Project

General

Profile

Bug #1396

AMPI intercomm_coll test break BigSim autobuild

Added by Sam White over 2 years ago. Updated over 2 years ago.

Status:
Merged
Priority:
Normal
Category:
AMPI
Target version:
Start date:
02/04/2017
Due date:
% Done:

0%


Description

https://charm.cs.illinois.edu/autobuild/cur/netlrts-linux-x86_64-bigsim.txt

The netlrts-linux-x86_64-bigsim build broke due to the new test not having a bgtest target.

make[3]: Entering directory `/scratch/autobuild/bluegene/charm/netlrts-linux-x86_64-bigsim/tests/ampi/intercomm_coll'
make[3]: *** No rule to make target `bgtest'.  Stop.
make[3]: Leaving directory `/scratch/autobuild/bluegene/charm/netlrts-linux-x86_64-bigsim/tests/ampi/intercomm_coll'
make[2]: *** [bgtest] Error 1
make[2]: Leaving directory `/scratch/autobuild/bluegene/charm/netlrts-linux-x86_64-bigsim/tests/ampi'
make[1]: *** [bgtest] Error 1
make[1]: Leaving directory `/scratch/autobuild/bluegene/charm/netlrts-linux-x86_64-bigsim/tests'
make: *** [bgtest] Error 2
fatal> error code 2 during remote> make bgtest TESTOPTS="++local ++no-va-randomization" 
Testing finished at Sat Feb 4 01:14:49 CST 2017
Returned from executing scripts/netlrts-linux-x86_64-bigsim/test on remote host
fatal> Test on remote host failed with fatal error (0)
Bad: Test on remote host failed with fatal error (0)

tests/ampi/Makefile contains a loop over all subdirectories, assuming that each test has a bgtest target, but tests/ampi/intercomm_coll/Makefile does not.

History

#1 Updated by Karthik Senthil over 2 years ago

  • Status changed from New to Implemented

#2 Updated by Sam White over 2 years ago

Autobuild failed inside intercomm_coll's bgtest now, during intercommunicator creation. I think we should not test BigSim on intercommunicators yet, until we can trust AMPI's intercomms more. So remove intercomm_coll from tests/ampi/Makefile's loop over its subdirectories.

Trace: traceroot: ./intercomm_coll
------------- Processor 0 Exiting: Called CmiAbort ------------
Reason: thread resumed, but callback data is still empty
[0] Stack Traceback:
  [0:0] CmiAbortHelper+0xb3  [0x69a3b5]
  [0:1] CmiAbort+0x2d  [0x69a3f0]
  [0:2] _ZNK10CkCallback17impl_thread_delayEv+0xe1  [0x5e2fb5]
  [0:3] _ZNK10CkCallback12thread_delayEv+0x23  [0x579125]
  [0:4] _ZN4ampi22createNewChildAmpiSyncEv+0x166  [0x54c6b4]
  [0:5] _ZN4ampi21intercommCreatePhase1Ei+0x34  [0x54de28]
  [0:6] _ZN12CkIndex_ampi51_call_redn_wrapper_intercommCreatePhase1_marshall15EPvS0_+0x78  [0x56dabc]
  [0:7] CkDeliverMessageFree+0x4e  [0x5cc6a6]
  [0:8] _ZN8CkLocRec11invokeEntryEP12CkMigratablePvib+0x285  [0x5ebb75]
  [0:9] _ZN8CkLocMgr10deliverMsgEP14CkArrayMessage9CkArrayIDmPK12CkArrayIndex11CkDeliver_ti+0x561  [0x5eec6d]
  [0:10] _ZN7CkArray7deliverEP14CkArrayMessage11CkDeliver_t+0x68  [0x5d33ce]
  [0:11]   [0x5cee25]
  [0:12] _Z15_processHandlerPvP11CkCoreState+0x198  [0x5cefbf]
  [0:13] _Z23BgProcessMessageDefaultP10threadInfoPc+0x195  [0x5a3e25]
  [0:14] _Z26BgProcessMessageFreezeModeP10threadInfoPc+0x167  [0x6ada6f]
  [0:15] _ZN14workThreadInfo9schedulerEi+0x403  [0x5ab043]
  [0:16] _ZN14workThreadInfo3runEv+0x475  [0x5ab60f]
  [0:17] _Z10run_threadP10threadInfo+0x41  [0x5a3e92]
  [0:18] CthStartThread+0x59  [0x698015]
  [0:19] +0x48200  [0x2aaaab53a200]

#3 Updated by Phil Miller over 2 years ago

  • Status changed from Implemented to Merged
  • translation missing: en.field_closed_date set to 2017-02-07 17:11:29.175779

Also available in: Atom PDF