Tuesday, 2020-01-14

« Monday, 2020-01-13 Index Wednesday, 2020-01-15 »
T42<Barnett %lastname%> (Photo, 471x539) https://irc.thaodan.de/.imgstore/CDut823Ixs.png [1]高仿鞋出售[2]微信群电报群群活跃度管理,水军[3]电报微信群推[4]电报/potato土豆/微信/Facebook/Twitter增粉[5]非小号,mytoken关注 [私聊][private]: @hyint03:23
Mister_Magistermal: on my moto x2 /dev/block/platform/msm_sdcc.1/mmcblk* is all broken symlink.07:53
Mister_Magisterwhat if it's not relative symlink problem but the symlink itself is broken?07:54
Mister_MagisterSYMLINK+="block/platform/$env{PLATFORM_FOLDER}/$env{PLATFORM_DEVICE}/$name" two times platform?07:56
Mister_Magisteroh wait platform_device is prably by_name and stuff08:00
Mister_Magistermal: absolute symlink: ENV{ID_PART_ENTRY_SCHEME}=="gpt", ENV{ID_PART_ENTRY_NAME}=="?*", RUN+="/bin/ln -s /dev/$name /dev/block/platform/$env{PLATFORM_FOLDER}/$env{PLATFORM_DEVICE}/$name"08:16
r0kk3rzKALUBE: did you request to join on gitlab?08:20
T42翁迪荔 %lastname% was added by: 翁迪荔 %lastname%08:42
Mister_Magisterwhy do we put .tar.bz2 inside zip? what sense does it make? could we just put .tar and then change unpack script to unpack it? would remove bzip2 requirement14:33
Mister_Magisterthat current twrp doesn't meet14:33
Mister_Magistermal:  update, on my 5z the symlink is proper16:07
Mister_Magisteraka valid16:07
malMister_Magister: maybe should show output of ls -lR /dev/block so I can how things were broken16:14
Mister_Magistermhm later16:14
Mister_Magistermal: that should be enough even :P https://imgur.com/a/lm0gEaB16:15
malso then find out which line in the udev rule causes that16:16
Mister_Magisteri know which line i even gave you line that makes absolute symlink :316:17
Mister_Magisterah yes16:19
Mister_Magisterby-name is broken16:19
Mister_Magisterevery single file :D16:19
malcan you show the config git repo for the device16:27
Mister_Magisterconfig… droid-config you mean? sure16:27
malyes of course16:28
Mister_Magisterbootcontrolhal: update_slot_attribute: Failed to get pentry/pentry_bak for xbl_a OHH WHAT NOW :<16:28
Mister_Magistermal: http://github.com/verdanditeam/droid-config-victara16:28
Mister_Magisterthis one had /dev/block/platform/blablabla/mmc* broken16:28
Mister_Magisteron the 5z both are good (By name too i was wrong)16:28
malyou need to give more information otherwise I cannot debug this, what is so different in that one device, udevadm might tell something16:30
Mister_Magisteryay got bootctl working! :316:39
Mister_Magistermal: it doesn't matter anyway don't trouble yourself16:39
Mister_Magistermal: got udev rules that fix bootctl! ENV{ID_PART_ENTRY_SCHEME}=="gpt", ENV{ID_PART_ENTRY_NAME}=="?*", RUN+="/bin/mkdir -p /dev/block/platform/$env{PLATFORM_FOLDER}/$env{PLATFORM_DEVICE}/by-name" ENV{ID_PART_ENTRY_SCHEME}=="gpt", ENV{ID_PART_ENTRY_NAME}=="?*", RUN+="/bin/ln -s /dev/block/$name /dev/block/platform/$env{PLATFORM_FOLDER}/$env{PLATFORM_DEVICE}/by-name/$env{ID_PART_ENTRY_NAME}" ENV{ID_PART_ENTRY_SCHEME}=="gpt", ENV{ID_16:41
Mister_MagisterPART_ENTRY_NAME}=="?*", RUN+="/bin/ln -s /dev/block/$name /dev/block/platform/$env{PLATFORM_FOLDER}/$env{PLATFORM_DEVICE}/$name"16:41
Mister_Magisterthat didn't work too well :/16:41
malwhich lines do those replace in the original udev rules16:42
Mister_MagisterHere fix for bootctl Error marking as having booted successfully (for logs reference) https://paste.opensuse.org/3394115516:42
Mister_Magisterreplace this two lines https://paste.opensuse.org/6900509916:43
Mister_Magistermal: ^16:43
malthe real question is why did the original ones fail16:44
malis either PLATFORM_DEVICE or PLATFORM_FOLDER empty?16:45
malwhich would lead to something//something in the path and then automatic symlinking would understand it incorrectly and add wrong relative path16:45
malhmm, but that cannot really happen I think16:46
malyou should verify that16:47
Mister_Magistermal: original ones are relative17:07
Mister_Magisterrelative don't work17:07
Mister_Magisterthey are correct, bootctl just can't read them properly17:08
Mister_Magisterthere is nothing we can do about it except for hacking it but what's the point17:08
malI'm confused now, so the original file created working symlink and it was some modifications you made that broke those?17:11
Mister_Magisterno17:12
malI'll ask in another way, does this file https://github.com/mer-hybris/droid-hal-configs/blob/master/sparse/lib/udev/rules.d/998-droid-system.rules create symlinks that point to non-existent targets?17:13
Mister_Magisterno17:13
malthat is what I asked in the above line17:13
Mister_Magistersame answer :P17:14
Mister_Magistersec17:14
malbut different question17:14
malthose questions were opposite ways of asking the same thing, you answered no to both17:14
malso what broken symlinks were you then talking about earlier17:15
malI don't understand where those come from17:15
Mister_Magisterjust give me second writting explanation17:15
Mister_Magistermal: on devices with hybris 15/16 there is bootctl to mark if slot is sucessfull or not. default binary reads symlink at /dev/block/platform/something/by-name/xbl_a which with relative symlink gives ../../blabla/sdXX (which is proper working symlink) and it returns error (it's a directory or something). android has absolute symlinks and those are working. Some people patched their gpt-utils but you said to make absolute symlink so17:16
Mister_Magisterthat's what i did. so now with absolute symlink bootctl is happy17:16
Mister_Magisterno broken symlinks17:17
malthen what was this about 09:53 < Mister_Magister> mal: on my moto x2 /dev/block/platform/msm_sdcc.1/mmcblk* is all broken symlink.17:18
Mister_Magisterbut that's for hybris 15/16 my moto z and asus 5z. The broken symlink i was talking about is different story and you don't need to worryt about it. my moto x2 which has 12.1 base i believe has broken every symlink under /dev/block/platform/ with default rules (absolute symlinks fix them ofc)17:18
malI was talking about that x2 case all the time17:18
Mister_Magisteroh sorry then17:18
Mister_Magisterbut it doesn't matter as phone is working fine17:19
malI was talking about broken symlinks17:19
Mister_Magisterit gives one .. too many17:19
Mister_Magisterbut it seriously doesn't matter17:19
malso did you check if either PLATFORM_DEVICE or PLATFORM_FOLDER is empty?17:19
malif it doesn't matter then why did you even talk about it, I wasted my time trying to figure it out17:20
Mister_Magistermal: told you not to worry about it17:20
Mister_Magisteru have no idea what platform_device is but yes one of them always on every device has to be empty because rule is /dev/block/platform/platform_folder/platform_device/mmcetc which on every device gives /dev/block/platform/xxxx/mmcetc17:20
Mister_Magisteri have no idea*17:20
Mister_Magisteryou can check on your own fairphone17:22
malyou only now said not to worry about it17:23
Mister_Magistermal: if i didn't annoy you enough can you tell me what i have to do to fix audio on hybris 16 or you don't know17:23
Mister_Magistermal: yes right, forget it17:23
Mister_Magisterjust in case you will be curious ;P17:24
malwithout logs i cannot help with audio issues, you should know it17:24
Mister_Magisterno i thought there is something necessary  that needs to be done i order to get audio17:25
Mister_Magisterlike binder plugin or something17:25
malnormal or call audio?17:25
Mister_Magisternormal17:25
malnothing special afaik17:25
Mister_Magisteraudioflinger-glue maybe? shouldn't be needed right. there was even somethnig replacing it17:26
Mister_Magistermaybe symlinks17:26
malsome need this for call audio https://github.com/mer-hybris/pulseaudio-modules-droid-hidl17:27
malit depends on what kind of setup your device has17:28
Mister_Magisterye ye that replaced audioflinger-glue17:28
Mister_Magistermaybe i will ask jusa when he is around17:28
malLOGS!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!17:29
Mister_Magistersorry sorry17:29
Mister_Magistersorry again17:29
* Mister_Magister is annoying as hell17:30
malMister_Magister: fp2 also has some broken symlinks in /dev/block/platform/msm_sdcc.1/mmc*17:36
Mister_Magistermal: :)17:49
Mister_Magistersee it's not important as phones are working17:49
Mister_Magisterjust interesting thing17:49
Mister_Magisterif your perfectionism on fp2 requires it you can fix it :P17:50
Mister_Magistermal: sorry here are logs :3 https://gist.github.com/Michal-Szczepaniak/06a3758f8d73ade082830e9f099534aa17:56
Mister_Magistereven got pa log :317:56
malmy assumption is firmware loading issue17:58
malyou need to show your device config repo again17:58
Mister_Magistersure17:58
Mister_Magisterhttps://github.com/verdanditeam/droid-config-Z01R let me update it17:58
Mister_Magistermal: here, updated :)18:01
malwhich android base?18:01
Mister_Magister1618:03
malyou seem to have some different set of .rc files than the faq mentions18:08
Mister_Magisteryeah because when i was porting it that entry didn't exist18:10
Mister_Magisteri was waaay ahead of time :P18:10
Mister_Magistermal: if you don't have any idea then i'm gonna move everything to droid-config repo and try clean image18:12
Mister_Magisterah right mal! new TWRPs don't have bzip2 so sfos installation fails. You can either build your own twrp or fix it. I was thinking… putting bz2 inside zip makes almost no sense so we could put just .tar and then unpack it using tar without bzip2 (gonna be faster even), or we can put bzip2 prebuild inside zip and then use it to unpack18:13
Mister_Magisterwondering what's your opinon18:13
malMister_Magister: first try the .rc files suggested in faq18:14
Mister_Magisteri need couple additional ones but i can remove some of my scripts sure18:14
malMister_Magister: your config submodule is quite old18:16
Mister_Magisterye whole port is :P18:16
malmissing some firmware loading fix for example because of that submodule18:17
Mister_Magisteroh18:17
Mister_Magisterinteresting18:17
Mister_Magisteri wanna move everything to obs anyway18:17
maland many audio related things, mostly audio routing but still18:17
Mister_Magisterlemme rebuild it real quick18:18
malso better update the whole thing, including dhd submodule (use upgrade-3.2.1 branch)18:18
maljust to be sure18:18
Mister_Magisteryes i want to move all my fixes from device to configs repo then i will push it to obs and then rebuild whole image freash okay?18:19
Mister_Magisterand what do you think about zip fix?18:19
malwhat zip fix?18:19
Mister_Magisterthe long message above18:20
malwell it might work that way18:21
Mister_Magisterbut which way would be better?18:21
Mister_Magisteri think dropping bzip218:22
Mister_Magisteror you have better idea?18:22
Mister_Magisterohh my configs fixed modem18:24
malwhat configs?18:26
Mister_Magisterdcd18:28
Mister_Magisteri updated it18:28
Mister_Magisterand now my ofono starts fine18:28
malso as usual updating submodule is a good idea18:29
Mister_Magisteryeah18:30
Mister_Magisterbut still no audio18:30
Mister_Magistertfa98xx_container_loaded(): Firmware init complete seems like it loads just fine18:31
Mister_Magisteryeah same thing. will debug it later18:32
malnew logs needed18:35
deathmistMister_Magister: no audio in general from pulseaudio or just calls that don't work?18:48
maldeathmist: no audio at all, logs suggested adsp failure but that was before the latest fixes so not sure what the logs say now18:50
deathmistI wonder if he looked at https://github.com/libhybris/libhybris/commit/d27c1a85703db8dea4539ceb4d869792fd78ee37 yet, that fixes all audio for me18:50
malmaybe not18:55
malI should finish some libhybris things I have been doing and then we could update the submodule18:55
KALUBEmal: Mister_Magister: just tested that udev fix and it works for me too! Bootctl is the last thing I needed to fix :D19:01
Mister_MagisterKALUBE: nice!19:04
Mister_Magistermal: you can assume nothing changed19:05
Mister_Magisterdeathmist: 2 dec is way too new so no i don't have that one. Thanks!19:05
Mister_Magisteri should add that udev fix to faq19:06
Mister_Magisternah it's too much to bother19:07
deathmistI even forked mer-hybris' libhybris and updated the submodule :p https://github.com/sailfishos-oneplus5/libhybris19:07
Mister_Magisterdeathmist: oh nice19:07
Mister_Magisterhmm when i was making it it was empty… now 83 repos… damn… https://github.com/verdanditeam19:16
KALUBEMister_Magister: I'm happy to make a PR for the faq if you want19:34
Mister_Magisterfeel free19:34
rinigusmal: when you'll be looking at libhybris, would you mind also to look into https://github.com/libhybris/libhybris/pull/433 ?19:39
malrinigus: I'll need to check that21:55
T42<edp_17> Good evening.22:05
T42<edp_17> I have a question before I continue my S7 attempt.22:05
T42<edp_17> I have applied these changes to my kernel (CM14.1) https://github.com/edp17/hero2lte_CONFIG_MODULES_patch22:06
T42<edp_17> Can be these the reason why the gui doesn't start?22:06
T42<edp_17> My other question is about the dhd sub-module. I had to update them for my hammerhead build and for dhd I needed to use the 'upgrade-3.2.1 branch'. Is this what I need for my S7 too or I can use the master?22:17
r0kk3rzuse the upgrade branch22:40
T42<edp_17> @r0kk3rz: Many thanks!22:52
r0kk3rzalso worth trying removing that patch22:55
r0kk3rzminimal changes is usually best for early stage porting22:55
T42<edp_17> This what I am also thinking about. I didn't find those in Driim's kernel, so probably I don't need (at this stage for sure). But when I found and it said was for loading modules, it was really tempting because I remember in hammerhead I needed to use the wifi driver as module otherwise it didn't remember the network.22:58
T42<edp_17> You are probably right, so I remove it and use the upgrade branch. Fingers crossed.22:59
T42<edp_17> Is there a way to remove a commit from the middle? Or I can only go back to the last right commit, skit the one I do not need and apply the rest?23:06
r0kk3rzat this stage the only mods you should do is the kernel config and fixup mountpoints23:13
r0kk3rzyou can do a rebase -i HEAD~6 and drop any commits you dont want23:14
r0kk3rzthe 6 is how many commits you want to go back23:14
T42<edp_17> Yeah, but I applied those commits at very early stage, so almost all defconfig and removing debug messages stuff were added after.23:15
T42<edp_17> Well, I have 9 commits and would like to keep the first 2, then skip 1 and keep the rest also. I don't think there is an easy way. I think the best if I go back to the last one prior that patch and manually apply the rest.23:16
r0kk3rzor23:17
r0kk3rzyou do what i said23:17
T42<edp_17> Yeah, I think there is no way to drop a commit somewhere in the middle but keep everything else. So, watch out, here I come. 😁23:20
T42<edp_17> Thanks.23:20
r0kk3rzthere is23:20
r0kk3rzlike i said23:20
r0kk3rzthe -i brings up a file, where you can tell it what to do23:21
r0kk3rzlike drop 2 commits in the middle, or squash, or rename, or whatever23:21
T42<edp_17> I see. Is the 6 includes the last commit or I need to count from there?23:22
r0kk3rzyou dont need to be super precise, go HEAD~11 if you want23:23
T42<edp_17> Ok, I did a git rebase -i HEAD~9 and dropped that commit. Now, how can I come out of that screen?23:26
r0kk3rzits vi, unless you changed the default editor23:27
r0kk3rzwhich i suggest you do if you dont know vi23:27
r0kk3rzso :wq23:27
T42<edp_17> No, not really.23:27
T42<edp_17> Ok, thanks.23:27
r0kk3rzyou can change it to nano if you want23:28
T42<edp_17> Thanks. Not now, because I go some conflicts to solve manually.23:28
« Monday, 2020-01-13 Index Wednesday, 2020-01-15 »

Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!