00:13 <JaMa> "file: upgrade 5.43 -> 5.44" from https://git.openembedded.org/openembedded-core/commit/?id=6e93f815e9439ad351c3b9a382accf7b3ba9a6e2 has also a bit surprising side-effect, one of recipes started to fail with: ERROR: QA Issue: /usr/share/nudgeScript/logVerifier22.py contained in package lib32-nudge-dev requires /usr/bin/python, but no providers found in RDEPENDS:lib32-nudge-dev? [file-rdeps]

00:14 <JaMa> and the same script is "Objective-C source, ASCII text" by file-5.43 and "Python script, ASCII text executable" by file-5.44

00:17 <JaMa> possibly thanks to https://github.com/file/file/commit/d9e0d13529b7d90e3b45330e9523d666da863895

02:08 <uniqdom> what happens in an embedded system if log size increase? my setup has /var/logs as a symbolic link to /var/volatile, which is a tmpfs.

02:09 <uniqdom> how do I know if the logs will rotate?

10:51 <RP> JaMa: sounds like a decent bugfix for file! :)

14:02 <JaMa> RP: kanavin: the issue with x86-64-v3 I've mentioned yesterday was actually caused by something in our DISTRO, we were still using QEMU_EXTRAOPTIONS_core2-64 for qemux86-64 builds, is there a reason not to define QEMU_EXTRAOPTIONS_x86-64-v3 = " -cpu Skylake-Client,check=false" in the tune file? Today I've tested that using this fixes the issue on our jenkins builders (now will try without any EXTRAOPTION

14:02 <JaMa> defined)

14:03 <RP> JaMa: sounds like an oversight

14:05 <JaMa> if it doesn't work correctly without EXTRAOPTION, I'll send patch to add it

14:11 <RP> JaMa: sounds good thanks

14:31 <JaMa> my Threadripper, doesn't like -cpu Skylake-Client (bunch of warnings without check=false, instead of Illegal Instruction seen with -cpu core2duo), the autodetected cpu seems to work fine, but explicitly setting -cpu might be better for future and reproducibility (that qemu always tries to emulate the same cpu, not based on host system)

14:32 <JaMa> some details from meson-qemuwrapper: https://pastebin.com/hyNV3QwF

15:15 <uniqdom> Hello, I'm having a problem while trying to build a recipe with latest git commit. I have SRCREV = "${AUTOREV}", PV = "0.1+git${SRCPV}", and SRC_URI = "git://git@hidden.com:20023/hidden/hidden.git;protocol=ssh;branch=master"

15:16 <uniqdom> it doesn't want to download the latest commit. it's like if it doesn't know that there is a new commit at all.

15:17 <uniqdom> if I put the commit hash in SRCREV, it updates the git code correctly

15:24 <RP> uniqdom: have you changed the default cache policy? Is the recipe reparsing each time you run bitbake?

15:27 <RP> denix: I've dropped meta-ti testing from langdale as the layer is no longer compatible

15:33 <uniqdom> RP: I haven't changed the default cache policy. yes, it is reparsing each time.

15:38 <RP> uniqdom: when it reparses it should be querying the revision so I'm not sure why that isn't working :(

18:39 <RP> zeddii: I did revisit the patches in https://git.yoctoproject.org/poky-contrib/log/?h=zedd/kernel but I think something is wrong. For example, https://git.yoctoproject.org/poky-contrib/commit/?h=zedd/kernel&id=0fd108c99e66a55dca19cef428d9716b9b18b3e5 changes meta-yocto-bsp and not the qemu machines

19:18 <uniqdom> I have added a few python module recipes autogenerated using pipoe. Now I'm struggling to get fcntl python3 module somehow.

19:55 <uniqdom> nvm, I found that there was a recipe for python3-fcntl by using oe-pkgdata-util

20:09 <RP> zeddii: you said you'd do a v2 but I suspect holidays and other stuff then happened! :)

22:16 <uniqdom_> is there a way to show the list of recipes included by a target like core_image_base?

22:51 <moto-timo> JaMa: this makes me wonder about best -cpu settings for my Ryzen host...

23:09 <hsv> bitbake is reading my bb file but not bothering to compile my simple program - https://paste.debian.net/plainh/f0323454

23:10 <hsv> withe LICENSE commented, bitbake complains do i know it's being read.

23:10 <hsv> *so

23:11 <hsv> i put a deliberate error in the .c file so it should fail, but the build succeeds.

23:22 <RP> hsv: what does the do_compile log show it doing?

23:26 <hsv> RP: sorry, where can i find that?

23:30 <RP> hsv: ${WORKDIR}/temp/log.do_compile

23:33 <hsv> WORKDIR being named after the layer, or bb file? either way it's not found.

23:36 <hsv> i wonder if it is the recipes dir, i called it recipes-programs (made up).

23:37 <RP> hsv: look it up for the recipe with bitbake-getvar or from the output of bitbake <recipe> -e

23:38 <RP> hsv: it is a bitbake variable

23:43 <hsv> bitbake-getvar -r <recipe> WORKDIR

23:44 <hsv> It shows a plausible path, but that doesn't exist.