C / C++ / MFC

Victor Nijegorodov25-Sep-23 5:32

Victor Nijegorodov

25-Sep-23 5:32

This memory is allocated on the stack and will be "released" after the loop finishes.

Calin Negru25-Sep-23 5:42

Calin Negru

25-Sep-23 5:42

The program knows I don’t need the integer after the loop is done, thanks.

Randor 25-Sep-23 5:36

Randor

25-Sep-23 5:36

It's been the same since C99

C99 block scope

C++ block scope

jschell25-Sep-23 7:25

25-Sep-23 7:25

I suspect maybe that is a bit different especially given the OP.

C, so before C++, would limit the scope to the method. C++ would have been based on the same. So if one has 3 for loops in one method with each in a block for a if statement, the compiler might or might not have limited it to the block.

What you refer to makes that explicit. The compiler thus must reuse the stack space.

The OP though is referring both the scope which is a method and, I believe, to each iteration of the for loop.

Randor 25-Sep-23 7:48

Randor

25-Sep-23 7:48

jschell wrote:
C, so before C++, would limit the scope to the method.

No, C89, C90 through until C98 had method scope.

jschell wrote:
So if one has 3 for loops in one method with each in a block for a if statement, the compiler might or might not have limited it to the block.

The ISO standards are listed at the bottom of this page. Could you point out what you are referring to?

I see what you are saying now. You are describing old C89 rules. Are you an embedded C programmer? Those scope rules were changed way back in 1999.

Yeah, modern compilers don't have method scope at all anymore. Some embedded compilers still do C89

modified 25-Sep-23 14:04pm.

jschell26-Sep-23 5:10

26-Sep-23 5:10

Randor wrote:
Could you point out what you are referring to?

I should have read your original post more carefully....

The following is part of what the OP posted.

"allocate memory for an int every frame."

To me the terms I underlined are significant.

Most compilers that I have ever seen, not just C/C++, use a 'stack frame' to manage the variables within a method.

The allocation, far as I can tell, is how the OP is referring to, because the post specifically uses those terms.

But it is still up to the compiler. As a matter of fact at least at one time compilers at one time made a big deal (advertising) that the method variables were managed as CPU 'register' values and were not put onto the stack frame at all. And that is definitely not in the specifications for C or C++.

I do know, because I looked at the assembly that compilers used to emit (and at times modified it) that compilers at one time did nothing more than allocate variables on the stack frame sequentially. I might even recall reading an article that a developer would need to manage variables more carefully to limit that. (I can perhaps recall the suggestion that all variables should be declared at the top of the method for that very reason.)

Now back to what you posted...

For what you posted the "scope" refers to where the variable is visible from. At the language level.

That does NOT specify how the compiler is to manage the variables.

Do you have a different specification, which would probably need to be after C99, that does state how the stack frame is to be managed? I say C99 since I was familiar with that one and I am rather certain that it says nothing at all about the stack frame.

I also looked through my books for "C++ Programming Language" and "C Programming Language" and found nothing at all about the stack frame. I did not expect to find it.

I did look through the Dragon book where I would expect this to be discussed. It doesn't use that term instead it uses the term 'activation record'. It discusses how the activation record can be managed by a stack.

Randor 26-Sep-23 5:38

Randor

26-Sep-23 5:38

The C99 standard is here: C99 Standard

Scope starts at 6.2.1

But you would probably want to read: 6.2.4 Storage durations of objects

I use to read all these many years ago, I stopped reading the specs after C11

At Microsoft we had a huge internal mailing list where everyone got to watch the compiler team go back and forth over the new language features. The gatekeeper of the STL library had the initials S.T.L. I always thought that was funny.

jschell27-Sep-23 5:23

27-Sep-23 5:23

Randor wrote:
6.2.4 Storage durations of objects

But 6.2.4 says nothing about how a stack frame is built.

Just as with your other reference it explains what the compiler must enforce but not how it must enforce it.

Following is the only thing that relates to the language

4 An object whose identifier is declared with no linkage and without the storage-class specifier statich asautomatic storage duration. 

5 For such an object that does not have a variable length array type, its lifetime extends from entry into the block with which it is associated until execution of that block ends in anyway.(Entering an enclosed block or calling a function suspends, but does not end, execution of the current block.) If the block is entered recursively, a new instance of the object is created each time. The initial value of the object is indeterminate. If an initialization is specified for the object, it is performed each time the declaration is reached in the execution of the block; otherwise, the value becomes indeterminate each time the declaration is reached.

The second paragraph is the only one even close to relevant and basically makes the same point as your other reference.

It does not specify how it is built on the stack frame.

Again a compiler writer could make a fully compliant compiler which used new slots on the stack frame for each block. Or it could reuse existing ones. Both implementations are compliant.

The referenced section allows a compiler writer to produce optimized code that reuses the slots. And they cannot be considered non-compliant if someone attempts to use a declared variable in a block outside the block (for example via a pointer.)

Consider exactly that case - using a pointer outside the block.
1. Compiler A uses new slots so the code works.
2. Compiler B reuses slots so the code doesn't work.

The user complains that Compiler B is non-compliant. The creators can tell them explicitly that they are using code that the spec does not support.

But it says nothing about Compiler A. Compiler A is NOT required to attempt to determine that a pointer belongs to a variable that goes out of scope. (Similar to having a method return a pointer to a local variable.) The compiler might choose to warn about that but is not required to do so.

Randor 27-Sep-23 6:57

Randor

27-Sep-23 6:57

Thanks for writing all that. The ironic part of this conversation is that the answer to the question at the top of this thread been the same since C99. $Shucks | :-\$

trønderen28-Sep-23 13:50

28-Sep-23 13:50

I'm still not clear about this.

If variables X and Y are defined in a block, must they both be available throughout the lifetime of that block?

If the compiler does a complete flow analysis, detecting that variable X is only used in the first half of the code, and variable Y only in the second half, with no overlapping use possible for any possible execution path, can then X and Y share the space?

For the running code, sharing would be OK. A debugger might display both X and Y as soon as the block is entered, and all the time until the block is left. If X and Y share a location, then Y would be incorrectly displayed for the first half, X for the second half. Does anything in the C++ standard forbid this? Does the language standard at all relate to tools like debuggers, or only to the executing program code itself?

jschell29-Sep-23 11:07

29-Sep-23 11:07

trønderen wrote:
If variables X and Y are defined in a block, must they both be available throughout the lifetime of that block?

Via the spec? No. The spec is just asserting that they are not 'available' outside the block.

In reality, in terms of the compiler emitted code? Probably.

Consider what difficulty the compiler has in determining what the scope of A is in the following. Keep in mind that the code can get much more complex.

{
   int A = 3;
   int *B = &A;
   int **C = &B;
   doit(C);
   
}

trønderen29-Sep-23 11:35

29-Sep-23 11:35

jschell wrote:
Consider what difficulty the compiler has in determining what the scope of A is in the following. Keep in mind that the code can get much more complex.
{
int A = 3;
int *B = &A;
int **C = &B;
doit(C);
}

This is the kind of examples that tend to give great respect for those people trying to optimize C/C++ code. And also a great pity with them.

Randor 29-Sep-23 12:04

Randor

29-Sep-23 12:04

trønderen wrote:
If variables X and Y are defined in a block, must they both be available throughout the lifetime of that block?

jschell wrote:
Via the spec? No.

Regarding C99 the lifetime is defined in 6.2.4 paragraph 2:

The lifetime of an object is the portion of program execution during which storage is
guaranteed to be reserved for it. An object exists, has a constant address,25) and retains
its last-stored value throughout its lifetime. 26) If an object is referred to outside of its
lifetime, the behavior is undefined. The value of a pointer becomes indeterminate when
the object it points to reaches the end of its lifetime.

Then paragraph 5 defines how that "lifetime" is applied in the block scope.

modified 4 days ago.

jschell2-Oct-23 5:09

2-Oct-23 5:09

Randor wrote:
Then paragraph 5 defines how that "lifetime" is applied in the block scope.

Just curious - what exactly do you think the scope is when a CPU register is used rather than a stack frame slot?

Randor 2-Oct-23 5:47

Randor

2-Oct-23 5:47

jschell wrote:
Just curious - what exactly do you think the scope is when a CPU register is used

All rules go out the window on the optimization pass. Could even optimize away the objects. But the initial code generation will follow the rules above if the compiler is "C99 compliant"

In fact if /LTCG is enabled entire functions could be removed/combined. Compiler optimization and link time code generation is an entirely different discussion.

trønderen2-Oct-23 8:01

2-Oct-23 8:01

To me, that sounds like one of two:

Either, the C/C++ scoping/life time rules apply to the source code only and how it 'conceptually' is executed, and does not relate to the compiled code and how it actually runs.

Or, the C/C++ standard does not, strictly speaking, allow optimization of the kinds that we are talking about here, such as allocating a variable in a register, reuse of stack locations within a single scope, or removing variables entirely.

I don't know which is correct. To me, it is the running code that matters, even if the correct interpretation of the standard is that it doesn't care about the generated code.

Randor 2-Oct-23 8:27

Randor

2-Oct-23 8:27

Yeah,

I guess the only way to answer a C/C++ language question is to point at the C/C++ standards that define the language.

It reminds me of this old blog post: This processor has no stack

trønderen2-Oct-23 12:17

2-Oct-23 12:17

The question is how the C/C++ standard defines its own scope. How far does the authority of the standard reach. How does it define which parts of the system must be compliant to the standard for the system to call itself standard compliant. Does it stop at the parser, or could an optimizer have any effect on whether the system is compliant.

This is more a "meta-question", not something you will find in the syntax description pages. If you find it at all in the standard text about the authority and scope of the standard itself, and which parts of a system must comply to the standard, it is probably in some introductory or concluding chapters. I do not expect it to be in the standard itself, and if it is there, I expect the wording to be so formalistic that you have to have some experience from working in the standards committee to get it right.

I did pick up a copy of a late-in-the-process draft documents, but the final document is too high priced that I am willing to pay that much for the spec of a language I use very rarely nowadays. I tried to read the drafts, but they were too huge for me to read every word - and in particular: to understand every word of it Smile | :)

Randor 2-Oct-23 12:54

Randor

2-Oct-23 12:54

trønderen wrote:
they were too huge for me to read every word - and in particular: to understand every word of it

Yeah, me too.

The last language spec I read was C11 and C++11 and haven't looked at C++14 or higher. I've learned some the new language features but not reading the specs anymore unless I have to.

trønderen21hrs 5mins ago

21hrs 5mins ago

Randor wrote:
It reminds me of this old blog post: This processor has no stack

Slight sidetrack: Fortran did not allow recursion until the Fortran 90 version; the memory usage was fixed. One of the arguments brought up to keep Fortran alive was that you never risked a stack overflow in Fortran; it was safer than stack languages. Besides: Since there were no stack management, the call overhead was reduced.

At least in its first standard version, Ada required recursive functions to be flagged as such in the program text, and that the maximum calling/recursion depth to be statically determinable at compile time, so that the maximum stack requirement can be calculated in advance.

CHILL also required recursive functions to be marked as such, but I am not sure if you were allowed to code an 'uncontrolled' recursion. At least, to manually search for stack overflow problems, you could limit the search to functions marked as recursive.

Randor 18hrs 8mins ago

Randor

18hrs 8mins ago

trønderen wrote:
Slight sidetrack

If you want to get further sidetracked you can watch Aaron Ballman go over all the C23 changes. He's on both the C and C++ committees. Video was posted yesterday.

I'm not going to be using any C23 anytime soon, but I try to keep up with all the language changes.

trønderen8hrs 22mins ago