**** BEGIN LOGGING AT Sun Aug 26 03:00:00 2012 Aug 26 03:00:30 but thumb is still quite experimental Aug 26 03:01:20 anyway, first step: flashing COMBINED fiasco image. good luck Aug 26 03:01:38 thumb seems complicated more chances of an error Aug 26 03:02:02 rather first first step: do a proper full backup, preferrably by backupmenu Aug 26 03:02:38 (thumb) yes, it's still experimental and thus not exactly a streamlined user experience Aug 26 03:02:47 for installation Aug 26 03:06:07 regarding backup, you probably don't even want to restore too much of your current system state. So maybe get the backup just for best practice and in case you mess something up or need to recover some data after flashing. Other than that purpose, your approach to redo all changes manually and write a proper log about what you did is better idea than restoring too much of what you wanted to get rid of Aug 26 03:50:07 i want my blu ecigs :-/ Aug 26 03:50:36 i can see it now. a total act of desperation forcing me to use usb host mode on my n900 just to charge the ecig pack Aug 26 04:00:58 i've not had a crash using thumb Aug 26 07:25:36 hi there.try to boot nitrdroid with uboot,but it cant recognize vmlinuz,bad format Aug 26 07:26:34 i'm stuck with bootloaders and now don't have an idea which one is better Aug 26 13:21:03 hm, what's the best way to make a simple debian chroot jail? Aug 26 13:21:08 i don't want the whole of easy debian Aug 26 13:22:37 depends on what you need Aug 26 13:23:10 man chroot Aug 26 13:25:43 or rather info chroot Aug 26 13:27:07 I would say using debootstrap Aug 26 13:29:47 NIN101: does it work if the host is not debian but only debian-ish? Aug 26 13:30:05 kerio: see info chroot Aug 26 13:30:17 it depends on what you want to run in your "jail" Aug 26 13:31:05 chroot is a pretty simple and easy to grok concept Aug 26 13:31:22 iirc I deboostraped debian sid from maemo Aug 26 13:38:13 kerio: otoh if you bindmount /* to /mychroot/*, you get a "jail" that's basically capable of doing everything and not restricted in any regard, so basically useless. The question is what you want to protect by your jail/chroot Aug 26 13:38:44 well, i want to protect the maemoness of my system Aug 26 13:39:29 and which components/devices of "server" system your binary running in chroot "client" needs access to Aug 26 13:39:51 debianness :) Aug 26 13:42:09 on N900 one of the major usecases for a chroot is to move rootfs / to eMMC where you got no problems with lack of free storage space Aug 26 13:42:27 well no, you would just move the rootfs there :) Aug 26 13:42:43 well, sorry. I can't help you there Aug 26 13:42:58 too few input to start thinking about your problem Aug 26 13:43:12 i *am* aware of how chroot jails work, i just wanted to know if there was a preferred way of doing it Aug 26 13:45:02 hmm, there *is* such a thing as "easychroot" Aug 26 13:45:08 (the thing easydebian uses) Aug 26 13:46:44 I would just go with debootstrap, it's one command and you get a minimal debian... Aug 26 13:47:48 NIN101: easychroot is just a bunch of scripts to use chroot jails apparently Aug 26 13:54:15 what else? Aug 26 13:54:40 there's simply not more to chroot than that Aug 26 13:55:03 DocScrutinizer05: oh god lol Aug 26 13:55:08 user ALL = NOPASSWD: ALL Aug 26 13:55:19 I know ;-P Aug 26 13:55:22 that's in "/etc/sudoers.d/chroot.sudoers" in the easy-chroot package Aug 26 13:55:24 ~jrtools Aug 26 13:55:25 jrtools is, like, http://wiki.maemo.org/User:Joerg_rw/tools Aug 26 13:55:27 no, no Aug 26 13:55:43 THIS SHIT ENABLES PASSWORDLESS SUDO WITHOUT WARNING Aug 26 13:55:48 what the fuck Aug 26 13:56:01 well, seems chroot.sudores took a bad example Aug 26 13:56:02 windows xp. Aug 26 13:56:16 i mean, this is in extras Aug 26 13:56:41 I thought rootsh looks pretty similar Aug 26 13:56:49 yeah Aug 26 13:56:55 but at least that gives out a warning Aug 26 13:57:01 i wouldn't've known about this crap Aug 26 13:57:11 fools with no idea about sudo and security Aug 26 13:57:28 also haha, easy-chroot installs its own /sbin/chroot Aug 26 13:57:33 as /sbin/qchroot Aug 26 13:57:37 probably because of messybox Aug 26 13:57:45 :-o Aug 26 13:57:50 oh no, nvm Aug 26 13:58:10 it's just a shell script that chroots with bells and whistles and insecurity Aug 26 13:58:22 mount -o bind /dev "$CHROOT/dev" Aug 26 13:58:25 oh HELL NAW Aug 26 13:58:45 well, that's where chroot details start Aug 26 13:58:58 yeah but you don't... do that Aug 26 13:59:15 do you want to grant your jailed binary access to real devices or not Aug 26 13:59:26 maybe not *all* of them Aug 26 13:59:39 i mean, i suppose it's actually fine Aug 26 13:59:53 once you have root you can mknod anyway Aug 26 14:01:34 guys, a stupid question, when typing with virtual keyboard and word prediction turned on, how to use what is predicted? Aug 26 14:08:02 (( maybe not *all* of them)) that's why I asked what you want to do in your chroot Aug 26 14:08:19 DocScrutinizer05: i don't actually know, i feel content with fremantle as it is Aug 26 14:08:21 \_o_/ Aug 26 14:08:33 well, I think you know pretty good how to set up a chroot for your needs Aug 26 14:08:51 :-) Aug 26 14:08:55 =D Aug 26 14:09:01 well, i actually kinda do Aug 26 14:09:19 the only problem is you should get to know your needs then Aug 26 14:10:02 i suppose so Aug 26 14:10:12 freemangordon, tap as if moving cursor Aug 26 14:11:23 kewl. /me never knew either Aug 26 14:11:50 freemangordon, test0,2ing <--tap here Aug 26 14:19:19 MohammadAG: thanks Aug 26 14:24:03 a common step for mankind, but a giant leap for you, man! (in memoriam Niel Armstrong) Aug 26 15:36:49 Anyone experienced constant shutdowns after 10 mins of uptime or less on N900? Aug 26 15:38:35 Suddenly, really suddenly... when I am in maemo or bootmenu it just shutdowns the whole time... Luckily backupmenu works on mass storage... Aug 26 15:39:16 hmh...nope Aug 26 15:39:28 MrPingu: overclocked? Aug 26 15:39:31 Nope Aug 26 15:39:39 full reflash and try again Aug 26 15:39:43 if it still does that, then it's borked Aug 26 15:39:46 but I have 36000 wakeups :D Aug 26 15:39:57 in powertop? Aug 26 15:39:59 otherwise, you messed up something that the watchdog expects Aug 26 15:40:01 yeah Aug 26 15:40:20 Well, let's tell the story from the start Aug 26 15:40:40 Yesterday night went to bed, woke up this morning -> Everything fine Aug 26 15:41:04 from 8 am to 12 it worked nicely Aug 26 15:41:18 then I wanted to check time around 14:00 Aug 26 15:41:32 phone not responding, battery dead Aug 26 15:42:21 Swapped battery with full one, 10 mins later phone off again Aug 26 15:42:34 checked battery in other phone was full for about 70% Aug 26 15:43:27 Guess I need a reflash but it's just odd Aug 26 15:43:52 restoring backup from backupmenu (1week old)didn't help either Aug 26 15:45:43 What can go wrong so suddenly? I mean I didn't even touch my phone... Aug 26 15:46:38 modem Aug 26 15:47:32 Sicelo: Any fix or do I need to reflash? Aug 26 15:49:01 i don't know. i was just mentioning that modem can sometimes be a culprit here. i wouldn't immediately suggest a reflash. see if you can't get people like DocScrutinizer05, freemangordon, etc Aug 26 15:51:16 MrPingu: you got KP51? Aug 26 15:51:23 cssu3 Aug 26 15:51:30 ? Aug 26 15:51:35 = KP51 from what I heard Aug 26 15:52:00 MrPingu: do you have something in dmesg? Aug 26 15:52:20 well, I totally lost track of this cssu-kernel stuff, actually there's no official cssu-kernel for plain CSSU-T/S Aug 26 15:52:37 I am on thumb ;) Aug 26 15:52:51 Anyway from what FMG told, cssu3 = KP51 Aug 26 15:52:54 I guess your problems are with KP, bme, bq24150.ko and related stuff Aug 26 15:53:19 we've seen several such reports lately Aug 26 15:53:20 maaaaaaaaan Aug 26 15:53:30 that's FUD-ish Aug 26 15:53:52 kerio: that's summary of events / reports of last few weeks Aug 26 15:54:21 DocScrutinizer05, I didn't play with bme bq245150.ko. Can it go insane without interaction? Aug 26 15:54:33 that's the huge mystery Aug 26 15:54:34 in theory, no Aug 26 15:54:35 damnit, now this brick doesn't even turn on Aug 26 15:54:38 the module is blacklisted Aug 26 15:55:20 oh nvm, act-dead state -.- Aug 26 15:55:28 haha Aug 26 15:55:43 yep, definetely bq24k.ko problem Aug 26 15:55:55 MrPingu: yeah, exactly in line with what we heard from other similar reports. Remove battery, reinsert after >60s, should reboot then Aug 26 15:56:05 DocScrutinizer05: after 60s? what's that time? Aug 26 15:56:08 bupbat fail? Aug 26 15:56:30 probably discharge time for Vbat buffer capacitors Aug 26 15:56:30 capacitors losing power? Aug 26 15:56:35 i see Aug 26 15:56:57 ah, same issue as Estel had Aug 26 15:57:02 I strongly suggest to roll back to KP50 Aug 26 15:57:28 :c Aug 26 15:58:12 Still 14000 wakeups, not a sane value Aug 26 15:58:24 causede by whom? Aug 26 15:58:32 -e Aug 26 16:00:12 How to see that? Aug 26 16:00:27 I'd bet it's I2C, bme, and/or kernel bq24150 Aug 26 16:00:33 MrPingu: powertop Aug 26 16:00:47 MrPingu: how do you see the 14k wakeups? Aug 26 16:01:01 I understand but how to see from what they come? Aug 26 16:01:40 once there's IRQs listed with count, and then there's list of processes with count Aug 26 16:01:56 DMA - 7126 Aug 26 16:02:10 I2c_omap - 3339 Aug 26 16:02:25 gp - 2386 Aug 26 16:02:42 what's gp? Aug 26 16:02:51 you tell me :P Aug 26 16:02:53 graphics processor? not sure Aug 26 16:03:20 or general purpose Aug 26 16:03:59 I2C is kinda high Aug 26 16:04:07 Anyway you are right i2c_omap is gone ape Aug 26 16:04:20 dma for sure too Aug 26 16:04:32 It's even listed twice in IRQ table ;p Aug 26 16:04:34 is that 56 or 57 Aug 26 16:04:36 ? Aug 26 16:04:50 there are two I2C busses Aug 26 16:04:56 The highest is 56 Aug 26 16:05:05 the one lower in the table is 57 Aug 26 16:05:47 56 | 2650 | INTC | i2c_omap Aug 26 16:05:48 12 | 601 | INTC | DMA Aug 26 16:05:50 37 | 493 | INTC | gp Aug 26 16:05:51 57 | 328 | INTC | i2c_omap Aug 26 16:06:38 PID# | Activity | Name | Function Entry (Expire) Aug 26 16:06:40 --------+------------+----------------+--------------------------- Aug 26 16:06:41 0 | 270 | | tick_nohz_restart_sched_tick (tick_sched_timer) Aug 26 16:07:29 37 | 68D| awk | cpufreq_governor_dbs (delayed_work_timer_fn) Aug 26 16:07:31 0 | 65 | | hrtimer_start (tick_sched_timer) Aug 26 16:07:32 548 | 36 | wl12xx | schedule_timeout (process_timeout) Aug 26 16:07:55 this is with device idle, on fastcharger Aug 26 16:08:19 Mine is on pc, on idle too Aug 26 16:08:53 so what are your top 3 PIDæactivity|name? Aug 26 16:09:13 C# | Ratio | Avg/dura | Frequency | Ratio Aug 26 16:09:14 --------+--------+----------+-----------+--------+ Aug 26 16:09:16 C0 | 2.4% | | 1150 MHz | nan% | Aug 26 16:09:17 C1 | 97.5% | 14.2ms | Aug 26 16:09:21 That's just not normal :O Aug 26 16:09:24 nah, not that Aug 26 16:10:08 PID USER PR NI VIRT RES SHR S PU %MEM TIME+ COMMAND Aug 26 16:10:10 2276 root 20 0 2296 1140 896 R 2.0 0.5 0:00.49 top Aug 26 16:10:11 10 root 20 0 0 0 0 S 0.7 0.0 0:03.52 omap2_mcspi Aug 26 16:10:13 1707 user 20 0 56196 10m 6728 S 0.3 4.5 0:07.02 hildon-home Aug 26 16:10:13 but yes, it's normal Aug 26 16:10:15 I know Aug 26 16:10:23 C1 normal? Aug 26 16:10:28 yep Aug 26 16:10:34 if on charger Aug 26 16:10:58 oh duh, it's on charger... Aug 26 16:11:10 forget a little detail xD Aug 26 16:11:26 MrPingu: please deliver top 3 from powertop, not frpm ps Aug 26 16:11:48 PID# | Activity | Name | Function Entry (Expire) Aug 26 16:11:50 --------+------------+----------------+--------------------------- Aug 26 16:11:51 38 | 36D| awk | cpufreq_governor_dbs (delayed_work_timer_fn) Aug 26 16:11:53 1272 | 27 | bme_RX-51 | sys_timer_settime (posix_timer_fn) Aug 26 16:11:54 0 | 23 | | tick_nohz_restart_sched_tick (tick_sched_timer) Aug 26 16:12:20 duh, pretty little activity there Aug 26 16:12:57 I'm missing 0 | 270 | | tick_nohz_restart_sched_tick (tick_sched_timer) Aug 26 16:13:33 somebody thought it was a good idea to mess with nohz in PK51? Aug 26 16:14:34 err nevermind, it's third in your list Aug 26 16:15:10 Seems device cooled down a bit... Aug 26 16:15:13 anyway for reboots, you should check dsme protocol Aug 26 16:15:59 dsme wdd: Error reading R&D mode flags, WD kicking enabled Aug 26 16:16:00 dsme wdd: Error opening WD /dev/twl4030_wdt: Device or resource busy Aug 26 16:16:02 dsme wdd: Error opening WD /dev/watchdog: Device or resource busy Aug 26 16:16:03 dsme wdd: no WD's opened; WD kicking disabled Aug 26 16:16:05 dsme wdd: Entering main loop, with 12 s interval Aug 26 16:16:06 is that normal? Aug 26 16:16:21 how did you get that? Aug 26 16:16:34 not really I think Aug 26 16:16:35 running dsme Aug 26 16:16:49 err, *running* dsme? Aug 26 16:17:09 please check /var/lib/dsme/stats/lifeguard_restarts Aug 26 16:18:07 Nokia-N900:/var/lib/dsme/stats# cat lifeguard_restarts Aug 26 16:18:09 /usr/bin/ohm-session-agent: 1 Aug 26 16:18:11 /usr/sbin/alarmd.wrapper : 1 Aug 26 16:18:12 /sbin/mce --force-syslog: 1 Aug 26 16:18:14 /usr/bin/camera-ui: 2 Aug 26 16:18:16 /usr/bin/mafw-dbus-wrapper mafw-gst-renderer: 1 Aug 26 16:18:18 /usr/bin/hildon-home: 12 Aug 26 16:18:19 /usr/bin/hildon-desktop: 3 Aug 26 16:18:21 /usr/bin/osso-abook-home-applet: 13 Aug 26 16:18:23 /usr/bin/hildon-status-menu: 18 Aug 26 16:18:25 /usr/bin/systemui: 19 Aug 26 16:18:27 /usr/sbin/browserd -d: 18 Aug 26 16:18:29 /usr/bin/hildon-input-method: 1 Aug 26 16:18:53 not really relevant, just check what counted up after next faulty reboot Aug 26 16:19:04 runniing dsme should give you those errors on an OK device. the watchdog is indeed busy (with first dsme instance) Aug 26 16:19:12 yep Aug 26 16:19:27 dsme isn't meant to get started twice Aug 26 16:19:41 :) Aug 26 16:20:12 anyway I don't see any bme related reboots in your lifeguard_restarts Aug 26 16:20:14 Now, if i remove the charger would it shutdown itself again? Aug 26 16:20:41 if it's on and active, it mustn't shut down on removing charger Aug 26 16:21:19 Well it happend me thrice today, put phone away -> then come back phone off Aug 26 16:21:44 That's how I ended up with this chat :P Aug 26 16:22:21 if it's kernel, mos likely you should have a log in mtd2 iirc Aug 26 16:22:37 MrPingu: please check syslog Aug 26 16:22:52 and mtd2 as Sicelo suggests Aug 26 16:24:31 Looked in mtd2 but didn't see anything special, at first glance Aug 26 16:24:42 MrPingu: search in syslog for last "syslogd 1.5.0#5maemo7+0m5: restart." Aug 26 16:25:01 syslog, sorry where could I find that? Aug 26 16:25:02 see what's the last few lines *before* that Aug 26 16:25:12 /var/log/syslog Aug 26 16:25:24 if you have ksyslogd installed Aug 26 16:25:32 nope Aug 26 16:25:38 too bad Aug 26 16:25:42 go install it Aug 26 16:25:49 on my way sir Aug 26 16:25:55 :-) Aug 26 16:25:56 tools or sdk repo? Aug 26 16:26:10 err, always tools if in question Aug 26 16:26:27 Ah, une moment Aug 26 16:26:44 sysklogd - System Logging Daemon Aug 26 16:27:14 500 https://downloads.maemo.nokia.com ./ Packages Aug 26 16:27:26 sysklogd: Aug 26 16:27:28 Installed: 1.5-5maemo7+0m5 Aug 26 16:28:25 apt-get install sysklogd Aug 26 16:28:56 probably even in extras-devel Aug 26 16:29:05 or extras Aug 26 16:29:18 I don't think so Aug 26 16:29:23 couldn't find it :P Aug 26 16:29:24 or nokia core repo Aug 26 16:30:30 http://maemo.org/packages/view/sysklogd/ Aug 26 16:31:03 Fremantle nokia-applications explicit armel Aug 26 16:31:17 Never heard of that repo Aug 26 16:31:20 Fremantle SDK free armel Aug 26 16:31:26 Yeah, saw it Aug 26 16:31:33 not in tools but sdk Aug 26 16:31:36 Fremantle SDK tools free armel Aug 26 16:32:05 Maemo 5 device SSU repository (>= PR1.2) Aug 26 16:32:07 hm, i don't have nokia-applications Aug 26 16:32:33 Me neither... Aug 26 16:32:55 BTW, I see some kernel-panics in mtd2 Aug 26 16:33:01 just quoting http://maemo.org/packages/view/sysklogd/ Aug 26 16:33:09 :nod: Aug 26 16:33:28 do they have timestamps? Aug 26 16:33:42 so you could actually tell when they happened Aug 26 16:34:15 No timestamps I believe Aug 26 16:34:24 but they do have some numbers in front... Aug 26 16:34:46 yeah, alas there's no decent tool to read out mtdoops Aug 26 16:35:16 at least I'm not aware of any Aug 26 16:36:16 ability to read is a clear advantage ;-P Aug 26 16:36:19 http://maemo.org/packages/view/sp-oops-extract/ Aug 26 16:36:48 Should I install that aswel? Aug 26 16:37:01 nfc, just this minute discovered it Aug 26 16:37:03 ;-D Aug 26 16:37:51 Installing syslogd now Aug 26 16:38:24 ;-DIroN900:~# sp-oops-extract /dev/mtd2 Aug 26 16:38:35 looks somewhat decent Aug 26 16:39:20 seems device is gone sane again Aug 26 16:39:33 I mean, more than 10 min uptime now without charger :) Aug 26 16:41:10 *might* be trackerd hogging IO and thus causing system critical process to freeze which in turn triggers DSME lifeguard WD Aug 26 16:41:49 might be, but I have lowmemorymode thicked + throttle to 10 Aug 26 16:41:54 but no idea how that naive explanation might fit into act_dead problem Aug 26 16:42:37 Wakeups are still somewhat high Aug 26 16:42:40 7700 Aug 26 16:43:07 yeah, kinda high Aug 26 16:43:23 where from do you read those? Aug 26 16:43:58 damn Aug 26 16:44:06 if you are speaking of the devil... Aug 26 16:44:19 screen went black Aug 26 16:44:27 device off... Aug 26 16:44:45 now is the time of syslog Aug 26 16:45:02 Installed at the right moment Aug 26 16:45:06 and comparing mtd2 to what it been before Aug 26 16:45:39 it skips the dots... Aug 26 16:45:44 and checking dsme lifeguard reset stats Aug 26 16:45:47 but asked for pin Aug 26 16:46:04 hildon-home coming up :) Aug 26 16:46:09 ls Aug 26 16:46:17 whoops Aug 26 16:46:54 Can't get to app launcher :o Aug 26 16:48:19 sorry, afk Aug 26 16:48:22 bbl Aug 26 16:48:47 MrPingu, whats the problem? Aug 26 16:48:56 I saw in backscroll, that it's something related to what I had? Aug 26 16:49:09 merlin1991, ping Aug 26 16:49:21 Just acting weird regarding battery, odd shutdowns Aug 26 16:49:30 any problems in turning it on again? Aug 26 16:49:35 sometimes Aug 26 16:49:37 (other than empty battery, if it's the case) Aug 26 16:49:47 symptoms? how it acts, when You can't turn it on? Aug 26 16:49:58 act-dead mode Aug 26 16:50:11 as for kp51, I don't have any problems with battery life, it was just about turning on in act_dead instead of noprmal, which may be some mystery about my device hardware-wise, after all Aug 26 16:50:13 or just reboots after the dots Aug 26 16:50:14 hm Aug 26 16:50:24 MrPingu: please don't mess with flasher options to "fix" it Aug 26 16:50:38 and random shutdowns while in maemo Aug 26 16:51:11 MrPingu, flasher options that DocScrutinizer05 have in mind are --set-rd-flag force-power-key - it make device to boot up as bootreason pwr_key always, so no act-dead Aug 26 16:51:11 DocScrutinizer05: Yes, sir! Aug 26 16:51:26 in my case, it works well, but of cours,e doesn't help finding cause of problem Aug 26 16:51:40 thats probably, why DocScrutinizer05 ask You for not doing it Aug 26 16:51:46 I had an extremely high number of wakeups Aug 26 16:51:58 ls Aug 26 16:52:08 but, You can always reset rd flags to normal, so if You need Your device reliable, you may turn this flag on, then, to help find problems, turn it off again Aug 26 16:52:11 yea, saw it Aug 26 16:52:16 well, in my case, no additional wakeups Aug 26 16:52:39 it's jsut that it booted in act_dead mode, and it was worse over time, i.e. harder and harder to turn it normal way Aug 26 16:52:46 DocScrutinizer05: Need to see my syslog? Aug 26 16:52:50 although, multimeter doesn't show anything special on usb lines Aug 26 16:53:05 and no other oddities during runtime. i've also, recently, did some crazy repair of Ir diode Aug 26 16:53:19 that lead to short between battery + and ground, fixed later Aug 26 16:53:23 Aug 26 18:53:03 Nokia-N900 kernel: [ 431.601623] HWRecoveryResetSGX: SGX Hardware Recovery triggere Aug 26 16:53:27 syslog is full of that Aug 26 16:53:29 shouldn't matter, but You can't filter out that something is still wrong Aug 26 16:53:39 MrPingu, forward it to Pali, for sure Aug 26 16:53:47 nothing like that in my syslog, at least for sure not spamming it Aug 26 16:53:57 also, never heard about it from any kp51 user Aug 26 16:54:09 DocScrutinizer05, BTw, if You are here, please, read my signature carefully, if you're reffering to it ;) Aug 26 16:54:26 [2012-08-26 18:24:42] MrPingu: search in syslog for last "syslogd 1.5.0#5maemo7+0m5: restart." [2012-08-26 18:25:02] see what's the last few lines *before* that Aug 26 16:54:39 at least be able to distinguish between "I am" and "I was", as it's present in my signature for more than a month (since I have stepped down from Council) Aug 26 16:54:55 thanks in advance ;) Aug 26 16:56:02 merlin1991, ping ping ping Aug 26 16:56:44 freemangordon, ping Aug 26 16:57:24 hey, Pali, MrPingu seems to have strange problems, that may be related to kp51 Aug 26 16:57:34 might be* Aug 26 16:57:44 describe it Aug 26 16:57:52 Aug 26 18:53:03 Nokia-N900 kernel: [ 431.601623] HWRecoveryResetSGX: SGX Hardware Recovery triggere Aug 26 16:57:56 syslog full of that Aug 26 16:57:59 battery dying very fast Aug 26 16:58:02 many wakeups Aug 26 16:58:16 +, just like with me, his device sometimes boot only in act_dead, despite being turned out by power button Aug 26 16:58:18 Now several graphic parts fail too Aug 26 16:58:34 but it's only one situation, where our symptoms are similalr - I don't have any problems *except* for act_dead Aug 26 16:58:38 started with kp51 Aug 26 16:58:40 probably Aug 26 16:58:53 I mean statusmenu can't be loaded except for some distorted colors Aug 26 16:59:00 I "fixed" it, temporary, by enabling rd flag force-power-button, so act_dead gtfo, but it's not "real" fix of course Aug 26 16:59:02 flash 2.6.28 omap1 and see if it happends too Aug 26 16:59:17 if not then it is not kernel related Aug 26 16:59:19 Can't, I am on thumb Aug 26 16:59:23 Pali, is there any change in kernel, that allows bq2415 to work properly? Aug 26 16:59:30 and is in effect even, if bq2415 isn't loaded? Aug 26 16:59:56 I did not look at bq2415x after v51 was released Aug 26 17:00:02 no time now... Aug 26 17:00:10 I know, i've reported tons of bugs against it in kp51 thread Aug 26 17:00:13 understandable Aug 26 17:00:25 but I'm only asking if during coding kp51, have You added something Aug 26 17:00:26 in kernel Aug 26 17:00:42 for bq2415, to allow it work properly, that is in effect even without this module loaded? Aug 26 17:01:08 I'm trying to check if my act_dead problems are in any way indeed related to kp51 Aug 26 17:01:46 use my RE getbootstate Aug 26 17:02:00 and see if my getbootstate tell act_dead too Aug 26 17:02:06 OK Aug 26 17:02:06 and if yes, look at code why Aug 26 17:02:17 where to expect logged output? Aug 26 17:02:19 sorry, now going offline, bye Aug 26 17:02:22 ok Aug 26 17:02:23 see ya Aug 26 17:03:24 errr, using getbootstate during act_dead to see if it shows act_dead? ;-) Aug 26 17:03:51 probably, he want me to check if his getboostate gets affected too Aug 26 17:03:52 well, that's just a tad tricky to accomplish Aug 26 17:03:55 sylog shows two restarts Aug 26 17:03:58 and if yes, it gives more debug info (I hope) Aug 26 17:04:11 one failed Aug 26 17:04:20 [2012-08-26 18:24:42] MrPingu: search in syslog for last "syslogd 1.5.0#5maemo7+0m5: restart." [2012-08-26 18:25:02] see what's the last few lines *before* that Aug 26 17:04:29 and one normal with those fsckd graphics Aug 26 17:04:54 Aug 26 18:44:17 Nokia-N900 ke_recv[1939]: prop_modified:1928: udi /org/freedesktop/Hal/devices/platform_slide modified button.state.value Aug 26 17:04:55 Aug 26 18:44:27 Nokia-N900 cellular: csd[1334]: ISI_SMS .966735> ind_reg_status(): Net registration (ind) status:1 rc:0 Aug 26 17:04:56 Aug 26 18:44:27 Nokia-N900 cellular: csd[1334]: ISI_SMS .970031> set_timeout(): Timeout 2253 s event type:-1 Aug 26 17:05:21 then it has to be a kerrnel panic/oops Aug 26 17:05:42 but why? never heard about someone else affected in such way Aug 26 17:05:46 Aug 26 19:00:55 Nokia-N900 kernel: [ 903.265808] HWRecoveryResetSGX: SGX Hardware Recovery triggered Aug 26 17:05:47 Aug 26 19:00:55 Nokia-N900 kernel: [ 903.890838] HWRecoveryResetSGX: SGX Hardware Recovery triggered Aug 26 17:05:49 Aug 26 19:01:01 Nokia-N900 kernel: [ 909.453277] HWRecoveryResetSGX: SGX Hardware Recovery triggered Aug 26 17:05:50 maybe sent it to Pali via PM on TMO? Aug 26 17:06:07 IRC is too loosy for such things, he won't read it, probably, if left here Aug 26 17:06:10 Aug 26 18:59:47 Nokia-N900 kernel: [ 835.968933] HWRecoveryResetSGX: SGX Hardware Recovery triggered Aug 26 17:06:12 Aug 26 19:00:01 Nokia-N900 kernel: [ 849.937652] HWRecoveryResetSGX: SGX Hardware Recovery triggered Aug 26 17:06:13 Aug 26 19:00:12 Nokia-N900 kernel: [ 860.328216] HWRecoveryResetSGX: SGX Hardware Recovery triggered Aug 26 17:06:15 Aug 26 19:00:12 Nokia-N900 browserd[2095]: SET_ACTIVE:0x3f8060, 0->1, is_layout_active:0, ChangeSusp:TRUE Aug 26 17:06:16 Aug 26 19:00:12 Nokia-N900 kernel: [ 861.195434] HWRecoveryResetSGX: SGX Hardware Recovery triggered Aug 26 17:06:18 Aug 26 19:00:13 Nokia-N900 browserd[2095]: SET_ACTIVE:0x3f8060, 1->0, is_layout_active:1, ChangeSusp:TRUE Aug 26 17:06:19 maybe this doesn't increase the amount of info it contains Aug 26 17:06:19 Aug 26 19:00:14 Nokia-N900 kernel: [ 862.460937] HWRecoveryResetSGX: SGX Hardware Recovery triggered Aug 26 17:06:21 Aug 26 19:00:15 Nokia-N900 browserd[2095]: SET_ACTIVE:0x3f8060, 0->1, is_layout_active:0, ChangeSusp:TRUE Aug 26 17:06:22 Aug 26 19:00:15 Nokia-N900 kernel: [ 864.132873] HWRecoveryResetSGX: SGX Hardware Recovery triggered Aug 26 17:06:24 Aug 26 19:00:16 Nokia-N900 kernel: [ 864.757965] HWRecoveryResetSGX: SGX Hardware Recovery triggered Aug 26 17:06:24 +q Aug 26 17:06:25 Aug 26 19:00:18 Nokia-N900 kernel: [ 866.406433] HWRecoveryResetSGX: SGX Hardware Recovery triggered Aug 26 17:06:27 Aug 26 19:00:18 Nokia-N900 browserd[2095]: SET_ACTIVE:0x3f8060, 1->0, is_layout_active:1, ChangeSusp:TRUE Aug 26 17:06:28 Aug 26 19:00:18 Nokia-N900 kernel: [ 866.804809] slide (GPIO 71) is now open Aug 26 17:06:30 Aug 26 19:00:18 Nokia-N900 ke_recv[1754]: prop_modified:1928: udi /org/freedesktop/Hal/devices/platform_slide modified button.state.value Aug 26 17:06:31 MrPingu: please stop it! Aug 26 17:06:32 Aug 26 19:00:19 Nokia-N900 kernel: [ 867.289276] HWRecoveryResetSGX: SGX Hardware Recovery triggered Aug 26 17:06:33 Aug 26 19:00:19 Nokia-N900 kernel: [ 867.914093] HWRecoveryResetSGX: SGX Hardware Recovery triggered Aug 26 17:06:34 Aug 26 19:00:19 Nokia-N900 browserd[2095]: SET_ACTIVE:0x3f8060, 0->1, is_layout_active:0, ChangeSusp:TRUE Aug 26 17:06:36 Aug 26 19:00:20 Nokia-N900 kernel: [ 868.539215] HWRecoveryResetSGX: SGX Hardware Recovery triggered Aug 26 17:06:37 he can'ty it gor triggerred Aug 26 17:06:37 Aug 26 19:00:20 Nokia-N900 browserd[2095]: SET_ACTIVE:0x3f8060, 1->0, is_layout_active:1, ChangeSusp:TRUE Aug 26 17:06:39 Aug 26 19:00:21 Nokia-N900 kernel: [ 869.296936] HWRecoveryResetSGX: SGX Hardware Recovery triggered Aug 26 17:06:41 Aug 26 19:00:22 Nokia-N900 browserd[2095]: SET_ACTIVE:0x3f8060, 0->1, is_layout_active:0, ChangeSusp:TRUE Aug 26 17:06:42 Aug 26 19:00:22 Nokia-N900 kernel: [ 870.961090] HWRecoveryResetSGX: SGX Hardware Recovery triggered Aug 26 17:06:44 he pasted huge output Aug 26 17:06:45 my whole log is spammed Aug 26 17:06:47 sorry Aug 26 17:06:48 BRB Aug 26 17:06:49 it got pasted line by line by his irc client Aug 26 17:06:50 got 2 eat Aug 26 17:07:01 MrPingu, ever heard about zerobin?:P Aug 26 17:07:06 or pastebin?:P Aug 26 17:07:36 MrPingu: you should make sure you got no undervolting, overclocking, and no smartreflex activated Aug 26 17:08:26 if those HWRecoveryResetSGX aren't triggered by any of those, odds are your device silicon is EOL Aug 26 17:08:55 (aka OC burnout, which of course can happen even without you ever doing OC, just way later) Aug 26 17:09:24 but hey, maybe it's thumb related Aug 26 17:09:33 also possible Aug 26 17:10:56 * DocScrutinizer05 away for dinner too Aug 26 17:12:56 :P hey, maybe it's related to anything :P Well, in last attempt, you might also want to backup everything, and do reflash, just to see if it's happen on stock too Aug 26 17:13:00 but don't worry, i' Aug 26 17:13:46 I'm quite confident, that you/we will find cause of it. Of course, "silicon EOL" is always possible - thus, always used as possible cause, when no one have freakin idea what is going on ;) But I have never seen situation, wshen it turned out to be the case for N900, so don't be much worried about that Aug 26 17:13:56 i.e. chances for such irreversible damage are loooow Aug 26 17:14:26 I'm out for some sunday afternun fun, feel free to ping me, will answer later Aug 26 17:14:35 BTw, merlin1991, my pings are still valid ;) Aug 26 17:18:58 btw merlin clearly stated he's not available until 1.9. Aug 26 17:21:03 * DocScrutinizer05 wonders why estel *always* has to argue like a turdhead, ignoring first 2 thirds of statement "if those HWRecoveryResetSGX aren't triggered by any of those, odds are your device silicon is EOL" as well as line before Aug 26 17:21:23 for what it's worth, sometimes i get a silly graphical corruption, if i start doing things too quickly when the n900 is just booting Aug 26 17:21:33 and that also prevents me from accessing the application launcher Aug 26 17:21:38 rebooting fixes it Aug 26 17:21:39 and just going offensive with ""silicon EOL" is always possible - thus, always used as possible cause, when no one have freakin idea what is going on " Aug 26 17:21:45 it happened on two N900s Aug 26 17:24:26 pisses me off when I tell 4 clearly defined and well pondered possible causes for gfx resets, and this dude first states "yeah, or *anything* else", just to continue "obviously you got no clue what's going on, but *ME* can assure all of you that it's not Si-EOL" Aug 26 17:25:38 kerio: yep, gfx resets are not all that uncommon, but they shouldn't happen too frequently Aug 26 17:26:02 DocScrutinizer05: what causes them? Aug 26 17:26:22 nobody knows, or otherwise we already would've fixed them Aug 26 17:26:22 just powervr silliness? Aug 26 17:26:39 probably some Si-Erratum Aug 26 17:27:01 some weird race condition occasionally hitting, on a hw level Aug 26 17:27:57 alas TI is far from disclosing all of those Si-Err, esp when there's no proper workaround/fix Aug 26 17:28:41 fukken TI Aug 26 17:28:46 yep Aug 26 17:30:36 kerio: however there's already some botch "fix" in form of the automatic reset of gfx core via kernel "watchdog" aiui. Too bad when that doesn't suffice to revert to normal operation Aug 26 17:30:41 hehe Aug 26 17:30:59 "sure, the engine sometimes locks up, but there's a fix, you just start it again" Aug 26 17:31:41 as long as that happens reliably and transparently in background, it's almost bearable Aug 26 17:32:35 I however don't see how gfx reboots could tear down or panic the whole linux core Aug 26 17:33:03 fukken kernel-power? :P Aug 26 17:33:27 nah, KP isn't that bad Aug 26 17:33:50 just a lot of the patches included never have seen proper review and thorough tests Aug 26 17:35:21 still I can't come up with a story how gfx reboots would cause inadverted kernel panic. But hey, that's the nature of bugs: you didn't think of this possible side effect until it bites your ass (otherwise you wouldn't have introduced the bug, no? ;-P) Aug 26 17:36:29 Back! Aug 26 17:36:49 maybe the kernel watchdog process for gfx isn't reentrant, so kernel OOPS when gfx 'segfaults' during wtachdog resetting it? Aug 26 17:38:00 MrPingu: [2012-08-26 19:07:36] MrPingu: you should make sure you got no undervolting, overclocking, and no smartreflex activated [2012-08-26 19:08:26] if those HWRecoveryResetSGX aren't triggered by any of those, odds are your device silicon is EOL [2012-08-26 19:08:55] (aka OC burnout, which of course can happen even without you ever doing OC, just way later) [2012-08-26 19:09:24] Aug 26 17:38:02 but hey, maybe it's thumb related Aug 26 17:38:26 DocScrutinizer05: ooh, could be smartreflex Aug 26 17:38:39 iirc it's enabled by default since kernel-power... 50? 51? Aug 26 17:38:47 DocScrutinizer05: HWRecoveryResetSGX was there since the begining of 201 Aug 26 17:38:48 sure thing. Nokia always told us SR is instable Aug 26 17:38:51 2010* Aug 26 17:39:01 well, by default-ish Aug 26 17:39:03 kerio: bullshit, SR is not enabled by default Aug 26 17:39:05 it's in the "default" profile Aug 26 17:39:08 freemangordon: yep, I know, and already mentioned that Aug 26 17:39:17 you have to set it to be the default Aug 26 17:39:37 kerio: SR has nothing to do woth SGX, it is on completely different power bus Aug 26 17:39:41 freemangordon: [2012-08-26 19:25:38] kerio: yep, gfx resets are not all that uncommon, but they shouldn't happen too frequently Aug 26 17:40:00 kerio: in short, stop spreading FUD Aug 26 17:40:03 k Aug 26 17:40:04 DocScrutinizer05: yep Aug 26 17:40:23 Reboot fixes those SGX errors Aug 26 17:40:23 i see a question mark on kerio's line Aug 26 17:40:25 and an iirc Aug 26 17:40:25 freemangordon: you're sure about that powerbus thing? *really* sure? Aug 26 17:40:30 making it more of question than a statement Aug 26 17:40:32 relax man Aug 26 17:40:38 * kerio spreads more fud over a piece of bread and then eats it Aug 26 17:40:56 and the only way to get rid of them is power off/on, a simple reboot won't help Aug 26 17:41:16 DocScrutinizer05: not 100%, but near that value :) Aug 26 17:41:18 I actually meant powercycle Aug 26 17:42:27 freemangordon: depending on hw circuitry incl PCB layout around twl4030 power regulator circuitry this assumption might be incorrect for N900 Aug 26 17:42:33 DocScrutinizer05: BTW I think it is an issue in the driver , not silicon errata Aug 26 17:42:46 and maybe that's exactly why Nokia said SR is flawed by hw Aug 26 17:42:59 DocScrutinizer05: well, I don;t think there is a way SGX and core to operate on one and the same voltage Aug 26 17:43:24 Wakeups however are still insane high Aug 26 17:43:33 as SGX frequency is constant ;) Aug 26 17:43:48 9463 - DMA Aug 26 17:43:48 no, but for example it's pretty easy to figure regulators for both are fed by same power rail so interference is quite possible Aug 26 17:44:12 could be, but SR does not change that Aug 26 17:44:15 2600 - mmc1 Aug 26 17:44:43 sure it does, it changes settings for one regulator and thus might cause interference to the other Aug 26 17:45:11 DocScrutinizer05: the same does OPP control even without SR Aug 26 17:45:35 maybe SR does more 'agressive' changes to regulator settings than non-SR OPP settings do Aug 26 17:46:15 well, this is too "maybe" so my bets are on a simple bug in the driver :) Aug 26 17:46:23 quite tricky to rule out anything along that line Aug 26 17:46:41 possible, even probable Aug 26 17:46:44 having in mind we have the same behaviour it stock kernel too Aug 26 17:46:46 What is exactly DMA and why is it taking so much wakeups? Aug 26 17:46:59 direct memory access Aug 26 17:47:10 good question, since mine isn't that spammy Aug 26 17:47:26 might be your PC to USB Aug 26 17:47:36 IOW usb data transfer Aug 26 17:47:37 uhm Aug 26 17:47:44 I am not attached to pc Aug 26 17:47:51 ooh, nfc then Aug 26 17:48:12 trackerd accessing eMMC? no idea Aug 26 17:48:15 me neither, im not as guru as you guys are :P Aug 26 17:49:02 basically DMA isn't an autonomous process, it needs some (linux) process to control it Aug 26 17:49:25 this linux process should show up un ps and/or (power)top Aug 26 17:50:23 freemangordon: (same behaviour) not a few dozen or even hundereds of gfx resets during >10 min Aug 26 17:50:34 at least I've never seen that Aug 26 17:51:11 <10min* Aug 26 17:51:18 Do the gurus need some data from my device or else I am going to reflash Aug 26 17:51:46 I need to have my device in a working state before tomorrow :P Aug 26 17:51:49 umm, wait another 5 min please. Even _my_ brain needs time to complete evaluations ;-) Aug 26 17:52:36 Sure, you got 2 hours ;) Aug 26 17:52:51 flashing doesn't take that long... Aug 26 17:52:55 at least save syslog and sp-oops-extract Aug 26 17:53:16 and a complete output of powertop Aug 26 17:53:24 MrPingu: one question, how did this start? Aug 26 17:53:29 out of the blue? Aug 26 17:53:30 NFC Aug 26 17:53:35 yes out of the blue Aug 26 17:53:50 did you try to remove the battery? Aug 26 17:53:59 I even swapped batteries Aug 26 17:54:02 [2012-08-26 17:40:20] Well, let's tell the story from the start Aug 26 17:54:04 [2012-08-26 17:40:40] Yesterday night went to bed, woke up this morning -> Everything fine Aug 26 17:54:05 [2012-08-26 17:41:04] from 8 am to 12 it worked nicely Aug 26 17:54:07 [2012-08-26 17:41:18] then I wanted to check time around 14:00 Aug 26 17:54:08 [2012-08-26 17:41:32] phone not responding, battery dead Aug 26 17:54:10 [ Aug 26 17:54:13 oh god paste Aug 26 17:54:15 thanks Aug 26 17:54:40 Yeah suddenly it had an flat battery Aug 26 17:54:46 swapped it for a full one Aug 26 17:55:06 but then I kept shutdowns within 10 mins Aug 26 17:55:21 sorry guys, I really need to care about my dinner now Aug 26 17:55:25 hmm, so you don't have any idea if rootfs is nog corrupted Aug 26 17:55:36 nope Aug 26 17:55:42 df -h Aug 26 17:56:03 leading to some broken graphics file which h-d or h-h tries to load in SGX Aug 26 17:56:13 MrPingu: did you try to change the theme? Aug 26 17:56:18 quite a good story indeed Aug 26 17:56:20 Not yet Aug 26 17:56:39 hey freemangordon :) Aug 26 17:56:47 hey zeq1 Aug 26 17:56:50 you can also try to delete application snapshots Aug 26 17:56:55 Hi DocScrutinizer05 Aug 26 17:56:58 hi zeq1 Aug 26 17:57:21 where? Aug 26 17:57:26 quite a good story indeedmrplease do a proper BM backup Aug 26 17:57:35 errrrr :-( Aug 26 17:57:36 MrPingu: /home/user/.cache/launch Aug 26 17:57:48 MrPingu: please do a proper BM backup Aug 26 17:58:03 and safe it off the phone Aug 26 17:58:11 save* Aug 26 17:58:17 I have a properbackup of 17 aug and one of 25 aug on my pc Aug 26 17:58:19 going afk for a while Aug 26 17:58:26 so yesterday Aug 26 17:58:35 do one of the current state, with backupmenu Aug 26 17:59:14 Yes, will say when it's done ;) Aug 26 17:59:22 If some other steps are needed Aug 26 17:59:41 add syslog and sp-oops-extract and powertop Aug 26 18:00:24 this should allow forensic analysis later Aug 26 18:01:19 then I suggest you reflash and test with standard config Aug 26 18:01:23 Now I start wondering if my backup of yesterday might be affected Aug 26 18:01:50 o/ Aug 26 18:02:52 I remember my phone was extremely hot while backup'ing Aug 26 18:04:21 nah Aug 26 18:04:36 probably unrelated Aug 26 18:04:47 DocScrutinizer05: Standard config, you mean omap kernel, and no CSSU? Aug 26 18:04:55 yep, please Aug 26 18:05:08 well, most probably some resource file (icon, .pvr, pcx, ) got corrupted Aug 26 18:05:22 then reinstall step by step, and see if you can spot what causes the effect Aug 26 18:05:48 DocScrutinizer05: let him try to clean .pvr cache and change the theme, it is a matter of seconds to be done Aug 26 18:05:52 freemangordon: might we detect that somehow? Aug 26 18:06:08 ^^^ Aug 26 18:06:23 e.g by comparing *.pvr of old and fresh install Aug 26 18:06:25 Right now backup'ing so can't do anything else... Aug 26 18:06:45 DocScrutinizer05: .pvr is done on-the-fly by h-d Aug 26 18:06:47 freemangordon: up to you :-) Aug 26 18:07:02 this is a part of smoke-and-mirror :) Aug 26 18:07:02 :nod: Aug 26 18:07:39 and actually is a maemo specific hack in clutter Aug 26 18:08:11 plus some SGX extension in GLES2 Aug 26 18:08:30 hacky all the way, so that could be a reason Aug 26 18:08:50 Want me to do an eMMC flash too? Aug 26 18:08:54 anyway, IF problems persist even after flashing, you might want to clean your battery contacts and get a new battery. Such symptoms *might* be caused by worn out LiIon cell or BSI borked - at least some of them, like sudden reboots and act_dead Aug 26 18:09:32 MrPingu: eMMC most likely unrelated, so no Aug 26 18:09:43 well, .pvr cache is on eMMC Aug 26 18:09:55 hmm, deleting it should suffice Aug 26 18:09:59 "/home/user/.cache/launch" Aug 26 18:10:10 iirc that is the only place Aug 26 18:10:12 rm /home/user/.cache/launch/* Aug 26 18:10:24 Get it Aug 26 18:10:46 no need to flash whole eMMC first instance Aug 26 18:10:47 One advantage, now I can use your apt-get wrapper from the start :P Aug 26 18:10:58 cya Aug 26 18:11:00 :-D Aug 26 18:11:28 Are you going? Bye! Aug 26 18:15:47 Yes, something is wrong for sure Aug 26 18:15:55 maybe fscked rootfs Aug 26 18:16:07 most probably :( Aug 26 18:16:24 let me take a pic Aug 26 18:17:56 on boot it shows missing images and such Aug 26 18:18:10 hehe Aug 26 18:18:37 same problem I got with a fake digitizer/frontcover Aug 26 18:22:25 http://imgur.com/GXp6N Aug 26 18:23:04 MrPingu: well, well, defffinitely your fottfs is corrupted :( Aug 26 18:23:10 damn Aug 26 18:23:17 rootfs that is Aug 26 18:23:40 Note: this image problem is the first time I got this Aug 26 18:24:03 I mean I have booted more than once today and I needed to load this screen because battery swap Aug 26 18:24:07 anything suspicious in dmesg log? Aug 26 18:24:17 related to NAND Aug 26 18:24:48 How to read whole dmesg log? Aug 26 18:24:56 zeq1: any progress with ipclite? Aug 26 18:25:06 MrPingu: dmesg Aug 26 18:25:10 I mean i can't go further up on busybox Aug 26 18:25:29 dmesg > you_name_it.log Aug 26 18:26:23 freemangordon: he got syslog meanwhile Aug 26 18:26:44 syslog >=dmsg Aug 26 18:27:04 yes, too noisy :) Aug 26 18:27:07 I see mtdoops isn dmesg Aug 26 18:27:28 care to pastebin? Aug 26 18:31:28 MrPingu: just unrelated "pro-tip": dmesg|less Aug 26 18:31:56 pastebin.com/nWjH4dq3 Aug 26 18:32:51 DocScrutinizer05: Thanks Aug 26 18:33:29 hmm, no SGX restarts now your images are missing :P Aug 26 18:33:52 SGX issue was fixed after powercycle Aug 26 18:34:55 please pastebin sp-oops-decode /dev/mtd2 Aug 26 18:35:51 errr sp-oops-extract Aug 26 18:36:43 DocScrutinizer05: makes no sense, his rootfs is FUBAR Aug 26 18:37:01 probably you're right Aug 26 18:37:05 21:20 http://imgur.com/GXp6N Aug 26 18:37:09 DocScrutinizer05: ^^^ Aug 26 18:37:31 Yes, it is :( Aug 26 18:37:33 :nod: Aug 26 18:38:07 Rootfs, I have always treated you very nice. How dare you do this to me? :'( Aug 26 18:38:24 MrPingu: you can check if your battery is loose, that caused me (and other people) lots of problems Aug 26 18:38:41 http://pastebin.com/nPUHRDk1 Aug 26 18:38:47 restarts/power-offs out of the blue Aug 26 18:38:48 [2012-08-26 20:08:54] anyway, IF problems persist even after flashing, you might want to clean your battery contacts and get a new battery. Such symptoms *might* be caused by worn out LiIon cell or BSI borked - at least some of them, like sudden reboots and act_dead Aug 26 18:39:04 cleaning won't help, he has to extend the springs Aug 26 18:39:13 :nod: Aug 26 18:39:45 battery isn't loose Aug 26 18:40:23 check your battery's contacts. You see bumps like half spheres. Do they touch each other? Aug 26 18:40:52 well, sorry I'm afk again Aug 26 18:41:04 MrPingu: which kernel do you use? Aug 26 18:41:10 cssu3 Aug 26 18:41:45 well, the last kernel opps is from KP47 :D Aug 26 18:41:47 Yes I see these bumps, they don't touch Aug 26 18:41:59 Are u sure? Aug 26 18:42:08 according to the log Aug 26 18:42:14 lemme check once again Aug 26 18:42:30 logs are saved through reflahs? Aug 26 18:42:45 don't believe that ;p Aug 26 18:43:12 last reflash was in may, KP49 atleast I think Aug 26 18:43:32 Got sane wakeups too now Aug 26 18:43:38 3000 Aug 26 18:43:51 "/dev/mtd2" is not flashed Aug 26 18:44:12 Doesn't it get emptied? Aug 26 18:44:12 MrPingu: yes, the last kernel oops is from KP47 Aug 26 18:44:17 no Aug 26 18:44:36 it is a circular buffer Aug 26 18:44:46 some kind of :) Aug 26 18:44:52 pretty insane: http://www.youtube.com/watch?v=1zXwOoeGzys Aug 26 18:46:17 it's not a kernel bug then ;) Aug 26 18:47:50 It sucks I need a reflash though Aug 26 18:48:11 MrPingu: well, it is Aug 26 18:48:18 it is? Aug 26 18:48:28 just doesn't result in oops Aug 26 18:48:34 https://bugs.maemo.org/show_bug.cgi?id=9150 Aug 26 18:48:35 04Bug 9150: Device doesn't respond via UI. syslog reports HWRecoveryResetSGX: SGX Hardware Recovery triggered, sgx_misr eating all CPU Aug 26 18:48:47 a pretty old one :) Aug 26 18:48:52 :( Aug 26 18:49:33 it took Nokia 2 years or so to mark it WONTFIX ;) Aug 26 18:50:12 and now I am hit by it -.- Aug 26 18:50:46 you are hit by a corrupted fs triggering the bug :P Aug 26 18:51:28 idc, it comes both down to an reflash ;p Aug 26 18:52:08 Do you think it's safe to restore yesterday's BM backup or is it possible that contains the corruption? Aug 26 18:52:09 yes "( Aug 26 18:52:24 well, it worths nothing to try it Aug 26 18:52:35 afaik, as I never used BM Aug 26 18:52:53 Me neither, actually Aug 26 18:53:06 better ask doc ;) Aug 26 18:53:26 he is BM fan and can help you Aug 26 18:53:32 He said it was safe ;) Aug 26 18:53:42 well, go on then :) Aug 26 18:53:58 BM doesn't mkfs Aug 26 18:54:13 so all you do is replace files via 'cp' Aug 26 18:54:40 if actually fs is corrupted, a BM restore without prior flashj won't help Aug 26 18:55:10 freemangordon: I had a little trouble getting fennec to build from romaxa's patchset. For some reason his patches didn't match the base rev. identified in the repo. Aug 26 18:55:16 a restore after flash can't cause additional damage though, beyond wasted time to do all flashing again Aug 26 18:55:33 zeq1: did you ask romaxa? Aug 26 18:56:02 You say *additional* :P Aug 26 18:56:24 brb Aug 26 18:56:34 well, you might restore files that got messed up/corrupted Aug 26 18:56:42 BM should really mkfs Aug 26 18:56:54 i mean, the ubifs Aug 26 18:56:54 indeed it should have an option for that Aug 26 18:57:09 at least BM can run fsck Aug 26 18:57:22 but I guess that's useless on ubifs Aug 26 18:57:26 attach the ubi volume, mkfs.ubifs Aug 26 18:58:50 I still fail to grok why system reboots after 10 minutes Aug 26 18:59:04 (or less) Aug 26 18:59:30 * DocScrutinizer05 suspects fs corruption not being the root cause of the problem, but maybe rather a consequence Aug 26 18:59:49 DocScrutinizer05: hmm, maybe I am wrong https://bugs.maemo.org/show_bug.cgi?id=7017#c17 Aug 26 18:59:50 04Bug 7017: SGX memory reset seems failed during reboot Aug 26 19:00:11 though besides the statement, there is no real info Aug 26 19:01:30 DocScrutinizer51: some WD is kicking, dsme most probably Aug 26 19:02:28 afaik after n tries in x seconds, if a process still fails to start, dsme reboots the device Aug 26 19:02:48 fails to start as crashes :D Aug 26 19:02:53 freemangordon: I need to check to see if romaxa has updated the repo since. If not I'll contact him. I did try to fix up the patches but there were still tricky build failures. Aug 26 19:03:26 Note: I got no reboots Aug 26 19:03:37 it's just the device wasn't "on" anymore Aug 26 19:03:46 sometimes it booted to act-dead Aug 26 19:03:51 sometimes it booted normal Aug 26 19:03:59 zeq1: ok. sorry I can't help, but I just finished VKB REing and want to play with the kernel a little bit. Your libc patches are welcome :P Aug 26 19:04:06 but when it booted normal it insane number of wakeups Aug 26 19:04:21 MrPingu: you have loose battery Aug 26 19:04:27 I can bet on that Aug 26 19:04:56 Ah yes, did I not make them available? Aug 26 19:05:13 I've still not had any problems fwiw Aug 26 19:05:17 the only way n900 to power-off out of the blue is through bad battery connection Aug 26 19:05:37 zeq1: don't remember, it was 2-3 weeks ago :D Aug 26 19:06:11 zeq1: anyway, I will pester you starting from tomorrow :P Aug 26 19:06:36 ok :) Aug 26 19:06:43 merlin1991 needs a proper .deb for toolchain ;) Aug 26 19:06:52 4.7.2 I mean Aug 26 19:07:06 I'd better get back to my gf, been busy decorating all day, lazy evening... Aug 26 19:07:21 yeah, not a bad idea Aug 26 19:07:27 freemangordon: or by battery brownout Aug 26 19:07:35 freemangordon: native toolchian deb or host (cross) toolchain deb Aug 26 19:07:42 (if you screw with bme) Aug 26 19:07:42 zeq1: cross Aug 26 19:07:55 ok Aug 26 19:08:02 the one installable in OS, not in SB Aug 26 19:08:24 I need to build the new one anyway Aug 26 19:08:44 or even better something similar to original SB installer (python one) Aug 26 19:08:48 zeq1: why? Aug 26 19:09:08 why use the latest Linaro? Aug 26 19:09:14 bug fixes :) Aug 26 19:09:36 hmm, I missed "the latest Linaro" part :P Aug 26 19:10:03 zeq1: but better keep 4.7.2 Aug 26 19:10:12 freemangordon: IDK, battery doesn't seem to be loose Aug 26 19:10:26 MrPingu: well, NFC then Aug 26 19:10:33 yes, that's the Linaro "stable" version (anyway afk for a while) Aug 26 19:11:01 Reflashing now, and moving on... Aug 26 19:19:21 MrPingu: you say the spheres of battery springs don't touch. They should though (at least almost) Aug 26 19:21:27 well, not actually touch, but a tiny bit too large a gap between them will cause poor contact to the blades of N900 Aug 26 19:22:36 It's a known problem that springs get out of shape when you insert anything too large into the battery contacts, or even just with time they get too weak Aug 26 19:24:50 anyway, bending them 'inwards' with a needle or pinpointed knife so much that the spheres touch should cure any such problems (be careful not to cause shortcircuits when doing so!) Aug 26 19:25:19 o/ Aug 26 19:29:36 bloody hell, I hate it when my credit card bounces Aug 26 19:30:10 I'm stuck without food tonight, because for some unknown reason, the bank decided to refuse any transaction. Aug 26 19:31:57 >tracked money Aug 26 19:32:00 mfw Aug 26 19:47:23 DocScrutinizer, well that's just the case. Was about to write I can't see if they touch. Almost none space between them ;) Aug 26 19:48:02 Sending cmt-mcusw image (5826 kB)... Aug 26 19:48:03 100% (5826 of 5826 kB, avg. 7970 kB/s) Aug 26 19:48:05 Sending request 0x14 failed: Invalid or incomplete multibyte or wide character Aug 26 19:48:06 Flashing cmt-mcusw... Sending request 0x50 failed: Protocol error Aug 26 19:48:15 WTF for problem do we have now? :( Aug 26 19:48:27 kerio: tell me about it Aug 26 19:49:15 lrn2cash and bitcoins Aug 26 19:49:19 which are about the same thing Aug 26 19:49:29 actually, bitcoins are even *less* traceable Aug 26 19:54:15 md5sums match Aug 26 19:57:05 kerio: lol, I'll be damned if I get conned in one of does stupid pyramid schemes Aug 26 19:57:22 teotwaki__: like the US dollar? Aug 26 19:57:41 kerio: and sorry, but the whole "traceable" thing, I LOL when I see the first "bitcoin credit card" on /.. Aug 26 19:57:49 nvm, other USB connection saved it Aug 26 20:31:33 I thought bitcoin is zombibied meanwhile Aug 26 20:32:01 or gone down like Reichmarks Aug 26 20:32:31 ~jrtools Aug 26 20:32:32 somebody said jrtools was http://wiki.maemo.org/User:Joerg_rw/tools Aug 26 20:36:38 BRB Aug 26 20:49:55 hi nandtester here Aug 26 20:50:02 hi sambo7 Aug 26 20:50:18 FINALLY I was able to revive my device:-) Aug 26 20:50:21 neat Aug 26 20:50:22 how? Aug 26 20:50:58 flasher-3.5 -h RX-51:2101 -F RX-51_2009SE_20.2010.36-2.203.1_PR_COMBINED_203_ARM.bin -c -S usb -f Aug 26 20:51:30 cold flashing didn't work for me Aug 26 20:51:47 at least what is described in the tutorial Aug 26 20:52:37 looks like cold-flashing to me Aug 26 20:52:40 and flasher-3.5 -h RX-51:2101 -F -c -S usb Aug 26 20:52:46 what was the problem btw? Aug 26 20:52:47 didn't work either Aug 26 20:53:11 I nandtested mtd0-mtd5 Aug 26 20:53:53 I also was trying to use Nokia phoenix...cold flashing was ok,but warm flashing failed Aug 26 20:54:17 in the beginning I was always using the global image Aug 26 20:55:16 but a few days ago I remembered that I have a device from Englang and I started to use the English image Aug 26 20:55:29 don't know if this made a difference Aug 26 20:56:01 nice. i think someone else had tol use a 'localized' image a while back Aug 26 20:56:21 phoenix didn't need some special cable? Aug 26 20:56:39 if coldflash restores CAL, then for sure localization matters Aug 26 20:57:46 btw >>flasher-3.5 -h RX-51:2101 -F RX-51_2009SE_20.2010.36-2.203.1_PR_COMBINED_203_ARM.bin -c -S usb -f<< and >>flasher-3.5 -h RX-51:2101 -F -c -S usb<< seem to differ only in additional -f Aug 26 20:57:58 err, mompls Aug 26 20:58:07 yeah Aug 26 20:59:08 and actually if coldflashing fails, the hw build revision might matter Aug 26 20:59:31 iirc you only took a shot into the blue with 2101 Aug 26 20:59:33 I have a lot of errors when I run dmesg Aug 26 21:00:01 hw-revision could not be read Aug 26 21:00:29 sure, since it's in CAL Aug 26 21:00:29 Found device RX-51, hardware revision 0000 Aug 26 21:01:02 I got the error messages when flashing Aug 26 21:01:04 Bootloader error log follows: ERROR: invalid header magic at addr 0x00000000: c0911a26 ERROR: invalid header magic at addr 0x00000000: ce9fb9e6 ERROR: Error processing request, stalling ERROR: EP0 STALL sent Error getting version of 'sw-release' Overriding revision information given by board. Aug 26 21:01:15 phoenix didn't need some special cable? Aug 26 21:01:48 phoenix can do what flasher does Aug 26 21:01:51 via USB Aug 26 21:01:54 not more Aug 26 21:02:18 for anything beyond, you need special 'cable' Aug 26 21:02:34 ah. Aug 26 21:02:55 * Sicelo still wishes to revive his WSOD-stricken N-Gage Aug 26 21:03:00 I found some tutorial in youtube & tried it...but I guess that like in flasher only the first part of the cold flashing was successful Aug 26 21:03:04 :c Aug 26 21:03:23 sambo7: your n900 will probably malfunction, though Aug 26 21:03:49 a bunch of informations were lost Aug 26 21:04:00 even if you bring it back to life Aug 26 21:04:08 I guess,it will... Aug 26 21:04:45 I already got some CAL errors messages in terminal... Aug 26 21:05:53 and when I go on product info I get Nokia N900 & Version: Aug 26 21:06:55 well, at least it fails gracefully Aug 26 21:06:57 does the phone work? Aug 26 21:07:27 haven't treid it yet... Aug 26 21:07:46 sambo7: i also have version unknown for product info Aug 26 21:07:49 :( Aug 26 21:08:13 Sicelo: on your n-gage? Aug 26 21:09:12 N900 Aug 26 21:10:09 sambo7: so probably time to try restoring CAL from an image, with nandwrite Aug 26 21:10:48 is nandwrite part of the mtd-tools? Aug 26 21:10:50 the phone should really not work Aug 26 21:10:51 kerio: i have an N-Gage that no shop here can flash. i have N900 too, and another working N-Gage :P Aug 26 21:10:53 yep Aug 26 21:10:54 pre-installed? Aug 26 21:11:34 there's nandread/write and there's mtd_debug Aug 26 21:11:50 for sure not preinstalled though Aug 26 21:12:34 will test if phone works first Aug 26 21:12:41 wifi is working Aug 26 21:15:24 well, sure you may test if phone works, but in the end you want to try and restore CAL somehow, since it can't get worse than it's now Aug 26 21:16:27 and at very least hw-version should get set to some sane value - otherwise you might expect problems on every simple reflash Aug 26 21:17:43 maybe you could find out about your hw revision via the device's serial number Aug 26 21:18:41 or with even more luck, hw rev is stored to modem filesystem as well Aug 26 21:19:01 and you *might* find out about it via pnatd Aug 26 21:19:35 (just handwaving, I never checked what you could find out from modem regarding hw rev) Aug 26 21:22:18 @Doc ... maybe you can give me some support on which commands to use in the terminal to advance... Aug 26 21:30:13 sorry,have to quit for now...maybe I'll be back later or tomorrow Aug 26 21:30:14 bye Aug 26 21:33:22 hello everybody Aug 26 21:33:57 I am a N9 user Aug 26 21:34:28 resently i try to write some script Aug 26 21:35:04 i want it auto run after system startup Aug 26 21:35:44 but i can't find the rc.local file Aug 26 21:36:21 #harmattan Aug 26 21:36:33 is N9 has some after file for configure autorun? Aug 26 21:36:47 see /topic Aug 26 21:37:32 thx for respond~ Aug 26 21:39:15 I still can't believe Armstrong is dead. Aug 26 21:39:19 maybe DocScrutinizer05, MohammadAG and others know .. but i'm sure they are in #harmattan as well Aug 26 21:40:02 teotwaki__: actually I already tried (in vain) to give a farewell bonmot Aug 26 21:40:37 there's a space missing :P Aug 26 21:40:40 it's "bon mot" Aug 26 21:40:56 but in this case, "mot juste" would be more appropriate Aug 26 21:41:18 for what I came up with, neither is appropriate Aug 26 21:41:44 "ouch!" probably is most appropriate ;-D Aug 26 21:42:46 Well, I'd simply say that the man who accomplished mankind's greatest leap just embarked on man's greatest journey. Aug 26 21:44:01 anyway, gotta go feed a colleague's cat. 'later Aug 26 21:44:18 o/ **** ENDING LOGGING AT Mon Aug 27 02:59:57 2012