This script finds cases of missing locks in the code.
There is some possibilty of false positives in cases
where a particular function is supposed to exit with
the lock held or there is any preceding function call
that releases the lock.
Suggested-by: Julia Lawall <julia.lawall@lip6.fr>
Signed-off-by: Himanshu Jha <himanshujha199640@gmail.com>
* Provide a single liner description at top using `///`
comments. This one liner is displayed when using `--verbose`
mode of coccicheck.
* Specify Condidence level property adhering to coccinelle.rst
section "Proposing new semantic patches".
* Add virtual patch rule to handle the patch case when coccicheck
is supplied with `--mode=patch` option.
* Add `depends on !(file in "ext")` to ignore reports from `ext/`
directory.
* Simplify rule to use disjunctions and use "exists" to match any
available control path.
Suggested-by: Julia Lawall <julia.lawall@lip6.fr>
Signed-off-by: Himanshu Jha <himanshujha199640@gmail.com>
* Provide a single liner description at top using `///`
comments. This one liner is displayed when using `--verbose`
mode of coccicheck.
* Specify Condidence level property adhering to coccinelle.rst
section "Proposing new semantic patches".
* Add virtual patch rule to handle the patch case when coccicheck
is supplied with `--mode=patch` option.
* Add `depends on !(file in "ext")` to ignore reports from `ext/`
directory.
* Simplify patch rule to reduce effort in reconstruction since
logically we are only *adding* void cast.
Signed-off-by: Himanshu Jha <himanshujha199640@gmail.com>
This script finds cases of NULL dereferences where a variable
is dereferenced under a NULL test, even though it is known
to be NULL.
Signed-off-by: Himanshu Jha <himanshujha199640@gmail.com>
* Use `expression` metavariable instead of `constant` to catch
missing cases such as:
diff -u -p a/subsys/net/ip/rpl.c b/subsys/net/ip/rpl.c
--- a/subsys/net/ip/rpl.c
+++ b/subsys/net/ip/rpl.c
@@ -686,7 +686,7 @@ static void new_dio_interval(struct net_
{
u32_t time;
- time = 1 << instance->dio_interval_current;
+ time = BIT(instance->dio_interval_current);
* Provide a single liner description at top using `///`
comments. This one liner is displayed when using `--verbose`
mode of coccicheck.
* Specify Condidence level property adhering to coccinelle.rst
section "Proposing new semantic patches".
* Add virtual patch rule to handle the patch case when coccicheck
is supplied with `--mode=patch` option.
* Edit patch rule to remove redundant parentheses in the output.
* Add `depends on !(file in "ext")` to ignore reports from `ext/`
directory.
Signed-off-by: Himanshu Jha <himanshujha199640@gmail.com>
Allow directly calling the flash partition extraction function and
remove passing any arguments that the function doesn't actually use.
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
Reworked flash partition extraction so we call it when we see a
node that is 'partiton@' instead of per property. We also handle the
'read-only' property in the flash partition extraction function which
lets us remove 'use-property-label' handling.
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
We don't have any cases that need 'use-property-label' related to
handling of reg extraction so lets remove it.
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
Add support to the extract script to handle 'pwms' client properties.
This is pretty much identical to how we hand 'gpios'.
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
New Qemu is emitting binary charachters which is disturbing generation
of XML, so make sure we only log printable charachters.
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
Update menuconfig and Kconfiglib to upstream revision 68426efeae6aa, to
add two minor menuconfig improvements and a small bugfix:
- Expected errors, like syntax errors in Kconfig files, no longer
generate a Python backtrace. Just the error message is shown.
- Entering a choice now places the cursor on the selected symbol, if
any.
- Having two consecutive empty 'if's (which could also appear e.g. due
to osource) crashed the menuconfig when entering show-all mode, due
to the 'if' removal logic in Kconfiglib failing to remove one of
them. All configurations of 'if's are now correctly removed.
This error was found while reading the code.
Some minor optimizations and some internal cleanup is included as well.
Signed-off-by: Ulf Magnusson <Ulf.Magnusson@nordicsemi.no>
Before dtc 1.4.7 we'd get something like the following for an gpios
property:
reg = <1 2 3 4>;
After dtc 1.4.7 we get:
cs-gpios = <0x05 0x0d 0x00>, < 0x06 0x00 0x00>;
We should handle both cases in the extract reg handling code. So if
we see a list of lists, we flatten it to a single list to normalize
the property.
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
We first check to see if we have a valid elf file before doing anything
else. Otherwise we'd could get script errors instead of just notifying
the user the file doesn't exist.
Signed-off-by: Kumar Gala <kumar.gala@gmail.com>
We need a minimum of --outdir, --objdump, and --nm set for the script to
work so mark these arguments as required.
Signed-off-by: Kumar Gala <kumar.gala@gmail.com>
Do a straight forward conversion from optparse to argparser. All
of our other python scripts use argparser already.
Signed-off-by: Kumar Gala <kumar.gala@gmail.com>
Update menuconfig (and Kconfiglib, just to sync) to upstream revision
bf1701b36634b, to add this commit:
menuconfig: Add jump-to for choices, menus, and comments
For choices, search the name and the prompt. This is the same as for
symbols, except names are optional (and rare) for choices.
For menus and comments, search the prompt (title text).
When jumping to a non-empty choice or menu, jump into it instead of
jumping to its menu node. If show-all mode is off and there are
visible items in the choice/menu, then jump to the first visible
node. Otherwise, enable show-all and jump to the first node.
Previously, only symbols could be jumped to.
Various other small fixes/improvements are included too:
- The "no range constraints" text was dropped from the input dialog
when settings int/hex symbols without an active 'range'. It might be
more confusing than helpful.
- A crash when pressing Ctrl-F (the view-help shortcut) with no matches
in the jump-to dialog was fixed
- Some gnome-terminal shoddiness was worked around to remove minor
jumpiness when reducing the height of the terminal
Signed-off-by: Ulf Magnusson <Ulf.Magnusson@nordicsemi.no>
Compat define is generated without value. This should not harm,
but since generated flag is a "Kconfig like" define, we migth need it
to hold a value.
Signed-off-by: Erwan Gouriou <erwan.gouriou@linaro.org>
Heuristic for populating an alias string for a subnode
of an alias node was wrong as there was no check if
alias already existed and string created was not in line
with device tree specfication (allowed chars: [0-9],[a-z], '-').
Fix these two points.
Signed-off-by: Erwan Gouriou <erwan.gouriou@linaro.org>
We can replace the method get_node_label_string in DTDirective with the
global function get_node_label. This way we have a single place that
labels for defines are being generated.
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
Don't based the compat to get_node_label, we can just figure out based
on the node address we are passed.
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
This puts the priviledged stack at the end of RAM.
This combines PR #10507 and #10542.
Fixes#10473Fixes#10474Fixes#10515
Signed-off-by: Adithya Baglody <adithya.nagaraj.baglody@intel.com>
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
Standard device tree properties like 'reg-names', 'interrupt-names', and
'phandle' should be handled by other extract mechanisms or ignored. So
filter them out before we even call extract_property
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
Some boards depend on environment variables so we want to make sure we
do not attempt to build boards requiring additional setup.
Add the section below into the board YAML file, sanitycheck will check
the environment and will only run tests on that board if the variables
are defined.
env:
- VAR1
- VAR2
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
We've never handled boolean properties well, if we had a boolean that
existed we'd generate a define of 'True' and if the boolean didn't exist
than nothing would be generated. So there was no easy way to tell that
the boolean wasn't defined.
Now if we mark a property as boolean in the yaml, we will generate a
define for it regardless if it exists or not. If it exists we'll
set the value to 1, and if it doesn't to 0.
Fixes#8376
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
Add a function that given a node address we get a list of compats back
instead of just the first. This is in prep for eDTS support.
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
Before dtc 1.4.7 we'd get something like the following for an reg
property:
reg = <1 2 3 4>;
After dtc 1.4.7 we get:
reg = <1 2>, <3 4>;
We should handle both cases in the extract reg handling code. So if
we see a list of lists, we flatten it to a single list to normalize
the property.
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
Python's open() by default using "system character encoding", which
may vary from system to system (e.g. LOCALE=C aka "ascii" on bare-OS
systems, something on Windows, etc.). But Zephyr files are utf-8, so
read them as such, by explicitly specifying encoding.
This is similar to changes earlier done to another script in 94620bd.
Signed-off-by: Paul Sokolovsky <paul.sokolovsky@linaro.org>
To be able to collect all coverage reports from the "native"
applications in the first CI instance, let's also place the
nrf52_bsim testcases first. The order now is:
unit_testing
native_posix
*_bsim
anything else
Signed-off-by: Alberto Escolar Piedras <alpi@oticon.com>
run tests based on sub-testcases, for example:
sanitycheck --sub-test net.app.app_tcp4_client_hostname_fail
will run all the tests in tests/net/app/net.app.
Useful for re-running tests that are reported in testrail and in reports
generated by sanitycheck.
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
Do not check for board/toolchain combination and if a certain toolchain
is configured for a board, run with whatever toolchain we have.
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
CMake will translate from the toolchain variant 'gccarmemb' to
'gnuarmemb', but sanitycheck does not. This causes inconsistent and
therefore confusing behaviour between CMake and sanitycheck.
Until gccarmemb is dropped support for, do the same translation with
sanitycheck.
Signed-off-by: Sebastian Bøe <sebastian.boe@nordicsemi.no>
scripts/dts_sanity/edts_ref.json_save has been pushed by error
in commmit a429104095
Clean it up.
Signed-off-by: Erwan Gouriou <erwan.gouriou@linaro.org>
The 'id' field was never used and tended to just have the compat of the
node. Lets remove it and removed some code in extract_dts_includes.py
related to it. Added a warning if 'id' is set in a yaml so we can
remove it going forward.
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
Add a new option as fixture in harness configurations for utilizing
sanitycheck to identify test cases that require external hardware
such as sensor, ble, networking for validation. The config will be
added to yaml files with unique fixture name to identify each hardware
and allow automation to trigger test execution on setup having the
specific fixture enabled. Also, remove the default required for type and
regex configs that is not essential in case of ztest based test cases.
Signed-off-by: Praful Swarnakar <praful.swarnakar@intel.com>
Update menuconfig (and Kconfiglib, just to sync) to upstream revision
256e5b3e38e92 to get the fix below in, for an issue in an external
project. Also update genrest.py with a similar fix (the genrest issue
was what originally prompted it).
menuconfig: Improve/fix promptless choice handling
The code assumed that all parent (interface) menus always have a
prompt, which is false for items in promptless choices. This led to
a crash e.g. when viewing the symbol information for a symbol within
a promptless choice.
Promptless choices with children can show up "legitimately" when
people define choices in multiple locations to add symbols, though
this is broken in the C tools.
Use standard_sc_expr_str(node.item) instead of the non-existing
prompt for promptless choices. That way they show up as
'<choice (name if any>)>', which is consistent with how they're
shown elsewhere.
This commit also changes how choice names are displayed in
show-name/show-all mode, to the standard_sc_expr_str() format.
Signed-off-by: Ulf Magnusson <Ulf.Magnusson@nordicsemi.no>
Format all cube packages to the same format before update to new
versions:
-Apply dos2unix
-Remove trailing white spaces
Signed-off-by: Erwan Gouriou <erwan.gouriou@linaro.org>
For now we've used range properties as a pass through. However range
properties can translate from one address space to another. This is
typically used one address spaces translate from one physical bus to
another (For example going from PCI address space to internal SoC memory
map). However, we can also use this for cases where we want to reduce
duplication (For example with ARMv8-M for secure v non-secure MMIO
registers).
'ranges' takes either the form of:
ranges; /* pass through translation */
or:
ranges = <child-bus-address parent-bus-address length>;
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
We use to walk to the full tree in search of addr/size cells. We should
only have to look at the parent of the node for these properties.
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
Given the path address to a node return the address/size cells. In
theory this should just be looking up the #{address,size}-cells in
the parent node, we'll make that cleanup next.
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
In case of having more than one reg, aliases would not generate
properly. Number of register at the end of define was missing.
Signed-off-by: Mieszko Mierunski <mieszko.mierunski@nordicsemi.no>
This script detects cases where ARRAY_SIZE can be used such as
where there is a division of sizeof the array by the sizeof its first
element or by any indexed element or the element type. It replaces the
division of the two sizeofs by ARRAY_SIZE helper macro.
Signed-off-by: Himanshu Jha <himanshujha199640@gmail.com>
'coccicheck' target is used to initiate Coccinelle checker. It
can be called via four different modes that generate different
output corresponding the mode.
The four modes are:
* 'patch' proposes a fix, when possible.
* 'report' generates a list in the following format:
file:line:column-column: message
* 'context' highlights lines of interest and their context in
a diff-like style. Lines of interest are indicated with '-'.
* 'org' generates a report in the Org mode format of Emacs.
The 'coccicheck' front-end can be called with suitable arguments:
* --mode=<mode>: specify the mode for processsing.
* --cocci=<path/to/foo.cocci>: specify the SmPL file to use.
* --verbose=<1>: enable verbose output.
Run `./scripts/coccicheck --help` for more detailed info
on various options available.
With the above arguments the coccinelle transformation engine
runs on the entire/part of the source code tree depending on
various options supplied to the coccicheck sanity checker.
More detailed documentation can be found at:
doc/application/coccinelle.rst
Cc: Julia Lawall <Julia.Lawall@lip6.fr>
Signed-off-by: Himanshu Jha <himanshujha199640@gmail.com>
When the code partition is not at the beginning of the flash, the load
offset is wrongly computed. The address in the device tree is already
relative to the beginning of the parent node, ie the beginning of the
flash memory space. There is therefore no need to subtract it.
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
