I have a function that I would like to provide an assembly implementation for
on amd64 architecture. For the sake of discussion let's just suppose it's an
Add function, but it's actually more complicated than this. I have the
assembly version working but my question concerns getting the godoc to display
correctly. I have a feeling this is currenty impossible, but I wanted to seek
advice.
Some more details:
The assembly implementation of this function contains only a few
instructions. In particular, the mere cost of calling the function is a
significant part of the entire cost.
It makes use of special instructions (BMI2) therefore can only be used
following a CPUID capability check.
The implementation is structured like this gist. At a high level:
In the generic (non-amd64 case) the function is defined by delegating to
addGeneric.
In the amd64 case the function is actually a variable, initially set to
addGeneric but replaced by addAsm in the init function if a cpuid
check passes.
This approach works. However the godoc output is crappy because in the
amd64 case the function is actually a variable. Note godoc appears to be
picking up the same build tags as the machine it's running on. I'm not sure
what godoc.org would do.
Alternatives considered:
The Add function delegates to addImpl. Then we pull some similar trick
to replace addImpl in the amd64 case. The problem with this is (in my
experiments) Go doesn't seem to be able to inline the call, and the assembly
is now wrapped in two function calls. Since the assembly is so small already
this has a noticable impact on performance.
In the amd64 case we define a plain function Add that has the useAsm
check inside it, and calls one of addGeneric and addAsm depending on the
result. This would have an even worse impact on performance.
So I guess the questions are:
Is there a better way to structure the code to achieve the performance I
want, and have it appear properly in documentation.
If there is no alternative, is there some other way to "trick" godoc?
See math.Sqrt for an example of how to do this.
Write a stub function with the documentation
Write a generic implementation as an unexported function.
For each architecture, write a function in assembler that jumps to the unexported generic implementation or implements the function directly.
To handle the cpuid check, set a package variable in init() and conditionally jump based on that variable in the assembly implementation.
If there is a function from an external package that is not used at all in my project, will the compiler remove the function from the generated machine code?
This question could be targeted at any language compiler in general. But, I think the behaviour may vary language to language. So, I am interested in knowing what does go compilers do.
I would appreciate any help on understanding this.
The language spec does not mention this anywhere, and from a correctness point of view this is irrelevant.
But know that the current version does remove certain constructs that the compiler can prove is not used and will not change the runtime behaviour of the app.
Quoting from The Go Blog: Smaller Go 1.7 binaries:
The second change is method pruning. Until 1.6, all methods on all used types were kept, even if some of the methods were never called. This is because they might be called through an interface, or called dynamically using the reflect package. Now the compiler discards any unexported methods that do not match an interface. Similarly the linker can discard other exported methods, those that are only accessible through reflection, if the corresponding reflection features are not used anywhere in the program. That change shrinks binaries by 5–20%.
Methods are a "harder" case than functions because methods can be listed and called with reflection (unlike functions), but the Go tools do what they can even to remove unused methods too.
You can see examples and proof of removed / unlinked code in this answer:
How to remove unused code at compile time?
Also see other relevant questions:
Splitting client/server code
Call all functions with special prefix or suffix in Golang
I am starting to work on a project and for one of the tasks I need to analyze the source code in order to gather information about the classes and their methods. More specifically, for each method I need to know which internal attributes and external objects (references) it uses throughout the entire method body.
I discussed it with my supervisors and they think that Bytecode manipulation libraries is the way to go. I already looked at BCEL, ASM and Javassist but I'm not sure which one I need to use. Do they all provide access to the method body where I can see all the instructions and get the information I need?
Any advice would be appreciate it. Thank you!
If you really “need to analyze the source code”, then libraries which allow to inspect the bytecode are not the way to go.
Otherwise, you really need to define your task precisely. Either, you are about to analyze classes, regardless of whether you will look at their source code or byte code, or you want to analyze source code and consider doing it by compiling first, followed by analyzing the compiled result. In the latter case, you have to compare the effort of both steps with alternative solution, which may, e.g. incorporate direct source code analysis.
Parsing byte code is rather easy, easier than analyzing source code, which is the reason why bytecode is produced prior to the execution of Java programs. To answer your concrete question, yes, all three libraries offer you a way to analyze the instructions and associated information. Which one is the best to fit your needs, is a question that is beyond the scope of Stackoverflow.
Whether analyzing the byte code helps, depends on your exact requirements. When it comes to field and method access, you may precisely get most of them using that approach. Only inlined compile-time constants lack their origins. When it comes to type use, you have to consider that not every source code artifact has an existing counterpart in the byte code, e.g. widening casts produce no actual code and and local variables usually don’t have a declared type (debugging information aside), but only an implied type which depends on how they are actually used. They also have no information about Generics, unless debugging information has been included.
I am learning Ruby and I'm having a major conceptual problem concerning typing. Allow me to detail why I don't understand with paradigm.
Say I am method chaining for concise code as you do in Ruby. I have to precisely know what the return type of each method call in the chain, otherwise I can't know what methods are available on the next link. Do I have to check the method documentation every time?? I'm running into this constantly running tutorial exercises. It seems I'm stuck with a process of reference, infer, run, fail, fix, repeat to get code running rather then knowing precisely what I'm working with during coding. This flies in the face of Ruby's promise of intuitiveness.
Say I am using a third party library, once again I need to know what types are allow to pass on the parameters otherwise I get a failure. I can look at the code but there may or may not be any comments or declaration of what type the method is expecting. I understand you code based on methods are available on an object, not the type. But then I have to be sure whatever I pass as a parameter has all the methods the library is expect, so I still have to do type checking. Do I have to hope and pray everything is documented properly on an interface so I know if I'm expected to give a string, a hash, a class, etc.
If I look at the source of a method I can get a list of methods being called and infer the type expected, but I have to perform analysis.
Ruby and duck typing: design by contract impossible?
The discussions in the preceding stackoverflow question don't really answer anything other than "there are processes you have to follow" and those processes don't seem to be standard, everyone has a different opinion on what process to follow, and the language has zero enforcement. Method Validation? Test-Driven Design? Documented API? Strict Method Naming Conventions? What's the standard and who dictates it? What do I follow? Would these guidelines solve this concern https://stackoverflow.com/questions/616037/ruby-coding-style-guidelines? Is there editors that help?
Conceptually I don't get the advantage either. You need to know what methods are needed for any method called, so regardless you are typing when you code anything. You just aren't informing the language or anyone else explicitly, unless you decide to document it. Then you are stuck doing all type checking at runtime instead of during coding. I've done PHP and Python programming and I don't understand it there either.
What am I missing or not understanding? Please help me understand this paradigm.
This is not a Ruby specific problem, it's the same for all dynamically typed languages.
Usually there are no guidelines for how to document this either (and most of the time not really possible). See for instance map in the ruby documentation
map { |item| block } → new_ary
map → Enumerator
What is item, block and new_ary here and how are they related? There's no way to tell unless you know the implementation or can infer it from the name of the function somehow. Specifying the type is also hard since new_ary depends on what block returns, which in turn depends on the type of item, which could be different for each element in the Array.
A lot of times you also stumble across documentation that says that an argument is of type Object, Which again tells you nothing since everything is an Object.
OCaml has a solution for this, it supports structural typing so a function that needs an object with a property foo that's a String will be inferred to be { foo : String } instead of a concrete type. But OCaml is still statically typed.
Worth noting is that this can be a problem in statically typed lanugages too. Scala has very generic methods on collections which leads to type signatures like ++[B >: A, That](that: GenTraversableOnce[B])(implicit bf: CanBuildFrom[Array[T], B, That]): That for appending two collections.
So most of the time, you will just have to learn this by heart in dynamically typed languages, and perhaps help improve the documentation of libraries you are using.
And this is why I prefer static typing ;)
Edit One thing that might make sense is to do what Scala also does. It doesn't actually show you that type signature for ++ by default, instead it shows ++[B](that: GenTraversableOnce[B]): Array[B] which is not as generic, but probably covers most of the use cases. So for Ruby's map it could have a monomorphic type signature like Array<a> -> (a -> b) -> Array<b>. It's only correct for the cases where the list only contains values of one type and the block only returns elements of one other type, but it's much easier to understand and gives a good overview of what the function does.
Yes, you seem to misunderstand the concept. It's not a replacement for static type checking. It's just different. For example, if you convert objects to json (for rendering them to client), you don't care about actual type of the object, as long as it has #to_json method. In Java, you'd have to create IJsonable interface. In ruby no overhead is needed.
As for knowing what to pass where and what returns what: memorize this or consult docs each time. We all do that.
Just another day, I've seen rails programmer with 6+ years of experience complain on twitter that he can't memorize order of parameters to alias_method: does new name go first or last?
This flies in the face of Ruby's promise of intuitiveness.
Not really. Maybe it's just badly written library. In core ruby everything is quite intuitive, I dare say.
Statically typed languages with their powerful IDEs have a small advantage here, because they can show you documentation right here, very quickly. This is still accessing documentation, though. Only quicker.
Consider that the design choices of strongly typed languages (C++,Java,C#,et al) enforce strict declarations of type passed to methods, and type returned by methods. This is because these languages were designed to validate that arguments are correct (and since these languages are compiled, this work can be done at compile time). But some questions can only be answered at run time, and C++ for example has the RTTI (Run Time Type Interpreter) to examine and enforce type guarantees. But as the developer, you are guided by syntax, semantics and the compiler to produce code that follows these type constraints.
Ruby gives you flexibility to take dynamic argument types, and return dynamic types. This freedom enables you to write more generic code (read Stepanov on the STL and generic programming), and gives you a rich set of introspection methods (is_a?, instance_of?, respond_to?, kind_of?, is_array?, et al) which you can use dynamically. Ruby enables you to write generic methods, but you can also explicity enforce design by contract, and process failure of contract by means chosen.
Yes, you will need to use care when chaining methods together, but learning Ruby is not just a few new keywords. Ruby supports multiple paradigms; you can write procedural, object oriend, generic, and functional programs. The cycle you are in right now will improve quickly as you learn about Ruby.
Perhaps your concern stems from a bias towards strongly typed languages (C++, Java, C#, et al). Duck typing is a different approach. You think differently. Duck typing means that if an object looks like a , behaves like a , then it is a . Everything (almost) is an Object in Ruby, so everything is polymorphic.
Consider templates (C++ has them, C# has them, Java is getting them, C has macros). You build an algorithm, and then have the compiler generate instances for your chosen types. You aren't doing design by contract with generics, but when you recognize their power, you write less code, and produce more.
Some of your other concerns,
third party libraries (gems) are not as hard to use as you fear
Documented API? See Rdoc and http://www.ruby-doc.org/
Rdoc documentation is (usually) provided for libraries
coding guidelines - look at the source for a couple of simple gems for starters
naming conventions - snake case and camel case are both popular
Suggestion - approach an online tutorial with an open mind, do the tutorial (http://rubymonk.com/learning/books/ is good), and you will have more focused questions.
I'm curious as to how to do a Clojure deftype that contains a reference to itself, e.g.
(deftype BinaryTree [^BinaryTree left ^BinaryTree right])
This doesn't work... however I see no intrinsic reason why it shouldn't be possible since the underlying Java class is perfectly capable of referring to itself.
What am I doing wrong here?
Mike.
Currently ^Class hints on fields (in opposition to ^primitive hints) are discarded, so there's no gain in trying to put them. This may change in the future.
However auto reference in a type definition (eg in method bodies, not in fields) somewhat works but the implementation is a bit of a hack. There's little incentive to fix auto-reference in the current java compiler given the promise of the rewrite of the compiler in Clojure.