Sunday, 2019-10-20

maldid you add the correct things to sparse, the rc files00:06
jellyponymal: No. There was no error, so i decided not to add it. Is it bad?00:10
jellyponymal: oh,sorry. Misunderstood. I though youre talking about straggler files00:13
jellyponyI added mount modules for system and vendor to sparse00:14
jellyponyThat all00:14
maljellypony: do you have these
malyou will need those00:19
jellyponymal: D:00:22
jellyponymal: I tell you more. My sparse only contains "/var"00:23
jellyponywith lib/environment/compositor/droid-hal-device.conf00:24
jellyponynothing else00:24
jellyponymal: What am I missing?00:25
maljellypony: get the files from my link and add to your sparse to correct place as seen in the link00:50
malyou might also need the ld.config.28.txt modification as mentioned in channel logs, but more tomorrow, it's very late now00:51
jellyponygood night :300:54
jellyponymal: actually what is your timezone? :000:58
r0kk3rzgmt+3 i think02:28
r0kk3rzi should really be more active and take over the night shift02:29
xreactx[m]__this seems to be the latest error... when I'm trying to flash it to my device, the SFOS installer fails.03:55
T42<BusterBg_18> Just flash your boot.img and extract the rootfs as instructed on hadk03:59
T42<Akatsu %lastname%> @mal seems to be ok07:31
*** OhYash1 is now known as ohyash07:55
deathmistmal: could it be that doesn't work because /system isn't mounted at this early stage? I'm going to do some more testing because mount binding the ld.config.28.txt is a much cleaner solution imo09:02
maldeathmist: /system is mounted earlier than droid-hal-init service11:26
malat least it should be11:26
T42<DylanVanAssche> mal: Installed the V4L fork of Qt Multimedia, still nothing in the Camera app.13:01
T42<DylanVanAssche> I only have some weird white/blue image in the Camera app. We have /dev/video0 (video render) and /dev/video1 (back camera). I think that the app picks the wrong input?13:01
malcould be13:09
mal@DylanVanAssche it works manually using gst-launch?13:11
mal@DylanVanAssche what do you see when you use GST_DEBUG to get more logs, do those show what it tries to use?13:14
T42<Akatsu %lastname%> i have question but not regarding sailfish, how can i get device to create bootdevice/by-name/* nodes? now i have no bootdevice under /dev/block13:16
malthose are usually created automatically using udev rules13:20
malat which state of the boot is your device?13:20
malwhich telnet port13:20
vknechtMal: I hope the new video-snapshot pr is to your liking :-)14:24
T42<DylanVanAssche> mal: hmmm it worked before, now it doesn't since gst-plugins-good doesn't build on OBS. I did that manually in the past but after reflashing that part is gone :(14:53
T42<DylanVanAssche> *manually14:53
T42<DylanVanAssche> mal: Will look at the V4L stuff later. I almost have the proximity sensor working under IIO. Well it works for MCE (display double tap works only when not covered). However, Messwerk doesn't work: 0.000 Lux and always no proximity detected.15:04
T42<DylanVanAssche> spiiroin: Do you see any issues with this IIO proximity adaptor code?
T42<DylanVanAssche> - The light sensor value reports an invalid value of 4294967295 (after adding this commit).17:13
T42<DylanVanAssche> - The proximity value is calculated correctly though. However, Messwerk shows -1 for the light sensor (error code I suppose) and proximity doesn't react.17:13
T42<DylanVanAssche> - journalctl:
T42<DylanVanAssche> I don't see any difference with the other proximity sensor implementations in sensorfw...17:13
malI would assume -1 is the same as 4294967295 i.e. unsigned vs signed?17:19
malthe opposite obviously17:20
mal@DylanVanAssche ^17:20
mal@DylanVanAssche can you also print the values of variables result, iioDevice.offset, iioDevice.scale and *.value_ for both als and prox17:25
T42<DylanVanAssche> mal: I would assume the same about that number...17:27
T42<DylanVanAssche> I will rebuild sensorfw to print those values too.17:27
*** jellypony1 is now known as jellypony17:56
T42<DylanVanAssche> mal: The light sensor is fixed if a manual scale override is applied of 1.0 (the proximity scale override the light scaling -> same sensor? )18:05
T42<DylanVanAssche> mal: With scale override fix:
T42<DylanVanAssche> Without scale fix:
T42<DylanVanAssche> However, the scale fix doesn't fix the proximity detection though18:15
mal@DylanVanAssche look at the offset with scale override18:30
malthe scale looks ok in both cases, 118:31
T42<DylanVanAssche> Yeah that wasn't the case before I tried the scale override.18:31
T42<adampigg> mal: time to look at camera fixes?18:31
mal@adampigg I'll see if I check those after sauna18:32
mal@DylanVanAssche can you show the iio config file in both cases18:33
mal@DylanVanAssche in the earlier log you had "proximity= true"18:34
T42<DylanVanAssche> It prints 'true' but if you check messwerk, it's not detected though (the print is coming from assigning true/false to withinProximity)18:34
T42<DylanVanAssche> mal: config:
mal@DylanVanAssche we need to first make sure sensorfw can show correct values in log18:36
T42<DylanVanAssche> mal: true, trying to assign a default value to offset and scale when the sensor has none. Otherwise these values can be uninitialized18:37
maldoes als need some scaling?18:39
malyou just need to use 1 by default18:39
T42<DylanVanAssche> mal: it has a scale property in IIO though18:39
malwell kernel should have initialized that18:39
T42<DylanVanAssche> `[root@Sailfish ~]# cat /sys/bus/iio/devices/iio\:device1/in_illuminance_scale18:40
T42<DylanVanAssche> 6.400000`18:40
T42<DylanVanAssche> proximity has the same scale18:40
malis illuminance the correct one?18:40
T42<DylanVanAssche> for the light sensor yes, the the proximity it has the same scale but the raw data path is different (in_proximity_raw)18:42
mal@DylanVanAssche just to understand how things work, are scales wrong or something? is the 6.4 wrong?18:46
T42<DylanVanAssche> mal: The scales are wrong and the offsets are wrong too. It worked fine if the proximity sensor support is not there though. Anyway, I'm tired, will investigate later further18:47
malare those scales defined in kernel driver or obtained from firmware?18:49
T42<DylanVanAssche> They are defined through a sysfs path: /sys/bus/iio/devices/DEVICE/in_illuminance_scale for example19:02
T42<DylanVanAssche> Sensorfw should read them when creating the sensor19:03
mal@DylanVanAssche it's kernel that creates those sysfs paths so where in kernel is the value set, or is it set using some data from i2c or something19:30
T42<DylanVanAssche> mal: I'm not so sure how these values are retrieved, but I think from the sensors over I2C or other connections19:31
malkernel driver source would tell that19:34
T42<DylanVanAssche> Okay, will take a look soonTM19:35
malcan you give a link to that on github19:37
T42<DylanVanAssche> To the driver source?19:40
T42<DylanVanAssche> stk3310 IIO driver:
T42<Umeaboy> Just because Samsung hasn't released full source code to build AOSP for the S10, does that automatically mean that we can't port Sailfish to it? I mean, the HADK tells the porter to build AOSP.19:47
T42<Umeaboy> I'm considering to purchase it.19:47
jellyponymal: Good evening! I made changes to patterns you suggested yesterday. Now droid hal running but i still cant get into gui. I had a few problem that had been solved by using chapter 13 of hadk. Unfortunately lipstick still fails. journalctl:
jellyponyis it because of jolla-startupwizard-pre-user-session.service failure?19:52
mal@Umeaboy there seems to be unofficial lineage os for that but no sources for the builds, only kernel sources19:52
maljellypony: you need to show more logs, dmesg and /usr/libexec/droid-hybris/system/bin/logcat19:53
jellyponydmesg no output ._.19:57
malhow is that possible19:58
jellypony /usr/libexec/droid-hybris/system/bin/logcat is not readable :019:59
jellyponyi mean it seems like it contains nontext data20:00
malthat is a command, not a file, same for both20:00
jellyponya knew about dmesg20:01
jellyponyits broken20:01
malthat's one of most basic things, not sure how it can be broken, are you sure you didn't do something strange to kernel configs?20:03
mallike copying whole defconfig from somewhere?20:03
jellyponylogcat output:
jellyponydmesg doesnt work in all the builds with different kernel configs20:05
jellyponydunno why20:05
malwhat different configs?20:05
jellyponymanually configured20:06
jellyponyi mean20:06
jellyponyI messed up few in the beginning20:06
malyou should show all the changes you made20:08
mal@DylanVanAssche have you tried printing the raw value string you read from buffer before you get result using strtol?20:09
mal@DylanVanAssche I mean buf after this
T42<Akatsu %lastname%> @Mister_Magister have you fixed this? I do have same issue with android base20:11
T42<Akatsu %lastname%> Fwd from mal: Mister_Magister: you have so much failing like "E QMI_FW  : QMUXD: WARNING qmi_qmux_if_pwr_up_init failed!" qmi is quite important for many things20:11
Mister_Magisterneed more context @Akatsu20:12
T42<Akatsu %lastname%>
jellyponymal: my kernel configs:20:16
jellyponydefault (from device tree):
Mister_Magister@Akatsu ahhh fxtec thingy. Sorry i'm afraid someone else fixed it. piggz fixed it with "Fixed with added firmware symlink /firmware > /vendor/firmware_mnt/" so you can ask him if that suggestion won't help20:17
maland that is by adding correct things to straggler files probably20:20
T42<DylanVanAssche> mal: Not yet, will check tomorrow20:23
T42<adampigg> @DylanVanAssche what do you have in /dev/iio* ?20:30
jellyponymal: sorry, incomplete logcat output . Here is20:33
jellyponycorrect one20:33
jellyponymal: output of logcat is too big D:20:38
maljellypony: show the first part of it20:47
maljellypony: about those kernel configs, seems to are different, one from out and one from kernel repo, very difficult to compare20:49
jellyponymal: got full output (wrote output to file which was transfered to pc via recovery):
jellyponymal: here is .config if it would be made with no changes at all
jellyponynow it's easier to compare21:12
jellyponySorry for my English, I'm not practicing much.21:20
maljellypony: you seem to have some strange change, you disabled CONFIG_FW_LOADER_USER_HELPER=y21:33
maland the fallback also21:34
maljellypony: also it's easier to see the kernel confrig changes if you give the defconfig files (original and modified)21:36
malnot the .config files21:36
T42<FagotMan> @clee [has anybody started working on getting Sailfis …], I'll "up" that question. Hope some will start working on 3xl 😃21:47
T42<FagotMan> I'll keep 200$ for some1 who can make it working if any1 is interested 😉21:47
piggzmal: added a fix for the resolution icon position also21:49
malpiggz: ok, I didn't look at those PRs today, trying to check tomorrow21:52
piggznp, im away now anyay :)21:53
jellyponymal: mer_kernel_check gives me warnings for these. afair it says these settings needed for lollipop devices.21:56
jellyponyso I disabled them21:56
jellyponywhy not?21:56
malplease show both the original and modified defconfig files (not the .config)22:00
maljellypony: we use fw loader in sailfish22:01
mallet's continue debugging the issues tomorrow22:09
jellyponymodified config:
jellyponyok :022:12
jellyponymal: good night :322:14
T42<Umeaboy> @mal [@Umeaboy there seems to be unofficial lineage …], Okey. What's missing that can't be found on
T42<Umeaboy> Can't we use the device tree from omnirom?22:19
malwe can in theory, just might a bit more work, not much but some22:22
r0kk3rzdepends on how much retooling they did with the make files23:34

Generated by 2.14.0 by Marius Gedminas - find it at!