| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
Comments:
<0> i commented on the comment.
<1> #define SAFE_FREE(ptr) if (ptr != NULL) { free(ptr); ptr = NULL; }
<1> allright.
<0> it is always valid to: free(malloc(anyvalue))
<2> phoenixx look at my paste i do plenty of checking :/
<0> but your debugger will whine if anyvalue is zero, so test for it yourself.
<2> the debugger isnt even active
<3> Then where did that message come from?
<2> malloc library prints it out heh
<1> there are only one checking for doing. heh. i will download this freeb**** version.
<2> i didnt setup this machine so i dont know much beyonh that
<2> hund thats certainly not in any of our programs :)
<1> where do you set the last_ptr?
<1> i saw. it's just a debug.
<2> yeah the last_ptr is just for debug
<1> okay, but useless if you dont set it.
<1> well. i cant see anything wrong there.
<1> at the moment.
<0> hylandrr: your nmalloc looks weird. nbytes as the name of the argument, but multiplied by the size of a pointer to void.
<2> its compiled with -DDEBUG ..same stuff happens with -DDEBUG
<1> and, i think you dont need malloc and realloc. only realloc is enough.
<0> hylandrr: in your nrealloc if nb is 0 you should free the block, and return a null pointer.
<2> twkm that should actually be malloc(nbytes); i pasted the wrong source revision
<1> you can realloc a non-pointer.
<0> hylandrr: since you are making some value judgement about ptr in nfree i would guess you need a stack or linked list.
<2> well ew linked lists :/
<4> linked lists are easy
<2> i want to avoid them
<2> as much as possinbl
<2> eyeah and slow jeffloc
<0> hylandrr: otherwise there is no last_ptr that is valid after two nmalloc's.
<4> not really
<2> last_ptr is only set during -DDEBUG to gcc tho, and this happens with or without it
<2> jeffloc my benches proved them for slow for what i need them for.
<0> hylandrr: but with debugging enabled you can only tolerate a single nmalloc -- provided the code were correct and something recorded last_ptr.
<1> rtos?
<2> twkm last_ptr isnt the issue here and it and the check for it is ONLY cmpiled in when gc is fed -DDEBUG, which it isnt
<0> hylandrr: anyway, the malloc library emitting a _warning_ about it is apprently done to try to help you.
<2> infact, removing it didnt do jack either
<0> hylandrr: but you don't need help so find out how to disable that nonsense.
<2> yeah well i suspect it is from reading 0 bytes off something
<2> which i expect
<2> and i looked at freebsd malloc manpage
<2> talked about setting envirnment vairables and /etc/malloc.conf
<2> to to on debugging
<2> but i want it off
<2> heh
<2> s/to to/to turn/
<0> if you want it off then surely you want to edit that file.
<0> also check for a MALLOC_OPTIONS environment variable.
<2> there is no such file
<2> there is no such environment variable
<0> and see if your program ***igns a value to _malloc_options.
<2> ive been up and down the GS
<2> ive used strace and it returns no suc h file or dir on he call to open it
<2> it doesnt
<2> ive grepped for it
<2> its frustrating me :-)
<2> twkm ive also tried giving V
<2> as malloc options
<0> well, it has the wording as if it is the standard free, but that isn't standard behavior.
<0> if i run int *p = malloc(0); free(p); free(0); i get no output at all.
<0> that's on 6.1 release.
<5> im in 6.1p1-release
<0> hylandrr: are you *certain* nobody has replaced your malloc functions, perhaps a preload library?
<5> I dont get any output either
<2> this is strange :/
<0> people would shoot if such a thing were in a release.
<1> are you using gcc hylandrr?
<2> yeah gcc, and fbsd 6.1
<1> libc version?
<0> even with MALLOC_OPTIONS=A i get no output.
<6> you malloc must be build with debugging
<5> the default for freebsd 6.1 i believe
<5> hmm let me see if i have read access to that configuration
<6> -DMALLOC_DEBUG or something
<6> at libc build time
<2> ok let me search
<2> WeirdMish that string is only found in source files, no Makefils etc
<5> check /usr/share/mk/
<0> hylandrr: anything including files that might not be noticed by you?
<2> k, searching, 1min
<2> usr/share/mk/ said nthing of it
<2> twkm only stdlib
<7> ty for voicingm e
<8> its an autovoice, my friend
<9> is there a way to send an integer with SIGUSR1?
<10> it's time to learn data structure :<... how long will it take?
<9> what data structure
<10> Data Structure in C...
<9> data structures are independent of the programming language used to implement them
<9> just know the common ones and for more advanced ones you can reference them
<11> GRRR python just exploded, I wanna download this torrent
<10> hm. it looks so complicated
<8> lol
<10> laughing at me?
<8> no, zid.
<10> lol
<8> what do you want to know about data structures?
<10> hm... i don't know.. it's one of my course for this term.. i have no idea what it's about
<8> its probably about storing data in your programs?
<9> so just learn what's in the course
<9> what term is that
<9> it's half way through the summer semester
<10> crosspondence :P
<10> but i still haven't got the papers yet
<12> zid_: How did Python explode?
<11> **** knows, but now java has totaly wrecked my machine
<11> I ****ing hate java
<11> it's crashed so badly that 'ps' is crashing looking at it
<12> Python explodes... Java wrecked your machine... Next maybe Perl will burn down your house!
<11> I may have to.. reboot
<12> And Ruby will steal your girlfriend!
<12> And JavaScript will have *** with your mom!
<11> brb I guess
<12> zid_: Did HTML rape your sister?
<11> No, that was the dog
<12> zid_: Your bad luck has rubbed off on me. I have an instance of iexplore.exe that's hung up that I can't kill.
<11> I fixed my java
<11> it was waiting on a file, which I have now installed
<12> Is there a Windows command to kill a process - but have it actually work no matter what?
<12> Now I have 3 IEXPLORE.EXE's that refuse to die.
<11> nope
<12> Well, that ****s.
<12> Windows is so reliable.
<12> I think fixing this might require a reboot.
<13> Reboots are good!
<8> reboot always helps
<13> Downtime is good!
<13> It lets the computers rest!
<8> Microsoft makes those freezes for your own good
<8> "Bah, who's gonna wait for Win2K to run now, I'll just turn off my computer and go to sleep instead."
<14> taskkill /f won't work?
<15> Teckla: sometimes they get blocked by some other app
<15> Teckla: I'd try killing explorer.exe .. afterwards, you have to restart it from inside the task mgr -- File -> New Task (Run)...
<12> Wow, Windows was so broken, it couldn't even shut itself down.
<14> you have to wait for the guy that owned your machine to shut it down
<16> when dealing with a multithreaded app, is it necessary to use a mutex to confirm access only when reading a variable, if another thread may be writing. or is it only an issue of two threads may be altering the variable
<12> feti_: You should use a mutex even when reading the variable.
<16> thanks.
<6> what's big problem, just call and order computer technician. it comes and reboots your computer just like this
<6> you should use a mutex even before locking a mutex!
<12> I should have installed the "save tabs" Firefox plug-in.
<12> Now I lost all my tabs.
<17> hmm
<17> hi teckla
<12> Hi niv_.
<17> hows it goin
Return to
#c
or
Go to some related
logs:
caught torrenting uplink
unarium in iran
#bsd
#politics
#nhl
#goal
(transport is unavailable) postfix/nqmgr
nubilescc
automate UnDMR
#beginner
|
|