**** BEGIN LOGGING AT Tue Jun 26 03:00:03 2018
Jun 26 06:19:38 <ubptgbot> qwevv pkndts was added by: qwevv pkndts
Jun 26 08:45:22 <ubptgbot> <vanyasem> superbox, waiting for your final review
Jun 26 08:45:30 <ubptgbot> <vanyasem> @UniversalSuperBox https://github.com/Halium/docs/pull/93
Jun 26 09:43:45 <ubptgbot> <Mattia990> Anyone knows what is usb-tethering service? Can i disable It safetly? (Because It slows my boot and phone take longer to load the gui)
Jun 26 09:49:17 <ubptgbot> <Flohack> @Mattia990, If you want to share the mobile internet via USB and RNDIS adapter
Jun 26 10:05:29 <ubptgbot> <Mattia990> @Flohack, Ah, so i can disable it using systemctl disable (or mask) usb-tethering?
Jun 26 10:13:29 <ubptgbot> <Flohack> @Mattia990, I think so, yes
Jun 26 10:16:35 <ubptgbot> <Mattia990> @Flohack, Ok, thanks!
Jun 26 13:16:51 <ubptgbot> <JBBgameich> On the reference rootfs the usb-tethering service is important, since it allows connecting to the phone using ssh.
Jun 26 14:16:53 <ubptgbot> <c_smith> Does anyone know where clang's location is defined? for some odd reason the executable location is defined as "/clang"...
Jun 26 14:17:13 <ubptgbot> <c_smith> which I'm pretty sure isn't normal behavior
Jun 26 14:18:57 <ubptgbot> <c_smith> and I've so far seen it on  mm-qcamera bits
Jun 26 14:20:10 <ubptgbot> <Mattia990> @vanyasem You forward the wrong message about pulseaudio in halium thoughts group
Jun 26 14:20:30 <ubptgbot> <vanyasem> what should i forward
Jun 26 14:20:33 <ubptgbot> <vanyasem> why is it wrong
Jun 26 14:21:14 <ubptgbot> <vanyasem> [Edit] what should i forward instead?
Jun 26 14:21:20 <ubptgbot> <vanyasem> i will fix it if i'm wrong
Jun 26 14:22:05 <ubptgbot> <Mattia990> @vanyasem, You forwarded this: https://github.com/halium-cancro instead of this: https://paste.ubuntu.com/p/9sZMjT8J9t/
Jun 26 14:22:28 <ubptgbot> <vanyasem> fixed
Jun 26 14:22:29 <ubptgbot> <vanyasem> thank you
Jun 26 14:22:48 <ubptgbot> <Mattia990> Thank to you
Jun 26 14:32:42 <ubptgbot> <c_smith> I'm guessing nobody knows why mka systemimage fails on not finding /clang?
Jun 26 14:33:15 <ubptgbot> <c_smith> because I'm sure the clang binary should not be in root
Jun 26 14:50:43 <ubptgbot> <c_smith> let's see if linking the system clang to /clang results in anything
Jun 26 14:50:55 <ubptgbot> <c_smith> good idea? probably not.
Jun 26 15:40:04 <ubptgbot> <c_smith> Well, I'm at a loss, no matter what version of clang I soft link to that directory, I get that error
Jun 26 15:42:41 <ubptgbot> <abhishek_01> @c_smith, Can you paste errors to paste.ubuntu.com
Jun 26 15:49:43 <ubptgbot> <c_smith> https://pastebin.com/j4vbZhiD
Jun 26 15:50:14 <ubptgbot> <c_smith> Still learning the arguments to pastebinit
Jun 26 16:36:40 <ubptgbot> <c_smith> I can try and paste it to the Ubuntu pastebin, I think I found the specific flash that posts it there
Jun 26 16:36:52 <ubptgbot> <c_smith> Though it'll be later
Jun 26 18:22:26 <ubptgbot> <bhushanshah> Tofe: re: #libhybris
Jun 26 18:23:15 <ubptgbot> <bhushanshah> It looks like libhybris is not hooking set_errno for some reason
Jun 26 18:24:08 <ubptgbot> <bhushanshah> Because as per my understanding, if libhybris hooks it, it can't end up in bionic
Jun 26 18:29:34 <Tofe> mmh you're quite right
Jun 26 18:29:48 <Tofe> I'll setup some hybris debug variables
Jun 26 18:29:49 <ubptgbot> <bhushanshah> https://github.com/libhybris/libhybris/blob/master/hybris/common/hooks.c#L1295
Jun 26 18:32:36 <Tofe> __errnno is hooked, but nothing about set_errno
Jun 26 18:33:22 <Tofe> could it be it was inlined ?
Jun 26 18:33:45 <ubptgbot> <bhushanshah> Maybe
Jun 26 18:43:56 <Tofe> bshah: https://github.com/CyanogenMod/android_bionic/blob/47cf20ae09cccbde2f0b262678e958c679ba21ff/libc/include/android/legacy_errno_inlines.h#L36 seems to confirm that it was exposed in the NDK as an inline function (I'm on armv7, Android 5.1 here)
Jun 26 18:46:29 <ubptgbot> <c_smith> So I've been running up against a wall with clang and running into the errors in the following paste no matter what I try, could anyone else take a look and possibly give me an idea of where to look at what to change at the very least?
Jun 26 18:46:32 <ubptgbot> <c_smith> https://pastebin.com/j4vbZhiD
Jun 26 18:47:40 <ubptgbot> <c_smith> Google-fu fails me on this
Jun 26 18:50:01 <ubptgbot> <bhushanshah> Tofe: our bionic have a patch to make it non-inline but I guess that doesn't help with blobs
Jun 26 18:51:13 <ubptgbot> <NotKit> @bhushanshah, I seen this happen before with TLS slots and Intel blobs
Jun 26 18:51:15 <ubptgbot> <NotKit> @bhushanshah, [Edit] I've seen this happen before with TLS slots and Intel blobs
Jun 26 18:51:19 <ubptgbot> <bhushanshah> https://github.com/ubports/android_bionic/commit/f3ea5ac36fe344dcea4bac546824385d3885b943
Jun 26 18:52:26 <ubptgbot> <bhushanshah> @NotKit, For context https://bpaste.net/show/ca78b9ceac49
Jun 26 18:53:17 <ubptgbot> <NotKit> you have Adreno quirks enabled, right?
Jun 26 18:53:54 <ubptgbot> <bhushanshah> That's question for Tofe I guess
Jun 26 18:54:35 <Tofe> well... mmh.. I don't remember... You mean in the libhybris configure flags ?
Jun 26 18:54:46 <ubptgbot> <bhushanshah> Yes
Jun 26 18:54:53 <Tofe> let me check
Jun 26 18:55:40 <ubptgbot> <bhushanshah> @NotKit, But well I do have it enabled for PlaMo
Jun 26 18:55:40 <Tofe> nope, no quirks in my options
Jun 26 18:55:51 <ubptgbot> <bhushanshah> Halium 7 target
Jun 26 18:55:52 <Tofe> ^ I should fix this...
Jun 26 18:57:51 <ubptgbot> <bhushanshah> @NotKit, Although adreno quirks is different thing IIRC
Jun 26 18:58:07 <ubptgbot> <NotKit> they hack malloc, yes
Jun 26 18:58:18 <ubptgbot> <NotKit> to avoid weird SailfishOS shader compilation error in browser
Jun 26 19:00:23 <ubptgbot> <bhushanshah> Right so well how did you solve issue with Intel blobs?
Jun 26 19:06:51 <Tofe> @bhushanshah: I guess it's the patch that shifts a bit more the TLS slots
Jun 26 19:07:28 <ubptgbot> <ilyaishere> Oh TLS slots again
Jun 26 19:07:35 <ubptgbot> <ilyaishere> Please, be careful :)
Jun 26 19:10:59 <ubptgbot> <bhushanshah> Tofe: so ultimately we need to look into one of solution @slp1605 proposed https://github.com/ubports/ubuntu-touch/issues/677
Jun 26 19:12:58 <ubptgbot> <NotKit> @bhushanshah, patching pthread_getspecific/pthread_setspecific to return key so it can be reached by ASM code as %gs + key * 4
Jun 26 19:13:33 <ubptgbot> <bhushanshah> Do you have link to patch @NotKit ?
Jun 26 19:13:52 <ubptgbot> <NotKit> it wouldn't help in your case, just an example of blob weirdness
Jun 26 19:13:54 <ubptgbot> <NotKit> http://imgur.com/zisT5YM - that was the issue
Jun 26 19:15:21 <ubptgbot> <NotKit> [Edit] it wouldn't help in your case, just an example of blob weirdness (but I think I lost it by now, since I didn't proceed further with Cherry Train + hybris)
Jun 26 19:16:08 <Tofe> That's a bit too much for me :)
Jun 26 19:17:19 <ubptgbot> <bhushanshah> Heh, same... :P
Jun 26 19:21:32 <Tofe> bshah: in slp's proposals, for option 2, what kind of performance penalty does a relocatable tls slot imply ? just an overhead when the thread is created ? more ?
Jun 26 19:24:34 <ubptgbot> <bhushanshah> I'm unsure
Jun 26 19:55:07 <ubptgbot> <abhishek_01> @NotKit, And in gallery afaik
Jun 26 19:55:10 <ubptgbot> <NotKit> in gallery it's BGRA format quirk
Jun 26 19:56:36 <ubptgbot> <abhishek_01> Ah💡
Jun 26 20:07:56 <ubptgbot> <c_smith> I guessing nobody has any tips they can give me about the build not finding clang at certain points?
Jun 26 20:08:40 <ubptgbot> <c_smith> I posted the entire output earlier, I can get the specific parts posted separately if it helps
Jun 26 20:11:17 <ubptgbot> <reivilibre> @c_smith, That doesn't seem like the entire output (assuming https://pastebin.com/j4vbZhiD ). Did you forget to include stderr?
Jun 26 20:11:45 <ubptgbot> <reivilibre> Actually, I'm seeing ANSI escape codes there too :/ Dodgy
Jun 26 20:12:48 <ubptgbot> <c_smith> Hmm. Pastebin it seems to not like this, give me a half hour to finish eating and I'll manually paste it
Jun 26 20:13:12 <ubptgbot> <c_smith> Not to self: check the output before paying the link
Jun 26 20:13:21 <ubptgbot> <c_smith> [Edit] Note to self: check the output before paying the link
Jun 26 20:13:37 <ubptgbot> <c_smith> [Edit] Note to self: check the output before pasting the link
Jun 26 20:13:41 <ubptgbot> <reivilibre> No worries, I assume you used a command-line pastebin tool?
Jun 26 20:13:56 <ubptgbot> <c_smith> Yep, pastebinit
Jun 26 20:14:33 <ubptgbot> <reivilibre> I'm thinking you probably did something like `mka <...> | pastebinit`
Jun 26 20:14:46 <ubptgbot> <c_smith> Correct
Jun 26 20:15:28 <ubptgbot> <reivilibre> Ah, in that case, only stdout ('standard out') will be put onto pastebin, whereas stderr ('standard error') would just wind its way up onto your terminal (but not on pastebin)
Jun 26 20:17:16 <ubptgbot> <c_smith> Hmmm... I've dealt with something similar before, would "mka systemimage 1&> | pastebinit theoretically work?
Jun 26 20:17:36 <ubptgbot> <c_smith> [Edit] Hmmm... I've dealt with something similar before, would "mka systemimage 1&> | pastebinit" theoretically work?
Jun 26 20:18:15 <ubptgbot> <reivilibre> I think `mka systemimage 2>&1 | pastebinit` would work … ( stderr = 2, stdout = 1, and `2>&1` redirects stderr into stdout)
Jun 26 20:18:42 <ubptgbot> <reivilibre> Your suggestion isn't familiar syntax to me, but that doesn't necessarily mean it's wrong
Jun 26 20:18:45 <ubptgbot> <c_smith> See to try, as I'm now thoroughly curious
Jun 26 20:18:56 <ubptgbot> <c_smith> [Edit] Time to try, as I'm now thoroughly curious
Jun 26 20:19:56 <ubptgbot> <reivilibre> You will also see e.g. `[0;31m` in the output, this is how colours are generated in the terminal; the program generates the ESCAPE character and then `[0;31m` to set the colour (that particular one probably means red..)
Jun 26 20:20:09 <ubptgbot> <reivilibre> Not strictly on-topic but just interesting knowledge
Jun 26 20:25:38 <ubptgbot> <c_smith> True, but it does look like that did it.
Jun 26 20:26:09 <ubptgbot> <c_smith> https://pastebin.com/TKSHWL7G
Jun 26 20:29:04 <ubptgbot> <reivilibre> If you see on line 578, you can see the commandline passed to `bash` which then tries to execute `/clang`..
Jun 26 20:29:22 <ubptgbot> <reivilibre> My guess would be an environment variable which hasn't been set
Jun 26 20:29:22 <ubptgbot> <c_smith> It LOOKS like it's looking for the Clang binary in the root directory of the OS, but linking any Clang (system or from the prebuilt directory) does absolutely nothing to change the error, so I'm not sure what to think
Jun 26 20:30:27 <ubptgbot> <reivilibre> Imagine a script which calls `${CC_DIR}/clang` and `CC_DIR` wasn't set, then the path would become just `/clang` . Hmm
Jun 26 20:30:45 <ubptgbot> <c_smith> that, that makes sense
Jun 26 20:31:03 <ubptgbot> <reivilibre> The only question is, who is doing that and what variable does it want :P
Jun 26 20:31:19 <ubptgbot> <c_smith> that's gonna take time track down for sure
Jun 26 20:31:31 <ubptgbot> <reivilibre> What device are you building? Are your manifests merged?
Jun 26 20:31:46 <ubptgbot> <reivilibre> I'm no expert but I can have a little poke around
Jun 26 20:31:53 <ubptgbot> <c_smith> cheeseburger, and manifests are current as of 6 hours ago
Jun 26 20:32:06 <ubptgbot> <reivilibre> You are porting this device yourself?
Jun 26 20:32:16 <ubptgbot> <c_smith> trying
Jun 26 20:32:23 <ubptgbot> <c_smith> gave up on this error for a couple weeks
Jun 26 20:32:34 <ubptgbot> <c_smith> figured I'd at least lend a hand again
Jun 26 20:33:18 <ubptgbot> <c_smith> and if I manage to, with help, port it, then better for OP5 users who want something like this
Jun 26 20:34:01 <ubptgbot> <c_smith> from what I know there's only 2 other people trying this, Marius (I hope I spelled his name correctly) and another guy on Github
Jun 26 20:34:30 <ubptgbot> <reivilibre> Are they stuck on this issue as well?
Jun 26 20:38:15 <ubptgbot> <c_smith> the guy on Github is likely there as he's going the route I went weeks ago (as well as now), Marius, no clue
Jun 26 20:38:53 <ubptgbot> <c_smith> took me all of an hour or so to get to this point, not counting time spent waiting
Jun 26 20:43:29 <ubptgbot> <c_smith> derp, forgot to sync 1 file: a make file that I removed all java-related things from that the halium tree setup script didn't touch
Jun 26 20:44:28 <ubptgbot> <c_smith> that is if it's in the main halium tree (which I don't think it is as it's board specific).
Jun 26 20:46:43 <ubptgbot> <c_smith> found the file, but as it's a vendor/<company>/<device>/Android.mk file, I'm pretty sure it's not going to have changes like these accepted.
Jun 26 20:48:28 <ubptgbot> <reivilibre> You could fork it and modify the changes yourself, if desperate :P
Jun 26 20:49:26 <ubptgbot> <c_smith> true. I'm just thinking it might make it easier for anyone working on this, but then again, probably only necessary to let them know that the java files are in Android.mk and not somewhere more sensible
Jun 26 20:49:40 <ubptgbot> <c_smith> easy enough to remove methinks
Jun 26 20:51:51 <ubptgbot> <reivilibre> Yeah sorry, I really don't know. But at least you have a pastebin output with some output on it now :P
Jun 26 20:52:08 <ubptgbot> <reivilibre> Maybe someone, who knows what they're on about, will have a chance
Jun 26 20:52:55 <ubptgbot> <c_smith> true story.
Jun 26 20:52:58 <ubptgbot> <c_smith> thanks for trying though
Jun 26 20:53:07 <ubptgbot> <reivilibre> yw, good luck with it :-)
Jun 26 20:56:22 <ubptgbot> <c_smith> I MAY have found a hint as to the chain to follow just now
Jun 26 20:56:29 <ubptgbot> <c_smith> looking at envsetup.sh
Jun 26 21:13:22 <ubptgbot> <c_smith> sooo I think I figured out the clang issue, breakfast DOES work and seems to set more than lunch, but for breakfast I have to set a specific name instead of just cheeseburger.
Jun 26 21:13:36 <ubptgbot> <c_smith> i.e. breakfast halium_cheeseburger-eng
Jun 26 21:15:13 <ubptgbot> <reivilibre> good on ya for being smart enough to try that. … Note I think `lunch` is intended for Halium/Android 5.1
Jun 26 21:15:22 <ubptgbot> <Lyud ☭♀☆🌈> @c_smith, probably vendorsetup.sh?
Jun 26 21:18:25 <ubptgbot> <c_smith> Olivier, yeah, I'm still learning how things work lol.  … Lyud, seems like it.
Jun 26 21:19:01 <ubptgbot> <reivilibre> @c_smith, I'm still learning how things work too, ;-)
Jun 26 21:19:59 <ubptgbot> <c_smith> hmmmm.... seems "hardware/qcom/diplay-caf/msm8998/liblight/lights_prv.cpp" is mssing.
Jun 26 21:20:09 <ubptgbot> <c_smith> the corresponding header file is there,
Jun 26 21:21:45 <ubptgbot> <c_smith> not exactly sure why Lineage would exlude that...
Jun 26 21:21:57 <ubptgbot> <c_smith> [Edit] not exactly sure why Lineage would exclude that...
Jun 26 21:25:39 <ubptgbot> <c_smith> hmmmm well it's there on Github, soooo repo sync
Jun 26 21:42:22 <ubptgbot> <c_smith> well, back to that error (referencing /clang), sooo breakfast did nothing to change that it seems, rather it restarted the build. >.<
Jun 26 21:47:42 <ubptgbot> <c_smith> Olivier, I think I may have found the variable needed, it's not the cc one, but SDCLANG
Jun 26 21:48:34 <ubptgbot> <reivilibre> Well done. Have you located where it should be set?
Jun 26 21:48:34 <ubptgbot> <c_smith> which makes sense as this phone is a Snapdragon device
Jun 26 21:48:39 <ubptgbot> <c_smith> not yet
Jun 26 21:48:52 <ubptgbot> <c_smith> still tracking down where it's specifically set
Jun 26 21:49:00 <ubptgbot> <reivilibre> I assume it's some vendor makefile being stupid, because I didn't have such issue
Jun 26 21:50:27 <ubptgbot> <c_smith> soooo I found where it SHOULD be set:
Jun 26 21:50:27 <ubptgbot> <c_smith> the boardconfig
Jun 26 21:51:05 <ubptgbot> <c_smith> at least what's in the Qualcomm developer documentation
Jun 26 22:36:26 <ubptgbot> <c_smith> and I THINK setting the sdclang path did it, we'll see. I've been wrong about that so far 4 or 5 times
Jun 26 22:37:39 <ubptgbot> <c_smith> hmmmm... this is new: an actual programming error
Jun 26 23:25:39 <ubptgbot> <c_smith> soooo I know nothing of C++ (most I know is Python and Bash), soooo I'm most likely not going to be able to figure out what's going on with this error, but I think I have pinned it down to a specific part of QCamera3HWI.cpp. I'll pastebin both the output and the part of the file I mentioned that is causing this
Jun 26 23:26:26 <ubptgbot> <c_smith> here's the output: http://paste.ubuntu.com/p/jhBfPswN6S/
Jun 26 23:29:52 <ubptgbot> <c_smith> and the offending section of QCamera3HWI.cpp: https://paste.ubuntu.com/p/RYZMNHFMKV/
Jun 26 23:30:06 <ubptgbot> <c_smith> line numbers added so people can find where it is if they want to jump in
Jun 26 23:55:01 <ubptgbot> <Lyud ☭♀☆🌈> Where did you get the device tree?
Jun 26 23:55:22 <ubptgbot> <Lyud ☭♀☆🌈> From LineageOS or from an unofficial source?
Jun 27 00:40:15 <ubptgbot> <c_smith> It's all official as much as possible
Jun 27 00:41:12 <ubptgbot> <c_smith> Meaning binary blobs are from TheMuppets, Android bits are losing, halium bits, well that's obvious
Jun 27 00:41:27 <ubptgbot> <c_smith> [Edit] Meaning binary blobs are from TheMuppets, Android bits are los, halium bits, well that's obvious
Jun 27 00:41:58 <ubptgbot> <c_smith> I don't think I have a single unofficial repo now that I think about it
Jun 27 02:04:01 <ubptgbot> <vanyasem> (Photo, 1280x960) https://irc.ubports.com/R0JMOLhH.png
Jun 27 02:04:02 <ubptgbot> <vanyasem> (Video, 31s) https://irc.ubports.com/cqsyRCvV.mp4
Jun 27 02:05:13 <ubptgbot> <vanyasem> Thanks to @NotKit (i owe you a beer) running Xorg on Nexus 5x (ArchLinuxArm rootfs)
Jun 27 02:07:38 <ubptgbot> <vanyasem> I will probably push a prebuilt rootfs with Xorg support configured later
Jun 27 02:29:59 <ubptgbot> <trueleo> Just crazy and my second … https://www.youtube.com/watch?v=92NKzAVlYyc
**** ENDING LOGGING AT Wed Jun 27 03:00:01 2018