**** BEGIN LOGGING AT Wed Apr 20 02:59:58 2016
Apr 20 03:55:08 <DocScrutinizer05> (wait a 10 min) *eventually* HAM will show a yellow notified like "one packet installed succkessfully"
Apr 20 03:55:13 <DocScrutinizer05> then it's done
Apr 20 03:55:44 <DocScrutinizer05> notifier*
Apr 20 04:50:40 <KotCzarny> wens, woohoo
Apr 20 04:50:45 <KotCzarny> m3 is nice feature wise
Apr 20 04:51:00 <KotCzarny> pity a83t is almost unsupported
Apr 20 06:20:17 <fishbulb> how close is that neo900 project?
Apr 20 16:46:46 <Sicelo> my n900 rebooted unexpectedly :/
Apr 20 16:47:09 <Sicelo> in syslog are a couple of:
Apr 20 16:47:13 <Sicelo> kernel: [31618.557556] HWRecoveryResetSGX: SGX Hardware Recovery triggered
Apr 20 16:48:20 <KotCzarny> disable hw in /etc/sgx.conf or whatever it is?
Apr 20 16:49:44 <Sicelo> i don't follow
Apr 20 16:50:32 <KotCzarny> what is the value of wsegl_usehwsync in /etc/powervr.d/hildon-desktop.ini ?
Apr 20 16:51:17 <Sicelo> 0
Apr 20 16:51:25 <KotCzarny> then its not it
Apr 20 16:59:26 <Sicelo> weird thing is that i was not even doing anything unusual on the N900. it was just connected to an ssh server, & i was not using it
Apr 20 16:59:49 <Sicelo> https://bugs.maemo.org/show_bug.cgi?id=9150
Apr 20 16:59:51 <povbot> 04Bug 9150: Device doesn't respond via UI. syslog reports HWRecoveryResetSGX: SGX Hardware Recovery triggered, sgx_misr eating all CPU
Apr 20 17:00:30 <Sicelo> i don't know if that sgx_misr process was running erratically, as i was not actively using device 
Apr 20 17:02:46 <KotCzarny> well, at least you know you are not alone
Apr 20 17:18:29 <drathir> anyone using genwall?
Apr 20 17:39:20 <Sicelo> anyone can explain how to understand/interpret /var/lib/dsme/stats/lifeguard_resets?
Apr 20 17:48:20 <sixwheeledbeast> Sicelo: I have had that in the past, unresolved bug IIRC.
Apr 20 17:48:28 <DocScrutinizer05> it's a tad obfuscated but I think it counts the totals of the different reset reasons
Apr 20 17:50:30 <DocScrutinizer05> yep, total of dsme process monitor restarts per process
Apr 20 17:51:20 <Sicelo> let me show mine.
Apr 20 17:51:57 <DocScrutinizer05> http://paste.opensuse.org/74724405
Apr 20 17:52:31 <Sicelo> http://paste.debian.net/439755/
Apr 20 17:54:12 <Sicelo> so i'm a little lost .. is it saying *all* those applications were reset?
Apr 20 17:54:13 <DocScrutinizer05> note that mine also has some intentional manual kills
Apr 20 17:54:20 <DocScrutinizer05> yes
Apr 20 17:54:37 <DocScrutinizer05> at some time since the system got installed
Apr 20 17:55:29 <DocScrutinizer05> note that I always restore via BM, so mine is basically at least since PR1.3 
Apr 20 17:55:53 <Sicelo> ah, so not useful info then if it's talking about history :p
Apr 20 17:56:02 <DocScrutinizer05> however camera-ui is somewhat real, I actually see lockups 
Apr 20 17:56:13 <DocScrutinizer05> delete it
Apr 20 17:56:50 <DocScrutinizer05> yes, it's crap it doesn't have history/timestamps
Apr 20 17:57:27 <Sicelo> my system has been going on for 5 years .. no reflsh/restore
Apr 20 17:57:44 <Sicelo> so that info could be from God-knows-when
Apr 20 17:58:26 <DocScrutinizer05> not bad then
Apr 20 17:59:42 <Sicelo> but lately i'm having weird things happening .. last week i would wake up to a completely hung device .. something to do with communication between mce & dbus
Apr 20 18:00:04 <Sicelo> is it possible to restart dbus without reboot?
Apr 20 18:00:21 <DocScrutinizer05> actually not "all tose apps were reset" but "restarted" after either quitting or going unresponsive
Apr 20 18:00:36 <DocScrutinizer05> hmm dbus, nope
Apr 20 18:00:40 <DocScrutinizer05> hardly
Apr 20 18:02:03 <Sicelo> that's unfortunate
Apr 20 18:02:50 <Sicelo> because most of the hiccups i have had on this device tend to involve dbus one way or other, usually with mce
Apr 20 18:03:27 <Sicelo> longest uptime is now only 10 days .. cannot manage beyond that
Apr 20 18:03:52 <Sicelo> although with this latest sgx problem, perhaps battery at fault too
Apr 20 18:04:13 <KotCzarny> sicelo, which branch you have?
Apr 20 18:04:26 <DocScrutinizer05> hmmm
Apr 20 18:05:21 <Sicelo> thumb
Apr 20 18:05:41 <KotCzarny> hmm
Apr 20 18:05:59 <KotCzarny> did it behave before you went to thumb?
Apr 20 18:06:33 <Sicelo> i believe so. even when i got to thumb :)
Apr 20 18:06:58 <Sicelo> this developed over time ... tbh it's hinting at a reflash,
Apr 20 18:07:06 <KotCzarny> :)
Apr 20 18:07:08 <Sicelo> which i'm not too keen on, yet
Apr 20 18:11:24 <kerio> Sicelo: restarting dbus makes everything stop working properly
Apr 20 18:11:38 <kerio> because the design is FUCKING AWFUL
Apr 20 18:11:38 <Sicelo> maemo bugs make for amusing reading sometimes :)
Apr 20 18:11:54 <Sicelo> kerio: doyou know how to even restart it?
Apr 20 18:12:00 <kerio> huh
Apr 20 18:12:04 <kerio> service dbus restart?
Apr 20 18:12:06 <kerio> or something like that
Apr 20 18:12:14 <KotCzarny>  /etc/init.d/dbus restart?
Apr 20 18:12:53 <Sicelo> i don't remember if i've tried that, and what happened thereafter
Apr 20 18:13:15 <KotCzarny> most likely you will get a reboot
Apr 20 18:13:36 <KotCzarny> because of watchdog killing things
Apr 20 18:14:51 * KotCzarny got me 128GB sd card for banana
Apr 20 18:15:08 <KotCzarny> gonna try upgrading a package or two for n900
Apr 20 18:15:39 <KotCzarny> especially that i can have working compile env for maemo in few hundred megs now
Apr 20 18:16:14 <KotCzarny> though binutils from lenny backports failed
Apr 20 18:19:34 <Sicelo> another N900 thing to understand is dsme ... is it documented somewhere?
Apr 20 18:20:21 <KotCzarny> http://wiki.maemo.org/Documentation/Maemo_5_Developer_Guide/Architecture/System_Software#Device_State_Management_Entity_.28DSME.29
Apr 20 18:22:36 <Sicelo> thanks
Apr 20 18:22:43 <KotCzarny> https://wiki.maemo.org/Free_Maemo#DSME
Apr 20 18:23:04 <KotCzarny> apparently source code for it is available
Apr 20 18:24:04 <KotCzarny> after 3 kills dsme will sw reset the device
Apr 20 18:24:15 <KotCzarny> (from random google snippets)
Apr 20 18:27:54 <Sicelo> anyway, got to make food. this N900 situation doesn't seem easily solvable for now. time to move to iPhone, lol
Apr 20 18:28:14 <KotCzarny> reflash before considering such drastic move
Apr 20 18:28:36 <KotCzarny> and try to beat 10 days on cssu-stable
Apr 20 18:29:21 <Sicelo> lol, just a joke. i have no interest in iphone
Apr 20 18:29:39 <DocScrutinizer05> https://bugs.maemo.org/show_bug.cgi?id=9150#c44  hw bug? interference from cmt TX into PVR?
Apr 20 18:29:41 <povbot> 04Bug 9150: Device doesn't respond via UI. syslog reports HWRecoveryResetSGX: SGX Hardware Recovery triggered, sgx_misr eating all CPU
Apr 20 18:30:23 <Sicelo> DocScrutinizer05: i've been reading it
Apr 20 18:30:41 <DocScrutinizer05> you didn't read my comment before
Apr 20 18:30:50 <Sicelo> which?
Apr 20 18:30:59 <DocScrutinizer05> hw bug? interference from cmt TX into PVR?
Apr 20 18:31:38 <DocScrutinizer05> could also be ripple on Vbatt due to cmt eating high amounts of current
Apr 20 18:32:27 <DocScrutinizer05> https://bugs.maemo.org/show_bug.cgi?id=9150#c45 is prolly a red herring
Apr 20 18:32:29 <povbot> 04Bug 9150: Device doesn't respond via UI. syslog reports HWRecoveryResetSGX: SGX Hardware Recovery triggered, sgx_misr eating all CPU
Apr 20 18:33:23 <Sicelo> for that bug, most of them seem to have a map program of sorts doing things in background. i wasn't using any maps program however
Apr 20 18:34:04 <Sicelo> Woody reported incoming SMS as being involved .. i didn't have that either
Apr 20 18:35:40 <DocScrutinizer05> >>  I might be able to provide a way to reproduce this bug reliably, perhaps using home router to create poor network conditions when the app is trying to load map images etc.<< in #48 is exactly to the point I guess
Apr 20 18:36:34 <DocScrutinizer05> when it's Vbatt ripple, an old battery will massively increase likelyhood of the bug
Apr 20 18:37:23 <Sicelo> i understand that. just wonder what would have 'rippled' me in this case though
Apr 20 18:37:42 <DocScrutinizer05> when it's EMI/RFI then the signal quality/level of 2G (particularly 2G, while 3G might be less susceptible) is relevant
Apr 20 18:38:10 <DocScrutinizer05> Sicelo: even mere phone does TX every now and then
Apr 20 18:38:20 <DocScrutinizer05> do yoz have 2G or 3G?
Apr 20 18:38:24 <DocScrutinizer05> you*
Apr 20 18:38:52 <DocScrutinizer05> how's your sigbal level?
Apr 20 18:39:11 <Sicelo> 3G. i was connected to wifi at the time, full strenght. also full strength on 3G
Apr 20 18:39:20 <DocScrutinizer05> hmmm
Apr 20 18:39:34 <KotCzarny> was it in the vicinity of red bar?
Apr 20 18:39:54 <DocScrutinizer05> then it's pretty unlikely that power bursts on cmt are the reason, neither TX RD
Apr 20 18:39:59 <DocScrutinizer05> RF
Apr 20 18:40:03 <Sicelo> not red. it became red maybe 20 minutes after the reboot
Apr 20 18:40:14 <DocScrutinizer05> err?
Apr 20 18:40:29 <KotCzarny> old battery might fluctuate more
Apr 20 18:42:33 <Sicelo> i'll monitor this closely, & try get hold of another battery
Apr 20 18:43:36 * Sicelo goes for a shower & food
Apr 20 18:53:05 <DocScrutinizer05> W*T*F?!  >>  it can add a /etc/powervr.d/ file[1] where it sets a suitable command buffer size for itself (I have no idea whether there's documentation on this, maybe Imagination www-site has something). [1] "strace -e open -f <application>" will tell which files SGX libs try to open.<<
Apr 20 18:57:06 <DocScrutinizer05> >>  I /don't/ see it anymore, still on PR1.2. Maybe coincidence, but I think it stopped when I stopped leaving GPS active 24/7.<< again a cmt domain thing that only possibly could have impact to system via Vbatt
Apr 20 19:25:05 <Linkandzelda> what can i do with a 2nd n900, anything cool?
Apr 20 19:25:33 <Sicelo> alternative OS, upstream kernel with Maemo, etc.
Apr 20 19:27:24 <Linkandzelda> hmm
Apr 20 19:33:00 <DocScrutinizer05> Linkandzelda: I once did an extremely cool setup with IRreco, using one N900 as the UI input device, and a 2nd N900 placed in front of my TV playing back the IR signal using lirc server, both communicating via WLAN
Apr 20 19:33:25 <DocScrutinizer05> you also can remote-access the camera of one N900 and show the image on 2nd
Apr 20 19:34:10 <Linkandzelda> oh, thats cool
Apr 20 19:34:10 <DocScrutinizer05> or you simply keep one device always-attached to your PC to always have a ssh session to maemo for testing etc
Apr 20 19:34:53 <DocScrutinizer05> many devels do the latter
Apr 20 19:36:51 <DocScrutinizer05> I also used a N900 with a shellscript and ssh to log ambient light changes (actually a display of an appliance) going off for 0.5s at (what I thought was) random intervals
Apr 20 19:37:55 <DocScrutinizer05> people used N900 as livestream camera to their website
Apr 20 20:56:41 <KotCzarny> Linkandzelda: connect one of the n900 to stereo system, install some server-remote type audio player and have an intelligent audio setup/internet radio
Apr 20 20:59:06 <Linkandzelda> not a bad idea indeed
Apr 20 21:02:21 <Sicelo> incidentally, i've been looking at icecast2 for n900. we've got the server in the repos
Apr 20 21:02:25 <Sicelo> and also a client, qice
Apr 20 21:06:53 <KotCzarny> it could also work as a video player in similar setup too, though video quality is low via cvbs cable
Apr 21 01:49:47 <DocScrutinizer05> Sicelo: so what's your status regarding https://bugs.maemo.org/show_bug.cgi?id=9150 ?
Apr 21 01:49:50 <povbot> 04Bug 9150: Device doesn't respond via UI. syslog reports HWRecoveryResetSGX: SGX Hardware Recovery triggered, sgx_misr eating all CPU
Apr 21 01:54:31 <DocScrutinizer05> Sicelo: ((another N900 thing to understand is dsme)) it's rather simple, except a few detail things around dbus. dsmetool --help
**** ENDING LOGGING AT Thu Apr 21 02:59:58 2016