tammranil has quit [Remote host closed the connection]
tammranil has joined #yocto
tammranil has quit [Remote host closed the connection]
tammranil has joined #yocto
mbulut has quit [Remote host closed the connection]
tammranil has quit [Remote host closed the connection]
tammranil has joined #yocto
Chaser has joined #yocto
Granjow has joined #yocto
<Granjow>
Hi! My kernel says that trace_printk() is being used and that it is a debug kernel and should not be used in production. How do I disable this?
GNUmoon has quit [Remote host closed the connection]
jmiehe has joined #yocto
GNUmoon has joined #yocto
gspbirel56 has quit [Ping timeout: 265 seconds]
jmiehe has quit [Client Quit]
gspbirel56 has joined #yocto
gspbirel56 has quit [Ping timeout: 246 seconds]
florian has joined #yocto
<olani>
I have a question about FIXMESTAGINGDIRHOST. For reasons I may bring up later I have 'devtool modified' autotools-native. But the generated sysroot is broken when used in dependent recipes.
<olani>
As far as I can tell, this is because the replacement of paths with the FIXMESTAGINGDIRHOST is done when the sstate is created. And sstate creation is disabled by externalsrc.bbclass.
<olani>
I can probably hack my way around this for now, but it seems like a bug.
<olani>
Why is this processing done by the sstate creator instead of the populate_sysroot task?
<olani>
This is all on styhead, and while our own layers are included they really shouldn't affect anything here.
<RP>
olani: the paths the data is created with/in is only known at creation time. At install time you only know where you're installing it to.
<RP>
olani: the replacement is therefore done in two halves, creation puts a marker in and then install replaces the marker
<olani>
Yeah, I think I get that. But if I look in the sysroot-destdir folder of a normal autotools-native build, the replacement is done there. In the devtooled case it is not. And the unpacked sysroot in say libtool-native does not have the marker, and so is not corrected.
<RP>
olani: It potentially sounds like a bug but it is hard to know exactly what you're doing
<olani>
RP: Should I file a bug and try to flesh out the details?
asgj-gom has joined #yocto
florian_kc has joined #yocto
gspbirel56 has joined #yocto
<RP>
olani: I think it would be worth it just to establish if there is an issue and where
<olani>
RP: OK, I'll do that
debbiemartin has joined #yocto
zpfvo has joined #yocto
mbulut has joined #yocto
<ak77>
ok. so packagegroup-core-weston RDEPENDS on weston-examples, I try to remove that runtime dependency (it's not really runtime dependency). how do I do that? RDEPENDS:packagegroup-core-weston:remove = "weston-examples" doesn't work
<olani>
ak77: You need to do that from your local.conf or equivalent, or in a packagegroup-core-weston.bbappend.
<olani>
ak77: I would recommend a bbappend.
<ak77>
olani: thank you
<RP>
olani: could you expand on "devtool modify automake does not work for me" in that bug - perhaps if we fix those issues, the rest would fall into place?
asgj-gom has quit [Quit: Client closed]
xmn has joined #yocto
zpfvo has quit [Remote host closed the connection]
mbulut has quit [Quit: Leaving]
jdiez has quit [Remote host closed the connection]
<olani>
RP: When I run 'devtool modify automake' I get an error:
<olani>
RP: ERROR: Command '['rm', '-rf', '/mnt/build/poky/fixmestagingdirhost/tmp/work/core2-64-poky-linux/automake/1.17/devtooltmp-2l27ojdj/workdir/sources-unpack']' returned non-zero exit status 1.
<olani>
RP: I don't get that for 'devtool modify quilt', so I think it's something special with automake. I tried to figure out what was going on, but it was blocking me so I created the workspace by hand instead.
<RP>
olani: I suspect something in the by hand approach didn't quite work
jdiez has joined #yocto
jdiez has quit [Client Quit]
<olani>
Yeah, you are probably right. The sysroots-component folder for quilt-native has the fixmepath* files. They are empty because quilt apparently did not need to be fixed.
<olani>
RP: I'm not sure how my automake workspace differs from quilts though.
jdiez has joined #yocto
<olani>
RP: I guess I need to find a package that is more similar to automake and see what is different.
rfuentess has quit [Remote host closed the connection]
<RP>
olani: some of the "toolchain" bootstrap components are a bit special
<qschulz>
do we have the new maintainer of the YP docs on IRC by any chance?
<qschulz>
abelloni: you'd probably be the one with this info :)
<olani>
RP: Turns out 'devtool modify automake-native' works. I suspect this is because automake-native is on automakes DEPENDS-path somehow. It still doesn't work though. I've updated bugzilla.
<qschulz>
RP: abelloni: <insert doc maintainer nick>: I think we should add an entry in the QA section for telling people how to read logs
<qschulz>
Maybe there's something we could do to improve the log output so it's more explicit what to do next
<qschulz>
but in the worse case maybe explaining how to actually start debugging this and not stop at the "failed with exit code" line would help people
<qschulz>
(and we could then simply point people at this QA entry instead of repeating ourselves (or rather, try to repeat ourselves less often :) )
<RP>
qschulz: very happy to see FAQ entries added
<RP>
qschulz: less happy about pages of text on log output
<qschulz>
RP: I think my idea was more "don't stop at that line, read a bit further down where you get a path to the actual error log file, its path usually looks like that <insert path>, this is where your debugging adventure starts"
<RP>
qschulz: works for me
<qschulz>
Because now the next mail that k anavin (don't want to ping him unnecessarily :) ) is going to send to that person is "look a bit more in the log and tell us what's in the log file"
<qschulz>
s/in the log and/in the bitbake output and/
<RP>
qschulz: agreed
agodard has joined #yocto
agodard has quit [Changing host]
agodard has joined #yocto
<mathieudb>
qschulz: agodard just joined if you want to talk about docs :)
<qschulz>
(I'm not asking you to write this new FAQ just to be clear; just asking for an opinion)
<agodard>
qschulz: right, I see, I agree. like what was suggested by RP on class appends, it makes a good pointer to give to people - and the FAQ seems to be a good place for that
asgj-gom has joined #yocto
Sree41 has joined #yocto
rob_w has quit [Quit: Leaving]
Sree41 has quit [Client Quit]
<asgj-gom>
Does anyone here know how I would host a custom opkg package repository? Like, is there a docker image available or do I just host a simple ftp server? Or should I ask in the openwrt irc instead?
<RP>
asgj-gom: should just be a simple http server if you create the index files
<RP>
asgj-gom: we have examples of it in the lib/oeqa/runtime/cases iirc
gspbirel56 has quit [Remote host closed the connection]
<RP>
khem: I suspect we will be affected too
goliath has quit [Quit: SIGSEGV]
druppy has joined #yocto
gspbirel56 has joined #yocto
Xagen has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
florian_kc has joined #yocto
kergoth has quit [Read error: Connection reset by peer]
florian_kc has quit [Ping timeout: 252 seconds]
Saur_Home44 has joined #yocto
Saur_Home4 has quit [Ping timeout: 256 seconds]
druppy has quit [Ping timeout: 252 seconds]
leon-anavi has joined #yocto
Xagen has joined #yocto
goliath has joined #yocto
xmn has joined #yocto
florian_kc has joined #yocto
Xagen has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
debbiemartin has quit [Quit: Connection closed for inactivity]
mbulut has quit [Remote host closed the connection]
mbulut has joined #yocto
mbulut_ has joined #yocto
<mbulut_>
does anyone recall a race condition like scenario with create-spdx in kirkstone or earlier?
<mbulut_>
i'm seeing build failures in CI that can't be reproduced locally
mbulut has quit [Ping timeout: 272 seconds]
<mbulut_>
on CI, it regularly fails while attempting to read a runtime dependency (kernel-5.15.52.spdx.json) of package X
<mbulut_>
the dependent package seems to vary but the dependency is always kernel-5.15.52.spdx.json
<mbulut_>
the file does not exist while executing do_create_runtime_spdx on package X
<mbulut_>
comparing the create-spdx class of kirkstone to that in scarthgap i see there was quite some movement in that area... not sure if that class is easily backportable though
<mbulut_>
(and i don't have evidence that there is actually a race going on, just guessing)
mbulut_ has quit [Remote host closed the connection]
<tgamblin>
Guest55: you could also find recipes in one of the main layers (openembedded-core, meta-openembedded, etc.) that don't have a maintainer and focus on submitting upgrade patches, adding ptests, etc. (you can do this even for recipes that do have maintainers, FTR)
Guest55 has quit [Ping timeout: 256 seconds]
<tgamblin>
or not
merit has quit [Ping timeout: 252 seconds]
warthog9 has quit [Ping timeout: 252 seconds]
merit has joined #yocto
warthog9 has joined #yocto
warthog9 has quit [Ping timeout: 252 seconds]
Chaser has quit [Quit: My Unrecognized Mac has gone to sleep. ZZZzzz…]