| Tekk_ | hey, is it normal for the system to just randomly peg itself on CPU every few seconds? | 00:38 |
|---|---|---|
| Tekk_ | Like every, I dunno, 15 seconds, lipstick suddenly jumps to 600% cpu usage, then drops down a couple seconds later. | 00:39 |
| chriadam | Tekk_: the lipstick process? | 00:41 |
| chriadam | or a related process like lipstick-security-ui or the lipstick bluetooth agent? | 00:41 |
| Tekk_ | Lipstick itself. | 00:41 |
| chriadam | under what circumstance is this occurring? during active use, or while idling at homescreen, or? | 00:42 |
| Tekk_ | Sitting with the screen off next to me. | 00:42 |
| chriadam | screen off?! | 00:43 |
| chriadam | hr | 00:43 |
| chriadam | m | 00:43 |
| Tekk_ | I'm actually trying to figure out why it's refusing to play along with rsync and sftp | 00:43 |
| Tekk_ | so I grabbed htop. | 00:43 |
| chriadam | ooi do you see the same behaviour when bluetooth is disabled? and wlan? | 00:43 |
| Tekk_ | Only thing I can think of is that whenever lipstick has to check for events, the phone's clocked so low that any usage at all hits all the processors at 600% | 00:43 |
| Tekk_ | I can't test it without wlan. Bluetooth is off. | 00:43 |
| chriadam | potentially. I wonder what the iphb timeout is.. | 00:44 |
| Tekk_ | I suggest this because hitting buttons in htop can also make htop hit 600 | 00:44 |
| chriadam | Tekk_: this is a bit out of my area of expertise. maybe spiiroin can help more with this question ^ (will be 8 hours or so until he's in the office, though, I guess). | 00:46 |
| Tekk_ | No worries, it's just a weird thing I noticed :) | 00:53 |
| Tekk_ | Bit more concerned with not being able to scp/sftp/rsync though. It's a bit weird | 00:58 |
| Tekk_ | I can ssh in, but if I try to do anything *but* normal ssh it freezes after debug2: channel 0: open confirm rwindow 0 rmax 32768 | 00:59 |
| spiiroin | Tekk_ chriadam: first thing that comes to my mind: if the device suspends, bunch of short-lived timers can trigger at once when device resumes for any reason. and lipstick has bunch of timers. | 05:32 |
| spiiroin | ... also some input devices might report "nothing-going" on after resume -> triggers input handlers too | 05:32 |
| Tekk_ | Makes sense. | 05:37 |
| Tekk_ | Plus like I saw earlier, I bet the 600 isn't actually much cpu usage if the cores are all clocked way down, considering how pressing down in htop while the phone was idled would cause a tiny spike like that :) | 05:38 |
| spiiroin | Tekk_: rsync & sftp problems might be something generic with wlan, ssh case here: https://together.jolla.com/question/173984/sailfishx-ssh-session-very-laggy-wlan/ | 05:39 |
| Tekk_ | spiiroin: Yeah, it turns out my router was in death spasms. | 05:42 |
| Tekk_ | It seems to have calmed down after I reflashed it. | 05:42 |
| Tekk_ | Thing gets cranky sometimes. It's over 10 years old. | 05:43 |
| spiiroin | ;-) | 05:44 |
| *** Renault_ is now known as Renault | 07:25 | |
| *** frinring_ is now known as frinring | 09:43 | |
| x2s | monich: Hi. Do you know if the baseband modem of the Xperia X has a permanently flashed firmware or if the firmware is loaded at boot time by the kernel driver? | 10:00 |
| monich | x2s: honestly, I don't know | 10:02 |
| monich | or it could be stored in a special partition | 10:03 |
| x2s | then I'll have to find the kernel driver and find out :) | 10:03 |
| x2s | and being loaded by the bootloader? Probably | 10:03 |
| x2s | who could I ask? Who knows such things? | 10:04 |
| * monich shrugs | 10:04 | |
| r0kk3rz | iirc there is some modem firmware which needs flashing each time | 10:05 |
| x2s | flashing each time? | 10:07 |
| x2s | r0kk3rz: can you explain, what you mean? | 10:10 |
| r0kk3rz | not really | 10:16 |
| r0kk3rz | x2s: is there a problem you're trying to solve? | 10:29 |
| x2s | r0kk3rz: yes. Several users (there's a whole thread about this) have the problem that their Xperia X isn't able to reconnect to the network, after loosing it (e.g. in an elevator). It's necessary to restart ofono (I put it into flightmode and back) | 10:51 |
| x2s | https://together.jolla.com/question/170394/xperia-x-loss-of-mobile-network/ | 10:52 |
| x2s | some say it is fixed after they flashed the latest android version on it and then went back to sfos | 10:52 |
| x2s | hrm. Either desmo deleted the last comment or it got deleted. He/she claimed that reflashing with android doesn't help either | 10:55 |
| r0kk3rz | could be superstition | 10:56 |
| r0kk3rz | also flightmode != 'restart ofono' | 10:57 |
| x2s | iirc then monich said something different | 10:59 |
| x2s | hrm. The comment is listed here: https://together.jolla.com/users/11916/desmo/?sort=recent but when I click on the link to it it's gone | 11:00 |
| x2s | oh. You have to click on "show more"... this breaks anchors. | 11:01 |
| x2s | and then there's the comment from Direc, saying that this happens to their Xperia X with Android Oreo as well | 11:03 |
| x2s | .oO( now, where to get 10 Xperia X to test this? I have a very reliable way to trigger this... ) | 11:04 |
| r0kk3rz | it does sound much lower than sfos | 11:10 |
| r0kk3rz | like modem <--> RIL | 11:10 |
| r0kk3rz | flight mode may also restart rild | 11:11 |
| x2s | .oO( I've missed customer calls because of this... very annoying ) | 11:14 |
| x2s | and my knowledge doesn't go far enough to analyze this myself. I'm willing to experiment and test | 11:15 |
| x2s | it's "deep enough", isn't it? | 11:15 |
| Thaodan | x2s: does the x has a modem partition? | 12:00 |
| Thaodan | than yes | 12:00 |
| Thaodan | flashed | 12:00 |
| Tekk_ | x2s: You mean you can receive calls? ;) | 12:00 |
| Tekk_ | (by the way, you don't have to put it into airplane mode, just turn off and on cellular data.) | 12:00 |
| Tekk_ | Receiving a call triggers it about 75% of the time for me. | 12:01 |
| x2s | Thaodan: that's the question. I have no idea. | 12:10 |
| x2s | Tekk_: yeeesss? ;P I love my SfX phone, but this not reconnecting after network loss problem annoys me | 12:11 |
| Tekk_ | Get called less :p | 12:12 |
| Tekk_ | Works for me | 12:12 |
| x2s | Thaodan: I never understood this smartphone stuff or Android things. I look at it from an embedded sw dev and longterm linux user and it's a big mess. | 12:12 |
| Tekk_ | x2s: Oh, that's easy | 12:12 |
| Tekk_ | It's all the price constraints of real embedded, but the users demand shiny features :) | 12:12 |
| Tekk_ | And you can always fix your mistakes later on because there are OTA updates, so why bother with QC? | 12:13 |
| x2s | oh yes, and lets do all in java, because it's great! | 12:15 |
| Thaodan | x2s: just do lsblk on the x | 12:33 |
| x2s | Thaodan: http://paste.debian.net/hidden/0cc957f6/ that looks aweful... | 13:00 |
| * Tekk_ thinks he may be smelling some alignment requirements somewhere | 13:00 | |
| x2s | Tekk_: yeah... sadly this isn't done automatically :/ | 13:41 |
| x2s | (and is part of the Android leftovers) | 13:42 |
| rydare | do i have to do all the updates to do Sailfish X | 21:23 |
| rydare | or just until 34.3.a or whatever for Linux compatibility | 21:23 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!