From: Greg Farnum <gfarnum@redhat.com>
Date: Tue, 2 Dec 2014 23:17:57 +0000 (-0800)
Subject: SimpleMessenger: allow RESETSESSION whenever we forget an endpoint
X-Git-Tag: v0.80.10~76^2
X-Git-Url: http://git-server-git.apps.pok.os.sepia.ceph.com/?a=commitdiff_plain;h=refs%2Fpull%2F3915%2Fhead;p=ceph.git

SimpleMessenger: allow RESETSESSION whenever we forget an endpoint

In the past (e229f8451d37913225c49481b2ce2896ca6788a2) we decided to disable
reset of lossless Pipes, because lossless peers resetting caused trouble and
they can't forget about each other. But they actually can: if mark_down()
is called.

I can't figure out how else we could forget about a remote endpoint, so I think
it's okay if we tell them we reset in order to clean up state. That's desirable
so that we don't get into strange situations with out-of-whack counters.

Fixes: #10080
Backport: giant, firefly, dumpling

Signed-off-by: Greg Farnum <gfarnum@redhat.com>
(cherry picked from commit 8cd1fdd7a778eb84cb4d7161f73bc621cc394261)
---

diff --git a/src/msg/Pipe.cc b/src/msg/Pipe.cc
index 54ce6799fa5..665e72c4f41 100644
--- a/src/msg/Pipe.cc
+++ b/src/msg/Pipe.cc
@@ -549,7 +549,7 @@ int Pipe::accept()
 	       << " > " << existing->connect_seq << dendl;
       goto replace;
     } // existing
-    else if (policy.resetcheck && connect.connect_seq > 0) {
+    else if (connect.connect_seq > 0) {
       // we reset, and they are opening a new session
       ldout(msgr->cct,0) << "accept we reset (peer sent cseq " << connect.connect_seq << "), sending RESETSESSION" << dendl;
       msgr->lock.Unlock();