Red Hat Bugzilla – Bug 479453
Circular locking in NFS
Last modified: 2014-08-31 19:29:05 EDT
Created attachment 328579 [details]
dmesg after compile attempt
Description of problem:
When attempting to build firewire-ipv4 on a freshly booted system, I get the attached error.
Version-Release number of selected component (if applicable):
This has been pretty reliable for a few kernel versions now.
Steps to Reproduce:
2.log in; cd n; rm *.o *.ko; make -f Makefile.stand
3.Observe error in dmesg
Attached circular locking error
Apparently this is a false alarm. Looks like peterz has a patch to silence the warnings that should be going upstream soon. Until then it appears that the warning can be ignored...
See discussion on linux-nfs upstream list for more info.
did that patch get merged? Worth keeping this open?
Yes. It did get merged:
Author: Peter Zijlstra <email@example.com>
Date: Mon Jan 12 13:02:11 2009 +0100
lockdep, mm: fix might_fault() annotation
Some code (nfs/sunrpc) uses socket ops on kernel memory while holding
the mmap_sem, this is safe because kernel memory doesn't get paged out,
therefore we'll never actually fault, and the might_fault() annotations
will generate false positives.
Reported-by: "J. Bruce Fields" <firstname.lastname@example.org>
Signed-off-by: Peter Zijlstra <email@example.com>
Signed-off-by: Ingo Molnar <firstname.lastname@example.org>
...closing as CURRENTRELEASE (please correct that if it's not the right resolution)...