Monday, 2019-10-21

*** sydney_u1tangle is now known as sydney_untangle02:13
T42<Umeaboy> Right.02:36
xreactx[m]__I tried flashing my hybris-boot and rootfs to my OnePlus 7 Pro... bricked device (now unbricked using qualcomm MSM Tool)03:39
xreactx[m]__mal: are there any other hybris-16.0 guides I should be aware of?03:40
xreactx[m]__mal: is hybris-16.0 even good enough to attempt a port? I'm still getting error 6 and 7 (not at the same time) when trying to install the zip in TWRP03:42
T42<BusterBg_18> How could you even brick your device? You are just flashing a boot.img and extracting some files to /data04:10
T42<birdzhang> xreactx[m]__: seems you break aboot04:39
*** Gizmokid2005 is now known as Gizmummy200505:18
r0kk3rzif you flash the wrong partition, sure you can brick things05:31
T42<adampigg> xreactx arnt those twrp errors related to recent twrp being unable to flash sailfish? ability to flash should have nothing to do with the stability of a hybris branch05:52
rinigusmorning! I managed to bypass boot issue that I had with xperia xz2 by changing kernel config. looks like its bootloader bug (details at
rinigusboot with sfos images proceeds now to sony logo and gets stuck at it for few minutes. after that device reboots and the same goes on. (similar to poco's port?)07:01
rinigusas I have twrp installed, I can reboot to that. question: where are logs stored in aosp9-based ports?07:01
r0kk3rzrinigus: check for telnet device07:51
rinigusr0kk3rz: don't have any records in dmesg not lsusb. but I can check out that again tonight07:52
r0kk3rzlame, that makes things more difficult07:55
r0kk3rzif you've got pstore logs enabled you might be able to get something07:56
rinigusr0kk3rz: on AOSP, pstore works. haven't seen it on SFOS image, but it could be I miss somewhere enabling it. but don't we have some logs in /data/.stowaways/sailfishos/init.log ? I wonder if it is expected to be on the same location in AOSP9 based port08:02
KALUBEHi all, I heard someone was interested in porting to the OnePlus 609:08
deathmistpketo: could you bring Jolla Store to OnePlus 5T? "Device model: OnePlus 5T (dumpling / dumpling)", the phone uniquely idetifies itself via IMEI1/2 numbers from packaging box and "About device" screen in Setting, for OBS builds I'm planning to merge them with existing cheeseburger ones somehow as they are practically identical in sources09:25
deathmistaand they stay constant across reboots as do MAC addresses :)09:25
r0kk3rzrinigus: not if it doesnt get that far09:50
pketodeathmist: added10:07
rinigusr0kk3rz: true, but there is hope. now do you know whether in xperia aosp9 ports we have the same /data/.stowaways as on "normal" ports? as far as I understand, its with lvm10:28
xreactx[m]__adampigg: I tried using several version of TWRP...11:54
xreactx[m]__BusterBG_18: I accidentally flashed the hybris boot and tried extracting the rootfs to see if it would work. I used a flash tool to get it to work.11:56
xreactx[m]__r0kk3rz: I immediately regretted what I did... But all is good now, minus the TWRP errors11:57
T42<adampigg> xreactx try 2.8.612:12
xreactx[m]__adampigg: the OnePlus 7 pro only has TWRP 3.3.x and above12:23
malcan't you manually extract the rootfs in telnet 23, assuming you get to that12:29
xreactx[m]__mal: I get "invalid tar format" error when I try to extract the rootfs13:54
xreactx[m]__Also when I flash lineage 16 the initial boot screen (OnePlus logo) changes with the new "powered by Android with Android 10 robot logo" at the bottom... Actually, as a side note and not done with SFOS, I also get errors when trying to install GAPPS on lineage 16 or other Android 9 roms13:59
jellyponymal: Good evening! I enabled FW_LOADER_USER_HELPER as you suggested yesterday and rebuilt hybris-boot. Nothing changed D:14:57
jellyponydmesg: still no output14:57
malyour defconfig looked very odd, I need to check it again14:58
jellypony - present15:01
jellypony - default15:01
xreactx[m]__Is there a new tar command for Android 9 and 10?15:23
T42<adampigg> mal, can i steal your time later?15:53
mal@adampigg it depends :)15:54
malfor reviews or something else?15:55
jellyponymal: Maybe it would be helpful. Nitrogen is nearly the same as whyred (except camera module afair). Even my device tree have been made out of whyred's. I seen messages here about porting sfos to this device. May be this guys had the similar issues as me and you could remember something.16:11
maljellypony: you are using new android base which is a bit different16:13
T42<adampigg> @mal [for reviews or something else?], Just reviews and testing :)16:14
T42<BusterBg_18> Nitrogen has a older base why not use that instead of 16?16:31
*** electro is now known as Guest6120216:47
Guest61202hi all16:50
jellyponyBusterBg_18: No it hasn't, I would need to build my own :O16:52
electroohi all16:53
electroomal, i have this spam, an idea to stop this ?17:06
electroodroid-hal-init: property_set("ro.boottime.time_daemon", "96244472673") failed: property already set17:06
electroodroid-hal-init: starting service 'time_daemon'...17:06
maldisable time_daemon17:09
malwe don't need it17:09
piggzyep, everyone disbales that17:09
piggzalso, fix your hci_up script17:10
T42<BusterBg_18> jellypony: I mean it does have lineage sources for oreo doesn't it?17:21
electroopiggz, it is this file -> droid-hcismd-up.sh17:22
electrooHow can i search text in files ?17:23
*** OhYash1 is now known as ohyash17:28
malelectroo: do you have the syntax issue in that?17:34
malmeaning does the script fail17:34
electroo/bin/sh /usr/bin/droid/droid-hcismd-up.sh17:34
electrooi share it to you17:35
malbut does it fail?17:38
electrooyes, it fail17:38
malelectroo: which android base? do you have smd bt driver in kernel?17:39
electrooi check17:39
electroohow can i check please ?17:39
electrooandroid base17:39
malwell that hybris-* branch did you use17:40
electrooi check, i come back so ...17:40
electroofor htc one m817:42
electroomal, CONFIG_BT_HCISMD=n17:44
malelectroo: hmm, in hybris-15.1 you can probably use bluebinder instead which means no need for that script at all, add this to kernel defconfig CONFIG_BT_HCIVHCI=y and add bluebinder to patterns17:45
electroookey thanks17:45
nitin03device poco f1...base lineage 16.0..alreay build sfos package flashed it...devices boots...gets stuck on device logo...17:46
malnitin03: quite expected for first boot, hybris-16.0 needs some additional changes17:47
nitin03my problem is i can't telnet into the device17:47
maldo you see anything in host dmesg about usb, anything about Mer there17:48
electrooi forgot the pattern path17:48
nitin03yes it shows for 2 or 3 seconds then it goes to android17:49
nitin03mer bootlader shows up...but after 2 or 3 seconds android device id shows up17:50
nitin03i also checked .config file in my out dir while building hybris boot and recovery image...everything is there except for audit....bcoz i set it to no17:53
nitin03and in kernel cmdline i set it and selinux to zero17:54
nitin03here is dmesg...
nitin03init log...
electroomal, my file ->
electrooi have bluez417:59
electroonot bluebinder17:59
electroowhat do you suggest ?17:59
nitin03mal can u look into this please18:06
malelectroo: bluez5 and bluebinder18:19
electroohumm, i don't can use bluez518:19
electrooi don't remember why but, i have a 32 bit cpu18:19
malelectroo: with bluebinder you can18:19
electrooo okey18:19
malor at least should be able to18:20
maljust try it and see how it works18:20
electrooi try this and compile so18:20
electrooi remove this so ! droid-config-m8-bluez418:21
electroofor bluez518:21
malyou change the last number to 518:21
maland then add a new line with bluebinder18:21
electrooi compile all18:22
electroomal, i have an error during compile hybris ->
electrooshould i validate this kernel config ? CONFIG_BT_HCIUART18:29
malhmm, check kernel/htc/msm8974/drivers/bluetooth/hci_vhci.c and see if it has #include <interrupt.h> or something at the top where includes are18:30
mallooks like it doesn't have it, so add #include <linux/interrupt.h> to a new line after for example poll.h18:32
malso do what I just said18:33
electrooit's a success18:34
malit's not unusual that some includes are missing from drivers which are rarely or never used in android18:35
electrooi should just compile the bluebinder and bluez5 package so18:35
electrooyes ^^18:35
electroowhat is the correct command to just compile two package18:36
electroorpm/dhd/helpers/ -c18:37
malbluebinder should be build automatically, just use rpm/dhd/helpers/ -b to repackage droid-hal and kernel, then rpm/dhd/helpers/ -c to build config packages because of the pattern changes18:41
malelectroo: I mean rpm/dhd/helpers/ -d for droid-hal18:41
malnot -b18:41
piggzmal: also requesting dilica-config.txt ;)18:43
malpiggz: oops, sorry18:44
electroomal, my device reboot after bluebinder_wait.sh19:07
electroomal, check the end of the journalctl ->
electrooFailed to start Simple proxy for using android binder based bluetooth through vhci..19:09
rinigusevening! I managed to get few messages in the pc logs (lsusb and dmesg) while booting xperia xz2/aosp9 based. paste:
rinigusgetting "ERROR: could not setup USB as usb0 or rndis0"19:11
riniguskernel config:
piggzmal: should a usb keyboard work in sfos?19:15
piggzoh, i fixed19:18
electrooelros34, an idea ? about reboot with bluebinder19:22
electroo@elros34,  an idea ? about reboot with bluebinder19:22
malpiggz: I think it could, I haven't tested much19:39
electroomal, have you an idea about this reboot problem ?19:41
electrooi have found the same problem for this guys Danct12 ->
vknechtrinigus, might want to try with this set of options (or combination), no quarantee tho...19:50
mal@adampigg cool19:52
electrooOo ^^19:53
malelectroo: what reboot problem?19:53
rinigusvknecht: thanks, will try! none taken :)19:53
electrooFailed to start Simple proxy for using android binder based bluetooth through vhci..19:53
electroocheck the last line of my journalctl19:53
piggzmal: the video might look better with a better silica config ;)19:54
malpiggz: :DDD19:54
piggzwas taking with p-o-v obvs tho :)19:54
piggzoh, u gotta try the PR, the reolution label is soo much better19:54
vknechtrinigus, also grep log for these, I noted them from discussion months ago but didn't help much for loire, except for mtp19:55
electroomal, bluebinder.service: Failed with result 'timeout'.19:56
electrooand my device restart19:56
rinigusvknecht: I think I did, but its always possible I missed something19:57
rinigusvknecht: as soon as I enable RNDIS, I get spammed with lots of compile errors as in
electroomy journalctl mal,
nitin03@mal here is my defconfig
nitin03init log
malelectroo: check what you get from "getprop | grep bluetooth" and also "getprop | grep bt"20:03
malnitin03: are you sure you correctly set selinux=0 to kernel commandline?20:04
nitin03audit=0 and selinux=020:04
malno audio=020:04
rinigusvknecht: I'll grep the channel logs20:04
electroomal :
malnitin03: also did you add the needed .rc files20:05
malalso there another thing that is needed20:06
nitin03@mal what other thing?20:07
malfixing ldconfig.28.txt20:07
nitin03where can i find this file?20:09
electroomal, do you know what happen ? maybe i need bluez4, no bluez520:10
electroomaybe it's not that the problem20:11
malnitin03: just a moment, I'm trying to prepare a service to use the file20:15
maltoo many people asking help all the time20:15
electroomal, when i flash the zip file, i must format the cache partition ?20:16 worries20:16
malwe don't use cache partition in sfos20:16
vknechtpiggz, wondering, did you share the twitter vid directly from pov ? tried that already ? from what I saw from share api, this could use some more parameters20:16
piggzvknecht: no, usb, mtp, laptop, twitter20:17
nitin03@mal is this the file
electroomal, mmc0: data txfr (0x00100000) error: -110 after 5369 ms20:25
malnitin03: yes, but you also need the content seen here put to /lib/systemd/system/system-etc-ld.config.28.txt.mount and symlink named /lib/systemd/system/ pointing to /lib/systemd/system/system-etc-ld.config.28.txt.mount20:25
malelectroo: that looks like file system corruption, maybe too many crashes20:25
electroowhat can i do to avoid this ?20:26
malelectroo: maybe your device doesn't have binderized bluetooth so you need to use bluez420:26
electrooi suppose yes20:26
malusually it's possible to have a binder wrapper that uses the old hal20:26
malbut lineage os porters probably haven't enabled it20:27
malpiggz: didn't you need the ld.config.28.txt also on fxtec20:27
malor was it some other device and I confuse things20:28
piggzmal: yes and no20:28
piggzyes, the problem exists20:28
piggzno, fixed a different way20:28
piggzwith bind mounts20:28
piggzor overlays20:28
malpiggz: that is a nice way to bind mount using systemd mount service20:29
malpiggz: not sure how you did it20:29
piggzsystemd service, not a mount unit20:29
malyes, that the other option, running mount command in there20:29
malpiggz: hmm, I don't seem to have any custom silica config on my tablet20:30
malapp grid looks fine with 1.0 pixel ratio20:30
piggzhmm, is everything not large?20:31
malsome parts are, most not20:31
malthe covers are a bit big, app list ok, top menu also a bit big20:31
nitin03@mal ok mal thank there anything i should also cover20:32
malnitin03: how new is the build?20:32
malI mean when did you do repo sync and apply-patches.sh20:33
nitin03i build yesterday20:33
malnitin03: maybe verify  in $ANDROID_ROOT/hybris-patches in your build environment that the first entry you see in git log is
malnitin03: did you understand where you put those files I mentioned earlier20:35
nitin03yes i did20:35
malso in correct subfolder in hybris/droid-configs/sparse20:36
nitin03also this patch is there and i applied it before compiling the boot image20:36
malok, good20:36
malthen you can probably just run " -c" and then build a new image20:37
malhopefully it does something20:37
nitin03i think i should do repo sync and make boot image again just to be sure20:37
malno need, there hasn't been any changes20:38
malrepo sync in this case is not as simple as usualy because of the applied patches20:38
electroomal, it's normal i have this file ? hadk/system/security/keystore/keystore.rc20:38
malelectroo: why do you ask?20:39
electroobecause i have read this file, i'm asking if it's me what i've made this20:39
electroobe appear, it can be a way to stop time_daemon20:40
piggzmal: this is what we do instead of linker change
malelectroo: system/ in $ANDROID_ROOT is from android20:41
electroookey20:42 thing that does fails is nfcd binder plugin package...but i build it separately, also resetting it to 1.0.4...other than that i never ran into any errors20:42
malelectroo: often services can be disabled in /usr/libexec/droid-hybris/system/etc/init/20:42
malnitin03: yes, known issue, the latest version of that plugin requires things that are not in current sfos release, just issues happen sometimes, fix is always to build a bit older version20:43
malwe cannot always prevent that from happening and hardcoding the version in is a bit cumbersome so we rarely do it20:44
malbuilding the old version is not that difficult to do anyway20:46
electroowhat ! /system/etc/init/keystore.rc20:46
malelectroo: what, you are overthinking the issue20:47
malelectroo: check how services have been disabled in /usr/libexec/droid-hybris/system/etc/init/ and do the same for time_daemon20:47
malhopefully that works20:47
nitin03i will get back to u tomorrow...after i test new sfos build20:47
piggzalso, use override keyword in service file to be sure :)20:48
jellyponymal: any ideas about my problem? I cant find anything relatable in the internet. It seems like not working dmesg is only possible if CONFIG_PRINTK=n, but it's not! wtf?20:48
maljellypony: I think some of the changes you made to kernel broke things20:49
jellyponymal: But none of my changes could possibly broke such basic future as logging .-.20:51
electroomal, have you a correct file for time_daemon.rc ?20:56
electroowhy i have this too : Sailfish droid-hal-init: property_set("ro.boottime.keystore", "64222000629") failed: property already set20:56
electrookeystore is android no ?20:56
maljellypony: several things that can be a problem, first of all you have moved around some lines which makes comparing difficult, second you have disabled selinux and haven20:56
maljellypony: several things that can be a problem, first of all you have moved around some lines which makes comparing difficult, second you have disabled selinux and haven't added selinux boot param stuff20:56
jellyponyI already did...20:57
malalso CONFIG_VT=y is missing, CONFIG_DUMMY=n is missing, CONFIG_AUDIT=y was removed20:57
maljellypony: not according to the defconfig in the link you gave earlier20:58
malalso other changes to the defconfig which I have no idea why you made those20:59
jellyponyoof :c21:03
jellyponyanyway thanks. I will try to remake t tomorrow21:05
maljellypony: so I would suggest you make a backup of the current defconfig, then copy the original over it build, run kernel checker and fix only based on that, remember to leave AUDIT=y and add the CONFIG_SECURITY_SELINUX_BOOTPARAM=y21:06
Thaodanhad anyone the error remote unknown command before when trying to boot an image with fastboot boot?21:13
Thaodani try to boot my image on my Xperia  and get the error21:14
Thaodantried two  different cables different pc's.21:14
Thaodanfastboot boot flash works tho21:14
malThaodan: fastboot seems a bit unstable sometimes21:17
malThaodan: some devices need signed boot image for fastboot boot to work21:18
electrooi must go21:19
electroogood night21:19
Thaodanmal: oh okay.  I thought it was me or my cable.  I'll report a bug at Sony if my new cable will not boot.21:27
malThaodan: it's not a problem with sony afaik, well can't be sure, but some issue that usb3 port and some cable are just problematic21:28
Thaodanhm I'll try to look for an usb2 port21:29
Thaodanthose are pretty rare nowadays21:29
malthat has another way to hack around the issue21:32
xreactx[m]__Could the TAR issue be the root of why the installer fails in TWRP?21:38
malit's known that some people have issues that tar causes twrp to run out of memory or just too active oom handler there and it gets killed22:03
malnot sure if the same issue22:04
xreactx[m]__I have 12gb of memory on my OnePlus 7 pro22:18
xreactx[m]__Set android_low_memory_killer=no?22:19
T42<NotKit> it is a bug of tar binary in some TWRP version it seems22:20
xreactx[m]__I tried both TWRP 3.3.1-3 and -422:22
xreactx[m]__NotKit: how can i check to verify if its a tar bug?23:28

Generated by 2.17.1 by Marius Gedminas - find it at!