kanavin has quit [Remote host closed the connection]
kanavin has joined #yocto
kroon has joined #yocto
<kroon>
RP, how can we be so sure that native sstate cache built in one build directory can be reused in another build directory, even when the binaries contain references to build-specific paths ? Is it because of testing, or the fact that the embedded build paths only exist during the recipe build itself anyway ?
adrian__ has joined #yocto
<RP>
kroon: the binaries are always moved to new sysroots and things like rm_work remove the originals so whilst we can't be 100% sure and we do find issues now and again, in general it would effectively be tested
<kroon>
RP, ok, I see
dkl has quit [Quit: %quit%]
dkl has joined #yocto
goliath has joined #yocto
dkl has quit [Quit: %quit%]
dkl has joined #yocto
dkl has quit [Client Quit]
dkl has joined #yocto
dkl has quit [Client Quit]
dkl has joined #yocto
pgowda_ has quit [Quit: Connection closed for inactivity]
pbergin__ has joined #yocto
pbergin__ has quit [Ping timeout: 256 seconds]
florian has joined #yocto
dtometzki has quit [Read error: Connection reset by peer]
WadeBerrier[m] has quit [Quit: You have been kicked for being idle]
Pan5ky has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
dvorkindmitry has joined #yocto
<dvorkindmitry>
where can I find the SRC_URI local files search path logic?
pbergin__ has quit [Ping timeout: 252 seconds]
vd has joined #yocto
tlwoerner has quit [Ping timeout: 252 seconds]
tlwoerner has joined #yocto
neothejoker[m] has joined #yocto
neothejoker[m] has left #yocto [#yocto]
sakoman has quit [Quit: Leaving.]
florian has joined #yocto
Pan5ky has joined #yocto
florian has quit [Ping timeout: 252 seconds]
florian has joined #yocto
florian has quit [Ping timeout: 252 seconds]
sakoman has joined #yocto
BCMM has quit [Quit: Konversation terminated!]
Pan5ky has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
florian has joined #yocto
adrian__ has quit [Ping timeout: 252 seconds]
florian has quit [Ping timeout: 268 seconds]
tlwoerner has quit [Ping timeout: 268 seconds]
<zeddii>
can anyone think of a reason why when I manually extract a rpm that is present in my rootfs assembly, I see the files I expect .. but yet they aren't in the rootfs/ ?
<zeddii>
some sort of sstate reuse that is preventing assembly from using it ?
<fray>
in the past I saw an issue where the rootfs didn't regenerate for some reason.. a simple -c cleanall on the rootfs fixed it
<fray>
I'd suggest you try that first
<fray>
Also if you enable package managemetn on the rootfs.. login, you can check if it should have been installed..
<zeddii>
it looks like the rootfs is being yanked out of state, yah. but I tried that. same thing ahppend second time.
<zeddii>
i.e. I *see* the rpm staged: ./oe-rootfs-repo/rpm/core2_64/kubernetes-cni-v1.23.1+gitdd1b0a124713105e94ed58ffc2115ee8c1fd9f72-r0.0.core2_64.rpm
<fray>
Ya, if you add package-management.. then you can get rpm to verify what is installed is actually on the disk
<fray>
you can easily query the DB as well...
<zeddii>
I always have packagemanagent on, and yah, the #@$@ thing is NOT on the rootfs, even though I have it in the rdpends of another rpm now.
<zeddii>
what's the quick way to get RPM to show me the depends again ?
<fray>
rpm -q <package> --requires
<zeddii>
I can run it on target, so that's easy enough
<fray>
note, package is just the name, not the filename
<fray>
to get a list of what is installed.. rpm -qa
amitk has joined #yocto
<fray>
you can also use rpm -qp <pacakge _FILE_> --requires to query what is on your build system
<zeddii>
so that's weird. I SEE my package in the requires
<zeddii>
I just created the package, so that is unlikely that it is provided by anything else.
<zeddii>
the packages-split in the build have what I'd expect it to install.
<zeddii>
never say never, let me try that query
<zeddii>
hmm. no what-provides ..
<fray>
command worked, but reported back nothing?
<zeddii>
naw, it said "rpm: --what-provides: unknown option"
<fray>
sec
amitk_ has quit [Ping timeout: 252 seconds]
<fray>
rpm -q --whatprovides (no middle -)
<zeddii>
aha. i was googling, should have guessed
<zeddii>
ahah. oddly, something else claims to provide it .. also a package I created, I'll have to check, it isn't breaking k3s, but only k8s.
<zeddii>
that's just what I needed. I'll keep a log of the tricks
<fray>
ya, that is usually hwat the trick is.. something else provides the thing and it breaks you
<zeddii>
I guess I could go with a preferred provider for the conlict, but I think I can just delete the generic one.
<fray>
preferred provider only affects build-time, not runtime.. So this could still happen.. definitely need to resolve it in the packaging side of things
<zeddii>
I nuked the kubernetes-cni rprovides from the cni package, and now my kubernetes build should provide it itself, that should fix it up.
<zeddii>
I see it rebuilding, so that's a positive sign
kroon has quit [Quit: Leaving]
<JosefHolzmayrThe>
NUKE IT FROM ORBIT!
<JosefHolzmayrThe>
...
<JosefHolzmayrThe>
what are we talking about anyway?
dtometzki has quit [Ping timeout: 252 seconds]
GNUmoon has quit [Ping timeout: 276 seconds]
Pan5ky has joined #yocto
Pan5ky has quit [Client Quit]
Pan5ky has joined #yocto
GNUmoon has joined #yocto
bq has quit [Quit: Floating point exception (core dumped)]
florian has joined #yocto
Pan5ky has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
tlwoerner has joined #yocto
tangofoxtrot has quit [Remote host closed the connection]