khem has quit [Quit: Connection closed for inactivity]
paulg has joined #yocto
<rburton>
manuel_: "yes". the QA on the autobuilder uses testimage/qemu. we use testimage/FVP. some use lava. Some use Labgrid. I'm sure there are others using something else.
<LetoThe2nd>
DiogenesMountain: no, what you want is a dev-distro
pvogelaar has joined #yocto
<kanavin>
RP: the busybox patch should probably not be taken
<kanavin>
'busybox: Use ISO 13400 timings in zcip'
<rburton>
manuel_: the short version is its a closed source qemu-like for arm.
<michalsieron>
rburton: because there is this recipe for sbsigntool, which depends on ccan, and ccan during do_configure tries to run a script from file created with mktemp
<michalsieron>
and I have /tmp mounted with noexec, so it fails and I cannot build it
<rburton>
michalsieron: don't mount /tmp noexec, because stuff like that happens
<manuel_>
rburton: Thanks. That explained it better than the entire ARM page did.
<rburton>
michalsieron: honestly we should probably check for that and fail early because that's not an uncommon pattern
<michalsieron>
It took me quite some time to even figure out why the build was failing
<michalsieron>
Because the script from ccan didn't exit early when running from /tmp failed, so do_configure succeeded and it only failed in do_compile
<rburton>
yeah, it's fun. anything using meson for example will write temporary scripts and binaries to /tmp (whereas autoconf uses the current directory).
<michalsieron>
That's more of a ccan issue, but I agree that a check in bitbake for /tmp mount options would be great
<rburton>
patches welcome! :)
<michalsieron>
rburton: at which point would such check be even made?
<rburton>
sanity.bbclass
<rburton>
does stuff like "is this file system case-sensitive" and "is the build directory on nfs"
GNUmoon has quit [Remote host closed the connection]
<chep>
Hi, I'm building a recipe which installs a script in `/usr/bin/script.sh`. This script is mandatory to build another recipe which depends on the first one. When I add `FILES:${PN}-dev += "/usr/bin/script.sh"` I can see the file in the -dev package (with dpkg -c) but it is not in the sysroot of my second recipe. Headers and lib are present in sysroot, only the script is missing. Can anyone tell me why ?
<rburton>
becase bindir isn't staged, you normally can't run target binaries on the host
<chep>
so what is the right way to do this ?
<rburton>
you can add bindir to SYSROOT_DIRS, or extend the recipe to have a native version which will stage the binaries and you can depend on that instead
<rburton>
michalsieron: what distro are you using? trying to isolate where the noexec comes fro
<chep>
rburton: ok, thank you
GNUmoon has joined #yocto
<michalsieron>
rburton: oh, that is probably linked to our server setup. but it is a RHEL
<rburton>
fedora doesn't use noexec, but maybe rhel is special
rob_w has joined #yocto
<michalsieron>
our other servers are ok in that regard. so far I've found only one, so as I said earlier, it may be that this specific host has different fstab version installed
GNUmoon has quit [Remote host closed the connection]
<rburton>
using a script i determined that none of the AB workers have noexec /tmp so i don't believe any of the mainstream distros do that by default (unless that's something that halstead changes when setting up a new one)
pbsds7 has joined #yocto
dkl_ has joined #yocto
stgloor has joined #yocto
dkl has quit [Read error: Connection reset by peer]
stgl has quit [Read error: Connection reset by peer]
pbsds has quit [Ping timeout: 264 seconds]
neofutur_ has quit [Ping timeout: 264 seconds]
pbsds7 is now known as pbsds
neofutur_ has joined #yocto
florian_kc has joined #yocto
stgloor is now known as stgl
<michalsieron>
rburton: you think a warning like that would suffice? "/tmp is mounted with noexec. Unexpected failures may occur!"
<rburton>
honestly i'd be tempted to make it fatal
<rburton>
send a patch and we can see if anyone screams :)
<michalsieron>
I would rather not break anyone's builds :D
<RP>
michalsieron: I'd make it fatal
<RP>
michalsieron: it is much better to tell people about these kinds of problems up front
rob_w has quit [Ping timeout: 246 seconds]
lexano has joined #yocto
vmeson has joined #yocto
alessioigor has quit [Quit: alessioigor]
Minvera has joined #yocto
xmn has quit [Ping timeout: 256 seconds]
linfax has quit [Ping timeout: 255 seconds]
martin_thingvold has joined #yocto
michalsieron has quit [Ping timeout: 250 seconds]
xmn has joined #yocto
vladest has quit [Ping timeout: 276 seconds]
Xagen has joined #yocto
dgriego has quit [Quit: Bye]
Vonter has quit [Ping timeout: 240 seconds]
dgriego has joined #yocto
rfuentess has quit [Remote host closed the connection]
<moto-timo>
when was addpylib first introduced? my git foo didn't find it yet
vladest has joined #yocto
<moto-timo>
nevermind, it helps to not be on ancient branch
<moto-timo>
sigh, I thought it was before mickledore
goliath has quit [Quit: SIGSEGV]
<RobertBerger>
What's the best practice with respect to (host/native) Python version compatibility when you write a bbclass with Python tasks?
<RobertBerger>
And changing the Python Version just for ONE layer, which does not work with the default Python version on Ubuntu 22 makes me a bit nervous
Vonter has joined #yocto
<RP>
RobertBerger: that is a bit too recent unfortunately :/
<RobertBerger>
There are already many moving targets, not sure I also want to use an "untested" Python version
<RobertBerger>
And apparently there are people who don't use the latest poky master, so that's even worse I guess.
<JPEW>
RobertBerger: That function is convenient, but it's not hard to roll it yourself, so I suggest doing that for compatablity
<mischief>
sigh. i try to set IMAGE_LINGUAS = "" from a .inc file but somehow including core-image after that results in the default from default-distrovars.inc and i dont get why
<mischief>
bitbake-getvar seems to be saying IMAGE_LINGUAS = "" is "set?" but it is not a soft assignment..
Kubu_work has quit [Quit: Leaving.]
vladest has quit [Remote host closed the connection]
jmd has joined #yocto
<mischief>
okay, bitbake-getvar is full of lies. if i set it to some value like "fnord", bitbake-getvar doesn't even print it
<mischief>
deleting build/cache build/tmp/cache seems to have sorted it. am i missing something special about variable dependencies with IMAGE_LINGUAS?
Vonter has quit [Ping timeout: 276 seconds]
Vonter has joined #yocto
amitk_ has quit [Ping timeout: 276 seconds]
nerdboy has joined #yocto
nerdboy has quit [Changing host]
nerdboy has joined #yocto
nerdboy has quit [Remote host closed the connection]
nerdboy has joined #yocto
nerdboy has quit [Changing host]
nerdboy has joined #yocto
vladest has joined #yocto
martin_thingvold has quit [Ping timeout: 264 seconds]
martin_thingvold has joined #yocto
rfs613 has quit [Ping timeout: 268 seconds]
amitk has joined #yocto
<khem>
depends on image you choose to depend on, some smallish images set it to nothing for saving some space,
<khem>
you can do something like IMAGE_LINGUAS:append = "..." wherever you are trying to add it
rfs613 has joined #yocto
rfs613- has joined #yocto
rfs613 has quit [Ping timeout: 256 seconds]
rfs613- is now known as rfs613
goliath has quit [Quit: SIGSEGV]
florian_kc has joined #yocto
gsalazar has quit [Remote host closed the connection]
Minvera has quit [Remote host closed the connection]
Minvera has joined #yocto
lexano has quit [Ping timeout: 276 seconds]
mischief has quit [Ping timeout: 268 seconds]
mischief has joined #yocto
nerdboy has quit [Ping timeout: 276 seconds]
lexano has joined #yocto
<DiogenesMountain>
LetoThe2nd: Thanks for the reply. Not sure that is possible in this case, but if it can't be done through images, then it can't be done through images.
nerdboy has joined #yocto
nerdboy has quit [Changing host]
nerdboy has joined #yocto
mischief has quit [Ping timeout: 256 seconds]
goliath has joined #yocto
mischief has joined #yocto
<LetoThe2nd>
DiogenesMountain: "it can't be done through images"
florian_kc has quit [Ping timeout: 276 seconds]
nerdboy has quit [Ping timeout: 256 seconds]
olani- has joined #yocto
Saur_Home has quit [Quit: Client closed]
Saur_Home has joined #yocto
jmd` has joined #yocto
jmd has quit [Remote host closed the connection]
jmd` has quit [Remote host closed the connection]
Ad0 has quit [Ping timeout: 264 seconds]
alperak has quit [Quit: Client closed]
Ad0 has joined #yocto
Chaser has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
amitk has quit [Ping timeout: 272 seconds]
nerdboy has joined #yocto
xmn has quit [Ping timeout: 260 seconds]
florian_kc has joined #yocto
simonew has quit [Quit: Konversation terminated!]
<vmeson>
on my system, llvm-native takes 1000s seconds in do_compile with little happening in parallel. Seems like a mesa dependency... if I knew more about mesa, I'd check if all ~2000 .c/cpp files are needed or if the mesa dependency could be avoided...
<vmeson>
can set the value of :term:`DISTRO_FEATURES_NATIVE` in your configuration such
<vmeson>
that ``opengl`` is excluded
ykrons has quit [Ping timeout: 256 seconds]
GNUmoon has joined #yocto
arielmrmx has joined #yocto
<rburton>
vmeson: mesa needs llvm for the shader compilers
Net147 has quit [Quit: Quit]
Net147 has joined #yocto
Net147 has quit [Changing host]
Net147 has joined #yocto
alessioigor has quit [Ping timeout: 256 seconds]
sakoman has quit [Quit: Leaving.]
sakoman has joined #yocto
olani- has quit [Ping timeout: 272 seconds]
tleb has quit [Ping timeout: 264 seconds]
tleb has joined #yocto
<paulg>
Exception: FileNotFoundError: [Errno 2] No such file or directory: 'mips64-poky-linux-strip '
<paulg>
ERROR: Task (poky/meta/recipes-kernel/linux/linux-yocto_6.6.bb:do_strip) failed with exit code '1'
<paulg>
Runs fine from a devshell, so it is time to hunt down that annoying trailing whitespace on the 1st line I posted. :-(
<RP>
paulg: is that an older release. It sounds familiar :/
<usvi>
how do I stop this sub-versioning(?) evilcorp-evilprod-update_1.5.0-kirkstone41-r0.1_armv7at2hf-neon.ipk that r0.1 . the 1 keeps changing up. I want to stop it
martin_thingvold has quit [Ping timeout: 268 seconds]
<paulg>
RP, latest master on everything - as of like 3h ago. Only "strange" thing I've done is revert the edgerouter delete from meta-yocto-bsp, 'cause I wanted to build a new image for the old turd.
<paulg>
checking a qemumips build now to compare...
<RP>
paulg: try making kernel.bbclass: strip = d.getVar('KERNEL_STRIP')
<RP>
strip = d.getVar('KERNEL_STRIP').strip()
<RP>
paulg: why you're awakening the dead, I'm not sure I want to know
<paulg>
Don't worry - I'm the one that sent the removal, so I won't be asking for re-addition!
<RP>
paulg: That was what was making it all the stranger! :)
<paulg>
It was one of those "I should do something with it, or it needs to go into e-waste." moments.
<paulg>
Got up to 8C here today which kicks in the Canadian "spring cleaning" mentality...
<RP>
paulg: it is warmer there than here!
<paulg>
RP, your addition of ".strip()" suffix seems to have fixed it (thanks) - don't ask me to explain it though!
<paulg>
why didn't that blow up on qemu mips (which is also 64 bit)
<paulg>
For the record, it would have taken me hours to track down that whitespace. :-P
<Saur_Home>
usvi: That is the PR server doing that. It increases the .X part whenever there are changes in the output compared to the last time you built the recipe with that version. It is to make sure two different builds get different versions, which is necessary if you are creating a package feed.
<usvi>
Saur_Home: ok thanks. what I want to really do is I want to grab the resulting ipk and encrypt it with a custom sh script. can I in anywhere get a handle of the actual ipk file which was build? like verbatim, including the r0.1 thingy. Of course I can parse it but would be convenient to postprocess it after everything else was done
<Saur_Home>
usvi: Are you just after that specifik package, or are you building an image and want all packages that go into it?
<usvi>
Saur_Home: only one is needed
goliath has quit [Quit: SIGSEGV]
<usvi>
I have been googling around with archive class but I'm just too stupid to spot what to do