**** BEGIN LOGGING AT Wed Apr 27 02:59:58 2011 Apr 27 03:58:29 build #17 of sibyte is complete: Success [build successful] Build details are at http://buildbot.openwrt.org:8010/builders/sibyte/builds/17 Apr 27 04:12:58 build #18 of pxcab is complete: Failure [failed compile_3] Build details are at http://buildbot.openwrt.org:8010/builders/pxcab/builds/18 Apr 27 04:31:13 hmmm....no luci....maybe defconfig is broken Apr 27 04:58:29 hmm. i keep getting asked for system type (ATHEROS_AR231X) during kernel config. Even though CONFIG_ATHEROS_AR231X=y. /me is confused. Apr 27 05:36:28 cshore: there shouldn't be any on a "pass through". Apr 27 05:58:37 build #19 of cobalt is complete: Failure [failed compile_6] Build details are at http://buildbot.openwrt.org:8010/builders/cobalt/builds/19 Apr 27 06:05:49 build #17 of au1000 is complete: Failure [failed shell shell_4 compile_6] Build details are at http://buildbot.openwrt.org:8010/builders/au1000/builds/17 Apr 27 06:43:38 i think i just ran into the bridge + ipv6 thing i've overheard people talking about Apr 27 06:50:12 russell--: annoying, isn't it? Apr 27 06:55:24 yes Apr 27 06:56:15 so, i was trying to run a 2.6.38.4 kernel, basically didn't have a bridge, wlan's didn't come up or were confused, etc. Apr 27 06:58:09 as a workaround compile ipv6 into the image (see my rfc patch on the list ;) Apr 27 06:58:51 i tried that (via make kernel_menuconfig) and the resulting image kernel panic'd for me ;) Apr 27 06:59:31 * russell-- goes to look at KanjiMonster's patch Apr 27 07:01:21 I have to admit, I didn't test 2.6.38, just 2.6.39 Apr 27 07:02:32 * russell-- doesn't mind a kernel panic, but it will require i get off the couch and go upstairs to power cycle ;-) Apr 27 07:07:02 i'm going to try 2.6.37.6 first ... seems like i saw something like this earlier, it might be unrelated. Apr 27 07:07:37 the ipv6/bridge thing was only introduced in 2.6.38, so anything earlier should be fine Apr 27 07:09:33 right Apr 27 07:09:56 so if 2.6.37.6 panics, then it's something else Apr 27 07:13:52 yeah, 2.6.37.6 panic'd too Apr 27 07:13:55 build #21 of s3c24xx is complete: Failure [failed compile_10] Build details are at http://buildbot.openwrt.org:8010/builders/s3c24xx/builds/21 Apr 27 07:14:06 maybe something batman-adv related Apr 27 07:22:37 http://pastebin.ca/2051119 Apr 27 07:23:51 * russell-- wonders why i'm seeing "Disabling lock debugging due to kernel taint" ... since madwifi went away, i don't know why there would be taint. Apr 27 07:24:40 russell--: you should enable the symbol table thingy for the kernel so the stack trace is actually readable ;) Apr 27 07:32:08 russell--: staging drivers also taint the kernel (the disabling of debugging isn't intended though; it's a bug) Apr 27 07:38:22 i recall seeing an option to enable a stack trace somehow without including the symbols in the image ... Apr 27 07:39:50 * russell-- found the symbol thing first ;-) Apr 27 07:41:03 heading to work now, bbl Apr 27 07:41:14 ok, thanks! Apr 27 07:52:29 aha, looks like maybe adhoc + ap related? http://pastebin.ca/2051127 Apr 27 07:52:44 nbd might want to look at that? Apr 27 07:53:43 hm Apr 27 07:54:29 not sure if this is an issue with ath5k or yet another brcm47xx stability issue Apr 27 07:54:37 my gut feeling says it's the latter Apr 27 07:55:20 oh, the other difference here is that i'm using a cm9 radio Apr 27 07:55:26 and 802.11a Apr 27 07:55:40 what did you use before? Apr 27 07:56:06 well, with the same version, i'm *not* seeing a problem with the stock radio on 2.4ghz Apr 27 07:56:55 the first time i saw this (or something very much like it) was r26738 Apr 27 07:57:00 i don't really trust the wgt634u at all Apr 27 07:57:06 the data bus error crap looks too familiar Apr 27 07:57:15 it showed up with madwifi a lot as well Apr 27 07:57:47 i've been happily running r26738 on stock wgts Apr 27 07:58:02 with ath5k, adhoc+ap, etc Apr 27 07:58:15 maybe only some cards are incompatible with this device Apr 27 07:58:19 pretty much the same configuration except a different radio and a different band Apr 27 07:58:27 yeah, possible Apr 27 07:58:59 what chipset does a stock wgt have? Apr 27 07:59:01 for wireless Apr 27 07:59:13 i think the cm9 has 5213A Apr 27 07:59:29 ath5k phy0: Atheros AR5213A chip found (MAC: 0x59, PHY: 0x43) Apr 27 07:59:40 that's stock Apr 27 07:59:52 hm, interesting Apr 27 08:00:07 maybe a power related issue Apr 27 08:00:48 from the panic log, with the cm9, i see: ath5k phy0: Atheros AR5213A chip found (MAC: 0x59, PHY: 0x43) Apr 27 08:01:32 i've never seen that before, was working fine with a slightly earlier rev of openwrt with madwifi Apr 27 08:03:01 ok, maybe more than slightly: r25265 with madwifi was working Apr 27 08:03:37 (roughly the end of january) Apr 27 08:09:34 * russell-- is trying with madwifi, otherwise the same Apr 27 08:19:12 yeah, madwifi works Apr 27 08:19:22 (for varying definitions of works) Apr 27 08:19:29 doesn't panic anyway Apr 27 08:25:43 going to try a more traditional configuration with ath5k Apr 27 08:28:58 okay, batman-adv and adhoc+ap is off the hook Apr 27 08:29:26 looking more like something that ath5k doesn't like about that radio Apr 27 08:30:47 * russell-- packs it in for nwo Apr 27 08:30:48 now Apr 27 08:53:18 russell--: unfortunately the trace doesn't give any indication as to why it crashes Apr 27 08:53:33 it looks like something is making either the card or the pci bus unstable Apr 27 08:53:52 and that makes it crash at a later point in time Apr 27 09:49:07 xMff: "- - config_get ieee80211d "$vif" ieee80211d" <- pgp-inline signing patches doesn't really work well (recent patch versions are supposed to be able to cope with it though) ;) Apr 27 10:12:41 KanjiMonster: true - I forgot Apr 27 12:22:08 build #18 of orion is complete: Failure [failed compile_7] Build details are at http://buildbot.openwrt.org:8010/builders/orion/builds/18 Apr 27 13:02:42 Kaloz: https://dev.openwrt.org/ticket/738 is fixed with https://dev.openwrt.org/changeset/24214/ (this is "your" ticket ;) Apr 27 13:03:27 KanjiMonster: hm? Apr 27 13:03:47 neither the ticket nor the changeset mention him anywhere Apr 27 13:04:12 ah, five years ago... Apr 27 13:04:42 yeah Apr 27 13:26:42 jow_laptop: should I close tickets I see are already fixed myself or should I notify the ticket owner (or you, if there is none)? Apr 27 13:27:15 just close them Apr 27 13:27:44 ok Apr 27 13:49:14 jow_laptop: I'll bug you a bit more: I see 4 open tickets for the wan/lan leds of the wr741nd, and the second newest is assigned to juhosg - close the others as duplicates, or close all except the oldest and reassign that one to juhosg? Apr 27 13:50:04 make them all a dupe to the altest Apr 27 13:50:07 *latest Apr 27 13:52:02 latest = newest? Apr 27 13:52:15 well the one already assigned Apr 27 13:52:29 ah, okay Apr 27 13:55:00 done Apr 27 13:55:06 amazing Apr 27 13:55:21 leds seem to be the most important feature of this unit Apr 27 13:57:16 as usual, only what's visible counts Apr 27 13:58:25 :) Apr 27 13:59:07 I usually disable the leds Apr 27 13:59:17 no need for bright flashy blue light in my bedroom Apr 27 14:01:36 yeah those blue leds are more annoying than others Apr 27 14:45:25 KanjiMonster, hey Apr 27 14:45:58 hi RealOpty Apr 27 14:46:07 did you get your bcm4321 to work? Apr 27 14:46:12 nope Apr 27 14:46:34 its saying that it cant assign address Apr 27 14:46:44 i think its talking about the MAC address Apr 27 14:46:59 some reason it detects the MAC as all 00s Apr 27 14:47:29 ah, I've seen that one before Apr 27 14:47:49 there's one or two tickets with all-zeroes mac addresses Apr 27 14:47:52 i have dmesg and hostapd info on other box, ill grab it Apr 27 14:48:04 b43 rejects all zeroes mac addresses Apr 27 14:48:13 yeah Apr 27 14:48:26 the problem is probably somewhere in the nvram->sprom parsing code Apr 27 14:48:32 i set the mac in the config to the one it should be Apr 27 14:48:42 then it won't find the phy Apr 27 14:48:48 ^ Apr 27 14:48:52 kept saying that :( Apr 27 14:49:09 brb Apr 27 14:51:30 KanjiMonster, http://paste.ubuntu.com/599817/ Apr 27 14:51:53 ssb: WARNING: Invalid SPROM CRC (corrupt SPROM) Apr 27 14:52:03 maybe thats something to do with it also?? Apr 27 14:52:54 KanjiMonster, also getting this too lol Apr 27 14:52:54 b43-phy0 ERROR: DMA RX reset timed out Apr 27 14:56:44 will_: did you modify the /etc/modules.d/30-b43 to be "b43 pio=1"? else it will always load with pio disabled Apr 27 14:56:57 but the mac address is probably the bigger problem Apr 27 14:57:21 (or the broken sprom) Apr 27 14:57:45 but I gotta go now, bbl Apr 27 14:58:35 KanjiMonster, thanks for the tip ttyl Apr 27 15:17:06 build #18 of uml is complete: Failure [failed compile_10] Build details are at http://buildbot.openwrt.org:8010/builders/uml/builds/18 Apr 27 15:20:43 build #18 of ar7 is complete: Failure [failed compile_6] Build details are at http://buildbot.openwrt.org:8010/builders/ar7/builds/18 Apr 27 15:21:40 nvram show Apr 27 15:21:53 RealOpty: ^ Apr 27 15:22:10 are the vars prefixed with pci/1/1/ or something? Apr 27 15:42:10 build #18 of sibyte is complete: Failure [failed compile_6] Build details are at http://buildbot.openwrt.org:8010/builders/sibyte/builds/18 Apr 27 15:53:23 xMff, yes most nvram has pci/1/1 Apr 27 15:53:41 is there a var containing a macaddr ? Apr 27 15:53:52 yes Apr 27 15:53:52 il0macaddr or et0macaddr Apr 27 15:54:06 pci/1/1/il0macaddr ? Apr 27 15:54:09 let me fire it up and check Apr 27 15:56:25 xMff, nope just a pci/1/1/macaddr=00:22:3f:7c:4b:76 Apr 27 15:56:36 hmm Apr 27 15:58:04 xMff, to get the wireless to show up in lspci i had to set cardbus=0 Apr 27 15:59:37 pci/1/1/sromrev=4 ??? Apr 27 16:00:05 please try this: Apr 27 16:00:39 nvram set il0macaddr=00:22:3f:7c:4b:76 set et0macaddr=00:22:3f:7c:4b:77 set il1macaddr=00:22:3f:7c:4b:78 commit; reboot Apr 27 16:00:56 check whether it changes anything Apr 27 16:02:36 b43-phy0 ERROR: DMA RX reset timed out Apr 27 16:02:37 ssb: Timeout waiting for bitmask 01800000 o Apr 27 16:02:37 n register 0F90 to clear. Apr 27 16:03:16 hmm Apr 27 16:04:00 i just noticed that after changing the cardbus setting the router model isnt detected by diag Apr 27 16:04:32 what model is that again? Apr 27 16:04:57 wnr834B v2 Apr 27 16:05:41 yeah Apr 27 16:05:54 diag.c uses carbus=1 to identify the unit Apr 27 16:06:05 if (!strcmp(boardtype,"0x0472") && !strcmp(getvar("cardbus"), "1")) { /* Netgear WNR834B V1 and V2*/ Apr 27 16:06:36 ^ Apr 27 16:06:42 just was reading it heh Apr 27 16:07:15 maybe my boarddata got erased and i need to restore it Apr 27 16:07:27 unlikelyx Apr 27 16:07:38 the wifi sprom is populated from nvram vars Apr 27 16:07:52 and yours are in an unusual format Apr 27 16:08:31 https://dev.openwrt.org/browser/trunk/target/linux/brcm47xx/patches-2.6.37/011-MIPS-BCM47xx-Fill-values-for-b43-into-ssb-sprom.patch?rev=24162 Apr 27 16:11:33 xMff, Ill be back in a few hours, will you be around? Apr 27 16:11:42 probably Apr 27 16:12:34 cool Apr 27 16:12:40 g2g wife is bugging me... Apr 27 16:27:20 build #14 of mpc52xx is complete: Failure [failed compile_4] Build details are at http://buildbot.openwrt.org:8010/builders/mpc52xx/builds/14 Apr 27 17:18:31 acoul * r26772 /trunk/target/linux/atheros/ (7 files in 2 dirs): linux/atheros: add 2.6.38 preliminary support Apr 27 17:34:42 build #22 of at91 is complete: Failure [failed compile_4] Build details are at http://buildbot.openwrt.org:8010/builders/at91/builds/22 Apr 27 17:49:10 cshore: anything? Apr 27 17:57:27 build #22 of ubicom32 is complete: Failure [failed compile_3] Build details are at http://buildbot.openwrt.org:8010/builders/ubicom32/builds/22 Apr 27 18:45:22 Hi, i need a little help with my custom module in my build. Apr 27 18:46:37 I have read what i could and have downloaded the OpenWRT source for my WRT54GL and have figured out how to add a module, but its not there after flashing the router? Apr 27 18:47:16 If i do a lsmod i see that it is installed. Apr 27 18:53:07 anyone here? Apr 27 18:56:17 "but its not there after flashing the router" ... "If i do a lsmod i see that it is installed" Apr 27 18:56:23 so whats the problem? Apr 27 18:57:25 Basically, i tried to add tcpdump as a module to make a custom build which looked successful. Apr 27 18:57:50 But after flashing the router with the custom build, it does not have tcpdump? Apr 27 18:58:13 you marked it as M ? Apr 27 18:58:25 No i marked it as a * Apr 27 18:58:51 you must have done something wrong then Apr 27 18:59:11 there is not more involved than menuconfig, mark as * and make Apr 27 18:59:20 I agree, but i did not see any error messages. Apr 27 18:59:29 that does not mean there werent any Apr 27 18:59:40 maybe your rootfs grew too large and the images in bin where not really updated Apr 27 18:59:55 check the timestamps Apr 27 19:00:06 Possibly. Apr 27 19:00:23 I see the files in the Bin folder of my build directory. Apr 27 19:00:51 But the only files i see in the Router are tcpdump.lst files. Apr 27 19:02:25 Ok, thanks, i will check size and have another go at it ;) Apr 27 19:02:34 I mean the bin/ folder in your buildroot Apr 27 19:03:19 Yes, i think you were referring to the size of the Tcpdump file before it tried to pull it into the Firmware build, right? Apr 27 19:03:37 I mean the final size of the firmware iamge Apr 27 19:03:56 if it exceeds a certain max size the image utils refuse to generate one Apr 27 19:04:13 I think the final size was about just over 2 MB Apr 27 19:04:31 I turned off a bunch of the additional stuff to save on space. Apr 27 19:06:23 just checked and the size is about 2 MB, when my unit has a 4 MB flash size. Apr 27 19:09:17 thanks :) Apr 27 19:17:02 jow * r26773 /trunk/package/relayd/files/relay.sh: [package] relayd: map "ipaddr" option to local ip (-L) Apr 27 19:17:39 jow * r26774 /branches/backfire/package/relayd/files/relay.sh: [backfire] merge r26773 Apr 27 19:57:10 hi friends, Is there a way to have max power of a 500mW card using madwifi? It's legal where I want to use it. Apr 27 19:59:52 I just can set max 19dBm on it Apr 27 20:18:29 xMff, still around? Apr 27 20:18:37 yep Apr 27 20:19:01 sweet Apr 27 20:19:12 so have any ideas on some more testing? Apr 27 20:19:50 whats the current state? Apr 27 20:20:08 DMA errors incorrect mac Apr 27 20:20:18 I did not really follow your endeavor Apr 27 20:20:25 kk Apr 27 20:20:25 dmesg ? Apr 27 20:20:39 i need to get it plug back in Apr 27 20:20:40 brb Apr 27 20:22:29 tornado: there is a power offset for higher powered cards in madwifi Apr 27 20:23:14 russell--, how can I set it? Apr 27 20:24:28 http://www.ubnt.com/downloads/SRXR_power_offset.pdf Apr 27 20:24:40 what radio do you have? Apr 27 20:26:07 I'm using a router station pro with an ar5414 based mini pci and openwrt OS Apr 27 20:28:03 tornado: which ar5414-based radio? Apr 27 20:29:40 russell--, sorry I can't remember the mini pci card brand Apr 27 20:30:00 russell--, should I do any change in source or there's a way to change it from console? Apr 27 20:30:18 did you read the link? Apr 27 20:31:08 russell--, yes it's just talking about the offset values, not about how to change them Apr 27 20:31:35 you set the lower value Apr 27 20:31:43 the radio is using a higher value Apr 27 20:32:00 e.g. you say 16 dBm, it uses 26 dBm Apr 27 20:32:18 make sense? Apr 27 20:32:43 you mean always it's doing it? with all cards? Apr 27 20:33:09 xMff, dmesg: http://paste.ubuntu.com/599979/ Apr 27 20:33:57 with those cards documented in the link Apr 27 20:34:47 russell--, let me check my card brand and model Apr 27 20:35:14 http://madwifi-project.org/wiki/Compatibility/Ubiquiti Apr 27 20:35:22 RealOpty: did you try to load b43 in pio mode? Apr 27 20:35:35 xMff, yes Apr 27 20:35:43 i will now make it default Apr 27 20:36:05 tornado: a 5414 might be an XR2 or XR5 (depending on the band) Apr 27 20:36:14 I wonder why it still tries dma Apr 27 20:36:35 russell--, I'm using a dbii networks card (F50-pro) Apr 27 20:37:00 it might use a similar offset scheme Apr 27 20:37:10 xMff, that dmesg wasnt using pio=1 Apr 27 20:37:26 tornado: hook it up to a spectrum analyzer and find out! Apr 27 20:38:06 russell--, yes, that was the last option ;) Apr 27 20:38:32 thanks for help Apr 27 20:38:55 russell--: have you considered trying ath5k on the cards instead of mwifi Apr 27 20:39:52 OutBackDingo: i have been, i think you are directing that remark to tornado instead ;-) Apr 27 20:39:52 OutBackDingo, that's a great option if you don't need turbo mode ;) Apr 27 20:40:53 OutBackDingo, I'm the fan of ath5k. but now I need turbo mode :) Apr 27 20:42:00 xMff, http://paste.ubuntu.com/599983/ Apr 27 20:42:28 dmesg with pio=1 via /etc/modules.d/30-b43 Apr 27 20:42:50 that looks better Apr 27 20:43:32 xMff, i think it has something to do with sprom broken somewhere Apr 27 20:43:38 i get this message: Could not set interface wlan0 flags: Cannot assign requested address Apr 27 20:43:39 did you check whether the mac in the wireless config matches the one from nvram ? Apr 27 20:43:44 the mac it set was all 00s Apr 27 20:44:23 ^- yes, i also set it and then of course it says that 'no PHY' Apr 27 20:45:38 ok Apr 27 20:45:47 hm Apr 27 20:47:08 xMff, shouldnt lines 8-9 and 80-81 be similar ? (dmesg) Apr 27 20:48:02 ssb: Core 4 found: PCI (cc 0x804, rev 0x08, vendor 0x4243) Apr 27 20:48:13 but on line 82 it says core 4 is USB ?????????? Apr 27 20:48:14 I think its two backplanes Apr 27 20:48:18 chained Apr 27 20:48:49 xMff, is that common? Apr 27 20:48:58 don't think so Apr 27 20:49:03 but its not unseen either Apr 27 20:49:09 ah ok Apr 27 20:49:45 ssb: Sonics Silicon Backplane found at address 0x18000000 Apr 27 20:49:45 ssb: Sonics Silicon Backplane found on PCI device 0000:00:01.0 Apr 27 20:49:56 xMff, to make the wireless card show up on lspci i had to 'nvram set cardbus=0' Apr 27 20:49:58 so the first is the system backplane providing pci Apr 27 20:50:06 and the second is another backplane on a pci device Apr 27 20:50:17 ah okie Apr 27 20:50:42 is it common to have to change that nvram setting on some brcm47xx devices? Apr 27 20:50:48 no Apr 27 20:52:05 xMff, if i had the original source code (well what all they decided to no leave out) would that assist any? Apr 27 20:52:23 probably not Apr 27 20:52:37 did you try boradcom-wl yet ? Apr 27 20:53:16 mmm let me look in the ticket, i think i did but there wasnt support for it Apr 27 20:53:28 I think there was at one point but they dropped it out Apr 27 20:54:10 openwrt brcm47xx trunk has a wl for kernel 2.6 Apr 27 20:55:15 xMff, has it always been there or added within the last 6 months? Apr 27 20:55:32 RealOpty: the ssb bus on the pci bus probably does not have a sprom like normal pci ssb devices Apr 27 20:56:00 this would explain why it has 00:00... as mac address Apr 27 20:56:10 Hauke: his nvram has vars prefixed with pci/1/1/ ... I think the SPROM has to be synthesized from those Apr 27 20:56:43 RealOpty: can you paste a full nvram dump? Apr 27 20:57:10 xMff: yes I think so too Apr 27 20:57:58 wonder how thgat can be added to SSB in an upstream-suitable manner Apr 27 20:58:52 ah wait, there is the fallback sprom facility already Apr 27 20:59:28 so parsing the pci/1/1/... vars has to be done in the platfrom setup code, then construct an ssb_sprom struct and pass that to ssb_arch_set_fallback_sprom() Apr 27 21:00:18 great, I think this is actually doable with not so much code Apr 27 21:01:05 maybe we can rework bcm47xx_fill_sprom() to accept a prefix Apr 27 21:01:11 is ssb complaining about beeing unable to parse the sprom on the pci bus? Apr 27 21:01:15 then we can reuse the "parsing" code as well Apr 27 21:01:19 yes Apr 27 21:01:28 my browser does not work because of an system upgrade ;-) Apr 27 21:01:51 ssb: SPROM offset is 0x1000 ssb: WARNING: Invalid SPROM CRC (corrupt SPROM) ssb: SPROM revision 0 detected. ssb: Unsupported SPROM revision 0 detected. Will extract v1 Apr 27 21:01:55 http://paste.ubuntu.com/600002/ <- nvram dump (sorry bout all the dd-wrt crap) Apr 27 21:02:33 probably similar to that ticket -> https://dev.openwrt.org/ticket/7272 Apr 27 21:02:56 indeed Apr 27 21:03:24 just before that error message it calls fallback = ssb_get_fallback_sprom(); Apr 27 21:03:39 yep, we just have to provide one Apr 27 21:03:45 similar to how brcm63xx does it Apr 27 21:04:13 xMff: yes Apr 27 21:04:54 RealOpty: do you run unmodified openwrt right now? Apr 27 21:05:04 RealOpty: or is it using the rt-n16 tickets? Apr 27 21:05:09 erm patches Apr 27 21:05:16 from the ticket Apr 27 21:05:19 fresh svn no patches of mine Apr 27 21:06:01 no patches from any tickes, just a fresh build Apr 27 21:06:12 okay Apr 27 21:06:51 hmm the cardbus thing is problematic Apr 27 21:07:04 xMff, speaking of Apr 27 21:07:12 if you set it to 0 you will also disable the board data protection in the flashmap Apr 27 21:08:00 yes, i think its only still working correctly cause it was =1 when i flashed, and it doesnt erase to make free space every time Apr 27 21:08:17 I suggest to set it back to 1 Apr 27 21:08:23 indeed Apr 27 21:08:38 then find out what triggers the difference in pci handling Apr 27 21:08:55 my lscpi looks diffrent than that of other logs of testing Apr 27 21:09:03 kk Apr 27 21:09:36 ah yeah, ssb comments look promising Apr 27 21:09:47 /* We do only have one cardbus device behind the bridge. */ Apr 27 21:09:51 ... apparently not Apr 27 21:11:23 ah ok and the cardbus state is indicated by the supplied invariants Apr 27 21:11:30 xMff, i dont know if this will help any but when i do lspci it only shows the wifi and another device. before it showed all this https://dev.openwrt.org/attachment/ticket/7630/netgear-8.09-dmesg-lspci-other.txt Apr 27 21:11:41 so since we already detect your model anyway we can just pretent to not have cardbus Apr 27 21:11:46 regardless of the nvram var Apr 27 21:12:06 xMff, u need a dmesg when cardbus=1 yes? Apr 27 21:12:11 no Apr 27 21:12:21 kk Apr 27 21:12:30 let me make a few test patches Apr 27 21:12:35 <3 Apr 27 21:13:13 xMff: I will add the fallback ssb sprom generation Apr 27 21:13:29 Hauke: okay and I will check the cardbus thing in the meanwhile Apr 27 21:13:43 xMff: ok that's nice Apr 27 21:14:40 I greatly appreciate the help. couldnt do it with out you guys :) Apr 27 21:18:10 Hauke: it would be really useful to have the "get_router()" helper in setup.c instead of the flashmap Apr 27 21:18:28 Hauke: I suspect there are going to be more such quirks with the newstyle bcm units Apr 27 21:18:52 yes Apr 27 21:18:59 quite probably Apr 27 21:19:27 * RealOpty needs newer hardware Apr 27 21:20:12 the parse_trx_partitions (or whatever it is called) should also be moved outside of the flashmap, since other flash support (serial, nand) will also need this Apr 27 21:21:30 xMff: if you need please move get_router() to setup.c Apr 27 21:22:05 as it has nothing to do with flash, expect it was needed there at first it is better places in setup.c Apr 27 21:22:53 RealOpty: http://luci.subsignal.org/~jow/980-wnr834b_no_cardbus_invariant.patch Apr 27 21:23:26 RealOpty: that goes into target/linux/brcm47xx/patches-2.6.37/ Apr 27 21:24:43 xMff, okie Apr 27 21:25:00 make target/linux/clean world V=99 Apr 27 21:31:46 Applying patch platform/980-wnr834b_no_cardbus_invariant.patch Apr 27 21:31:46 patching file arch/mips/bcm47xx/setup.c Apr 27 21:32:03 bout half done compiling Apr 27 21:32:45 rebooting Apr 27 21:34:10 don't forget to reset the nvram var if you didn't so yet Apr 27 21:35:37 omg ty Apr 27 21:35:54 i have no box to help me jtag this if i erase that data lol Apr 27 21:37:09 nbd, from #linux-wireless i got pointed generally to this: http://marc.info/?l=linux-wireless&m=129845633919601&w=2 Apr 27 21:37:22 re my kernel panic'ing Apr 27 21:38:52 that thread seems AHB related Apr 27 21:39:03 i don't see what that has to do with your pci issues Apr 27 21:39:51 yeah Apr 27 21:39:58 also this: http://marc.info/?l=linux-wireless&m=127733695709302&w=2 Apr 27 21:40:02 but with no answer Apr 27 21:40:44 hm instead of moving get_router() into setup.c one should maybe merge the diag.c code there and get rid of the kmod? Apr 27 21:40:54 i guess that previous thread was pointing iomem problems in openwrt patches Apr 27 22:07:13 xMff, well that did make some progress. but not the kind to fix it :( Apr 27 22:07:18 dmesg is coming Apr 27 22:10:44 xMff, http://paste.ubuntu.com/600033/ Apr 27 22:11:05 fixed the having to set cardbus=0 Apr 27 22:11:22 good Apr 27 22:11:49 indeed Apr 27 22:12:02 xMff, i included the lspci and wireless Apr 27 22:12:09 config Apr 27 22:17:58 RealOpty: I've another one in a few minutes Apr 27 22:18:23 :) Apr 27 22:18:35 can your unit be recovered with tftp? Apr 27 22:19:24 only if the board data isnt erased Apr 27 22:20:24 brb Apr 27 22:21:20 http://luci.subsignal.org/~jow/981-extend_bcm47xx_fill_sprom.patch & http://luci.subsignal.org/~jow/982-wnr834b_provide_fallback_sprom.patch Apr 27 22:28:14 back Apr 27 22:28:39 see above Apr 27 22:28:44 indeed Apr 27 22:28:48 working on it now Apr 27 22:28:53 build #18 of lantiq is complete: Failure [failed compile_6] Build details are at http://buildbot.openwrt.org:8010/builders/lantiq/builds/18 Apr 27 22:32:34 shit, a typo Apr 27 22:33:06 doing clean now Apr 27 22:33:26 i can just edit the file? Apr 27 22:34:58 hm yes Apr 27 22:35:14 first of all do this: make target/linux/{clean,prepare} V=99 QUILT=1 Apr 27 22:35:47 or maybe not, just try the patch for now Apr 27 22:35:50 I updated the 981 one Apr 27 22:36:10 * RealOpty installs quilt Apr 27 22:36:35 download the 981 again? Apr 27 22:37:15 xMff, u made changes to 981 on your server since ive downloaded? Apr 27 22:37:21 yes Apr 27 22:37:23 kk Apr 27 22:37:32 it lacked a comma Apr 27 22:38:56 Compiling now Apr 27 22:39:20 after this run you may switch to http://wiki.openwrt.org/doc/devel/patches#iteratively.modify.patches.without.cleaning.the.source.tree Apr 27 22:39:52 kk Apr 27 22:40:41 I just discovered the -X option for ssh the other day..... thats an awesome feature. now i can transcode videos and manage files on my wifes machine without her yelling for her computer back Apr 27 22:41:15 heh Apr 27 22:45:10 ive been using freeNX on my builder machine. Apr 27 22:45:36 mainly cause i can use my old school xbox with gentoox + nx client. Apr 27 22:46:20 xMff, build error Apr 27 22:46:27 show me Apr 27 22:47:16 http://paste.ubuntu.com/600042/ Apr 27 22:49:19 I see Apr 27 22:49:41 edit build_dir/linux-brcm47xx/linux-2.6.37.6/arch/mips/bcm47xx/setup.c Apr 27 22:50:16 go to line 65 Apr 27 22:50:25 via quilt? Apr 27 22:50:33 replace "const char *name" with "char *name" Apr 27 22:50:57 yeah Apr 27 22:51:06 * RealOpty setsup quilt conf Apr 27 22:51:14 did you run make target/linux/{clean,prepare} V=99 QUILT=1 yet? Apr 27 22:51:34 nope Apr 27 22:54:12 xMff, just got quilt installed, now i added my config to .quiltrc, now run that command yes? Apr 27 22:54:18 yes Apr 27 22:55:21 afterwards, best in a different terminal, go to build_dir/linux-brcm47xx/linux-2.6.37.6 Apr 27 22:56:02 then quilt push -a ? Apr 27 22:56:11 first quilt top Apr 27 22:56:14 kk Apr 27 22:56:17 what does it tell ? Apr 27 22:56:25 999-something ? Apr 27 22:57:12 xMff, yes Apr 27 22:57:19 wl_exports is on top Apr 27 22:57:27 ok. quilt pop platform/981-extend_bcm47xx_fill_sprom.patch Apr 27 22:57:45 quilt edit arch/mips/bcm47xx/setup.c Apr 27 22:58:55 kk fixing code Apr 27 22:59:43 after you saved and quit the editor, run quilt refresh, quilt push -a Apr 27 23:00:08 go to your other terminal and run "make target/linux/{compile,install} V=99" Apr 27 23:00:25 this time it will recompile everything again, next time only the changed source files Apr 27 23:01:15 kk Apr 27 23:01:40 compiling Apr 27 23:01:53 * RealOpty needs to spend some time learning quilt Apr 27 23:11:08 error Apr 27 23:11:37 xMff, http://paste.ubuntu.com/600047/ Apr 27 23:13:06 ah ok Apr 27 23:13:15 go to the quilt terminal Apr 27 23:13:24 quilt pop platform/981-extend_bcm47xx_fill_sprom.patch Apr 27 23:13:36 oops, one too far Apr 27 23:13:39 quilt push Apr 27 23:13:47 should be at 982 then Apr 27 23:14:21 Hi all Apr 27 23:15:12 i have here a Thomson SpeedTouch 510 and i want o port GNU/Linux to it Apr 27 23:15:57 OpenWrt supports that device? Apr 27 23:19:26 making diner bbs Apr 27 23:19:35 i have also here a di-500b Apr 28 00:03:59 build #19 of ps3 is complete: Failure [failed compile_3] Build details are at http://buildbot.openwrt.org:8010/builders/ps3/builds/19 Apr 28 00:04:58 angelox_123: depends on what chipset it is Apr 28 00:06:26 angelox_123: the speedtouch 510 has only 2 MB flash, openwrt does not support that few space Apr 28 00:12:31 and on di-500b ? Apr 28 00:13:47 couldn't find info on this one Apr 28 00:15:12 sorry Apr 28 00:15:18 it's dsl-500b Apr 28 00:16:08 can you open the case and identify the chips` Apr 28 00:16:10 ? Apr 28 00:16:52 about the dsl-500b no,but the speedtouch i can Apr 28 00:17:52 would be needed for the dsl-500b Apr 28 00:18:22 maybe tomorrow i'll can open it :D,btw my speedtouch uses a bcm6348 as processor... Apr 28 00:18:58 yes, openwrt supports bcm63xx cpus but the avilable space is too tight Apr 28 00:19:06 it just won't fit into two MB Apr 28 00:19:19 there's no how to setup a NFS? Apr 28 00:20:18 there is, but whats the point of an embewdded device that needs a companion server? Apr 28 00:20:56 you can try to build a minimal openwrt with nfs support and then load the rest somehow Apr 28 00:21:01 but its a big hackery Apr 28 00:21:31 thanks,i'd try only to do some tests... Apr 28 00:21:42 i'll try get my dsl-500b tomorrow and check it Apr 28 00:22:11 anyway,what tools i would need to do that change? Apr 28 00:22:11 ok Apr 28 00:22:44 what change? Apr 28 00:23:09 i mean,i don't understand how OpenWrt works,so what would i need to make it work on my device Apr 28 00:23:48 first you should attach a serial cable to gain access to the bootloader Apr 28 00:24:27 for the speedtouch you may need to add a board definition to the brcm63xx kernel sources Apr 28 00:24:34 is there a how to use a usb instead serial? Apr 28 00:24:36 for the dsl-500b I can't tell Apr 28 00:24:54 google for usb ttl serial Apr 28 00:25:29 hum ok thank you very much...i'll check it tomorrow and ask again here if it's possible :) Apr 28 00:25:53 build #14 of adm5120 is complete: Exception [exception interrupted] Build details are at http://buildbot.openwrt.org:8010/builders/adm5120/builds/14 Apr 28 00:25:54 build #21 of brcm47xx is complete: Exception [exception interrupted] Build details are at http://buildbot.openwrt.org:8010/builders/brcm47xx/builds/21 Apr 28 00:25:56 build #20 of cobalt is complete: Exception [exception interrupted] Build details are at http://buildbot.openwrt.org:8010/builders/cobalt/builds/20 Apr 28 00:25:58 build #19 of orion is complete: Exception [exception interrupted] Build details are at http://buildbot.openwrt.org:8010/builders/orion/builds/19 Apr 28 00:26:16 angelox_123: http://wiki.openwrt.org/doc/hardware/port.serial Apr 28 00:26:55 thanks russell Apr 28 00:30:37 build #19 of pxcab is complete: Failure [failed compile_3] Build details are at http://buildbot.openwrt.org:8010/builders/pxcab/builds/19 Apr 28 00:38:29 ping philipp64|laptop Apr 28 00:48:38 xMff, Apr 28 00:48:53 what command was next? pop? Apr 28 00:48:53 RealOpty, ? Apr 28 00:49:13 go to the 982 patch Apr 28 00:49:36 quilt pop platform/982-wnr834b_provide_fallback_sprom.patch Apr 28 00:50:26 kkok Apr 28 00:50:58 Removing patch platform/999-wl_exports.patch Apr 28 00:50:58 Restoring arch/mips/bcm47xx/nvram.c Apr 28 00:50:58 Restoring arch/mips/mm/cache.c Apr 28 00:50:58 Now at patch platform/982-wnr834b_provide_fallback_sprom.patch Apr 28 00:51:14 ok Apr 28 00:51:22 quilt edit arch/mips/bcm47xx/setup.c Apr 28 00:52:21 find the line "iv->has_cardbus_slot = 0;" Apr 28 00:52:39 move it below "ssb_arch_set_fallback_sprom(&bcm47xx_fallback_sprom);" Apr 28 00:52:57 ok Apr 28 00:52:59 save, quit, quilt refresh; quilt push -a Apr 28 00:53:10 make target/linux/{compile,install} V=99 Apr 28 00:55:45 Refreshed patch platform/982-wnr834b_provide_fallback_sprom.patch Apr 28 00:55:59 $ quilt push -a Apr 28 00:55:59 Applying patch platform/999-wl_exports.patch Apr 28 00:55:59 patching file arch/mips/bcm47xx/nvram.c Apr 28 00:55:59 patching file arch/mips/mm/cache.c Apr 28 00:55:59 Now at patch platform/999-wl_exports.patch Apr 28 00:56:23 ^- does that look correct? Apr 28 00:56:26 yes Apr 28 00:56:32 kk Apr 28 00:56:35 compiling Apr 28 01:17:55 xMff, i think its working. checking some things now Apr 28 01:19:45 the wifi? Apr 28 01:19:55 yes Apr 28 01:19:58 nice Apr 28 01:20:00 it detected the mac Apr 28 01:20:08 im putting it in pio mode now Apr 28 01:20:16 still dma issues? Apr 28 01:20:19 yeah Apr 28 01:20:27 ok Apr 28 01:20:58 but this round the wifi light went completly off. im taking it as a good sign Apr 28 01:21:08 completely Apr 28 01:28:30 can you post a dmesg? Apr 28 01:29:21 yes Apr 28 01:31:03 xMff, my laptop detects the wifi now. but cant get an ip address. also the light (wifi) on the router is off Apr 28 01:31:05 http://paste.ubuntu.com/600087/ Apr 28 01:31:11 ^ dmesg Apr 28 01:32:01 ok Apr 28 01:32:18 for the wifi part we have to wait for mb or someone else to check Apr 28 01:32:40 in the meanwhile you could try kmod-brcm-wl now that the sprom stuff works Apr 28 01:33:18 cool Apr 28 01:33:24 xMff, thanks alot bro Apr 28 01:33:49 excited to see it working on openwrt Apr 28 01:42:51 I ahve to sleep, bbl Apr 28 01:44:15 night Apr 28 01:48:45 build #22 of brcm63xx is complete: Failure [failed compile_6] Build details are at http://buildbot.openwrt.org:8010/builders/brcm63xx/builds/22 **** ENDING LOGGING AT Thu Apr 28 02:59:57 2011