Bug Descriptions

FindBugs Bug Descriptions

This document lists the standard bug patterns reported by  version 0.8.5.

Summary

序号

等级

Description

Category

14

A

EC: Call to equals()    with null argument

调用Equal方法的参数为空

Correctness

15

A

EC:    Call to equals() comparing different interface types

调用Equal方法去比较不同的接口类型

Correctness

16

A

EC: Call    to equals() comparing different types

调用Equal方法去比较不同的数据类型

Correctness

17

A

ES:    Comparison of String objects using == or !=

使用== 或者 != 去比较字符串或对象

Correctness

29

B

IJU:    TestCase implements setUp but doesn't call super.setUp()

测试实例实现了setUp()但没有调用super.setUp()

Correctness

36

A

NP: Null    pointer dereference in method

方法中存在空指针

Correctness

37

A

NP:    Null pointer dereference in method on exception path

方法中存在某异常路径中的空指针

Correctness

38

A

NP:    Possible null pointer dereference in method

方法中存在可能在运行时出现的空指针异常

Correctness

39

A

NP: Possible null pointer    dereference in method on exception path

方法中存在可能在某些异常路径下出现的空指针异常

Correctness

41

C

Nm: Class defines    equal(); should it be equals()?

类中定义了equal方法,但是没有重写java.lang.Object中equal方法,这也有可能是有意这么做的

Correctness

42

B

Nm: Confusing    method names

一些方法名只有大小写的区别,容易混绕

Correctness

43

B

Nm:    Confusing method name

方法名和其超类的类名相同

Correctness

44

C

Nm: Class    defines hashcode(); should it be hashCode()?

类中定义了hashCode()方法,但是没有重写java.lang.Object中的hashCode()方法,这也有可能是有意这么做的

Correctness

45

C

Nm: Class    defines tostring(); should it be toString()?

类中定义了tostring()方法,但是没有重写java.lang.Object中的tostring()方法,这也有可能是有意这么做的

Correctness

46

A

Nm: Very    confusing method names

一些方法名只有大小写的区别,特别容易混绕

Correctness

47

A

ODR:    Method may fail to close database resource

方法可能没有关闭数据源

Correctness

48

A

ODR: Method may fail to    close database resource on exception

方法在异常时可能无法关闭数据源

Correctness

49

A

OS: Method    may fail to close stream

方法可能没有关闭流

Correctness

50

A

OS:    Method may fail to close stream on exception

方法在异常时可能无法关闭流

Correctness

53

C

RCN: Redundant comparision    to null of previously checked value

预测值的几个比较的值中有null值

Correctness

54

C

RCN: Redundant comparison to null

几个比较的值中有null值

Correctness

57

C

RV:    Method ignores return value

方法忽略了返回值

Correctness

59

A

SI: Static initializer for class    creates instance before all static final fields assigned

类的静态初始化器实例化一个该类对象要在所有的常量属性定义之前

Correctness

72

C

UwF:    Unwritten field

永远不会写到的字段,所有读取将返回默认值

Correctness

83

A

MS: Field    isn't final but should be

字段应该为常量的却不是常量

Malicious code vulnerability

86

A

IS2:    Inconsistent synchronization

类中字段访问不同步

Multithreaded correctness

88

A

ML:    Method synchronizes on an updated field

方法从一个可变字段对象的引用中同步。这是不太可能有用的,因为不同的线程可以在不同的对象上同步。

Multithreaded correctness

92

C

Ru: Invokes    run on a thread (did you mean to start it instead?)

当一个类继承Runnable接口时候,还去调用run方法,实际上只需要调用Start方法就可以启动新的线程,

Multithreaded correctness

93

C

SC:    Constructor invokes Thread.start()

构造函数调用Thread.start()

Multithreaded correctness

102

C

Dm: Method    invokes dubious new String(String) constructor; just use the argument

方法中调用String的构造函数来新建一个字符串,直接使用其参数即可

Performance

103

A

Dm:    Method invokes dubious String.equals(""); use String.length() ==    0 instead

方法中调用了不确定的String.equals(""),使用String.length() == 0来代替

Performance

104

A

Dm:    Method invokes toString() method on a String; just use the String

方法中用一个字符串去调用方法toString(),用String就可以了

Performance

105

C

Dm:    Method invokes dubious new String() constructor; just use ""

方法中新建一个空字符串用new String()。直接用“”就可以了

Performance

106

C

FI: Empty finalizer    should be deleted

finalizer()方法是没有用的,所以应该予以删除

Performance

114

C

UPM:    Private method is never called

这个私有方法从来没有被调用到

Performance

115

C

UrF: Unread    field

读不到的字段,应该从类中删除

Performance

116

C

UuF: Unused    field

读不到的字段,应该从类中删除

Performance

Descriptions

AM: Creates an  empty jar file entry

The code calls putNextEntry(),  immediately followed by a call to closeEntry(). This  results in an empty JarFile entry. The contents of the entry should be  written to the JarFile between the calls to putNextEntry() and closeEntry().

AM: Creates an  empty zip file entry

The code calls putNextEntry(),  immediately followed by a call to closeEntry(). This  results in an empty ZipFile entry. The contents of the entry should be  written to the ZipFile between the calls to putNextEntry() and closeEntry().

BIT: Incompatible bit masks

This method compares an expression of the form (a &  C) to D, which will always compare unequal due to the specific values of  constants C and D. This may indicate a logic error or typo.

BIT: Incompatible bit masks

This method compares an expression of the form (a &  0) to 0, which will always compare equal. This may indicate a logic error or  typo.

BIT: Incompatible bit masks

This method compares an expression of the form (a | C) to  D. which will always compare unequal due to the specific values of constants  C and D. This may indicate a logic error or typo.

Typically, this bug occurs because the code wants to  perform a membership test in a bit set, but uses the bitwise OR operator  ("|") instead of bitwise AND ("&").

BOA: Class  overrides a method implemented in super class Adapter wrongly

This method overrides a method found in a parent class,  where that class is an Adapter that implements a listener defined in the  java.awt.event or javax.swing.event package. As a result, this method will  not get called when the event occurs.

BRSA: Method  attempts to access a result set field with index 0

A call to getXXX or updateXXX methods of a result set was  made where the field index is 0. As ResultSet fields start at index 1, this  is always a mistake.

CN: Class implements Cloneable but  does not define or use clone method

Class implements Cloneable but does not define or use the  clone method.

CN: clone method does  not call super.clone()

This class defines a clone() method that does not call  super.clone(), and is not final. If this class ("A") is  extended by a subclass ("B"), and the subclass B  calls super.clone(), then it is likely that B's clone() method will  return an object of type A, which violates the standard contract for  clone().

If all clone() methods call super.clone(), then they are  guaranteed to use Object.clone(), which always returns an object of the  correct type.

Co: Covariant compareTo()  method defined

This class defines a covariant version of compareTo().  To correctly override the compareTo() method in  the Comparable interface, the parameter of compareTo() must have  type java.lang.Object.

DE: Method might drop exception

This method might drop an exception.  In general,  exceptions should be handled or reported in some way, or they should be  thrown out of the method.

DE: Method might ignore  exception

This method might ignore an exception.  In general,  exceptions should be handled or reported in some way, or they should be  thrown out of the method.

Dm: Method invokes System.exit(...)

Invoking System.exit shuts down the entire Java virtual machine. This should only been done when it is appropriate. Such calls make it hard or impossible for your code to be invoked by other code. Consider throwing a RuntimeException instead.

EC: Call to equals() with null  argument

This method calls equals(Object), passing a null value as  the argument. According to the contract of the equals() method, this call  should always return false.

EC: Call to equals()  comparing different interface types

This method calls equals(Object) on two references of  unrelated interface types (neither is a subtype of the other). According to  the contract of equals(), objects of different classes should always compare  as unequal. Note that it is possible that the program contains classes that  implement both interfaces, so the comparison may be valid. However, it is  worth inspecting closely.

EC: Call to equals()  comparing different types

This method calls equals(Object) on two references of  different types. According to the contract of equals(), objects of different  classes should always compare as unequal; therefore, it is likely that the  result of this comparision will always be false at runtime.

ES: Comparison of  String objects using == or !=

This code compares java.lang.String  objects for reference equality using the == or != operators. Unless both  strings are either constants in a source file, or have been interned using  the String.intern() method, the same string value may be represented by two different  String objects. Consider using the equals(Object) method  instead.

Eq: Covariant equals()  method defined

This class defines a covariant version of equals().   To correctly override the equals() method in java.lang.Object, the parameter of equals() must have type java.lang.Object.

Eq: Covariant equals()  method defined, Object.equals(Object) inherited

This class defines a covariant version of the equals()  method, but calls the normal equals(Object) method defined in the  base java.lang.Object class.  The class should probably define a non-covariant  version of equals().  (I.e., a method with the signature boolean equals(java.lang.Object).

FI: Explicit invocation  of finalizer

This method contains an explicit invocation of the finalize()  method on an object.  Because finalizer methods are supposed to be  executed once, and only by the VM, this is a bad idea.

FI: Finalizer does not  call superclass finalizer

This finalize() method does not make a call to its superclass's finalize()  method.  So, any finalizer actions defined for the superclass will not  be performed.  Add a call to super.finalize().

FI: Finalizer nullifies  superclass finalizer

This empty finalize() method  explicitly negates the effect of any finalizer defined by its  superclass.  Any finalizer actions defined for the superclass will not  be performed.  Unless this is intended, delete this method.

HE: Class defines  equals() but not hashCode()

This class overrides equals(Object), but  does not override hashCode().  Therefore, the class may violate the invariant that equal  objects must have equal hashcodes.

HE: Class defines  equals() and uses Object.hashCode()

This class overrides equals(Object), but does  not override hashCode(), and inherits the implementation of hashCode()  from java.lang.Object (which returns the identity hash code, an arbitrary value  assigned to the object by the VM).  Therefore, the class is very likely  to violate the invariant that equal objects must have equal hashcodes.

If you don't want to define a hashCode method, and/or  don't believe the object will ever be put into a HashMap/Hashtable, define  the hashCode() method to throw UnsupportedOperationException.

HE: Class defines hashCode()  but not equals()

This class defines a hashCode() method but  not an equals() method.  Therefore, the class may violate the invariant that  equal objects must have equal hashcodes.

HE: Class  defines hashCode() and uses Object.equals()

This class defines a hashCode() method but  inherits its equals() method from java.lang.Object (which defines equality by comparing object references).   Although this will probably satisfy the contract that equal objects must have  equal hashcodes, it is probably not what was intended by overriding the hashCode()  method.  (Overriding hashCode() implies that the object's  identity is based on criteria more complicated than simple reference  equality.)

HE: Class  inherits equals() and uses Object.hashCode()

This class inherits equals(Object) from  an abstract superclass, and hashCode() from from java.lang.Object (which returns the identity hash code, an arbitrary value  assigned to the object by the VM).  Therefore, the class is very likely  to violate the invariant that equal objects must have equal hashcodes.

If you don't want to define a hashCode method, and/or  don't believe the object will ever be put into a HashMap/Hashtable, define  the hashCode() method to throw UnsupportedOperationException.

IC: Initialization  circularity

A circularity was detected in the static initializers of  the two classes referenced by the bug instance.  Many kinds of  unexpected behavior may arise from such circularity.

IJU: TestCase implements  setUp but doesn't call super.setUp()

Class is a JUnit TestCase and implements the setUp  method. The setUp method should call super.setUp(), but doesn't.

IMSE: Dubious catching of  IllegalMonitorStateException

IllegalMonitorStateException is generally only thrown in  case of a design flaw in your code (calling wait or notify on an object you  do not hold a lock on).

It: Iterator next() method  can't throw NoSuchElement exception

This class implements the java.util.Iterator  interface.  However, its next() method is not capable of  throwing java.util.NoSuchElementException.  The next() method should be changed so it throws NoSuchElementException if is called when there are no more elements to return.

MF: Class defines field  that obscures a superclass field

This class defines a field with the same name as a  visible instance field in a superclass. This is confusing, and may indicate  an error if methods update or access one of the fields when they wanted the  other.

MF: Method defines a  variable that obscures a field

This method defines a local variable with the same name  as a field in this class or a superclass. This may cause the method to read  an uninitialized value from the field, leave the field uninitialized, or  both.

MWN: Mismatched notify()

This method calls Object.notify() or Object.notifyAll()  without obviously holding a lock on the object.  Calling notify() or  notifyAll() without a lock held will result in an IllegalMonitorStateException being thrown.

MWN: Mismatched wait()

This method calls Object.wait() without obviously holding  a lock on the object.  Calling wait() without a lock held will result in  an IllegalMonitorStateException being thrown.

NP: Null pointer dereference in  method

A null pointer is dereferenced here.  This will lead  to a NullPointerException when the code is executed.

NP: Null pointer  dereference in method on exception path

A pointer which is null on an exception path is  dereferenced here.  This will lead to a NullPointerException  when the code is executed.  Note that because FindBugs currently does  not prune infeasible exception paths, this may be a false warning.

Also note that FindBugs considers the default case of a  switch statement to be an exception path, since the default case is often  infeasible.

NP: Possible null pointer  dereference in method

A reference value dereferenced here might be null at runtime.   This may lead to a NullPointerException when the code is executed.

NP: Possible  null pointer dereference in method on exception path

A reference value which is null on some exception control  path is dereferenced here.  This may lead to a NullPointerException when the code is executed.  Note that because FindBugs  currently does not prune infeasible exception paths, this may be a false  warning.

Also note that FindBugs considers the default case of a  switch statement to be an exception path, since the default case is often  infeasible.

NS: Questionable use of  non-short-circuit logic

This code seems to be using non-short-circuit logic  (e.g., & or |) rather than short-circuit logic (&& or ||).  Non-short-circuit logic causes both sides of the expression to be evaluated  even when the result can be inferred from knowing the left-hand side. This  can be less efficient and can result in errors if the left-hand side guards  cases when evaluating the right-hand side can generate an error.

Nm: Class defines equal(); should  it be equals()?

This class defines a method equal(Object).  This method does not override the equals(Object) method in java.lang.Object, which is probably what was intended.

Nm: Confusing method names

The referenced methods have names that differ only by  capitalization.

Nm: Confusing method  name

This method has the same name as the superclass of the  class it is defined in.

Nm: Class defines  hashcode(); should it be hashCode()?

This class defines a method called hashcode().   This method does not override the hashCode() method in java.lang.Object, which is probably what was intended.

Nm: Class defines  tostring(); should it be toString()?

This class defines a method called tostring().   This method does not override the toString() method in java.lang.Object, which is probably what was intended.

Nm: Very confusing method  names

The referenced methods have names that differ only by  capitalization.

ODR: Method may  fail to close database resource

The method creates a database resource (such as a  database connection or row set), does not assign it to any fields, pass it to  other methods, or return it, and does not appear to close the object on all  paths out of the method.  Failure to close database resources on all  paths out of a method may result in poor performance, and could cause the  application to have problems communicating with the database.

ODR:  Method may fail to close database resource on exception

The method creates a database resource (such as a  database connection or row set), does not assign it to any fields, pass it to  other methods, or return it, and does not appear to close the object on all  exception paths out of the method.  Failure to close database resources  on all paths out of a method may result in poor performance, and could cause  the application to have problems communicating with the database.

OS: Method may fail to close  stream

The method creates an IO stream object, does not assign  it to any fields, pass it to other methods, or return it, and does not appear  to close the stream on all paths out of the method.  This may result in  a file descriptor leak.  It is generally a good idea to use a finally  block to ensure that streams are closed.

OS: Method may  fail to close stream on exception

The method creates an IO stream object, does not assign  it to any fields, pass it to other methods, or return it, and does not appear  to close it on all possible exception paths out of the method.  This may  result in a file descriptor leak.  It is generally a good idea to use a finally  block to ensure that streams are closed.

PZLA: Consider  returning a zero length array rather than null

It is often a better design to return a length zero array  rather than a null reference to indicate that there are no results (i.e., an  empty list of results). This way, no explicit check for null is needed by  clients of the method.

On the otherhand, using null to indicate "there is  no answer to this question", then it is probably appropriate. For  example, File.listFiles() returns an empty list if given a directory containing no files,  and returns null if the file is not a directory.

RC: Suspicious reference  comparison

This method compares two reference values using the == or  != operator, where the correct way to compare instances of this type is  generally with the equals() method. Examples of classes which should  generally not be compared by reference are java.lang.Integer,  java.lang.Float, etc.

RCN:  Redundant comparision to null of previously checked value

This method contains a redundant comparison of a  reference value to null. Two types of redundant comparison are reported:

  • Both       values compared are definitely null
  • One       value is definitely null and the other is definitely not null

This particular warning generally indicates that a value  known not to be null was checked against null. While the check is not  necessary, it may simply be a case of defensive programming.

RCN:  Redundant comparison to null

This method contains a redundant comparison of a  reference value to null. Two types of redundant comparison are reported:

  • Both       values compared are definitely null
  • One       value is definitely null and the other is definitely not null

This particular warning represents two specific kinds of  redundant comparisions:

  1. A       value was dereferenced, and later explicitly compared to null: this       often indicates an error in the method
  2. A       literal null value was explicitly compared to null: this may indicate       that the method was modified without complete understanding of the       invariants of the method

RR: Method ignores results of  InputStream.read()

This method ignores the return value of one of the  variants of java.io.InputStream.read() which can return multiple bytes.  If the return value is not  checked, the caller will not be able to correctly handle the case where fewer  bytes were read than the caller requested.  This is a particularly  insidious kind of bug, because in many programs, reads from input streams  usually do read the full amount of data requested, causing the program to  fail only sporadically.

RR: Method ignores results of  InputStream.skip()

This method ignores the return value of java.io.InputStream.skip() which can skip multiple bytes.  If the return value is not  checked, the caller will not be able to correctly handle the case where fewer  bytes were skipped than the caller requested.  This is a particularly  insidious kind of bug, because in many programs, skips from input streams  usually do skip the full amount of data requested, causing the program to  fail only sporadically. With Buffered streams, however, skip() will only skip  data in the buffer, and will routinely fail to skip the requested number of  bytes.

RV: Method ignores  return value

The return value of this method should be checked.

SA: Self assignment  of field

This method contains a self assignment of a field; e.g.

 
  int x;
  public void foo() {
    x = x;
  }

Such assignments are useless, and may  indicate a logic error or typo.

SI: Static  initializer for class creates instance before all static final fields  assigned

The class's static initializer creates an instance of the  class before all of the static final fields are assigned.

SIO: Unnecessary  type check done using instanceof operator

Type check performed using the instanceof operator where  it can be statically determined whether the object is of the type requested.

SW:  Certain swing methods should only be invoked from the Swing event thread

(From JDC  Tech Tip): The Swing methods show(), setVisible(), and pack() will create  the associated peer for the frame. With the creation of the peer, the system  creates the event dispatch thread. This makes things problematic because the  event dispatch thread could be notifying listeners while pack and validate  are still processing. This situation could result in two threads going  through the Swing component-based GUI -- it's a serious flaw that could  result in deadlocks or other related threading issues. A pack call causes  components to be realized. As they are being realized (that is, not  necessarily visible), they could trigger listener notification on the event  dispatch thread.

Se: Non-transient  non-serializable instance field in serializable class

This Serializable class defines a non-primitive instance  field which is neither transient, Serializable, or java.lang.Object, and does not appear to implement the Externalizable interface or the readObject() and writeObject() methods.  Objects of this class will not be deserialized  correctly if a non-Serializable object is stored in this field.

Se:  serialVersionUID isn't final

This class defines a serialVersionUID  field that is not final.  The field should be made final if it is  intended to specify the version UID for purposes of serialization.

Se:  serialVersionUID isn't long

This class defines a serialVersionUID  field that is not long.  The field should be made long if it is intended  to specify the version UID for purposes of serialization.

Se:  serialVersionUID isn't static

This class defines a serialVersionUID  field that is not static.  The field should be made static if it is  intended to specify the version UID for purposes of serialization.

Se: Class is  Serializable but its superclass doesn't define a void constructor

This class implements the Serializable  interface and its superclass does not. When such an object is deserialized,  the fields of the superclass need to be initialized by invoking the void  constructor of the superclass. Since the superclass does not have one,  serialization and deserialization will fail at runtime.

Se:  Class is Externalizable but doesn't define a void constructor

This class implements the Externalizable  interface, but does not define a void constructor. When Externalizable  objects are deserialized, they first need to be constructed by invoking the  void constructor. Since this class does not have one, serialization and  deserialization will fail at runtime.

SnVI: Class is  Serializable, but doesn't define serialVersionUID

This class implements the Serializable  interface, but does not define a serialVersionUID  field.  A change as simple as adding a reference to a .class object will  add synthetic fields to the class, which will unfortunately change the  implicit serialVersionUID (e.g., adding a reference to String.class will generate a static field class$java$lang$String). Also, different source code to bytecode compilers may use  different naming conventions for synthetic variables generated for references  to class objects or inner classes. To ensure interoperability of Serializable  across versions, consider adding an explicit serialVersionUID.

UCF: Useless control  flow in method

This method contains a useless control flow  statement.  Often, this is caused by inadvertently using an empty  statement as the body of an if statement, e.g.:

 
    if (argv.length == 1);
        System.out.println("Hello, " + argv[0]);

UI: Usage of  GetResource may be unsafe if class is extended

Calling this.getClass().getResource(...) could  give results other than expected if this class is extended by a class in  another package.

UR: Uninitialized read of  field in constructor

This constructor reads a field which has not yet been  assigned a value.  This is often caused when the programmer mistakenly  uses the field instead of one of the constructor's parameters.

UwF: Unwritten field

This field is never written.  All reads of it will  return the default value. Check for errors (should it have been  initialized?), or remove it if it is useless.

EI: Method may expose internal  representation by returning reference to mutable object

Returning a reference to a mutable object value stored in  one of the object's fields exposes the internal representation of the  object.  If instances are accessed by untrusted code, and unchecked  changes to the mutable object would compromise security or other important  properties, you will need to do something different. Returning a new copy of  the object is better approach in many situations.

FI: Finalizer  should be protected, not public

A class's finalize() method should have  protected access, not public.

MS: Method may expose  internal static state by storing a mutable object into a static field

This code stores a reference to an externally mutable  object into a static field. If unchecked changes to the mutable object would  compromise security or other important properties, you will need to do  something different. Storing a copy of the object is better approach in many  situations.

MS: Field isn't final and  can't be protected from malicious code

A mutable static field could be changed by malicious code  or by accident from another package. Unfortunately, the way the field is used  doesn't allow any easy fix to this problem.

MS: Public static method may  expose internal representation by returning array

A public static method returns a reference to an array  that is part of the static state of the class. Any code that calls this  method can freely modify the underlying array. One fix is to return a copy of  the array.

MS: Field should be both  final and package protected

A mutable static field could be changed by malicious code  or by accident from another package. The field could be made package  protected and/or made final to avoid this vulnerability.

MS: Field is a mutable array

A final static field references an array and can be  accessed by malicious code or by accident from another package. This code can  freely modify the contents of the array.

MS: Field is a mutable  Hashtable

A final static field references a Hashtable and can be  accessed by malicious code or by accident from another package. This code can  freely modify the contents of the Hashtable.

MS: Field should be moved  out of an interface and made package protected

A final static field that is defined in an interface  references a mutable object such as an array or hashtable. This mutable  object could be changed by malicious code or by accident from another  package. To solve this, the field needs to be moved to a class and made  package protected to avoid this vulnerability.

MS: Field should be package  protected

A mutable static field could be changed by malicious code  or by accident. The field could be made package protected to avoid this  vulnerability.

MS: Field isn't final but  should be

A mutable static field could be changed by malicious code  or by accident from another package. The field could be made final to avoid  this vulnerability.

2LW: Wait with two locks  held

Waiting on a monitor while two locks are held may cause  deadlock.   Performing a wait only releases the lock on the object being  waited on, not any other locks.   This not necessarily a bug, but is  worth examining closely.

DC: Possible double check of  field

This method may contain an instance of double-checked  locking.  This idiom is not correct according to the semantics of the  Java memory model.  For more information, see the web page http://www.cs.umd.edu/~pugh/java/memoryModel/DoubleCheckedLocking.html.

IS2: Inconsistent  synchronization

The fields of this class appear to be accessed  inconsistently with respect to synchronization.  This bug report  indicates that the bug pattern detector judged that

  1. The       class contains a mix of locked and unlocked accesses,
  2. At       least one locked access was performed by one of the class's own methods,       and
  3. The       number of unsynchronized field accesses (reads and writes) was no more       than one third of all accesses, with writes being weighed twice as high       as reads

A typical bug matching this bug pattern is forgetting to  synchronize one of the methods in a class that is intended to be thread-safe.

You can select the nodes labeled "Unsynchronized  access" to show the code locations where the detector believed that a  field was accessed without synchronization.

Note that there are various sources of inaccuracy in this  detector; for example, the detector cannot statically detect all situations  in which a lock is held.  Also, even when the detector is accurate in  distinguishing locked vs. unlocked accesses, the code in question may still  be correct.

This description refers to the "IS2" version of  the pattern detector, which has more accurate ways of detecting locked vs.  unlocked accesses than the older "IS" detector.

LI: Incorrect lazy  initialization of static field

This method contains an unsynchronized lazy  initialization of a non-volatile static field. Because the compiler or  processor may reorder instructions, threads are not guaranteed to see a  completely initialized object, if the method  can be called by multiple threads. You can make the field  volatile to correct the problem. For more information, see the Java Memory Model web  site.

ML: Method  synchronizes on an updated field

This method synchronizes on an object references from a  mutable field. This is unlikely to have useful semantics, since different  threads may be synchronizing on different objects.

NN: Naked notify in method

A call to notify() or notifyAll() was made without any (apparent) accompanying modification to  mutable object state.  In general, calling a notify method on a monitor  is done because some condition another thread is waiting for has become  true.  However, for the condition to be meaningful, it must involve a  heap object that is visible to both threads.

This bug does not necessarily indicate an error, since  the change to mutable object state may have taken place in a method which  then called the method containing the notification.

No: Using notify()  rather than notifyAll() in method

This method calls notify() rather than notifyAll().  Java monitors are often used for multiple  conditions.  Calling notify() only wakes up one thread,  meaning that the thread woken up might not be the one waiting for the  condition that the caller just satisfied.

RS: Class's readObject()  method is synchronized

This serializable class defines a readObject() which is synchronized.  By definition, an object created by  deserialization is only reachable by one thread, and thus there is no need  for readObject() to be synchronized.  If the readObject() method  itself is causing the object to become visible to another thread, that is an  example of very dubious coding style.

Ru: Invokes run on a thread (did  you mean to start it instead?)

This method explicitly invokes run() on  an object.  In general, classes implement the Runnable  interface because they are going to have their run() method invoked  in a new thread, in which case Thread.start() is the  right method to call.

SC: Constructor invokes  Thread.start()

The constructor starts a thread. This is likely to be  wrong if the class is ever extended/subclassed, since the thread will be  started before the subclass constructor is started.

SP: Method spins on field

This method spins in a loop which reads a field.   The compiler may legally hoist the read out of the loop, turning the code  into an infinite loop.  The class should be changed so it uses proper  synchronization (including wait and notify calls).

UG: Unsynchronized get  method, synchronized set method

This class contains similarly-named get and set methods  where the set method is synchronized and the get method is not.  This  may result in incorrect behavior at runtime, as callers of the get method  will not necessarily see a consistent state for the object.  The get  method should be made synchronized.

UW: Unconditional wait in  method

This method contains a call to java.lang.Object.wait() which is not guarded by conditional control flow.  If the  condition that the method intends to wait for has already happened, the  thread could wait indefinitely.

VO: A volatile  reference to an array doesn't treat the array elements as volatile

This declares a volatile reference to an array, which  might not be what you want. With a volatile reference to an array, reads and  writes of the reference to the array are treated as volatile, but the array  elements are non-volatile. To get volatile array elements, you will need to  use one of the atomic array classes in java.util.concurrent (provided in Java  5.0).

WS: Class's writeObject()  method is synchronized but nothing else is

This class has a writeObject() method  which is synchronized; however, no other method of the class is synchronized.

Wa: Wait not in loop in method

This method contains a call to java.lang.Object.wait() which is not in a loop.  If the monitor is used for multiple  conditions, the condition the caller intended to wait for might not be the  one that actually occurred.

Dm: Method invokes dubious  Boolean constructor; use Boolean.valueOf(...) instead

Creating new instances of java.lang.Boolean  wastes memory, since Boolean objects are immutable and there are only two useful values of  this type.  Use the Boolean.valueOf() method to create Boolean objects instead.

Dm: Explicit garbage collection;  extremely dubious except in benchmarking code

Code explicitly invokes garbage collection. Except for  specific use in benchmarking, this is very dubious.

In the past, situations where people have explicitly  invoked the garbage collector in routines such as close or finalize methods  has led to huge performance black holes. Garbage collection can be expensive.  Any situation that forces hundreds or thousands of garbage collections will  bring the machine to a crawl.

Dm: Method invokes dubious new  String(String) constructor; just use the argument

Using the java.lang.String(String) constructor  wastes memory because the object so constructed will be functionally indistinguishable  from the String passed as a parameter.  Just use the argument String  directly.

Dm: Method invokes  dubious String.equals(""); use String.length() == 0 instead

An object is compared to the empty String object using  the equals() method here. Checking that the String object's length is zero  may be faster, and removes String constants from the class file.

Dm: Method invokes  toString() method on a String; just use the String

Calling String.toString() is just a redundant  operation. Just use the String.

Dm: Method invokes dubious  new String() constructor; just use ""

Creating a new java.lang.String  object using the no-argument constructor wastes memory because the object so  created will be functionally indistinguishable from the empty string constant  "".  Java guarantees that identical string constants will be  represented by the same String object.  Therefore, you should just use the empty string  constant directly.

FI: Empty finalizer should be deleted

Empty finalize() methods are useless, so  they should be deleted.

FI: Finalizer does nothing but call  superclass finalizer

The only thing this finalize() method  does is call the superclass's finalize() method, making it  redundant.  Delete it.

ITA: Method uses  toArray() with zero-length array argument

This method uses the toArray() method of a collection  derived class, and passes in a zero-length prototype array argument. It is  more efficient to use

myCollection.toArray(new Foo[myCollection.size()])

If the array passed in is big enough to  store all of the elements of the collection, then it is populated and  returned directly. This avoids the need to create a second array (by  reflection) to return as the result.

SBSC:  Method concatenates strings using + in a loop

The method seems to be building a String using  concatenation in a loop. In each iteration, the String is converted to a  StringBuffer/StringBuilder, appended to, and converted back to a String. This  can lead to a cost quadractic in the number of iterations, as the growing  string is recopied in each iteration.

Better performance can be obtained by using a  StringBuffer (or StringBuilder in Java 1.5) explicitly.

For example:

 
  // This is bad
  String s = "";
  for (int i = 0; i < field.length; ++i) {
    s = s + field[i];
  }
 
  // This is better
  StringBuffer buf = new StringBuffer();
  for (int i = 0; i < field.length; ++i) {
    buf.append(field[i]);
  }
  String s = buf.toString();

SIC: Should be a  static inner class

This class is an inner class, but does not use its  embedded reference to the object which created it.  This reference makes  the instances of the class larger, and may keep the reference to the creator  object alive longer than necessary.  If possible, the class should be  made into a static inner class.

SIC: Could be  refactored into a named static inner class

This class is an inner class, but does not use its  embedded reference to the object which created it.  This reference makes  the instances of the class larger, and may keep the reference to the creator  object alive longer than necessary.  If possible, the class should be  made into a static inner class.  Since anonymous inner classes cannot be marked as static, doing this will  requiring refactoring the inner class so that it is a named inner class.

SIC:  Could be refactored into a static inner class

This class is an inner class, but does not use its  embedded reference to the object which created it except during construction  of the inner object.  This reference makes the instances of the class  larger, and may keep the reference to the creator object alive longer than  necessary.  If possible, the class should be made into a static inner class. Since the reference to  the outer object is required during construction of the inner instance, the  inner class will need to be refactored so as to pass a reference to the outer  instance to the constructor for the inner class.

SS: Unread field: should  this field be static?

This class contains an instance final field that is  initialized to a compile-time static value. Consider making the field static.

UPM: Private  method is never called

This private method is never called. Although it is  possible that the method will be invoked through reflection, it is more  likely that the method is never used, and should be removed.

UrF: Unread field

This field is never read.  Consider removing it from  the class.

UuF: Unused field

This field is never used.  Consider removing it from  the class.

 
 

 

 

你可能感兴趣的:(Bug Descriptions)