**** BEGIN LOGGING AT Tue Oct 24 03:00:01 2017 Oct 24 06:36:23 hi all Oct 24 06:38:12 hi Oct 24 06:45:13 was there a way to show recipes that will be built before building a target? Oct 24 06:45:17 thanks Oct 24 07:46:27 hello Oct 24 07:48:28 in a team of yocto users, is it possible to share the storage space so that you can use a local machine for processor, memory, etc. and the shared storage for build storage space? Oct 24 07:49:04 opposed to our current status where we have for each build machine a storage that is not enough for all users .... Oct 24 08:25:36 Hello Oct 24 08:27:09 Hi Oct 24 08:31:32 I'm trying to write a recipe which would install a proprietary python wheel package to target. I think I can't use use pypi as it would try to fetch the source from the internets. Also setuptools seems to be useless as the .whl files do not contain a setup.py... Any ideas what would be the best way to do this? Oct 24 08:32:22 i have here a kernel cmdline issue on yocto 2.2; kernel cmd line is: rootfstype=ubifs ro; readonly, but the poky distro mounts the root alwas as rw; the cmdline ro is ignored; can anybody tell me where to look for this cmdline evaluation? thx Oct 24 08:33:02 If I wanted to install the package to host I could do it like this: pip install my-propriotary-stuff.whl Oct 24 08:33:10 cat /proc/cmdline is what you expect? Oct 24 08:34:18 test Oct 24 08:34:28 or even better, more /rpoc/cmdline Oct 24 08:34:40 So should I put similar command to the do_compile? Any ideas are welcome. Oct 24 08:35:17 \/proc\/cmdline is OK, but ro is ignored by yocto Oct 24 08:35:37 what about /etc/fstab ? Oct 24 08:35:42 Costin1: bitbake -s ? Oct 24 08:36:32 rootfstype=ubifs ro chk_data_crc ubi.mtd=0 root=ubi0:rootfs console=ttyS0,115200n8 mtdparts=davinci_nand.0:86m(Rootfs0),86m(Rootfs1),-(UBI) vt.global_cursor_default=0 internpower mem=95M Oct 24 08:37:56 if you don't have it already, create /etc/fstab with the required entry/entries, Klopilot Oct 24 08:38:49 fstab: //dev//root // auto defaults 1 1 Oct 24 08:39:09 replace defaults with ro Oct 24 08:39:09 thanks i will have a look at fstab Oct 24 08:39:17 thanks Oct 24 08:39:18 or whatever you need Oct 24 08:39:33 example: ro,user Oct 24 08:41:39 can you tell me in wich recipe fstab is controlled? Oct 24 08:41:47 no Oct 24 08:42:05 usually base-files but to be sure do something like: rpm -qf /etc/fstab Oct 24 08:50:12 thank you, cornel! Oct 24 09:36:43 otavio: Hi, I based a custom board BSP on linux wandboard_imx_4.1.15_1.1.0_ga branch, and found out that nxp released a 4.9 imx kernel. Do you know if there will be a linux-fslc-imx_4.9 ? Oct 24 09:49:02 If I have IMAGE_INSTALL_append = ".." and IMAGE_INSTALL_append_my-machine = "...", they will be both picked up right? Oct 24 09:50:37 sveinse: should be, yes Oct 24 09:50:50 (given that you are building for my-macine) Oct 24 09:51:02 of course. thanks. Oct 24 13:01:55 nayfe: there is already; i did not yet merged stable fixes on it Oct 24 13:05:11 otavio: nice! you know when it will be released? no pressure :) Oct 24 13:40:41 otavio: ha, i was looking not looking in right github, you mean this branch : https://github.com/Freescale/linux-fslc/tree/4.9-1.0.x-imx ? Oct 24 13:41:34 nayfe: well, honestly I am not too interested on 4.9 as I am using 4.1 or 4.13 for all our customers Oct 24 13:41:58 nayfe: so I'll likely use this to host community fixes Oct 24 13:42:13 nayfe: and avoid duplication of work Oct 24 13:42:54 otavio: nayfe: sorry to jump in, but are you guys talking about GCC versions...? Oct 24 13:42:56 otavio: ok, i suppose 4.1 for IPU/VPU support, and 4.13 for others ? Oct 24 13:43:14 Ah, kernels. Oct 24 13:43:22 Adam_: talking about linux version for freescale SOCs Oct 24 13:43:28 nayfe: ty Oct 24 13:44:21 nayfe: 4.13 has etnaviv and vpu support Oct 24 13:44:32 nayfe: ipu is not complete yet Oct 24 13:44:42 Adam_: Linux kernel Oct 24 13:45:15 nayfe: we added support to use mainline and we are moving many customers to it Oct 24 13:45:49 otavio: ty - sorry I was trying to build GCC 4.9 toolchain on Morty (by default it uses 6.9). Oct 24 13:46:06 Adam_: why? Oct 24 13:46:21 otavio: ok I see, I use a lot of gstreamer IPU/VPU stuff, so best option is probably to wait for 4.13 IPU support Oct 24 13:46:39 otavio: I have a bunch of custom apps that require an old version of GCC.... Oct 24 13:46:42 nayfe: not sure; gst works very well on mainline Oct 24 13:46:49 Adam_: time to upgrade? Oct 24 13:47:14 nayfe: we will move to 4.14 very soon Oct 24 13:47:32 otavio: I know. But it's hard to convince the app guys. I assume gcc 4.9 + morty is not a combo that's been tested? Oct 24 13:48:50 RP1, can this http://lists.openembedded.org/pipermail/openembedded-core/2017-October/143537.html make it into yocto pyro please? Oct 24 13:48:51 otavio: cpu time consuming is a big concern for us, so IPU helped in 4.1.15 ... Oct 24 13:49:43 otavio: you mean mainstream without IPU is better than 4.1 with IPU ? Oct 24 13:50:09 tront: just that patch? Oct 24 13:50:40 RP1, yes, just the one Oct 24 13:50:45 nayfe: try it Oct 24 13:50:49 nayfe: it is Oct 24 13:51:25 RP1: are you taking 2.5 patches already? Oct 24 13:51:46 otavio: not yet, no Oct 24 13:51:53 RP1: Ok Oct 24 13:54:00 otavio: ok thanks, i'll give it a try, but it means quite a lot of rework for our gstreamer plugins. Is there a place to look after work in progress ? meta-freescale mailinglist ? Oct 24 13:56:36 RP1, thanks a lot Oct 24 13:57:07 tront: thankfully its an easier patch to review but we can;'t normally do that.... Oct 24 13:57:48 what is the usual procedure? Oct 24 13:58:31 tront: Armin queues, tests, I merge the batch Oct 24 13:59:01 hello, I am using Kubuntu 16.04 and when I run bitbake recovery -g -u depexp I get: "Failed to connect to Mir: Failed to connect to server socket: No such file or directory" Oct 24 13:59:07 I see; noted Oct 24 13:59:12 tront: it takes time as the tests take time and tracking down the failures takes time. We need more people Oct 24 13:59:15 thanks again for making this exception now Oct 24 13:59:40 was never actually able to view the dependency graph using depexp, not even when I was using Ubuntu 14.04 stock Oct 24 14:00:55 there was also some kind of error on 14.04 too, though I do not remember what it was exactly Oct 24 14:01:56 Hey guys, Is there an easy way to call a function on the final image after its created? We would like to copy images to our build server once bitbake has created the final image. We've been looking at IMAGE_POSTPROCESS_COMMAND but the final image has not been created yet. Oct 24 14:01:57 can anyone give me a good example on how to properly view a dependency graph for a recipe on Ubuntu 16.04? Oct 24 14:04:26 Hi All ! I have problem with "devtool modify". I remember that i tried it last week, and now when i try to bitbake, he is looking to another conf folder inside the dir that i create for devtool modify. how to reverse it ? Thanks by advance Oct 24 14:04:46 Monty_: I suppose you could create a recipe that pulls in your final image recipe as a dependency and does whatever more tasks you need, though that sounds ugly Oct 24 14:09:30 No idea? Oct 24 14:11:40 Monty_: then again (after re-reading your sentence), should you not be building the images on the build server in the first place? Oct 24 14:12:27 and let your Continuos Integration system handle whatever build artifacts are produced Oct 24 14:16:28 I get the oslution Oct 24 14:16:40 the workspace made by devtool is added in bblayers.conf Oct 24 14:16:43 just need to remove there Oct 24 14:16:55 ( cleaning tmp/ for nothing :s ) Oct 24 14:28:23 The "build server" is more of a cloud backup - I've confused the question by giving it a bad name. It doesn't make sense to build the images there. There's no easy/clean way then? Oct 24 14:30:09 Monty_: I suppose you could just write a bash script that runs bitbake, waits for it to finish successfully and does the upload? Oct 24 14:33:37 Yer. That's what we're probably going to do. Just doesn't feel like we're using yocto/bitbake correctly Oct 24 14:35:31 Monty_: I am not that experienced in Yocto myself, but as far as I know the two main CI solutions are either Yocto Autobuilder or Jenkins... I assume the proper way to handle artifact uploads to somewhere would be via the mechanisms those provide Oct 24 14:36:19 though I am really not familiar with either... perhaps someone could elaborate on the matter? Oct 24 14:39:44 if you want to run something after image construction you can do do_image[postfuncs] += "my_function" Oct 24 14:46:10 joshuagl: really useful to know, though "postfuncs" is mentioned only once in the Mega-manual and in a slightly different context Oct 24 14:48:19 eduardas_m: https://www.yoctoproject.org/docs/latest/bitbake-user-manual/bitbake-user-manual.html#variable-flags is probably the best bit of docs for it Oct 24 14:51:58 rburton: thank you... also, since you are here, could you please tell me how to successfully run depexp on Ubuntu 16.04? I get: "Failed to connect to Mir: Failed to connect to server socket: No such file or directory" Oct 24 14:52:18 lol mir Oct 24 14:52:28 I simply run bitbake recovery -g -u depexp Oct 24 14:52:45 presumably you need to know what environment variables mir needs to work, and then ensure they don't get wiped like $DISPLAY does Oct 24 14:52:51 or stop using mir Oct 24 14:53:31 rburton: yeah, I know... something is off in Kubuntu 16.04, but the funny thing is that I'm running regular Plasma 5 on X11 Oct 24 14:53:40 not Mir Oct 24 14:54:48 rburton: $DISPLAY gets removed when sourcing the environment for bitbake? Oct 24 14:54:57 I did not know that Oct 24 14:54:58 DISPLAY doesn't no Oct 24 14:55:05 well, it does, but we add it back for depexp Oct 24 14:55:17 but presumably mir doesn't use DISPLAY as that's an X variable Oct 24 14:56:04 trying it out now. However the docs say do_image_* starts the image generation process. do_image_complete completes the image generation process. do_image_complete task performs post-processing on the image through the IMAGE_POSTPROCESS_COMMAND. Oct 24 14:57:14 image_post_process_command didn't work as the image hadn't been created yet Oct 24 15:05:33 http://www.yoctoproject.org/docs/current/ref-manual/ref-manual.html#image-generation-dev-environment - there doesn't appear to be any tasks/commands that happen after the image has been generated. IMAGE_POSTPROCESS_COMMAND acts on the final image output files Oct 24 15:05:53 rburton: simply running export DISPLAY=:0 after sourcing environment for Yocto does not help in running depexp Oct 24 15:06:26 here is the full error: https://pastebin.com/PNYPxYij Oct 24 15:08:48 there's no depexp in your bitbake, you should use taskexp instead Oct 24 15:08:55 I think the mir stuff may be a red herring Oct 24 15:09:06 "FATAL: Unable to import extension module "depexp" from bb.ui. Valid extension modules: knotty, ncurses or taskexp" Oct 24 15:09:22 depexp was dropped and replaced with taskexp a release or so ago Oct 24 15:09:51 joshuagl: thank you, I was not aware of that... I am using Yocto Pyro Oct 24 15:10:46 the very release in which it happened :-) Oct 24 15:10:46 http://www.yoctoproject.org/docs/latest/ref-manual/ref-manual.html#migration-2.3-bitbake-changes Oct 24 15:16:31 joshuagl: yes, taskexp works for me :) Oct 24 15:17:11 \o/ Oct 24 15:18:10 I had problem adding strace in IMX6. Oct 24 15:18:16 I've added in the image recipe strace. Oct 24 15:18:27 I've verified that in bitbake -e i have strace in image_install Oct 24 15:18:37 but when i boot my card.. no strace in /usr/bin Oct 24 15:19:59 ChrysD: what does your rootfs manifest say? Oct 24 15:20:17 does it say the strace package is included? Oct 24 15:20:28 can you remember me where i can find this manifest? Oct 24 15:20:57 ChrysD: in tmp/deploy/images Oct 24 15:21:42 strace cortexa9hf-vfp-neon 4.9-r0 Oct 24 15:21:45 with the .manifest in filename Oct 24 15:23:02 So it means that it should be installed Oct 24 15:23:04 ? Oct 24 15:23:21 ChrysD: I would really assume it to be installed... are you sure you booted the same image? Oct 24 15:23:33 yeah sure Oct 24 15:23:45 i've remove my tmp/ sstate and everything Oct 24 15:23:48 so it's the first image i made Oct 24 15:26:36 ChrysD: I add strace myself on i.MX6 via IMAGE_INSTALL += " strace", so I can not think of why it might not work... the package management tools on the board (rpm or opkg) also say the package is missing? Oct 24 15:27:09 only things it says is command not found Oct 24 15:27:17 and it's not in /usr/bin Oct 24 15:27:21 Same with ltrace... Oct 24 15:28:08 ChrysD: I also append tools-debug to IMAGE_FEATURES Oct 24 15:28:10 I added devregs for eexample, and the command works Oct 24 15:31:17 ChrysD: you remember you have to not forget the space before the package name in IMAGE_INSTALL += " strace" or else it won't be parsed? I am running out of ideas Oct 24 15:31:33 i do IMAGE_INSTALL += " Oct 24 15:31:36 strace \ Oct 24 15:31:40 ltrace \ Oct 24 15:31:44 blblabla\ Oct 24 15:31:48 and it works Oct 24 15:31:54 i mean for somes of them Oct 24 15:31:59 but not all of the package Oct 24 15:32:07 and i have a space between the package name and " \ " Oct 24 15:32:16 ChrysD: does only strace not work? Oct 24 15:32:30 eduardas_m : i have 5 packages, 2 of them dont work Oct 24 15:32:34 ltrace and strace Oct 24 15:32:38 both are in the manifest Oct 24 15:32:39 if it is the first one in the list, it must have a space before the name Oct 24 15:32:45 both are in IMAGE_INSTALL with bitbake -e Oct 24 15:33:02 ChrysD: yes, that confuses me also Oct 24 15:34:00 IMAGE_INSTALL +=" \ imx-test \ i2c-tools \ devregs \ strace \ ltrace \ " Oct 24 15:34:07 ChrysD: you should be able to check if the rootfs has the strace binary, just inspect the tar.gz if you have that built as an image fstype. Oct 24 15:34:25 and strace ltrace doesn't work but they are the two last.... Oct 24 15:34:40 ChrysD: this is my section with packages if this helps: https://pastebin.com/i5f46cVk Oct 24 15:34:40 ChrysD: since having the package installed doesn't mean you have to have the binary ;) Oct 24 15:35:14 ok Oct 24 15:35:47 eduardas_m : that's mine : https://pastebin.com/qzBGRfmm Oct 24 15:35:50 ChrysD: you can also explore the rootfs in a non-archived form in /tmp Oct 24 15:36:05 making a tar.gz is not necessary Oct 24 15:36:34 in tmp/work I mean Oct 24 15:36:47 sure, if you have the build in tmp available, but will not be there if you blew it away and rebuilt from sstate Oct 24 15:38:15 How can I check? Oct 24 15:39:06 I add tools-debug and I have strace Oct 24 15:39:13 Q. Is it possible to respond to an old mailing list thread of OE? Oct 24 15:39:23 ChrysD: for example, my rootfs can be found in /home/eduardas/BUILD_3/pyro_builds/build_rc/tmp/work/var_som_mx6-fslc-linux-gnueabi/fod-final-image/1.0-r0/rootfs/ Oct 24 15:39:46 eduardas_m : yeah i know that Oct 24 15:40:06 me there : /Yocto/build/tmp/work/imx6qhera-poky-linux-gnueabi/ov5640-image-debug/1.0-r0/rootfs Oct 24 15:40:32 and now i have in /usr/bin strace Oct 24 15:40:38 after adding tools debug as image features Oct 24 15:40:57 ChrysD: ltrace too? Oct 24 15:41:02 yeah Oct 24 15:41:12 and if i don't put debug tools Oct 24 15:41:16 i have no ltrace and ftrace Oct 24 15:41:19 strace* Oct 24 15:41:56 Thanks nrossi and eduardas for your support, but still don't understand why can't be installed manually. Oct 24 15:42:10 ChrysD: kind of weird, since I would have expected to be able to add those one by one... had debug-tools enabled for other reasons Oct 24 15:42:15 Eduardas, as you are working on IMX6, have you got ftrace working on your board? Oct 24 15:43:41 i have kernel panic now Oct 24 15:43:43 ahahah Oct 24 15:43:46 ChrysD: to be completely honest, I am a Linux noob... I have no experience with any kind of tracing tools... I just add that stuff to use "some time in the future".. I know, disappointing Oct 24 15:44:04 for "when I need it" Oct 24 15:44:14 The same, but now it's time to me to use them Oct 24 15:44:15 ahahah Oct 24 15:44:49 ChrysD: ok, good luck... got to go Oct 24 15:45:35 nrossi : i have my rootfs on tmp/ different from the one on my image.... Oct 24 15:46:56 ChrysD: different to the rootfs tarball or the image you have booted? Oct 24 15:47:11 the image i have booted Oct 24 15:47:32 ChrysD: how is it different exactly? Oct 24 15:47:52 nrossi : in /usr/bin/ Oct 24 15:48:05 nrossi : things are missing Oct 24 15:48:30 ChrysD: what sort of things? binaries, symlinks? Oct 24 15:48:39 nrossi : binaries like strace Oct 24 15:49:09 ChrysD: did you not mention above you got it working? Oct 24 15:49:26 no i have it on my /usr/bin in tmp/ Oct 24 15:49:28 not on my board Oct 24 15:49:36 ah ok, i miss understood. Oct 24 15:49:53 ChrysD: then your issue is image/disk related Oct 24 15:50:06 I have it there /build/tmp/work/imx6qhera-poky-linux-gnueabi/ov5640-image-debug/1.0-r0/rootfs Oct 24 15:50:17 ov5640-image-debug is the image name that i put after bitbake Oct 24 15:50:22 ChrysD: how are you writing the image to your target? sd card? Oct 24 15:50:28 sdcard yeah Oct 24 15:50:58 ChrysD: are you using a sdcard image that is generated by the build or manually partitioning/etc Oct 24 15:50:59 ? Oct 24 15:51:06 by the build Oct 24 15:51:26 ChrysD: when you write the image to disk do you just 'dd' it? Oct 24 15:51:48 nrossi : i use win32 Oct 24 15:51:57 ChrysD: ouch... Oct 24 15:51:58 nrossi win32 disk isntaller* Oct 24 15:52:28 I have a linux server which compile bitbake and everything. But i put my sdcard into my window machine. Oct 24 15:52:34 So i do a Winscp of the .sdcard Oct 24 15:52:39 then a win32 Oct 24 15:53:04 I could do an ssh dd if necessary i guess Oct 24 15:53:25 But it's the first time that i put things in IMAGE_INSTALL but don't get it in the final image Oct 24 15:53:54 ChrysD: your issue might be that the new disk image is not being written to the disk properly? Oct 24 15:54:13 mhhh Oct 24 15:54:42 ChrysD: have you double checked the image file being written is the same as the on from the build? (md5sum) Oct 24 15:54:57 no, i triple checked Oct 24 15:55:07 :s Oct 24 15:55:18 If i add other packages Oct 24 15:55:28 like gstreamer, or qt or whatever, no problem Oct 24 15:55:28 ChrysD: i see, have you tried formatting the disk, and then using the win32 imager? Oct 24 15:55:39 can i somehow set the sstate_dir via environment variables instead of local.conf? Oct 24 15:55:59 i add other debugging tools , no problem. I get only that with strace and lstrace Oct 24 15:56:11 ChrysD: oh, so its just strace playing up, that is odd Oct 24 15:56:31 nrossi : i'm in a 3.14.28 core image minimal Oct 24 15:57:02 nrossi : and the ov5640-image-debug is just an image which add debugging tools, or other package + kernel module Oct 24 15:57:42 ChrysD: did you check if strace was in the rootfs tarball? Oct 24 15:57:50 ChrysD: or just in tmp/ Oct 24 15:57:58 just in tmp/ Oct 24 15:58:38 ChrysD: if you build the tarball, check in it, if its there then its definitely image related problem Oct 24 15:59:07 peacememories: check BB_ENV_EXTRAWHITE Oct 24 15:59:27 i'm checking nrossi Oct 24 16:01:23 nrossi : it's in the ov5640-image-debug-imx6qhera-Blablablabla.rootfs.tar.bz2 Oct 24 16:01:58 ChrysD: ok, lets check if in the sdcard image. This is a little tricky, i assume you have root on your build machine? Oct 24 16:02:09 nrossi : i have root on my build machine Oct 24 16:02:37 ChrysD: get losetup installed if not already Oct 24 16:03:22 nrossi i have it Oct 24 16:06:14 ChrysD: losetup -P /dev/loop0 <.sdcard file> Oct 24 16:06:50 When i do that, it send me to the help page of losetup Oct 24 16:07:13 you mean -p or -P ? Oct 24 16:07:14 oh drop the -P, it might be in newer versions only Oct 24 16:07:34 ok Oct 24 16:07:38 it works Oct 24 16:07:48 then to replace the -P, run partprobe /dev/loop0 Oct 24 16:08:11 you should now have some /dev/loop0pX nodes Oct 24 16:08:19 mount the one that is the rootfs Oct 24 16:08:24 and explore for strace Oct 24 16:09:14 [17:51] nrossi : i use win32 [17:51] ChrysD: ouch... Oct 24 16:09:59 that's how i can sum up Oct 24 16:10:02 ChrysD: ? Oct 24 16:10:06 it's there Oct 24 16:10:19 on the loopbacked image? Oct 24 16:10:34 yeah Oct 24 16:10:52 that's why i copied your " ouch... " ahah Oct 24 16:10:56 ChrysD: ok so the solution here, is to take your windows machine and send it back to microsoft :P Oct 24 16:11:11 or maybe to throught it through the window Oct 24 16:11:20 that's why is a windows machine no? Oct 24 16:11:29 that would hurt the window, don't make your windows feel the pain of windows :P Oct 24 16:11:31 throw it* Oct 24 16:11:52 We use lot of windows tools to, the best solution was to have two computer :s Oct 24 16:12:24 window's dd tools have already been iffy, last time i was on windows i gave up and got a linux machine for sd card imaging Oct 24 16:12:33 s/already/always/ Oct 24 16:13:28 ChrysD: btw, to proper clean up that /dev/loop0 mount, losetup -d Oct 24 16:14:24 ChrysD: also if you can always use your board to do you sd card imaging :) Oct 24 16:14:58 ChrysD: have one card with known tools working that boots to ramfs, or use another sdcard reader on usb or something Oct 24 16:17:56 nrossi : thank you for the tips of the losetup Oct 24 16:18:30 nrossi : what you mean by " that boots to ramfs" ? Oct 24 16:19:47 ChrysD: boot it with the rootfs in ram, ramfs or ramdisk. It's quite easy to do with u-boot, get a .cpio and load it to a memory address and pass it to bootm's second arg (bootm (or bootz if your using zImage) Oct 24 16:20:47 sorry needs to be a uboot wrapped cpio, so ".cpio.gz.u-boot" is probably what you are after Oct 24 16:21:55 I think he is already booted in th e ram Oct 24 16:22:00 as i can change the kernel Oct 24 16:22:11 while the board is booted Oct 24 16:22:29 i already tried to do an scp of the kernel Oct 24 16:22:44 oh maybe you mean rootfs is not on the ram Oct 24 16:22:50 in the* Oct 24 16:23:14 i said nothing. Oct 24 16:23:23 ChrysD: you are using a rootfs from the sdcard at the moment yes? e.g. root=/dev/mmcblk0p2 Oct 24 16:23:31 yeah Oct 24 16:24:05 ChrysD: that would prevent you from removing the sd card, but if you put the rootfs in ram, then once booted the sd is not needed, you can remove and replace with the sdcard you want to load the image on Oct 24 16:29:14 nrossi : you mean if i want to make rootfs changes Oct 24 16:29:54 ChrysD: just an idea if you cant attach you sd card to your build machine Oct 24 16:30:32 nrossi : yeah yeah i have understand. So i put the rootfs in the ram so that i can change it in my sdcard. Oct 24 16:30:53 nrossi : the best would be to have the rootfs in nfs Oct 24 16:31:03 ChrysD: yep exactly, nfs rootfs works too Oct 24 16:31:22 nrossi : I need to work into my linux workflow ... Oct 24 16:32:11 nrossi : i copied the /git folrder of the linux folder for kernel developpement and i took the run.do_compile / run .do_configure and run.do_compile_kernelmodules ( i modified path ) so that i can work into kernel quickly. Oct 24 16:33:24 ChrysD: have a look at EXTERNALSRC, saves having to work out of the oe build Oct 24 16:40:18 nrossi : ok thanks Oct 24 16:42:36 nrossi : i don't understand how win32disk installer coudl have a problem Oct 24 16:42:45 he jus tneed to write the binaryahah Oct 24 16:44:37 ChrysD: depends on the sd reader drivers, i've seen mention of issues before Oct 24 16:46:51 nrossi : i have to go, thanks for the help , i've appreciate, also about learning losetup =) If you have time, do you mind explain me what's your workflow ? Oct 24 16:46:58 nrossi : i mean next time :D Oct 24 16:47:16 ChrysD: sure, ping me if im on Oct 24 16:47:27 nrossi : thanks a lot, bye ! Oct 24 17:56:51 nayfe: I think the changes are going to be small; check the mailing list or contact me in private Oct 24 17:57:03 nayfe: we can assist you if need Oct 24 19:18:07 hi Oct 24 19:19:39 what is the best way to change a kernel recipe? Oct 24 21:01:56 * armpit arrg.. builds driving me crazy Oct 24 21:09:58 armpit: Oct 24 21:10:05 orly? :-P Oct 24 23:58:28 ls Oct 24 23:58:30 ls Oct 25 00:01:13 maybe if i add "-l" it will work better and offset my dumbasss factor of typing it in the wrong window... **** ENDING LOGGING AT Wed Oct 25 03:00:02 2017