libera/#maemo-leste/ Monday, 2024-10-07

Wizzupok, did that00:34
sicelo:-)00:37
siceloand second workqueue method seems to be working beautifully on my end00:37
sicelobut must sleep now00:37
Wizzupgreat, gn00:42
inkyhttps://mikrophone.net/15:10
inkylets hope they succeed and we'll have a maemo port.15:10
Wizzupmicrocontroller doesn't sound like something that would run linux15:18
uvos__not only that15:44
uvos__never mind that it cant run linux, its specs are laugable even for a uc15:44
uvos__16kB Data SRAM15:44
uvos__thats small 8bit uc territory15:45
uvos__specs are kinda simmilar to something like an old nokia 3210 style feature phone for referance15:48
uvos__its allso a pretty wierd and seemingly senslessly idelogically motivated move to use a riscv cpu as your main processor15:51
uvos__but then an esp32 for wifi, wich is almost certenly more powerfull than your riscv soc15:51
uvos__you could just use the esp32 with hw wifi for evertthing and be better off15:52
sicelohttps://github.com/maemo-leste/droid4-linux/pull/13/16:34
sicelohere's updated MR ... it builds up on the previous commit16:35
sicelofreemangordon: please also review16:35
siceloi've tested this extensively on my Leste N900, and seems to work with 95% reliability.16:36
siceloit's not 100% because there's definitely something wrong somewhere ... maybe hardware? or ke-recv? The time it takes charger chip to refresh its internal registers keeps increasing somewhat, especially when connected to a PC. seems much more reliable on a charger or power bank.16:37
siceloi set the delay to 9s in the MR. i'm stable on 8s actually, but hoping one more sec will improve the detection a bit more. 7s didn't help much when connecting to a PC. might have worked fine from power bank but i didn't have access to one at the time16:42
sicelowill send it upstream for further review in a couple of days (hope arno11 will have given it a go too)16:52
Wizzupugh hildon-thumbnailer packaging is kind of a mess21:14
uvosWizzup: the audio problem on pasteur is simply down to the fact that pasteur variants have no modem audio21:18
uvosnaturally the setup of the cpu<->modem and cpcap<->modem dais fails21:18
Wizzupuvos: ok, so we need to modify the dts for that?21:21
Wizzupthat makes sense btw21:21
uvosyeah we need to move all the modem audio related stuff from motorola-mapphone-common to motorola-mapphone-handset21:22
uvosam working on it21:22
Wizzupcheck :)21:23
uvosactually mz616 might have modem audio21:27
uvossince it shares a modem setup with xt91021:27
uvosbut i cant test that theory21:27
uvosWizzup: btw21:31
uvosfbcon=rotate:1 needs to go in boot.cfg on mz61721:31
uvosthats just silly21:31
Wizzupwhen you say that's just silly, do you mean it's not important but good to do?21:33
uvosit needs to be removed21:33
uvossomeone copy pasted that from d421:33
uvosbut mz61x is native landscape21:34
uvosso we are makeing it portrait21:34
uvosok21:40
uvosdts changes made21:40
uvosi have audio21:40
uvosbut only via the headphones21:40
uvosthe speakers are probubly to large to drive directly from cpcap so they added an external amp that needs to be enabled21:41
uvoswhat do you know the mz617 signal map has a gpio the d4s dosent21:46
uvosexternal_amp_en     19121:46
uvossounds very suspicous21:46
Wizzupah needs to go, yes21:46
Wizzupuvos: sweet @ headphones!21:46
Wizzupuvos: suspicious as in likely the right value? :D21:47
uvosalmost cerntenly thats the right gpio yeah21:50
uvosim just struggeling here to remember how the gpio numbers from the stock kernel corrspond to the gpios on the mainline kernel21:52
WizzupI have some vague memory but it's probably more helpful if I don't help :D21:59
Wizzupas in it's probably not helpful enough21:59
uvosyeah i got that far allready22:00
uvosits &gpio5 3122:00
tmlinduvos: only am3 gpio banks start at 0 i recall, so should be &gpio6 31 on mapphones22:25
tmlindnice if you get the audio working, ttyl22:25
Wizzupuvos: btw with the headphones, do you also hit ret?22:36
uvosWizzup: nope22:38
uvosnot related22:39
Wizzupok, too bad :D22:42
uvosyeah it works23:19
Wizzupsweet :)23:19
uvosbut i have yet to figure out how to make dapm do what i want (again)23:19
uvosso i just pinned the amp on23:19
uvosunfortionatly it has an annoying buzz when cpcap is inactive but the amp is enabled23:20
uvosso i need to figure this out23:20
Wizzuphmm23:21
uvosin snd/soc/codecs/cpcap.c function cpcap_input_right_mux_put_enum i have no idea how to associate the parameters with cpcap_spkr_l_mux_enum and so on23:25
uvosreally all we need to do there is set the amp gpio high when the mux is set to output to speaker on either channel23:25
uvosbut snd/soc is complicated as usual23:25
inky> you could just use the esp32 with hw wifi for evertthing and be better off23:27
inkytrue23:27
uvosWizzup: here be wip branch if you are interested: https://github.com/IMbackK/droid4-linux/tree/mz617wip23:32
uvosWizzup: btw please kick mapphone-kexecboot-config23:38
uvosi removed the framebuffer rotation on mz61723:38
uvosanyhow ttyl23:38
Wizzupty, ttyl, will do this tonight23:47

Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!