Loading...

XML

Word

Printable

Type: Enhancement
Resolution: Fixed
Priority: P3
Fix Version/s: 9
Affects Version/s: 9
Component/s: hotspot
Labels:
- performance

Subcomponent:
compiler
Resolved In Build:
b112

When instantiating arrays from Java, we have to zero the backing storage to match JLS requirements. In some cases, like with the subsequent arraycopy, we are able to remove zeroing. However, in a generic case where a complicated processing is done after allocation, compilers are unable to figure out the array is covered completely. See ~~JDK-8150463~~ for a motivational example of this.

It might make sense to allow new Unsafe method that will return uninitialized arrays to trusted Java code.

Patches:
http://cr.openjdk.java.net/~shade/8150465/webrev.jdk.03/
http://cr.openjdk.java.net/~shade/8150465/webrev.hs.04/

Notes:
http://cr.openjdk.java.net/~shade/8150465/notes.txt

blocks

JDK-8148604 JEP 280, Switch to more optimal concatenation strategy

Resolved

JDK-8150463 StringConcat MH_INLINE_SIZED_EXACT should skip storage initialization

Resolved

relates to

JDK-8028422 Creating uninitialized primitive arrays can boost Java performance

Closed

JDK-8259339 AllocateUninitializedArray C2 intrinsic fails with void.class input

Resolved

JDK-8188835 AllocateUninitializedArray fails with "IllegalStateException: Wrong class: expected = class [I, but got class [C "

Closed

JDK-8150669 C1 intrinsic for Class.isPrimitive

Resolved

(1 relates to)

Assignee:: Aleksey Shipilev

Reporter:: Aleksey Shipilev

Votes:: 1 Vote for this issue

Watchers:: 4 Start watching this issue

Created:: 2016-02-23 10:03

Updated:: 2021-01-06 09:10

Resolved:: 2016-03-03 14:28

Details

Description

Attachments

Issue Links

Activity

People

Dates