[00:00] <Bartman007> I was going to order one when they were 30USD, missed my chance though.
[00:02] <{Nico}> Bartman007: 40 on wallmart, alas they don't ship to international
[00:02] <Bartman007> {Nico}: where are you located?
[00:03] <{Nico}> France
[00:05] <Bartman007> hmmm, their website doesn't show any airlink products...
[00:08] <groz> mbm, are you using dhcp ?
[00:09] <mbm> groz: yep
[00:09] <groz> you dhcp client ups the adapter on a new lease, that breaks the association, gotta re-auth
[00:09] <groz> it probably doesn't do that on a dhcp renew
[00:09] <mbm> no, something else is wrong -- wpa_supplicant isn't even starting anymore
[00:10] <groz> hmm, ya, you need openwrt desktop edition too
[00:10] <groz> ok, i gotta run, time to go make some holes in the sky
[00:10] <groz> be back either late tomorrow nite, or, sometime tuesday
[00:11] <{Nico}> groz: have nice flights
[00:12] <mbm> I hate the chain of hotplug scripts and dynamically generated config files that make ubuntu a pain to debug
[00:12] <nbd> heh, chain of hotplug scripts, dynamically generate config files ... reminds me of openwrt :)
[00:12] <florian_91_44> lol
[00:12] <nbd> but that's mostly my fault :)
[00:13] <florian_91_44> have to go sleeping see you tomorrow
[00:13] <nbd> see you tomorrow
[00:14] <mbm> yeah, and I just debugged the openwrt networking scripts :P
[00:14] <mbm> and that was an absolute pain in the ass -- there isn't even any documentation for that
[00:14] <db90h> http://www.bitsum.com/images/wl530g.png <- on that image, the solder points on the bottom look like maybe they might be JTAG? is there an easy way to determine which ones are whcih.. trial and error maybe? I'm actually looking for a serial port... many of hte ones on the bottom are running at 3.3v, that's about all I know. The two above the CPU I haven't probed, for fear I short something while trying to probe them
[00:15] <db90h> by which ones are which, i mean if they are JTAG, whcih go to which EJTAG signal line..
[00:16] <db90h> maybe trace them to the CPU inputs? .. I am not experienced with electronics at all, just know what little I've done with the wrt54g(s)'s..
[00:16] <mbm> ugh, who's been mangling the resistors in that pic?
[00:16] <db90h> that's the FCC photo
[00:16] <db90h> not mine, thankfully ;)
[00:16] <Bartman007> lol
[00:16] <mbm> jtag and serial go standard headers, not resistor banks
[00:17] <db90h> standard headers?
[00:17] <nbd> mbm: iirc on some devices you can only find them on resistor banks
[00:17] <mbm> db90h: it'll look like a set of jumpers (or holes to put them)
[00:17] <db90h> am i at least right that those circular brown/yellow sports are solder points?
[00:17] <nbd> no idea
[00:18] <db90h> mbm: unfortunatelly, there is no such thing on this device.. i've examined it closely in detail.. i reckon they didn't have room, it's a pocket router
[00:18] <mbm> the yellow pads are test points, the brown and yellow crud is flux from someone's bad soldering job
[00:19] <db90h> yea, i reckon the units they send the FCC are development units or something
[00:19] <db90h> i've seen such things on other FCC pics
[00:19] <mbm> test points are usually only used to measure the across components and verify the volts/ohms
[00:19] <db90h> darn
[00:20] <db90h> well those test points are the only hope i see
[00:20] <db90h> so i guess there isn't much hope ;)
[00:20] <mbm> yeah, fcc gets a lot of early prototype units that have had extra wires added and parts moved as part of the early debugging
[00:21] <mbm> I've seen plenty of boards with very thin wires added to patch bugs in the board
[00:21] <db90h> oh well, thanks for looking.. i am going to have to give up on a serial port i think.. i wouldn't want to try to solder on wires to those test points anyway
[00:22] Action: Exiles would if he had one
[00:22] <db90h> lol, you're always soliciting for free shit Exiles
[00:25] <Exiles> :)
[00:25] <Exiles> can you blame me? I'm willing to "try" stuff in return for Free stuff
[00:25] <db90h> someday maybe someone will make your day ;)
[00:25] <db90h> well i'm off .. cyaz
[01:18] <{Nico}> 'night all
[03:37] <Bartman007> if anyone wants to have a go at it, buildroot-ng is breaking on ldso during the uClibc compilation... www.padded-cell.net/openwrt/armeb_uclibc_error
[03:37] <Bartman007> if not, I'll submit the output in a bug report
[05:12] <[mbm]> florian_91_44: I remembered what the dnsmasq error was
[05:12] <[mbm]> there was a case where the isp used multiple nameservers
[05:13] <[mbm]> and one of the nameservers was either misconfigured or only resolved certain domains
[05:13] <[mbm]> dnsmasq would query that server and cache the fact that the domain didn't exist
[05:13] <[mbm]> without even consulting the other server to see if it could find the name
[08:59] <florian_91_44> [mbm]: thanks for the dnsmasq tip
[09:02] <florian> [mbm]: this option "no-negcache" should be set by default don't you think so ?
[09:06] <[mbm]> if it works
[09:07] <florian> sure if it works
[09:07] <florian> by the way [mbm] do you know how if bootloader arguments are passed the same way for any bootloader or are they (logically) bootloader specific ?
[09:09] <florian> from a prom_init point of view
[09:09] <[mbm]> many of the platforms have the kernel boot args hardcoded into the kernel
[09:10] <florian> actually, my problem is that with 2.4 kernels, prom_init prototype was prom_init(int argc, char **argv), that is no longer the case with 2.6 kernel, but I don't find where this code has been moved
[09:11] <florian> all the more as I found cmdline code but that platform and/or bootloader specific
[09:11] <[mbm]> ;)
[09:12] <florian> which bootloader runs on aruba apxx ?
[09:13] <[mbm]> cfe
[09:14] <[mbm]> hmm well it depends on the platform
[09:14] <[mbm]> it's all modified and branded as apboot
[09:15] <florian> ok, sounds normal to me
[10:25] <CIA-17> florian * r4693 /packages/net/wdiag/ (. Makefile): Add wdiag to -ng, closes #709
[10:41] <florian> [mbm]: what are fw_arg0 in the generic mips code ?
[10:42] <florian> arch/mips/mips-boards/generic/init.c
[10:43] Action: [mbm] doesn't have a copy of the code handy
[10:44] <florian> ok no pb
[10:44] <florian> I just wonder if it is generic code or board specific actually
[11:33] <florian> nbd: ping ?
[11:51] <nbd> pong
[11:52] <florian> nbd: ok to commit ? http://pastebin.ca/151701 ? this would seriously ease kmod test
[11:55] <nbd> hmm... i'd prefer if it was committed when it's actually working
[11:55] <florian> ok no pb
[11:56] <florian> I was just wondering if we have the right to do this kind of stuff ? Well it's basically the same thing as for the wl binary and module
[11:58] <nbd> no idea
[12:12] <[mbm]> think that the header on line 10 should be adjusted .. the makefile might be gpl but broadcom's driver isn't
[14:22] <florian> ok
[15:49] <florian> nbd: do you know what bcmprocfs is used for ?
[15:49] <florian> with bcm63xx boards
[15:50] <nbd> no idea
[15:51] <florian> I think it provides entry in /proc for the proprietary devices
[19:15] <Bartman007> Kaloz: my uClibc issue was resolved.
[21:19] <Bartman007> armeb wpa_supplicant issue. www.padded-cell.net/openwrt/wpa_supplicant_error
[21:57] <florian__> nbd: I have made a tarball which contains all binary modules for bcm63xx, including atm, wifi, enet, usb ...
[21:57] <florian__> nothing important yet, just to ease testing on my side
[21:57] <nbd> ok
[21:58] <nbd> which sdk version is it from?
[22:00] <florian__> comtrend CT-536B+-A101-302JAZ-C01_R05-open-source.tar.gz
[22:00] <florian__> I should a FROM in this tarball
[22:02] <CIA-17> florian * r4694 /branches/whiterussian/openwrt/package/quagga/ (5 files in 3 dirs): Add quagga isisd, closes #735
[22:02] <db90h> is there any technical reason why renice isn't included in openwrt's busybox?
[22:03] <Bartman007> most people don't need to renice their processes on a router.
[22:04] <db90h> right, i was just curious if that was the only reason for its ommission.. i'll infer so from your response, thanks
[22:05] <Bartman007> db90h: that was just a guess since I'm not on the dev team.
[22:12] <florian__> nbd: I have fixed up bootloader detection, this just works :)
[22:12] <nbd> great
[22:13] <nbd> so does mtd work again?
[22:15] <florian__> humm not yet
[22:16] <florian__> but I have not flashed the box with a fresh jffs2 system
[22:16] <florian__> so I just wonder which partition to use, but it should not be a problem so far
[22:17] <florian__> good news, the atmapi module loads just ouf of the box
[22:18] <florian__> same thing for bcmprocfs
[22:19] <florian__> I will commit the bootloader detection code back
[00:00] --- Tue Aug 29 2006
[00:02] <{Nico}> Bartman007: 40 on wallmart, alas they don't ship to international
[00:02] <Bartman007> {Nico}: where are you located?
[00:03] <{Nico}> France
[00:05] <Bartman007> hmmm, their website doesn't show any airlink products...
[00:08] <groz> mbm, are you using dhcp ?
[00:09] <mbm> groz: yep
[00:09] <groz> you dhcp client ups the adapter on a new lease, that breaks the association, gotta re-auth
[00:09] <groz> it probably doesn't do that on a dhcp renew
[00:09] <mbm> no, something else is wrong -- wpa_supplicant isn't even starting anymore
[00:10] <groz> hmm, ya, you need openwrt desktop edition too
[00:10] <groz> ok, i gotta run, time to go make some holes in the sky
[00:10] <groz> be back either late tomorrow nite, or, sometime tuesday
[00:11] <{Nico}> groz: have nice flights
[00:12] <mbm> I hate the chain of hotplug scripts and dynamically generated config files that make ubuntu a pain to debug
[00:12] <nbd> heh, chain of hotplug scripts, dynamically generate config files ... reminds me of openwrt :)
[00:12] <florian_91_44> lol
[00:12] <nbd> but that's mostly my fault :)
[00:13] <florian_91_44> have to go sleeping see you tomorrow
[00:13] <nbd> see you tomorrow
[00:14] <mbm> yeah, and I just debugged the openwrt networking scripts :P
[00:14] <mbm> and that was an absolute pain in the ass -- there isn't even any documentation for that
[00:14] <db90h> http://www.bitsum.com/images/wl530g.png <- on that image, the solder points on the bottom look like maybe they might be JTAG? is there an easy way to determine which ones are whcih.. trial and error maybe? I'm actually looking for a serial port... many of hte ones on the bottom are running at 3.3v, that's about all I know. The two above the CPU I haven't probed, for fear I short something while trying to probe them
[00:15] <db90h> by which ones are which, i mean if they are JTAG, whcih go to which EJTAG signal line..
[00:16] <db90h> maybe trace them to the CPU inputs? .. I am not experienced with electronics at all, just know what little I've done with the wrt54g(s)'s..
[00:16] <mbm> ugh, who's been mangling the resistors in that pic?
[00:16] <db90h> that's the FCC photo
[00:16] <db90h> not mine, thankfully ;)
[00:16] <Bartman007> lol
[00:16] <mbm> jtag and serial go standard headers, not resistor banks
[00:17] <db90h> standard headers?
[00:17] <nbd> mbm: iirc on some devices you can only find them on resistor banks
[00:17] <mbm> db90h: it'll look like a set of jumpers (or holes to put them)
[00:17] <db90h> am i at least right that those circular brown/yellow sports are solder points?
[00:17] <nbd> no idea
[00:18] <db90h> mbm: unfortunatelly, there is no such thing on this device.. i've examined it closely in detail.. i reckon they didn't have room, it's a pocket router
[00:18] <mbm> the yellow pads are test points, the brown and yellow crud is flux from someone's bad soldering job
[00:19] <db90h> yea, i reckon the units they send the FCC are development units or something
[00:19] <db90h> i've seen such things on other FCC pics
[00:19] <mbm> test points are usually only used to measure the across components and verify the volts/ohms
[00:19] <db90h> darn
[00:20] <db90h> well those test points are the only hope i see
[00:20] <db90h> so i guess there isn't much hope ;)
[00:20] <mbm> yeah, fcc gets a lot of early prototype units that have had extra wires added and parts moved as part of the early debugging
[00:21] <mbm> I've seen plenty of boards with very thin wires added to patch bugs in the board
[00:21] <db90h> oh well, thanks for looking.. i am going to have to give up on a serial port i think.. i wouldn't want to try to solder on wires to those test points anyway
[00:22] Action: Exiles would if he had one
[00:22] <db90h> lol, you're always soliciting for free shit Exiles
[00:25] <Exiles> :)
[00:25] <Exiles> can you blame me? I'm willing to "try" stuff in return for Free stuff
[00:25] <db90h> someday maybe someone will make your day ;)
[00:25] <db90h> well i'm off .. cyaz
[01:18] <{Nico}> 'night all
[03:37] <Bartman007> if anyone wants to have a go at it, buildroot-ng is breaking on ldso during the uClibc compilation... www.padded-cell.net/openwrt/armeb_uclibc_error
[03:37] <Bartman007> if not, I'll submit the output in a bug report
[05:12] <[mbm]> florian_91_44: I remembered what the dnsmasq error was
[05:12] <[mbm]> there was a case where the isp used multiple nameservers
[05:13] <[mbm]> and one of the nameservers was either misconfigured or only resolved certain domains
[05:13] <[mbm]> dnsmasq would query that server and cache the fact that the domain didn't exist
[05:13] <[mbm]> without even consulting the other server to see if it could find the name
[08:59] <florian_91_44> [mbm]: thanks for the dnsmasq tip
[09:02] <florian> [mbm]: this option "no-negcache" should be set by default don't you think so ?
[09:06] <[mbm]> if it works
[09:07] <florian> sure if it works
[09:07] <florian> by the way [mbm] do you know how if bootloader arguments are passed the same way for any bootloader or are they (logically) bootloader specific ?
[09:09] <florian> from a prom_init point of view
[09:09] <[mbm]> many of the platforms have the kernel boot args hardcoded into the kernel
[09:10] <florian> actually, my problem is that with 2.4 kernels, prom_init prototype was prom_init(int argc, char **argv), that is no longer the case with 2.6 kernel, but I don't find where this code has been moved
[09:11] <florian> all the more as I found cmdline code but that platform and/or bootloader specific
[09:11] <[mbm]> ;)
[09:12] <florian> which bootloader runs on aruba apxx ?
[09:13] <[mbm]> cfe
[09:14] <[mbm]> hmm well it depends on the platform
[09:14] <[mbm]> it's all modified and branded as apboot
[09:15] <florian> ok, sounds normal to me
[10:25] <CIA-17> florian * r4693 /packages/net/wdiag/ (. Makefile): Add wdiag to -ng, closes #709
[10:41] <florian> [mbm]: what are fw_arg0 in the generic mips code ?
[10:42] <florian> arch/mips/mips-boards/generic/init.c
[10:43] Action: [mbm] doesn't have a copy of the code handy
[10:44] <florian> ok no pb
[10:44] <florian> I just wonder if it is generic code or board specific actually
[11:33] <florian> nbd: ping ?
[11:51] <nbd> pong
[11:52] <florian> nbd: ok to commit ? http://pastebin.ca/151701 ? this would seriously ease kmod test
[11:55] <nbd> hmm... i'd prefer if it was committed when it's actually working
[11:55] <florian> ok no pb
[11:56] <florian> I was just wondering if we have the right to do this kind of stuff ? Well it's basically the same thing as for the wl binary and module
[11:58] <nbd> no idea
[12:12] <[mbm]> think that the header on line 10 should be adjusted .. the makefile might be gpl but broadcom's driver isn't
[14:22] <florian> ok
[15:49] <florian> nbd: do you know what bcmprocfs is used for ?
[15:49] <florian> with bcm63xx boards
[15:50] <nbd> no idea
[15:51] <florian> I think it provides entry in /proc for the proprietary devices
[19:15] <Bartman007> Kaloz: my uClibc issue was resolved.
[21:19] <Bartman007> armeb wpa_supplicant issue. www.padded-cell.net/openwrt/wpa_supplicant_error
[21:57] <florian__> nbd: I have made a tarball which contains all binary modules for bcm63xx, including atm, wifi, enet, usb ...
[21:57] <florian__> nothing important yet, just to ease testing on my side
[21:57] <nbd> ok
[21:58] <nbd> which sdk version is it from?
[22:00] <florian__> comtrend CT-536B+-A101-302JAZ-C01_R05-open-source.tar.gz
[22:00] <florian__> I should a FROM in this tarball
[22:02] <CIA-17> florian * r4694 /branches/whiterussian/openwrt/package/quagga/ (5 files in 3 dirs): Add quagga isisd, closes #735
[22:02] <db90h> is there any technical reason why renice isn't included in openwrt's busybox?
[22:03] <Bartman007> most people don't need to renice their processes on a router.
[22:04] <db90h> right, i was just curious if that was the only reason for its ommission.. i'll infer so from your response, thanks
[22:05] <Bartman007> db90h: that was just a guess since I'm not on the dev team.
[22:12] <florian__> nbd: I have fixed up bootloader detection, this just works :)
[22:12] <nbd> great
[22:13] <nbd> so does mtd work again?
[22:15] <florian__> humm not yet
[22:16] <florian__> but I have not flashed the box with a fresh jffs2 system
[22:16] <florian__> so I just wonder which partition to use, but it should not be a problem so far
[22:17] <florian__> good news, the atmapi module loads just ouf of the box
[22:18] <florian__> same thing for bcmprocfs
[22:19] <florian__> I will commit the bootloader detection code back
[00:00] --- Tue Aug 29 2006