echo_salik | after resinstalling target. I get this: /lib/ld-linux-armhf.so.3: No such file or directory when i run: sb2 -t $VENDOR-$DEVICE-$PORT_ARCH gcc main.c -o test && ./test | 02:20 |
---|---|---|
echo_salik | but this runs fine: sb2 -t $VENDOR-$DEVICE-$PORT_ARCH gcc main.c -o test | 02:21 |
echo_salik | * but this runs fine: sb2 -t $VENDOR-$DEVICE-$PORT_ARCH ./test | 02:22 |
echo_salik | oh i am supposed to run it like that... | 02:22 |
echo_salik | r0kk3rz: remember that sync.h error i had initially... its back... | 04:13 |
echo_salik | does the target have linux headers? i guess linux/sync.h is in headers | 04:22 |
echo_salik | ok this does have headers... | 04:27 |
echo_salik | Q: how can I make a folder be part of the make environment... | 09:03 |
echo_salik | like if i run #include <myfile.h> it would be read | 09:04 |
r0kk3rz | it should be in the standard include paths | 09:05 |
echo_salik | r0kk3rz: any specific way to find what ar ethose? | 09:06 |
r0kk3rz | so you fixed your hal package build? | 09:07 |
echo_salik | nope... trying to include the linux/sync.h file from google's repo so that instead it might build | 09:08 |
r0kk3rz | yeah dont do that | 09:08 |
echo_salik | oh... would mess it up more or build successfully while failing silently? | 09:10 |
echo_salik | would that* | 09:10 |
r0kk3rz | its just not even wrong | 09:10 |
echo_salik | any ideas as to what i shoud do? | 09:12 |
r0kk3rz | that should be provided by your droid-hal-devel package | 09:12 |
r0kk3rz | so check you have something in your local repo, and check whats in it | 09:12 |
echo_salik | kernel repo? | 09:13 |
r0kk3rz | nope | 09:14 |
r0kk3rz | different type of repo, i dont mean git | 09:14 |
r0kk3rz | hok, so part of the build process is packing lots of stuff into RPM packages, and putting them into your local package repository | 09:14 |
r0kk3rz | which you can find under droid-local-repo | 09:14 |
echo_salik | oh the rpm? | 09:14 |
r0kk3rz | so all that stuff you built in habuild gets packaged up like that | 09:15 |
echo_salik | oh... | 09:15 |
echo_salik | so i should search by extracting the repo and using `find`? | 09:16 |
r0kk3rz | yeah that would be a good idea | 09:16 |
echo_salik | thanks | 09:18 |
echo_salik | i have sync.h in a different location | 09:19 |
r0kk3rz | ? | 09:19 |
echo_salik | now why would it do that -.- | 09:19 |
echo_salik | instead of linux folder it is in sync folder | 09:19 |
echo_salik | um... how to i share images here? or is it not allowed? | 09:20 |
r0kk3rz | so, backtrack a second | 09:20 |
r0kk3rz | what have you just done | 09:20 |
echo_salik | making a hastebin... | 09:21 |
echo_salik | https://hastebin.com/utokeqosuw.http | 09:23 |
r0kk3rz | ok so that makes sense | 09:26 |
r0kk3rz | theres your problem, and the fix will be somewhere in your kernel tree | 09:26 |
echo_salik | so my next question: do i have to fork libhybris to make the change from linux to sync folder? or is there a simpler way... (please be a simpler way) | 09:26 |
echo_salik | kernel tree? | 09:27 |
r0kk3rz | yeah | 09:27 |
r0kk3rz | you know, where your kernel source code is | 09:27 |
echo_salik | oh so i should just move that sync file from sync to linux? | 09:28 |
r0kk3rz | dont move it | 09:29 |
r0kk3rz | symlink it perhaps | 09:29 |
echo_salik | oh | 09:29 |
echo_salik | ok... | 09:29 |
echo_salik | ok... sync.h is in linux dir -.- | 09:30 |
r0kk3rz | look in the out folder | 09:31 |
echo_salik | ./target/product/rolex/obj/KERNEL_OBJ/usr/include/linux/sync.h | 09:33 |
echo_salik | result of fine | 09:33 |
echo_salik | *find | 09:33 |
r0kk3rz | seems reasonable | 09:38 |
echo_salik | yeah... could there be a problem any other place? | 09:38 |
echo_salik | could ther ebe a problem with the spec? | 09:40 |
r0kk3rz | the spec should be ok, but make sure your dhd is up to date | 09:41 |
echo_salik | my device folder stucture is like this: device/xiaomi/rolex but my kernel is like: kernel/xiaomi/msm8917 | 09:42 |
r0kk3rz | thats fine | 09:42 |
echo_salik | r0kk3rz: its uptodate | 09:50 |
echo_salik | r0kk3rz: https://github.com/echosalik/droid-hal-rolex | 09:55 |
echo_salik | can i fork libhybris and change the linux/sync.h to sync/sync.h to see if that patch works? | 09:57 |
r0kk3rz | you dont need to fork it | 09:58 |
r0kk3rz | but i wouldnt | 09:58 |
echo_salik | sync.h might not be the only header misplaced... if others are as well it might narrow down where the problem might be? | 10:00 |
echo_salik | r0kk3rz: ok so i have created a patch instead of forking it and added it in buildmw function with an if condition. | 10:43 |
echo_salik | it might either work or fail... both might tell something | 10:44 |
echo_salik | r0kk3rz: ok mow my rpm is missing sw_sync.h | 11:50 |
echo_salik | so RPM is built in target... | 11:52 |
echo_salik | also i noticed some of my changes in the config are not applied. im seeing warnings in mer-kernel-check | 12:02 |
r0kk3rz | that might be ok | 12:08 |
r0kk3rz | not all flags exist in all kernels | 12:08 |
r0kk3rz | so long as they are only warnings, should be ok | 12:08 |
echo_salik | r0kk3rz: no but i have set those | 12:08 |
echo_salik | oh so it can show warnings even if the kernel has no option for that? | 12:09 |
echo_salik | if i run kernel check on my defconfig there are no warnings | 12:13 |
r0kk3rz | yeah dont do that | 12:26 |
echo_salik | .config file is not the same as defcofig? | 12:28 |
echo_salik | oh so .config is generated by make by parsing through the defconfig | 12:32 |
echo_salik | r0kk3rz: as a last try i am keeping sw_sync.h (copy from kernel tree) in the same dir as sync.c and changing it's include | 14:10 |
echo_salik | r0kk3rz: ok there are a few header files that are not in the devel-rolex rpm and the error causing file is libsync | 14:13 |
echo_salik | there is already a libsync.so in the devel rpm... so why is it needed to make another? | 14:15 |
r0kk3rz | its not | 14:16 |
r0kk3rz | it using the header so you can link to it ofc | 14:16 |
echo_salik | so the sync.h is used by libsync in hybris so that it can use libsync.so? | 14:18 |
mal | what are you doing? is that issue during libhybris build? | 14:30 |
mal | are you using which android base? | 14:31 |
echo_salik | mal: running build_packages.sh and yes duing libhybris build | 14:35 |
echo_salik | mal: im using lineage 15.1 | 14:35 |
mal | that needs a different libhybris version | 14:36 |
echo_salik | oh | 14:36 |
mal | https://github.com/krnlyng/libhybris-1/tree/android8-initial | 14:36 |
mal | remember to clone correctly so the submodule is fetched also | 14:37 |
mal | 15.1 needs many custom things which are not documented anywhere | 14:37 |
echo_salik | oh... | 14:38 |
echo_salik | so i should clone this in hybris/mw/ ? | 14:38 |
mal | obviously | 14:39 |
echo_salik | just confirming... | 14:39 |
mal | you then build it using build_packages.sh -b somepath | 14:40 |
echo_salik | ok... | 14:41 |
echo_salik | thanks mal | 14:42 |
echo_salik | and thanks r0kk3rz for clearing stuff up and bearing my annoyance... | 14:42 |
mal | I have too tired to document the 15.1 build process, I have it all in my head | 14:44 |
echo_salik | maybe i can be of some help? | 14:45 |
echo_salik | i doubt i'll be... but still... | 14:45 |
taixzo | mal: anything I can do, not knowing anything about porting, to help with the matissewifi image? | 14:51 |
mal | I just have way too many projects ongoing that I need to finish some to make some sense to everything | 14:53 |
* vknecht sends moral support over the air | 15:06 | |
vknecht | hope you had the occasion for some holidays this summer ? | 15:07 |
echo_salik | hey mal, when running build_packages.sh with -b hybris/mw/libhybris/ im getting Installed (but unpackaged) file(s) found i have added those files in spec but i still get those errors. | 15:19 |
mal | what files | 15:20 |
echo_salik | making a hastebin | 15:21 |
echo_salik | https://hastebin.com/unayolewif.coffeescript | 15:21 |
mal | then you didn't add those correctly to spec | 15:28 |
mal | you can check the original spec to see how those were there https://github.com/mer-hybris/libhybris/blob/master/rpm/libhybris.spec | 15:29 |
echo_salik | oh so i have to use this %{_libdir} | 15:32 |
echo_salik | instead of using /usr/lib | 15:32 |
echo_salik | hey mal, i still have the same error: this is my new spec straggler: https://hastebin.com/uguhodavuc.shell | 15:39 |
mal | NOT TO THAT SPEC | 15:39 |
echo_salik | oh hybris.spec... | 15:40 |
mal | echo_salik: I gave you the libhybris spec with the exact things you add to the one in the other branch and you do something different, that is some messed up logic | 15:40 |
mal | straggler files is a thing specific to droid-hal spec | 15:44 |
mal | and only for problems related to build of those rpms | 15:44 |
echo_salik | mal: my logic was based on what the HADK document in 7.2.2 that i need to add it in hal-droid spec | 15:44 |
echo_salik | i though you gave that to me as a template | 15:44 |
echo_salik | i still have errors: File not found: /home/salik/hadk/hybris/mw/libhybris/installroot/usr/lib/libwifi.so.1; File not found: /home/salik/hadk/hybris/mw/libhybris/installroot/usr/lib/libwifi.so.1.0.0 | 15:46 |
echo_salik | wait... there isnt a folder there | 15:46 |
mal | echo_salik: but that section was about droid-hal, not about libhybris | 15:47 |
echo_salik | its mapping to the wrong location... installroot is in hadk... | 15:47 |
mal | are you really sure you are using the correct branch of libhybris? | 15:47 |
echo_salik | mal: yeah... spec thing made me mix both | 15:48 |
mal | note that the repo was named libhybris-1, unless you manually made it go to libhybris it will be named differently | 15:48 |
mal | pretty sure you are messing thing up somehow | 15:49 |
echo_salik | git remote -v returns: origin https://github.com/krnlyng/libhybris-1.git (fetch) | 15:50 |
mal | how about git branch | 15:50 |
echo_salik | * android8-initial | 15:51 |
echo_salik | hey mal someone named branek had the same error on 2018-06-28 you gave him a pastebin link but that link has now expired | 15:56 |
echo_salik | this was this question: what package provide libwifi? | 15:57 |
mal | how about removing the libwifi stuff from the spec | 15:59 |
mal | it should have clear | 15:59 |
echo_salik | ok... | 16:00 |
echo_salik | mal the libwifi part is alreay commented... | 16:02 |
echo_salik | with an # sign | 16:02 |
echo_salik | i just redownloaded the spec file and removed the libwifi but same error | 16:12 |
vknecht | echo_salik, what do your spec files look like now ? | 16:28 |
echo_salik | vknecht: https://hastebin.com/nilurinija.http | 16:41 |
* echo_salik sings: 99 little bugs in the code 99 little bugs... take one down patch it around... 127 little bugs in the code... | 16:59 | |
mal | echo_salik: I just did a clean build of that libhybris branch against aosp 8.1 base without any errors, unless lineageos base is in some way different | 17:02 |
mal | echo_salik: how exactly did you clone it? like this "git clone https://github.com/krnlyng/libhybris-1.git -b android8-initial --recurse-submodules" | 17:04 |
echo_salik | mal: yep | 17:09 |
echo_salik | exactly like that | 17:09 |
echo_salik | do i have to use asop base? | 17:11 |
mal | no, you can't use the same | 17:11 |
mal | show the whole build log from the libhybris.log | 17:12 |
echo_salik | in a moment | 17:12 |
echo_salik | mal: https://hastebin.com/wipomimuyi.coffeescript | 17:13 |
mal | how can that be if you have the libwifi commented out as it is by default in that branch | 17:15 |
mal | what local changes have you done currently? | 17:16 |
mal | run git diff | 17:16 |
mal | this whole thing should have been trivial to fix, no idea how you manage to break the build | 17:16 |
echo_salik | mal: as you said... we fail in mysterious ways... | 17:21 |
echo_salik | mal: https://hastebin.com/pevimehame.sql | 17:26 |
echo_salik | steps: downloaded spec from that link, deleted wifi listings | 17:26 |
mal | those changes make very little sense, you removed libvibrator? | 17:28 |
vknecht | echo_salik, what's the exact command you used ? | 17:29 |
echo_salik | mal: nope... but isnt that diff with respect to that hybris-l repo? | 17:30 |
echo_salik | vknecht: rpm/dhd/helpers/build_packages.sh -b hybris/mw/libhybris my exact command | 17:31 |
mal | echo_salik: based on that git fif you did remove it | 17:32 |
mal | *git diff | 17:33 |
echo_salik | wait that's not the whole git diff | 17:35 |
echo_salik | my bad: https://hastebin.com/agerikicim.shell | 17:36 |
echo_salik | what i mean to say is this git diff is referring all the changes that the new spec that i downloaded has... the only changes i manually made are removing wifi so things | 17:37 |
echo_salik | i had not commited the libspec that i downloaded... so its showing those changes too | 17:38 |
echo_salik | i not good at conveying what i mean... | 17:38 |
mal | so why did you add libwifi back there manually, just look at your git diff | 17:39 |
mal | I give up, you are telling one thing but git diff says something completely different | 17:39 |
echo_salik | ok... what i'll do again is redownload that spec... commit it... then remove libwifi stuff then show you the git diff... | 17:41 |
mal | what the hell are you doing, LIBWIFI IS ALREADY REMOVED FROM THE DAMN SPEC | 17:43 |
mal | I'm not going to help anymore today, you have all the information you need, just fix it yourself | 17:43 |
r0kk3rz | echo_salik: you dont need to commit anything, it seems like you misunderstand a few git fundamentals | 17:47 |
echo_salik | mal: you gave me this file: https://github.com/mer-hybris/libhybris/blob/master/rpm/libhybris.spec, please look at line: 347 | 17:47 |
mal | echo_salik: WE WERE TALKING ABOUT LIBNFC WHEN I GAVE YOU THAT | 17:48 |
echo_salik | r0kk3rz: mal gave me ^ link, it has libwifi.so uncommented | 17:48 |
mal | NOT LIBWIFI | 17:48 |
mal | there a thing called context, look it up from a dictionary | 17:49 |
echo_salik | mal you gave me that spec, i am using that spec... how the hell am i supposed to know you mean a different spec when you told me to use THAT spec!? | 17:50 |
mal | well with normal logic if your error was about libnfc then you search for things related to thatt from what I gave you, but that would require too much logic it seems | 17:51 |
mal | I even said to look for the things in the error from the spec | 17:53 |
mal | I will not comment anymore here because soon I will be kicked out for bad language | 17:54 |
echo_salik | i'll leave the chat sorry for everything | 17:55 |
mal | oops :DD | 17:57 |
r0kk3rz | oh dear mal :) | 18:24 |
mal | maybe I should just stay silent for a few weeks | 19:00 |
r0kk3rz | mal: just chill out, if people are frustrating you then ignore them | 19:11 |
r0kk3rz | i certainly do that some times, especially if i think they have all the information they need | 19:11 |
mal | I just get annoyed so easily nowaday so better I stop helping completely | 19:17 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!