| |
| |
| |
|
Page: 1 2 3
Comments:
<0> it's REAL quiet toniht
<0> +g
<1> and hot
<2> i'm listening to vh1 rocks
<2> it isn't quiet
<1> oh and it isn't hot then
<0> hmm ok
<0> sorted
<0> anything happen today?
<2> ["Everything that has a beginning, has an end.."]
<2> what about infinite loops?
<2> crap
<0> very true
<0> #c == lies
<3> have you ever executed an infinite loop?
<4> The Cray-3 is so fast it can execute an infinite loop in under 2 seconds!
<0> har
<4> [jargon file]
<2> megyek bye
<0> hmyeah
<0> laters
<5> hey
<5> can i do pointer arithmetic on void * pointers?
<5> or do I have cast it to char * (or sth) first?
<6> yes, by casting them to the proper types
<7> which begs the question, why is it a void*.
<6> "a generic container".... ?
<5> twkm: because I'm tied to the prototype of function variable with parameter of void *.
<5> (write(2) replacement)
<6> isn't write(2) using char *
<5> not in glibc, and I think then not in POSIX.
<5> OK, then, let's cast >:>
<7> then i suggest you use a local variable of the type you find most appropriate to the ongoing use.
<5> twkm: it's an option, but I need pointer arithmetic only once, and in the other places i can do without it. So I think single cast is OK
<7> i wouldn't.
<7> but it's your code.
<8> hi guys. how come every single time I try to compile the simplest code with GCC4, it's broken
<8> no compiler errors, no linker errors, and it even LOADS fine, but when I run it it either segfaults or just hangs
<8> compiling the same code with gcc3.6 creates flawless binaries...
<8> problem is, I've got Fedora Core 5 on my desktop, and it provides GCC4 (and afaik ONLY gcc4), which makes me kinda sad...
<9> patson: first thing to do: get rid of fc5 and replace it with debian
<8> debu_: nice try, but debian goes with gcc4 by default also
<9> patson: but still fc is problematic "#*!#
<8> as does ubuntu, and mandrake, and suse, and virtually everyone except for gentoo, which relies on a working compiler
<8> debu_: I haven't got any problems with the distribution... just the compiler and toolchain
<10> patson: you sure it's the compiler and not fedora core?
<10> you DO KNOW that debian's gcc-4 is not exactly same as fedora's
<11> file a bug report at fedora bugs?
<8> cmouse: yes, I know that... and ubuntu's GCC4 does the same crap also
<10> ok...
<8> again, using GCC 3 works perfectly
<11> compile it from source then
<10> makes one wonder how exactly they got fedora core running =)
<11> it doesn't even take long if you configure it just for c
<10> it's compiled using that gcc-4
<8> yes
<8> domen: compile GCC3 with GCC4? :)
<10> patson: are you using CFLAGS or LDFLAGS?
<8> CFLAGS with -O3
<8> but that's about it
<10> so nothing like -fiamricer?
<8> the rest is -lLIBRARY
<8> no
<10> k
<12> heh, just blame the compiler, it must be at fault when sw crashes :)
<12> I've never had any problems with gcc4 and really doubt it would compile C incorrectly
<12> at least for long
<8> oh? :)
<10> patson: care to share the source of your "simplest" binary?
<12> gcc had couple of users that would notice if it is broken
<8> hmm... not yet
<12> s/had/has/
<13> Simplest code?
<8> not yet. :)
<13> You are probably stepping outside the boundaries of C standard.
<12> blaming compiler is almost as silly as at my work where they blamed CVS :)
<8> it'll take some doing, isolating the code that borks
<10> CVS is stupid
<10> patson: gdb
<10> valgring
<10> valgrind
<10> etc.
<8> functions that usually fail are standard libsdl funcs
<12> cvs might be stupid but it doesn't cause compile errors
<8> like MapRGB...
<14> hi all
<14> is possible to destroy/unlock a locked pthread_mutex from another thread ?
<15> don't know about destroy but if you could unlock other people's mutexes, it would be pretty stupid.
<14> yes i know. I've just a problem for the termination of my app, i would to force unlock or destroy the mutex to make the thread exit
<15> "Attempting to destroy a locked mutex results in undefined behavior."
<14> uhm
<14> ok thx, i think another way
<15> you'd better :-)
<14> yes
<14> whaen i catch CTRL+C signal, is important to make all thread exit naturally or i just should exit() ?
<15> that probably depends on what they were doing and can you handle unflushed buffers or whatnot.
<16> moin
<14> hi lorindol
<14> k, thanks auris
<17> How is it possible to call a certain function of another binary, that is working as a child process of the callers parent process. The child program doesnt contain a main entry, so "execvp (PROGRAM, args);" won't work.
<10> m_0_r_0_n: yes.
<10> m_0_r_0_n: dlopen()
<17> ?
<10> dlsym()
<10> (*sym)(params);
<17> hu, thanks, have to check this :)
<15> a child process without main?
<17> yes, it is working as an extension of the parent process
<15> don't mix terminology.
<17> ?
<10> Auris: look at this nick
<10> Auris: then rethink
<18> Hehe.
<17> you are real idiots, you dont have an idea what I am working for
<15> of course we don't. at least I am not telepathic.
<15> how do you load your extension?
<17> That was my question. I ve to understand the mechanism of an openmediasystem that is using this technis to control calls, each child process that is created is one caller
<17> So the main process is asking for events and creates child processes if needed
<15> ah. well, I at least understood that you have your extension up and running already and you want to call a function in it.
<17> Yes, the system is already running
<17> But I want to understand what is under the hood
<17> I will check the advises of cmouse
<10> probably uses threading...
<10> dlopen(), dlsym(), pthread_create()
<19> please, this channel is about c++ ?
<1> no, #c++ is about c++
<20> if he has to ask that i think he belongs in #c#
<1> or #javascript
<20> :P
<2> re
<14> is there a usleep corresponmdent for windows ?
<4> #include <windows.h>
<4> [...]
<4> Sleep(msec);
<7> Sleep
<14> yes, but usleep
<14> micro sleep :)
<7> there is no exact complement.
<14> ok seems there is some multimedia api
<12> select()
<14> oo thx pil
<14> fine
<2> bye
<21> hy
<22> ciao a kuole
<23> wadd se fagg?
<21> kuole?
<22> mhm mean ciao tai kuole
<22> ;]
<22> hi sm_albanian_chick:)
<24> noob question: what is the easiest (best?) way to divide two integers and you want to have exact result on 2 decimals. For example when a integer is 2 and b integer is 3 a/b results in 0.00.
Return to
#c
or
Go to some related
logs:
#worldchat
#html
#worldchat
fresks of cock
#worldchat
#worldchat
#worldchat
#sicilia
#worldchat
#london
|
|