osd: clear PG_STATE_CLEAN when repair object

there is a race be found, when we repair object on clean state,
we queue a DoRecovery peering event, but before the peering event
dequeue,a snaptrim event on the missing object's snap dequeue,
then we will get pass the context< SnapTrimmer >().can_trim()
and go to get the context of the missing object(snapdir)

we can avoid this by clear clean state when we found missing..

Fixes: https://tracker.ceph.com/issues/41348
Signed-off-by: Zengran Zhang <zhangzengran@sangfor.com.cn>
(cherry picked from commit 521f095c6505bbee7570fb3c01b32436bdbf65a4)

Conflicts:
	src/osd/PrimaryLogPG.cc
- assert() instead of ceph_assert(), and Feature PR #26942 ("Improvements to
  auto repair") is not being backported

diff --git a/src/osd/PrimaryLogPG.cc b/src/osd/PrimaryLogPG.cc
index d1b5953e22e9ed720ecc7eecdfa4fe8266cfb520..68c578685c27b487aacd2f449656c808c006a6cd 100644 (file)
--- a/src/osd/PrimaryLogPG.cc
+++ b/src/osd/PrimaryLogPG.cc
@@ -14524,6 +14524,7 @@ int PrimaryLogPG::rep_repair_primary_object(const hobject_t& soid, OpRequestRef
   if (!eio_errors_to_process) {
     eio_errors_to_process = true;
     assert(is_clean());
+    state_clear(PG_STATE_CLEAN);
     queue_peering_event(
         CephPeeringEvtRef(
          std::make_shared<CephPeeringEvt>(