mrpelotazo has quit [Read error: Connection reset by peer]
mrpelotazo has joined #yocto
amitk has joined #yocto
xmn has quit [Ping timeout: 268 seconds]
enok has joined #yocto
thomas_34 has joined #yocto
<thomas_34>
Hey good morning. I have a include files, which defines variables and tasks for a lot (20) of recipes. At which point makes it sense to convert that into a bbclass?
<thomas_34>
Can someone explain me whats the difference in general between .inc-Files and bbclass? What is the bigger picture here?
hnez has joined #yocto
alessioigor has joined #yocto
alperak has joined #yocto
bharathbck has joined #yocto
<mcfrisk_>
thomas_34: or a recipe which the 20 other recipes DEPEND at build time to use the headers
<mcfrisk_>
bblasses are consumed by recipes and build tools, mostly. An .inc is a way for multiple recipes to share common stuff. bbclass can be used across layers, .inc needs to be in the same directory as recipes.
goliath has joined #yocto
<thomas_34>
mcfrisk_, "An .inc is a way for multiple recipes to share common stuff." Is that also valid doing by bbclass?
<thomas_34>
Or are bbclasses limited to only share common functions and not variables?
<mcfrisk_>
thomas_34: bblasses can share anything which ends up in recipes, variables, funcions, asks etc etc
<mcfrisk_>
but, a set of header files sounds like a SW package and thus would to me belong to a recipe.
<thomas_34>
Ähm.. wait. Where does these "header files" come from?
<thomas_34>
From me: "I have include files, which defines variables"... I mean I have .inc files with bitbake variables defined, which I reuse in lot of recipes
<mcfrisk_>
thomas_34: ah my bad, too little coffee
<mcfrisk_>
include files sounded too much like .h
enok has quit [Ping timeout: 272 seconds]
<thomas_34>
Ahh okay, no. I'm entirely within bitbake ;)
<mcfrisk_>
if the data is recipe variables etc for a small set of recipes, then an .inc. for a lot of recipes across layers, a bbclass
<mcfrisk_>
but the things could also be in distro configuration, e.g. DISTRO_FEATURES, or in the machine config like MACHINE_FEATURES
<thomas_34>
Ahhh... okay. The the "border" for .inc files are layers? Is that a "practical" distinction?
<mcfrisk_>
how recipes find .inc, only recipes within same directory will find it. For bbclasses, all recipes from all layers will find them.
<thomas_34>
Okay thank you. Now I get the bigger picture I guess
<mcfrisk_>
no problem, hope this helps
rob_w has joined #yocto
olani has quit [Ping timeout: 264 seconds]
ptsneves has joined #yocto
linfax has joined #yocto
gsalazar has joined #yocto
mckoan|away is now known as mckoan
<mckoan>
good morning
ablu has quit [Read error: Connection reset by peer]
enok has joined #yocto
ablu has joined #yocto
leon-anavi has joined #yocto
Kubu_work has joined #yocto
<thomas_34>
Can I access a "recipe-variable" within a bbclass? Just asking if that should be possible, before I search for my error :S
<michalsieron>
hello there, I tried using `INCLUDE_PYCS = "0"` to remove `*.pyc` files from generated packages (from python3 recipe). However, it seems that many `.pyc` files are still being included. I guess it's because some packages in `python3-manifest.json` have empty list for the `.pyc` files (`"cached": []`)? Is there some nicer way of dealing with those
<michalsieron>
files than using `find+rm` in .bbappend?
enok has quit [Ping timeout: 255 seconds]
Saur_Home has quit [Quit: Client closed]
Saur_Home has joined #yocto
LocutusOfBorg has quit [Read error: Connection reset by peer]
LocutusOfBorg has joined #yocto
LocutusOfBorg has quit [Read error: Connection reset by peer]
jsandman has quit [Remote host closed the connection]
hnez has joined #yocto
starblue has quit [Ping timeout: 260 seconds]
tomzy has joined #yocto
starblue has joined #yocto
tomzy is now known as tomzy-0
Ad0 has joined #yocto
tomzy-0 has quit [Ping timeout: 250 seconds]
DvorkinDmitry has joined #yocto
<DvorkinDmitry>
is it OK to put INITRAMFS_IMAGE = definition to image file?
<mcfrisk_>
DvorkinDmitry: I wondered the same few days ago, conclusion was that no, it's not enough for image recipe to know which INITRAMFS_IMAGE to build. I kept this in distro config then.
<DvorkinDmitry>
mcfrisk_, thank you
enok has quit [Ping timeout: 264 seconds]
tomzy has joined #yocto
xmn has joined #yocto
michalsieron has quit [Quit: Client closed]
TehRiizQ has joined #yocto
enok has joined #yocto
bharathbck has quit [Ping timeout: 250 seconds]
tomzy has quit [Quit: Client closed]
lexano has joined #yocto
ldywicki has joined #yocto
<rburton>
DvorkinDmitry: initramfs_image is for a kernel, so no
<DvorkinDmitry>
rburton, where is it better to place except local.conf?
<rburton>
your distro or machine conf
<DvorkinDmitry>
rburton, thank you
Saur_Home has quit [Quit: Client closed]
Saur_Home has joined #yocto
tomzy has joined #yocto
florian_kc has quit [Ping timeout: 255 seconds]
prabhakalad has quit [Quit: Konversation terminated!]
<rburton>
RP: why would that be an unclosed function when there's just leading whitespace (or is the bitbake parser actually that strict and it needs the closing brace to be on a line of its own)
<thomas_34>
rburton Thank you :) I can test this, when its merged
<thomas_34>
RP, I dont know. Maybe its related that python is kinda strict about line indention etc...
<thomas_34>
*indentation
TehRiizQ has quit [Ping timeout: 250 seconds]
rm5248 has joined #yocto
GillesM has joined #yocto
GillesM has quit [Client Quit]
Jones42 has joined #yocto
thomas_34 has quit [Ping timeout: 250 seconds]
phako_ has quit [Quit: Client closed]
ptsneves has quit [Ping timeout: 255 seconds]
<RP>
rburton: the closing brace needs to be unintended to ensure it is different from whitespace indented python. Yes, it could work differently but it currently doesn't
<rburton>
fair
<Jones42>
Hi, at the Chemnitzer Linux-Tage last month, Marco Felsch from Pengutronix mentioned something about upstreaming their own image_types_verity.bbclass . Has someone already seen some patches flying around regarding that? I tried to search the mailing lists, but without success
<rburton>
that's right. you're copying from WORKDIR (which is where unpack puts files, and where the build happens) into D, which is the root of the package contents
amitk has quit [Remote host closed the connection]
<Guest1337>
well im confused. i want to put all the services to /etc/systemd/system and the shell script into /usr/bin
<rm5248>
I haven't tried to compile with llvm before, so I don't have much insight there. It /could/ be a cmake problem though, I know some people were complaining about cmake the other week and prefer meson for cross-compiling
simone has joined #yocto
<gmorell>
yeah the program in question theoretically can compile with gcc but the docs are super not helpful
<gmorell>
I'm just going to grab a bin from the releases
<gmorell>
and drop it in the correct place
usvi has quit [Remote host closed the connection]
simonew has joined #yocto
simone has quit [Ping timeout: 272 seconds]
simonew has quit [Remote host closed the connection]