[00:01] <tziOm> even tho CONFIG_BLK_DEV_PDC202XX_OLD=y is set in .config, I see no reference whatsoever in dmesg/logs about PDC..
[00:02] <tziOm> what I get is:
[00:02] <tziOm> Uniform Multi-Platform E-IDE driver Revision: 7.00beta4-2.4
[00:02] <tziOm> ide: Assuming 33MHz system bus speed for PIO modes; override with idebus=xx
[00:02] <tziOm> .. nothing more
[00:05] <crazy_imp> thats the same as i get
[00:06] <tziOm> no PCI: enabling device...
[00:06] <crazy_imp> i also tried the pdc202xx_new, same
[00:07] <tziOm> must they be modules? ... no...
[00:08] <tziOm> anyone?
[00:08] <crazy_imp> also build build in the kernel, same, nothing is other
[00:08] <tziOm> you dont get any messages when you insmod it?
[00:08] <crazy_imp> tziOm: do you have rc5 on it atm? what does lspci says?
[00:09] <crazy_imp> i tells me: no such device if i try to insert it with isnmod
[00:09] <tziOm> no, I have kami
[00:10] <crazy_imp> ok, then i will put rc5 on it myself and take a look what lspci says
[00:12] <crazy_imp> tziOm: do you have any ideas what the 19 solderpads are, at the end of the board, behind the ideconnector?
[00:13] <tziOm> nope, havent checked...
[00:13] <tziOm> some io prolly
[00:14] <crazy_imp> yes, but for what :D
[00:14] <h3sp4wn> what about the modem that those boards have ?
[00:16] <crazy_imp> i'm a little bit confused about the sentry5 chipset, lspci tells me that there are 3 ethernet controllers on it, but it only has lan and wlan, so what is the third for?
[00:18] <florian__> nbd, {Nico} how do we proceed for packages ?
[00:19] <nbd> port all of them to buildroot-ng and make sure that they're structured and categorized
[00:19] <crazy_imp> looks like i have to read many datasheets to understand the device :D
[00:19] <nbd> freeze trunk at the moment
[00:19] <florian__> ok
[00:20] <florian__> alphabetical order ? most useful order ?
[00:20] <tziOm> nbd, do you know anything about the wlhdd and the CONFIG_BLK_DEV_PDC202XX_OLD that you would like to share?
[00:21] <nbd> florian__: maybe it would make sense to make a packages/old directory, svn cp in all the stuff from trunk and then slowly move and convert it over in the order of relevance
[00:21] <nbd> florian__: that has the additional sideffect of tracking the renames
[00:21] <nbd> tziOm: not at the moment. but i will test it soon, because i need it for a project
[00:22] <florian__> nbd: ok
[00:22] <tziOm> read my above.. even tho compiled into kernel... it seems not there...
[00:23] <tziOm> does it need PCMCIA? no...
[00:23] <tziOm> nbd, and what about the "PCI: Disabled\nPCI: Fixing up bus 0" message? thats correct?
[00:25] <nbd> might be wrong, no idea
[00:25] <tziOm> root@OpenWrt:/# cat /proc/ide/drivers
[00:25] <tziOm> ide-disk version 1.17
[00:25] <tziOm> ide-default version 0.9.newide
[00:25] <nbd> i'll have a short glance at the system code...
[00:26] <florian__> nbd: did you get your hand on your presentation ?
[00:27] <nbd> not yet
[00:27] <florian__> ok, the presentation has been accepted, so now I will gather some informations here ;)
[00:30] <CIA-4> nbd * r4003 / (2 files in 2 dirs): fix typo in brcm-2.4 system code
[00:30] <nbd> tziOm: try svn update; make target/linux-clean world
[00:32] <tziOm> nbd, target/linux/brcm-2.4/patches/001-bcm47xx.patch
[00:32] <nbd> yes
[00:32] <nbd> that's where i fixed the typo
[00:32] <tziOm> wth was it?
[00:33] <nbd> a wrong device id
[00:33] <nbd> was 0x4710, while it should have been 0x4310
[00:33] <tziOm> shit..
[00:33] <tziOm> willl it do anything with the wl module too maby then, sucker! ;)
[00:34] <nbd> sucker?
[00:37] <tziOm> any command to make ping exit true when it has made X successful pings (and gotten replies?)
[00:38] <nbd> *sigh*
[00:39] <nbd> there are two parts to it
[00:39] <nbd> part 1: type 'man ping'
[00:39] <nbd> part 2: understand the text
[00:39] <nbd> no, wait
[00:39] <nbd> 3 parts
[00:39] <nbd> part 3: actually do it
[00:40] <tziOm> ping -c1 192.168.1.6 >/dev/null 2>&1|| echo loop
[00:42] Action: nbd goes to sleep
[00:42] <tziOm> nbd, works like a charm
[00:42] <tziOm> thanks duud
[00:43] <nbd> ide as well?
[00:43] <tziOm> atleast the pdc20265 seems
[00:43] <tziOm> dunno if it found any disk tho.
[00:44] <tziOm> seems not..
[00:45] <tziOm> Uniform Multi-Platform E-IDE driver Revision: 7.00beta4-2.4
[00:45] <tziOm> ide: Assuming 33MHz system bus speed for PIO modes; override with idebus=xx
[00:45] <tziOm> PDC20265: IDE controller at PCI slot 01:01.0
[00:45] <tziOm> PCI: Enabling device 01:01.0 (0004 -> 0007)
[00:45] <tziOm> PDC20265: chipset revision 2
[00:45] <tziOm> PDC20265: not 100% native mode: will probe irqs later
[00:45] <tziOm> PDC20265: ROM enabled at 0x000d0000
[00:45] <tziOm> PDC20265: (U)DMA Burst Bit DISABLED Primary PCI Mode Secondary PCI Mode.
[00:45] <tziOm> PDC20265: FORCING BURST BIT 0x00->0x01 ACTIVE
[00:45] <tziOm> ide0: BM-DMA at 0x0140-0x0147, BIOS settings: hda:pio, hdb:pio
[00:45] <tziOm> ide1: BM-DMA at 0x0148-0x014f, BIOS settings: hdc:DMA, hdd:pio
[00:45] <crazy_imp> so it works now?
[00:46] <tziOm> that works so far.. but it does not find my connected disk seems..
[00:46] <tziOm> no devs in /dev or /ide
[00:46] <tziOm> devide
[00:46] <crazy_imp> kmod-ide installed and modules loaded?
[00:46] <nbd> maybe you forgot something in your kernel config
[00:46] <nbd> crazy_imp: he refuses to use modules
[00:46] <tziOm> dont think so, no
[00:47] <tziOm> CONFIG_BLK_DEV_IDEDISK=y
[00:50] <tziOm> root@OpenWrt:/# cat /proc/ide/drivers
[00:50] <tziOm> ide-disk version 1.17
[00:50] <tziOm> ide-default version 0.9.newide
[00:51] <tziOm> any clues?
[00:51] Action: nbd is gone: sleep
[00:53] <wbx> iwlist wl0 scanning
[00:53] <wbx> Unable to handle kernel paging request at virtual address c0170000, epc == c016beb0, ra == c016bf98
[00:53] <wbx> hmm. known bug?
[00:54] <nbd> try grabbing the latest wl from linksys
[00:54] <nbd> and see if it causes the same problem
[00:54] <nbd> wlcompat hasn't been tested much with the new driver
[01:00] <h3sp4wn> Is there a newer nas than the one in white russian ?
[01:06] <h3sp4wn> I was trying to extract the nas from the latest wl500-gd firmware (using hexdump) but if its pointless I will stop trying
[02:18] <CIA-4> nico * r4004 /branches/buildroot-ng/openwrt/package/wlcompat/ (Makefile src/Makefile src/wlcompat.c wlcompat.c): move source file to src/ and let the kernel tree build the modules.
[03:20] <crodler> how can i write on a loop-mounted cramfs image?
[04:05] <CIA-4> nico * r4005 /branches/buildroot-ng/openwrt/package/switch/src/ (switch-adm.c switch-core.c switch-robo.c): fix module params on 2.6.17, suppress warnings.
[13:17] <CIA-4> kaloz * r4006 /branches/buildroot-ng/openwrt/target/linux/aruba-2.6/patches/ (000-aruba.patch 010-ar2313_enet.patch): fix the Aruba ethernet drivers for 2.6.17
[14:32] <unverbraucht> i'm trying to find the uhci problems on kamikaze (brcm) 2.6
[14:33] <unverbraucht> can anybody give me a pointer where to start?
[14:33] <unverbraucht> i'd first diff the uhci against a vanilla 2.6.17 uhci_hcd.c
[14:33] <unverbraucht> and hope that anything shows up there :)
[14:46] <unverbraucht> nbd: michael sent out a patch up-stream with your bcm43xx hostap changes
[14:46] <unverbraucht> https://lists.berlios.de/pipermail/bcm43xx-dev/2006-June/001972.html
[14:46] <unverbraucht> :)
[14:49] <Kaloz> unverbraucht: they are NDL's changes :p
[14:49] <unverbraucht> oh, sorry about that :)
[14:49] <unverbraucht> just remembered the nick started with an n... :/
[15:52] <crazy_imp> nbd: the ide bug for the wl-hdd isn't solved yet, or?
[16:01] <nbd> crazy_imp: what bug exactly?
[16:01] <crazy_imp> this one: insmod pdc202xx_new.o
[16:01] <crazy_imp> insmod: init_module: pdc202xx_new: No such device
[16:03] <nbd> you need to use _old
[16:04] <crazy_imp> but with the _old i had the same problem, btw. where is the _new? i didn't kicked it out of the config *confused*
[16:04] <nbd> is this kamikaze or whiterussian?
[16:07] <crazy_imp> nbd: kamikaze (r4006) with 2.4 kernel, the 2.6 kernel drops the wl-hdd in an reboot loop :/
[16:08] <nbd> the bug in kamikaze is solved
[16:08] <nbd> do svn update; make target/linux-clean world
[16:10] <crazy_imp> ok, i hope it solves the bug :)
[16:11] <nbd> crazy_imp: it did solve the bug for tziOm
[16:11] <nbd> it was a simple 1-character-typo in a device id
[16:11] <crazy_imp> btw, the pdc20265r can control a raid :D
[16:12] <crazy_imp> ok, lets see if i can confirm it
[16:12] <crazy_imp> where is the difference between the wl-hdd, and the wl-hdd2.5 only the cpu?
[16:13] <nbd> i didn't know those two were different at al
[16:13] <crazy_imp> the toh list it as different, the wl-hdd is a little bit faster
[16:19] <crazy_imp> lspci gives me more 3 more devices as the last build before :D
[16:39] <tziOm> nbd, any news on the ide issue?
[16:41] <crazy_imp> :D
[16:41] <tziOm> crazy_imp, hey
[16:41] <crazy_imp> wait some minutes and i can say yes or no :D
[16:42] <tziOm> i say no
[16:42] <tziOm> no updates to svn atleast
[16:42] <crazy_imp> i wish i can say yes if it's ready
[16:43] <crazy_imp> the r4006 is a new try for me now, forgot to clean it up before building a new one
[16:43] <crazy_imp> tzi0m: do you have a wl-hdd or a wl-hdd2.5?
[16:44] <tziOm> its a 2.5" hdd in it...
[16:44] <tziOm> didnt know it existed 3.5 versions..
[16:44] <crazy_imp> how many bogomips?
[16:44] <nbd> 200
[16:44] <tziOm> BogoMIPS : 82.94
[16:44] <nbd> hmm.. you got the slow one
[16:44] <crazy_imp> tzi0m: thats the same as my wl-hdd2.5 has
[16:45] <nbd> mine's also using 2.5"
[16:45] <nbd> but it has 200 mhz
[16:45] <crazy_imp> no
[16:45] <tziOm> yeah? cool.
[16:45] <crazy_imp> 125mhz cpu
[16:45] <tziOm> could one clock it?
[16:45] <nbd> what do you mean no?
[16:45] <nbd> mine does have 200 mhz
[16:45] <crazy_imp> CPU revision is: 00024000
[16:45] <crazy_imp> CPU: BCM4710 rev 0 at 125 MHz
[16:46] <nbd> i think there are several revs of the same product
[16:46] <nbd> and i got the faster version
[16:47] <tziOm> nbd, have you tested the ide controller on it? I seem to have luck with getting the PDC20265 working now, but no luck with detecting the ide drive.. actually no info about ide devs atall...
[16:47] <nbd> how much did you change the config/
[16:47] <nbd> ?
[16:47] <tziOm> just this:
[16:47] <nbd> :)
[16:47] <tziOm> PDC20265: FORCING BURST BIT 0x00->0x01 ACTIVE
[16:47] <tziOm> ide0: BM-DMA at 0x0140-0x0147, BIOS settings: hda:pio, hdb:pio
[16:47] <tziOm> ide1: BM-DMA at 0x0148-0x014f, BIOS settings: hdc:DMA, hdd:pio
[16:48] <tziOm> nbd, not much changes in the ide part.. just in kernel no modules.
[16:48] <tziOm> nbd, does it work at your place?
[16:48] <nbd> i haven't tested yet
[16:48] <nbd> but last time i checked, it worked
[16:48] <nbd> on whiterussian
[16:48] <crazy_imp> nbd: my rev is r1.15, i think i will create a wl-hdd2.5 page in the wiki if i have time :)
[16:54] <crazy_imp> time to flash :D
[16:56] <tziOm> nbd, are you working on getting 2.6.17 in?
[16:56] <nbd> 2.6.17 is already in
[16:56] <nbd> but it won't work on the wl-hdd
[16:56] <tziOm> nbd, do you know how the bcm43xx drivers behave?
[16:56] <nbd> no
[16:57] <nbd> before i integrate it, i need to rewrite parts of the system code
[16:57] <tziOm> no..? maby I should check it out.. how can I choose it? Do I need make DEVELOPER=1 or something?
[16:57] <nbd> no
[16:57] <nbd> you can't use it yet
[16:57] <tziOm> nbd, lots of work to be done?
[16:57] <nbd> not that much
[16:58] <nbd> but still: you can't use it yet
[16:58] <nbd> you have to wait
[16:58] <nbd> and don't ask me about it every day
[16:58] <nbd> because at some point, i'll simply start ignoring you completely
[16:58] <tziOm> hmm...
[16:58] <tziOm> what version is used as the brcm-2.6 now then?
[16:59] <nbd> that's it. you've reached the quota of answers from me for now
[16:59] <tziOm> nbd, you have some major issues Im not educated to deal with
[16:59] <nbd> no, you seem to need some lessons in being less annoying
[17:00] <tziOm> I find you a bad team-player.. we are on the same side you know.. doing similar stuff mol
[17:00] <tziOm> .
[17:00] <nbd> i hate it when people throw lots of questions at me without even *trying* to figure things out themselves
[17:00] <nbd> and i told you that before
[17:00] <nbd> it doesn't have anything to do with being a bad team player
[17:00] <nbd> you're simply wasting my time
[17:01] <crazy_imp> nbd: it's still building the pdc202xx_new
[17:02] <nbd> checking...
[17:02] <nbd> in the default config only _OLD is marked as m
[17:04] <crazy_imp> hm, i can't remember that i have changed it
[17:05] <nbd> go into target/linux/brcm-2.4 and run svn status
[17:06] <crazy_imp> ok, wait
[17:07] <crazy_imp> drops me: M config (so its my fault?)
[17:08] <nbd> looks like it
[17:08] <nbd> do svn revert config
[17:09] <crazy_imp> and now the same procedure like before (svn update; make target/linux-clean world)?
[17:09] <nbd> no need to do svn update
[17:09] <crazy_imp> (without the update)
[17:09] <nbd> yeah
[17:09] <crazy_imp> ok
[17:09] <tziOm> or svn diff
[17:16] <crazy_imp> if i look at the board of my wl-hdd, the flashchip has 8 lines (thats the amount i can see, maybe there are more, because there are 19 pads) to the solderpads i'm thinking about the hole time, is it jtag? some brings a voltage of 3,3v one comes along with 5,2v (its the same as the power for the hdd) and one with 0,25v and some doesn't shows voltage against the ground
[17:22] <nbd> crazy_imp: maybe it's also the same bus that you can use on the wl-500g to connect an uart for serial access
[17:23] <crazy_imp> ok, will take a look
[17:48] <tziOm> crazy_imp, might be a modem in there somewhere..
[17:48] <tziOm> 00:03.0 Modem: Broadcom Corporation BCM47xx V.92 56k modem
[17:48] <crazy_imp> i know
[17:49] <nbd> tziOm: i think that's just a misclassification in the system code
[17:50] <crazy_imp> thats a point to fetch all the datasheets about the chips and look for a modem :D
[17:50] <tziOm> s/ misclassification in the/nother typo in my/
[17:50] <tziOm> ;)
[17:50] <nbd> no
[17:51] <nbd> this is something that broadcom had left in their code
[17:51] <nbd> so not exactly my fault
[18:01] <common> buildroot-ng got no xscale option yet
[18:10] <tziOm> common, what are you?
[18:14] <tziOm> nbd, when are you thinking of testing your wlhdd?
[18:16] <nbd> when i have time
[18:22] <tziOm> nbd, hmm.. I notice my laptop is printing this line: Probing IDE interface ide0...
[18:23] <tziOm> I dont see anything like this in the wlhdd...
[18:23] <tziOm> any way to make this happend?
[18:23] <crazy_imp> nbd: pdc202xx_old, i can load it now, but it isn't loaded by the modules.d/20-ide (it's there in the first line) if i kick the other ide modules, i couldn't load it (unsolved symbols) so i think it should be reorderd (or is there something against it?). but still doesn't find my hdd, tried to plug it in while the modules are loaded, no changes, no dmesg record
[18:24] <nbd> i'll look into it
[18:24] <nbd> just wait a while
[18:24] <nbd> i'm busy with a few other things right now
[18:25] <frop> nbd: ping
[18:25] <nbd> pong
[18:25] <crazy_imp> ok, i still have to learn for school, so its good if i don't get new things to play around :D
[18:25] <frop> i've just tried to flash AR7 without that 003-net_driver_cpmac.patch hack
[18:25] <frop> ....passing "setenv MAC_PORT 0" to ADAM"
[18:25] <frop> 2
[18:26] <frop> it change from "external PHY" to "internal PHY"
[18:28] <tziOm> nbd, could it be drivers/ide/ide-probe.c related..? seems like it even tho it finds controller and everything is ok, it doesnt find the disk connected since it does not probe for it seems..
[18:31] <tziOm> more exactly probe_hwif
[18:40] <tziOm> hmm.. seems one could put the PDC20265 in raid mode...
[18:43] <nbd> tziOm: what for?
[18:43] <tziOm> dunno.. cheap controller seems
[18:43] <nbd> raid mode with only one port
[18:43] <nbd> wow, that's useful
[18:43] <tziOm> could grab more there probably
[18:44] <tziOm> just not connected or perhaps on extra solderspot
[18:46] <crazy_imp> tzi0m: forget about extra solderpads, there aren't enought for ide, and you would need a 2. external psu for the drive or you have to delay the start of the second...
[18:49] <tziOm> maby ide works now... hmm..
[19:13] <tziOm> there must be bugs in the driver source
[19:13] <tziOm> everything is configured correctly from my side
[19:57] <tziOm> nbd, there?
[19:57] <nbd> no
[19:57] Action: nbd is away: hacking
[19:58] <tziOm> nbd, doing some testing here.. with whiterussian for simplicity... and found that it is the ide-detect stuff (ide-detect.o module) that does what I am missing... is it possible to enable this in kernel? (No...) is this something that is hacked away somehow?
[19:59] <tziOm> nbd, shouldnt need to have the ide stuff in modules.. havent got that on other systems of mine.. and there the ide-detectioon goes as should
[20:00] <nbd> i'll look into it when i have time and feel like it
[20:01] <tziOm> nbd - atleast I get exactly the same dmesg output in whiterussian... as in kami .. until I load the ide-detect thats when it starts working and the dmesg output is different
[20:02] <tziOm> nbd, if you have any clues of whatitmightbe, can you please hint.. so I can look at it _now_!?
[20:02] <nbd> i cannot tell you until i look at the code myself
[20:05] <[mbm]> ...
[20:11] <tziOm> [mbm], what about you?
[20:11] <[mbm]> ?
[20:11] <tziOm> do you know howto fix the ide issues im having?
[20:11] <nbd> tziOm: seriously. stop bugging people all the time
[20:13] <CIA-4> nbd * r4007 /branches/buildroot-ng/openwrt/target/linux/rb532-2.6/ (config patches/100-rb5xx_support.patch): fix rb532 for 2.6.17
[20:14] <frop> /voice CIA-4
[20:14] <nbd> :)
[20:14] <frop> what's rb532??
[20:15] <nbd> routerboard 523 from mikrotik
[20:15] <nbd> 532
[20:15] <frop> ok
[20:15] <frop> i've not think to it
[20:15] <frop> ...there're plan on "when" buildroot-old will get into trash?
[20:16] <nbd> as soon as most of the important packages are ported to the new system and the remaining issues are fixed
[20:16] <nbd> target/linux still needs to be redone partially
[20:16] <frop> ...apart "when ng it'll be ready"...
[20:16] <frop> nbd ahah...i was sure of your response!
[20:16] <frop> :)
[20:16] <nbd> and the build system should have a way of suppressing the trash that's scrolling during the build
[20:16] <nbd> like the old one has
[20:17] <nbd> but with a better structure
[20:17] <frop> eheh --not-verbose
[20:17] <nbd> yeah
[20:17] <frop> ...and how about rc6?
[20:18] <frop> latest: rc5 (huh? still? yeah...)
[20:18] <nbd> do you have the backlog of yesterday's developer meeting?
[20:18] <frop> no...i wasn't here...at home at least...
[20:18] <frop> there's something onto website?
[20:18] <frop> Kernel .config support (IKCONFIG) [N/y/?] n
[20:18] <frop> Kernel->user space relay support (formerly relayfs) (RELAY) [N/y/?] (NEW)
[20:19] <frop> building 4006
[20:19] <nbd> http://openwrt.org/logs/openwrt-devel.log.20060618
[20:19] <frop> :)
[20:19] <nbd> frop: the transition to 2.6.17 is not fully complete yet, some configs still need to be updated
[20:20] <frop> k...i'll wait for some revision too...
[20:20] <frop> s
[20:20] <nbd> which target are you building for?
[20:21] <frop> brcm-2.6
[20:21] <frop> [22:35] <[mbm]> so .. rc6
[20:21] <frop> [22:35] <nbd> rc6...
[20:21] <frop> [22:35] <{Nico}> hmm, rc6
[20:21] <frop> [22:35] <Kaloz> huh, another rc? ;)
[20:21] <frop> ahaha
[20:23] <tziOm> get this error in rc5 (first this: ext2fs_check_if_mount: No such file or directory while determining whether /dev/discs/disc0/part1 is mounted.) .. then this, which causes it to fail: Creating journal (32768 blocks): mke2fs: No such file or directory while trying to create journal
[20:25] <dragorn> sounds like it's expecting devfs
[20:25] <tziOm> might be simple.. hold..
[20:25] <dragorn> or a udev configured differently
[20:26] <tziOm> guessing on mtab, but strange error messages..
[20:26] <dragorn> both those errors are pretty obviously "that device node doesn't exist"
[20:26] <dragorn> so figure out either why your ide nodes don't appear, or fix your fstab
[20:26] <tziOm> yep, was the /etc/mtab that didnt exist .. linked it to /proc/mounts
[20:26] <dragorn> or change your udev rules, since devfs is dead and buried in modern kernels
[20:27] <dragorn> thats fine, linking it to /proc/mounts is perfectly standard
[20:27] <dragorn> and has nothing to do with dev nodes not existing
[20:27] <[mbm]> frop: yeah the rc's have become a bit of a joke
[20:27] <frop> [23:15] <nbd> still better than writing a bloated python monster :)
[20:27] <tziOm> wth doesnt that link exist by default?
[20:27] <frop> AHAHA
[20:27] Action: frop quote nbd
[20:27] <frop> [mbm] i think that a 0.9 will be a better "name" too
[20:28] <CIA-4> kaloz * r4008 /branches/buildroot-ng/openwrt/target/linux/aruba-2.6/config: update Aruba config for 2.6.17
[20:28] <frop> it gives to tha user a "sense" of "stability"
[20:28] <[mbm]> frop: well, it was originally scheduled for 1.0 but we thought it was silly to call that 1.0 when we'd done so much work on kamikaze and buildroot-ng
[20:28] <tziOm> you shouldnt do anything before you have .17 working with bcm43xx..
[20:28] <dragorn> [mbm]: This is why I gave up numbering releases
[20:29] <[mbm]> frop: so now we've bumped that to 0.9 with 1.0 being a merge with all the new stuff
[20:29] <dragorn> [mbm]: Year, month, subrel num :P
[20:29] <dragorn> [mbm]: otherwise i'd be on like version 19.
[20:29] <[mbm]> dragorn: yeah, much better scheme .. (wish I'd have thought about that sooner)
[20:29] <frop> [mbm] it's good for me...
[20:29] <dragorn> [mbm]: I hit the same versioning thrash. "Well, this is a huge update, so it should be another major rev number, but I'm already at version 4, durr...."
[20:30] <[mbm]> dragorn: wait, you mean you're not on revision 20060k?
[20:31] <dragorn> or version 0.3.99997? :P
[20:31] <[mbm]> if I ever do another version it'll be 0.3
[20:31] <[mbm]> then 0.3.1.4.1 ..
[20:32] <nbd> 0.1.3.3.7
[20:33] <nbd> or 0.1.33.7
[20:33] <nbd> or better: 0.1.33-7 :)
[20:33] <[mbm]> nbd: you don't get it .. the 0.pi numbering scheme never runs out of sub revisions
[20:35] <dragorn> until they ban pi for containing every copyrighted work in the universe
[20:36] <nbd> dragorn: nah, they wouldn't do that. they'd just go after everybody that knows the offsets :)
[20:36] <[mbm]> in other words an infitite number of monkeys and typewriters will never produce pi
[21:00] <kender> hey
[21:09] <[mbm]> 'lo
[21:12] <florian__> such a pity someone update kernel to 2.6.17 without testing all archs
[21:18] <florian__> do you think we could work on a common generic kernel configuration, for instance everything related to netfilter is generic
[21:18] <florian__> almos same thing for drivers
[21:19] <florian__> it would be just easier to only merge new arch-specific options rather than checking the whole config file per target ihmo
[21:19] <nbd> yeah, i already suggested something like that
[21:20] <nbd> although we need to have different netfilter configurations for different targets
[21:20] <florian__> right
[21:20] <nbd> but we should do it in a way so that we have scripts for converting the kernel config and splitting it up into the generic parts
[21:20] <florian__> then we could have several templates : netfilter-router, netfilter-bridge
[21:20] <nbd> so that it's still easy to make changes without having to hand-edit everything
[21:20] <[mbm]> maybe we should get rid of the policy of making all modules and packaging them and just dynamically generate the kernel config
[21:21] <nbd> yeah
[21:21] <nbd> makes sense
[21:21] <florian__> right
[21:21] <nbd> although we really need to make sure that the kernel build handles it properly
[21:21] <nbd> 2.6 should be fine
[21:21] <nbd> but i don't know about 2.4
[21:21] <florian__> 2.6 well manages dependencies
[21:21] <nbd> 2.4's kernel build system behaves in a very weird way, occasionally
[21:22] <florian__> we will probably get rid of 2.4 ?
[21:22] <nbd> not too soon
[21:22] <florian__> what's blocking?
[21:22] <florian__> bcm43xx in AP mode ?
[21:22] <florian__> among other things
[21:22] <nbd> broadcom on 2.6 still has a long way to go until it can work as well as the 2.4 stuff
[21:22] <nbd> so we should continue maintainig the 2.4 port for a while
[21:23] <nbd> anyway, we can exclude 2.4 from the generic config stuff
[21:23] <nbd> and just write that for 2.6
[21:23] <nbd> makes things easier
[21:24] <nbd> we should be able to do it without breaking 2.4 or doing much extra work around its weirdness
[21:24] <florian__> any new developper would like to join the package porting effort ?
[21:32] <florian__> humm au1000-2.6 is broken
[21:32] <h3sp4wn> florian: Does it just entail rewriting the Makefiles in the buildroot-ng style ? (I could do that)
[21:32] <florian__> almost
[21:33] <florian__> by the way, your yafc package is not available to me :( server #500
[21:33] <h3sp4wn> That server has been decommisioned unfortunately
[21:34] <florian__> ah no pb
[21:36] <h3sp4wn> What is the other thing / things it entails ?
[21:37] <florian__> that's almost it
[21:40] <malbon> florian__: are you talking about /packages ?
[21:47] <[mbm]> ip
[21:49] <florian__> yep
[21:53] <malbon> florian__: ok. quick question, how do I make packages compile (no I've not tried yet)
[21:53] <malbon> florian__: I am thinking of giving some packages a go, but buildroot-ng confuses me... ;)
[21:53] <nbd> malbon: what part confuses you?
[21:54] <malbon> nbd: how /packages relates to my tree
[21:55] <nbd> it doesn't relate to that tree (yet)
[21:56] <malbon> so I just port them into my normal buildroot-ng tree, but check them into packages?
[21:56] <nbd> yes
[21:56] <malbon> (after testing of course) :P
[21:56] <nbd> of course
[21:56] <malbon> ok, that's not too hard.
[21:57] <nbd> in the future i want to have a script that resolves dependencies and copies packages from /packages into the buildroot
[21:57] <malbon> yes, something that does it automagically would be good.
[21:57] <nbd> shouldn't be too hard to write
[21:58] <nbd> we only need to come up with a proper categorization scheme for the stuff in /packages/
[21:58] <nbd> i already checked some libraries into libs/
[21:58] <nbd> i think we can keep that
[21:59] <malbon> yes I saw. I was going to give things like hostapd and chillispot a go.
[21:59] <CIA-4> nbd * r4009 /branches/buildroot-ng/openwrt/ (20 files in 7 dirs): remove the wl.o build from the kernel patch and move the driver along with wl, nas and wlc into packages/broadcom-wl
[22:05] <tziOm> may I ask you about webif.. is this something developed for this project? can You point me to sources?
[22:06] <nbd> tziOm: whiterussian svn
[22:07] <tziOm> is it possible to co only webif?
[22:10] <florian__> yep
[22:17] <florian__> whiterussian is still feature freeze right ?
[22:21] <nbd> depends
[22:22] <nbd> the plan is to bring the current tree into a useful state for a 0.9 release
[22:22] <CIA-4> nbd * r4010 /branches/buildroot-ng/openwrt/target/linux/image/brcm/Makefile: fix broadcom trx alignment
[22:23] <florian__> ok I just wanted to close the nfs-kernel-server ticket
[22:23] <nbd> make it a package for buildroot-ng
[22:23] <nbd> that should be enough
[22:23] <florian__> right
[22:30] <CIA-4> nbd * r4011 /branches/buildroot-ng/openwrt/package/broadcom-wl/Makefile: add modules.d file for wl
[22:31] <CIA-4> nbd * r4012 /branches/buildroot-ng/openwrt/package/wlcompat/Makefile: fix wlcompat depends/default
[22:48] <nbd> cool. the wl.o 'apsta' mode actually works
[22:48] <nbd> i managed to set it up manually with wlc
[22:48] <nbd> now i just need to finish my magic shell scripting code to configure that thing
[22:49] <[mbm]> wait, you can actually be ap and sta at once?
[22:49] <nbd> yes
[22:49] <[mbm]> nice
[22:49] <florian__> lol whoa
[22:49] <nbd> but it's a special mode, so you can't just set arbitrary modes for virtual interfaces
[22:50] <nbd> and this time when i changed several runtime parameters with wlc, it didn't even oops or spit out weird things
[22:50] <nbd> so the newer one from linksys seems to be a bit more reliable than the last one
[22:50] <malbon> does it have vif's with different mac addresses?
[22:51] <nbd> unfortunately not
[22:51] <malbon> shame
[22:51] <nbd> which makes it really hard to use different ssids with different encryption on windows
[22:51] <nbd> because the windows 802.11 stack is too stupid to handle that
[22:51] <florian__> can you configure the ethernet address with ifconfig hw ether ?
[22:51] <nbd> florian__: no, this limitation is by design
[22:51] <nbd> florian__: it doesn't keep track of different bssids for different interfaces
[22:52] <florian__> ok
[22:52] <florian__> even under linux, I scan intermitently 1 or the other ssid
[22:52] <malbon> it's like horse riding 100 miles to shoot yourself in the foot. crazy
[22:52] <nbd> it's broadcom
[22:53] <nbd> what do you expect? :)
[22:53] <malbon> a reasonable ammount of workyness would be nice.
[22:53] <nbd> hehe
[22:53] <nbd> well, maybe the open source driver can get that feature first
[22:53] <nbd> :)
[22:53] <nbd> i know for sure that the template ram can send out beacons with different bssids
[22:53] <malbon> hopefully. I think it probably will.
[22:53] <florian__> they are doing great progress on newer mimo dirvers
[22:54] <nbd> because if you hit a weird case in the driver where it doesn't configure the template ram, you will get a BRCM_TEST_SSID on a different bssid
[22:54] <nbd> that reminds me ... i should change the broadcom-wl package so that it also builds the mimo version
[22:58] <[mbm]> is it just me or is there very little work done on making linux into an embeded solution?
[22:58] <[mbm]> seems like anything other than an x86 arch, or anything involving cross compiling is just asking for problems
[22:59] <nbd> right
[23:00] <florian__> [mbm]: what makes you think that ?
[23:01] <[mbm]> florian__: experience?
[23:02] <florian__> [mbm]: what would you recommend for embedded systems ?
[23:11] <CIA-4> nbd * r4013 /branches/buildroot-ng/openwrt/package/broadcom-wl/ (Makefile src/kmod/Makefile): build mimo version of the broadcom wl driver as well
[23:45] <CIA-4> nbd * r4014 /branches/buildroot-ng/openwrt/package/base-files/default/etc/functions.sh: add a list append function to functions.sh
[23:52] <[mbm]> grr
[23:52] <[mbm]> new kenrel forgot -ffreestanding
[23:52] <[mbm]> which causes some strcpy errors
[00:00] --- Tue Jun 20 2006
[00:02] <tziOm> what I get is:
[00:02] <tziOm> Uniform Multi-Platform E-IDE driver Revision: 7.00beta4-2.4
[00:02] <tziOm> ide: Assuming 33MHz system bus speed for PIO modes; override with idebus=xx
[00:02] <tziOm> .. nothing more
[00:05] <crazy_imp> thats the same as i get
[00:06] <tziOm> no PCI: enabling device...
[00:06] <crazy_imp> i also tried the pdc202xx_new, same
[00:07] <tziOm> must they be modules? ... no...
[00:08] <tziOm> anyone?
[00:08] <crazy_imp> also build build in the kernel, same, nothing is other
[00:08] <tziOm> you dont get any messages when you insmod it?
[00:08] <crazy_imp> tziOm: do you have rc5 on it atm? what does lspci says?
[00:09] <crazy_imp> i tells me: no such device if i try to insert it with isnmod
[00:09] <tziOm> no, I have kami
[00:10] <crazy_imp> ok, then i will put rc5 on it myself and take a look what lspci says
[00:12] <crazy_imp> tziOm: do you have any ideas what the 19 solderpads are, at the end of the board, behind the ideconnector?
[00:13] <tziOm> nope, havent checked...
[00:13] <tziOm> some io prolly
[00:14] <crazy_imp> yes, but for what :D
[00:14] <h3sp4wn> what about the modem that those boards have ?
[00:16] <crazy_imp> i'm a little bit confused about the sentry5 chipset, lspci tells me that there are 3 ethernet controllers on it, but it only has lan and wlan, so what is the third for?
[00:18] <florian__> nbd, {Nico} how do we proceed for packages ?
[00:19] <nbd> port all of them to buildroot-ng and make sure that they're structured and categorized
[00:19] <crazy_imp> looks like i have to read many datasheets to understand the device :D
[00:19] <nbd> freeze trunk at the moment
[00:19] <florian__> ok
[00:20] <florian__> alphabetical order ? most useful order ?
[00:20] <tziOm> nbd, do you know anything about the wlhdd and the CONFIG_BLK_DEV_PDC202XX_OLD that you would like to share?
[00:21] <nbd> florian__: maybe it would make sense to make a packages/old directory, svn cp in all the stuff from trunk and then slowly move and convert it over in the order of relevance
[00:21] <nbd> florian__: that has the additional sideffect of tracking the renames
[00:21] <nbd> tziOm: not at the moment. but i will test it soon, because i need it for a project
[00:22] <florian__> nbd: ok
[00:22] <tziOm> read my above.. even tho compiled into kernel... it seems not there...
[00:23] <tziOm> does it need PCMCIA? no...
[00:23] <tziOm> nbd, and what about the "PCI: Disabled\nPCI: Fixing up bus 0" message? thats correct?
[00:25] <nbd> might be wrong, no idea
[00:25] <tziOm> root@OpenWrt:/# cat /proc/ide/drivers
[00:25] <tziOm> ide-disk version 1.17
[00:25] <tziOm> ide-default version 0.9.newide
[00:25] <nbd> i'll have a short glance at the system code...
[00:26] <florian__> nbd: did you get your hand on your presentation ?
[00:27] <nbd> not yet
[00:27] <florian__> ok, the presentation has been accepted, so now I will gather some informations here ;)
[00:30] <CIA-4> nbd * r4003 / (2 files in 2 dirs): fix typo in brcm-2.4 system code
[00:30] <nbd> tziOm: try svn update; make target/linux-clean world
[00:32] <tziOm> nbd, target/linux/brcm-2.4/patches/001-bcm47xx.patch
[00:32] <nbd> yes
[00:32] <nbd> that's where i fixed the typo
[00:32] <tziOm> wth was it?
[00:33] <nbd> a wrong device id
[00:33] <nbd> was 0x4710, while it should have been 0x4310
[00:33] <tziOm> shit..
[00:33] <tziOm> willl it do anything with the wl module too maby then, sucker! ;)
[00:34] <nbd> sucker?
[00:37] <tziOm> any command to make ping exit true when it has made X successful pings (and gotten replies?)
[00:38] <nbd> *sigh*
[00:39] <nbd> there are two parts to it
[00:39] <nbd> part 1: type 'man ping'
[00:39] <nbd> part 2: understand the text
[00:39] <nbd> no, wait
[00:39] <nbd> 3 parts
[00:39] <nbd> part 3: actually do it
[00:40] <tziOm> ping -c1 192.168.1.6 >/dev/null 2>&1|| echo loop
[00:42] Action: nbd goes to sleep
[00:42] <tziOm> nbd, works like a charm
[00:42] <tziOm> thanks duud
[00:43] <nbd> ide as well?
[00:43] <tziOm> atleast the pdc20265 seems
[00:43] <tziOm> dunno if it found any disk tho.
[00:44] <tziOm> seems not..
[00:45] <tziOm> Uniform Multi-Platform E-IDE driver Revision: 7.00beta4-2.4
[00:45] <tziOm> ide: Assuming 33MHz system bus speed for PIO modes; override with idebus=xx
[00:45] <tziOm> PDC20265: IDE controller at PCI slot 01:01.0
[00:45] <tziOm> PCI: Enabling device 01:01.0 (0004 -> 0007)
[00:45] <tziOm> PDC20265: chipset revision 2
[00:45] <tziOm> PDC20265: not 100% native mode: will probe irqs later
[00:45] <tziOm> PDC20265: ROM enabled at 0x000d0000
[00:45] <tziOm> PDC20265: (U)DMA Burst Bit DISABLED Primary PCI Mode Secondary PCI Mode.
[00:45] <tziOm> PDC20265: FORCING BURST BIT 0x00->0x01 ACTIVE
[00:45] <tziOm> ide0: BM-DMA at 0x0140-0x0147, BIOS settings: hda:pio, hdb:pio
[00:45] <tziOm> ide1: BM-DMA at 0x0148-0x014f, BIOS settings: hdc:DMA, hdd:pio
[00:45] <crazy_imp> so it works now?
[00:46] <tziOm> that works so far.. but it does not find my connected disk seems..
[00:46] <tziOm> no devs in /dev or /ide
[00:46] <tziOm> devide
[00:46] <crazy_imp> kmod-ide installed and modules loaded?
[00:46] <nbd> maybe you forgot something in your kernel config
[00:46] <nbd> crazy_imp: he refuses to use modules
[00:46] <tziOm> dont think so, no
[00:47] <tziOm> CONFIG_BLK_DEV_IDEDISK=y
[00:50] <tziOm> root@OpenWrt:/# cat /proc/ide/drivers
[00:50] <tziOm> ide-disk version 1.17
[00:50] <tziOm> ide-default version 0.9.newide
[00:51] <tziOm> any clues?
[00:51] Action: nbd is gone: sleep
[00:53] <wbx> iwlist wl0 scanning
[00:53] <wbx> Unable to handle kernel paging request at virtual address c0170000, epc == c016beb0, ra == c016bf98
[00:53] <wbx> hmm. known bug?
[00:54] <nbd> try grabbing the latest wl from linksys
[00:54] <nbd> and see if it causes the same problem
[00:54] <nbd> wlcompat hasn't been tested much with the new driver
[01:00] <h3sp4wn> Is there a newer nas than the one in white russian ?
[01:06] <h3sp4wn> I was trying to extract the nas from the latest wl500-gd firmware (using hexdump) but if its pointless I will stop trying
[02:18] <CIA-4> nico * r4004 /branches/buildroot-ng/openwrt/package/wlcompat/ (Makefile src/Makefile src/wlcompat.c wlcompat.c): move source file to src/ and let the kernel tree build the modules.
[03:20] <crodler> how can i write on a loop-mounted cramfs image?
[04:05] <CIA-4> nico * r4005 /branches/buildroot-ng/openwrt/package/switch/src/ (switch-adm.c switch-core.c switch-robo.c): fix module params on 2.6.17, suppress warnings.
[13:17] <CIA-4> kaloz * r4006 /branches/buildroot-ng/openwrt/target/linux/aruba-2.6/patches/ (000-aruba.patch 010-ar2313_enet.patch): fix the Aruba ethernet drivers for 2.6.17
[14:32] <unverbraucht> i'm trying to find the uhci problems on kamikaze (brcm) 2.6
[14:33] <unverbraucht> can anybody give me a pointer where to start?
[14:33] <unverbraucht> i'd first diff the uhci against a vanilla 2.6.17 uhci_hcd.c
[14:33] <unverbraucht> and hope that anything shows up there :)
[14:46] <unverbraucht> nbd: michael sent out a patch up-stream with your bcm43xx hostap changes
[14:46] <unverbraucht> https://lists.berlios.de/pipermail/bcm43xx-dev/2006-June/001972.html
[14:46] <unverbraucht> :)
[14:49] <Kaloz> unverbraucht: they are NDL's changes :p
[14:49] <unverbraucht> oh, sorry about that :)
[14:49] <unverbraucht> just remembered the nick started with an n... :/
[15:52] <crazy_imp> nbd: the ide bug for the wl-hdd isn't solved yet, or?
[16:01] <nbd> crazy_imp: what bug exactly?
[16:01] <crazy_imp> this one: insmod pdc202xx_new.o
[16:01] <crazy_imp> insmod: init_module: pdc202xx_new: No such device
[16:03] <nbd> you need to use _old
[16:04] <crazy_imp> but with the _old i had the same problem, btw. where is the _new? i didn't kicked it out of the config *confused*
[16:04] <nbd> is this kamikaze or whiterussian?
[16:07] <crazy_imp> nbd: kamikaze (r4006) with 2.4 kernel, the 2.6 kernel drops the wl-hdd in an reboot loop :/
[16:08] <nbd> the bug in kamikaze is solved
[16:08] <nbd> do svn update; make target/linux-clean world
[16:10] <crazy_imp> ok, i hope it solves the bug :)
[16:11] <nbd> crazy_imp: it did solve the bug for tziOm
[16:11] <nbd> it was a simple 1-character-typo in a device id
[16:11] <crazy_imp> btw, the pdc20265r can control a raid :D
[16:12] <crazy_imp> ok, lets see if i can confirm it
[16:12] <crazy_imp> where is the difference between the wl-hdd, and the wl-hdd2.5 only the cpu?
[16:13] <nbd> i didn't know those two were different at al
[16:13] <crazy_imp> the toh list it as different, the wl-hdd is a little bit faster
[16:19] <crazy_imp> lspci gives me more 3 more devices as the last build before :D
[16:39] <tziOm> nbd, any news on the ide issue?
[16:41] <crazy_imp> :D
[16:41] <tziOm> crazy_imp, hey
[16:41] <crazy_imp> wait some minutes and i can say yes or no :D
[16:42] <tziOm> i say no
[16:42] <tziOm> no updates to svn atleast
[16:42] <crazy_imp> i wish i can say yes if it's ready
[16:43] <crazy_imp> the r4006 is a new try for me now, forgot to clean it up before building a new one
[16:43] <crazy_imp> tzi0m: do you have a wl-hdd or a wl-hdd2.5?
[16:44] <tziOm> its a 2.5" hdd in it...
[16:44] <tziOm> didnt know it existed 3.5 versions..
[16:44] <crazy_imp> how many bogomips?
[16:44] <nbd> 200
[16:44] <tziOm> BogoMIPS : 82.94
[16:44] <nbd> hmm.. you got the slow one
[16:44] <crazy_imp> tzi0m: thats the same as my wl-hdd2.5 has
[16:45] <nbd> mine's also using 2.5"
[16:45] <nbd> but it has 200 mhz
[16:45] <crazy_imp> no
[16:45] <tziOm> yeah? cool.
[16:45] <crazy_imp> 125mhz cpu
[16:45] <tziOm> could one clock it?
[16:45] <nbd> what do you mean no?
[16:45] <nbd> mine does have 200 mhz
[16:45] <crazy_imp> CPU revision is: 00024000
[16:45] <crazy_imp> CPU: BCM4710 rev 0 at 125 MHz
[16:46] <nbd> i think there are several revs of the same product
[16:46] <nbd> and i got the faster version
[16:47] <tziOm> nbd, have you tested the ide controller on it? I seem to have luck with getting the PDC20265 working now, but no luck with detecting the ide drive.. actually no info about ide devs atall...
[16:47] <nbd> how much did you change the config/
[16:47] <nbd> ?
[16:47] <tziOm> just this:
[16:47] <nbd> :)
[16:47] <tziOm> PDC20265: FORCING BURST BIT 0x00->0x01 ACTIVE
[16:47] <tziOm> ide0: BM-DMA at 0x0140-0x0147, BIOS settings: hda:pio, hdb:pio
[16:47] <tziOm> ide1: BM-DMA at 0x0148-0x014f, BIOS settings: hdc:DMA, hdd:pio
[16:48] <tziOm> nbd, not much changes in the ide part.. just in kernel no modules.
[16:48] <tziOm> nbd, does it work at your place?
[16:48] <nbd> i haven't tested yet
[16:48] <nbd> but last time i checked, it worked
[16:48] <nbd> on whiterussian
[16:48] <crazy_imp> nbd: my rev is r1.15, i think i will create a wl-hdd2.5 page in the wiki if i have time :)
[16:54] <crazy_imp> time to flash :D
[16:56] <tziOm> nbd, are you working on getting 2.6.17 in?
[16:56] <nbd> 2.6.17 is already in
[16:56] <nbd> but it won't work on the wl-hdd
[16:56] <tziOm> nbd, do you know how the bcm43xx drivers behave?
[16:56] <nbd> no
[16:57] <nbd> before i integrate it, i need to rewrite parts of the system code
[16:57] <tziOm> no..? maby I should check it out.. how can I choose it? Do I need make DEVELOPER=1 or something?
[16:57] <nbd> no
[16:57] <nbd> you can't use it yet
[16:57] <tziOm> nbd, lots of work to be done?
[16:57] <nbd> not that much
[16:58] <nbd> but still: you can't use it yet
[16:58] <nbd> you have to wait
[16:58] <nbd> and don't ask me about it every day
[16:58] <nbd> because at some point, i'll simply start ignoring you completely
[16:58] <tziOm> hmm...
[16:58] <tziOm> what version is used as the brcm-2.6 now then?
[16:59] <nbd> that's it. you've reached the quota of answers from me for now
[16:59] <tziOm> nbd, you have some major issues Im not educated to deal with
[16:59] <nbd> no, you seem to need some lessons in being less annoying
[17:00] <tziOm> I find you a bad team-player.. we are on the same side you know.. doing similar stuff mol
[17:00] <tziOm> .
[17:00] <nbd> i hate it when people throw lots of questions at me without even *trying* to figure things out themselves
[17:00] <nbd> and i told you that before
[17:00] <nbd> it doesn't have anything to do with being a bad team player
[17:00] <nbd> you're simply wasting my time
[17:01] <crazy_imp> nbd: it's still building the pdc202xx_new
[17:02] <nbd> checking...
[17:02] <nbd> in the default config only _OLD is marked as m
[17:04] <crazy_imp> hm, i can't remember that i have changed it
[17:05] <nbd> go into target/linux/brcm-2.4 and run svn status
[17:06] <crazy_imp> ok, wait
[17:07] <crazy_imp> drops me: M config (so its my fault?)
[17:08] <nbd> looks like it
[17:08] <nbd> do svn revert config
[17:09] <crazy_imp> and now the same procedure like before (svn update; make target/linux-clean world)?
[17:09] <nbd> no need to do svn update
[17:09] <crazy_imp> (without the update)
[17:09] <nbd> yeah
[17:09] <crazy_imp> ok
[17:09] <tziOm> or svn diff
[17:16] <crazy_imp> if i look at the board of my wl-hdd, the flashchip has 8 lines (thats the amount i can see, maybe there are more, because there are 19 pads) to the solderpads i'm thinking about the hole time, is it jtag? some brings a voltage of 3,3v one comes along with 5,2v (its the same as the power for the hdd) and one with 0,25v and some doesn't shows voltage against the ground
[17:22] <nbd> crazy_imp: maybe it's also the same bus that you can use on the wl-500g to connect an uart for serial access
[17:23] <crazy_imp> ok, will take a look
[17:48] <tziOm> crazy_imp, might be a modem in there somewhere..
[17:48] <tziOm> 00:03.0 Modem: Broadcom Corporation BCM47xx V.92 56k modem
[17:48] <crazy_imp> i know
[17:49] <nbd> tziOm: i think that's just a misclassification in the system code
[17:50] <crazy_imp> thats a point to fetch all the datasheets about the chips and look for a modem :D
[17:50] <tziOm> s/ misclassification in the/nother typo in my/
[17:50] <tziOm> ;)
[17:50] <nbd> no
[17:51] <nbd> this is something that broadcom had left in their code
[17:51] <nbd> so not exactly my fault
[18:01] <common> buildroot-ng got no xscale option yet
[18:10] <tziOm> common, what are you?
[18:14] <tziOm> nbd, when are you thinking of testing your wlhdd?
[18:16] <nbd> when i have time
[18:22] <tziOm> nbd, hmm.. I notice my laptop is printing this line: Probing IDE interface ide0...
[18:23] <tziOm> I dont see anything like this in the wlhdd...
[18:23] <tziOm> any way to make this happend?
[18:23] <crazy_imp> nbd: pdc202xx_old, i can load it now, but it isn't loaded by the modules.d/20-ide (it's there in the first line) if i kick the other ide modules, i couldn't load it (unsolved symbols) so i think it should be reorderd (or is there something against it?). but still doesn't find my hdd, tried to plug it in while the modules are loaded, no changes, no dmesg record
[18:24] <nbd> i'll look into it
[18:24] <nbd> just wait a while
[18:24] <nbd> i'm busy with a few other things right now
[18:25] <frop> nbd: ping
[18:25] <nbd> pong
[18:25] <crazy_imp> ok, i still have to learn for school, so its good if i don't get new things to play around :D
[18:25] <frop> i've just tried to flash AR7 without that 003-net_driver_cpmac.patch hack
[18:25] <frop> ....passing "setenv MAC_PORT 0" to ADAM"
[18:25] <frop> 2
[18:26] <frop> it change from "external PHY" to "internal PHY"
[18:28] <tziOm> nbd, could it be drivers/ide/ide-probe.c related..? seems like it even tho it finds controller and everything is ok, it doesnt find the disk connected since it does not probe for it seems..
[18:31] <tziOm> more exactly probe_hwif
[18:40] <tziOm> hmm.. seems one could put the PDC20265 in raid mode...
[18:43] <nbd> tziOm: what for?
[18:43] <tziOm> dunno.. cheap controller seems
[18:43] <nbd> raid mode with only one port
[18:43] <nbd> wow, that's useful
[18:43] <tziOm> could grab more there probably
[18:44] <tziOm> just not connected or perhaps on extra solderspot
[18:46] <crazy_imp> tzi0m: forget about extra solderpads, there aren't enought for ide, and you would need a 2. external psu for the drive or you have to delay the start of the second...
[18:49] <tziOm> maby ide works now... hmm..
[19:13] <tziOm> there must be bugs in the driver source
[19:13] <tziOm> everything is configured correctly from my side
[19:57] <tziOm> nbd, there?
[19:57] <nbd> no
[19:57] Action: nbd is away: hacking
[19:58] <tziOm> nbd, doing some testing here.. with whiterussian for simplicity... and found that it is the ide-detect stuff (ide-detect.o module) that does what I am missing... is it possible to enable this in kernel? (No...) is this something that is hacked away somehow?
[19:59] <tziOm> nbd, shouldnt need to have the ide stuff in modules.. havent got that on other systems of mine.. and there the ide-detectioon goes as should
[20:00] <nbd> i'll look into it when i have time and feel like it
[20:01] <tziOm> nbd - atleast I get exactly the same dmesg output in whiterussian... as in kami .. until I load the ide-detect thats when it starts working and the dmesg output is different
[20:02] <tziOm> nbd, if you have any clues of whatitmightbe, can you please hint.. so I can look at it _now_!?
[20:02] <nbd> i cannot tell you until i look at the code myself
[20:05] <[mbm]> ...
[20:11] <tziOm> [mbm], what about you?
[20:11] <[mbm]> ?
[20:11] <tziOm> do you know howto fix the ide issues im having?
[20:11] <nbd> tziOm: seriously. stop bugging people all the time
[20:13] <CIA-4> nbd * r4007 /branches/buildroot-ng/openwrt/target/linux/rb532-2.6/ (config patches/100-rb5xx_support.patch): fix rb532 for 2.6.17
[20:14] <frop> /voice CIA-4
[20:14] <nbd> :)
[20:14] <frop> what's rb532??
[20:15] <nbd> routerboard 523 from mikrotik
[20:15] <nbd> 532
[20:15] <frop> ok
[20:15] <frop> i've not think to it
[20:15] <frop> ...there're plan on "when" buildroot-old will get into trash?
[20:16] <nbd> as soon as most of the important packages are ported to the new system and the remaining issues are fixed
[20:16] <nbd> target/linux still needs to be redone partially
[20:16] <frop> ...apart "when ng it'll be ready"...
[20:16] <frop> nbd ahah...i was sure of your response!
[20:16] <frop> :)
[20:16] <nbd> and the build system should have a way of suppressing the trash that's scrolling during the build
[20:16] <nbd> like the old one has
[20:17] <nbd> but with a better structure
[20:17] <frop> eheh --not-verbose
[20:17] <nbd> yeah
[20:17] <frop> ...and how about rc6?
[20:18] <frop> latest: rc5 (huh? still? yeah...)
[20:18] <nbd> do you have the backlog of yesterday's developer meeting?
[20:18] <frop> no...i wasn't here...at home at least...
[20:18] <frop> there's something onto website?
[20:18] <frop> Kernel .config support (IKCONFIG) [N/y/?] n
[20:18] <frop> Kernel->user space relay support (formerly relayfs) (RELAY) [N/y/?] (NEW)
[20:19] <frop> building 4006
[20:19] <nbd> http://openwrt.org/logs/openwrt-devel.log.20060618
[20:19] <frop> :)
[20:19] <nbd> frop: the transition to 2.6.17 is not fully complete yet, some configs still need to be updated
[20:20] <frop> k...i'll wait for some revision too...
[20:20] <frop> s
[20:20] <nbd> which target are you building for?
[20:21] <frop> brcm-2.6
[20:21] <frop> [22:35] <[mbm]> so .. rc6
[20:21] <frop> [22:35] <nbd> rc6...
[20:21] <frop> [22:35] <{Nico}> hmm, rc6
[20:21] <frop> [22:35] <Kaloz> huh, another rc? ;)
[20:21] <frop> ahaha
[20:23] <tziOm> get this error in rc5 (first this: ext2fs_check_if_mount: No such file or directory while determining whether /dev/discs/disc0/part1 is mounted.) .. then this, which causes it to fail: Creating journal (32768 blocks): mke2fs: No such file or directory while trying to create journal
[20:25] <dragorn> sounds like it's expecting devfs
[20:25] <tziOm> might be simple.. hold..
[20:25] <dragorn> or a udev configured differently
[20:26] <tziOm> guessing on mtab, but strange error messages..
[20:26] <dragorn> both those errors are pretty obviously "that device node doesn't exist"
[20:26] <dragorn> so figure out either why your ide nodes don't appear, or fix your fstab
[20:26] <tziOm> yep, was the /etc/mtab that didnt exist .. linked it to /proc/mounts
[20:26] <dragorn> or change your udev rules, since devfs is dead and buried in modern kernels
[20:27] <dragorn> thats fine, linking it to /proc/mounts is perfectly standard
[20:27] <dragorn> and has nothing to do with dev nodes not existing
[20:27] <[mbm]> frop: yeah the rc's have become a bit of a joke
[20:27] <frop> [23:15] <nbd> still better than writing a bloated python monster :)
[20:27] <tziOm> wth doesnt that link exist by default?
[20:27] <frop> AHAHA
[20:27] Action: frop quote nbd
[20:27] <frop> [mbm] i think that a 0.9 will be a better "name" too
[20:28] <CIA-4> kaloz * r4008 /branches/buildroot-ng/openwrt/target/linux/aruba-2.6/config: update Aruba config for 2.6.17
[20:28] <frop> it gives to tha user a "sense" of "stability"
[20:28] <[mbm]> frop: well, it was originally scheduled for 1.0 but we thought it was silly to call that 1.0 when we'd done so much work on kamikaze and buildroot-ng
[20:28] <tziOm> you shouldnt do anything before you have .17 working with bcm43xx..
[20:28] <dragorn> [mbm]: This is why I gave up numbering releases
[20:29] <[mbm]> frop: so now we've bumped that to 0.9 with 1.0 being a merge with all the new stuff
[20:29] <dragorn> [mbm]: Year, month, subrel num :P
[20:29] <dragorn> [mbm]: otherwise i'd be on like version 19.
[20:29] <[mbm]> dragorn: yeah, much better scheme .. (wish I'd have thought about that sooner)
[20:29] <frop> [mbm] it's good for me...
[20:29] <dragorn> [mbm]: I hit the same versioning thrash. "Well, this is a huge update, so it should be another major rev number, but I'm already at version 4, durr...."
[20:30] <[mbm]> dragorn: wait, you mean you're not on revision 20060k?
[20:31] <dragorn> or version 0.3.99997? :P
[20:31] <[mbm]> if I ever do another version it'll be 0.3
[20:31] <[mbm]> then 0.3.1.4.1 ..
[20:32] <nbd> 0.1.3.3.7
[20:33] <nbd> or 0.1.33.7
[20:33] <nbd> or better: 0.1.33-7 :)
[20:33] <[mbm]> nbd: you don't get it .. the 0.pi numbering scheme never runs out of sub revisions
[20:35] <dragorn> until they ban pi for containing every copyrighted work in the universe
[20:36] <nbd> dragorn: nah, they wouldn't do that. they'd just go after everybody that knows the offsets :)
[20:36] <[mbm]> in other words an infitite number of monkeys and typewriters will never produce pi
[21:00] <kender> hey
[21:09] <[mbm]> 'lo
[21:12] <florian__> such a pity someone update kernel to 2.6.17 without testing all archs
[21:18] <florian__> do you think we could work on a common generic kernel configuration, for instance everything related to netfilter is generic
[21:18] <florian__> almos same thing for drivers
[21:19] <florian__> it would be just easier to only merge new arch-specific options rather than checking the whole config file per target ihmo
[21:19] <nbd> yeah, i already suggested something like that
[21:20] <nbd> although we need to have different netfilter configurations for different targets
[21:20] <florian__> right
[21:20] <nbd> but we should do it in a way so that we have scripts for converting the kernel config and splitting it up into the generic parts
[21:20] <florian__> then we could have several templates : netfilter-router, netfilter-bridge
[21:20] <nbd> so that it's still easy to make changes without having to hand-edit everything
[21:20] <[mbm]> maybe we should get rid of the policy of making all modules and packaging them and just dynamically generate the kernel config
[21:21] <nbd> yeah
[21:21] <nbd> makes sense
[21:21] <florian__> right
[21:21] <nbd> although we really need to make sure that the kernel build handles it properly
[21:21] <nbd> 2.6 should be fine
[21:21] <nbd> but i don't know about 2.4
[21:21] <florian__> 2.6 well manages dependencies
[21:21] <nbd> 2.4's kernel build system behaves in a very weird way, occasionally
[21:22] <florian__> we will probably get rid of 2.4 ?
[21:22] <nbd> not too soon
[21:22] <florian__> what's blocking?
[21:22] <florian__> bcm43xx in AP mode ?
[21:22] <florian__> among other things
[21:22] <nbd> broadcom on 2.6 still has a long way to go until it can work as well as the 2.4 stuff
[21:22] <nbd> so we should continue maintainig the 2.4 port for a while
[21:23] <nbd> anyway, we can exclude 2.4 from the generic config stuff
[21:23] <nbd> and just write that for 2.6
[21:23] <nbd> makes things easier
[21:24] <nbd> we should be able to do it without breaking 2.4 or doing much extra work around its weirdness
[21:24] <florian__> any new developper would like to join the package porting effort ?
[21:32] <florian__> humm au1000-2.6 is broken
[21:32] <h3sp4wn> florian: Does it just entail rewriting the Makefiles in the buildroot-ng style ? (I could do that)
[21:32] <florian__> almost
[21:33] <florian__> by the way, your yafc package is not available to me :( server #500
[21:33] <h3sp4wn> That server has been decommisioned unfortunately
[21:34] <florian__> ah no pb
[21:36] <h3sp4wn> What is the other thing / things it entails ?
[21:37] <florian__> that's almost it
[21:40] <malbon> florian__: are you talking about /packages ?
[21:47] <[mbm]> ip
[21:49] <florian__> yep
[21:53] <malbon> florian__: ok. quick question, how do I make packages compile (no I've not tried yet)
[21:53] <malbon> florian__: I am thinking of giving some packages a go, but buildroot-ng confuses me... ;)
[21:53] <nbd> malbon: what part confuses you?
[21:54] <malbon> nbd: how /packages relates to my tree
[21:55] <nbd> it doesn't relate to that tree (yet)
[21:56] <malbon> so I just port them into my normal buildroot-ng tree, but check them into packages?
[21:56] <nbd> yes
[21:56] <malbon> (after testing of course) :P
[21:56] <nbd> of course
[21:56] <malbon> ok, that's not too hard.
[21:57] <nbd> in the future i want to have a script that resolves dependencies and copies packages from /packages into the buildroot
[21:57] <malbon> yes, something that does it automagically would be good.
[21:57] <nbd> shouldn't be too hard to write
[21:58] <nbd> we only need to come up with a proper categorization scheme for the stuff in /packages/
[21:58] <nbd> i already checked some libraries into libs/
[21:58] <nbd> i think we can keep that
[21:59] <malbon> yes I saw. I was going to give things like hostapd and chillispot a go.
[21:59] <CIA-4> nbd * r4009 /branches/buildroot-ng/openwrt/ (20 files in 7 dirs): remove the wl.o build from the kernel patch and move the driver along with wl, nas and wlc into packages/broadcom-wl
[22:05] <tziOm> may I ask you about webif.. is this something developed for this project? can You point me to sources?
[22:06] <nbd> tziOm: whiterussian svn
[22:07] <tziOm> is it possible to co only webif?
[22:10] <florian__> yep
[22:17] <florian__> whiterussian is still feature freeze right ?
[22:21] <nbd> depends
[22:22] <nbd> the plan is to bring the current tree into a useful state for a 0.9 release
[22:22] <CIA-4> nbd * r4010 /branches/buildroot-ng/openwrt/target/linux/image/brcm/Makefile: fix broadcom trx alignment
[22:23] <florian__> ok I just wanted to close the nfs-kernel-server ticket
[22:23] <nbd> make it a package for buildroot-ng
[22:23] <nbd> that should be enough
[22:23] <florian__> right
[22:30] <CIA-4> nbd * r4011 /branches/buildroot-ng/openwrt/package/broadcom-wl/Makefile: add modules.d file for wl
[22:31] <CIA-4> nbd * r4012 /branches/buildroot-ng/openwrt/package/wlcompat/Makefile: fix wlcompat depends/default
[22:48] <nbd> cool. the wl.o 'apsta' mode actually works
[22:48] <nbd> i managed to set it up manually with wlc
[22:48] <nbd> now i just need to finish my magic shell scripting code to configure that thing
[22:49] <[mbm]> wait, you can actually be ap and sta at once?
[22:49] <nbd> yes
[22:49] <[mbm]> nice
[22:49] <florian__> lol whoa
[22:49] <nbd> but it's a special mode, so you can't just set arbitrary modes for virtual interfaces
[22:50] <nbd> and this time when i changed several runtime parameters with wlc, it didn't even oops or spit out weird things
[22:50] <nbd> so the newer one from linksys seems to be a bit more reliable than the last one
[22:50] <malbon> does it have vif's with different mac addresses?
[22:51] <nbd> unfortunately not
[22:51] <malbon> shame
[22:51] <nbd> which makes it really hard to use different ssids with different encryption on windows
[22:51] <nbd> because the windows 802.11 stack is too stupid to handle that
[22:51] <florian__> can you configure the ethernet address with ifconfig hw ether ?
[22:51] <nbd> florian__: no, this limitation is by design
[22:51] <nbd> florian__: it doesn't keep track of different bssids for different interfaces
[22:52] <florian__> ok
[22:52] <florian__> even under linux, I scan intermitently 1 or the other ssid
[22:52] <malbon> it's like horse riding 100 miles to shoot yourself in the foot. crazy
[22:52] <nbd> it's broadcom
[22:53] <nbd> what do you expect? :)
[22:53] <malbon> a reasonable ammount of workyness would be nice.
[22:53] <nbd> hehe
[22:53] <nbd> well, maybe the open source driver can get that feature first
[22:53] <nbd> :)
[22:53] <nbd> i know for sure that the template ram can send out beacons with different bssids
[22:53] <malbon> hopefully. I think it probably will.
[22:53] <florian__> they are doing great progress on newer mimo dirvers
[22:54] <nbd> because if you hit a weird case in the driver where it doesn't configure the template ram, you will get a BRCM_TEST_SSID on a different bssid
[22:54] <nbd> that reminds me ... i should change the broadcom-wl package so that it also builds the mimo version
[22:58] <[mbm]> is it just me or is there very little work done on making linux into an embeded solution?
[22:58] <[mbm]> seems like anything other than an x86 arch, or anything involving cross compiling is just asking for problems
[22:59] <nbd> right
[23:00] <florian__> [mbm]: what makes you think that ?
[23:01] <[mbm]> florian__: experience?
[23:02] <florian__> [mbm]: what would you recommend for embedded systems ?
[23:11] <CIA-4> nbd * r4013 /branches/buildroot-ng/openwrt/package/broadcom-wl/ (Makefile src/kmod/Makefile): build mimo version of the broadcom wl driver as well
[23:45] <CIA-4> nbd * r4014 /branches/buildroot-ng/openwrt/package/base-files/default/etc/functions.sh: add a list append function to functions.sh
[23:52] <[mbm]> grr
[23:52] <[mbm]> new kenrel forgot -ffreestanding
[23:52] <[mbm]> which causes some strcpy errors
[00:00] --- Tue Jun 20 2006