@# Quotes DB     useful, funny, interesting





Google
 
Web www.quotesdb.info
Undernet  |  EFnet  |  Quakenet  |  Freenode  |  Dalnet  |  Ircnet  |  Galaxynet
Page: 1 2



Comments:

<0> is there a way to interrupt startup scripts during boot, like ctrl-c with bsd
<0> for when something like nfs hangs
<1> Is there a performance penalty involved when using a modular kernel driver as opposed to one that is compiled in?
<2> robink: nope
<3> Yo
<3> what is a reasonable number of max processes for a desktop machine?
<2> per user?
<2> i use 128
<2> but i guess that might be not enough for some java enabled things
<3> balrog-kun: I see... what could be enough? 256?
<2> dunno
<3> balrog-kun: I'd like to implement a builtin fork protection into the kernel
<2> few programs will work without functional fork() :)
<3> balrog-kun: Oh, it will be functional... I just want to kill abusers...
<3> balrog-kun: Is 1024 enough for the whole OS?
<2> for a desktop, i'd guess yeah



<2> would be enough for my desktop
<3> Good to know :)
<3> balrog-kun: The current kernel is too vulnerable to bomb forking
<3> balrog-kun: I've seen ulimit (bash's), but it doesn't seem to apply to root
<2> right, no kind of limits apply to root
<3> Yes, this is a big problem
<3> unless they're hard-coded into the kernel, it's impossible to fix the problem
<3> If I remember FreeBSD does this
<3> I think the default is 4096 but I'm not sure
<2> i wouldn't like my root to have a limit on number of processes
<2> i use max_proc_per_user=64 on my freebsd box
<3> balrog-kun: problem is that it can be changed under a script
<3> balrog-kun: well, some things need a limit... like the stack size
<3> oh well
<3> balrog-kun: How can I track a process?
<2> what do you mean by track?
<3> balrog-kun: locate it
<2> like where it is?
<2> where the executable is?
<3> balrog-kun: as in "find all processes with the name 'bash'"
<2> ps i guess :)
<3> if this can be done recursively, nice
<3> balrog-kun: at kernel level, I mean
<2> dunno, see the sources of ps
<3> balrog-kun: damn... they're integrated into binutils, I think
<2> yup
<2> there is a couple of functions like find_pid in kernel/pid.c
<3> nice :)
<3> let me see
<3> balrog-kun: it takes an enum pid_type
<3> Why doesn't the kernel use an alias for its own types?
<3> typedef pid_type type;
<3> oops
<3> typedef pid_type pid_type;
<3> would be a bit easier to write
<3> er
<3> typedef enum pid_type pid_type;
<3> that
<2> dunno
<2> cause it's written in C
<3> balrog-kun: so?
<3> what I wrote *is* valid C :P
<2> yep, but i don't think that's what typedef is for in C
<3> of course it is
<2> i mean yeah, but only when it gives you some advantage
<3> balrog-kun: it does
<2> you can use that typedef or not use it
<3> yeah, you can use it or not, but typedef'ing it makes things easier
<2> enum pid_type is looks simple enough and it helps to avoid some mistakes
<3> like what?
<2> not sure
<3> that's because you actually would be prone to less mistakes if you typedef'ed it
<2> hmm dunno
<2> it would also make the namespace more dense
<2> i mean you'd have all types in one namespace
<3> ok...
<3> balrog-kun: Have you coded any fork code before?
<3> The API is pretty unkown to me
<3> the major idea is in my mind, but I'm kinda lost
<2> if you don't mean code that uses fork() then no
<3> balrog-kun: I'm asking about the code implementing the kernel forking



<3> not fork()
<2> no i haven't
<3> Btw, have you coded anything into the kernel?
<2> kindof but it's not in the linus tree
<2> it's in omap tree
<3> Omap?
<2> omap is an architecture
<3> Really?
<3> Never heard of this one
<2> it's one of the trees that are very slowly merged into linus tree
<2> embedded
<3> Nice :D
<3> I see that struct task_struct implements list_head
<3> how can I use list_head so it'll point to the head of task_struct?
<4> is anyone have tried to update the kernel version to 2.6.15.6?
<5> anyone know anyway to wake the display remotely without a keyboard or mouse hooked up? :)
<6> hello guys ,i need help with a framerbuffer .. can any help me???
<2> hyakki: like when the screensaver in X turned it off?
<5> yea i guess its called console blanking
<5> dunno how to wake it from ssh :)
<2> probably run a second X session and kill it
<2> X :3 than ^C
<5> dont trust X it can crash the pc easy :)
<5> or rather the video drivers
<2> what about svgalib? :)
<2> are you using a framebuffer?
<5> nope
<2> hmm then maybe there's some file in /proc that lets you control that
<5> yea thats what i was thinking
<2> running X normally does the trick
<5> anyone know a quick way to zerofill a nearly dead drive?
<5> want to erase my data before it fails :)
<5> dd if=/dev/zero of=/dev/hdd died with error when it hit a bad sector
<2> add conv=noerror or something (don't remember)
<5> ill try it
<5> you know how to use the hardware erase ..think its a part of smart
<5> or hdparm shows this 28min for SECURITY ERASE UNIT.
<5> just dunno how to use it heh
<2> doesn't show anything like this for my HDs :)
<5> hm its under Security with hdparm -iI output
<2> oh, i've got Security: not enabled there
<5> damn
<4> any idea why " mount: error 6 mounting ext3"?
<7> what would cause lag for my mouse and xmms when compiling stuff? It also lags the first time I open a program like open office after I reboot, it never did that before when I used a genkernel, but since I made my own with menuconfig it happens now, is there something I did with my hard drive options?
<8> maybe
<9> how can i handle with more than 11 logical partitions on my scsi disk? man makedev says i can only have sda5~sda15, but i wanna use all 256 logical partitions my extended partition supports
<10> im on debian, i need to load the module B when is loaded the module A, but if i write it on /lib/modules/..../modules.dep , if i make a "depmod -a" it are deleted, any idea ? maybe using /etc/modules.conf ?
<9> Thanatermesis, no idea
<3> yo
<11> yo!
<3> neil_m: have you ever used list_head?
<3> Hey
<3> I've been checking list_head's implementation
<3> so the address is always given by any_list->h, where h is of type list_head?
<12> :c
<13> Hi. When closing an fd while doing epoll_wait nothing happens, but what happenes when epoll_ctl EPOLL_CTL_ADD an fd while epoll_wait and then write somthing to that fd? Will epoll_wait return? Does edge-triggered mode make a difference here?
<14> Hi, anyone here has ever tried to implement RSA in the linux kernel? I am working on a porject that demands it so would like some hints as to where to start from and how to go about it. Please give some suggestions
<15> hi all
<15> who all have commit permission to the kernel tree ?
<15> only linus ?
<16> thatd be silly
<16> the devs have acess
<15> and those are ?
<16> probably will say it somewhere on kernel.org
<15> ok.
<16> if you look at the changelogs you can see some of them
<15> oh, yes.
<2> the people in changelog are mostly people without direct access to the repository
<2> they are the authors of the patches, not the commiters
<17> most oughta be committed though :)
<17> im not vertain but I think rusty is still the kernel patch monkey
<17> arrgh certain rather
<15> who is rusty ?


Name:

Comments:

Please enter the result of the sum 63 + 46 (to avoid spam):






Return to #kernel
or
Go to some related logs:

libdevread3
#gimp
link to filesystem ubuntu
mplayer full screeen
#ubuntu
#perl
openembedded makedevs
www.flashguvenlik.com
ubuntu downgrade apt-get mythtv
Graffiti FontsT 3.0



Home  |  disclaimer  |  contact  |  submit quotes