Saturday, 2020-03-14

*** jrayhawk_ is now known as jrayhawk02:06
riniguspiggz: libhybris main can be used, they merged the patches06:30
piggzrinigus: ta07:58
T42<adampigg> mal: how can i disable rpmlint errors that id rather ignore
T42<adampigg> and dont say the solution is not to install the vendor stuff :)09:20
mal@adampigg did you minimize the build or is that supposed to be the whole android side build?16:03
mal@adampigg that droid-hal rpm is huge16:05
malif you are going package the whole system/vendor stuff the go with the "official" way, not like that16:06
T42<adampigg> Its included for some patched lib...could be minimized more, but i byoassed the rpmlint error for now16:07
T42<adampigg> Its fine :)16:07
mal@adampigg which lib?16:11
malyou know you can just build that, not all16:11
T42<adampigg> Yeah, ill fix it later...i may also remove the qx1000 repo16:40
piggzmal: so, is there a way to specify additional files to be included in the install?17:29
piggzatm, I hacked the inc file to install whole of /vendor17:29
piggzbug/improvement for ..... the mer kernel check command should use habuild_device if set, not device17:36
piggzhang on, it does, im just using the wrong spec17:37
malpiggz: you need to be a bit more specific so we can fix it properly17:55
piggzmal: so, i need to patch a vendor library ... im happy to install the files into /usr/libexec/droidhybris/vendor/....... and then bind mount over the system file....17:56
piggzi just need to know how to include just that lib in the rpm and not copy the whole of vendor/17:57
malpiggz: I meant that if you have the minimized build but then just run "make foo" for the lib foo you patch, does it work?17:57
piggzill have to check ... i aready removed a lot from the build, but obviously not enough!17:59
piggzi might make clean and see if its much smaller17:59
piggzas it is bigger than the last build i made17:59
malI would do normal minimizing and then try that, I used it on some devices to build extra patched libs17:59
mallike patched hwcomposer on 2011 xperias18:00
maljust "make hybris-hal some_lib"18:00
piggzmal: ok....additionally, the dhd inc file doest copy stuff from vendor at all so i modified that18:02
malpiggz: just curious, what kind of thing do you need to patch in vendor?18:13
piggzlib msm recovery for bootctl18:14
piggzlike rinigus did18:14
piggzi think there is also another way using udev18:14
piggzbut, in the future i may also need to patch and install some camera libs18:15
piggzso, it would be nice if there was a way to install stuff from vendor18:15
T42<NotKit> piggz, also some RIL-related libs18:16
T42<NotKit> to fix data18:16
malisn't there some other way to fix those than to patch the libs18:17
T42<NotKit> maybe, but that is the easiest compared to say doing fake netd18:18
T42<NotKit> (since we can recompile them)18:18
mallike ?18:19
piggzlol maybe ... thats a new one to me18:21
malwell mentioning problems to me might help sometimes :)18:22
piggzwell, its taken me until now to get a phone that will go on the network, so i have a revived interest18:24
piggzmy build seems relatively minimal, 5500 targets instead of the 81000 it started with!18:32
piggzmy build seems relatively minimal, 5500 targets instead of the 81000 it started with!19:24
piggzmal: what is a safe version of ofono ril binder plugin?19:24
malpiggz: 3.2.1 has 1.0.1019:26
malpiggz: although I see no new dependency changes after that19:29
piggzmal: ok, ill try master .... what is the full version of 3.2.1?19:29
malpiggz: full version?19:30
piggzthe 4th number19:31
*** Simdol is now known as DeltoidDelta19:36
DevelLevelHi, I am new to working with Sailfish OS so sorry for my inexperience. I have successfully built Sailfish OS for my the device, the SONY xperia z5 compact/suzuran. But the issue I am facing is that my zip image are unable to boot. I am stuck at the SONY splash screen. I’ve tried taking a pmsg-ramoops, but the file only shows one character which is unreadable: []. I am building with Lineage os 14.1 as a base and th20:32
DevelLevele official lineage os device trees. And here is the output of a kernel check: I’ve tried to fix as many ERRORs and WARNINGS as possible but these WARNINGS are left. I don't know how to proceed hence why I am reaching out here. Is there any way for me to get another sort of logfile or solve this issue? Where should I start looking?20:32
malbased on those you have done some not so good configuration options, you should use CONFIG_SECURITY_SELINUX_BOOTPARAM and keep CONFIG_AUDIT=y, as mentioned in hadk pdf chapter 9.3.120:37
DevelLevelI will try... The only problem I had with SECURITY_SELINUX_BOOTPARAM is that i have it enabled. But it does not show as enabled in the kernel check. Here are my defconfig addons:
DevelLevelI should also add that when I am flashing the Sailfish OS zip am i getting this: mount: failed to mount /dev/block/bootdevice/by-name/userdata at /data: device or resource busy. But the instalation completes with no errors...20:46
malDevelLevel: note that those are all in capital letters21:01
malyou added those new ones not like that21:01
malshow what changes you made to fixup-mountpoints21:01
DevelLevelops I will recompile the kernel with CAPITAL letters...21:04
maljust look how the other ones in the file look like and do it like those21:04
DevelLevelHere you go:
malcan you give me a link to your device repo, of lineageos for example21:09
malhow did you determine the values for fixup-mountpoints?21:12
DevelLevelFrom chapter 5.3 in HADK21:13
malbased on this you should be using block/bootdevice/by-name/ instead of block/platform/soc.0/f9824900.sdhci/by-name/21:13
malso you need to run make hybris-hal again in the correct environment like you did before and after that " -d" and then build new image21:15
DevelLevelOk, thank you verry much for the help. I will see what I can do. BTW I added some more configs related to selinux and change them to CAPITAL letters. The BOOTPARAM warning is gone. But i were unable to activate AUDITY21:16
malyou already had audit there21:17
DevelLevelI know but i added some dependencies but it did not work...21:18
DevelLevelI got the same: mount: failed to mount /dev/block/bootdevice/by-name/userdata at /data: device or resource busy. Issue but this time i got a white light when booting. I think i took a step forward getting this to work!!! Thank you verry much mal21:55
T42<edp_17> Good evening.22:26
T42<edp_17> I know it looks annoying but I don't know what am doing wrong with the bluetooth. It still doesn't work only if I manually run the brcm_patchram_plus command.22:28
T42<edp_17> The config inf /etc/firmware/ and bluetooth-rfkill-event.service and journalctl is here:
T42<edp_17> When I click on the bluetooth, there is a bluetooth_rfkill_event but the bt is not turning on.22:29
T42<edp_17> The parameters in the /etc/firmware/bcm4330b1.conf are correct, so I am puzzled.22:30
T42<edp_17> I have tried to add the brcm_patchram_plus command into the bluetooth-rfkill-event.service as ExecStartPre, then as ExecStartPost but none of them worked.22:32
T42<elros34> Have you tried to run bluetooth-rfkill-event in debug mode (-d option)?22:33
malwhat does command rfkill show in terminal22:33
maldebug is also a good idea22:34
T42<edp_17> rfkill output:
T42<edp_17> I'll try the debug.22:34
T42<edp_17> I have changed the 'ExecStart' to '/usr/sbin/bluetooth_rfkill_event -d $DEBUG $BTMODULE $CONFIGFILE' in 'bluetooth-rfkill-event.service' but didn't give me more in the journalctl.22:40
T42<edp_17> I have tried to run the 'bluetooth_rfkill_event -d -c /etc/firmware/bcm4330b1.conf'  but that didn't show anything. Then I checked the journalctl again:
T42<edp_17> Still nothing, it only shows that I switched on/off the bluetooth in Settings.22:44
T42<edp_17> the content of '/etc/sysconfig/bluetooth-rfkill-event-hciattach' :
T42<edp_17> I went through 10.5 in hadq-faq but except the symlinks in /etc/firmware, I have everything else.22:49
malthat log had more than before22:50
malthe rfkill event: idx 0 type BLUETOOTH(2) op CHANGE(2) soft 0 hard 022:50
T42<edp_17> I don't really know what this is. I guess it shows when I switch the bt off or on in Settings.22:52
T42<elros34> for me it doesn't looks like a full log, so reboot and grab full log again22:52
malthe message looks like it doesn't find the switch because rfkill add wasn't seen22:54
malbefore it sees rfkill change22:54
T42<edp_17> Yes, it is not a full log. I'll reboot and give a full log.22:55
T42<edp_17> I have found this too: 'rfkill_switch_add(bluetooth_rfkill_event.c:1076): Skipping over unsupported rfkill switch 'bcm4330 Bluetooth'22:56
malchange that string to "bcm4330 Bluetooth\n" and build the bluetooth-rfkill-event rpm again and install it on device23:01
T42<edp_17> I cannot upload the full journalctl because it is too large. 😔23:01
malno need23:01
maljust do what I just told23:01
T42<edp_17> @mal : thanks, I try.23:02
T42<edp_17> Should I rebuild the broadcom-bluetooth too?23:04
T42<edp_17> Okay, here it is what I have done. I changed that line as advised then rebuilt and installed bluetooth-rfkill-event. (First I needed to remove bluetooth-rfkill-event-configs-mer)23:15
T42<edp_17> Then I did a systemctl daemon-reload and went to Settings and tried to turn bt on then off (or vice verse)23:16
T42<edp_17> Then I ran 'bluetooth_rfkill_event -d -c /etc/firmware/bcm4330b1.conf' and also went to Settings and tried on/off the bt (or vice verse)23:17
T42<edp_17> Now there is more in the log:
T42<edp_17> I see, there are some invalid parameters in the conf.23:19
maldid you reboot23:24
T42<edp_17> Not yet. I am doing.23:24
malyou have some mistakes in the config file23:26
T42<edp_17> What I also did was: commented out the no2bytes and patchram keywords in the .conf and recreated the symlink into /etc/firmware.23:26
malreplace word patchram dl_patch23:27
malreplace word patchram with dl_patch23:27
T42<edp_17> I replaced the word and rebooted. When I try to switch the bt on in settings:
maldid it work? based on log it did something23:37
T42<edp_17> No, it didn't. Based on the log it turned the bt on then off again.23:38
maldoes the line 16 in that log look like the command you used manually?23:39
malit shows the command it runs23:40
T42<edp_17> No, I only rebooted and tried to switch the bt on/off in Settings.23:41
malread what I wrote23:42
malyou said earlier that you can get bluetooth working manually by running some command, is the command on line 16 of that last log the same as the manual command23:43
T42<edp_17> Yes, except the '--bd_addr 6a:1a:8a:ca:2d:a6' bit because that was not in the manual command when I ran that.23:45
malso why doesn't that work then23:46
malif it's the same command23:46
T42<edp_17> I have repeated the test.23:49
T42<edp_17> Rebooted.23:49
T42<edp_17> Started journalctl.23:49
T42<edp_17> In Settings turned bt on.23:50
T42<edp_17> It has turned of for a second, then off again.23:50
T42<edp_17> Journal:
T42<edp_17> Is it possible that a pulseaudio setting does prevent the bt to stay on?23:51
malyou need to compare that log to what you see if you disable the bluetooth-rfkill-event service and instead run the manual command23:53
malbut it's getting late now, more debugging some other time23:54
T42<edp_17> Okay, thanks for your time again.23:55
T42<edp_17> When I disabled the bluetooth-rfkill-event service and ran the command manually, the bt turned on and stayed on. No I check logs.23:57

Generated by 2.17.1 by Marius Gedminas - find it at!