| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8
Comments:
<0> anyone ever heard of chenbro server cases
<0> they got a.. really cheap 2U case with hot swappable containers
<0> it makes me wonder if it's going to crumble when used
<1> it's biodegradable too!
<0> it might be that scares me
<0> .tw based company
<0> specs look good, maybe too good
<0> I have a bad feeling I'd get it and holes/hardware doesn't line up
<1> yeah that ****s
<2> Yeah I've heard of chenbro
<2> they are alright, my old work used to import them
<2> not the best quality, but not utter ****
<3> man, irssi is being a son of a bitch
<3> it can't connect to freenode unless i type the IP directly
<4> sadly I can't seem to exploit the buffer overflow in infocmp
<5> should i memset() before free() ?
<6> Uhh, no?
<6> Content doesn't matter for free
<5> cant information leack
<7> doesn't malloc clear the allocated memory on most implementations?
<7> if so that means anyone who allocates the same memory after you free it will zero it out automatically
<5> =o
<7> if not, what the heck are you writing that requires that much security
<8> Comp, no, actually. It's not obligated to. calloc does.
<8> But the questions is still silly.
<8> question, even.
<5> silly?
<8> Yes. What exactly are you trying to do?
<9> malloc doesn't clear memory it allocates
<7> anyways... what's so damn important about the data you're storing that makes you have to clear it out before freeing?
<10> if often makes sense to
<5> paranoia
<5> lul
<10> malloc errors can make anyone paranoid, easily
<10> :-)
<10> my theory is, it's better be paranoid and double-protect against malloc/free errors, than not
<5> -_-
<8> Aeon, how exactly does clearning memory before you free it protect you from errors?
<7> I don't see how it would
<10> it's long story
<7> clearing it after malloc would help though, indeterminate starting values are never cool
<8> I mean, I can think of a scenario, but it's rather stupid.
<8> Comp, that's what calloc is for.
<7> yeah... I never officially learned most of this stuff, so I didn't know about calloc actually
<8> (You're trying to make sure you actually have the amount of memory you want to free, so you expect the memset to dump core if you don't, instead of a silent failure on free.)
<10> btw it got official name
<10> "memory poisoning"
<10> now, i don't mean memset()int whole freed memory
<10> )that would have cryptographic meaning BTW)
<10> but only ***igning NULL to pointers in the freed memory
<10> i consider it good prcice
<10> <--- Mish
<10> think about accidenatl use of freed memory
<10> if you "poison" pointers before free (by ***igning NULL or some invalid value,
<10> then you catch such errors sooner
<10> the sooner you catch errors
<7> the better
<10> the more time you save in debugging
<10> if you generally spend much time in debugging
<10> then this makes sense
<10> if you generally spend 0 time in debugging then it might not make sense to you
<8> Aeon, yeah, but that's not what we're talking about here.
<10> ah
<10> whatever
<8> Exactly.
<3> unf unf unf
<3> minix is working nice in bochs using networking now that i have the network running over a bridged if_tap device :)
<3> piggybacking onto an existing physical device using raw packets didn't work well earlier
<11> Too early.
<11> Vratha: Nice.
<12> It's gone midday here
<11> 0606 here
<3> aedinius: thanks :) had to do it on bochs under freebsd, b/c i can't get bochs running on mac os x
<3> it bitches that it can't init some sort of carbon libs
<11> Did you try Q.app? (Qemu)
<3> and parallels desktop's emulated NIC doesn't work well with minix for some reason
<3> not under OS X, but under fbsd Qemu's NICs didn't work either
<11> Ah, okay then.
<3> i tried building qemu from darwinports, but it kept failing with gcc-3.3 complaining it couldn't execute cc1
<11> Weird
<3> yeah :-/
<11> 3.3?!
<3> cc1 is under a weird location anyway
<3> yeah, i think it's using the apple build tools
<11> Apple build tools should be 4.0 =]
<3> which is gcc-3.3 and 4.0, afaik
<3> *shrug*... i don't know how to tell "port" to build with another compiler of my choice
<11> What does gcc -v say?
<3> it's using gcc-3.3 explicitly on that build
<11> Odd.
<11> Anyway.
<3> but anyway, gcc -v is 4.0.1
<11> http://www.maconintel.com/news.php?article=105
<3> oooh, nice; i should give that a shot :)
<11> Yeah
<11> It'll run near native it says
<11> Can't go wrong there
<3> word up on that
<8> Join the army, they said. See the world, they said.
<3> so is it using virtualization to achieve that now? b/c i thought emulation was slow
<3> and bochs is a CPU+stuff emulator
<7> I wonder how they optimized it enough that an emulator can run at "near native" speeds
<11> Yes, virtualization =]
<12> KidArt: It's been a long time since I read those
<7> (since they used scare quotes as well)
<7> oh
<3> aedinius: oh, they must have rewritten quite a chunk of bochs to do virtualization then :)
<11> Wait.
<11> The website doesn't agree with that link.
<11> "WinTel is our popular Cocoa graphical user interface used to control the included powerful underlying open-source "Qemu" x86 virtualization/emulation software."
<11> And it costs monay.
<3> bah, money :(
<3> i already bought parallels desktop
<3> at the very worst, i can just run X11 and tunnel bochs from my FBSD desktop to my macbook
<3> which sorta ****s... i just wish parallels desktop would fix the problem
<3> er, just "parallels"
<3> but anyway, i'll try Q.app and see if it works
<3> nice, qemu is way faster on my macbook than on my 1.4 ghz tbird
<3> well, doesn't matter anyway, b/c it still doesn't find the ethernet card produced by qemu
<13> anyone knows of a very fast algorithm written in C that checks if a number is a prime? "Given an input number n, we see if any integer m from 2 to n-1 divides n. If n is divisible by any m then n is composite, otherwise it is prime." doesnt do the trick for me
<12> Well, that's the definition of prime
<12> "does 937 divide by 2? No. By 3? .." and so on
<5> heh
<13> i've written algo to do that but its not as fast as i need it to be
<12> There are no really fast prime-determination algorithms. If there were, we'd know much bigger prime numbers
<12> Factorisation is basically the only way
<8> Factorization isn't the only way.
<8> There are fast primality testing algorithms.
<12> He lies!
<8> Rabin-Miller, say.
<13> i've heard of fast primality test algos,
<8> BoR0, look into the Rabin-Miller algorithm. It's probabilistic, but you shouldn't mind.
<13> thanks, will do
<8> Two9A, as to the biggest known prime numbers, afaik those are Mersenne primes, and detecting them relies on some special properties, has nothing to do with general primality testing.
<12> Well, the Mersennes are rather specific, unfortunately
<8> How's that unfortunate?
<14> hi all
<12> Mersenne can't get you all the primes :P
<14> anyone here have experience with inpout32.dll in mfc?
<14> KidArt: you do?
<8> ultrach, was talking to Two9A
<14> oh ok
<8> ultrach, and you ought to try #winprog
<14> ok thanks
<9> factorization is the only way to be sure
<8> I think there's non-probabilitstic fast primality-testing too. But Rabin-Miller is good enough in practice.
Return to
#c
or
Go to some related
logs:
#freebsd
#beginner
#nhl
#sendmail
#beginner
#beginner
#csharp
ntu settings for xbox live
#politics
<3 pwn
|
|