Lose XHCI Device on HP Ivybridge While Resuming on Battery

Lose XHCI Device on HP Ivybridge While Resuming on Battery

[Tom Goetz]

After resuming more than once on battery these HP Ivybridge laptops, the XHCI devices stop working. Have anyone seen this before? I wanted to check before diving in deeper. Let me know if you have any ideas. Thanks!

Facts
	- Xen 4.0.3, Linux 3.2.7 PVOPs
	- Happens on HP Ivybridge. Doesn't happen on very similar HP Sandybridge Clash system.
	- Happens on battery, but not on AC.
	- Doesn't happen on first suspend/resume. First resume may be long. Don't have enough sample to be sure of this.
	- CPU power governor doesn't effect the issue. Happens irregardless of governor.
	- Seems to effect MSI devices only.
	- IRQ changes appear in good resume case. Not just bad.
 
I've diffed lspci output across a working resume on AC and across a bad resume on battery. There seem to be the differences unique to the bad case:

 00:14.0 USB controller: Intel Corporation Panther Point USB xHCI Host Controller (rev 04) (prog-if 30 [XHCI])
 Subsystem: Hewlett-Packard Company Device 179b
-Flags: bus master, medium devsel, latency 0, IRQ 326
+Flags: medium devsel, IRQ 327                             <-------------- lose bus master, latency 0
 Memory at 94720000 (64-bit, non-prefetchable) [size=64K]
 Capabilities: [70] Power Management version 2
 Capabilities: [80] MSI: Enable+ Count=1/8 Maskable- 64bit+
 Kernel driver in use: xhci_hcd
 Kernel modules: xhci-hcd

--- 

 00:19.0 Ethernet controller: Intel Corporation 82579LM Gigabit Network Connection (rev 04)
 Subsystem: Hewlett-Packard Company Device 179b
-Flags: bus master, fast devsel, latency 0, IRQ 327
-Memory at 94700000 (32-bit, non-prefetchable) [size=128K]
-Memory at 9473a000 (32-bit, non-prefetchable) [size=4K]
-I/O ports at 4060 [size=32]
+Flags: fast devsel, IRQ 20                           <-------------- lose bus master, latency 0
+Memory at 94700000 (32-bit, non-prefetchable) [disabled] [size=128K]   <------ becomes disabled
+Memory at 9473a000 (32-bit, non-prefetchable) [disabled] [size=4K]     <------ becomes disabled
+I/O ports at 4060 [disabled] [size=32]                                 <------ becomes disabled
 Capabilities: [c8] Power Management version 2
-Capabilities: [d0] MSI: Enable+ Count=1/1 Maskable- 64bit+
+Capabilities: [d0] MSI: Enable- Count=1/1 Maskable- 64bit+    <---- toggle enable polarity
 Capabilities: [e0] PCI Advanced Features
 Kernel driver in use: e1000e
 Kernel modules: e1000e

---

 00:1a.0 USB controller: Intel Corporation Panther Point USB Enhanced Host Controller #2 (rev 04) (prog-if 20 [EHCI])
 Subsystem: Hewlett-Packard Company Device 179b
-Flags: bus master, medium devsel, latency 0, IRQ 16
+Flags: medium devsel, IRQ 16                           <-------------- lose bus master, latency 0
 Memory at 94739000 (32-bit, non-prefetchable) [size=1K]
 Capabilities: [50] Power Management version 2
 Capabilities: [58] Debug port: BAR=1 offset=00a0
 Capabilities: [98] PCI Advanced Features
 Kernel driver in use: ehci_hcd
 Kernel modules: ehci-hcd

syslog looks like this for both good and bad resumes:

Mar  5 15:48:51 newt kernel: [  289.730303] xhci_hcd 0000:00:14.0: can't derive routing for PCI INT A
Mar  5 15:48:51 newt kernel: [  289.730311] xhci_hcd 0000:00:14.0: PCI INT A: no GSI - using ISA IRQ 10
Mar  5 15:48:51 newt kernel: [  289.730319] xen: registering gsi 10 triggering 0 polarity 1
Mar  5 15:48:51 newt kernel: [  289.730327] xen_map_pirq_gsi: returning irq 10 for gsi 10
Mar  5 15:48:51 newt kernel: [  289.730332] xen: --> pirq=10 -> irq=10 (gsi=10)
Mar  5 15:48:51 newt kernel: [  289.730337] Already setup the GSI :10
Mar  5 15:48:51 newt kernel: [  289.730445] xhci_hcd 0000:00:14.0: setting latency timer to 64
Mar  5 15:48:51 newt kernel: [  289.730455] xhci_hcd 0000:00:14.0: xHCI Host Controller
Mar  5 15:48:51 newt kernel: [  289.730687] xhci_hcd 0000:00:14.0: new USB bus registered, assigned bus number 3
Mar  5 15:48:51 newt kernel: [  289.730826] xhci_hcd 0000:00:14.0: cache line size of 64 is not supported
Mar  5 15:48:51 newt kernel: [  289.730858] xhci_hcd 0000:00:14.0: irq 10, io mem 0x94720000
Mar  5 15:48:51 newt kernel: [  289.731470] xHCI xhci_add_endpoint called for root hub
Mar  5 15:48:51 newt kernel: [  289.731477] xHCI xhci_check_bandwidth called for root hub
Mar  5 15:48:51 newt kernel: [  289.731551] hub 3-0:1.0: USB hub found
Mar  5 15:48:51 newt kernel: [  289.731568] hub 3-0:1.0: 4 ports detected
Mar  5 15:48:51 newt kernel: [  289.843501] xhci_hcd 0000:00:14.0: xHCI Host Controller
Mar  5 15:48:51 newt kernel: [  289.843722] xhci_hcd 0000:00:14.0: new USB bus registered, assigned bus number 4
Mar  5 15:48:51 newt kernel: [  289.844015] xHCI xhci_add_endpoint called for root hub
Mar  5 15:48:51 newt kernel: [  289.844021] xHCI xhci_check_bandwidth called for root hub
Mar  5 15:48:51 newt kernel: [  289.844084] hub 4-0:1.0: USB hub found
Mar  5 15:48:51 newt kernel: [  289.844104] hub 4-0:1.0: 4 ports detected

Re: Lose XHCI Device on HP Ivybridge While Resuming on Battery

[Tom Goetz]

On Mar 5, 2012, at 12:09 PM, xen-devel-request@lists.xen.org wrote:

> Send Xen-devel mailing list submissions to
> 	xen-devel@lists.xen.org
> 
> 
> 
> Today's Topics:
> 
>   2. Lose XHCI Device on HP Ivybridge While Resuming on	Battery
>      (Tom Goetz)
> 
> 
> 
> Message: 2
> Date: Mon, 5 Mar 2012 11:11:10 -0500
> From: Tom Goetz <tom.goetz@virtualcomputer.com>
> To: xen-devel@lists.xen.org
> Subject: [Xen-devel] Lose XHCI Device on HP Ivybridge While Resuming
> 	on	Battery
> Message-ID: <5D6CE9D4-9859-47F7-9885-7A8CAD00AF64@virtualcomputer.com>
> Content-Type: text/plain; charset=us-ascii
> 
> After resuming more than once on battery these Ivybridge laptop, the XHCI devices stop working. Have anyone seen this before? I wanted to check before diving in deeper. Let me know if you have any ideas. Thanks!
> 

I have more information on this. Here's the sequence of events:

1. XHCI driver removed with SUSPEND_MODULES
2. suspend
3. resume
4. XHCI driver loaded by SUSPEND_MODULES
5. Device is enabled by driver.
6. Device disabled by runtime PM.
7. Plugging in a USB device does not result in the device being enabled.

This works on the same machine in Ubuntu 12.04 Beta 1 (Linux 3.2.0) native which is very similar to our Ubuntu 12.04 based (Linux 3.2.5 & 3.2.9) build. So either this was broken between 3.2.0 and 3.2.5 or this is Xen related. I've included some tracing of resume and runtime pm suspend of the device. A USB flash stick was plugged in at the end of the log where nothing happens:


Mar  8 21:57:47 newt kernel: [  407.323444] __pm_runtime_resume[879] 0000:00:14.0 RPM_GET_PUT 2

Mar  8 21:57:47 newt kernel: [  407.323451] rpm_resume[512] 0000:00:14.0 before

Mar  8 21:57:47 newt kernel: [  407.323461] __pm_runtime_idle[811] 0000:00:14.0 RPM_GET_PUT 1

Mar  8 21:57:47 newt kernel: [  407.323477] __pm_runtime_resume[879] 0000:00:14.0 RPM_GET_PUT 2

Mar  8 21:57:47 newt kernel: [  407.323487] rpm_resume[512] 0000:00:14.0 before

Mar  8 21:57:47 newt kernel: [  407.323489] xhci_hcd 0000:00:14.0: remove, state 4

Mar  8 21:57:47 newt kernel: [  407.323530] rpm_resume[512] 0000:00:14.0 before

Mar  8 21:57:47 newt kernel: [  407.323532] __pm_runtime_idle[811] 0000:00:14.0 RPM_GET_PUT 2

Mar  8 21:57:47 newt kernel: [  407.323731] xhci_hcd 0000:00:14.0: USB bus 4 deregistered

Mar  8 21:57:47 newt kernel: [  407.323820] xhci_hcd 0000:00:14.0: remove, state 4

Mar  8 21:57:47 newt kernel: [  407.323862] rpm_resume[512] 0000:00:14.0 before

Mar  8 21:57:47 newt kernel: [  407.323873] __pm_runtime_idle[811] 0000:00:14.0 RPM_GET_PUT 3

Mar  8 21:57:47 newt kernel: [  407.324123] xhci_hcd 0000:00:14.0: USB bus 3 deregistered

Mar  8 21:57:47 newt kernel: [  407.324232] pci_disable_device[1345] 0000:00:14.0

Mar  8 21:57:47 newt kernel: [  407.324244] xhci_hcd 0000:00:14.0: can't derive routing for PCI INT A

Mar  8 21:57:47 newt kernel: [  407.324246] pci_disable_device[1353] 0000:00:14.0 busmaster off

Mar  8 21:57:57 newt kernel: [  409.053506] rpm_resume[512] 0000:00:14.0 before

Mar  8 21:57:57 newt kernel: [  413.422310] pci 0000:00:14.0: restoring config space at offset 0xf (was 0x100, writing 0x10a)

Mar  8 21:57:57 newt kernel: [  413.422354] pci 0000:00:14.0: restoring config space at offset 0x4 (was 0x4, writing 0x94720004)

Mar  8 21:57:57 newt kernel: [  413.422368] pci 0000:00:14.0: restoring config space at offset 0x1 (was 0x2900000, writing 0x2900002)

Mar  8 21:57:57 newt kernel: [  413.424862] __pm_runtime_idle[811] 0000:00:14.0 RPM_GET_PUT 2

Mar  8 21:57:58 newt kernel: [  414.122464] __pm_runtime_idle[811] 0000:00:14.0 RPM_GET_PUT 1

Mar  8 21:57:59 newt kernel: [  415.758316] __pci_enable_device_flags[1129] 0000:00:14.0

Mar  8 21:57:59 newt kernel: [  415.758323] __pci_enable_device_flags[1133] 0000:00:14.0

Mar  8 21:57:59 newt kernel: [  415.758342] xhci_hcd 0000:00:14.0: can't derive routing for PCI INT A

Mar  8 21:57:59 newt kernel: [  415.758348] xhci_hcd 0000:00:14.0: PCI INT A: no GSI - using ISA IRQ 10

Mar  8 21:57:59 newt kernel: [  415.758378] __pci_enable_device_flags[1145] 0000:00:14.0

Mar  8 21:57:59 newt kernel: [  415.758481] pci_set_master[2620] 0000:00:14.0

Mar  8 21:57:59 newt kernel: [  415.758485] __pci_set_master[2593] 0000:00:14.0

Mar  8 21:57:59 newt kernel: [  415.758510] __pci_set_master[2606] 0000:00:14.0 enabling pci_write_config_word ret 0

Mar  8 21:57:59 newt kernel: [  415.758519] xhci_hcd 0000:00:14.0: setting latency timer to 64

Mar  8 21:57:59 newt kernel: [  415.758529] xhci_hcd 0000:00:14.0: xHCI Host Controller

Mar  8 21:57:59 newt kernel: [  415.758771] xhci_hcd 0000:00:14.0: new USB bus registered, assigned bus number 3

Mar  8 21:57:59 newt kernel: [  415.758924] xhci_hcd 0000:00:14.0: cache line size of 64 is not supported

Mar  8 21:57:59 newt kernel: [  415.758957] xhci_hcd 0000:00:14.0: irq 10, io mem 0x94720000

Mar  8 21:57:59 newt kernel: [  415.759002] pci_enable_msi_block[699] 0000:00:14.0

Mar  8 21:57:59 newt kernel: [  415.759027] pci_enable_msi_block[703] 0000:00:14.0

Mar  8 21:57:59 newt kernel: [  415.759036] pci_enable_msi_block[709] 0000:00:14.0

Mar  8 21:57:59 newt kernel: [  415.759062] pci_enable_msi_block[716] 0000:00:14.0

Mar  8 21:57:59 newt kernel: [  415.759066] pci_enable_msi_block[724] 0000:00:14.0

Mar  8 21:57:59 newt kernel: [  415.821636] xhci_hcd 0000:00:14.0: xHCI Host Controller

Mar  8 21:57:59 newt kernel: [  415.821879] xhci_hcd 0000:00:14.0: new USB bus registered, assigned bus number 4

Mar  8 21:57:59 newt kernel: [  415.971575] __pm_runtime_idle[811] 0000:00:14.0 RPM_GET_PUT 1

Mar  8 21:57:59 newt kernel: [  416.403188] pm_runtime_allow[1156] 0000:00:14.0 1

Mar  8 21:57:59 newt kernel: [  416.403195] pm_runtime_allow[1159] 0000:00:14.0 IDLE

Mar  8 21:57:59 newt kernel: [  416.403202] rpm_idle[206] 0000:00:14.0 before

Mar  8 21:57:59 newt kernel: [  416.403208] rpm_check_suspend_allowed[145] 0000:00:14.0 children

Mar  8 21:57:59 newt kernel: [  416.403214] rpm_idle[210] 0000:00:14.0 !conditions

Mar  8 21:58:01 newt kernel: [  418.441612] rpm_suspend[474] usb3 parent 0000:00:14.0

Mar  8 21:58:01 newt kernel: [  418.441613] rpm_idle[206] 0000:00:14.0 before

Mar  8 21:58:01 newt kernel: [  418.441615] rpm_check_suspend_allowed[145] 0000:00:14.0 children

Mar  8 21:58:01 newt kernel: [  418.441616] rpm_idle[210] 0000:00:14.0 !conditions

Mar  8 21:58:01 newt kernel: [  418.441630] rpm_suspend[474] usb4 parent 0000:00:14.0

Mar  8 21:58:01 newt kernel: [  418.441631] rpm_idle[206] 0000:00:14.0 before

Mar  8 21:58:01 newt kernel: [  418.441635] pm_runtime_work[708] 0000:00:14.0 RPM_REQ_IDLE

Mar  8 21:58:01 newt kernel: [  418.441636] rpm_idle[206] 0000:00:14.0 before

Mar  8 21:58:01 newt kernel: [  418.441639] pci_pm_runtime_idle[1064] 0000:00:14.0 before

Mar  8 21:58:01 newt kernel: [  418.441640] pci_pm_runtime_idle[1074] 0000:00:14.0 pm_runtime_suspend

Mar  8 21:58:01 newt kernel: [  418.441642] __pm_runtime_suspend[849] 0000:00:14.0 before

Mar  8 21:58:01 newt kernel: [  418.441643] rpm_suspend[323] 0000:00:14.0 before

Mar  8 21:58:01 newt kernel: [  418.441644] rpm_suspend[419] 0000:00:14.0 RPM_SUSPENDING

Mar  8 21:58:01 newt kernel: [  418.441646] pci_pm_runtime_suspend[1016] 0000:00:14.0 before

Mar  8 21:58:01 newt kernel: [  418.441648] hcd_pci_runtime_suspend[564] 0000:00:14.0 before

Mar  8 21:58:01 newt kernel: [  418.441649] suspend_common[389] 0000:00:14.0 before

Mar  8 21:58:01 newt kernel: [  418.441688] pci_disable_device[1345] 0000:00:14.0

Mar  8 21:58:01 newt kernel: [  418.441697] pci_disable_device[1353] 0000:00:14.0 busmaster off

Mar  8 21:58:01 newt kernel: [  418.441850] suspend_common[436] 0000:00:14.0 after

Mar  8 21:58:01 newt kernel: [  418.441852] hcd_pci_runtime_suspend[569] 0000:00:14.0 afyer ret 0

Mar  8 21:58:01 newt kernel: [  418.441853] pci_pm_runtime_suspend[1018] 0000:00:14.0 after

Mar  8 21:58:01 newt kernel: [  418.441917] xhci_hcd 0000:00:14.0: PME# enabled

Mar  8 21:58:01 newt kernel: [  418.461530] pci_pm_runtime_suspend[1037] 0000:00:14.0 done

Mar  8 21:58:01 newt kernel: [  418.461544] rpm_suspend[474] 0000:00:14.0 parent pci0000:00

Mar  8 21:58:01 newt kernel: [  418.461553] __pm_runtime_suspend[853] 0000:00:14.0 after

Mar  8 21:58:01 newt kernel: [  418.461555] pci_pm_runtime_idle[1076] 0000:00:14.0 after

Re: Lose XHCI Device on HP Ivybridge While Resuming on Battery

[Konrad Rzeszutek Wilk]

On Mon, Mar 05, 2012 at 11:11:10AM -0500, Tom Goetz wrote:
> After resuming more than once on battery these HP Ivybridge laptops, the XHCI devices stop working. Have anyone seen this before? I wanted to check before diving in deeper. Let me know if you have any ideas. Thanks!
> 
> Facts
> 	- Xen 4.0.3, Linux 3.2.7 PVOPs
> 	- Happens on HP Ivybridge. Doesn't happen on very similar HP Sandybridge Clash system.
> 	- Happens on battery, but not on AC.

Did you guys try to limit the C-state? say max_cstate=2 ?

> 	- Doesn't happen on first suspend/resume. First resume may be long. Don't have enough sample to be sure of this.
> 	- CPU power governor doesn't effect the issue. Happens irregardless of governor.
> 	- Seems to effect MSI devices only.
> 	- IRQ changes appear in good resume case. Not just bad.
>  
> I've diffed lspci output across a working resume on AC and across a bad resume on battery. There seem to be the differences unique to the bad case:
> 
>  00:14.0 USB controller: Intel Corporation Panther Point USB xHCI Host Controller (rev 04) (prog-if 30 [XHCI])
>  Subsystem: Hewlett-Packard Company Device 179b
> -Flags: bus master, medium devsel, latency 0, IRQ 326
> +Flags: medium devsel, IRQ 327                             <-------------- lose bus master, latency 0
>  Memory at 94720000 (64-bit, non-prefetchable) [size=64K]
>  Capabilities: [70] Power Management version 2
>  Capabilities: [80] MSI: Enable+ Count=1/8 Maskable- 64bit+
>  Kernel driver in use: xhci_hcd
>  Kernel modules: xhci-hcd
> 
> --- 
> 
>  00:19.0 Ethernet controller: Intel Corporation 82579LM Gigabit Network Connection (rev 04)
>  Subsystem: Hewlett-Packard Company Device 179b
> -Flags: bus master, fast devsel, latency 0, IRQ 327
> -Memory at 94700000 (32-bit, non-prefetchable) [size=128K]
> -Memory at 9473a000 (32-bit, non-prefetchable) [size=4K]
> -I/O ports at 4060 [size=32]
> +Flags: fast devsel, IRQ 20                           <-------------- lose bus master, latency 0
> +Memory at 94700000 (32-bit, non-prefetchable) [disabled] [size=128K]   <------ becomes disabled
> +Memory at 9473a000 (32-bit, non-prefetchable) [disabled] [size=4K]     <------ becomes disabled
> +I/O ports at 4060 [disabled] [size=32]                                 <------ becomes disabled
>  Capabilities: [c8] Power Management version 2
> -Capabilities: [d0] MSI: Enable+ Count=1/1 Maskable- 64bit+
> +Capabilities: [d0] MSI: Enable- Count=1/1 Maskable- 64bit+    <---- toggle enable polarity
>  Capabilities: [e0] PCI Advanced Features
>  Kernel driver in use: e1000e
>  Kernel modules: e1000e
> 
> ---
> 
>  00:1a.0 USB controller: Intel Corporation Panther Point USB Enhanced Host Controller #2 (rev 04) (prog-if 20 [EHCI])
>  Subsystem: Hewlett-Packard Company Device 179b
> -Flags: bus master, medium devsel, latency 0, IRQ 16
> +Flags: medium devsel, IRQ 16                           <-------------- lose bus master, latency 0
>  Memory at 94739000 (32-bit, non-prefetchable) [size=1K]
>  Capabilities: [50] Power Management version 2
>  Capabilities: [58] Debug port: BAR=1 offset=00a0
>  Capabilities: [98] PCI Advanced Features
>  Kernel driver in use: ehci_hcd
>  Kernel modules: ehci-hcd
> 
> syslog looks like this for both good and bad resumes:
> 
> Mar  5 15:48:51 newt kernel: [  289.730303] xhci_hcd 0000:00:14.0: can't derive routing for PCI INT A
> Mar  5 15:48:51 newt kernel: [  289.730311] xhci_hcd 0000:00:14.0: PCI INT A: no GSI - using ISA IRQ 10
> Mar  5 15:48:51 newt kernel: [  289.730319] xen: registering gsi 10 triggering 0 polarity 1
> Mar  5 15:48:51 newt kernel: [  289.730327] xen_map_pirq_gsi: returning irq 10 for gsi 10
> Mar  5 15:48:51 newt kernel: [  289.730332] xen: --> pirq=10 -> irq=10 (gsi=10)
> Mar  5 15:48:51 newt kernel: [  289.730337] Already setup the GSI :10
> Mar  5 15:48:51 newt kernel: [  289.730445] xhci_hcd 0000:00:14.0: setting latency timer to 64
> Mar  5 15:48:51 newt kernel: [  289.730455] xhci_hcd 0000:00:14.0: xHCI Host Controller
> Mar  5 15:48:51 newt kernel: [  289.730687] xhci_hcd 0000:00:14.0: new USB bus registered, assigned bus number 3
> Mar  5 15:48:51 newt kernel: [  289.730826] xhci_hcd 0000:00:14.0: cache line size of 64 is not supported
> Mar  5 15:48:51 newt kernel: [  289.730858] xhci_hcd 0000:00:14.0: irq 10, io mem 0x94720000
> Mar  5 15:48:51 newt kernel: [  289.731470] xHCI xhci_add_endpoint called for root hub
> Mar  5 15:48:51 newt kernel: [  289.731477] xHCI xhci_check_bandwidth called for root hub
> Mar  5 15:48:51 newt kernel: [  289.731551] hub 3-0:1.0: USB hub found
> Mar  5 15:48:51 newt kernel: [  289.731568] hub 3-0:1.0: 4 ports detected
> Mar  5 15:48:51 newt kernel: [  289.843501] xhci_hcd 0000:00:14.0: xHCI Host Controller
> Mar  5 15:48:51 newt kernel: [  289.843722] xhci_hcd 0000:00:14.0: new USB bus registered, assigned bus number 4
> Mar  5 15:48:51 newt kernel: [  289.844015] xHCI xhci_add_endpoint called for root hub
> Mar  5 15:48:51 newt kernel: [  289.844021] xHCI xhci_check_bandwidth called for root hub
> Mar  5 15:48:51 newt kernel: [  289.844084] hub 4-0:1.0: USB hub found
> Mar  5 15:48:51 newt kernel: [  289.844104] hub 4-0:1.0: 4 ports detected
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xen.org
> http://lists.xen.org/xen-devel

Re: Lose XHCI Device on HP Ivybridge While Resuming on Battery

[Tom Goetz]

On Mar 13, 2012, at 12:25 PM, Konrad Rzeszutek Wilk wrote:

> On Mon, Mar 05, 2012 at 11:11:10AM -0500, Tom Goetz wrote:
>> After resuming more than once on battery these HP Ivybridge laptops, the XHCI devices stop working. Have anyone seen this before? I wanted to check before diving in deeper. Let me know if you have any ideas. Thanks!
>> 
>> Facts
>> 	- Xen 4.0.3, Linux 3.2.7 PVOPs
>> 	- Happens on HP Ivybridge. Doesn't happen on very similar HP Sandybridge Clash system.
>> 	- Happens on battery, but not on AC.
> 
> Did you guys try to limit the C-state? say max_cstate=2 ?

Hi Konrad. Thanks for the reply. C-states don't matter. Don't know if you've seen my latest post.

http://lists.xen.org/archives/html/xen-devel/2012-03/msg00904.html

This seems to be something to do with not getting a wakeup interrupt for the xHCI device when it's in D3 if it has previously been in S3 and resumed. I'm starting to debug the Xen -> Dom0 interrupt path for that now.

Tom