| sharks | So elros34 pointed out this one last night https://sailfishos.wiki/books/hadk/page/hadk-hot#bkmrk-%2A-for-android-8-10-d? | 00:02 |
|---|---|---|
| sharks | And then subsequently pointed out my kernel had "such nice commitless kernel sources" | 00:02 |
| sharks | Which probably explains why I can't find the two commits I'm looking for in order to revert them. | 00:03 |
| sharks | The alternative solution mentioned is to "force do_skip_initramfs = 0" in https://github.com/sharks-dev/android_kernel_samsung_msm8916/blob/lineage-17.1/init/initramfs.c? | 00:03 |
| sharks | I can't find any mention of do_skip_initramfs in that file. Is the solution as simple as sticking `static int do_skip_initramfs = 0;` at the top of the file? | 00:03 |
| sharks | Beginning to think I've bitten off more than I can chew. | 00:03 |
| ecrn | no, you probably have some othe issue because there is no skipping initramfs possible in that way in your sources | 00:07 |
| ecrn | and no worries, I debugged my device "blind" for about two days | 00:16 |
| ecrn | before I even knew the kernel works | 00:16 |
| sharks | Bugger, okay. Thanks for your help. Must be something else preventing init. | 02:12 |
| ecrn | what I may suggest in such a case is trying to alter the behavior in some way, i.e., since in your case the device hangs without any output, maybe try putting 'reboot -f' somewhere early in the init-script | 02:17 |
| ecrn | to see if it reaches it | 02:17 |
| ecrn | also enabling the watchdog on init could force a delayed reboot | 02:22 |
| ecrn | so that you can check /proc/last_kmsg from TWRP for example | 02:23 |
| ecrn | sharks: well | 02:36 |
| ecrn | there is one idea | 02:36 |
| sharks | Just tried reboot -f, no dice. | 02:36 |
| ecrn | you used msm8916_sec_defconfig as a base for your config | 02:37 |
| ecrn | why is that? | 02:37 |
| ecrn | on the lineage device tree for the device gt510wifi (which I assume is the same as gt5note10wifi) they use a different config | 02:38 |
| ecrn | sharks: https://github.com/retiredtab/android_device_samsung_gt510wifi/blob/lineage-18.1/BoardConfig.mk#L11 | 02:38 |
| ecrn | try applying your changes to the msm8916_sec_gt510wifi_eur_defconfig | 02:39 |
| ecrn | and not msm8916_sec__defconfig | 02:39 |
| ecrn | msm8916_sec_defconfig ^ | 02:39 |
| sharks | Huh. I don't know how I missed that. Do I just make the changes there and `make hybris-boot` again, or do I need to tell the HADK SDK somehow to use the correct defconfig? | 02:42 |
| sharks | No, that doesn't seem right. msm8916_sec_gt510wifi_eur_defconfig contains what looks like device-specific configurations. msm8916_sec_defconfig must still be included as well. Besides, when I make a change in msm8916_sec_defconfig, it's reflected in mer_verify_kernel_config | 02:45 |
| ecrn | yep, you are right, I am not familiar with the qualcomm kernel culture it seems. they pass several defconfigs when configuring, including the msm8916_sec_defconfig | 02:47 |
| ecrn | so as long as there was the msm8916_sec_gt510wifi_eur_defconfig passed, you are good | 02:48 |
| sharks | Well that makes two of us, I'm not familiar with it either! | 02:48 |
| sharks | Stupid question - I still have a handful of warnings (but no errors) in mer_verify_kernel_config. They should not be preventing the kernel from booting, should they? They might stop SFOS from working, or working properly, but I should at least be getting init.log / dmesg outputs | 02:54 |
| ecrn | what warnings exactly? | 03:00 |
| sharks | These ones: https://paste.opensuse.org/pastes/86a8ee0c503e | 03:09 |
| sharks | I mean there's nothing stopping me from fixing them, I just didn't think they'd be what was holding me up at this point. | 03:09 |
| ecrn | probably | 03:16 |
| ecrn | where was the 'reboot -f' placed? on very beginning? | 03:17 |
| sharks | Very beginning yes. | 03:18 |
| T42 | <birdzhang> you can ignore warnings, it's normal | 03:24 |
| sharks | That's kinda what I thought, yeah. | 03:28 |
| ecrn | did you try the lineage build from the tree that you use? | 04:40 |
| T42 | <elros34> you should ignore warnings especially if you have not booting kernel, sometimes even errors. Show your changes from init-script. Did you check whether hybris-boot.img looks symilar to lineage's boot.img with some tool like abootimg or mkbootimg or something like that? | 06:56 |
| sharks | elros - yes, both images look very similar with abootimg: https://paste.opensuse.org/pastes/0f497e416537 | 07:50 |
| sharks | I'm beginning to wonder if it would be worth trying to start from postmarketOS rather than lineageOS? | 07:52 |
| T42 | <elros34> what about boot image header version? I don't see much point in trying postmarketos before reall debugging your current issue. Where are you init-script changes? Have you tried hybris-boot without any defconfig changes to rule out any kconfig breakage? | 12:11 |
| T42 | <elros34> Do you really wait more than 1 minute after first boot according to hadk-hot? | 12:12 |
| T42 | <elros34> also provide your generated .config | 12:20 |
| T42 | <A_T_R> 4.6.0.15 is the latest sdk version? | 13:03 |
| T42 | <btchdRPR> hey all!! | 13:55 |
| T42 | <btchdRPR> I have an OnePlus 7T pro phone here that is in need of a port :) | 13:55 |
| T42 | <btchdRPR> | 13:55 |
| T42 | <btchdRPR> I have sources for Lineage 18.1 available, but unsure if that's the best point to start with. | 13:55 |
| T42 | <btchdRPR> | 13:55 |
| T42 | <btchdRPR> Also, do we need to build a full system image? Or do it like the standalone kernel method like ubpoets? | 13:55 |
| T42 | <btchdRPR> | 13:55 |
| T42 | <btchdRPR> Newbie - please go easy on me :) | 13:55 |
| T42 | <btchdRPR> thanks | 13:55 |
| T42 | <btchdRPR> as I assume android 14 base isn't going to be supported yet | 14:00 |
| mal | android 14 base is supported | 14:15 |
| T42 | <btchdRPR> oh wow, i assumed it isn't because of droidian/ubtouch not having halium14. | 14:22 |
| T42 | <btchdRPR> | 14:22 |
| T42 | <btchdRPR> do you need to build a full system image? because i have a nice a14 target (redmi k70 pro) but it has no port of lineage/anything (re @SailfishFreenodeIRCBridgeBot: <mal>android 14 base...) | 14:22 |
| T42 | <Verevka86> based on lineage-21 there is no manifest in mer-hybris | 14:24 |
| mal | oh, I forgot to push the manifest | 14:26 |
| mal | I have done two ports with hybris-21 already | 14:27 |
| T42 | <btchdRPR> wew. | 14:28 |
| T42 | <btchdRPR> | 14:28 |
| T42 | <btchdRPR> i assume then the first step would be to port lineage to the device? | 14:28 |
| mal | well there is the halium way to port also | 14:30 |
| T42 | <A_T_R> latest sdk version 3.11 ? | 15:28 |
| T42 | <A_T_R> can i use with mix versions of sdk and target,toolings? | 15:41 |
| T42 | <A_T_R> 1.Sailfish_OS-4.6.0.13-Platform_SDK_Chroot-i486.tar.bz2 | 15:41 |
| T42 | <A_T_R> 2.Sailfish_OS-5.0.0.29-Sailfish_SDK_Target-aarch64.tar.7z | 15:41 |
| T42 | <A_T_R> 3.Sailfish_OS-5.0.0.29-Sailfish_SDK_Tooling-i486.tar.7z | 15:41 |
| T42 | <A_T_R> Clean install | 15:41 |
| T42 | <A_T_R> for hybris-17.1 branch | 16:13 |
| T42 | <A_T_R> this is okay `repo sync -c --no-clone-bundle --no-tags --optimized-fetch --prune --force-sync -j8` ? | 16:13 |
| T42 | <Verevka86> <mal> Can you push it? I would love to try it on my device. | 18:59 |
| mal | I'll check that it's properly minimized first | 19:20 |
| T42 | <elros34> @A_T_R probably old sdk will be fine but you can update it manually too | 19:28 |
| T42 | <elros34> I assume all repos for sfos 5.0 are available but I did not check it | 19:29 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!