**** BEGIN LOGGING AT Wed Feb 28 03:00:02 2018 Feb 28 03:31:14 naddy: you might be removing sysvinit from DISTRO_FEATURES Feb 28 04:31:25 Hi all Feb 28 04:34:16 I'm getting this error trying to boot a image in a zynq board: "Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(179,2)" using a SD card, the whole log is here: https://0bin.net/paste/KqdlYfegvUzIGr1x#-XSyrywU0gEDvZeuSA1VMLi1pG/D3WYFUEDZtH+qXz6 Feb 28 04:35:41 I created the partiion #2 using fdisk and and the tar -zxvf the core-image-minimal-zedboard-zynq7.tar.gz on it. I tried with a ext3, and ext4 partition Feb 28 07:39:29 New news from stackoverflow: Unable to clone meta-openembedded Feb 28 09:25:39 Now I'm receiving a Waiting for root device /dev/mmcblk0p2 message and then it hangs.. any idea? Feb 28 09:34:40 : do you have a second partion if yes check the type and check /etc/fstab Feb 28 09:35:42 it's a ext4, it's created Feb 28 09:36:06 And I have my fs in the partition2 Feb 28 09:39:54 New news from stackoverflow: How can I fix binutils fetch error || I can't print the vprintf log by journalctl Feb 28 10:05:30 I'm trying to do IMAGE_NAME = "${IMAGE_BASENAME}-${MACHINE}-${BUILDNAME}" Feb 28 10:05:30 and now I just can't get the vardeps/checksum things right... Feb 28 10:22:40 I guess BUILDNAME is an extra special variable... ? Feb 28 10:24:26 : whats in your /etc/fstab on /dev/mmcblk0p2 Feb 28 10:25:25 welhm, it's not even looking into the partition... it can be empty or with a file system Feb 28 10:28:11 your /dev/mmcblk0p2 should contain your rootfs Feb 28 10:28:27 welhm, yes it does Feb 28 10:28:30 oke Feb 28 10:29:51 welhm, it's my /etc/fstab https://0bin.net/paste/umeCIioyjrG5OLh-#MuKtjUlPK8yHKA2zGAytaDEQYYKxG3XdAyGOzuHXLgI Feb 28 10:31:32 are you using uboot ? Feb 28 10:33:01 if yes, is there somthing like: root=/dev/mmcblk0p2 in your uEnv.txt Feb 28 10:33:19 Yes Feb 28 10:33:26 and yes Feb 28 10:36:42 sd-card unlocked ? Feb 28 10:39:34 Hi, I'm trying to add networkd to my distro. This is what I've done: configure systemd as default initmanager - add my own recipe and a systemd.bbappend file but this one does not appear in the list of the appends. I'm sure I'm missing something but I cant figure out the point ! Feb 28 10:43:44 : ( i'm a bit of a noob regarding recipes) but as far i know the systemd.bbappend needs to be named like the systemd.bb file ( like systemd_229.bbappend Feb 28 10:45:15 the filename is systemd_%.bbappend so, as far as I understand well, this file always applies Feb 28 10:50:20 in fact, I do not know where to store the bbappend file. Currently it is stored in recipes-core/systemd but I'm not sure it is the correct... Feb 28 10:50:38 can be anywhere Feb 28 10:57:04 uhm... any idea why it is not used then ? Feb 28 10:57:40 can be that its blocked by a other network entety Feb 28 11:01:23 like initv can block systemd Feb 28 11:04:59 you have. PACKAGECONFIG_append = networkd in your bbappend ? Feb 28 11:20:46 Can anybody please help me to create toolchain for windows? Feb 28 11:38:12 rahul12789: add meta-mingw to bblayers and set SDKMACHINE to one of the options in meta-mingw, such as SDKMACHINE = "i686-mingw32" or x86_64-mingw32 Feb 28 11:39:16 Thanks I'm doing same Feb 28 11:39:27 then bitbake myimagename -c populate_sdk Feb 28 11:39:34 Ya Feb 28 11:39:45 RP: fired a mut, bit smaller than next and might actually build Feb 28 11:40:06 rburton: I have a fix for my bitbake runqeueue patch Feb 28 11:40:12 cool Feb 28 11:40:12 rburton: go stuff isn't ready Feb 28 11:40:20 no, discovered :) Feb 28 11:41:11 rburton: toolchain will be in .sh script right? will it work in windows If I'm extracting in linux and copiying to windows? Feb 28 11:42:24 rburton: repushed -next which I think is good Feb 28 11:43:03 rahul12789: you'll need to unpack on windows but you need sh etc anyway - its called a mingw toolchain for a reason Feb 28 11:43:29 Fine Feb 28 11:44:22 rburton: After unpacking that toolchain, I need to configure it into eclipse yocto adt plugin right? Feb 28 11:44:35 sounds right. never used eclipse Feb 28 11:45:04 rburton: Then what are you using ? makefiles? Feb 28 11:45:38 i work on oe itself, not writing apps using the toolchain Feb 28 11:47:21 rburton: Fine understood :) Feb 28 12:14:02 Please help me with this error: Kernel panic - not syncing: VFS: Unable to mount root fs on unknown(179,2)... it's a zedboard Feb 28 12:14:36 Tried with morty and rocko braches Feb 28 12:20:20 this is the serial output: https://0bin.net/paste/2MdVDgrGhk89+jbj#m7jfllJLuaFM-NUYlMjAnrJy7kkpJGO0ad5BIBVbppQ Feb 28 12:23:08 @welhm: sorry for late answerr, i was outside .. In the bbappend file, there a line PACKAGECONFIG += " networkd resolved timesyncd hostnamed .... and much other" Feb 28 12:29:22 : i think you need to add rootfstype=ext4 to uEnv.txt Feb 28 12:29:42 welhm, tried that... didn0t work Feb 28 12:31:33 if you replace /dev/root / auto defaults 1 1 | with: /dev/root / ext4 defaults 1 1 Feb 28 12:37:16 welhm, the error is happening before it mounts the SD partition, so, it's not even looking inside of the partition Feb 28 12:37:30 if you add a do_install_append() to the .bbappend and add there a bb.warn than you must see the warning when building systemd and you are sure that the script is run Feb 28 12:37:54 Guys, shat is the difference in Yocto - meta-extsdk-toolchain vs do_populate_sdk_ext for core image? Why do we need meta-extsdk-toolchain at all? Feb 28 12:40:32 New news from stackoverflow: What is the difference in Yocto - meta-extsdk-toolchain vs do_populate_sdk_ext for core image? Feb 28 12:41:35 yocti: This is the qestion I asked a couple of minutes before!! Feb 28 12:41:36 zks: Error: "This" is not a valid command. Feb 28 12:41:56 > yocti: This is the qestion I asked a couple of minutes before!! Feb 28 12:42:52 Guys! Anyone here? )) Feb 28 12:46:52 zks: if you want a SDK, use populate_sdk or populate_sdk_ext Feb 28 12:47:14 (yocti is a bot) Feb 28 12:47:19 rburton: What is meta-extsdk-toolchain then for? Feb 28 12:48:27 rburton: Thanks, man BTW! )) Feb 28 12:51:04 there is somthing preventing the /dev/mmcblk0p2 is ro it should be rw Feb 28 12:51:16 can be driver or the sw on the sd-card Feb 28 12:54:47 I'm not shure if im right. but as far i know the ext sdk is an extensible sdk Feb 28 12:56:29 hello, i'm trying to use gdb on the target machine (poky 2.2) but i don't have the debug symbols for the standard c++ library (libstdc++ i believe) even tho i have the IMAGE_FEATURES dbg-pkgs and INHIBIT_PACKAGE_STRIP = "1" Feb 28 12:56:37 Is there anything else to enable? Feb 28 12:59:50 zks: its part of the sdk Feb 28 13:49:44 hello everyone Feb 28 13:50:34 rburton: Thanks man! Feb 28 13:51:19 You can answer this question on stackoverflow to boost your profile Feb 28 13:51:32 https://stackoverflow.com/questions/49029703/what-is-the-difference-in-yocto-meta-extsdk-toolchain-vs-do-populate-sdk-ext-f Feb 28 13:54:58 hi i am new to yocto/poky/openembedded. I try to build "complete" iptables support into my build. So i added iptables to my local.conf in IMAGE_INSTALL_append. The required kernel-modules (which are listed in the iptables.bb under RRECOMMENDS) are not build into the image automatically however. Is there a way to do this, apart from adding every single module in the kernel menuconfig? Feb 28 14:45:37 RP: I think you missed v3 of my patch "icecc.bbclass: Ignore more icecream vars in hashes" on master-next (4d9ddb9).... would it be easiest for me to submit a new patch to fix it up on top? Feb 28 14:46:47 JPEWhacker: thanks for spotting, I can fix Feb 28 15:04:21 you mean, by adding tools-sdk? Feb 28 15:59:01 : I'm sure the file is not processed because i intentionally introduce an error and bitbake does not complain.. Feb 28 16:11:15 New news from stackoverflow: Amend boot cmdline in custom image build **** BEGIN LOGGING AT Wed Feb 28 16:15:33 2018 Feb 28 16:39:18 hi, is anyone build u-boot for xilinx-ultrazed , I'm getting error from u-boot [PMUFW is not found], and pmu-firmware recipe is being skipped due to uncompatible host, do we need vivado tool-chain to generate proper boot image? Feb 28 17:12:16 Is it too late to get a backport into Pyro? Feb 28 17:30:24 JPEWhacker: nope Feb 28 17:40:19 rburton: I am seeing some issues with go programs segfaulting with glibc 2.27 :( Feb 28 17:40:23 on aarch64 Feb 28 17:40:39 Can we have one architecture for the planet please Feb 28 17:41:56 khem, one ring to bind them all? Feb 28 17:42:08 khem: yes, if it is also open source Feb 28 17:42:21 RISC-V, I believe? Feb 28 17:43:31 * armpit haha aarch64 is RISCY ; ) Feb 28 17:51:23 the distro I'm working with has USERADD_ERROR_DYNAMIC set to 'error'. How do I add a new user ID? Manually? Or is there a better method? Feb 28 18:08:18 kanavin1: Definitely, you have the right ideas Feb 28 18:09:11 I am going to give go 1.10 patches a go Feb 28 18:09:17 and see if that helps Feb 28 19:13:11 I have a bunch of users and groups to add, any pointers are appreciated. Feb 28 19:15:33 Use useradd/groupadd on the target, copy files back to my layer? disable the error and copy new users from the build output? Feb 28 19:18:33 RP, the pyro-next night lsb failure looks to be kvm related. I just reproduced in on ubuntu 17 Feb 28 19:54:00 nathani_: Are you still looking for a (one time) method to automate adding the IDs? Feb 28 19:54:26 indeed Feb 28 19:56:02 khem: well we either take glibc 2.27 this week or wait until 2.6 Feb 28 19:56:44 I don't know of anything, sorry, but.... is supect it is something that could be scripted with the bitbake's python interface (tinfoil). Feb 28 19:57:22 nathani_: easy way is to set the variable to something other than error Feb 28 19:57:35 iirc you get errors when you use them in an image then Feb 28 19:57:52 That too :) Feb 28 19:59:29 ok, so i set to warn, and now I'm in the middle of merging the output back, but it's a pain Feb 28 20:00:23 nathani_: don't use useradd-static? Feb 28 20:01:03 there's an argument that oe-core should ship mapping for all the recipes it ships Feb 28 20:01:21 a selftest and a file for core seems sensible Feb 28 20:02:32 That would have been nice :) Feb 28 20:02:37 rburton: I just reproduced the segfault with go-helloworld so should be debuggable Feb 28 20:02:42 cool Feb 28 20:06:08 khem: if you can have a look at poky-contrib:ross/glibc, i plan to submit that to the list shortly Feb 28 20:06:15 all green on the AB \o/ \o/ Feb 28 20:42:53 rburton: ok found the problem and fix too Feb 28 20:43:18 the fix can go after glibc 2.27 merge Feb 28 20:46:41 rburton: sent the fix to ml Feb 28 20:47:56 rburton: patch looks ok Feb 28 22:30:54 How can I get paid support? Feb 28 22:43:12 armpit: hmm, why don't we see it with master? Feb 28 22:43:44 its the 4.1 kernel maybe Feb 28 22:44:44 possible the meltdown changes. I was going to move the kernel hashes to double check.. Feb 28 22:47:01 now I have a simple testcase to work with Feb 28 23:00:18 armpit: it does sound like you're narrowing it down which is good Feb 28 23:04:39 JoseBravo: contact one of the many companies listed on the yoctoproject page Feb 28 23:07:09 rburton, is that section of the page hidden? can't find it Feb 28 23:07:36 JoseBravo: https://www.yoctoproject.org/ecosystem/yocto-project-participants Feb 28 23:08:50 rburton, thanks! Mar 01 00:10:03 * armpit damn..it is the kernel **** ENDING LOGGING AT Thu Mar 01 02:59:59 2018