[rvm-research] Find holes in a block of SegregatedFreeListSpace
|
View:
New views
5 Messages
—
Rating Filter:
Alert me
|
 |
[rvm-research] Find holes in a block of SegregatedFreeListSpace
by Dingwen Yuan
::
Rate this Message:
Sorry, I hit a space and the above mail was mistakenly sent. Again, the code is like this:
@Inline private void auditBlock (Address block, int sizeClass, Extent blockSize) { Extent cellBytes = Extent.fromIntSignExtend( cellSize[sizeClass]); I
call this function to scan blocks in "available list" at the RELEASE
phase of GC directly after space.release() was called. But I found
there should be some bug in it, but cannot tell where. Could anyone
help me with this?Address cursor = block.plus(blockHeaderSize[sizeClass]); Address end = block.plus(blockSize); while(cursor.LT(end)) { ObjectReference current = VM.objectModel.getObjectFromStartAddress(cursor); if (current == null || !this.isCellLive(current)) { holes++; } cursor = cursor.plus(cellBytes); } } Thanks in advance. Dingwen TU Darmstadt ------------------------------------------------------------------------------ Crystal Reports - New Free Runtime and 30 Day Trial Check out the new simplified licensing option that enables unlimited royalty-free distribution of the report engine for externally facing server and web deployment. http://p.sf.net/sfu/businessobjects _______________________________________________ Jikesrvm-researchers mailing list Jikesrvm-researchers@... https://lists.sourceforge.net/lists/listinfo/jikesrvm-researchers |
|
|
Re: [rvm-research] Find holes in a block of SegregatedFreeListSpace
by Steve Blackburn
::
Rate this Message:
Hi Dingwen,
Sorry I have not replied earlier. I'm afraid the specifics of the surrounding code are not at the front of my mind right now and it's pretty tough to debug your code with so little context. Can you be a bit more specific about how it is failing? For what it's worth, I strongly recommend the use of assertions, particularly in GC code---figure out what your invariants and preconditions are and then thrown assertions in liberally to ensure your assumptions are holding true. The assertions cost nothing in a production build (they are turned off and removed statically---no runtime checks). --Steve On 10/06/2009, at 6:32 PM, Dingwen Yuan wrote: > Sorry, I hit a space and the above mail was mistakenly sent. Again, > the code is like this: > > @Inline > private void auditBlock (Address block, int sizeClass, > Extent blockSize) { > > Extent cellBytes = Extent.fromIntSignExtend( > cellSize[sizeClass]); > Address cursor = block.plus(blockHeaderSize[sizeClass]); > Address end = block.plus(blockSize); > > while(cursor.LT(end)) { > ObjectReference current = > VM.objectModel.getObjectFromStartAddress(cursor); > if (current == null || !this.isCellLive(current)) { > holes++; > } > cursor = cursor.plus(cellBytes); > } > } > > I call this function to scan blocks in "available list" at the > RELEASE phase of GC directly after space.release() was called. But I > found there should be some bug in it, but cannot tell where. Could > anyone help me with this? > > Thanks in advance. > > Dingwen > > TU Darmstadt > ------------------------------------------------------------------------------ > Crystal Reports - New Free Runtime and 30 Day Trial > Check out the new simplified licensing option that enables unlimited > royalty-free distribution of the report engine for externally facing > server and web deployment. > http://p.sf.net/sfu/businessobjects_______________________________________________ > Jikesrvm-researchers mailing list > Jikesrvm-researchers@... > https://lists.sourceforge.net/lists/listinfo/jikesrvm-researchers ------------------------------------------------------------------------------ Crystal Reports - New Free Runtime and 30 Day Trial Check out the new simplified licensing option that enables unlimited royalty-free distribution of the report engine for externally facing server and web deployment. http://p.sf.net/sfu/businessobjects _______________________________________________ Jikesrvm-researchers mailing list Jikesrvm-researchers@... https://lists.sourceforge.net/lists/listinfo/jikesrvm-researchers |
|
|
Re: [rvm-research] Find holes in a block of SegregatedFreeListSpace
by Dingwen Yuan
::
Rate this Message:
Hallo Steve,
Thank you for the reply. Some context is like this. I added some code to scan the mature space of GenMS for holes in the available list. GenMS.collectionPhase public final void collectionPhase(short phaseId) { if (traceFullHeap()) { ........... if (phaseId == RELEASE) { if (Options.verbose.getValue() > 0) Log.writeln("main release"); matureTrace.release(); msSpace.release(); holes = 0; msSpace.linearScan(); Log.write("holes: "); Log.writeln(holes); super.collectionPhase(phaseId); return; } } } SegregatedFreeListSpace.linearScan public final void linearScan(boolean print) { for (int sizeClass = 0; sizeClass < sizeClassCount(); sizeClass++) { if (VM.VERIFY_ASSERTIONS) { VM.assertions._assert(this.flushedBlockHead.get(sizeClass).isZero()); VM.assertions._assert(this.consumedBlockHead.get(sizeClass).isZero()); } Address block = this.availableBlockHead.get(sizeClass); Extent blockSize = Extent.fromIntSignExtend(BlockAllocator.blockSize(blockSizeClass[sizeClass])); while (!block.isZero()) { Address next = BlockAllocator.getNext(block); this.auditBlock(block, sizeClass, blockSize); block = next; } } } } SegregatedFreeListSpace.auditBlock @Inline private void auditBlock (Address block, int sizeClass, Extent blockSize) { Address end = block.plus(blockSize); Address cursor = block.plus(blockHeaderSize[sizeClass]); Extent cellExtent = Extent.fromIntSignExtend(cellSize[sizeClass]); while (cursor.LT(end)) { ObjectReference current = VM.objectModel.getObjectFromStartAddress(cursor); if (current.isNull() || !isCellLive(current)) { holes++; } cursor = cursor.plus(cellExtent); } } And I also did a very simple experiment on it. The experiment code added 20 Integer to a static arrayList. Of course, all these objects will be promoted to mature space after gc. At the end of the code I called 10 System.gc() in a sequence and I also set -X:gc:fullHeapSystemGC=true so that full gc was be triggered. The holes of these 10 gcs are holes: 0 holes: 1367 holes: 348 holes: 1359 holes: 1011 holes: 1359 holes: 348 holes: 1359 holes: 1011 holes: 1359 This doesn't conform to my expectation. I supposed that the second to the tenth "holes" should be 0 when the first is already 0. And in addition, there seems to be some pattern in the numbers -- 1359 = 1011+348. My guess is that this may relate to the markState of MarkSweepSpace which changes from gc to gc. BTW, the version I used is 2.9.2 cheers, Dingwen 2009/6/11 Steve Blackburn <Steve.Blackburn@...> Hi Dingwen, ------------------------------------------------------------------------------ Crystal Reports - New Free Runtime and 30 Day Trial Check out the new simplified licensing option that enables unlimited royalty-free distribution of the report engine for externally facing server and web deployment. http://p.sf.net/sfu/businessobjects _______________________________________________ Jikesrvm-researchers mailing list Jikesrvm-researchers@... https://lists.sourceforge.net/lists/listinfo/jikesrvm-researchers |
|
|
Re: [rvm-research] Find holes in a block of SegregatedFreeListSpace
by Robin Garner
::
Rate this Message:
Hi Dingwen,
You might find the MMTk Test Harness (http://www.jikesrvm.org/The+MMTk+Test+Harness) useful in debugging this problem. Regards, Robin Dingwen Yuan wrote: > Hallo Steve, > > Thank you for the reply. Some context is like this. I added some code > to scan the mature space of GenMS for holes in the available list. > > GenMS.collectionPhase > public final void collectionPhase(short phaseId) { > > if (traceFullHeap()) { > ........... > if (phaseId == RELEASE) { > if (Options.verbose.getValue() > 0) > Log.writeln("main release"); > matureTrace.release(); > > msSpace.release(); > > holes = 0; > msSpace.linearScan(); > Log.write("holes: "); > Log.writeln(holes); > > super.collectionPhase(phaseId); > return; > } > } > } > > SegregatedFreeListSpace.linearScan > > public final void linearScan(boolean print) { > for (int sizeClass = 0; sizeClass < sizeClassCount(); sizeClass++) { > > if (VM.VERIFY_ASSERTIONS) { > > VM.assertions._assert(this.flushedBlockHead.get(sizeClass).isZero()); > > VM.assertions._assert(this.consumedBlockHead.get(sizeClass).isZero()); > } > > Address block = this.availableBlockHead.get(sizeClass); > Extent blockSize = > Extent.fromIntSignExtend(BlockAllocator.blockSize(blockSizeClass[sizeClass])); > > while (!block.isZero()) { > Address next = BlockAllocator.getNext(block); > this.auditBlock(block, sizeClass, blockSize); > block = next; > } > } > } > } > > SegregatedFreeListSpace.auditBlock > @Inline > private void auditBlock (Address block, int sizeClass, > Extent blockSize) { > > Address end = block.plus(blockSize); > > Address cursor = block.plus(blockHeaderSize[sizeClass]); > Extent cellExtent = Extent.fromIntSignExtend(cellSize[sizeClass]); > > while (cursor.LT(end)) { > ObjectReference current = > VM.objectModel.getObjectFromStartAddress(cursor); > if (current.isNull() || !isCellLive(current)) { > holes++; > } > cursor = cursor.plus(cellExtent); > } > } > > And I also did a very simple experiment on it. The experiment code > added 20 Integer to a static arrayList. Of course, all these objects > will be promoted to mature space after gc. At the end of the code I > called 10 System.gc() in a sequence and I also set > -X:gc:fullHeapSystemGC=true so that full gc was be triggered. The > holes of these 10 gcs are > holes: 0 > holes: 1367 > holes: 348 > holes: 1359 > holes: 1011 > holes: 1359 > holes: 348 > holes: 1359 > holes: 1011 > holes: 1359 > > This doesn't conform to my expectation. I supposed that the second to > the tenth "holes" should be 0 when the first is already 0. And in > addition, there seems to be some pattern in the numbers -- 1359 = > 1011+348. My guess is that this may relate to the markState of > MarkSweepSpace which changes from gc to gc. > > BTW, the version I used is 2.9.2 > > cheers, > > Dingwen > > > 2009/6/11 Steve Blackburn <Steve.Blackburn@... > <mailto:Steve.Blackburn@...>> > > Hi Dingwen, > > Sorry I have not replied earlier. > > I'm afraid the specifics of the surrounding code are not at the front > of my mind right now and it's pretty tough to debug your code with so > little context. Can you be a bit more specific about how it is > failing? > > For what it's worth, I strongly recommend the use of assertions, > particularly in GC code---figure out what your invariants and > preconditions are and then thrown assertions in liberally to ensure > your assumptions are holding true. The assertions cost nothing in a > production build (they are turned off and removed statically---no > runtime checks). > > --Steve > > On 10/06/2009, at 6:32 PM, Dingwen Yuan wrote: > > > Sorry, I hit a space and the above mail was mistakenly sent. Again, > > the code is like this: > > > > @Inline > > private void auditBlock (Address block, int sizeClass, > > Extent blockSize) { > > > > Extent cellBytes = Extent.fromIntSignExtend( > > cellSize[sizeClass]); > > Address cursor = block.plus(blockHeaderSize[sizeClass]); > > Address end = block.plus(blockSize); > > > > while(cursor.LT(end)) { > > ObjectReference current = > > VM.objectModel.getObjectFromStartAddress(cursor); > > if (current == null || !this.isCellLive(current)) { > > holes++; > > } > > cursor = cursor.plus(cellBytes); > > } > > } > > > > I call this function to scan blocks in "available list" at the > > RELEASE phase of GC directly after space.release() was called. But I > > found there should be some bug in it, but cannot tell where. Could > > anyone help me with this? > > > > Thanks in advance. > > > > Dingwen > > > > TU Darmstadt > > > ------------------------------------------------------------------------------ > > Crystal Reports - New Free Runtime and 30 Day Trial > > Check out the new simplified licensing option that enables unlimited > > royalty-free distribution of the report engine for externally facing > > server and web deployment. > > > http://p.sf.net/sfu/businessobjects_______________________________________________ > > Jikesrvm-researchers mailing list > > Jikesrvm-researchers@... > <mailto:Jikesrvm-researchers@...> > > https://lists.sourceforge.net/lists/listinfo/jikesrvm-researchers > > > ------------------------------------------------------------------------------ > Crystal Reports - New Free Runtime and 30 Day Trial > Check out the new simplified licensing option that enables unlimited > royalty-free distribution of the report engine for externally facing > server and web deployment. > http://p.sf.net/sfu/businessobjects > _______________________________________________ > Jikesrvm-researchers mailing list > Jikesrvm-researchers@... > <mailto:Jikesrvm-researchers@...> > https://lists.sourceforge.net/lists/listinfo/jikesrvm-researchers > > > ------------------------------------------------------------------------ > > ------------------------------------------------------------------------------ > Crystal Reports - New Free Runtime and 30 Day Trial > Check out the new simplified licensing option that enables unlimited > royalty-free distribution of the report engine for externally facing > server and web deployment. > http://p.sf.net/sfu/businessobjects > ------------------------------------------------------------------------ > > _______________________________________________ > Jikesrvm-researchers mailing list > Jikesrvm-researchers@... > https://lists.sourceforge.net/lists/listinfo/jikesrvm-researchers > ------------------------------------------------------------------------------ Crystal Reports - New Free Runtime and 30 Day Trial Check out the new simplified licensing option that enables unlimited royalty-free distribution of the report engine for externally facing server and web deployment. http://p.sf.net/sfu/businessobjects _______________________________________________ Jikesrvm-researchers mailing list Jikesrvm-researchers@... https://lists.sourceforge.net/lists/listinfo/jikesrvm-researchers |
|
|
Re: [rvm-research] Find holes in a block of SegregatedFreeListSpace
by Dingwen Yuan
::
Rate this Message:
Hi Robin,
I think I am not able be work with Test Harness in version 2.9.2. cheers, Dingwen 2009/6/12 Robin Garner <robin.garner@...> Hi Dingwen, ------------------------------------------------------------------------------ Crystal Reports - New Free Runtime and 30 Day Trial Check out the new simplified licensing option that enables unlimited royalty-free distribution of the report engine for externally facing server and web deployment. http://p.sf.net/sfu/businessobjects _______________________________________________ Jikesrvm-researchers mailing list Jikesrvm-researchers@... https://lists.sourceforge.net/lists/listinfo/jikesrvm-researchers |
| Free embeddable forum powered by Nabble | Forum Help |
