vladest has quit [Remote host closed the connection]
seninha has quit [Quit: Leaving]
seninha has joined #yocto
seninha has quit [Quit: Leaving]
roussinm has joined #yocto
<roussinm>
I think what we are doing right now is broken but lets see. We share sstate-cache for our build-runner. We have 2 runner x86
<roussinm>
_64, but different cpus. And I think that native-recipe could in theory compile with cpu instruction that are not comptatible with the other cpu? Am I crazy?
<roussinm>
python or cmake as an example.
AKN has joined #yocto
starblue has quit [Ping timeout: 256 seconds]
<AKN>
Hi been trying to get brcmfmac drivers working in custom yocto image for beaglebone ai past two weeks
<AKN>
It was working for the same configuration a couple of months ago, now it doesn't even load the firmware files
<AKN>
kindly help me out.
starblue has joined #yocto
<roussinm>
AKN: error message from dmesg?
camus has joined #yocto
prabhakarlad has quit [Ping timeout: 252 seconds]
<AKN>
roussinm: No error message, It doesn't even tries to load the firmware
<roussinm>
The driver loads?
camus1 has joined #yocto
<AKN>
roussinm: No
<roussinm>
AKN: ok I'm not sure how the driver binds to the device, does it required a DTS or something else?
camus has quit [Ping timeout: 276 seconds]
camus1 is now known as camus
<AKN>
roussinm: I have build the same image couple of months ago with same dts files firmware no & revision were loaded
<AKN>
roussinm: Now it does even load the firmware
<AKN>
roussinm: *doesn't even load the firmware
<roussinm>
AKN: hmm without the hardware it's really hard, from what you are telling me and from my experience, it feels like it doesn't bind on the device. I guess the probe of the driver isn't called?
<roussinm>
Maybe it's called but errors out.
<AKN>
roussinm: I am building the image for beaglebone ai , using meta-ti repo
<roussinm>
AKN: Sure. But if the driver isn't loading I would looks at the driver's _probe function to see if it's ever called or if it errors out, but if you don't see error messages it feels like the _probe isn't even called.
<roussinm>
modprobe the driver?
<roussinm>
manually.
<roussinm>
AKN: Do you know if the driver is built-in or a module?
<AKN>
roussinm: module
<AKN>
roussinm: brcmfmac43455
<roussinm>
ok then you should try to manually load the driver to see if binds to the device.
<AKN>
roussinm: Sure
nemik has quit [Ping timeout: 256 seconds]
nemik has joined #yocto
nemik has quit [Ping timeout: 256 seconds]
nemik has joined #yocto
sakoman has quit [Quit: Leaving.]
AKN has quit [Ping timeout: 244 seconds]
jclsn has quit [Ping timeout: 244 seconds]
jclsn has joined #yocto
amitk has joined #yocto
AKN has joined #yocto
otavio has quit [Ping timeout: 248 seconds]
nemik has quit [Ping timeout: 256 seconds]
nemik has joined #yocto
nemik has quit [Ping timeout: 256 seconds]
nemik has joined #yocto
AKN has quit [Ping timeout: 248 seconds]
otavio has joined #yocto
otavio_ has joined #yocto
otavio has quit [Read error: Connection reset by peer]
_whitelogger has joined #yocto
atomic_null[m] has quit [*.net *.split]
futureposer[m] has quit [*.net *.split]
ahs3[m] has quit [*.net *.split]
mootikins[m] has quit [*.net *.split]
denix has quit [*.net *.split]
RP has quit [*.net *.split]
denix has joined #yocto
alicef has quit [Quit: install gentoo]
willykin[m] has joined #yocto
RP has joined #yocto
alicef has joined #yocto
mohamed-dhiamtir has quit [*.net *.split]
clifonlintic[m] has quit [*.net *.split]
derRichard has quit [*.net *.split]
Tartarus has quit [*.net *.split]
iokill has quit [*.net *.split]
elfenix[m] has quit [*.net *.split]
yocton has quit [*.net *.split]
CosmicPenguin has quit [*.net *.split]
halstead has quit [*.net *.split]
awafaa has quit [*.net *.split]
mkorpershoek has quit [*.net *.split]
darknighte has quit [*.net *.split]
flynn378 has quit [*.net *.split]
stkw0 has quit [*.net *.split]
Ch^W has quit [*.net *.split]
yudjinn[m] has quit [*.net *.split]
iokill has joined #yocto
yocton has joined #yocto
darknighte has joined #yocto
stkw0 has joined #yocto
Ch^W has joined #yocto
mkorpershoek has joined #yocto
flynn378 has joined #yocto
halstead has joined #yocto
CosmicPenguin has joined #yocto
derRichard has joined #yocto
Lcvette[m] has joined #yocto
awafaa has joined #yocto
mait[m] has joined #yocto
dwagenk has joined #yocto
khem has joined #yocto
ejoerns[m] has joined #yocto
atomic_null[m] has joined #yocto
futureposer[m] has joined #yocto
camus has quit [Remote host closed the connection]
camus has joined #yocto
mootikins[m] has joined #yocto
ahs3[m] has joined #yocto
clifonlintic[m] has joined #yocto
mohamed-dhiamtir has joined #yocto
elfenix[m] has joined #yocto
Tartarus has joined #yocto
madisox_ has joined #yocto
rburton_ has joined #yocto
yudjinn[m] has joined #yocto
jbronder has joined #yocto
zbr_ has joined #yocto
rburton has quit [Ping timeout: 272 seconds]
willykin[m] has quit [Ping timeout: 272 seconds]
rburton_ is now known as rburton
mborges has quit [Ping timeout: 272 seconds]
jquaresma[m] has quit [Ping timeout: 272 seconds]
ramacassis[m] has quit [Ping timeout: 272 seconds]
jsbronder has quit [Ping timeout: 272 seconds]
zibri has quit [Ping timeout: 272 seconds]
madisox has quit [Ping timeout: 272 seconds]
sgw has quit [Ping timeout: 272 seconds]
madisox_ is now known as madisox
AKN has joined #yocto
sgw has joined #yocto
camus1 has joined #yocto
camus has quit [Remote host closed the connection]
camus1 is now known as camus
mborges has joined #yocto
goliath has joined #yocto
kroon has joined #yocto
jquaresma[m] has joined #yocto
nemik has quit [Ping timeout: 248 seconds]
nemik has joined #yocto
ramacassis[m] has joined #yocto
willykin[m] has joined #yocto
nemik has quit [Ping timeout: 246 seconds]
nemik has joined #yocto
OnkelUlla has joined #yocto
alessioigor has joined #yocto
alessioigor has quit [Client Quit]
kroon has quit [Quit: Leaving]
kroon has joined #yocto
rob_w has joined #yocto
jclsn has quit [Ping timeout: 276 seconds]
<deuteron>
How can I remove -src and -dbg packages from an SDK? I tried TOOLCHAIN_TARGET_TASK:remove but it doesn't do the trick.
frieder has joined #yocto
jclsn has joined #yocto
xenador77 has quit [Remote host closed the connection]
mckoan|away is now known as mckoan
Schlumpf has joined #yocto
nemik has quit [Ping timeout: 246 seconds]
hcg has joined #yocto
nemik has joined #yocto
nemik has quit [Ping timeout: 248 seconds]
nemik has joined #yocto
gsalazar has joined #yocto
GNUmoon has quit [*.net *.split]
granjow has joined #yocto
mvlad has joined #yocto
vladest has joined #yocto
Baehrune has joined #yocto
granjow has quit [Ping timeout: 240 seconds]
prabhakarlad has joined #yocto
nemik has quit [Ping timeout: 240 seconds]
nemik has joined #yocto
jclsn has quit [Ping timeout: 268 seconds]
nemik has quit [Ping timeout: 264 seconds]
nemik has joined #yocto
nemik has quit [Ping timeout: 248 seconds]
nemik has joined #yocto
ptsneves has joined #yocto
jclsn has joined #yocto
goliath has quit [Quit: SIGSEGV]
nemik has quit [Ping timeout: 264 seconds]
goliath has joined #yocto
nemik has joined #yocto
xenador77 has joined #yocto
leon-anavi has joined #yocto
nemik has quit [Ping timeout: 264 seconds]
<LetoThe2nd>
yo dudX
nemik has joined #yocto
<bigendian>
good cest morning all
<bigendian>
i prepared a layout by a wks file, not clear btw why /dev/root mounted on /, by df is always 91% in use
<bigendian>
checking by "du", direcories seems to take 10% maybe
twnqx has joined #yocto
jclsn has quit [Ping timeout: 268 seconds]
ptsneves has quit [Ping timeout: 252 seconds]
kayterina[m] has joined #yocto
nemik has quit [Ping timeout: 264 seconds]
nemik has joined #yocto
granjow has joined #yocto
<granjow>
Hello! I'm having troubles adding a menu entry to the GRUB menu which is shown for the live install. I tried manually adding an entry with a custom recipe which inherits grub-efi-cfg and defined a build_efi_cfg, but that does not generate grub.cfg anymore. How do I correctly add a GRUB entry to the live installation?
<kayterina[m]>
hello, can I have a link to devtool's source code please?
<dvorkindmitry>
In my user-space recipe I need to build branch0 if I have kernel 5.4 and branch1 if I have kernel 5.10. How to do that?
<LetoThe2nd>
kayterina[m]: i just happen to know, tbh. but generally: it can be called as a binary, so obviously at least somewhere a file of the name "devtool" must exist in the tree. then you can just search for it, using "find", and then wordk your way from there.
nemik has joined #yocto
<LetoThe2nd>
dvorkindmitry: what triggers the kernel version? use the same for the recipe.
<dvorkindmitry>
LetoThe2nd, preffered_version
<LetoThe2nd>
dvorkindmitry: set where?
<dvorkindmitry>
LetoThe2nd, in machine settings file at machine/
<LetoThe2nd>
dvorkindmitry: then why not set SRCBRANCH for the recipe there too? its hacky (such doesn't belong into the machine, rather the distro), but its the closest that you'll get.
<dvorkindmitry>
LetoThe2nd, I need to switch between two branches depending on one or another kernel version automatically
<LetoThe2nd>
dvorkindmitry: thats just not how it works.
<dvorkindmitry>
LetoThe2nd, ok. what would you suggest? my app sets kernel-dependant settings
<LetoThe2nd>
dvorkindmitry: a recipe does not trigger on things that another recipe does. the config can set things. so if the config can set a preferred version for the kernel, why can't it also set the branch? if you want more magic, then you can use some form of variable expansion based on the preferred kernel version that you set, but it just doesn't happen in the recipe.
<LetoThe2nd>
dvorkindmitry: i already suggested something, by the way.
<LetoThe2nd>
generally it would be better to trigger such off DISTRO, but hey, whatever works for you.
granjow has quit [Ping timeout: 268 seconds]
vladest has quit [Remote host closed the connection]
vladest has joined #yocto
starblue has quit [Ping timeout: 276 seconds]
starblue has joined #yocto
florian has joined #yocto
Baehrune has joined #yocto
florian_kc has joined #yocto
ptsneves has joined #yocto
selff has joined #yocto
ptsneves has quit [Quit: Client closed]
AKN has quit [Ping timeout: 276 seconds]
KanjiMonster has joined #yocto
dz1 has joined #yocto
adamtrhon has joined #yocto
<selff>
hello everyone. im having a problem when switching between hotspot and wpa. wpa is reading the wrong file when trying to switch. my configs are like this:
<selff>
when i do "systemctl start hostapd", the hotspot stand up successfully and just reads "hostapd.network". when i try switch to wpa, wpa is reading "hostapd.network" and "wlan.network". i couldnt find a solution.
granjow has joined #yocto
<granjow>
Could now add the custom GRUB line by copying the modified grub-efi-cfg.bbclass to my layer directory and prepending the path to BBPATH.
granjow has left #yocto [#yocto]
granjow has joined #yocto
nemik has quit [Ping timeout: 264 seconds]
nemik has joined #yocto
tlwoerner has joined #yocto
nemik has quit [Ping timeout: 240 seconds]
nemik has joined #yocto
AKN has joined #yocto
Guest82 has joined #yocto
<KanjiMonster>
hi. I'm building kirstone for an arm cortex-a9 platform, and python3-cryptography is causing python3 to coredump with SIGILL on load. using gdb it shows it crashes on a neon instruction. DEFAULTUNE is just "cortex-a9", so there should be no neon
<KanjiMonster>
the source that was compiled is rust, so my first instinct would be blaming the rust (cross-)compiler, but I'm way out of my league there to debug this, so I could use a helping hand
xmn has quit [Quit: ZZZzzz…]
nemik has quit [Ping timeout: 246 seconds]
nemik has joined #yocto
nemik has quit [Ping timeout: 240 seconds]
nemik has joined #yocto
granjow has left #yocto [#yocto]
<Guest82>
Hi all, I've built a kernel module in my layer recipe based on imx_rpmsg_tty. It autoloads and runs just fine, but issues a warning that it taints the kernel.
<Guest82>
It seems that any out-of-tree module will taint the kernel when googling for information. Did I miss a step or is this normal behaviour?
<manuel1985>
Guest82: I think that's normal behaviour.
selff has quit [Quit: Client closed]
Guest96 has joined #yocto
<Guest96>
Hi All
<Guest96>
I am also using yocto build system for my project
<Guest96>
Now i stuck with creating master recipe(Ex:core-image-minimal.bb)
Guest96 has quit [Client Quit]
te-johan has joined #yocto
<roussinm>
I think what we are doing right now is broken but let's see. We share sstate-cache for our build-runner. We have 2 runners x86_64, but different CPUs. And I think that native-recipe could in theory compile CPU instructions that are not compatible with the other CPU? Am I crazy? Python or CMake as an example.
<rburton>
roussinm: no, host architecture is part of the key, so different host arch won't share sstate
<rburton>
roussinm: the autobuilder has a single sstate cache on NFS for ~20 machines, some x86-64 some arm, different host distributions, different target distributions and machines
<roussinm>
rburton: ok so a runner when executing a python3 native on a host get illegal instruction. but on the other runner it passes.
<roussinm>
during a do_configure task.
<te-johan>
Hi, i'm getting pseudo abort errors due to ld.so.conf. Seems like the file is created twice, the second time ld.so.conf.prelink is created and renamed to ld.so.conf. When I rebuild my image I get a pseudo abort because the database contains the inode of the file before it was renamed. Does this sound familiar to anyone?
<rburton>
roussinm: fun! so the code is doing -mtune=native somewhere and that *will* break.
<rburton>
roussinm: what recipe?
<roussinm>
mozjs on hardknott, it execute some python code.
<rburton>
do you know what it executes?
<roussinm>
rburton: TMPDIR="${B}/lcl_tmp" ${S}/js/src/configure ${EXTRA_OECONF} this line inside the do_configure task.
<rburton>
i meant, can you identify what is actually crashing
RP has quit [Ping timeout: 248 seconds]
<deuteron>
How can I remove -src and -dbg packages from an SDK? I want just ${PN} and ${PN}-dev in the SDK, but I want the packages for the -src and -dbg packages. I tried TOOLCHAIN_TARGET_TASK:remove but it doesn't do the trick. Any ideas?
RP has joined #yocto
<RP>
rburton: python dependency creep in core is depressing :(
<rburton>
yes
<rburton>
RP: give it a year and half of these tools will die off, hopefully. i was told last week that 'poetry is winning' but recipe stats say otherwise so far.
<RP>
rburton: I guess there is a cycle to these things
<roussinm>
rburton: run.do_configure.9144: line 163: 10078 Illegal instruction (core dumped)
<roussinm>
is that line 163 of the script do_configure or the python from inside mozilla?
<rburton>
the configure, hopefully :)
m4ho has quit [Ping timeout: 256 seconds]
<roussinm>
so that's the line TMPDIR="${B}/lcl_tmp" ${S}/js/src/configure ${EXTRA_OECONF}, but I have no idea which instruction
<rburton>
dmesg will tell you what actually crashed
xmn has joined #yocto
kroon has quit [Quit: Leaving]
<roussinm>
rburton: I wish, it looks like I have no message from the runner, they are inside a docker, and looks like I don't see any kernel.
Guest82 has quit [Quit: Client closed]
<roussinm>
erros.
<rburton>
roussinm: what distros is the docker running?
<roussinm>
debian stretch
Schlumpf has quit [Quit: Client closed]
seninha has joined #yocto
nemik has quit [Ping timeout: 256 seconds]
nemik has joined #yocto
<rburton>
roussinm: can you pastebin the log.do_configure please
Baehrune has quit [Ping timeout: 240 seconds]
nemik has quit [Ping timeout: 268 seconds]
nemik has joined #yocto
<rburton>
rburton: even better, the log.do_compile for python3-native
xenador77 has quit [Remote host closed the connection]
xenador77 has joined #yocto
Lemur-for-rent has joined #yocto
<rburton>
roussinm: just checking: you didn't do something like set -mtune=native in the compiler flags anywhere did you, because that would cause all sorts of problems.
<rburton>
well, might be. but it thinks its finished.
florian has quit [Quit: Ex-Chat]
<roussinm>
well the configure scripts from mozilla execute some python
te-johan has quit [Quit: Client closed]
<rburton>
ah yes its a sh which calls python which calls sh
florian_kc has quit [Ping timeout: 240 seconds]
sakoman has joined #yocto
rob_w has quit [Quit: Leaving]
<roussinm>
rburton: when you said the sstate takes the host arch in consideration, it takes the instruction set extensions too? It's part of the hash?
<rburton>
no, just the triple
<rburton>
host code should never be compiled for a specific tune
<rburton>
the default in gcc is really generic
<roussinm>
it shouldn't for sure. But I read somewhere where python requires avx2 now.
<roussinm>
so is it crazy to think that python3-native choose to build with avx2 because it's avaible on the host, but then execute on my other machine that doesn't have avx2 and fails?
<rburton>
that shouldn't happen
<rburton>
it *could* but python shouldn't
paulg has quit [Ping timeout: 240 seconds]
hcg has quit [Quit: Client closed]
nemik has quit [Ping timeout: 256 seconds]
nemik has joined #yocto
nemik has quit [Ping timeout: 264 seconds]
nemik has joined #yocto
<roussinm>
rburton: i guess i could objdump the python3 binary and check if there is avx2 instructions...
AKN has quit [Read error: Connection reset by peer]
<roussinm>
that can't work, python3 binary too small.
vmeson has quit [Quit: Konversation terminated!]
<rburton>
roussinm: you'll likely want the libpython.so anyway
<roussinm>
probably ya
nemik has quit [Ping timeout: 240 seconds]
nemik has joined #yocto
<vin[m]>
I see mentions of using bitbake as a server and communicating over some rpc (xml rpc but migrating to something else). Is there some documentation on how to set this up.
nemik has quit [Ping timeout: 246 seconds]
nemik has joined #yocto
vmeson has joined #yocto
<roussinm>
rburton: no usage of the ymm register. So probably no avx
<roussinm>
runner cpus ares: Xeon-X5675 and a Xeon E5-2630 v2
<RP>
if we know the configure site for the avx2 bits we could disable that in python3-native
<rburton>
maaaaybe. shouldn't be using those instructions. easily tested if you can grab the python from one machine and run it on another
<rburton>
i had a 2630 iirc, and yes it's quite old. couldn't build clear linux as their libc was tuned to assume avx...
adamtrhon has quit [Ping timeout: 246 seconds]
hcg has joined #yocto
nemik has quit [Ping timeout: 256 seconds]
nemik has joined #yocto
nemik has quit [Ping timeout: 240 seconds]
nemik has joined #yocto
Minvera has joined #yocto
jclsn has quit [Ping timeout: 244 seconds]
prabhakarlad has joined #yocto
jclsn has joined #yocto
mckoan is now known as mckoan|away
seninha has joined #yocto
jclsn has quit [Ping timeout: 244 seconds]
nemik has quit [Ping timeout: 256 seconds]
nemik has joined #yocto
nemik has quit [Ping timeout: 240 seconds]
nemik has joined #yocto
seninha has quit [Quit: Leaving]
<roussinm>
rburton: from sstate-cache which task I should unzip, the only one I find is populate-sysroot and license stuff. Does populate-sysroot task make sens to unzip?
hcg has quit [Quit: Client closed]
jclsn has joined #yocto
xenador77 has quit [Remote host closed the connection]
<roussinm>
rburton: I extract the recipe to the host, but when I try to execute the python3.9 execute from the sysroot, it tells me No such file or directory
Wouter0100 has quit [Remote host closed the connection]
Wouter0100 has joined #yocto
fitzsim has quit [Ping timeout: 240 seconds]
frieder has quit [Remote host closed the connection]
<rburton>
roussinm: you'll need to fixup paths and stuff. much easier to do in a bitbake context
<Lemur-for-rent>
So how does that translate into the best way to modify it with yocto? The `ROOTFS_POSTPROCESS_COMMAND`?
<Lemur-for-rent>
Is there like a recipe local equivalent of that?
<Lemur-for-rent>
I guess I could expand the install task and run a sed like command assuming that the unit file will be in place after the original install task is done
<gchamp_>
personally I'd create a bbappend for systemd, then use a patch to change systemd-networkd-wait-online.service.in before systemd gets compiled and installed: https://docs.yoctoproject.org/singleindex.html#patching
<Lemur-for-rent>
* could expand the install task in a bbappend
seninha_ has quit [Quit: Leaving]
dvorkindmitry has joined #yocto
seninha has joined #yocto
<Lemur-for-rent>
gchamp_: Fair enough, thanks. It's just kind of difficult to reason with patches when reading them the first time
goliath has joined #yocto
<gchamp_>
Lemur-for-rent: agreed. What you suggested would work too.
jclsn has quit [Ping timeout: 255 seconds]
florian_kc has joined #yocto
fitzsim has joined #yocto
mvlad has quit [Remote host closed the connection]
<Lemur-for-rent>
Hm...
<Lemur-for-rent>
Seem like it can't fine the file in the `do_install_append` stage
Minvera has quit [Remote host closed the connection]
seninha has quit [Ping timeout: 240 seconds]
<roussinm>
Lemur-for-rent: is it a new file that you have next to your bbappend?
Lemur-for-rent has quit [Ping timeout: 252 seconds]
nemik has quit [Ping timeout: 246 seconds]
nemik has joined #yocto
prabhakarlad has joined #yocto
<roussinm>
rburton: so looking at a core dump: 0x7ffa735e163c <ffi_prep_cif+12> vmovq %r8,%xmm1 that's the invalid instruction, from C. So conclusion something is happening during configuration so that it can configure with avx.
florian_kc has joined #yocto
roussinm has quit [Quit: WeeChat 3.0]
roussinm has joined #yocto
mrnuke has quit [Ping timeout: 256 seconds]
mrnuke has joined #yocto
<roussinm>
rburton: don't really know if it's a proper fix but: EXTRA_OECONF_class-native += "--with-gcc-arch=generic" to the recipe, fixed the problem for us. What do you think?
<rburton>
sounds like a proper fix to me
<rburton>
i wonder if that's new behaviour to libffi
<rburton>
good work chasing it down
<rburton>
can you send a patch?
<roussinm>
rburton: I will send a patch, but probably tmr.
leonanavi has quit [Remote host closed the connection]
nemik has quit [Ping timeout: 240 seconds]
nemik has joined #yocto
nemik has quit [Ping timeout: 240 seconds]
nemik has joined #yocto
<RP>
roussinm: sounds like a good fix to me too fwiw
<roussinm>
RP: Thanks for the additional feedback.