irclog2html for #elinux on 20051118

00:01.34*** join/#elinux joe_bleau (n=joe_blea@ppp-208-191-54-232.dsl.fyvlar.swbell.net)
00:29.24*** join/#elinux noclouds (n=mhfan@60.166.48.88)
01:04.51*** join/#elinux andersee (n=andersee@codepoet.org)
01:04.51*** join/#elinux joe_bleau (n=joe_blea@ppp-208-191-54-232.dsl.fyvlar.swbell.net)
01:04.51*** join/#elinux pigeon (n=pigeon@adsl-60-232.swiftdsl.com.au) [NETSPLIT VICTIM]
01:04.51*** join/#elinux T`2 (n=total@tor/session/x-eebbac9ff8ba1dd1) [NETSPLIT VICTIM]
01:04.51*** join/#elinux George (i=george@kde/developer/gwright) [NETSPLIT VICTIM]
01:04.51*** join/#elinux Evanrude (n=david@wsip-68-15-251-34.dl.dl.cox.net) [NETSPLIT VICTIM]
01:04.51*** join/#elinux GPSFan (n=Ken@65.121.49.208)
01:04.52*** join/#elinux weo1 (n=weo@p5499D0CC.dip.t-dialin.net) [NETSPLIT VICTIM]
01:04.52*** join/#elinux file (n=jcolp@mctnnbsa24w-142167049176.nb.aliant.net)
01:04.52*** join/#elinux catastr0p (n=catastr0@pcp04232047pcs.plyntv01.mi.comcast.net)
01:04.52*** join/#elinux CIA-4 (n=CIA@flapjack.navi.cx) [NETSPLIT VICTIM]
01:04.52*** join/#elinux markl__ (i=mark@dsl093-225-127.slc1.dsl.speakeasy.net) [NETSPLIT VICTIM]
01:04.52*** join/#elinux markl_ (i=mark@dsl093-225-127.slc1.dsl.speakeasy.net) [NETSPLIT VICTIM]
01:04.52*** join/#elinux Khem (n=Khem@gateway-1237.mvista.com) [NETSPLIT VICTIM]
01:04.52*** join/#elinux kergoth (n=kergoth@c-24-118-219-25.hsd1.mn.comcast.net)
01:04.52*** mode/#elinux [+ovo andersee file kergoth] by irc.freenode.net
01:09.56*** join/#elinux hufnus (n=slonsiki@DSL135-071.LABridge.com)
01:51.03*** part/#elinux joe_bleau (n=joe_blea@ppp-208-191-54-232.dsl.fyvlar.swbell.net)
02:16.06*** join/#elinux Soopaman- (n=soopaman@67.71.84.89)
02:17.00*** join/#elinux sjhill (n=sjhill@spvl-02-110.dsl.netins.net)
02:19.00*** join/#elinux joe_bleau (n=joe_blea@adsl-65-70-230-240.dsl.fyvlar.swbell.net)
02:57.43*** join/#elinux we0 (n=weo@p5499D5A7.dip.t-dialin.net)
04:33.19*** join/#elinux MonMotha (n=monmotha@cpe-71-65-34-156.indy.res.rr.com)
04:34.26MonMothaok,  I need to build a 2.4.27 kernel using gcc 3.3, but apparently gcc 3.3 is outputing assembly that my binutils doesn't like.  Any ideas?
04:34.51MonMothathis is for x86
04:51.13T0mWah
04:51.17T0mWsigh
04:51.36T0mWhere goes /me into perl hashes, again... :-/
04:51.53MonMothaT0mW: ag, you've been wrestling with those for how long>
04:51.55MonMotha?
04:52.20T0mWa while, I understand them somewhat, they seem powerfull
04:52.26T0mWI just cannot seem to "get it"
04:53.42MonMothaT0mW: I've got a binary module for 2.4.27 that I need to get to load on a newer 2.4 kernel.  What do you think my odds are?  I uses USB mostly
04:54.42T0mWI've got to organize (tree out): component package, component value, component name into a associative struture (hash?) where I can dump all 0805 SMD component names which are 10K, then the 0.1uf, etc.
04:55.10T0mWMonMotha: you could tell the module loader to ignore the module version, it may work.
04:55.21MonMothait's already being forced, so that may already be the cae
04:55.22MonMothacase
04:55.23T0mWMonMotha: in theory it should work
04:56.02T0mWIf the interfaces of other stuff hasn't changed then it should work
04:56.16T0mWs/interfaces/params/
04:56.46MonMothayeah, that's what I'm hoping
04:57.14T0mWall it would need to do is call the appropriate kernel routines.  the module loader should resolve those references to match the current kernel hooks
04:57.24MonMothaI'm successively patching up to newer versions to try to get it to even compile with my newer toolchain
04:57.36T0mWyou've got source?
04:57.40T0mWfor the module?
04:57.48MonMothano, of the kernel
04:57.51MonMothaI have no source to the module
04:57.54T0mWah
04:57.57MonMothaactually, mind if I msg you?
04:58.03T0mWgo ahead
05:29.51*** join/#elinux samscol (n=donkey@203.121.86.58)
05:33.36samscolif i wanted to make a x86 bootable image, could i do something like dd if=zero of=myimage ; dd if=stage1 of=myimage ; fdisk myimage ; mkfs -o offset where offset gets us to partition 1
05:36.10T0mW?
05:36.28samscolT0mW: i guess that's a no? :)
05:37.06T0mWI was just trying to figure out what you wanted to do. build a bootable image or put it on the harddrive
05:37.11T0mW?
05:37.39samscoli'd like to create an image, ie: a file that i can later just dd on to whatever media i like ,hdd , CF, etc
05:37.44T0mWdd is copying blocks of data, you can name a device as the target
05:38.15T0mWex: dd if=<myfile> of=/dev/hda  OR  dd if=<myfile> of=/dev/hda1
05:38.29samscolyes, i understand dd
05:38.34T0mWok
05:39.01T0mWwhat are you putting this on? what system?
05:39.11samscolx86
05:39.14T0mWdoh
05:39.17T0mWx86 what
05:39.39T0mWfor example I use ARM
05:39.56samscoli486
05:40.01samscoldoes that matter?
05:40.09T0mWbut, I have a TuxScreen, some LPD boards, a WebPal, etc., etc.
05:40.17T0mWall are "ARM"
05:40.28T0mWx86 is not x86
05:40.30samscolok. give me one second. i'll rephrase my question. it's not an arch dependent thing
05:40.42T0mWit is x86 something or something x86
05:41.02samscoli want to create a file. this file will have the image of a fully bootable operting system.
05:41.06T0mWbut it is arch dependant
05:41.11samscolso the first 512 bytes will have a bootloader
05:41.17samscolin this case stage1 of grub
05:41.31T0mWembedded systems cannot use the same type of bootloading techniques as a stock PC computer
05:41.31samscolthen somewhere in that first 512 bytes, i think there will also be an MBR
05:41.48samscolgimme a chance to elaborate
05:41.54T0mWok
05:42.09samscolso far we have stage1 + partition table inside this file
05:42.09T0mWpartition table does actual drive boot
05:42.15T0mWright
05:42.27samscolthen we need filesystems on the partitions
05:42.35samscoli would assume, i could do something like
05:42.37T0mWthen it picks up the boot code from sector zero (logical zero)
05:42.56T0mWhowever, lilo + grub replace the partition table boot sector.
05:43.01T0mWIIRC
05:43.20T0mWyou don't get to the sector boot, IIRC
05:43.30samscolcat stage1[+partitiontable] partition1image partition2image > bootableimage
05:43.49samscolto make stage1+partitiontable is where i am having trouble
05:44.51T0mWyou're beyond me, I don't do pc boots, I use linux as a dev system for embedded systems and use either a custom written bootloader or blob
05:44.53samscoldoes the partition table not live inside the first 512 bytes or right after it?
05:45.07T0mWfirst 1K is partition table
05:45.16T0mWfirst 512 is used
05:45.23samscolfirst 1024? so the second 512 has the partition table?
05:45.45T0mWaround 0x1a0 is the table, bytes before and after it are the boot
05:45.48samscoland the first 512 is stage1
05:46.04samscol1a0 is 416 bytes
05:46.44T0mWthe partition table actually is part of the first physical track, track zero is for the exclusive use of the parition boot: part of that contains a table of partitions
05:47.41T0mWthe actual table starts at 0x1be
05:47.51T0mWthat is the first entry
05:48.11T0mWthe table has no formal length
05:48.24T0mWas you can do extended partition table entries
05:48.51T0mWoriginally, the table was fixed at four entries
05:49.00T0mWfour partitions
05:49.50samscola quick question on a side note
05:49.51T0mWsamscol: try this: dd if=/dev/hda of=partition.img bs=1k count=1
05:50.10T0mWthat will get you the first 4k bytes of track zero of the drive into a file
05:50.16samscoli've done dd if=/dev/zero of=512img bs=512 count=1024*1024
05:50.26T0mWthen use midnight commander (mc) to examine the data
05:50.30samscolso i've got myself a 512MB empty image
05:50.34T0mWnot /dev/zero
05:50.41samscoland now i want to copy stage1 onto this 512img
05:50.48T0mW/dev/zero only fills your file with 0's
05:51.16samscolcan i dd if=stage1 of=512img bs=512 count=1 without truncating back down to 512
05:51.32T0mWsamrobb: then use cat, cat will do binary data
05:51.38samscolwhat i want to do is take stage 1 and put it on the first sector of the file 512img
05:51.58samscolso if do cat stage1 > 512img? won't that truncate?
05:52.12samscolor are you saying do cat stage1 512img-512 > 512img ?
05:52.18T0mWright
05:52.45samscolso dd can't do a write to an offset and range inside a file?
05:52.53T0mWno it cannot offset
05:53.06samscolwhat does dd seek=? do then?
05:53.06T0mWwait a minute
05:53.28samscolhmm, there's also a notrunc to not truncate
05:53.34samscolthat's what i want
05:53.39T0mWseek=BLOCKS
05:53.48samscolyeah, i'm good at reading :-)
05:53.56T0mWBLOCKS == bs (block size)
05:54.40T0mWyou can prolly do it with dd, dd is pretty powerful
05:55.05T0mWjust that I've never bothered messing with the boot partition stuff
05:55.30samscollet's do it together then. i think i'm almost there
05:55.43T0mWnaw, I've got perl hashes to write
05:55.45T0mW:-(
05:55.56samscolok. i'll give it a try and let you know how i do
05:58.20samscolok. i've used fdisk to put a partition table on there
05:58.33samscolnow i need to figure out what offset i should dd the filesystem to
06:05.44MonMothawell, it would appear that all my modules loaded right up
06:07.09MonMothaactually, it would help if I told it to use the new kernel...
06:07.15T0mWheh
06:07.35T0mWI got to get the perl book and study the hashes of hashes chapter
06:08.40MonMothaT0mW: I did get 2.4.32 to compile without incident
06:09.33T0mWMonMotha: any secret as to how to do it
06:09.45T0mWMonMotha: what were you doing wrong?
06:10.14T0mWah
06:10.18MonMothaT0mW: dunno, 2.4.32 did fine
06:10.20T0mWI see my problem
06:10.25MonMothamust have been patched to handle the new binutils
06:10.39MonMotha2.4.27 is fairly old at this point (almost 2 years)
06:11.02T0mWMonMotha: that could be it
06:11.15T0mWMonMotha: later 2.4 is "fixed" to work
06:11.30MonMothayup
06:11.34T0mWMonMotha: throw Alan a pat on the back
06:12.53samscolif i fdisked an image, is there a way to figure out at what offset a particular partition should be based on the fdisk output?
06:13.46MonMothayeah, he does a pretty good job maintaining 2.4
06:18.57T0mWah, hash of hashes now works
06:18.59T0mWgood
06:28.52*** part/#elinux joe_bleau (n=joe_blea@adsl-65-70-230-240.dsl.fyvlar.swbell.net)
06:54.08MonMothagah, stupid binary only modules
06:54.32samscolMonMotha: nvidia?
06:54.54MonMothaoh hell no, they at least have a working wrapper
06:55.39MonMothaknow much about getting modules to load when they don't want to?  I've got some unresolved symbols that clearly could be matched up.  They have this random stuff on the end of them; the symbol clearly exists and the API shouldn't have changed
06:59.10MonMothasome of them do match
06:59.18MonMothaI'm not sure what the extra stuff is
06:59.30kergoththats module symbol versioning
06:59.32kergothits a crc
06:59.59kergothprobably have a kernel without modversions and modules with, or vice versa, or just building modules against kernel headers that dont match the running kernel in general
07:00.01MonMothakergoth: ok, if it doesn't match does that mean the structure changed and is incompatible?
07:00.22kergothgenerally, yeah
07:00.26MonMothahum...ok
07:00.35MonMothaI guess I get to hack up 2.4.27 until I can build it again then
07:00.36kergoththe idea is to avoid loading a module into an incompatible kernel, whether its the same kernel version or not
07:01.10MonMothaI have a binary module that I need to get loaded
07:01.39kergothwell, it isnt entirely binary.  the wrapper isnt, and that mustve been built against the wrong kernel sources
07:02.16MonMothakergoth: all I have is the resultant binary.  I'm guessing there was no wrapper at all, just someone who built the module against whatever they had laying around
07:02.37kergothick
07:02.42kergothgood luck ;)
07:02.46MonMothaI now need to get it to load into a kernel that I can actually do something with (no source to their kernel, of course)
07:03.07MonMothait appears to be stock 2.4.27, but I can't get 2.4.27 to build on my modern system
07:03.11MonMothayeah, that's kinda what I thought too
07:03.30MonMothaall this for a freakin' copy protection dongle.  May be easier to just hack out the copy protection :)
07:03.53kergothcould do something creepy like making a wrapper module yourself to translate their modversioned symbols into the ones your kernel has
07:03.59kergothand cross your fingers and pray
07:04.01kergoth:P
07:04.29MonMothadoes it having that CRC on there mean it was built with modversions on?
07:04.37kergothyeah
07:04.47kergothcheck /proc/ksyms to see if your running kernel does
07:05.03MonMothawell, I'm rebuilding the kernel, th ough I guess I can boot theirs
07:05.56MonMothathe last one there is just sad
07:06.05kergothno doubt.  that was a redhat.
07:06.09kergothforget which version
07:06.32MonMothasounds like something redhat would do
07:07.17kergoth"okay sir.,. we'll need to install the kernel sources, copy the config into place, and do a make oldconfig; make dep, then rebuild our drive" .. "what?"
07:07.19MonMothaI guess I can try to patch 2.4.27 up to where I can build it.  I was only h aving one real problem
07:07.40samscolanyone know what it means when grub boots only as far as saying GRUB
07:07.48MonMothait means it's messed up :)
07:08.01samscolit's not listed in the grub error messages page
07:08.08kergothi think that means it couldnt load stage2 or something
07:08.18kergothif it shows grub, obviously stage1 loaded, at least partly..
07:08.29samscolyup
07:08.40samscoli've got a stage 1.5 as well
07:08.46kergothfun
07:08.54samscoldoes stage1 read the partition table?
07:09.03MonMothakergoth: you know a bit about GPL issues, right?
07:09.35kergotha bit.  not as much as tim, but.. :)
07:20.52*** join/#elinux T` (n=total@c-24-6-52-222.hsd1.ca.comcast.net)
07:29.05MonMothaso great, I'm patching stuff that is waaaay above my head, deep within the bowels of the linux process switching code
07:29.07MonMothafun stuff
07:34.14*** join/#elinux darkschneider (n=gab@213-140-6-96.ip.fastwebnet.it)
08:30.49*** join/#elinux gregiadf (n=gregiadf@202.75.186.154)
09:00.32*** join/#elinux noclouds (n=mhfan@60.166.48.88)
09:28.13*** join/#elinux ade|desk (n=adavey@194.200.143.249)
10:01.13*** join/#elinux CosmicPenguin (n=nobody@aus-ext-proxy01.amd.com)
10:01.13*** mode/#elinux [+v CosmicPenguin] by ChanServ
10:09.13*** join/#elinux samscol (n=donkey@203.121.86.58)
10:37.33*** join/#elinux noclouds (n=mhfan@60.166.48.88)
11:31.43*** join/#elinux T`2_ (n=total@tor/session/x-91d9540ab6495971)
13:01.09*** join/#elinux prpplague (n=billybob@216.110.99.43)
13:23.43*** join/#elinux Crofton__|laptop (n=balister@66-207-66-26.black.dmt.ntelos.net)
13:47.44*** join/#elinux prpplague (n=billybob@72.22.129.90)
14:14.15*** join/#elinux GPSFan (n=Ken@65.121.49.208)
14:26.13*** join/#elinux sjhill (n=sjhill@eth13.com-link.com)
14:35.08*** join/#elinux prpplague (n=billybob@72.22.129.90)
14:54.55CosmicPenguinMorning
15:01.30*** join/#elinux bzzbzz (n=franco@HSE-Montreal-ppp3464509.sympatico.ca)
15:30.08*** join/#elinux vrm (n=vrm@249.55.101-84.rev.gaoland.net)
15:44.11*** join/#elinux thraxisp (n=thraxisp@ottgate.precidia.com)
15:49.55*** join/#elinux eggers (n=eggers@pixpat.austin.ibm.com)
16:15.36*** join/#elinux ade|desk (n=adavey@194.200.143.249)
16:28.05*** join/#elinux Crofton__|laptop (n=balister@hc6521c88.dhcp.vt.edu)
16:48.37*** join/#elinux hufnus (n=slonsiki@DSL135-071.LABridge.com)
16:51.32*** join/#elinux Genesis (n=genesis@dev.aimao.org)
16:54.03*** join/#elinux cb_hm (n=cbrake2@cpe-24-166-121-47.neo.res.rr.com)
17:00.20*** join/#elinux Genesis (n=genesis@dev.aimao.org)
17:00.46*** join/#elinux markl__ (i=mark@dsl093-225-127.slc1.dsl.speakeasy.net)
17:01.39Genesisbonsoir
17:04.31*** join/#elinux prpplague (n=billybob@72.22.129.90)
17:04.59prpplagueCrofton_ you get your fpu stuff squared away?
17:11.13Crofton__|laptopWell, I really need to do some tests to see how bad FP performance is on soft float CPUS
17:11.51Crofton__|laptopI have some code that will do some FP calculations and need to pay attention to make sure they don't swamp the machine
17:12.05prpplagueahh ok
17:12.31prpplagueCrofton__|laptop: not like you probably don't have a million other things to do, but...........
17:12.33Crofton__|laptopof course, I have so much other stsuff to do first :)
17:12.38Crofton__|laptopheh
17:12.53prpplagueCrofton__|laptop: if you get time after you do the tests, why not write a page on the elinux wiki
17:13.15Crofton__|laptopI should
17:13.44Crofton__|laptopBasically I should so some simple math tests so I can say how much slower FP math is versus int
17:13.49Crofton__|laptopat least for simple cases
17:13.59Crofton__|laptopIt's not like I am talking logs are anything
17:14.03Crofton__|laptoper taking
17:14.38prpplagueCrofton__|laptop: well you can always just start out with some basic info
17:14.45Crofton__|laptopyeah
17:15.14prpplagueCosmicPenguin: hope your day is going better than mine
17:15.19Crofton__|laptopI have a lib that converts a frequency in FP to sum integer that sets up a NCO
17:15.28Crofton__|laptopnumerically controlled oscillator
17:15.32prpplagueah
17:15.36prpplague~fridays
17:15.43prpplague~friday
17:15.44ibotFriday sucks, because it should be the weekend, but you're stuck working anyway.
17:17.46prpplagueCosmicPenguin: chandler has gotten introuble at school every day this week
17:21.46*** join/#elinux T`2 (n=total@pdpc/supporter/student/T)
17:22.08prpplaguekergoth: hey old chap
17:22.22prpplaguekergoth: whats cookin these days in your part of the world?
17:35.09CosmicPenguinprpplague: acting out?
17:35.44prpplagueCosmicPenguin: yea, talking, pestering the girls, miss behaving
17:35.58prpplagueCosmicPenguin: enough to make me drink extra tonight, hehe
17:36.05CosmicPenguinSo basically he's being a kid
17:36.38prpplagueCosmicPenguin: hmm, well not sure other kids get in trouble as much as he does
17:36.56prpplagueCosmicPenguin: he's about to give me an ucler
17:37.16CosmicPenguinKids will do that to you - I'll be we were all terrors when we were young
17:37.17CosmicPenguinI know I was
17:38.04Crofton_http://www.cnn.com/2005/EDUCATION/11/16/girl.on.corner.ap/index.html
17:39.18prpplaguehehe
17:39.32Crofton__|laptopI'm not sure it is a great idea
17:39.47Crofton__|laptopbut sometimes you just need to get their attention
17:40.02prpplagueyea we've been having problems with chandler doing his homework as well
17:40.17Crofton_I never did mine :)
17:40.22prpplagueof course, i'm not a big homework supporter anyway
17:40.23Crofton_Hated high school
17:40.29prpplaguehe me either
17:40.34Crofton_I could get away with it though
17:40.41Crofton_cost me once I hit college though
17:42.18Crofton_I made an A in the hardest math class I took by doing my homework
17:42.31Crofton_but it didn't stick :)
17:42.50prpplaguehehe
17:43.53CosmicPenguinstupid remote X
18:10.45CosmicPenguinwow - a busy day
18:13.33prpplagueCosmicPenguin: you can say that again
18:15.50CosmicPenguinwow - a busy day
18:15.53CosmicPenguinhehe
18:21.18prpplaguehehe
18:21.44prpplagueCosmicPenguin: lordy lordy
18:37.19CosmicPenguinWow - stuff is actually working, and working well
18:37.24CosmicPenguinwho knew?
18:40.11*** join/#elinux Genesis (n=genesis@dev.aimao.org)
18:41.30Genesisbonsoir
19:22.50*** join/#elinux zinosat (n=Davide@adsl-ull-93-128.44-151.net24.it)
20:16.54*** join/#elinux GPSFan (n=Ken@65.121.49.208) [NETSPLIT VICTIM]
20:28.18*** join/#elinux T`2 (n=total@pdpc/supporter/student/T)
21:13.41*** join/#elinux file[laptop] (n=jcolp@mctnnbsa24w-142167049176.nb.aliant.net)
21:20.48file[laptop]kurg blah
21:27.23CosmicPenguinyeah
21:27.25CosmicPenguinyou tell em
21:27.28CosmicPenguinbastards!
21:27.32file[laptop]death!
21:30.00*** part/#elinux Khem (n=Khem@gateway-1237.mvista.com)
21:38.42*** join/#elinux bzzbzz (n=franco@HSE-Montreal-ppp3464509.sympatico.ca)
21:51.14CosmicPenguinhmmm
22:22.35*** join/#elinux Crofton_|laptop (n=balister@66-207-66-26.black.dmt.ntelos.net)
23:35.11T0mWboard #1 assembled
23:35.26T0mWbreaktime before board #2!
23:36.55*** join/#elinux joe_bleau (n=joe_blea@adsl-69-151-117-209.dsl.fyvlar.swbell.net)

Generated by irclog2html.pl by Jeff Waugh - find it at freshmeat.net! Modified by Tim Riker to work with blootbot logs, split per channel, etc.