Yet, at the old days when compiler output was directly consumed by hardware CPU and there was no chance to optimize something later, everything had to be done during the compilation. At that time various C++ compilers competed among themselves to produce the fastest code, the most optimized one. The competition had to be quite hard, as often they tried to optimize too much and sometimes even ''overoptimized''. I remember that from time to time I was getting some mysterious error in my program that vanished away as soon as (usually after many hours of debugging) I realized what can be the cause and I disabled some optimization switches.

Yet, at the old days when compiler output was directly consumed by hardware CPU and there was no chance to optimize something later, everything had to be done during the compilation. At that time various C++ compilers competed among themselves to produce the fastest code, the most optimized one. The competition had to be quite hard, as often they tried to optimize too much and sometimes even ''overoptimized''. I remember that from time to time I was getting some mysterious error in my program that vanished away as soon as (usually after many hours of debugging) I realized what can be the cause and I disabled some optimization switches.

For a while I believed that problems of this kind cannot happen to [[JavaC]], however I was probably wrong. Recently I needed to prevent an object to be garbage collected from memory and wrote following code:

For a while I believed that problems of this kind cannot happen to [[JavaC]], however I was probably wrong. Recently I needed to prevent an object to be [[Garbage Collection|garbage collected]] from memory and wrote following code:

<source lang="java" snippet="compiler.surprises.intro"/>

<source lang="java" snippet="compiler.surprises.intro"/>

<source lang="java" snippet="compiler.surprises.error"/>

<source lang="java" snippet="compiler.surprises.error"/>

So far, so good. This code behaves exactly as expected. It leads to conclusion that if you have a variable defined in a method body, and it has a reference to your object, the object cannot be garbage collected, until the method execution ends. OK, now guess: will the following test succeed or fail?

So far, so good. This code behaves exactly as expected. It leads to conclusion that if you have a variable defined in a method body, and it has a reference to your object, the object cannot be [[Garbage Collection|garbage collected]], until the method execution ends. OK, now guess: will the following test succeed or fail?

<source lang="java" snippet="compiler.surprises.surprise"/>

<source lang="java" snippet="compiler.surprises.surprise"/>

To my biggest surprise the reference can really be garbage collected, even there is a local variable pointing to it! This is an example of surprising (''over'')optimization of [[JavaC]] or [[HotSpot]]. It turns out that, in spite of being declared for the whole method, the variable is not used outside of the '''if''' block and as such the [[JavaC]] allocates its space on stack only for the execution of the '''if''' branch. This is quite surprising behaviour. An easy to fix one, yet surprising one:

To my biggest surprise the reference can really be [[Garbage Collection|garbage collected]], even there is a local variable pointing to it! This is an example of surprising (''over'')optimization of [[JavaC]] or [[HotSpot]]. It turns out that, in spite of being declared for the whole method, the variable is not used outside of the '''if''' block and as such the [[JavaC]] allocates its space on stack only for the execution of the '''if''' branch. This is quite surprising behaviour. An easy to fix one, yet surprising one:

<source lang="java" snippet="compiler.surprises.fix"/>

<source lang="java" snippet="compiler.surprises.fix"/>

<source lang="java" snippet="compiler.surprises.fix.final"/>

<source lang="java" snippet="compiler.surprises.fix.final"/>

However the same code without '''final''' works as well. It is enough to initialize the variable in both branches of the '''if''' statement to prevent garbage collection of the reference held by the ''retValue'' variable:

However the same code without '''final''' works as well. It is enough to initialize the variable in both branches of the '''if''' statement to prevent [Garbage Collection]] of the reference held by the ''retValue'' variable:

<source lang="java" snippet="compiler.surprises.fix.init"/>

<source lang="java" snippet="compiler.surprises.fix.init"/>

I believe that original motivation for [[APITypes:CompilerOptimizations|compiler optimizations]] is to speed program execution without affecting its behaviour. However this often is just a distant dream as from time to time the [[APITypes:CompilerOptimizations|optimizations]] change execution semantics and as soon as that happen they start to be part of the API of our languages and their libraries!

I believe that original motivation for [[APITypes:CompilerOptimizations|compiler optimizations]] is to speed program execution without affecting its behaviour. However this often is just a distant dream as from time to time the [[APITypes:CompilerOptimizations|optimizations]] change execution semantics and as soon as that happen they start to be part of the API of our languages and their libraries!

<source lang="java" snippet="compiler.surprises.fix"/>

<source lang="java" snippet="compiler.surprises.fix"/>

The fix is easy, however the consequences of my finding are really horrible for predictability of our code. [[NetBeans]] may rely on the expected behaviour (e.g. having an uninitialized local variable is enough) quite a lot. From time to time our tests are failing and it may be due to this randomness. Usually everything is OK, but from time to time, on machines with ''too powerful'' virtual machines, too many cores, too low memory, etc. the GC can kick in while the method is running and release the reference, causing our tests to fail because of an unexpected situation.

The fix is easy, however the consequences of my finding are really horrible: This means that compiler optimization are not as invisible as they should be. People can rely on or be hurt by them. They can influence predictability of our code, they can change our code to do something else than programmer would expect. This may be flaw of the compiler or of the language designer, yet [[NetBeans]] probably rely on the expected behaviour (e.g. having an uninitialized local variable is enough to hold a reference while a method is being executed) quite a lot. We know that from time to time our tests are failing unexpectedly and unexplainably and it may be due to this randomness. Usually everything is OK, but from time to time, on machines with ''too powerful'' virtual machines, too many cores, too low memory, etc. the GC can kick in while the method is running and release the reference, causing our tests to fail because of an unexpected situation.

 

 

Liam noted at [http://weblogs.java.net/blog/jst/archive/2008/10/the_better_comp.html ljnelson's blog note] that it is enough to make the variable '''final''' and the problem goes away. True, '''final''' helps, I've just tried that:

Liam noted at [http://weblogs.java.net/blog/jst/archive/2008/10/the_better_comp.html ljnelson's blog note] that it is enough to make the variable '''final''' and the problem goes away. True, '''final''' helps, I've just tried that:

<source lang="java" snippet="compiler.surprises.fix.init"/>

<source lang="java" snippet="compiler.surprises.fix.init"/>

This very likely means that the compiler puts the variable onto the stack of the topmost block where it is guaranteed to be fully initialized. That is why we [http://openide.netbeans.org/issues/show_bug.cgi?id=150492 need a hint] to warn developers about declaration of non-fully initialized non-primitive variables, as those can be source of the memory leaks.

This very likely means that the compiler puts the variable onto the stack of the topmost block where it is guaranteed to be fully initialized. That is why we [http://openide.netbeans.org/issues/show_bug.cgi?id=150492 need a hint] to warn developers about declaration of non-fully initialized non-primitive variables, as those can be source of the memory leaks.

 

<source lang="java" snippet="compiler.surprises.fix"/>

<source lang="java" snippet="compiler.surprises.fix"/>

The fix is easy, however the consequences of my finding are really horrible. [[NetBeans]] may rely on the expected behaviour (e.g. having an uninitialized local variable is enough) quite a lot. From time to time our tests are failing and it may be due to this randomness. Usually everything is OK, but from time to time, on machines with ''too powerful'' virtual machines, too many cores, too low memory, etc. the GC can kick in while the method is running and release the reference, causing our tests to fail because of an unexpected situation.

The fix is easy, however the consequences of my finding are really horrible for predictability of our code. [[NetBeans]] may rely on the expected behaviour (e.g. having an uninitialized local variable is enough) quite a lot. From time to time our tests are failing and it may be due to this randomness. Usually everything is OK, but from time to time, on machines with ''too powerful'' virtual machines, too many cores, too low memory, etc. the GC can kick in while the method is running and release the reference, causing our tests to fail because of an unexpected situation.

Maybe we will need to perform complete audit of [[NetBeans]] sources to eliminate use of uninitialized local variables. And all of this just because [[APITypes:CompilerOptimizations|compiler optimizations]] seem to become thing that external API users can depend on. [[APITypes:CompilerOptimizations|Compiler optimizations]] seem to be part of the API of our libraries!

Maybe we will need to perform complete audit of [[NetBeans]] sources to eliminate use of uninitialized local variables. And all of this just because [[APITypes:CompilerOptimizations|compiler optimizations]] seem to become thing that external API users can depend on. [[APITypes:CompilerOptimizations|Compiler optimizations]] seem to be part of the API of our libraries!

<source lang="java" snippet="compiler.surprises.fix.final"/>

<source lang="java" snippet="compiler.surprises.fix.final"/>

However the same code without '''final''' works as well. It is enough to initialize the variable in both branches of the '''if''' statement to prevent garbage collection of the reference hold by the ''retValue'' variable:

However the same code without '''final''' works as well. It is enough to initialize the variable in both branches of the '''if''' statement to prevent garbage collection of the reference held by the ''retValue'' variable:

<source lang="java" snippet="compiler.surprises.fix.init"/>

<source lang="java" snippet="compiler.surprises.fix.init"/>

This very likely means that the compiler puts the variable into the topmost block where it is guaranteed to be fully initialized. That is why we [http://openide.netbeans.org/issues/show_bug.cgi?id=150492 need a hint] to warn developers about declaration of non-fully initialized non-primitive variables, as those can be source of the memory leaks.

This very likely means that the compiler puts the variable onto the stack of the topmost block where it is guaranteed to be fully initialized. That is why we [http://openide.netbeans.org/issues/show_bug.cgi?id=150492 need a hint] to warn developers about declaration of non-fully initialized non-primitive variables, as those can be source of the memory leaks.

Do you remember the time when we were still coding in C++ and we used real compilers, producing not just [[ByteCode]], but real machine code, executed by the target CPU? I do remember, at least a bit. Because for a certain time, while studying my University, I made my living developing an implementation of [[wikipedia::SQL|SQL database]] for [[wikipedia::Novell NetWare|Novell]].

Do you remember the time when we were still coding in C++ and we used real compilers, producing not just [[ByteCode]], but real machine code, executed by the target CPU? I do remember, at least a bit. I was developing an implementation of [[wikipedia::SQL|SQL database]] for [[wikipedia::Novell NetWare|Novell]] for a certain time while studying my University.

This used to be the time when compilers needed to perform optimizations. These days are gone now, the [[JavaC]] just emits [[ByteCode]] and only later, when it is really executed the [[HotSpot]] virtual machine perform optimizations. The [[DynamicCompilation]] makes this possible. The [[JavaC]] does not need to optimize anything, everything can be done later, with much greater knowledge about the execution environment.

This used to be the time when compilers needed to perform optimizations. These days are gone now, the [[JavaC]] just emits [[ByteCode]] and only later, when it is really executed the [[HotSpot]] virtual machine performs optimizations. The [[DynamicCompilation]] makes this possible. The [[JavaC]] does not need to optimize anything, everything can be done later, with much greater knowledge about the execution environment.

Yet, at the old days when compiler output was directly consumed by hardware CPU and there was no chance to optimize something later, everything had to be done during the compilation. At that time various C++ compilers competed among themselves to produce the fastest code, the most optimized one. The competition had to be quite hard, as often they tried to optimize too much and sometimes even ''overoptimized''. I remember that from time to time I was getting some mysterious error in my program that vanished away as soon as (usually after many hours of debugging) I realized what can be the cause and I disabled some optimization switches.

Yet, at the old days when compiler output was directly consumed by hardware CPU and there was no chance to optimize something later, everything had to be done during the compilation. At that time various C++ compilers competed among themselves to produce the fastest code, the most optimized one. The competition had to be quite hard, as often they tried to optimize too much and sometimes even ''overoptimized''. I remember that from time to time I was getting some mysterious error in my program that vanished away as soon as (usually after many hours of debugging) I realized what can be the cause and I disabled some optimization switches.

<source lang="java" snippet="compiler.surprises.error"/>

<source lang="java" snippet="compiler.surprises.error"/>

So far, so good. This code behaves exactly as expected. It leads to conclusion that if you have a variable defined in a method body, and it has a reference to your object, the object cannot be garbage collected, until the method execution ends. OK, now guess: will the following test succeed of fail?

So far, so good. This code behaves exactly as expected. It leads to conclusion that if you have a variable defined in a method body, and it has a reference to your object, the object cannot be garbage collected, until the method execution ends. OK, now guess: will the following test succeed or fail?

<source lang="java" snippet="compiler.surprises.surprise"/>

<source lang="java" snippet="compiler.surprises.surprise"/>

The fix is easy, however the consequences of my finding are really horrible. [[NetBeans]] may rely on the expected behaviour (e.g. having an uninitialized local variable is enough) quite a lot. From time to time our tests are failing and it may be due to this randomness. Usually everything is OK, but from time to time, on machines with ''too powerful'' virtual machines, too many cores, too low memory, etc. the GC can kick in while the method is running and release the reference, causing our tests to fail because of an unexpected situation.

The fix is easy, however the consequences of my finding are really horrible. [[NetBeans]] may rely on the expected behaviour (e.g. having an uninitialized local variable is enough) quite a lot. From time to time our tests are failing and it may be due to this randomness. Usually everything is OK, but from time to time, on machines with ''too powerful'' virtual machines, too many cores, too low memory, etc. the GC can kick in while the method is running and release the reference, causing our tests to fail because of an unexpected situation.

Maybe we will need to to perform complete audit of [[NetBeans]] sources to eliminate use of uninitialized local variables. And all of this just because [[APITypes:CompilerOptimizations|compiler optimizations]] seem to become thing that external API users can depend on. [[APITypes:CompilerOptimizations|Compiler optimizations]] seem to be part of the API of our libraries!

Maybe we will need to perform complete audit of [[NetBeans]] sources to eliminate use of uninitialized local variables. And all of this just because [[APITypes:CompilerOptimizations|compiler optimizations]] seem to become thing that external API users can depend on. [[APITypes:CompilerOptimizations|Compiler optimizations]] seem to be part of the API of our libraries!

== Additional Findings ==

== Additional Findings ==

<source lang="java" snippet="compiler.surprises.fix.init"/>

<source lang="java" snippet="compiler.surprises.fix.init"/>

This very likely means that the compiler puts the variable into the topmost block where it is guaranteed to be fully initialized. That is why we need a hint to warn developers about declaration of non-fully initialized non-primitive variables, as those can be source of the memory leaks.

This very likely means that the compiler puts the variable into the topmost block where it is guaranteed to be fully initialized. That is why we [http://openide.netbeans.org/issues/show_bug.cgi?id=150492 need a hint] to warn developers about declaration of non-fully initialized non-primitive variables, as those can be source of the memory leaks.

Maybe we will need to to perform complete audit of [[NetBeans]] sources to eliminate use of uninitialized local variables. And all of this just because [[APITypes:CompilerOptimizations|compiler optimizations]] seem to become thing that external API users can depend on. [[APITypes:CompilerOptimizations|Compiler optimizations]] seem to be part of the API of our libraries!

Maybe we will need to to perform complete audit of [[NetBeans]] sources to eliminate use of uninitialized local variables. And all of this just because [[APITypes:CompilerOptimizations|compiler optimizations]] seem to become thing that external API users can depend on. [[APITypes:CompilerOptimizations|Compiler optimizations]] seem to be part of the API of our libraries!

Do you remember the time when we were still coding in C++ and we used real compilers, producing not just [[ByteCode]], but real machine code, executed by the target CPU? I do remember, at least a bit. Because for a certain time, while studying my University, I made my living developing an implementation of [[wikipedia::SQL|SQL database]] for [[wikipedia::Novell|Novell NetWare]].

Do you remember the time when we were still coding in C++ and we used real compilers, producing not just [[ByteCode]], but real machine code, executed by the target CPU? I do remember, at least a bit. Because for a certain time, while studying my University, I made my living developing an implementation of [[wikipedia::SQL|SQL database]] for [[wikipedia::Novell NetWare|Novell]].

This used to be the time when compilers needed to perform optimizations. These days are gone now, the [[JavaC]] just emits [[ByteCode]] and only later, when it is really executed the [[HotSpot]] virtual machine perform optimizations. The [[DynamicCompilation]] makes this possible. The [[JavaC]] does not need to optimize anything, everything can be done later, with much greater knowledge about the execution environment.

This used to be the time when compilers needed to perform optimizations. These days are gone now, the [[JavaC]] just emits [[ByteCode]] and only later, when it is really executed the [[HotSpot]] virtual machine perform optimizations. The [[DynamicCompilation]] makes this possible. The [[JavaC]] does not need to optimize anything, everything can be done later, with much greater knowledge about the execution environment.