What are you talking about? His array size is 3. So you can access only elements arr[0], arr[1] and arr[2]. In his code he is doing something like *(array_ptr+1*sizeof(int)) which is equivalent to arr[4] assuming integers are 4 bytes long. This is wrong. *(array_ptr+2*sizeof(int)) accesses the element arr[8] which is also wrong.

but every time T2 sends a message, he has to to start a timer. if the timer expires before the ACK is received, he has to send that message again.

i've never used a state machine and non-blocking mode lol but if the sending window is really big, maybe an ACK arrives when the thread is still sending stuff. what do you think?

N = 4 send 0 send 1 send 2 send 3 receive ACK 0 send 4 receive ACK 1 send 5 ACK 2 and 3 maybe are lost, so the timer ends and he needs to send them again

starting a new timer thread loop, providing a interface to add/erase timer handler, and that thread loop runs due timer handler repeatly.

yes, default is blocking, so waiting will timeout(may forever)

c - Making recvfrom() function non-blocking - Stack Overflow https://stackoverflow.com/questions/15941005/making-recvfrom-function-non-blocking

a general way is set fd to NON_BLOCK, and using select or epoll with timeout argument.

a general way is set fd to NON_BLOCK, and using select or epoll with timeout argument.

yeah, but then i don't know which timer has ended, because i need N timers, one for every msg sent

the server will send explicitly "i MSG" the client will send explicitly "ACK i" so no problem for that

You do it with: array[ index ] or you can use a pointer: int array[3] = { 1 , 2 , 3 }; int * array_ptr = array; Then access it with: *( array_ptr ) //will return 1 *( array_ptr + 1 * sizeof(int) ) //will return 2 *( array_ptr + 2 * sizeof(int) ) //will return 3

It doesn't create a new range. It is a view on top of your existing range. Nothing is copied.

Yeah but most of them uses the calling convention of System V ABI Also it makes easier to integrate C programs with assembly But compilers like Clang and GCC on optimisation level -O3 will not follow it precisely and practically they don't usually speed up your program that much (depending on the program)

Yeah but most of them uses the calling convention of System V ABI Also it makes easier to integrate C programs with assembly But compilers like Clang and GCC on optimisation level -O3 will not follow it precisely and practically they don't usually speed up your program that much (depending on the program)

Yeah but most of them uses the calling convention of System V ABI Also it makes easier to integrate C programs with assembly But compilers like Clang and GCC on optimisation level -O3 will not follow it precisely and practically they don't usually speed up your program that much (depending on the program)

Yeah but most of them uses the calling convention of System V ABI Also it makes easier to integrate C programs with assembly But compilers like Clang and GCC on optimisation level -O3 will not follow it precisely and practically they don't usually speed up your program that much (depending on the program)

> But compilers like Clang and GCC on optimisation level -O3 will not follow it precisely No, if they don't follow it precisely, it's a broken compiler.

Wrong, see https://en.cppreference.com/w/cpp/language/as_if If what you were saying was true, any compiler with the ability to inline function calls was "broken".

if a *function call* is inlined, it's not a function call, so it doesn't follow the calling convention w.r.t. function call because the code is fully merged into a single subroutine what is the point of your objection?

that the original statement of "But compilers like Clang and GCC on optimisation level -O3 will not follow it precisely" is accurate. In practice, if you're not dealing with exported symbols, the compiler can and will generate *anything*.

You missed & in name[i]

that the original statement of "But compilers like Clang and GCC on optimisation level -O3 will not follow it precisely" is accurate. In practice, if you're not dealing with exported symbols, the compiler can and will generate *anything*.

Will not follow how? I agree that it can generate anything, *but* it must not violate the ABI. If it does, then it's definitely buggy.

Hence why i mentioned exported symbols. as long as the compiler (& to some extent the linker) can prove that whatever you're dealing with, it can generate any code, even breaking the ABI. (< as in, if you look at the layout of the data / calling conventions / whatever, it doesn't follow anything standardized.)

I would love to get more explanation on this part: > even breaking the ABI I can't see the safe guarantee behind this, because if it's breaking the ABI, how come your program run properly? A calling convention exists to make each routine run without stepping on each others' toes. If you break this you don't have any way to call functions, syscalls, and anything that's ruled inside the ABI.

Hey, but wait... I think I remember you once discussed this with me on the lore. So yeah, he is right. Basically the compiler does some kind of escape analysis and if it can prove that the symbol is used internally, then it doesn't have to follow the defined calling convention. You, yourself once said: "The compiler thinks x() is static and will not clobber %r8d, even the ABI says %r8d will be clobbered by a function call. So i think it should be backported to the stable kernel, it's still a fix.". And you showed me an example on that day. Look at this: https://lore.kernel.org/lkml/CAOG64qPgTv5tQNknuG9d-=oL2EPQQ1ys7xu2FoBpNLyzv1qYzA@mail.gmail.com/

Yup, this explains what i meant way better than i could've done. do you mind me DMing you some example i just dug up?

I cannot post images here.

Hey, but wait... I think I remember you once discussed this with me on the lore. So yeah, he is right. Basically the compiler does some kind of escape analysis and if it can prove that the symbol is used internally, then it doesn't have to follow the defined calling convention. You, yourself once said: "The compiler thinks x() is static and will not clobber %r8d, even the ABI says %r8d will be clobbered by a function call. So i think it should be backported to the stable kernel, it's still a fix.". And you showed me an example on that day. Look at this: https://lore.kernel.org/lkml/CAOG64qPgTv5tQNknuG9d-=oL2EPQQ1ys7xu2FoBpNLyzv1qYzA@mail.gmail.com/

Hey, but wait... I think I remember you once discussed this with me on the lore. So yeah, he is right. Basically the compiler does some kind of escape analysis and if it can prove that the symbol is used internally, then it doesn't have to follow the defined calling convention. You, yourself once said: "The compiler thinks x() is static and will not clobber %r8d, even the ABI says %r8d will be clobbered by a function call. So i think it should be backported to the stable kernel, it's still a fix.". And you showed me an example on that day. Look at this: https://lore.kernel.org/lkml/CAOG64qPgTv5tQNknuG9d-=oL2EPQQ1ys7xu2FoBpNLyzv1qYzA@mail.gmail.com/

Did this really get backported to stable kernel anyway? Lol.

is it necessary to point address of array?

How to print palindrome number (it should be prime number as well) between given range??

https://pastebin.com/embed_js/rfAQandR

Changes to head in insert will not affect head in main. You have to pass a pointer to head instead i.e. a node**

what do you mean i am passing a pointer to head

Changes to head in insert will not affect head in main. You have to pass a pointer to head instead i.e. a node**

head will be a null pointer in main after insert returns. Here is an example void fun(int* ptr){ ptr = 0; } int main(){ int a = 10; int* p = &a; fun(p); } Here p will not become a null pointer after fun returns.

Hi, How to safe programming in c ?!

Hi, How to safe programming in c ?!

Save?

Hi, How to safe programming in c ?!

Hi, How to safe programming in c ?!

use wisely scanf()

how to save spaces?

where to save?

save a string containing spaces with scanf

Read this

please can someone help me on my projects I am a student and I am currently on several projects in c++ as -jsinfinity: a javascript interpreter oriented for software creation (made with the v8 library) and -jstore: a store built around my jsinfinity interpreter which would make it possible to launch high-level applications without even having to download them, to reduce the weight occupied by the applications and etc. I need you to help me fix some bugs and tell me how I can configure my projects so that anyone can help me more easily on my projects also I have never worked on online projects with other people I would like to improve thank you for helping me Here are the links to my two repositories: contact me inbox if you want to help me

Is friend function distroy the concept of oops ?

What do you mean by destroy oops? If you mean break encapsulation ..Read this https://isocpp.org/wiki/faq/friends#friends-and-encap