entil | gotta ask about the description "(fixed in latest repos)" - does that mean if someone wants to install on the fp5, it's actually better to get a testing/devel/something image off obs instead of the official download? or is this about adding a more experimental repository and installing packages from there? | 05:21 |
---|---|---|
entil | https://build.sailfishos.org/project/show/nemo:devel:hw:fairphone:fp5 or https://repo.sailfishos.org/obs/nemo:/testing:/hw:/fairphone:/fp5/sailfishos_5.0.0.68/ or what have you | 05:24 |
entil | am I screwed? got a good deal on an fp5, fastboot-flashed latest android 14, and sailfish, but then I got carried away and accidentally re-locked the bootloader, and the phone won't accept "fastboot flashing unlock" anymore? | 11:30 |
entil | `fastboot oem allow-flashing` -> FAILED (remote: 'unknown command') | 11:36 |
Pan[m] | <entil> "`fastboot oem allow-flashing` ->..." <- Use "fastboot flashing unlock" | 11:47 |
entil | fastboot flashing unlock | 11:48 |
entil | FAILED (remote: 'Flashing Unlock is not allowed | 11:48 |
entil | ') | 11:48 |
entil | fastboot: error: Command failed | 11:48 |
entil | that was the first thing I did | 11:49 |
entil | "ignore flashing ability and flashing regardless (FP5)" yeah did that too | 11:53 |
entil | like that's the one tip floating around, but why would editing the script ever lead to anything but "Flashing is not allowed in Lock State" from the device | 11:54 |
T42_ | <DerRenard> How does FP5 perform unlocking? Like nexus/pixel via fastboot only, or with specific code/application like xiaomi does? | 12:13 |
T42_ | <DerRenard> At least, if FP has something like EDL mode, you can try to flash it thru this. But maybe you'll need to have a service account to do this from EDL state | 12:14 |
entil | the initial unlocks were fastboot, as was the re-locking, and that put it in a state where it can't be unlocked | 12:43 |
entil | I tried for half an hour to get adb/fastboot/this thing https://xdaforums.com/t/guide-tool-reboot-to-edl-mode-from-fastboot-no-more-test-point-method-kenzo.3398718/ going on a windows machine, but I have no idea how to work it, the device wasn't found despite installing google android usb drivers | 12:44 |
mal | entil: did you follow this https://www.fairphone.com/en/bootloader-unlocking-code-for-fairphone/ | 12:44 |
entil | mal: sure, the issue came when I had flashed sfos and re-locked, the phone said it's been corrupted, and won't boot, and won't allow `fastboot flashing unlock` | 12:45 |
mal | hmm | 12:45 |
entil | to me this feels like when fairphone says "may prevent your Fairphone from booting correctly if a non-Fairphone operating system (OS) is installed." it's the understatement of the year | 12:45 |
mal | why would you try to relock bootloader when sfos is installed? | 12:46 |
entil | worst thing I would expect to happen is "oh, didn't boot, lemme unlock again" and it would work, not possibly an rma situation | 12:46 |
entil | mal: to get rid of the bootup nag | 12:46 |
entil | had no reason to think it would cause a situation this bad :D | 12:47 |
mal | so you can't flash anyting now? | 12:47 |
entil | seems that way, I tried hacking the flash script from fp themselves and it just rejects every image file the same way, so all that advice is useless too | 12:48 |
entil | recovery mode in the fastboot menu (on the phone) is also useless, as that apparently lives in the now-replaced android parts ... maybe? anyway, that recovery thing just restarts the bootloader eventually | 12:49 |
Pan[m] | entil: Maybe switch slots and see if there is still something there? | 12:51 |
entil | slots? what? how? | 12:52 |
Mister_Magister | bro doesn't know his phone is a/b slot | 12:53 |
entil | not allowed in locked state | 12:53 |
Mister_Magister | that makes sense | 12:54 |
Mister_Magister | but can you boot sfos? | 12:54 |
entil | no, that's the problem, it ends up in the fastboot menu, due to the "corruption" it talked about | 12:55 |
Pan[m] | entil: Oh, then you will probably have to send your phone to fairphone u.u | 12:55 |
entil | nothing would please me more than be able to tell it "ok, boot slot b" or whatever, but I don't see anything in the fastboot help to accomplish that, at least not as it stands now | 12:55 |
Mister_Magister | maybe try booting couple times and it will switch slots | 12:55 |
Mister_Magister | after 5 failed boot attemtps it should switch slots | 12:56 |
Pan[m] | entil: fastboot set_active other | 12:56 |
mal | pretty sure both slots have sfos boot image | 12:56 |
entil | tried rebooting without even a usb cable way more than 5 times just now, no visible change | 12:57 |
entil | Pan[m]: well, sure, but it says it's not allowed in lock mode | 12:57 |
entil | guess the only thing that could possibly help, unless the sfos flash script does things differently, is to instruct it to boot from somewhere else, but it's probably prohibited in lock mode if at all possible otherwise | 12:58 |
mal | entil: which fastboot version are you using? | 12:58 |
Mister_Magister | as they say "you done fucked up" | 12:58 |
entil | 34.0.5-debian | 12:58 |
entil | Mister_Magister: I'm upset at myself, sure, but I don't appreciate fairphone's documentation and I'll suggest they make the warning a bit more explicit about what can happen, not just "may not boot properly" | 12:59 |
Mister_Magister | on op6 you could use edl/firehose to reflash it, not sure if firehose is available for fairphone, not sure how fair it is | 13:00 |
entil | at least that windows binary from 2016 did sweet fuck all, but it could just be that messing around with windows, one that doesn't see the phone at all, went over my pain threshold | 13:00 |
entil | I could cook up that usb cable if the loader exists, but as the cables are apparently different for different phones, the last thing I want is electrical damage | 13:01 |
mal | wondering why it says 'unknown command' for "fastboot oem allow-flashing" some places say that should have worked | 13:03 |
Mister_Magister | op6 has button combination to enter edl | 13:03 |
mal | entil: does "fastboot reboot fastboot" change anything | 13:06 |
mal | wait | 13:06 |
entil | https://github.com/bkerler/edl/issues/477#issuecomment-1985305224 this does not sound great, though I'm not sure if I know how to read the table behind the link | 13:06 |
entil | mal: "Failed to boot into userspace fastboot; one or more components might be unbootable" - afaiui that's the "Recovery mode" in the fastboot menu, dependent on android | 13:07 |
mal | yeah | 13:07 |
mal | just in case try to grab latest fastboot from https://developer.android.com/tools/releases/platform-tools | 13:08 |
mal | and see if that allow-flashing would luckily work | 13:09 |
entil | "oem allow-flashing"? that was never a known command | 13:11 |
entil | "flashing unlock" still says "not allowed" | 13:11 |
mal | https://xdaforums.com/t/fairphone-5-unbrick-android-14-bootloader-only.4711932/ this talks about "fastboot oem allow-flashing" | 13:11 |
entil | mal: the last message is SOL | 13:14 |
entil | I wonder what my edl chances might be, "Your device needs to have a usb pid of 0x9008 in order to make the edl tool work" is this lsusb output stuff? | 13:15 |
entil | "entered the usb pid 0x900E" after that sounds more like a state than a product id | 13:17 |
mal | well try any way with latest platform tools, nothing really to lose | 13:18 |
entil | I did, that was 8 minutes ago, "still says 'not allowed'" | 13:19 |
entil | I'll see what happens if I run that edl tool, once I have it installed, but I'm bound to be even more useless with it than not | 13:19 |
mal | so allow-flashing is still unknown? | 13:20 |
entil | simple instructions like "1) strip usb cable 2) connect wires 3) plug in 4) run fix_things.sh" is fine and dandy, but risking the wrong short-circuit and/or something is danger zone | 13:20 |
entil | mal: yeah, and it's probably not up to fastboot, but the phone, like I can have whatever version of fastboot ever but if the phone doesn't expose that command, it's not there - the success was iirc said for fp4/a13 or something else than this | 13:20 |
entil | edl has these instructions like "Xiaomi: Press and hold vol dwn + pwr, in fastboot mode connect usb." - Run "./fastpwn oem edl". | 13:23 |
entil | probably refers to a xiaomi phone, not this chip in lsusb output, fastpwn oem edl says Protocol error | 13:24 |
entil | "Press and hold vol up+dwn, connect usb. For some, only use vol up." also did nothing, but I'll have another shot at it; even better would be to verify this thing can ever actually work | 13:24 |
mal | I asked one fairphone employee about that, we'll see when I'll get and answer | 13:25 |
mal | *an answer | 13:25 |
entil | ooh nice of you, thank you! | 13:27 |
entil | I held vol-up and -down pressed for half a minute or something, screen went black, it may actually have entered some secret mode, but edl with no loader and fastpwn do nothing, so it might also have crapped itself | 13:28 |
entil | hmm, this is interesting, after having those vol buttons pressed, fastboot itself is saying protocol error, I do wonder if I managed to put it in an edl mode after all, but the edl script doesn't find the phone | 13:43 |
entil | hope I didn't make it worse | 13:43 |
entil | mal: not asking you to snitch on your friend, but is this a person who might know about these matters and maybe even get back despite the weekend? I'm sorta pressured to do something else, and with this "Protocol Error" I'm fresh out of ideas, so I disconnected the phone and powered it down... Reason I'm not asking about snitching is if I should mail fairphone officially as well or wait a while? | 13:55 |
rinigus | I had ofono crashing with IMS when there is a second call incoming while I was on the call. with debug, I traced it down to handling of calls array in binder_voicecall_ext_calls_changed | 15:18 |
rinigus | ref: https://github.com/mer-hybris/ofono-binder-plugin/blob/c8e3b0389f94de9e1d7b9824123f1839e636518a/src/binder_voicecall.c#L2081 | 15:18 |
rinigus | it assumes that I am providing a pointer with array data that is also NULL terminated. this is possible, but not like that by default. just have to use g_ptr_array_new_null_terminated on array creation. | 15:18 |
rinigus | fix: https://github.com/rinigus/ofono-binder-plugin-ext-qti/commit/f78df8577aff673380ba50aa75f53ad78e2f1e46 | 15:18 |
rinigus | looking at the code, I think hidl version and one used by mtk should get similar fix. | 15:18 |
entil | left the phone turned off and got this feeling I should see if something's magickally happened, let it try to boot five times out of superstition first, and the only change is I'm not getting that protocol error anymore, only "Flashing Unlock is not allowed" | 15:39 |
mal | entil: should know things | 15:39 |
entil | mal: very nice, and I calm myself knowing that I'm in no real hurry to get this, except my profound and growing hatred and despisal for the 10iv, but I'm "fine" in the phone department | 15:40 |
mal | entil: which are the main issues for you on 10iv? | 15:41 |
entil | and now I don't need to worry my even-more-half-assed attempts with EDL didn't cause some permanent mode of "Protocol Error", but that it's as finely bricked | 15:41 |
entil | mal: charges slow as balls, reboots after disconnecting any and all chargers I've used, no camera support, bluetooth is touch-and-go, and it can't decide if it wants to work over mobile internet or only tethered to my work phone, and sometimes it struggles when switching to my home wifi | 15:42 |
entil | mal: epäilyttävästi puhelinta muistuttava kasa paskaa | 15:42 |
mal | hehe | 15:42 |
entil | oh yeah, and the need for bluetooth because of the call echoing | 15:43 |
mal | about the usb disconnect issue, how old android base do you have on it, I mean how long ago did you flash sfos to it, I think on x10iv or v flashing latest android and the sfos back fixed such issues, might be some firmware bug in older android versions or something | 15:44 |
mal | is the slow charging always or only on some type of chargers | 15:45 |
entil | not snitching on jolla people but first I heard "thanks for the logs, we'll look into it!" and later "oh yeah that's a tough one, I switched to c2 because of it, I passed the word onward" - I don't care :P android appsupport says android 13 api level 33, not sure what we can deduce from that, I can't remember when I took this into use, but about as soon as sfos was released | 15:46 |
mal | which logs in this case? | 15:46 |
entil | my gut is it's always slow, but I can only prove it by using this sick anker I got that can push and pull electricity like nothing I've seen, and it charges the phone at 2.4W | 15:46 |
mal | https://github.com/sonyxperiadev/kernel/commit/d1bdd007f35a7cf0f882235e2d9c5e77f0b5a420 | 15:47 |
entil | mal: I'd have to dig around, I'm beyond caring at this point, I trust your sfos distro way more than anything jolla offers for the 10iv | 15:47 |
mal | need to check if that has made it to 5.0 yet | 15:47 |
mal | well I have developed both :) | 15:47 |
mal | it's just that android base quality differs quite a lot sometimes | 15:48 |
entil | I don't doubt it and thank you for your service, but I will not risk flashing a new android onto a working phone and messing it up, and sure not gonna reinstall sfos and play the backup-restore game | 15:48 |
mal | looks like that usb fix is not in 5.0 yet | 15:50 |
entil | the only good thing I can think of to say about the 10iv is that, especially with appsupport turned off and no browser left open, it has pretty solid battery life | 15:50 |
entil | for the fp5, not considering watching videos or any of that super important, I verified all my android needs are served by my work phone, and I'll sell the 10iv to my company and ... unfortunately I need google android on it anyway ... so I'm set, the fp5 could also be fast enough that I can spend idle time running imgorganizer way better than on the 10iv | 15:52 |
mal | if you sent logs about the usb reboot issue, those haven't revealed anything | 15:52 |
entil | but just want to make phone calls without the handsfree and take a photo or two without using my work phone and sending it over signal | 15:52 |
mal | so is that x10iv now unused for you? just thinking if you want to test flashing back latest official android to it? | 15:53 |
mal | and see if the usb issue disappears | 15:53 |
entil | it was /sys/fs/pstore | 15:56 |
entil | the 10iv will be available for d*cking about when the fp5 is in "production use" | 15:56 |
entil | right now I'm spooked enough that I'll wait for your fairphone buddy to say something, rma the phone if needed, and then I can reinstall the 10iv for funsies, but it will unfortunately end up a google android at some point | 15:58 |
mal | yeah, pstore didn't reveal anything in that usb reboot issue | 15:58 |
entil | apropos nothing, I wonder what jolla's licensing for aas is as they can distribute it for free now for these phones | 15:59 |
mal | it's our product | 15:59 |
entil | it was said they can't sell it, but if they can distribute it for free for an installation that's f.. oh :o | 16:00 |
entil | I thought it was licensed | 16:00 |
entil | well sell it for the fp5 goddamnit | 16:00 |
entil | no one seems to care about waydroid anyway | 16:00 |
entil | like, please, what do you want for it, 50 euro? | 16:01 |
T42_ | <elros34> why is so important jola sell appsupport for fp5, can't you buy version for xperia similar enough to fp5 and install it? | 16:19 |
entil | afaik you can't install it on unofficial distros | 16:31 |
entil | neither technically nor legally | 16:32 |
T42_ | <Mister_Magister> you can technically | 16:32 |
T42_ | <Mister_Magister> aliendalvik has been run on many unofficial devices | 16:32 |
entil | fine, don't take my money, then ;P | 16:34 |
entil | but I don't know what the market is for voluntary subscriptions wrt community ports, at least it's a bit of a shady area because it's the volunteers doing the work but there usually isn't a good way to fund that work, and without funding jolla, there's nothing to port - but if aas is a commercial product in its own right, there could be a market opportunity for jolla to tap into money from | 16:36 |
entil | community ports as well | 16:36 |
entil | somewhat off-topic, just know that I'd gladly pay jolla for aas service but if it's somewhat kosher to just install it, I can do that too, and buy a license out of guilt if it ever becomes a possibility | 16:37 |
entil | mal: oh yeah, about that android version, checked just unfocused enough that I missed it also says "Security patch level 8/5/24" so maybe that can be a clue as to how ancient my base is | 16:40 |
rinigus | what is an expected support for ims calls in terms of putting them on hold or swapping between active and waiting calls? | 17:40 |
rinigus | does jolla's ims plugin allow to swap between calls and putting calls on hold? | 17:41 |
mal | entil: maybe fastboot getvar all might tell something | 17:45 |
T42_ | <DerRenard> Hello. Yesterday I got stuck after renaming (regenerating) all Mi439 to mi439. Now I'm stuck at the point of final building (or packing?) | 18:24 |
T42_ | <DerRenard> Here's the log. <mal> suggested me to make one new variable in the ~/.hadk .env with HABUILD_DEVICE = Mi439 and redefine DEVICE with mi439. Here is my log during rpm packing (final stage before getting flashable zip) | 18:24 |
T42_ | <DerRenard> https://pastebin.com/uWhd1Esd | 18:24 |
mal | you can see the error about droid-hal-mi439-devel installation failing, install it manually | 18:29 |
T42_ | <DerRenard> Can I get rid of all of those? | 18:37 |
T42_ | <DerRenard> ``` | 18:37 |
T42_ | <DerRenard> conflicts with file from package | 18:37 |
T42_ | <DerRenard> droid-hal-Mi439-devel-0.0.6-202509262301.aarch64 (@System) | 18:37 |
T42_ | <DerRenard> ``` | 18:37 |
T42_ | <DerRenard> | 18:37 |
T42_ | <DerRenard> I mean remove all packages with capital M in them? | 18:37 |
mal | yes | 18:39 |
T42_ | <DerRenard> Stupid question - how? | 18:39 |
mal | using the sb2 command | 18:40 |
mal | easiest way is to use zypper via sb2 as I showed earlier and the zypper command at the end being "zypper in droid-local-repo/droid-hal-mi439-devel-0*.rpm -droid-hal-Mi439-devel" | 18:41 |
mal | oops | 18:41 |
mal | "zypper in droid-local-repo/$DEVICE/droid-hal-mi439-devel-0*.rpm -droid-hal-Mi439-devel" | 18:41 |
T42_ | <DerRenard> I created this command. Will this work as intended? | 18:42 |
T42_ | <DerRenard> ``` | 18:42 |
T42_ | <DerRenard> sb2 -t $VENDOR-$DEVICE-$PORT_ARCH -R -m sdk-install zypper rm droid-config-Mi439-1*``` | 18:42 |
T42_ | <DerRenard> I did this earlier. This one command | 18:42 |
T42_ | <DerRenard> ``` | 18:42 |
T42_ | <DerRenard> sb2 -t $VENDOR-$DEVICE-$PORT_ARCH -R -m sdk-install zypper in --force droid-local-repo/$DEVICE/droid-config-mi439-1*.rpm``` | 18:43 |
T42_ | <DerRenard> (re @SailfishFreenodeIRCBridgeBot: <mal>"zypper in droi...) | 18:43 |
mal | yes, replace the zypper part of that with the one I just gave | 18:43 |
T42_ | <DerRenard> Okay, done. Start the process again? | 18:46 |
mal | maybe check "sb2 -t $VENDOR-$DEVICE-$PORT_ARCH -R -m sdk-install zypper se droid-hal-Mi439" | 18:49 |
mal | and see if it reports any other packages installed | 18:49 |
T42_ | <DerRenard> ``` | 18:50 |
T42_ | <DerRenard> S | Name | Summary | Type | 18:50 |
T42_ | <DerRenard> ---+--------------------------+-----------------------------------------------+-------- | 18:51 |
T42_ | <DerRenard> i | droid-hal-Mi439 | Droid HAL package for Mi439 | package | 18:51 |
T42_ | <DerRenard> i | droid-hal-Mi439-detritus | Straggler files for Mi439 hw | package | 18:51 |
T42_ | <DerRenard> i+ | droid-hal-mi439-devel | Development files for droid-hal device: mi439 | package | 18:51 |
T42_ | <DerRenard> i | droid-hal-Mi439-users | Create users and groups for device: Mi439 | package``` | 18:51 |
T42_ | <DerRenard> (re @SailfishFreenodeIRCBridgeBot: <mal>and see if it r...) | 18:51 |
T42_ | <DerRenard> https://pastebin.com/TSEDPNJm | 18:51 |
mal | so do the same for those others, do you understand how? | 18:51 |
T42_ | <DerRenard> yep (re @SailfishFreenodeIRCBridgeBot: <mal>so do the same ...) | 18:51 |
T42_ | <DerRenard> What? https://pastebin.com/K61AM8Np | 18:53 |
T42_ | <DerRenard> Uh... It scares me | 18:53 |
mal | answer 1 | 18:58 |
mal | it will get installed back later | 18:58 |
mal | that droid-config package | 18:58 |
mal | wait | 18:59 |
mal | you have wrong command | 18:59 |
mal | the package names in the zypper command need to match | 18:59 |
mal | you are trying to install droid-hal-mi439-devel-0*.rpm but remove droid-hal-Mi439-detritus | 18:59 |
T42_ | <DerRenard> I changed it but it shows the same (re @SailfishFreenodeIRCBridgeBot: <mal>you are trying ...) | 19:01 |
mal | don't be scared if it removes something | 19:02 |
T42_ | <DerRenard> fine then | 19:02 |
T42_ | <DerRenard> okay, i removed everything contained capital letter in it | 19:08 |
T42_ | <DerRenard> I started to get this error | 19:20 |
T42_ | <DerRenard> ``` | 19:20 |
T42_ | <DerRenard> PlatformSDK renard@sfosBUILD:~/hadk$ rpm/dhd/helpers/build_packages.sh --version | 19:20 |
T42_ | <DerRenard> dirname: missing operand | 19:20 |
T42_ | <DerRenard> Try 'dirname --help' for more information. | 19:20 |
T42_ | <DerRenard> * Building rpm/droid-hal-version-mi439.spec | 19:20 |
T42_ | <DerRenard> ./rpm/dhd/helpers/util.sh: line 341: //droid-hal-version-mi439.log: Permission denied | 19:20 |
T42_ | <DerRenard> !! building of package failed | 19:20 |
T42_ | <DerRenard> ``` | 19:20 |
T42_ | <DerRenard> | 19:20 |
T42_ | <DerRenard> What's wrong? | 19:20 |
mal | does "echo $ANDROID_ROOT" print the correct path? | 19:22 |
T42_ | <DerRenard> yes | 19:24 |
T42_ | <DerRenard> ``` | 19:24 |
T42_ | <DerRenard> PlatformSDK renard@sfosBUILD:~/hadk$ echo $ANDROID_ROOT/ | 19:24 |
T42_ | <DerRenard> /home/renard/hadk/``` | 19:24 |
mal | wait, did you create that droid-hal-version package as instructed in chapter 7.1? | 19:24 |
T42_ | <DerRenard> yes | 19:24 |
mal | in which folder do you have it? | 19:24 |
mal | it should be in hybris/droid-hal-version-$DEVICE | 19:25 |
T42_ | <DerRenard> Ah. It has older name with capital leter | 19:26 |
T42_ | <DerRenard> Ah. It has older name with capital letter (edited) | 19:26 |
mal | so fix the folder and spec (both filename and the content) | 19:27 |
T42_ | <DerRenard> Done (re @SailfishFreenodeIRCBridgeBot: <mal>so fix the fold...) | 19:29 |
T42_ | <DerRenard> same error (re @DerRenard: I started to get thi...) | 19:36 |
T42_ | <DerRenard> after I create the right directory, files in it and rerun all prev commands | 19:37 |
mal | did you fix the content of droid-hal-version spec= | 19:39 |
T42_ | <DerRenard> Uh? (re @SailfishFreenodeIRCBridgeBot: <mal>did you fix the...) | 19:40 |
T42_ | <DerRenard> Oh! It's done | 19:41 |
T42_ | <DerRenard> ``` | 19:41 |
T42_ | <DerRenard> PlatformSDK renard@sfosBUILD:~/hadk$ rpm/dhd/helpers/build_packages.sh --version | 19:41 |
T42_ | <DerRenard> * Building rpm/droid-hal-version-mi439.spec | 19:41 |
T42_ | <DerRenard> * Building of droid-hal-version-mi439 finished successfully | 19:41 |
T42_ | <DerRenard> ----------------------DONE! Now proceed on creating the rootfs------------------``` | 19:41 |
T42_ | <DerRenard> it looks like i forget about change the name of the spec file and left it like `...-hammerhead.spec`... | 19:42 |
mal | yeah, better fix those now before you forget | 19:42 |
T42_ | <DerRenard> I wonder why script isn't reporting what exactly goes wrong. If it hasn't located the file report about it. There's a lot of places where something could went wrong | 19:46 |
mal | error handling in the script is not the best | 19:48 |
T42_ | <DerRenard> I remember how I was shocked about better error handling in clang compared to gcc. Gcc is chaotic, clang is more structured and informative (re @SailfishFreenodeIRCBridgeBot: <mal>error handling ...) | 19:51 |
T42_ | <DerRenard> I got zip! Yes. Now how do i need to flash it? Can i use LOS recovery? (my phone has a lack of TWRP support, i used to use LOS recovery for flashing) | 19:53 |
T42_ | <DerRenard> And when i was flashing LOS, I flashed dtbo, vbmeta before the recovery. Do i need them? | 19:55 |
mal | make sure dtbo matches wwhat you built | 20:06 |
T42_ | <DerRenard> what? (re @SailfishFreenodeIRCBridgeBot: <mal>make sure dtbo ...) | 20:06 |
T42_ | <DerRenard> I have downloaded a lineageos 21 (which base i used) and installed it on the phone for all this time i was building sfos. Where located all android-base files from HABUILD i dont know, so i cant took them | 20:08 |
mal | well you have to flash correct android version first | 20:08 |
T42_ | <DerRenard> It's already been installed (all - dtbo,vbmeta,recovery, los 21) (re @SailfishFreenodeIRCBridgeBot: <mal>well you have t...) | 20:08 |
T42_ | <DerRenard> So just flash sfos zip in the recovery? | 20:09 |
mal | yes, if it works | 20:10 |
T42_ | <DerRenard> It report "failed to extract filesystem"... | 20:13 |
T42_ | <DerRenard> and something about userdata if i've seen it correctly | 20:14 |
T42_ | <DerRenard> Tried to boot hybris-boot through fastboot. Fastboot didn't turned off the screen, but command shows no devices connected. Is it okay when booting hybris-boot from out/target/... ? | 20:21 |
mal | it should be fine, not all devices support "fastboot boot" | 20:24 |
T42_ | <DerRenard> But what's wrong with my installation? Why it cannot install sfos zip? | 20:25 |
T42_ | <DerRenard> To be more precise, it failed when tried to mount `/dev/block/bootdevice/by-name/userdata at /data` | 20:27 |
T42_ | <DerRenard> Oh. More info. It report | 20:29 |
T42_ | <DerRenard> ERROR recovery : footer is wrong | 20:29 |
T42_ | <DerRenard> ERROR recovery : signature verification fasiled | 20:29 |
T42_ | <DerRenard> I'm using LOS recovery from my firmware (los 21) | 20:29 |
mal | try to format userdata first to ext4 using fastboot | 20:30 |
T42_ | <DerRenard> What? How? (re @SailfishFreenodeIRCBridgeBot: <mal>try to format u...) | 20:30 |
mal | or via recovery is possible | 20:30 |
mal | fastboot format:ext4 userdata | 20:31 |
T42_ | <DerRenard> This can be reverted? (by flashing stock miui or los) | 20:32 |
mal | yes, if you flash back android it should make it whatever android wants | 20:33 |
T42_ | <DerRenard> uhhh | 20:33 |
T42_ | <DerRenard> ``` | 20:33 |
T42_ | <DerRenard> /usr/lib/android-sdk/platform-tools/mke2fs failed with status 1 | 20:33 |
T42_ | <DerRenard> fastboot: error: Cannot generate image for userdata``` | 20:34 |
T42_ | <DerRenard> what is wrong? | 20:34 |
T42_ | <roundedrectangle> so, quick update: removing the default.prop part in droid-hal-device helped me with my leeco s2 port and i got further. but, when publishing the port, how should i publish these changes? do i need to fork the droid-hal-device and change the submodule to point to my fork? | 20:38 |
mal | @roundedrectangle which android version do you use as base? | 20:45 |
T42_ | <roundedrectangle> lineage 17.1, android 10 | 20:45 |
mal | @DerRenard probably too old fastboot, try download new one from https://developer.android.com/tools/releases/platform-tools | 20:46 |
mal | @roundedrectangle so even the android build doesn't have default.prop, you could make a pull request to check existence of default.prop in https://github.com/mer-hybris/droid-hal-device/blob/master/droid-hal-device.inc#L692C7-L692C20 | 20:47 |
T42_ | <DerRenard> I found SHRP recovery and it says it has support of A14, so i'll try this first (re @SailfishFreenodeIRCBridgeBot: <mal>@DerRenard prob...) | 20:48 |
T42_ | <roundedrectangle> ok thank you | 20:49 |
T42_ | <DerRenard> how can i format userdata from recovery' terminal? | 20:53 |
T42_ | <DerRenard> same error (re @DerRenard: To be more precise, ...) | 21:19 |
T42_ | <DerRenard> i can not flash it through lineage recovery even after manually format userdata to ext4 | 21:20 |
T42_ | <DerRenard> SHRP does bootloop if i flash SFOS thru it | 21:20 |
T42_ | <DerRenard> any ideas what can i do now? (re @SailfishFreenodeIRCBridgeBot: <mal>yes, if you fla...) | 21:20 |
T42_ | <elros34> so you boot to recovery, format userdata and then directly install sailfish? Show whole log from recovery | 21:28 |
T42_ | <DerRenard> Userdata i format from fastboot, after that i boot in los recovery (i dont know how to show logs from it) and flash zip via sideload (re @elros34: so you boot to recov...) | 21:30 |
T42_ | <DerRenard> Any twrp/shrp does bootloop | 21:33 |
T42_ | <DerRenard> Even on clear los 21 | 21:33 |
T42_ | <elros34> you dont have format data function in recovery? adb shell and then pull log | 21:33 |
T42_ | <DerRenard> It doesn't support it (re @elros34: you dont have format...) | 21:34 |
T42_ | <DerRenard> ADB is available only through sideloading | 21:34 |
T42_ | <DerRenard> It doesn't support to pull logs out of it | 21:34 |
T42_ | <DerRenard> This recovery allows you only to install an update or perform a factory reset | 21:35 |
T42_ | <DerRenard> This is a nice for installing magisk, so it did work for me. But now I cannot do anything with it to install sfos | 21:36 |
T42_ | <DerRenard> As far as I know only way to show you logs is send a picture but I can't send picture in this chat | 21:37 |
T42_ | <elros34> dead end. | 21:37 |
T42_ | <elros34> :) so you have only 2 lines of error? | 21:38 |
T42_ | <DerRenard> No, but this is most relevant part of it (re @elros34: :) so you have only ...) | 21:39 |
T42_ | <DerRenard> I can send photo to dm if you want to see it | 21:39 |
T42_ | <elros34> upload it somewhere and I can't believe you you have no adb pull/shell support, it's disastear | 21:40 |
T42_ | <DerRenard> wait. i'll google is los recovery support it. maybe i didn't saw this feature before | 21:41 |
T42_ | <DerRenard> oh, it has this feature! | 21:42 |
T42_ | <DerRenard> how to pull logs? | 21:42 |
T42_ | <elros34> adb pull | 21:42 |
T42_ | <DerRenard> `adb: pull requires an argument` (re @elros34: adb pull) | 21:42 |
T42_ | <elros34> path to log whatever it is either you see it on screen or find in lineage docs | 21:43 |
T42_ | <elros34> guess /tmp/recovery.log | 21:43 |
T42_ | <DerRenard> https://pastebin.com/d3HRmjw8 | 21:45 |
T42_ | <DerRenard> size of log is... insane | 21:45 |
T42_ | <DerRenard> compare to what i have seen on screen | 21:45 |
T42_ | <DerRenard> here (re @elros34: path to log whatever...) | 21:47 |
T42_ | <elros34> totally unexpected: tar: exec bzip2: No such file or directory | 21:50 |
T42_ | <DerRenard> what? ahahahah (re @elros34: totally unexpected: ...) | 21:50 |
T42_ | <DerRenard> how | 21:50 |
T42_ | <DerRenard> this even possible ;D | 21:50 |
T42_ | <DerRenard> linux without bzip2? haha | 21:50 |
T42_ | <elros34> https://sailfishos.wiki/link/20#bkmrk-%C2%A0%C2%A0%C2%A0%C2%A0%C2%A0%C2%A0%C2%A0-%2A-twrp-is-kn | 21:51 |
T42_ | <DerRenard> can it be installed without any recovery? | 21:53 |
T42_ | <DerRenard> just using fastboot? | 21:53 |
T42_ | <elros34> not zip image you created | 21:54 |
T42_ | <elros34> hybris-boot is not strict recovery so I would say yes it can be installed without recovery | 21:54 |
T42_ | <DerRenard> fine... any ideas how? | 21:55 |
T42_ | <elros34> copy via los recovery image somewhere sdcard or some partition except data so it will be not lost and flash hybris-boot | 21:57 |
T42_ | <elros34> probably you can unpack zip on pc and copy only tarball, it will be faster | 21:57 |
T42_ | <DerRenard> it is present even without unpacking (re @elros34: probably you can unp...) | 21:59 |
T42_ | <DerRenard> in the output folder nearby zip | 21:59 |
T42_ | <DerRenard> but where to place it? | 21:59 |
T42_ | <elros34> like I said, sdcard maybe /cache or even /data if it's really formated without encryption | 22:01 |
T42_ | <DerRenard> pushed | 22:02 |
T42_ | <DerRenard> to cache | 22:02 |
T42_ | <DerRenard> what next | 22:02 |
T42_ | <DerRenard> reboot to fastboot or stay in recovery? (re @elros34: like I said, sdcard ...) | 22:05 |
T42_ | <elros34> just flash normal way via fastboot | 22:06 |
T42_ | <DerRenard> what to flash? hybris-boot? (re @elros34: just flash normal wa...) | 22:06 |
T42_ | <elros34> yes still same hybris-boot | 22:07 |
T42_ | <DerRenard> flased | 22:07 |
T42_ | <DerRenard> flashed (edited) | 22:07 |
T42_ | <elros34> telnet to it at 23 port | 22:09 |
T42_ | <DerRenard> ah? (re @elros34: telnet to it at 23 p...) | 22:09 |
T42_ | <DerRenard> how/ | 22:10 |
T42_ | <DerRenard> how? (edited) | 22:10 |
T42_ | <elros34> it's all explained in hadk and hadk-hot | 22:10 |
T42_ | <DerRenard> I found. I need to leave fastboot and start telnet, right? (re @elros34: it's all explained i...) | 22:12 |
T42_ | <elros34> no you need to just power on device so it will boot hybris-boot and start telnet | 22:13 |
T42_ | <DerRenard> i did this, but telnet cannot connect to anything (re @elros34: no you need to just ...) | 22:13 |
T42_ | <elros34> be more verbose | 22:14 |
T42_ | <DerRenard> I flashed hybris-boot image | 22:15 |
T42_ | <DerRenard> Did fastboot reboot | 22:15 |
T42_ | <DerRenard> enter `telnet 192.168.2.15 2323` | 22:15 |
T42_ | <DerRenard> Nothing is connected (re @elros34: be more verbose) | 22:15 |
T42_ | <DerRenard> also tried simple 23 at the end | 22:15 |
T42_ | <DerRenard> nothing | 22:15 |
T42_ | <elros34> did you follow all steps from hadk about operating blindly on device? Does dmesg shows your device? | 22:15 |
T42_ | <DerRenard> yes, dmesg shows it, but it disconnects after rime (re @elros34: did you follow all s...) | 22:16 |
T42_ | <DerRenard> yes, dmesg shows it, but it disconnects after time (edited) (re @elros34: did you follow all s...) | 22:16 |
T42_ | <elros34> what time what does it shows. Without detailed info I can't guess what is wrong. Did you read also hadk-hot it has all common issues with telnet | 22:18 |
T42_ | <DerRenard> wait. it's old ones. no, it doesn't shows anything connected device during boot (re @elros34: did you follow all s...) | 22:18 |
T42_ | <DerRenard> no, dmesg see nothing during boot... why... i flashed the correct hybris-boot, but nothing. how | 22:23 |
T42_ | <DerRenard> output for iSerial | 22:24 |
T42_ | <DerRenard> ``` | 22:24 |
T42_ | <DerRenard> iSerial 1 0000:00:14.0 | 22:24 |
T42_ | <DerRenard> iSerial 0 | 22:24 |
T42_ | <DerRenard> iSerial 0 | 22:24 |
T42_ | <DerRenard> iSerial 3 E0D55E6CE776E720384907E0 | 22:24 |
T42_ | <DerRenard> iSerial 1 0000:00:14.0``` | 22:24 |
T42_ | <elros34> doesn't tell me anything so I will ask again did you read hadk-hot? poweroff device, flash hybris-boot, run dmesg -w on pc, power on device and wait at least 1 minute then show log | 22:28 |
T42_ | <DerRenard> yes (re @elros34: doesn't tell me anyt...) | 22:32 |
T42_ | <DerRenard> all of this | 22:32 |
T42_ | <DerRenard> not working for me at all | 22:32 |
T42_ | <DerRenard> I wait in around 5 mins in total | 22:33 |
T42_ | <elros34> so you have no single line of log and no device reboot? | 22:39 |
T42_ | <DerRenard> yes (re @elros34: so you have no singl...) | 22:40 |
T42_ | <DerRenard> at all | 22:40 |
T42_ | <elros34> so if you really followed my steps then its much harder issue. Either device hangs very early in init-script or boot image is incorrect. YOu could compare lineage boot one and hybris-boot.img with IIRC unpack_bootimg | 22:45 |
T42_ | <DerRenard> Yes, i followed your directions strictly. In fastboot it shows in dmesg, in booting - nothing (re @elros34: so if you really fol...) | 22:47 |
T42_ | <elros34> which device repo excatly do you use? | 22:47 |
T42_ | <DerRenard> uh? kernel or what? (re @elros34: which device repo ex...) | 22:48 |
T42_ | <elros34> android_device_* | 22:48 |
T42_ | <DerRenard> ``` | 22:49 |
T42_ | <DerRenard> <project path="device/xiaomi/Mi439" | 22:49 |
T42_ | <DerRenard> name="LineageOS/android_device_xiaomi_Mi439" | 22:49 |
T42_ | <DerRenard> revision="lineage-21" /> | 22:49 |
T42_ | <DerRenard> | 22:49 |
T42_ | <DerRenard> <project path="kernel/xiaomi/msm8937" | 22:49 |
T42_ | <DerRenard> name="LineageOS/android_kernel_xiaomi_msm8937" | 22:49 |
T42_ | <DerRenard> revision="lineage-21" /> | 22:49 |
T42_ | <DerRenard> | 22:49 |
T42_ | <DerRenard> <project path="device/xiaomi/mithorium-common" | 22:49 |
T42_ | <DerRenard> name="LineageOS/android_device_xiaomi_mithorium-common" | 22:49 |
T42_ | <DerRenard> revision="lineage-21" /> | 22:50 |
T42_ | <DerRenard> | 22:50 |
T42_ | <DerRenard> <project path="vendor/xiaomi/Mi439" | 22:50 |
T42_ | <DerRenard> name="TheMuppets/proprietary_vendor_xiaomi_Mi439" | 22:50 |
T42_ | <DerRenard> revision="lineage-21" /> | 22:50 |
T42_ | <DerRenard> | 22:50 |
T42_ | <DerRenard> <project path="vendor/xiaomi/mithorium-common" | 22:50 |
T42_ | <DerRenard> name="TheMuppets/proprietary_vendor_xiaomi_mithorium-common" | 22:50 |
T42_ | <DerRenard> revision="lineage-21" />``` | 22:50 |
T42_ | <DerRenard> (re @elros34: android_device_*) | 22:50 |
T42_ | <DerRenard> local manifest | 22:50 |
T42_ | <elros34> from my limited understanding you device doesn't use some latest features in boot image BOARD_BOOTIMG_HEADER_VERSION := 1 still comparision with lineage boot could be usefull | 22:58 |
T42_ | <DerRenard> Okay. What should i do next? And what can i do? If i cannot flash it, cannot see any logs, cannot do anything but wondering... it seems impossible to do anything (re @elros34: from my limited unde...) | 23:01 |
T42_ | <elros34> still same I told you compare lineage boot.img and sailfish hybris-boot.img with split_bootimg | 23:02 |
T42_ | <elros34> you can still do a lot of things like building boot image without defconfig changes modyfing init-script and maybe other | 23:03 |
T42_ | <DerRenard> What shall i looking for? (re @elros34: still same I told yo...) | 23:04 |
T42_ | <DerRenard> Wow. to make any changes i need to know what to change (re @elros34: you can still do a l...) | 23:04 |
T42_ | <DerRenard> i don't understand what is happening right now | 23:04 |
T42_ | <DerRenard> i can't flash ready to use zip. wtf | 23:05 |
T42_ | <elros34> I will not write wall of text with all possible steps. Focus on one thing; make sure hybris-boot.img is correct by comparing whether it has same parameters as lineage one | 23:05 |
T42_ | <elros34> if it is correct then you can try other things | 23:05 |
T42_ | <elros34> forget about zip now. Even if you could flash it it will not help, you can't boot simple init script now | 23:07 |
T42_ | <DerRenard> So, i need to look thru boot config in los boot and compare it to hybris one? Okay, i can do this (if I'd know where to look for this config files), but i will understand nothing. | 23:27 |
T42_ | <DerRenard> | 23:27 |
T42_ | <DerRenard> All i understand after this whole process that I understand nothing. I changed some file here, some files there. I don't understand where this files needed, how the building is functioning, what is happening inside. I'm green in it. Remember one thing, i read only hadk, and did compile of automatic lineage to my device. This ALL of mine experience. (Ah, and did some flashing before with precompiled zip firmware). | 23:27 |
T42_ | <DerRenard> | 23:27 |
T42_ | <DerRenard> A lot of software you told me to use I met here for the first time. | 23:27 |
T42_ | <DerRenard> | 23:27 |
T42_ | <DerRenard> Yes, I had a lot of experience in manipulating linux, but neither in development nor building kernels/something else. I am regular user who used to use terminal for simple task and can do some react webpage, that's all. | 23:27 |
T42_ | <DerRenard> | 23:27 |
T42_ | <DerRenard> Aren't there any more newbie resources where whole process of debugging this shit will described more understandable (also notice I'm not English native, this also adds a layer of understandability) | 23:27 |
T42_ | <DerRenard> | 23:27 |
T42_ | <DerRenard> I thought that if i got any compiled zip, i could just install it and see at least gui. Maybe with not working net/touch/whatever, but working system. What i got? Black undebugable screen without any understanding what is going on. Wow. Cool result... I have built sfos... But got no luck to try it (re @elros34: I will not write wal...) | 23:27 |
T42_ | <elros34> split_bootimg is a tool probably somewhere in your android sources. You run it with boot image as parameter. If you have problem with compaing both linea and sfos boot image outputs you can just upload it here nothing scary. I am also not native ENglish speaker and I am not verbose type of pearson becausefor common issue I expect porter to really try figure out details itself first (search in irc logs which almost always have excatly the s | 23:36 |
T42_ | <elros34> sailfish porting docs are not easy and are notoriously outdated. It is what it is with limited resources jolla have, I update hadk-hot from time to time to ease porting just a little but of course is not enough | 23:41 |
T42_ | <elros34> Normally zip should give at least some logs and access but some devices are more picky. I also do not know why you have no sign of life from booting kernel but based on experience documented in this very channel first thing to check whether your generated image is correct | 23:44 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!