On Tue, Jul 30, 2013 at 05:50:02PM -0700, Ethan Jackson wrote:
> > There are definitely ways around the thread safety analysis. My
> > feeling is that we should be pretty strict about it except in modules
> > where contention is a real problem.
> >
> >> Why does stp need a recursive mutex? I don'
> There are definitely ways around the thread safety analysis. My
> feeling is that we should be pretty strict about it except in modules
> where contention is a real problem.
>
>> Why does stp need a recursive mutex? I don't see any natural
>> recursion here.
So looking back at this, I'd like t
> Some of the functions take the mutex gratuitously, e.g. because they
> access immutable state. One example is stp_get_name(). I guess that
> these are not a big deal because this is not going to be performance
> critical, but at some point we'll hit some performance-critical case
> where we don
On Fri, Jul 26, 2013 at 06:07:03PM -0700, Ethan Jackson wrote:
> Signed-off-by: Ethan Jackson
In stp_unref(), the list_remove() needs to be guarded by the mutex.
Some of the annotations are only on the function prototypes and not
the definitions. Although the compiler is OK with this, it would
Signed-off-by: Ethan Jackson
---
lib/stp.c | 360 +---
lib/stp.h |3 +-
ofproto/ofproto-dpif.c |2 +-
tests/test-stp.c |2 +-
4 files changed, 280 insertions(+), 87 deletions(-)
diff --git a/lib/stp.c b/lib/
