A huge pile of SimpleMessenger refactors, cleanups, and fixes (wip-msgr)
was just merged into master, and this code was changed around quite a bit.
Did you have a specific test case that was triggering the problem? If so,
do you mind running it on the current master?
Most of the cleanups/fixes were motivated by testing with socket failure
injection enabled (e.g., 'ms inject socket failures = 500' to simulate a
failure on every ~500th socket read or write). This is being added to the
rados regression test suite.
Thanks!
sage
On Tue, 14 Aug 2012, Alexandre Oliva wrote:
> Don't wait for a signal if the dispatch queue is non-empty already,
> and support a non-NULL pipe for local responses. If we were to
> require pipe to be NULL for that, the response would be lost.
>
> Signed-off-by: Alexandre Oliva <oliva@xxxxxxxxxxxxxxxxx>
> ---
> src/msg/SimpleMessenger.cc | 4 ++--
> 1 files changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/src/msg/SimpleMessenger.cc b/src/msg/SimpleMessenger.cc
> index 53e36cd..973219f 100644
> --- a/src/msg/SimpleMessenger.cc
> +++ b/src/msg/SimpleMessenger.cc
> @@ -373,7 +373,7 @@ void SimpleMessenger::DispatchQueue::entry()
> }
> lock.Lock();
> }
> - if (!stop)
> + if (queued_pipes.empty() && !stop)
> cond.Wait(lock); //wait for something to be put on queue
> }
> lock.Unlock();
> @@ -2477,7 +2477,7 @@ void SimpleMessenger::submit_message(Message *m, Connection *con, const entity_a
> }
>
> // local?
> - if (!pipe && my_inst.addr == dest_addr) {
> + if (my_inst.addr == dest_addr) {
> if (!destination_stopped) {
> // local
> ldout(cct,20) << "submit_message " << *m << " local" << dendl;
> --
> 1.7.7.6
>
> --
> Alexandre Oliva, freedom fighter http://FSFLA.org/~lxoliva/
> You must be the change you wish to see in the world. -- Gandhi
> Be Free! -- http://FSFLA.org/ FSF Latin America board member
> Free Software Evangelist Red Hat Brazil Compiler Engineer
> --
> To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
> the body of a message to majordomo@xxxxxxxxxxxxxxx
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>
>
--
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
