<yoCuda>
But it is not clear how to do it, I only get the compilation of CUDA samples, but how do you add the driver and other cuda libraries?
<yoCuda>
hope someone has done it before!
<yoCuda>
I tried to research this topic, but I am disappointed from the little interest in this topic. Although, Nvidia Cards are being used to drive more applications than ever. (e.g. AI applications)
<yoCuda>
hope someone had tackled the problem and resolved it!
<yoCuda>
e.g. if I have the correct kernel headers/gcc and toolchain installed, can I use the runfile installers for driver and cuda regardless of my sysroot?
Vonter has quit [Read error: Connection reset by peer]
Vonter has joined #yocto
Schlumpf has joined #yocto
camus1 has joined #yocto
camus has quit [Ping timeout: 272 seconds]
camus1 is now known as camus
rob_w has joined #yocto
goliath has joined #yocto
mckoan|away is now known as mckoan
<mckoan>
good morning
frieder has joined #yocto
<mckoan>
yoCuda: this is the Yocto Project channel therefore it is supposed you are using meta-cuda, isn't it?
<yoCuda>
mckoan morning, no I am not using meta -tegra. I am running an x86 machine with Tesla T4 card, and i just want to install its drivers and cuda libraries.
<mckoan>
yoCuda: probably you are in the wrong channel
<yoCuda>
what channel do u recommend?
crawlwer_ has joined #yocto
crawlwer_ has quit [Remote host closed the connection]
zpfvo has joined #yocto
leon-anavi has joined #yocto
zyga__ is now known as zyga
Vonter has quit [Ping timeout: 255 seconds]
Schlumpf has quit [Quit: Client closed]
alex88 has quit [Ping timeout: 268 seconds]
alex88 has joined #yocto
Schlumpf has joined #yocto
florian has joined #yocto
Schlumpf has quit [Quit: Client closed]
Schlumpf has joined #yocto
alex88 has quit [Ping timeout: 256 seconds]
alex88 has joined #yocto
crawlwer_ has joined #yocto
crawlwer_ has quit [Remote host closed the connection]
JaMa has joined #yocto
florian_kc has joined #yocto
camus1 has joined #yocto
camus has quit [Remote host closed the connection]
camus1 is now known as camus
Vonter has joined #yocto
Vonter_ has joined #yocto
Vonter has quit [Ping timeout: 255 seconds]
prabhakarlad has quit [Quit: Client closed]
yocton has joined #yocto
prabhakarlad has joined #yocto
rob_w has quit [Remote host closed the connection]
rob_w has joined #yocto
camus1 has joined #yocto
Schlumpf has quit [Quit: Client closed]
camus has quit [Ping timeout: 272 seconds]
camus1 is now known as camus
BCMM has joined #yocto
xmn has quit [Quit: ZZZzzz…]
tnovotny has joined #yocto
mihai has joined #yocto
LocutusOfBorg has quit [Quit: ZNC 1.8.2+deb1+bionic2 - https://znc.in]
LocutusOfBorg has joined #yocto
fotte has joined #yocto
Schlumpf has joined #yocto
<fotte>
Hello everyone. I cant find the documentation for configuring my package through local.conf. I am looking for an explanation of the correct format of ${VARNAME}_pn-${PN} (e.g. DEBUG_BUILD_pn-linux-mainline) - can someone point me in the right direction?
<qschulz>
I think it's the recipe name after _pn-
<fotte>
I believe so too, but a doc would be great. I currently have the issue, that DEBUG_BUILD_pn-linux-mainline should be set, and therefore a config-fragment should be added to the kernel-config. But this does not happen. And i cannot see 'DEBUG_BUILD*' in the environment (e.g. by running bitbake virtual/kernel -e)
<qschulz>
seems like it's missing from the bitbake doc
<qschulz>
are you sure linux-mainline is taken when you build virtual/kernel?
<qschulz>
just to be sure, use linux-mainline directly, at least it's one less possible issue
camus has quit [Ping timeout: 276 seconds]
camus1 is now known as camus
yoCuda has quit [Quit: Client closed]
<fotte>
Had a tainted virtual/kernel, did a cleanall and now i wait for a result. DEBUG_BUILD is now part of my environment, I can see the operations which are executed to buildup the src_uri variable. Is it possible to show the result of these operations?
likewise has joined #yocto
<qschulz>
fotte: bitbake -e should show this?
bjobjo has quit [Quit: leaving]
kroon has joined #yocto
<fotte>
qschulz: Indeed! However the output is unexpected. I am using a .bbappend file: https://pastebin.com/gWyzX12F the output of bitbake -e an operation containing 'my-debug'. It is part of the pre-expansion value, but it is not part of the expanded SRC_URI
<qschulz>
fotte: you need to look for a line starting with DEBUG_BUILD= in bitbake -e
<qschulz>
DEBUG_BUILD_pn-linux-mainline="1" is likely to appear for any recipe since it's part of the a configuration file
<fotte>
qschulz: Now i am confused. Maybe I should state my assumptions: in local.conf: FOOVAR_pn-barrecipe = "my_value", when parsing barrecipe I can check for FOOVAR by using: SRC_URI += "${@bb.utils.contains('FOOVAR', 'my_value', ' ', 'file://...', d)}
rob_w has quit [Quit: Leaving]
jwillikers has joined #yocto
<qschulz>
fotte: what I meant is that build bitbake virtual/kernel -e | grep DEBUG_BUILD should actually be build bitbake virtual/kernel -e | grep DEBUG_BUILD=
<qschulz>
fotte: also... it probably works since my-debug.cfg is not in your SRC_URI?
<qschulz>
if you set DEBUG_BUILD_pn-linux-mainline="0" in your local.conf, does it appear?
<fotte>
craziest thing. Setting DEBUG_BUILD_pn-linux-mainline = "0", does work - now my-debug is part of SRC_URI, and is listed as a configuration fragment. But why?!
<qschulz>
fotte: because you inverted the logic
<qschulz>
c.f. link posted above
<fotte>
aaarrrgss shoot me. wrong order of arguments to the contains function… Thank you so much!
frieder_ has quit [Ping timeout: 276 seconds]
<qschulz>
fotte: my pleasure :)
frieder has joined #yocto
<qschulz>
rburton: quite unlucky for the moment with containers supposedly able to compile poky with rootless podman. Tried pyrex, issue with SSH agent forwarding. Just tried crops/poky, even worse, unable to write in the workdir. I guess it's time for some elbow grease :)
likewise has quit [Remote host closed the connection]
likewise has joined #yocto
Schlumpf has quit [Quit: Client closed]
prabhakarlad has quit [Quit: Ping timeout (120 seconds)]
<kroon>
Is there a freescale/nxp dedicated yocto forum or mailing list somewhere ? meta-freescale mailing list seems abandoned
hpsy has joined #yocto
eduardas has joined #yocto
<JPEW>
qschulz: pyrex *should* work, but it's not a very common use case, so it probably needs a better test case if you can figure out what's wrong and fix it :/
<qschulz>
JPEW: yeah, still planning to investigate a bit, but felt like crops was nicer to use in jenkins?
<JPEW>
qschulz: ah ya. Probably. Pyrex is more for developers than CI, although we use it on CI FWIW, so it's not impossible
<likewise>
kroon: for what architecture? i.MX?
<qschulz>
JPEW: yeah, probably going to end up using pyrex for myself, and crops on jenkins. At least that's the plan :)
jwillikers has quit [Remote host closed the connection]
<kroon>
likewise, I tried emailing "meta-freescale@lists.yoctoproject.org" but got a reply that the mail couldn't be delievered, "meta-freescale wasn't found at lists.yoctoproject.org."
jwillikers has joined #yocto
yates has joined #yocto
<yates>
when i put 'FILES_glibc-locale += "/usr/lib/locale/"' in my glibc-locale_2.32.bbappend, it appears that it is getting replaced by bitbake from the build message 'WARNING: Variable key FILES_${PN} (${bindir}/* ...) replaces original key FILES_glibc-locale (/usr/lib/locale/).'
<yates>
how do i get my FILES_glibc-locale to "stick"?
<yates>
this is in a "bitbake core-image-minimal"
<yates>
the end result being that i get a QA error that /usr/lib/locale/ is not shipped in any package
<yates>
please help. as you know i've been fighting this problem for weeks and i'm at my wit's end with it.
<yates>
that was the output of bitbake -c populate_sdk_ext core-image-minimal -e | grep FILES | fpaste
<qschulz>
yates: bitbake -e on glibc recipe ;)
<yates>
qschulz: good idea
<yates>
hang on - i'm checking out my BBFILES setting. it could be hosered
fotte has quit [Ping timeout: 255 seconds]
zyga_ has joined #yocto
zyga has quit [Ping timeout: 268 seconds]
<yates>
rburton: what did you say the idiom is for BBFILES? something like /*/recipes-*/*.bbappend ?
<rburton>
have a look at literally any layer
<rburton>
recipes-*/*/*.bb recipes-*/*/*.bbappend
<yates>
k
zyga_ is now known as zyga
Tokamak has joined #yocto
camus1 has joined #yocto
camus has quit [Ping timeout: 276 seconds]
camus1 is now known as camus
v2d has joined #yocto
Tokamak_ has joined #yocto
<yates>
if a build folder has a conf/bblayers.conf with a BBFILES += "${LAYERDIR}/..." defined, will that BBFILES get concatenated to any BBFILES += "${LAYERDIR}/..." defined in a layer's conf/layer.conf?
<yates>
is it valid for the build folder to have a BBFILES definition?
Tokamak has quit [Ping timeout: 258 seconds]
tnovotny has quit [Quit: Leaving]
<rburton>
the default BBFILES in bblayers.conf is "". LAYERDIR doesn't make any sense in the context of bblayers.conf. What layer?
<rburton>
maybe look at what a fresh poky does for bblayers.conf and eg meta-poky/conf/layer.conf.
sakoman has joined #yocto
dlan has joined #yocto
Tokamak has joined #yocto
xmn has joined #yocto
Tokamak_ has quit [Ping timeout: 258 seconds]
Neur0mante has joined #yocto
frieder has quit [Ping timeout: 272 seconds]
leon-anavi has quit [Quit: Leaving]
<yates>
l
<yates>
k
frieder has joined #yocto
<yates>
ok my BBFILES ?= "" in my build/conf/bblayers.conf
frieder has quit [Remote host closed the connection]
<yates>
and i changed glibc-locale to ${PN}. that got rid of the "WARNING ... replaces original key" warning, but i still get the exact sameQA error
roussinm has joined #yocto
<yates>
let me recheck the -e output
kayterina has quit [Remote host closed the connection]
jmiehe has joined #yocto
florian has quit [Quit: Ex-Chat]
florian_kc has quit [Ping timeout: 272 seconds]
mithro_ is now known as mithro
kostrak has joined #yocto
Vonter_ is now known as Vonter
Neur0mante has quit [Quit: Client closed]
camus has quit [Quit: camus]
camus has joined #yocto
Tokamak has quit [Ping timeout: 258 seconds]
dev1990 has joined #yocto
eduardas has quit [Quit: Konversation terminated!]
v2d has quit [Quit: Client closed]
zpfvo has quit [Remote host closed the connection]
<yates>
that ^^^ is the output from bitbake glibc-locale -e | grep FILES >~/files0.txt
<yates>
(i.e., files0.txt)
<yates>
again i do not know which of these may be relevent. this line looks suspicious: override[glibc-locale]:rename from FILES_${PN} data.py:104 [expandKeys]
kroon has quit [Quit: Leaving]
fancer_ has quit []
fancer has joined #yocto
camus1 has joined #yocto
camus has quit [Ping timeout: 256 seconds]
camus1 is now known as camus
nerdboy has quit [Remote host closed the connection]
nerdboy has joined #yocto
<yates>
rburton: do you have any more thoughts?
<BuZZ-T>
can anyone give me hint how to debug devtool? when i try "devtool modify virtual/kernel" i get an error printed out "IndexError: list index out of range"
<yates>
ok so my FILES is working, see the very last path here: FILES_glibc-locale="/usr/bin/* /usr/sbin/* /usr/libexec/* /usr/lib/lib*.so.* /etc /com /var /bin/* /sbin/* /lib/*.so.* /lib/udev /usr/lib/udev /lib/udev /usr/lib/udev /usr/share/glibc /usr/lib/glibc/* /usr/share/pixmaps /usr/share/applications /usr/share/idl /usr/sha
<BuZZ-T>
okay.. the problem was that in the kernel Makefile (./build/tmp/work-shared/ohp-vf/kernel-source/Makefile) was defined as an empty string.
<BuZZ-T>
when i set it manually to something the devtool modify is working
<yates>
rburton: ?
<BuZZ-T>
hmm.. sorry it has not worked :/
jwillikers has quit [Remote host closed the connection]
<yates>
BuZZ-T: sounds like you're having about as much fun as I am...
<BuZZ-T>
upgrading between yocto versions is always fun :P but fortunately its all open source.. so i need to dig a little bit arround :)
<yates>
yeah, it's only code...
<yates>
1000s and 1000s of lines of code, but code...
<BuZZ-T>
at least i found out whats wrong.. but not why it was like that :)
<yates>
BuZZ-T: i wish you luck! you will find it soon!
<BuZZ-T>
so refetching kernel sources.. that take a few minutes
<yates>
you're not using DL_DIR?
<BuZZ-T>
sure i do, but the cleanall was my last attempt to get it working
<BuZZ-T>
somehow the created kernel sources doesn't have EXTRAVERSION set.. and cleansstate brought another error
<yates>
i've eyed yocto's state and associated automatic (?) dependencies with a grain of salt for years now
jwillikers has joined #yocto
<yates>
if anyone is still listening, i suspect my glibc-locale QA problem is not fixed with a FILES_${PN} = blah because there is no glibc-locale package per se; the glibc-locale packaging splits it into a whole bunch of other packages.
<yates>
(${PN} == glibc-locale)
<yates>
i tried "FILES_glibc-gconvs = blah" because i saw the glibc-gconv package get generated in the do_package routine (because i hacked it with a bb.note())
<yates>
that did not work, but it is appreantly because of another deeper problem with the csky architecture setup
<yates>
in other words, i think i'm edging closer to the final solution
likewise has quit [Remote host closed the connection]
Guest3848 has joined #yocto
Guest3848 has quit [Client Quit]
glumanda has joined #yocto
camus1 has joined #yocto
leonanavi has joined #yocto
camus has quit [Ping timeout: 276 seconds]
camus1 is now known as camus
pbergin has joined #yocto
leon-anavi has quit [Ping timeout: 276 seconds]
pbergin has quit [Client Quit]
glumanda has quit [Quit: WeeChat 3.2]
v2d has quit [Quit: Client closed]
v2d has joined #yocto
v2d has quit [Quit: Client closed]
v2d has joined #yocto
BuZZ-T has quit [Quit: BuZZ-T has left the building...]
<jordemort>
off chance but anyone ever had any problems running GPGME (specifically, ostree) inside an initramfs? is it because i don't have a real init?
<jordemort>
figured it out, machine thought it was 1970 and the key wasn't valid yet
prabhakarlad has quit [Quit: Client closed]
prabhakarlad has joined #yocto
florian_kc has quit [Ping timeout: 255 seconds]
<paulg>
good old dead battery/epoch problem never gets old.
<paulg>
I guess 1970 finally goes away in 2038.
goliath has quit [Quit: SIGSEGV]
<RP>
paulg: it is still "0" isn't it? :)
<RP>
Has anyone else seen boot time socket activation issues with systemd and ssh which may or may not be arm specific?
<RP>
(ssh connection made early in boot just hangs)
<paulg>
not the classic "no entropy on new rootfs" problem?
florian_kc has joined #yocto
<RP>
lukma: FWIW I fixed that eSDK pseudo issue
<RP>
paulg: connection hanging for 300s then new connections ok?
<paulg>
if your only entropy is coming from network interrupts, then it can look like that....
leonanavi has quit [Quit: Leaving]
<RP>
paulg: it has the virtio rng too I think :/
<paulg>
in theory, if you have a console, you should see it sitting there pining for the fijords while it waits for entropy to create the host key on initial boot, I would think.
<paulg>
but I'm a kernel guy, so what do I know?
<RP>
paulg: The logs show it reaching a serial login with systemd saying network enabled but the kernel logs look truncated to me :/
<paulg>
if you reboot with the same "already booted once" filesystem, is the problem gone?
<RP>
paulg: history shows that this is somehow load related so rebooting with the original bnot booted filesystem works
<paulg>
RP, is it exactly 300s - and then fail (connection refused, or ... ?) ; or "feels like five minutes" and then lets you in, or .... ?
<RP>
paulg: our timeout kicks in and it gives up
<paulg>
ah, right - automated tests like the LTP poo I was swimming in. Got it.
<RP>
paulg: looking at the logs suggests openssh keygen didn't finish. The image that did work (1 out of the three) uses dropbear instead
<RP>
I think I need to check the virtio rng passthrough is actually there
<paulg>
yeah, that sounds about right - and makes me think of the "no entropy for keygen" issue as soon as I heard it.
<RP>
paulg: it is good thinking
<RP>
CONFIG_HW_RANDOM_VIRTIO=y and it is on the qemu commandline
amitk_ has quit [Ping timeout: 272 seconds]
<RP>
paulg: everything says this should all be enabled. Starting to wonder if it was just resource starved that keygen took longer than 5 mins :(