<rburton>
Guest87: sounds like your code needs updating for more pedantic c++ stdlib
<Guest87>
rburton i was also able to use memcpy with <string.h> in kirkstone, dunfell, zeus. then it seems that it will be sufficient to include <cstring> in my code where these operations are performed. i guess it's because of the compiler update, isn't it?
<rburton>
yes
<rburton>
newer compilers are more pedantic/strict
florian_kc has quit [Ping timeout: 272 seconds]
<Guest87>
if i need to use both string.h and cstring in the same class, will it be a problem? does the compiler set the priority itself? i can find out by trying. it like, i want some ready answer :D
<Guest87>
rburton even these informations enough, thank you.
Schlumpf has joined #yocto
vladest has quit [Quit: vladest]
vladest has joined #yocto
pidge has quit [Ping timeout: 240 seconds]
pidge_ has joined #yocto
DvorkinDmitry has joined #yocto
<DvorkinDmitry>
my image build fails with "Postinstall scriptlets of ['oac-php'] have failed. If the intention is to defer them to first boot,
<DvorkinDmitry>
then please place them into pkg_postinst_ontarget_${PN} ()". I'm using simple do_install() {} in my .bb file. log.do_rootfs says: re.error: invalid group reference 19 at position 1. what may be wrong?
<rburton>
i wonder if the @ is upsetting the regex matching
pabigot has quit [Ping timeout: 245 seconds]
<kayterina[m]>
Hello, I have a recipe and its debug package does not have an ELF interpreter set. I have to set it manually with patchelf. The package is in "/packages-split/<recipe>-dbg/usr/bin/.debug/"
<kayterina[m]>
Has anyone seen this before? Am I looking into something yocto-specific or something in the cmake files of the recipe?
<rburton>
kayterina[m]: almost definitely the cmake files
tokamak has joined #yocto
<rburton>
but .debug/ files are not meant to be ran, they're just the debug segments
pabigot has joined #yocto
<kayterina[m]>
Hm, I would swear it run when I put the interpreter there. Then where should the debug files be built?
<rburton>
oh you can never run anything in .debug/
<rburton>
its literally just the debug symbols
<rburton>
what do you mean by "debug files"?
<kayterina[m]>
the non stripped executable
<rburton>
there is no non-stripped executable. install the -dbg package and gdb will load the .debug/ file to get the symbols.
<DvorkinDmitry>
rburton, what would you recommed? before it was working good
<rburton>
DvorkinDmitry: figure out where the bad regex group is coming from
<DvorkinDmitry>
if I write like this, SYSTEMD_SERVICE:${PN} = "php-fpm\@9000.service", it says it can't find the file
<rburton>
i'd try with master if you can, because there were fixes to the regex
<rburton>
if you look at the git log for that class you can probably apply them locally to test
<DvorkinDmitry>
rburton, I['ve been using dunfell for a long time before and it was working...
<DvorkinDmitry>
it has been corrupted in 2aa82324d43467e7c8bfbbb59570ee3306264b75 commit
<rburton>
DvorkinDmitry: please file abug or ideally a patch, and for even more bonus points a test for oe-selftest :)
amitk__ has joined #yocto
amitk has quit [Ping timeout: 245 seconds]
grma has quit [Ping timeout: 252 seconds]
xmn has joined #yocto
<vvn>
I see that some bsp layers try not to define too much machines and let the use customize base configs. Would you guys recommend the end user to define their own machines to describe their products, or is it better to have machine tweaks in their distros?
<rburton>
inheriting a reference machine and customising it makes sense
<vvn>
taking the example of the beaglebone or raspberry pi, what I'm not comfortable with is having to rebuild a few (machine-) packages like the kernel while in fact the machine-specific bits didn't really change (for example the machine config only chooses a specific dtb)
<vvn>
but I guess this is something that can be improved at the package level? (fine tuning it more specifically than MACHINE_ARCH?)
Xagen has joined #yocto
camus has quit [Quit: camus]
sakoman has joined #yocto
kpo_ has quit [Ping timeout: 272 seconds]
amitk_ has quit [Ping timeout: 245 seconds]
alessioigor has quit [Quit: alessioigor]
alessioigor has joined #yocto
<LetoThe2nd>
jonmason: i'm obviously too stupid to figure a bit of kas out. the docs say you can set a layer-path to direct to a sublayer in a nested directory structure. you don't happen to have an example of that set up somewhere?
amitk has joined #yocto
amitk has quit [Ping timeout: 250 seconds]
<derRichard>
LetoThe2nd: you mean pointing kas to a layer at a given path instead a git/svn/hg repo?
<LetoThe2nd>
derRichard: no.... pointing to a local layer in the repository that also holds the kas.ymls is common. but all layers that i've ever used it with are at the top level. now i need a blablah/meta-blah, essentiall.y
<jonmason>
LetoThe2nd: we're lazy and just use the local reference, but it should just work
<jonmason>
let me try something out to verify
<LetoThe2nd>
jonmason: gah, i just tried the obvious and it seems to work.
leon-anavi has quit [Remote host closed the connection]
florian_kc has joined #yocto
otavio has joined #yocto
florian_kc has quit [Ping timeout: 272 seconds]
<tlwoerner>
anyone else seeing this when doing "bitbake -c menuconfig linux-yocto"?
<tlwoerner>
scripts/Kconfig.include:50: Sorry, this assembler is not supported.
<khem>
Guest87: so please include <cstring> instead of string.h
<rburton>
tlwoerner: jonmason said that happened
<rburton>
clearly we need to somehow selftest menuconfig
<rburton>
zeddii: ^^^
<jonmason>
It was working yesterday, but not on zeddii's kernel branch
<jonmason>
it could've been my env, so I didn't complain at the time
marc1 has quit [Ping timeout: 246 seconds]
marc1 has joined #yocto
zelgomer has joined #yocto
<zelgomer>
VAR:override:append versus VAR:append:override, is there any good reason to prefer one over the other?
goliath has joined #yocto
sakoman has quit [Quit: Leaving.]
amitk__ has quit [Ping timeout: 245 seconds]
alessioigor has quit [Quit: alessioigor]
alessioigor has joined #yocto
alessioigor has quit [Quit: alessioigor]
alessioigor has joined #yocto
nemik has quit [Ping timeout: 272 seconds]
nemik has joined #yocto
<JaMa>
zelgomer: yes, they work very differently and in most cases you want VAR:append:override, try it and look at bitbake-getvar output to understand the difference