| |
| |
| |
|
Page: 1 2
Comments:
<0> because thats just a vague recollection which could be completely misleading :)
<0> given you have the patch & the dicumentation that come with it, you are in a better position then I too confirm/deny my guess
<0> arrgh
<0> need caffiene followed by rest
<1> gb, now i have the patching-process logged
<1> ill pastbin it
<1> okay ?
<0> y***ine: I cant read it, maybe sumone else can
<0> wot did the docs say?
<1> there is only 2 steps that have ben skipped
<1> sorry 3
<0> did it say too patch from 2.6.16?
<0> you know this is a no skipping zone dont you?
<2> hi
<0> all these druggies
<2> can anyone tell me if the order in which network device drivers are loaded, ever changes ?
<2> i.e. do eth0 and eth1 ever get swapped ?
<0> Can?Perhaps;Want too;Perhaps not :)
<0> Xteven: depends on many things
<0> but short answer is yes
<2> ok
<2> just out of curiosity, what determines the order in which the drivers get loaded ?
<0> Xteven: its dependent on many things, no one thing determines the order
<0> in general the first module loaded gets eth0
<0> but that's not the case if it's inbuilt or you have multiple devices using the same module
<2> yes, but thats not what I mean
<2> suppose I compile all my drivers in the kernel, no modules
<2> what determines the order in which hardware is probed ?
<0> how bout saying what you mean then :P
<2> sorry :)
<0> ok then it's determined on irq & io address
<2> support I have a 3com and intel card and I compile both drivers fixed in the kernel
<2> oh ?
<2> interesting
<2> so the bus is scanned and then drivers are loaded ?
<0> it's easiest too control the order by leaving the nics as modules
<2> thats true
<2> I've considered that
<0> noone in their right mind would build them into the kernel
<2> the debian installer does not create a /etc/modules automatically
<0> but I know lots of people in their left mind :)
<2> which means, if I compile the drivers as modules, and I install a new kernel
<2> then no modules get loaded and I have no network
<2> :)
<3> hi, how do i flush the arp cache using only kernel commands?
<0> so make your /etc/modules
<2> gb: do you know of a tool to do it automagically ?
<0> Xteven: too throw in another spanner, if you're using pcmcia then it's not dependent on modules, but on the card slot :)
<2> luckily, I don't have pcmcia stuff :) only servers
<0> Xteven: my servers run on notebooks :P
<0> so like I said before many variables
<2> interesting :P builtin UPS
<0> Xteven: nope, I got them cause they're stuffed :)
<2> I think I'll create a standard /etc/modules for all my servers, which includes loads all the drivers
<2> that way, they're always in the same order ;)
<1> gb, here i log the interessting part of the patching process : http://pastebin.com/651437
<0> I usually run them off solar though
<1> i know guys this is a bit long but i would be glad if anyone of you can have a look on it
<0> y***ine: like I said I cant look at web stuff from here, no pasting ability with ease
<0> if your here in 12 hours I might take a gander then
<1> i get this : fs/sysfs/file.c:352: error: conflicting type qualifiers for 'sysfs_file_operations' while compiling
<1> gb thanks i just hope someon of the guys here would have time for that
<2> gb: you have solar panels ?
<2> wow that would be nice :)
<0> Xteven: read my away please :)
<2> ok sorry
<0> Xteven: my mistake, I should know better then too summon peoples curiosity
<0> y***ine: a look at some of the idle times in here will give you some idea on when you'll have someone look at it :)
<0> even then, not everyone reads their scrollback
<1> merci gb ;)
<1> re
<0> hmm no bobbens
<0> time for food I think
<1> :)
<0> x/me should be asleep
<0> arrgh
<4> again ?
<0> but yes more caffiene might help
<0> Kjetil: of course
<4> If you keep this up there won't be any typists left
<0> Kjetil: why do you say that?
<0> I find shooting my typist with caffiene always helps :)
<0> makes me nice n sleepy :)
<4> haha
<5> hi
<5> anyone actually awake?
<6> could anybody tell me wether there is a point in filing a bug from this : http://pastebin.com/651853 ? (2.4.32)
<7> java?
<6> java there ? nope. not even installed
<6> the process that failed most probably was some postfix process
<1> i get this : fs/sysfs/file.c:352: error: conflicting type qualifiers for 'sysfs_file_operations' while compiling my kernel 2.6-16.2 patched with patch-2.6.17-rc1
<1> here i log the interessting part of the patching process : http://pastebin.com/651437
<5> why are you running release candidate kernels if you can't diagnose patch issues?
<5> =/
<6> kumquat, that might be because he was asked to test wether some issue exists in latst rc. at least that is usually why i am running rcs :)
<5> ah
<5> that would make sense I suppose
<1> kumquat, its the only one that supprt my wireless card
<1> Richlv, no see above
<5> oh.
<5> stupid broadcom
<5> heh.
<5> well, gl...
<5> I'm using Gentoo... afaik we already have the 2.6.17 patchset ready to fly with zero mucking around to get it working...
<1> kumquat, i have a bcm4306
<5> hm
<5> where does all the code for dealing with copy-on-write pages live?
<5> ah
<8> hi all .. why is NORET_TYPE defined at all in the kernel /?
<8> seeing that -- http://lxr.linux.no/source/include/linux/linkage.h#L38
<8> and what is FASTCALL ?
<8> http://lxr.linux.no/shttp://lxr.linux.no/source/include/asm-i386/linkage.h#L6ource/include/asm-i386/linkage.h#L6
<8> ewww
<8> http://lxr.linux.no/source/include/asm-i386/linkage.h#L6
<5> fastcall is a calling convention
<8> kumquat: that means ?
<5> it defines the stack/register setup for a subroutine call
<8> and what advantages does it offer ?
<5> speec.
<5> speed evne
<8> even even
<8> :D
<5> blah, I can't type. bbs, cleaning keyboard :)
<8> also, can i get something like lxr.linux.no for my pc ? being on the net hours for it ****s
<8> and building / installing lxr is too tedious
<8> any other options i have ?
Return to
#kernel
or
Go to some related
logs:
#web
ERROR 1268 (HY000): Can't drop one or more of the requested users
#python
rafl perl blog
#gentoo
reflect failed to find getAwtLockFunctions.
riceandmeat
gconfd GConf server is not in use
apt-get libmikmod
#math
|
|