Re: [yocto] Working on Dependent Recipes in eSDK
> ChenQi wrote on 10/22/2018 07:04:32 PM: > > Hi Aaron, > > The main pain for now is that `devtool sdk-install ' does > not work as expected, right? > I just checked the codes of sdk-install, and found that it does not fit > the current RSS design. > I'd like to check with you if this is the major concern from your team? > Do you have some other concerns? > I'll open enhancements/bugs on bugzilla and hopefully will work on them > in the 2.7 development cycle. > > Best Regards, > Chen Qi > > ... snip ... Well, I'm not quite sure how "devtool sdk-install " relates after "devtool modify " has been called. My developers use devtool to install dependent recipes into the eSDK when they start working on something. Often they find that they need to make a change in the library code that a dependent recipe provides. In effect, they are developing a patch for the dependent recipe. So they use devtool to say they're going to modify that dependent recipe, and they go and make the changes they want. Then what? How do they get those changes, that patch, available to the eSDK? Any changes made to the library code provided by the dependent recipe are not available to the eSDK, except through devtool commands. In effect there are two different worlds at play, the eSDK sysroots versus the recipe sysroots. The devtool commands seem to make use of the recipe sysroots, but in my mind that defeats the purpose of an eSDK. So I made a plug-in for devtool were I can force the eSDK sysroots to be updated. I'm not sure how that works, but it seems to propagate the workspace (recipe sysroot land) to the eSDK, so changes that developers make are now available for use. I want my developers to use the toolchain (cmake, make, gcc, etc) provided by the eSDK and by things installed into it (devtool sdk-install), rather than having to use devtool as their iterative build command. Ultimately, I would like to be able to skip my devtool plug-in, as it requires the use of "devtool build " and "devtool sdk-sysroot" each time a dependent recipe is modified. But I am not sure how that would work, other than the devtool build command should be updating the eSDK sysroots itself. -- ___ yocto mailing list yocto@yoctoproject.org https://lists.yoctoproject.org/listinfo/yocto
Re: [yocto] Working on Dependent Recipes in eSDK
> Paul Eggleton wrote on 10/18/2018 12:30:07 PM: > > I think we would *like* this to be possible, from what I can tell I think the > issue is that the copies of the library in the sysroots of other dependent > recipes aren't being updated. The eSDK (and devtool) were originally written > when there weren't recipe-specific sysroots, and this stuff was a bit simpler. > > Moving back to the full build system won't automatically fix this, though it > does make workarounds possible - you'd need to rebuild the dependent recipes > to get their sysroots updated. > > Cheers, > Paul > ... snip ... I put together a devtool plugin to almost do what I want: > from devtool import exec_build_env_command, DevtoolError > > def sdk_sysroot(args, config, basepath, workspace): > """Entry point for the devtool sdk-sysroot command""" > > try: > exec_build_env_command(config.init_path, basepath, 'bitbake build-sysroots', watch=True) > except bb.process.ExecutionError as e: > raise DevtoolError('Failed to bitbake build-sysroots:\n%s' % (str(e))) > > def register_commands(subparsers, context): > """Register devtool subcommands from the sdk plugin""" > if context.fixed_setup: > parser_sdk_sysroot = subparsers.add_parser( > 'sdk-sysroot', > help='Update the SDK sysroot', > description='Updates the SDK sysroot so that changes made to recipes in the workspace will be available,', > group='sdk') > parser_sdk_sysroot.set_defaults(func=sdk_sysroot) This allows me to "devtool modify ", edit the code, "devtool build ", "devtool sdk-sysroot", and use the new code from the dependent recipe immediately. This makes me wonder if "bitbake build-sysroots" should not be a part of "devtool build", like it is for "devtool sdk-install" (where I stole it from). -- ___ yocto mailing list yocto@yoctoproject.org https://lists.yoctoproject.org/listinfo/yocto
Re: [yocto] Working on Dependent Recipes in eSDK
> Alexander Kanavin wrote on 10/18/2018 11:30:17 AM: > > I don’t think esdk is designed for this kind of ‘full stack’ > development. You can go back to working directly with yocto layers, > and skip the esdk altogether. Devtool will work exactly same (minus > the sdk specific commands). > > Alex I'm trying to get the developers away from using bitbake and devtool to build their code, as those tools have overhead and don't easily allow the developers to iterate quickly or allow them to run other make targets for unit testing, etc. I really want the developers at the "make" level for the majority of their time. When it comes time for recipes and image creation, then they'll use devtool. > > On 18 Oct 2018, at 19.37, aaron_wri...@selinc.com wrote: > > > > My team recently started using the eSDK for day-to-day development. > > Previously, they were using the bitbake command to build, which works, but > > has overhead. Yesterday three different people came up to me a question > > about workflow with the eSDK, and they all had to do with dependent > > recipes in the eSDK. > > > > For example, a developer would sit down to make something new that relies > > on a library provided by a recipe in the eSDK. So they "devtool > > sdk-install needed-library", and get to work. While working, they > > discover, I need to add something to this library I depend on so that it > > does what I want in the way I want. So they "devtool modify > > needed-library" and change the source of that library to do what they > > want, and build that library using its build system (CMake or Autotools). > > Then they return to working on the new something and find that none of > > their changes to the library they depend on are available. > > > > At this point, they are lost. They just start trying things to make it > > work, like "devtool build needed-library", which successfully completes, > > but doesn't help. Some people also try to use the build system of the > > library they modified to install it into the machine sysroot of the eSDK > > with a command like, "make install > > DESTDIR=/tmp/sysroots/", and that actually does help. > > They can now use the changes they made to the library in the new something > > they are working on. However, when they use the "devtool sdk-install" > > command again, their changes to the needed library that they installed > > into the machine sysroot are reverted. > > > > So--what is the workflow they are supposed to use? > > > > Pushing the changes to the library upstream, updating the recipe, and > > waiting for a new eSDK is not practical given that the developers may have > > to iterate on the changes to the library quite a bit before they get it > > just right for their new something. > > > > Is there a way to update the eSDK sysroots with recipes that are in the > > workspace? Could "devtool build" also update the eSDK sysroots? Or is > > there a different way to get this done? > > -- > > ___ > > yocto mailing list > > yocto@yoctoproject.org > > https://urldefense.proofpoint.com/v2/url? > u=https-3A__lists.yoctoproject.org_listinfo_yocto=DwIFaQ=- > _uRSsrpJskZgEkGwdW- > sXvhn_FXVaEGsm0EI46qilk=GBJubmJUwwrr1B4TJFRqESMrRsDipvY66vPwm4- > gtNQ=- > C-70rjnGksaogzIkgE5MdoY3IfA3x62MZVOwbZmw4c=T5PxZ3bNB429M76sWwBqgrtMsMhB- > kz1BXkdY8hOIRY= -- ___ yocto mailing list yocto@yoctoproject.org https://lists.yoctoproject.org/listinfo/yocto
[yocto] Working on Dependent Recipes in eSDK
My team recently started using the eSDK for day-to-day development. Previously, they were using the bitbake command to build, which works, but has overhead. Yesterday three different people came up to me a question about workflow with the eSDK, and they all had to do with dependent recipes in the eSDK. For example, a developer would sit down to make something new that relies on a library provided by a recipe in the eSDK. So they "devtool sdk-install needed-library", and get to work. While working, they discover, I need to add something to this library I depend on so that it does what I want in the way I want. So they "devtool modify needed-library" and change the source of that library to do what they want, and build that library using its build system (CMake or Autotools). Then they return to working on the new something and find that none of their changes to the library they depend on are available. At this point, they are lost. They just start trying things to make it work, like "devtool build needed-library", which successfully completes, but doesn't help. Some people also try to use the build system of the library they modified to install it into the machine sysroot of the eSDK with a command like, "make install DESTDIR=/tmp/sysroots/", and that actually does help. They can now use the changes they made to the library in the new something they are working on. However, when they use the "devtool sdk-install" command again, their changes to the needed library that they installed into the machine sysroot are reverted. So--what is the workflow they are supposed to use? Pushing the changes to the library upstream, updating the recipe, and waiting for a new eSDK is not practical given that the developers may have to iterate on the changes to the library quite a bit before they get it just right for their new something. Is there a way to update the eSDK sysroots with recipes that are in the workspace? Could "devtool build" also update the eSDK sysroots? Or is there a different way to get this done? -- ___ yocto mailing list yocto@yoctoproject.org https://lists.yoctoproject.org/listinfo/yocto
[yocto] How to patch local source files
I've been using "devtool modify/update-recipe" in order to generate patches in bbappends in my layer. However, I've run into a problem that seems like it should be easy, but it isn't. I want to patch a local source file. As in, a file referenced in SRC_URI that isn't the source code but more like a configuration file or init script. Using the devtool method I get a copy of the file in my layer, but that seems like that would cause maintenance problems down the road when upstream changes the file as I won't automatically pick up the changes. This seems to be the intended behavior though, as this patch[1] says, "We don't want to create patches against the local source files but rather update them directly. Thus, 'oe-local-file' directory is ignored in patch generation when doing update-recipe..." This seems completely opposite of what I want. Why would I want a copy of the file if I changed one line, causing maintenance issues in the future? Am I understanding this right? Is there a workflow that allows for patching local source files? [1] https://urldefense.proofpoint.com/v2/url?u=https-3A__patchwork.openembedded.org_patch_104075_=DwIBAg=zVFQZQ67ypsA9mYKSCqWmQHiVkCCaN-Gb60_N6TVnLk=IZ3mUDA8X9CZR4hNctKngDDYs_BhK3qVNLG-nWGbW2E=8pcye74JGQLQcBV99bh9bpRb7bX3WqHFUiIq157bDY4=XLqetQZfaf9jTNco2uT4jCN6FbQ1tHD9F_ZJQvBouJI= -- ___ yocto mailing list yocto@yoctoproject.org https://lists.yoctoproject.org/listinfo/yocto
[yocto] [multilib] 64-bit and 32-bit package install the same executable
I've noticed that for my custom image both the 32-bit and the 64-bit versions of the same package are getting installed. For example, it has both shadow-base and lib32-shadow-base installed. When I check the /bin/su.shadow executable I find it is 32-bit. Is it expected that the lib32 version of the package will "win" when both packages install the same /bin/su.shadow file? Is this something that is always true, so that I can always assume, when two packages install the same executable, that it will end up being 32-bit? Why does the packaging system, PACKAGE_CLASSES = "package_ipk", not give an error when both packages install the same file? -- ___ yocto mailing list yocto@yoctoproject.org https://lists.yoctoproject.org/listinfo/yocto
Re: [yocto] eSDK and Multilib
> Don't install both into the SDK? Interesting, I assumed the populate_sdk_ext was building both lib32 and non-lib32 versions of the recipes on its own. I'll have to double check my image definition for recipes that are having both variants pulled in. Thanks for the idea. > If you must, oe_multlib_header might help. > > Ross > >> I'm running pyro, and I'm struggling to create an eSDK for multilib >> custom image. I keep running into FileExistsError errors during the >> do_sdk_depends task. >> >> It seems as though the lib32 and non-lib32 versions of a recipe are >> both creating a -dev package that tries to install the same headers. >> >> Am I missing an easy solution to this? -- ___ yocto mailing list yocto@yoctoproject.org https://lists.yoctoproject.org/listinfo/yocto
[yocto] eSDK and Multilib
I'm running pyro, and I'm struggling to create an eSDK for multilib custom image. I keep running into FileExistsError errors during the do_sdk_depends task. It seems as though the lib32 and non-lib32 versions of a recipe are both creating a -dev package that tries to install the same headers. Am I missing an easy solution to this?-- ___ yocto mailing list yocto@yoctoproject.org https://lists.yoctoproject.org/listinfo/yocto
[yocto] How to use iconv natively in do_compile?
I'm on pyro, and I have a do_compile step that converts some files to UTF8. I use iconv to do this. According to `oe-pkgdata-util find-path /usr/bin/iconv`, iconv is provided by the glibc-utils package. So I tried adding glibc-native or glibc to my DEPENDS, but that didn't work. I also tried virtual/libiconv-native without any luck. virtual/libiconv is listed in ASSUME_PROVIDED but iconv isn't in HOSTTOOLS. I'm not 100% sure how those variables are supposed to work. What do I put in my DEPENDS so that iconv is natively available during do_compile? -- ___ yocto mailing list yocto@yoctoproject.org https://lists.yoctoproject.org/listinfo/yocto
Re: [yocto] BKM for Importing Python Modules
"Burton, Ross"wrote on 11/30/2017 02:47:43 AM: > Python functions in classes are executed by the Python that is > executing bitbake, which doesn't have access to the modules in the > native sysroot (and won't, as any compiled modules in there are > linked against the native python, not the host python). So adding to sys.path works, but only accidentially? sys.path.append( d.getVar("STAGING_DIR_NATIVE", True) + d.getVar("PYTHON_SITEPACKAGES_DIR", True)) > You can solve this by either mandating python-cryptography on the > host, or writing a standalone script that is executed using > nativepython (which is the python built by python-native). Not a terrible idea. I'll just have to pull lots of stuff out of d and pass them as arguments. Thanks. > Ross > > On 30 November 2017 at 00:09, wrote: > I have a python function that I've added to PACKAGEFUNCS in a > bbclass, and I want to use python-cryptography in that function. > However, sys.path doesn't seem setup to allow this import. I'm > curious to know the best known method for using python modules that > are outside of /lib directory. Should I just append > STAGING_LIBDIR_NATIVE to sys.path? > > python do_stuff() { > import cryptography # <-- this fails > > # Other stuff... > } > > python () { > if (not bb.data.inherits_class("native", d) and > not bb.data.inherits_class("cross", d)): > > d.appendVar("PACKAGEFUNCS", " do_stuff") > d.appendVarFlag( > "do_package", > "depends", > " python-cryptography-native:do_populate_sysroot") > } > -- -- ___ yocto mailing list yocto@yoctoproject.org https://lists.yoctoproject.org/listinfo/yocto
[yocto] BKM for Importing Python Modules
I have a python function that I've added to PACKAGEFUNCS in a bbclass, and I want to use python-cryptography in that function. However, sys.path doesn't seem setup to allow this import. I'm curious to know the best known method for using python modules that are outside of /lib directory. Should I just append STAGING_LIBDIR_NATIVE to sys.path? python do_stuff() { import cryptography # <-- this fails # Other stuff... } python () { if (not bb.data.inherits_class("native", d) and not bb.data.inherits_class("cross", d)): d.appendVar("PACKAGEFUNCS", " do_stuff") d.appendVarFlag( "do_package", "depends", " python-cryptography-native:do_populate_sysroot") }-- ___ yocto mailing list yocto@yoctoproject.org https://lists.yoctoproject.org/listinfo/yocto
[yocto] devtool sdk-install
I'm confused as to how the devtool sdk-install command is supposed to work. Take for example: devtool sdk-install cxxtest This installs all the /usr/lib and /usr/include stuff into the sysroot, but doesn't install the /usr/bin stuff required to actually use cxxtest. So should I install nativesdk-cxxtest instead? Still doesn't work, but I was just curious about the nativesdk- prefix. cxxtest-native doesn't work either. Does the cxxtest recipe need a SYSROOT_DIRS_append = " ${bindir}" in order to get the /usr/bin files installed in the eSDK when someone installs it? PS - often devtool sdk-install will act like it is working and then print an error at the end: NOTE: Tasks Summary: Attempted 105 tasks of which 71 didn't need to be rerun and all succeeded. ERROR: Failed to install nativesdk-cxxtest - unavailable PPS - a google search for devtool sdk-install returns nothing, so hopefully the mailing list can help.-- ___ yocto mailing list yocto@yoctoproject.org https://lists.yoctoproject.org/listinfo/yocto
Re: [yocto] How to use same user in two recipes?
> Hi > > I want to use the same user in two different recipe. > I create the user in recipe A.bb and it works fine. > Now I want to use the same user in B.bb so I add DEPENDS = "A" > thinking that sysroot will be > populated with A.bb users. > > But when I bitbake the recipe, user is not found. > I conclude that I have to use USERADD_* in each recipe, right? > > What happen if I use different USERADD_PARAM_${PN} in each recipe? > > I also try to use EXTRA_USERS_PARAMS but without success? > > Is there a way to populate sysroot with users? > > Thanks > > -- > Fabien > -- When I have this issue I just make a separate recipe that creates the user, and have all the other recipes depend on it. That way the user is always created with the same parameters, without possibility of getting out of sync.-- ___ yocto mailing list yocto@yoctoproject.org https://lists.yoctoproject.org/listinfo/yocto
[yocto] populate_sdk_ext => "Cannot install package nativesdk-git-perltools"
I'm trying to create an extended SDK for a multilib environment and I run into this error during the buildtools-tarball-1.0-r0 do_populate_sdk task: ...snip lots of configuring lines... Configuring nativesdk-buildtools-perl-dummy. Collected errors: * check_conflicts_for: The following packages conflict with nativesdk-perl: * check_conflicts_for: nativesdk-buildtools-perl-dummy * * opkg_solver_install: Cannot install package nativesdk-git-perltools. ...snip lots of configuring lines... I'm using morty: commit 493b1c9aeaeb60440085bb3692e9362a87553bce Author: Richard PurdieDate: Tue Oct 10 11:01:38 2017 +0100 I'm not sure what to look for to solve this issue. Any ideas?-- ___ yocto mailing list yocto@yoctoproject.org https://lists.yoctoproject.org/listinfo/yocto
Re: [yocto] local fetcher and devtool don't work well together
Paul Eggletonwrote on 10/04/2017 06:47:50 PM: > Hi Aaron, > > On Thursday, 5 October 2017 11:54:32 AM NZDT aaron_wri...@selinc.com wrote: > > I have a recipe with: > > > > SRC_URI = "file://a/b/c/d;subdir=src" > > S = "${WORKDIR}/src/a/b/c/d" > > > > First off, ${S} must be set to the full path because the basepath uri > > parameter[1] doesn't work. So this doesn't work: > > > > SRC_URI = "file://a/b/c/d;subdir=src;basepath=a/b/c/d" > > S = "${WORKDIR}/src" > > > > I wish it would, but that's ok. I can deal with the long ${S} path as it > > compiles and works fine with bitbake. > > My question would be what are you trying to achieve here? Should you be using > the externalsrc class perhaps? I've waffled between externalsrc and the local fetcher, but the local fetcher seems to behave the best. What I'm trying to do is a little odd, but we are trying to use yocto with the recipes, meta layer, and source code in the same repository. This makes the continuous integration server much simpler, and developers can see the results of their commit/push in the context of the whole product being put together with yocto. This might not be the best idea, and yocto certainly seems to be based on the fact that the source is remote, but things are generally working well for us; at least continuous integration server wise. Right now I'm trying to figure out how to get developers developing. > https://urldefense.proofpoint.com/v2/url? > u=http-3A__www.yoctoproject.org_docs_current_dev-2Dmanual_dev-2Dmanual.html-23building-2Dsoftware-2Dfrom-2Dan-2Dexternal-2Dsource=DwICAg=zVFQZQ67ypsA9mYKSCqWmQHiVkCCaN- > Gb60_N6TVnLk=rUGbzRLPRvFX3fKfyNvK9Xgz1MfBXwYtoWGtKfkzG3k=o9VVMczGF- > R7- > EC8L6cShpCsU-3GWfug-0DbDpHLamU=JFZPoF89tSSrPhrhqmYSV5VrbkGpTM4jsqr1ij9sBa4= > > > However, it doesn't work with > > devtool: > > > > $ devtool modify -n d ~/mysrc/a/b/c/d > > NOTE: Creating workspace layer in <... snip > > ...>/yocto/poky/build/workspace > > NOTE: Enabling workspace layer in bblayers.conf > > Parsing recipes..done. > > NOTE: Recipe d now set up to build from ~/mysrc/a/b/c/d/b/c/d > > > > Where is the extra 'b/c/d' coming from in where it will be built from? > > This directory is reflected in the bbappends created the workspace: > > > > FILESEXTRAPATHS_prepend := "${THISDIR}/${PN}:" > > FILESPATH_prepend := "~/mysrc/a/b/c/d/b/c/d/oe-local-files:" > > > > Is it possible to have a ${S} of "${WORKDIR}/src" in this situation? > > How can I get devtool to behave and not duplicate the path? > > The ~ shouldn't be in the value of FILESPATH, it ought to be expanded, so > that's something we need to fix. Don't worry about the ~, that was just something I passed in the email to obscure my path. Sorry. > With regard to the path doubling of the > subdirectories, it's assuming that because you have that subdirectory path in > S then you need that underneath the specified source directory (since that is > usually true if you were for example pointing to a checkout of a git > repository, or an extracted source tarball). You are correct. I can't believe I didn't think of that before. Thank you so much. That gets me going at least. > devtool hasn't been written to handle this usage (which is unusual -typically > the recipe is pointing at remote source). It would be good to get an > understanding of your use case here so we can figure out where to go from > here. > > Cheers, > Paul > > -- > > Paul Eggleton > Intel Open Source Technology Centre -- ___ yocto mailing list yocto@yoctoproject.org https://lists.yoctoproject.org/listinfo/yocto
[yocto] local fetcher and devtool don't work well together
I have a recipe with: SRC_URI = "file://a/b/c/d;subdir=src" S = "${WORKDIR}/src/a/b/c/d" First off, ${S} must be set to the full path because the basepath uri parameter[1] doesn't work. So this doesn't work: SRC_URI = "file://a/b/c/d;subdir=src;basepath=a/b/c/d" S = "${WORKDIR}/src" I wish it would, but that's ok. I can deal with the long ${S} path as it compiles and works fine with bitbake. However, it doesn't work with devtool: $ devtool modify -n d ~/mysrc/a/b/c/d NOTE: Creating workspace layer in <... snip ...>/yocto/poky/build/workspace NOTE: Enabling workspace layer in bblayers.conf Parsing recipes..done. NOTE: Recipe d now set up to build from ~/mysrc/a/b/c/d/b/c/d Where is the extra 'b/c/d' coming from in where it will be built from? This directory is reflected in the bbappends created the workspace: FILESEXTRAPATHS_prepend := "${THISDIR}/${PN}:" FILESPATH_prepend := "~/mysrc/a/b/c/d/b/c/d/oe-local-files:" inherit externalsrc # NOTE: We use pn- overrides here to avoid affecting multiple variants in the case where the recipe uses BBCLASSEXTEND EXTERNALSRC_pn-backtrace = "~/mysrc/a/b/c/d/b/c/d" If I fix the path in the workspace bbappends, it does compile and work file. Is it possible to have a ${S} of "${WORKDIR}/src" in this situation? How can I get devtool to behave and not duplicate the path? PS - This also doesn't work with a normal 'devtool modify d'. [1] - https://urldefense.proofpoint.com/v2/url?u=http-3A__www.yoctoproject.org_docs_latest_bitbake-2Duser-2Dmanual_bitbake-2Duser-2Dmanual.html-23bb-2Dthe-2Dunpack=DwIBAg=zVFQZQ67ypsA9mYKSCqWmQHiVkCCaN-Gb60_N6TVnLk=IZ3mUDA8X9CZR4hNctKngDDYs_BhK3qVNLG-nWGbW2E=wH_YZj_cv9gBOhJqSKTr4fiuM5UCTphWx9zfY2ceudg=2z9LBCu2xDjBhB51eToRNH7-wsLrFY4NPKy4_cwuRI0= -- ___ yocto mailing list yocto@yoctoproject.org https://lists.yoctoproject.org/listinfo/yocto
[yocto] Yocto eSDK nativesdk vs native vs normal
I'm having a hard time figuring out what specific package to install when I'm in the eSDK environment. Say I want to install cmake, do I: devtool sdk-install cmake or: devtool sdk-install cmake-native or: devtool sdk-install nativesdk-cmake CMake is just one example, any other package would cause the same confusion for me. Does it depend on the package; tools for use in the eSDK vs libraries for the target?-- ___ yocto mailing list yocto@yoctoproject.org https://lists.yoctoproject.org/listinfo/yocto
[yocto] Yocto eSDK, Editing/Adding Recipes, and VCS
I've been trying to document some workflows for my developers using the Yocto eSDK for our image, but I am coming up empty when it comes to editing or adding recipes. I can use devtool to edit a recipe like so: devtool edit-recipe -a my-recipe But any changes I make are hidden away in the esdk/layers/poky/meta-mylayer directory, which isn't managed by a VCS (git in this instance). So there's this modified file in a directory--great. What is the developer supposed to do with it? Copy it over to a VCS managed directory of meta-mylayer, commit, and push? I feel like I'm missing something with the devtool workflow. Can anyone point me in the right direction? -- ___ yocto mailing list yocto@yoctoproject.org https://lists.yoctoproject.org/listinfo/yocto
[yocto] Minimal eSDK with PKGDATA causes licensing issues
I added these lines to my eSDK image I'm trying to build: SDK_EXT_TYPE = "minimal" SDK_INCLUDE_TOOLCHAIN = "1" SDK_INCLUDE_PKGDATA = "1" But bitbake comes back with lots of license issues (mpv, gstreamer, libvdpau, etc). These are all the commercial licensed recipes. I don't want PKGDATA for recipes that I haven't already listed in LICENSE_FLAGS_WHITELIST to end up the eSDK, so I don't care about these errors, but they stop the build in its tracks. How can I exclude commercial licensed (or any incompatible license) recipes from my eSDK?-- ___ yocto mailing list yocto@yoctoproject.org https://lists.yoctoproject.org/listinfo/yocto