WinDbg & 1394

NTDEV
1

I’m having trouble getting WinDbg set up for 1394 debugging. I understand
about the first-time failure stuff. I gotten 1394 working many times
before.

Here’s the message I get during the first run:

Using 1394 for debugging
Could not update 1394 device drivers - error e000020b
Failed to open 1394 channel 32

I’m running the latest WinDbg on Vista x86. The WinDBG installer and WinDbg
run as administrator and I’ve tried various combinations of compatibility
emulation. Any ideas? Does it matter which brand of 1394 controller is
present?

-Shawn

2

What version of WinDbg are you using? What is the host OS?

>> xxxxx@gmail.com 2007-05-07 19:08 >>>
I’m having trouble getting WinDbg set up for 1394 debugging. I
understand
about the first-time failure stuff. I gotten 1394 working many times
before.

Here’s the message I get during the first run:

Using 1394 for debugging
Could not update 1394 device drivers - error e000020b
Failed to open 1394 channel 32

I’m running the latest WinDbg on Vista x86. The WinDBG installer and
WinDbg
run as administrator and I’ve tried various combinations of
compatibility
emulation. Any ideas? Does it matter which brand of 1394 controller
is
present?

-Shawn

Questions? First check the Kernel Driver FAQ at
http://www.osronline.com/article.cfm?id=256

To unsubscribe, visit the List Server section of OSR Online at
http://www.osronline.com/page.cfm?name=ListServer

3

Shawn Brooks wrote:

I’m having trouble getting WinDbg set up for 1394 debugging. I
understand about the first-time failure stuff. I gotten 1394 working
many times before.

Here’s the message I get during the first run:

Using 1394 for debugging
Could not update 1394 device drivers - error e000020b
Failed to open 1394 channel 32

I’m running the latest WinDbg on Vista x86. The WinDBG installer and
WinDbg run as administrator and I’ve tried various combinations of
compatibility emulation. Any ideas? Does it matter which brand of
1394 controller is present?

You’re absolutely certain this machine has a 1394 controller? E000020B
is ERROR_NO_SUCH_DEVINST.

–
Tim Roberts, xxxxx@probo.com
Providenza & Boekelheide, Inc.

4

>> What version of WinDbg are you using? What is the host OS?

I’m running the latest WinDbg on Vista x86.

> You’re absolutely certain this machine has a 1394 controller? E000020B
is ERROR_NO_SUCH_DEVINST.

Yeah, I noticed that too. Yes, there is a 1394 controller on board. It
shows up in Device Manager without error.

5

I’ve had problems connecting 6.7.0.5 to Vista (x64) using 1394, and
Microsoft has quasi-confirmed this. My issues (failure to break in) are
not obviously related to yours, so I haven’t anything to suggest other
than to try the previous version of WinDbg (6.6.7.5).

mm

>> xxxxx@gmail.com 2007-05-07 19:55 >>>
> What version of WinDbg are you using? What is the host OS?

I’m running the latest WinDbg on Vista x86.

> You’re absolutely certain this machine has a 1394 controller?
E000020B
is ERROR_NO_SUCH_DEVINST.

Yeah, I noticed that too. Yes, there is a 1394 controller on board.
It
shows up in Device Manager without error.

Questions? First check the Kernel Driver FAQ at
http://www.osronline.com/article.cfm?id=256

To unsubscribe, visit the List Server section of OSR Online at
http://www.osronline.com/page.cfm?name=ListServer

6

I tried 6.6.7.5. Same problem.

On 5/7/07, Martin O’Brien wrote:
>
> I’ve had problems connecting 6.7.0.5 to Vista (x64) using 1394, and
> Microsoft has quasi-confirmed this. My issues (failure to break in) are
> not obviously related to yours, so I haven’t anything to suggest other
> than to try the previous version of WinDbg (6.6.7.5).
>

7

no ideas then. sorry.

>> xxxxx@gmail.com 2007-05-07 20:26 >>>
I tried 6.6.7.5. Same problem.

On 5/7/07, Martin O’Brien wrote:
>
> I’ve had problems connecting 6.7.0.5 to Vista (x64) using 1394, and
> Microsoft has quasi-confirmed this. My issues (failure to break in)
are
> not obviously related to yours, so I haven’t anything to suggest
other
> than to try the previous version of WinDbg (6.6.7.5).
>

—
Questions? First check the Kernel Driver FAQ at
http://www.osronline.com/article.cfm?id=256

To unsubscribe, visit the List Server section of OSR Online at
http://www.osronline.com/page.cfm?name=ListServer

8

Are you running with admin rights? Are you sure that WinDbg is running with those elevated rights? On first install, WinDbg needs to install the 1394 drivers for the debugger. (Not the 1394 bus driver, the 1394 protocol driver for the debugger.) After that first install, you *might* be able to run WinDbg as a normal user, I don’t know.

From: xxxxx@lists.osr.com [mailto:xxxxx@lists.osr.com] On Behalf Of Shawn Brooks
Sent: Monday, May 07, 2007 7:09 PM
To: Windows System Software Devs Interest List
Subject: [ntdev] WinDbg & 1394

I’m having trouble getting WinDbg set up for 1394 debugging. I understand about the first-time failure stuff. I gotten 1394 working many times before.

Here’s the message I get during the first run:

Using 1394 for debugging
Could not update 1394 device drivers - error e000020b
Failed to open 1394 channel 32

I’m running the latest WinDbg on Vista x86. The WinDBG installer and WinDbg run as administrator and I’ve tried various combinations of compatibility emulation. Any ideas? Does it matter which brand of 1394 controller is present?

-Shawn
— Questions? First check the Kernel Driver FAQ at http://www.osronline.com/article.cfm?id=256 To unsubscribe, visit the List Server section of OSR Online at http://www.osronline.com/page.cfm?name=ListServer

9

What does the target 1394 say in Device manager? The vendor’s driver should be yellow banged.

–
The personal opinion of
Gary G. Little
“Shawn Brooks” wrote in message news:xxxxx@ntdev…
>> What version of WinDbg are you using? What is the host OS?

I’m running the latest WinDbg on Vista x86.

>> You’re absolutely certain this machine has a 1394 controller? E000020B
is ERROR_NO_SUCH_DEVINST.

Yeah, I noticed that too. Yes, there is a 1394 controller on board. It shows up in Device Manager without error.

10

Is this a PCMCIA based host controller? I have had a LOT of problems with several of those. In fact, I have one Adaptec PCMCIA 1394 card that works great and another Adaptec card that doesn’t work at all.

I would suggest trying another card if that is the case.

Bill M.
“Shawn Brooks” wrote in message news:xxxxx@ntdev…
I tried 6.6.7.5. Same problem.

On 5/7/07, Martin O’Brien wrote:
I’ve had problems connecting 6.7.0.5 to Vista (x64) using 1394, and
Microsoft has quasi-confirmed this. My issues (failure to break in) are
not obviously related to yours, so I haven’t anything to suggest other
than to try the previous version of WinDbg (6.6.7.5).

11

>> Are you running with admin rights? Are you sure that WinDbg is running
with those elevated rights? On first install, WinDbg needs to install the
1394 drivers for the debugger. (Not the 1394 bus driver, the 1394 protocol
driver for the debugger.)

Yes to both questions. I’ve turned UAC off and use “run as administrator”
for both the WinDBG install app and WinDBG itself.

> What does the target 1394 say in Device manager? The vendor’s
driver should be yellow banged.

The 1394 chip is from VIA and is operational. It’s not banged. I don’t
understand why you say it should be banged. Other systems I have WinDBG on
don’t have banged-out 1394 adapters.

> Is this a PCMCIA based host controller? I have had a LOT of problems
with several of those. In fact, I have one Adaptec PCMCIA 1394 card that
works great and another Adaptec card that doesn’t work at all.

Hey Bill. No, the 1394 chip is on the motherboard. However, it isn’t the
ubiquitous TI 1394 chip, so I’m wondering if that might be the key to the
problem. I’m going to scrounge a TI-based 1394 PCI card and see if that
works tonight. I’m wondering if the WinDbg driver installer simply doesn’t
recognize the VIA chip.

BTW, thanks for all the responses.

12

Yeah…it is probably just a chip problem…but that is a SWAG (silly wild ass guess). My take is to find a controller that works and never venture from it :slight_smile: Sorry to not be more help.

Bill M.
“Shawn Brooks” wrote in message news:xxxxx@ntdev…
>> Are you running with admin rights? Are you sure that WinDbg is running with those elevated rights? On first install, WinDbg needs to install the 1394 drivers for the debugger. (Not the 1394 bus driver, the 1394 protocol driver for the debugger.)

Yes to both questions. I’ve turned UAC off and use “run as administrator” for both the WinDBG install app and WinDBG itself.

>> What does the target 1394 say in Device manager? The vendor’s driver should be yellow banged.

The 1394 chip is from VIA and is operational. It’s not banged. I don’t understand why you say it should be banged. Other systems I have WinDBG on don’t have banged-out 1394 adapters.

>> Is this a PCMCIA based host controller? I have had a LOT of problems with several of those. In fact, I have one Adaptec PCMCIA 1394 card that works great and another Adaptec card that doesn’t work at all.

Hey Bill. No, the 1394 chip is on the motherboard. However, it isn’t the ubiquitous TI 1394 chip, so I’m wondering if that might be the key to the problem. I’m going to scrounge a TI-based 1394 PCI card and see if that works tonight. I’m wondering if the WinDbg driver installer simply doesn’t recognize the VIA chip.

BTW, thanks for all the responses.

13

>> What does the target 1394 say in Device manager? The vendor’s driver should be yellow banged.

?
The 1394 chip is from VIA and is operational. It’s not banged.
I don’t understand why you say it should be banged. Other systems
I have WinDBG on don’t have banged-out 1394 adapters.

I think he means on the target machine, not the debugger machine. You certainly don’t want the 1394 board on your debugger machine disabled or non-functional.

Yes to both questions. I’ve turned UAC off and use “run as administrator” for both the WinDBG install app and WinDBG itself.

Ok, good. Now open device manager on your debugger (host) machine. Set the view to “Devices by connection”. Find your 1394 adapter. Under it, you should see several devices enumerated by the 1394 protocol stack. One of them should be “1394 Windows Debug Driver(Kernel Mode)”. Is that present? Is it working?

From: xxxxx@lists.osr.com [mailto:xxxxx@lists.osr.com] On Behalf Of Shawn Brooks
Sent: Tuesday, May 08, 2007 2:56 PM
To: Windows System Software Devs Interest List
Subject: Re: [ntdev] WinDbg & 1394

> Are you running with admin rights? Are you sure that WinDbg is running with those elevated rights? On first install, WinDbg needs to install the 1394 drivers for the debugger. (Not the 1394 bus driver, the 1394 protocol driver for the debugger.)

Yes to both questions. I’ve turned UAC off and use “run as administrator” for both the WinDBG install app and WinDBG itself.

> What does the target 1394 say in Device manager? The vendor’s driver should be yellow banged.

The 1394 chip is from VIA and is operational. It’s not banged. I don’t understand why you say it should be banged. Other systems I have WinDBG on don’t have banged-out 1394 adapters.

> Is this a PCMCIA based host controller? I have had a LOT of problems with several of those. In fact, I have one Adaptec PCMCIA 1394 card that works great and another Adaptec card that doesn’t work at all.

Hey Bill. No, the 1394 chip is on the motherboard. However, it isn’t the ubiquitous TI 1394 chip, so I’m wondering if that might be the key to the problem. I’m going to scrounge a TI-based 1394 PCI card and see if that works tonight. I’m wondering if the WinDbg driver installer simply doesn’t recognize the VIA chip.

BTW, thanks for all the responses.
— Questions? First check the Kernel Driver FAQ at http://www.osronline.com/article.cfm?id=256 To unsubscribe, visit the List Server section of OSR Online at http://www.osronline.com/page.cfm?name=ListServer

14

Did you check the version after you installed in help about ? From your
error message it looks I have other problems than you but I also cannot get
the latest WinDbg to work with firewire connections on Vista while it has
always been working. After some desperate moves like reversing the firewire
cable, disabling USB legacy emulation in the Bios, getting a card with real
TI chips I decided to uninstall the latest Windbg and reinstall the old
version bit it still displays 6.7, the latest version … Now need to run
some registry and file monitor to find out how to really uninstall Windbg to
allow me to install the old one again, I already have too much work to do
and instead I would like to spend my time on something more fruitful.

/Daniel

“Shawn Brooks” wrote in message news:xxxxx@ntdev…
I tried 6.6.7.5. Same problem.

On 5/7/07, Martin O’Brien wrote:
I’ve had problems connecting 6.7.0.5 to Vista (x64) using 1394, and
Microsoft has quasi-confirmed this. My issues (failure to break in) are
not obviously related to yours, so I haven’t anything to suggest other
than to try the previous version of WinDbg (6.6.7.5).

15

Well, I inadvertently fixed the problem–at least for a while. I booted up
this morning and noticed the “safely remove device” icon on my tray. It’s
normally not there, so I clicked it and discovered it was the 1394 device in
question being controlled by the WinDBG driver. WinDBG loaded right up and
I got to work debugging.

I rebooted the machine the and now it’s not working again–with the original
error.

I really don’t know what to make of it.

On 5/9/07, Daniel Terhell wrote:
>
> Did you check the version after you installed in help about ? From your
> error message it looks I have other problems than you but I also cannot
> get
> the latest WinDbg to work with firewire connections on Vista while it has
> always been working. After some desperate moves like reversing the
> firewire
> cable, disabling USB legacy emulation in the Bios, getting a card with
> real
> TI chips I decided to uninstall the latest Windbg and reinstall the old
> version bit it still displays 6.7, the latest version … Now need to run
> some registry and file monitor to find out how to really uninstall Windbg
> to
> allow me to install the old one again, I already have too much work to do
> and instead I would like to spend my time on something more fruitful.
>
> /Daniel
>
>
> “Shawn Brooks” wrote in message news:xxxxx@ntdev…
> I tried 6.6.7.5. Same problem.
>
>
> On 5/7/07, Martin O’Brien wrote:
> I’ve had problems connecting 6.7.0.5 to Vista (x64) using 1394, and
> Microsoft has quasi-confirmed this. My issues (failure to break in) are
> not obviously related to yours, so I haven’t anything to suggest other
> than to try the previous version of WinDbg (6.6.7.5).
>
>
> —
> Questions? First check the Kernel Driver FAQ at
> http://www.osronline.com/article.cfm?id=256
>
> To unsubscribe, visit the List Server section of OSR Online at
> http://www.osronline.com/page.cfm?name=ListServer
>

16

> emulation. Any ideas? Does it matter which brand of 1394 controller is

present?

It must be OHCI1394-compatible on target, so, it’s better to have Texas
Instruments chip.

Boot the target without debugger, go to Device Manager and look at what driver
was loaded for 1394 chip. If it is OHCI1394.SYS - then you’re fine.

–
Maxim Shatskih, Windows DDK MVP
StorageCraft Corporation
xxxxx@storagecraft.com
http://www.storagecraft.com

17

> >> Are you running with admin rights? Are you sure that WinDbg is running

with those elevated rights? On first install, WinDbg needs to install the
1394 drivers for the debugger. (Not the 1394 bus driver, the 1394 protocol
driver for the debugger.)

The usual Windows 1394 stack on host + possibly some WinDbg’s kernel-mode
module to talk to 1394bus from user mode (out-of-the-box 1394bus does not
support it).

KD1394.DLL on target.

–
Maxim Shatskih, Windows DDK MVP
StorageCraft Corporation
xxxxx@storagecraft.com
http://www.storagecraft.com