buZz | i think i'll try a reinstall from that 20220410 droid4 image again | 00:57 |
---|---|---|
buZz | and reinstall android while i'm at it :P | 00:58 |
buZz | or maybe eh, whatsitsface | 00:58 |
Wizzup | buZz: what broke? | 01:00 |
buZz | well, deb-src, and the dns issue on gprs still, i wonder if a reinstall and update fully -before- inserting sim might help | 01:00 |
buZz | the android broke because i think i reused on of the partitions as swap once :P | 01:01 |
buZz | one* | 01:01 |
buZz | gee, on android i can still get 3G? | 02:01 |
buZz | 5.1Mbps connection in my nerdcave | 02:01 |
Wizzup | that seems surprising | 02:04 |
Wizzup | the modem is really it's own os | 02:04 |
Wizzup | fwiw I also get 3g on my leste d4 np | 02:04 |
Wizzup | buZz: can you describe the dns issue and file a bug so that I can take a look? | 02:04 |
buZz | yeah , where do you want the bug? | 02:05 |
buZz | just https://github.com/maemo-leste/bugtracker/issues i guess :P | 02:06 |
buZz | Wizzup: actually, i bet its just part of https://github.com/maemo-leste/bugtracker/issues/507 | 02:08 |
Wizzup | so if a dns server is not provided by static ip - e.g. cellular, it should fall back to 9.9.9.9 | 02:11 |
buZz | not sure if thats the 'should' | 02:12 |
buZz | if there were already scripts to fix this on previous maemo, why not just run em again? | 02:12 |
buZz | well, i guess improvement is good | 02:12 |
Wizzup | if only the world was as simple as that :) | 02:12 |
buZz | ^_^ | 02:12 |
Wizzup | so for me this has been working with and without tor, with and without wireguard, with and without wifi/gprs | 02:13 |
Wizzup | so I would need to get more info from you on what is going wrong | 02:13 |
buZz | hmhm, well , resolv.conf points to 127.0.0.1 , dnsmasq -is- running | 02:14 |
buZz | just not sure why dnsmasq doesnt know where to look for cellular dhcp connections | 02:14 |
buZz | for wifi it works | 02:14 |
buZz | and if i enable tor, dns will go over tor aswell? , which seems to fix it | 02:14 |
buZz | also just overwriting resolv.conf fixes it | 02:14 |
buZz | oh! | 02:15 |
buZz | after i flashed android, booted it, powered down, booted into leste again | 02:15 |
buZz | i get 3G in leste | 02:15 |
buZz | lol | 02:15 |
buZz | i guess it flashes the baseband aswell? | 02:16 |
buZz | yeah nice, 5.8Mbps even :) | 02:18 |
Wizzup | right 127.0.0.1 is not the point | 02:18 |
Wizzup | the point is what dnsmasq is trying to read | 02:18 |
buZz | hmhm | 02:18 |
buZz | any logs i could grab? | 02:18 |
Wizzup | I don't think android flashes it, you probably forgot to issue the commands to change the modem region | 02:18 |
Wizzup | yes, but it's too late for me now | 02:19 |
Wizzup | what matters is what resolvconf contains | 02:19 |
buZz | oh quite possible :D | 02:19 |
buZz | ah, ok | 02:19 |
Wizzup | I don't remember where it is stored but can look tomorrow | 02:20 |
Wizzup | but if you know how resolvconf works.. | 02:20 |
buZz | running the scripts in /etc/gprs/ doesnt help, fyi :) | 02:29 |
* Wizzup zzz | 02:31 | |
buZz | nn ;) | 02:31 |
buZz | ah, also , if i just run dhclient -v -4 wwan3 , it does get a working dns into resolvconf/dnsmasq | 02:37 |
buZz | oh? maybe i found it, uninstalling the tor daemon :) | 02:41 |
buZz | hm nope | 03:11 |
lel | buzztiaan opened a pull request: https://github.com/maemo-leste/mce/pull/55 (some small typo) | 03:24 |
buZz | unrelated though | 03:24 |
buZz | it must something with udhcpc and how its grabbing dns data to push towards resolvconf, resolvconf only seems to be getting 0.0.0.0 | 03:45 |
lel | buzztiaan opened a pull request: https://github.com/maemo-leste/libicd-network-ipv4/pull/4 (prevent setting 0.0.0.0 as dns) | 04:17 |
buZz | \o | 04:17 |
buZz | almost 4:20 | 04:17 |
mighty17[m] | https://github.com/maemo-leste-upstream-forks/mesa/blob/maemo/beowulf-devel/debian/patches/0001-Fix-meson-build-system-and-build-pvr-driver.patch#L36 why are we removing pvr from meson? | 08:23 |
mighty17[m] | tmlind: i applied the mesa patch, the wlroots patch, but still it fails with the same error, am i missing something? | 09:23 |
mighty17[m] | ``` | 09:23 |
mighty17[m] | (phoc:2789): phoc-wlroots-CRITICAL **: 12:52:51.960: [backend/drm/renderer.c:246] Failed to pick render format for plane 32 | 09:23 |
mighty17[m] | (phoc:2789): phoc-wlroots-CRITICAL **: 12:52:51.960: [backend/drm/drm.c:725] connector LVDS-1: Failed to initialize renderer: initial page-flip failed | 09:23 |
mighty17[m] | (phoc:2789): phoc-wlroots-CRITICAL **: 12:52:51.961: [backend/drm/drm.c:813] connector LVDS-1: Failed to initialize renderer for plane | 09:23 |
buZz | does anyone here know, the deb-src stuff, did i set it up correctly? cant seem to get any sources | 13:33 |
buZz | 22:11:29 < buZz> pastebin.com/ErzdcnrF | 13:33 |
buZz | 22:11:53 < buZz> isnt that setup correctly? | 13:33 |
buZz | hmmmm, 40 -bytes- ?? https://maedevu.maemo.org/leste/dists/beowulf-devel/droid4/source/ | 13:35 |
buZz | i guess thats the source of my issues :P | 13:36 |
buZz | ~badum | 13:36 |
Wizzup | hmm | 13:36 |
buZz | *tsssh* | 13:36 |
Wizzup | maybe again disk space problems? | 13:36 |
buZz | others do seem to have more https://maedevu.maemo.org/leste/dists/beowulf-devel/pinephone/source/ | 13:36 |
buZz | Wizzup: oh, maybe it gets overwritten on every build on phoenix? | 13:36 |
buZz | failed build cause of diskspace -> no sources collected -> empty tarball pushed? | 13:37 |
Wizzup | so does apt-get source osso-xterm work for you? | 13:37 |
buZz | cant we make some space on phoenix? delete some old build perhaps | 13:37 |
buZz | or archive em | 13:37 |
buZz | osso-xterm does work! | 13:38 |
buZz | omap-linux does not , which is what i really wanted to edit :P | 13:38 |
buZz | really wanna try to see if i can get cpcap to behave slightly better | 13:39 |
Wizzup | I did do that often @ delete old stuff | 13:39 |
Wizzup | buZz: yea so really you should git clone the kernel repo but I agree we have a problem | 13:40 |
Wizzup | so for the disk space, I have another place we can run the machine | 13:40 |
Wizzup | but it will take -my- (or someone else's) time to migrate it all | 13:40 |
Wizzup | like, install nginx, rsync packages, set up irc bot logger, get hostkeys in jenkins | 13:40 |
Wizzup | and then we need the dns change | 13:40 |
buZz | hmhm, and i can just plonk the built kernel in /boot ? | 13:41 |
buZz | oh i guess i can add it to kexecboot even | 13:41 |
buZz | nice | 13:42 |
buZz | i'll go try that route | 13:42 |
buZz | Wizzup: did you see i PR'd a workaround for the DNS issues? | 13:42 |
buZz | 04:17:34 < lel> buzztiaan opened a pull request: https://github.com/maemo-leste/libicd-network-ipv4/pull/4 (prevent setting 0.0.0.0 as dns) | 13:43 |
Wizzup | let me look | 13:43 |
buZz | also one @ mce but that was just a typo that annoyed me in the logs ;) | 13:43 |
Wizzup | buZz: yes @ boot, well also you need modules | 13:43 |
buZz | right | 13:43 |
Wizzup | buZz: ah that's a good fix | 13:43 |
buZz | they dont need to be on /boot though, right? just in the rootfs? | 13:44 |
buZz | no initrd needed? | 13:44 |
Wizzup | wonder why udhcpc does that | 13:44 |
Wizzup | no initramfs no | 13:44 |
Wizzup | correct @ modules | 13:44 |
buZz | sweet | 13:44 |
Wizzup | but if the change you want to make is in dts only it's much easier | 13:44 |
buZz | hmhm, well | 13:44 |
buZz | i first wanna experiment with it for a while | 13:45 |
buZz | so, repo isnt maemo-leste/droid4-linux anymore? or just pending a rename | 13:46 |
buZz | oh man, thats a lot of branches :D haha | 13:46 |
Wizzup | yes | 13:47 |
Wizzup | do you have a cross compiler set up? | 13:47 |
Wizzup | you don't want to do the dev on the d4 | 13:47 |
buZz | why not? :D | 13:47 |
buZz | i have time | 13:48 |
bencoh | because it's slow | 13:48 |
Wizzup | it's really, really slow | 13:48 |
bencoh | and painful | 13:48 |
Wizzup | and cross compiling is painless | 13:48 |
buZz | thats ok, i once did a gentoo stage1 install on a 486sx | 13:48 |
Wizzup | like it's really easy at this point :) | 13:48 |
buZz | hehe well, i might have it setup :P | 13:48 |
buZz | but, which branch is the current one? | 13:49 |
buZz | i guess the maemo-5.15 one? | 13:49 |
Wizzup | no | 13:50 |
Wizzup | sorry, just let me make my coffee | 13:50 |
buZz | np, take yr time ^_^ | 13:51 |
Wizzup | and yes we need to rename droid4-linux | 13:52 |
buZz | ah yeah, ok | 13:52 |
Wizzup | I think this one: https://github.com/maemo-leste/droid4-linux/tree/wip/n900/maemo-5.15-cleaned-up | 13:52 |
Wizzup | and yes we need to clean that up :) | 13:53 |
lel | MerlijnWajer closed a pull request: https://github.com/maemo-leste/libicd-network-ipv4/pull/4 (prevent setting 0.0.0.0 as dns) | 13:54 |
buZz | w000t | 13:55 |
Wizzup | are we sure this is coming from udhcpc though? | 13:57 |
buZz | ah gee 'wip/n900/maemo-5.15-cleaned-up' is the whole branchname :D | 13:57 |
Wizzup | this happened with gprs for you, right? | 13:57 |
Wizzup | buZz: yes | 13:57 |
buZz | Wizzup: i added syslog output to the udhcpd script there, and its showing '0.0.0.0' as input | 13:58 |
buZz | only on gprs&umts indeed | 13:58 |
buZz | seemingly udhdpc fails to interpret the dns servers from the telco (youfone/kpn) and pushes 0.0.0.0 forward | 13:58 |
Wizzup | I am surprised udhcpc would run at all on gprs/umts, I think we just get the ip from ofono and pass it as static gconf to ipv4 module | 13:58 |
buZz | i dont see any way to affect or debug the parsing further | 13:59 |
Wizzup | did syslog show udhcpc running? | 13:59 |
buZz | well, -that script- that i did a PR on runs on connecting to 2/3G | 13:59 |
Wizzup | right | 13:59 |
buZz | which is a udhcpc script | 13:59 |
Wizzup | I think it's our script | 13:59 |
Wizzup | that runs in various parts | 13:59 |
buZz | if its used outside of udhcpc , that might be a problem | 13:59 |
buZz | s/a/the/ | 13:59 |
Wizzup | not just on udhcpc but also for static dhcp conf | 13:59 |
Wizzup | https://github.com/maemo-leste/libicd-network-ofono/commit/d462a8a5e121fc36c8e7b818e254d88ad4306345 | 14:00 |
Wizzup | (yes that contains the word hack often) | 14:00 |
buZz | yeah, but its not calling 'that script' ? | 14:01 |
Wizzup | yes it is, per icd module order | 14:01 |
Wizzup | here: https://github.com/maemo-leste/libicd-network-ofono/commit/d462a8a5e121fc36c8e7b818e254d88ad4306345#diff-00691051bb184e3f26ba941c1e259dba8911b3d1b89dad9ff013fe530f5976d3R146 | 14:01 |
buZz | so link_up_cb() calls 50_ipv4_network_setup ? | 14:02 |
Wizzup | indirectly, yes | 14:02 |
buZz | without being udhcpc that might cause issues yeah | 14:02 |
Wizzup | it might be a provider thing | 14:02 |
Wizzup | since I don't get it in .nl but you do | 14:02 |
buZz | hmhm | 14:02 |
Wizzup | not kpn, right? | 14:02 |
buZz | youfone is a MVNO on kpn | 14:03 |
Wizzup | hm | 14:03 |
Wizzup | maybe we should have an issu to deal with this in a better way | 14:04 |
lel | MerlijnWajer closed an issue: https://github.com/maemo-leste/bugtracker/issues/507 (icd2/busybox-udhcpc: Respect DNS (resolv.conf) from DHCP upon connection) | 14:04 |
buZz | hmhm | 14:05 |
buZz | well, i'm sure eventually we'll get more ppl running into this :) | 14:05 |
Wizzup | yes | 14:06 |
buZz | either way, this fix allows the 9.9.9.9 to get used by me | 14:07 |
Wizzup | right, I merged it, just now wondering if it's the fix in the right place :) | 14:12 |
buZz | its a workaround, not a fix :) | 14:12 |
buZz | the issue caused another issue, it work arounds that | 14:12 |
buZz | we still need/should find where dns -should- be coming from | 14:13 |
Wizzup | well in all non-gprs cases I am sure it work | 14:13 |
Wizzup | s | 14:13 |
Wizzup | unless gconf contains bad ips for dns | 14:13 |
Wizzup | (which 0.0.0.0 could be) | 14:13 |
buZz | this workaround will work in all cases, 0.0.0.0 should never be set as dns | 14:13 |
Wizzup | it's building now | 14:15 |
Wizzup | (see ##leste-ci) | 14:15 |
Wizzup | uvos: do we have an issue for the call problems? I am a bit confused because I think tmlind still uses it daily, so maybe it's also related to how we do UCM or something | 14:16 |
Wizzup | uvos: in any case if we have an issue maybe it'll be easier to tackle it soon | 14:16 |
buZz | dev_info_once(ddata->dev, "Can not find battery nvmem device. Assuming generic lipo battery\n"); | 14:23 |
buZz | hmhm, this is what i'm getting | 14:23 |
buZz | (on a droid4 with replaced battery) | 14:23 |
buZz | Wizzup: grabs the update \p and works | 14:25 |
buZz | i wonder what this is ; if (strcmp(dev_name(dev), "89-500029ba0f73") == 0) | 14:26 |
buZz | https://paste.debian.net/1239678/ <- seems weird to me | 14:27 |
buZz | its strcmp()'ing on a NULL thats sent to it? | 14:27 |
Wizzup | where is this from | 14:27 |
buZz | oh eh, wait, what | 14:27 |
Wizzup | ah | 14:28 |
buZz | drivers/power/supply/cpcap-battery.c | 14:28 |
Wizzup | I think that just checks for a genuine battery maybe | 14:28 |
buZz | i wonder if it would ever finish this without 'Can not xxxx battery nvmem' for anyone? | 14:28 |
Wizzup | yes | 14:29 |
buZz | i dont understand it perhaps, but battery_id seems to not get set to anything? | 14:29 |
Wizzup | } else if (nvmem_device_read(nvmem, 2, 1, &battery_id) < 0) { | 14:31 |
Wizzup | that sets the battery id | 14:31 |
buZz | ah | 14:31 |
buZz | but, thats not even hitting anything on this droid4 | 14:31 |
Wizzup | you might have a special battery from dreamer | 14:31 |
buZz | maybe i can try putting back the old battery, see if that triggers something else | 14:31 |
buZz | i do, but its unused still | 14:31 |
buZz | this is from that german guy with newoldstock? | 14:32 |
Wizzup | in any case if you want to tweak the max voltage I am sure there's a better way to do it | 14:32 |
buZz | yeah , uvos iirc said there should be a way to overwrite it just through sysfs | 14:33 |
buZz | but that doesnt seem to work or exist | 14:33 |
buZz | 20:29:53 < uvos> the detrection code is not in leste anymore | 14:33 |
buZz | but , we might have just been looking at the wrong branch ;) | 14:34 |
Wizzup | he probably has a patch ready | 14:34 |
buZz | wonder if i can see the nvmem stuff easily myself | 14:34 |
buZz | stuff like cpcap_battery_unkown_data annoys me so :D | 14:37 |
buZz | lol | 14:37 |
Wizzup | fix it up :) | 14:37 |
buZz | hehe , i might ;) | 14:37 |
buZz | hmm, there is some bus/nvmem in sysfs but doesnt seem to hold any devices :( | 14:43 |
Wizzup | I don't know how this works, sry | 14:45 |
buZz | np | 14:46 |
buZz | rubber duckies dont need to apologize ;) | 14:46 |
buZz | hehe | 14:46 |
bencoh | :] | 14:47 |
buZz | i -think- i could get cpcap-battery driver to export the nvmem as a userspace device in sysfs | 14:50 |
buZz | similar to https://www.mail-archive.com/linux-kernel@vger.kernel.org/msg1438721.html | 14:50 |
sicelo | I doubt they'll take it in power/supply subsystem. What would it do there (in sysfs)? | 14:56 |
buZz | nothing, just be readable by users/programs | 14:56 |
buZz | like a batteryinfo panel could then tell you if you have a EB41, HW4X, BW8X, etc | 14:57 |
buZz | hmmm cpcap didnt have depends on NVMEM in Kconfig | 15:38 |
buZz | this might matter, not sure | 15:38 |
buZz | i guess just some fields to export the data might be better than the whole nvmem | 15:43 |
buZz | but i do see many drivers that use nvmem that just expose the whole thing | 15:43 |
buZz | ¯\_(ツ)_/¯ | 15:43 |
buZz | Wizzup: what about this? https://github.com/maemo-leste/libicd-network-ofono/blob/4643b264e8b12fce865ecbbc09296cea49e28ad2/src/iap.c#L162 | 15:49 |
buZz | doesnt that mean the DNS data only gets copied on first provisioning? | 15:49 |
buZz | hmm, i guess thats not what this means | 15:49 |
buZz | i cant imagine having a static ip on 2/3G :P | 15:49 |
Wizzup | static as opposed to dhcp | 15:53 |
Wizzup | not as opposed to dynamic | 15:53 |
Wizzup | and yes this should run on every connect | 15:53 |
Wizzup | maybe you can check your gconf for the dns values | 15:53 |
Wizzup | ipv4_dns1 and so on | 15:54 |
buZz | rofl , noticed it was ~25% battery, connected my charger | 15:55 |
buZz | doing so triggered it to feel its lowpower, and booooom shutdown | 15:55 |
buZz | -_- | 15:56 |
* buZz getting kinda annoyed by this feature | 15:56 | |
Wizzup | is your battery calibrated? | 15:57 |
Wizzup | it's possible that it quickly drops in voltage a lot, and I think there's a threshold for how long it needs to be below a certain voltage | 15:57 |
buZz | where is the monitoring happening? from the kernel driver? | 16:00 |
Wizzup | mce I think | 16:01 |
freemangordon | buZz: did you manage to connect through gprs on d4? | 16:12 |
freemangordon | as I have a feeling something changed in ofono and it does not report any internet contexts | 16:12 |
freemangordon | it used to, based on mobile-broadband-provider-info data | 16:12 |
freemangordon | or was it sim-pin-query? | 16:13 |
* freemangordon checks | 16:13 | |
freemangordon | hmm, what package was that? | 16:14 |
Wizzup | hm | 16:16 |
Wizzup | what are you looking for | 16:16 |
buZz | freemangordon: yes, 2G and 3G both work | 16:20 |
buZz | i cant select between the two i think, but it was 2G all the time until i booted android with the sim | 16:21 |
Wizzup | freemangordon: it also works for me | 16:21 |
buZz | oh | 16:21 |
buZz | i disabled pin entry requirement | 16:21 |
freemangordon | ok, seems I am missing something in the VM | 16:27 |
freemangordon | will upgrade VM and will try again later on | 16:30 |
buZz | you have a VM with modem? :D | 16:31 |
buZz | oh, ofono has a phonesim i guess | 16:31 |
sicelo | you can also passthrough your modem (to the vm) ;-) | 16:34 |
buZz | run qemu on the droid4 and passthrough original modem? :P | 16:35 |
Wizzup | I don't think it even supports kvm | 16:35 |
Wizzup | there are plenty of usb modems that can work | 16:35 |
buZz | :) maybe something fun to try once, plonk a 5g usb modem on | 16:41 |
buZz | need to visit DHL, brb | 16:41 |
Wizzup | I have done it in the past and documented how to do it, for development purposes | 16:41 |
Wizzup | most of what we need to fix now is in ofono on the droid4 itself | 16:42 |
Wizzup | not so much in our code | 16:42 |
freemangordon | buZz: no, USB passthrough | 16:53 |
freemangordon | May 2 18:02:03 localhost icd2 0.98[24329]: [OFONO] Got modem data | 17:02 |
freemangordon | May 2 18:02:03 localhost icd2 0.98[24329]: [OFONO] Got APN: | 17:02 |
freemangordon | May 2 18:02:03 localhost icd2 0.98[24329]: [OFONO] Got Ctx: (nil) | 17:02 |
freemangordon | Wizzup: ^^^ | 17:02 |
Wizzup | does it see a sim? | 17:03 |
Wizzup | what does ofono report over dbus? | 17:03 |
Wizzup | is this when connecting to an apn, or just icd2 startup? | 17:03 |
Wizzup | what action did you do/trigger? | 17:03 |
freemangordon | https://pastebin.com/bWFq7QsR | 17:04 |
freemangordon | I tried to connect via connui dialog | 17:04 |
freemangordon | I see "Telenor BG" there | 17:04 |
freemangordon | clicking it results in "internet connection failed. try again?" error | 17:05 |
freemangordon | which bus to monitor? seesion or system? | 17:05 |
freemangordon | *session | 17:05 |
Wizzup | that unfortunately is expected | 17:05 |
Wizzup | it doesn't wait for ofono to activate | 17:06 |
Wizzup | if you try again after the error usually it works since it's active | 17:06 |
Wizzup | this needs to be fixed in the icd2 plugin, with a timeout when it doesn't get active | 17:06 |
Wizzup | ofono just returns right away on make active call, and caller has to wait for status changes on dbus | 17:06 |
freemangordon | hmm it seems my gcon db is broken | 17:06 |
freemangordon | *gconf db | 17:07 |
freemangordon | Wizzup: I don;t have gprs icon name set up, is that normal? | 17:08 |
freemangordon | no icon here either https://github.com/maemo-leste/libicd-network-ofono/blob/master/etc/libicd-network-ofono.schemas | 17:08 |
Wizzup | uh, I need to remember... | 17:08 |
* freemangordon checks on the device | 17:09 | |
freemangordon | on fremantle that is | 17:09 |
Wizzup | user@maindroid:~$ gconftool -R /system/osso/connectivity/network_type/GPRS | grep icon statusbar_icon_name = general_packetdata icon_name = general_packetdata statusbar_dimmed_icon_name = statusarea_internetconn_packetdata_dimmed | 17:10 |
sicelo | buZz: yes, qemu is able to pass it through as well. | 17:10 |
freemangordon | ugh | 17:10 |
freemangordon | Wizzup: connui-iapsettings-gprs is needed as well | 17:11 |
freemangordon | do we have a metapackage for all this? | 17:11 |
Wizzup | it is in hildon-connectivity-mobile | 17:11 |
freemangordon | hmm, it is not installed in the VM it seems | 17:12 |
Wizzup | probably a good idea to install it | 17:16 |
freemangordon | Wizzup: hmm, I think it is another isse, not that context is not activated | 17:16 |
freemangordon | see pastebin | 17:16 |
freemangordon | it somehow gets NULL ctx | 17:16 |
Wizzup | I think this is what I said before | 17:17 |
Wizzup | where it does not wait | 17:17 |
freemangordon | no, APN is empty | 17:17 |
Wizzup | then it is not provisioned | 17:17 |
freemangordon | so ofono_modem_get_context_by_apn returns NULL | 17:17 |
Wizzup | I think you wrote that code at some point | 17:17 |
freemangordon | it is not | 17:17 |
Wizzup | I don't remember where it is | 17:17 |
freemangordon | I think it is in pin query | 17:17 |
freemangordon | but I don;t remember the package :) | 17:17 |
Wizzup | that is connui-cellular but I don't think it is in there | 17:18 |
freemangordon | hmm, where it is then? | 17:18 |
Wizzup | libicd-network-ofono probably | 17:18 |
freemangordon | dammit | 17:18 |
freemangordon | ok, lemme check my commits | 17:18 |
Wizzup | see a76b7a7c68fdf63565189dd9e87265b92af65e34 | 17:18 |
freemangordon | https://github.com/maemo-leste/libicd-network-ofono/commit/a76b7a7c68fdf63565189dd9e87265b92af65e34 | 17:19 |
freemangordon | Wizzup: ok, I think the issue is that this sim has been provisioned some time ago, but APN was somehow left empty | 17:23 |
bencoh | ofono retrives the APN from the SIM? | 17:23 |
Wizzup | freemangordon: right, that is possible, also we need to reprovision in different countries sometimes I think | 17:23 |
Wizzup | not sure | 17:23 |
Wizzup | bencoh: I don't think the apn, but operators probably | 17:24 |
freemangordon | apns come from ofono, I think it gets them from mobile-broadband-provider-info | 17:24 |
Wizzup | right, using operator codes | 17:24 |
freemangordon | the operator name comes from SIM or from network, if we are registered | 17:25 |
Wizzup | apn for sure does not though | 17:26 |
Wizzup | that's why we need the provider info | 17:26 |
freemangordon | Wizzup: ok, what about if APN is empty, shall I treat that IAP/SIM as not provisioned? | 17:26 |
Wizzup | I think so | 17:28 |
freemangordon | ok | 17:28 |
bencoh | hmm, so networks are supposed to register to, say, android or iOS, and provide hardcoded network configuration? | 17:28 |
Wizzup | I think apn is always required | 17:28 |
Wizzup | I don't think networks register to the OS | 17:28 |
Wizzup | I think the OS gets some info from sim, and then reads known apn values and prefills them for the user | 17:28 |
bencoh | is the sim supposed to update itself then? | 17:29 |
bencoh | which reminds me, some providers are moving to VoLTE only | 17:29 |
freemangordon | bencoh: SIM is under operators control, I bet they can do OTA updates :) | 17:29 |
sicelo | bencoh: there is OTA provisioning | 17:29 |
bencoh | sounds about right, okay :) | 17:30 |
bencoh | so basically, ofono/whatever should be able to read it from sim as well | 17:30 |
Wizzup | bencoh: sad day @ volte | 17:30 |
sicelo | bencoh: not afaik. the sim doesn't keep that kind of stuff. by OTA provisioning, i meant - operator is able to send your device a provisioning message. | 17:32 |
bencoh | ah | 17:32 |
sicelo | for linux users, i don't think we have anything for those kinds of messages, so we use mobile-provider... | 17:32 |
bencoh | I see | 17:32 |
freemangordon | sicelo: well, n900 can handle those | 17:35 |
freemangordon | I will look at how it does it, when it comes to it, someday :) | 17:35 |
bencoh | :) | 17:40 |
sicelo | yeah, true | 17:43 |
freemangordon | Wizzup: yeah, this time I hit the bug you mentioned | 17:51 |
freemangordon | Wizzup: do you want me to do a PR or shall I push directly? | 17:52 |
Wizzup | freemangordon: feel free to push | 17:55 |
freemangordon | ok | 17:55 |
Wizzup | I will be afk for ~15 mins | 17:55 |
freemangordon | I will push/build in the meanwhile :) | 17:55 |
freemangordon | omg, exactly 3 years later :) | 17:57 |
freemangordon | hmm: | 18:05 |
freemangordon | /etc/maemo-dhcp.d/50_ipv4_network_setup: 74: [: missing ] | 18:05 |
freemangordon | /etc/maemo-dhcp.d/50_ipv4_network_setup: 74: [: missing ] | 18:05 |
freemangordon | /etc/maemo-dhcp.d/50_ipv4_network_setup: 93: kill: Illegal option -S | 18:05 |
freemangordon | Wizzup: so, dos conversations registers itself with telepathy? | 18:12 |
freemangordon | *does | 18:12 |
Wizzup | freemangordon: the '[' error looks weird | 18:19 |
Wizzup | freemangordon: is that from buzz' fix that I merged today to -devel? | 18:19 |
Wizzup | freemangordon: @ telepathy, yes and no | 18:19 |
freemangordon | I doubt | 18:20 |
freemangordon | unless it is in -devel already | 18:20 |
Wizzup | you doubt what? | 18:20 |
freemangordon | sec | 18:20 |
freemangordon | yeah, looks like it | 18:21 |
Wizzup | ok | 18:22 |
freemangordon | it is exactly on line 74 which is part of that PR | 18:22 |
Wizzup | he probably assumes bash? | 18:22 |
freemangordon | mhm | 18:22 |
Wizzup | buZz: can you fix? | 18:22 |
freemangordon | could be something else as well | 18:23 |
freemangordon | like empty variables | 18:23 |
freemangordon | have to go afk for a while, will try to repro later on | 18:23 |
Wizzup | you didn't merge to master btw | 18:25 |
freemangordon | neither did you :p | 18:33 |
freemangordon | This branch is 11 commits ahead of master. | 18:33 |
freemangordon | will do though | 18:33 |
freemangordon | done | 18:34 |
Wizzup | oh, maybe I didn't because of my hacky commits | 18:37 |
Wizzup | :) | 18:37 |
buZz | oh, i did assume bash, i'll fix it | 18:40 |
Wizzup | ty | 18:43 |
buZz | grabbed my second droid4 while on the move, still with original battery | 19:02 |
buZz | swapped sim and sdcard, everything works the same , -including- unable to read nvmem in battery and thus failing to set the right voltages | 19:02 |
buZz | so eh, what -is- the sh we have? dash? busybox? zsh? | 19:04 |
buZz | Wizzup: funny enough, i cannot configure APN, setting it in the internetconnections will get me a nonworking connection, only emptystring is valid | 19:05 |
buZz | so i feel ~something is overriding it , or changing the string anyway | 19:05 |
buZz | ok, its dash | 19:08 |
buZz | freemangordon: is your /bin/sh also dash? | 19:08 |
Wizzup | buZz: that is normal on debian so likely | 19:24 |
Wizzup | but let's be compatible with busybox sh too | 19:24 |
buZz | i just notice i also get the failsafe on my home wifi now :P heh, lets see if that change fixes it | 19:37 |
buZz | while resolvconf/dnsmasq get 9.9.9.9 , avahi-daemon does spot the correct dns :P | 19:40 |
freemangordon | buZz: whatever comes with leste :) | 19:49 |
buZz | freemangordon: does http://space.nurdspace.nl/~buzz/50_ipv4_network_setup fix it for you? | 19:54 |
buZz | i'm not getting a error on either version :P | 19:54 |
buZz | and cant remember having done anything special to sh/dash/bash , but i dont know | 19:54 |
buZz | at least with this version i'm getting the right dns on wif | 19:56 |
buZz | i | 19:56 |
buZz | and 9.9.9.9 on 3G | 19:57 |
freemangordon | hmm, why do we set 9.9.9.9 for gprs? DNS is provided by MNO | 20:11 |
freemangordon | for example: | 20:11 |
freemangordon | Settings = { Interface=wwan0 Method=static Address=10.215.142.176 Netmask=255.255.255.224 Gateway=10.215.142.177 DomainNameServers=85.118.197.5,85.118.197.69, } | 20:11 |
buZz | where are you getting that string? | 20:11 |
freemangordon | /usr/share/ofono/scripts/list-contexts | 20:12 |
buZz | the 9.9.9.9 is a failsafe when no dns was sent or setup | 20:12 |
freemangordon | install ofono-scripts | 20:12 |
freemangordon | a stupid question - how to check my current DNS setting? | 20:12 |
buZz | cat /run/dnsmasq/resolv.conf | 20:12 |
freemangordon | nameserver 9.9.9.9 | 20:13 |
freemangordon | not good IMO | 20:13 |
buZz | and /run/resolvconf/interface/wwan3.udhcpc etc | 20:13 |
buZz | indeed | 20:13 |
buZz | something is passing 0.0.0.0 through and i have no clue what | 20:13 |
buZz | list-contexts also gives me a emptystring APN :) | 20:14 |
freemangordon | it is LTE here | 20:14 |
buZz | nice :) what speeds do you get? | 20:15 |
freemangordon | ~10Mbits | 20:15 |
buZz | sweet | 20:15 |
freemangordon | but this is USB passthrough in VM | 20:15 |
freemangordon | buZz: anyway, the script you have provided no longer errors | 20:16 |
freemangordon | please do a PR | 20:16 |
buZz | VMeamo :) | 20:16 |
buZz | alrighty! | 20:16 |
freemangordon | also, could you fix "/etc/maemo-dhcp.d/50_ipv4_network_setup: 93: kill: Illegal option -S" while at it? | 20:16 |
freemangordon | sh kill wants -s $signal | 20:17 |
buZz | oh, surething | 20:17 |
freemangordon | not -$signal | 20:17 |
freemangordon | thanks | 20:17 |
buZz | -s SIGHUP or -s HUP? | 20:17 |
freemangordon | no idea | 20:17 |
buZz | ^_^ | 20:17 |
freemangordon | just try it | 20:17 |
freemangordon | I guess it is -s SIGHUP | 20:17 |
freemangordon | but... a guess ;) | 20:17 |
buZz | SIGHUP | 20:18 |
buZz | :) kill -l lists them | 20:18 |
lel | buzztiaan opened a pull request: https://github.com/maemo-leste/libicd-network-ipv4/pull/5 (small mistake!) | 20:20 |
Wizzup | buZz: can you 'git commit --amend' and make the git commit desc more clear | 20:30 |
Wizzup | something like ' | 20:30 |
Wizzup | 'etc/50_ipv4_network_setup: make sh compliant' | 20:30 |
lel | buzztiaan edited a pull request: https://github.com/maemo-leste/libicd-network-ipv4/pull/5 (made etc/50_ipv4_network_setup more sh compliant) | 20:49 |
Wizzup | buZz: can you also git push -f | 21:43 |
Wizzup | the commit still says 'small mistake!' | 21:43 |
Wizzup | https://github.com/maemo-leste/libicd-network-ipv4/pull/5/commits | 21:43 |
buZz | eh well, i'll just nuke it and resubmit, 1moment | 21:55 |
buZz | github isnt taking it :P | 21:56 |
Wizzup | buZz: hmmm | 22:19 |
Wizzup | buZz: what are you trying? | 22:20 |
lel | buzztiaan opened a pull request: https://github.com/maemo-leste/libicd-network-ipv4/pull/6 (made etc/50_ipv4_network_setup more sh compliant) | 22:25 |
buZz | was eating ;) | 22:25 |
Wizzup | buZz: I mean I could help with making the old one work | 22:26 |
Wizzup | but ok | 22:26 |
Wizzup | thanks | 22:26 |
lel | MerlijnWajer closed a pull request: https://github.com/maemo-leste/libicd-network-ipv4/pull/6 (made etc/50_ipv4_network_setup more sh compliant) | 22:29 |
buZz | Wizzup: i made the pr through github, there was no local git involved ;) | 22:41 |
Wizzup | buZz: you didn't 'git push' at any time? :D | 22:47 |
buZz | indeed ;) | 22:48 |
Wizzup | building now | 22:49 |
sicelo | i wonder if n900 doesn't mind being hilighted so much :-) | 22:56 |
buZz | :) | 23:13 |
buZz | could always /nick macrocrab then ;) | 23:13 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!