dl9pf changed the topic of #yocto to: Welcome to the Yocto Project | Learn more: http://www.yoctoproject.org | Join the community: http://www.yoctoproject.org/community | Channel logs available at https://www.yoctoproject.org/irc/ and https://libera.irclog.whitequark.org/yocto/ | Having difficulty on the list, or with someone on the list? Contact YP community mgr Nicolas Dechesne (ndec)
Spooster has quit [Remote host closed the connection]
<yates> abelloni: thanks, but the question is, which recipe? i'm builing core-image-minimal. do i create a .bbappend and add it there? what if i want it for all images? is there a better spot?
<yates> i.e., a common place where all recipes will pick it up?
<yates> i've made my own kernel recipe - should it be there? (seems not)
<yates> it is tied to the kernel and architecture
goliath has quit [Quit: SIGSEGV]
jpuhlman_ has joined #yocto
jpuhlman__ has quit [Ping timeout: 252 seconds]
WadeBerrier[m] has joined #yocto
yates has quit [Ping timeout: 265 seconds]
camus has joined #yocto
otavio has quit [Remote host closed the connection]
sakoman has quit [Quit: Leaving.]
vmeson has quit [Ping timeout: 240 seconds]
vmeson has joined #yocto
Spooster has joined #yocto
Spooster has quit [Remote host closed the connection]
mattsm has quit [Killed (copper.libera.chat (Nickname regained by services))]
mattsm has joined #yocto
mranostaj has quit [Remote host closed the connection]
<alex88> ya
<alex88> wops
vmeson has quit [*.net *.split]
WadeBerrier[m] has quit [*.net *.split]
xmn has quit [*.net *.split]
vmeson has joined #yocto
WadeBerrier[m] has joined #yocto
paulg has quit [Ping timeout: 252 seconds]
mranostaj has joined #yocto
dtometzki has quit [Ping timeout: 240 seconds]
ChanServ has quit [shutting down]
davidinux has joined #yocto
kpo_ has quit [Read error: Connection reset by peer]
kpo_ has joined #yocto
ChanServ has joined #yocto
rob_w has joined #yocto
zyga-mbp has joined #yocto
florian_kc has joined #yocto
mckoan|away is now known as mckoan
<mckoan> good morning
LetoThe2nd has joined #yocto
<LetoThe2nd> yo dudx
zpfvo has joined #yocto
<mckoan> LetoThe2nd: hey!
leon-anavi has joined #yocto
ant_ has joined #yocto
ant_ has quit [Ping timeout: 268 seconds]
pbergin has joined #yocto
<LetoThe2nd> mckoan: howdy
tperrot has quit [Quit: leaving]
tprrt has joined #yocto
tprrt has quit [Client Quit]
tprrt has joined #yocto
tprrt is now known as tperrot
goliath has joined #yocto
florian_kc is now known as florian
tnovotny has joined #yocto
florian has quit [Read error: Connection reset by peer]
florian__ has joined #yocto
florian__ is now known as florian
pbergin has quit [Ping timeout: 268 seconds]
JaMa has joined #yocto
ilunev has joined #yocto
kpo_ has quit [Read error: Connection reset by peer]
pbergin has joined #yocto
florian_kc has joined #yocto
florian__ has joined #yocto
florian has quit [Read error: Connection reset by peer]
florian_kc has quit [Ping timeout: 252 seconds]
<fabatera[m]> <v0n "How do I install the initramfs i"> I guess you want it within the kernel image. https://www.yoctoproject.org/docs/3.1.2/mega-manual/mega-manual.html#var-INITRAMFS_IMAGE_BUNDLE
pbergin has quit [Ping timeout: 268 seconds]
<fabatera[m]> Hi guys, I'm not sure but this looks wrong:
<fabatera[m]> > grep INITRAMFS_IMAGE -r poky/meta/lib/oeqa/selftest/cases/imagefeatures.py
<fabatera[m]> INITRAMFS_IMAGE = "core-image-minimal"
<fabatera[m]> Shouldn't it be "core-image-minimal-initramfs" ?
prabhakarlad has joined #yocto
<RP> rburton: any interest in trying to fix the hanging ltp test that only happens on arm?
<rburton> grrr
<rburton> link?
<RP> rburton: it was after http://git.yoctoproject.org/cgit.cgi/poky/commit/?id=99c023029100528cb2a2ccf4ce0c9c6c4bbe2055 the proc01 test hangs again. I think abelloni broke the hanging build doing some debugging
<RP> rburton: I suspect if we run a few builds we'd get another hang
<RP> rburton: the question is whether we disable it or try and fix it
<RP> we'd wondered if it was the other bug but probably not that one. The x86 one probably was
vmeson has quit [Ping timeout: 265 seconds]
<RP> rburton: https://github.com/linux-test-project/ltp/commits/master/testcases/kernel/fs/proc looks a bit messy. I suspect cat /proc/kmsg > /dev/null on qemuarm64 can probably reproduce
<RP> rburton: I set a couple of builds away on the infra and a local build running to see if I could see anything
michael has joined #yocto
vmeson has joined #yocto
michael has quit [Quit: Leaving]
davidinux has quit [Ping timeout: 258 seconds]
davidinux has joined #yocto
Fanfwe has quit [Ping timeout: 268 seconds]
Fanfwe has joined #yocto
<override> can anyone help me understand whats up with devtool https://paste.ee/p/apKkp
pbergin has joined #yocto
Guest28 has joined #yocto
georgem has joined #yocto
florian__ is now known as florian
<RP> rburton: confirmed that building a qemuarm64 core-image-sato, booting it and running "cat /proc/kmsg > /dev/null" hangs
<mckoan> override: devtool (and recipetool) is not always able to do everyting, you should continue editing your recipe manually then
<mckoan> s/everyting/everything
<override> mckoan - think it hands up before giving me a general template to edit even. Ill look into it a little more
<mckoan> override: you have it in build/workspace/recipes/python3-uvicorn/python3-uvicorn_0.14.0.bb
rob_w has quit [Remote host closed the connection]
<Guest28> opps. ok I'll check it out in that case
<override> sorry - I'm setting up irssi and would get rid of guest28 when Im done..
pbergin has quit [Ping timeout: 250 seconds]
bps has joined #yocto
paulg has joined #yocto
<rburton> RP: that's... special
<rburton> RP: kernel or cat hanging?
Guest28 has quit [Quit: Connection closed]
<paulg> poor cat.
<paulg> it was probably just doing its constitutional duty.
LetoThe2nd has quit [Quit: Connection closed for inactivity]
<RP> rburton: not sure but it should at least be debuggable. I was trying to get strace on it but not there yet
<rburton> yeah, that's a nice small test case
<rburton> i'll take it once i've dug into this mte thing, just need enough info to hand over to the toolchain people
rob_w has joined #yocto
Spooster has joined #yocto
pbergin has joined #yocto
<v0n> abelloni: everything's bundled inside the rootfs in /boot, so I'd put the initramfs image there as well
angolini has joined #yocto
prabhakarlad has quit [Quit: Client closed]
halstead has quit [Quit: Leaving]
halstead has joined #yocto
<angolini> Hi, I have a patch for wireguard-module that might be useful upstream (this is the recipe http://cgit.openembedded.org/meta-openembedded/tree/meta-networking/recipes-kernel/wireguard/wireguard-module_1.0.20210219.bb) could you, please, confirm which would be the mailing list to send it? Thanks in advance
<RP> rburton: FWIW strace shows it hanging in read() and never gets any data
<override> any python recipe writing folks around?
pbergin has quit [Ping timeout: 265 seconds]
<RP> rburton: my desktop does that too so there is something more to this :/
sakoman has joined #yocto
<vmeson> angolini: wireguard is in meta-networking: https://layers.openembedded.org/layerindex/branch/master/recipes/?q=wireguard so patches go to openembedded-devel@lists.openembedded.org
<angolini> thanks!
<RP> rburton: might be non-blocking isn't working
<rburton> RP: best kernel testing ever
<vmeson> override: lots of people here are "python recipe writing folks", do you have a follow-up question? ;-)
rob_w has quit [Quit: Leaving]
prabhakarlad has joined #yocto
<override> vmeson, trying to write a recipe for this module called uvicorn, and having a hard time finding packages for its rdepends
<override> does that make sense?
<override> vmeson:
LetoThe2nd has joined #yocto
<vmeson> I'm not familiar with these packages but the seem to be similar: https://stackshare.io/stackups/gunicorn-vs-uvicorn
<vmeson> (or maybe not...)
<vmeson> override: anyway, if there are missing dependencies, you'll just have to add them to meta-python or elsewhere. Do you know about: https://layers.openembedded.org/layerindex/branch/master/recipes/
jsbronder has quit [Quit: WeeChat 2.9]
vmeson has quit [Quit: Konversation terminated!]
jsbronder has joined #yocto
<override> shoot - just when I had vmenson;s reply typed out they quit
vmeson has joined #yocto
<override> wb vmeson
<override> think id have to include uvicorn
<override> but what sucks is that I have to get gunicorn in too becuase I saw some other module that depends on it..
florian has quit [Ping timeout: 250 seconds]
gourve_l has quit [Ping timeout: 252 seconds]
florian has joined #yocto
elfenix has quit [Remote host closed the connection]
MysticMice has joined #yocto
<MysticMice> Hey, all. I am trying to setup a ADS7846 based touch controller, targeting a RPi4. When building rpi's kernel, I don't really see the module there - i'd expect a ads7846.c with actual driver code. When I look into the tree, I can see a ads7846.mod.c that doesn't really resemble a actual linux driver.
wesm has joined #yocto
<MysticMice> Not sure if this is related to the meta for the rpi's kernel (the traditional meta-raspberrypi), but, i'd expect to see something like a mainline kernel tree.
<mckoan> MysticMice: did you run make modules ?
<MysticMice> Am I wrong to make such assuption?
<MysticMice> @mckoan no, nothing expect setting up the metas and running the target image
hpsy has joined #yocto
<mckoan> MysticMice: you need to set the MACHINE_ESSENTIAL_EXTRA_RRECOMMENDS variable to include kernel modules
<mckoan> MysticMice: or try MACHINE_EXTRA_RRECOMMENDS += "kernel-module-ads7846"
<MysticMice> mckoan alright, doing it. Should be added to something like local.conf and rerun bitbake, or do I need a more forceful build for it?
<zedd> of course, if that module isn't configured in the rpi kernel config you are building, that rrecommend won't do anything.
<mckoan> MysticMice: the right place is the machine fine, however a quick test can be done editing local.conf
<mckoan> s/fine/file
ilunev has quit [Quit: Textual IRC Client: www.textualapp.com]
tnovotny has quit [Quit: Leaving]
<MysticMice> I see. The weird is that the ads7846.ko is generated
davidinux has quit [Ping timeout: 252 seconds]
gourve_l has joined #yocto
<Ch^W> khem: We upgraded our SDK to be based on hardknott (and gcc 10) and now we get consistent build failures on gobject-introspection. Not sure what else is going to break after that, but gobject-introspection does not appear build any longer.
<Ch^W> It builds perfectly fine under gcc 8.
<Ch^W> I also tried with buildtools-extended in case it was something I was doing wrong with the way we built gcc 10. Still same result.
<paulg> zedd, I checked out v5.4 for a trial run, and it doesn't even have the files for https://lists.yoctoproject.org/g/linux-yocto/message/9988 so I'm assuming that means "not applicable" on that baseline...
<zedd> yah. I'd say so.
goliath has quit [Quit: SIGSEGV]
<zedd> any idea if they've been send to greg for upstream -stable inclusion ?
prabhakarlad has quit [Quit: Client closed]
<paulg> "They aren't (yet?) a part of linux-stable but I did let stable know that
<paulg> they backport to v5.10 essentially hands free. "
* paulg notes zedd doesn't read past the 1st sentence.
<paulg> translation - not in stable-queue ; I let sashal know on IRC but didn't hear anything back
<paulg> maybe they are gun shy about anything that adds a new Kconfig? Given the other stuff that gets backported, I seriously doubt it though.
<paulg> in any case it has been hanging around forever and I got sick of looking at the turd.
halstead_ has joined #yocto
pbergin has joined #yocto
florian_kc has joined #yocto
<zedd> hah. I never look at 0/n first, I look at the patches, and then see if 0/N matches my thoughts on them :D
davidinux has joined #yocto
florian has quit [Ping timeout: 252 seconds]
zpfvo has quit [Remote host closed the connection]
davidinux has quit [Client Quit]
davidinux has joined #yocto
Guest71 has joined #yocto
<moto-timo> kanavin: the alternative to the patch to test for 'tools-sdk' in IMAGE_FEATURES is going to be completely dropping those two test cases...
<moto-timo> kanavin: I think the only reason they have been passing is that all the ptest images on AB have had 'tools-sdk' set (globally)
<moto-timo> kanavin: but in effect, the patch will drop those two tests anyway, right?
<moto-timo> s/drop/skip/
meego has joined #yocto
mckoan is now known as mckoan|away
<khem> Ch^W: what errors do you see ? and what is your build host OS
rber|res has joined #yocto
Guest71 has quit [Quit: Client closed]
hpsy1 has joined #yocto
prabhakarlad has joined #yocto
<override> wow, this taking care of python dependencies is no easy task. Does yocto generally not promote using python for embedded systems, or is it just the module im after (uvicorn) ?
hpsy has quit [Ping timeout: 252 seconds]
<rburton> yocto promotes you actually knowing what software is being used, and not just running pip on the target
<rburton> when you ship something and need to produce a license manifest or rebuild an identical image a year later, you'll appreciate it
MysticMice_ has joined #yocto
<override> when I cant find a package using pypi or the oe-pkgdata-util where else can I look at?
MysticMice_ has quit [Client Quit]
MysticMice has quit [Quit: Going offline, see ya! (www.adiirc.com)]
hpsy1 has quit [Ping timeout: 265 seconds]
<rburton> if its not in pypi then i'd worry!
<rburton> layers.openembedded.org is the master index of layers and their recipes
<rburton> meta-python has many
<rburton> remember oe-pkgdata-util only finds stuff you've already built, so if you're missing the depends then you most likely haven't built it
<Ch^W> khem: Build host was created from Yocto hardknott (we build our own SDK distro).
<Ch^W> I tried buildtools-extended from Dunfell, but same result. Our last "known good" build was under GCC 8 in thud, but there are no buildtools-extended for that release. It looks like 3.0.4 is the oldest set of buildtools-extended.
<Ch^W> If it is only me, I can look further, but this started happening right after we deployed a Hardknott SDK that went from GCC 8 to GCC 10.
<rburton> Ch^W: there's no errors in that log
halstead has quit [Quit: Leaving]
<Ch^W> rburton: Sorry, I think the build is set to break on warnings. I can paste in more.
<rburton> it doesn't say warnings-are-errors
halstead_ is now known as halstead
zyga-mbp has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
hpsy has joined #yocto
<Ch^W> Ok, gist updaed with the full output of gobject-introspection build. I am continuing to look at logs.
zyga-mbp has joined #yocto
florian_kc has quit [Read error: Connection reset by peer]
florian_kc has joined #yocto
<rburton> rerun ninja in a devshell to figure out what exactly is failing, as it's not gcc
<rburton> unless bison warnings are actually errors
<Ch^W> It is odd that such a mature project has shift/reduce errors in the grammar. Perhaps I was just getting distracted by that.
<override> devtool doesnt really take care of the package PACKAGECONFIG option in local.conf or is there a devtool option im missing?
<override> PACKAGECONFIG variable .. Im guessing I gotta go in local.conf and set it up myself?
<kergoth> most don't change packageconfig at all. what are you expecting devtool to do about it?
bunk has joined #yocto
<override> kergoth just expecting it to populate all keys/values the recipe files expects. like I see devtool making files with stuff like PACKAGECONFIG[:python_version < "3.8"] = ",,,python3-typing_extensions"
<kergoth> yes.. devtool adds the optional configuration variables, its expected you'll have to modify that manually, but devtool can't read your mind to know whicho f those you actually want to enable
<kergoth> you'd set PACKAGECONFIG ?= "some-default-value" in the recipe based on what you want the default recipe behavior to be
<kergoth> then the user would either be happy with that or alter how it builds in their distro or local.conf, but that doesnt' ahve anything to do with devtool
<override> not expecting devtool to be a mind reader .. think I get the idea tho, thanks
florian__ has joined #yocto
zyga-mbp has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
florian_kc has quit [Read error: Connection reset by peer]
florian has joined #yocto
hpsy has quit [Ping timeout: 265 seconds]
florian__ has quit [Ping timeout: 268 seconds]
rcw has joined #yocto
<rburton> Override FWIW that packageconfig example is malformed
<override> rburton .. im reading up on packageconfig .. apparently I have to initialzie the list of features ... PACKAGECONFIG ??= "f1 f2 f3 ..."
leon-anavi has quit [Quit: Leaving]
meego has quit [Quit: Leaving...]
bunk has quit [Quit: leaving]
<override> rburton are you saying devtool malformed packageconfig?
<override> couldnt have asked for anything else to thrown into the mix, love it.
<override> couldnt have asked for anything else to thrown into the mix, love it.
otavio has joined #yocto
<Ch^W> Hmmm, running ninja in devshell shows me that bison is segfaulting...
* Ch^W goes further down the rabbit hole...
pbergin has quit [Quit: Leaving]
<override> cna anyone please help me understand why PACKACONFIG is getting formed the way it is, and how should I be fixing it? https://paste.ee/p/tGBgd
<kergoth> override: devtool output is *generated* and the comments in the recipe specifically say you need to review it to make sure it's correct.
<override> I get that.. clearly I'm using devtool cus im just starting out with yocto - im trying to see what I can do to make PACKAGECONFIG not come out malformed like that
<override> Im trying to see if keys in PACKAGECONFIG can even have spaces etc
<rburton> override: obviously devtool got very confused. can you file a bug?
<override> and ive read up on the manual a little bit, I get the broader picture. Just putting it out there. Not trying to use anyone as a private consultant or anything, feel free to help when at your convinience..
<override> rburton: im listing down all the bugs I have to file. Just that I gotta get this recipe out first.
<rburton> for now just delete all those PACKAGECONFIG lines as they're just nonsense
<override> Not easy for me to qualify this stuff as a bug - I have to go through the manual and all first to figure out the lay of the land..
<override> ok, thanks rburton for giving this a look.
<override> Ive been struggling with this for over a weekend a few days now. Summer weekend of my life that ill never get back, so I really appreciate all your help
<rburton> throw something into bugzilla and tell me the #, i'll fix up the components if needed
<override> send over your bugzilla link
<rburton> just tell me the # here
<override> the # for what
<rburton> the bug you create
<override> ok - give me the link to your bugzilla so I can create one?
<rburton> bugzilla.yoctoproject.org
<rburton> (sorry)
<rburton> RP: woo -7+0 on the master cve run
<override> no worries
chrfle has joined #yocto
<override> rburton renaming bb.parsefailed to .bb isnt doing the trick, what am I missing?
hpsy has joined #yocto
<v0n> Shouldn't systemd-initramfs install /etc/initrd-release?
<Ch^W> So far the primary difference looks like a terminal issue. bison.real is the one used in the build root, and it expects libtextstyle.so.0. The segfault happens on one of the free() calls in term_ostream_create
Guest71 has joined #yocto
goliath has joined #yocto
<Ch^W> Ok, found it.
<Ch^W> There is a segfault in bison.real when --color=no is missing.
<Ch^W> Looks like a bison bug of some sort.
<Ch^W> I think if the terminal does not support colors, it complains, but the complaint leads to a sefgault.
Vonter has quit [Ping timeout: 265 seconds]
<rburton> niiiice
<rburton> thanks bison
<rburton> can you turn that off in bison at build time?
williamh has joined #yocto
williamh has quit [Client Quit]
<Ch^W> It is a wrapper of some sort, so I am going to see if a patch will add the --color=no argument.
williamh89 has joined #yocto
williamh89 has quit [Client Quit]
williamh89 has joined #yocto
yates has joined #yocto
<yates> ok guys, i sure could use some help, and i'll tell you up front this is going to be a PITA
<yates> on second thought - let me do this later. my current environment does not allow me to provide the details.
<kanavin> moto-timo, answered on the mailing list
<kergoth> override, rburton: looks like the devtool python project setup.py parsing for optional features needs to interpret and process the python-version specific conditionals
yocti has joined #yocto
yocti has quit [Client Quit]
yocti has joined #yocto
<rburton> yes
<rburton> Ch^W: i was thinking can it be told to not do ncurses stuff at all
<Ch^W> rburton: Hmmmm, well I am not sure. I only found this when building gobject-introspection.
<Ch^W> rburton: But let me kook.
<Ch^W> *look
<Ch^W> And even if it can, this is very clearly a bug they should look at. I will submit a detailed issue to the bison project.
williamh89 has quit [Ping timeout: 246 seconds]
<Ch^W> rburton: So here is the weird thing... bison-native triggers that ncurses stuff because ldd shows that it is looking for libgettextstyle.so.0, but the bison we install into our SDK (non-native), does not look for htat library.
<Ch^W> So it is like the native recipe is pulling in stuff it does not need?
<RP> rburton: yes, that was nice :)
<Ch^W> bison_3.7.5.bb -> inherit autotools gettext texinfo
<Ch^W> Last change to that was in 2014.
<Ch^W> And from the changelog, that looks like an "in the future" thing.
<Ch^W> So clearly the bison project needs to know about this. And I will tell them. But I am curious what the collective wisdom says to do here? If removing gettext from the inherit list solves it, then go with that? Or patch --color=no into the wrapper?
Guest5496 has joined #yocto
Guest5496 has quit [Client Quit]
ant_ has joined #yocto
florian has quit [Ping timeout: 265 seconds]
williamh89 has joined #yocto
<Ch^W> Hmmm, ok patch it is... Removing gettext from the inherit annoys the QA checker.
florian has joined #yocto
<williamh89> does anything have an idea how I can get my DSI connector recognized in X11? for example, when I run xrandr I get: default connected 720x1280+0+0 0mm x 0mm
<williamh89> 720x1280 0.00*
<williamh89> but im sort of expecting: DSI-1 connected primary 720x1280+0+0 (normal left inverted right x axis y axis) 0mm x 0mm
zyga-mbp has joined #yocto
zyga-mbp has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
camus1 has joined #yocto
zyga-mbp has joined #yocto
camus has quit [Ping timeout: 258 seconds]
camus1 is now known as camus
zyga-mbp has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
LetoThe2nd has quit [Quit: Connection closed for inactivity]
risca has quit [Ping timeout: 264 seconds]
roussinm has joined #yocto
risca has joined #yocto
Guest71 has quit [Quit: Client closed]
xmn has joined #yocto
sakoman has quit [Quit: Leaving.]
* Ch^W gives bison the stink eye...
<Ch^W> Can _anyone_ build gobject-introspection-native with the Hardknott complement of dev tools? Or is it just me?
<Ch^W> Once past the command-line fun, it segfaults again on the scannerparser.y shift/reduce warnings.
Vonter has joined #yocto
<v0n> Can WKS_FILE be set from an image recipe?
rodrjassoccom[m] has joined #yocto
florian has quit [Ping timeout: 252 seconds]
Spooster has quit [Remote host closed the connection]
camus has quit [Ping timeout: 258 seconds]
camus has joined #yocto
rewitt2 has joined #yocto
rewitt1 has quit [Remote host closed the connection]
hpsy1 has joined #yocto
hpsy has quit [Ping timeout: 265 seconds]
Vonter has quit [Ping timeout: 252 seconds]
goliath has quit [Quit: SIGSEGV]
risca has quit [Ping timeout: 258 seconds]
risca has joined #yocto
risca has quit [Ping timeout: 252 seconds]
rewitt3 has joined #yocto