**** BEGIN LOGGING AT Tue Aug 13 02:59:56 2019 Aug 13 06:30:24 Morning! Aug 13 07:26:36 Morning! Aug 13 09:14:11 morning Aug 13 09:34:18 bshah: ping Aug 13 09:49:08 Herrie: pong Aug 13 10:06:43 JaMa: Morning Aug 13 10:06:57 bshah: We can confirm that we also have issues with oFono, voicecall on Android 7 Aug 13 10:07:14 We fixed the issue we had on Android 5 (incorrect patching of voicecall). But Android 7 not working Aug 13 10:07:30 Tofe will debug the rilmodem driver a bit to see if he can track if down Aug 13 10:54:16 bshah: do you already have some debug clues? Aug 13 11:57:49 Tofe: not much unfortunately Aug 13 11:58:08 ok; I just didn't want to redo an investigation :) Aug 13 13:06:18 Tofe: I had a quick go with 1.30 just for the fun of it, seems all patches can be dropped that are in 1.29 (except for 0001-mbim-add-an-optional-TEMP_FAILURE_RETRY-macro-copy.patch), just need to update ell to 0.21 for it to build successfully due to gpio used by quectel.c Aug 13 13:14:07 Tofe: There's not a whole lot of things that are of interest to us in 1.30, just this one caught my eye: https://git.kernel.org/pub/scm/network/ofono/ofono.git/commit/?id=dc7fb359cbffb7787d3a03eb76f9460bd1498eb2 Aug 13 13:32:13 Herrie: ok, good to know, maybe we'll keep 1.29 or 1.30 instead of Mer's... I'll do my debug on 1.29, it seems it's the cleanest output Aug 13 13:33:06 For the commit you're pointed to, I didn't see any ofono crash lately, so I guess we're not really impacted by the fix Aug 13 17:21:14 Tofe: Well any fix we can probably upstream and upgrading to 1.30 or newer in future should be easy Aug 13 17:21:56 It said possible crash, so I guess it's not that common Aug 13 17:22:02 Maybe in some edge cases Aug 13 18:19:33 Damn, not sure I'm getting somewhere with this debug Aug 13 18:20:31 But it's probably related to a change in Android's RIL which the rilmodem driver doesn't handle well Aug 13 18:21:13 After all, it's based on a quite old code from UbuntuPorts, which might not have been maintained that much regarding newer Android versions Aug 13 18:22:06 I wonder if it wouldn't be faster to strip Mer's driver from the mce bits Aug 13 18:25:12 Still, I haven't yet compared the debug traces between tissot and hammerhead. They *should* be similar. Aug 13 18:26:11 Tofe: Well I doubt Mer reinvented the wheel from scratch Aug 13 18:26:22 They reworked the code quite a lot Aug 13 18:26:31 They probably took Rilmodem at some point and changed stuff Aug 13 18:26:39 Yes Aug 13 18:26:59 Not sure why they insist on their fork and not upstream Aug 13 18:27:11 Most things could be (made) generic Aug 13 18:27:24 Seems oFono has some more maintainers now as well Aug 13 18:27:26 I think the intrication with mce and sailfish-manager is now too deep Aug 13 18:27:46 Not only Marcel Holtman anymore Aug 13 18:29:03 I mean it seems actively maintained upstream now with daily merges etc Aug 13 18:37:20 Tofe: There's also https://git.merproject.org/mer-core/libgrilio Aug 13 18:37:29 Just I have no clue what it's being used for exactly and how Aug 13 18:37:41 MEr's documentation is sparse at best Aug 13 18:39:11 Which isn't the first time ;) Aug 13 18:42:09 it's a lib to handle ril I/O (with a bit of glib help) -> lib g ril io Aug 13 18:43:29 right now I'm just trying to figure out how to fetch these damn debug logs from ofono :p Aug 13 18:43:50 (that is, if they are sent...) Aug 13 18:49:00 Tofe: You seen: https://together.jolla.com/question/175286/howto-enable-debug-mode-with-ofono/ ? Aug 13 18:49:14 ooooh Aug 13 18:49:49 I was so damn sure I was starting it with -d Aug 13 18:50:01 And here also: https://wiki.kubuntu.org/ConnMan/Debugging Aug 13 18:50:06 env OFONO_AT_DEBUG=1 Aug 13 18:50:11 exec ofonod -d Aug 13 18:50:11 ok, better now Aug 13 19:08:03 Hammerhead: RIL_REQUEST_GET_SIM_STATUS: card_state=1,universal_pin_state=0,gsm_umts_index=0,cdma_index=-1,ims_index=-1,num_apps=1 Aug 13 19:08:03 versus Aug 13 19:08:03 Tissot: RIL_REQUEST_GET_SIM_STATUS: card_state=0,universal_pin_state=0,gsm_umts_index=-1,cdma_index=-1,ims_index=-1,num_apps=0 Aug 13 19:09:06 card_state isn't good news here... I'll try to see what it means in detail. But it's the very beginning of the ofono driver init in both case, nothing is sent to RIL before that Aug 13 19:10:37 RIL_CARDSTATE_ABSENT 0 and RIL_CARDSTATE_PRESENT 1 ... meh... Aug 13 19:12:02 Could the issue be on Android's container side?... Aug 13 19:16:20 Always a possibility of course Aug 13 19:17:51 https://github.com/UnholyDevs/hardware_interfaces/blob/f597430b918a183bca6d1877a062e199995ecd97/radio/1.1/IRadio.hal#L46 Aug 13 20:03:53 Not sure if this is of any use? Not sure it's in upstream oFono, don't have PC to check: https://git.merproject.org/mer-core/ofono/commit/a135d0ea52fb64ea0bce1a2887c9308e2f6b641a Aug 13 20:04:43 This is an interesting background read on RIL https://programming.vip/docs/rild-chapter-5-a-complete-request-process.html Aug 13 20:10:05 Herrie: it's not upstream :/ Aug 13 20:14:02 Well it might be a lead then Aug 13 20:26:51 The values are basically the same, just not expressed the same way Aug 13 20:27:06 I'll try again tomorrow **** ENDING LOGGING AT Wed Aug 14 02:59:56 2019