**** BEGIN LOGGING AT Fri Apr 29 02:59:56 2022 Apr 29 05:49:49 there's a quite tricky part in BT regarding audio iirc Apr 29 05:53:22 grr Apr 29 05:53:30 basically analog audio gets fed to BT chip instead of headphone amp? Apr 29 05:53:33 my memory seems foggy Apr 29 09:26:52 FM audio? That seems to go to the audio codec, so i guess that part's easy enough Apr 29 09:27:42 It's just that for some reason, the fm part doesn't operate when the bt side is off. However, the product brief says it does. Apr 29 09:28:52 So either we suffer because ours is Nokia custom, or it's the way the chip is connected/controlled Apr 29 18:48:33 or the kernel drivers(?) are "stupid"? I know they share one interface for the chip, maybe layering / dependency conflicts? Apr 29 18:49:01 the chip isn't Nokia custom afaik Apr 29 18:50:09 I'd suspect "killswitch" for BT TX is simply too brute force, shutting down the whole chip as much as possible Apr 29 19:42:48 I hope it's not the drivers :-) Apr 29 19:43:43 I guess you're right about it not being custom. I found that the bcm2048 is used on a couple of nokia feature phones, one or two lg phones as well as a portable media player Apr 29 19:44:31 Maybe I will make time to go through the driver code, and see what happens if we skip the Bluetooth check (it's in the driver) Apr 29 19:46:07 Pity there's zero documentation for it ... I'd try manual i2c Apr 29 19:55:02 joerg: but i remember on tmo, the dev who made the first working radio application for n900 (Martin Grimme, I think) specifically mentioned that there was no way for it to work without bt on Apr 29 20:12:36 https://www.broadcom.com/site-search?q=bcm2048 Apr 29 20:24:10 sicelo: sorry, it was FMTX that's connected to headphones lines https://i.imgur.com/zGcDxwv.png Apr 29 20:27:34 BTFM chip (bcm2048) has FM_AUDIO_PR FM_AUDIO_PL that go straight to audio codec, unrelated to BT audio PCM interface Apr 29 20:29:07 Yes Apr 29 20:29:54 It's the activation of the fm function that's bound to bt Apr 29 20:34:49 The product brief says, "fm control through either bsc compatible interface or bluetooth hci" Apr 29 20:35:52 there been a http://www.broadcom.com/products/Bluetooth/Bluetooth-RF-Silicon-and-Software-Solutions/BCM2048 but it's gone Apr 29 20:35:59 maybe archive.com? Apr 29 20:37:45 Isn't that the 2-page product brief? Seems Broadcom only publishes briefs :-) Apr 29 20:40:14 https://web.archive.org/web/20100107153109/http://www.broadcom.com/products/Bluetooth/Bluetooth-RF-Silicon-and-Software-Solutions/BCM2048 Apr 29 20:53:37 no idea since I can't check anymore, since... it's gone Apr 29 20:54:44 I had a bookmark for this, no other bookmarks related to BCM2048 - I guess whan I didn't find anything back when, it's futile to get locked in to a search session like I just did Apr 29 20:54:57 excuse me, what was the *question*? Apr 29 20:56:09 found https://www.nokiarepairchicago.us/n85/bluetooth-and-fm-rds-radio-module.html and https://html.datasheetq.com/pdf-html/306820/Broadcom/1page/BCM2048.html?lang=en Apr 29 20:56:52 https://www.embedded.com/fm-radio-joins-with-bluetooth-on-same-silicon/ Apr 29 20:59:43 http://www.mjdic.com/proclass-read-id-1075910.html Apr 29 21:02:41 and of course https://elinux.org/N900 Apr 29 21:04:10 https://elixir.bootlin.com/linux/v4.8/source/drivers/staging/media/bcm2048/radio-bcm2048.c Apr 29 21:04:16 aaaaand I'm out Apr 29 21:04:26 sorry Apr 29 21:12:20 the weirdest thing is Broadcom outright denies now that chip ever existed Apr 29 21:14:14 a "nice" example why eventually you have to bin all your dsigns and start anew from scratch since building the old design simply became impossible due to sourcing problems Apr 29 21:16:53 Fun :-) Apr 29 21:25:39 i guess this is what you were referring to earlier - "There is no direct bypass from the FMRX hardware to the sound system (or more precisely, there is, but it cannot be used since it does not seem to filter high frequency sounds that may damage the speakers) so the sound has to be processed by the CPU." good to know this actually Apr 29 22:51:32 how to fix the /topic to deal with infobot no more? Apr 29 22:54:45 *sigh* btdt err dtat, particularly I asked http://reisenweber.net/irclogs/libera/_maemo/_maemo.2022-01-01.log.html#t2022-01-01T18:43:13 Apr 29 22:58:45 http://reisenweber.net/irclogs/libera/_maemo/_maemo.2021-01-16.log.html#t2021-01-16T18:08:09 ff has a few of them Apr 29 23:16:14 I don't see an issue with a new bot even if some of the db is lost. Apr 29 23:18:16 all of the db is lost, so the purpose and usefulness of a bot is pretty limited Apr 29 23:18:42 might come in handy: http://reisenweber.net/irclogs/libera/_maemo/search?q=%28%3Fi%29%5Einfobot+%2B.*unbanip Apr 29 23:19:31 replace "unbanip" by whatever searchterm you are interested in and hope infobot used to know about it Apr 29 23:20:08 most useful ones can be re entered the data is in the logs. Apr 29 23:20:15 *** The channel topic is "Welcome to #maemo http://maemo.org/intro/ | "firmware" see ~lf (or ~tabletsdev ~combined ~emmc ~flasher) sdk: ~sb | http://wiki.maemo.org/Community_Council | Source: http://wiki.maemo.org/Sources | Chanlog: http://reisenweber.net/irclogs/libera/_maemo (legacy: http://mg.pov.lt/maemo-irclog )| #maemo-ssu is where the (few) devels and maintainers of the maemo-future meet". Apr 29 23:20:44 sixwheeledbeast: sure Apr 29 23:22:40 actually my local logs might have like 95% of the relevant factoids ;-) in infobot query log where I defined them **** ENDING LOGGING AT Sat Apr 30 02:59:56 2022