W10 Driver load fails with Problem Code 0x27 (39) Status (0xc0000034)

[Rick Davis, Abaco Systems]

Greetings all!



I have a bit of a mystery to solve, and could use some guidance. We have a (video codec) PCIe board which works well on Windows XP and Windows 7. I've created a Windows 10-specific driver, but it refuses to install, yielding:



Problem code 0x27

Problem status 0xC0000034



which translates loosely to "Object Name not found."



The generation of the new driver differs in several ways from prior builds, e.g.

• Some kernel calls tweaked for the post Windows 8.x era

• Built and (test) signed in Visual Studio 2015 (vs. prior WDK 7006.xxx toolstack)

• Targeted directly at Windows 10's KMDF version 1.15 (so no coinstaller accompanies it.)

• Built as a Universal Driver (though doesn't install when built as a Desktop driver either.)

• stampinf and an INX file used to yield the deployable INF file

The driver install fails whether via dpinst, pnputil, "Update driver" in the Device Manager, or under the Kernel Debugging Tools for Windows in Visual Studio 2015. The messages from this later method are perhaps the most illuminating:




WDTF_TEST : Install driver package onto device


Start Test


3/16/2016 05:57:34.170 PM


WDTF_TARGETS : - Query("HardwareIDs='PCI\VEN_11B5&DEV_0037' OR CompatIDs='PCI\VEN_11B5&DEV_0037'")


Message


3/16/2016 05:57:34.204 PM


WDTF_TARGETS : Target: Abaco Systems GESPA Device Driver Disk - 0037 PCI\VEN_11B5&DEV_0037&SUBSYS_858010EE&REV_01\4&3F76365&0&0009


Message


3/16/2016 05:57:34.219 PM


WDTF_TARGETS : - GetInterfaces("DriverSetup")


Message


3/16/2016 05:57:34.234 PM


WDTF_TARGETS : Target: Abaco Systems GESPA Device Driver Disk - 0037 PCI\VEN_11B5&DEV_0037&SUBSYS_858010EE&REV_01\4&3F76365&0&0009


Message


3/16/2016 05:57:34.248 PM


WDTF_DRIVER_SETUP_DEVICE : - UpdateDriver()


Message


3/16/2016 05:57:34.262 PM


WDTF_DRIVER_SETUP_DEVICE : Target: Abaco Systems GESPA Device Driver Disk - 0037 PCI\VEN_11B5&DEV_0037&SUBSYS_858010EE&REV_01\4&3F76365&0&0009


Message


3/16/2016 05:57:34.277 PM


WDTF_TARGETS : - Query("INF::FileName='oem13.inf'")


Message


3/16/2016 05:57:47.903 PM


WDTF_TEST : Failed to install driver package on any device

File: , Line: -1

Error Type:

Error Code: 0x0

Error Text: Error 0x00000000


Error


3/16/2016 05:57:47.934 PM


WDTF_DRIVER_SETUP_SYSTEM : - SnapTriageLogs()


Message


3/16/2016 05:57:47.959 PM


WDTF_DRIVER_SETUP_SYSTEM : Target: BLACKBIRD


Message


3/16/2016 05:57:47.973 PM


WDTF_TEST : Install driver package onto device

Result: Fail


End Test


3/16/2016 05:57:48.004 PM


The messages from pnputil differ, but the net result is the same: pnputil reports the driver installation was successful:



F:\W10\MsSignedW10x64>pnputil -i -a gefspa.inf

Microsoft PnP Utility



Processing inf : gefspa.inf

Successfully installed the driver on a device on the system.

Driver package added successfully.

Published name : oem2.inf





Total attempted: 1

Number successfully imported: 1



Driver startup fails the same way, with the error code 39, and the accompanying “Object Name not found” message.



My guess is that there's a failure to dynamically load / link the driver module (gefspa.sys) but I haven't figured out a way to clarify what, exactly, is upsetting the loader. Depend.exe output for my gefspa.sys looks identical to drivers that actually work.



Left to my own devices, my next step is to build, install, and "debug" one of the sample drivers (e.g. echo) and contrast how that exercise differs from gefspa, but any suggestions / experiments to try would be welcomed. Thanks!












Rick Davis Windows System Software GE Intelligent Platforms

Continue reading...