=== bjf is now known as bjf-afk [01:52] I'm desparately trying to get netconsole to work properly, I've followed https://wiki.ubuntu.com/KernelTeam/Netconsole "to the 't'" and I am unable to get netconsole to log between my netbook and laptop. [01:53] I've also tried a crossover between the two, manually setting the IP's [01:53] I cannot get my netbook to show that its receiving data [01:57] I'm getting a kernel panic and I'd like to try to track it down, nothing is showing up in the logs [01:58] h00k: does the device you're trying to debug have a serial port? [01:58] jk-: negative, thats why I'm trying netconsole [01:58] 'k [01:59] you're using netcat to receive? [01:59] I was trying netcat, also syslog-ng [02:00] yeah, stick with netcat for now, less to go wrong [02:00] alright [02:00] syslog-ng baleeted [02:00] so you're getting the 'netconsole' init messages in dmesg on the sender? [02:00] :) [02:01] (from step 5) [02:02] jk-: I do see the netconsole messages, the last one being reported "device eth0 not up yet, forcing it" [02:02] ok [02:02] oop [02:02] and then I see netconsole: network logging started [02:04] so, apparently this is working. (this is since I've tried it last on a crossover cable between the two) [02:04] and the MAC address it shows is correct for your receiving machine? (do an 'ip link show dev eth0' on the receiver to check its MAC) [02:05] oh, it's working then? [02:05] I mean, netconsole reports as logging on the sender machine [02:05] yep, ok [02:06] the MAC shows as correct (as reported by ifconfig eth0): HWaddr 00:22:15:75:be:99 [02:07] so to generate some output from the kernel: [02:07] echo '?' | sudo dd of=/proc/sysrq-trigger [02:08] nothing showing up in the netcat? [02:08] well, I'm on the receiving device now (wifi) [02:08] I...can't...have wireless/crossover going at the same time...can I? [02:09] hm, could you explain your setup a little more? you're using wifi on the receiving machine at the same time? [02:09] no, right now I have disconnected the crossover cable to go online and get some more information [02:10] ah, ok. [02:10] when I'm actually going to test it, I disconnect my netbook from wireless, plug the crossover in, and do it that way [02:10] you can use both at the same time [02:10] I've tried to generate kernel activity [02:10] orly [02:10] okay [02:10] I've tried to generate kernel activity by plugging in a USB device [02:10] just means you have to be a little more careful with the setup :) [02:10] yeah, that should work too. [02:10] oh look at that. [02:10] I didn't realize I could [02:10] hang on, lemme netcat [02:11] you need to make sure that eth0 is up and configured correctly though [02:11] eth0 is connected as my static IP [02:11] to the sender [02:11] ip link show dev eth0 ? [02:11] and: [02:11] ip addr show dev eth0 [02:12] 2: eth0: mtu 1500 qdisc pfifo_fast state UP qlen 1000 link/ether 00:22:15:75:be:99 brd ff:ff:ff:ff:ff:ff [02:12] ok [02:13] so, lemme try that to generate some output [02:13] hang on, do the second command too: [02:13] ip addr show dev eth0 [02:13] I have netcat running: netcat -l -p 6666 -u 192.168.0.2 | tee ~/netconsole.log [02:13] k [02:14] http://pastebin.ubuntu.com/302476/ [02:15] cool, could you pastebin the 'dmesg | grep netconn' on the sender too? [02:17] if it's too tricky (ie, the machine isn't on a network), don't worry [02:18] hang on, I can get it [02:18] sorry, 'netcon' not 'netconn' [02:19] already fix'd [02:19] surprised in this little using it it hasn't panicked yet ;) [02:20] http://pastebin.ubuntu.com/302480/ [02:20] (usb drive) [02:21] I mean, from what I can see, everything /should/ be working. [02:21] yeah, all looks good [02:21] on the receiver: [02:21] sudo tcpdump -ni eth0 [02:22] then generate some console output on the sender [02:22] for example? [02:22] previous example above? [02:22] yeah [02:22] echo '?' | sudo dd of=/proc/sysrq-trigger [02:23] echo '?' | sudo dd of=/proc/sysrq-trigger [02:23] yeah [02:23] then pastebin the output of tcpdump, if there is any [02:24] http://pastebin.ubuntu.com/302483/ [02:24] so, apparently that is working [02:24] cool [02:24] er, wait [02:24] note: 21:23:35.760788 IP 192.168.0.2 > 192.168.0.1: ICMP 192.168.0.2 udp port 6666 unreachable, length 59 [02:25] your netcat isn't running then? [02:25] er. [02:25] * h00k facepalms [02:26] there, it ran, netcat reported:invalid connection to [192.168.0.2] from (UNKNOWN) [192.168.0.1] 6666 [02:26] once I generated the '?' and dd'd it to /proc/sysrq-trigger [02:27] hm, you must have a different netcat to me [02:27] jk-: I did this: netcat -l -p 6666 -u 192.168.0.2 | tee ~/netconsole.log [02:27] that IP being my netbook (receiver's IP) [02:27] correct? [02:27] yeah, my netcat won't accept -l and -p at the same time [02:28] ah, that was copy/pasted from the ubuntu wiki [02:28] what is good to use..then? [02:29] netcat is good, just gotta figure out the correct options :) [02:29] is this an ubuntu machine (the receiver)? [02:29] yes [02:29] -l is listen mode, p specifies port [02:30] u is udp mode [02:30] [jk@pororo ~]$ nc -l -p 6666 -u 192.168.0.2 [02:30] nc: cannot use -p and -l [02:30] 0.0 [02:30] there are two flavours of netcat though [02:30] ah, okay [02:31] [v1.10-38] [02:31] ok, sudo netstat -apu [02:31] see if you can find netcat listening on the right port there [02:31] http://pastebin.ubuntu.com/302486/ [02:31] (it's netcat-openbsd vs netcat-traditional) [02:32] ok, no netcat there :/ [02:33] http://pastebin.ubuntu.com/302487/ [02:33] okay, there we goo. [02:33] jk-: thanks in advance for all of your help. [02:33] cool, working now? [02:33] it's listening on 6666, did you change anything? [02:33] no, I didn't [02:33] hm [02:34] and what happens if you do the "echo '?' ...." now ? [02:34] jk-: netcat dies: invalid connection to [192.168.0.2] from (UNKNOWN) [192.168.0.1] 6666 [02:34] same :/ [02:36] i'm not sure why netcat needs that IP address [02:36] maybe take it out and use "-s 192.168.0.2" instead? [02:36] I don't know either, perhaps I should get rid of it. [02:36] .. or yeah, take it out altogether [02:37] http://pastebin.ubuntu.com/302490/ [02:37] ta-da! [02:37] woot [02:37] in just using this mode: netcat -l -p 6666 -u | tee ~/netconsole.log [02:37] cool, looks better to me :) [02:37] thanks, jk-. I really appreciate you taking the time. [02:37] no problem :) [02:38] now see if you can pop the bug :) [02:38] so, the instructions on the wiki said to have the ip in the netcat listener, too. [02:38] now lets see if I can get this sucker to crash [02:49] Can anyone tell me the new way to regenerate the control file? [02:49] used to be: touch debian/rules.d/control.stub.in; fakeroot debian/rules clean [02:49] but doesn't seem to work for me with karmic [02:54] pturing: just a guess, but touch s/debian/debian.master/ ? [02:54] https://wiki.ubuntu.com/KernelTeam/AbstractedDebian [02:54] ie, touch debian.master/rules.d/control.stub.in [02:57] ok thanks for the link [02:59] jk-: thank you again for the help, I'll part the channel and perhaps be back, depending on what I find [02:59] h00k: no problem, good luck. [03:00] thanks. [03:07] btw, do you know the reason why Ubuntu builds loopback device support in now, while debian still has it as a module? === hzhang__ is now known as panda|phenom [04:07] jk-: I think its my proc [04:12] h00k: how so? [04:19] jk-: standby [04:20] jk-: http://pastebin.ubuntu.com/302528/ [04:21] eep [04:21] jk-: appears to be processor? [04:21] jk-: I think...not exactly sure how to decypher this [04:25] jk-: so, I'm happy that this wworked. [04:25] jk-: thank you very much, again. [04:26] any kernel output before your last paste? [04:26] no, that was it [04:26] jk-: prior to it was the '?' generated by myself [04:26] then I was watching "Fight Club" to try to get it to kernel panic [04:27] it says to run it through mcelog --ascii [04:27] and sure enough, it did. [04:27] ah, ok [04:27] yeah, i'm doing that, its not showing any output, I should learn what its doing [04:27] but yeah, does look like something hardware-related. what machine is this? [04:28] jk-: Dell XPS M1530 [04:29] jk-: Intel T7250 processor, 3gb RAM, 200gb 7200rpm HD w/freefall sensor, nvidia 8600mgt video car [04:29] d [04:31] mcelog --ascii, is this something that I would run when i try to get it to die again? [04:33] no, you use it to parse that kernel output [04:33] mcelog --ascii < your-kernel-log.txt [04:33] oh, oh, i get it. [04:33] , manpage [04:34] "Note that when the panic comes from a different machine than where mcelog is running on you might need to specify the correct architecture ( --k8 or --p4 or --core2 )" [04:38] jk-: I ran it through with the --core2 flag, it just appears to echo the log back. [04:40] this is interesting because its specifically saying 64bit kernels [04:40] h00k: hm, i don't know much about the MCE stuff, sorry :( [04:40] I am on 32bit [04:41] jk-: I'll check ##hardware, they appear to be arguing about gfx cards. === cjwatson_ is now known as cjwatson === lamont` is now known as lamont === bjf-afk is now known as bjf [17:03] Theres an upstream package linked on the ubuntu-meta package https://launchpad.net/ubuntu/+source/linux-meta that's making the upstream description a bit weird. If thats supposed to be there, then so be it, but I'm pretty sure that it should be removed.