Wednesday, 2022-07-13

T42<Umeaman> Are you perhaps awake, @Miklos?03:30
T42<Umeaman> I did what you told me to and I still have the same result.03:30
T42<Umeaman> Here's the error:
T42<edp_17> Have you checked in out/target/product/$DEVICE/obj/KERNEL_OBJ/.config whether that flag is n? If not then because of its dependencies, it got turned back on by the ccompiler.05:54
T42<edp_17> Just set it to #CONFIG_SUNRPC is not set how Elros recommended.06:02
T42<elros34> show 'git diff' of all your defconfig changes, that will show all optional filesystems you enabled06:09
T42<edp_17> Can I get some help please? There is no /system mounted so only telnet 23 works (not 2323)07:13
T42<edp_17> Based on what I saw on the android base :
T42<edp_17> I've created fixum-mount points :
T42<edp_17> (I used jfltexx as an example)07:14
T42<edp_17> On device not much things have mounted and / is missing in fstab :
T42<elros34> /system partition have nothing to telnet 23/2323. Did you mean /data? Reason for not mounted data is in init.log07:16
T42<elros34> /etc/fstab never have anything useful on sfos07:17
T42<edp_17> Oh, okay.07:18
T42<edp_17> Yes, /data has failed to mount "mounting /data//.stowaways/sailfishos on /target failed: No such file or directory". init.log :
T42<edp_17> mounting /dev/block/platform/msm_sdcc.1/by-name/userdata on /data failed: No such file or directory07:19
T42<TheVancedGamer> There's your problem07:22
T42<TheVancedGamer> bootdevice/ is created by ueventd07:22
T42<elros34> should be clear what is wrong, if you read notes then it's explicitly mention to not assume 'block/bootdevice/by-name/' but always check your device fstab07:22
T42<elros34> in your case it will contains platform/msm_sdcc.1/07:23
T42<edp_17> On android? (re @elros34: should be clear what...)07:23
T42<elros34> on androidor just in device repo, it doesn't matter, same fstab file07:24
T42<TheVancedGamer> Even then device fstab can rely on some cmdlines07:24
T42<TheVancedGamer> Like androidboot.bootdevice=platform/msm_sdcc.1/07:24
T42<TheVancedGamer> that then maps bootdevice/ to that directory07:25
T42<edp_17> Ok, I got it. Is it possible to fix this on device without rebuilding droid-hal?07:25
T42<elros34> at least you need to build hybris-boot and flash it, then you can manually fix all systemd mount units07:26
T42<edp_17> where, in init.rc?07:39
T42<elros34> /usr/lib/systemd/system/*.mount07:40
T42<edp_17> Thanks!07:46
T42<edp_17> There is no /user/lib. :) I guess there will be once I flashed the fixed hybris-boot, right?07:48
T42<elros34> yes07:49
T42<edp_17> Thanks.07:50
piggzlbt: ping08:55
T42<edp_17> OBS is building VERY slow. :(09:05
piggz@edp_17 probbaly cos i keep kicking off long kernel builds! D09:06
T42<edp_17> long kernel builds?09:46
T42<edp_17> It's been over 2 hours now. It says "Buildtime: about 2 hours " This is for droid-hal. 😳09:47
deathmistThaodan: does need more reviews and testing still?10:14
Thaodandeathmist: merged10:36
T42<edp_17> OBS really sucks today. It's been building one package for more than 5 hours!! now. :(13:07
T42<Umeaman> It's getting older as we do. Expect things to go slower. :)13:17
T42<edp_17> Yeah, but not this slow. :] It's building that only package since morning.14:23
T42<edp_17> I though, qucikly build a new image to install. What I was thinking. :)14:24
mal@edp_17 which package?15:11
T42<edp_17> mal : droid-hal15:12
maljust a warning, obs will be busy for the rest of today, was just added there and quite a lot of packages will be built15:12
T42<edp_17> It's been 'building' for 7 hours, then I triggered the service and it's restarted. Since, it is 'scheduled' :(15:13
malyeah, it will hopefully get its turn at some point but might take some time15:14
T42<edp_17> I realized that I need to wait. :) Thanks.15:17
malmaybe I will wait some time before triggering some more builds15:18
mal this page tells how many builds are in queue15:19
mal1272 arm builds in waiting queue and 2600 in blocked queue :)15:20
malprobably not everything will get rebuilt in the end if nothing changed in the dependency packages15:20
T42<edp_17> Too late, already triggered some others and one of them completed fine within half an hour or so.15:21
T42<edp_17> Chum is occupying most of the jobs.15:26
malthere is a second chum project I didn't yet trigger15:33
malI will do that later15:33
T42<Umeaman> It was reenabled. I'm gonna comment out those 2 lines. (re @edp_17: Have you checked in ...)15:36
T42<Umeaman> Or even all three.15:36
T42<Umeaman> CONFIG_SUNRPC=y15:36
T42<Umeaman> CONFIG_SUNRPC_GSS=y15:36
T42<edp_17> I've got similar situation few times.15:37
T42<edp_17> Once you are happy with your kernel, might be worth to overwrite defconfig with content of .config. That way you can be sure that you only have it in there which is relevant.15:38
T42<Umeaman> Yeah.15:39
T42<Umeaman> You mean so that they match each other?15:39
T42<edp_17> In one of my kernels the defconfig was messy and this way I've managed to clear it out. Of course if you are happy with your defconfig you don't have to do that.15:42
T42<elros34> I suggest regenerate defconfig also before making any changes17:02
T42<elros34> @lbt: hi,  could you remove this leftover form deleted project:
T42<NIPER> How to install unofficial android apps support!18:00
T42<NIPER> How to install unofficial android apps support? (edited)18:00
T42<NIPER> Sailfish
T42<NIPER> LG nexus 5 (hammerhead)18:00
lbttypically those directories will be cleaned out as publishing happens18:41
T42<elros34> @lbt with 58 -> 64 sub project that didn't happen. I have removed and created 68, will w18:53
T42<elros34> will see if that will help18:53
mal@edp_17 that droid-hal build is odd, it should be fast since it just copies the rpms, but there is nothing in log so I don't know what is happening19:51
mallbt: any idea what could be happening here that build should be just copying prebuilt rpms but seems stuck somehow19:53
T42<edp_17> mal : thanks for looking into it. I haven't checked the status of that since 5pm. I hoped that would be completed by now. I hope it will be sorted.20:06
T42<elros34> I think triggering rebuild usually helps for "Status: Updating" stuck20:17
mal@edp_17 for my packages the usual build time like 1-2 minutes20:19
mal@edp_17 try triggering that build again20:20
T42<edp_17> mal : i know. Usually that is the build time for my packages too. That's why I 'complained' about it so many times today. :)20:36
T42<edp_17> Will try to trigger.20:36
T42<edp_17> mal : No matter what I try, droid-hal package remains "scheduled" :(21:36
T42<edp_17> Same problem with my i9100 and jfltexx projects too.21:39
T42<edp_17> It seems i9100 and jfltexx builds are progressing but the hammerhead still stuck on 'scheduled'22:10
T42<Umeaman> I always delete out/ in ~/hadk22:24
T42<Umeaman> That should regenerate a new defconfig-22:24
T42<Umeaman> That should regenerate a new defconfig (edited)22:24
T42<elros34> why would you always delete out/ directory? This way you rebuild whole android/kernel part from for nothing22:26
T42<edp_17> Is OBS down? 😳22:29
T42<Umeaman> Elros: Uhm, to make a new .config based on my changed defconfig. Won't that work?22:37
T42<edp_17> For that you don't need to delete nothing from out/22:39
T42<edp_17> It will re-generate .config there from your changed defconfig.22:39
T42<edp_17> I am trying to OTA upgrade and getting this error:
T42<edp_17> I've checked that url (repo) does exist and looks okay. Any idea why and how to solve it22:40
lbtI'm doing some admin work - there may be some downtime...22:43
lbtedp_17: try again22:43
T42<elros34> @Umeaman it's pointless, where did you get that idea? .config in out/ is always regenerated from your defconfig when you run make hybris-hal. No single file needs to be deleted22:45
T42<Umeaman> Okey.22:45
T42<Umeaman> Just trying my own theories.22:45
T42<edp_17> lbt : Oh, I see. Thanks, it's working again. Except my droid-hal package in hammerhead project. That's still in 'scheduled' status. (Have just tried to restart but it always is going back to 'scheduled')22:49
T42<edp_17> @elros34 : Good evening. Do you have idea for this? (re @edp_17: I am trying to OTA u...)22:50
lbtit's going down again ... :)22:53
lbtthe system is really busy so it's hard to check - may need to wait until the build storm is over22:54
T42<edp_17> lbt : Okay. Could that certificate error be a result of what your are doing?22:57
mal@edp_17 let's check your package tomorrow, now it could be scheduled just because there is no time to build it22:57
T42<edp_17> mal : Okay, thanks. The other two builds have completed, so I guess this one also will tomorrow or eventually. :)22:58
T42<edp_17> How should I know what to add into CPUCHECK_STRING in /var/lib/flash-partition/device-info ?23:00
T42<edp_17> Ok, I got it.23:09

Generated by 2.17.1 by Marius Gedminas - find it at!