| genericv2 | purged 'out/' and trying again. | 00:00 |
|---|---|---|
| genericv2 | Same error again on a fully clean build. | 00:26 |
| genericv2 | But i'll continue tomorrow. I am logging out now, thanks for helping T42_ mal | 00:27 |
| genericv2 | hi again, i'll continue now. Any idea to previous initramfs generation error? | 11:07 |
| genericv2 | Anything to check on trees etc. | 11:08 |
| T42_ | <elros34> best way would be if you show whole make output from yesterday after cleaning build output but if you don't have it anymore then just what you have. | 11:11 |
| genericv2 | No problem i can build again. | 11:12 |
| genericv2 | I'll redirect the entire output to a file | 11:12 |
| T42_ | <elros34> show what you have now first | 11:12 |
| genericv2 | https://paste.opensuse.org/pastes/804c43ad2887 | 11:17 |
| genericv2 | Not the entire output but must be enough. | 11:18 |
| mal | just to be sure, which branch of hybris-boot are you using? | 11:25 |
| genericv2 | hybris-20.0 | 11:27 |
| genericv2 | and i am using lineage-20 trees for my phone. | 11:27 |
| mal | I meant the repo in hybris/hybris-boot | 11:30 |
| genericv2 | I don't know if i missed a part, hybris/hybris-boot doesn't have a branch but repo command is clones android13 branch i think. | 11:33 |
| genericv2 | <project path="hybris/hybris-boot" name="mer-hybris/hybris-boot" revision="android13" /> | 11:33 |
| genericv2 | HABUILD_SDK [sagit] twoxa@suse:~/android/hybris/hybris-boot$ git branch => * (no branch) | 11:34 |
| T42_ | <elros34> git branch -a? | 11:35 |
| genericv2 | remotes/github/android13 | 11:36 |
| genericv2 | remotes/m/hybris-20.0 -> github/android13 | 11:36 |
| T42_ | <elros34> so there is no out/target/product/sagit/obj/ROOT/hybris-boot_intermediates/boot-initramfs.gz? | 11:38 |
| mal | could cpio be missing somehow? | 11:39 |
| genericv2 | T42_ yeah, it does not exist. | 11:40 |
| T42_ | <elros34> in HABUILD not in platform sdk | 11:40 |
| genericv2 | cpio is installed, checked again now. | 11:40 |
| genericv2 | In habuild. | 11:40 |
| genericv2 | HABUILD_SDK [sagit] twoxa@suse:~/android$ cpio => cpio: You must specify one of -oipt options. Try 'cpio --help' or 'cpio --usage' for more information. | 11:41 |
| T42_ | <elros34> please show git diff in hybris-boot | 11:43 |
| mal | wondering which compression format that uses | 11:45 |
| genericv2 | did 'git diff | tee gitdiff' | 11:46 |
| genericv2 | https://paste.opensuse.org/pastes/0d801595386a | 11:47 |
| genericv2 | And this is the full log of a clean build: https://paste.opensuse.org/pastes/6e6d1b11ee44 | 11:49 |
| genericv2 | mal, where can i check the compression format | 11:51 |
| T42_ | <elros34> you could search for BOARD_RAMDISK_USE in device/ | 11:52 |
| genericv2 | Nope, nothing related to it on my trees. It using the default probably. | 11:54 |
| mal | I'll test a bit, just a moment | 11:55 |
| genericv2 | Okay i am waiting. Thank you. | 11:56 |
| mal | it's possible that lz4 support patch could be broken, you could try to revert that in hybris-boot | 11:57 |
| mal | the last commit in that repo | 11:57 |
| mal | make sure you don't lose your fixup changes when doing that | 11:58 |
| genericv2 | Reverted that lz4 patch, trying again now. | 12:01 |
| mal | I can reproduce that issue, trying to fix, it's the lz4 support patch causing the issues | 12:02 |
| genericv2 | Yeah reverting that commit fixed the issue | 12:03 |
| genericv2 | HABUILD_SDK [sagit] twoxa@suse:~$ ls android/out/target/product/sagit/obj/ROOT/hybris-recovery_intermediates/ | 12:03 |
| genericv2 | initramfs recovery-initramfs.gz | 12:03 |
| genericv2 | Thank you very much for helping me :) | 12:03 |
| genericv2 | #### build completed successfully (07:05 (mm:ss)) #### | 12:08 |
| mal | something else to note, when using the android13 branch of droid-configs-device submodule it has issues unless you make some changes to it or build new version of mce | 12:17 |
| mal | I think I found a fix to the lz4 support issue, testing with both kind of setups now | 13:02 |
| genericv2 | Hi again, now i'll work on the kernel but i have a question. Android App Support is paid in official Sony devices. Is my unofficial build going to have Android app support? Or i need pay for it? | 16:11 |
| T42_ | <elros34> neither you can try to use free alternative waydroid https://github.com/sailfishos-open/waydroid | 16:15 |
| genericv2 | So i can't use sailfishos's android app implemention :( Thats bad. Actually i only use telegram as a messaging app and no other special android apps but that android api thingy was cool. I'll use waydroid then. | 16:21 |
| T42_ | <elros34> there are many native telegram apps abandoned and maintained | 16:22 |
| genericv2 | Yeah i saw one of them. And i liked it, i love sfos design. I'll use one of the telegram clients and waydroid for banking apps etc, if i succeed porting it of course😅 I'll continue now. | 16:28 |
| genericv2 | Stuck again :( Kernel complication gives an error about this file: https://github.com/LineageOS/android_kernel_xiaomi_msm8998/blob/lineage-20/drivers/video/console/dummycon.c | 18:32 |
| genericv2 | I can fix the complication error by editing lines 118, 119 and 120 but it can't boot recovery. When CONFIG_VT is enabled, phone reboots immediately. And when it's disabled, the phone waits 30~ seconds and reboots. | 18:32 |
| genericv2 | Here is the console output with original dummycon.c + Enabled CONFIG_VT: https://paste.opensuse.org/pastes/bea42037aae2 I asked so many questions, sorry if i am annoying. | 18:32 |
| T42_ | <elros34> you are supposed to have config_cummy disabled then I think you shouldnt have any errors in kernel about that | 18:42 |
| deathmist | genericv2: qcom kernels (specifically older ones) are known to be broken with VT=y, you'll have to revert some commits in your kernel source similar to https://github.com/ubports/porting-notes/wiki/Halium-9#qualcomm-devices-with-linux-kernel-44-panic-with-config_vt-enabled | 18:43 |
| T42_ | <elros34> lineage 15 or something like that but not 20 | 18:44 |
| T42_ | <elros34> oh looks like it really might have this commit https://github.com/LineageOS/android_kernel_xiaomi_msm8998/commit/642b725c9cc052c2c1f2eaf44c34d7ffa8b4290f | 18:45 |
| genericv2 | Got it, i'll try older kernels. | 18:45 |
| deathmist | ?? | 18:45 |
| deathmist | no | 18:45 |
| deathmist | you need to revert some commits | 18:46 |
| deathmist | which break the VT driver | 18:46 |
| genericv2 | Also CONFIG_DUMMY is set to 'n' | 18:46 |
| genericv2 | ah saw it now, okay | 18:46 |
| deathmist | or are you porting on an older hybris version? I've not been following the discussion really | 18:46 |
| genericv2 | Nope i am porteeing on hybris 20 | 18:46 |
| genericv2 | porting* | 18:47 |
| T42_ | <elros34> so to be clear do you have any kernel compilation issues with config_vt=y config_dummy=n? | 18:47 |
| genericv2 | Yes, i have. I am these (https://paste.opensuse.org/pastes/bea42037aae2) errors with vt=y and dummy=n | 18:48 |
| genericv2 | What did i just wrote | 18:49 |
| T42_ | <elros34> ok so lets not mix different issues for now | 18:49 |
| genericv2 | Yeah, now reverting commits. Need a moment. | 18:50 |
| T42_ | <elros34> no | 18:50 |
| T42_ | <elros34> these commits have nothing to your issue | 18:50 |
| genericv2 | what i need to do then now. | 18:54 |
| T42_ | <elros34> can you show your git diff of your defconfig | 18:55 |
| T42_ | <elros34> these lines are from 2.6 kernel so looks like somebody forget to update something: https://github.com/LineageOS/android_kernel_xiaomi_msm8998/blame/lineage-20/drivers/video/console/dummycon.c#L118 | 18:56 |
| genericv2 | Please wait a minute, i'll send it | 19:00 |
| genericv2 | https://paste.opensuse.org/pastes/77a32806b938 | 19:07 |
| T42_ | <elros34> I think it should be safe to modify these 3 functions like here: https://github.com/jameshilliard/linux-grsec/blob/4.4.162-grsec/drivers/video/console/dummycon.c | 19:07 |
| genericv2 | lets try | 19:07 |
| T42_ | <elros34> did you really only make these few changes? mer-kernel-check doesn't complain about anything else? Do you have CONFIG_AUTOFS4_FS set in .config? | 19:09 |
| T42_ | <elros34> I do not like this minimized defconfigs, you never really know what is really set | 19:11 |
| genericv2 | I normally checked all warning too, started clean after seeing it not booting. But these are the only errors i get | 19:11 |
| T42_ | <elros34> not all warnings are important and some are useless | 19:12 |
| genericv2 | Yeah i know. Lot of them are just about NFS | 19:12 |
| genericv2 | Build completed with no errors after dummycon.c modification | 19:14 |
| T42_ | <elros34> if kernel builds fine then reenable config_vt temporary, build and flash hybris-boot | 19:14 |
| T42_ | <elros34> sorry disable not reenable | 19:14 |
| T42_ | <elros34> It is so old issue I didn't expect it in lineage 20 | 19:14 |
| T42_ | <elros34> this kernel is probably not in line with common lineage changes | 19:15 |
| genericv2 | Built again with VT=n (and DUMMY=n). Will flash now | 19:20 |
| genericv2 | Didint able to boot to recovery again, but it didin't crash. | 19:21 |
| genericv2 | VT=y makes it crash immediately | 19:22 |
| T42_ | <elros34> follow host dmesg, see whether it reboots or stays on, you will not see anything on screen yet | 19:22 |
| genericv2 | It's not reconnecting to usb while it's stuck at mi logo | 19:26 |
| T42_ | <elros34> what about all usb stuff explained in hadk? did you check for common boat loader messages and telnet? | 19:28 |
| genericv2 | Just asking to be sure, hybris recovery must boot when there is nothing else installed about sfos/hybris right? | 19:30 |
| T42_ | <elros34> hybris-boot/recover both will boot without sfos rootfs and both with show absolutely nothing on screen | 19:31 |
| genericv2 | Okay i am not doing something wrong then. | 19:32 |
| genericv2 | Can't telnet as there is no usb connection | 19:33 |
| genericv2 | Not sure where to check for bootloader messages. I'll read documention. | 19:33 |
| genericv2 | may /proc/last_kmsg help? | 19:34 |
| T42_ | <elros34> it may but focus on basics so hadk 9.2 then you run dmesg -w, flash hybris-boot.img and power on /connect device | 19:35 |
| mal | check which usb config options you have enabled in defconfig | 19:35 |
| T42_ | <elros34> so far we don't even know whether you have any hybris-boot messages in dmesg or not and this is always first step | 19:36 |
| mal | yeah, good point | 19:36 |
| genericv2 | I always tried recovery. I'll backup my current boot partiton and work with hybris-boot.img now. | 19:37 |
| genericv2 | mal: i'll check now | 19:37 |
| genericv2 | https://paste.opensuse.org/pastes/6c3c3a34c4c8 here is everything about usb in defconfig | 19:39 |
| genericv2 | Yeah, there simply no sign of life in dmesg | 19:42 |
| T42_ | <elros34> device reboots or stays on? | 19:43 |
| genericv2 | It stays up for 30-60 seconds and reboots | 19:44 |
| genericv2 | (VT=y makes it reboot instantly as i said) | 19:44 |
| genericv2 | there is some interesting logs on lsusb -v by the way | 19:44 |
| T42_ | <elros34> so 30 or 60s? Have you checkedd for /data/.stowaways/sailfishos/init.log in recovery | 19:45 |
| genericv2 | from 30 to 60, changes time by time | 19:45 |
| T42_ | <elros34> what about first boot after flashing hybris-boot.img because this is most important | 19:45 |
| T42_ | <elros34> What about CONFIG_AUTOFS4_FS I asked previously? | 19:46 |
| genericv2 | I am not sure if it can write files to /data because the storage is encrypted right now. | 19:48 |
| genericv2 | i'll flash hybris-boot and count how many seconds it takes to reboot nowe | 19:48 |
| T42_ | <elros34> didn't you read hadk-hot, you must remove data encryption | 19:49 |
| genericv2 | CONFIG_AUTOFS4_FS is set to 'y' | 19:49 |
| mal | also userdata has to be ext4, afaik | 19:49 |
| T42_ | <elros34> no, I do not think so this was ever an issue | 19:50 |
| mal | ok | 19:50 |
| genericv2 | Oops, i missed that probably. This is my dailydriver so didin't wiped userdata before starting to port. I'll backup everything and wipe data. | 19:50 |
| T42_ | <elros34> you may take care of this config_vt and kernel changes now too to ease debugging or make small change in hybris-boot script instead | 19:51 |
| genericv2 | so no need to wipe data for now? Would be great because i was planning to wipe it after making some progress | 19:53 |
| T42_ | <elros34> it can ease debugging but its not necessary to get initramfs working, you may try to see if you have usb configs mal was talking about https://github.com/mlehtima/android_kernel_fairphone_sm7225/blob/hybris-20.0/arch/arm64/configs/vendor/fairphone/FP4.config#L154 | 19:54 |
| mal | yeah, if device uses gsi usb then it needs those for sure | 19:55 |
| genericv2 | Oh okay then. If we are unable to make initramfs work, i'll wipe it. | 19:55 |
| genericv2 | Added those lines to kernel. Complication gave a lot of errors. This kernel is a mess | 19:58 |
| genericv2 | kernel/xiaomi/msm8998/drivers/usb/gadget/function/rndis.c file is broken. Another linux 2.x code probably. oof | 20:11 |
| T42_ | <elros34> you can you show full error? | 20:12 |
| genericv2 | https://paste.opensuse.org/pastes/f7e4f1e016af | 20:14 |
| T42_ | <elros34> yeah this common misleading error, I think disabling gsi should help : https://irclogs.sailfishos.org/logs/%23sailfishos-porters/%23sailfishos-porters.2019-07-29.log.html#t2019-07-29T22:46:07-4 | 20:18 |
| genericv2 | Yeah disabling GSI helped with that error. Now another one came: https://paste.opensuse.org/pastes/2124a34fca74 | 20:27 |
| T42_ | <elros34> trey without that CONFIG_USB_CONFIGFS_RMNET_BAM? | 20:29 |
| genericv2 | AA | 20:34 |
| genericv2 | [11708.566594] usb 1-12: USB disconnect, device number 35 | 20:34 |
| genericv2 | [11717.547838] usb 1-12: new high-speed USB device number 36 using xhci_hcd | 20:34 |
| genericv2 | [11717.694793] usb 1-12: New USB device found, idVendor=18d1, idProduct=d001, bcdDevice= 4.04 | 20:34 |
| genericv2 | [11717.694798] usb 1-12: New USB device strings: Mfr=1, Product=2, SerialNumber=3 | 20:34 |
| genericv2 | [11717.694799] usb 1-12: Product: Refusing | 20:34 |
| genericv2 | [11717.694800] usb 1-12: Manufacturer: Mer Boat Loader | 20:34 |
| genericv2 | [11717.694801] usb 1-12: SerialNumber: Mer | 20:34 |
| genericv2 | It's alive :) | 20:34 |
| genericv2 | i'll try to telnet to it, seems like its not rebooting | 20:34 |
| genericv2 | Yeah it's not rebooting. We made progress | 20:36 |
| genericv2 | SerialNumber: Mer Debug telnet on port 23 on usb0 192.168.2.15 - also running udhcpd | 20:38 |
| T42_ | <elros34> so did you managed to telnet? | 20:39 |
| genericv2 | Telnet is getting stuck at "Trying 192.168.2.15" I am trying to fix it. | 20:39 |
| T42_ | <elros34> do you use correct 23 port? | 20:40 |
| genericv2 | Yeah i am using port 23. | 20:40 |
| genericv2 | same at both my host suse and build env ubuntu | 20:42 |
| T42_ | <elros34> follow hadk-hot tips for telnet issues | 20:43 |
| genericv2 | i hope i don't need to force reboot the phone somehow, my power button is broken | 20:45 |
| T42_ | <elros34> ehh | 20:46 |
| genericv2 | xD | 20:47 |
| genericv2 | Command ip link is not detecting my phone | 20:49 |
| genericv2 | this is the problem | 20:49 |
| T42_ | <elros34> whats the output of ifconfig -a | 20:50 |
| mal | is the ip set for the usb interface | 20:50 |
| genericv2 | Nope, no 'usb0' in ifconfig aswell | 20:51 |
| genericv2 | It sees the thing as a "Portable media player", my favorite bootloader | 20:51 |
| mal | sometimes the interface is not usb0 | 20:51 |
| T42_ | <elros34> then device is not powered anymore | 20:51 |
| T42_ | <elros34> connect/disconnnect phone and see what interface appears if ifconfig | 20:52 |
| genericv2 | It must be alive. Disconnecting and reconnecting updates dmesg. And nothing on ifconfig | 20:52 |
| T42_ | <elros34> even with -a argument, really nothing? | 20:53 |
| genericv2 | Yeah, a big nothing except my network adapter | 20:54 |
| genericv2 | A | 20:54 |
| T42_ | <elros34> do you still get New USB device found on every connection? | 20:54 |
| genericv2 | My laptop found it | 20:54 |
| genericv2 | The phone is connected as a network interface now | 20:54 |
| genericv2 | okay i am in | 20:55 |
| genericv2 | i can telnet to phone with my laptop | 20:55 |
| genericv2 | KDE or Suse is did something probably | 20:56 |
| T42_ | <elros34> check if you can mount data partition manually without removing encryption | 20:57 |
| genericv2 | Data is already mounted | 21:01 |
| genericv2 | '/dev/sda17 on /data type ext4' | 21:01 |
| genericv2 | i'll can continue porting now, right? | 21:03 |
| T42_ | <elros34> you need to fix config_vt issue | 21:04 |
| genericv2 | ah true, just forget about that | 21:05 |
| genericv2 | What i need to do now, i forget everything while trying to boot into mer | 21:06 |
| genericv2 | Applied patches in ubports documention | 21:30 |
| genericv2 | Still crashes with VT=y | 21:30 |
| genericv2 | One of the patched failed tho | 21:30 |
| genericv2 | Applied failing patch by-hand. Everything must be okay this time. Starting build again. | 21:44 |
| genericv2 | Complication error. Time to give up. Thanks everyone | 21:48 |
| T42_ | <b100dian> When you say complication instead of compiling you sound less than your potential:p | 21:52 |
| genericv2 | haha xD, i just learnt english by myself. I am mixing words sometimes. | 21:54 |
| genericv2 | i need to logoff now, Unfortunately i am delaying sfos for now. I want it so can't even give up. | 21:56 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!