alessioigor has quit [Remote host closed the connection]
alessioigor has joined #yocto
alessioigor has quit [Client Quit]
schtobia has quit [Quit: Bye!]
alessioigor has joined #yocto
schtobia has joined #yocto
florian__ has joined #yocto
mbulut has joined #yocto
alessioigor has quit [Remote host closed the connection]
Net147 has joined #yocto
Net147 has quit [Changing host]
Net147 has joined #yocto
alessioigor has joined #yocto
lthadeus has quit [Ping timeout: 268 seconds]
alessioigor has quit [Client Quit]
Guest74 has joined #yocto
lthadeus has joined #yocto
alessioigor has joined #yocto
Net147 has quit [Read error: Connection reset by peer]
alessioigor has quit [Remote host closed the connection]
alessioigor has joined #yocto
Net147 has joined #yocto
Net147 has quit [Changing host]
Net147 has joined #yocto
Guest74 has quit [Quit: Client closed]
alessioigor has quit [Remote host closed the connection]
alessioigor has joined #yocto
olani has joined #yocto
alessioigor has quit [Remote host closed the connection]
Net147 has quit [Quit: Quit]
Net147 has joined #yocto
Net147 has quit [Changing host]
Net147 has joined #yocto
gsalazar has joined #yocto
alessioigor has joined #yocto
alessioigor has quit [Remote host closed the connection]
alessioigor has joined #yocto
gsalazar has quit [Ping timeout: 245 seconds]
alessioigor has quit [Remote host closed the connection]
alessioigor has joined #yocto
prabhakar has joined #yocto
prabhakarlad has joined #yocto
Net147 has quit [Read error: Connection reset by peer]
Net147 has joined #yocto
Net147 has quit [Changing host]
Net147 has joined #yocto
alessioigor has quit [Remote host closed the connection]
Net147 has quit [Ping timeout: 252 seconds]
camus has quit [Quit: camus]
alessioigor has joined #yocto
camus has joined #yocto
camus has quit [Ping timeout: 252 seconds]
alessioigor has quit [Remote host closed the connection]
alessioigor has joined #yocto
starblue has quit [Ping timeout: 255 seconds]
Net147 has joined #yocto
Net147 has quit [Changing host]
Net147 has joined #yocto
alessioigor has quit [Remote host closed the connection]
starblue has joined #yocto
alessioigor has joined #yocto
alessioigor has quit [Remote host closed the connection]
alessioigor has joined #yocto
tnovotny_ has joined #yocto
tnovotny has quit [Read error: Connection reset by peer]
tnovotny has joined #yocto
tnovotny_ has quit [Ping timeout: 245 seconds]
alessioigor has quit [Quit: alessioigor]
alessioigor has joined #yocto
lthadeus has quit [Remote host closed the connection]
alessioigor has quit [Quit: alessioigor]
dmoseley has joined #yocto
alessioigor has joined #yocto
camus has joined #yocto
tnovotny has quit [Ping timeout: 260 seconds]
zpfvo has quit [Ping timeout: 276 seconds]
zpfvo has joined #yocto
me1 has joined #yocto
zpfvo has quit [Ping timeout: 252 seconds]
me1 has quit [Client Quit]
zpfvo has joined #yocto
Tyaku has quit [Quit: Lost terminal]
neverpanic has joined #yocto
xmn has joined #yocto
sgw has joined #yocto
alperak has joined #yocto
Daanct12 has quit [Quit: WeeChat 4.1.1]
<alperak>
for a recipe and I need to give the value __STDC_VERSION__ to a compile option in cmake. can I get this from somewhere or have to give it hardcoded?
rob_w has quit [Remote host closed the connection]
tnovotny has joined #yocto
alessioigor has quit [Quit: alessioigor]
alessioigor has joined #yocto
<RP>
kanavin_: I did think about a simple experiment for this sigs issue. Run a build with a clean sstate and run the test. Then backdate the timestamps on all the sstate files. Then rerun the tests and check it updates all the stamps
prabhakarlad has quit [Ping timeout: 250 seconds]
alperak has quit [Quit: Client closed]
alperak has joined #yocto
<kanavin_>
RP: I can try that
gsalazar has joined #yocto
prabhakar has quit [Ping timeout: 268 seconds]
joekale has joined #yocto
gsalazar has quit [Ping timeout: 260 seconds]
<kanavin_>
RP: but why backdate the timestamps separately? if I note the clock before running the test for the second time, I'd be able to tell if the timestamps are updated?
alessioigor has quit [Quit: alessioigor]
alessioigor has joined #yocto
<RP>
kanavin_: fair enough. I guess I was just thinking it would make it more obvious
<wmills_>
rburton: Last week you said you thought you had seen the hashserve timeout warnings. If that happens again please record the conditions in this bug report: https://bugzilla.yoctoproject.org/show_bug.cgi?id=15306
<rburton>
yeah will do, i saw that
vladest has quit [Ping timeout: 276 seconds]
<kanavin_>
RP: collect_spdx_deps is not updated, but everything else is
<RP>
wmills_: I know zeddii was running into this. If you're able I've put a potential problematic set of commits into the bug which it would be good to narrow things down to if possible
<RP>
kanavin_: I wonder why that one was somehow special
<zeddii>
yah. I was seeing the exact same thing. I'll update with what I tried for debug.
gsalazar has joined #yocto
tnovotny has quit [Quit: Leaving]
Vonter has joined #yocto
goliath has quit [Quit: SIGSEGV]
<wmills_>
RP: I did put the good commit and bad commit in the bug report but : 1) they are 1 month apart and 2) are from poky.
<wmills_>
I guess I could do the branch point for nanbield. I assume the issue is after that
vladest has joined #yocto
<JPEW>
wmills_: What's your build host?
speeder has joined #yocto
<wmills_>
see the bug report it is all recorded in more detail than I can repeat bit by bit here
<speeder>
hello, I am newbie. I need to use "perf", seemly it is part of the SDK, but the SDK seemly is a tool to compile yocto... or I am wrong and the sdk is also part of poky itself? (as in... install things inside poky, not just on my dev machine)
<JPEW>
I wonder if the code restructuing has resulting in some sneaky extra sync() calls with sqlite
vladest1 has joined #yocto
<wmills_>
speeder: if you want the full sdk look for the core-images that end in _sdk. You can also just add the tools you need to your image of choice. Look at the image defs for the _sdk images for cluses how to do that
vladest has quit [Ping timeout: 245 seconds]
vladest1 is now known as vladest
<wmills_>
JPEW or anyone, how bug is the timeout I have not looked. One question I am wondering is this: Is the longer build times with the bug a cause or an effect of the timeout
<JPEW>
The longer build time is an effect of the timeout
<JPEW>
So we need to fix the timeout and that will fix the build times
<wmills_>
Can I set hashserve to none or something to test that?
grsandeep85 has joined #yocto
<JPEW>
wmills_: Ya, but you'll get pretty poor sstate usage without it
<JPEW>
wmills_: Also, I'm pretty sure you won't get timeouts if there is no contact with the server :0
<JPEW>
:)
<wmills_>
JPEW: These builds are starting with zero sstate so I don't think I would lose much. Yes the idea would be to compare the build times between the "good" and "bad" commit w/o the effect of the timeouts
ray-san has quit [Ping timeout: 264 seconds]
<JPEW>
wmills_: Ya, in that case if you set `BB_SIGNATURE_HANDLER = "OEBasicHash"` in local.conf it will disable hash equivalence
<rburton>
speeder: add dbg-pkgs to IMAGE_FEATURES to get the symbols, and perf to IMAGE_INSTALL to install perf. you don't need special images.
<rburton>
and yes there is overlap between the usage of 'sdk'
Vonter has quit [Quit: WeeChat 4.1.1]
tgamblin_ has joined #yocto
<JPEW>
Ah, I think I know the problem
tgamblin has quit [Read error: Connection reset by peer]
tgamblin_ is now known as tgamblin
<wmills_>
JPEW: if you push a test patch somewhere I can test it. I have not replaced that HD yet just for this purpsoe.
<JPEW>
Yep
<JPEW>
wmills_: poky ok or do you want bitbake?
<wmills_>
poky is actually prefered
<JPEW>
wmills_: K
<wmills_>
fewer changes to what I am doing already
<wmills_>
JPEW: Im off for a few hours, I will check back later
Saur_Home has quit [Quit: Client closed]
Saur_Home has joined #yocto
<JPEW>
wmills_, rburton, zeddii: 264b8d564848590c1a4670b7468e0ca5b5092387 in poky-contrib should be the fix. Patch is also on the ML
kpo has joined #yocto
* zeddii
tests
<joekale>
Hello! I was reading this blog post from lucaceresoli https://bootlin.com/blog/simplest-yocto-layer-meta-kiss/ and was curious if there was a good reference for the "bad" Yocto/OE coding and the better way of doing some of the things he mentions and more.
<rburton>
JPEW: awesome
ecdhe_ is now known as ecdhe
mbulut has quit [Ping timeout: 240 seconds]
jsbronder has quit [Quit: WeeChat 3.8]
jsbronder has joined #yocto
<rburton>
vmeson: thanks for updating the wiki to mark mickledore as EOL. I just got to disable our scheduled mickledore builds :)
rfuentess has quit [Remote host closed the connection]
alessioigor has quit [Quit: alessioigor]
alessioigor has joined #yocto
mckoan is now known as mckoan|away
bhstalel has quit [Ping timeout: 250 seconds]
amitk_ has joined #yocto
Chocky has quit [Quit: Leaving.]
Chocky has joined #yocto
Chocky1 has joined #yocto
Chocky has quit [Client Quit]
Chocky1 has quit [Client Quit]
Chocky has joined #yocto
flom84 has joined #yocto
Vonter has joined #yocto
Chocky has quit [Client Quit]
Chocky has joined #yocto
florian__ has quit [Ping timeout: 245 seconds]
khem has joined #yocto
zpfvo has quit [Quit: Leaving.]
florian has quit [Quit: Ex-Chat]
sakman has quit [Ping timeout: 264 seconds]
alessioigor has quit [Quit: alessioigor]
alessioigor has joined #yocto
jmd has joined #yocto
<RP>
JPEW: that looks like something we should fast track!
<JPEW>
RP: Ya. It makes perfect sense to explain the problem, so I'm pretty confident that is the correct fix
alessioigor has quit [Quit: alessioigor]
alessioigor has joined #yocto
flom84 has quit [Remote host closed the connection]
sakman has joined #yocto
<RP>
JPEW, kanavin_: I've been looking at this hash issue a little bit more. I'm starting to think there is a timing race where a hash equivalence is added part way through a build. The "trail" of siginfo files left if an x86_64 worker changes the hashes in an aarch64 build, mid build may get 'intetesting' ?
* RP
can kind of see how it is happening but hard to articulate
amitk_ has quit [Quit: leaving]
<kanavin_>
RP: I was planning to see how x86_64 worker can influence native builds on aarch64 workers, but I'm for now rewriting the logic in printdiff as written in the last comment in the bug. It's going well, just needs more testing and fixing up the tests, and confirming that it does indeed resolve the issue where needed siginfo files never even appear.
<RP>
kanavin_: I hadn't noticed tht comment, it is an interesting idea. I need to think about that! :)
<RP>
I would like to better understand the case this breaks down, even if just to work out a test case
mbulut has joined #yocto
gsalazar_ has joined #yocto
gsalazar_ has quit [Ping timeout: 252 seconds]
<vmeson>
rburton: np. it still feels odd to have most of the releases only supported for 6-7 months and have the LTS releaes supported for 4 years...
<khem>
vmeson: I see three types of users one who are rolling with main dev tree, second who are rolling at a slower pace and need a stable base, third who are even slower or perhaps do not want to upgrade and want a stable release until their products EOL and basically thats what you see. 4 years is what seemed to fit the Yocto project members usecase most
alperak has quit [Quit: Client closed]
alperak has joined #yocto
alessioigor has quit [Quit: alessioigor]
alessioigor has joined #yocto
Chaser has quit [Quit: Chaser]
frieder has quit [Remote host closed the connection]
mbulut has quit [Ping timeout: 240 seconds]
ederibaucourt has quit [Ping timeout: 255 seconds]
florian__ has joined #yocto
Marmottus11 has joined #yocto
jsandman1 has joined #yocto
bluelightning_ has joined #yocto
shivamurthy_ has joined #yocto
wmills__ has joined #yocto
awafaa_ has joined #yocto
smooge- has joined #yocto
kergoth__ has joined #yocto
sudip_ has joined #yocto
rhadye_ has joined #yocto
landgraf1 has joined #yocto
JaMa has quit [Killed (silver.libera.chat (Nickname regained by services))]
wmills has quit [Ping timeout: 252 seconds]
rhadye has quit [Ping timeout: 252 seconds]
awafaa has quit [Ping timeout: 252 seconds]
bluelightning has quit [Ping timeout: 252 seconds]
sudip has quit [Ping timeout: 252 seconds]
shivamurthy has quit [Ping timeout: 252 seconds]
landgraf has quit [Ping timeout: 252 seconds]
kergoth_ has quit [Ping timeout: 252 seconds]
neverpanic has quit [Ping timeout: 252 seconds]
Marmottus1 has quit [Ping timeout: 252 seconds]
smooge has quit [Ping timeout: 252 seconds]
jsandman has quit [Ping timeout: 252 seconds]
JaMa has joined #yocto
Marmottus11 is now known as Marmottus1
rhadye_ is now known as rhadye
bluelightning_ is now known as bluelightning
jsandman1 is now known as jsandman
smooge- is now known as smooge
shivamurthy_ is now known as shivamurthy
wmills__ is now known as wmills
awafaa_ is now known as awafaa
kergoth__ is now known as kergoth_
neverpanic has joined #yocto
<wmills_>
JPEW: I clean build from your commit gets 0 warnings. The build time is 8869 Sec which is a lot better but not great compared to the old commit. That is probably a different issue.
Chocky has quit [Read error: Connection reset by peer]
Chocky1 has joined #yocto
Chocky1 has quit [Client Quit]
Chocky has joined #yocto
Chocky has quit [Client Quit]
Chocky has joined #yocto
Chocky1 has joined #yocto
Chocky has quit [Client Quit]
Chocky1 has quit [Client Quit]
Chocky has joined #yocto
Chocky has quit [Client Quit]
Chocky has joined #yocto
Chocky1 has joined #yocto
Chocky has quit [Client Quit]
Chocky1 has quit [Client Quit]
Chocky has joined #yocto
Vonter has quit [Ping timeout: 256 seconds]
Vonter has joined #yocto
Chocky1 has joined #yocto
Chocky has quit [Client Quit]
<halstead>
khem: All the fedora workers are running kernel 6.3.3 now.
<zeddii>
wmills: I'm also seeing a slower build, but no warnings.
gsalazar has quit [Ping timeout: 245 seconds]
<gmorell>
hmm, I'm attempting to get this perl package (munin) built for a cross compile and I'm having issues where the architectures of the builder and target are stepping on each other with libmodule-build-perl-native
<gmorell>
all the resource I've found in regards to building perl are out of date
<gmorell>
and I was wondering if there was something obvious I was missing
<rburton>
perl and cross is "fun" and its entirely possible that munim just ignores the entire problem and gets it wrong
<rburton>
paging moto-timo who may have some insights as he can read perl
<moto-timo>
gmorell: I don't see "munin" on metacpan.org... i see a couple related packages... but anyway, look at other recipes in oe-core and meta-perl for some of the additional variables that may need to be set like EXTRA_CPANFLAGS or EXTRA_PERLFLAGS
<moto-timo>
gmorell: we'll need more info to help further (pastebin the recipe, errors/log)... but no promises
<gmorell>
so is the move here that I add recipes for those that don't exist w/in m-oe/meta-perl
<moto-timo>
gmorell: that errror is because libmodule-build-perl is not in Perl 5.x itself any longer, so you need DEPENDS libmodule-build-perl-native
<moto-timo>
gmorell: not RDEPENDS (that's _runtime_ not _build time_)
<moto-timo>
oh you have that
<moto-timo>
I'd first try moving the inherit line earlier (before DEPENDS and RDEPENDS) and I don't think you need anything except libmodule-build-perl-native in DEPENDS (the rest is in the classes already). Look at other recipes in meta-perl for the order of the variables... you've got things like SRC_URI etc a bit later than is usually done.
<alperak>
clear
<moto-timo>
gmorell: also, which release of Yocto Project are you working on? dunfell? kirkstone? master?
<gmorell>
kirkstone
jmd has left #yocto [ERC 5.4 (IRC client for GNU Emacs 28.2)]
<gmorell>
no change post-shuffle, updated paste, still complaining about: "loadable library and perl binaries are mismatched (got handshake key 0xeb00080, needed 0xcd00080) "
<moto-timo>
'bitbake -c devshell munin' and output of 'which perl'?
alessioigor has quit [Quit: alessioigor]
<moto-timo>
it sounds like you are getting host contamination... which is probably something bad in their Build.PL script maybe.
<moto-timo>
gmorell: so you can see what is available in meta-perl and the rest you will need to create recipes for
<moto-timo>
gmorell: I know a bunch of those look familiar, but not all.
<gmorell>
yeah that's what I was expecting, but looking at the existing recipes they don't seem super complex, just a bit of work
<moto-timo>
gmorell: exactly... and sadly my work to add perl modules to recipetool/devtool is stalled by how metacpan.org metadata is stored (nested ElasticSearch is not queryable)
<gmorell>
ye, I figured it out eventually, had to add some bbappends to enable the -native
<moto-timo>
gmorell: the "it's in my head but it should be documented" approach is to search metacpan.org for whatever it is and see if it is a sub-module or a stand alone.
<moto-timo>
gmorell: please send a patch for the -native if that was lacking.
<gmorell>
moto-timo: I will do so once I get this going
<moto-timo>
gmorell: in some recipes you'll see I have much more verbosely added RPROVIDES... but that was tedious and probably brittle (folks, including me, do not check everything when a new AUH upgrade is available)
<moto-timo>
so libcgi-perl could have helpfully had RPROVIDES libcgi-cookie-perl
<RP>
wmills_, JPEW, zeddii: I merged that hashserv fix FWIW
<RP>
JPEW: X would have rehashed due to the quilt-native equivalence
<JPEW>
RP: Ah, OK, the rehashing happens before perlcross-native rebuilds?
<RP>
JPEW: yes, on X due to quilt-native having an equivalence
<JPEW>
RP: Yes, that makes sense then
alimon has quit [Ping timeout: 255 seconds]
tgamblin_ has joined #yocto
<RP>
JPEW: I think I'm a little happier knowing how it could happen. Next question is what to do about it :/
tgamblin has quit [Ping timeout: 252 seconds]
<JPEW>
RP: Well... would it really have an adverse effect on end users? I was sort of under the impressing the test was "peeking under the hood" to try and figure things out and hashequiv was confusing the test
Chocky1 has quit [Ping timeout: 252 seconds]
<RP>
JPEW: the idea is the user should be able to use the tools to understand what changed or why sstate wasn't being used. This stops that a bit unless we can mitigate it, which there is one potential idea for
<RP>
JPEW: the tests were designed as "change something" and see if we can report what changed
Chocky has joined #yocto
Chocky has quit [Client Quit]
Chocky1 has joined #yocto
alimon has joined #yocto
<RP>
JPEW: still not 100% convinced it was this but I guess it is a possible way it could happen :/