<Tyaku_>
In fact, as my image B had a require to image A, then if I set INCOMPATIBLE_LICENSE it is also taken by the softwares in image B, But for the image B I don't care about licence.
<Tyaku_>
Finally this is what I am doing:
<Tyaku_>
In the main image (A), I set the INCOMPATIBLE_LICENSE
<Tyaku_>
In the DEV image (B), I set WHITELIST_GPL-3.0 += "gdbserver"
vladest has quit [Quit: vladest]
vladest1 has joined #yocto
luc4 has joined #yocto
xmn has quit [Quit: ZZZzzz…]
vladest1 is now known as vladest
florian__ has joined #yocto
xmn has joined #yocto
florian__ has quit [Ping timeout: 252 seconds]
dvergatal has quit [Ping timeout: 256 seconds]
dvergatal has joined #yocto
vladest has quit [Quit: vladest]
vladest1 has joined #yocto
vladest1 has quit [Client Quit]
vladest has joined #yocto
Saur_Home has quit [Quit: Client closed]
Saur_Home has joined #yocto
rfuentess has quit [Remote host closed the connection]
leon-anavi has joined #yocto
rfuentess has joined #yocto
vladest has quit [Ping timeout: 252 seconds]
mbulut_ has joined #yocto
yannd has quit [Remote host closed the connection]
florian_kc has joined #yocto
g0hl1n has joined #yocto
ptsneves has joined #yocto
Guest66 has quit [Quit: Client closed]
<LetoThe2nd>
rburton: possibly stupid me, but shouldn't generic-arm64 be able to run on qemu too?
<RP>
LetoThe2nd: with the firmware. I was tempted to suggest we should add that...
<LetoThe2nd>
RP: ok guess I missed some important part of the presentation then. yay for virtual events alongside daily business.
<RP>
LetoThe2nd: there is a qemu machine in meta-arm which does this
luc4 has quit [Quit: Konversation terminated!]
<LetoThe2nd>
ah?
<LetoThe2nd>
RP: gotcha, thanks!
Chaser has quit [Remote host closed the connection]
Chaser has joined #yocto
vladest has joined #yocto
alperak has quit [Quit: Client closed]
<kanavin_>
RP: I need a bit of help understanding how siginfo files work. If I issue:
<kanavin_>
so for aarch, the sstate object is there, but the signature is not
<kanavin_>
Is this all working as expected? I find it royally confusing.
Saur_Home has quit [Quit: Client closed]
Saur_Home has joined #yocto
bhstalel has quit [Quit: Client closed]
prabhakarlad has quit [Ping timeout: 250 seconds]
florian_kc is now known as florian
prabhakar has quit [Ping timeout: 252 seconds]
<ray-san>
hi, can someone point me to a recipe which provides the file "/usr/lib/dri/radeonsi_drv_video.so"? i guess, i need that to get gpu acceleration in chromium working. i'm somewhat lost
lthadeus has joined #yocto
<RP>
kanavin_: the difference would be hash equivalence
alessioigor has quit [Quit: alessioigor]
alessioigor has joined #yocto
<kanavin_>
RP: I thought so. Then it must be that printdiff interacts badly with it.
<RP>
kanavin_: I think the issue is that the new hash is only available really late so we can change the filename but not the contents
<kanavin_>
RP: I don't follow, can you explain?
<RP>
kanavin_: siginfo files should always be there for corresponding objects. The fact it isn't is probably part of the issue and a sign it is being deleted
<RP>
kanavin_: when hash equivalence is found, it "ripples" through the dependency tree. For determinism (and sanity), the hash written into the file is the computed hash, not the unihash from hash equivalence
lthadeus has quit [Quit: Leaving]
<RP>
I'd also note that the siginfo code that writes these files predates hash equivalence
lthadeus has joined #yocto
<kanavin_>
RP: for 4b50 hash both signature and object are there:
<kanavin_>
(buildbot-venv) pokybuild@ubuntu2204-arm-1:/srv/autobuilder/autobuilder.yocto.io/pub/sstate$ ls universal/4b/50/sstate\:perlcross-native\:*
<rburton>
LetoThe2nd: yeah you want qemu-generic-arm64. normal generic-arm64 doesn't work because qemu doesn't come with firmware, you need to build it. that immediately removes the "generic" bit.
<kanavin_>
so it was there, and got pruned
<RP>
kanavin_: the timing sounds right too. That means we're not updating the timestamps the deletion script is using
<RP>
rburton: would firmware for qemu not be quite appropriate for that target though? Maybe as an optional thing which you can add into the build ?
<LetoThe2nd>
rburton: yeah understood. fun fact, built that and wondered why it still gave me an error. then realized that instead of tabbing i actually had typed the command, and rumqemu, well, ...
<rburton>
RP: well that's what qemu-generic-arm64 does. it just includes the generic-arm64 machine and sets up the firmware build.
<Rich_1234>
ray-san: You will most likely have to look at your BSP vendors layer and the graphics package they have. I wanted to get gpu acceleration in chrome recently on a TI AM68 and was told they would add support for it at the end of Q1 2024
<RP>
rburton: I'd still probably just include it in the one machine but I'm probably evil/wrong
<rburton>
we originally decided that felt very icky
<RP>
rburton: I think what would swing the balance for me is allowing plain genericarm64 to work on qemu. Not really my call though
* RP
wonders if we should be enabling PACKAGE_MINIDEBUGINFO as default
<ray-san>
Rich_1234: my platform is genericx86-64, so i guess, here is somewhat the right place?
<rburton>
ray-san: i'd imagine mesa, but you'd need to read the mesa documentation to determine what options you need to turn on
<rburton>
RP: the firmware recipes are of course in meta-arm as they're arm-specific
<ray-san>
rburton: i've turned pretty much every (sensible) option for mesa on, seems like none of the _drv_video files is built
<rburton>
ray-san: did you read the _mesa_ documentation and determine that the _recipe_ options cover what you need?
<ray-san>
rburton: do you mean from the original package maintainers? i am trying to figure out how those files get built at my debian host system, but i'm somewhat lost, mesa seems to be a monster of a package
<ray-san>
rburton: yeah, if i try to get the sources via "apt-get source mesa-va-drivers", it downloads the mesa source tarball, so i guess at least mesa is correct
<rburton>
you need to be sure you've turned on gallium+va
<rburton>
looks like thats a magic dependency in mesa, enabled if libva is present at build time
<ray-san>
i have this line in my local.conf "PACKAGECONFIG:append:pn-mesa = ' gallium-llvm'"
<rburton>
also add 'va'
<ray-san>
ok, i try
mrybczyn has joined #yocto
GNUmoon has quit [Remote host closed the connection]
GNUmoon has joined #yocto
perceval has joined #yocto
<ray-san>
rburton: looks like that did it, chromium doesn't run with hardware acceleration yet, but i have to recompile it with use-vaapi
bhstalel has joined #yocto
<rburton>
i definitely need a tip jar
<rburton>
moto-timo: just posted python3-validators to oe-devel but forgot to tag it as meta-python, just in case you miss it and khem leaves it to you :)
<rburton>
if i'm going to review and fix random recipes on SO i at least want to get the recipe public
<ray-san>
a lot of stuff is rebuilt, i guess this is a good sign :D
lthadeus has quit [Quit: Leaving]
<LetoThe2nd>
rburton: what is the boot flow of qemu-generic-arm64? straight to kernel? I see a /boot partition, though.
jmiehe has joined #yocto
Mike23 has joined #yocto
mrybczyn has quit [Quit: Client closed]
Net147 has quit [Ping timeout: 264 seconds]
<rburton>
LetoThe2nd: qemu - edk2 - grub - kernel
<LetoThe2nd>
rburton: thx
<rburton>
it emulates a systemready-sr machine using the sbsa-ref qemu machine
<LetoThe2nd>
rburton: well you can guess why i'm asking
gsalazar has joined #yocto
JerryM has joined #yocto
mrybczyn has joined #yocto
prabhakarlad has joined #yocto
michaelgalassi has joined #yocto
Mike23 has quit [Quit: Client closed]
Net147 has joined #yocto
Net147 has quit [Changing host]
Net147 has joined #yocto
bhstalel has quit [Quit: Client closed]
Mike23 has joined #yocto
dvergatal has quit [Ping timeout: 252 seconds]
goliath has joined #yocto
dvergatal has joined #yocto
<kanavin_>
sbom on zeus for a medical device question: face meet palm
<kanavin_>
I hope someone can write a better answer than mine
<JaMa>
hehe
dvergatal has quit [Ping timeout: 256 seconds]
<kanavin_>
I have to say there's a kind of poetic justice in that
alessioigor has quit [Quit: alessioigor]
<kanavin_>
neglect updating yocto ---> lose supplier contract because you can't provide sbom
alessioigor has joined #yocto
g0hl1n has quit [Quit: Client closed]
dvergatal has joined #yocto
<JaMa>
I fear they will hack something ugly together (like the original product) to keep the customer happy (and oblivious to how ugly and wrong the sbom they supplied is)
<JaMa>
if they didn't catch any of the well known security issues in zeus, then incorrect sbom won't be noticed as well
<kanavin_>
that's why there should be an independent audit too
<kanavin_>
sbom is only good if it can be verified
<JaMa>
and with independent audit, the missing or incorrect sbom might not be the worst issue they find :)
Saur_Home has quit [Quit: Client closed]
Saur_Home has joined #yocto
dvergatal has quit [Ping timeout: 268 seconds]
dvergatal has joined #yocto
alperak has quit [Quit: Client closed]
alperak has joined #yocto
bhstalel has joined #yocto
mrybczyn has quit [Quit: Client closed]
Chaser has joined #yocto
perceval has quit [Ping timeout: 256 seconds]
<rburton>
michaelo: i got distracted during some of your slides. what was the thing about storing the cover letter in the branch?
Daanct12 has quit [Quit: WeeChat 4.1.1]
<fray>
git branch --edit-description if I remember correctly
<rburton>
aha format-patch has cover-from-description, i never knew that
<rburton>
yes, useful
g0hl1n has joined #yocto
andreascian has joined #yocto
mbulut_ has quit [Ping timeout: 264 seconds]
jmiehe has quit [Quit: jmiehe]
lexano has quit [Ping timeout: 268 seconds]
Chaser has quit [Quit: Chaser]
rstreif has quit [Ping timeout: 260 seconds]
Chaser has joined #yocto
perceval has joined #yocto
Minvera has joined #yocto
prabhakarlad has quit [Quit: Client closed]
lexano has joined #yocto
prabhakarlad has joined #yocto
prabhakar has joined #yocto
wmills_ has joined #yocto
bunk has joined #yocto
prabhakarlad has quit [Quit: Client closed]
JaMa has quit [Read error: Connection reset by peer]
prabhakarlad has joined #yocto
JaMa has joined #yocto
<RP>
rburton: you can describe all your branches! :)
<rburton>
lol yeah right
lthadeus has joined #yocto
Xagen has joined #yocto
vladest has quit [Ping timeout: 268 seconds]
<chep>
Hi, I'm trying to create a new fetcher based on gitsm. I created a new file my_gitsm with a class MyGitSM(GitSM) which only reimplement the `supports` method. Then I created a bbclass to add my fetcher. When I try to fetch, I got this error : FetchError("Recipe uses a floating tag/branch 'master' for repo 'XXXX' without a fixed SRCREV yet doesn't call bb.fetch2.get_srcrev() (use SRCPV in PV for OE).", None)
<chep>
in my recipe I have PV = "git_${SRCPV}"
<chep>
what do I miss ?
<JaMa>
did you set SRCREV as well?
egueli-AV has quit [Quit: Leaving]
<Saur_Home>
chep: If you are using master or nanbield, the ${SRCPV} variable has been removed. bitbake now expects to find a + in PV if it should append the SHA-1 for the package version. E.g., use `PV = "1.2.3+git"` instead.
<chep>
JaMa: no but with gitsm it works and I'm just overriding `supports`
<chep>
Saur_Home: I'm on kirkstone
<chep>
JaMa: sorry yes it is set to ${AUTOREV}
alimon has quit [Ping timeout: 256 seconds]
<chep>
And logs are in a directory called git_\$\{SRCPV\}-r0 it's like SRCPV doesn't exist
yudjinn has joined #yocto
Guest67 has joined #yocto
luc4 has joined #yocto
bunk has quit [Quit: leaving]
<rburton>
can i just shout out to alperak for absolutely flooding meta-oe with upgrade patches! 👏
<alperak>
rburton It's an honor to hear this from you. :) Thank you!
<chep>
If I set a fixed SRCREV, I got the same error message…
alimon has joined #yocto
bhstalel has quit [Ping timeout: 250 seconds]
ptsneves has quit [Read error: Connection reset by peer]
Guest67 has quit [Quit: Client closed]
andreascian has quit [Ping timeout: 250 seconds]
lthadeus has quit [Ping timeout: 268 seconds]
lthadeus has joined #yocto
vladest has joined #yocto
Vonter has joined #yocto
bhstalel has joined #yocto
Mike23 has quit [Quit: Client closed]
zpfvo has quit [Ping timeout: 256 seconds]
zpfvo has joined #yocto
alessioigor has quit [Quit: alessioigor]
alessioigor has joined #yocto
<g0hl1n>
hi, what would be the best/fastest way to get the "expanded" (in this case $METADATA_REVISION) DISTRO_VERSION without triggering a build and parsing it from the "Build Configuration" output? Thanks :-)
<kanavin_>
RP: I'm embarrassed to ask, but how do I make queries to hash equiv server from the command line? I.e. "I have a hash, is there an equivalent one"?
zpfvo has quit [Ping timeout: 256 seconds]
zpfvo has joined #yocto
<RP>
kanavin_: I'm not sure, I don't think bitbake-hashclient can support it. JPEW?
<Saur_Home>
g0hl1n: Simplest is `bitbake-getvar --value DISTRO_VERSION`, but that of course uses tinfoil and thus parses the recipes if needed. However, without parsing the recipes, it would be hard to get a correct value in case it depends on other variables.
<JaMa>
chep: I guess your modified fetcher is missing something, there was something like supports_srcrev function which it needs to support etc, don't remember the detail and cannot look now, but that's where I would debug the fetcher code
Kubu_work has quit [Quit: Leaving.]
<RP>
kanavin_: I don't really understand how we can get there :/
<kanavin_>
RP: I suspect (but I don't have enough knowledge) that printdiff wants 'straight' signatures, and won't follow hashequiv shortcuts the way bitbake does in a build.
<JPEW>
The client should probably implement that
<JPEW>
But it doesn't today
<JPEW>
I'll put it on the TODO list
<RP>
kanavin_: That is a good theory. We could add some function calls to make it "resolve" those hashes?
<kanavin_>
RP: here's a guess: something forced quilt-native to rebuild, but the output was the same and thus registered in hashequiv. perlcross-native only depends on that, and so didn't rebuild, and wait straigh to populate_sysroot
<kanavin_>
makes sense, or still doesn't hold?
<RP>
kanavin_: it can't go to populate_sysroot without running compile/install though
<RP>
kanavin_: about to get pulled into meetings :(
cambrian_invader has joined #yocto
<kanavin_>
RP: if perlcross-native only needs quilt-native, and quilt-native's output has not changed, then why would it want to re-run compile/install, and not just take populate_sysroot. I guess I'll try to carry on with JPEW :)
<RP>
if it just takes populate_sysroot, how does the artefact get generated though?
hcg has quit [Quit: Client closed]
<yocton>
Is there a bug triage meeting now or it got cancelled by the Yocto summit?
<g0hl1n>
Saur_Home: thanks. That's what I've been looking for. Yes, parsing of the recipes would be needed. That's true. Thanks again :-)
* JPEW
needs to read the bug
<yocton>
(Nobody on the Zoom, I'll go with canceled :) )
<kanavin_>
RP First, quilt-native was built with input hash X and output A, and perlcross-native was build on top of that with input hash Y + A and output B.
<JPEW>
yocton: correct
<kanavin_>
RP Then quilt-native was rebuilt with input hash Z and same output A, thus perlcross-native would still have input hash Y + A, and thus there is no need to rebuild it, and we can take B from cache.
rfuentess has quit [Remote host closed the connection]
luc4 has quit [Ping timeout: 268 seconds]
<rburton>
I hate myself, but has anyone managed to write a Maven class?
<khem>
ehmm
<rburton>
well, yeah
<khem>
I remember a project where they were trying to replace bitbake
<khem>
with maven so it must be a good tool
flom84 has joined #yocto
Vonter has quit [Quit: WeeChat 4.1.1]
sakoman has quit [Ping timeout: 276 seconds]
<rburton>
its bazel for java!
<rburton>
make of that what you will
perceval has quit [Quit: Leaving]
Vonter has joined #yocto
<khem>
scala guys were using it when I had to deal with scala long time ago
andreascian has joined #yocto
tnovotny has quit [Remote host closed the connection]
<JPEW>
kanavin_: Isn't quilt-native ignored is signature calcualtions?
alperak54 has joined #yocto
mbulut_ has joined #yocto
alperak has quit [Ping timeout: 250 seconds]
alperak54 has quit [Client Quit]
alperak has joined #yocto
<kanavin_>
JPEW, it is yes. we really need to ask hashequiv about this.
<JPEW>
Ok, hang on
<mischief>
is there a way to skip the 'remote tracking branch' conversion in do_kernel_checkout?
sakoman has joined #yocto
<chep>
JaMa: my fetcher inherits from gitsm and do nothing more… I also tried to copy all gitsm code. I'll look for supports_srcrev, thx
flom84 has quit [Quit: Leaving]
JaMa has quit [Ping timeout: 264 seconds]
sakman_ has joined #yocto
sakman has quit [Ping timeout: 264 seconds]
andreascian has quit [Ping timeout: 250 seconds]
sakman_ is now known as sakman
florian has quit [Quit: Ex-Chat]
<JPEW>
kanavin_: Sent a patch to the ML and CC'd you
bhstalel has quit [Ping timeout: 250 seconds]
mckoan is now known as mckoan|away
JaMa has joined #yocto
lthadeus has quit [Remote host closed the connection]
GNUmoon has quit [Remote host closed the connection]
GNUmoon has joined #yocto
Tyaku_ has quit [Quit: Lost terminal]
zpfvo has quit [Ping timeout: 256 seconds]
zpfvo has joined #yocto
Guest67 has joined #yocto
<kanavin_>
JPEW, how does one use it? what is a 'method'?
<JPEW>
kanavin_: It's the mechansim used to calculate the hashes... basically a namespace. In your case, oe.sstatesig.OEOuthashBasic
<kanavin_>
JPEW, why basic, shouldnt it be OEEquivHash?
<kanavin_>
or I'm just confused
<JPEW>
*Outhash*Basic not OEBasicHash
<RP>
kanavin_: it needs to match the string used in the code in that function
<JPEW>
The output hash calculation method is not the same as the signature generation.... they both just happen to have a "Basic" option. It's a little confusing if you don't notice that
<RP>
kanavin_: basically the hashserver has namespaces
vladest has joined #yocto
Chaser has quit [Remote host closed the connection]
Chaser has joined #yocto
nerdboy has quit [Remote host closed the connection]
alessioigor has quit [Quit: alessioigor]
nerdboy has joined #yocto
nerdboy has joined #yocto
nerdboy has quit [Changing host]
alessioigor has joined #yocto
Vonter has quit [Ping timeout: 252 seconds]
Vonter has joined #yocto
andreascian has joined #yocto
<kanavin_>
RP JPEW I've written into the bug what the tool says, but not sure what's the next thing I could try :-/
<kanavin_>
I mean, I could just add hashequiv disabling to the test, but we should find out the scenario that played out here
mbulut_ has quit [Ping timeout: 256 seconds]
Saur_Home has quit [Quit: Client closed]
Saur_Home has joined #yocto
jmd has joined #yocto
leon-anavi has quit [Quit: Leaving]
Guest67 has quit [Quit: Client closed]
andreascian has quit [Quit: Client closed]
<alperak>
Is the autotools class run autoreconf after configure, right? Is there a way to run autoreconf before configure?
Chaser has quit [Remote host closed the connection]
Chaser has joined #yocto
AndyT has joined #yocto
mbulut_ has joined #yocto
Chaser has quit [Remote host closed the connection]
<alperak>
why is this error happening? we found the solution but I don't understand it.
<alperak>
thanks btw.
<rburton>
maybe we should add that to autotools.bbclass but there's only a few recipes which need it...
Chaser has quit [Quit: Chaser]
vladest1 has joined #yocto
vladest has quit [Ping timeout: 252 seconds]
vladest1 is now known as vladest
vladest has quit [Client Quit]
vladest1 has joined #yocto
vladest1 is now known as vladest
u4ia has quit [Quit: WeeChat 4.0.1]
prabhakarlad has quit [Quit: Client closed]
alessioigor has quit [Quit: alessioigor]
alessioigor has joined #yocto
xmn has quit [Quit: ZZZzzz…]
gsalazar has quit [Ping timeout: 268 seconds]
<rburton>
alperak: ok i've fixed it locally in autotools.bbclass so your workaround will be needed for a short amount of time hopefully
JaMa has joined #yocto
frieder has quit [Remote host closed the connection]
g0hl1n has quit [Quit: Client closed]
amitk_ has quit [Ping timeout: 268 seconds]
JerryM has quit [Quit: Konversation terminated!]
prabhakarlad has joined #yocto
florian__ has quit [Ping timeout: 252 seconds]
gsalazar has joined #yocto
gsalazar has quit [Ping timeout: 276 seconds]
JaMa has quit [Ping timeout: 256 seconds]
behanw has quit [Quit: Connection closed for inactivity]
gsalazar has joined #yocto
prabhakarlad has quit [Quit: Client closed]
prabhakar has quit [Quit: Connection closed]
prabhakar has joined #yocto
prabhakarlad has joined #yocto
Rich_1234_home has joined #yocto
<Rich_1234_home>
Anyone have an estimate on how long `bin/bitbake-selftest` takes? Running on a pretty slow laptop
florian__ has joined #yocto
JaMa has joined #yocto
<alperak>
good night everyone
alessioigor has quit [Quit: alessioigor]
alperak has quit [Quit: Client closed]
g0hl1n has joined #yocto
michaelgalassi has quit [Quit: Client closed]
<vmeson>
Rich_1234_home: did it finish yet? ;-)
<Rich_1234_home>
vmeson yep :) just disabled net tests and took 273 seconds for 637 tests
<vmeson>
Rich_1234_home: k and good. Let me test my shiny new Dell i7 laptop: BB_SKIP_NETTESTS="yes" time ../../poky.git/bitbake/bin/bitbake-selftest
<Rich_1234_home>
This is on a tiny Starlabs starlite (not the new tablet loking one)
<vmeson>
Ran 637 tests in 236.399s -- this system maxes out one core at ~5GHz! and the test did end up turning on the fan briefly.
<Rich_1234_home>
not bad not bad, running an Intel pentium silver @1.1ghz here 8)
<Rich_1234_home>
i'm trying to send in my first patch (maybe soon)
<Rich_1234_home>
maybe a dumb question but if I want to patch a bitbake test, do I branch off of poky and then use the oe patch tester or just branch off bitbake
joekale has joined #yocto
<zeddii>
still no timeouts
Rich_1234_home has quit [Quit: Client closed]
Vonter has quit [Ping timeout: 276 seconds]
Rich_1234_home has joined #yocto
Vonter has joined #yocto
Kubu_work has joined #yocto
gsalazar has quit [Ping timeout: 268 seconds]
Kubu_work has quit [Quit: Leaving.]
joekale has quit [Ping timeout: 245 seconds]
Xagen has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
vladest has quit [Quit: vladest]
vladest1 has joined #yocto
vladest1 is now known as vladest
<RP>
Rich_1234: with the nettests disabled, bitbake-selftest should be fastish, particularly compared to many other tests
<RP>
kanavin_: I strongly believe we need to get to the bottom of this as real world users are going to hit this
<Rich_1234_home>
I have a patch made, I sent myself the test email. It didn't fail the patch tester however the commit message is all in the subject header instead of in the line below. i.e. subject is "[PATCH] bitbake: tests: Fix duplicate test_underscore_override test --- Found a duplicate test, added _2 suffix to one, 74 tests now pass up from 73." How do i move
<Rich_1234_home>
the --- part into the body?
<RP>
Rich_1234: "git commit --amend" and split the lines?
<Rich_1234_home>
RP: yeah you were right, had a --- instead of blank line. Looks like it is ready to go now 8)
<Rich_1234_home>
Ok that is now submitted (y)
<RP>
Rich_1234: looks good, thanks!
<Rich_1234_home>
RP: no worries :)
<Rich_1234_home>
but have a good evening all, I am off
<RP>
Rich_1234_home: I will put it into testing overnight :)
<Rich_1234_home>
nice :)
Rich_1234_home has quit [Ping timeout: 250 seconds]