**** BEGIN LOGGING AT Wed Aug 04 02:59:57 2010 Aug 04 04:50:32 build #64 of ps3 is complete: Failure [failed compile_6] Build details are at http://tksite.gotdns.org:8010/builders/ps3/builds/64 Aug 04 07:24:57 nbd: latest ath9k on my 1043nd runs fine so far. then again, the previous versions were fine too Aug 04 07:30:39 hello Aug 04 07:32:33 <_trine> has the switch been fixed ? Aug 04 07:34:19 when did it broke? Aug 04 07:34:36 <_trine> oh don't start that Aug 04 08:01:47 _trine: nope Aug 04 08:01:52 _trine being worked on Aug 04 08:03:18 <_trine> ta Aug 04 08:03:41 <_trine> is jushog back yet Aug 04 08:10:42 did he go away ? Aug 04 08:13:16 <_trine> yes I understand he has been on vacation Aug 04 08:14:01 <_trine> but that's something that's been mentioned to me so I have not had it from the horses mouth so to speak Aug 04 08:37:01 hi all. I have a Mikrotik RB-750G I'd really like to load openwrt on but it's not yet supported (afaict). Is anybody working on this device atm? Is there anything I can do to help? I believe it's a cross between an RB750 and RB450G, both of which are apparently working with openwrt trunk. Aug 04 10:00:58 _trine: i'll commit the fix today Aug 04 10:01:07 hopefully it'll take care of the regression you were seein Aug 04 10:04:35 nbd: \o/ Aug 04 10:04:41 even tomorrow is ok :) Aug 04 10:04:49 blogic: he's talking about a different switch Aug 04 10:05:20 but i want to take care of the one that you're using today as well ;) Aug 04 10:06:46 lol Aug 04 10:06:48 even better Aug 04 10:07:01 already did some cleanups on swconfig yesterday Aug 04 10:10:51 hcg * r22477 /trunk/target/linux/omap35xx/gumstix/base-files/lib/wifi/mac80211.sh: omap35xx: Add temporary fix for marvell wlan MAC address detection Aug 04 10:24:45 nbd: are you currently working on the ar8216 driver? Aug 04 10:28:20 i'm still thinking about what to do with it Aug 04 10:28:31 at some point i want to split ar8216 and ar8316 stuff Aug 04 10:28:43 because i think the ar8216 specific workarounds are just too messy to push upstream Aug 04 10:28:49 whereas ar8316 can be cleaned up properly Aug 04 10:38:12 hi Aug 04 10:39:53 nbd: except for the routerboard-workaround (which is probably also needed for the fritzbox 7390, if openwrt ever supports it - there its also the ethernet/switch driver which initializes the switch) Aug 04 10:40:17 i'll see if i can decode those values using the docs Aug 04 10:46:14 ah, great Aug 04 10:50:29 what soc is in fritzbox 7390 ? Aug 04 11:00:54 blogic: ikanos fusiv vx180, according to freetz Aug 04 11:20:53 Is anyone working to get openwrt running on mikrotik RB750G? Is there anything I can do to help/kickstart it? Aug 04 11:25:38 hcg * r22478 /trunk/target/linux/omap35xx/gumstix/defconfig.gumstix: gumstix: Change defconfig Aug 04 11:36:47 eremos: get a serial connection, try to boot a self compiled openwrt image (best through netboot or similar), post logs (especially if it's not working) ;) Aug 04 11:37:13 trouble is this device lacks a serial port Aug 04 11:37:21 afaik, there is noone working on a rb750g Aug 04 11:37:40 so I'm not sure how they got the rb750 working as that's the same but without the gigabit Aug 04 11:37:48 not even pads on the board? Aug 04 11:38:10 it's got a couple of pads but I don't know how that works Aug 04 11:38:15 or if it's really JTAG Aug 04 11:38:55 as far as I can tell the 750g is a cross between a 750 and a 450g Aug 04 11:39:13 someone's posted the kernel patches that includes drivers for 750g Aug 04 11:39:18 just needs to be integrated .. I think Aug 04 11:39:53 would a kernel boot at all if the hardware is unrecognised? Aug 04 11:40:47 afaict it's got most of the same hardware as 750 and 450g so I would think it should boot Aug 04 12:01:53 eremos: the kernel would boot, but it wouldn't setup the ethernet interface(s), so you couldn't tell wether it sucessfully booted or not Aug 04 12:02:35 don't suppose you know who wrote the 750 support and how it was done? Aug 04 12:02:44 by hacking in a serial port or did they just get lucky and it booted? :) Aug 04 12:03:48 no, sorry Aug 04 12:04:44 what if I bought one to send to the devs? Aug 04 12:04:51 it's on the list of requested devices Aug 04 12:06:16 is the squashfs image created by the openwrt build system a standard squashfs image? if I try to mount it using mount openwrt-atheros-root.squashfs ./dir/ -t squashfs -o loop - it fails Aug 04 12:27:08 nico * r22479 /packages/libs/libtorrent/Makefile: [packages] libtorrent: apply fix from [22233] to armeb as well Aug 04 12:31:00 <{Nico}> cshore: ping Aug 04 12:33:00 <{Nico}> cshore: FS mod_fax still broken on ppc due to build failure in spandsp >> http://openwrt.pastebin.com/diyAEPT1 Aug 04 12:35:17 acoul * r22480 /packages/net/bird/Makefile: net/bird: update bird to version 1.2.4 Aug 04 12:35:55 <{Nico}> cshore: and FS fails on x86 now due to unsupported "-fgnu89-inline" option in gcc-4.1.2 >> http://openwrt.pastebin.com/a6GLgjGz Aug 04 12:36:43 <{Nico}> cshore: i'll try to fix both so that we have FS for all targets in 10.03.1 Aug 04 12:40:28 why are we still using gcc-4.1.2 ? Aug 04 12:43:53 eremos: it should help, but doesn't guarantee success. working without serial or jtag access really sucks Aug 04 12:44:09 I would imagine Aug 04 13:26:40 lars * r22481 /trunk/target/linux/xburst/patches-2.6.35/054-mmc.patch: [xburst] Fix mmc card detection Aug 04 13:29:13 acoul * r22482 /trunk/package/iptables/ (Makefile patches/030-no-libnfnetlink.patch): package/iptables: upgrade to iptables version 1.4.9 Aug 04 13:31:48 lars * r22483 /trunk/target/linux/xburst/ (5 files in 4 dirs): [xburst] Adjust codec driver module name for 2.6.35 Aug 04 13:32:35 lars * r22484 /trunk/target/linux/xburst/Makefile: [xburst] Set -march=mips32 Aug 04 13:33:59 lars * r22485 /trunk/target/linux/xburst/ (6 files in 6 dirs): [xburst] Drop 2.6.32 patches Aug 04 14:07:15 nico * r22486 /packages/net/freeswitch/Makefile: [packages] freeswitch: remove ozmod_ss7_boost from mod-openzap (no build anymore), adjust copyright, bump release number Aug 04 16:16:45 hm, did somebody recently try to boot 47xx? Aug 04 16:17:03 It gets to "Please be patient, while OpenWrt loads ..." and hangs there Aug 04 16:21:52 current stable boots Aug 04 16:38:52 but it doesn't seem to work in STA mode Aug 04 16:40:51 it simply doesn't fire up wpa_supplicant for whatever reason. Aug 04 16:41:26 config is correct. That worked on the other device until it had a hardware failure Aug 04 16:42:45 odd Aug 04 16:42:49 wpad-mini is installed? Aug 04 16:43:03 (or similar) Aug 04 16:45:32 I could not find any wpad package, so I installed wpa-supplicant Aug 04 16:46:22 is this the stable release or custom build? Aug 04 16:46:32 kamikaze latest Aug 04 16:46:47 trunk does not boot at all Aug 04 16:47:00 odd Aug 04 16:47:35 try sh -x /sbin/wifi > log.txt 2>&1 and then throw that file into a pastebin Aug 04 16:48:22 http://pastebin.org/447623 Aug 04 16:48:58 http://pastebin.org/447624 Aug 04 16:49:27 http://pastebin.org/447625 Aug 04 16:49:47 ah Aug 04 16:49:53 this is the proprietary wl Aug 04 16:49:58 i was expecting b43 or something Aug 04 16:50:14 that one's not supposed to be using wpa_supplicant Aug 04 16:50:16 it needs 'nas' Aug 04 16:50:20 ah Aug 04 16:50:30 Well stable does not have b43, right? Aug 04 16:50:35 it does Aug 04 16:51:10 is that a simple package or how does that work? Aug 04 16:51:31 yes, it's a package. i thought it was in the default images in backfire Aug 04 16:51:44 though not in the 10.03 release Aug 04 16:51:54 only in the branch Aug 04 16:52:04 ah wait. backfire is latest stable? Aug 04 16:52:10 yes Aug 04 16:52:27 what did you flash? Aug 04 16:52:28 Ok. There really really needs to be a hint in the download section to what is latest. Aug 04 16:52:32 kamikaze Aug 04 16:52:49 ah, so the two year old stuff Aug 04 16:53:04 front page shows 10.03 as the newest one Aug 04 16:53:06 so does the channel topic Aug 04 16:54:14 Well, the subdirs have a LATEST-IS hint Aug 04 16:55:05 It's really annoying if you need the fucking device working (which is the case for me) and I first have to read stuff to find out what is latest and whatever Aug 04 16:55:44 RTAI also only uses release names for their releases and it is a pain to find out what the hell is the latest or even next to last release. Aug 04 16:56:03 yeah Aug 04 16:59:00 ah ok that boots b43 by default. That's nice Aug 04 17:09:18 hmpf, so now it correctly authenticates to the AP, but it fails to fetch an IP address via udhcpc Aug 04 17:09:26 udhcpc is running, though Aug 04 17:09:36 jow * r22487 /trunk/package/dnsmasq/files/dnsmasq.init: Aug 04 17:09:36 [PATCH] Allow UCI interface names in /etc/config/dhcp Aug 04 17:09:36 Signed-off-by: Gabriel Kerneis Aug 04 17:12:51 jow * r22488 /packages/utils/usb-modeswitch-data/Makefile: Aug 04 17:12:52 [PATCH] [usb-modeswitch-data] Update to latest available debian source. Aug 04 17:12:52 This fixes the build of usb-modeswitch-data since the older version was no longer available at the specified URL. Aug 04 17:12:52 Signed-off-by: Ithamar R. Adema Aug 04 17:14:16 jow * r22489 /packages/net/dsl-qos-queue/Makefile: Aug 04 17:14:16 [PATCH] [dsl-qos-queue] Add libipq to DEPENDS, and reference shared library, since the static library is no longer being built. Aug 04 17:14:16 This fixes the build of this package and makes it work again. Aug 04 17:14:16 Signed-off-by: Ithamar R. Adema Aug 04 17:15:32 Crap I'm gonna use a patchcable now. This does not work at all. Aug 04 17:16:50 jow * r22490 /trunk/package/dnsmasq/files/dnsmasq.init: [package] dnsmasq: add -c or --cache-size option to dnsmasq via cachesize UCI option (#7625) Aug 04 17:18:58 jow * r22491 /packages/utils/lm-sensors/Makefile: Aug 04 17:18:58 [PATCH] lm-sensors package upgrade/update Aug 04 17:18:58 Update the lm-Sensors Makefile to new the release (3.1.2). Enable Aug 04 17:18:58 package selection on other platforms. Aug 04 17:18:58 Signed-off-by: Luka Perkov Aug 04 17:22:01 jow * r22492 /packages/utils/joe/Makefile: Aug 04 17:22:01 [PATCH] update joe package, support for all emulation modes. Aug 04 17:22:01 UPDATE: Aug 04 17:22:01 - include support for all emulation modes jmacs jpico etc. Aug 04 17:22:01 - see patches/002-builtinrc.patch Aug 04 17:27:03 nbd * r22493 /trunk/package/mac80211/patches/550-ath9k_tsf_fix.patch: ath9k: fix rx tsf parsing Aug 04 17:28:35 Does somebody else run backfire in STA mode with DHCP enabled? Aug 04 17:29:10 nbd * r22494 /branches/backfire/package/mac80211/ (16 files in 2 dirs): mac80211: merge changes from latest trunk into backfire Aug 04 17:29:34 mb__: I used to do, but not anymore Aug 04 17:29:53 backfire branch is likely to work much better than backfire release Aug 04 17:30:09 mb__: there is a problem with it, as soon as you apply settings in luci (or by running wifi), it will loose it's IP and never get it back Aug 04 17:30:27 yeah, i noticed the same Aug 04 17:30:32 the problem is that udhcpc is not killed properly Aug 04 17:30:41 I sent a patch for it to openwrt-devel about 2 weeks ago Aug 04 17:30:44 and somebody changed stuff so that it's not restarted properly if it's running already Aug 04 17:30:52 Well, my problem is that udhcpc is running correctly but it is unable to fetch the information for whatever reason. Aug 04 17:30:59 Wireless is correctly authenticated Aug 04 17:31:04 * nbd looks at the list Aug 04 17:31:39 nbd: [PATCH] base-files: kill existing udhcpc instance Aug 04 17:31:48 nbd: sent 19/07 Aug 04 17:31:54 yeah, found it Aug 04 17:32:32 not sure if that's the best way, but it worked for me Aug 04 17:33:15 Ok, I will start with checking out backfire SVN and building it. That will take some time... Aug 04 17:34:07 stintel: looks good, i'll do some tests with that Aug 04 17:34:24 nbd: ok Aug 04 18:18:28 nbd your last ath9k fix ... what exactly is it for (don't get smart from the commit log ...) Aug 04 18:20:45 its for fixing rx tsf parsing obviously Aug 04 18:20:49 * xMff runs... Aug 04 18:26:00 lol Aug 04 18:26:15 mhh there is something strange with the uboot-envtools ... Aug 04 18:26:53 i've patched them to also eat binarydumps of uboot config partitions ... now i added an environment variable via uboot direct and then via fw_setenv on the binary dump ... Aug 04 18:27:30 after that the binary_dump is readable by fw_printenv but the mtd config is not anymore! ... there seems something different in the crc calculation *grr* Aug 04 18:27:37 at least on dockstar Aug 04 18:32:00 ahh forgot something in my simulation *hrhr* ... xMff do you know with what a nandflash sector is overwritten on erase? 0xff? Aug 04 18:32:58 yes, FF iirxc Aug 04 18:33:10 nbd: Ok, I'm running SVN now. And it also does not work. It's really strange. If I config a static IP address I'm also unable to ping the AP. Aug 04 18:33:24 than i should do this as well on the binary dump ... *hehe Aug 04 18:33:27 It correctly authenticates with WPA-PSK, though, nbd Aug 04 18:34:16 nbd: There's some serious breakage or misconfiguration in the networking. Aug 04 18:35:50 nbd: It only affects the wireless, though. If I put the cable into the switch, it starts to work. Aug 04 18:36:02 The wireless is bridged with the switch Aug 04 18:36:10 in sta mode? Aug 04 18:37:07 yeah Aug 04 18:37:14 try to unbridge it Aug 04 18:37:26 Well, I need it bridged for my application Aug 04 18:37:44 erm, bridged client mode with b43 is not possible Aug 04 18:37:53 with all mac80211 drivers in fact Aug 04 18:38:34 ehm, why? Aug 04 18:38:41 802.11 limitation Aug 04 18:38:55 it works with proprietary broadcom Aug 04 18:39:02 yes, proprietary broadcom does arp-nat Aug 04 18:39:36 ok, then back to proprietary... Aug 04 18:39:37 and stinks therefore! Aug 04 18:39:39 ;) Aug 04 18:40:12 This really sucks ass. I just wanted to print a sheet of paper and it took me 5 hours so far due to brokeness... Aug 04 18:40:49 So, the application is: I need to connect a LAN printer to the WLAN. Is there any way to do that without bridging? Aug 04 18:41:08 use a routed setup with multiple subnets Aug 04 18:41:21 maybe with some snat/dnat trickery if no broadcast is involved Aug 04 18:41:28 or yes, simple routing Aug 04 18:42:46 Well, put the printer into another net and reconfigure the whole net routing? Doesn't smell too inviting. Aug 04 18:43:30 then just use some brcm-2.4 release Aug 04 18:44:12 i did it when migrating from broadcom (was a arp-nat setup aswell) to mac80211 ... it is not too big i think - depending on your net Aug 04 18:45:54 Well, I recently reconfigured the whole net to bridging. It was a horrible mess routing with multiple subnets before. I don't really wanna go back there. Aug 04 18:46:14 I guess I could live with one additional subnet for the printer, though. Aug 04 18:47:53 aaargh, and the build machine is unstable, too. Aug 04 18:48:28 xMff: is wds an option here? Aug 04 18:48:35 depends on the ap Aug 04 18:50:26 mb__: i have that running with wds Aug 04 18:50:35 if you are in control of the ap.. Aug 04 18:52:13 Yeah hm Aug 04 18:55:29 loswillios: I do not want the device on the printer to act as an AP, though. Is that possible with WDS? Aug 04 18:55:36 then it's easy Aug 04 18:55:38 yes Aug 04 18:56:05 if ap and sta are mac80211 (b43, ath9k, ...) you can use ap-sta wds Aug 04 18:56:06 implementing a decent replacement for the client bridging crap is on my todo list Aug 04 18:56:23 if both are brcm-2.4 you need ap2ap wds Aug 04 18:56:32 mb__: https://forum.openwrt.org/viewtopic.php?id=16615 Aug 04 18:56:39 of one is brcm-2.4 and the other mac80211, no wds is possible Aug 04 18:56:54 leave out the client ap part Aug 04 18:57:52 if sta is brcm-2.4 you can use the client bridge, regardless of the ap Aug 04 18:57:56 Yeah this is a sane mac80211 architecture. Aug 04 18:58:07 ;-) Aug 04 18:58:48 you basically just configure normal ap-sta (as you had already?) and add "option wds 1" on both ap and sta Aug 04 18:58:58 what xMff says Aug 04 18:59:22 this should enable bridging Aug 04 18:59:24 loswillios: So does it require any change on the AP at all? The config in the link looks pretty standard-AP Aug 04 18:59:27 on the client side Aug 04 18:59:39 mb__: "option wds 1" Aug 04 19:00:02 ah. You dunno how the option is called in hostapd? :) Aug 04 19:00:07 a regular 802.11 STA does not transmit the source MAC address. and standard APs don't accept 4-address-frames from clients Aug 04 19:00:17 mb__: wds_sta=1 Aug 04 19:00:30 cool, thanks Aug 04 19:00:40 it'll create a separate interface for each incoming wds sta Aug 04 19:00:45 and put that into the bridge Aug 04 19:00:56 dunno if all of my hostapd patches have been merged by jouni yet, though Aug 04 19:03:50 nbd: Ah, I had an old hostapd config template that didn't have the WDS stuff, yet. Therefore I didn't find it ;) Aug 04 19:19:42 loswillios: Hm, doesn't seem to work. But maybe my hostapd is too old. Aug 04 19:20:35 you need a somewhat recent mac80211 as well Aug 04 19:20:56 the client bridge replacement will probably be ready next week or so Aug 04 19:21:10 it'll use proxyarp + routing + broadcast relay Aug 04 19:21:36 along with a slight mangling of dhcp packet to force the dhcp server to send responses as broadcast frames Aug 04 19:21:54 nbd: How recent? will .35 do? Aug 04 19:21:57 yeah Aug 04 19:22:01 k Aug 04 19:22:10 Was going to update anyway Aug 04 19:22:31 at least i think it should be enough, the important bugfixes were made a long time ago Aug 04 19:22:43 (unless you're using 11n, then you probably need something newer) Aug 04 19:22:49 ok, yeah. I'm currently running .33.something Aug 04 19:22:50 i only use bleeding edge when it comes to wireless ;) Aug 04 19:23:06 that's usually more stable than the releases Aug 04 19:23:29 yeah I used to run compat for quite some time. Aug 04 19:23:45 But at some point stable simply worked and I didn't see a reason for installing compat Aug 04 19:24:21 with ath9k that process took a little longer Aug 04 19:24:38 but i think it's finally stable (aside from some incompatibilities with intel stuff) Aug 04 19:24:53 I'm running ath5k on the AP Aug 04 19:25:22 haven't really tested that much yet Aug 04 19:25:50 i still run my madwifi fork on my legacy equipment Aug 04 19:26:11 :) Aug 04 19:27:13 though i don't use much legacy equipment any more these days Aug 04 19:27:29 i like the performance of 802.11n Aug 04 19:27:59 Hm Wikipedia says that WDS does only work with WEP and not WPA. Is that true? Aug 04 19:28:15 there is no real standard for WDS Aug 04 19:28:20 there are some variants that don't support crypto Aug 04 19:28:28 but mac80211 just does AP and STA with 4-address-frames Aug 04 19:28:33 that works with crypto Aug 04 19:28:39 Ah, right Aug 04 19:28:59 at some point i'll also fix up the other wds mode Aug 04 19:29:12 i don't think anybody has really tested this yet Aug 04 19:29:17 at least last time i checked it was broken Aug 04 19:29:21 that was a long time ago Aug 04 19:34:09 hey nbd ... are this rtl8366xx you did today important to me? ^^ Aug 04 19:34:28 no Aug 04 19:34:30 just cleanup Aug 04 19:34:40 no functional changes Aug 04 19:35:49 kk Aug 04 19:49:40 fuuuuuuuuuuuuu.... I was wondering why tarball extract takes ages: Aug 04 19:49:42 mb@quimby:~/kernel$ tar xj linux-2.6.35.tar.bz2 Aug 04 19:50:00 j vs. f :) Aug 04 20:37:35 what's a good way to connect to test points (tiny vias). I should buy some kind of clip or something so I can use the serial on the new device I got (it's smaller than the palm of my hand, and has usb and RJ45 and wireless...it's by monoprice and I want to see what's inside it) Aug 04 20:39:43 cshore: Well, any reason why not solder wires to it? Aug 04 20:40:07 it's tiny, I'm concerned about damaging it attempting it Aug 04 20:40:30 the vias are smaller than pin sized Aug 04 20:40:42 Well, I would be more concerned by damaging it with mechanical clamps... Aug 04 20:41:12 Are these real vias or pads? Aug 04 20:41:59 they're through the board, but you they're small - you couldn't put a header pin though it, it's about half or less that size Aug 04 20:43:18 yeah ok. Well. I doubt you could possibly mechanically contact them somehow with clamps... Aug 04 20:43:48 If they are gold plated, it's trivial to solder. Use leaded (!) solder. It's trivial with that. Aug 04 20:45:31 And as wire don't take normal wire, but that fine wire used in coils Aug 04 20:45:54 hmmmm.....my soldering skills are really pathetic and this is a micro-size board...right, definitely can't use regular wire Aug 04 20:46:49 Well, practice on another board that you don't need anymore. Aug 04 20:46:52 I wish I had one of those pencil (i.e. really small tip) soldering irons Aug 04 20:46:55 yeah, good idea Aug 04 20:47:18 yeah you need a pencil tip. It will be hard with a screwdrivertip or bigger Aug 04 20:48:01 yeah, I'll have to buy one...need to look up prices and availability...any good places to buy from? Aug 04 20:48:36 I'm german and I buy that stuff from reichelt.dde Aug 04 20:48:38 .de Aug 04 20:49:19 I would recommend a temperature-regulated soldering station. They are expensive, but it's way harder to damage traces by overheating. Aug 04 20:49:41 I have a rather cheap one, though. Aug 04 20:49:57 Something in the 50 EUR range. Aug 04 20:50:58 yeah, it looks like it's time to bite the bullet and get something decent. I've been making do with an old Radio Shack soldering iron, but that was for regular sized boards Aug 04 20:52:55 good equipment is half the skill ;) Aug 04 20:56:30 i finally got all my connections soldered last night. now i got a serial and jtag port :) Aug 04 20:58:01 time to get the router running openwrt :) Aug 04 20:58:20 at least functioning :) Aug 04 21:44:14 nbd irgendwas am trunk hat die ath9k performance gedrückt meiner meinung nach Aug 04 21:44:49 leider kann ich das nicht mehr genauer analysieren ... ab morgen ist meine bessere hälfte wieder da und dann ist erstmal für 3 wochen ruhe mit geeken :) Aug 04 21:44:58 damn wrong language *G* Aug 04 21:46:33 nbd forget it ... i think i'm unable to use iperf correct ... transfering a big file is as fast as ever ... Aug 04 21:47:32 jow * r22495 /trunk/package/ppp/ (3 files in 2 dirs): [package] pppd: only remove own default route on shutdown (#7694) Aug 04 21:48:09 jow * r22496 /branches/backfire/package/ppp/ (3 files in 2 dirs): [backfire] merge r22495 Aug 04 21:50:51 jow * r22497 /trunk/package/ppp/patches/330-retain_foreign_default_routes.patch: [package] ppp: whitespace cleanup in previous commit Aug 04 21:51:27 jow * r22498 /branches/backfire/package/ppp/patches/330-retain_foreign_default_routes.patch: [backfire] merge r22497 Aug 05 00:11:19 ping ;p Aug 05 01:03:20 cshore, ping :) Aug 05 01:03:34 i think it was you who told me how to test gpio before? Aug 05 02:02:46 Realopty: pong Aug 05 02:02:57 cshore, :) Aug 05 02:06:52 what's up? Aug 05 02:08:37 gpio. how do i find out my LEDs gpio numbers? Aug 05 02:09:05 if you don't have a datasheet you drive each gpio and see which ones light up Aug 05 02:09:19 how about for reset button? Aug 05 02:09:24 on some devices that can cause a short circuit Aug 05 02:09:29 :| Aug 05 02:09:32 same thing, but dirin instead of dirout Aug 05 02:09:56 I did it to my device without problems Aug 05 02:10:13 but I can't guaranteed it's not going to be one for you Aug 05 02:10:17 the dd-wrt source code says that the gpio for power is 3 Aug 05 02:10:24 im about to test it ;p Aug 05 02:10:30 indeed Aug 05 02:10:30 ah, then try that one :) Aug 05 02:10:43 if you have info, use it! Aug 05 02:10:50 for sure Aug 05 02:10:52 much better than random testing Aug 05 02:10:59 gpioctl dirin 3 Aug 05 02:11:08 its just that dd-wrt code is so diffrent ;p Aug 05 02:11:13 do that twice (if the first one gives a warning) Aug 05 02:11:16 how do i get that package installed on openwrt? Aug 05 02:11:19 gpio numbers don't change Aug 05 02:11:27 opkg install? Aug 05 02:11:34 package name sorry lol Aug 05 02:11:40 gpioctl Aug 05 02:11:44 lol TY Aug 05 02:12:07 btw so far ive made some progess on porting this device to openwrt. :) Aug 05 02:12:12 resolved the WAN bug. Aug 05 02:12:14 excellent Aug 05 02:12:40 and im now able to flash from factory firmware with out having to flash dd-wrt first :D Aug 05 02:13:40 found an opensource code that will generate the headers for a .chk file (now i dont have to use the binary that came with the factory firmware. that didnt even wanna work!) Aug 05 02:13:42 I might actually pick your brains soon....I've never actually done a port...I just figured out imagetags for my device, but florian did the boarids and stuff (though I know how to do that now, I think) Aug 05 02:13:51 excellent Aug 05 02:14:33 well, I figured out imagetags for other devices too (that the .chk generator) Aug 05 02:14:44 and the gpios on a couple of devices Aug 05 02:14:48 cshore, do you know where the file location is for the file that says "trx makethis/image" or "make a bin"? Aug 05 02:15:05 brb Aug 05 02:15:36 you mean the firmware-tools (or is it firmware-utils?) in /tools ? Aug 05 02:19:15 cshore, yes and no Aug 05 02:19:22 ive placed the source in that folder Aug 05 02:19:43 but theres a script somewhere that generates the *.bins Aug 05 02:20:15 you mean the makefile that calls one of the programs in /tools/firmware-tools? Aug 05 02:20:28 yup Aug 05 02:20:42 ah, its probably in a makefile Aug 05 02:21:05 /target/linux//image Aug 05 02:21:24 sorry /target/linux//image/Makefile Aug 05 02:26:18 cshore, thanks. e Aug 05 02:26:22 -e Aug 05 02:26:35 -e? Aug 05 02:26:57 didnt mean to put e in the last sentence ;p Aug 05 02:27:04 oh delete e...I was thinking command line switch Aug 05 02:27:06 lol Aug 05 02:27:28 cshore, do you know the file that says make /tools/frimware-utils/trx ? Aug 05 02:28:17 /tools/firmware-tools/Makefile or /tools/firmware-tools/src/Makefile (don't remember exactly) Aug 05 02:28:44 kk Aug 05 02:29:21 yup thats the one Aug 05 02:29:25 cshore, thanks so much :) Aug 05 02:29:28 np Aug 05 02:29:43 at least someone is getting useful work done today Aug 05 02:30:14 lol :) Aug 05 02:33:49 I hate wall warts Aug 05 02:35:15 why can't all devices take a regular plug and if a transformer thingy is needed, put it somewhere other than at the end of the power cable where it screws with your power bar / UPS space Aug 05 02:35:27 Aug 05 02:36:41 I'd like the give the inventor of the wall wart a good smack upside the head ! *mutter* *mutter* Aug 05 02:37:54 exactly. that annoys me too Aug 05 02:38:14 and what if u need to mod the transformer!?! Aug 05 02:38:28 then its a real pain Aug 05 02:38:41 mod the tranformer? that just sounds like a bad idea Aug 05 02:39:09 :P sometimes parts fail. Aug 05 02:39:13 ah, right Aug 05 02:39:32 cshore, ok so im trying to tell openwrt to make this .chk file Aug 05 02:39:35 Transformers are one thing I'm not brave enough to mess with Aug 05 02:39:37 and im quite stuck Aug 05 02:39:37 ok Aug 05 02:39:43 what's happening? Aug 05 02:39:56 cause i know nothing about makefiles (lol) Aug 05 02:39:59 anyways Aug 05 02:40:20 take a look into trunk_netgear/target/linux/brcm47xx/image Aug 05 02:40:52 i think that i need to add the line into : Aug 05 02:41:11 define Image/Build Aug 05 02:41:14 under that Aug 05 02:41:38 yes Aug 05 02:41:50 ah okies :) Aug 05 02:42:12 $(STAGING_DIR_HOST)/bin/mkchk Aug 05 02:42:24 :) thought so. just wasnt too sure Aug 05 02:42:55 but umm do u think i need to do somekinda make clean for the firmware-utils? so it will compile the mkchkimg ? Aug 05 02:43:05 ive added the code to the makefile Aug 05 02:43:28 you will need Aug 05 02:43:29 make tools/firmware-utils/{clean,compile,install} Aug 05 02:43:38 with the actual {} Aug 05 02:44:14 just do the line you said, yeah? Aug 05 02:44:20 yep Aug 05 02:44:27 badass Aug 05 02:46:17 out of curiousity are you younger (i.e. 20ish or less)? (I'm an old fart, but the way you talk doesn't sound like my generation) Aug 05 02:47:18