| *** nephros_ is now known as nephros[c2] | 06:41 | |
| *** nephros[c2] is now known as nephros_ | 06:47 | |
| *** nephros_ is now known as nephros | 06:49 | |
| *** nephros is now known as nephros[c2] | 06:50 | |
| *** nephros is now known as nephros[c2] | 07:04 | |
| *** nephros[c2] is now known as nephros | 07:11 | |
| rinigus | @b100dian: I have just finished the build and I think you could close https://github.com/sonyxperiadev/bug_tracker/issues/846 | 07:15 |
|---|---|---|
| *** nephros is now known as nephros[c2] | 07:15 | |
| rinigus | the merged PR fixed it | 07:16 |
| T42 | <b100dian> rinigus: good job | 08:58 |
| T42 | <b100dian> I like that references to r62 appear and the discussions | 08:59 |
| rinigus | @b100dian: looks like there is some progress :) | 09:04 |
| *** nephros_ is now known as nephros[c2] | 10:11 | |
| *** ggabriel is now known as Guest8062 | 16:27 | |
| shebang | hi | 17:54 |
| rinigus | @b100dian: flashed aosp13 with disabled fpc (commented out starting of it) and temp was already fine. | 19:44 |
| rinigus | tested wifi and it crashed again. this time have a logcat from crash. | 19:45 |
| rinigus | could be that it is since I had android 14 stock before and switched to asop13. at least I hope so | 19:45 |
| T42 | <b100dian> rinigus: nice! any change in camera after that vintf patch? | 19:46 |
| rinigus | there was some recommendation to use android13 as a base in this case. but hopefully we will get patches for aosp14 soon | 19:46 |
| T42 | <b100dian> Is this 5.15 right? | 19:47 |
| T42 | <b100dian> I am trying to build mer-kernel-check'd hybris-recovery for 5.15 without much sfossdk, just hybris-boot and deps | 19:47 |
| T42 | <b100dian> and to my surprise, `fastboot boot hybris-recovery.img` is not supported | 19:48 |
| rinigus | no, camera still off. I tried to define TARGET_USES_QTI_CAMERA, but it didn't help. even there are some camera blobs in SW_xxx, but couldn't make it work | 19:48 |
| T42 | <b100dian> rinigus: it would be a good time to grab some logcat from camera too I guess? Maybe that helps the existing GSM + camera nagara bug | 19:49 |
| T42 | <b100dian> (in my experience, logs are so large you need a github gist or the like for them, no paste service copes) | 19:49 |
| T42 | <b100dian> so, back to hybris-recovery.img booting: I flashed it, and it didn't boot because your phone cannot be checked / is broken etc. (don't remember the exact message) | 19:50 |
| T42 | <b100dian> so I reverted to the stock boot - which I can grab from the device because Lineage GSI has adb root option | 19:50 |
| rinigus | although, to boot and pass android logo, I had to add one more catch (...) in device/sony/common/hardware/health/CycleCountBackupRestore.cpp as stoi exception managed to sneak through std::invalid_argument | 19:50 |
| T42 | <b100dian> Hmm.. how did you figure that out, do we have /sys/fs/pstore logs? | 19:51 |
| rinigus | on aosp13, adb was enabled from boot. so, I could grab while booting and saw this (just a sec) | 19:52 |
| rinigus | https://paste.mozilla.org/oK2WQ7pN/raw | 19:53 |
| rinigus | it had a message in debug: | 19:53 |
| rinigus | Abort message: 'terminating with uncaught exception of type St16invalid_argument: stoi: no conversion' | 19:53 |
| rinigus | and stack. from that was easy to find | 19:54 |
| T42 | <b100dian> weird, I didn't have that - although I kindof dismissed stuck to logo builds - but the initial 5.15 kernel build was booting | 19:55 |
| T42 | <b100dian> It's good to know that adb shell is avalable early though | 19:55 |
| rinigus | will check the kernel. should be 5.15. | 19:56 |
| rinigus | Linux version 5.15.94 | 19:56 |
| rinigus | as for why boot hybris recovery doesn't work, not sure. there was no boot attempt even? | 19:59 |
| T42 | <b100dian> ah, back tot that - there may be other problems (such as mismatching dtbo or modules w.r.t stock) but my first guess would be vbmeta for "device corrupted" or something like that | 20:01 |
| T42 | <b100dian> rinigus: so, question, did you have to build/flash a vbmeta previously on tama (or empty vbmeta?) | 20:02 |
| rinigus | yes, we had. looking for script | 20:03 |
| rinigus | procedure was as in https://github.com/sailfishos-sony-tama/main/blob/hybris-10/flashing.md , look towards end | 20:04 |
| T42 | <b100dian> rinigus: but the way you build that vbmeta, is it just an empty one or .. ? because building it from make vbmetaimage would probably mean building boot.img (not hybris-boot one) and hashing that.. | 20:10 |
| rinigus | @b100dian: it was built using aosp system part. see https://github.com/sailfishos-sony-tama/main/blob/hybris-10/hadk-sony-xz2.md#build-systemimage-vendorimage | 20:12 |
| rinigus | and then used just that | 20:13 |
| T42 | <b100dian> thanks, will look into this | 20:14 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!