Re: [PATCH v2 07/14] locks: convert to i_lock to protect i_flock list

2013-06-13 Thread Jeff Layton
On Thu, 13 Jun 2013 10:41:40 -0400 "J. Bruce Fields" wrote: > On Tue, Jun 11, 2013 at 07:09:01AM -0400, Jeff Layton wrote: > > Having a global lock that protects all of this code is a clear > > scalability problem. Instead of doing that, move most of the code to be > > protected by the i_lock

Re: [PATCH v2 07/14] locks: convert to i_lock to protect i_flock list

2013-06-13 Thread J. Bruce Fields
On Tue, Jun 11, 2013 at 07:09:01AM -0400, Jeff Layton wrote: > Having a global lock that protects all of this code is a clear > scalability problem. Instead of doing that, move most of the code to be > protected by the i_lock instead. > > The exceptions are the global lists that

Re: [PATCH v2 07/14] locks: convert to i_lock to protect i_flock list

2013-06-13 Thread J. Bruce Fields
On Tue, Jun 11, 2013 at 07:09:01AM -0400, Jeff Layton wrote: Having a global lock that protects all of this code is a clear scalability problem. Instead of doing that, move most of the code to be protected by the i_lock instead. The exceptions are the global lists that file_lock-fl_link sits

Re: [PATCH v2 07/14] locks: convert to i_lock to protect i_flock list

2013-06-13 Thread Jeff Layton
On Thu, 13 Jun 2013 10:41:40 -0400 J. Bruce Fields bfie...@fieldses.org wrote: On Tue, Jun 11, 2013 at 07:09:01AM -0400, Jeff Layton wrote: Having a global lock that protects all of this code is a clear scalability problem. Instead of doing that, move most of the code to be protected by

[PATCH v2 07/14] locks: convert to i_lock to protect i_flock list

2013-06-11 Thread Jeff Layton
Having a global lock that protects all of this code is a clear scalability problem. Instead of doing that, move most of the code to be protected by the i_lock instead. The exceptions are the global lists that file_lock->fl_link sits on. Those still need a global lock of some sort, so wrap just

[PATCH v2 07/14] locks: convert to i_lock to protect i_flock list

2013-06-11 Thread Jeff Layton
Having a global lock that protects all of this code is a clear scalability problem. Instead of doing that, move most of the code to be protected by the i_lock instead. The exceptions are the global lists that file_lock-fl_link sits on. Those still need a global lock of some sort, so wrap just