Mercurial > hg > graal-compiler
comparison src/share/vm/runtime/globals.hpp @ 546:05c6d52fa7a9
6690928: Use spinning in combination with yields for workstealing termination.
Summary: Substitute a spin loop for most calls to yield() to reduce the stress on the system.
Reviewed-by: tonyp
author | jmasa |
---|---|
date | Sun, 08 Feb 2009 13:18:01 -0800 |
parents | 23673011938d |
children | dca06e7f503d |
comparison
equal
deleted
inserted
replaced
545:58054a18d735 | 546:05c6d52fa7a9 |
---|---|
1653 "Sleep time when sleep is used for yields") \ | 1653 "Sleep time when sleep is used for yields") \ |
1654 \ | 1654 \ |
1655 develop(uintx, WorkStealingYieldsBeforeSleep, 1000, \ | 1655 develop(uintx, WorkStealingYieldsBeforeSleep, 1000, \ |
1656 "Number of yields before a sleep is done during workstealing") \ | 1656 "Number of yields before a sleep is done during workstealing") \ |
1657 \ | 1657 \ |
1658 develop(uintx, WorkStealingHardSpins, 4096, \ | |
1659 "Number of iterations in a spin loop between checks on " \ | |
1660 "time out of hard spin") \ | |
1661 \ | |
1662 develop(uintx, WorkStealingSpinToYieldRatio, 10, \ | |
1663 "Ratio of hard spins to calls to yield") \ | |
1664 \ | |
1658 product(uintx, PreserveMarkStackSize, 1024, \ | 1665 product(uintx, PreserveMarkStackSize, 1024, \ |
1659 "Size for stack used in promotion failure handling") \ | 1666 "Size for stack used in promotion failure handling") \ |
1660 \ | 1667 \ |
1661 product_pd(bool, UseTLAB, "Use thread-local object allocation") \ | 1668 product_pd(bool, UseTLAB, "Use thread-local object allocation") \ |
1662 \ | 1669 \ |