[
https://issues.apache.org/jira/browse/CASSANDRA-9954?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14651041#comment-14651041
]
Robert Stupp commented on CASSANDRA-9954:
-----------------------------------------
Background information (byte-code level).
All loops, branches, cases, jumps, gotos etc use _labels_ in the byte code.
The trick that this patch uses is to inject a check-call after each label.
The check itself is quite cheap.
Example "amok UDF code":
{code:java}
while (true)
{
if (System.currentTimeMillis() % 12345 == 99999)
break;
}
return null;
{code}
This would compile to this byte-code (_FRAME_ and _LINENUMBER_ byte-code
omitted for readability):
{noformat}
L0
INVOKESTATIC java/lang/System.currentTimeMillis ()J
LDC 12345
LREM
LDC 99999
LCMP
IFNE L0
L1
GOTO L2
L2
ACONST_NULL
ARETURN
L3
...
{noformat}
The patch injects the following byte-code after each label:
{noformat}
INVOKESTATIC org/apache/cassandra/cql3/functions/JavaUDF.checkTimeout ()Z
IFEQ continueLabel
ACONST_NULL
ARETURN
{noformat}
resulting in:
{noformat}
L0
INVOKESTATIC org/apache/cassandra/cql3/functions/JavaUDF.checkTimeout ()Z
IFEQ LCONTINUE1
ACONST_NULL
ARETURN
LCONTINUE1
INVOKESTATIC java/lang/System.currentTimeMillis ()J
LDC 12345
LREM
LDC 99999
LCMP
IFNE L0
L1
INVOKESTATIC org/apache/cassandra/cql3/functions/JavaUDF.checkTimeout ()Z
IFEQ LCONTINUE2
ACONST_NULL
ARETURN
LCONTINUE2
GOTO L2
L2
INVOKESTATIC org/apache/cassandra/cql3/functions/JavaUDF.checkTimeout ()Z
IFEQ LCONTINUE2
ACONST_NULL
ARETURN
LCONTINUE2
ACONST_NULL
ARETURN
L3
...
{noformat}
> Improve Java-UDF timeout detection
> ----------------------------------
>
> Key: CASSANDRA-9954
> URL: https://issues.apache.org/jira/browse/CASSANDRA-9954
> Project: Cassandra
> Issue Type: Improvement
> Reporter: Robert Stupp
> Assignee: Robert Stupp
> Fix For: 3.x
>
>
> CASSANDRA-9402 introduced a sandbox using a thread-pool to enforce security
> constraints and to detect "amok UDFs" - i.e. UDFs that essentially never
> return (e.g. {{while (true)}}.
> Currently the safest way to react on such an "amok UDF" is to _fail-fast_ -
> to stop the C* daemon since stopping a thread (in Java) is just no solution.
> CASSANDRA-9890 introduced further protection by inspecting the byte-code. The
> same mechanism can also be used to manipulate the Java-UDF byte-code.
> By manipulating the byte-code I mean to add regular "is-amok-UDF" checks in
> the compiled code.
> EDIT: These "is-amok-UDF" checks would also work for _UNFENCED_ Java-UDFs.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
