3 Syntax and Operators
BitBake files have their own syntax. The syntax has similarities to several other languages but also has some unique features. This section describes the available syntax and operators as well as provides examples.
3.1 Basic Syntax
This section provides some basic syntax examples.
3.1.1 Basic Variable Setting
The following example sets VARIABLE
to “value”. This assignment
occurs immediately as the statement is parsed. It is a “hard”
assignment.
VARIABLE = "value"
As expected, if you include leading or trailing spaces as part of an assignment, the spaces are retained:
VARIABLE = " value"
VARIABLE = "value "
Setting VARIABLE
to “” sets
it to an empty string, while setting the variable to “ “ sets it to a
blank space (i.e. these are not the same values).
VARIABLE = ""
VARIABLE = " "
You can use single quotes instead of double quotes when setting a variable’s value. Doing so allows you to use values that contain the double quote character:
VARIABLE = 'I have a " in my value'
Note
Unlike in Bourne shells, single quotes work identically to double quotes in all other ways. They do not suppress variable expansions.
3.1.2 Modifying Existing Variables
Sometimes you need to modify existing variables. Following are some cases where you might find you want to modify an existing variable:
Customize a recipe that uses the variable.
Change a variable’s default value used in a
*.bbclass
file.Change the variable in a
*.bbappend
file to override the variable in the original recipe.Change the variable in a configuration file so that the value overrides an existing configuration.
Changing a variable value can sometimes depend on how the value was originally assigned and also on the desired intent of the change. In particular, when you append a value to a variable that has a default value, the resulting value might not be what you expect. In this case, the value you provide might replace the value rather than append to the default value.
If after you have changed a variable’s value and something unexplained occurs, you can use BitBake to check the actual value of the suspect variable. You can make these checks for both configuration and recipe level changes:
For configuration changes, use the following:
$ bitbake -e
This command displays variable values after the configuration files (i.e.
local.conf
,bblayers.conf
,bitbake.conf
and so forth) have been parsed.Note
Variables that are exported to the environment are preceded by the string “export” in the command’s output.
For recipe changes, use the following:
$ bitbake recipe -e \| grep VARIABLE="
This command checks to see if the variable actually makes it into a specific recipe.
3.1.3 Line Joining
Outside of functions, BitBake joins any line ending in a backslash character (”") with the following line before parsing statements. The most common use for the “" character is to split variable assignments over multiple lines, as in the following example:
FOO = "bar \
baz \
qaz"
Both the “" character and the newline
character that follow it are removed when joining lines. Thus, no
newline characters end up in the value of FOO
.
Consider this additional example where the two assignments both assign
“barbaz” to FOO
:
FOO = "barbaz"
FOO = "bar\
baz"
Note
BitBake does not interpret escape sequences like “n” in variable
values. For these to have an effect, the value must be passed to some
utility that interprets escape sequences, such as
printf
or echo -n
.
3.1.4 Variable Expansion
Variables can reference the contents of other variables using a syntax that is similar to variable expansion in Bourne shells. The following assignments result in A containing “aval” and B evaluating to “preavalpost”.
A = "aval"
B = "pre${A}post"
Note
Unlike in Bourne shells, the curly braces are mandatory: Only ${FOO}
and not
$FOO
is recognized as an expansion of FOO
.
The “=” operator does not immediately expand variable references in the right-hand side. Instead, expansion is deferred until the variable assigned to is actually used. The result depends on the current values of the referenced variables. The following example should clarify this behavior:
A = "${B} baz"
B = "${C} bar"
C = "foo"
*At this point, ${A} equals "foo bar baz"*
C = "qux"
*At this point, ${A} equals "qux bar baz"*
B = "norf"
*At this point, ${A} equals "norf baz"\*
Contrast this behavior with the Immediate variable expansion (:=) operator.
If the variable expansion syntax is used on a variable that does not
exist, the string is kept as is. For example, given the following
assignment, BAR
expands to the literal string “${FOO}” as long as
FOO
does not exist.
BAR = "${FOO}"
3.1.5 Setting a default value (?=)
You can use the “?=” operator to achieve a “softer” assignment for a variable. This type of assignment allows you to define a variable if it is undefined when the statement is parsed, but to leave the value alone if the variable has a value. Here is an example:
A ?= "aval"
If A
is
set at the time this statement is parsed, the variable retains its
value. However, if A
is not set, the variable is set to “aval”.
Note
This assignment is immediate. Consequently, if multiple “?=” assignments to a single variable exist, the first of those ends up getting used.
3.1.6 Setting a weak default value (??=)
It is possible to use a “weaker” assignment than in the previous section by using the “??=” operator. This assignment behaves identical to “?=” except that the assignment is made at the end of the parsing process rather than immediately. Consequently, when multiple “??=” assignments exist, the last one is used. Also, any “=” or “?=” assignment will override the value set with “??=”. Here is an example:
A ??= "somevalue"
A ??= "someothervalue"
If A
is set before the above statements are
parsed, the variable retains its value. If A
is not set, the
variable is set to “someothervalue”.
Again, this assignment is a “lazy” or “weak” assignment because it does not occur until the end of the parsing process.
3.1.7 Immediate variable expansion (:=)
The “:=” operator results in a variable’s contents being expanded immediately, rather than when the variable is actually used:
T = "123"
A := "test ${T}"
T = "456"
B := "${T} ${C}"
C = "cval"
C := "${C}append"
In this example, A
contains “test 123”, even though the final value
of T
is “456”. The variable B
will end up containing “456
cvalappend”. This is because references to undefined variables are
preserved as is during (immediate)expansion. This is in contrast to GNU
Make, where undefined variables expand to nothing. The variable C
contains “cvalappend” since ${C}
immediately expands to “cval”.
3.1.8 Appending (+=) and prepending (=+) With Spaces
Appending and prepending values is common and can be accomplished using the “+=” and “=+” operators. These operators insert a space between the current value and prepended or appended value.
These operators take immediate effect during parsing. Here are some examples:
B = "bval"
B += "additionaldata"
C = "cval"
C =+ "test"
The variable B
contains “bval additionaldata” and C
contains “test
cval”.
3.1.9 Appending (.=) and Prepending (=.) Without Spaces
If you want to append or prepend values without an inserted space, use the “.=” and “=.” operators.
These operators take immediate effect during parsing. Here are some examples:
B = "bval"
B .= "additionaldata"
C = "cval"
C =. "test"
The variable B
contains “bvaladditionaldata” and C
contains
“testcval”.
3.1.10 Appending and Prepending (Override Style Syntax)
You can also append and prepend a variable’s value using an override style syntax. When you use this syntax, no spaces are inserted.
These operators differ from the “:=”, “.=”, “=.”, “+=”, and “=+” operators in that their effects are applied at variable expansion time rather than being immediately applied. Here are some examples:
B = "bval"
B_append = " additional data"
C = "cval"
C_prepend = "additional data "
D = "dval"
D_append = "additional data"
The variable B
becomes “bval additional data” and C
becomes “additional data cval”.
The variable D
becomes “dvaladditional data”.
Note
You must control all spacing when you use the override syntax.
It is also possible to append and prepend to shell functions and BitBake-style Python functions. See the “Shell Functions” and “BitBake-Style Python Functions” sections for examples.
3.1.11 Removal (Override Style Syntax)
You can remove values from lists using the removal override style syntax. Specifying a value for removal causes all occurrences of that value to be removed from the variable.
When you use this syntax, BitBake expects one or more strings. Surrounding spaces and spacing are preserved. Here is an example:
FOO = "123 456 789 123456 123 456 123 456"
FOO_remove = "123"
FOO_remove = "456"
FOO2 = " abc def ghi abcdef abc def abc def def"
FOO2_remove = "\
def \
abc \
ghi \
"
The variable FOO
becomes
“ 789 123456 “ and FOO2
becomes “ abcdef “.
Like “_append” and “_prepend”, “_remove” is applied at variable expansion time.
3.1.12 Override Style Operation Advantages
An advantage of the override style operations “_append”, “_prepend”, and
“_remove” as compared to the “+=” and “=+” operators is that the
override style operators provide guaranteed operations. For example,
consider a class foo.bbclass
that needs to add the value “val” to
the variable FOO
, and a recipe that uses foo.bbclass
as follows:
inherit foo
FOO = "initial"
If foo.bbclass
uses the “+=” operator,
as follows, then the final value of FOO
will be “initial”, which is
not what is desired:
FOO += "val"
If, on the other hand, foo.bbclass
uses the “_append” operator, then the final value of FOO
will be
“initial val”, as intended:
FOO_append = " val"
Note
It is never necessary to use “+=” together with “_append”. The following sequence of assignments appends “barbaz” to FOO:
FOO_append = "bar"
FOO_append = "baz"
The only effect of changing the second assignment in the previous example to use “+=” would be to add a space before “baz” in the appended value (due to how the “+=” operator works).
Another advantage of the override style operations is that you can combine them with other overrides as described in the “Conditional Syntax (Overrides)” section.
3.1.13 Variable Flag Syntax
Variable flags are BitBake’s implementation of variable properties or attributes. It is a way of tagging extra information onto a variable. You can find more out about variable flags in general in the “Variable Flags” section.
You can define, append, and prepend values to variable flags. All the standard syntax operations previously mentioned work for variable flags except for override style syntax (i.e. “_prepend”, “_append”, and “_remove”).
Here are some examples showing how to set variable flags:
FOO[a] = "abc"
FOO[b] = "123"
FOO[a] += "456"
The variable FOO
has two flags:
[a]
and [b]
. The flags are immediately set to “abc” and “123”,
respectively. The [a]
flag becomes “abc 456”.
No need exists to pre-define variable flags. You can simply start using them. One extremely common application is to attach some brief documentation to a BitBake variable as follows:
CACHE[doc] = "The directory holding the cache of the metadata."
3.1.14 Inline Python Variable Expansion
You can use inline Python variable expansion to set variables. Here is an example:
DATE = "${@time.strftime('%Y%m%d',time.gmtime())}"
This example results in the DATE
variable being set to the current date.
Probably the most common use of this feature is to extract the value of
variables from BitBake’s internal data dictionary, d
. The following
lines select the values of a package name and its version number,
respectively:
PN = "${@bb.parse.BBHandler.vars_from_file(d.getVar('FILE', False),d)[0] or 'defaultpkgname'}"
PV = "${@bb.parse.BBHandler.vars_from_file(d.getVar('FILE', False),d)[1] or '1.0'}"
Note
Inline Python expressions work just like variable expansions insofar as the “=” and “:=” operators are concerned. Given the following assignment, foo() is called each time FOO is expanded:
FOO = "${@foo()}"
Contrast this with the following immediate assignment, where foo() is only called once, while the assignment is parsed:
FOO := "${@foo()}"
For a different way to set variables with Python code during parsing, see the “Anonymous Python Functions” section.
3.1.15 Unsetting variables
It is possible to completely remove a variable or a variable flag from BitBake’s internal data dictionary by using the “unset” keyword. Here is an example:
unset DATE
unset do_fetch[noexec]
These two statements remove the DATE
and the do_fetch[noexec]
flag.
3.1.16 Providing Pathnames
When specifying pathnames for use with BitBake, do not use the tilde (“~”) character as a shortcut for your home directory. Doing so might cause BitBake to not recognize the path since BitBake does not expand this character in the same way a shell would.
Instead, provide a fuller path as the following example illustrates:
BBLAYERS ?= " \
/home/scott-lenovo/LayerA \
"
3.2 Exporting Variables to the Environment
You can export variables to the environment of running tasks by using
the export
keyword. For example, in the following example, the
do_foo
task prints “value from the environment” when run:
export ENV_VARIABLE
ENV_VARIABLE = "value from the environment"
do_foo() {
bbplain "$ENV_VARIABLE"
}
Note
BitBake does not expand $ENV_VARIABLE
in this case because it lacks the
obligatory {}
. Rather, $ENV_VARIABLE
is expanded by the shell.
It does not matter whether export ENV_VARIABLE
appears before or
after assignments to ENV_VARIABLE
.
It is also possible to combine export
with setting a value for the
variable. Here is an example:
export ENV_VARIABLE = "variable-value"
In the output of bitbake -e
, variables that are exported to the
environment are preceded by “export”.
Among the variables commonly exported to the environment are CC
and
CFLAGS
, which are picked up by many build systems.
3.3 Conditional Syntax (Overrides)
BitBake uses OVERRIDES to control what
variables are overridden after BitBake parses recipes and configuration
files. This section describes how you can use OVERRIDES
as
conditional metadata, talks about key expansion in relationship to
OVERRIDES
, and provides some examples to help with understanding.
3.3.1 Conditional Metadata
You can use OVERRIDES
to conditionally select a specific version of
a variable and to conditionally append or prepend the value of a
variable.
Note
Overrides can only use lower-case characters. Additionally, underscores are not permitted in override names as they are used to separate overrides from each other and from the variable name.
Selecting a Variable: The
OVERRIDES
variable is a colon-character-separated list that contains items for which you want to satisfy conditions. Thus, if you have a variable that is conditional on “arm”, and “arm” is inOVERRIDES
, then the “arm”-specific version of the variable is used rather than the non-conditional version. Here is an example:OVERRIDES = "architecture:os:machine" TEST = "default" TEST_os = "osspecific" TEST_nooverride = "othercondvalue"
In this example, the
OVERRIDES
variable lists three overrides: “architecture”, “os”, and “machine”. The variableTEST
by itself has a default value of “default”. You select the os-specific version of theTEST
variable by appending the “os” override to the variable (i.e.TEST_os
).To better understand this, consider a practical example that assumes an OpenEmbedded metadata-based Linux kernel recipe file. The following lines from the recipe file first set the kernel branch variable
KBRANCH
to a default value, then conditionally override that value based on the architecture of the build:KBRANCH = "standard/base" KBRANCH_qemuarm = "standard/arm-versatile-926ejs" KBRANCH_qemumips = "standard/mti-malta32" KBRANCH_qemuppc = "standard/qemuppc" KBRANCH_qemux86 = "standard/common-pc/base" KBRANCH_qemux86-64 = "standard/common-pc-64/base" KBRANCH_qemumips64 = "standard/mti-malta64"
Appending and Prepending: BitBake also supports append and prepend operations to variable values based on whether a specific item is listed in
OVERRIDES
. Here is an example:DEPENDS = "glibc ncurses" OVERRIDES = "machine:local" DEPENDS_append_machine = "libmad"
In this example,
DEPENDS
becomes “glibc ncurses libmad”.Again, using an OpenEmbedded metadata-based kernel recipe file as an example, the following lines will conditionally append to the
KERNEL_FEATURES
variable based on the architecture:KERNEL_FEATURES_append = " ${KERNEL_EXTRA_FEATURES}" KERNEL_FEATURES_append_qemux86=" cfg/sound.scc cfg/paravirt_kvm.scc" KERNEL_FEATURES_append_qemux86-64=" cfg/sound.scc cfg/paravirt_kvm.scc"
Setting a Variable for a Single Task: BitBake supports setting a variable just for the duration of a single task. Here is an example:
FOO_task-configure = "val 1" FOO_task-compile = "val 2"
In the previous example,
FOO
has the value “val 1” while thedo_configure
task is executed, and the value “val 2” while thedo_compile
task is executed.Internally, this is implemented by prepending the task (e.g. “task-compile:”) to the value of OVERRIDES for the local datastore of the
do_compile
task.You can also use this syntax with other combinations (e.g. “
_prepend
”) as shown in the following example:EXTRA_OEMAKE_prepend_task-compile = "${PARALLEL_MAKE} "
3.3.2 Key Expansion
Key expansion happens when the BitBake datastore is finalized. To better understand this, consider the following example:
A${B} = "X"
B = "2"
A2 = "Y"
In this case, after all the parsing is complete, BitBake expands
${B}
into “2”. This expansion causes A2
, which was set to “Y”
before the expansion, to become “X”.
3.3.3 Examples
Despite the previous explanations that show the different forms of variable definitions, it can be hard to work out exactly what happens when variable operators, conditional overrides, and unconditional overrides are combined. This section presents some common scenarios along with explanations for variable interactions that typically confuse users.
There is often confusion concerning the order in which overrides and various “append” operators take effect. Recall that an append or prepend operation using “_append” and “_prepend” does not result in an immediate assignment as would “+=”, “.=”, “=+”, or “=.”. Consider the following example:
OVERRIDES = "foo"
A = "Z"
A_foo_append = "X"
For this case,
A
is unconditionally set to “Z” and “X” is unconditionally and
immediately appended to the variable A_foo
. Because overrides have
not been applied yet, A_foo
is set to “X” due to the append and
A
simply equals “Z”.
Applying overrides, however, changes things. Since “foo” is listed in
OVERRIDES
, the conditional variable A
is replaced with the “foo”
version, which is equal to “X”. So effectively, A_foo
replaces
A
.
This next example changes the order of the override and the append:
OVERRIDES = "foo"
A = "Z"
A_append_foo = "X"
For this case, before
overrides are handled, A
is set to “Z” and A_append_foo
is set
to “X”. Once the override for “foo” is applied, however, A
gets
appended with “X”. Consequently, A
becomes “ZX”. Notice that spaces
are not appended.
This next example has the order of the appends and overrides reversed back as in the first example:
OVERRIDES = "foo"
A = "Y"
A_foo_append = "Z"
A_foo_append = "X"
For this case, before any overrides are resolved,
A
is set to “Y” using an immediate assignment. After this immediate
assignment, A_foo
is set to “Z”, and then further appended with “X”
leaving the variable set to “ZX”. Finally, applying the override for
“foo” results in the conditional variable A
becoming “ZX” (i.e.
A
is replaced with A_foo
).
This final example mixes in some varying operators:
A = "1"
A_append = "2"
A_append = "3"
A += "4"
A .= "5"
For this case, the type of append
operators are affecting the order of assignments as BitBake passes
through the code multiple times. Initially, A
is set to “1 45”
because of the three statements that use immediate operators. After
these assignments are made, BitBake applies the “_append” operations.
Those operations result in A
becoming “1 4523”.
3.5 Functions
As with most languages, functions are the building blocks that are used to build up operations into tasks. BitBake supports these types of functions:
Shell Functions: Functions written in shell script and executed either directly as functions, tasks, or both. They can also be called by other shell functions.
BitBake-Style Python Functions: Functions written in Python and executed by BitBake or other Python functions using
bb.build.exec_func()
.Python Functions: Functions written in Python and executed by Python.
Anonymous Python Functions: Python functions executed automatically during parsing.
Regardless of the type of function, you can only define them in class
(.bbclass
) and recipe (.bb
or .inc
) files.
3.5.1 Shell Functions
Functions written in shell script and executed either directly as functions, tasks, or both. They can also be called by other shell functions. Here is an example shell function definition:
some_function () {
echo "Hello World"
}
When you create these types of functions in
your recipe or class files, you need to follow the shell programming
rules. The scripts are executed by /bin/sh
, which may not be a bash
shell but might be something such as dash
. You should not use
Bash-specific script (bashisms).
Overrides and override-style operators like _append
and _prepend
can also be applied to shell functions. Most commonly, this application
would be used in a .bbappend
file to modify functions in the main
recipe. It can also be used to modify functions inherited from classes.
As an example, consider the following:
do_foo() {
bbplain first
fn
}
fn_prepend() {
bbplain second
}
fn() {
bbplain third
}
do_foo_append() {
bbplain fourth
}
Running do_foo
prints the following:
recipename do_foo: first
recipename do_foo: second
recipename do_foo: third
recipename do_foo: fourth
Note
Overrides and override-style operators can be applied to any shell function, not just tasks.
You can use the bitbake -e
recipename command to view the final
assembled function after all overrides have been applied.
3.5.2 BitBake-Style Python Functions
These functions are written in Python and executed by BitBake or other
Python functions using bb.build.exec_func()
.
An example BitBake function is:
python some_python_function () {
d.setVar("TEXT", "Hello World")
print d.getVar("TEXT")
}
Because the Python “bb” and “os” modules are already imported, you do not need to import these modules. Also in these types of functions, the datastore (“d”) is a global variable and is always automatically available.
Note
Variable expressions (e.g. ${X}
) are no longer expanded within Python
functions. This behavior is intentional in order to allow you to freely set
variable values to expandable expressions without having them expanded
prematurely. If you do wish to expand a variable within a Python function,
use d.getVar("X")
. Or, for more complicated expressions, use d.expand()
.
Similar to shell functions, you can also apply overrides and override-style operators to BitBake-style Python functions.
As an example, consider the following:
python do_foo_prepend() {
bb.plain("first")
}
python do_foo() {
bb.plain("second")
}
python do_foo_append() {
bb.plain("third")
}
Running do_foo
prints the following:
recipename do_foo: first
recipename do_foo: second
recipename do_foo: third
You can use the bitbake -e
recipename command to view
the final assembled function after all overrides have been applied.
3.5.3 Python Functions
These functions are written in Python and are executed by other Python code. Examples of Python functions are utility functions that you intend to call from in-line Python or from within other Python functions. Here is an example:
def get_depends(d):
if d.getVar('SOMECONDITION'):
return "dependencywithcond"
else:
return "dependency"
SOMECONDITION = "1"
DEPENDS = "${@get_depends(d)}"
This would result in DEPENDS
containing dependencywithcond
.
Here are some things to know about Python functions:
Python functions can take parameters.
The BitBake datastore is not automatically available. Consequently, you must pass it in as a parameter to the function.
The “bb” and “os” Python modules are automatically available. You do not need to import them.
3.5.4 BitBake-Style Python Functions Versus Python Functions
Following are some important differences between BitBake-style Python functions and regular Python functions defined with “def”:
Only BitBake-style Python functions can be tasks.
Overrides and override-style operators can only be applied to BitBake-style Python functions.
Only regular Python functions can take arguments and return values.
Variable flags such as
[dirs]
,[cleandirs]
, and[lockfiles]
can be used on BitBake-style Python functions, but not on regular Python functions.BitBake-style Python functions generate a separate
${
T}/run.
function-name.
pid script that is executed to run the function, and also generate a log file in${T}/log.
function-name.
pid if they are executed as tasks.Regular Python functions execute “inline” and do not generate any files in
${T}
.Regular Python functions are called with the usual Python syntax. BitBake-style Python functions are usually tasks and are called directly by BitBake, but can also be called manually from Python code by using the
bb.build.exec_func()
function. Here is an example:bb.build.exec_func("my_bitbake_style_function", d)
Note
bb.build.exec_func()
can also be used to run shell functions from Python code. If you want to run a shell function before a Python function within the same task, then you can use a parent helper Python function that starts by running the shell function withbb.build.exec_func()
and then runs the Python code.To detect errors from functions executed with
bb.build.exec_func()
, you can catch thebb.build.FuncFailed
exception.Note
Functions in metadata (recipes and classes) should not themselves raise
bb.build.FuncFailed
. Rather,bb.build.FuncFailed
should be viewed as a general indicator that the called function failed by raising an exception. For example, an exception raised bybb.fatal()
will be caught insidebb.build.exec_func()
, and abb.build.FuncFailed
will be raised in response.
Due to their simplicity, you should prefer regular Python functions over
BitBake-style Python functions unless you need a feature specific to
BitBake-style Python functions. Regular Python functions in metadata are
a more recent invention than BitBake-style Python functions, and older
code tends to use bb.build.exec_func()
more often.
3.5.5 Anonymous Python Functions
Sometimes it is useful to set variables or perform other operations programmatically during parsing. To do this, you can define special Python functions, called anonymous Python functions, that run at the end of parsing. For example, the following conditionally sets a variable based on the value of another variable:
python () {
if d.getVar('SOMEVAR') == 'value':
d.setVar('ANOTHERVAR', 'value2')
}
An equivalent way to mark a function as an anonymous function is to give it the name “__anonymous”, rather than no name.
Anonymous Python functions always run at the end of parsing, regardless of where they are defined. If a recipe contains many anonymous functions, they run in the same order as they are defined within the recipe. As an example, consider the following snippet:
python () {
d.setVar('FOO', 'foo 2')
}
FOO = "foo 1"
python () {
d.appendVar('BAR',' bar 2')
}
BAR = "bar 1"
The previous example is conceptually equivalent to the following snippet:
FOO = "foo 1"
BAR = "bar 1"
FOO = "foo 2"
BAR += "bar 2"
FOO
ends up with the value “foo 2”, and
BAR
with the value “bar 1 bar 2”. Just as in the second snippet, the
values set for the variables within the anonymous functions become
available to tasks, which always run after parsing.
Overrides and override-style operators such as “_append
” are applied
before anonymous functions run. In the following example, FOO
ends
up with the value “foo from anonymous”:
FOO = "foo"
FOO_append = " from outside"
python () {
d.setVar("FOO", "foo from anonymous")
}
For methods you can use with anonymous Python functions, see the “Functions You Can Call From Within Python” section. For a different method to run Python code during parsing, see the “Inline Python Variable Expansion” section.
3.5.6 Flexible Inheritance for Class Functions
Through coding techniques and the use of EXPORT_FUNCTIONS
, BitBake
supports exporting a function from a class such that the class function
appears as the default implementation of the function, but can still be
called if a recipe inheriting the class needs to define its own version
of the function.
To understand the benefits of this feature, consider the basic scenario
where a class defines a task function and your recipe inherits the
class. In this basic scenario, your recipe inherits the task function as
defined in the class. If desired, your recipe can add to the start and
end of the function by using the “_prepend” or “_append” operations
respectively, or it can redefine the function completely. However, if it
redefines the function, there is no means for it to call the class
version of the function. EXPORT_FUNCTIONS
provides a mechanism that
enables the recipe’s version of the function to call the original
version of the function.
To make use of this technique, you need the following things in place:
The class needs to define the function as follows:
classname_functionname
For example, if you have a class file
bar.bbclass
and a function nameddo_foo
, the class must define the function as follows:bar_do_foo
The class needs to contain the
EXPORT_FUNCTIONS
statement as follows:EXPORT_FUNCTIONS functionname
For example, continuing with the same example, the statement in the
bar.bbclass
would be as follows:EXPORT_FUNCTIONS do_foo
You need to call the function appropriately from within your recipe. Continuing with the same example, if your recipe needs to call the class version of the function, it should call
bar_do_foo
. Assumingdo_foo
was a shell function andEXPORT_FUNCTIONS
was used as above, the recipe’s function could conditionally call the class version of the function as follows:do_foo() { if [ somecondition ] ; then bar_do_foo else # Do something else fi }
To call your modified version of the function as defined in your recipe, call it as
do_foo
.
With these conditions met, your single recipe can freely choose between the original function as defined in the class file and the modified function in your recipe. If you do not set up these conditions, you are limited to using one function or the other.
3.6 Tasks
Tasks are BitBake execution units that make up the steps that BitBake
can run for a given recipe. Tasks are only supported in recipes and
classes (i.e. in .bb
files and files included or inherited from
.bb
files). By convention, tasks have names that start with “do_”.
3.6.1 Promoting a Function to a Task
Tasks are either shell functions or
BitBake-style Python functions
that have been promoted to tasks by using the addtask
command. The
addtask
command can also optionally describe dependencies between
the task and other tasks. Here is an example that shows how to define a
task and declare some dependencies:
python do_printdate () {
import time
print time.strftime('%Y%m%d', time.gmtime())
}
addtask printdate after do_fetch before do_build
The first argument to addtask
is the name
of the function to promote to a task. If the name does not start with
“do_”, “do_” is implicitly added, which enforces the convention that all
task names start with “do_”.
In the previous example, the do_printdate
task becomes a dependency
of the do_build
task, which is the default task (i.e. the task run
by the bitbake
command unless another task is specified explicitly).
Additionally, the do_printdate
task becomes dependent upon the
do_fetch
task. Running the do_build
task results in the
do_printdate
task running first.
Note
If you try out the previous example, you might see that the
do_printdate
task is only run the first time you build the recipe with the
bitbake
command. This is because BitBake considers the task “up-to-date”
after that initial run. If you want to force the task to always be
rerun for experimentation purposes, you can make BitBake always
consider the task “out-of-date” by using the
[nostamp]
variable flag, as follows:
do_printdate[nostamp] = "1"
You can also explicitly run the task and provide the -f option as follows:
$ bitbake recipe -c printdate -f
When manually selecting a task to run with the bitbake recipe
-c task
command, you can omit the “do_” prefix as part of the task
name.
You might wonder about the practical effects of using addtask
without specifying any dependencies as is done in the following example:
addtask printdate
In this example, assuming dependencies have not been
added through some other means, the only way to run the task is by
explicitly selecting it with bitbake
recipe -c printdate
. You
can use the do_listtasks
task to list all tasks defined in a recipe
as shown in the following example:
$ bitbake recipe -c listtasks
For more information on task dependencies, see the “Dependencies” section.
See the “Variable Flags” section for information on variable flags you can use with tasks.
Note
While it’s infrequent, it’s possible to define multiple tasks as
dependencies when calling addtask
. For example, here’s a snippet
from the OpenEmbedded class file package_tar.bbclass
:
addtask package_write_tar before do_build after do_packagedata do_package
Note how the package_write_tar
task has to wait until both of
do_packagedata
and do_package
complete.
3.6.2 Deleting a Task
As well as being able to add tasks, you can delete them. Simply use the
deltask
command to delete a task. For example, to delete the example
task used in the previous sections, you would use:
deltask printdate
If you delete a task using the deltask
command and the task has
dependencies, the dependencies are not reconnected. For example, suppose
you have three tasks named do_a
, do_b
, and do_c
.
Furthermore, do_c
is dependent on do_b
, which in turn is
dependent on do_a
. Given this scenario, if you use deltask
to
delete do_b
, the implicit dependency relationship between do_c
and do_a
through do_b
no longer exists, and do_c
dependencies are not updated to include do_a
. Thus, do_c
is free
to run before do_a
.
If you want dependencies such as these to remain intact, use the
[noexec]
varflag to disable the task instead of using the
deltask
command to delete it:
do_b[noexec] = "1"
3.6.3 Passing Information Into the Build Task Environment
When running a task, BitBake tightly controls the shell execution environment of the build tasks to make sure unwanted contamination from the build machine cannot influence the build.
Note
By default, BitBake cleans the environment to include only those things exported or listed in its whitelist to ensure that the build environment is reproducible and consistent. You can prevent this “cleaning” by setting the BB_PRESERVE_ENV variable.
Consequently, if you do want something to get passed into the build task environment, you must take these two steps:
Tell BitBake to load what you want from the environment into the datastore. You can do so through the BB_ENV_WHITELIST and BB_ENV_EXTRAWHITE variables. For example, assume you want to prevent the build system from accessing your
$HOME/.ccache
directory. The following command “whitelists” the environment variableCCACHE_DIR
causing BitBake to allow that variable into the datastore:export BB_ENV_EXTRAWHITE="$BB_ENV_EXTRAWHITE CCACHE_DIR"
Tell BitBake to export what you have loaded into the datastore to the task environment of every running task. Loading something from the environment into the datastore (previous step) only makes it available in the datastore. To export it to the task environment of every running task, use a command similar to the following in your local configuration file
local.conf
or your distribution configuration file:export CCACHE_DIR
Note
A side effect of the previous steps is that BitBake records the variable as a dependency of the build process in things like the setscene checksums. If doing so results in unnecessary rebuilds of tasks, you can whitelist the variable so that the setscene code ignores the dependency when it creates checksums.
Sometimes, it is useful to be able to obtain information from the original execution environment. BitBake saves a copy of the original environment into a special variable named BB_ORIGENV.
The BB_ORIGENV
variable returns a datastore object that can be
queried using the standard datastore operators such as
getVar(, False)
. The datastore object is useful, for example, to
find the original DISPLAY
variable. Here is an example:
origenv = d.getVar("BB_ORIGENV", False)
bar = origenv.getVar("BAR", False)
The previous example returns BAR
from the original execution
environment.
3.7 Variable Flags
Variable flags (varflags) help control a task’s functionality and dependencies. BitBake reads and writes varflags to the datastore using the following command forms:
variable = d.getVarFlags("variable")
self.d.setVarFlags("FOO", {"func": True})
When working with varflags, the same syntax, with the exception of overrides, applies. In other words, you can set, append, and prepend varflags just like variables. See the “Variable Flag Syntax” section for details.
BitBake has a defined set of varflags available for recipes and classes. Tasks support a number of these flags which control various functionality of the task:
[cleandirs]
: Empty directories that should be created before the task runs. Directories that already exist are removed and recreated to empty them.[depends]
: Controls inter-task dependencies. See the DEPENDS variable and the “Inter-Task Dependencies” section for more information.[deptask]
: Controls task build-time dependencies. See the DEPENDS variable and the “Build Dependencies” section for more information.[dirs]
: Directories that should be created before the task runs. Directories that already exist are left as is. The last directory listed is used as the current working directory for the task.[lockfiles]
: Specifies one or more lockfiles to lock while the task executes. Only one task may hold a lockfile, and any task that attempts to lock an already locked file will block until the lock is released. You can use this variable flag to accomplish mutual exclusion.[noexec]
: When set to “1”, marks the task as being empty, with no execution required. You can use the[noexec]
flag to set up tasks as dependency placeholders, or to disable tasks defined elsewhere that are not needed in a particular recipe.[nostamp]
: When set to “1”, tells BitBake to not generate a stamp file for a task, which implies the task should always be executed.Caution
Any task that depends (possibly indirectly) on a
[nostamp]
task will always be executed as well. This can cause unnecessary rebuilding if you are not careful.[number_threads]
: Limits tasks to a specific number of simultaneous threads during execution. This varflag is useful when your build host has a large number of cores but certain tasks need to be rate-limited due to various kinds of resource constraints (e.g. to avoid network throttling).number_threads
works similarly to the BB_NUMBER_THREADS variable but is task-specific.Set the value globally. For example, the following makes sure the
do_fetch
task uses no more than two simultaneous execution threads: do_fetch[number_threads] = “2”Warning
Setting the varflag in individual recipes rather than globally can result in unpredictable behavior.
Setting the varflag to a value greater than the value used in the
BB_NUMBER_THREADS
variable causesnumber_threads
to have no effect.
[postfuncs]
: List of functions to call after the completion of the task.[prefuncs]
: List of functions to call before the task executes.[rdepends]
: Controls inter-task runtime dependencies. See the RDEPENDS variable, the RRECOMMENDS variable, and the “Inter-Task Dependencies” section for more information.[rdeptask]
: Controls task runtime dependencies. See the RDEPENDS variable, the RRECOMMENDS variable, and the “Runtime Dependencies” section for more information.[recideptask]
: When set in conjunction withrecrdeptask
, specifies a task that should be inspected for additional dependencies.[recrdeptask]
: Controls task recursive runtime dependencies. See the RDEPENDS variable, the RRECOMMENDS variable, and the “Recursive Dependencies” section for more information.[stamp-extra-info]
: Extra stamp information to append to the task’s stamp. As an example, OpenEmbedded uses this flag to allow machine-specific tasks.[umask]
: The umask to run the task under.
Several varflags are useful for controlling how signatures are calculated for variables. For more information on this process, see the “Checksums (Signatures)” section.
[vardeps]
: Specifies a space-separated list of additional variables to add to a variable’s dependencies for the purposes of calculating its signature. Adding variables to this list is useful, for example, when a function refers to a variable in a manner that does not allow BitBake to automatically determine that the variable is referred to.[vardepsexclude]
: Specifies a space-separated list of variables that should be excluded from a variable’s dependencies for the purposes of calculating its signature.[vardepvalue]
: If set, instructs BitBake to ignore the actual value of the variable and instead use the specified value when calculating the variable’s signature.[vardepvalueexclude]
: Specifies a pipe-separated list of strings to exclude from the variable’s value when calculating the variable’s signature.
3.8 Events
BitBake allows installation of event handlers within recipe and class
files. Events are triggered at certain points during operation, such as
the beginning of operation against a given recipe (i.e. *.bb
), the
start of a given task, a task failure, a task success, and so forth. The
intent is to make it easy to do things like email notification on build
failures.
Following is an example event handler that prints the name of the event
and the content of the FILE
variable:
addhandler myclass_eventhandler
python myclass_eventhandler() {
from bb.event import getName
print("The name of the Event is %s" % getName(e))
print("The file we run for is %s" % d.getVar('FILE'))
}
myclass_eventhandler[eventmask] = "bb.event.BuildStarted
bb.event.BuildCompleted"
In the previous example, an eventmask has been
set so that the handler only sees the “BuildStarted” and
“BuildCompleted” events. This event handler gets called every time an
event matching the eventmask is triggered. A global variable “e” is
defined, which represents the current event. With the getName(e)
method, you can get the name of the triggered event. The global
datastore is available as “d”. In legacy code, you might see “e.data”
used to get the datastore. However, realize that “e.data” is deprecated
and you should use “d” going forward.
The context of the datastore is appropriate to the event in question. For example, “BuildStarted” and “BuildCompleted” events run before any tasks are executed so would be in the global configuration datastore namespace. No recipe-specific metadata exists in that namespace. The “BuildStarted” and “BuildCompleted” events also run in the main cooker/server process rather than any worker context. Thus, any changes made to the datastore would be seen by other cooker/server events within the current build but not seen outside of that build or in any worker context. Task events run in the actual tasks in question consequently have recipe-specific and task-specific contents. These events run in the worker context and are discarded at the end of task execution.
During a standard build, the following common events might occur. The following events are the most common kinds of events that most metadata might have an interest in viewing:
bb.event.ConfigParsed()
: Fired when the base configuration; which consists ofbitbake.conf
,base.bbclass
and any globalINHERIT
statements; has been parsed. You can see multiple such events when each of the workers parse the base configuration or if the server changes configuration and reparses. Any given datastore only has one such event executed against it, however. If BB_INVALIDCONF is set in the datastore by the event handler, the configuration is reparsed and a new event triggered, allowing the metadata to update configuration.bb.event.HeartbeatEvent()
: Fires at regular time intervals of one second. You can configure the interval time using theBB_HEARTBEAT_EVENT
variable. The event’s “time” attribute is thetime.time()
value when the event is triggered. This event is useful for activities such as system state monitoring.bb.event.ParseStarted()
: Fired when BitBake is about to start parsing recipes. This event’s “total” attribute represents the number of recipes BitBake plans to parse.bb.event.ParseProgress()
: Fired as parsing progresses. This event’s “current” attribute is the number of recipes parsed as well as the “total” attribute.bb.event.ParseCompleted()
: Fired when parsing is complete. This event’s “cached”, “parsed”, “skipped”, “virtuals”, “masked”, and “errors” attributes provide statistics for the parsing results.bb.event.BuildStarted()
: Fired when a new build starts. BitBake fires multiple “BuildStarted” events (one per configuration) when multiple configuration (multiconfig) is enabled.bb.build.TaskStarted()
: Fired when a task starts. This event’s “taskfile” attribute points to the recipe from which the task originates. The “taskname” attribute, which is the task’s name, includes thedo_
prefix, and the “logfile” attribute point to where the task’s output is stored. Finally, the “time” attribute is the task’s execution start time.bb.build.TaskInvalid()
: Fired if BitBake tries to execute a task that does not exist.bb.build.TaskFailedSilent()
: Fired for setscene tasks that fail and should not be presented to the user verbosely.bb.build.TaskFailed()
: Fired for normal tasks that fail.bb.build.TaskSucceeded()
: Fired when a task successfully completes.bb.event.BuildCompleted()
: Fired when a build finishes.bb.cooker.CookerExit()
: Fired when the BitBake server/cooker shuts down. This event is usually only seen by the UIs as a sign they should also shutdown.
This next list of example events occur based on specific requests to the server. These events are often used to communicate larger pieces of information from the BitBake server to other parts of BitBake such as user interfaces:
bb.event.TreeDataPreparationStarted()
bb.event.TreeDataPreparationProgress()
bb.event.TreeDataPreparationCompleted()
bb.event.DepTreeGenerated()
bb.event.CoreBaseFilesFound()
bb.event.ConfigFilePathFound()
bb.event.FilesMatchingFound()
bb.event.ConfigFilesFound()
bb.event.TargetsTreeGenerated()
3.9 Variants - Class Extension Mechanism
BitBake supports two features that facilitate creating from a single recipe file multiple incarnations of that recipe file where all incarnations are buildable. These features are enabled through the BBCLASSEXTEND and BBVERSIONS variables.
Note
The mechanism for this class extension is extremely specific to the implementation. Usually, the recipe’s PROVIDES , PN , and DEPENDS variables would need to be modified by the extension class. For specific examples, see the OE-Core native , nativesdk , and multilib classes.
BBCLASSEXTEND
: This variable is a space separated list of classes used to “extend” the recipe for each variant. Here is an example that results in a second incarnation of the current recipe being available. This second incarnation will have the “native” class inherited.BBCLASSEXTEND = "native"
BBVERSIONS
: This variable allows a single recipe to build multiple versions of a project from a single recipe file. You can also specify conditional metadata (using the OVERRIDES mechanism) for a single version, or an optionally named range of versions. Here is an example:BBVERSIONS = "1.0 2.0 git" SRC_URI_git = "git://someurl/somepath.git" BBVERSIONS = "1.0.[0-6]:1.0.0+ 1.0.[7-9]:1.0.7+" SRC_URI_append_1.0.7+ = "file://some_patch_which_the_new_versions_need.patch;patch=1"
The name of the range defaults to the original version of the recipe. For example, in OpenEmbedded, the recipe file
foo_1.0.0+.bb
creates a default name range of1.0.0+
. This is useful because the range name is not only placed into overrides, but it is also made available for the metadata to use in the variable that defines the base recipe versions for use infile://
search paths (FILESPATH).
3.10 Dependencies
To allow for efficient parallel processing, BitBake handles dependencies at the task level. Dependencies can exist both between tasks within a single recipe and between tasks in different recipes. Following are examples of each:
For tasks within a single recipe, a recipe’s
do_configure
task might need to complete before itsdo_compile
task can run.For tasks in different recipes, one recipe’s
do_configure
task might require another recipe’sdo_populate_sysroot
task to finish first such that the libraries and headers provided by the other recipe are available.
This section describes several ways to declare dependencies. Remember, even though dependencies are declared in different ways, they are all simply dependencies between tasks.
3.10.1 Dependencies Internal to the .bb
File
BitBake uses the addtask
directive to manage dependencies that are
internal to a given recipe file. You can use the addtask
directive
to indicate when a task is dependent on other tasks or when other tasks
depend on that recipe. Here is an example:
addtask printdate after do_fetch before do_build
In this example, the do_printdate
task
depends on the completion of the do_fetch
task, and the do_build
task depends on the completion of the do_printdate
task.
Note
For a task to run, it must be a direct or indirect dependency of some other task that is scheduled to run.
For illustration, here are some examples:
The directive
addtask mytask before do_configure
causesdo_mytask
to run beforedo_configure
runs. Be aware thatdo_mytask
still only runs if its input checksum has changed since the last time it was run. Changes to the input checksum ofdo_mytask
also indirectly causedo_configure
to run.The directive
addtask mytask after do_configure
by itself never causesdo_mytask
to run.do_mytask
can still be run manually as follows:$ bitbake recipe -c mytask
Declaring
do_mytask
as a dependency of some other task that is scheduled to run also causes it to run. Regardless, the task runs afterdo_configure
.
3.10.2 Build Dependencies
BitBake uses the DEPENDS variable to manage
build time dependencies. The [deptask]
varflag for tasks signifies
the task of each item listed in DEPENDS
that must complete before
that task can be executed. Here is an example:
do_configure[deptask] = "do_populate_sysroot"
In this example, the do_populate_sysroot
task
of each item in DEPENDS
must complete before do_configure
can
execute.
3.10.3 Runtime Dependencies
BitBake uses the PACKAGES, RDEPENDS, and RRECOMMENDS variables to manage runtime dependencies.
The PACKAGES
variable lists runtime packages. Each of those packages
can have RDEPENDS
and RRECOMMENDS
runtime dependencies. The
[rdeptask]
flag for tasks is used to signify the task of each item
runtime dependency which must have completed before that task can be
executed.
do_package_qa[rdeptask] = "do_packagedata"
In the previous
example, the do_packagedata
task of each item in RDEPENDS
must
have completed before do_package_qa
can execute.
Although RDEPENDS
contains entries from the
runtime dependency namespace, BitBake knows how to map them back
to the build-time dependency namespace, in which the tasks are defined.
3.10.4 Recursive Dependencies
BitBake uses the [recrdeptask]
flag to manage recursive task
dependencies. BitBake looks through the build-time and runtime
dependencies of the current recipe, looks through the task’s inter-task
dependencies, and then adds dependencies for the listed task. Once
BitBake has accomplished this, it recursively works through the
dependencies of those tasks. Iterative passes continue until all
dependencies are discovered and added.
The [recrdeptask]
flag is most commonly used in high-level recipes
that need to wait for some task to finish “globally”. For example,
image.bbclass
has the following:
do_rootfs[recrdeptask] += "do_packagedata"
This statement says that the do_packagedata
task of
the current recipe and all recipes reachable (by way of dependencies)
from the image recipe must run before the do_rootfs
task can run.
BitBake allows a task to recursively depend on itself by referencing itself in the task list:
do_a[recrdeptask] = "do_a do_b"
In the same way as before, this means that the do_a
and do_b
tasks of the current recipe and all
recipes reachable (by way of dependencies) from the recipe
must run before the do_a
task can run. In this
case BitBake will ignore the current recipe’s do_a
task circular dependency on itself.
3.10.5 Inter-Task Dependencies
BitBake uses the [depends]
flag in a more generic form to manage
inter-task dependencies. This more generic form allows for
inter-dependency checks for specific tasks rather than checks for the
data in DEPENDS
. Here is an example:
do_patch[depends] = "quilt-native:do_populate_sysroot"
In this example, the do_populate_sysroot
task of the target quilt-native
must have completed before the do_patch
task can execute.
The [rdepends]
flag works in a similar way but takes targets in the
runtime namespace instead of the build-time dependency namespace.
3.11 Functions You Can Call From Within Python
BitBake provides many functions you can call from within Python functions. This section lists the most commonly used functions, and mentions where to find others.
3.11.1 Functions for Accessing Datastore Variables
It is often necessary to access variables in the BitBake datastore using Python functions. The BitBake datastore has an API that allows you this access. Here is a list of available operations:
Operation |
Description |
---|---|
|
Returns the value of variable “X”. Using “expand=True” expands the value. Returns “None” if the variable “X” does not exist. |
|
Sets the variable “X” to “value” |
|
Adds “value” to the end of the variable “X”. Acts like |
|
Adds “value” to the start of the variable “X”. Acts like
|
|
Deletes the variable “X” from the datastore. Does nothing if the variable “X” does not exist. |
|
Renames the variable “X” to “Y”. Does nothing if the variable “X” does not exist. |
|
Returns the value of variable “X”. Using “expand=True” expands the value. Returns “None” if either the variable “X” or the named flag does not exist. |
|
Sets the named flag for variable “X” to “value”. |
|
Appends “value” to the named flag on the variable “X”. Acts like
|
|
Prepends “value” to the named flag on the variable “X”. Acts like
|
|
Deletes the named flag on the variable “X” from the datastore. |
|
Sets the flags specified in the |
|
Returns a |
|
Deletes all the flags for the variable “X”. Does nothing if the variable “X” does not exist. |
|
Expands variable references in the specified string
expression. References to variables that do not exist are left as is. For
example, |
3.11.2 Other Functions
You can find many other functions that can be called from Python by
looking at the source code of the bb
module, which is in
bitbake/lib/bb
. For example, bitbake/lib/bb/utils.py
includes
the commonly used functions bb.utils.contains()
and
bb.utils.mkdirhier()
, which come with docstrings.
3.12 Task Checksums and Setscene
BitBake uses checksums (or signatures) along with the setscene to determine if a task needs to be run. This section describes the process. To help understand how BitBake does this, the section assumes an OpenEmbedded metadata-based example.
These checksums are stored in STAMP. You can examine the checksums using the following BitBake command:
$ bitbake-dumpsigs
This command returns the signature data in a readable
format that allows you to examine the inputs used when the OpenEmbedded
build system generates signatures. For example, using
bitbake-dumpsigs
allows you to examine the do_compile
task’s
“sigdata” for a C application (e.g. bash
). Running the command also
reveals that the “CC” variable is part of the inputs that are hashed.
Any changes to this variable would invalidate the stamp and cause the
do_compile
task to run.
The following list describes related variables:
BB_HASHCHECK_FUNCTION: Specifies the name of the function to call during the “setscene” part of the task’s execution in order to validate the list of task hashes.
BB_SETSCENE_DEPVALID: Specifies a function BitBake calls that determines whether BitBake requires a setscene dependency to be met.
BB_SETSCENE_VERIFY_FUNCTION2: Specifies a function to call that verifies the list of planned task execution before the main task execution happens.
BB_STAMP_POLICY: Defines the mode for comparing timestamps of stamp files.
BB_STAMP_WHITELIST: Lists stamp files that are looked at when the stamp policy is “whitelist”.
BB_TASKHASH: Within an executing task, this variable holds the hash of the task as returned by the currently enabled signature generator.
STAMP: The base path to create stamp files.
STAMPCLEAN: Again, the base path to create stamp files but can use wildcards for matching a range of files for clean operations.
3.13 Wildcard Support in Variables
Support for wildcard use in variables varies depending on the context in
which it is used. For example, some variables and file names allow
limited use of wildcards through the “%
” and “*
” characters.
Other variables or names support Python’s
glob syntax,
fnmatch
syntax, or
Regular Expression (re)
syntax.
For variables that have wildcard suport, the documentation describes which form of wildcard, its use, and its limitations.