IRP_MJ_CLEANUP assumption

Is there a chance for a non paging I/O write to still be processed after cleanup has completed ? In post cleanup , can I assume the file will not be modified using the same FileObject ? (again , excluding paging I/O) .

That's my assumption as well -- after post-cleanup returns, only paging IO writes are legal. Tagging along in case someone has a case that breaks this assumption.

And IRP_MJ_QUERY_FILE_INFORMATION - as I rediscovered for the third time last week

Reads also.
Who made a '20 character" minimum post rule :wink:

It came with the new software by default!

But I think it makes sense, because it stops people from posting replies that contain nothing but:

This.

Which (I admit) hasn't ever been a big issue in this forum anyhow.

Well... you asked :rofl:

Any time I need to remember what's allowed post-Cleanup I go to the FAT source code for the check it does (and have a laugh about some of these things being allowed "for Lou" :joy:):

Windows-driver-samples/filesys/fastfat/verfysup.c at main · microsoft/Windows-driver-samples · GitHub

Of course other file systems might allow other things, but FAT is always a good baseline.

1 Like

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.