[Ksummit-2008-discuss] topic proposal

Roland Dreier rdreier at cisco.com
Wed Aug 27 08:19:13 PDT 2008


 > > Take a look at, say, drivers/net/mlx4/cmd.c.  That uses a semaphore to
 > > count how many firmware commands are outstanding, and block until a
 > > command completes when the maximum number of commands are already
 > > queued.  Which of your cases does that fall into?  It's not a mutex,
 > > it's not a completion, and I really don't think it's anything that's
 > > wild or un-maintainable.

 > The poll_sem looks like a mutex, and the event_sem falls into the same
 > case as what I outlined to Greg, which would be a completion (with some
 > modifications to completions) ..

Except that poll_sem can't be converted to a mutex because it is
locked/unlocked from different threads.  And it's not clear to me what
the advantage of changing event_sem to a counting completion would be,
even if such objects were implemented -- how does a counting completion
differ from a semaphore really?

 - R.


More information about the Ksummit-2008-discuss mailing list