[PATCH v2 0/4] sysctl: Remove sentinel elements from fs dir

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



From: Joel Granados <j.granados@xxxxxxxxxxx>

What?
These commits remove the sentinel element (last empty element) from the
sysctl arrays of all the files under the "fs/" directory that use a
sysctl array for registration. The merging of the preparation patches
(in https://lore.kernel.org/all/ZO5Yx5JFogGi%2FcBo@xxxxxxxxxxxxxxxxxxxxxx/)
to mainline allows us to just remove sentinel elements without changing
behavior (more info here [1]).

These commits are part of a bigger set (here
https://github.com/Joelgranados/linux/tree/tag/sysctl_remove_empty_elem_V5)
that remove the ctl_table sentinel. We make the review process easier by
chunking the commits into manageable pieces. Each chunk can be reviewed
separately without noise from parallel sets.

Sending the "fs/*" chunk now that the "drivers/" has been mostly
reviewed [6]. After this and the "kernel/*" are reviewed we only have 2 more
chunks ("net/*" and miscellaneous) to complete the sentinel removal.
Hurray!!!

Why?
By removing the sysctl sentinel elements we avoid kernel bloat as
ctl_table arrays get moved out of kernel/sysctl.c into their own
respective subsystems. This move was started long ago to avoid merge
conflicts; the sentinel removal bit came after Mathew Wilcox suggested
it to avoid bloating the kernel by one element as arrays moved out. This
patchset will reduce the overall build time size of the kernel and run
time memory bloat by about ~64 bytes per declared ctl_table array. I
have consolidated some links that shed light on the history of this
effort [2].

Testing:
* Ran sysctl selftests (./tools/testing/selftests/sysctl/sysctl.sh)
* Ran this through 0-day with no errors or warnings

Size saving after this patchset:
    * bloat-o-meter
        - The "yesall" config saves 1920 bytes [4]
        - The "tiny" config saves 576 bytes [5]
    * If you want to know how many bytes are saved after all the chunks
      are merged see [3]

Base commit:
tag: sysctl-6.7-rc1 (8b793bcda61f)

Comments/feedback greatly appreciated

Best

Joel

---
Changes in v2:
- changed commit message from "aio: *" to "fs: *"
- We now register fsverity_sysctl_table with one call instead of
  selecting a call based CONFIG_FS_VERITY_BUILTIN_SIGNATURES
- Link to v1: https://lore.kernel.org/r/20231107-jag-sysctl_remove_empty_elem_fs-v1-0-7176632fea9f@xxxxxxxxxxx

[1]
We are able to remove a sentinel table without behavioral change by
introducing a table_size argument in the same place where procname is
checked for NULL. The idea is for it to keep stopping when it hits
->procname == NULL, while the sentinel is still present. And when the
sentinel is removed, it will stop on the table_size. You can go to 
(https://lore.kernel.org/all/20230809105006.1198165-1-j.granados@xxxxxxxxxxx/)
for more information.

[2]
Links Related to the ctl_table sentinel removal:
* E-mail threads that summarize the sentinel effort
  https://lore.kernel.org/all/ZO5Yx5JFogGi%2FcBo@xxxxxxxxxxxxxxxxxxxxxx/
  https://lore.kernel.org/all/ZMFizKFkVxUFtSqa@xxxxxxxxxxxxxxxxxxxxxx/
* Replacing the register functions:
  https://lore.kernel.org/all/20230302204612.782387-1-mcgrof@xxxxxxxxxx/
  https://lore.kernel.org/all/20230302202826.776286-1-mcgrof@xxxxxxxxxx/
* E-mail threads discussing prposal
  https://lore.kernel.org/all/20230321130908.6972-1-frank.li@xxxxxxxx
  https://lore.kernel.org/all/20220220060626.15885-1-tangmeng@xxxxxxxxxxxxx

[3]
Size saving after removing all sentinels:
  These are the bytes that we save after removing all the sentinels
  (this plus all the other chunks). I included them to get an idea of
  how much memory we are talking about.
    * bloat-o-meter:
        - The "yesall" configuration results save 9158 bytes
          https://lore.kernel.org/all/20230621091000.424843-1-j.granados@xxxxxxxxxxx/
        - The "tiny" config + CONFIG_SYSCTL save 1215 bytes
          https://lore.kernel.org/all/20230809105006.1198165-1-j.granados@xxxxxxxxxxx/
    * memory usage:
        In memory savings are measured to be 7296 bytes. (here is how to
        measure [7])

[4]
add/remove: 0/0 grow/shrink: 0/30 up/down: 0/-1920 (-1920)
Function                                     old     new   delta
xfs_table                                   1024     960     -64
vm_userfaultfd_table                         128      64     -64
test_table_unregister                        128      64     -64
test_table                                   576     512     -64
root_table                                   128      64     -64
pty_table                                    256     192     -64
ocfs2_nm_table                               128      64     -64
ntfs_sysctls                                 128      64     -64
nlm_sysctls                                  448     384     -64
nfs_cb_sysctls                               192     128     -64
nfs4_cb_sysctls                              192     128     -64
namei_sysctls                                320     256     -64
locks_sysctls                                192     128     -64
inotify_table                                256     192     -64
inodes_sysctls                               192     128     -64
fsverity_sysctl_table                        128      64     -64
fs_stat_sysctls                              256     192     -64
fs_shared_sysctls                            192     128     -64
fs_pipe_sysctls                              256     192     -64
fs_namespace_sysctls                         128      64     -64
fs_exec_sysctls                              128      64     -64
fs_dqstats_table                             576     512     -64
fs_dcache_sysctls                            128      64     -64
fanotify_table                               256     192     -64
epoll_table                                  128      64     -64
dnotify_sysctls                              128      64     -64
coredump_sysctls                             256     192     -64
coda_table                                   256     192     -64
cachefiles_sysctls                           128      64     -64
aio_sysctls                                  192     128     -64
Total: Before=429912331, After=429910411, chg -0.00%

[5]
add/remove: 0/0 grow/shrink: 0/9 up/down: 0/-576 (-576)
Function                                     old     new   delta
root_table                                   128      64     -64
namei_sysctls                                320     256     -64
inodes_sysctls                               192     128     -64
fs_stat_sysctls                              256     192     -64
fs_shared_sysctls                            192     128     -64
fs_pipe_sysctls                              256     192     -64
fs_namespace_sysctls                         128      64     -64
fs_exec_sysctls                              128      64     -64
fs_dcache_sysctls                            128      64     -64
Total: Before=1886645, After=1886069, chg -0.03%

[6]
https://lore.kernel.org/all/20231002-jag-sysctl_remove_empty_elem_drivers-v2-0-02dd0d46f71e@xxxxxxxxxxx

[7]
To measure the in memory savings apply this on top of this patchset.

"
diff --git a/fs/proc/proc_sysctl.c b/fs/proc/proc_sysctl.c
index c88854df0b62..e0073a627bac 100644
--- a/fs/proc/proc_sysctl.c
+++ b/fs/proc/proc_sysctl.c
@@ -976,6 +976,8 @@ static struct ctl_dir *new_dir(struct ctl_table_set *set,
        table[0].procname = new_name;
        table[0].mode = S_IFDIR|S_IRUGO|S_IXUGO;
        init_header(&new->header, set->dir.header.root, set, node, table, 1);
+       // Counts additional sentinel used for each new dir.
+       printk("%ld sysctl saved mem kzalloc \n", sizeof(struct ctl_table));

        return new;
 }
@@ -1199,6 +1201,9 @@ static struct ctl_table_header *new_links(struct ctl_dir *dir, struct ctl_table_
                link_name += len;
                link++;
        }
+       // Counts additional sentinel used for each new registration
+       //
+               printk("%ld sysctl saved mem kzalloc \n", sizeof(struct ctl_table));
        init_header(links, dir->header.root, dir->header.set, node, link_table,
                    head->ctl_table_size);
        links->nreg = nr_entries;
"
and then run the following bash script in the kernel:

accum=0
for n in $(dmesg | grep kzalloc | awk '{print $3}') ; do
    echo $n
    accum=$(calc "$accum + $n")
done
echo $accum

To: Luis Chamberlain <mcgrof@xxxxxxxxxx>
To: willy@xxxxxxxxxxxxx
To: josh@xxxxxxxxxxxxxxxx
To: Kees Cook <keescook@xxxxxxxxxxxx>
To: David Howells <dhowells@xxxxxxxxxx>
To: Alexander Viro <viro@xxxxxxxxxxxxxxxxxx>
To: Christian Brauner <brauner@xxxxxxxxxx>
To: Benjamin LaHaise <bcrl@xxxxxxxxx>
To: Eric Biederman <ebiederm@xxxxxxxxxxxx>
To: Trond Myklebust <trond.myklebust@xxxxxxxxxxxxxxx>
To: Anna Schumaker <anna@xxxxxxxxxx>
To: Chuck Lever <chuck.lever@xxxxxxxxxx>
To: Jeff Layton <jlayton@xxxxxxxxxx>
To: Neil Brown <neilb@xxxxxxx>
To: Olga Kornievskaia <kolga@xxxxxxxxxx>
To: Dai Ngo <Dai.Ngo@xxxxxxxxxx>
To: Tom Talpey <tom@xxxxxxxxxx>
To: Jan Kara <jack@xxxxxxx>
To: Amir Goldstein <amir73il@xxxxxxxxx>
To: Matthew Bobrowski <repnop@xxxxxxxxxx>
To: Anton Altaparmakov <anton@xxxxxxxxxx>
To: Namjae Jeon <linkinjeon@xxxxxxxxxx>
To: Mark Fasheh <mark@xxxxxxxxxx>
To: Joel Becker <jlbec@xxxxxxxxxxxx>
To: Joseph Qi <joseph.qi@xxxxxxxxxxxxxxxxx>
To: Iurii Zaikin <yzaikin@xxxxxxxxxx>
To: Eric Biggers <ebiggers@xxxxxxxxxx>
To: "Theodore Y. Ts'o" <tytso@xxxxxxx>
To: Chandan Babu R <chandan.babu@xxxxxxxxxx>
To: "Darrick J. Wong" <djwong@xxxxxxxxxx>
To: Jan Harkes <jaharkes@xxxxxxxxxx>
To: coda@xxxxxxxxxx
Cc: linux-cachefs@xxxxxxxxxx
Cc: linux-kernel@xxxxxxxxxxxxxxx
Cc: linux-fsdevel@xxxxxxxxxxxxxxx
Cc: linux-aio@xxxxxxxxx
Cc: linux-mm@xxxxxxxxx
Cc: linux-nfs@xxxxxxxxxxxxxxx
Cc: linux-ntfs-dev@xxxxxxxxxxxxxxxxxxxxx
Cc: ocfs2-devel@xxxxxxxxxxxxxxx
Cc: fsverity@xxxxxxxxxxxxxxx
Cc: linux-xfs@xxxxxxxxxxxxxxx
Cc: codalist@xxxxxxxxxxxxxxx
---

Signed-off-by: Joel Granados <j.granados@xxxxxxxxxxx>

---
Joel Granados (4):
      cachefiles: Remove the now superfluous sentinel element from ctl_table array
      fs: Remove the now superfluous sentinel elements from ctl_table array
      sysctl:  Remove the now superfluous sentinel elements from ctl_table array
      coda:  Remove the now superfluous sentinel elements from ctl_table array

 fs/aio.c                           | 1 -
 fs/cachefiles/error_inject.c       | 1 -
 fs/coda/sysctl.c                   | 1 -
 fs/coredump.c                      | 1 -
 fs/dcache.c                        | 1 -
 fs/devpts/inode.c                  | 1 -
 fs/eventpoll.c                     | 1 -
 fs/exec.c                          | 1 -
 fs/file_table.c                    | 1 -
 fs/inode.c                         | 1 -
 fs/lockd/svc.c                     | 1 -
 fs/locks.c                         | 1 -
 fs/namei.c                         | 1 -
 fs/namespace.c                     | 1 -
 fs/nfs/nfs4sysctl.c                | 1 -
 fs/nfs/sysctl.c                    | 1 -
 fs/notify/dnotify/dnotify.c        | 1 -
 fs/notify/fanotify/fanotify_user.c | 1 -
 fs/notify/inotify/inotify_user.c   | 1 -
 fs/ntfs/sysctl.c                   | 1 -
 fs/ocfs2/stackglue.c               | 1 -
 fs/pipe.c                          | 1 -
 fs/proc/proc_sysctl.c              | 1 -
 fs/quota/dquot.c                   | 1 -
 fs/sysctls.c                       | 1 -
 fs/userfaultfd.c                   | 1 -
 fs/verity/init.c                   | 1 -
 fs/xfs/xfs_sysctl.c                | 2 --
 lib/test_sysctl.c                  | 2 --
 29 files changed, 31 deletions(-)
---
base-commit: 8b793bcda61f6c3ed4f5b2ded7530ef6749580cb
change-id: 20231107-jag-sysctl_remove_empty_elem_fs-dbdcf6581fbe

Best regards,
-- 
Joel Granados <j.granados@xxxxxxxxxxx>





[Index of Archives]     [Linux Ext4 Filesystem]     [Union Filesystem]     [Filesystem Testing]     [Ceph Users]     [Ecryptfs]     [NTFS 3]     [AutoFS]     [Kernel Newbies]     [Share Photos]     [Security]     [Netfilter]     [Bugtraq]     [Yosemite News]     [MIPS Linux]     [ARM Linux]     [Linux Security]     [Linux Cachefs]     [Reiser Filesystem]     [Linux RAID]     [NTFS 3]     [Samba]     [Device Mapper]     [CEPH Development]

  Powered by Linux