| budrz | lineage kernel sources: https://github.com/LineageOS/android_kernel_google_msm-4.9/tree/lineage-18.1 | 00:12 |
|---|---|---|
| T42 | <elros34> AFAIK this should be no problem in android 11 but you can check cat /proc/cmdline in lineage (not recovery) to make sure there is no skip_initramfs but more likely it's something else | 00:23 |
| budrz | here is my kernel sources: https://gitlab.com/budrzproj/android_kernel_google_bonito | 00:33 |
| budrz | ok, I'll check | 00:33 |
| budrz | there is no skip_initramfs | 00:37 |
| budrz | the boot.img didn't work. same messages as before | 00:43 |
| budrz | so are you suggesting reverting to the originial defconfig, except maybe some of the config "ERROR"s fixed from the hybris kernel config check? | 00:48 |
| budrz | I'll pick this up tomorrow | 00:55 |
| T42 | <elros34> yes revert to original, if that will lead to the same results then maybe you are making abootimg incorrectly. You have working lineage kernel to compare | 06:21 |
| T42 | <elros34> again check what is in host dmesg, because I still do not know | 06:24 |
| voidanix[m] | broken vibration, camera and video playback all at once means gstreamer is broken right? | 15:26 |
| voidanix[m] | here's a jolla-gallery with GST_DEBUG https://paste.opensuse.org/76419158 and ngfd+gdb https://paste.opensuse.org/62219180 just in case | 15:27 |
| voidanix[m] | urgh, figured out my issue: https://github.com/mer-hybris/hybris-patches/commit/6931424e20da1bb93bb0cf4d07ff7102874a81e8#diff-d304b8ff3fcf1d700749c7f1d67533a0aba477b37d1693998dfaaa7e566e14a3R16 | 19:53 |
| voidanix[m] | ping @Thaodan: does using the commit above require another commit somewhere else? | 19:55 |
| T42 | <elros34> according to description yes: some commit in droidmedia | 19:57 |
| voidanix[m] | that one is already merged, it seems there is something for dhd though | 20:01 |
| Thaodan | voidanix[m]: there's one pr in dhd for packaging that you need to use if you build this locally | 21:04 |
| budrz | elros34: I tried creating the image with the original defconfig and modified to remove the errors from the mer-kernel-config check script but still no progress | 21:36 |
| budrz | dmesg has the same messages as before | 21:37 |
| budrz | so I guess maybe I'm making abootimg incorrectly | 21:37 |
| T42 | <elros34> again: I still do not know what are messages in your host dmesg at first boot | 21:38 |
| T42 | <elros34> you showed flood of logs, I have no idea when you flashed kernel, when it try to boot and when it switch slot and starts android | 21:38 |
| T42 | <elros34> anyway in android sources you should have unpack_bootimg or something like that, if I remember correctly this tool prints a lot of info about abootimg which you can use to compare with lineage's boot.img | 21:40 |
| budrz | there's unpack_bootimg.py in system/tools/mkbootimg/ | 21:44 |
| budrz | is that it? | 21:44 |
| T42 | <elros34> probably, try and you will see | 21:44 |
| budrz | elros34: here is the output of the image created using the original defconfig: https://gitlab.com/budrzproj/sailfish-log-files/-/blob/main/boot-img-unpack-br-w-error.log | 21:56 |
| budrz | and one with the errors removed from the kernel check script: https://gitlab.com/budrzproj/sailfish-log-files/-/blob/main/boot-img-unpack-br-wo-error.log | 21:57 |
| budrz | there's not much of a difference: https://gitlab.com/budrzproj/sailfish-log-files/-/blob/main/hybris-w-wo-error.diff | 21:57 |
| T42 | <elros34> they will be the same, compare with working one so:lineage's boot.img | 21:57 |
| budrz | ok, hang on | 21:58 |
| budrz | here is the lineage output: https://gitlab.com/budrzproj/sailfish-log-files/-/blob/main/boot-img-unpack-lineage.log | 22:01 |
| budrz | and the diff between lineage and the image with out errors from the config script: https://gitlab.com/budrzproj/sailfish-log-files/-/blob/main/hybris-lineage-boot-img.diff | 22:02 |
| budrz | there's much more of a difference here... | 22:03 |
| T42 | <elros34> not really I was hoping for dtb differences but no. | 22:04 |
| budrz | seems like I'm missing the androidboot.super_partition command line argument though (among other things). not sure if adding that would help | 22:05 |
| T42 | <elros34> Have you tried flashing some empty vbmeta image? no idea whether this is needed for your device. Try adding 'sleep 40': here https://github.com/mer-hybris/hybris-boot/blob/master/init-script#L345 then mka hybris-boot and flash hybris-boot.img (can be without any defconfig changes). Provide dmesg from first boot after flashing and reboot also timing. | 22:23 |
| T42 | <elros34> or maybe even in line 343 | 22:27 |
| budrz | there is a vbmeta.img for lineage and the stock rom | 23:04 |
| budrz | I will try adding the sleep command to the init script. can the vbmeta image be from lineage? | 23:08 |
| T42 | <elros34> I am not sure about it but AFAIK it should be empty vbmeta flashed so this kind of verification is disabled | 23:10 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!