There is no need to include <linux/rbtree.h> in init_task.h.
Move it to the right place, in kernel/cred.c which uses RB_ROOT_CACHED.
This is a follow-up of commit 4e7e3adbba52 ("Expand various INIT_* macros
and remove") which moved things from init_task.h to init_task.c
Signed-off-by: Christophe JAILLET <[email protected]>
---
Let see if build-bots agree with me!
This patch depends on [1] and [2].
All these patches are related to init_task.h simplification.
They are unrelated and can be applied separately, but as they modify more
or less the same place in the same file, there may be some merge conflict.
[1]: https://lore.kernel.org/all/1a3d5bd51b7807471a913f8fa621e5a4ecd08e6a.1670100520.git.christophe.jaillet@wanadoo.fr/
[2]: https://lore.kernel.org/all/94c8f2123a8833b61d84a662ec35f9c070cdf4dd.1670147823.git.christophe.jaillet@wanadoo.fr/
---
include/linux/init_task.h | 1 -
init/init_task.c | 1 +
2 files changed, 1 insertion(+), 1 deletion(-)
diff --git a/include/linux/init_task.h b/include/linux/init_task.h
index 6d2c4ea4c97a..e18a6c6f6fd9 100644
--- a/include/linux/init_task.h
+++ b/include/linux/init_task.h
@@ -9,7 +9,6 @@
#include <linux/ipc.h>
#include <linux/user_namespace.h>
#include <linux/seqlock.h>
-#include <linux/rbtree.h>
#include <linux/refcount.h>
#include <linux/sched/autogroup.h>
#include <net/net_namespace.h>
diff --git a/init/init_task.c b/init/init_task.c
index ff6c4b9bfe6b..2392c0a67fb7 100644
--- a/init/init_task.c
+++ b/init/init_task.c
@@ -2,6 +2,7 @@
#include <linux/init_task.h>
#include <linux/export.h>
#include <linux/mqueue.h>
+#include <linux/rbtree.h>
#include <linux/sched.h>
#include <linux/sched/sysctl.h>
#include <linux/sched/rt.h>
--
2.34.1
[+Cc Ingo, who had been working on lots of include improvements.]
The Cc list is odd -- it appears that get_maintainers.pl is broken on
init_task.c:
$> ./scripts/get_maintainer.pl init/init_task.c
Mark Rutland <[email protected]> (commit_signer:1/1=100%)
"Paul E. McKenney" <[email protected]> (commit_signer:1/1=100%)
Marco Elver <[email protected]> (commit_signer:1/1=100%,authored:1/1=100%,removed_lines:5/5=100%)
[email protected] (open list)
What's going on here?
On Sun, Dec 04, 2022 at 12:39PM +0100, Christophe JAILLET wrote:
> There is no need to include <linux/rbtree.h> in init_task.h.
> Move it to the right place, in kernel/cred.c which uses RB_ROOT_CACHED.
>
> This is a follow-up of commit 4e7e3adbba52 ("Expand various INIT_* macros
> and remove") which moved things from init_task.h to init_task.c
>
> Signed-off-by: Christophe JAILLET <[email protected]>
> ---
> Let see if build-bots agree with me!
>
> This patch depends on [1] and [2].
> All these patches are related to init_task.h simplification.
>
> They are unrelated and can be applied separately, but as they modify more
> or less the same place in the same file, there may be some merge conflict.
>
> [1]: https://lore.kernel.org/all/1a3d5bd51b7807471a913f8fa621e5a4ecd08e6a.1670100520.git.christophe.jaillet@wanadoo.fr/
> [2]: https://lore.kernel.org/all/94c8f2123a8833b61d84a662ec35f9c070cdf4dd.1670147823.git.christophe.jaillet@wanadoo.fr/
> ---
> include/linux/init_task.h | 1 -
> init/init_task.c | 1 +
> 2 files changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/include/linux/init_task.h b/include/linux/init_task.h
> index 6d2c4ea4c97a..e18a6c6f6fd9 100644
> --- a/include/linux/init_task.h
> +++ b/include/linux/init_task.h
> @@ -9,7 +9,6 @@
> #include <linux/ipc.h>
> #include <linux/user_namespace.h>
> #include <linux/seqlock.h>
> -#include <linux/rbtree.h>
> #include <linux/refcount.h>
> #include <linux/sched/autogroup.h>
> #include <net/net_namespace.h>
> diff --git a/init/init_task.c b/init/init_task.c
> index ff6c4b9bfe6b..2392c0a67fb7 100644
> --- a/init/init_task.c
> +++ b/init/init_task.c
> @@ -2,6 +2,7 @@
> #include <linux/init_task.h>
> #include <linux/export.h>
> #include <linux/mqueue.h>
> +#include <linux/rbtree.h>
> #include <linux/sched.h>
> #include <linux/sched/sysctl.h>
> #include <linux/sched/rt.h>
> --
> 2.34.1
>
On Sun, 2022-12-04 at 14:55 +0100, Marco Elver wrote:
> [+Cc Ingo, who had been working on lots of include improvements.]
>
> The Cc list is odd -- it appears that get_maintainers.pl is broken on
> init_task.c:
>
> $> ./scripts/get_maintainer.pl init/init_task.c
> Mark Rutland <[email protected]> (commit_signer:1/1=100%)
> "Paul E. McKenney" <[email protected]> (commit_signer:1/1=100%)
> Marco Elver <[email protected]> (commit_signer:1/1=100%,authored:1/1=100%,removed_lines:5/5=100%)
> [email protected] (open list)
>
> What's going on here?
No one claims to be a maintainer of any init/ files
($ git grep "^F:.*init" MAINTAINERS)
so by default it merely shows the people that have patched the
particular file in the last year.
You could perhaps see this better by adding --no-git --no-gitfallback
to the get_maintainer.pl command line
$ ./scripts/get_maintainer.pl --nogit --nogit-fallback init/init_task.c
[email protected] (open list)
see the --help option for details
$ ./scripts/get_maintainer.pl --help
usage: ./scripts/get_maintainer.pl [options] patchfile
./scripts/get_maintainer.pl [options] -f file|directory
version: 0.26
MAINTAINER field selection options:
--email => print email address(es) if any
--git => include recent git *-by: signers
--git-all-signature-types => include signers regardless of signature type
or use only (Signed-off-by:|Reviewed-by:|Acked-by:) signers (default: 0)
--git-fallback => use git when no exact MAINTAINERS pattern (default: 1)
--git-chief-penguins => include (Linus Torvalds)
--git-min-signatures => number of signatures required (default: 1)
--git-max-maintainers => maximum maintainers to add (default: 5)
--git-min-percent => minimum percentage of commits required (default: 5)
--git-blame => use git blame to find modified commits for patch or file
--git-blame-signatures => when used with --git-blame, also include all commit signers
--git-since => git history to use (default: 1-year-ago)
--hg-since => hg history to use (default: -365)
--interactive => display a menu (mostly useful if used with the --git option)
--m => include maintainer(s) if any
--r => include reviewer(s) if any
--n => include name 'Full Name <[email protected]>'
--l => include list(s) if any
--moderated => include moderated lists(s) if any (default: true)
--s => include subscriber only list(s) if any (default: false)
--remove-duplicates => minimize duplicate email names/addresses
--roles => show roles (status:subsystem, git-signer, list, etc...)
--rolestats => show roles and statistics (commits/total_commits, %)
--file-emails => add email addresses found in -f file (default: 0 (off))
--fixes => for patches, add signatures of commits with 'Fixes: <commit>' (default: 1 (on))
--scm => print SCM tree(s) if any
--status => print status if any
--subsystem => print subsystem name if any
--web => print website(s) if any
Output type options:
--separator [, ] => separator for multiple entries on 1 line
using --separator also sets --nomultiline if --separator is not [, ]
--multiline => print 1 entry per line
Other options:
--pattern-depth => Number of pattern directory traversals (default: 0 (all))
--keywords => scan patch for keywords (default: 1)
--sections => print all of the subsystem sections with pattern matches
--letters => print all matching 'letter' types from all matching sections
--mailmap => use .mailmap file (default: 1)
--no-tree => run without a kernel tree
--self-test => show potential issues with MAINTAINERS file content
--version => show version
--help => show this help information
Default options:
[--email --tree --nogit --git-fallback --m --r --n --l --multiline
--pattern-depth=0 --remove-duplicates --rolestats]
Notes:
Using "-f directory" may give unexpected results:
Used with "--git", git signators for _all_ files in and below
directory are examined as git recurses directories.
Any specified X: (exclude) pattern matches are _not_ ignored.
Used with "--nogit", directory is used as a pattern match,
no individual file within the directory or subdirectory
is matched.
Used with "--git-blame", does not iterate all files in directory
Using "--git-blame" is slow and may add old committers and authors
that are no longer active maintainers to the output.
Using "--roles" or "--rolestats" with git send-email --cc-cmd or any
other automated tools that expect only ["name"] <email address>
may not work because of additional output after <email address>.
Using "--rolestats" and "--git-blame" shows the #/total=% commits,
not the percentage of the entire file authored. # of commits is
not a good measure of amount of code authored. 1 major commit may
contain a thousand lines, 5 trivial commits may modify a single line.
If git is not installed, but mercurial (hg) is installed and an .hg
repository exists, the following options apply to mercurial:
--git,
--git-min-signatures, --git-max-maintainers, --git-min-percent, and
--git-blame
Use --hg-since not --git-since to control date selection
File ".get_maintainer.conf", if it exists in the linux kernel source root
directory, can change whatever get_maintainer defaults are desired.
Entries in this file can be any command line argument.
This file is prepended to any additional command line arguments.
Multiple lines and # comments are allowed.
Most options have both positive and negative forms.
The negative forms for --<foo> are --no<foo> and --no-<foo>.