irclog2html for blob on 2002.04.15

07:23:29mmattenmorning
09:28:26seletzmorning
14:41:38KuEdEhi guys
14:42:41KuEdEPerhaps anyone knows about the source of the "Mounting Rootfs panic"..... could it be a bad toolchain ?
14:55:45KuEdEPerhaps anyone knows about the source of the "Mounting Rootfs panic"..... could it be a bad toolchain ?
14:59:06prpplagueKuEdE: what is the root device your are trying to use?
15:01:17KuEdEits a ramdisk  root=/dev/ram
15:01:53KuEdEre :-))
15:02:00KuEdE<KuEdE> its a ramdisk  root=/dev/ram
15:02:13KuEdE01:00
15:02:19prpplaguedid you make sure to do a rdev on your kernel image?
15:02:48KuEdE:-D   .. oups .. whats rdev ?
15:05:19prpplagueKuEdE: when you build a kernel, the root device on the system you build is used as the root device, if you intended on using another device you need to use the rdev command on the image to change the root dev
15:05:39prpplagueyou should be able to do a man on rdev and get some additional information
15:05:45KuEdEthx.... found out @ same moment :-))
15:05:55KuEdEill do ..
15:06:03prpplagueKuEdE: i'm not saying this is the problem, but its one thing to check
15:07:09KuEdEyeah.. its another point i can look after before i get another toolchain...
15:07:37KuEdE... ill do that..have a smoke and tell ya if it works
15:07:47KuEdEbut anyways... thx !!
16:00:59KuEdErdev'ing is like (in my case)  setting CONFIG_CMDLINE="/dev/ram"..... i tried it but without success...  perhaps ill get the lart toolchain instead using the suse ones
16:01:44KuEdEi mean CONFIG_CMDLINE="root=/dev/ram"
16:03:01KuEdE4(0prpplague4) 0 i hope im right.... or are there another things which i got to manipulate with rdev ?
16:20:02RussKuEdE: don't use colors excessively
16:23:53seletzhi
16:35:12prpplaguehey seletz
16:35:19prpplaguedoh
16:36:26prpplaguei wonder is seletz will return
16:51:55prpplagueseletz: you made hotel reservations for ottawa yet?
16:53:39seletzprpplague: yes, i got the confirmation today
16:53:47prpplagueseletz: which hotel?
16:54:01KuEdEprpplague: got another idea ?       Is there a possibility that my toolchain is bad (i use the SuSE one )   .......
16:54:07seletzprpplague: well, the one recomended on the ols site
16:54:13prpplagueseletz: ohh
16:54:25seletzprpplague: hmm?
16:54:34prpplagueseletz: that one seem a little expensive
16:55:10prpplagueKuEdE: sorry, no suggestions off the top of my head
16:55:23seletzprpplague: well, these are suites, and they mean "+kitchen +sofa (read: 2nd bed!!) etc"
16:55:39seletzprpplague: so the price is for 2 persons.
16:55:39prpplagueKuEdE: my next guess would be check whether the device is compiled into the kernel
16:55:52prpplagueseletz: ya, its only me
16:56:27prpplagueseletz: btw, did you know there is a beer festival on the 21-23?
16:56:38seletzprpplague: i checked out some other nearby hotels, but they're comparable expensive.
16:56:47seletzprpplague: in ottawa?
16:56:51prpplagueseletz: ya
16:57:18seletzprpplague: doh, booked my flight already. 25 of june arrival.
16:57:51prpplagueseletz: ya i've already booked too, but i'm going to change mine, its actually cheaper too
16:58:05prpplagueseletz: gotta run, bbl
16:58:13seletzprpplague: see ya
16:58:13prpplagueKuEdE: good luck
16:58:24seletzKuEdE: what problem do you have?
16:58:33seletzjust curious
16:58:43KuEdEthere compiled in..... im sure....  ill try tomorrow ... BTW: i use the "lart-ramdisk-1.2-lmbench.gz".....
16:59:14KuEdEseletz:  my kernel doesnt like the ramdisk....
16:59:19seletzscratching behind his ear
16:59:22seletzwell,
16:59:32seletzsee ya guys...
16:59:38KuEdEUnable to mount root fs 01:00.....
16:59:44seletzah
16:59:58seletzuse root=/dev/ram0
17:00:08seletzbe sure to compile in initrd support
17:00:13KuEdEall done
17:00:18seletzhmm
17:00:34seletzhow big is your rd?
17:00:40seletzdecompressed i mean
17:00:50KuEdEits a new sa1110 hw port im doing... so i never can say that HW is 100% ok
17:01:05seletzhmm, the kernel starts up?
17:01:11KuEdEbut it worx well on WinshitCE
17:01:19seletzhmm
17:01:27KuEdEi mean... WinCE woks well ON this hardware ;-)
17:01:57seletzmaybe you want to take a look at my initial ramdisk, its quite minimal
17:02:01KuEdEmy .config is very similar to lart's one
17:02:08KuEdEyeah...
17:02:11seletzhttp://www.eletztrick.de
17:02:18KuEdE:-)
17:02:23seletzsearch for cramfs howto
17:02:42seletzthere should some links in the download section too
17:03:20KuEdEthx.... perhaps i forgot something whats  clear to a "senior" arm developer
17:03:20seletzyou have to be wary about the _decompressed_ ramdisk size.
17:03:35seletzKuEdE: is this a new port (i.e. new board/machine)
17:03:35KuEdE32megs ram i got
17:04:17KuEdEits a sa1110 based thing...
17:04:32seletzKuEdE: well, doesnt depend on how much ram you got. initial ramdisks are _fixed_ _size_ /COMPILE TIME)
17:05:22seletzKuEdE: well, the question is: Is it a new kernel port (i.e. a new machine number and so on) or are you "just" using an existing kernel
17:05:34KuEdEnew number
17:05:45KuEdEmade a port of blob too...
17:05:59KuEdEits still not registered at arm.uk
17:06:08seletzKuEdE: which hw? Name?
17:06:10KuEdEi took a high fantsy number
17:06:46KuEdEyeah.. i know... but first i want to bring it working before sending diffs
17:06:50seletzKuEdE: anyway: make sure that in your fixup function your ramdisk size is set up correctly
17:07:48seletzKuEdE: IIRC you must parse ATAG tags by hand and setup the ramdisk. Check out arch/arm/assabet.c, search for ....
17:07:51seletzlooking
17:08:27KuEdEin lart.c theres no fixup ... so why do I need a fixup for mine ?
17:08:44seletzKuEdE: setup_initrd()
17:08:54KuEdEsorry for bothering with dumbass questions...
17:08:55seletzKuEdE: lart is _minimal_
17:09:18seletzKuEdE: my custom board is minimal too, but i do need a fixup routine because:
17:09:28seletzKuEdE: - i need some specific interupts
17:09:39seletzKuEdE: - i have HW registers to set up correctly
17:09:48seletzKuEdE:  - configure the initial RD
17:09:53KuEdEi future ill need too...   setup external DAC for Contrast.. Backlite
17:09:53seletz:)
17:10:07seletzKuEdE: look at system3.c
17:10:12KuEdEok
17:10:39seletzKuEdE: thats my board, so expecht some bugs :)
17:10:45seletzs/expecht/expect/
17:10:56KuEdE:-O
17:11:04seletzKuEdE: i have some contr/bright/ setting via PWM in there
17:11:35seletzKuEdE: i use a CSTN display, works quite well (what can you expect from CSTN, tho)
17:12:45seletzKuEdE: my apporach was: look at assabet. Copy it. Cut ugly stuff out. goto 1. Look at neponset code. merge them.
17:12:54seletzKuEdE: got a SA1111 too?
17:12:57KuEdEnope
17:13:05seletzKuEdE: lucky ...
17:13:12KuEdEbut i got a usb host (isp1161)
17:13:21seletzKuEdE: SA1111 is crap IMHO
17:13:31KuEdEi heard in the mailinglists...
17:13:38seletzKuEdE: supported by linux already?
17:13:48KuEdEu mean 1161 ?
17:13:53seletzyup
17:14:14KuEdEtheres a V0.8 of this driver.. (not merged in the official kernel tree)   ...
17:14:30seletzwell, ok
17:14:34KuEdEi read ..it wuz developed on a PC ISA card
17:15:18KuEdEbut i heard from WinCE guys that the ISP1161 is very hard to bring to work
17:15:24KuEdEso...
17:15:37KuEdEsorry my friends... i gotta.. go..
17:15:43seletzKuEdE: one last thing: be sure to send diffs as soon as possible, Russell King is _very_ good in finding bugs and ugly (read: wrong, not linux-like, Not The Way To Do It) code.
17:15:54KuEdEk...
17:16:01seletzKuEdE: It helped me a _lot_
17:16:25seletzKuEdE: dont worry if the kernel boots not correctly
17:16:33KuEdEill check in tomorrow..... ill try these thing u told me..
17:16:37KuEdEBut 1 question:
17:16:41KuEdElast one
17:16:55seletzKuEdE: just read the docs at armlinux website at last thrice :)
17:17:02seletzKuEdE: specially the patches section
17:17:03KuEdEu say lart is minimal.. but how to run linux when initrd fails.. ?
17:17:27KuEdEim a little bit confused..
17:19:02seletzKuEdE: I dont know exactly wether or not you have to call the setup_initrd() directly or wether or not its setup inirectly by having correct kernel parameters (and BLOB set up your ATAG tags).
17:19:25KuEdEahh..
17:19:26seletzKuEdE: i have 3 lines in my fixup function and all is set :)
17:19:57seletzKuEdE: one thing: Are you sure that your kernel expects the ramdisk at the right physical ram address?
17:20:15KuEdEhmm.. another good question...
17:20:17seletzlook at this:
17:20:31|static void __init fixup_system3(struct machine_desc *desc,
17:20:31|struct param_struct *params, char **cmdline, struct meminfo *mi)
17:20:31|{
17:20:31|DPRINTK( "%s\n", "START" );
17:20:32|ROOT_DEV = MKDEV(RAMDISK_MAJOR,0);
17:20:32|setup_ramdisk( 1, 0, 0, 8192 );
17:20:36|setup_initrd( 0xc0800000, 8*1024*1024 );
17:20:41seletz}
17:21:28seletzthe setup_initrd() call has the phys. ram addr of the initrd as a parameter
17:21:45KuEdEyep... ive seen this in your system3.c  and in simpad.c
17:21:53seletzKuEdE: the fixup function runs when no VM is running IIRC
17:22:07KuEdEbut now... :-(((  sorry... i gotta go... my bud's waiting
17:22:18KuEdEk
17:22:22KuEdEill save this s
17:22:22seletzKuEdE: the assabet.c does parse the ATAG
17:22:33seletzok, i'll leave too
17:22:35seletzsee ya
17:22:47seletz(join #kernelnewbies!!)
17:26:52KuEdEill save this chat log......seems very informative ;-)))
17:27:03KuEdEcu guys  and thx again
19:26:02Russseletz_home: /dev/ram is the wrong root for a ramdisk iirc
19:26:06Russthat would be a ramfs
19:26:49Russnm
19:26:53RussNotes: You may have "/dev/ram" or "/dev/ramdisk" or both.
22:47:11prpplaguehowdy all
22:49:43prpplagueBZFlag: seen my order for tux+jtag yet?
22:51:50Russ|werkprpplague: my board house screwed up on substitutions and forgot a part
22:51:52Russ|werkgrrrr
22:52:40prpplagueargh
22:52:50prpplaguei hate it when that happenes
22:54:26prpplagueRuss|werk: on my tcsx-1 board, i had laid out a set of jumpers to run the processor at 50, 66 and 75mhz, but the board house couldn't figure out what two resistors were for, so they droped them, so now it will only run at 75mhz
22:55:15prpplagueRuss|werk: what kind of board is it?
23:17:34Russ|werksa1100
23:17:55Russ|werkI caught them before they placed all the zero ohms
23:18:10Russ|werkeven though it clearly explained them in the bom
23:39:29prpplaguegood that you caught them
23:48:17Russ|werkthey also could not find the transpower jacks I asked for, but said they found some sutible replacements
23:48:45Russ|werkI repeatadly checked with them that they are exact matches, including the turn ratio on the built in filters
23:49:12Russ|werkand on the board I now have straight through ethernet jacks with no filters

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