Discussion in 'Windows 7' started by Enigma256, Feb 6, 2011.
You need to login to view this posts content.
I just did an audit (though not a very thorough one), and it looks like winusb.sys is affected too, so I've edited the first post to reflect that. However, it looks like these are the only three files affected by this sort of installer bug, so I guess that's good...
Thanks for this very useful information.
I would bet that on day of official release of SP1, there will be a critical update pushed to Windows Update to address this. At least you would hope so.
Just thougt i'd add, I used the sp1 installer and it updated all my usb drivers. (I just checked after reading your post) I suggest people check before they do this.
I think you are reading the driver version in the device manager. That will report 7601 because that version comes from whateve the INF says. The driver file details in the device manager, for reasons that I elaborated on in the other thread, do not always report versions correctly. I am referring to the versions of the driver files themselves, located in system32\drivers, and checked using Explorer (and not the device manager).
I manage multiple live machines, different hardware, some 32-bit, some 64-bit. Every single one of them exhibits this problem. This problem is also reliably and consistently reproducible on some test VPCs that I have, and I have a friend who reports the exact same problem.
Strange, of the 3 files i did not find "winusb.sys" in system32\drivers folder; however it is found in winsxs repository. Is there anyone else having this observation? The system is a fresh install of win7 pro from the official integrated x86 iso.
That's normal. Not all hardware configurations need or use winusb.sys, so it may not be active on your system. This is why my batch file has an "if exist" conditional check for winusb.sys.
Ohh, i see. The rig is just a netbook utilized as a test bed for new releases of win7.
on my laptop the files you've talked about are all the newer sp1 versions, but after checking on my desktop they are still rtm versions.
Good share Enigma. I experienced those usb controller issues myself.
A very detailed initial post.
i tried/tested your observations and on:
x2 virtual systems with SP1.exe applied = 1 file was copied/updated on each,
on a real laptop with SP1.exe applied 3 files copied/updated.
i compared file info with a real PC with Nvidia controller (with the KB976972 update applied), but no SP1 used.
And your finding hold true.
Thank you, for finding issue and resolution.
On my system,
Biostar MB with NVIDIA GeForce 6150 / nForce 430 chipset,
Windows7 Ultimate x86 ITA updated to Service Pack 1 by windows6.1-KB976932-X86.exe file,
usbport.sys and usbehci.sys not updated, still 6.1.7600.16445 version
winusb.sys NOT present (???).
On a Virtualbox machine, hosted on the same PC
Windows7 Ultimate Service Pack 1 x86 ITA (installed with an original ISO slipstreamed with the Service Pack 1 file)
usbport.sys and usbehci.sys updated to 6.1.7601.17514 version
winusb.sys NOT present (???).
Good, just did the batch file method, very useful post.
See post #7; this is normal--not all systems need/use winusb.sys, and the batch file does take this into account.
Thanks for that info; so it looks like self-slipstreamed ISOs are okay...
1. 01.25 install windows6.1-KB976932-x86-INTL.exe
2. 01.28 clean up
3. working with usb as usual
4. 02.07 read your thread, check my files - all three NEW version.
5. But before SP1 installed all hotfixes from Solor's repository.
No problems here.
All files were properly updated.
Is the described bug applying to this http://forums.mydigitallife.net/thr...p-image-causing-system-corruption.?highlight= issue posted by me earlier?
It was Standard extended host PCI to USB adapter not starting automatically after doing sp1 cleanup steps, the version reported by devmgr was 6.1.7601.17514 (21.6.2006).
Now I did clean reinstall and the issue doesnot occure anymore.
Yea, it sounds like a possibility. Especially the part about "notifications about not working at full speed", since those were the error messages that I got that tipped me off to something being seriously wrong.
Hmm, I wonder if the bug with the installer only happens with certain USB controllers. I have seen it happen with 100% reproducible consistency with Intel, NVIDIA, and the emulated controller used inside VirtualBox. I wonder if you and the other people who report not encountering the bug are using a different controller? The other possibility is, when Windows reinstalls a controller that was deleted from the device manager, it essentially repaired the problem. Maybe there is another way to trigger a similar sort of effect, and you happened to have triggered it somehow? (though, if there is another way, I have no idea what it might be, since I tried several ways to try to get Windows to fix itself and none of them worked)
Windows7 Ultimate x86 ITA updated to Service Pack 1 by windows6.1-KB976932-X86.exe
all three files are updated in windows\system32\drivers
The only file you are correct about is the usbehci.sys, the others are all SP1 versions, Devmgr reports the incorrect version for these files for some reason.
The same incorrect version reporting occurs for partmgr when you check the drivers for each hdd, SP0 partmgr.sys version is reported as installed but only the SP1 partmgr.sys version can be found on the system.