Re-ordering the API names moved a trailing '|' causing the pattern to match
*every* #define in the input.
Signed-off-by: Keith Packard <keithp@keithp.com>
The `NEW_TYPEDEFS` rule consistently generates false
positives when implementing POSIX standard APIs.
It makes sense to disable this check for the POSIX haders
rather than requiring merge superpowers constantly. That
way, we can merge as per usual after sufficient approvals
rather than waiting for someone with merge superpowers to
override / manually merge.
Signed-off-by: Yong Cong Sin <ycsin@meta.com>
All code in the Zephyr core must use only the Zephyr C library API
according to rules A.4 and A.5. Such code is not permitted to request API
extensions from the C library via any of the API request mechanisms.
This addition to checkpatch.pl verifies that patches don't #define
any of these:
__STRICT_ANSI__
_POSIX_SOURCE
_POSIX_C_SOURCE
_XOPEN_SOURCE
_ISOC99_SOURCE
_ISOC11_SOURCE
_ATFILE_SOURCE
_GNU_SOURCE
_BSD_SOURCE
_SVID_SOURCE
_DEFAULT_SOURCE
Reference: #49922
Signed-off-by: Keith Packard <keithp@keithp.com>
The `USE_NEGATIVE_ERRNO` rule consistently generates false
positives when working with certain areas of POSIX.
It makes sense to disable this check for the POSIX API
rather than requiring merge superpowers constantly. That
way, we can merge as per usual after sufficient approvals
rather than waiting for someone with merge superpowers to
override / manually merge.
Signed-off-by: Christopher Friedt <cfriedt@meta.com>
The current docs are incorrect as comma separated paths
are not working. Providing a separate --exclude argement
per path does however work.
Signed-off-by: Håkon Øye Amundsen <haakon.amundsen@nordicsemi.no>
checkpatch identifies a label only when a terminating colon
immediately follows an identifier.
Bitfield definitions can appear to be labels so ignore any
spaces between the identifier terminating colon and any digit
that may be used to define a bitfield length.
Miscellanea:
o Improve the initial checkpatch comment
o Use the more typical '&&' instead of 'and'
o Require the initial label character to be a non-digit
(Can't use $Ident here because $Ident allows ## concatenation)
o Use $sline instead of $line to ignore comments
o Use '$sline !~ /.../' instead of '!($line =~ /.../)'
Link: https://lkml.kernel.org/r/b54d673e7cde7de5de0c9ba4dd57dd0858580ca4.camel@perches.com
Signed-off-by: Joe Perches <joe@perches.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Manikishan Ghantasala <manikishanghantasala@gmail.com>
Cc: Alex Elder <elder@ieee.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Stephanos Ioannidis <stephanos.ioannidis@nordicsemi.no>
Both CONFIG_BIG_ENDIAN and CONFIG_LITTLE_ENDIAN can be used in the code,
so reflect that in the error message in checkpatch.
Signed-off-by: Carles Cufi <carles.cufi@nordicsemi.no>
With the current clang-format rules, we have that multi-line macros that
contain empty lines (usual in driver definition macros) are formatted
with spaces only, no leading tabs. For example:
```c
#define MYDRIVER_DEFINE(i) \
struct mydriver_data data##i; \
\ /* starts with spaces */
struct mydriver_config config##i; \
...
```
This patch makes checkpatch ignore such cases, so that clang-format can
be used in tree while keeping checkpatch happy.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Some macros, e.g., FOR_EACH_FIXED_ARG, require a separator argument that
needs to be in parentheses, e.g., (||). This should not trigger a
checkpatch warning.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Checkpatch was checking for macro concats (##), however
we often use UTIL_CAT to allow for multiple macro expansion passes
to succeed.
Signed-off-by: Tom Burdick <thomas.burdick@intel.com>
checkpatch requires a Makefile to be present at the top-level directory.
Remove this requirement.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
This squashes and applies the Linux diffs to scripts/checkpatch.pl
between Linux commit 16fbf79b0f83bc75 ("Linux 5.6-rc7") and
9123e3a74ec7b93 ("Linux 5.9-rc1") except for commits identified below.
The last 1000 commits to Zephyr master were compared for checkpatch
output differences between the previous Zephyr version and this
version. One new diagnostic about function declarations with an empty
parameter-list was introduced (FUNCTION_WITHOUT_ARGS). The text of an
existing diagnostic was changed (DT_SPLIT_BINDING_PATCH). The text of
LONG_LINE diagnostics was enhanced to provide the actual line length.
Linux commit dfa05c28ca7ffc0a ("checkpatch: remove email address
comment from email address comparisons") was removed because
differences in the scripts resulted in false signed-off-by check
diagnostics when a full name included characters not in Basic Latin,
due to changes in how the author name was extracted. Earlier upstream
changes not integrated into Zephyr may be required.
Linux commit b95692f8b3000166 ("checkpatch: prefer fallthrough; over
fallthrough comments") was removed because Zephyr doesn't support the
upstream pseudo keyword.
Linux commit bdc48fa11e46f867 ("checkpatch/coding-style: deprecate
80-column warning") was edited to not actually change the 80-column
maximum line limit as this change has not been mooted for Zephyr.
Linux commit ced69da1db0b57bb ("checkpatch: fix CONST_STRUCT when
const_structs.checkpatch is missing") was edited to the CONST_STRUCT
file as that's not supported in Zephyr.
Signed-off-by: Peter A. Bigot <pab@pabigot.com>
Now that we are standardizing C99 integer types we can revert the commit
that warned about C99 type usage.
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
This commit reverts the device tree binding vendor prefix check regular
expression change that was introduced by the following commit:
5b10fac97e
The changed regular expression fails to detect the correctly specified
vendor prefixes in the `vendor-prefixes.txt`.
Signed-off-by: Stephanos Ioannidis <root@stephanos.io>
Linux uses for_each in macros that produce loops; Zephyr uses
FOR_EACH. Update the corresponding checks to match Zephyr's spelling.
Signed-off-by: Peter Bigot <peter.bigot@nordicsemi.no>
This takes the linux diffs between
a8c964eacb21288b2dbfa9d80cee5968a3b8fb21 and
16fbf79b0f83bc752cee8589279f1ebfe57b3b6e and applies them to the
Zephyr copy. Three changes did not apply cleanly:
* linux added a comment to the line that enables C99 comments;
Zephyr disallows them.
* linux uses vendor-prefixes.yaml; zephyr uses the older .txt file
* manual addition of colon in a check before BRACKET_SPACE error
Signed-off-by: Peter Bigot <peter.bigot@nordicsemi.no>
There is no code to handle __syscall_inline so it is better removing
it from doxygen and checkpatch.
Signed-off-by: Flavio Ceolin <flavio.ceolin@intel.com>
import linux-v5.1:Documentation/devicetree/bindings/vendor-prefixes.txt
Use vendor-prefixes.txt to check vendor prefixes
used in compatible strings and property names.
Signed-off-by: Antony Pavlov <antonynpavlov@gmail.com>
This was never a long-term solution, more of a gross hack
to get test cases working until we could figure out a good
end-to-end solution for memory domains that generated
appropriate linker sections. Now that we have this with
the app shared memory feature, and have converted all tests
to remove it, delete this feature.
To date all userspace APIs have been tagged as 'experimental'
which sidesteps deprecation policies.
Signed-off-by: Andrew Boie <andrew.p.boie@intel.com>
This is generating lots of false positives and we keep overriding them
manually, so downgrade and mark it as a warning.
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
This is pulled from commit a8c964eacb21288b2dbfa9d80cee5968a3b8fb21 of
the Linux kernel with local zephyr related modifications.
Fixes#4135.
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
To define a system call, it's now sufficient to simply tag the inline
prototype with "__syscall" or "__syscall_inline" and include a special
generated header at the end of the header file.
The system call dispatch table and enumeration of system call IDs is now
automatically generated.
Signed-off-by: Andrew Boie <andrew.p.boie@intel.com>
