JDK-8302147 : Speed up compiler/jvmci/compilerToVM/IterateFramesNative.java
  • Type: Enhancement
  • Component: hotspot
  • Sub-Component: compiler
  • Affected Version: 21
  • Priority: P4
  • Status: Resolved
  • Resolution: Fixed
  • Submitted: 2023-02-09
  • Updated: 2023-02-22
  • Resolved: 2023-02-15
The Version table provides details related to the release that this issue/RFE will be addressed.

Unresolved : Release in which this issue/RFE will be addressed.
Resolved: Release in which this issue/RFE has been resolved.
Fixed : Release in which this issue/RFE has been fixed. The release containing this fix may be available for download as an Early Access Release or a General Availability Release.

To download the current JDK release, click here.
JDK 21
21 b10Fixed
Related Reports
Relates :  
Description
https://github.com/openjdk/jdk/blame/master/test/hotspot/jtreg/compiler/jvmci/compilerToVM/IterateFramesNative.java

Takes about 27 seconds, and is one of the slowest tier1 compiler tests.

Can we speed this test up? Or at least move it to a higher tier?
Comments
Changeset: 46bcc490 Author: Emanuel Peter <epeter@openjdk.org> Date: 2023-02-15 07:36:53 +0000 URL: https://git.openjdk.org/jdk/commit/46bcc4901ea66589ec4c6904f8bab9156422312a
15-02-2023

A pull request was submitted for review. URL: https://git.openjdk.org/jdk/pull/12496 Date: 2023-02-09 15:58:05 +0000
09-02-2023

Talked with Andreas. He says we could try these things: -XX:CompileOnly may only make it slower -> try removing it. Lower iterations, and compile threashold. (see CompilerWhiteBoxTest.THRESHOLD) We just have to make sure to keep this in the commandline, and that it passes with it: -Dcompiler.jvmci.compilerToVM.IterateFramesNative.checkCompiled=true Run like this: time ~/Documents/jtreg/bin/jtreg -va -s -jdk:/home/emanuel/Documents/fork5-jdk/build/linux-x64-debug/jdk/ -nativepath:/home/emanuel/Documents/fork5-jdk/build/linux-x64-debug/images/test/hotspot/jtreg/native -javaoptions:"" /home/emanuel/Documents/fork5-jdk/open/test/hotspot/jtreg/compiler/jvmci/compilerToVM/IterateFramesNative.java It seems we check the compilation after the first go already. So we could just run it once, or at most like 5 times: diff --git a/test/hotspot/jtreg/compiler/jvmci/compilerToVM/IterateFramesNative.java b/test/hotspot/jtreg/compiler/jvmci/compilerToVM/IterateFramesNative.java index ae09e95b024..cc2eff066e4 100644 --- a/test/hotspot/jtreg/compiler/jvmci/compilerToVM/IterateFramesNative.java +++ b/test/hotspot/jtreg/compiler/jvmci/compilerToVM/IterateFramesNative.java @@ -104,7 +104,7 @@ public class IterateFramesNative { } private void test() { - for (int i = 0; i < CompilerWhiteBoxTest.THRESHOLD + 1; i++) { + for (int i = 0; i < 5; i++) { testNativeFrame("someString", i); } }
09-02-2023

[~aw] Dear Andreas, I am currently trying to speed up slow-running tier1-compiler-tests. Is there a way to make it less resource consuming? If not, would it be ok to move it to a higher tier, so that we run it less often?
09-02-2023