man/man2/rt_sigqueueinfo.2: tfix (tid => tgid)

Signed-off-by: Askar Safin <safinaskar@zohomail.com>
Message-ID: <20250130050625.3356602-1-safinaskar@zohomail.com>
Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man5/proc_pid_limits.5: tfix

Reported-by: "Tomerius, Kai" <Kai.Tomerius@elektrobit.com>
Fixes: 07c3768e9db8 (2023-08-15; "proc.5, proc_pid_limits.5: Split /proc/PID/limits from proc(5)")
Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man7/pathname.7: Pathnames are opaque C strings

On Mon, Jan 27, 2025 at 07:27:59PM +0100, наб wrote:
> Skimming the thread: UNIX paths are sequences of non-NUL bytes.
>
> It is never correct to expect to be able to have a (parse, unparse)
> operation pair for which unparse(parse(x)) = x for path x.
>
> It's obviously wrong to reject a pathname just because you dont like it.
>
> Thus, when displaying a path, either (a) dump it directly to the output
> (the user has configured their display device to understand the paths they use),
> or if that's not possible (b) setlocale(LC_ALL, "") + mbrtowc() loop
> and render the result (applying usual ?/� substitutions for mbrtowc()
> errors makes sense here).
>
> There are very few operations on paths that are actually reasonable
> to do, ever; those are: appending stuff, prepending stuff
> (this is just appending stuff with the arguments backwards),
> and cleaving at /es;
> the "stuff" better be copied whole-sale from some other path
> or an unprocessed argument (or, sure, the PFCS).
>
> If you're getting bytes to append to a path, do that directly.
>
> If you're getting characters to append to a path,
> then wctomb(3) is the only non-invalid solution,
> since that (obviously) turns characters into bytes in the current
> locale, which (ex def) is the operation desired.
>
> I don't understand what the UTF-32 dance is supposed to be.
>
> If you're recommending transcoding paths, don't.
>
> To re-iterate: paths are not character sequences.
> They do not represent characters.
> You can't meaningfully coerce them thusly without loss of precision
> (this is ok to do for display! and nothing else).
> If at any point you find yourself turning wchar_t -> char
> you are doing something wrong;
> if you find yourself doing char -> wchar_t for anything beside display
> you should probably reconsider.
>
> This is different under Win32 of course. But that concerns us naught.

Suggested-by: наб <nabijaczleweli@nabijaczleweli.xyz>
Cc: Jason Yundt <jason@jasonyundt.email>
Cc: Florian Weimer <fweimer@redhat.com>
Cc: "G. Branden Robinson" <branden@debian.org>
Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man7/pathname.7: EXAMPLES: Use a non-ASCII filename

The ASCII filename was too simple to actually show anything useful.

Signed-off-by: Alejandro Colomar <alx@kernel.org>

share/mk/build/: Ignore known errors with pathname.7

Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man7/pathname.7: Add file documenting pathnames

The goal of this new manual page is to help people create programs that
do the right thing even in the face of unusual paths.  The information
that I used to create this new manual page came from these sources:

Link: <https://unix.stackexchange.com/a/39179/316181>
Link: <https://sourceware.org/pipermail/libc-help/2024-August/006737.html>
Link: <https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/tree/fs/ext4/ext4.h?h=v6.12.9#n2288>
Link: <man:unix(7)>
Link: <https://unix.stackexchange.com/q/92426/316181>
Cc: Florian Weimer <fweimer@redhat.com>
Signed-off-by: Jason Yundt <jason@jasonyundt.email>
Message-ID: <20250121133523.24606-1-jason@jasonyundt.email>
Cc: "G. Branden Robinson" <branden@debian.org>
[alx: wfix, ffix, and other tweaks]
Signed-off-by: Alejandro Colomar <alx@kernel.org>

Changes.old: tfix

Reported-by: Brian Inglis <Brian.Inglis@SystematicSW.ab.ca>
Signed-off-by: Alejandro Colomar <alx@kernel.org>

share/mk/build/pdf/book/prepare.pl: Fix duplicate overview-panel entries

Signed-off-by: Alejandro Colomar <alx@kernel.org>

Changes.old: tfix

Signed-off-by: Alejandro Colomar <alx@kernel.org>

Start of man-pages-NEXT: Move Changes to Changes.old

Signed-off-by: Alejandro Colomar <alx@kernel.org>

lsm: Released 6.10

Signed-off-by: Alejandro Colomar <alx@kernel.org>

Changes: Ready for 6.10

Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man7/unicode.7: pfix

Reported-by: Wyatt Carpenter <wyattscarpenter@gmail.com>
Signed-off-by: Alejandro Colomar <alx@kernel.org>

SPONSORS: Publish the sponsors of this project

These sponsors started sponsoring this project in 2024-11-01.
Thanks a lot for that!

For anyone interested in sponsoring this project, please contact the
maintainer and/or mailing list.

Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man7/man-pages.7: Stop telling contributors to write titles in all caps

Recently, I submitted my first patch to the Linux man-pages project.  In
my patch, I had created a new manual page.  On the manual page’s title
line, I had written the title of my new page in all caps because
man-pages(7) said that I should write it that way.  It turns out that
man-pages(7) was wrong and that the title on the title line should have
matched the title in the manual page’s filename [1][2].  This commit
corrects man-pages(7) so that it does not tell contributors to use all
caps when writing titles on title lines.

Link: [1] <https://lore.kernel.org/linux-man/rph24kz36vysoeix4qoxxxcwq3c3fskws2vmxkkgcb2lpits3f@ysm7ug66svzh/T/#mc84250a6634d7f776118879021331001cceccbe5>
Link: [2] <https://lore.kernel.org/linux-man/mog6nnwzwl2dmlrec5b7l76wbxlsnklvdulok644x6o557trib@3zwtoz47r4x3/T/#mf71422d0e159210a7ca2798f2bba50a668e1410e>
Message-ID: <20250114211427.160509-1-jason@jasonyundt.email>
Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man2/fork.2: Add _exit(2) to example

The _exit(2) function is a better choice for exiting a child in many
cases.  Most prominently it avoids calls of functions registered with
atexit(3) by the parent.

There are valid reasons to call exit(3) and the example is actually one
of them: flush FILE-based output.  Since atexit(3) is never called, we
could just stay with exit(3).

Signed-off-by: Tobias Stoeckmann <tobias@stoeckmann.org>
Message-ID: <tngwcffbrzbfkj6vrxgxpekrp3bzuftdy2mzow56xyfkrcna2w@nbgr2ourerxo>
Link: <https://github.com/shadow-maint/shadow/pull/1171>
Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man3/: SYNOPSIS: Use typeof() to improve readability of function pointer types

Due to different spacing, I had missed these ones in the previous
commits of this kind.

Suggested-by: Jorenar <dev@jorenar.com>
Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man7/user_namespaces.7: tfix

Fixes: 8c74a1cea495 (2016-07-07; "user_namespaces.7: Clarify details of CAP_SYS_ADMIN and cgroup v1 mounts")
Signed-off-by: Alexander Stepchenko <geochip@altlinux.org>
Message-ID: <bc17315eff26eb31ecc78a2c44b89dfb077813df.1736444255.git.geochip@altlinux.org>
Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man2/statx.2: Document STATX_DIO_READ_ALIGN

Document the new STATX_DIO_READ_ALIGN flag and the new
stx_dio_read_offset_align field guarded by it.

Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: Christian Brauner <brauner@kernel.org>
Cc: Jan Kara <jack@suse.cz>
Cc: Chandan Babu R <chandan.babu@oracle.com>
Cc: Hongbo Li <lihongbo22@huawei.com>
Cc: Ryusuke Konishi <konishi.ryusuke@gmail.com>
Cc: <linux-nilfs@vger.kernel.org>
Cc: <linux-fsdevel@vger.kernel.org>
Cc: <linux-xfs@vger.kernel.org>
Reviewed-by: "Darrick J. Wong" <djwong@kernel.org>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Message-ID: <20250109083226.GA22264@lst.de>
Signed-off-by: Alejandro Colomar <alx@kernel.org>

src/bin/diffman, man/man1/diffman: Remove program

It was useful for implementing duffman(1), but now it doesn't use it
anymore.  Drop this program.

Users can do this instead:

$ MAN_KEEP_FORMATTING=1 diff -u <(man page1) <(man page2);

Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man1/diffman-git.1: Add manual page

Signed-off-by: Alejandro Colomar <alx@kernel.org>

src/bin/: Rename duffman => diffman-git

Signed-off-by: Alejandro Colomar <alx@kernel.org>

src/bin/duffman: Don't pipe to less(1)

Let the user decide to pipe or not.

Signed-off-by: Alejandro Colomar <alx@kernel.org>

src/bin/duffman: Always read(1) -r

In src/bin/duffman line 45:
| while read f; do \
        ^--^ SC2162 (info): read without -r will mangle backslashes.

Reported-by: shellcheck(1)
Signed-off-by: Alejandro Colomar <alx@kernel.org>

src/bin/duffman: Handle errors correctly

In src/bin/duffman line 30:
cd $(git rev-parse --show-toplevel);
   ^-- SC2046 (warning): Quote this to prevent word splitting.
     ^---------------------------^ SC2312 (info): Consider invoking this command separately to avoid masking its return value (or use '|| true' to ignore).

Reported-by: shellcheck(1)
Signed-off-by: Alejandro Colomar <alx@kernel.org>

src/bin/duffman: Allow diffing two git references

Signed-off-by: Alejandro Colomar <alx@kernel.org>

src/bin/duffman: Allow diffing git references, not just the working directory

Signed-off-by: Alejandro Colomar <alx@kernel.org>

src/bin/duffman: Diff against HEAD, not the system pages

Signed-off-by: Alejandro Colomar <alx@kernel.org>

src/bin/duffman: wsfix

Signed-off-by: Alejandro Colomar <alx@kernel.org>

src/bin/duffman, scripts/bash_aliases: Move duffman() to a separate script

Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man2/stat.2: Linux 6.11 allows using NULL with AT_EMPTY_PATH.

Cc: Dan Albert <danalbert@google.com>
Signed-off-by: Elliott Hughes <enh@google.com>
Message-ID: <CAJgzZoqAOpJajmAnr-i9h3sPC8F_Uu0A+3eg4nkP+xTAV5fPGg@mail.gmail.com>
Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man2/setns.2: Add missing info about time ns

Only singlethreaded processes can setns into time ns.

Link: <https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/kernel/time/namespace.c?h=v6.12#n309>
Signed-off-by: Michal Clapinski <mclapinski@google.com>
Message-ID: <20250107135700.3995936-1-mclapinski@google.com>
Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man2/signal.2: VERSIONS: The typedef-less definition of signal(2) isn't hard to read

At least if you use typeof().

Acked-by: Jorenar <dev@jorenar.com>
Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/: Use typeof() to improve readability of function pointer types

An exception of good taste is structure members.  There, for alignment
reasons, traditional syntax seems more appropriate usually.

Suggested-by: Jorenar <dev@jorenar.com>
Cc: Martin Uecker <uecker@tugraz.at>
Signed-off-by: Alejandro Colomar <alx@kernel.org>

share/mk/: $CLANG_CFLAGS: Add -Wno-language-extension-token

We use typeof(), which triggers that.  It's now standard, so this is a
false positive, IMO.

Signed-off-by: Alejandro Colomar <alx@kernel.org>

src/bin/diffman: Simplify temporary file names

Now that we're using labels, the file name is unimportant.

Signed-off-by: Alejandro Colomar <alx@kernel.org>

src/bin/diffman: Use diff(1) labels for nicer output

Signed-off-by: Alejandro Colomar <alx@kernel.org>

src/bin/diffman, man/man1/diffman.1: Use man(1) to format the pages

This allows using the environment variables that man(1) understands.

Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man3/: EXAMPLES: wsfix

Add a space after a cast.

Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man3/: EXAMPLES: Remove unnecessary parentheses

Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man3/sem_open.3: SYNOPSIS: This is a variadic function

Specify the prototype consistently with open(2).

Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man2/io_submit.2: Document RWF_NOAPPEND

Document flag introduced in Linux v6.9.

Cc: Rich Felker <dalias@libc.org>
Cc: Christian Brauner <brauner@kernel.org>
Signed-off-by: John Garry <john.g.garry@oracle.com>
Message-ID: <20241126090847.297371-3-john.g.garry@oracle.com>
Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man2/readv.2: Document RWF_NOAPPEND

Document flag introduced in Linux v6.9.

Cc: Rich Felker <dalias@libc.org>
Cc: Christian Brauner <brauner@kernel.org>
Signed-off-by: John Garry <john.g.garry@oracle.com>
Message-ID: <20241126090847.297371-2-john.g.garry@oracle.com>
[alx: wfix, srcfix]
Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man2/statx.2: Update STATX_WRITE_ATOMIC filesystem support

Linux v6.13 will include atomic write support for xfs and ext4, so
update STATX_WRITE_ATOMIC commentary to mention that.

Cc: "Darrick J. Wong" <djwong@kernel.org>
Cc: <ritesh.list@gmail.com>
Signed-off-by: John Garry <john.g.garry@oracle.com>
Message-ID: <20241203145359.2691972-1-john.g.garry@oracle.com>
[alx: ffix]
Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man3/pthread_cond_init.3: Remove EINTR error code

In pthreads.7 it says:

"For each of the pthreads functions that can return an error,
POSIX.1-2001 specifies that the function can never fail with the error
EINTR."

Link: <https://lore.kernel.org/linux-man/Z3W_qgawqyEB-QrA@comp../>
Cc: Florian Weimer <fweimer@redhat.com>
Signed-off-by: Arkadiusz Drabczyk <arkadiusz@drabczyk.org>
Message-ID: <20250103173816.6409-2-arkadiusz@drabczyk.org>
Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man7/signal.7: Don't mention that pthread API can return EINTR

In pthreads.7 it says:

"For each of the pthreads functions that can return an error,
POSIX.1-2001 specifies that the function can never fail with the error
EINTR."

Link: <https://lore.kernel.org/linux-man/Z3W_qgawqyEB-QrA@comp../>
Cc: Florian Weimer <fweimer@redhat.com>
Signed-off-by: Arkadiusz Drabczyk <arkadiusz@drabczyk.org>
Message-ID: <20250103173816.6409-1-arkadiusz@drabczyk.org>
Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man3/: Don't use 'length' to refer to buffer size

Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man3/stdio_ext.3: wfix

For consistency with wide characters, say characters instead of bytes.

Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man3/{stdin,stdio}.3: SYNOPSIS: Consistently declare as 'extern' variables

And leave two spaces between the type name and the variable name.

Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man2/ioctl_pipe.2: SYNOPSIS: Fix $1, which is not an array parameter

Link: <https://lore.kernel.org/linux-man/20241214180423.2thsuyyfosrlyajb@devuan/T/#u>
Reported-by: Alejandro Colomar <alx@kernel.org>
Suggested-by: Cyril Hrubis <chrubis@suse.cz>
Cc: David Howells <dhowells@redhat.com>
Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man2/madvise.2: wfix

Cc: Jann Horn <jannh@google.com>
Suggested-by: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
Message-ID: <20241206113418.14327-1-lorenzo.stoakes@oracle.com>
Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man3/string.3: SYNOPSIS: Simplify

Replace the prototypes by manual-page references, and use generic text
which doesn't use the names of the parameters.  It's just an overview.

Signed-off-by: Alejandro Colomar <alx@kernel.org>

src/bin/diffman: Add support for mdoc(7) pages

Signed-off-by: Alejandro Colomar <alx@kernel.org>

src/bin/diffman: Run the entire groff pipeline

Signed-off-by: Alejandro Colomar <alx@kernel.org>

src/bin/diffman: Fix bug at parsing options

We don't want multiple cases to match the option.

The '*)' was matching any option.

Fixes: f2dfd79facf0 (2024-12-13, "src/bin/diffman: Add default case")
Signed-off-by: Alejandro Colomar <alx@kernel.org>

src/bin/diffman, man/man1/diffman.1: -s: Add support for diff(1)'s -s flag

Signed-off-by: Alejandro Colomar <alx@kernel.org>

src/bin/mansect: Use pipefail and some safety measures

Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man3/getline.3: wfix

Fixes: ef53ef760638 (2024-12-18, "man/man3/getline.3: Clarify ERRORS")
Signed-off-by: Alejandro Colomar <alx@kernel.org>

CONTRIBUTING.d/patches/subject: Use the full path always

This is easier to do, since auto-completion will do it.  Also, it's more
consistent.  I started using full paths a few commits ago.

Signed-off-by: Alejandro Colomar <alx@kernel.org>

CONTRIBUTING: Add reference to CONTRIBUTING.d/git

Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man3/getopt.3: Remove _<PID>_GNU_nonoption_argv_flags_ description

Per <https://sourceware.org/git/?p=glibc.git;a=commitdiff;h=bf079e19f50d64aa5e05b5e17ec29afab9aabb20>:
-  this was set by bash 2.0 (1996-12)
   (implemented in glibc b07c5668f672125074dd5b0b658145e1544120be)
-  it's no longer set by bash 2.01 (1997-06) because it was bad
-  glibc disabled this with no way to enable it in 2001-08
   (518a0dd201c48a5c15d73c1919c304a9f8f5e3c1)
-  glibc removed it in 2017 bf079e19f50d64aa5e05b5e17ec29afab9aabb20

So this was experienced by people for 5 months at best,
and could remotely be experienced for 3 years
(if you somehow wanted this bad behaviour and added it into your shell),
over 20 years ago.
No modern reader (or, frankly, non-modern reader)
has ever used this, or could use it, really.

Link to the removal note for completeness only.

Link: <https://sourceware.org/git/?p=glibc.git;a=commitdiff;h=bf079e19f50d64aa5e05b5e17ec29afab9aabb20>
Cc: Eugene Syromyatnikov <evgsyr@gmail.com>
Signed-off-by: Ahelenia Ziemiańska <nabijaczleweli@nabijaczleweli.xyz>
Message-Id: <dwfybzlb5ydbsc4puo6igj7nm7iregquv6hxhhqb53bwrvqswb@tarta.nabijaczleweli.xyz>
[alx: Add break points in URI; minor tweaks to commit message too]
Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man3/getline.3: Clarify ERRORS

Signed-off-by: Elliott Hughes <enh@google.com>
Message-ID: <CAJgzZoruFUg6X=JUNJXCbBOKs13SX=dsNFNdTMct2VecUFG=ww@mail.gmail.com>
Signed-off-by: Alejandro Colomar <alx@kernel.org>

CONTRIBUTING.d/git: mkdir(1) parent directory before writing to file

Reported-by: Elliott Hughes <enh@google.com>
Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man3/strverscmp.3: This is NOT the ordering used by ls -v

Compare, given:
#include <stdlib.h>
#include <stdio.h>
#include <string.h>
int compar(const char **l, const char **r) {
return strverscmp(*l, *r);
}
int main(int argc, char ** argv) {
qsort(argv + 1, argc - 1, sizeof(*argv), compar);
for(int i = 1; i <  argc; ++i)
puts(argv[i]);
}
yields:
$ /bin/ls -v1 a*  # coreutils ls
a-1.0a
a-1.0.1a
$ ../vers a*      # as above
a-1.0.1a
a-1.0a
$ ls -v1 a*       # voreutils ls @ 5781698 with strverscmp()-equivalent sorting
a-1.0.1a
a-1.0a
compare also the results for real data like
netstat-nat-1.{0,1{,.1},2,3.1,4{,.{1,2,3,4,5,6,7,8,9,10}}}.tar.gz

Thus, coreutils ls -v (and sort -V) does NOT use strverscmp(3);
it uses a modified Debian version comparison algorithm with additional
suffix processing and `ls -v`-specific exceptions.

Signed-off-by: Ahelenia Ziemiańska <nabijaczleweli@nabijaczleweli.xyz>
Message-Id: <myuppkwnltqtxduoop7g7wfuyou5cdo6sotocrvyztmqnazvph@tarta.nabijaczleweli.xyz>
Signed-off-by: Alejandro Colomar <alx@kernel.org>

scripts/bash_aliases: duffman(): Silence output of cp(1)

Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man2/get_mempolicy.2: SYNOPSIS: ffix

Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man2/: SYNOPSIS: Use array notation

Signed-off-by: Alejandro Colomar <alx@kernel.org>

src/bin/diffman: Silence shellcheck(1) diagnostic

Signed-off-by: Alejandro Colomar <alx@kernel.org>

src/bin/diffman: Add default case

SC2249 (info): Consider adding a default *) case, even if it just exits with error.

In principle, getopts(1) shouldn't return other stuff, but it doesn't
hurt to be cutious.

Reported-by: shellcheck(1)
Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/: SYNOPSIS: Use typeof() to improve readability of function pointer types

Suggested-by: Jorenar <dev@jorenar.com>
Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man3/makecontext.3: SYNOPSIS: Add ellipsis to function type

The old syntax with empty parentheses has been removed from the language
in C23.

Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/man3/isalpha.3: tfix

Fixes: ba687b00ecb3 (2023-07-30, "man3/: srcfix")
Signed-off-by: Alejandro Colomar <alx@kernel.org>

proc_pid_fdinfo.5: Reduce indent for most of the page

When /proc/pid/fdinfo was part of proc.5 man page the indentation made
sense.  As a standalone man page the indentation doesn't need to be so
far over to the right.  Remove the initial tagged pragraph, move the
"since Linux 2.6.22" to a new HISTORY subsection.

Suggested-by: "G. Branden Robinson" <g.branden.robinson@gmail.com>
Message-Id: <20241206073828.1119464-1-irogers@google.com>
[alx: ffix]
Signed-off-by: Alejandro Colomar <alx@kernel.org>

scripts/bash_aliases: duffman: Pass any arguments (options) to diffman(1)

Signed-off-by: Alejandro Colomar <alx@kernel.org>

src/bin/diffman, diffman.1: -U: Add support for diff(1)'s -U option

Signed-off-by: Alejandro Colomar <alx@kernel.org>

src/bin/diffman: -w: Add support for diff(1)'s -w flag

Signed-off-by: Alejandro Colomar <alx@kernel.org>

ld.so.8: tfix

Signed-off-by: Anhad Singh <andypython@protonmail.com>
Message-Id: <20241206071814.55913-1-andypython@protonmail.com>
Signed-off-by: Alejandro Colomar <alx@kernel.org>

madvise.2: Add description of MADV_GUARD_INSTALL, MADV_GUARD_REMOVE

Lightweight guard region support has been added to Linux 6.13, which
adds MADV_GUARD_INSTALL and MADV_GUARD_REMOVE flags to the madvise()
system call.  Therefore, update the manpage for madvise() and describe
these operations.

Reviewed-by: Jann Horn <jannh@google.com>
Signed-off-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
Message-Id: <20241205104125.67518-1-lorenzo.stoakes@oracle.com>
Signed-off-by: Alejandro Colomar <alx@kernel.org>

CONTRIBUTING.d/patches/description: Add link to history of inter-sentence space

Cc: "G. Branden Robinson" <branden@debian.org>
Signed-off-by: Alejandro Colomar <alx@kernel.org>

madvise.2: MADV_SOFT_OFFLINE requests can report EBUSY

If the page could not be offlined madvise will report EBUSY.
This might occur if the page is currently in use or locked.

Signed-off-by: Tyonnchie Berry <tyberry@redhat.com>
Message-Id: <Z0XzU9R9Kx0RoeUG@redhat.com>
Acked-by: "Luis Claudio R. Goncalves" <lgoncalv@redhat.com>
[alx: wfix, ffix, and other tweaks]
Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/: wfix

Signed-off-by: Alejandro Colomar <alx@kernel.org>

process_madvise.2: Describe 6.13 behaviour permitting all madvise flags

Since Linux 6.13 it has become possible to use all madvise flags when
targeting the calling process.

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
Message-Id: <20241129164422.89837-1-lorenzo.stoakes@oracle.com>
Signed-off-by: Alejandro Colomar <alx@kernel.org>

readlink.2: SYNOPSIS: Use array notation

Signed-off-by: Alejandro Colomar <alx@kernel.org>

landlock_create_ruleset.2: SYNOPSIS: Add missing include

Signed-off-by: Alejandro Colomar <alx@kernel.org>

share/mk/: $MANINTROPAGES: Remove redundant grep(1) call

Signed-off-by: Alejandro Colomar <alx@kernel.org>

share/mk/: $MANEXT: Use extended regex

This makes the line fit in 80 columns, and also makes it simpler to
read.

Signed-off-by: Alejandro Colomar <alx@kernel.org>

scripts/bash_aliases: duffman(): Add function to diff all modified pages

Signed-off-by: Alejandro Colomar <alx@kernel.org>

src/bin/sortman, sortman.1, share/mk/: Move sortman script to src/bin/, and add manual page

Signed-off-by: Alejandro Colomar <alx@kernel.org>

man1/: SYNOPSIS: ffix

Use SY/YS, and other formatting improvements.

Signed-off-by: Alejandro Colomar <alx@kernel.org>

src/bin/diffman, diffman.1: Add program and its manual page

This program diffs manual pages.  It's useful for reviewing changes to a
manual page:

$ diffman membarrier ./man2/membarrier.2 | less -R;

Signed-off-by: Alejandro Colomar <alx@kernel.org>

etc/checkpatch/checkpatch.conf: Ignore MACRO_ARG_PRECEDENCE

This diagnostic is useful for preventing unsafe macros, but other linter programs
provide a smaller rate of false positives, so let's turn this one off.

Signed-off-by: Alejandro Colomar <alx@kernel.org>

etc/clang-tidy/config.yaml: -clang-diagnostic-sign-conversion

This diagnostic is too noisy, and isn't very useful.  Turn it off.

Signed-off-by: Alejandro Colomar <alx@kernel.org>

getgroups.2: EXAMPLES: Add example program

Signed-off-by: Alejandro Colomar <alx@kernel.org>

posix_fadvise.2: POSIX_FADV_NOREUSE now supported.

POSIX_FADV_NOREUSE is now supported in Linux.  Update text regarding
former no op behavior.  Indicate the readahead policy and treatment of
file pages read with this flag.

Link: <https://lore.kernel.org/linux-mm/20221230215252.2628425-2-yuzhao@google.com/>
Signed-off-by: T.J. Alumbaugh <talumbau@google.com>
Signed-off-by: Yuanchu Xie <yuanchu@google.com>
Message-Id: <20241120045214.1294799-1-yuanchu@google.com>
Acked-by: Yu Zhao <yuzhao@google.com>
Signed-off-by: Alejandro Colomar <alx@kernel.org>

alloca.3: Clarify that alloca(3)ted space is deallocated when the caller returns

Reported-by: Pali Rohár <pali@kernel.org>
Signed-off-by: Alejandro Colomar <alx@kernel.org>

mount_namespaces.7: Use correctly the terms "mount" and "mount point"

On Sun, Nov 17, 2024 at 16:12:24 +0100, Michael Kerrisk wrote:
>
> A "mount" is a tuple consisting of:
> * a mount ID
> * a source (e.g., a device)
> * a target or "mount point" (i.e. a path name)
> * the ID of the parent of this mount
> * other stuff (e.g., options)

Reported-by: Helge Kreutzmann <debian@helgefjell.de>
Cc: Jakub Wilk <jwilk@jwilk.net>
Acked-by: "Michael T. Kerrisk" <mtk.manpages@gmail.com>
Signed-off-by: Alejandro Colomar <alx@kernel.org>

cgroups.7: grfix

Signed-off-by: Alejandro Colomar <alx@kernel.org>

man/: Terminology consistency reforms (n, size, length)

Use 'length' for the lenght of a string.
Use 'n' for the number of elements.
Use 'size' for the number of bytes.  (And in wide-character string
functions, 'size' also refers to the number of wide characters.)

The change is quite large, and I might have made some mistakes.
But overall, this should improve consistency in use of these terms.

Signed-off-by: Alejandro Colomar <alx@kernel.org>

getdents.2: ffix

Reported-by: Helge Kreutzmann <debian@helgefjell.de>
Signed-off-by: Alejandro Colomar <alx@kernel.org>

wcwidth.3: Rename function parameter

Reported-by: Helge Kreutzmann <debian@helgefjell.de>
Signed-off-by: Alejandro Colomar <alx@kernel.org>