currentSite should be ThreadLocal because not having this screws up a lot of the instrumentation-based dynamic call graphs we generate for the shootout benchmarks.
I have also conditionally changed how the string of the currentSite is created based on the output format that Julian came up with for the dynamic call graph. This support is necessary, because the Java std libraries are not instrumented. Therefore, they would appear as if calls from them show up from nowhere in the log that WALA generates for the dynamic call graph. This fix make those calls originate from a fake library BLOB node in the call graph.
Of course, doing nothing isn't always the right behavior. Sometimes a
previously-unhandled value is truly unexpected and one should fail by
throwing an exception. It may not always be clear whether an
exception or doing nothing is the right choice. For some `switch`
statements affected by this commit, I initially guessed that throwing
an exception was the right default behavior, but was proven wrong when
doing so caused WALA regression test failures. That's strong evidence
that the unmatched values were not really unexpected, but merely
should have been handled by doing nothing as before.
Previously each of these `switch` statements would implicitly do
nothing if an unanticipated `enum` value came along. My impression is
that each of these `switch` statements is supposed to be exhaustive,
such that an unexpected (unhandled) value should never appear. If one
does, we should recognize it and complain loudly.
Of course, sometimes the right behavior for previously-unhandled
values is to do nothing. It may not always be clear whether an
exception or doing nothing is the right choice. For this commit,
WALA's regression tests still pass even with the possibility of
throwing an exception for unexpected values. If we assume that the
test suite is thorough, that tells me that throwing an exception is
the right policy for each `switch` statement that I'm changing here.
Eclipse was warning that these `switch` statements had no `default`
cases. Each did have some default behavior, but implemented outside
the `switch`. By moving the default behavior into a `default` case
within the `switch`, we eliminate a static warning with no change
whatsoever to the run-time behavior.
In general, my approach was to try to eliminate each unused parameter
using Eclipse's "Change Method Signature" refactoring. That did not
always succeed: a parameter may be unused in some base class method,
but then be used in subclass's override of that method. In cases
where refactoring to eliminate a parameter failed, I instead annotated
the parameter with '@SuppressWarnings("unused")' to silence the
warning.
Note: this group of changes creates a significant risk of
incompatibility for third-party WALA code. Some removed parameters
change externally-visible APIs. Furthermore, these changes do not
necessarily lead to Java compilation errors. For example, suppose
third-party code subclasses a WALA class or interface, overrides a
method, but does not annotate that method as @Override. Removing a
parameter means that the third-party method no longer overrides. This
can quietly change code behavior without compile-time errors or
warnings. This is exactly why one should use @Override wherever
possible, but we cannot guarantee that third-party WALA users have
done that.
Previously some of these were accessing such fields through a subclass
of the declaring class. That creates an unnecessary extra inter-class
dependency lower in the type hierarchy than necessary.
Also, suppress this warning in an automated test input where the
indirect static accesses are explicitly intentional.
If a method is private, there's no risk that a subclass elsewhere
might be overriding it and depending on dynamic dispatch to choose the
right implementation. So all of these private methods can safely be
declared static without risk of regression in either WALA code or
unseen third-party code.
Removing an unused field sometimes means removing constructor code
that used to initialize that field. Removing that initialization code
sometimes leaves whole constructor arguments unused. Removing those
unused arguments can leave us with unused code to compute those
arguments in constructors' callers, and so on. This commit tries to
clean all of this up, working backward from the unused fields that an
earlier commit already removed. Hopefully I have avoided removing
upstream code that had other important side effects, but it wouldn't
hurt for a WALA expert to review this change carefully.
This fixes the remaining 34 Eclipse "Resource '...' should be managed
by try-with-resource" warnings that were still left after the previous
commit.
Unlike the fixes in that previous commit, the changes here are *not*
plugging potential resource leaks. However, in many cases that is
simply because the code before the close() call cannot currently throw
exceptions. If exceptions became possible in the future, leaks could
result. Using try-with-resource preemptively avoids that.
Furthermore, in code that was already dealing with exceptions, the
try-with-resource style is usually considerably simpler.
This fixes 33 out of 37 Eclipse "Potential resource leak: '...' may
not be closed" warnings. It also fixes 3 out of 37 Eclipse "Resource
'...' should be managed by try-with-resource" warnings, although that
was not the main focus of this effort.
The remaining 4 warnings about potential resource leaks all involve a
leaked JarFile instance that is passed to a JarFileModule constructor
call. JarFileModile never attempts to close its underlying JarFile;
this code is written as though JarFile cleanup were the caller's
responsibility. However, the JarFile often cannot be closed by the
code that creates the JarFileModule either, since the JarFile needs to
remain open while the JarFileModule is in use, and some of these
JarFileModules stay around beyond the lifetime of the code that
created them. Truly fixing this would essentially require making
JarFileModule implement Closeable, which in turn would probably
require that Module implement Closeable, which in turn would require
changes to lots of code that deals with Module instances to arrange
for them to be properly closed. That's more invasive than I'm
prepared to take on right now.
Instead, rely on Java's ability to infer type parameters in many
contexts. This removes 665 Eclipse warnings.
Note: a few of these changes are to files under "test" subdirectories.
Presumably those are files that serve as test inputs rather than being
part of WALA code proper. As far as I can tell, these changes do not
break any WALA tests. But if any of those tests were specifically
intended to exercise WALA on code with non-inferred generic type
parameters, then I really should be leaving those alone.
a) serializable added for use by Android services
b) test classes refactored to allow Android variants to use JUnit 3
2) shrike instrumentation now uses java.lang.instrument
a) refactoring
b) online variants of call graph tracing
com.ibm.wala.cast.js.rhino.test/harness-src/com/ibm/wala/cast/js/test/TestPrototypeCallGraphShapeRhino.java
com.ibm.wala.cast.js.test/harness-src/com/ibm/wala/cast/js/test/TestPrototypeCallGraphShape.java
com.ibm.wala.cast.js.test.data/examples-src/pages/prototype.html
work (not yet finished) on fixes to property accesses for JavaScript:
com.ibm.wala.cast/source/java/com/ibm/wala/cast/ipa/callgraph/AstSSAPropagationCallGraphBuilder.java
com.ibm.wala.cast.java/src/com/ibm/wala/cast/java/ipa/callgraph/AstJavaSSAPropagationCallGraphBuilder.java
com.ibm.wala.cast.js/source/com/ibm/wala/cast/js/ipa/callgraph/JSSSAPropagationCallGraphBuilder.java
currently unused tests to remind me to fix bugs:
com.ibm.wala.cast.js.test/harness-src/com/ibm/wala/cast/js/test/TestSimpleCallGraphShape.java
com.ibm.wala.cast.js.test.data/examples-src/tests/loops.js
com.ibm.wala.cast.js.test.data/examples-src/tests/primitive_strings.js
fixes to exception handler code generation in JavaScript:
com.ibm.wala.cast.js.rhino/source/com/ibm/wala/cast/js/translator/RhinoToAstTranslator.java
com.ibm.wala.cast.js.test.data/examples-src/tests/try.js
com.ibm.wala.cast.js.test/harness-src/com/ibm/wala/cast/js/test/TestSimpleCallGraphShape.java
fixes to make the system build on both juno and luna
com.ibm.wala.cast.js.test.data/pom.xml
pom.xml
targets/e42/e42.target
targets/e44/e44.target
targets/pom.xml
com.ibm.wala.core.tests/META-INF/MANIFEST.MF
com.ibm.wala.dalvik.test/META-INF/MANIFEST.MF
com.ibm.wala.ide.jdt.test/META-INF/MANIFEST.MF
com.ibm.wala.ide.jdt/source/com/ibm/wala/cast/java/translator/jdt/FakeExceptionTypeBinding.java
com.ibm.wala.ide.jdt/source/com/ibm/wala/ide/util/JavaEclipseProjectPath.java
com.ibm.wala.ide.jsdt.tests/META-INF/MANIFEST.MF
com.ibm.wala.ide.jsdt.tests/src/com/ibm/wala/ide/jsdt/tests/AbstractJSProjectScopeTest.java
com.ibm.wala.ide/src/com/ibm/wala/ide/util/EclipseProjectPath.java
com.ibm.wala.ide/src/com/ibm/wala/ide/util/ProgressMonitorDelegate.java
beginnings of "pointer analysis" on top of field-based analysis
com.ibm.wala.cast.js/source/com/ibm/wala/cast/js/callgraph/fieldbased/flowgraph/FlowGraph.java
com.ibm.wala.cast.js/source/com/ibm/wala/cast/js/callgraph/fieldbased/flowgraph/vertices/PropVertex.java
com.ibm.wala.cast.js/source/com/ibm/wala/cast/js/callgraph/fieldbased/flowgraph/vertices/RetVertex.java
com.ibm.wala.cast.js/source/com/ibm/wala/cast/js/callgraph/fieldbased/flowgraph/vertices/VarVertex.java
com.ibm.wala.cast.js/source/com/ibm/wala/cast/js/callgraph/fieldbased/flowgraph/vertices/VertexFactory.java
com.ibm.wala.core/src/com/ibm/wala/ipa/callgraph/propagation/PointerAnalysis.java
com.ibm.wala.core/src/com/ibm/wala/ipa/callgraph/propagation/cfa/ExceptionReturnValueKey.java
fixes for crashes in correlartion tracking
com.ibm.wala.cast.js/source/com/ibm/wala/cast/js/ipa/callgraph/correlations/extraction/ClosureExtractor.java
fixes for Dalvik IR generation
com.ibm.wala.core/src/com/ibm/wala/cfg/BytecodeCFG.java
com.ibm.wala.core/src/com/ibm/wala/cfg/ShrikeCFG.java
com.ibm.wala.core/src/com/ibm/wala/ssa/SSACFG.java
com.ibm.wala.dalvik.test/source/com/ibm/wala/dalvik/drivers/APKCallGraphDriver.java
com.ibm.wala.dalvik.test/source/com/ibm/wala/dalvik/test/callGraph/JVMLDalvikComparison.java
com.ibm.wala.dalvik/src/com/ibm/wala/dalvik/classLoader/DexCFG.java
com.ibm.wala.dalvik/src/com/ibm/wala/dalvik/dex/instructions/UnaryOperation.java
com.ibm.wala.dalvik/src/com/ibm/wala/dalvik/ssa/AbstractIntRegisterMachine.java
com.ibm.wala.dalvik/src/com/ibm/wala/dalvik/ssa/DexSSABuilder.java
fixes to stack map generation when instrumenting for Java 7
com.ibm.wala.shrike/src/com/ibm/wala/shrike/cg/DynamicCallGraph.java
com.ibm.wala.shrike/src/com/ibm/wala/shrikeBT/ConstantInstruction.java
com.ibm.wala.shrike/src/com/ibm/wala/shrikeBT/analysis/Analyzer.java
com.ibm.wala.shrike/src/com/ibm/wala/shrikeBT/analysis/ClassHierarchy.java
com.ibm.wala.shrike/src/com/ibm/wala/shrikeBT/analysis/Verifier.java
com.ibm.wala.shrike/src/com/ibm/wala/shrikeBT/shrikeCT/ClassInstrumenter.java
com.ibm.wala.shrike/src/com/ibm/wala/shrikeCT/StackMapConstants.java
com.ibm.wala.shrike/src/com/ibm/wala/shrikeCT/StackMapTableReader.java
com.ibm.wala.shrike/src/com/ibm/wala/shrikeCT/StackMapTableWriter.java
2) make instrumentor preserve the names of jar entries and classes as
they are input, rather than recomputing class names when writing the
output jar. This usually makes no difference, but can preserve broken
structures when the input jar file has mismatches between class names
and its entry names.
Julian Dolby
Karim Ali
Ben Liblit
Julian Dolby
Juergen Graf
Manu Sridharan
msridhar1
sjfink
dolby-oss