| |
| |
| |
|
Page: 1 2 3
Comments:
<0> have you watched the video?
<0> (that goes with the news story, that is)
<0> the article just mentions that Hans wasn't home when the warrant was served, but the video says police believe he fled the country, and have been unable to locate or contact him
<1> that will put a whole new quality on ReiserFS-Linux flames
<2> yeah
<3> hahah
<3> thats awesome
<1> anyway, I don't want to be Raiser's children now
<4> morning
<1> morning
<5> hello everyone :)
<5> and hi undesktop here too :)
<1> yeah, hi :-)
<2> cool, fixed the nfs file handle hashing problem
<5> hi geist :)
<2> still basically ****s. you have this 32 (or 64) byte opaque blob you need to track
<2> finding a decent hash algorithm for something you have no real knowledge of is pretty hard
<2> i found that summing up all the words and then folding the top 16 bytes into the bottom 16 bits is decent enough
<1> maybe it's already equally distributed
<2> it isn't
<1> hm, bad
<1> does it still contain inodes and so?
<2> here's a smattering of them from my nas box (runs linux on ext3)
<2> 01000001003a0000035640480358005260c80574000000000000000000000000
<2> 01000002003a00000356404803580053208f1beb035800520000000000000000
<2> 01000001003a000003564048053d0011208f1bf0000000000000000000000000
<2> 01000002003a00000356404803580054208f1e8e035800520000000000000000
<2> 01000002003a00000356404803580055208f1f24035800520000000000000000
<2> 01000002003a00000356404803580056208f20ab035800520000000000000000
<1> uh pretty bad as hash heh
<2> well, yeah and no. actually summing the parts isn't too bad
<2> since there's probably an inode in the middle of that
<2> and inodes tend to not collide (for obvious reasons)
<2> i dont have the traceback stil, but when i was first testing it i was testing against another server who's disk was basically empty
<2> so the handles looked very similar
<2> but if you ls -i the source files, i noticed that their inodes were all like 750, 751, 752, ...
<6> Does anyone know under which circumstance bochs fires an irq7?
<2> that's usually for the parallel port, right?
<2> or does it ***ign 'pci' devices to it?
<6> yes, but I don#t use the parallel port
<6> and I don't think so
<2> i dunno. is it a problem?
<6> Well no, it's not a problem, I'm just curious...
<2> it's pretty common for devices coming out of boot to have pending irqs
<2> it's usually not a problem because if you aren't ready to take an irq you mask it at the interrupt controller
<6> eh wait, it's not that this irq happens once and than not again, it's coming currently 3times at my os'es startup and sometimes when pressing a key or moving the mouse... Do you think I should mask that irq, because I prefer putting a notice on the screen that a not handled irq has been fired...
<2> oh most definitely. you should never accept and IRQ you arne't prepared to deal with
<2> best cause it's harmless like you just found
<2> worst case it just locks your system up because it wont stop (most level triggered pci devices would do that to you)
<6> oh... ok, thanks for pointing that out to me. :)
<2> and it's not really bad per se. if you dont have a driver for a device you should just ignore it basically
<6> yup, i'll do that
Return to
#osdev
or
Go to some related
logs:
matplotlab
php_mysql msjava
#perl
dazjorz #css
#postfix
#suse
xvncviewer: ConnectToTcpAddr: connect: Connection refused
uninstall postgresql deian
mod_python import local module
nusoap $_SESSION soapclient
|
|