| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
Comments:
<0> kthx^, you might be better off including errno.h , then the int errno will have an error number
<0> most libc functions set it when they get an error
<0> can be helpfull
<1> at least on posix. not on win32.
<0> he isn't using win32 anymore. that's why he's having problems
<1> ahh.
<2> lol
<2> it gave me a 0 on return
<0> lol ??
<2> thats means success
<0> that means theres no error
<2> thats what i dont GET
<3> GET / HTTP/1.0?
<0> then it's closing the file
<2> its not a file
<0> I think he might be trying to port some kind of evil irc bot
<3> is that what he's doing?
<2> lol
<2> ircd
<0> on unix/linux everything is a file
<2> bots dont use close that much
<3> ircd already runs on unix.
<2> this is from scratch
<2> 52kb
<3> from scratch on win32?
<2> yeah
<3> how are you testing it?
<2> shell
<3> what client are you using, i mean?
<2> mirc?
<0> why would you want to port the windows version when there are versions already to run?
<2> coz this is from srcatch
<2> its is MINE
<2> ;D
<0> oh
<2> learning experience
<0> learning to handle frustration
<2> yeah haha
<2> HAHAH
<2> YES WOOT
<2> omfg
<2> =)
<2> shutdown(iClient.iSocket,SHUT_WR);
<2> ^^ the solution
<2> </3
<2> and to think all the professionals here didnt tell me ;]
<0> still can't see why you'd need to use shutdown
<1> lots of things are crappy.
<0> I never do and my server stuff works
<1> shut_wr usually provokes connected clients to close, so you can just wait a little for the read==0 then close as well.
<4> calc shark2
<5> <shark2> I can write to stdin, read stdout, repeat
<6> I did tell you about shutdown, kthx^
<2> yeah
<2> i used a diff how first time tho
<2> In function `ParseClient(Client*)':: undefined reference to `ParseData(char**, Client*)'
<2> what would be causing this errror?
<3> what, besides the obvious things?
<3> or including the obvious things?
<2> i have included everything
<2> decalred it properly ;/
<2> lawl i got it
<7> ok, i get -inf as a return value from printf(); how can i do an if() comparison to the float number if its either inf or -inf ?
<4> Boro: isinf()
<7> how can i know if its positive or negative ?
<4> man ininf
<4> oops
<4> man isinf
<7> ah, i couldnt find it on cppreference, let me try man
<7> ok, thank you
<2> if (iPoll.revents == POLLHUP) <--- that should detect when a client closes a socket?
<2> if i use recv i recv a disconnection but if i use poll + recv i dont recv a disconnection ;/
<2> =[
<4> kkkkthx^: no, disconnection normally comes as POLLIN
<4> you recv it, and get EOF or -1
<1> read/recv normally returns 0 on normal disconnect.
<4> yeah, s/EOF/0 for eof/
<8> i dont know if here is the right place to ask about windowZe. but okay.. i need to get the mac-address from a network interface. i can't use netbios, and querying ndis with DeviceIoControl makes it incompatible with non nt-based systems.. as well using SendARP, cause iphlpapi doens't exists in such systems.
<9> If you don't get an answer, try #winprog
<8> ah, thanks.
<10> howdy hund
<9> Hello Joe.
<9> Boyoboyoboyoboy.
<9> Ritchie made it you know, albeit not alone.
<11> hey, is there any way, under freebsd(phk malloc i guess) to check if a pointer p***ed to free() actually is inatialized (ie in the case of say malloc(0), the c spec doesnt guarantie a NULL pointer to be returned)
<9> You'll have to use a memory profiler.
<11> so there is no way for me in C to simply check the pointer (void *) for wether its initialized?
<9> I'm not so sure if I know what you mean.
<1> not safely.
<11> consider this: you have a call to malloc() with a size depending on the length of something, and at times the call might be malloc(0); (although never a negative number)
<1> if you hate portibilty you could check if the control blocks are where they should be.
<11> when malloc(0) occurs malloc() doesnt return NULL (atleast on freebsd 6, haventested others)
<9> He did specify freebsd.
<11> (and this is according to c spec)
<1> p = malloc(something); if (something && !p) { /* oops, failed! */ abort(); }
<11> i already have such a check but theres still a unitializationg problem there
<9> hylandrr: ANSI C says that malloc(0) either gives a NULL pointer or a unique pointer.
<1> either of which may be safely free'd.
<1> i.e., free(malloc(foo)) is safe for any value of foo.
<1> (though you might run out of swap!)
<0> I always thought freeint a NULL pointer made an exception
<9> "If ptr is a null pointer, no action shall occur."
<11> hund: this is from the freebsd mailinglist, on a question someone posted
<11> // The C standard does not require malloc(0) to return NULL;
<11> > // but whatever it returns MUST NOT be dereferenced.
<1> right. so don't.
<9> hylandrr: Dereferenced is a whole other issue.
<9> hylandrr: Dereferencing is when you use * or [].
<1> p = malloc(something); if (something) { if (!p) { /* failed! */ abort(); } ... }
<11> yes but read what it saysl it doesnt guarantie that it returns a null pointer
<11> a // The C standard does not require malloc(0) to return NULL;
<11> err
<11> mispaste
<11> hund i know that
<1> notice that i test if there was an allocation requested.
<9> Yes, but it's guaranteed to be *unique*.
<11> hund: and how do i check that?:P
<9> How do you check what?
<9> You might want to zoom out and tell us the real problem at hand.
<11> malloc(0); returns neither a NULL pointer and neither something i can p*** to free()
<11> thats the problem.
<8> well. you can free a malloc(0) return value (the pointer to ZERO size buffer).
<11> now how do i check for this case?:)
<11> phoenixx on freebsd that leads to a verbose error
<9> Why do you need to check? Just do it.
<8> did you tested it?
<9> No, malloc(0) always gives something you can p*** to free().
<11> phoenixx yes
<8> do it in a while loop
<8> in a simple program and test again.
<11> hund i told you my problem like 3 times now.
<11> phoenixx ive tested this for 4 days now:P
<9> And yet you don't listen.
<11> hund you dont understand my problem it seems, perhaps im not making myself clear.
<11> pastebin
<11> hold.
<8> good choice :)
<9> hylandrr: http://rafb.net/paste/results/expug694.html
<9> hylandrr: That's perfectly conforming.
<8> put it on a while loop for m***ive test ;p
<11> hund: od in free(): warning: chunk is already free
<11> heh
<1> that's just your memory debugging being stupid. if you care you can compare foo against 0.
<11> twkm that was hunds testprogram
<11> :P
<1> i understand.
<11> http://pastebin.ca/77827 heres the memory functions i use
<1> you commented on it.
<1> i commented on the comment.
Return to
#c
or
Go to some related
logs:
9ale7 shhab
#firebird
#windows
GTA SANDANDRS
cygwin crap
#gamedev
mr zulo gay
#beginner
flash buttons.fla
#politics
|
|