Last minute fixes for v32r0

• c685d714: fix stack-use-after-scope from address sanitizer
• b887e28c: fix regression introduced with !904 (merged)

changed milestone to %v32r0

assigned to @clemenci

in that case, made add a constructor which takes an rvalue ref to EventContext, and then explicitly = delete it so that trying to construct an AlgContext with a temporary doesn't compile...

specifically:

diff --git a/GaudiKernel/GaudiKernel/IAlgContextSvc.h b/GaudiKernel/GaudiKernel/IAlgContextSvc.h
index 8398cbba0..ef6c8578c 100644
--- a/GaudiKernel/GaudiKernel/IAlgContextSvc.h
+++ b/GaudiKernel/GaudiKernel/IAlgContextSvc.h
@@ -82,6 +82,7 @@ namespace Gaudi {
        *  @param context current event context
        */
       AlgContext( IAlgorithm* alg, IAlgContextSvc* svc, const EventContext& context );
+      AlgContext( IAlgorithm* alg, IAlgContextSvc* svc, EventContext&& context ) = delete;
 
       /** constructor from the service and the algorithm
        *  Internally invokes IAlgContextSvc::setCurrentAlg

Thanks for the idea! At some point (for another case) I was looking for a way to prevent passing temporaries, and this is the answer :)

I'll do it another day.

for some reason this discussion does not require a resolution (did you comment directly on the commit? I answer to myself: yes, the discussion header says that the comment was on the commit and not on the diff)... I'll have to create the issue by hand :(

resolved all discussions

mentioned in issue #72 (closed)

resolved all discussions

mentioned in issue #73 (closed)

mentioned in commit lhcb/Phys@f9d37e92

mentioned in merge request lhcb/Phys!563 (closed)

mentioned in commit lhcb/Phys@215c1f70

mentioned in merge request lhcb/Hlt!625 (closed)

• [2019-05-17 00:03] Validation started with lhcb-dd4hep#180
• [2019-05-17 00:04] Validation started with lhcb-sanitizers#248
• [2019-05-17 00:04] Validation started with lhcb-lcg-dev3#895
• [2019-05-17 00:04] Validation started with lhcb-test-throughput2#10
• [2019-05-17 00:06] Validation started with lhcb-lcg-dev4#903
• [2019-05-17 00:07] Validation started with lhcb-soa-track#80
• [2019-05-17 00:09] Validation started with lhcb-gaudi-head#2255
• [2019-05-17 00:10] Validation started with lhcb-tdr-test#548
• [2019-05-17 08:07] Validation started with lhcb-gaudi-merge#752
• [2019-05-18 00:03] Validation started with lhcb-test-throughput2#11
• [2019-05-18 00:03] Validation started with lhcb-lcg-dev3#896
• [2019-05-18 00:04] Validation started with lhcb-sanitizers#249
• [2019-05-18 00:04] Validation started with lhcb-dd4hep#181
• [2019-05-18 00:05] Validation started with lhcb-lcg-dev4#904
• [2019-05-18 00:06] Validation started with lhcb-gaudi-head#2256
• [2019-05-18 00:06] Validation started with lhcb-soa-track#81
• [2019-05-18 00:09] Validation started with lhcb-tdr-test#549
• [2019-05-18 00:36] Validation started with lhcb-gaudi-head#2256

added 1 commit

• b887e28c - Fix propagation of RootInTES from parents to children

Compare with previous version

changed the description

The two failing DaVinci tests are fixed. The failure in https://lhcb-nightlies.cern.ch/logs/tests/nightly/lhcb-gaudi-head/2256/x86_64-centos7-gcc8-opt/DaVinci/ is due to another MR (lhcb/LHCb!1931 (merged))

merged

Yes, I suspected that too, thanks.

mentioned in commit lhcb/Phys@4fdc7472