<zeddii>
moto-timo: is this a config that I can test ? I can see what's up with it tomorrow
<moto-timo>
zeddii: it turns out WDAT isn't supported on the target platform anyway, so no efibootguard for this project
<zeddii>
moto-timo: aha. I ran the config_analysis task on the option, if I had the right BSP, it would have told me more, but I have qemux86-64 in that build,
<moto-timo>
zeddii: this was meta-intel intel-skylake-64 linux-intel-rt 6.6 in scarthgap
<zeddii>
but in linux-yocto-dev at least, it only depends on ACPI and WATCHDOG
<moto-timo>
I ran menuconfig and then diffconfig and that's where CONFIG_ACPI_WATCHDOG=y barfed out
<zeddii>
it actually selects ACPI_WATCHDOG in 6.11+, so it doesn't depend on it per-se, but does try and enable it, but could still show up in the .config even without ACPI_WATCHDOG
<moto-timo>
even though I was just (re) setting CONFIG_WDAT_WDT=m which should already have been there from intel-x86.cfg inheritance
<moto-timo>
apparently WDAT table has to be supported at BIOS level ... and it isn't on this skylake platform from 2016/2017 era
<moto-timo>
if this was newer than skylake I would still be able to use ITCO_WDT, but v4 is not supported by efibootguard
<moto-timo>
and efivarfs is not supported on -rt kernel
<Guest46>
is there anyone who is familier with phys. I have upgraded my linux from 4.9 to 6.136. My phy is connected to mac over mii through smi interface. Nothing needs to be configured so using generic linux driver. I have used the same device tree as in older kernel. In Uboot I am able to read registers through mii dump where the link is shown do be
<Guest46>
established. In linux link is never established. The dmesg comparison between old and new kernel has a difference in it. I am using nxp's soc. Only changes being made in "arch/arm/mach-imx/mach-imx7d.c" in "static void __init imx7d_enet_clk_sel(void)" to set clock to external oscillator. I have also tried with kernel 6.6.23 without any luck. Any
<RP>
seems on fedora41, pseudo can't see what tar is doing :(
dkc has quit [Remote host closed the connection]
dkc has joined #yocto
jpuhlman has quit [Ping timeout: 248 seconds]
jpuhlman has joined #yocto
Guest77 has quit [Quit: Client closed]
<RP>
if I build tar-native and use that it works. If I use the host's tar, it does not
luc4 has joined #yocto
<RP>
Looks like it is acl functions in glibc we're missing
ThomasRoos has joined #yocto
<ThomasRoos>
Hi, gettting an error only when building arm32 (raspberrypi-2) in a docker container:
<ThomasRoos>
2024-10-02T12:47:32.7929040Z ../xorgproto-2023.2/meson.build:22:0: ERROR: Executables created by c compiler arm-poky-linux-gnueabi-gcc -mthumb -mfpu=neon-vfpv4 -mfloat-abi=hard -mcpu=cortex-a7 -fstack-protector-strong -O2 -D_FORTIFY_SOURCE=2 -Wformat -Wformat-security -Werror=format-security -D_TIME_BITS=64 -D_FILE_OFFSET_BITS=64
<ThomasRoos>
--sysroot=/codebuild/output/src3044033969/src/actions-runner/_work/meta-aws-demos-private/meta-aws-demos-private/build/tmp/work/cortexa7t2hf-neon-vfpv4-poky-linux-gnueabi/xorgproto/2023.2/recipe-sysroot are not runnable.
<ThomasRoos>
-> any ideas?
davidinux has quit [Quit: WeeChat 4.1.1]
mbulut has quit [Ping timeout: 252 seconds]
mbulut has joined #yocto
<rburton>
ThomasRoos: the machine probably doesn't set qemu-user
ThomasRoos has quit [Ping timeout: 256 seconds]
tepperson has joined #yocto
<tepperson>
for a native recipe, how would I update the paths or environment variables for binaries installed by the recipe?
rob_w has quit [Remote host closed the connection]
reatmon_ has quit [Remote host closed the connection]
reatmon_ has joined #yocto
<rburton>
tepperson: you mean hardcoded paths in the binaries?
<rburton>
ThomasRoos: can you share the entire configure log for that recipe?
<tepperson>
rburton: I mean like I am building a native recipe that installs into ${STAGING_DIR_NATIVE}/blabla/bin and I need ${STAGING_DIR_NATIVE}/blabla/bin added to the PATH environment variable
<rburton>
tepperson: easier to just put it into the right place?
<rburton>
ensure the binaries are in ${bindir}, sorted
<rburton>
EXTRANATIVEPATH can be used if the binaries end up in STAGING_DIR_NAME/usr/bin/somename
<rburton>
if you _can't_ just put the binaries in the right place then just add STAGING_DIR_NATIVE/blaba/bin to PATH in the recipes that need to know where the binaries are
<rburton>
ThomasRoos: i'm guessing your machine explicitly disables the qemu-usermode MACHINE_FEATURE. its typically backfilled into the configuration.
<rburton>
ThomasRoos: very odd. have a look at the generated meson.cross file in the xorgproto work directory. it should have an exe-wrapper assignment. maybe its set but qemu crashes?
<rburton>
oh
<rburton>
yeah
<rburton>
that can happen
<rburton>
something like qemu does fun memory allocation games and docker breaks it
<ThomasRoos>
raspberry-64 machine works just fine...
<qschulz>
"CPython now has experimental support for running in a free-threaded mode, with the global interpreter lock (GIL) disabled. This is an experimental feature and therefore is not enabled by default"
<tgamblin>
yeah, RealPython has upped their enshittification game
<qschulz>
FWIW, I find eht official python docs to be so good that I only want to read stuff from there
MathieuD1 is now known as mathieudb
mathieudb has quit [Changing host]
mathieudb has joined #yocto
jmd has quit [Remote host closed the connection]
Hazza has joined #yocto
tgamblin has quit [Read error: Connection reset by peer]
<RP>
I think this is fork() vs vfork()
Haxxa has quit [Ping timeout: 265 seconds]
<RP>
pseudo has no vfork wrapper but is probably shouldn't have
<RP>
well, can't
tgamblin has joined #yocto
mckoan is now known as mckoan|away
amitk has quit [Ping timeout: 265 seconds]
Kubu_work has quit [Quit: Leaving.]
zpfvo has quit [Remote host closed the connection]
<tepperson>
is there a guide for writing native recipes somewhere?
<Crofton>
BBCLASSEXTEND += "native" hopefully
<rburton>
tepperson: there's very little special. either inherit native if the recipe is _explictly and only_ native, or just bbclassextend a normal recipe
<rburton>
use the variables and they typically just work
<rburton>
or ask a more concrete question
<tepperson>
rburton: for install, files get installed to ${D}, like ${D}/usr/blabla ?
<tepperson>
rburton: would ${D}${prefix}/blabla be ok?
<rburton>
if you insist on putting binaries in a special directory to make it harder to run them
<tepperson>
rburton: I am trying to write a native recipe for quartus lite and it seems to do odd things
rfuentess has quit [Remote host closed the connection]
<RP>
tepperson: the paths in the native case do look a little odd at first glance as we build and install them to <workdir>/<sysroot>/usr/bin, not the /usr/bin. That means prefix has a longer value than you'd expect, not just /usr. It also means there is path duplication for ${D}${prefix}
<RP>
that is all expected/normal and just disconcerting when you first see it
florian has quit [Quit: Ex-Chat]
florian_kc has quit [Ping timeout: 264 seconds]
alcroito has joined #yocto
<alcroito>
Hi. I'm trying to find out if yocto collects copyright assignments to embed them into its generated spdx SBOM. Looking through some of the source code, so far I think it doesn't. Is that a correct assessment?
<rburton>
JPEW: ^
<RP>
alcroito: we collect up source references. You'd have to process the sources to collect any copyright assignments
<JPEW>
alcroito: No copyrights
<alcroito>
Ok. Thank you for confirm. And for collected licenses, is it correct that the main source is the LICENSE key from each recipe?
<RP>
JPEW: there should be broken mingw builds on ubuntu2404-vk-2 and ubuntu2404-vk-3 btw
<JPEW>
alcroito: We scan sources for the SPDX license expression
<JPEW>
RP: Ya, I'm logged in.... where are the failed builds?
<RP>
(mv build-renamed to build to match it as built)
mischief has quit [Quit: WeeChat 4.4.2]
<alcroito>
JPEW: Interesting that the source files are scanned for spdx license expressions, but not spdx license copyrights texts. I guess it might be related to the fact that there's a DB of licenses, but copyrights are arbitrary?
<JPEW>
I don't think there is a standard way of expressing copyright like there is for licenses
mischief has joined #yocto
<JPEW>
(at least AFAIK)
<RP>
JPEW: there is, it is just ugly and people don't really see what it adds
<JPEW>
Fair
mischief has quit [Client Quit]
<RP>
we could scan for that and just Copyright I guess
<alcroito>
at least reuse spec mentions a common way to specify copyrights https://reuse.software/spec-3.2/ via SPDX-FileCopyrightText
<JPEW>
SPDX-SnippetCopyrightText would be fairly simple to add support for. Look for where SPDX-License-Identifier is used
<alcroito>
I obviously can't say authoritatively, but at least all the Qt framework sources use "Copyright"
Kubu_work has joined #yocto
<JPEW>
Ya, I can't say I've ever seen SPDX-SnippetCopyrightText
<RP>
I've seen the proposal/spec and wondered what it adds compared to simply "Copyright"
<JPEW>
RP: Unambiguous regex matching
<JPEW>
But not much else
<alcroito>
Just Copyright is fine to, according to spec. I guess they added it for grepping, as mentioned
<alcroito>
One final question, i briefly watched the FOSSDEM sbom presentations, and there was a brief mention about trying to integrate 3rd party spdx sboms into the yocto generated ones. I assume there's no integration point for that at the moment, and no immediate plans to introduce that?
<JPEW>
alcroito: No. It's a dream ATM :)
<alcroito>
I figured. :'(
<JPEW>
It's a more reasonable dream with SPDX 3, but still just a dream
<alcroito>
By the way, do you know what would be a good place to hang out around to hear news about spdx sboms in general? Would that be the spdx.org mailing lists only?
druppy has quit [Remote host closed the connection]
druppy has joined #yocto
norvil has joined #yocto
<norvil>
hi! I'm working on a beagleboardx15 and I'm trying to enable libgpiod to be able to use gpio pins, so I added this at the local.conf: PREFERRED_VERSION_libgpiod = "1.6.3"
<norvil>
but when I open the image on the beagle, i can´t use it. It doesn't show any erros, the leds just don't turn on
<norvil>
does anyone know how to fix it or what packages am I missing??
jmd has joined #yocto
steelswords94 has quit [Read error: Connection reset by peer]
steelswords94 has joined #yocto
druppy has quit [Ping timeout: 252 seconds]
<tepperson>
i have a script that runs the locale command during the build, what is the proper way to make that available to a recipe?
florian_kc has joined #yocto
<JPEW>
RP: Sorry, how do I rename the directory as the pokybuild user? Am I missing some permission to sudo or su to that users?
<RP>
JPEW: "sudo -iu pokybuild" will let you become that user?
<JPEW>
RP: it wants my password when I do that
<RP>
halstead: ?
<RP>
halstead: can you help JPEW with ubuntu2404-vk-2 access?
<halstead>
RP: JPEW: Yes.
<halstead>
JPEW: You are missing the groups needed for that. I'll see why.
<halstead>
JPEW: Yeah the builders group is commented out from setting up your account on the hashserv. I'll fix it everywhere. two minutes.
<halstead>
JPEW: Please log out and reconnect then try again.
<JPEW>
halstead: That fixed it. Thanks
<halstead>
JPEW: You're welcome. Hopefully it didn't slow you down too much.
<khem>
qschulz: yeah its disabled and I am also curious about JIT and its impact on bitbake with log running builds
steelswords94 has quit [Read error: Connection reset by peer]
steelswords94 has joined #yocto
jmd has quit [Remote host closed the connection]
cabazon has joined #yocto
hcg has joined #yocto
xmn has joined #yocto
ptsneves has joined #yocto
Vonter has quit [Ping timeout: 244 seconds]
<hcg>
I wonder if anyone could supply some advice on an issue we have currently - we are using fitImages as out boot images. We have a single kernel, 2 devicetrees and 2 config sections. We have multiple different products using the same core SoC, but each product has it's own baseboard which can have many additional devices. The 2 devicetrees are
<hcg>
bascally the main devicetree we normally boot with and the secondary is a devicetree used for recovery boot. The recovery boot requires a different devicetree to the primary boot. Giving the known that we have multiple devices using the same base core image and devicetree, we now provide additional devices, and these are provided in Yocto recipes -
<hcg>
In each of these recipes, we provide reference to the real hardware, but then we also apply devicetree entries to the rank...
dmoseley has joined #yocto
dmoseley_ has quit [Ping timeout: 265 seconds]
norvil has quit [Quit: Client closed]
hcg has quit [Ping timeout: 256 seconds]
ptsneves has quit [Quit: ptsneves]
Hazza has quit [Quit: Haxxa flies away.]
Haxxa has joined #yocto
<tepperson>
is there a recipe to build locale for native use in yocto?
<RP>
I'm still struggling to debug this pseudo issue after spending a day beating it up :(