I am running Windows XP Home hosted via vmware running on Fedora Core 6 linux. Formerly, I was able to run the LogCard Utility (V5.01) and connect to the PM3 running firmware V82 without incident. After upgrading the PM3 to firmware V97, however, I can no longer connect. I instead get an error (reported by XP) to the effect that the USB device is damaged and no connection is possible. This happens regardless of what cable and/or USB port I use.
Everything worked fine when I set it up on a native XP Home machine (running the exact same version of XP Home). Also, all other USB devices continue to work as usual on my vmware-instance of XP so everthing seems fine at that end as well.
Questions: Has anyone else tried this combination and had a similar problem? Does anyone have it working? Is it possible to get the USB device file for the PM3 so I can attempt to reinstall it? Any suggestions would be greatly appreciated.
Daniel
Can't connect to PM3 (linux/vmware/XP Home)
Parishd:
There were updates to the firmware in the USB handling between 82 and 97 to make better compatibility with Mac OS X systems. I have no idea what is causing the errors for you at this time, the setup you have is not formally supported.
The "device file" (driver?) is the built in HID driver.
I would also check to see that your kernel is up to date. There are kernel updates that were approved a year or so ago by Linus in the HID section. These fixes related to Linux not behaving properly with long report types.
You can email me directly and we can work with the developers and see if we can figure this out.
There were updates to the firmware in the USB handling between 82 and 97 to make better compatibility with Mac OS X systems. I have no idea what is causing the errors for you at this time, the setup you have is not formally supported.
The "device file" (driver?) is the built in HID driver.
I would also check to see that your kernel is up to date. There are kernel updates that were approved a year or so ago by Linus in the HID section. These fixes related to Linux not behaving properly with long report types.
You can email me directly and we can work with the developers and see if we can figure this out.
The firmware upgrades to the PM fixed problems with device initialization on Linux. Before the upgrades, Linux would completely ignore the device. Now that the PM3 is linux-capable, it's probably preventing VMWare from accessing it. If the Linux kernel opens the monitor exclusively, then VMWare can see the hardware, but it can't open the device to communicate with it.
You could try a modprobe -r usbhid to unload the Linux HID driver after plugging in the monitor, but before starting VMWare. It will break your USB mouse, but it will identify whether it's the Linux kernel that has your PM locked up.
If you find that it IS the Linux kernel that's causing VMWare problems, it might be possible to write a small C program to detach the HID driver from just the PM3/4 and not your mouse. I think I remember the libhid people having something like this. You might look into the command-line program: libhid-detach-device.
Mike
You could try a modprobe -r usbhid to unload the Linux HID driver after plugging in the monitor, but before starting VMWare. It will break your USB mouse, but it will identify whether it's the Linux kernel that has your PM locked up.
If you find that it IS the Linux kernel that's causing VMWare problems, it might be possible to write a small C program to detach the HID driver from just the PM3/4 and not your mouse. I think I remember the libhid people having something like this. You might look into the command-line program: libhid-detach-device.
Mike
c2scott, Thanks for the offer to help and I know this is an unusual configuration and not officially supported. While not the latest, I'm at kernel 2.6.18.2 and on the linux side all usb devices (I've tried) behave as they should. I was thinking of a device file for the windows side, not linux, since this is the one that is reporting the error. I'll follow up by email after I've had a chance to look into Mike's suggestions.
Mike, The PM3 is in fact recognized and reported properly by linux. While I believe this is necessary in order that the vmware instance of XP to recognize it, I will test this out later this morning. In fact, vmware does have a problem with hi-speed usb devices, which I disable when using usb2 devices with rmmod ehci_hcd, thus forcing them to use the slower usb1 mode. This has worked for all devices I've tried other than the PM3. Also, (at least my) 2.4.18 kernel uses the modules ohci_hcd and ehci_hcd for usb; there is no usbhid. Was this introduced after 2.4.18?
Thanks again to you both and I will be back later today after I've had a chance to investigate a bit further.
Daniel
Mike, The PM3 is in fact recognized and reported properly by linux. While I believe this is necessary in order that the vmware instance of XP to recognize it, I will test this out later this morning. In fact, vmware does have a problem with hi-speed usb devices, which I disable when using usb2 devices with rmmod ehci_hcd, thus forcing them to use the slower usb1 mode. This has worked for all devices I've tried other than the PM3. Also, (at least my) 2.4.18 kernel uses the modules ohci_hcd and ehci_hcd for usb; there is no usbhid. Was this introduced after 2.4.18?
Thanks again to you both and I will be back later today after I've had a chance to investigate a bit further.
Daniel
I'm familiar with Debian distributions around the 2.6.9-2.6.11 kernel release. The usbhid driver should be around in almost all 2.6.x kernels. Odds are your kernel has it compiled-in rather than a module. A grep of dmesg should indicate whether it's loading it or not. If it's not being loaded as a module then trying to remove it won't help.
While Linux should detect both a FW82 and a FW97 PM3, there will be differences in its level of support. It might have detected it and initialized it as a USB device on the bus, but not loaded the additional HID driver necessary to communicate with it. This left the device present and open for VMWare to access using basic USB protocols and the Win32 drivers.
I would look into VMWare too and see if it has any options for USB devices. You might be able to indicate that the device be 'detached' from the kernel driver vs an HID mouse which should NOT be detached from the kernel.
Mike
While Linux should detect both a FW82 and a FW97 PM3, there will be differences in its level of support. It might have detected it and initialized it as a USB device on the bus, but not loaded the additional HID driver necessary to communicate with it. This left the device present and open for VMWare to access using basic USB protocols and the Win32 drivers.
I would look into VMWare too and see if it has any options for USB devices. You might be able to indicate that the device be 'detached' from the kernel driver vs an HID mouse which should NOT be detached from the kernel.
Mike
After looking into this a bit further, I'm reasonably certain that this is a problem with vmware 5.5.3, which I already know to have issues with at least the hi-speed usb and possibly with usb1 as well, although this is the first time I have not been able to find a workaround.
To summarize, the PM3 is recognized as a usb1 device by Fedora Core 6 linux running kernel 2.6.18.2 and is properly reported by the diagnostic utilities. Vmware also recognizes the device and lists it in the '/VM/Removable Devices' menu. Regardless of being properly recognized on the linux side, XP always sees the device then reports that it has malfunctioned, and no connection is established.
Scott, I need to look into the vmware side more before I'm comfortable taking advantage of your offer to look into it. Thanks!
And Mike, thanks for your suggestions.
Daniel
To summarize, the PM3 is recognized as a usb1 device by Fedora Core 6 linux running kernel 2.6.18.2 and is properly reported by the diagnostic utilities. Vmware also recognizes the device and lists it in the '/VM/Removable Devices' menu. Regardless of being properly recognized on the linux side, XP always sees the device then reports that it has malfunctioned, and no connection is established.
Scott, I need to look into the vmware side more before I'm comfortable taking advantage of your offer to look into it. Thanks!
And Mike, thanks for your suggestions.
Daniel