Details
-
Bug
-
Status: Closed
-
Major
-
Resolution: Fixed
-
3.0.2 JBIG2
-
None
Description
We use pdfbox-jbig2 3.02 in a multi-threaded environment, and the SoftReferenceCache implementation is a problem because of its usage of an unsynchronized WeakHashMap.
Today, we found two threads blocking each other with the same stack trace:
pool-95-thread-1 - priority:5 - threadId:0x00007f882c109800 - nativeId:0x3d81 - nativeId (decimal):15745 - state:RUNNABLE stackTrace: java.lang.Thread.State: RUNNABLE at java.util.WeakHashMap.get(WeakHashMap.java:403) at org.apache.pdfbox.jbig2.util.cache.SoftReferenceCache.get(SoftReferenceCache.java:40) at org.apache.pdfbox.jbig2.JBIG2ImageReader.read(JBIG2ImageReader.java:243) at javax.imageio.ImageReader.read(ImageReader.java:939) at sun.reflect.GeneratedMethodAccessor3824.invoke(Unknown Source) at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43) at java.lang.reflect.Method.invoke(Method.java:498) at org.icepdf.core.pobjects.graphics.images.JBig2Decoder.decodeJbig2(JBig2Decoder.java:161) at org.icepdf.core.pobjects.graphics.images.JBig2Decoder.decode(JBig2Decoder.java:80) at org.icepdf.core.pobjects.graphics.images.ImageStream.getImage(ImageStream.java:94) at org.icepdf.core.pobjects.graphics.images.references.ImageStreamReference.call(ImageStreamReference.java:74) at org.icepdf.core.pobjects.graphics.images.references.ImageStreamReference.call(ImageStreamReference.java:40) at java.util.concurrent.FutureTask.run(FutureTask.java:266) at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624) at java.lang.Thread.run(Thread.java:748) Locked ownable synchronizers: - <0x0000000600019a48> (a java.util.concurrent.ThreadPoolExecutor$Worker)
I couldn't manage to write a unit test to reproduce the problem unfortunately. But the patch is straightforward and attached to this ticket.