Disable SVE when detecting Java version by mpdude · Pull Request #4765 · apache/solr

mpdude · 2026-08-19T17:11:54Z

I am trying to start Solr in a Docker container in a virtual arm64 Linux machine on Apple Silicon hardware using Parallels Desktop virtualization. Seems this configuration does not support SVE vector length detection.

When bin/solr runs java -version, this makes the JVM emit a notice:

Starting Solr
Your current version of Java is too old to run this version of Solr.
We found major version , using command '/opt/java/openjdk/bin/java -version', with response:
OpenJDK 64-Bit Server VM warning: Unable to get SVE vector length on this system. Disabling SVE. Specify -XX:UseSVE=0 to shun this warning.
openjdk version "25.0.3" 2026-04-21 LTS
OpenJDK Runtime Environment Temurin-25.0.3+9 (build 25.0.3+9-LTS)
OpenJDK 64-Bit Server VM Temurin-25.0.3+9 (build 25.0.3+9-LTS, mixed mode, sharing)

Please install latest version of Java 21 or set JAVA_HOME properly.

Version detection trips over this extra warning line. So, improve version parsing logic to go over warning lines and lock to the first version line.

HoustonPutman · 2026-08-19T18:08:36Z

That's weird. When I run the solr:10 image via Docker Desktop on Apple Silicon (which will make a linux vm), I see no issue.

$ java -version
openjdk version "25.0.3" 2026-04-21 LTS
OpenJDK Runtime Environment Temurin-25.0.3+9 (build 25.0.3+9-LTS)
OpenJDK 64-Bit Server VM Temurin-25.0.3+9 (build 25.0.3+9-LTS, mixed mode, sharing)

Starting Solr
[0.001s][warning][pagesize] UseLargePages disabled, no large pages configured and available on the system.
WARNING: Using incubator modules: jdk.incubator.vector
WARNING: A terminally deprecated method in sun.misc.Unsafe has been called
WARNING: sun.misc.Unsafe::arrayBaseOffset has been called by com.lmax.disruptor.RingBufferFields (file:/opt/solr-10.0.0/server/lib/ext/disruptor-3.4.4.jar)
WARNING: Please consider reporting this to the maintainers of class com.lmax.disruptor.RingBufferFields
WARNING: sun.misc.Unsafe::arrayBaseOffset will be removed in a future release
2026-08-19 18:02:56.622 INFO  (main) [] o.e.j.s.Server jetty-12.0.27; built: 2025-09-10T23:47:49.595Z; git: 3569a3e83ad136ee44e26b370b74c1c5e9f33e61; jvm 25.0.3+9-LTS
2026-08-19 18:02:56.740 INFO  (main) [] o.e.j.e.w.StandardDescriptorProcessor NO JSP Support for /solr, did not find org.apache.jasper.servlet.JspServlet
2026-08-19 18:02:56.749 INFO  (main) [] o.e.j.s.h.ContextHandler Started oeje10w.WebAppContext@85ec632{solr-jetty-context,/solr,b=file:///opt/solr-10.0.0/server/solr-webapp/webapp/,a=AVAILABLE,h=oeje10s.SessionHandler@1c05a54d{STARTED}}{/opt/solr-10.0.0/server/solr-webapp/webapp}

risdenk · 2026-08-19T18:51:08Z

I think we should just make this regex better instead?

mpdude · 2026-08-20T12:13:22Z

It's a virtual machine run by Parallels Desktop, that probably comes into play as well 🫤.

Either way, I'll improve the version parsing regex and hope that it's an acceptable change.

I am trying to start Solr in a Docker container in a virtual `arm64` Linux machine on Apple Silicon hardware. To my knowledge, this configuration does (currently?) not support SVE vector length detection. When `bin/solr` runs `java -version`, this makes the JVM emit a notice: ``` Starting Solr Your current version of Java is too old to run this version of Solr. We found major version , using command '/opt/java/openjdk/bin/java -version', with response: OpenJDK 64-Bit Server VM warning: Unable to get SVE vector length on this system. Disabling SVE. Specify -XX:UseSVE=0 to shun this warning. openjdk version "25.0.3" 2026-04-21 LTS OpenJDK Runtime Environment Temurin-25.0.3+9 (build 25.0.3+9-LTS) OpenJDK 64-Bit Server VM Temurin-25.0.3+9 (build 25.0.3+9-LTS, mixed mode, sharing) Please install latest version of Java 21 or set JAVA_HOME properly. ``` Version detection trips over this extra warning line. So, improve version parsing logic to go over warning lines and lock to the first `version` line.

mpdude · 2026-08-20T12:19:17Z

Changed to lock to the first version line instead of using head -1; rebased, squashed and updated initial PR description.

risdenk · 2026-08-23T01:02:12Z

  exit 1
 else
-  JAVA_VER_NUM=$(echo "$JAVA_VER" | grep -v '_OPTIONS' | head -1 | awk -F '"' '/version/ {print $2}' | sed -e's/^1\.//' | sed -e's/[._-].*$//')
+  JAVA_VER_NUM=$(echo "$JAVA_VER" | grep -v '_OPTIONS' | awk -F '"' '/ version "/ {print $2; exit}' | sed -e's/^1\.//' | sed -e's/[._-].*$//')


have these been run against a bunch of different jdk versions to see if this works?

mpdude · 2026-08-25T13:20:57Z

I haven't actually run the changed script against different Java/JVM versions since I don't have them at hand.

However, AI researched a few typical -version outputs for me, and here is a bash script that verifies the updated grep/awk expression against those patterns.

Does that help?

#!/bin/bash
# Standalone test for the JAVA_VER_NUM extraction logic in solr/bin/solr (line ~178).
# Run: bash test-java-version-parsing.sh

set -u

extract() {
  echo "$1" | grep -v '_OPTIONS' | awk -F '"' '/ version "/ {print $2; exit}' | sed -e's/^1\.//' | sed -e's/[._-].*$//'
}

pass=0
fail=0

check() {
  local desc="$1" input="$2" expected="$3"
  local actual
  actual=$(extract "$input")
  if [[ "$actual" == "$expected" ]]; then
    printf 'PASS  %-55s got %s\n' "$desc" "$actual"
    pass=$((pass+1))
  else
    printf 'FAIL  %-55s expected %s, got "%s"\n' "$desc" "$expected" "$actual"
    fail=$((fail+1))
  fi
}

check "Java 8, old 1.x scheme" \
'java version "1.8.0_412"
Java(TM) SE Runtime Environment (build 1.8.0_412-b08)
Java HotSpot(TM) 64-Bit Server VM (build 25.412-b08, mixed mode)' \
"8"

check "Java 11 OpenJDK (Temurin)" \
'openjdk version "11.0.24" 2024-07-16
OpenJDK Runtime Environment Temurin-11.0.24+8 (build 11.0.24+8)
OpenJDK 64-Bit Server VM Temurin-11.0.24+8 (build 11.0.24+8, mixed mode)' \
"11"

check "Java 17 IBM Semeru/OpenJ9 + JDK_JAVA_OPTIONS noise line" \
'picked up JDK_JAVA_OPTIONS: -Xshare:off
openjdk version "17.0.11" 2024-04-16
IBM Semeru Runtime Open Edition 17.0.11.0 (build 17.0.11+9)
Eclipse OpenJ9 VM 17.0.11.0 (build openj9-0.44.0, JRE 17 Linux amd64-64-Bit Compressed References 20240416_547)' \
"17"

check "Java 21 OpenJDK (Temurin, LTS)" \
'openjdk version "21.0.4" 2024-07-16 LTS
OpenJDK Runtime Environment Temurin-21.0.4+7 (build 21.0.4+7-LTS)
OpenJDK 64-Bit Server VM Temurin-21.0.4+7 (build 21.0.4+7-LTS, mixed mode, sharing)' \
"21"

check "Java 21 GraalVM" \
'java version "21.0.2" 2024-01-16
Java(TM) SE Runtime Environment Oracle GraalVM 21.0.2+13.1 (build 21.0.2+13-jvmci-23.1-b30)
Java HotSpot(TM) 64-Bit Server VM Oracle GraalVM 21.0.2+13.1 (build 21.0.2+13-jvmci-23.1-b30, mixed mode, sharing)' \
"21"

check "Java 25, SVE warning line BEFORE version line (this PR's bug report)" \
'OpenJDK 64-Bit Server VM warning: Unable to get SVE vector length on this system. Disabling SVE. Specify -XX:UseSVE=0 to shun this warning.
openjdk version "25.0.3" 2026-04-21 LTS
OpenJDK Runtime Environment Temurin-25.0.3+9 (build 25.0.3+9-LTS)
OpenJDK 64-Bit Server VM Temurin-25.0.3+9 (build 25.0.3+9-LTS, mixed mode, sharing)' \
"25"

echo
echo "For comparison, the OLD logic (head -1 | awk '/version/') on the SVE-warning case:"
old_result=$(echo 'OpenJDK 64-Bit Server VM warning: Unable to get SVE vector length on this system. Disabling SVE. Specify -XX:UseSVE=0 to shun this warning.
openjdk version "25.0.3" 2026-04-21 LTS' | grep -v '_OPTIONS' | head -1 | awk -F '"' '/version/ {print $2}' | sed -e's/^1\.//' | sed -e's/[._-].*$//')
echo "  -> \"${old_result}\" (empty = the reported bug: JAVA_VER_NUM ends up blank)"

echo
echo "$pass passed, $fail failed"
[[ $fail -eq 0 ]]

github-actions Bot added the start-scripts label Aug 19, 2026

risdenk reviewed Aug 19, 2026

View reviewed changes

mpdude force-pushed the patch-1 branch from 2627b02 to 8f313a4 Compare August 20, 2026 12:18

risdenk reviewed Aug 23, 2026

View reviewed changes

Navigation Menu

Uh oh!

Disable SVE when detecting Java version - #4765

Disable SVE when detecting Java version#4765
mpdude wants to merge 1 commit into
apache:mainfrom
mpdude:patch-1

mpdude commented Aug 19, 2026 •

edited

Loading

Uh oh!

HoustonPutman commented Aug 19, 2026

Uh oh!

risdenk Aug 19, 2026

Uh oh!

mpdude commented Aug 20, 2026

Uh oh!

mpdude commented Aug 20, 2026

Uh oh!

risdenk Aug 23, 2026

Uh oh!

mpdude commented Aug 25, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Uh oh!

Conversation

mpdude commented Aug 19, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

HoustonPutman commented Aug 19, 2026

Uh oh!

risdenk Aug 19, 2026

Choose a reason for hiding this comment

Uh oh!

mpdude commented Aug 20, 2026

Uh oh!

mpdude commented Aug 20, 2026

Uh oh!

risdenk Aug 23, 2026

Choose a reason for hiding this comment

Uh oh!

mpdude commented Aug 25, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

mpdude commented Aug 19, 2026 •

edited

Loading