GitLab

An inline function can have an attribute, as in include/linux/log2.h,
and kernel-doc handles this already for simple cases.  However,
some attributes have arguments (e.g. the "target" attribute).
Handle those too.

Furthermore, attributes could be at the beginning of a function
declaration, before the return type.  To correctly handle this case,
you need to strip spaces after the attributes; otherwise, dump_function
is left confused.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Acked-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>

A prototype like

    /**
     * foo - sample definition
     * @bar: a parameter
     */
    int foo(int (*bar)(int x,
                       int y));

is currently producing

    .. c:function:: int foo (int (*bar) (int x,                    int y)

       sample definition

    **Parameters**

    ``int (*)(int x,                    int y) bar``
      a parameter

Collapse the spaces so that the output is nicer.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Acked-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>

kernel-doc supports documenting struct members "inline" since
a4c6ebed

 ("scripts/kernel-doc Allow struct arguments documentation
in struct body"). This requires the inline kernel-doc comments to have
the opening and closing comment markers (/** and */ respectively) on
lines of their own, even for short comments. For example:

	/**
	 * struct foo - struct documentation
	 */
	struct foo {
		/**
		 * @bar: member documentation
		 */
		int bar;
	};

Add support for one line inline comments:

	/**
	 * struct foo - struct documentation
	 */
	struct foo {
		/** @bar: member documentation */
		int bar;
	};

Note that mixing of the two in one doc comment is not allowed; either
both comment markers must be on lines of their own, or both must be on
the one line. This limitation keeps both the comments more uniform, and
kernel-doc less complicated.

Cc: Daniel Vetter <daniel@ffwll.ch>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>

Without this patch we get warnings for named variable arguments.

    warning: No description found for parameter '...'
    warning: Excess function parameter 'args' description in 'alloc_ordered_workqueue'
Signed-off-by: Silvio Fricke <silvio.fricke@gmail.com>
Reviewed-by: Jani Nikula <jani.nikula@intel.com
Signed-off-by: Jonathan Corbet <corbet@lwn.net>

Add yet another regex to kernel-doc to trap @param() references separately
and not produce corrupt RST markup.
Signed-off-by: Jonathan Corbet <corbet@lwn.net>

As far as I can tell, the handling of "..." arguments has never worked
right, so any documentation provided was ignored in favor of "variable
arguments."  This makes kernel-doc handle "@...:" as documented.  It does
*not* fix spots in kerneldoc comments that don't follow that convention,
but they are no more broken than before.
Signed-off-by: Jonathan Corbet <corbet@lwn.net>

When using a typedef function like this one:
	typedef bool v4l2_check_dv_timings_fnc (const struct v4l2_dv_timings * t, void * handle);

The Sphinx C domain expects it to create a c:type: reference,
as that's the way it creates the type references when parsing
a c:function:: declaration.

So, a declaration like:

	.. c:function:: bool v4l2_valid_dv_timings (const struct v4l2_dv_timings * t, const struct v4l2_dv_timings_cap * cap, v4l2_check_dv_timings_fnc fnc, void * fnc_handle)

Will create a cross reference for :c:type:`v4l2_check_dv_timings_fnc`.

So, when outputting such typedefs in RST format, we need to handle
this special case, as otherwise it will produce those warnings:

	./include/media/v4l2-dv-timings.h:43: WARNING: c:type reference target not found: v4l2_check_dv_timings_fnc
	./include/media/v4l2-dv-timings.h:60: WARNING: c:type reference target not found: v4l2_check_dv_timings_fnc
	./include/media/v4l2-dv-timings.h:81: WARNING: c:type reference target not found: v4l2_check_dv_timings_fnc

So, change the kernel-doc script to produce a RST output for the
above typedef as:
	.. c:type:: v4l2_check_dv_timings_fnc

	   **Typedef**: timings check callback

	**Syntax**

	  ``bool v4l2_check_dv_timings_fnc (const struct v4l2_dv_timings * t, void * handle);``
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>

Improve the parser to handle typedefs like:

	typedef bool v4l2_check_dv_timings_fnc(const struct v4l2_dv_timings *t, void *handle);
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>

Right now, for a struct, kernel-doc produces the following output:

	.. c:type:: struct v4l2_prio_state

	   stores the priority states

	**Definition**

	::

	  struct v4l2_prio_state {
	    atomic_t prios[4];
	  };

	**Members**

	``atomic_t prios[4]``
	  array with elements to store the array priorities

Putting a member name in verbatim and adding a continuation line
causes the LaTeX output to generate something like:
	item[atomic_t prios\[4\]] array with elements to store the array priorities

Everything inside "item" is non-breakable, with may produce
lines bigger than the column width.

Also, for function members, like:

        int (* rx_read) (struct v4l2_subdev *sd, u8 *buf, size_t count,ssize_t *num);

It puts the name of the member at the end, like:

        int (*) (struct v4l2_subdev *sd, u8 *buf, size_t count,ssize_t *num) read

With is very confusing.

The best is to highlight what really matters: the member name.
is a secondary information.

So, change kernel-doc, for it to produce the output on a different way:

	**Members**

	``prios[4]``

	  array with elements to store the array priorities

Also, as the type is not part of LaTeX "item[]", LaTeX will split it into
multiple lines, if needed.

So, both LaTeX/PDF and HTML outputs will look good.

It should be noticed, however, that the way Sphinx LaTeX output handles
things like:

	Foo
	   bar

is different than the HTML output. On HTML, it will produce something
like:

	**Foo**
	   bar

While, on LaTeX, it puts both foo and bar at the same line, like:

	**Foo** bar
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>

This reverts commit a88b1672

.

From the origin comit log::

  The RST cpp:function handler is very pedantic: it doesn't allow any
  macros like __user on it

Since the kernel-doc parser does NOT make use of the cpp:domain, there
is no need to change the kernel-doc parser eleminating the address_space
tags.
Signed-off-by: Markus Heiser <markus.heiser@darmarIT.de>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>

changeset b7e67f6c

 ("doc-rst: linux_tv: supress lots of warnings")
were meant to touch only on media files, but it also touched
at this script by mistake. Revert such change.
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

The RST cpp:function handler is very pedantic: it doesn't allow any
macros like __user on it:

	Documentation/media/kapi/dtv-core.rst:28: WARNING: Error when parsing function declaration.
	If the function has no return type:
	  Error in declarator or parameters and qualifiers
	  Invalid definition: Expecting "(" in parameters_and_qualifiers. [error at 8]
	    ssize_t dvb_ringbuffer_pkt_read_user (struct dvb_ringbuffer * rbuf, size_t idx, int offset, u8 __user * buf, size_t len)
	    --------^
	If the function has a return type:
	  Error in declarator or parameters and qualifiers
	  If pointer to member declarator:
	    Invalid definition: Expected '::' in pointer to member (function). [error at 37]
	      ssize_t dvb_ringbuffer_pkt_read_user (struct dvb_ringbuffer * rbuf, size_t idx, int offset, u8 __user * buf, size_t len)
	      -------------------------------------^
	  If declarator-id:
	    Invalid definition: Expecting "," or ")" in parameters_and_qualifiers, got "*". [error at 102]
	      ssize_t dvb_ringbuffer_pkt_read_user (struct dvb_ringbuffer * rbuf, size_t idx, int offset, u8 __user * buf, size_t len)
	      ------------------------------------------------------------------------------------------------------^

So, we have to remove it from the function prototype.
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>

While trying to make gpu docs warning free I stumbled over one output
which wasn't following proper compiler error output standards. Fix it
up for more quickfix awesomeness.

Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Jani Nikula <jani.nikula@intel.com>
Cc: linux-doc@vger.kernel.org
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>

The conversion script added some comments at the end.
They point to the original DocBook files, with will be
removed after the manual fixes. So, they'll be pointing
to nowere. So, remove those comments.

They'll be forever stored at the Kernel tree. So, if
someone wants the references, it is just a matter of
looking at the backlog.
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

The c language parser checks if there are duplicated object
definitions. That causes lots of warnings like:
	WARNING: duplicate C object description of ioctl

Let's remove those by telling Sphinx that the language for
those objects are c++. The look of the descriptions will
be close, and the warnings will be gone.

Please notice that we had to keep a few of them as C, as
the c++ parser seems to be broken when it finds an enum.

Yet, this reduced from 219 warnings to 143, with is
a good thing.
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

Scan all input files for EXPORT_SYMBOLs along with the explicitly
specified export files before actually parsing anything.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

If the kernel-doc comments for functions are not in the same file as the
EXPORT_SYMBOL statements, the -export and -internal output selections do
not work as expected. This is typically the case when the kernel-doc
comments are in header files next to the function declarations and the
EXPORT_SYMBOL statements are next to the function definitions in the
source files.

Let the user specify additional source files in which to look for the
EXPORT_SYMBOLs using the new -export-file FILE option, which may be
given multiple times.

The pathological example for this is include/net/mac80211.h, which has
all the kernel-doc documentation for the exported functions defined in a
plethora of source files net/mac80211/*.c.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

Reduce duplication in follow-up work. No functional changes.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

Signed-off-by: Jani Nikula <jani.nikula@intel.com>

Since

commit 32217761


Author: Jani Nikula <jani.nikula@intel.com>
Date:   Sun May 29 09:40:44 2016 +0300

    kernel-doc: concatenate contents of colliding sections

we started getting (more) errors on duplicate section names, especially
on the default section name "Description":

include/net/mac80211.h:3174: warning: duplicate section name 'Description'

This is usually caused by a slightly unorthodox placement of parameter
descriptions, like in the above case, and kernel-doc resetting back to
the default section more than once within a kernel-doc comment.

Ignore warnings on the duplicate section name automatically assigned by
kernel-doc, and only consider explicitly user assigned duplicate section
names an issue.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

No functional changes.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

Lots of kerneldoc entries use "example:" or "note:" as section headers.
Until such a time as we can make them use proper markup, make them work as
intended.
Signed-off-by: Jonathan Corbet <corbet@lwn.net>

Opt-in since this wreaks the rst output and must be removed
by consumers again. This is useful to adjust the linenumbers
for included kernel-doc snippets in shinx. With that sphinx
error message will be accurate when there's issues with the
rst-ness of the kernel-doc comments.

Especially when transitioning a new docbook .tmpl to .rst this
is extremely useful, since you can just use your editors compilation
quickfix list to accurately jump from error to error.

v2:
- Also make sure that we filter the LINENO for purpose/at declaration
  start so it only shows for selected blocks, not all of them (Jani).
  While at it make it a notch more accurate.
- Avoid undefined $lineno issues. I tried filtering these out at the
  callsite, but Jani spotted more when linting the entire kernel.
  Unamed unions and similar things aren't stored consistently and end
  up with an undefined line number (but also no kernel-doc text, just
  the parameter type). Simplify things and filter undefined line
  numbers in print_lineno() to catch them all.

v3: Fix LINENO 0 issue for kernel-doc comments without @param: lines
or any other special sections that directly jump to the description
after the "name - purpose" line. Only really possible for functions
without parameters. Noticed by Jani.

Cc: Jani Nikula <jani.nikula@intel.com>
Cc: linux-doc@vger.kernel.org
Cc: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

state3 = prototype parsing, so name them accordingly.

Cc: Jani Nikula <jani.nikula@intel.com>
Cc: linux-doc@vger.kernel.org
Cc: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

Further up in the state machinery we switch from STATE_NAME to
STATE_DOCBLOCK when we match /$doc_block/. Which means this block of
code here is entirely unreachable, unless there are multiple DOC:
sections within a single kernel-doc comment.

Getting a list of all the files with more than one DOC: section using

$ git grep -c " * DOC:" | grep -v ":1$"

and then doing a full audit of them reveals there are no such comment
blocks in the kernel.

Supporting multiple DOC: sections in a single kernel-doc comment does
not seem like a recommended way of doing things anyway, so nuke the code
for simplicity.

Cc: Jani Nikula <jani.nikula@intel.com>
Cc: linux-doc@vger.kernel.org
Cc: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
[Jani: amended the commit message]
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

If the documentation comment does not have params or sections, the
section heading may leak from the previous documentation comment.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

If there are multiple sections with the same section name, the current
implementation results in several sections by the same heading, with the
content duplicated from the last section to all. Even if there's the
error message, a more graceful approach is to combine all the
identically named sections into one, with concatenated contents.

With the supported sections already limited to select few, there are
massively fewer collisions than there used to be, but this is still
useful for e.g. when function parameters are documented in the middle of
a documentation comment, with description spread out above and
below. (This is not a recommended documentation style, but used in the
kernel nonetheless.)

We can now also demote the error to a warning.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

kernel-doc currently identifies anything matching "section header:"
(specifically a string of word characters and spaces followed by a
colon) as a new section in the documentation comment, and renders the
section header accordingly.

Unfortunately, this turns all uses of colon into sections, mostly
unintentionally. Considering the output, erroneously creating sections
when not intended is always worse than erroneously not creating sections
when intended. For example, a line with "http://example.com" turns into
a "http" heading followed by "//example.com" in normal text style, which
is quite ugly. OTOH, "WARNING: Beware of the Leopard" is just fine even
if "WARNING" does not turn into a heading.

It is virtually impossible to change all the kernel-doc comments, either
way. The compromise is to pick the most commonly used and depended on
section headers (with variants) and accept them as section headers.

The accepted section headers are, case insensitive:

 * description:
 * context:
 * return:
 * returns:

Additionally, case sensitive:

 * @return:

All of the above are commonly used in the kernel-doc comments, and will
result in worse output if not identified as section headers. Also,
kernel-doc already has some special handling for all of them, so there's
nothing particularly controversial in adding more special treatment for
them.

While at it, improve the whitespace handling surrounding section
names. Do not consider the whitespace as part of the name.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

Yes, for our purposes the type should contain typedef.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

The latter isn't special to rst.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

If a param description spans multiple lines, check any leading
whitespace in the first continuation line, and remove same amount of
whitespace from following lines.

This allows indentation in the multi-line parameter descriptions for
aesthetical reasons while not causing accidentally significant
indentation in the rst output.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

Handle whitespace on the first line of param text as if it was the empty
string. There is no need to add the newline in this case. This improves
the rst output in particular, where blank lines may be problematic in
parameter lists.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

Move away from field lists, and simply use **strong emphasis** for
section headings on lines of their own. Do not use rst section headings,
because their nesting depth depends on the surrounding context, which
kernel-doc has no knowledge of. Also, they do not need to end up in any
table of contexts or indexes.

There are two related immediate benefits. Field lists are typically
rendered in two columns, while the new style uses the horizontal width
better. With no extra indent on the left, there's no need to be as fussy
about it. Field lists are more susceptible to indentation problems than
the new style.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

The inline member markup allows whitespace lines before the actual
documentation starts. Strip the leading blank lines. This improves the
rst output.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

Current approach leads to two blank lines, while one is enough.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

No functional changes.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

The use of these is confusing in the script, and per this grep, they're
not used anywhere anyway:

$ git grep " \* [%$&][a-zA-Z0-9_]*:" -- *.[ch] | grep -v "\$\(Id\|Revision\|Date\)"

While at it, throw out the constants array, nothing is ever put there
again.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

Let the user use @foo, &bar, %baz, etc. in the first kernel-doc purpose
line too.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

This bit is already done by xml_unescape() above.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

Link "&foo->bar", "&foo->bar()", "&foo.bar", and "&foo.bar()" to the
struct/union/enum foo definition. The members themselves do not
currently have anchors to link to, but this is better than nothing, and
promotes a universal notation.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>