Patch: update instrumentation order to fix coverage

[kmnls]

Reason:
JaCoCo produces wrong coverage because the coverage instrumentation happens after the hooks are set.
This instrumentation should follow the JaCoCo algorithm but in fact they produce different results.
Both algorithms inject control points into the end of each BB where the INVOKE happens.
BBs where there are no INVOKEs are not marked.
Let's check the case: BBs originally did not have any INVOKE but hooks are injected traceCmp callback.
CoverageRecorder happily adds a control point into such a BB but JaCoCo will never know about this.
As the result, all points after the mentioned will be misplaced in JaCoCo coverage.

Sample for test:

package com.example;

import java.util.Random;

public class Junit5Example1 {

public static boolean earlyReturn = false;

public static boolean logic(String input) {
    long random = new Random(42).nextLong();
    boolean cmp = input.startsWith("magicstring" + random);
    if (earlyReturn) {
        return true;
    }
    if (cmp
            && input.length() > 35
            && input.charAt(35) == 'C') {
        mustNeverBeCalled();
    }
    return true;
}

private static void mustNeverBeCalled() {
    throw new Error("mustNeverBeCalled has been called");
}

}

All the lines after if (earlyReturn) will be colored wrongly because this BB will have an additional coverage point after traceCmp.
wrong.Junit5Example1.java.zip
right.Junit5Example1.java.zip

[fmeum]

Looks good, thanks!

[kmnls]

@fmeum Could you please merge?

[kmnls]

'Build all targets and run all tests / build_and_test (windows-2019, 8) (pull_request) '
is constantly failing with
'ERROR: C:/users/runneradmin/_bazel_runneradmin/2gbstkcc/external/io_bazel_rules_kotlin/src/main/kotlin/BUILD.bazel:37:12: Building external/io_bazel_rules_kotlin/src/main/kotlin/build.jar () [for tool] failed: failed to delete output files before executing action'
Probably it may be ignored?

[fmeum]

@kmnls Yes, that's not due to this PR. I will keep trying.

[fmeum]

@kmnls The CI failure should be fixed, but there is a test that seems to fail for reasons related to this PR. Could you take a look?

[fmeum]

The test times out with:

== Java Exception: java.lang.IllegalStateException: Expected "11", got "10"
	at com.example.CoverageFuzzer.assertEquals(CoverageFuzzer.java:189)
	at com.example.CoverageFuzzer.assertCoverageDump(CoverageFuzzer.java:174)
	at com.example.CoverageFuzzer.fuzzerTearDown(CoverageFuzzer.java:92)
	at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
	at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:77)
	at java.base/jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
	at java.base/java.lang.reflect.Method.invoke(Method.java:568)
	at com.code_intelligence.jazzer.driver.FuzzTargetRunner.shutdown(FuzzTargetRunner.java:371)
	at java.base/java.lang.Thread.run(Thread.java:833)

It looks like you might just have to update the count in

jazzer/tests/src/test/java/com/example/CoverageFuzzer.java

Line 174 in 5a3496b

assertEquals(11, countHits(classToCoverCoverage.getProbes()));

.