Wednesday, 2020-10-07

*** kiwi_29 <kiwi_29!> has joined #yocto00:08
*** kiwi_29 <kiwi_29!> has quit IRC00:13
*** cp- <cp-!> has quit IRC00:41
*** cp- <cp-!> has joined #yocto00:42
*** JonathanCrockett <JonathanCrockett!> has quit IRC00:42
*** sakoman <sakoman!> has joined #yocto00:53
*** rcw <rcw!~rcwoolley@> has quit IRC01:11
*** roussinm <roussinm!> has quit IRC01:12
*** shoragan <shoragan!~shoragan@debian/developer/shoragan> has quit IRC01:14
*** OnkelUlla <OnkelUlla!> has quit IRC01:15
*** shoragan <shoragan!~shoragan@debian/developer/shoragan> has joined #yocto01:18
*** OnkelUlla <OnkelUlla!> has joined #yocto01:21
*** moto-timo <moto-timo!~ttorling@fsf/member/moto-timo> has quit IRC01:36
*** goliath <goliath!> has quit IRC01:52
*** JonathanCrockett <JonathanCrockett!> has joined #yocto02:55
*** stephano <stephano!> has quit IRC03:15
*** JonathanCrockett <JonathanCrockett!> has quit IRC03:18
*** JonathanCrockett <JonathanCrockett!> has joined #yocto03:28
*** JonathanCrockett <JonathanCrockett!> has quit IRC03:55
*** stacktrust <stacktrust!> has quit IRC04:10
*** stacktrust <stacktrust!> has joined #yocto04:12
*** nerdboy <nerdboy!~sarnold@gentoo/developer/nerdboy> has quit IRC04:13
*** nerdboy <nerdboy!~sarnold@> has joined #yocto04:16
*** nerdboy <nerdboy!~sarnold@gentoo/developer/nerdboy> has joined #yocto04:17
*** feddischson <feddischson!> has joined #yocto04:36
*** AndersD <AndersD!> has joined #yocto04:39
OutBackDingokhem: is this you?
*** jobroe <jobroe!> has joined #yocto04:52
*** moto-timo <moto-timo!> has joined #yocto05:01
*** moto-timo <moto-timo!~ttorling@fsf/member/moto-timo> has joined #yocto05:01
* moto-timo managed to migrate raw ext4 to new LVM drive including docker weechat container. Yay.05:01
*** RPI_IMX6 <RPI_IMX6!> has joined #yocto05:03
RPI_IMX6Where can I change it ? TUNE_FEATURES        = "arm vfp cortexa7 neon thumb callconvention-hard"   Thanks!05:04
*** thaytan <thaytan!> has quit IRC05:21
*** thaytan <thaytan!> has joined #yocto05:22
khemRPI_IMX6:what do you want to change in that list ?05:23
khemusually DEFAULTTUNE variable imapacts this list but it not the only one05:24
*** beneth` <beneth`!~beneth@2001:41d0:c:a71:1000:22::> has joined #yocto05:25
khemOutBackDingo: yes05:31
*** manuel1985 <manuel1985!> has joined #yocto05:43
RPI_IMX6khem:  I bitbaked the image and got  some useful IPKs on " NXP i.MX6 based Solo to Quad Core SOM Cortex A9" . I tried the IPKs on my Cortex A7 board and I could  not install it . I think I maybe can change the Tuning Feature to make it work on other board.05:44
khemRPI_IMX6: yes you can, where you can use lowest common denominator05:48
*** dreyna_ <dreyna_!~dreyna@2601:646:4201:e280:d1d3:bd1c:bb8b:d26b> has quit IRC05:49
*** kpo_ <kpo_!> has quit IRC05:49
*** kpo_ <kpo_!> has joined #yocto05:50
khemperhaps enforce DEFAULTTUNE = "armv7athf-neon"05:51
khemvia distro policy for armv7 machines05:51
*** pohly <pohly!> has joined #yocto05:51
RPI_IMX6khem: Thanks. Do you know where to change the similar setting in DOCKER file? The current project is running on Raspberry PI. I wand to make it work on IMX6.05:54
*** agust <agust!> has joined #yocto05:54
OutBackDingokhem: you think its enough for a backport of the  --branch release/core45
*** georgem_home <georgem_home!uid210681@gateway/web/> has quit IRC06:03
moto-timoRP: toaster is definitely currently broken. I will file a bug with proper links and evidence in the morning. Blocker :(06:07
OutBackDingokhem: nevermind just found this
moto-timoRP: the good news is it breaks toaster bare-metal and container equally. So the love is shared amongst all the children.06:08
moto-timoRP: toaster_ui.log is the culprit now, not cooker log (except XMLRPC server exits VERY fast in cooker log)06:10
*** rcoote <rcoote!> has joined #yocto06:11
*** dreyna_ <dreyna_!> has joined #yocto06:11
*** creich <creich!> has quit IRC06:16
moto-timoRP: I suspect his logic is failing, but it might be lack of cookerdata being instantiated instead
*** creich <creich!> has joined #yocto06:17
moto-timoRP: this CANNOT work now, because cookerdata is not populated yet... and I don't know yet where to populate cookerdata properly
*** chris_ber <chris_ber!~quassel@> has joined #yocto06:22
*** frsc <frsc!> has joined #yocto06:22
*** chris_ber <chris_ber!~quassel@> has quit IRC06:32
*** rob_w <rob_w!~rob@unaffiliated/rob-w/x-1112029> has joined #yocto06:35
*** RPI_IMX6 is now known as wxin06:44
*** mihai- <mihai-!~mihai@unaffiliated/mihai> has joined #yocto06:44
*** fl0v0 <fl0v0!~fvo@> has joined #yocto06:45
*** jmiehe <jmiehe!> has joined #yocto06:46
*** chris_ber <chris_ber!~quassel@> has joined #yocto06:47
*** ebail <ebail!> has joined #yocto06:49
*** wxin <wxin!> has quit IRC06:49
*** frwi <frwi!~frwi@> has joined #yocto06:50
*** LetoThe2nd <LetoThe2nd!uid453638@gateway/web/> has joined #yocto06:55
*** ebail <ebail!> has quit IRC06:56
*** chris_ber <chris_ber!~quassel@> has quit IRC06:58
*** frwi <frwi!~frwi@> has quit IRC06:58
*** frwi <frwi!~frwi@> has joined #yocto06:58
*** leon-anavi <leon-anavi!~Leon@> has joined #yocto07:00
*** chris_ber <chris_ber!~quassel@> has joined #yocto07:01
*** frwi <frwi!~frwi@> has quit IRC07:03
*** frwi <frwi!~frwi@> has joined #yocto07:04
*** lukma <lukma!> has joined #yocto07:05
LetoThe2ndelGayo dudX07:05
LetoThe2nd"yo dudX"07:05
moto-timoGuten Morgen und Guten Nacht LetoThe2nd07:06
moto-timoIch bin schlaffen07:06
LetoThe2ndmoto-timo: hehe. fun typo: "schlafen" = sleep, "schlaffen" = "be sagging"07:06
moto-timomy German is terrible, sorry07:07
moto-timobut that is hilarious07:07
LetoThe2ndmakes for a good laugh, and laughing != terrible. all is well.07:07
moto-timoI could use a laugh EVERY SINGLE DAY of #2020 shitshow07:08
moto-timo"You won't believe what happened!"07:08
LetoThe2ndi think that applies to most of us.07:08
moto-timoUmm. Yes. I would. #202007:08
LetoThe2ndanyways, have a good and refreshing night!07:09
moto-timoOn Saturday I will brew an Oktoberfest. Late, but still. It will be delicious.07:09
moto-timoSomething like 40 liters...07:09
moto-timoGuten Nacht!07:10
moto-timoBis morgen.07:10
*** dStruct <dStruct!~matt@unaffiliated/dstruct> has quit IRC07:11
*** Ox861726f6c64 <Ox861726f6c64!~Ox861726f@gateway/tor-sasl/ox861726f6c64> has joined #yocto07:12
*** awaisb <awaisb!~quassel@> has joined #yocto07:12
*** abelal <abelal!~quassel@> has quit IRC07:13
*** ebail <ebail!> has joined #yocto07:13
*** dStruct <dStruct!~matt@unaffiliated/dstruct> has joined #yocto07:13
LetoThe2ndI hereby declare this "Van Halen blasting day" in #yocto07:15
*** mdp <mdp!sid49840@gateway/web/> has quit IRC07:19
*** mdp <mdp!sid49840@gateway/web/> has joined #yocto07:19
*** rubdos_ <rubdos_!~rubdos@2a02:578:859d:700:8b44:5716:382d:a7da> has joined #yocto07:21
*** rubdos <rubdos!~rubdos@2a02:578:859d:700:8b44:5716:382d:a7da> has quit IRC07:22
RPmoto-timo: interesting. I can't quite spot what is breaking :/07:27
*** gsalazar <gsalazar!5e3dbd6b@gateway/web/cgi-irc/> has joined #yocto07:29
*** gsalazar <gsalazar!5e3dbd6b@gateway/web/cgi-irc/> has quit IRC07:32
*** gsalazar <gsalazar!5e3dbd6b@gateway/web/cgi-irc/> has joined #yocto07:34
*** frsc <frsc!> has quit IRC07:35
*** frwi <frwi!~frwi@> has quit IRC07:39
*** frwi <frwi!~frwi@> has joined #yocto07:39
*** rob_w_ <rob_w_!> has joined #yocto07:39
*** rob_w <rob_w!~rob@unaffiliated/rob-w/x-1112029> has quit IRC07:41
*** frsc <frsc!> has joined #yocto07:41
*** rob_w <rob_w!~rob@unaffiliated/rob-w/x-1112029> has joined #yocto07:44
rburtonhm i wonder what the best way to determine if a recipe is available to be built in oe-selftest is07:45
*** mckoan|away is now known as mckoan07:46
*** rob_w_ <rob_w_!> has quit IRC07:46
mckoangood morning07:46
*** dev1990 <dev1990!> has joined #yocto07:53
*** fbre <fbre!91fdde45@> has joined #yocto07:55
fbreHi! I'm somewhere in any .bbappend file of my layer. Is there a global variable where the kernel arguments are available which u-boot passes to the kernel on startup?07:56
LetoThe2ndfbre: nope, why should they be?07:56
fbreBecause, if I had a yocto image which I don't want to touch again, I have the chance to break the boot-up in u-boot and parametrize the system by u-boot kernel arguments.08:00
qschulzfbre: bootargs environment variable does that in U-Boot08:01
fbreyes, and I want to access that later in my .bbappend script08:02
fbreI want the /init script to get those u-boot arguments08:04
*** megabread <megabread!~megabread@2a01:4b00:e031:2600:2879:43f8:8781:2957> has joined #yocto08:05
LetoThe2ndfbre: then either set it globally through some form of distro/machine and have both react to it, or rethink your solution.08:06
qschulzfbre: whatever you put into bootargs will be passed to the kernel boot arguments, you don't need to do anything special08:06
LetoThe2ndconcerning init, is INIT_MANAGER = "" a valid approach?08:07
qschulzfbre: another possibility is to use, extlinux.conf or something similar, save it somewhere modifiable in your medium storage, load it in U-Boot and "execute" it08:08
qschulzfbre: that way, you can  modify the extlinux.conf file from a recipe if you want08:08
*** rob_w_ <rob_w_!> has joined #yocto08:09
*** rob_w <rob_w!~rob@unaffiliated/rob-w/x-1112029> has quit IRC08:12
fbreaaaah, I found out, I can parse /proc/cmdline at runtime to get the u-boot boot parameters to the kernel. No need to construct something in .bbappend scripts. :)08:14
jmieheHey, I'm yocto'ing again and want to update/replace a file installed in a lower layer. Am I right to create a bbappend for the recipe that originally installs the file, then delete said file in the bbappend, then I should be free to re-install this file?08:16
LetoThe2ndjmiehe: should work08:17
gsalazarjmiehe: depending on the file you might not even need to delete the original one. It should be overwritten just by listing it in the SRC_URI (as long as it has the same name)08:19
qschulzfbre: there might be some wrapper available somewhere to avoid parsing by hand the cmdline but I can't remember if that was a dream of mine or something that exists08:19
qschulzgsalazar: though if you install it manually in do_install, your do_install will fail because the file does not exist in WORKDIR anymore08:20
fbreqschulz: hahaha a dream...08:21
LetoThe2ndpipe dream?08:21
qschulzgsalazar: (note this will only happen after a cleansstate because the WORKDIR (where files from SRC_URI are put) isn't cleaned between re-runs IIRC)08:21
gsalazarqschulz: good to know, usually it's the approach I take to override configuration files for some utilities (ie nginx)08:22
fbreI asked here because there is a function do_install() in my .bbappend script which calls this:    install -m 0755 ${WORKDIR}/ ${D}/init      and I thought I can also pass the u-boot boot arguments to that /init script by a magic yocto env var or something08:23
jmiehegsalazar: It's a (text) config file, if that matters08:24
qschulzgsalazar: wait wait, I got everything mixed up, I thought you were saying removing it from SRC_URI. So... removing a file from SRC_URI to not install it => incorrect/risky08:24
qschulzgsalazar: if you have a bbappend with a file to override, just FILESEXTRAPATHS_prepend := will be enough, no need to "re-add" it to SRC_URI08:25
gsalazarqschulz: that's what I meant08:25
qschulzgsalazar: if it's a file to add in a bbappend, FILESEXTRAPATHS_prepend := + SRC_URI08:25
qschulzgsalazar: my bad :)08:25
qschulzfbre: no, recipe data is local and usually bootargs are defined in U-Boot binary (hardcoded or some logic is done to create it) so no :/08:26
qschulzfbre: I repeat though, you could create an extlinux.conf file that is programmatically set in Yocto depending on your machine/distro etc...08:26
qschulz(you can create files in do_compile/do_install by hand if need be)08:27
gsalazarjmiehe: it's how qschulz said, you just need a bbappend with a FILESEXTRAPATHS_prepend := "<path to file>" where you you add the files you want to override08:28
qschulzfbre: note that some code change might be required in u-boot for this to work ;) (but it's not uncommon to use extlinux.conf invendor BSPs)08:28
qschulzjmiehe: though... please respect subdirectories. If your original file is in </path/to/recipe/>files/<machine>/myfile.txt, in your bbappend directory you want it in files/<machine>/ too08:29
*** rob_w_ <rob_w_!> has quit IRC08:39
*** rob_w <rob_w!~rob@unaffiliated/rob-w/x-1112029> has joined #yocto08:39
*** florian_kc <florian_kc!~florian_k@Maemo/community/contributor/florian> has joined #yocto08:40
*** CTer <CTer!> has joined #yocto08:43
*** dreyna_ <dreyna_!> has quit IRC08:56
*** CTer <CTer!> has quit IRC09:00
RPTwo totally green master-next builds :)09:09
LetoThe2ndRP: css trickery? post-its on screen? a sharpie?09:11
*** fbre <fbre!91fdde45@> has quit IRC09:12
RPLetoThe2nd: I've wondered which :)09:13
qschulzLetoThe2nd: someone tempered with the autobuilder, I don't see any other option.09:14
qschulzRP: \o/09:14
LetoThe2ndqschulz: it wasn't me! just saying!09:16
LetoThe2ndRP: without digging the pseudo internals, what came to me: shouldn't inotify be able to trigger on those unexpected deletes?09:17
RPLetoThe2nd: in theory but something has to be running to listen. We'd simply run the system out of watches09:18
LetoThe2ndRP: even with all the new filters?09:18
RPLetoThe2nd: bitbake monitors just the metadata and we're pushing the boundaries of what most systems will support09:19
RPDoing that once per recipe won't scale sadly09:19
LetoThe2ndRP: hm, i think i get it. was just what came to my mind after thinking about your explanation.09:20
RPLetoThe2nd: you'd have to have something active and listening for every recipe, for every unusual entry in the database for that recipe09:21
LetoThe2ndRP: we just should use mongodb instead of a filesystem!09:27
*** florian_kc is now known as florian09:29
*** T_UNIX <T_UNIX!uid218288@gateway/web/> has joined #yocto09:30
*** gendevbot <gendevbot!~devbot@> has quit IRC09:37
*** gendevbot <gendevbot!~devbot@> has joined #yocto09:42
*** ebail <ebail!> has quit IRC09:46
*** xtron <xtron!~xtron@> has joined #yocto09:51
xtronstrange behavior, I've _append "ima" in DISTRO_FEATURES in *.conf and is set as REQUIRED_DISTRO_FEATUERS = "ima" in then why it is complaining about -> ERROR: Nothing PROVIDES 'ima-evm-utils-native'09:54
xtronima-evm-utils-native was skipped: missing required distro feature 'ima' (not in DISTRO_FEATURES)09:54
LetoThe2ndxtron: guess: missing space in front of "ima"09:55
LetoThe2nd-> append semantics09:55
xtronspace is not the issue, I think this is related to native build of the recipe09:58
xtronwhen I take environment of the ima-evm-utils-native recipe, it shows the error, non-native recipe doesn't show this error09:59
JaMayes, I think there was a change not to use DISTRO_FEATURES in native recipes (with some features used by qemu being the only exception)09:59
*** ebail <ebail!> has joined #yocto10:00
xtronJaMa, can you point to that change, if you can recall10:01
JaMaoe-core 0e1f8fa0535dab63742f575b52ec9fa891368ca2 is part of what I recall10:03
*** zandrey <zandrey!~zandrey@> has joined #yocto10:08
*** Klanticus <Klanticus!~quassel@> has joined #yocto10:10
*** mckoan <mckoan!~marco@unaffiliated/mckoan> has quit IRC10:11
*** mckoan <mckoan!~marco@unaffiliated/mckoan> has joined #yocto10:12
*** mckoan is now known as mckoan|away10:13
xtronJaMa, thanks for sharing, now checking these commits10:13
* RP wonders how much fun these 9 patches will cause10:18
* RP puts in master to get wider testing10:18
JaMaRP: :)10:20
RPI've put a link to the wiki, in the pseudo abort message so we can put up tips and other info there10:20
xtronwhat we need to do to skip REQUIRED_DISTRO_FEATURES requirement for native build, setting DISTRO_FEATURES_NATIVE = "" doesn't help10:33
xtronok, DISTRO_FEATURES_NATIVE is to set features for native build, then why REQUIRED_DISTRO_FEATURES is applying to native build, correct me if I'm wrong10:36
*** Dracos-Carazza <Dracos-Carazza!~Dracos-Ca@> has quit IRC10:37
qschulzxtron: my understanding of the docs is that DISTRO_FEATURES_NATIVE are added to REQUIRED_DISTRO_FEATURES when building for native?10:37
*** Dracos-Carazza <Dracos-Carazza!~Dracos-Ca@> has joined #yocto10:38
*** Ox861726f6c64 <Ox861726f6c64!~Ox861726f@gateway/tor-sasl/ox861726f6c64> has quit IRC10:40
JaMaRP: pseudo is already triggering ubuntu report error from /usr/bin/install, should we send those to canonical to confuse them? :)10:41
qschulzxtron: REQUIRED_DISTRO_FEATURES_class-native = "" ?10:43
xtronso if a recipe is extended to native, should we need to explicitly set the REQUIRED_DISTRO_FEAUTRES_class-native to "", or we should provide the native distro feature as well?10:44
xtronqschulz, yeah, so better approach is to set to "", as we aren't creating a native distro10:45
JaMaFYI "sudo systemctl disable apport" is now useful on ubuntu systems while pseudo issues are being worked on (apport is annoying, because it steals focus whenever it creates new report window)10:46
JaMaeither I have very good reproducer setup or very bad system :)10:47
*** ejo <ejo!2e1312b6@> has joined #yocto10:50
ejoHi! Does anyone have experience with installing binaries from a tar.gz? I'm installing a prebuilt java from bellsoft, but yocto removes some strings from the binary, some time after do_install. when running java it ends with a segfault. If I copy over a fresh java binary with scp it runs fine10:56
ejois there a way to stop Yocto from removing these strings from the binary file?10:57
JaMaejo: is it removed by stripping debug symbols in do_package? you can disable it10:57
LetoThe2ndINHIBIT_STRIP = "1" or somthing like that.10:58
ejoI'm not sure when the stripping happens, I'll look into INHIBIT_STRIP, thanks!10:59
JaMaI'm not sure what "yocto removes some strings" refers to :)10:59
ejoit removes strings from the java binary, here's an excerpt from the diff:11:01
ejo-;JFAF8F-JDK_JAVA_OPTIONS-_JAVA_LAUNCHER_DEBUG-_JAVA_OPTIONS-NOTE: Picked up %s: %s-openjdk-java-11.0.8+10-LTS--J-ms8m11:01
JaMayou can use "file" on the file in image directory and then the same file in package directory, if the later says "stripped" and looks like what you see on target, then this might be the case11:01
RPJaMa: that could be fun :/11:01
ejofile says stripped11:02
RPJaMa: is that an existing TMPDIR? this change may not react well to reusing an old TMPDIR now I think about it11:03
JaMaRP: I've around 70 failures in small image (mostly from oe-core recipes including glibc do_stash_locale), now rebuilding from clean TMPDIR11:03
JaMayes, it was existing TMPDIR, now from empty it looks a bit better (1000 tasks to go)11:04
*** nagua <nagua!> has joined #yocto11:04
JaMaRP: maybe it's wroth bumping abi_version to enforce TMPDIR cleanup?11:05
JaMaI mean OELAYOUT_ABI11:06
JaMaclean build didn't end much better, 57 tasks failed, all except qtbase failed in do_package (qtbase in do_install)11:08
JaMawill try to find some time to debug it after work11:08
naguaHey everyone, I recently experienced differences between compiling a cmake project via a yocto generated SDK vs building it with bitbake. I traced it back to poky/meta/cmake.bbclass usage of -DCMAKE_NO_SYSTEM_FROM_IMPORTED=1. This line was introduced in 79d45bf56d6 due to an issue with compatibility of cmake vs gcc. The bug was fixed in a recent cmake version, so I think this line can be removed11:10
naguaagain to align cmake usage inside the SDK and recipes.11:10
JaMaRP: there is small typo in "[PATCH 3/9] pseudo: Abort on mismatch patch" if you're changing this area later "pseudo_diag("abort()ing pseudi client by server request."11:11
naguaThe cmake issue in question:
JaMaI don't remember what "qmake -install qinstall" does internally, but now it seems to abort pseudo on every single call (e.g. in that qtbase.do_install)11:14
RPJaMa: ah, yes. I'll fix that next time I'm in there11:17
*** kiwi_29 <kiwi_29!> has joined #yocto11:17
JaMaFYI: is my old reproducer which is now 100% reliable11:18
*** georgem_home <georgem_home!uid210681@gateway/web/> has joined #yocto11:20
*** kiwi_29 <kiwi_29!> has quit IRC11:22
RPJaMa: when you say 100% reliable, you mean its failing always?11:23
JaMayes, every "qmake -install qinstall" call seems to fail in both pseudo-test recipes and now qtbase.do_install as well11:24
RPJaMa: sounds like its doing something interesting we need to debug :/11:25
JaMabefore todays changes in master it was triggering user-host-contaminated QA issues in various qmake built recipes, but pretty rarely and even the didn't trigger it always11:26
ejoSo none of the files are stripped now, but I still get a different md5sum for the file in the image directory and the file in the package directory, any ideas why?11:26
RPJaMa: now we're at least deterministic? :)11:26
JaMaI'll try to look into it after work, but I might be braindead already by then11:26
RPJaMa: I see you've put all the reproducer pieces nicely together so I can have a quick look, see if I can see the issue11:27
JaMaIf you're interested, feel free, I wanted to check and clean it a bit first (as now it doesn't need to call it 999 times to reproduce it)11:28
JaMaand I think I've more qtbase-native changes somewhere to debug what's happening internally, but couldn't find the branch with it yesterday :)11:29
*** rubdos_ is now known as rubdos11:29
JaMasome context for qmake qinstall
*** Ox861726f6c64 <Ox861726f6c64!~Ox861726f@gateway/tor-sasl/ox861726f6c64> has joined #yocto11:38
RPJaMa: I see loads of aborts during do_install which the Makefile ignores :)11:41
RPJaMa: there is something odd going on with /proc/self which reminds me of a patch on the list11:48
RPJaMa: doesn't help but was worth a shot!11:51
*** kanavin_home <kanavin_home!~ak@2a02:2450:1011:56f:b652:c378:b428:fdf> has joined #yocto11:52
*** leon-anavi <leon-anavi!~Leon@> has quit IRC11:52
*** leon-anavi <leon-anavi!~Leon@> has joined #yocto11:53
*** psnsilva <psnsilva!> has joined #yocto11:59
*** frsc <frsc!> has quit IRC12:01
*** frsc <frsc!> has joined #yocto12:18
RPJaMa: I think your test case may be broken :/12:21
RPJaMa: if I remove the mv command, I don't see any pseudo issues. The mv is quite a race nightmare as its unclear whether N+1 will have been installed in some parallel make yet or not?12:21
RPpseudo is getting confused since it can't know whether the end result is the original target file or the N+1 file12:22
JaMait was there to make it worse nightmare (when it was triggered rarely), but I assume qtbase.do_install doesn't so anything as nasty as these mv12:23
RPJaMa:  if I take the mv out, it seems to behave ok12:24
JaMathis is from qtbase.do_install earlier today
RPJaMa: what does pseudo.log say for the failure?12:25
JaMaand then number of qinstall calls matches with aborts12:25
JaMamartin@jama:/OE/build/luneos-master/webos-ports$ grep -c "qmake -install qinstall" /OE/build/luneos-master/webos-ports/tmp-glibc/work/core2-64-webos-linux/qtbase/5.15.0+gitAUTOINC+f6fe4bbab7-r0/temp/log.do_install.275690912:25
JaMamartin@jama:/OE/build/luneos-master/webos-ports$ grep -c pseudi /OE/build/luneos-master/webos-ports/tmp-glibc/work/core2-64-webos-linux/qtbase/5.15.0+gitAUTOINC+f6fe4bbab7-r0/temp/log.do_install.275690912:25
RPJaMa: now that is interesting, its req ""12:26
RPJaMa: We've seen this once already and it was host distro specific12:26
RPJaMa: the fix last time was
JaMathis is on ubuntu-20.0412:27
RPJaMa: it was 20.04 which broke last time with this too12:27
RPJaMa: can you add PSEUDO_DEBUG=nfoPcvdDyerpswikVx to the specific recipe and rerun it, then share the pseudo.log. I just need the section around one of these req "" lines12:28
RPJaMa: you can just add an export PSEUDO_DEBUG = "nfoPcvdDyerpswikVx"12:29
RPJaMa: actually, the log I need may end up in the task log12:29
RPJaMa: you can just set that export on the task thats doing it to reduce the amount of logs too12:30
*** kayterina <kayterina!kayterina-@gateway/shell/> has quit IRC12:33
*** hmw1 <hmw1!hmwmatrixo@gateway/shell/> has quit IRC12:33
*** xicopitz[m] <xicopitz[m]!xicopitzma@gateway/shell/> has quit IRC12:34
*** henriknj <henriknj!hnjematrix@gateway/shell/> has quit IRC12:34
*** nrossi1 <nrossi1!nrossimatr@gateway/shell/> has quit IRC12:34
*** clementp[m]1 <clementp[m]1!cperonmatr@gateway/shell/> has quit IRC12:34
*** silviof <silviof!silv-iomat@gateway/shell/> has quit IRC12:34
*** omar[m] <omar[m]!omarmalekm@gateway/shell/> has quit IRC12:34
*** yangm <yangm!yanyetanot@gateway/shell/> has quit IRC12:34
*** nrossi <nrossi!nrossimatr@gateway/shell/> has joined #yocto12:40
RPJaMa: basically that "" should not be there, it means the magic to translate an fd into a path hasn't worked. Previously it would use inode mappings but that is rather risky and is what we've disabled12:45
*** zandrey <zandrey!~zandrey@> has quit IRC12:49
*** maudat <maudat!> has joined #yocto12:52
*** yangm <yangm!yanyetanot@gateway/shell/> has joined #yocto12:57
*** apsychogirl[m] <apsychogirl[m]!apsychogir@gateway/shell/> has joined #yocto12:57
*** kayterina <kayterina!kayterina-@gateway/shell/> has joined #yocto12:57
*** henriknj <henriknj!hnjematrix@gateway/shell/> has joined #yocto12:57
*** silviof <silviof!silv-iomat@gateway/shell/> has joined #yocto12:57
*** clementp[m] <clementp[m]!cperonmatr@gateway/shell/> has joined #yocto12:57
*** hmw1 <hmw1!hmwmatrixo@gateway/shell/> has joined #yocto12:57
*** xicopitz[m] <xicopitz[m]!xicopitzma@gateway/shell/> has joined #yocto12:57
*** omar[m] <omar[m]!omarmalekm@gateway/shell/> has joined #yocto12:57
*** goliath <goliath!> has joined #yocto13:00
*** phlifish_ <phlifish_!~megabread@2a01:4b00:e031:2600:2879:43f8:8781:2957> has joined #yocto13:03
JaMaRP: pseudo.log and last 20K lines from log.do_install (whole file has 150M)
*** nrossi <nrossi!nrossimatr@gateway/shell/> has quit IRC13:03
*** nagua <nagua!> has quit IRC13:03
*** ebail <ebail!> has quit IRC13:03
*** megabread <megabread!~megabread@2a01:4b00:e031:2600:2879:43f8:8781:2957> has quit IRC13:03
*** mdp <mdp!sid49840@gateway/web/> has quit IRC13:03
*** LetoThe2nd <LetoThe2nd!uid453638@gateway/web/> has quit IRC13:04
*** agust <agust!> has quit IRC13:04
*** pohly <pohly!> has quit IRC13:04
*** nerdboy <nerdboy!~sarnold@gentoo/developer/nerdboy> has quit IRC13:04
*** stacktrust <stacktrust!> has quit IRC13:04
*** shoragan <shoragan!~shoragan@debian/developer/shoragan> has quit IRC13:04
*** vmeson <vmeson!> has quit IRC13:04
*** w00die <w00die!~w00die@> has quit IRC13:04
*** Bunio_FH <Bunio_FH!> has quit IRC13:04
*** cengiz_io <cengiz_io!~cengiz_io@> has quit IRC13:04
*** laurittr <laurittr!> has quit IRC13:04
*** awafaa <awafaa!sid716@gateway/web/> has quit IRC13:04
*** curlybracket <curlybracket!> has quit IRC13:04
*** tkoskine <tkoskine!> has quit IRC13:04
*** shoragan <shoragan!~shoragan@debian/developer/shoragan> has joined #yocto13:04
*** LetoThe2nd <LetoThe2nd!uid453638@gateway/web/> has joined #yocto13:04
*** cengiz_io <cengiz_io!~cengiz_io@> has joined #yocto13:04
*** mdp <mdp!sid49840@gateway/web/> has joined #yocto13:04
*** w00die <w00die!~w00die@> has joined #yocto13:04
*** stacktrust <stacktrust!> has joined #yocto13:04
*** Bunio_FH <Bunio_FH!> has joined #yocto13:05
*** dp <dp!> has quit IRC13:07
*** nagua <nagua!> has joined #yocto13:09
*** ebail <ebail!> has joined #yocto13:09
*** agust <agust!> has joined #yocto13:09
*** pohly <pohly!> has joined #yocto13:09
*** nerdboy <nerdboy!~sarnold@gentoo/developer/nerdboy> has joined #yocto13:09
*** vmeson <vmeson!> has joined #yocto13:09
*** laurittr <laurittr!> has joined #yocto13:09
*** awafaa <awafaa!sid716@gateway/web/> has joined #yocto13:09
*** curlybracket <curlybracket!> has joined #yocto13:09
*** tkoskine <tkoskine!> has joined #yocto13:09
*** PaowZ__ <PaowZ__!~Vince@> has joined #yocto13:11
*** PaowZ_ <PaowZ_!~Vince@> has quit IRC13:15
*** PaowZ__ <PaowZ__!~Vince@> has quit IRC13:17
*** stephano <stephano!> has joined #yocto13:22
*** ecdhe <ecdhe!~ecdhe@unaffiliated/ecdhe> has joined #yocto13:32
paulbarkerRP: I'm running a few builds with master-next today, I'll let you know if anything blows up13:35
*** |Sno| <|Sno|!> has quit IRC13:42
*** sno <sno!> has joined #yocto13:44
*** psnsilva_ <psnsilva_!~psnsilva@2001:818:dae7:b100:a057:d35a:f3c0:9a41> has joined #yocto13:45
*** psnsilva <psnsilva!> has quit IRC13:47
*** ericch <ericch!> has joined #yocto13:49
*** leon-anavi <leon-anavi!~Leon@> has quit IRC13:52
*** leon-anavi <leon-anavi!~Leon@> has joined #yocto13:53
*** PaowZ_ <PaowZ_!~Vince@> has joined #yocto13:56
*** sagner <sagner!> has joined #yocto13:57
*** thaytan <thaytan!> has quit IRC13:59
*** rcw <rcw!~rcwoolley@> has joined #yocto13:59
*** thaytan <thaytan!> has joined #yocto14:00
RPJaMa: interesting. I have a patch in mind you could try, seems the statx calls are breaking14:02
*** Sandrita <Sandrita!b8a36cf4@gateway/web/cgi-irc/> has joined #yocto14:02
*** andycooper <andycooper!uid246432@gateway/web/> has joined #yocto14:03
RPJaMa: - the statx bit is the bit which might help you14:05
RPpaulbarker: thanks!14:05
RPJaMa: my system doesn't have statx which may explain a few things!14:07
rburtonRP: what variables specifically are used to evaluate eg SRC_URI_qemux86 overrides14:10
rburtondevtool is building a replacement OVERRIDES without eg qemux86 in that SRC_URI is still happening14:10
RPrburton: only SRC_URI and OVERRIDES14:13
*** jobroe <jobroe!> has quit IRC14:15
JaMaRP: thanks, will try, I've added target qtbase to and now running build with 20.04 and 18.04 (before I was seing some changes in qinstall behavior due to renameat2, but statx is also detected by qtbase)14:16
RPJaMa: the patch name in the recipe is wrong btw, sorry :/14:16
rburtonRP: in which case, arghhghghgh14:24
*** sgw1 <sgw1!> has joined #yocto14:49
*** King_InuYasha is now known as Conan_Kudo14:54
*** Conan_Kudo is now known as King_InuYasha14:54
* zeddii thinks it is a rm -rf tmp kind of day14:54
tgamblina fine start to any morning14:55
LetoThe2ndzeddii: sounds like a plan!14:57
*** AndersD <AndersD!> has quit IRC14:59
*** roussinm <roussinm!> has joined #yocto15:01
zeddiiglibc, glibc-locale and now libgcc have failed. definitely time to take the rm -rf timeout.15:03
kergothi wipe tmp before at least half my builds, possibly more. sstate is my friend.15:06
* kergoth yawns15:06
*** King_InuYasha is now known as Conan_Kudo15:11
*** Conan_Kudo is now known as King_InuYasha15:11
*** JonathanCrockett <JonathanCrockett!> has joined #yocto15:21
*** frsc <frsc!> has quit IRC15:23
JPEWkergoth: ;)15:27
kergothhuh. rsync is faster?15:28
kergothI use so i can kick off the next build while the previous is being removed, i don't mind the performance impact, just want to get the next one going15:29
JPEWkergoth: Much faster AFAICT15:31
JPEWI think rm -rf does some sort of directory listing before deleting15:31
JPEWrsync just compares files as they come15:31
*** nagua <nagua!> has quit IRC15:32
LetoThe2ndwhats a good way to remove init systems completely? is INIT_MANAGER = "" a thing?15:32
JaMaRP: qtbase.do_install failed on 20.04 and didn't fail on 18.04 ubuntu, with your statxfix.patch (and proc.patch) added it built for 20.04 as well now15:33
kergothJPEW: very interesting, thanks, i'll try that15:33
RPJaMa: ok, cool. My guess about the issue was right then :)15:35
RPJaMa: I'll get that cleaned up and queued15:35
*** frsc <frsc!> has joined #yocto15:36
*** rcoote <rcoote!> has quit IRC15:38
qschulzLetoThe2nd: rm -rf /sbin/init in ROOTFS_POSTPROCESS_COMMAND15:40
LetoThe2ndqschulz: hehehe15:40
*** Dracos-Carazza_ <Dracos-Carazza_!~Dracos-Ca@> has joined #yocto15:44
*** Dracos-Carazza <Dracos-Carazza!~Dracos-Ca@> has quit IRC15:45
*** leon-anavi <leon-anavi!~Leon@> has quit IRC15:52
*** leon-anavi <leon-anavi!~Leon@> has joined #yocto15:53
*** ejo <ejo!2e1312b6@> has quit IRC15:55
JPEWLetoThe2nd: It seems like there should be *some* way to do that for e.g. an initrd15:56
*** kpo_ <kpo_!> has quit IRC15:56
*** kpo_ <kpo_!> has joined #yocto15:57
*** jmiehe <jmiehe!> has quit IRC16:02
*** Konsgnx <Konsgnx!> has joined #yocto16:04
*** ebail <ebail!> has quit IRC16:06
*** Delphine34 <Delphine34!81006503@> has joined #yocto16:12
Delphine34Hello everyone16:14
*** rubdos <rubdos!~rubdos@2a02:578:859d:700:8b44:5716:382d:a7da> has quit IRC16:15
*** rubdos <rubdos!~rubdos@2a02:578:859d:700:8b44:5716:382d:a7da> has joined #yocto16:16
*** manuel1985 <manuel1985!> has quit IRC16:16
*** chris_ber <chris_ber!~quassel@> has quit IRC16:17
RPJaMa: those are now in master-next16:20
RPJPEW, kergoth: I can see a case for adding something like this to bitbake...16:22
RPWe have a janitor process in ab-helper which does this too16:22
*** florian <florian!~florian_k@Maemo/community/contributor/florian> has quit IRC16:22
JaMaRP: thanks, is there a reason why these patches aren't in oe-core branch in pseudo repo?16:23
RPJaMa: just testing atm really, I was just thinking about that16:24
qschulzDelphine34: o/16:24
RPJaMa: I'll probably add them soon16:24
*** fl0v0 <fl0v0!~fvo@> has quit IRC16:25
JaMathe image build seems much better with this, there are some failures, but it got to building qtwebengine, so it's taking significantly longer this time :)16:26
qschulzI'm debugging swupdate eating up all my RAM rn... The first thing is Ive a 300MB rootfs but because of IMAGE_OVERHEAD_FACTOR being 1.3, my ext4 ends up being 400MB big... I've a whole partition on eMMC for that ext4... do I actually need this factor to be anything else than 1.0?16:27
RPJaMa: I guess I wanted to post the patches for wider review before adding to the branch (which they are getting now)16:27
JaMaand I just jinxed it with 58 failed tasks16:28
qschulzafk til tmrw but if anyone has some tips on this IMAGE_OVERHEAD_FACTOR thingy, i'll read it tomorrow morning :)16:30
frayqschulz block size can affect this, which is one of the reasons for the overhead16:31
frayi.e. if you have a disk with 10 million 1 byte files..  it's not 10 MB.. but the system will likely need a lot more16:31
frayblock size or inode size is what I really meant16:32
*** manuel1985 <manuel1985!> has joined #yocto16:32
JaMaRP: a lot of failures are do_package error 134, but log.do_package doesn't show anything and pseudo.log has just one mismatch16:40
JaMapath mismatch [651 links]: ino 96733650 db '/OE/build/luneos-master/webos-ports/tmp-glibc/work/core2-64-webos-linux/glibc-locale/2.32-r0/image/usr/share/licenses/glibc-locale/generic_GPLv2' req '/OE/build/luneos-master/webos-ports/openembedded-core/meta/files/common-licenses/GPL-2.0'.16:40
*** feddischson <feddischson!> has quit IRC16:40
RPJaMa: ah. That is probably a path we need to add to PSEUDO_IGNORE_PATHS16:42
JaMait might be related to LICENSE_CREATE_PACKAGE functionality as I don't see it with nodistro builds I think16:43
*** sagner <sagner!> has quit IRC16:44
RPJaMa: very likely. I'd like to understand which code is writing to meta/files/common-licenses16:44
*** creich <creich!> has quit IRC16:45
yannI'm trying to reproduce the problem with llvm-config when using multilib, using qemu_x86-64 target, but just adding "require conf/multilib.conf" to local.conf triggers "No tuning found for lib32 multilib", any idea where the problem is ?16:45
*** creich <creich!> has joined #yocto16:46
RPJaMa: actually, no, its the question of what wrote the other file, which would be the LICENSE_CREATE_PACKAGE code16:46
*** kiwi_29 <kiwi_29!> has joined #yocto16:56
*** vineela <vineela!~vtummala@> has joined #yocto17:01
*** xtron <xtron!~xtron@> has quit IRC17:05
*** geheimnis` <geheimnis`!~geheimnis@> has quit IRC17:14
*** geheimnis` <geheimnis`!~geheimnis@> has joined #yocto17:20
*** Delphine34 <Delphine34!81006503@> has quit IRC17:23
*** manuel1985 <manuel1985!> has quit IRC17:26
*** Klox04809 <Klox04809!> has quit IRC17:31
*** Klox04809 <Klox04809!> has joined #yocto17:32
*** kpo_ <kpo_!> has quit IRC17:36
*** Ida <Ida!9a489981@> has joined #yocto17:36
*** kpo_ <kpo_!> has joined #yocto17:36
*** Klox04809 <Klox04809!> has quit IRC17:39
*** mihai- <mihai-!~mihai@unaffiliated/mihai> has quit IRC17:40
*** Klox04809 <Klox04809!> has joined #yocto17:43
*** psnsilva_ <psnsilva_!~psnsilva@2001:818:dae7:b100:a057:d35a:f3c0:9a41> has quit IRC17:46
armpitoooh, i got an abort17:47
*** psnsilva <psnsilva!~psnsilva@2001:818:dae7:b100:a057:d35a:f3c0:9a41> has joined #yocto17:50
JaMaRP: with LICENSE_CREATE_PACKAGE disabled was able to build luneos image now; with webosose I was able to still reproduce abort with webruntime (chromium) build but that already had issues with host-user-contaminated so it was probably doing something fishy already17:51
*** leon-anavi <leon-anavi!~Leon@> has quit IRC17:52
*** leon-anavi <leon-anavi!~Leon@> has joined #yocto17:53
JPEWRP, kergoth: Ya. We have a local script that does a clean of the build directory, excluding a few key directories (e.g. "conf"). It uses rsync, and even has an option to do the clean in the background so you can start a new build right away while it's cleaning17:57
*** manuel1985 <manuel1985!> has joined #yocto17:57
*** phlifish_ <phlifish_!~megabread@2a01:4b00:e031:2600:2879:43f8:8781:2957> has quit IRC17:58
*** linums <linums!~linums@> has joined #yocto18:08
linumsHi guys, do any of you have experience how to launch core-image-weston?18:09
linumsI can't get it to start weston18:09
khemqschulz: you only need overhead factor if you need to generate some files etc.18:09
khemlinums: it should launch weston on boot, describe what you are doing18:10
*** manuel1985 <manuel1985!> has quit IRC18:10
linumsI just simply tried to bake the core-image-weston (since my own recipe was not starting as expected)18:11
linumsAnd this here does the same ,that after weston launch starts18:11
linumsI get the no drm device found error18:11
khemlinums: what machine are you building for18:12
linumsThis is an x86-60 fujitsu machine18:12
linumsWith an lvds screen on it18:12
khemDoes it work if you build for qemux86-64 ?18:13
khemno drm device found, seems to convey that your graphics driver or mesa is not compiled correctly perhaps18:14
*** linums <linums!~linums@> has quit IRC18:17
*** linums <linums!> has joined #yocto18:19
linumsI've built x11 image for this machine and that worked18:21
linumsWith my image18:21
linumsIs this not the same driver in the background?18:22
khemI dont know18:22
JPEWlinums: Do you have any DRI nodes in /dev/dri/ ?18:22
*** ssajal <ssajal!> has joined #yocto18:23
linumsWell, I don't think so18:23
*** linums <linums!> has quit IRC18:23
*** linums <linums!~linums@> has joined #yocto18:24
JPEWlinums: Usually /dev/dri/card0 .... I;m not sure how X works, but weston requires DRM/KMS support in the kernel18:25
*** sgw <sgw!> has joined #yocto18:25
*** beneth` <beneth`!~beneth@2001:41d0:c:a71:1000:22::> has left #yocto18:25
*** sgw1 <sgw1!> has quit IRC18:27
*** linums <linums!~linums@> has quit IRC18:28
*** manuel1985 <manuel1985!> has joined #yocto18:28
*** linums <linums!> has joined #yocto18:29
linumsSo, I have no dri directory18:29
linumsWhat does it mean18:29
linums-sorry somehow I was disconnected-18:29
JPEWlinums: You need to enabled DRM/KMS support for your GPU in the kernel (if available)18:30
linumsi thought that such modification is done in the image core-image-wayland :(18:30
linumsThis is the reason why I've tried that one18:30
JPEWWell, different GPU have different drivers, and I don't think we enable them all18:31
*** vineela <vineela!~vtummala@> has quit IRC18:32
linumsBut don't than how come x11 work?18:33
*** zkrx <zkrx!> has quit IRC18:33
JPEWMaybe it's using an EGL acclerated framebuffer?18:33
JPEWe.g. /dev/fb018:33
JPEWI think you can run weston on the framebuffer, but I don't know if it will have GPU acceleration if you do that18:34
*** LetoThe2nd <LetoThe2nd!uid453638@gateway/web/> has quit IRC18:34
*** T_UNIX <T_UNIX!uid218288@gateway/web/> has quit IRC18:35
*** zkrx <zkrx!> has joined #yocto18:37
*** frsc <frsc!> has quit IRC18:51
*** linums <linums!> has quit IRC18:56
*** linums <linums!~linums@> has joined #yocto18:56
linumsJPEW: I guess than I have to add some driver to my image to build19:01
linumsI know that under ubuntu distro this machine runs wayland19:01
linumsSo it is capable, so I have some misconfiguration19:02
*** Ida <Ida!9a489981@> has quit IRC19:04
*** linums <linums!~linums@> has quit IRC19:09
*** linums <linums!> has joined #yocto19:10
*** lxc <lxc!> has joined #yocto19:12
lxcwhat would be the way to disable building of a bootloader? For kernel it seems as I can do PREFERRED_PROVIDER_virtual/kernel ?= "linux-dummy" but what about boot?19:12
*** sagner <sagner!~ags@2a02:169:3df5::edf> has joined #yocto19:14
*** beneth` <beneth`!~beneth@2001:41d0:c:a71:1000:22::> has joined #yocto19:18
*** Ida <Ida!81004ec3@> has joined #yocto19:24
*** Ida <Ida!81004ec3@> has quit IRC19:31
*** vineela <vineela!vtummala@nat/intel/x-bhhqdjoijprxmqoy> has joined #yocto19:31
rewittlxc: Why do you not want a kernel or bootloader? Just curious.19:35
*** dreyna <dreyna!> has joined #yocto19:36
*** manuel1985 <manuel1985!> has quit IRC19:43
*** leon-anavi <leon-anavi!~Leon@> has quit IRC19:45
*** leon-anavi <leon-anavi!~Leon@> has joined #yocto19:45
*** psnsilva <psnsilva!~psnsilva@2001:818:dae7:b100:a057:d35a:f3c0:9a41> has quit IRC19:47
*** psnsilva <psnsilva!~psnsilva@2001:818:dae7:b100:a057:d35a:f3c0:9a41> has joined #yocto19:47
*** linums <linums!> has quit IRC19:48
*** manuel1985 <manuel1985!> has joined #yocto19:50
*** linums <linums!> has joined #yocto19:51
*** kiwi_29 <kiwi_29!> has quit IRC19:52
JPEWHmm, SERIAL_CONSOLES_CHECK breaks builds when using read-only rootfs....19:55
*** leon-anavi <leon-anavi!~Leon@> has quit IRC19:56
*** Konsgnx <Konsgnx!> has quit IRC19:56
*** blauskaerm <blauskaerm!~blauskaer@> has joined #yocto19:57
zeddiievery time I add a new go package, it makes me sad20:02
* zeddii starts jumping through hoops20:02
*** sagner <sagner!~ags@2a02:169:3df5::edf> has quit IRC20:03
* armpit time to go20:07
zeddiiis my sadness too much for you!?20:07
* zeddii sobs20:07
* zeddii starts symlink'ing random directories together until it admits the code is where it belongs20:08
armpitnice.. using a bigger hammer20:08
JPEWzeddii: symlinks are so old fashioned.... bind-mounts!20:14
zeddiidang. this application doesn't have a vendor/ directory with it's dependencies. this is going to be painful.20:17
zeddiimy symlink quest has failed.20:17
* zeddii cries a bit more20:18
* paulg hands zeddii a bag of slightly used symlinks20:22
RPpaulg: we're good at creating those in significant excess :)20:24
manuel1985Which task does `bitbake <image-name>` run, when I don't specify one through the `-c` switch?20:25
neverpanicmanuel1985: "build"20:26
*** dreyna_ <dreyna_!> has joined #yocto20:30
*** rcwoolley <rcwoolley!~rcwoolley@> has joined #yocto20:30
kergothmanuel1985: see also BB_DEFAULT_TASK in meta/conf/bitbake.conf20:31
*** junland_ <junland_!~junland@> has joined #yocto20:31
*** junland <junland!~junland@> has quit IRC20:31
*** ndec <ndec!sid219321@linaro/ndec> has quit IRC20:31
*** ndec <ndec!sid219321@linaro/ndec> has joined #yocto20:32
*** dreyna <dreyna!> has quit IRC20:33
*** rcw <rcw!~rcwoolley@> has quit IRC20:33
khemkergoth: if I change BB_DEFAULT_TASK to say do_coverity scan, will it then run this task and its deps only for every recipe in image depchain ?20:35
RPkhem: as long as do_coverity depends on do_build20:36
kergothit'll run that task for the recipe or recipes you specify, not their deps, unless that task is configured to recrdeptask20:36
kergoth--runall= may be a better bet20:37
paulbarkerkhem: You'd want `bitbake <thing> --runall=do_coverity_scan` for that I believe. Changing the default task wouldn't cause it to run recursively20:37
manuel1985I'm just watching one of LetoThe2nds tutorials where he installs a simpleechoserver in his image. It all works nicely when he runs qemu without slirp, but with slirp the simpleechoserver doesn't respond. I'm experiencing the same here right now. Are there some common pitfalls when using slirp?20:37
RPkergoth: was just looking up that flag :)20:37
khemyeah runall is what I use20:37
RPkhem: I'd second looking at runall20:37
paulbarkerRP: master-next builds are good so far. I've tested with meta-sancloud/meta-ti/meta-arm/etc included, changed DISTRO and triggered a rebuild, cleaned glibc to force most things to be rebuilt20:38
paulbarkerI'm going to leave a world build running overnight, I expect many failures but hopefully none due to pseudo aborts20:38
RPpaulbarker: sounds good :)20:38
RPSadly the autobuilder has lit up red :(20:40
paulbarkerIt's also a good test of my freshly re-installed build machine. Gotta put that 2TB NMVe drive through its paces20:40
RPso there goes any chance I thought of a break :(20:40
*** lxc <lxc!> has quit IRC20:43
*** rob_w <rob_w!~rob@unaffiliated/rob-w/x-1112029> has quit IRC20:47
RPAh, its that /proc patch20:47
ecdheDoes anyone use docker containers for multiple projects with a shared cache?20:47
RPssajal: around?20:48
ecdheHow disimilar to the projects have to be before it doesn't make sense to share the cache any more?20:48
paulbarkerecdhe: I tend to split caches based on DISTRO and on Yocto Project branch. I do use docker/podman for builds and just bind mount in the shared cache & downloads directories20:49
JPEWecdhe: sstate cache? We use docker container for build and share one sstate20:49
kergothi use a single sstate cache across the board and clear out every week or two based on last access time20:49
kergothi.e. sstates older than a week get wiped20:50
JPEWYa, we have a cron job to clear ours every friday. More so that it doesn't grow unbounded than anything20:50
RPecdhe: you can share one across everything, its preference and ease of cleaning more than anything20:50
JPEWecdhe: Even if the only thing you can share are -native things, it's still helpful20:51
RPTo be really clear with everyone, the proc patch for pseudo in master-next appears to be "toxic" somehow with my recent changes20:51
kergoththe only time i limit my sstate is if i'm working on something that invalidates everything, and i know it'll change multiple times as i work, then i isolate to a local SSTATE_DIR but still pull from the shared one with SSTATE_MIRRORS20:51
*** ssajal <ssajal!> has quit IRC20:51
RPkergoth: there are even selftests which do that :)20:52
RPkergoth: one broke with the pseudo changes, that is how I found the cache invalidation issue20:52
kergothi also force SSTATE_DIR to an empty dir under tmpdir when i'm using -S printdiff to check changes rather than compare against sstate, to force it to only use STAMPS_DIR for comparison20:52
* armpit what if i want to run the odd selftests instead of the even ones20:53
RPkergoth: I wish we could find a way to make that code do what users actually want20:53
RParmpit: the odd ones are specially for you ;-)20:54
kergothi think part of the issue is that it does two jobs, sort of. "why isn't sstate being used" and "what did my change just do" aren't always one and the same, the latter is not unlike what bitbake-whatchanged just did.. of course, the former has other annoyances when mirrors are involved20:54
RPkergoth: we should therefore perhaps split it into two modes...20:55
kergothmight make the behavior more consistent at least20:55
*** beneth` <beneth`!~beneth@2001:41d0:c:a71:1000:22::> has left #yocto20:57
*** ssajal <ssajal!> has joined #yocto20:59
*** pohly <pohly!> has quit IRC20:59
*** linums <linums!> has quit IRC21:02
*** kiwi_29 <kiwi_29!> has joined #yocto21:03
*** linums <linums!~linums@> has joined #yocto21:03
*** kiwi_29 <kiwi_29!> has quit IRC21:07
*** psnsilva <psnsilva!~psnsilva@2001:818:dae7:b100:a057:d35a:f3c0:9a41> has quit IRC21:16
*** psnsilva <psnsilva!~psnsilva@2001:818:dae7:b100:a057:d35a:f3c0:9a41> has joined #yocto21:16
manuel1985Hmmm... Can one set MACHINE to multiple values?21:17
JPEWmanuel1985: I don't think so21:17
kergothmanuel1985: nope. if you want to build for multiple machines, use multiconfig. if you want to make one machine inherit from another, use require + MACHINEOVERRIDES21:19
*** psnsilva_ <psnsilva_!~psnsilva@2001:818:dae7:b100:a057:d35a:f3c0:9a41> has joined #yocto21:19
*** psnsilva <psnsilva!~psnsilva@2001:818:dae7:b100:a057:d35a:f3c0:9a41> has quit IRC21:22
manuel1985JPEW, kergoth: Thanks21:22
*** asteriusio <asteriusio!> has quit IRC21:26
*** asteriusio <asteriusio!> has joined #yocto21:31
*** linums <linums!~linums@> has quit IRC21:33
*** linums <linums!~linums@> has joined #yocto21:33
*** maudat <maudat!> has quit IRC21:35
RPVictor's speedup for 32 bit mips is ingenious21:43
JPEWHah, thats is fantastic21:45
RPok, we have to invent some hardware but for the speed gain, I don't care :)21:50
*** Ox861726f6c64 <Ox861726f6c64!~Ox861726f@gateway/tor-sasl/ox861726f6c64> has quit IRC21:52
*** linums <linums!~linums@> has quit IRC22:04
*** kamel <kamel!> has quit IRC22:04
*** linums <linums!> has joined #yocto22:05
*** roussinm <roussinm!> has quit IRC22:06
*** mrc3 <mrc3!~mrc3@linaro/mrc3> has quit IRC22:06
*** iokill <iokill!> has quit IRC22:06
*** Saur <Saur!pkj@nat/axis/x-fnprpckyscifdmgp> has quit IRC22:06
*** iokill <iokill!> has joined #yocto22:07
*** roussinm <roussinm!> has joined #yocto22:07
*** mrc3 <mrc3!~mrc3@linaro/mrc3> has joined #yocto22:08
*** radsquirrel <radsquirrel!> has quit IRC22:09
*** eLmankku <eLmankku!> has quit IRC22:09
*** radsquirrel <radsquirrel!> has joined #yocto22:09
*** Saur <Saur!pkj@nat/axis/x-hocnrpuviscwizbb> has joined #yocto22:10
*** kamel_b <kamel_b!> has joined #yocto22:10
*** eLmankku <eLmankku!> has joined #yocto22:10
*** agust <agust!> has quit IRC22:11
*** tkoskine <tkoskine!> has quit IRC22:25
*** psnsilva_ <psnsilva_!~psnsilva@2001:818:dae7:b100:a057:d35a:f3c0:9a41> has quit IRC22:29
*** psnsilva_ <psnsilva_!~psnsilva@2001:818:dae7:b100:a057:d35a:f3c0:9a41> has joined #yocto22:29
*** psnsilva_ <psnsilva_!~psnsilva@2001:818:dae7:b100:a057:d35a:f3c0:9a41> has quit IRC22:30
*** psnsilva <psnsilva!~psnsilva@2001:818:dae7:b100:a057:d35a:f3c0:9a41> has joined #yocto22:31
*** tkoskine <tkoskine!> has joined #yocto22:46
*** ericch <ericch!> has quit IRC23:04
kergothJPEW: Any objection to changing the bindir of cross-canadian for mingw to avoid the subdir under bindir? Placing them under a subdir means they fail to find the .dlls that live in the root of bindir.23:07
*** Sandrita <Sandrita!b8a36cf4@gateway/web/cgi-irc/> has quit IRC23:09
*** Klanticus <Klanticus!~quassel@> has quit IRC23:13
* armpit open a bug ; )23:21
*** stephano <stephano!> has quit IRC23:22
*** rangergord <rangergord!rangergord@gateway/shell/xshellz/x-xypjtmpjklwmhzgj> has quit IRC23:25
*** rangergord <rangergord!rangergord@gateway/shell/xshellz/x-dfznvnbotmmxfvca> has joined #yocto23:30
*** linums <linums!> has quit IRC23:39
*** linums <linums!~linums@> has joined #yocto23:39

Generated by 2.17.2 by Marius Gedminas - find it at!