| f1c4d0e3a27e9b39916750147ecdea1418fcc231 |
|
02-Dec-2014 |
Mathieu Chartier <mathieuc@google.com> |
Try normal allocation if large object allocation fails
If a large object allocation fails, we now try the normal allocators.
Bug: 18504942
Change-Id: I18b9759d6af885556941542c57fec584f18197f1
|
| 27f5ae830c5418fa92094608a6e9f693ea88bb69 |
|
21-Aug-2014 |
Mathieu Chartier <mathieuc@google.com> |
Check pending exception result in AllocObjectWithAllocator.
Possible previous bug:
Allocation fails due to OOM and the collector transitions.
This caused us to incorrectly retry the allocation with a pending
exception. We now return null if there is a pending exception.
Bug: 17164348
Change-Id: I22eab472afb2fdea6e800963ccb35ec0755ba0e6
|
| 59fe711f88191cd8ca1a386c4fa0d2f9e484af50 |
|
14-Jul-2014 |
Mathieu Chartier <mathieuc@google.com> |
Fix infinite loop when calling SetStatus after OOM.
There was a problem where we would call SetStatus when we had an OOM
error. This results in attempting to find the ExceptionInInitializer
class which if not loaded does more allocations resulting in an
infinite loop.
Also some cleanup addressing other comments.
Bug: 16082350
(cherry picked from commit fd22d5bada15d95b5ea8ab5a4dda39077e1a54ee)
Change-Id: Ie291eb0f52ba9c63f24591fae691dd9f393e6ccb
|
| fd22d5bada15d95b5ea8ab5a4dda39077e1a54ee |
|
14-Jul-2014 |
Mathieu Chartier <mathieuc@google.com> |
Fix infinite loop when calling SetStatus after OOM.
There was a problem where we would call SetStatus when we had an OOM
error. This results in attempting to find the ExceptionInInitializer
class which if not loaded does more allocations resulting in an
infinite loop.
Also some cleanup addressing other comments.
Bug: 16082350
Change-Id: I5c1e638a03ddf700ab4e9cad9a3077d2b1b26c43
|
| 14cc9be4adc652071979395d337d1380763844fa |
|
11-Jul-2014 |
Mathieu Chartier <mathieuc@google.com> |
Faster TLAB allocator.
New TLAB allocator doesn't increment bytes allocated until we allocate
a new TLAB. This increases allocation performance by avoiding a CAS.
MemAllocTest:
Before GSS TLAB: 3400ms.
After GSS TLAB: 2750ms.
Bug: 9986565
Change-Id: I1673c27555330ee90d353b98498fa0e67bd57fad
|
| ffddfdf6fec0b9d98a692e27242eecb15af5ead2 |
|
03-Jun-2014 |
Tim Murray <timmurray@google.com> |
DO NOT MERGE
Merge ART from AOSP to lmp-preview-dev.
Change-Id: I0f578733a4b8756fd780d4a052ad69b746f687a9
|
| c179016fe188bef09487e777aa0fd861f5cdf067 |
|
23-May-2014 |
Mathieu Chartier <mathieuc@google.com> |
Add reserve area to allocation stacks.
This fixes an issue with heap verification which was caused when
the allocation stack overflowed. This resulted in heap verification
failures since we were storing the newly allocated object in a
handle scope without having it be live either in the live bitmap
or allocation stack. We now push the object in the reserve area
before we do a GC due to allocation stack overflow.
Change-Id: I83b42c4b3250d7eaab1b49e53066e21c8656a740
|
| 3e5cf305db800b2989ad57b7cde8fb3cc9fa1b9e |
|
21-May-2014 |
Ian Rogers <irogers@google.com> |
Begin migration of art::Atomic to std::atomic.
Change-Id: I4858d9cbed95e5ca560956b9dabd976cebe68333
|
| eb8167a4f4d27fce0530f6724ab8032610cd146b |
|
08-May-2014 |
Mathieu Chartier <mathieuc@google.com> |
Add Handle/HandleScope and delete SirtRef.
Delete SirtRef and replaced it with Handle. Handles are value types
which wrap around StackReference*.
Renamed StackIndirectReferenceTable to HandleScope.
Added a scoped handle wrapper which wraps around an Object** and
restores it in its destructor.
Renamed Handle::get -> Get.
Bug: 8473721
Change-Id: Idbfebd4f35af629f0f43931b7c5184b334822c7a
|
| 4cd662e54440f76fc920cb2c67acab3bba8b33dd |
|
04-Apr-2014 |
Hiroshi Yamauchi <yamauchi@google.com> |
Fix Object::Clone()'s pre-fence barrier.
Pass in a pre-fence barrier object that sets in the array length
instead of setting it after returning from AllocObject().
Fix another potential bug due to the wrong default pre-fence barrier
parameter value. Since this appears error-prone, removed the default
parameter value and make it an explicit parameter.
Fix another potential moving GC bug due to a lack of a SirtRef.
Bug: 13097759
Change-Id: I466aa0e50f9e1a5dbf20be5a195edee619c7514e
|
| 624468cd401cc1ac0dd70c746301e0788a597759 |
|
01-Apr-2014 |
Hiroshi Yamauchi <yamauchi@google.com> |
Make the support code for read barriers a bit more general.
Add an option for Baker in addition to Brooks.
Bug: 12687968
Change-Id: I8a31db817ff6686c72951b6534f588228e270b11
|
| 3e41780cb3bcade3b724908e00443a9caf6977ef |
|
20-Mar-2014 |
Hiroshi Yamauchi <yamauchi@google.com> |
Refactor the garbage collector driver (GarbageCollector::Run).
Bug: 12687968
Change-Id: Ifc9ee86249f7938f51495ea1498cf0f7853a27e8
|
| 38e68e9978236db87c9008bbe47db80525d2fa16 |
|
07-Mar-2014 |
Hiroshi Yamauchi <yamauchi@google.com> |
Use the card table to speed up the GSS collector.
Scan only dirty cards, as opposed to the whole space, to find
references from the non-moving spaces to the bump pointer spaces at
bump pointer space only collections.
With this change, the Ritz MemAllocTest speeds up by 8-10% on host and
2-3% on N4. The Ritz EvaluateFibonacci speeds up by 8% and its average
pause time is reduced by 43% on N4.
Bug: 11650816
Change-Id: I1eefe75776bc37e24673b301ffa65a25f9bd4cde
|
| 5ccd498d4aa450b0381344724b072a932709a59a |
|
11-Mar-2014 |
Hiroshi Yamauchi <yamauchi@google.com> |
Put the post zygote non-moving space next to the malloc space.
This change fixes the following problem with the GSS collector: if the
post zygote non-moving space happens to be located after the bump
pointer spaces, the bump pointer space is treated wrongfully as an
immune space. The reasons are 1) the immune space is represented by a
simple address range and 2) the GSS collector treats the post zygote
non-moving space as an immune space at a bump pointer space only
collection.
In addition, this change makes it a reality that all the non-moving
spaces are adjacent, which we tend to assume in the code (eg. the
notion of the immune space represented by a simple address range.)
This should help avoid potential confusions in the future.
Fix a DCHECK failure where usable_size isn't set properly when
-XX:UseTLAB is used.
Change-Id: I585920e433744a390f87e9a25bef6114b2a4623f
|
| c645f1ddb7c40bea6a38eda4b3f83f6b6dec405b |
|
07-Mar-2014 |
Mathieu Chartier <mathieuc@google.com> |
Add more VerifyObject calls.
Added verify object calls to SirtRef, IndirectReferenceTable,
ReferenceTable.
Removed un-needed verify object in ScopedObjectAccess / DecodeJObject
since object sources are handled.
Bug: 12934910
Change-Id: I55a46a8ea61fed2a77526eda27fd2cce97a9b125
|
| f517f1a994fab72ba484bbbac6911e315f59f6cd |
|
07-Mar-2014 |
Mathieu Chartier <mathieuc@google.com> |
Restore obj after RequestConcurrentGC.
RequestConcurrentGC can cause thread suspension, this means that
another thread could transition the heap or cause moving GC.
Bug: 12934910
Change-Id: I5c07161e2e849d7acbdf939f1c24e1ba361a1d6a
|
| bd0a65339a08dc28c6b56d2673f1f13b6bddd7aa |
|
27-Feb-2014 |
Mathieu Chartier <mathieuc@google.com> |
Enable large object space for command line runs.
Added a dynamic large_object_threshold_ variable which is max int
when the large object space is disabled.
No longer clear timing logger timings after zygote creation since
it is not required and removes the boot semispace timings.
Change-Id: I693865f4699cc32381199377239854c6ec42f37e
|
| a55cf41c9d1da7ee8b2f63974dedfb484042dd03 |
|
27-Feb-2014 |
Ian Rogers <irogers@google.com> |
Ensure usable space data is zeroed in arrays.
Address other outstanding review comments.
Change-Id: Iaffe04de080772a0d0c5fd973bcac0e23c8c3e25
|
| 6fac447555dc94a935b78198479cce645c837b89 |
|
26-Feb-2014 |
Ian Rogers <irogers@google.com> |
Make allocations report usable size.
Work-in-progress to allow arrays to fill usable size. Bug: 13028925.
Use C++11's override keyword on GCC >= 2.7 to ensure that we override GC and
allocator methods.
Move initial mirror::Class set up into a Functor so that all allocated objects
have non-zero sizes. Use this property to assert that all objects are never
larger than their usable size.
Other bits of GC related clean-up, missing initialization, missing use of
const, hot methods in .cc files, "unimplemented" functions that fail at
runtime in header files, reducing header file includes, move valgrind's space
into its own files, reduce number of array allocation routines.
Change-Id: Id5760041a2d7f94dcaf17ec760f6095ec75dadaa
|
| 9d04a20bde1b1855cefc64aebc1a44e253b1a13b |
|
31-Jan-2014 |
Hiroshi Yamauchi <yamauchi@google.com> |
(Experimental) Add Brooks pointers.
This feature is disabled by default.
Verified that the Brooks pointers are installed correctly by using the
CMS/SS collectors.
Change-Id: Ia9be9814ab6e29169ac85edc4792ce8c81d552a9
|
| 4e30541a92381fb280cd0be9a1763b713ee4d64c |
|
19-Feb-2014 |
Mathieu Chartier <mathieuc@google.com> |
Fix and optimize verify object.
VerifyObject no longer resides in heap. You can now enable
VerifyObject for non-debug builds. VerifyStack is still slow, so it
is now guarded by its own flag.
Fixed the image writer to not use verification at places where
verification fails due to invalid reads.
Fixed RosAlloc to use SizeOf which doesn't call verify object.
Added a flag paremeter to some of the mirror getters / setters to
be able to selectively disable VerifyObject on certain calls.
Optimized the GC to not verify each object multiple times during
object scanning if verify object is enabled.
Added 3 verification options: verify reads, verify this, and verify
writes so that you can select how much verification you want for
mirror getters and setters.
Removed some useless DCHECKs which would slow debug builds without
providing any benefits.
TODO: RosAlloc verification doesn't currently work with verify
objects.
Bug: 12934910
Bug: 12879358
Change-Id: Ic61033104dfc334543f89b0fc0ad8cd4f4015d69
|
| f5b0e20b5b31f5f5465784adcf2a204dcd69c7fd |
|
12-Feb-2014 |
Hiroshi Yamauchi <yamauchi@google.com> |
Thread-local allocation stack.
With this change, Ritz MemAllocTest gets ~14% faster on N4.
Bug: 9986565
Change-Id: I2fb7d6f7c5daa63dd4fc73ba739e6ae4ed820617
|
| e6da9af8dfe0a3e3fbc2be700554f6478380e7b9 |
|
16-Dec-2013 |
Mathieu Chartier <mathieuc@google.com> |
Background compaction support.
When the process state changes to a state which does not perceives
jank, we copy from the main free-list backed allocation space to
the bump pointer space and enable the semispace allocator.
When we transition back to foreground, we copy back to a free-list
backed space.
Create a seperate non-moving space which only holds non-movable
objects. This enables us to quickly wipe the current alloc space
(DlMalloc / RosAlloc) when we transition to background.
Added multiple alloc space support to the sticky mark sweep GC.
Added a -XX:BackgroundGC option which lets you specify
which GC to use for background apps. Passing in
-XX:BackgroundGC=SS makes the heap compact the heap for apps which
do not perceive jank.
Results:
Simple background foreground test:
0. Reboot phone, unlock.
1. Open browser, click on home.
2. Open calculator, click on home.
3. Open calendar, click on home.
4. Open camera, click on home.
5. Open clock, click on home.
6. adb shell dumpsys meminfo
PSS Normal ART:
Sample 1:
88468 kB: Dalvik
3188 kB: Dalvik Other
Sample 2:
81125 kB: Dalvik
3080 kB: Dalvik Other
PSS Dalvik:
Total PSS by category:
Sample 1:
81033 kB: Dalvik
27787 kB: Dalvik Other
Sample 2:
81901 kB: Dalvik
28869 kB: Dalvik Other
PSS ART + Background Compaction:
Sample 1:
71014 kB: Dalvik
1412 kB: Dalvik Other
Sample 2:
73859 kB: Dalvik
1400 kB: Dalvik Other
Dalvik other reduction can be explained by less deep allocation
stacks / less live bitmaps / less dirty cards.
TODO improvements: Recycle mem-maps which are unused in the current
state. Not hardcode 64 MB capacity of non movable space (avoid
returning linear alloc nightmares). Figure out ways to deal with low
virtual address memory problems.
Bug: 8981901
Change-Id: Ib235d03f45548ffc08a06b8ae57bf5bada49d6f3
|
| b122a4bbed34ab22b4c1541ee25e5cf22f12a926 |
|
20-Nov-2013 |
Ian Rogers <irogers@google.com> |
Tidy up memory barriers.
Change-Id: I937ea93e6df1835ecfe2d4bb7d84c24fe7fc097b
|
| 692fafd9778141fa6ef0048c9569abd7ee0253bf |
|
30-Nov-2013 |
Mathieu Chartier <mathieuc@google.com> |
Thread local bump pointer allocator.
Added a thread local allocator to the heap, each thread has three
pointers which specify the thread local buffer: start, cur, and
end. When the remaining space in the thread local buffer isn't large
enough for the allocation, the allocator allocates a new thread
local buffer using the bump pointer allocator.
The bump pointer space had to be modified to accomodate thread
local buffers. These buffers are called "blocks", where a block
is a buffer which contains a set of adjacent objects. Blocks
aren't necessarily full and may have wasted memory towards the
end. Blocks have an 8 byte header which specifies their size and is
required for traversing bump pointer spaces.
Memory usage is in between full bump pointer and ROSAlloc since
madvised memory limits wasted ram to an average of 1/2 page per
block.
Added a runtime option -XX:UseTLAB which specifies whether or
not to use the thread local allocator. Its a NOP if the garbage
collector is not the semispace collector.
TODO: Smarter block accounting to prevent us reading objects until
we either hit the end of the block or GetClass() == null which
signifies that the block isn't 100% full. This would provide a
slight speedup to BumpPointerSpace::Walk.
Timings: -XX:HeapMinFree=4m -XX:HeapMaxFree=8m -Xmx48m
ritzperf memalloc:
Dalvik -Xgc:concurrent: 11678
Dalvik -Xgc:noconcurrent: 6697
-Xgc:MS: 5978
-Xgc:SS: 4271
-Xgc:CMS: 4150
-Xgc:SS -XX:UseTLAB: 3255
Bug: 9986565
Bug: 12042213
Change-Id: Ib7e1d4b199a8199f3b1de94b0a7b6e1730689cad
|
| c528dba35b5faece51ca658fc008b688f8b690ad |
|
26-Nov-2013 |
Mathieu Chartier <mathieuc@google.com> |
Enable moving classes.
Slight reduction in Zygote size, memory savings are in the noise.
Before: Zygote size: 8739224
After: Zygote size: 8733568
Fixed a bug where we didn't set the concurrent start bytes after
switching the allocator from bump pointer to ROSAlloc in the
zygote. This caused excessive memory usage.
Added the method verifiers as roots to fix an issue caused by
RegTypes holding a Class*.
Added logic to clear card table in the SemiSpace collector, this
reduces DalvikOther from ~2400k -> ~1760k when using the SemiSpace
collector.
Added a missing lock to the timing loggers which caused a rare
one time crash in std::set.
Bug: 11771255
Bug: 8499494
Bug: 10802951
Change-Id: I99d2b528cd51c1c5ed7012e3220b3aefded680ae
|
| 7bf82af01ec250a4ed2cee03a0e51d179fa820f9 |
|
07-Dec-2013 |
Mathieu Chartier <mathieuc@google.com> |
Fix memory usage regression and clean up collector changing code.
Memory usage regressed since we didn't properly update
concurrent_start_bytes_ when changing collectors.
Bug: 12034247
Change-Id: I1c69e71cd2919e0d3bf75485a4ac0b0aeca59278
|
| e4e23c0b0cfebd01f17732b6686a74f9e437e978 |
|
06-Dec-2013 |
Hiroshi Yamauchi <yamauchi@google.com> |
Fix valgrind-test-art-host-gtest-object_test.
This had been failing with the error message:
object_test F 2360 2360 art/runtime/gc/heap-inl.h:139] Check failed: !running_on_valgrind_
Then, this failing DCHECK was removed in a refactoring in
cbb2d20bea2861f244da2e2318d8c088300a3710, I believe.
This change adds back a DCHECK that's equivalent to the old one and
fixes ObjectTest.CheckAndAllocArrayFromCode by replacing a
CheckAndAllocArrayFromCode call with a
CheckAndAllocArrayFromCodeInstrumented call.
Bug: 11670287
Change-Id: Ica2f707b9a9ff48ef973b9e326a4d9786c1781f8
|
| 95a659fca9cb89d5cf592ad5808d38f0117a583e |
|
22-Nov-2013 |
Hiroshi Yamauchi <yamauchi@google.com> |
Fix a libartd.so boot crash when kMovingCollector is true.
Bug: 11828452
Change-Id: I208a99d7ee7cb6af37046387e97156e3b240cda6
|
| 1febddf359ae500ef1bb01ab4883b076fcb56440 |
|
20-Nov-2013 |
Mathieu Chartier <mathieuc@google.com> |
Set array length before fence in allocation code path.
Could not delete SetLength since it is required by
space_test.
Bug: 11747779
Change-Id: Icf1ead216b6ff1b519240ab0d0ca30d68429d5b6
|
| cbb2d20bea2861f244da2e2318d8c088300a3710 |
|
15-Nov-2013 |
Mathieu Chartier <mathieuc@google.com> |
Refactor allocation entrypoints.
Adds support for switching entrypoints during runtime. Enables
addition of new allocators with out requiring significant copy
paste. Slight speedup on ritzperf probably due to more inlining.
TODO: Ensuring that the entire allocation path is inlined so
that the switch statement in the allocation code is optimized
out.
Rosalloc measurements:
4583
4453
4439
4434
4751
After change:
4184
4287
4131
4335
4097
Change-Id: I1352a3cbcdf6dae93921582726324d91312df5c9
|
| cf58d4adf461eb9b8e84baa8019054c88cd8acc6 |
|
26-Sep-2013 |
Hiroshi Yamauchi <yamauchi@google.com> |
A custom 'runs-of-slots' memory allocator.
Bug: 9986565
Change-Id: I0eb73b9458752113f519483616536d219d5f798b
|
| 590fee9e8972f872301c2d16a575d579ee564bee |
|
13-Sep-2013 |
Mathieu Chartier <mathieuc@google.com> |
Compacting collector.
The compacting collector is currently similar to semispace. It works by
copying objects back and forth between two bump pointer spaces. There
are types of objects which are "non-movable" due to current runtime
limitations. These are Classes, Methods, and Fields.
Bump pointer spaces are a new type of continuous alloc space which have
no lock in the allocation code path. When you allocate from these it uses
atomic operations to increase an index. Traversing the objects in the bump
pointer space relies on Object::SizeOf matching the allocated size exactly.
Runtime changes:
JNI::GetArrayElements returns copies objects if you attempt to get the
backing data of a movable array. For GetArrayElementsCritical, we return
direct backing storage for any types of arrays, but temporarily disable
the GC until the critical region is completed.
Added a new runtime call called VisitObjects, this is used in place of
the old pattern which was flushing the allocation stack and walking
the bitmaps.
Changed image writer to be compaction safe and use object monitor word
for forwarding addresses.
Added a bunch of added SIRTs to ClassLinker, MethodLinker, etc..
TODO: Enable switching allocators, compacting on background, etc..
Bug: 8981901
Change-Id: I3c886fd322a6eef2b99388d19a765042ec26ab99
|
| dfb325e0ddd746cd8f7c2e3723b3a573eb7cc111 |
|
30-Oct-2013 |
Ian Rogers <irogers@google.com> |
Don't use UTF16 length as length for MUTF8.
Bug 11367555.
Change-Id: Ia0b07072a1a49d435c3b71ed9a668b316b7ff5d8
|
| 3b4c18933c24b8a33f38573c2ebcdb9aa16efeb5 |
|
13-Sep-2013 |
Hiroshi Yamauchi <yamauchi@google.com> |
Split the allocation path into 'instrumented' and 'uninstrumented'
ones.
The instrumented path is equivalent to the existing allocation path
that checks for three instrumentation mechanisms (the debugger
allocation tracking, the runtime allocation stats collection, and
valgrind) for every allocation. The uinstrumented path does not
perform these checks. We use the uninstrumented path by default and
enable the instrumented path only when any of the three mechanisms is
enabled. The uninstrumented version of Heap::AllocObject() is inlined.
This change improves the Ritz MemAllocTest by ~4% on Nexus 4 and ~3%
on Host/x86.
Bug: 9986565
Change-Id: I3e68dfff6789d77bbdcea98457b694e1b5fcef5f
|