| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10
Comments:
<0> however, lik eI said
<0> I have seperate functions for stuff
<0> like, schedling is its own
<0> I have one for saving thread states
<0> etc
<0> I can call them at any time.
<0> privided im in dpl0
<0> Systemcall inteface is useful.
<1> OK, I see
<1> Zeii: so, set the stack frame's CS,DS e.t. with different privilege level, the new task can be ***igned to run in kernel or user space ?
<0> well, yes
<0> if you set the CS to say....
<0> the Selector for Userspace | 3
<0> youll be requesting it
<0> basically
<0> CODESEG | DPL
<0> after the | is the requesting dpl.
<0> codeseg is the code segment for the dpl you want
<1> Zeii: sorry for delay, I just have a phone call
<1> Zeii: I understand you :-) Thank you
<1> very happy to discuess with you
<0> no problem man
<0> I need to go for a walk, RSI is really .. hurting my wrists
<0> afk
<2> wob
<2> mop
<3> morning mur
<2> gm
<4> [12:28:20] <0> I have a Scheduler, a main scheduler
<4> [12:28:28] <0> then I have a crazy Clock interrupt handler, which is like a scheduler itself.
<4> [12:28:34] <0> keeping track of quantums and such
<4> [12:28:51] <0> I think your solution is okay :)
<4> [12:29:43] Cheery [n=Cheery@a81-197-54-146.elisa-laajakaista.fi] has joined #osdev
<4> [12:30:01] <1> Zeii: OK, I see
<4> [12:30:28] <0> man I hate lilo ><
<4> [12:30:32] <1> Zeii: but really, everytime for sechedule, a IRQ0 has to be invoked, I think it is expensive
<4> [12:31:22] <0> it depends on your Microkernels design
<4> [12:31:36] <0> Systemcalsl arent painful in my system
<4> [12:31:39] <0> however, lik eI said
<4> [12:31:45] <0> I have seperate functions for stuff
<4> [12:31:47] <0> like, schedling is its own
<4> [12:31:52] <0> I have one for saving thread states
<4> [12:31:52] <0> etc
<4> [12:31:55] <0> I can call them at any time.
<4> [12:32:03] <0> privided im in dpl0
<4> [12:32:10] <0> Systemcall inteface is useful.
<4> [12:32:15] <1> OK, I see
<4> [12:32:39] gr00ber [n=gr00ber@135.80-202-180.nextgentel.com] hat IRC verl***en: "Leaving"
<4> [12:33:13] <1> Zeii: so, set the stack frame's CS,DS e.t. with different privilege level, the new task can be ***igned to run in kernel or user space ?
<4> [12:36:25] <0> well, yes
<4> [12:36:29] <0> if you set the CS to say....
<4> [12:36:35] <0> the Selector for Userspace | 3
<4> [12:36:40] <0> youll be requesting it
<4> [12:36:41] <0> basically
<2> Mathis, what are you again up to?
<4> [12:36:44] KillerX [n=anant@gentoo/student/KillerX] hat IRC verl***en: "Leaving."
<4> [12:36:45] <0> CODESEG | DPL
<4> [12:36:52] <0> after the | is the requesting dpl.
<4> [12:37:00] <0> codeseg is the code segment for the dpl you want
<4> [12:40:06] colyli [n=coly@61.51.107.139] hat IRC verl***en: Remote closed the connection
<4> [12:40:20] colyli [n=coly@61.51.107.139] has joined #osdev
<4> [12:44:26] KillerX [n=anant@gentoo/student/KillerX] has joined #osdev
<4> [12:56:38] linus__^ [n=nb@host110-183-dynamic.6-87-r.retail.telecomitalia.it] has joined #osdev
<2> Mathis, stop.
<4> [13:10:09] <1> Zeii: sorry for delay, I just have a phone call
<4> [13:10:16] <1> Zeii: I understand you :-) Thank you
<4> [13:10:23] <1> very happy to discuess with you
<4> [13:12:25] <0> no problem man
<4> [13:12:31] <0> I need to go for a walk, RSI is really .. hurting my wrists
<4> [13:16:25] <0> afk
<4> [13:18:09] wobster [n=wobster@p50927628.dip.t-dialin.net] has joined #osdev
<5> :DD
<1> poor Mathis
<5> now we could have a fiesta!
<5> #osdev.idiots--;
<2> /kb Cheery
<2> ;)
<5> ha! Don't try double fiesta ;)
<6> mur: woot - its the first time i see you kicking people ;P
<2> ronny_, stop sleeping then :P
<6> mur: its not my fault - its all about timezones
<6> hmm
<7> timezones == evil
<6> timezones rock - it keeps the time synced with the daylight
<6> all that wouldnt matter if we used the global internt time someone developes some years ago
<5> An another ridiculous comment, mwk.
<5> how something you can't change can be evil? :)
<8> ideally we would have a time signature that indicates both a shared numeric reference as well as a local daylight reference
<2> ronny_, yeas, WHY DO YOU SLEEP
<8> perhaps a combination of number and letter
<2> Its only 1204124@
<8> where the letter could indicate the sun's position
<2> dude it's ALREADY 1204100@!!!
<2> go to sleep
<2> hehe
<6> mur: im still human - god jost wouldnt make me an ethereal caffeine driven entity
<5> mrd: that sounds interesting.
<8> but then i guess letters aren't as universal as numbers
<6> letters are as universal - as you can easy index them
<2> ronny_, you wish :P
<6> nah - being ethereal gets boring after a few hundred decades
<5> hours counted as certain sequences of letters or names, making an objective time measurement so that it's relative for people.
<6> hmm
<5> XOZ is morning for people on one area, and night for others on another.
<5> lol
<8> suffixing with letters would also be nice, since the letter could be optional if the writer feels the location or daylight has significance or not
<8> though it wouldnt dispose of timezones would it
<5> ok, but now, there's 24 hours in day, what letter would mean what part of day it is?
<8> you could keep the hour:min numeric standard, but only one hour exists everywhere at the same time. different time zones would have a different letter to indicate the sun's position
<5> mrd, actually yes it would, you could resolve the timezone time from location and the letter.
<5> I wonder, what is more important information, how long time has been spent or what time of day it was?
<8> depends on the context
<8> but if you separate the two, people would choose the appropriate identifier to indicate their meaning; so instead of saying 'it happend around 3' they could say 'it happend around H' heh
<5> hmm, so.. each timezone would have a letter and there would be yet a global time?
<8> ah, no each timezone would have a different letter at any given time, due to the sun's relative position
<5> ah, ok.
<5> what letters would be used for such suffixing and how would they go?
<5> so that there wouldn't be i-1 in the pattern. :)
<7> hm...
<8> good question
<7> what about starting with A on 180
<7> and just going 'round the world?
<5> then it'd be i1 = A, 180(pallo) i(n+1) = in + 1
<5> yes, maybe it could go. :)
<7> date is non-continous on 180 so letters are non-continous there too
<7> this doesn't accomodate weird situations, though... like timezones that are n hours, 30 minutes behind UTC
<5> so greenwich midday would be A ?
<5> while in here A would be two hours after midday.
<5> I guess it wouldn't be too continuous. :/
<0> What is Mathis problem anyway?
<0> hey you guys
<0> can I make a bootloader stage 1, out of multiple .asm sources?
<9> Zeii: Yes. And sorry for not giving feedback yet. Didn't have time to dig through the sources yet :\
<0> its okay.
<0> Serishema eventually picked the creap out of it then said, its pretty good and isnt vulnerable to the things she said.
<0> which was cool
<0> lol
<9> :D
<0> once she got the design
<0> Im going to try and port my Intel syntax bootloader, to AT&T syntax :P
<0> part of hte new Ion initiative.
<9> Eww. :s
<9> That sounds.. like hard work ;)
<0> well, it shouldnt be
<0> btw, JimPanic, you have any experience with sVN?
<0> Im trying to setup an SVN thing
<0> but ive not done it before.
<9> Zeii: What host OS?
<0> Linux
<7> fun thing
Return to
#osdev
or
Go to some related
logs:
#qmail
#perl
#math
modprobing in ubuntu
virtual_transport_maps
#gentoo
sles 9.3 online-update rpm --import
apt-get You must have the Lame MP3 encoding library installed to compile Myth.
gentoo alsasound' not found; aborting
english word for apartment
|
|