goodnight, is there any open-source C project that i can contribute on github? any suggestion would be great

Guys, who has a book about masm64 with C++, please share. Sometimes it's better to use C++ with masm64.

That is what it means when we tall about the standard. But in practice you can almost always predict the outcome of a program when you know the exact compiler/architecture and how the program executed last time. But sure, sometimes you are right. Especially when it comes to multithreded code and using memory after it has been freed the behavior can be really hard to predict at times. But not all undefined behavior is like that. Unfortunately i dont have an example here right now. But often undefined behavior jusz means that the comittee didnt define what should happen, but if you know your compiler does it always a certain way, you can know what the code will do even if it is undefined behavior :p

goodnight, is there any open-source C project that i can contribute on github? any suggestion would be great

You are confusing undefined behavior with implementation defined behavior i guess.

https://en.cppreference.com/w/cpp/language/ub Take a look at the "unspecified behavior" point. That definition was what I was beeing taught. But yeah, i think we have a diffetent definition. Note for the future: make sure we have the same definitions of things before arguing about them ^^

i have found this register uint32_t hmagic[3] asm("r12");

Register keyword though applicable in C doesn't have any meaning in C++. Even in C, you should not be using it unless you are programming in embedded environments. Your compiler will know better about optimizations like this and you should learn to trust it.

yes but if i am making something like embedded environments how i do it in c++

C++ doesn't support it. So you can't. If your compiler supports it as an extension, you can use it.

C++ doesn't support it. So you can't. If your compiler supports it as an extension, you can use it.

C++ doesn't support it. So you can't. If your compiler supports it as an extension, you can use it.

C++ doesn't support it. So you can't. If your compiler supports it as an extension, you can use it.

C++ doesn't support it. So you can't. If your compiler supports it as an extension, you can use it.

i have used restrict why it doesnt work

Do you know what restrict is used for?

i searched

Do you know what restrict is used for?

Do you know what restrict is used for?

as i understanded its to tell the compiler that a pointer will point to something that will be pointed by only one pointer during the pointer lifetime

Yes. So why do you think the compiler should not read the memory pointed to by the pointer? Restrict is just used to help with compiler optimizations. If the compiler can figure out that the pointed to memory was not changed, then it may not load the memory again. But if you change the pointed to memory using that pointer, then the compiler can no longer do this optimization and may have to read the pointed to value from memory again. So this depends on your code.

wait so how should i make it optimize

By not changing the memory using that pointer and by increasing the optimization level of your compiler.

By not changing the memory using that pointer and by increasing the optimization level of your compiler.

By not changing the memory using that pointer and by increasing the optimization level of your compiler.

By not changing the memory using that pointer and by increasing the optimization level of your compiler.

i am not changing the memory

First of all your code is wrong in the sense that you are violating the restrict keyword. You are breaking strict aliasing rules by using a cast to uint32_t * pointer even though it's is just for a read. This is not correct. Moreover, restrict is a hint to the compiler. The compiler can ignore it. At best you can increase the optimization level of your compiler to see if it works. Increase the optimization level to O3.

but how i dont violate the strict aliasing rule

(uint32_t*) ch_info This is how. If you must do something like this in C, using a union is the recommended way

is it how to violate it

(uint32_t*) ch_info This is how. If you must do something like this in C, using a union is the recommended way

but i am doing in that way because it is a buffer of many bytes and i read them as int32

(uint32_t*) ch_info This is how. If you must do something like this in C, using a union is the recommended way

This is Undefined Behavior. Read about strict aliasing rules and how you can break it without even knowing about it. https://stackoverflow.com/questions/98650/what-is-the-strict-aliasing-rule

The link above also suggests using a union like I suggested something back.

and how i make an union like that

Read the link above

Read the link above

you mean this union { Msg msg; unsigned int asBuffer[sizeof(Msg)/sizeof(unsigned int)]; };

Yes something like this but in your case the members would be uint32_t and an array of uint8_t of size 4

can you show me the example

I can't code on my phone. So unfortunately no. But your union would look like this typedef union { uint32_t a; uint8_t b[4]; } Buffer; This would work in C but not in C++. In C++, accessing a union member that was not the last one set is Undefined Behavior.

I can't code on my phone. So unfortunately no. But your union would look like this typedef union { uint32_t a; uint8_t b[4]; } Buffer; This would work in C but not in C++. In C++, accessing a union member that was not the last one set is Undefined Behavior.

and i am reading from a bigger buffer that i dont know how big it is

Why can't you declare a dynamic array of uint32_t then, instead of uint8_t and then trying to reinterpret that uint8_t array as a uint32_t? How is your buffer populated? You ask many questions without giving complete details of what it is that you are trying to do.

i am reading from a file the array contains bytes of data that i need to read as int32 int16 and 8

How do you know when yo read an int16 or an int32?

How do you know when yo read an int16 or an int32?

And assuming you know how to read it correctly, you can do this: uint32_t a; fileStream.read(reinterpret_cast<char*>(&a), sizeof(a)); This is safe.

And assuming you know how to read it correctly, you can do this: uint32_t a; fileStream.read(reinterpret_cast<char*>(&a), sizeof(a)); This is safe.

yes but then i have to call read every time i want to read something

Calling read multiple times the way I suggested is the only guaranteed to work solution. Any other solution like populating a buffer of bytes and then reading in different values by casting pointers is not a portable solution. Nor is it guaranteed to work on machines with strict alignment requirements As far as your question about restrict keyword is concerned, it is an optimization hint like I suggested earlier. Try increasing your compiler's optimization level (for gcc and clang you can pass -O3) to see if the compiler can optimize a restricted pointer by storing it in a register. It may or may not work. You can't force the compiler to do it. Maybe another way to force it would be to use const_cast on the pointer to the buffer of bytes to make it a pointer to a const buffer. That might force the compiler to use a register. But regardless of all these, it doesn't change the fact that this is Undefined Behavior in C++ though it may work.

so is it also bad to do this uint32_t info_hdr[3]; read_bytes (( char * ) info_hdr, 12 ); if ( info_hdr[0] != 0x5453494C || info_hdr[2] != 0x4F464E49 { return 2; }

This is fine depending on what does read_bytes does? Where is it reading from? If it is reading from a char* buffer populated by reading from a file, then you would essentially be doing the same thing underneath though it is masked by read_bytes in this case The C++ standard library guarantees that reading a stored int from a file into a char buffer aliased into another int works. But this is not guaranteed to work when you do the buffering and aliasing yourself.

how to implement so that, depending on the entered number, more numbers can be entered, and their number will be equal to the entered first number

how to implement so that, depending on the entered number, more numbers can be entered, and their number will be equal to the entered first number

how to implement so that, depending on the entered number, more numbers can be entered, and their number will be equal to the entered first number

int count{}; std::cin >> count; for (int i = 0; i < count; ++i) { int number{}; std::cin >> number; }

<3 thx bro

Ok