RVM

Modularize threading system (native thread support)

Details

  • Number of attachments :
    0

Description

Currently we only support M-to-N threading, that is mapping M Java threads on N virtual processors executing on N pthreads. This causes a number of issues when interacting with native code and we have a number of complicated solutions to this problem. The thread system should be redesigned to support the mapping of M Java threads to M pthreads.

Issue Links

duplicates

Improvement - An improvement or enhancement to an existing feature or task. RVM-64 Flexible thread model

  • Major - Major loss of function.
  • Closed - The issue is considered finished, the resolution is correct. Issues which are not closed can be reopened.
relates to

New Feature - A new feature of the product, which has yet to be developed. RVM-546 Support for Harmony's thread library

  • Major - Major loss of function.
  • Closed - The issue is considered finished, the resolution is correct. Issues which are not closed can be reopened.
1. Only allow clean/expected transitions in thread execStatus Sub-task Open Open Unassigned
 
2. Implement Thread.setPriority Sub-task Open Open Unassigned
 
3.
 Implement PowerPC syscalls Sub-task Closed Closed Ian Rogers
 
4.
 Rationalize createVM code Sub-task Closed Closed Filip Pizlo
 
5.
 Native Threads: exception in bootimagewriter on ppc32-aix and ppc64-linux Sub-task Closed Closed Daniel Frampton
 
6.
 Native Threads: uninterruptibility violations building prototype-opt image on ia32-linux Sub-task Closed Closed Unassigned
 
7.
 Stack maps seem to be broken Sub-task Closed Closed Unassigned
 
8.
 mutator contexts are not being collected after thread death Sub-task Closed Closed Daniel Frampton
 
9.
 Threads occasionally remain IN_JAVA_TO_BLOCK and never reach a safepoint during a GC stop-the-world request Sub-task Closed Closed Filip Pizlo
 
10.
 Javadoc for a lot of RVMThread methods is missing, in particular methods for entering/exiting the "blocked"states. Sub-task Closed Closed Filip Pizlo
 
11.
 Rename the "NATIVE" states to something better - perhaps "IN_GC_SAFE"? Sub-task Closed Closed Filip Pizlo
 
12.
 Add comments and assertions to the block adapter code. Sub-task Closed Closed Filip Pizlo
 
13.
 Add comments relating to the use and purpose or, and rationale for acctLock Sub-task Closed Closed Filip Pizlo
 
14.
 Clean up lock classes: HeavyCondLock, Latch, NoYieldPointsCondLock, and SpinLock Sub-task Closed Closed Filip Pizlo
 
15.
 TimerThread is broken Sub-task Closed Closed Filip Pizlo
 
16. Decide what 'available' processors means on native threads Sub-task Open Open Unassigned
 
17.
 Update userguide to decribe native threads Sub-task Closed Closed Filip Pizlo
 
18. Figure out why native threads stability is proportional to the number of available processors Sub-task Open Open Unassigned
 
19. Figure out if OSR's use of contextRegisters is correct and/or necessary, given their changed meaning in native threads Sub-task Open Open Unassigned
 
20.
 handleHandshakeRequest should be called from enterNativeBlockedImpl Sub-task Closed Closed Unassigned
 
21.
 Adjust sampling mechanism in AOS to account for native threads Sub-task Closed Closed David Grove
 
22.
 NPE killing Eclipse performance runs Sub-task Closed Closed Daniel Frampton
 
23.
 Apache Harmony class library completely broken by native thread code Sub-task Closed Closed Daniel Frampton
 
24. Improve tracing support in RVMThread Sub-task Open Open Unassigned
 
25.
 join() with a timeout doesn't time out Sub-task Closed Closed Filip Pizlo
 

Activity

People

  • Assignee:
    Filip Pizlo
    Reporter:
    Ian Rogers
Vote (0)
Watch (0)

Dates

  • Created:
    Updated:
    Resolved: