| freemangordon | Wizzup: I assumed you will spin a new build | 00:24 |
|---|---|---|
| freemangordon | I have fixed the power usage | 00:24 |
| Wizzup | freemangordon: ok, I will do it tomorrow | 02:39 |
| freemangordon | heh, tracker-sparql is so smart... it caches up to 100 queries | 10:37 |
| freemangordon | so, eventually, at some point memory usage will stop increasing :D | 10:37 |
| freemangordon | unfortunately there is no way to disable the cache | 10:38 |
| freemangordon | and 100 is hardcoded | 10:38 |
| freemangordon | meh | 10:38 |
| Wizzup | what is that cache for? | 11:08 |
| freemangordon | sql queries | 11:09 |
| freemangordon | umm... | 11:09 |
| freemangordon | sql3 prepared statements | 11:09 |
| freemangordon | it keeps a cache of LRU queries and the prepared statemets | 11:10 |
| freemangordon | which is not that bad if you have 16G | 11:10 |
| Wizzup | well, we can just decrease this number since we build it oursleves anyway | 11:11 |
| freemangordon | but now, even after I have massaged it a bit, in the VM it uses ~230MB RES for ~ 1800 songs | 11:11 |
| freemangordon | well, yeah, but I wanted to avoid that if possible | 11:12 |
| Wizzup | wait so it also caches the results of the queries? | 11:14 |
| freemangordon | no | 11:15 |
| freemangordon | at least that's my understanding | 11:15 |
| freemangordon | as it does sqlite3_reset() on cursor close | 11:16 |
| freemangordon | but, it uses queries as hash table key | 11:16 |
| freemangordon | I don;t know what is left after sqlite3_reset() | 11:17 |
| freemangordon | https://www.sqlite.org/c3ref/reset.html | 11:17 |
| freemangordon | but obviously a lot, so 20 of those take ~180 MB | 11:17 |
| freemangordon | crowdfail, eh? | 11:47 |
| Wizzup | hm? | 11:51 |
| Wizzup | going to build new 6.6 now | 11:51 |
| freemangordon | Wizzup: google "crowdstrike bsod" | 11:53 |
| freemangordon | ;) | 11:53 |
| freemangordon | I guess most of the windows machines are now there | 11:54 |
| Wizzup | Ah | 11:54 |
| freemangordon | no, really, how is a single company allowed to have such a power to me is a mistery | 11:55 |
| freemangordon | not to say that pushing production updates on friday is crazy | 11:55 |
| Wizzup | there's many more companies with such power now :D | 11:55 |
| freemangordon | yeah | 11:55 |
| freemangordon | if it is not those guys then it will be some other "antivirus software" vendor | 11:56 |
| freemangordon | the whole "windows security" is... | 11:56 |
| Wizzup | of android, or ios :D | 12:03 |
| Wizzup | or* | 12:03 |
| Wizzup | freemangordon: when this works I think we can move from experimental to devel, agree? | 12:12 |
| arno11 | Wizzup: for d4, sounds good, but for n900 there is still the emmc swap issue making it buggy and unusable after few hours. | 12:17 |
| arno11 | the easiest way could be to use sd swapfile by default | 12:17 |
| arno11 | (temporary) | 12:17 |
| Wizzup | yes, I have the pr open | 12:19 |
| Wizzup | although I guess that's not for the other swap location | 12:19 |
| Wizzup | we have the postinst rules for n900 in leste-config to update fstab, so we could do that | 12:20 |
| arno11 | yep but i suppose first boot will be longer if it has to create the swapfile or ? | 12:21 |
| arno11 | (for other swap location: i didn't notice any issue after 3.5 days with sd swap and no reboot) | 12:27 |
| Wizzup | arno11: don't we have swap partition? | 12:34 |
| Wizzup | maybe we don't | 12:35 |
| arno11 | we don't | 12:35 |
| arno11 | indeed, sounds good to have | 12:36 |
| Wizzup | well we can't just repartition :) | 12:36 |
| arno11 | otherwise we have to figure out what's going on with emmc | 12:39 |
| freemangordon | I think that's the proper thing to do | 12:41 |
| arno11 | yep and it still works perfectly with 6.1 | 12:41 |
| freemangordon | Wizzup: I think 6.6 is not ready for devel | 12:42 |
| freemangordon | `because of reboot/shutdown issue | 12:42 |
| arno11 | what's the issue btw ? | 12:44 |
| Wizzup | freemangordon: what is the issue? | 12:48 |
| uvos__ | Wizzup: the device hangs on shutdown/reboot | 12:49 |
| uvos__ | i did not see this problem early in the 6.6 cycle | 12:50 |
| uvos__ | not sure what changed | 12:50 |
| uvos__ | but it dose hang here now | 12:50 |
| freemangordon | does here 19 of 20 times | 12:50 |
| uvos__ | for sure this dident happen at all when i initally rebased the kernel | 12:50 |
| uvos__ | back when cpufreq did not work | 12:51 |
| uvos__ | so either it came in with stable updates | 12:51 |
| uvos__ | or our additional patches since then | 12:51 |
| Wizzup | did someone connect serial and monitor kernel messages? | 13:02 |
| freemangordon | I don't have serial, afaik | 13:05 |
| Wizzup | oh | 13:23 |
| Wizzup | if you have a usb ttyl serial like pl2303 or so it's very easy to make | 13:23 |
| freemangordon | don;t have time now anyways | 13:23 |
| Wizzup | in any case. new kernel is in repos | 13:25 |
| Wizzup | I would argue it's ok for devel since it's devel, but I won't anymore then | 13:26 |
| freemangordon | well, we know there is issue | 13:28 |
| freemangordon | which is not minor though | 13:29 |
| freemangordon | so, ... | 13:29 |
| Wizzup | but audio works ;) | 13:31 |
| Wizzup | I haven't seen the device ever hang with 6.6 I think, but I will check now with serial | 13:31 |
| freemangordon | well, yeah, autio works :) | 13:31 |
| freemangordon | *audio | 13:32 |
| Wizzup | 'Will now restart.' and it restarts fine | 13:32 |
| freemangordon | you are lucky | 13:32 |
| freemangordon | lemme upgrade | 13:32 |
| Wizzup | lucky every time eh :D | 13:32 |
| freemangordon | mhm | 13:32 |
| freemangordon | here it hangs every time | 13:32 |
| freemangordon | well, reboots once in a while, but still | 13:33 |
| Wizzup | I have serial now, although I think due to how we disconnect the kernel messages from tty I might not see them on restart | 13:33 |
| Wizzup | how do you restart? | 13:33 |
| Wizzup | or power off? | 13:33 |
| freemangordon | sudo reboot | 13:34 |
| freemangordon | and power-off reboots sometimes | 13:34 |
| Wizzup | that has always ben the case for me (that power off sometimes reboots) | 13:34 |
| freemangordon | but reboot hangs | 13:34 |
| Wizzup | not for me, again | 13:34 |
| Wizzup | (just rebooted) | 13:34 |
| Wizzup | I might need to disable droid4-pm to get kernel messages upon shutdown if we think it is necessary | 13:35 |
| Wizzup | but right now I can't reproduce | 13:35 |
| freemangordon | upgrading the kernel, lets see after the upgrade | 13:35 |
| Wizzup | this is on the droid 4 yeah? | 13:35 |
| freemangordon | mhm | 13:35 |
| Wizzup | trying switch off from ui now | 13:36 |
| freemangordon | iphb was not rebuild upon upgrade | 13:36 |
| Wizzup | switch off worked fine too | 13:37 |
| Wizzup | freemangordon: got any console errors? I don't think I changed anything since that worked | 13:37 |
| freemangordon | no | 13:37 |
| Wizzup | ok, let me see if it is loaded for me then | 13:38 |
| Wizzup | I didn't see any errors | 13:38 |
| Wizzup | 'dpkg-reconfigure linux-image-omap' shows uninstall working ok, install also working ok | 13:41 |
| Wizzup | but it wasn't loaded on my first boot somehow | 13:41 |
| Wizzup | maybe the order in which headers are updated/installed vs when configure runs or something | 13:41 |
| arno11 | freemangordon: did you downgrade before upgrading ? (for iphb) | 13:41 |
| freemangordon | no | 13:41 |
| freemangordon | but it should work regardless | 13:42 |
| arno11 | (that's what i did to avoid errors) | 13:42 |
| arno11 | ok | 13:42 |
| Wizzup | as stated, could be that headers aren't installed timely enough by dpkg | 13:42 |
| Wizzup | there is no problem with kernel heads location or dkms | 13:42 |
| Wizzup | (as in, rerunning works fine) | 13:42 |
| freemangordon | no, it is not the headers | 13:43 |
| freemangordon | it should fail if it does not find the headers | 13:43 |
| freemangordon | it is something else | 13:43 |
| freemangordon | so, how long does it take to reboot your device? | 13:43 |
| Wizzup | what do you want timed excatly? | 13:43 |
| Wizzup | from when I type 'reboot' to when I see motorola logo? | 13:44 |
| freemangordon | as here it eventually rebooted, but I suspect it was WD | 13:44 |
| freemangordon | yes | 13:44 |
| freemangordon | poweroff hung it | 13:44 |
| Wizzup | 28s | 13:44 |
| freemangordon | like, I will have to reset it | 13:45 |
| Wizzup | seems like I can't help since I can't reproduce | 13:45 |
| freemangordon | :( | 13:45 |
| Wizzup | (with my serial cable I mean) | 13:45 |
| freemangordon | maybe bring me one | 13:45 |
| Wizzup | I can bring the cable on Monday, or make a few | 13:45 |
| Wizzup | ok | 13:45 |
| freemangordon | yes, please | 13:45 |
| freemangordon | new, old, does not matter | 13:46 |
| Wizzup | https://www.olimex.com/Products/Components/Cables/USB-Serial-Cable/USB-SERIAL-M/ | 13:46 |
| freemangordon | do you have SIm card in that device? | 13:46 |
| Wizzup | you need this + microusb cable | 13:46 |
| Wizzup | yes | 13:46 |
| Wizzup | I will order these then, since I also lack extra myself | 13:46 |
| freemangordon | hmm, leme see, I have one from pinephone I guess | 13:46 |
| Wizzup | yes but you'd have to destroy it | 13:46 |
| Wizzup | rx / tx goes to usb d+ and d-, gnd to gnd | 13:47 |
| freemangordon | is it usb-to-ttl what I need? | 13:47 |
| Wizzup | yes, I always use the pl2303 | 13:47 |
| freemangordon | I have one here, lemme see what it is exactly | 13:48 |
| Wizzup | olimex ships quickly so I think it'd best to just buy some and I'll bring my cable too next week | 13:48 |
| freemangordon | cp2102 | 13:51 |
| freemangordon | Wizzup: this https://elimex.bg/product/74867-kit-k2118-konvertor-usb-kam-uart-ttl-cp2102-6pn | 13:52 |
| Wizzup | if it is ttl level that is fine too, that's what matters | 13:52 |
| freemangordon | how do I power that? I need external 5V? | 13:52 |
| Wizzup | you don't connect power to the droid serial | 13:53 |
| Wizzup | power for the module comes from usb | 13:53 |
| freemangordon | ok | 13:53 |
| freemangordon | where is droid4 serial? | 13:53 |
| Wizzup | microusb port | 13:53 |
| Wizzup | so you need the olimex thing linked above and a microusb cable you can sacrifice | 13:53 |
| Wizzup | open them both up, and then solder three out of the four wires together, and close it up | 13:54 |
| freemangordon | I have one cut in 2 mieces :) | 13:54 |
| freemangordon | *pieces | 13:54 |
| freemangordon | and ? | 13:54 |
| freemangordon | what is the wiring? | 13:54 |
| Wizzup | gnd to gnd, and rx / tx to d- / d+, don't know which to which, tony has it on his website | 13:56 |
| Wizzup | http://muru.com/linux/d4/ | 13:56 |
| Wizzup | The wire connections are: | 13:56 |
| Wizzup | d-device uart txusb serial rx ftdi yellow, pl2303 white | 13:56 |
| Wizzup | d+device uart rxusb serial tx ftdi orange, pl2303 green | 13:56 |
| Wizzup | gnddevice groundusb serial gnd | 13:56 |
| Wizzup | vbusdo not connect | 13:56 |
| Wizzup | I just tried rx/tx and swapped when it didn't work and then soldered when it did work | 13:56 |
| freemangordon | ok, lemme try | 13:57 |
| freemangordon | do I have to do anything on the device to enable that serial console? | 13:57 |
| freemangordon | hmm, they say 3.3V | 13:58 |
| Wizzup | no, you just need to bash keys to get it out of pm and respond | 13:58 |
| Wizzup | but you sometimes have to plug in a few times | 13:58 |
| freemangordon | ok | 13:58 |
| Wizzup | I have to go, back later | 14:00 |
| freemangordon | ok | 14:00 |
| freemangordon | does not seem to work | 14:09 |
| freemangordon | yeah, it needs 115200 speed | 14:14 |
| * sicelo remembers that he still hasn't gotten around to making serial for N900 | 14:18 | |
| freemangordon | the fuck! now it works | 14:21 |
| freemangordon | I mean - reboot/shutdown | 14:22 |
| freemangordon | ok, no idea why, but now I see no issues with poweroff/reboot | 14:36 |
| Wizzup | maybe you're just waiting longer now? | 14:41 |
| freemangordon | no | 14:42 |
| freemangordon | Wizzup: what to do to be able to see powerdown console messages? | 14:43 |
| uvos | freemangordon: you need to enable the serial console in cmdline (/boot/boot/boot.cfg) and disable kernel console detach in the droid4-pm script | 15:30 |
| uvos | note that the latter has massive pm impiclations, it compleatly breaks all low power states in the omap4 | 15:30 |
| freemangordon | тханкс | 16:28 |
| freemangordon | thanks | 16:28 |
| dsc_ | :D | 16:39 |
| Wizzup | freemangordon: so were you able to reproduce? | 17:30 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!