dreamer | lol @ dremear | 08:35 |
---|---|---|
buZz | :) | 12:25 |
freemangordon | Wizzup: | 12:54 |
freemangordon | Jun 7 13:47:28 localhost wpa_supplicant[27089]: wlan0: Reject scan trigger since one is already pending | 12:54 |
freemangordon | Jun 7 13:47:28 localhost iap_conndlg 2.88+0m5[1233]: connui_dbus_mcall_reply(): Failed DBUS method call with error 'fi.w1.wpa_supplicant1.Interface.ScanError': Scan request rejected | 12:54 |
Wizzup | freemangordon: is that when scanning for just one AP? | 12:54 |
freemangordon | yes | 12:54 |
Wizzup | Maybe try again after a few seconds? | 12:54 |
Wizzup | I think only one scan can be active at once | 12:55 |
freemangordon | seems plugin scans as well in background | 12:55 |
Wizzup | It does when icd2 tells it to scan | 12:55 |
Wizzup | It doesn't just scan out of it's own, afaik | 12:55 |
freemangordon | ok, will see what's going on | 12:55 |
Wizzup | I had this in my wpa_supplicant.conf for a while but I don't think it helps: | 12:56 |
Wizzup | autoscan=periodic:-1 | 12:56 |
Wizzup | there is some more that we can do on the control socket with wpa, about auto scanning | 12:58 |
Wizzup | but I think you'll have to deal with this problem regardless | 12:58 |
Danct12 | Good afternoon | 13:00 |
Wizzup | hi | 13:00 |
Danct12 | How are you? | 13:01 |
Wizzup | Pretty decent, how about you? | 13:08 |
Wizzup | freemangordon: the above (about autoscan) is just about wpa_supplicant deciding to scan on its own, but I think that is mostly not happening (anymore) | 13:26 |
Wizzup | but if you have some conndlg open, that schedules scans every now and then | 13:26 |
Danct12 | Wizzup, I'm good. Decided to recharge my Fxtec Pro1 for some fun today | 14:16 |
Danct12 | I wonder when we'll have a filebrowser on Leste | 14:18 |
uvos | Danct12: liri-files works fine and is touch optimized and in extras. | 14:21 |
Wizzup | Danct12: hamsterfiler would be pretty simple to port, too | 14:25 |
Wizzup | freemangordon also said he might port the 'original' one some time too\ | 14:25 |
bencoh | hamsterfile <3 | 14:32 |
bencoh | I don't really see a reason to port the stock filer once we have hamsterfile :) | 14:32 |
Wizzup | I've personally never used it but I think it shouldn't be too hard to port | 14:33 |
bencoh | it depends on qt4 | 14:34 |
Wizzup | yeah, it needs porting to qt5 | 14:35 |
Wizzup | but it's by the same author whose other qt4 stuff I already ported to qt5 :p | 14:35 |
bencoh | did devuan drop qt4 altogether? | 14:35 |
Wizzup | we didn't bother including the maemo qt4 port | 14:35 |
bencoh | alright | 14:35 |
bencoh | was it that significant btw? | 14:36 |
Wizzup | mostly mean that the author uses similar constructs that need porting in different sw | 14:36 |
Wizzup | so I'm familiar with the work required, is all I meant | 14:36 |
Wizzup | oh, I misread | 14:36 |
bencoh | yeah, I was referring to the maemo qt4 port, sorry (terrible phrasing / timing) | 14:36 |
Wizzup | I don't know, feel free to take a look at it :-p | 14:37 |
bencoh | :) | 14:37 |
Wizzup | I don't remember why we made the decision at the time, but I think the idea was just to look forward and port stuff over | 14:37 |
bencoh | yeah I guess so | 14:37 |
bencoh | application-wise qt4 would have allowed to just rebuild old apps "as-is", but I see your point | 14:37 |
Wizzup | yeah, we can still do it, although qt4 will be dropped soon from debian I think | 14:38 |
bencoh | ah | 14:38 |
bencoh | in that case ... | 14:38 |
Wizzup | tbh what would be really useful is someone to help me with the QML stuff for openmediaplayer | 14:38 |
Wizzup | I did a lot of the porting but stuck on qml | 14:38 |
Wizzup | ah, need to run, bbl | 14:38 |
bencoh | I kinda walked away from QML when I realized it had some serious power-related issues | 14:39 |
bencoh | (it might be fixed in qt5) | 14:39 |
uvos | qml remains and allways will be a pig | 14:43 |
freemangordon | Wizzup: as soon as you click on the "hidden wlan", connui dialog gets closed | 15:00 |
freemangordon | but I'll double-check who is initiating the scan | 15:01 |
freemangordon | but yeah, most-probably it is icd2's timer expired while waiting for me to pres something in the dialog | 15:02 |
freemangordon | Wizzup: hmm, shall I do spmething special when I scna for a hidden wlan? | 15:18 |
freemangordon | *scan | 15:18 |
freemangordon | Where I am supposed to get the result from? | 15:20 |
Wizzup | freemangordon: I think you need to provide the ssid | 16:26 |
freemangordon | Wizzup: I am https://pastebin.com/UzW6pLEL | 18:07 |
freemangordon | and I am getting "Scanning" and "ScanDone" signals | 18:08 |
freemangordon | then what? | 18:08 |
Wizzup | freemangordon: I think you need to monitor for 'NetworkAdded' signals | 18:27 |
Wizzup | as wpaicd.c does (which I suggested we could share the code from) | 18:27 |
freemangordon | I see | 18:28 |
Wizzup | you might also be able to just get a list of all the registered Bss'es after scandone | 18:28 |
freemangordon | hmm, hidden wlan does not appear there | 18:34 |
freemangordon | neither I see NetworkAdded signal | 18:35 |
freemangordon | actually I see BSSID with empty SSID after that scan | 18:37 |
freemangordon | Wizzup: anyu clue about that comment https://chromium.googlesource.com/chromiumos/platform/flimflam/+/refs/heads/0.11.257.B90/plugins/newwifi.c#1289 ? | 18:41 |
Wizzup | Not really | 18:56 |
Wizzup | this is connman ? | 18:56 |
gordon1 | that commit causes n900 to panic if you boot kernel with flipper open fb2c599f056640d289b2147fbe6d9eaee689f1b2 ARM: omap3: enable off mode automatically | 18:59 |
gordon1 | which is strange and weird coincidence | 19:00 |
gordon1 | but probably problem is somewhere else in off mode code, just it gets triggered during boot with this commit | 19:02 |
gordon1 | any particular place i need to report it? | 19:08 |
sicelo | linux-omap ML, I guess | 19:20 |
sicelo | Anyway, weird ... mine didn't have issues with that commit, and booting with cover open | 19:21 |
sicelo | If that's what you mean by "flipper" | 19:22 |
gordon1 | i mean flip open the keyboard | 19:23 |
gordon1 | if phone is closed it boots ok, if you open the keyboard - it might panic, and if you flip open the keyboard right at the moment when kernel just initialized the framebuffer and continue booting rest of the stuff - it 100% guaranteed panic | 19:25 |
uvos | yeah linux omap ml | 19:51 |
uvos | also tmlind ^^^ | 19:51 |
uvos | any ideas? | 19:51 |
uvos | can you extract the panic? | 19:52 |
uvos | via serial or so? | 19:52 |
uvos | the trace i mean | 19:52 |
gordon1 | nope | 20:01 |
gordon1 | that's why i opted out to bisecting it | 20:01 |
gordon1 | i mean i could try usbtty thing from u-boot if that would work as emergency console | 20:01 |
gordon1 | or i'm open to ideas how else that could be done | 20:02 |
gordon1 | i mean technically i can laser cut a jig for pogo pins, but that's going to happen only next weeked or even later | 20:03 |
gordon1 | should be ok if i would use 3V UART for RX only? | 20:07 |
uvos | yes | 20:07 |
gordon1 | ST_SDRC ST_OMAPCTRL | 20:34 |
gordon1 | two bits reported by idlest1_core | 20:34 |
gordon1 | no idea what it means | 20:35 |
gordon1 | but i have pretty much nothing running and i did UART suspend as per sicelo suggestion | 20:36 |
gordon1 | do i need to turn on "opportunistic sleep" btw, or is CPU PM managed on some different level? | 20:45 |
L29Ah | gordon1: just smash a few pogo pins with a blob of PCL into N900 :) | 20:48 |
L29Ah | would still need an adjustable PSU instead of the battery tho | 20:48 |
gordon1 | i have my pogo pins in the same place where i have a laser cutter, so it doesn't matter | 20:48 |
gordon1 | yeah, exactly | 20:49 |
sicelo | Or can use the battery in same way sre did (that's what I will do) | 20:52 |
L29Ah | cut the plastic housing and insert it at an angle? | 20:53 |
freemangordon | Wizzup: oh, shit, SSID should not be zero terminated :( | 21:30 |
Wizzup | freemangordon: ah :-) | 21:30 |
Wizzup | gordon1: if others can reproduce your problem, I could try to reproduce it on my n900 serial | 21:31 |
gordon1 | Wizzup: you can try to use this config on 5.12.9 and try to boot it with keyboard open or flip keyboard open right after the moment when framebuffer is initialized http://ix.io/3p5z | 21:33 |
gordon1 | took my sources from right here https://cdn.kernel.org/pub/linux/kernel/v5.x/linux-5.12.9.tar.xz | 21:33 |
Wizzup | I can't try it right now, but I can try it a bit later | 21:34 |
gordon1 | sure, no rush | 21:35 |
sicelo | (I always boot with keyboard open ... since I didn't take time to make boot.scr. so I always boot with manually typing "run sdboot" each time) | 21:37 |
gordon1 | i modified u-boot to do run sdboot first in bootmenu, easier than to find out why saveenv is not present | 21:43 |
sicelo | Right. So the point is, booting with keyboard open, I didn't have that problem with 5.9-5.12 ... of course, maybe other things at play, who knows | 21:49 |
sicelo | Thanks for following up on this :-) | 21:50 |
* sicelo appreciates any N900 work | 21:50 | |
Wizzup | sicelo: can you check if you have the problem with 5.12.9, or did you include that one? | 21:51 |
Wizzup | it's also possible that using a different u-boot somehow causes trouble | 21:51 |
sicelo | I think my last tests were up to 5.12.7 | 21:51 |
gordon1 | cat arch/arm/boot/zImage arch/arm/boot/dts/omap3-n900.dtb >zImage && bash | 21:53 |
gordon1 | ./scripts/mkuboot.sh -A arm -O linux -C none -T kernel -a 0x80008000 -e 0x80008000 -n 'pmos' -d zImage uImage | 21:53 |
gordon1 | that's how i assemble image, if that is something that might be in play | 21:53 |
sicelo | Same | 21:53 |
gordon1 | and that's u-boot commit 25db3403094a57320644544eb0f38b1670a6716b | 21:54 |
gordon1 | oops | 21:54 |
gordon1 | wrong one | 21:54 |
gordon1 | that one 89be8e31ccd1c53b010385ed0807eb00f0eec06a | 21:54 |
sicelo | I mostly use fremantle old u-boot, and also did use wizzup's uboot from maedevu. No issue with both | 21:54 |
Wizzup | might be worth trying somehow | 21:55 |
Wizzup | uvos: posting it tomorrow | 22:04 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!