS2-Connector (USB-Mouse/USB-Keyboard).
> When I unload the usb module on the host, I can not handle the system
> anymore. 
> 
> I can start qemu/kvm remote over a serial-console. But then I cannot
> use the guest system without keyboard/mouse. (login and so) 
> 
> "Han, Weidong" told to use IRQF_SHARED in reqeust_irq(). In which
> code? 
> 
> Both tests are only made to get an idea how it works with other
> devices. 
> 
> My primary efforts are to use my onboard Gigabit with VT-d direct on
> the Windows guest-system. This device don't share the IRQ with an
> other device. But I have the problems with: 1) slow down the complete
> system and generate this Unbalanced IRQ 21 messages. 2) Packet
> statistic under windows shows on this device no package traffic in
> live-network. (no ping etc. possible)  
> 
> Thanks for your hints.
> 
> Gregor

Re: Test with VT-d patches

[Yang, Sheng]

On Thursday 04 September 2008 17:12:28 Han, Weidong wrote:
> gg@seh.de wrote:
> > Hello,
> >
> > I try to explain the current state on my pc.
> >
> > 1) The patch KVM 1/2 was applied with the following change.
> >    The files "drivers/pci/iova.h" and "drivers/pci/intel-iommu.h"
> >    already exists. And I get both files rejected.
> >    So I take the headers from the KVM 1/2 Patchfile and use these a
> >    "iova.h" and "intel-iommu.h". Kernel compiles without errors.
> >
> > 2) What known bug is in the latest userspace patchfile? (Told by
> > Yang, Sheng)
>

I think it's mentioned by Ben that userspace patch refer a kvm fd before it 
was initialized correctly. (patch at 2008-08-26, Hope I don't miss any update 
on this... ) 

> in add_assigned_device(), you can simply comment out some lines between
> following #ifdef and #endif as follows.
>
> +#ifdef KVM_CAP_IOMMU
> //+	r = kvm_check_extension(kvm_context, KVM_CAP_IOMMU);
> //+	if (r)
> +		assigned_devices[nr_assigned_devices].dma |=
> +			KVM_DEV_ASSIGN_ENABLE_IOMMU;
> +
> //+	r = get_param_value(dma, sizeof dma, "dma", arg);
> //+	if (r && !strncmp(dma, "none", 4)) {
> //+		assigned_devices[nr_assigned_devices].dma &=
> //+			~KVM_DEV_ASSIGN_ENABLE_IOMMU;
> +	}
> +#endif
>

Yeah, as this workaround.

> Randy (Weidong)
>
> >> Hello,
> >>
> >> * On Wednesday 03 September 2008 14:07:36 gg@seh.de wrote:
> >>> Hi,
> >>>
> >>> i make some more tests with
> >>> 1) an old APCI1500/PCI card no linux driver support -> so i don't
> >>> need to unload the module. 2) my second network-card Realtek 10/100
> >>> MBit.
> >>>
> >>> Both don't work at all. Only my first gigabit-onboard-network-card
> >>> starts.
> >>>
> >>> Here the output from userspace/dmesg:
> >>>
> >>>
> >>> 1) // Applied Micro Circuits Corp. APCI1500 Signal processing
> >>> controller
> >>>
> >>> Warning: No DNS servers found
> >>> Registered host PCI device 03:00.0 ("03:00.0") as guest device
> >>> 00:03.0 assigned_dev_update_irq: Input/output error
> >>> assigned_dev_update_irq: Input/output error
> >>
> >> This means the devices shares the irq with some other device in the
> >> system. See the "lspci -v" output for details.
> >>
> >>> 2) // 03:02.0 Ethernet controller: Realtek Semiconductor Co., Ltd.
> >>> RTL-8139/8139C/8139C+ (rev 10) Registered host PCI device 03:02.0
> >>> ("03:02.0") as guest device 00:03.0 assigned_dev_update_irq:
> >>> Input/output error
> >>>
> >>> [  369.195971] pci 0000:03:02.0: PCI INT A -> GSI 18 (level, low)
> >>> -> IRQ 18 [  369.326026] kvm_vm_ioctl_assign_irq: couldn't allocate
> >>> irq for pv device
> >>
> >> Same error.
> >>
> >>> Sould i update the BIOS if possible/available?
> >>
> >> That's not necessary. We don't support assigning devices that share
> >> the irq on
> >>
> >> the host with some other device. You can try inserting one device at
> >> a time and try different PCI slots.
> >>
> >>> Could it be a problem with the already assigned patches? (See the
> >>> first E-Mail :The VTD [PATCH1/2] seems already be applied. )
> >>
> >> No patches have already been applied to any tree. You'll definitely
> >> have to apply the 1/2 patch as well.
> >
> > Ok. On both PCI-Cards (1) APCI1500 Signal processing controller and
> > (2) Realtek NIC
> > the IRQ is shared with my USB-UHCI-Controller.
> > On this PC there is no PS2-Connector (USB-Mouse/USB-Keyboard).
> > When I unload the usb module on the host, I can not handle the system
> > anymore.

Well, another choice is use VNC from anyother machine. 

> >
> > I can start qemu/kvm remote over a serial-console. But then I cannot
> > use the guest system without keyboard/mouse. (login and so)
> >
> > "Han, Weidong" told to use IRQF_SHARED in reqeust_irq(). In which
> > code?

kvm_vm_ioctl_assign_irq() in x86.c. But for you are using keyboard and mouse, 
I think it would cause chaos? Maybe you can tell USB 2.0 controller apart 
from USB 1.1, and just unload 2.0 ones. The USB 1.1 is UHCI while 2.0 is 
EHCI, the drivers can be made as modules.

> >
> > Both tests are only made to get an idea how it works with other
> > devices.
> >
> > My primary efforts are to use my onboard Gigabit with VT-d direct on
> > the Windows guest-system. This device don't share the IRQ with an
> > other device. But I have the problems with: 1) slow down the complete
> > system and generate this Unbalanced IRQ 21 messages. 2) Packet
> > statistic under windows shows on this device no package traffic in
> > live-network. (no ping etc. possible)

No clue, if you ensure that you've got latest version of patch... I think it 
was posted by Amit at 2008-08-26.

BTW: Can you try a new Linux guest if possible? 

-- 
regards
Yang, Sheng

> >
> > Thanks for your hints.
> >
> > Gregor

Re: Test with VT-d patches

[gg]

Hello,

with the hint/patches from "Han, Weidong" I cound correct my system-build and now the NIC works fine under Windows.
Only the "Unbalanced enables for IRQ 21" (my pci-passed throug Gigabit NIC)commes nearly endless. (not importend for me)

I am happy now.
Thanks all,

Gregor


>On Thursday 04 September 2008 17:12:28 Han, Weidong wrote:
>> gg@seh.de wrote:
>> > Hello,
>> >
>> > I try to explain the current state on my pc.
>> >
>> > 1) The patch KVM 1/2 was applied with the following change.
>> >    The files "drivers/pci/iova.h" and "drivers/pci/intel-iommu.h"
>> >    already exists. And I get both files rejected.
>> >    So I take the headers from the KVM 1/2 Patchfile and use these a
>> >    "iova.h" and "intel-iommu.h". Kernel compiles without errors.
>> >
>> > 2) What known bug is in the latest userspace patchfile? (Told by
>> > Yang, Sheng)
>>
>
>I think it's mentioned by Ben that userspace patch refer a kvm fd before it 
>was initialized correctly. (patch at 2008-08-26, Hope I don't miss any update 
>
>on this... ) 
>
>> in add_assigned_device(), you can simply comment out some lines between
>> following #ifdef and #endif as follows.
>>
>> +#ifdef KVM_CAP_IOMMU
>> //+	r = kvm_check_extension(kvm_context, KVM_CAP_IOMMU);
>> //+	if (r)
>> +		assigned_devices[nr_assigned_devices].dma |=
>> +			KVM_DEV_ASSIGN_ENABLE_IOMMU;
>> +
>> //+	r = get_param_value(dma, sizeof dma, "dma", arg);
>> //+	if (r && !strncmp(dma, "none", 4)) {
>> //+		assigned_devices[nr_assigned_devices].dma &=
>> //+			~KVM_DEV_ASSIGN_ENABLE_IOMMU;
>> +	}
>> +#endif
>>
>
>Yeah, as this workaround.
>
>> Randy (Weidong)
>>
>> >> Hello,
>> >>
>> >> * On Wednesday 03 September 2008 14:07:36 gg@seh.de wrote:
>> >>> Hi,
>> >>>
>> >>> i make some more tests with
>> >>> 1) an old APCI1500/PCI card no linux driver support -> so i don't
>> >>> need to unload the module. 2) my second network-card Realtek 10/100
>> >>> MBit.
>> >>>
>> >>> Both don't work at all. Only my first gigabit-onboard-network-card
>> >>> starts.
>> >>>
>> >>> Here the output from userspace/dmesg:
>> >>>
>> >>>
>> >>> 1) // Applied Micro Circuits Corp. APCI1500 Signal processing
>> >>> controller
>> >>>
>> >>> Warning: No DNS servers found
>> >>> Registered host PCI device 03:00.0 ("03:00.0") as guest device
>> >>> 00:03.0 assigned_dev_update_irq: Input/output error
>> >>> assigned_dev_update_irq: Input/output error
>> >>
>> >> This means the devices shares the irq with some other device in the
>> >> system. See the "lspci -v" output for details.
>> >>
>> >>> 2) // 03:02.0 Ethernet controller: Realtek Semiconductor Co., Ltd.
>> >>> RTL-8139/8139C/8139C+ (rev 10) Registered host PCI device 03:02.0
>> >>> ("03:02.0") as guest device 00:03.0 assigned_dev_update_irq:
>> >>> Input/output error
>> >>>
>> >>> [  369.195971] pci 0000:03:02.0: PCI INT A -> GSI 18 (level, low)
>> >>> -> IRQ 18 [  369.326026] kvm_vm_ioctl_assign_irq: couldn't allocate
>> >>> irq for pv device
>> >>
>> >> Same error.
>> >>
>> >>> Sould i update the BIOS if possible/available?
>> >>
>> >> That's not necessary. We don't support assigning devices that share
>> >> the irq on
>> >>
>> >> the host with some other device. You can try inserting one device at
>> >> a time and try different PCI slots.
>> >>
>> >>> Could it be a problem with the already assigned patches? (See the
>> >>> first E-Mail :The VTD [PATCH1/2] seems already be applied. )
>> >>
>> >> No patches have already been applied to any tree. You'll definitely
>> >> have to apply the 1/2 patch as well.
>> >
>> > Ok. On both PCI-Cards (1) APCI1500 Signal processing controller and
>> > (2) Realtek NIC
>> > the IRQ is shared with my USB-UHCI-Controller.
>> > On this PC there is no PS2-Connector (USB-Mouse/USB-Keyboard).
>> > When I unload the usb module on the host, I can not handle the system
>> > anymore.
>
>Well, another choice is use VNC from anyother machine. 
>
>> >
>> > I can start qemu/kvm remote over a serial-console. But then I cannot
>> > use the guest system without keyboard/mouse. (login and so)
>> >
>> > "Han, Weidong" told to use IRQF_SHARED in reqeust_irq(). In which
>> > code?
>
>kvm_vm_ioctl_assign_irq() in x86.c. But for you are using keyboard and mouse, 
>
>I think it would cause chaos? Maybe you can tell USB 2.0 controller apart 
>from USB 1.1, and just unload 2.0 ones. The USB 1.1 is UHCI while 2.0 is 
>EHCI, the drivers can be made as modules.
>
>> >
>> > Both tests are only made to get an idea how it works with other
>> > devices.
>> >
>> > My primary efforts are to use my onboard Gigabit with VT-d direct on
>> > the Windows guest-system. This device don't share the IRQ with an
>> > other device. But I have the problems with: 1) slow down the complete
>> > system and generate this Unbalanced IRQ 21 messages. 2) Packet
>> > statistic under windows shows on this device no package traffic in
>> > live-network. (no ping etc. possible)
>
>No clue, if you ensure that you've got latest version of patch... I think it 
>was posted by Amit at 2008-08-26.
>
>BTW: Can you try a new Linux guest if possible? 
>
>-- 
>regards
>Yang, Sheng
>
>> >
>> > Thanks for your hints.
>> >
>> > Gregor

Re: Test with VT-d patches

[Yang, Sheng]

On Wednesday 03 September 2008 16:37:36 gg@seh.de wrote:
> Hi,
>
> i make some more tests with
> 1) an old APCI1500/PCI card no linux driver support -> so i don't need to
> unload the module. 2) my second network-card Realtek 10/100 MBit.
>
> Both don't work at all. Only my first gigabit-onboard-network-card starts.
>
> Here the output from userspace/dmesg:

Hi Gregor

I have a question here. Which version of patch you used? For in fact, Amit's 
latest userspace patch can't work due to a bug. So seems you didn't use 
Amit's latest userspace patch? What's about kernel part?

-- 
regards
Yang, Sheng
> 1) // Applied Micro Circuits Corp. APCI1500 Signal processing controller
>
> Warning: No DNS servers found
> Registered host PCI device 03:00.0 ("03:00.0") as guest device 00:03.0
> assigned_dev_update_irq: Input/output error
> assigned_dev_update_irq: Input/output error
> *** glibc detected *** qemu-system-x86_64: double free or corruption (out):
> 0x00000000013edfb0 ***
>
> ======= Backtrace: =========
> /lib/libc.so.6[0x7f18c6c609e8]
> /lib/libc.so.6(cfree+0x76)[0x7f18c6c63036]
> qemu-system-x86_64[0x470d0c]
> qemu-system-x86_64[0x418104]
> qemu-system-x86_64[0x469b62]
> qemu-system-x86_64[0x417fed]
> qemu-system-x86_64[0x4f937d]
> qemu-system-x86_64[0x5221ff]
> qemu-system-x86_64[0x522af3]
> qemu-system-x86_64[0x4f9906]
> qemu-system-x86_64[0x4f9c20]
> /lib/libpthread.so.0[0x7f18c76353ea]
> /lib/libc.so.6(clone+0x6d)[0x7f18c6ccdb9d]
> ======= Memory map: ========
> 00400000-00595000 r-xp 00000000 08:03 719950                            
> /usr/local/bin/qemu-system-x86_64 00794000-00795000 r--p 00194000 08:03
> 719950                             /usr/local/bin/qemu-system-x86_64
> 00795000-007a5000 rw-p 00195000 08:03 719950                            
> /usr/local/bin/qemu-system-x86_64 007a5000-00b9b000 rw-p 007a5000 00:00 0
> 00b9b000-00b9c000 rwxp 00b9b000 00:00 0
> 00b9c000-00bb7000 rw-p 00b9c000 00:00 0
> 012e7000-01613000 rw-p 012e7000 00:00 0                                 
> [heap] 40730000-48c31000 rwxp 40730000 00:00 0
> 48c31000-48c32000 ---p 48c31000 00:00 0
> 48c32000-49432000 rw-p 48c32000 00:00 0
> 49432000-49433000 ---p 49432000 00:00 0
> 49433000-49436000 rw-p 49433000 00:00 0
> 7f1894000000-7f1894021000 rw-p 7f1894000000 00:00 0
> 7f1894021000-7f1898000000 ---p 7f1894021000 00:00 0
> 7f1899ae0000-7f1899af6000 r-xp 00000000 08:03 908548                    
> /lib/libgcc_s.so.1 7f1899af6000-7f1899cf5000 ---p 00016000 08:03 908548    
>                 /lib/libgcc_s.so.1 7f1899cf5000-7f1899cf6000 r--p 00015000
> 08:03 908548                     /lib/libgcc_s.so.1
> 7f1899cf6000-7f1899cf7000 rw-p 00016000 08:03 908548                    
> /lib/libgcc_s.so.1 7f1899d04000-7f1899e1e000 rw-s 00000000 00:08 589836    
>                 /SYSV00000000 (deleted) 7f1899e1e000-7f189ae1f000 rw-p
> 7f1899e1e000 00:00 0
> 7f189af19000-7f189af1e000 r-xp 00000000 08:03 706832                    
> /usr/lib/libXfixes.so.3.1.0 7f189af1e000-7f189b11d000 ---p 00005000 08:03
> 706832                     /usr/lib/libXfixes.so.3.1.0
> 7f189b11d000-7f189b11e000 rw-p 00004000 08:03 706832                    
> /usr/lib/libXfixes.so.3.1.0 7f189b11e000-7f189b127000 r-xp 00000000 08:03
> 707183                     /usr/lib/libXcursor.so.1.0.2
> 7f189b127000-7f189b327000 ---p 00009000 08:03 707183                    
> /usr/lib/libXcursor.so.1.0.2 7f189b327000-7f189b328000 rw-p 00009000 08:03
> 707183                     /usr/lib/libXcursor.so.1.0.2
> 7f189b328000-7f189b375000 rw-p 7f18c7f13000 00:00 0
> 7f189b3a9000-7f189b3b0000 r-xp 00000000 08:03 707187                    
> /usr/lib/libXrandr.so.2.1.0 7f189b3b0000-7f189b5af000 ---p 00007000 08:03
> 707187                     /usr/lib/libXrandr.so.2.1.0
> 7f189b5af000-7f189b5b0000 rw-p 00006000 08:03 707187                    
> /usr/lib/libXrandr.so.2.1.0 7f189b5b0000-7f189b5b9000 r-xp 00000000 08:03
> 707141                     /usr/lib/libXrender.so.1.3.0
> 7f189b5b9000-7f189b7b8000 ---p 00009000 08:03 707141                    
> /usr/lib/libXrender.so.1.3.0 7f189b7b8000-7f189b7b9000 r--p 00008000 08:03
> 707141                     /usr/lib/libXrender.so.1.3.0
> 7f189b7b9000-7f189b7ba000 rw-p 00009000 08:03 707141                    
> /usr/lib/libXrender.so.1.3.0 7f189b7ba000-7f189b7ca000 r-xp 00000000 08:03
> 706836                     /usr/lib/libXext.so.6.4.0
> 7f189b7ca000-7f189b9ca000 ---p 00010000 08:03 706836                    
> /usr/lib/libXext.so.6.4.0 7f189b9ca000-7f189b9cc000 rw-p 00010000 08:03
> 706836                     /usr/lib/libXext.so.6.4.0
> 7f189b9cc000-7f189b9d1000 r-xp 00000000 08:03 706262                    
> /usr/lib/libXdmcp.so.6.0.0 7f189b9d1000-7f189bbd0000 ---p 00005000 08:03
> 706262                     /usr/lib/libXdmcp.so.6.0.0
> 7f189bbd0000-7f189bbd1000 rw-p 00004000 08:03 706262                    
> /usr/lib/libXdmcp.so.6.0.0 7f189bbd1000-7f189bbd3000 r-xp 00000000 08:03
> 706256                     /usr/lib/libXau.so.6.0.0
> 7f189bbd3000-7f189bdd2000 ---p 00002000 08:03 706256                    
> /usr/lib/libXau.so.6.0.0 7f189bdd2000-7f189bdd3000 rw-p 00001000 08:03
> 706256                     /usr/lib/libXau.so.6.0.0
> 7f189bdd3000-7f189bdee000 r-xp 00000000 08:03 706260                    
> /usr/lib/libxcb.so.1.0.0 7f189bdee000-7f189bfed000 ---p 0001b000 08:03
> 706260                     /usr/lib/libxcb.so.1.0.0
> 7f189bfed000-7f189bfee000 r--p 0001a000 08:03 706260                    
> /usr/lib/libxcb.so.1.0.0 7f189bfee000-7f189bfef000 rw-p 0001b000 08:03
> 706260                     /usr/lib/libxcb.so.1.0.0
> 7f189bfef000-7f189bff0000 r-xp 00000000 08:03 706258                    
> /usr/lib/libxcb-xlib.so.0.0.0 7f189bff0000-7f189c1ef000 ---p 00001000 08:03
> 706258                     /usr/lib/libxcb-xlib.so.0.0.0
> 7f189c1ef000-7f189c1f0000 r--p 00000000 08:03 706258                    
> /usr/lib/libxcb-xlib.so.0.0.0 7f189c1f0000-7f189c1f1000 rw-p 00001000 08:03
> 706258                     /usr/lib/libxcb-xlib.so.0.0.0
> 7f189c1f1000-7f189c2f4000 r-xp 00000000 08:03 706254                    
> /usr/lib/libX11.so.6.2.0 7f189c2f4000-7f189c4f4000 ---p 00103000 08:03
> 706254                     /usr/lib/libX11.so.6.2.0
> 7f189c4f4000-7f189c4f5000 r--p 00103000 08:03 706254                    
> /usr/lib/libX11.so.6.2.0 7f189c4f5000-7f189c4f9000 rw-p 00104000 08:03
> 706254                     /usr/lib/libX11.so.6.2.0
> 7f189c4f9000-7f18c5e6c000 rw-p 7f189c4f9000 00:00 0
> 7f18c5e6c000-7f18c5e77000 r-xp 00000000 08:03 907564                    
> /lib/libnss_files-2.8.90.so 7f18c5e77000-7f18c6076000 ---p 0000b000 08:03
> 907564                     /lib/libnss_files-2.8.90.so
> 7f18c6076000-7f18c6077000 r--p 0000a000 08:03 907564                    
> /lib/libnss_files-2.8.90.so 7f18c6077000-7f18c6078000 rw-p 0000b000 08:03
> 907564                     /lib/libnss_files-2.8.90.so
> 7f18c6078000-7f18c608c000 r-xp 00000000 08:03 709384                    
> /usr/lib/libdirect-1.0.so.0.1.0 7f18c608c000-7f18c628b000 ---p 00014000
> 08:03 709384                     /usr/lib/libdirect-1.0.so.0.1.0
> 7f18c628b000-7f18c628d000 rw-p 00013000 08:03 709384                    
> /usr/lib/libdirect-1.0.so.0.1.0 7f18c628d000-7f18c6295000 r-xp 00000000
> 08:03 709386                     /usr/lib/libfusion-1.0.so.0.1.0
> 7f18c6295000-7f18c6494000 ---p 00008000 08:03 709386                    
> /usr/lib/libfusion-1.0.so.0.1.0 7f18c6494000-7f18c6495000 rw-p 00007000
> 08:03 709386                     /usr/lib/libfusion-1.0.so.0.1.0
> 7f18c6495000-7f18c6500000 r-xp 00000000 08:03 709385                    
> /usr/lib/libdirectfb-1.0.so.0.1.0 7f18c6500000-7f18c6700000 ---p 0006b000
> 08:03 709385                     /usr/lib/libdirectfb-1.0.so.0.1.0
> 7f18c6700000-7f18c6703000 rw-p 0006b000 08:03 709385                    
> /usr/lib/libdirectfb-1.0.so.0.1.0 7f18c6703000-7f18c6705000 r-xp 00000000
> 08:03 907558                     /lib/libdl-2.8.90.so
> 7f18c6705000-7f18c6905000 ---p 00002000 08:03 907558                    
> /lib/libdl-2.8.90.so 7f18c6905000-7f18c6906000 r--p 00002000 08:03 907558  
>                   /lib/libdl-2.8.90.so 7f18c6906000-7f18c6907000 rw-p
> 00003000 08:03 907558                     /lib/libdl-2.8.90.so
> 7f18c6907000-7f18c69e0000 r-xp 00000000 08:03 708797                    
> /usr/lib/libasound.so.2.0.0 7f18c69e0000-7f18c6be0000 ---p 000d9000 08:03
> 708797                     /usr/lib/libasound.so.2.0.0
> 7f18c6be0000-7f18c6be3000 r--p 000d9000 08:03 708797                    
> /usr/lib/libasound.so.2.0.0 7f18c6be3000-7f18c6be7000 rw-p 000dc000 08:03
> 708797                     /usr/lib/libasound.so.2.0.0
> 7f18c6be7000-7f18c6d50000 r-xp 00000000 08:03 907555                    
> /lib/libc-2.8.90.so 7f18c6d50000-7f18c6f4f000 ---p 00169000 08:03 907555   
>                  /lib/libc-2.8.90.so 7f18c6f4f000-7f18c6f53000 r--p
> 00168000 08:03 907555                     /lib/libc-2.8.90.so
> 7f18c6f53000-7f18c6f54000 rw-p 0016c000 08:03 907555                    
> /lib/libc-2.8.90.so 7f18c6f54000-7f18c6f59000 rw-p 7f18c6f54000 00:00 0
> 7f18c6f59000-7f18c6f90000 r-xp 00000000 08:03 907594                    
> /lib/libncurses.so.5.6 7f18c6f90000-7f18c718f000 ---p 00037000 08:03 907594
>                     /lib/libncurses.so.5.6 7f18c718f000-7f18c7194000 rw-p
> 00036000 08:03 907594                     /lib/libncurses.so.5.6
> 7f18c7194000-7f18c71fd000 r-xp 00000000 08:03 709392                    
> /usr/lib/libSDL-1.2.so.0.11.1 7f18c71fd000-7f18c73fc000 ---p 00069000 08:03
> 709392                     /usr/lib/libSDL-1.2.so.0.11.1
> 7f18c73fc000-7f18c73ff000 rw-p 00068000 08:03 709392                    
> /usr/lib/libSDL-1.2.so.0.11.1 7f18c73ff000-7f18c742b000 rw-p 7f18c73ff000
> 00:00 0
> 7f18c742b000-7f18c742d000 r-xp 00000000 08:03 907574                    
> Aborted (core dumped)
>
>
>
>
> 2) // 03:02.0 Ethernet controller: Realtek Semiconductor Co., Ltd.
> RTL-8139/8139C/8139C+ (rev 10) Registered host PCI device 03:02.0
> ("03:02.0") as guest device 00:03.0 assigned_dev_update_irq: Input/output
> error
>
> [  369.195971] pci 0000:03:02.0: PCI INT A -> GSI 18 (level, low) -> IRQ 18
> [  369.326026] kvm_vm_ioctl_assign_irq: couldn't allocate irq for pv device
> [  369.326363] qemu-system-x86[6780]: segfault at 84bd2620 ip
> 00000000004a29b6 sp 0000000049efce18 error 4 in
> qemu-system-x86_64[400000+195000] [  369.436754] pci 0000:03:02.0: PCI INT
> A disabled
>
>
>
> Sould i update the BIOS if possible/available?
>
> Could it be a problem with the already assigned patches? (See the first
> E-Mail :The VTD [PATCH1/2] seems already be applied. )
>
>
> Gregor

RE: Test with VT-d patches

[Han, Weidong]

I think the assigned NIC shares IRQ with other devices. You can verify
it at request_irq() in kvm_vm_ioctl_assign_irq() (arch/x86/kvm/x86.c).
"assigned_dev_update_irq: Input/output error" should be caused by
request_irq() failure. In my machine, add-on NIC shares IRQ with usb.
You can check it in /proc/interrupts. If it shares IRQ with usb, you can
pass "IRQF_SHARED" to request_irq(), or add "nousb" option in host grub
to don't load drivers for usb. 

BTW, NIC assignment works for linux and windows XP guests on my machine.


Randy (Weidong)

gg@seh.de wrote:
> Hi,
> 
> i make some more tests with
> 1) an old APCI1500/PCI card no linux driver support -> so i don't
> need to unload the module. 2) my second network-card Realtek 10/100
> MBit. 
> 
> Both don't work at all. Only my first gigabit-onboard-network-card
> starts. 
> 
> Here the output from userspace/dmesg:
> 
> 
> 1) // Applied Micro Circuits Corp. APCI1500 Signal processing
> controller 
> 
> Warning: No DNS servers found
> Registered host PCI device 03:00.0 ("03:00.0") as guest device 00:03.0
> assigned_dev_update_irq: Input/output error
> assigned_dev_update_irq: Input/output error
> *** glibc detected *** qemu-system-x86_64: double free or corruption
> (out): 0x00000000013edfb0 *** 
> 
> ======= Backtrace: =========
> /lib/libc.so.6[0x7f18c6c609e8]
> /lib/libc.so.6(cfree+0x76)[0x7f18c6c63036]
> qemu-system-x86_64[0x470d0c]
> qemu-system-x86_64[0x418104]
> qemu-system-x86_64[0x469b62]
> qemu-system-x86_64[0x417fed]
> qemu-system-x86_64[0x4f937d]
> qemu-system-x86_64[0x5221ff]
> qemu-system-x86_64[0x522af3]
> qemu-system-x86_64[0x4f9906]
> qemu-system-x86_64[0x4f9c20]
> /lib/libpthread.so.0[0x7f18c76353ea]
> /lib/libc.so.6(clone+0x6d)[0x7f18c6ccdb9d]
> ======= Memory map: ========
> 00400000-00595000 r-xp 00000000 08:03 719950                         
> /usr/local/bin/qemu-system-x86_64 00794000-00795000 r--p 00194000
> 08:03 719950                            
> /usr/local/bin/qemu-system-x86_64 00795000-007a5000 rw-p 00195000
> 08:03 719950                            
> /usr/local/bin/qemu-system-x86_64 007a5000-00b9b000 rw-p 007a5000
> 00:00 0 00b9b000-00b9c000 rwxp 00b9b000 00:00 0 00b9c000-00bb7000
> rw-p 00b9c000 00:00 0 012e7000-01613000 rw-p 012e7000 00:00 0        
> [heap] 40730000-48c31000 rwxp 40730000 00:00 0 48c31000-48c32000 ---p
> 48c31000 00:00 0 48c32000-49432000 rw-p 48c32000 00:00 0
> 49432000-49433000 ---p 49432000 00:00 0
> 49433000-49436000 rw-p 49433000 00:00 0
> 7f1894000000-7f1894021000 rw-p 7f1894000000 00:00 0
> 7f1894021000-7f1898000000 ---p 7f1894021000 00:00 0
> 7f1899ae0000-7f1899af6000 r-xp 00000000 08:03 908548                 
> /lib/libgcc_s.so.1 7f1899af6000-7f1899cf5000 ---p 00016000 08:03
> 908548                     /lib/libgcc_s.so.1
> 7f1899cf5000-7f1899cf6000 r--p 00015000 08:03 908548                 
> /lib/libgcc_s.so.1 7f1899cf6000-7f1899cf7000 rw-p 00016000 08:03
> 908548                     /lib/libgcc_s.so.1
> 7f1899d04000-7f1899e1e000 rw-s 00000000 00:08 589836                 
> /SYSV00000000 (deleted) 7f1899e1e000-7f189ae1f000 rw-p 7f1899e1e000
> 00:00 0 7f189af19000-7f189af1e000 r-xp 00000000 08:03 706832         
> /usr/lib/libXfixes.so.3.1.0 7f189af1e000-7f189b11d000 ---p 00005000
> 08:03 706832                     /usr/lib/libXfixes.so.3.1.0
> 7f189b11d000-7f189b11e000 rw-p 00004000 08:03 706832                 
> /usr/lib/libXfixes.so.3.1.0 7f189b11e000-7f189b127000 r-xp 00000000
> 08:03 707183                     /usr/lib/libXcursor.so.1.0.2
> 7f189b127000-7f189b327000 ---p 00009000 08:03 707183                 
> /usr/lib/libXcursor.so.1.0.2 7f189b327000-7f189b328000 rw-p 00009000
> 08:03 707183                     /usr/lib/libXcursor.so.1.0.2
> 7f189b328000-7f189b375000 rw-p 7f18c7f13000 00:00 0
> 7f189b3a9000-7f189b3b0000 r-xp 00000000 08:03 707187                 
> /usr/lib/libXrandr.so.2.1.0 7f189b3b0000-7f189b5af000 ---p 00007000
> 08:03 707187                     /usr/lib/libXrandr.so.2.1.0
> 7f189b5af000-7f189b5b0000 rw-p 00006000 08:03 707187                 
> /usr/lib/libXrandr.so.2.1.0 7f189b5b0000-7f189b5b9000 r-xp 00000000
> 08:03 707141                     /usr/lib/libXrender.so.1.3.0
> 7f189b5b9000-7f189b7b8000 ---p 00009000 08:03 707141                 
> /usr/lib/libXrender.so.1.3.0 7f189b7b8000-7f189b7b9000 r--p 00008000
> 08:03 707141                     /usr/lib/libXrender.so.1.3.0
> 7f189b7b9000-7f189b7ba000 rw-p 00009000 08:03 707141                 
> /usr/lib/libXrender.so.1.3.0 7f189b7ba000-7f189b7ca000 r-xp 00000000
> 08:03 706836                     /usr/lib/libXext.so.6.4.0
> 7f189b7ca000-7f189b9ca000 ---p 00010000 08:03 706836                 
> /usr/lib/libXext.so.6.4.0 7f189b9ca000-7f189b9cc000 rw-p 00010000
> 08:03 706836                     /usr/lib/libXext.so.6.4.0
> 7f189b9cc000-7f189b9d1000 r-xp 00000000 08:03 706262                 
> /usr/lib/libXdmcp.so.6.0.0 7f189b9d1000-7f189bbd0000 ---p 00005000
> 08:03 706262                     /usr/lib/libXdmcp.so.6.0.0
> 7f189bbd0000-7f189bbd1000 rw-p 00004000 08:03 706262                 
> /usr/lib/libXdmcp.so.6.0.0 7f189bbd1000-7f189bbd3000 r-xp 00000000
> 08:03 706256                     /usr/lib/libXau.so.6.0.0
> 7f189bbd3000-7f189bdd2000 ---p 00002000 08:03 706256                 
> /usr/lib/libXau.so.6.0.0 7f189bdd2000-7f189bdd3000 rw-p 00001000
> 08:03 706256                     /usr/lib/libXau.so.6.0.0
> 7f189bdd3000-7f189bdee000 r-xp 00000000 08:03 706260                 
> /usr/lib/libxcb.so.1.0.0 7f189bdee000-7f189bfed000 ---p 0001b000
> 08:03 706260                     /usr/lib/libxcb.so.1.0.0
> 7f189bfed000-7f189bfee000 r--p 0001a000 08:03 706260                 
> /usr/lib/libxcb.so.1.0.0 7f189bfee000-7f189bfef000 rw-p 0001b000
> 08:03 706260                     /usr/lib/libxcb.so.1.0.0
> 7f189bfef000-7f189bff0000 r-xp 00000000 08:03 706258                 
> /usr/lib/libxcb-xlib.so.0.0.0 7f189bff0000-7f189c1ef000 ---p 00001000
> 08:03 706258                     /usr/lib/libxcb-xlib.so.0.0.0
> 7f189c1ef000-7f189c1f0000 r--p 00000000 08:03 706258                 
> /usr/lib/libxcb-xlib.so.0.0.0 7f189c1f0000-7f189c1f1000 rw-p 00001000
> 08:03 706258                     /usr/lib/libxcb-xlib.so.0.0.0
> 7f189c1f1000-7f189c2f4000 r-xp 00000000 08:03 706254                 
> /usr/lib/libX11.so.6.2.0 7f189c2f4000-7f189c4f4000 ---p 00103000
> 08:03 706254                     /usr/lib/libX11.so.6.2.0
> 7f189c4f4000-7f189c4f5000 r--p 00103000 08:03 706254                 
> /usr/lib/libX11.so.6.2.0 7f189c4f5000-7f189c4f9000 rw-p 00104000
> 08:03 706254                     /usr/lib/libX11.so.6.2.0
> 7f189c4f9000-7f18c5e6c000 rw-p 7f189c4f9000 00:00 0
> 7f18c5e6c000-7f18c5e77000 r-xp 00000000 08:03 907564                 
> /lib/libnss_files-2.8.90.so 7f18c5e77000-7f18c6076000 ---p 0000b000
> 08:03 907564                     /lib/libnss_files-2.8.90.so
> 7f18c6076000-7f18c6077000 r--p 0000a000 08:03 907564                 
> /lib/libnss_files-2.8.90.so 7f18c6077000-7f18c6078000 rw-p 0000b000
> 08:03 907564                     /lib/libnss_files-2.8.90.so
> 7f18c6078000-7f18c608c000 r-xp 00000000 08:03 709384                 
> /usr/lib/libdirect-1.0.so.0.1.0 7f18c608c000-7f18c628b000 ---p
> 00014000 08:03 709384                    
> /usr/lib/libdirect-1.0.so.0.1.0 7f18c628b000-7f18c628d000 rw-p
> 00013000 08:03 709384                    
> /usr/lib/libdirect-1.0.so.0.1.0 7f18c628d000-7f18c6295000 r-xp
> 00000000 08:03 709386                    
> /usr/lib/libfusion-1.0.so.0.1.0 7f18c6295000-7f18c6494000 ---p
> 00008000 08:03 709386                    
> /usr/lib/libfusion-1.0.so.0.1.0 7f18c6494000-7f18c6495000 rw-p
> 00007000 08:03 709386                    
> /usr/lib/libfusion-1.0.so.0.1.0 7f18c6495000-7f18c6500000 r-xp
> 00000000 08:03 709385                    
> /usr/lib/libdirectfb-1.0.so.0.1.0 7f18c6500000-7f18c6700000 ---p
> 0006b000 08:03 709385                    
> /usr/lib/libdirectfb-1.0.so.0.1.0 7f18c6700000-7f18c6703000 rw-p
> 0006b000 08:03 709385                    
> /usr/lib/libdirectfb-1.0.so.0.1.0 7f18c6703000-7f18c6705000 r-xp
> 00000000 08:03 907558                     /lib/libdl-2.8.90.so
> 7f18c6705000-7f18c6905000 ---p 00002000 08:03 907558                 
> /lib/libdl-2.8.90.so 7f18c6905000-7f18c6906000 r--p 00002000 08:03
> 907558                     /lib/libdl-2.8.90.so
> 7f18c6906000-7f18c6907000 rw-p 00003000 08:03 907558                 
> /lib/libdl-2.8.90.so 7f18c6907000-7f18c69e0000 r-xp 00000000 08:03
> 708797                     /usr/lib/libasound.so.2.0.0
> 7f18c69e0000-7f18c6be0000 ---p 000d9000 08:03 708797                 
> /usr/lib/libasound.so.2.0.0 7f18c6be0000-7f18c6be3000 r--p 000d9000
> 08:03 708797                     /usr/lib/libasound.so.2.0.0
> 7f18c6be3000-7f18c6be7000 rw-p 000dc000 08:03 708797                 
> /usr/lib/libasound.so.2.0.0 7f18c6be7000-7f18c6d50000 r-xp 00000000
> 08:03 907555                     /lib/libc-2.8.90.so
> 7f18c6d50000-7f18c6f4f000 ---p 00169000 08:03 907555                 
> /lib/libc-2.8.90.so 7f18c6f4f000-7f18c6f53000 r--p 00168000 08:03
> 907555                     /lib/libc-2.8.90.so
> 7f18c6f53000-7f18c6f54000 rw-p 0016c000 08:03 907555                 
> /lib/libc-2.8.90.so 7f18c6f54000-7f18c6f59000 rw-p 7f18c6f54000 00:00
> 0 7f18c6f59000-7f18c6f90000 r-xp 00000000 08:03 907594               
> /lib/libncurses.so.5.6 7f18c6f90000-7f18c718f000 ---p 00037000 08:03
> 907594                     /lib/libncurses.so.5.6
> 7f18c718f000-7f18c7194000 rw-p 00036000 08:03 907594                 
> /lib/libncurses.so.5.6 7f18c7194000-7f18c71fd000 r-xp 00000000 08:03
> 709392                     /usr/lib/libSDL-1.2.so.0.11.1
> 7f18c71fd000-7f18c73fc000 ---p 00069000 08:03 709392                 
> /usr/lib/libSDL-1.2.so.0.11.1 7f18c73fc000-7f18c73ff000 rw-p 00068000
> 08:03 709392                     /usr/lib/libSDL-1.2.so.0.11.1
> 7f18c73ff000-7f18c742b000 rw-p 7f18c73ff000 00:00 0
> 7f18c742b000-7f18c742d000 r-xp 00000000 08:03 907574                 
> Aborted (core dumped)                                       
> 
> 
> 
> 
> 2) // 03:02.0 Ethernet controller: Realtek Semiconductor Co., Ltd.
> RTL-8139/8139C/8139C+ (rev 10) 
> Registered host PCI device 03:02.0 ("03:02.0") as guest device 00:03.0
> assigned_dev_update_irq: Input/output error
> 
> [  369.195971] pci 0000:03:02.0: PCI INT A -> GSI 18 (level, low) ->
> IRQ 18 [  369.326026] kvm_vm_ioctl_assign_irq: couldn't allocate irq
> for pv device [  369.326363] qemu-system-x86[6780]: segfault at
> 84bd2620 ip 00000000004a29b6 sp 0000000049efce18 error 4 in
> qemu-system-x86_64[400000+195000] [  369.436754] pci 0000:03:02.0:
> PCI INT A disabled  
> 
> 
> 
> Sould i update the BIOS if possible/available?
> 
> Could it be a problem with the already assigned patches? (See the
> first E-Mail :The VTD [PATCH1/2] seems already be applied. ) 
> 
> 
> Gregor

Re: Test with VT-d patches

[Amit Shah]

Hello,

* On Wednesday 03 September 2008 14:07:36 gg@seh.de wrote:
> Hi,
>
> i make some more tests with
> 1) an old APCI1500/PCI card no linux driver support -> so i don't need to
> unload the module. 2) my second network-card Realtek 10/100 MBit.
>
> Both don't work at all. Only my first gigabit-onboard-network-card starts.
>
> Here the output from userspace/dmesg:
>
>
> 1) // Applied Micro Circuits Corp. APCI1500 Signal processing controller
>
> Warning: No DNS servers found
> Registered host PCI device 03:00.0 ("03:00.0") as guest device 00:03.0
> assigned_dev_update_irq: Input/output error
> assigned_dev_update_irq: Input/output error

This means the devices shares the irq with some other device in the system. 
See the "lspci -v" output for details.

> 2) // 03:02.0 Ethernet controller: Realtek Semiconductor Co., Ltd.
> RTL-8139/8139C/8139C+ (rev 10) Registered host PCI device 03:02.0
> ("03:02.0") as guest device 00:03.0 assigned_dev_update_irq: Input/output
> error
>
> [  369.195971] pci 0000:03:02.0: PCI INT A -> GSI 18 (level, low) -> IRQ 18
> [  369.326026] kvm_vm_ioctl_assign_irq: couldn't allocate irq for pv device

Same error.

> Sould i update the BIOS if possible/available?

That's not necessary. We don't support assigning devices that share the irq on 
the host with some other device. You can try inserting one device at a time 
and try different PCI slots.

> Could it be a problem with the already assigned patches? (See the first
> E-Mail :The VTD [PATCH1/2] seems already be applied. )

No patches have already been applied to any tree. You'll definitely have to 
apply the 1/2 patch as well.

Re: Test with VT-d patches

[Luca Tettamanti]

On Tue, Sep 2, 2008 at 4:17 PM,  <gg@seh.de> wrote:
> Hi,
>
> here comes a small part of the dmesg output. Qemu/KVM produces now a CPU usage
> of about 90%.
>
>
> Sep  2 11:27:35 ubuntu klogd: [  335.057707] ------------[ cut here ]------------
> Sep  2 11:27:35 ubuntu klogd: [  335.057711] WARNING: at kernel/irq/manage.c:180 __enable_irq+0x34/0x80()
> Sep  2 11:27:35 ubuntu klogd: [  335.057713] Unbalanced enable for IRQ 21
[...]
> This messages comes endless.
> Something with IRQs?

Hum, seems that interrupt has already been enabled... did you load the
driver for the NIC in the host? With pass-through the device is
"owned" by the guest.

Luca

RE: Test with VT-d patches

[Han, Weidong]

After I destroyed the guest by clicking qume "x" button, and assigned
the same assigned device to guest, I met this same issue. The reason is
the assigned device doesn't exit normally, then its status is incorrect.
Doing FLR should solve the issue. We will add FLR mechanism for it
later.

For Gregor's issue, I suspect your device is also in a incorrect status
before assignment.

Randy (Weidong)

gg@seh.de wrote:
> Hi,
> 
> here comes a small part of the dmesg output. Qemu/KVM produces now a
> CPU usage 
> of about 90%.
> 
> 
> Sep  2 11:27:35 ubuntu klogd: [  335.057707] ------------[ cut here
> ]------------ 
> Sep  2 11:27:35 ubuntu klogd: [  335.057711] WARNING: at
> kernel/irq/manage.c:180 __enable_irq+0x34/0x80() 
> Sep  2 11:27:35 ubuntu klogd: [  335.057713] Unbalanced enable for
> IRQ 21 
> Sep  2 11:27:35 ubuntu klogd: [  335.057714] Modules linked in:
> kvm_intel kvm ipv6 binfmt_misc i915 drm rfcomm l2cap bluetooth ppdev
> cpufreq_ondemand cpufreq_conservative cpufreq_stats freq_table
> cpufreq_powersave cpufreq_userspace sbs wmi container sbshc video
> output pci_slot bay battery iptable_filter ip_tables x_tables ac lp
> loop snd_hda_intel snd_pcm_oss snd_mixer_oss snd_pcm snd_seq_dummy
> snd_seq_oss snd_seq_midi snd_rawmidi snd_seq_midi_event snd_seq
> snd_timer snd_seq_device iTCO_wdt serio_raw psmouse
> iTCO_vendor_support snd soundcore snd_page_alloc parport_pc parport
> button shpchp pci_hotplug intel_agp joydev evdev dcdbas ext3 jbd
> mbcache sg sd_mod crc_t10dif sr_mod cdrom ata_generic usbhid hid
> pata_acpi ahci 8139too ehci_hcd libata uhci_hcd 8139cp mii scsi_mod
> usbcore dock thermal processor fan fuse Sep  2 11:27:35 ubuntu klogd:
> [  335.057756] Pid: 6609, comm: qemu-system-x86 Tainted: G        W
> 2.6.27-rc4 #5             
> Sep  2 11:27:35 ubuntu klogd: [  335.057758]
> Sep  2 11:27:35 ubuntu klogd: [  335.057758] Call Trace:
> Sep  2 11:27:35 ubuntu klogd: [  335.057761]  [<ffffffff802469bc>]
> warn_slowpath+0xbc/0xf0 
> Sep  2 11:27:35 ubuntu klogd: [  335.057769]  [<ffffffffa0450ba6>] ?
> walk_shadow+0x96/0xe0 [kvm] 
> Sep  2 11:27:35 ubuntu klogd: [  335.057777]  [<ffffffffa0455190>] ?
> paging64_page_fault+0x220/0x270 [kvm] 
> Sep  2 11:27:35 ubuntu klogd: [  335.057785]  [<ffffffffa0456a78>] ?
> x86_decode_insn+0x8a8/0xe20 [kvm] 
> Sep  2 11:27:35 ubuntu klogd: [  335.057792]  [<ffffffffa04536c0>] ?
> paging64_shadow_walk_entry+0x0/0x210 [kvm] 
> Sep  2 11:27:35 ubuntu klogd: [  335.057795]  [<ffffffff80294a04>]
> __enable_irq+0x34/0x80 
> Sep  2 11:27:35 ubuntu klogd: [  335.057797]  [<ffffffff80294aa2>]
> enable_irq+0x52/0x80 
> Sep  2 11:27:35 ubuntu klogd: [  335.057805]  [<ffffffffa044acad>]
> kvm_assigned_dev_ack_irq+0x2d/0x40 [kvm] 
> Sep  2 11:27:35 ubuntu klogd: [  335.057812]  [<ffffffffa045ad4a>]
> kvm_notify_acked_irq+0x3a/0x60 [kvm] 
> Sep  2 11:27:35 ubuntu klogd: [  335.057820]  [<ffffffffa045ac01>]
> kvm_pic_read_irq+0x91/0x140 [kvm] 
> Sep  2 11:27:35 ubuntu klogd: [  335.057827]  [<ffffffffa045ae8f>]
> kvm_cpu_get_interrupt+0x5f/0x70 [kvm] 
> Sep  2 11:27:35 ubuntu klogd: [  335.057831]  [<ffffffffa0475748>]
> vmx_intr_assist+0x178/0x190 [kvm_intel] 
> Sep  2 11:27:35 ubuntu klogd: [  335.057838]  [<ffffffffa04503c8>]
> kvm_arch_vcpu_ioctl_run+0x3a8/0x810 [kvm] 
> Sep  2 11:27:35 ubuntu klogd: [  335.057841]  [<ffffffff8026193a>] ?
> enqueue_hrtimer+0x8a/0x140 
> Sep  2 11:27:35 ubuntu klogd: [  335.057848]  [<ffffffffa04455d2>]
> kvm_vcpu_ioctl+0x2f2/0x5b0 [kvm] 
> Sep  2 11:27:35 ubuntu klogd: [  335.057856]  [<ffffffffa0446d75>] ?
> kvm_vm_ioctl+0xd5/0x280 [kvm] 
> Sep  2 11:27:35 ubuntu klogd: [  335.057858]  [<ffffffff8020a675>] ?
> __switch_to+0x265/0x490 
> Sep  2 11:27:35 ubuntu klogd: [  335.057860]  [<ffffffff802ef7c6>]
> vfs_ioctl+0x36/0xb0 
> Sep  2 11:27:35 ubuntu klogd: [  335.057862]  [<ffffffff804e6259>] ?
> thread_return+0x3d/0x654 
> Sep  2 11:27:35 ubuntu klogd: [  335.057864]  [<ffffffff802efac3>]
> do_vfs_ioctl+0x283/0x2f0 
> Sep  2 11:27:35 ubuntu klogd: [  335.057866]  [<ffffffff802e2078>] ?
> fget_light+0x68/0xb0 
> Sep  2 11:27:35 ubuntu klogd: [  335.057868]  [<ffffffff802efbd1>]
> sys_ioctl+0xa1/0xb0 
> Sep  2 11:27:35 ubuntu klogd: [  335.057870]  [<ffffffff8020c7da>]
> system_call_fastpath+0x16/0x1b 
> Sep  2 11:27:35 ubuntu klogd: [  335.057872]
> Sep  2 11:27:35 ubuntu klogd: [  335.057873] ---[ end trace
> 5c1916f7999c68bb ]-- 
> 
> 
> 
> 
> This messages comes endless.
> Something with IRQs?
> 
> Regards,
> 
> Gregor Glomm

Re: Test with VT-d patches

[gg]

Hi, 


>On Tue, Sep 2, 2008 at 4:17 PM,  <gg@seh.de> wrote:
>> Hi,
>>
>> here comes a small part of the dmesg output. Qemu/KVM produces now a CPU 
>usage
>> of about 90%.
>>
>>
>> Sep  2 11:27:35 ubuntu klogd: [  335.057707] ------------[ cut here 
>]------------
>> Sep  2 11:27:35 ubuntu klogd: [  335.057711] WARNING: at 
>kernel/irq/manage.c:180 __enable_irq+0x34/0x80()
>> Sep  2 11:27:35 ubuntu klogd: [  335.057713] Unbalanced enable for IRQ 21
>[...]
>> This messages comes endless.
>> Something with IRQs?
>
>Hum, seems that interrupt has already been enabled... did you load the
>driver for the NIC in the host? With pass-through the device is
>"owned" by the guest.

No.
The driver was not loaded. 
It is not possible to start qemu with a pci-device used by the host.
(I try this with the second network device.)

Can this interrupt be shared with an other IRQ?

Gregor

Re: Test with VT-d patches

[Yang, Sheng]

On Wednesday 03 September 2008 02:40:06 gg@seh.de wrote:
> Hi,
>
> >On Tue, Sep 2, 2008 at 4:17 PM,  <gg@seh.de> wrote:
> >> Hi,
> >>
> >> here comes a small part of the dmesg output. Qemu/KVM produces now a CPU
> >
> >usage
> >
> >> of about 90%.
> >>
> >>
> >> Sep  2 11:27:35 ubuntu klogd: [  335.057707] ------------[ cut here
> >
> >]------------
> >
> >> Sep  2 11:27:35 ubuntu klogd: [  335.057711] WARNING: at
> >
> >kernel/irq/manage.c:180 __enable_irq+0x34/0x80()
> >
> >> Sep  2 11:27:35 ubuntu klogd: [  335.057713] Unbalanced enable for IRQ
> >> 21
> >
> >[...]
> >
> >> This messages comes endless.
> >> Something with IRQs?
> >
> >Hum, seems that interrupt has already been enabled... did you load the
> >driver for the NIC in the host? With pass-through the device is
> >"owned" by the guest.
>
> No.
> The driver was not loaded.
> It is not possible to start qemu with a pci-device used by the host.
> (I try this with the second network device.)
>
> Can this interrupt be shared with an other IRQ?

Oh, no, not currently...

The device's IRQ should not be shared with other, we haven't implement shared 
IRQ logic, but soon. 
-- 
regards
Yang, Sheng
>
> Gregor

Test with VT-d patches

[gg]

Hi,

i am interested in the use of the new VT-d hardware feature.
My Dell-PC "OPTIPLEX" is capable for this.
The Linux-System was an Ubuntu-8.10 (AMD64) with the current
Linux-Kernel from the KVM-Kernel GIT repository. (2.6.27-rc4)
To use VT-d i download kvm-74 and take the patches from "Amit Shah".
1) The KVM/userspace [PATCH1/1] was applied without errors.
2) The VTD [PATCH1/2] seems already be applied.
3) The VTD [PATCH2/2] was applied without errors.

Now I use the command line option -pcidevice dev=00:03.19 to pass the
Intel Pro Gigabit Network Device to my WindowsXP Guest-System.

Qemu told me something like: "passing 00:03.19 as device 00:03.00 to the guest system". Windows starts normally fast. But then its slow rapidly down. My mouse stops for about 10sec and then goes again for 1sec.

Windows remember the new hardware correctly. And after i install the new driver, the system seems to go a little bit faster.

But the problem was, that no ping or other network packages was send/received. 

What can i do to find the problem? (Debugging?)

Regards,

Gregor Glomm

Re: Test with VT-d patches

[Yang, Sheng]

On Tuesday 02 September 2008 16:44:18 gg@seh.de wrote:
> Hi,
>
> i am interested in the use of the new VT-d hardware feature.
> My Dell-PC "OPTIPLEX" is capable for this.
> The Linux-System was an Ubuntu-8.10 (AMD64) with the current
> Linux-Kernel from the KVM-Kernel GIT repository. (2.6.27-rc4)
> To use VT-d i download kvm-74 and take the patches from "Amit Shah".
> 1) The KVM/userspace [PATCH1/1] was applied without errors.
> 2) The VTD [PATCH1/2] seems already be applied.
> 3) The VTD [PATCH2/2] was applied without errors.
>
> Now I use the command line option -pcidevice dev=00:03.19 to pass the
> Intel Pro Gigabit Network Device to my WindowsXP Guest-System.
>
> Qemu told me something like: "passing 00:03.19 as device 00:03.00 to the
> guest system". Windows starts normally fast. But then its slow rapidly
> down. My mouse stops for about 10sec and then goes again for 1sec.
>
> Windows remember the new hardware correctly. And after i install the new
> driver, the system seems to go a little bit faster.
>
> But the problem was, that no ping or other network packages was
> send/received.
>
> What can i do to find the problem? (Debugging?)

Hi Gregor

Can you have a look at your dmesg and post it here? I think we can got some 
clue. 

Thanks!
--
regards
Yang, Sheng
>
> Regards,
>
> Gregor Glomm
> --
> To unsubscribe from this list: send the line "unsubscribe kvm" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

RE: Test with VT-d patches

[Han, Weidong]

Hi Gregor,

You'd better set the loglevel=8 to print debug information. Can you also
try linux guest?

Randy (Weidong)

Yang, Sheng wrote:
> On Tuesday 02 September 2008 16:44:18 gg@seh.de wrote:
>> Hi,
>> 
>> i am interested in the use of the new VT-d hardware feature.
>> My Dell-PC "OPTIPLEX" is capable for this.
>> The Linux-System was an Ubuntu-8.10 (AMD64) with the current
>> Linux-Kernel from the KVM-Kernel GIT repository. (2.6.27-rc4)
>> To use VT-d i download kvm-74 and take the patches from "Amit Shah".
>> 1) The KVM/userspace [PATCH1/1] was applied without errors.
>> 2) The VTD [PATCH1/2] seems already be applied.
>> 3) The VTD [PATCH2/2] was applied without errors.
>> 
>> Now I use the command line option -pcidevice dev=00:03.19 to pass the
>> Intel Pro Gigabit Network Device to my WindowsXP Guest-System.
>> 
>> Qemu told me something like: "passing 00:03.19 as device 00:03.00 to
>> the guest system". Windows starts normally fast. But then its slow
>> rapidly down. My mouse stops for about 10sec and then goes again for
>> 1sec. 
>> 
>> Windows remember the new hardware correctly. And after i install the
>> new driver, the system seems to go a little bit faster.
>> 
>> But the problem was, that no ping or other network packages was
>> send/received. 
>> 
>> What can i do to find the problem? (Debugging?)
> 
> Hi Gregor
> 
> Can you have a look at your dmesg and post it here? I think we can
> got some clue.
> 
> Thanks!
> --
> regards
> Yang, Sheng
>> 
>> Regards,
>> 
>> Gregor Glomm
>> --
>> To unsubscribe from this list: send the line "unsubscribe kvm" in
>> the body of a message to majordomo@vger.kernel.org
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html