I'm having an issue with a new bitbake recipe. My recipe depends on files produced by a net-snmp recipe from openembedded, but net-snmp doesn't seem to be building before trying to compile my recipe.
In my .bb file, I have:
DEPENDS = " \
net-snmp \
...
EXTRA_OEMAKE += " \
BINDIR=${STAGING_BINDIR} \
...
do_compile() {
oe_runmake foo
}
Then my makefile includes
CFLAGS=`${BINDIR}/net-snmp-config --cflags` ...
When I run bitbake foo, do_compile throws the error
/bin/sh: [some-path-to-net-snmp-config]: No such file or directory
If I check the staging directory, net-snmp-config is indeed missing.
But if I then run bitbake -C compile net-snmp foo, foo builds successfully, and net-snmp-config is in the staging directory.
What am I doing wrong here?
Edit: I've just noticed that even if foo builds successfully, net-snmp is suspiciously absent from my final rootfs image.
If I add RDEPENDS="net-snmp" to my recipe, I get the error:
satisfy_dependencies_for: Cannot satisfy the following dependencies for packagegroup-core-boot:
| * net-snmp *
I think I found the solution.
Based on this blog entry, I added the line
do_compile[depends] += "net-snmp:do_install"
Now my recipe builds successfully every time.
I was also trying to use RDEPENDS incorrectly. Since net-snmp outputs several packages, I needed to specify the individual packages in RDEPENDS in order to include them in my final build.
Related
This follows on from
How do I strip and objcopy a built .so file in the Yocto bitbake compile step?
This leads back to considerable background information.
As mentioned in the previous question, I am seeking to build OCA, which I have
as a non-Yocto makefile-based project, in Yocto. The project, which builds
fine outside of Yocto, and even in Yocto, is quite complex. The issue is that it
is not cross-compiling for my target, which is aarch64 armv8-a. It is building
successfully, but for my host machine, which is x86-64. Then Yocto sensibly
refuses to package it, saying "Unable to recognise the format of the input
file".
I changed the compile flags in my makefile to -march=armv8-a, but got the error
"cc1plus: error: bad value ('armv8-a') for '-march=' switch" which seems to
mean that I can't use the host's installed gcc for cross-compiling, but rather a
cross-compiler is needed. I previously custom-added two additional layers, a
sample helloworld and mDNS (see previous questions for lots of background), and
they all cross-compiled fine, so I know that Yocto is basically set up to do it.
What is the method to get the cross-compilation to happen? Do I need to do a lot
of pervasive changes to my project's makefile system? It may not ever have been
designed with Yocto in mind.
I am looking into this: "cross-compile library recipe in yocto":
cross-compile library recipe in yocto
which seems to have some relevant information. Edit: it was only standard stuff that I had seen before; nothing about how cross-compilation gets invoked instead of the host machine's gcc.
Edit: My updated recipe, with the FILES_${PN} added and make changed to
oe_runmake.
DESCRIPTION = "OCA"
PRIORITY = "optional"
SECTION = "protocols"
LICENSE = "MIT"
LIC_FILES_CHKSUM = "file://${COMMON_LICENSE_DIR}/MIT;md5=0835ade698e0bcf8506ecda2f7b4f302"
SRC_URI = "file://oca-1.2.7"
S = "${WORKDIR}/oca-1.2.7/Src"
# Need to override S because BitBake expects the source to be in a dir called
# oca-1.2.7 in the work dir, but it's actually additionally under Src/.
# Need a do_compile, since OCA has a makefile with a non-standard name,
# makefileOCA. Also needs non-standard flags, -f and linuxRelease.
do_compile() {
export CAP_HOME="${WORKDIR}/oca-1.2.7"
oe_runmake -f makefileOCA linuxRelease
}
do_install() {
install -d ${D}${libdir}
cp ../Obj/linuxApp/Release/OcaProtoController.so ${D}${libdir}/OcaProtoController.so
chmod 0755 ${D}${libdir}/OcaProtoController.so
}
FILES_${PN} += "${libdir}/OcaProtoController.so"
Edit: I found some info:
https://www.yoctoproject.org/docs/1.8.1/adt-manual/adt-manual.html
https://www.yoctoproject.org/docs/1.8.1/adt-manual/adt-manual.html#setting-up-the-cross-development-environment
I commented-out the setting of CC and LD in makeOCA.inc.
I did the cleaning actions of deleting the tmp folder from my build-wayland
build dir, and did "bitbake -c cleansstate oca". Then I did "time bitbake oca".
Now it looks like it is using the cross compiler.
The first error I got now is:
~/Yocto/imx-yocto-bsp/build-wayland/tmp/work/aarch64-poky-linux/oca/1.2.7-r0/oca-1.2.7/Obj/linuxApp/Release/OcaAgentProxies.a
aarch64-poky-linux-ld: cannot find crus: No such file or directory
So this is the next question: Can you tell me what "crus" means in "LDFLAGS = crus $#"?
I want copy folders and it's content to yocto during image build process. For this process I am using following recipe
SUMMARY = "Installation Recipe"
DESCRIPTION = "It installs folder"
HOMEPAGE = ""
LICENSE = "CLOSED"
MY_FILES1 = "/home/jane/d1fold"
MY_FILES2 = "/home/jane/d2fold"
inherit allarch
do_install() {
install -d ${D}/home/root
cp -R ${MY_FILES1}/* ${D}/home/root
cp -R ${MY_FILES2} ${D}/home/root
}
FILES_${PN} += " /home/root"
But I receive following error ERROR: QA Issue: weaved: Recipe inherits the allarch class, but has packaged architecture-specific binaries [arch]. How can I resolve this error?
This error means that you are trying to install architecture-specific binaries (compiled for x86, arm64 etc), while inheriting allarch class. From yocto reference manual:
The allarch class is inherited by recipes that do not produce architecture-specific output.
This is an obvious contradiction.
What are you trying to do? Creating of recipe that only installs some files seems like wrong architecture decision. And why do you want to inherit allarch?
You are just coping files to rootfs. So you no need use inherit allarch. remove that and compile.
bitbake cleanall
Removes all output files, shared state cache, and downloaded source files for a target
It is not clear or documented if it cleans all build time dependencies as well
If you want to clean everything do,
bitbake world -c cleanall --continue
The --continue will ignore any dependency errors while cleaning. Continue as much as possible after an error.
No, cleanall does not clean dependencies. eg
bitbake -c cleanall core-image-minimal
only removes the output of that named recipe.
What i usually do to clean "everything" is running cleanall on the receipe "world":
bitbake -c cleanall world
If that fails because of unresolvable packages like that:
ERROR: Nothing PROVIDES 'sg3-utils' (but /home/blubb/meta-freescale/recipes-devtools/utp-com/utp-com_git.bb DEPENDS on or otherwise requires it).
I just add the packages temporary to the ASSUME_PROVIDED variable like this :
bitbake -c cleanall world --ignore-deps=python-nativedtc-native --ignore-deps=sg3-utils
If nothing provides this packages it is unlikely that they where ever build.
Please read the mega-manual section do_cleanall .
do_cleanall removes:
all output files
shared state (sstate) cache
and downloaded source files for a target (i.e. the contents of DL_DIR).
You can run this task using BitBake as follows:
$ bitbake -c cleanall <recipe-name>
If recipe name is not passed to cleanall task it does not work.
Removes all output files, shared state (sstate) cache, and downloaded source files for a target (i.e. the contents of DL_DIR). Essentially, the do_cleanall task is identical to the do_cleansstate task with the added removal of downloaded source files.
You can run this task using BitBake as follows:
$ bitbake -c cleanall recipe
Typically, you would not normally use the cleanall task. Do so only if you want to start fresh with the do_fetch task.
Other folks have already answered that bitbake does not automatically clean dependencies, but you can create an Inter-task dependency (https://www.yoctoproject.org/docs/3.1/bitbake-user-manual/bitbake-user-manual.html#inter-task-dependencies) to clean your dependencies if needed by adding a command to the recipe:
do_task[depends] = "recipe:task"
We've extended bitbake to build native recipes and automatically run unit tests during a build. In that case we need to clean the native recipe when cleaning the target so you could add:
do_clean[depends] = "${PN}-native:do_clean"
do_cleanall[depends] = "${PN}-native:do_cleanall"
do_cleansstate[depends] = "${PN}-native:do_cleansstate"
That solution falls a bit short because the native recipes will attempt to clean ${PN}-native-native, so you'll need a conditional to not apply if it's already native:
do_clean[depends] += "${#'' if bb.data.inherits_class('native', d) else '${PN}-native:do_clean'}"
do_cleanall[depends] += "${#'' if bb.data.inherits_class('native', d) else '${PN}-native:do_cleanall'}"
do_cleansstate[depends] += "${#'' if bb.data.inherits_class('native', d) else '${PN}-native:do_cleansstate'}"
I have been using Yocto to create Linux builds for an ARM board.
I had been cross compiling add on applications manually. Now we are in a place where we would like a nice integrated build so I started adding custom recipes to yocto.
I have been struggling with the ARM build (a x86 build with the same code seems fine).
Even a basic 'hello world' pretty much cut and paste from the development manual does not work (http://www.yoctoproject.org/docs/current/dev-manual/dev-manual.html#new-recipe-writing-a-new-recipe)
Here is the recipe:
SUMMARY = "Simple helloworld application"
SECTION = "examples"
LICENSE = "MIT"
LIC_FILES_CHKSUM = "file://${COMMON_LICENSE_DIR}/MIT;md5=0835ade698e0bcf8506ecda2f7b4f302"
RPROVIDES_${PN} = "helloworld"
FILES_${PN} += "${bindir}"
SRC_URI = "file://helloworld.c"
S = "${WORKDIR}"
do_compile() {
${CC} helloworld.c -o helloworld ${LDFLAGS}
}
do_install() {
install -d ${D}${bindir}
install -m 0755 helloworld ${D}${bindir}
}
Here is the error:
ERROR: helloworld not found in the base feeds (smarc_samx6i cortexa9t2hf-vfp-neon-mx6qdl cortexa9hf-vfp-neon-mx6qdl cortexa9t2hf-vfp-neon cortexa9t2hf-vfp cortexa9hf-vfp-neon cortexa9hf-vfp armv7at2hf-vfp-neon armv7ahf-vfp-neon armv7at2hf-vfp armv7ahf-vfp armv6thf-vfp armv6hf-vfp armv5tehf-vfp armv5ehf-vfp armv5thf-vfp armv5hf-vfp noarch any all).
ERROR: Function failed: do_rootfs
Any suggestions as to what would be causing this error?
The package does build properly; the problem seems to be isolated to finding it for the rootfs.
Thanks!
EDIT:
I have a solution that seems to work, although it is not ideal long term.
Changing the package name under IMAGE_INSTALL from helloworld to helloworld-0.0.1 resolves the issue. Obviously I would rather not hard code the version of each package in the top level recipe and other packages do not require this, so hopefully there is another solution.
EDIT 2:
Renaming the recipe and removing the version string also resolves the issue. Once again, this does not seem ideal long term.
OK, after some further testing I discovered this was a naming issue with the recipe.
It was named helloworld-0.0.1.bb (the same format with the other recipes I had put together driving me to try this simple test).
If anyone else encounters this simply replacing the '-' with a '_' resolves this.
1.Rename your recipe name e.g hello-0.1.bb to hello_0.1.bb
2.Add below line at last only:
FILES_${PN} = "${bindir}/*"
Abvoe line helps you copy your binary to rootfs.
I'm using bitbake to compile rygel for yocto from the meta-openembedded layer, with the plugins mpris and gst-launch.
As per the PACKAGECONFIG docs, I created a bbappend file and added the following:
EXTRA_OECONF = "--disable-introspection \
--disable-tracker-plugin \
--with-media-engine=gstreamer \
--enable-mpris-plugin \
--enable-gst-launch-plugin"
PR = "r1"
It compiles and installs, but has no plugins.
The append shows up when I run bitbake-layers show-appends, so at least bitbake is finding it. After running bitbake the directory tmp/work/core2-64-poky-linux/rygel/0.26.1-r1/image/usr/lib/rygel-2.6/plugins/ is populated. Then when I run the image /usr/lib/rygel-2.6/ contains an engines dir and nothing else.
Any idea where I'm going wrong?
I don't think your read all the way down to "If you want to change an existing PACKAGECONFIG block, you can do so one of two ways:".
From a bbappend, just do
PACKAGECONFIG_append = " mpris gst-launch"
In the recipe do_install, they remove some of the engines and plugins files. This might be the reason you do not see them in your image.
do_install_append() {
# Remove .la files for loadable modules
rm -f ${D}/${libdir}/rygel-${LIBV}/engines/*.la
rm -f ${D}/${libdir}/rygel-${LIBV}/plugins/*.la
}
your compiling plugins successfully and not able to see in board(rootfs)? if yes please add below line in your .bbappend file. '
FILES_${PN} += "${libdir}/*"
this will add all your compiled plugins to your rootfs image.