[#RVM-498] ppc-linux32 dying on SPEC jbb 2005 with 4 threads

RVM

ppc-linux32 dying on SPEC jbb 2005 with 4 threads

Created: 18/May/08 05:18 AM Updated: 14/Oct/08 07:55 AM

Component/s:

Instruction Architecture: PowerPC, Runtime: Threads and Concurrency

Affects Version/s:

3.0, 3.0.1

Fix Version/s:

3.0.2

Time Tracking:

Not Specified

Description

« Hide

With 1 exception we're consistently dying in production when SPEC jbb 2005 reaches 4 threads on rvmppclnx64:

All

Comments

Work Log

Change History

Sort Order:

[ Permlink | « Hide ]

David Grove added a comment - 19/May/08 08:22 AM

this machine is actually running ppc32-linux tests.

[ Permlink | « Hide ]

David Grove added a comment - 04/Jul/08 03:03 PM

Confirmed that we still see this crash; appears in last night's run http://jikesrvm.anu.edu.au/cattrack/results/rvmppclnx64.anu.edu.au/core/4275/production/default/SPECjbb2005/SPECjbb2005

[ Permlink | « Hide ]

Ian Rogers added a comment - 23/Jul/08 04:52 AM

Rescheduling for 3.0.1 as no time to investigate currently.

[ Permlink | « Hide ]

Ian Rogers added a comment - 15/Aug/08 05:17 AM

Still occurring:
http://jikesrvm.anu.edu.au/cattrack/results/piccolo.watson.ibm.com/core-ppc32/4794/production/Performance/SPECjbb2005/SPECjbb2005/default/Output.txt

[ Permlink | « Hide ]

David Grove added a comment - 14/Oct/08 07:55 AM

I think this is mostly fixed (piano/piccolo runs usually pass) but until cattrack gets back to being usable I want to leave it open. But its definitely not a regression over 3.0, so we don't have to hold 3.0.1 for it.