<kanavin>
rburton, any idea why GNU had to reinvent test harnesses with autotest (obviously not conforming to any standard for test results, and not supporting installable tests)
<kanavin>
other than GNU being GNU
<rburton>
that reason
<LetoThe2nd>
NIAG
<kanavin>
rburton, I just advised the poor chap that wants to add tar ptest on the subject (on oe-core list)
rfuentess has quit [Read error: Connection reset by peer]
rfuentess has joined #yocto
varjag has joined #yocto
tgamblin has quit [Ping timeout: 240 seconds]
starblue has quit [Ping timeout: 244 seconds]
starblue has joined #yocto
rfuentess has quit [Read error: Connection reset by peer]
tgamblin has joined #yocto
rfuentess has joined #yocto
prabhakarlad has quit [Quit: Client closed]
l3s8g has quit [Ping timeout: 240 seconds]
MrFrank has quit [Server closed connection]
MrFrank has joined #yocto
u1106 has quit [Server closed connection]
u1106 has joined #yocto
<qschulz>
rburton: back again with GitLab Runner questions :)
<qschulz>
kas requires USER_ID and GROUP_ID environment variables to be passed right?
<rburton>
qschulz: if you're not using k8s then you don't need the FF_ lines and the others are all just niceness. you don't _need_ anything to use a kas container beyond the image: line.
Danct12 has joined #yocto
<qschulz>
rburton: well... you gotta mount SSTATE_DIR and DL_DIR within the container ?
l3s8g has joined #yocto
<rburton>
sure, we do that with CACHE_DIR
<qschulz>
but for it to persist you have to have it in the toml file of the gitlab-runner no?
rfuentess has quit [Read error: Connection reset by peer]
rfuentess has joined #yocto
<rburton>
yeah that's setup outside of the gitlab-ci.yml
amsobr has joined #yocto
<qschulz>
that's what I'm after :)
<rburton>
i can't share that bit as i can't even see it, devops manage it. we've an EFS mount in the containers running on k8s
<qschulz>
rburton: just send them a carefully crafted webp image to see it ;)
<qschulz>
rburton: but thanks, just trying to figure out what I need in that gitlab-runner toml file for it to work. I have permissions issues right now :)
<rburton>
maybe ask jonmason, he runs gitlab-runner directly on his machines iirc
<rburton>
argh cve report of doom
<qschulz>
tlwoerner: do you have a rock5b? working on adding rk3588 support to meta-rockchip
<qschulz>
tlwoerner: unfortunately our products only support very downstream kernel and u-boot but I could check rock5b boots mainline with meta-rockchip
rfuentess has quit [Read error: Connection reset by peer]
rfuentess has joined #yocto
marex has quit [Server closed connection]
marex has joined #yocto
<jonmason>
qschulz: I have a quick how-to I can send you for setting up gitlsb ci
varjag has quit [Quit: ERC (IRC client for Emacs 27.1)]
Vonter has quit [Ping timeout: 240 seconds]
Vonter has joined #yocto
alessioigor has quit [Quit: alessioigor]
alessioigor has joined #yocto
varjag has joined #yocto
Danct12 has quit [Quit: WeeChat 4.0.4]
<amsobr>
i have an http sstate mirror and want to set up a hash equivalence server. Since sstate is served over http, should I have any special care to keep the remote sstate up to date?
<amsobr>
I thought I might drop the sstate from the local build trees (shared between themselves with SSTATE_DIR) into the http root from time to time but have no idea if that'd make sense
Schlumpf has quit [Quit: Client closed]
rfuentess has quit [Remote host closed the connection]
lexano has joined #yocto
beneth has joined #yocto
Kubu_work has joined #yocto
<qschulz>
jonmason: would be much appreciated :)
<ptsneves>
amsobr: you should sync your sstate cache from time to time, depending on the activity your project sees. If you want to do the sync manually you can keep an eye out for the sstate cache match rate. Also keep in mind that there are no integrity checks on sstate cache artifacts so you should make sure your mirror is https and not tamperable
Guest98 has quit [Ping timeout: 245 seconds]
alessioigor has quit [Quit: alessioigor]
alessioigor has joined #yocto
rfuentess has joined #yocto
brazuca has joined #yocto
Guest98 has joined #yocto
alessioigor has quit [Quit: alessioigor]
alessioigor has joined #yocto
Danct12 has joined #yocto
prabhakarlad has joined #yocto
Minvera has joined #yocto
<amsobr>
ptsneves: OK, thanks. And to sync should I just copy the local sstate artefacts from the local end from time to time (or when match rate drops) or is there anything else that should be looked for?
<amsobr>
And is there a way of purging the sstate from old/obsolete entries?
<amsobr>
I'm sure this has been asked before but I haven't been able to find definitive answers to this...
<RP>
amsobr: check the docs, it is documented
<ptsneves>
amsobr: when you find please paste the doc link
<RP>
michaelo: I think "duplicate" is the wrong phrase here, maybe obsolete ?
xmn has joined #yocto
<RP>
ptsneves, amsobr: ^^^
<amsobr>
RP: thanks. I had actually read that one before but ruled it out because it mentions "duplicate" files
<RP>
amsobr: we need to fix the wording, a patch would be very welcome
Guest29 has joined #yocto
<RP>
michaelo: another vote to fix the wording! :)
<ptsneves>
RP: I liked the find easy trick and i think never used. Thanks
<amsobr>
ah, my bad, I had actually seen the kirkstone (should really read master more often!) and the `find` note isn't there. Got it!
<RP>
amsobr: I knew this was in the docs as we did recently improve it!
Guest98 has quit [Ping timeout: 245 seconds]
rob_w has quit [Remote host closed the connection]
GillesM has joined #yocto
rfuentess has quit [Read error: Connection reset by peer]
rfuentess has joined #yocto
<rburton>
kanavin: fyi i have a cairo upgrade to 1.17.8 as a step towards 1.18.0
varjag has quit [Quit: ERC (IRC client for Emacs 27.1)]
rfuentess has quit [Read error: Connection reset by peer]
Guest29 has quit [Ping timeout: 245 seconds]
rfuentess has joined #yocto
rfuentess has quit [Read error: Connection reset by peer]
rfuentess has joined #yocto
<ptsneves>
RP: would a patch that includes the sha256 sum in the state cache file name be acceptable? The reason I would like it is so that failures fetching do not lead to broken sstate files, which is the case with wget running on bad networks. It also would harden against sstate cache tampering
c_sutton has quit [Read error: Connection reset by peer]
<ptsneves>
"It also would harden against sstate cache tampering" scratch this part
c_sutton has joined #yocto
astlep55040 has quit [Server closed connection]
rfuentess has quit [Read error: Connection reset by peer]
rfuentess has joined #yocto
astlep55040 has joined #yocto
rfuentess has quit [Read error: Connection reset by peer]
rfuentess has joined #yocto
otavio has joined #yocto
rfuentess has quit [Remote host closed the connection]
rfuentess has joined #yocto
frieder has quit [Ping timeout: 255 seconds]
rsalveti has joined #yocto
l3s8g has quit [Remote host closed the connection]
l3s8g has joined #yocto
<RP>
ptsneves: that can't work. How would you know the name of the file to fetch?
<michaelo>
RP, amsobr: happy to fix the wording here. I will backport to kirkstone too.
<michaelo>
thanks!
<RP>
ptsneves: note that we do support signing of the files which would solve your problem I suspect
<ptsneves>
RP: we support signing the sstate cache files? Wow did not know
<ptsneves>
RP: Ach you are right regarding the files to fetch. But signing would indeed help.
<ptsneves>
Where can i find info about it?
frieder has joined #yocto
rfuentess has quit [Read error: Connection reset by peer]
<kanavin>
they promise to do 1.18.0 today, so let's see
rfuentess has joined #yocto
Schlumpf has joined #yocto
<RP>
kanavin: your "discussion" was highlighted to a few of us. I think your point was reasonable if a little direct :)
<kanavin>
RP: I managed hands off keyboard this time!
<RP>
kanavin: I think rburton is trying to get any issues we might encounter raised so 1.18.0 might be better
<kanavin>
RP: there's a vintage alex patchbomb coming :)
<RP>
kanavin: 32 bit issues? or something else?
<kanavin>
no, just a gigantic rollup of all the easy version updates
<kanavin>
the harder ones already went in
vladest has quit [Ping timeout: 264 seconds]
<RP>
kanavin: ok. I'm trying my best to keep up with the patches and get some issues fixed/debugged
alessioigor has quit [Quit: alessioigor]
alessioigor has joined #yocto
<RP>
In looking at the details of a bug, I think I've found there is a significant file overlap bug in sstate where errors haven't been reported and we have significant bitrot :(
tgamblin has quit [Ping timeout: 246 seconds]
goliath has quit [Quit: SIGSEGV]
<RP>
bitrot isn't so bad, it is mainly SPDX :/
<reatmon>
We are trying to debug some pseudo errors that have popped up, but are having some issues related to the interaction with child processes. In the file: meta/lib/oe/package_manager/ipk/__init__.py under list_pkgs() it calls opkg to get the package list. But the error check is on both the return code AND if there is anything on cmd_stderr. But if you turn on pseudo debugging then there is always output on stderr. Is there an easy
<reatmon>
way to disable debugging in this one call?
<rburton>
sounds like the code should just look at the return code anyway
<khem>
RP: per-recipe image testing, do we have some docs on it ?
Kubu_work has quit [Quit: Leaving.]
kscherer has joined #yocto
brazuca has quit [Quit: Client closed]
florian_kc has joined #yocto
nerdboy has quit [Ping timeout: 255 seconds]
slimak has joined #yocto
nerdboy has joined #yocto
nerdboy has joined #yocto
nerdboy has quit [Changing host]
vladest has quit [Remote host closed the connection]
vladest has joined #yocto
florian_kc has quit [Ping timeout: 255 seconds]
brazuca has joined #yocto
slimak has quit [Ping timeout: 244 seconds]
dvergatal has joined #yocto
ptsneves has joined #yocto
<halstead>
khem: I interrupted your meta-oe during maintenance and caused it to error out. Please queue up a new one as needed.
ptsneves has quit [Ping timeout: 240 seconds]
l3s8g has quit [Remote host closed the connection]
alessioigor has quit [Quit: alessioigor]
alessioigor has joined #yocto
l3s8g has joined #yocto
florian_kc has joined #yocto
lighteagle has joined #yocto
florian_kc is now known as florian
<khem>
halstead: np thanks for letting me know.
Haxxa has quit [Quit: Haxxa flies away.]
brazuca has quit [Quit: Client closed]
Haxxa has joined #yocto
Guest10 has joined #yocto
<Guest10>
I have a system running from the kirkstone release, kernel 5.15 - I am seeing shown here: https://pastebin.com/17QqaxDM , both regulatory.db and regulatory.db.p7s exist in /lib/firmware/ - has anyone come across a solution to this message?
<Guest10>
My understanding is that since kernel 4.15, the crda user space tool is no longer need and the direct firmware loading _should_ be the accepted path - I'm noticing I cannot update the wireless regdom using `iw reg set XX` , trying to figure out why this is the case.
<ad__>
hi, have a recipe in meta-noto that can be built by bitbake, but not found if added to image
<Guest10>
yeah, I'm trying to figure out how to manually verify those signatures :)
<khem>
its perhaps being regenerated without signing
<khem>
look at build
<Guest10>
I suppose I could disable the signing (in the kernel) and see if that gets rid of the error - that could point to the signature being incorrect. I'm not sure what to look for in the build other than what is in that recipe I linked. That doesn't appear to do any explicit signing
l3s8g has quit [Ping timeout: 260 seconds]
brazuca has joined #yocto
<RP>
khem: I don't know what you mean?
<khem>
I mean core-image-ptest-<recipe>
<khem>
which is spun from core-image-ptest-all
<RP>
khem: ah, no real docs :(
<RP>
khem: it is basically just using BBCLASSEXTEND
<khem>
No worries, I think its a neat stuff, I will do some writeup
<RP>
khem: can you tell if /mnt/b/yoe/master/build/tmp/deploy/spdx/by-hash/*/sstate:os-release:all-yoe-linux:1.0:r0:allarch:11:/os-release.spdx.json is present?
<khem>
yes its a symlink - /mnt/b/yoe/master/build/tmp/deploy/spdx/by-hash/allarch/sstate:os-release:all-yoe-linux:1.0:r0:allarch:11:/os-release.spdx.json -> ../../../allarch/packages/os-release.spdx.json
<khem>
and symlink does not exist :(
<RP>
khem: missing a level of ../ ?
<khem>
its because I cleaned tmpdir
Vonter has quit [Ping timeout: 240 seconds]
<khem>
hmm level seems ok
<RP>
khem: I think it may be "allarch" vs "all"
Vonter has joined #yocto
<khem>
all just has gcc-source and llvm-source
<khem>
yeah so whats difference between all and allarch.
<khem>
RP: I am consistently seeing one failure in mdadm on AB I see ptests all pass for it, so it might be clang or something but do you remember anything related to ptests and mdadm