Re: WinUSB resume
- From: "Brandon Wilson [MSFT]" <brandonw@xxxxxxxxxxxxxxxxxxxx>
- Date: Tue, 11 Aug 2009 10:58:34 -0700
It sounds like you aren't using UMDF at all, so the host automatic host restart would not be a part of the picture. Conceptually there are two types of stacks you could use WinUSB 1) WinUSB->UMDF framework->Your UMDF Driver Code or 2) WinUSB->Your user-mode service/application. In the second case which it seems like you are using, UMDF is not involved.
I'm not really an expert in USB/WinUSB so I'm not sure what the issue is. I'll leave it to one of the more knowledgeable people to chime in here.
"Dean" <Dean@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote in message news:B8AA42B5-1E32-4C09-8EC1-C4608E201720@xxxxxxxxxxxxxxxx
Hi, again. I do more test. If I add a 1.5K resistance between power and.
ground, I monitor the resume work fine. WinUSB driver won't restart my usb
device anymore.
Resistance current in suspend mode resume from suspend
0K 254uA fail
1K 6mA OK
1.5K 3.6mA OK
1.6K 2.6mA fail
2K 2.2mA fail
I don't know why, can anyone explain?
"Dean" wrote:
Thank you very much. I'm using WinUSB as my usb device driver only, I don't
write any driver source code. I'm afraid there is a bug in WinUSB's resume
path that caused a crash in the host process. Can you teach me how to debug
WinUSB?
I'd better describe more about my test.
Platform:
Windows Vista sp2, WinDDK 7600.16385.0, Intel Core2 Duo CPU E8300, 4GB RAM,
32 bit operating system.
WinUSB settings:
HKR,,"SystemWakeEnabled",0x00010001,1
HKR,,"DeviceIdleEnabled",0x00010001,1
HKR,,"DeviceIdleIgnoreWakeEnable",0x00010001,1
HKR,,"UserSetDeviceIdleEnabled",0x00010001,1
HKR,,"DefaultIdleState",0x00010001,0x1
HKR,,"DefaultIdleTimeout",0x00010001,2000
Scenario:
Plug in a usb device to my PC, I monitor the current of the usb device is
25mA. More than 2 seconds later, the usb device selective suspend, the
current is power down to 254uA. I try a bulk transfer command
WinUsb_ReadPipe(), the usb device has been detached and attached
automatically. I can hear
Windows new device arrival notification sound, it appear as a restart to the
end user.
"Brandon Wilson [MSFT]" wrote:
> UMDF has a feature where if the user-mode driver crashes, it will
> automatically restart the device and re-load the the driver. If there > was a
> bug in your driver's resume path that caused a crash in the host > process,
> framework would automatically recover and it would just appear as a > restart
> to the end user. USB resume is supported in UMDF/WinUSB so it is > likely
> that the host process is crashing and automatically being restarted.
>
> If a debugger is attached to the host process, the crash should be > caught
> there so you can debug further. The WdfVerifier tool in the Win7 WDK
> provides a nice GUI interface to configure all the necessary values for
> debugging.
- Follow-Ups:
- Re: WinUSB resume
- From: Dean
- Re: WinUSB resume
- References:
- WinUSB resume
- From: Dean
- Re: WinUSB resume
- From: Brandon Wilson [MSFT]
- Re: WinUSB resume
- From: Dean
- Re: WinUSB resume
- From: Dean
- WinUSB resume
- Prev by Date: Re: How to make a raw PDO device accessible from user mode code in con
- Next by Date: Re: OT: MSDN managed answers on this newsgroup
- Previous by thread: Re: WinUSB resume
- Next by thread: Re: WinUSB resume
- Index(es):
Relevant Pages
|