Note: Object Lifetime RAII not using/with block scope RAII
It seems like its possible using an extra gc category, short lived objects(check gc category somewhat frequently), long lived objects(check gc category less frequently), and resource objects(check gc category very frequently). Or possibly with an extra reference counting gc for resource objects.
It seems like the using/with style can have some benefits by promoting a more functional style(forgive me if I'm wrong and this is not the functional style) of I/O discouraging lots of I/O spread out over the place vs the flexibility of object based RAII (because it's easier). But some problems probably require hard to track resource lifetimes.
Are there reasons besides avoiding gc complexity and speed, that this has not been done on mainstream languages?(I understand that some languages use reference counting as part of gc in their main implementations and as such, RAII may work there, but as I believe their spec doesn't specify reference counting for some type of objects/or all objects and that other implementations used by people don't have reference counting, limiting use of object lifetime RAII in those languages.
P.S.:Do they have c++ type RAII in perl?
Many languages make it a lot easier to write a custom inner block processor† than it was traditionally in C++ (this may have been addressed in the current drafts of the latest standard). When you have these, much of the requirement of the use of RAII for exact resource handling becomes much less pressing; you can do something like this:
using (Transaction t = makeTX()) {
// blah
}
instead of:
{
Transaction t = makeTX();
// blah
}
There's not a huge difference really except that when you have multiple nested using constructs it's much clearer what the order of resource release is. (It's also IMO easier to do special handling in the case where an exception is thrown, useful for things like transactions where you'd want to roll back on error, but I don't expect everyone to agree with me there.) Also note that there are many different ways of writing using constructs, some much more heavyweight than others, but we don't really need to explore the differences here.
Given that exact resource handling is dealt with in this different way, there's a lot less demand for the C++ RAII style and it is viable to use Garbage Collection (GC) instead, as that handles complex cases (i.e., anywhere where it is difficult to tie object lifetime to a specific scope) far more easily. To be fair, there are cases when you need exact resource management with a non-trivial lifetime, but those cases are nasty for everyone.
Perl uses Garbage Collection and has cheap subroutine blocks, as do most other scripting languages in one form or another (because the division between code and data is looser in scripting languages than more traditional compiled langs). The only big scripting language that I'm aware of that doesn't use GC is Tcl, and that's because the value system there is guaranteed loop-free for technical semantic reasons and so reference counting is sufficient. Code blocks are still very cheap there though.
If we look at mainstream compiled languages (i.e., not scripting languages) then we really see a divide in about 1990. Languages from before then (including C++) tend to not assume garbage collection (with some exceptions such as Lisp, Smalltalk and the functional programming languages) whereas languages from after that point (notably Java and C#) do assume GC. I guess there was a substantial philosophical shift about that point, probably coupled to some clever implementations that dealt with the most egregious problems in GC before that point. When you have GC, you simply don't think of RAII as a solution; it's very rooted in C++'s model of the world.
† I just made that term up.
Related
I came across this term when exploring Rust.
I saw different kinds of explanations regarding this and still don't quite get the ideas.
In The Embedded Rust Book, it said
Type states are also an excellent example of Zero Cost Abstractions
the ability to move certain behaviors to compile time execution or analysis.
These type states contain no actual data, and are instead used as
markers.
Since they contain no data, they have no actual representation in
memory at runtime:
Does it mean the runtime is faster because there is no memory in runtime?
Appreciate it if anyone can explain it in an easy to understand way.
Zero Cost Abstractions means adding higher-level programming concepts, like generics, collections and so on do not come with a run-time cost, only compile time cost (the code will be slower to compile). Any operation on zero-cost abstractions is as fast as you would write out matching functionality by hand using lower-level programming concepts like for loops, counters, ifs and using raw pointers.
Or another way to view this is that using zero-cost abstraction tools, functions, templates, classes and such come with "zero cost" for the performance of your code.
Zero cost abstractions are ones that bear no runtime costs in execution speed or memory usage.
By contrast, virtual methods are a good example of a costly abstraction: in many OO languages the type of the method's caller is determined at runtime which requires maintaining a lookup table (runtime memory usage) and then actually performing the lookup (runtime overhead per method call, likely at least an extra pointer dereference) with the runtime type to determine which version of the method to call. Another good example would be garbage collection: in return for being able to not worry about the details of memory allocation you pay with GC pauses.
Rust though mostly tries to have zero cost abstractions: ones that let you have your cake and eat it too. Ones that compiler can safely and correctly convert to forms that bear no extra indirection/memory usage. In fact, the only thing that I'm aware of (somebody more knowledgeable correct me if I'm wrong) that you really pay for at runtime in Rust is bounds checking.
The concept of zero cost abstractions originally came from the functional world. However, the terminology comes from C++. According to Bjarne Stroustrup,
In general, C++ implementations obey the
zero-overhead principle: What you don’t use, you don’t pay for. And further: What you do use, you couldn’t hand code any better.
This quote along with most answers fail to deliver the idea in its entirety because the context in which these were said isn't explicitly stated.
If there was only one programming language in the world: be it Rust or C++, zero-cost abstractions would be indistinguishable from most other compiler optimizations. The implication here is that there are countless other languages that let you do the same things as Rust or C++, but with a nonzero and often runtime-specific cost.
The topic of multi-threaded access to Lisp objects came up in another post at https://stackoverflow.com/posts/comments/97440894?noredirect=1, but as a side issue, and I am hoping for further clarification.
In general, Lisp functions (and special forms, macros, etc) seem to naturally divide into accessors and modifiers of objects. Modifiers of shared objects are clearly problematic in multi-threaded applications, since updates occurring at the same time can interfere with each other (requiring protective locks, atomic operations, etc).
But the question of potential accessor interference seems less clear. Of course, any accessor could be written to include latent modifying code, but I would like to think that the basic Lisp accessor operations (as specified in CLHS and implemented for the various platforms) do not. However, I suspect there could be a very few exceptions for reasons of efficiency—exceptions that would be good to be aware of if otherwise used in multi-threaded code without protection. (The kind of exceptions I’m talking about are not operations like maphash which can be used as both an accessor and modifier.)
It would be helpful if anyone with implementation experience could point to at least one built-in access-only operation (say in SBCL or other source) that includes potentially troublesome modification. I know guarantees are hard to come by, but heuristic guidance is useful too.
Any code that does that would be a bug in an implementation that supports multithreading. SBCL protects functions that are not thread-safe with the famous *world-lock*.
If you have a real reason to want an immutable structure, use defconstant with a read-only defstruct.
(defstruct number (value :read-only t))
(defconstant +five+ (make-number 5))
If you were designing a programming language that features automatic memory management, would using reference counting allow for determinism guarantees that are not possible with a garbage collector?
Would there be a different answer to this question for functional vs. imperative languages?
Would using reference counting allow for determinism guarantees that are not possible with a garbage collector?
The word guarantee is a strong one. Here are the guarantees you can provide with reference counting:
Constant time overhead at an assignment to adjust reference counts.
Constant time to free an object whose reference count goes to zero. (The key is that you must not decrement that object's children right away; instead you must do it lazily when the object is used to satisfy a future allocation request.)
Constant time to allocate a new object when the relevant free list is not empty. This guarantee is conditional and isn't worth much.
Here are some things you can't guarantee with reference counting:
Constant time to allocate a new object. (In the worst case, the heap may be growing, and depending on the system the delay to organize new memory may be considerable. Or even worse, you may fill the heap and be unable to allocate.)
All unreachable objects are reclaimed and reused while maintaining constant time for other operations. (A standard reference counter can't collect cyclic garbage. There are a variety of ingenious workarounds, but generally they invalidate constant-time guarantees for simple operations.)
There are now some real-time garbage collectors that provide pretty interesting guarantees about pause times, and in the last 5 years there have been pretty interesting developments in both reference counting and garbage collection. From where I sit as an informed outsider, there's no obvious winner.
Some of the best recent work on reference counting is by David Bacon of IBM and by Erez Petrank of Technion. If you want to learn what a sophisticated, modern reference-counting system can do, look up their papers. Among other things, they are using multiple processors in amazing ways.
For information about memory management and real-time guarantees more generally, check out the International Symposium on Memory Management.
Would there be a different answer to this question for functional vs. imperative languages?
Because you asked about guarantees, no. But for memory management in general, the performance tradeoffs are quite different for an imperative language (lots of mutation but low allocation rates), an impure functional language (hardly any mutation but high allocation rates), and a pure, lazy functional language (lots of mutation—all those thinks being updated—and high allocation rates).
would using reference counting allow for determinism guarantees that are not possible with a garbage collector?
I don't see how. The process of lowering the reference count of an object is not time-bounded, as that object may be the single root for an arbitrary large object graph.
The only way to approach the problem of GC for real-time systems is by using either a concurrent collector or an incremental one - and no matter if the system uses reference counting or not; in my opinion your distinction between reference counting and "collection" is not precise anyway, e.g. systems which utilize reference counting might still occasionally perform some memory sweep (for example, to handle cycles).
You might be interested in IBM's Metronome, and I also know Microsoft has done some research in direction of good, real-time memory management.
If you look at the RTSJ spec (JSR-1), you'll see they did an end-run around the problem by providing for no-heap realtime threads. By having a separate category of thread that isn't allowed to touch any object that might require the thread to be stopped for garbage collection, JSR-1 side stepped the issue. There aren't many RTSJ implementations right now, but the area of realtime garbage collection is a hot topic in that community.
For real time programming, does reference counting have an advantage over garbage collection in terms of determinism?
Yes. The main advantage of reference counting is simplicity.
If you were designing a programming language that features automatic memory management, would using reference counting allow for determinism guarantees that are not possible with a garbage collector?
A GC like Baker's Treadmill should attain the same level of guarantees regarding determinism that reference counting offers.
Would there be a different answer to this question for functional vs. imperative languages?
Yes. Reference counting alone does not handle cycles. Some functional languages make it impossible to create cycles by design (e.g. Erlang and Mathematica) so they trivially permit reference counting alone as an exact approach to GC.
In real time programming garbage collection could be harmful, because you don't know when the garbage collector will collect... so yes, reference counting is definitely better in this context.
As a side note, usually in real time system only some parts needs real time processing, so you could avoid garbage collection just in sensitive components. A real world example is a C# program running on a Windows CE target.
From some involvement in various projects migrating significant chunks of code from C++ (with various smart pointer classes, including reference counted) to garbage collected Java/C#, I observe that the biggest pain-points all seem to be related to classes with non-empty destructors (particularly when used for RAII). This is a pretty big flag that deterministic cleanup is expected.
The issue is surely much the same for any language with objects; I don't think hybrid OO-functional languages like Scala or Ocaml enjoy any particular advantages in this area. Situation might be different for more "pure" functional languages.
In the future, will managed runtimes provide additional protections against subtle data corruption issues?
Managed runtimes such as Java and the .NET CLR reduce or eliminate the possibility of many memory corruption bugs common in native languages like C#. Nonetheless, they are surprisingly not immune from all memory corruption problems. One intuitively expects that a method that validates its input, has no bugs, and robustly handles exceptions will always transform its object from one valid state to another, but this is not the case. (It is more accurate to say that it is not the case using prevailing programming conventions--object implementors need to go out of their way to avoid the problems I describe.)
Consider the following scenarios:
Threading. The caller might share the object with other threads and make concurrent calls on it. If the object does not implement locking, the fields might be corrupted. (Perhaps--unless notified that the object is thread-safe--runtimes should use an interlock on every method call to throw an exception if any method on the same object executing concurrently on another thread. This would be a protection feature and, just like other well-accepted safety features of managed runtimes, it has some cost.)
Re-entrancy. The method makes a callout to an arbitrary function (such as an event handler) that ultimately calls methods on the object that are not designed to be called at that point. This is even trickier than thread safety and many class libraries do not get this right. (Worse yet, class libraries are known to poorly document what re-entrancy is allowed.)
For all of these cases, it can be argued that thorough documentation is a solution. However, documentation also can prescribe how to allocate and deallocate memory in unmanaged languages. We know from experience (e.g., with memory allocation) that the difference between documentation and language/runtime enforcement is night and day.
What can we expect from languages and runtimes in the future to protect us from these problems and other subtle problems like them?
I think languages and runtimes will keep moving forward, keep abstracting away issues from the developer, and keep making our lives easier and more productive.
Take your example - threading. There are some great new features on the horizon in the .NET world to simplify the threading model we use daily. STM.NET may eventually make shared state much, much safer to handle, for example. The parallel extensions in .NET 4 make life very easy for threading compared to current technologies.
I think that transactional memory is promising for addressing some of these issues. I'm not sure if this answers your question in some way but this is an interesting topic in any event:
http://en.wikipedia.org/wiki/Software_transactional_memory
There was an episode of Software Engineering Radio on the topic a year or so ago maybe.
First of all, "managed" is a bit of a misnomer: languages like OCaml, Haskell, and SML achieve such protections and safety while being fully compiled. All relevant "management" occurs at compile time through static analysis, which aids optimization and speed.
Anyway, to answer your question: if you look at languages like Erlang and Haskell, state is isolated and immutable by default. With kind of system, threading and reentrancy is safe by default, and because you have to go out of your way to break these rules, it is obvious to see where unsafe code can arise.
By starting with safe defaults but leaving room for advanced unsafe usage, you get the best of both worlds. It seems reasonable that future systems that are safe by your definition may follow some of these practices as well.
What can we expect in the future?
Nothing. Thread-state and re-entrancy are not problems I see tools/runtimes solving. Instead I think in the future people will move to styles that avoid programming with mutable state to bypass these issues. Languages and libraries can help make these styles of programming more attractive, but the tools are not the solution - changing the way we write code is the solution.
I'm designing a language. First, I want to decide what code to generate. The language will have lexical closures and prototype based inheritance similar to javascript. But I'm not a fan of gc and try to avoid as much as possible. So the question: Is there an elegant way to implement closures without resorting to allocate the stack frame on the heap and leave it to garbage collector?
My first thoughts:
Use reference counting and garbage collect the cycles (I don't really like this)
Use spaghetti stack (looks very inefficient)
Limit forming of closures to some contexts such a way that, I can get away with a return address stack and a locals' stack.
I won't use a high level language or follow any call conventions, so I can smash the stack as much as I like.
(Edit: I know reference counting is a form of garbage collection but I am using gc in its more common meaning)
This would be a better question if you can explain what you're trying to avoid by not using GC. As I'm sure you're aware, most languages that provide lexical closures allocate them on the heap and allow them to retain references to variable bindings in the activation record that created them.
The only alternative to that approach that I'm aware of is what gcc uses for nested functions: create a trampoline for the function and allocate it on the stack. But as the gcc manual says:
If you try to call the nested function through its address after the containing function has exited, all hell will break loose. If you try to call it after a containing scope level has exited, and if it refers to some of the variables that are no longer in scope, you may be lucky, but it's not wise to take the risk. If, however, the nested function does not refer to anything that has gone out of scope, you should be safe.
Short version is, you have three main choices:
allocate closures on the stack, and don't allow their use after their containing function exits.
allocate closures on the heap, and use garbage collection of some kind.
do original research, maybe starting from the region stuff that ML, Cyclone, etc. have.
This thread might help, although some of the answers here reflect answers there already.
One poster makes a good point:
It seems that you want garbage collection for closures
"in the absence of true garbage collection". Note that
closures can be used to implement cons cells. So your question
seem to be about garbage collection "in the absence of true
garbage collection" -- there is rich related literature.
Restricting problem to closures does not really change it.
So the answer is: no, there is no elegant way to have closures and no real GC.
The best you can do is some hacking to restrict your closures to a particular type of closure. All this is needless if you have a proper GC.
So, my question reflects some of the other ones here - why do you not want to implement GC? A simple mark+sweep or stop+copy takes about 2-300 lines of (Scheme) code, and isn't really that bad in terms of programming effort. In terms of making your programs slower:
You can implement a more complex GC which has better performance.
Just think of all the memory leaks programs in your language won't suffer from.
Coding with a GC available is a blessing. (Think C#, Java, Python, Perl, etc... vs. C++ or C).
I understand that I'm very late, but I stumbled upon this question by accident.
I believe that full support of closures indeed requires GC, but in some special cases stack allocation is safe. Determining these special cases requires some escape analysis. I suggest that you take a look at the BitC language papers, such as Closure Implementation in BitC. (Although I doubt whether the papers reflect the current plans.) The designers of BitC had the same problem you do. They decided to implement a special non-collecting mode for the compiler, which denies all closures that might escape. If turned on, it will restrict the language significantly. However, the feature is not implemented yet.
I'd advise you to use a collector - it's the most elegant way. You should also consider that a well-built garbage collector allocates memory faster than malloc does. The BitC folks really do value performance and they still think that GC is fine even for the most parts of their operating system, Coyotos. You can migitate the downsides by simple means:
create only a minimal amount of garbage
let the programmer control the collector
optimize stack/heap use by escape analysis
use an incremental or concurrent collector
if somehow possible, divide the heap like Erlang does
Many fear garbage collectors because of their experiences with Java. Java has a fantastic collector, but applications written in Java have performance problems because of the sheer amount of garbage generated. In addition, a bloated runtime and fancy JIT compilation is not really a good idea for desktop applications because of the longer startup and response times.
The C++ 0x spec defines lambdas without garbage collection. In short, the spec allows non-deterministic behavior in cases where the lambda closure contains references which are no longer valid. For example (pseudo-syntax):
(int)=>int create_lambda(int a)
{
return { (int x) => x + a }
}
create_lambda(5)(4) // undefined result
The lambda in this example refers to a variable (a) which is allocated on the stack. However, that stack frame has been popped and is not necessarily available once the function returns. In this case, it would probably work and return 9 as a result (assuming sane compiler semantics), but there is no way to guarantee it.
If you are avoiding garbage collection, then I'm assuming that you also allow explicit heap vs. stack allocation and (probably) pointers. If that is the case, then you can do like C++ and just assume that developers using your language will be smart enough to spot the problem cases with lambdas and copy to the heap explicitly (just like you would if you were returning a value synthesized within a function).
Use reference counting and garbage collect the cycles (I don't really like this)
It's possible to design your language so there are no cycles: if you can only make new objects and not mutate old ones, and if making an object can't make a cycle, then cycles never appear. Erlang works essentially this way, though in practice it does use GC.
If you have the machinery for a precise copying GC, you could allocate on the stack initially and copy to the heap and update pointers if you discover at exit that a pointer to this stack frame has escaped. That way you only pay if you actually do capture a closure that includes this stack frame. Whether this helps or hurts depends on how often you use closures and how much they capture.
You might also look into C++0x's approach (N1968), though as one might expect from C++ it consists of counting on the programmer to specify what gets copied and what gets referenced, and if you get it wrong you just get invalid accesses.
Or just don't do GC at all. There can be situations where it's better to just forget the memory leak and let the process clean up after it when it's done.
Depending on your qualms about GC, you might be afraid of the periodic GC sweeps. In this case you could do a selective GC when an item falls out of scope or the pointer changes. I'm not sure how expensive this would be though.
#Allen
What good is a closure if you can't use them when the containing function exits? From what I understand that's the whole point of closures.
You could work with the assumption that all closures will be called eventually and exactly one time. Now, when the closure is called you can do the cleanup at the closure return.
How do you plan on dealing with returning objects? They have to be cleaned up at some point, which is the exact same problem with closures.
So the question: Is there an elegant way to implement closures without resorting to allocate the stack frame on the heap and leave it to garbage collector?
GC is the only solution for the general case.
Better late than never?
You might find this interesting: Differential Execution.
It's a little-known control stucture, and its primary use is in programming user interfaces, including ones that can change dynamically while in use. It is a significant alternative to the Model-View-Controller paradigm.
I mention it because one might think that such code would rely heavily on closures and garbage-collection, but a side effect of the control structure is that it eliminates both of those, at least in the UI code.
Create multiple stacks?
I've read that the last versions of ML use GC only sparingly
I guess if the process is very short, which means it cannot use much memory, then GC is unnecessary. The situation is analogous to worrying about stack overflow. Don't nest too deeply, and you cannot overflow; don't run too long, and you cannot need the GC. Cleaning up becomes a matter of simply reclaiming the large region that you pre-allocated. Even a longer process can be divided into smaller processes that have their own heaps pre-allocated. This would work well with event handlers, for example. It does not work well, if you are writing compiler; in that case, a GC is surely not much of a handicap.