SIGSEGV in 19.0.1+11 (alpine-musl)
henrik242 opened this issue · 3 comments
henrik242 commented
Using bellsoft/liberica-openjdk-alpine-musl:19@sha256:7ed452eb660c529f1fc6a835909010c9c9abe90ed2438a9d595a06ebaa49aa5f
Getting error:
> [linux/arm64 5/15] RUN cd /tmp && /opt/jdk-19/bin/javac SSLTest.java && /opt/jdk-19/bin/java -cp /tmp SSLTest accounts.google.com api.sendgrid.com kinesis.eu-west-1.amazonaws.com:
#0 1.086 #
#0 1.086 # A fatal error has been detected by the Java Runtime Environment:
#0 1.087 #
#0 1.087 # SIGSEGV (0xb) at pc=0x0000005502231528, pid=8, tid=24
#0 1.094 #
#0 1.096 # JRE version: OpenJDK Runtime Environment (19.0.1+11) (build 19.0.1+11)
#0 1.096 # Java VM: OpenJDK 64-Bit Server VM (19.0.1+11, mixed mode, tiered, compressed oops, compressed class ptrs, g1 gc, linux-aarch64)
#0 1.096 # Problematic frame:
#0 1.096 # V [libjvm.so+0x927528] PhaseCFG::schedule_node_into_block(Node*, Block*)+0x48
#0 1.125 #
#0 1.127 # No core dump will be written. Core dumps have been disabled. To enable core dumping, try "ulimit -c unlimited" before starting Java again
#0 1.128 #
#0 1.139 # An error report file with more information is saved as:
#0 1.139 # /tmp/hs_err_pid8.log
#0 1.388
#0 1.389 [error occurred during error reporting (), id 0xb, SIGSEGV (0xb) at pc=0x0000005502401f28]
#0 1.389
#0 1.391 #
#0 1.392 # If you would like to submit a bug report, please visit:
#0 1.392 # https://bell-sw.com/support
#0 1.392 #
#0 1.400 [thread 12 also had an error]
#0 1.401 qemu: uncaught target signal 6 (Aborted) - core dumped
#0 1.414 Aborted
------
WARNING: No output specified for docker-container driver. Build result will only remain in the build cache. To push result image into registry use --push or to load image into docker use --load
Dockerfile:30
--------------------
29 | COPY SSLTest.java /tmp/
30 | >>> RUN cd /tmp && \
31 | >>> $JAVA_HOME/bin/javac SSLTest.java && \
32 | >>> $JAVA_HOME/bin/java -cp /tmp SSLTest accounts.google.com api.sendgrid.com kinesis.eu-west-1.amazonaws.com
33 |
--------------------
error: failed to solve: process "/bin/sh -c cd /tmp && $JAVA_HOME/bin/javac SSLTest.java && $JAVA_HOME/bin/java -cp /tmp SSLTest accounts.google.com api.sendgrid.com kinesis.eu-west-1.amazonaws.com" did not complete successfully: exit code: 134
SSLTest.java contents:
import javax.net.ssl.SSLSocket;
import javax.net.ssl.SSLSocketFactory;
import java.io.*;
public class SSLTest {
public static void main(String[] args) {
if (args.length < 1) {
System.out.println("Usage: SSLTest host [host...]");
System.exit(1);
}
int failures = 0;
SSLSocketFactory factory = (SSLSocketFactory) SSLSocketFactory.getDefault();
for (int i = 0; i < args.length; i++) {
try {
SSLSocket socket = (SSLSocket) factory.createSocket(args[i], 443);
socket.getOutputStream().write(1);
InputStream in = socket.getInputStream();
while (in.available() > 0) {
System.out.print(in.read());
}
System.out.println("Verified SSL connection to " + args[i]);
} catch (Exception e) {
System.err.println("Could not connect to " + args[i] + ": " + e.toString());
failures++;
}
}
System.exit(failures);
}
}
Weird thing: This worked Dec 5th 2022 with the exact same liberica image.
henrik242 commented
I see some differences in the working build from Dec 5th and now.
Failing:
#8 [linux/amd64 1/15] FROM bellsoft/liberica-openjdk-alpine-musl:19@sha256:7ed452eb660c529f1fc6a835909010c9c9abe90ed2438a9d595a06ebaa49aa5f
#8 sha256:c87dca686b5dddc2363bd9cc87a09a8ea4ac2a09117771312ea3a528bd4e3057 76.71MB / 76.71MB 1.1s done
#8 sha256:213ec9aee27d8be045c6a92b7eac22c9a64b44558193775a1a7f626352392b49 2.81MB / 2.81MB 0.2s done
Working:
#8 [linux/arm64 1/15] FROM bellsoft/liberica-openjdk-alpine-musl:19@sha256:7ed452eb660c529f1fc6a835909010c9c9abe90ed2438a9d595a06ebaa49aa5f
#8 sha256:9b18e9b68314027565b90ff6189d65942c0f7986da80df008b8431276885218e 2.71MB / 2.71MB 0.0s done
#8 sha256:6ce5c5863ba1f43dd93f2e3e882d73c427440e11896b2247b2314dec5a24af93 76.38MB / 76.38MB 0.6s done
voitylov commented
could you post hs_err log here?
henrik242 commented
Frustratingly I'm not able to reproduce this today, while yesterday it failed both on CI (linux host) and locally (macOS host).
The image signatures from above was a false positive, I was looking at the wrong places in the logs.
I'll reopen if it happens again.