Is coinstater hard to debug from Vista?

NTDEV
1

Hi, folks,

Recently I found that the DIF_INSTALLREQUEST request of coinstaller is hard to debug from Windows Vista and later. We can’t debug into our specific codes of processing DIF_INSTALLREQUEST with WinDBG.exe, but we are sure that the codes take effect. And the weird thing is that the codes for DIF_REMOVE request could be debugged without any problem.

Has anybody encountered this symptom?

Thanks a bunch in advance.

All Best Regards,

Joseph

2

Have you tried coding a ‘hard’ breakpoint into the DllMain of your
CoInstaller and using a kernel debug session instead of attaching in UM?

Good Luck,
Dave Cattley

-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of luming419@163.com
Sent: Friday, October 29, 2010 2:27 AM
To: Windows System Software Devs Interest List
Subject: [ntdev] Is coinstater hard to debug from Vista?

Hi, folks,

Recently I found that the DIF_INSTALLREQUEST request of coinstaller is hard
to debug from Windows Vista and later. We can’t debug into our specific
codes of processing DIF_INSTALLREQUEST with WinDBG.exe, but we are sure that
the codes take effect. And the weird thing is that the codes for DIF_REMOVE
request could be debugged without any problem.

Has anybody encountered this symptom?

Thanks a bunch in advance.

All Best Regards,

Joseph

NTDEV is sponsored by OSR

For our schedule of WDF, WDM, debugging and other seminars visit:
http://www.osr.com/seminars

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

3

Thank you David. I will have a try on what you said.

So far, what I know is that the breakpoint for DIF_REMOVE could be debugged, while the breakpoint for DIF_INSTALLDEVICE not on Windows Server 2008 R2 x64. Do you know why the latest coinstaller is designed this way by Microsoft?

Thanks a lot.

Joseph