| T42 | <MaxRegnerasif> Hi Guys, ist there a Port for Xperia 10 iv? Thanks | 09:10 |
|---|---|---|
| Caffeine | Heya peoples, I'm currently porting the Pixel XL and I've hit a roadblock. I'm at "5.4 Building Relevant Bits of your Android base" in the HADK documentation where I'm currently building hybris-hal and droidmedia. I run "make -j$(nproc --all) hybris-hal droidmedia" and basically it gives me an error saying this "FAILED: ninja: 'ueventd.qcom.rc', needed by 'hybris-hal', missing and no known rule to make it". | 21:48 |
| Caffeine | It's looking for a file that currently doesn't exist cause it's been renamed by the lineage porters | 21:49 |
| mal | are you sure you have all needed device repos and correct branches of those | 21:49 |
| Caffeine | Lineage-18.1 device, and kernel | 21:50 |
| mal | can you link to the device repo | 21:50 |
| Caffeine | Of course, give me a second please | 21:50 |
| mal | did you apply patches? | 21:50 |
| Caffeine | I did. | 21:50 |
| Caffeine | https://github.com/LineageOS/android_device_google_marlin | 21:51 |
| Caffeine | https://github.com/LineageOS/android_kernel_google_marlin | 21:51 |
| Caffeine | these are the repos in question | 21:51 |
| mal | yeah, seems those are only ones needed | 21:51 |
| Caffeine | (again, both using the "lineage-18.1" branch | 21:51 |
| Caffeine | Very strange then, it's still looking for "ueventd.qcom.rc" | 21:52 |
| Caffeine | https://github.com/LineageOS/android_device_google_marlin/blob/lineage-18.1/ueventd.common.rc | 21:52 |
| Caffeine | This is the renamed file | 21:53 |
| Caffeine | I renamed it back to see what would happen, but it still gives me the same thing | 21:53 |
| Caffeine | The annoying thing is is that it's not telling me where it's looking for it. | 21:54 |
| mal | https://github.com/mer-hybris/hybris-boot/blob/master/Android.mk#L299 | 21:54 |
| mal | hmm, do you have vendor repo? assuming your device needs one | 21:54 |
| Caffeine | Oh! I have an entire story to tell you about that.. :skull: | 21:55 |
| Caffeine | It has a script called "extract-files.sh" which extracts all of the vendor files from a device. | 21:56 |
| Caffeine | Which is what I did obviously | 21:56 |
| Caffeine | but it didn't pull a lot of them | 21:56 |
| Caffeine | give me a second | 21:56 |
| mal | usually easier to use https://github.com/TheMuppets/proprietary_vendor_google/tree/lineage-18.1 | 21:56 |
| Caffeine | I was unaware of this, thanks for showing me that. | 21:57 |
| Caffeine | I'll save it | 21:57 |
| mal | themuppets is the usual place to find vendor repos | 21:57 |
| T42 | <elros34> maybe this line breaks it: https://github.com/LineageOS/android_device_google_marlin/blob/lineage-18.1/common/base.mk#L241 | 21:57 |
| Caffeine | I feel like I've seen it before | 21:57 |
| Caffeine | Oh wow! That was quick | 21:58 |
| mal | yeah, change that to ueventd.commom.rc | 21:58 |
| Caffeine | I'll see what happens if I change that | 21:58 |
| Caffeine | Will I have to reclone? | 21:58 |
| mal | no | 21:59 |
| mal | just change it in your local copy and then try building | 21:59 |
| Caffeine | Cheers | 22:00 |
| Caffeine | lets see if it works | 22:00 |
| Caffeine | FAILED: ninja: 'ueventd.common.rc', needed by 'hybris-hal', missing and no known rule to make it | 22:02 |
| Caffeine | damn :P | 22:02 |
| Caffeine | I need to read this file again | 22:02 |
| mal | you said you renamed the file | 22:03 |
| Caffeine | I changed it back before compiling | 22:03 |
| mal | ok | 22:03 |
| mal | grep for ueventd in device repo | 22:03 |
| Caffeine | sure | 22:03 |
| Caffeine | This'll take a while | 22:05 |
| mal | why? | 22:06 |
| Caffeine | I don't use grep often, is it just "grep ueventd"? | 22:07 |
| Caffeine | found it | 22:09 |
| T42 | <elros34> more like 'grep -rn ueventd $ANDROID_ROOT/device/' | 22:09 |
| Caffeine | https://pastebin.com/RvbeXqNK | 22:10 |
| Caffeine | This is what it showed up with | 22:10 |
| T42 | <elros34> you were supposed to change: base.mk:241:INIT += ueventd.qcom.rc to > 'ueventd.common.rc | 22:12 |
| Caffeine | Oh this is a seperate one that I cloned | 22:12 |
| Caffeine | I should've done the one in hadk | 22:12 |
| Caffeine | https://pastebin.com/hRSmC1X9 | 22:13 |
| Caffeine | my bad | 22:13 |
| T42 | <elros34> so it still complain now when you make hybris-hal? | 22:15 |
| Caffeine | It does | 22:16 |
| T42 | <elros34> so remove that line and let's see what will blow up | 22:16 |
| Caffeine | That's an idea. | 22:16 |
| Caffeine | I started a clean build before you sent that, the build didn't get far before getting to the same point again. Worth a try. | 22:18 |
| Caffeine | Well, it hasn't blown up so far | 22:19 |
| Caffeine | FAILED: ninja: 'fstab.common', needed by 'hybris-hal', missing and no known rule to make it | 22:19 |
| Caffeine | Alright, something different | 22:20 |
| Caffeine | It's in the root of the marlin device folder | 22:22 |
| Caffeine | what!? | 22:22 |
| mal | Caffeine: I think maybe remove the two lines from base.mk | 22:28 |
| mal | because those end up in PRODUCT_PACKAGES list and that is checked in hybris-boot Android.mk | 22:28 |
| Caffeine | Okay, remove fstab.common? | 22:29 |
| mal | yes | 22:29 |
| Caffeine | all good, thanks | 22:30 |
| Caffeine | I feel like this is going to end up being a lot of removing lines | 22:30 |
| Caffeine | FAILED: ninja: 'vendor/google/marlin/proprietary/bin/move_widevine_data.sh', needed by 'out/target/product/marlin/system/bin/move_widevine_data.sh', missing and no known rule to make | 22:30 |
| mal | just be sure those files really still end up in out | 22:30 |
| Caffeine | Alright! | 22:31 |
| mal | so clone the repo I linked to vendor/google | 22:31 |
| Caffeine | sure | 22:31 |
| mal | for that latest issue | 22:31 |
| Caffeine | Where should I clone it to | 22:32 |
| Caffeine | Oh I see | 22:33 |
| Caffeine | nevermind | 22:33 |
| T42 | <elros34> look at repo url and error then try to guess :) | 22:33 |
| Caffeine | Just gotta wait for it to clone lol | 22:34 |
| Caffeine | then I think I put the marlin folder where the old one was in hadk/vendor/google/marlin | 22:35 |
| T42 | <elros34> I have impression that marlin was one of the first problematic device with very early bootloop because of broken tty kthread commit in kernel and I still see this commit in lineage-18.. after all these years | 22:36 |
| Caffeine | Oh dear | 22:36 |
| Caffeine | This shouldn't deter me, right | 22:36 |
| Caffeine | :/ | 22:36 |
| T42 | <elros34> not at all. Didn't you find some old sailfish image or nobody ever released it? | 22:37 |
| Caffeine | I didn't find one | 22:37 |
| Caffeine | So I'm going to pick it up, hopefully. | 22:37 |
| Caffeine | I'm new to porting Sailfish you see, so I'm only just getting my start | 22:37 |
| Caffeine | I've ported PostmarketOS to one of my phones, but this is a lot different. | 22:38 |
| Caffeine | Also, may I just say thanks for the help so far. I very much appreciate it :) | 22:42 |
| T42 | <elros34> "move tty_port workqueue to be a kthread " https://github.com/LineageOS/android_kernel_google_marlin/commits/lineage-18.1?after=52bcca9b8e2d97a61c289f294b56bf4fbed1f12d+69&branch=lineage-18.1&path%5B%5D=drivers&path%5B%5D=tty&qualified_name=refs%2Fheads%2Flineage-18.1. Maybe also "tty: add tty_port_set_policy function" for clean git revert | 22:42 |
| Caffeine | What does this mean | 22:44 |
| T42 | <elros34> revert it otherwise you will have bootloop without any logs except via uart | 22:45 |
| Caffeine | Oh, okay | 22:46 |
| Caffeine | FAILED: ninja: 'libhwc2_compat_layer', needed by 'hybris-hal', missing and no known rule to make it | 22:48 |
| Caffeine | this is weird | 22:48 |
| Caffeine | I can't seem to find "libgwc2_compat_layer" | 22:48 |
| T42 | <elros34> you didn't clone libhybris according to hadk-faq | 22:49 |
| Caffeine | but it's here | 22:49 |
| Caffeine | oh well | 22:49 |
| Caffeine | how do I mess that up! | 22:50 |
| Caffeine | I'll do that now! | 22:51 |
| T42 | <elros34> more sketchy notes to read: https://etherpad.wikimedia.org/p/hadk-hot | 22:55 |
| Caffeine | Man, I messed up again | 22:57 |
| Caffeine | I cloned master and realized it wasn't the android11 branch | 22:58 |
| Caffeine | fatal: previous rebase directory .git/rebase-apply still exists but mbox given. | 22:58 |
| Caffeine | and I deleted it and cloned it again | 22:58 |
| Caffeine | (should've switched branch) | 22:58 |
| mal | what are you talking about android11 branch? | 22:58 |
| Caffeine | libhybris | 22:58 |
| mal | use master | 22:59 |
| Caffeine | ughhhh | 22:59 |
| mal | android11 is some old leftover from development | 22:59 |
| mal | I should cleanup the old branches | 22:59 |
| Caffeine | my tree is a mess | 22:59 |
| Caffeine | I hate this | 22:59 |
| T42 | <elros34> hadk is not walk in the park:) | 23:00 |
| Caffeine | I will survive | 23:01 |
| Caffeine | I changed branch | 23:01 |
| Caffeine | repatched | 23:01 |
| Caffeine | and it's doing it! | 23:01 |
| mal | I think I will cleanup the libhybris branches tomorrow, found several old ones | 23:01 |
| T42 | <elros34> but what are you patching? you only needed to clone libhybris | 23:02 |
| mal | yeah, libhybris should just work | 23:02 |
| mal | just recursively clone master branch from https://github.com/mer-hybris/libhybris | 23:03 |
| Caffeine | oh well, then this is just gonna be a messy thing then *shrug* | 23:03 |
| Caffeine | it's compiling now, so I'm not complaining | 23:03 |
| Caffeine | Oh I didn't revert that thing | 23:03 |
| Caffeine | damnit | 23:04 |
| T42 | <elros34> don't worry you will build kernel many times:) | 23:04 |
| Caffeine | Well at least I ain't hurting anyone else by hurting myself | 23:05 |
| Caffeine | > Maybe also "tty: add tty_port_set_policy function" for clean git revert | 23:10 |
| Caffeine | so try reverting to this commit? | 23:10 |
| Caffeine | #### build completed successfully (16:30 (mm:ss)) #### | 23:17 |
| Caffeine | Lets goo! | 23:17 |
| T42 | <elros34> IIRC yes revert these 2 commits | 23:17 |
| Caffeine | how do I revert two :thinking: | 23:17 |
| Caffeine | wait, how do I do that when it's using repo? | 23:19 |
| T42 | <elros34> git revert <commit hash> | 23:25 |
| Caffeine | Okay, reverted both | 23:27 |
| Caffeine | eventually | 23:27 |
| Caffeine | https://pastebin.com/amXGWdgd | 23:27 |
| Caffeine | I think this worked? | 23:27 |
| Caffeine | yeah it did | 23:28 |
| Caffeine | nice | 23:28 |
| Caffeine | I'll just recompile the kernel rq | 23:28 |
| Caffeine | Oops, forgot to fix the merge conflict :sob: | 23:31 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!