summaryrefslogtreecommitdiff
path: root/Documentation/filesystems
diff options
context:
space:
mode:
authorRandy Dunlap <rdunlap@infradead.org>2019-01-06 19:19:29 -0800
committerJonathan Corbet <corbet@lwn.net>2019-01-07 15:24:16 -0700
commit9f63df26beea78aae0500d1225a99bbd905dd157 (patch)
tree1883198ac737c87ed21439f22c85378a46c4b880 /Documentation/filesystems
parentbfeffd155283772bbe78c6a05dec7c0128ee500c (diff)
Documentation/filesystems: fix title underline lengths in path-lookup.rst
Fix Sphinx warnings in path-lookup.rst: Documentation/filesystems/path-lookup.rst:347: WARNING: Title underline too short. Documentation/filesystems/path-lookup.rst:358: WARNING: Title underline too short. [...] Signed-off-by: Randy Dunlap <rdunlap@infradead.org> Acked-by: NeilBrown <neilb@suse.com> Signed-off-by: Jonathan Corbet <corbet@lwn.net>
Diffstat (limited to 'Documentation/filesystems')
-rw-r--r--Documentation/filesystems/path-lookup.rst24
1 files changed, 12 insertions, 12 deletions
diff --git a/Documentation/filesystems/path-lookup.rst b/Documentation/filesystems/path-lookup.rst
index 9d6b68853f5b..80e22eda4132 100644
--- a/Documentation/filesystems/path-lookup.rst
+++ b/Documentation/filesystems/path-lookup.rst
@@ -344,7 +344,7 @@ In particular it is held while scanning chains in the dcache hash
table, and the mount point hash table.
Bringing it together with ``struct nameidata``
---------------------------------------------
+----------------------------------------------
.. _First edition Unix: http://minnie.tuhs.org/cgi-bin/utree.pl?file=V1/u2.s
@@ -355,7 +355,7 @@ converts a "name" to an "inode". ``struct nameidata`` contains (among
other fields):
``struct path path``
-~~~~~~~~~~~~~~~~~~
+~~~~~~~~~~~~~~~~~~~~
A ``path`` contains a ``struct vfsmount`` (which is
embedded in a ``struct mount``) and a ``struct dentry``. Together these
@@ -366,13 +366,13 @@ step. A reference through ``d_lockref`` and ``mnt_count`` is always
held.
``struct qstr last``
-~~~~~~~~~~~~~~~~~~
+~~~~~~~~~~~~~~~~~~~~
This is a string together with a length (i.e. _not_ ``nul`` terminated)
that is the "next" component in the pathname.
``int last_type``
-~~~~~~~~~~~~~~~
+~~~~~~~~~~~~~~~~~
This is one of ``LAST_NORM``, ``LAST_ROOT``, ``LAST_DOT``, ``LAST_DOTDOT``, or
``LAST_BIND``. The ``last`` field is only valid if the type is
@@ -381,7 +381,7 @@ components of the symlink have been processed yet. Others should be
fairly self-explanatory.
``struct path root``
-~~~~~~~~~~~~~~~~~~
+~~~~~~~~~~~~~~~~~~~~
This is used to hold a reference to the effective root of the
filesystem. Often that reference won't be needed, so this field is
@@ -510,7 +510,7 @@ potentially interesting things about these dentries corresponding
to three different flags that might be set in ``dentry->d_flags``:
``DCACHE_MANAGE_TRANSIT``
-~~~~~~~~~~~~~~~~~~~~~~~
+~~~~~~~~~~~~~~~~~~~~~~~~~
If this flag has been set, then the filesystem has requested that the
``d_manage()`` dentry operation be called before handling any possible
@@ -529,7 +529,7 @@ filesystem, which will then give it a special pass through
``d_manage()`` by returning ``-EISDIR``.
``DCACHE_MOUNTED``
-~~~~~~~~~~~~~~~~
+~~~~~~~~~~~~~~~~~~
This flag is set on every dentry that is mounted on. As Linux
supports multiple filesystem namespaces, it is possible that the
@@ -542,7 +542,7 @@ If this flag is set, and ``d_manage()`` didn't return ``-EISDIR``,
and a new ``dentry`` (both with counted references).
``DCACHE_NEED_AUTOMOUNT``
-~~~~~~~~~~~~~~~~~~~~~~~
+~~~~~~~~~~~~~~~~~~~~~~~~~
If ``d_manage()`` allowed us to get this far, and ``lookup_mnt()`` didn't
find a mount point, then this flag causes the ``d_automount()`` dentry
@@ -698,7 +698,7 @@ With that little refresher on seqlocks out of the way we can look at
the bigger picture of how RCU-walk uses seqlocks.
``mount_lock`` and ``nd->m_seq``
-~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
We already met the ``mount_lock`` seqlock when REF-walk used it to
ensure that crossing a mount point is performed safely. RCU-walk uses
@@ -727,7 +727,7 @@ results would have been the same. This ensures the invariant holds,
at least for vfsmount structures.
``dentry->d_seq`` and ``nd->seq``
-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
In place of taking a count or lock on ``d_reflock``, RCU-walk samples
the per-dentry ``d_seq`` seqlock, and stores the sequence number in the
@@ -774,7 +774,7 @@ getting a counted reference to the new dentry before dropping that for
the old dentry which we saw in REF-walk.
No ``inode->i_rwsem`` or even ``rename_lock``
-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
A semaphore is a fairly heavyweight lock that can only be taken when it is
permissible to sleep. As ``rcu_read_lock()`` forbids sleeping,
@@ -796,7 +796,7 @@ locking. This neatly handles all cases, so adding extra checks on
rename_lock would bring no significant value.
``unlazy walk()`` and ``complete_walk()``
--------------------------------------
+-----------------------------------------
That "dropping down to REF-walk" typically involves a call to
``unlazy_walk()``, so named because "RCU-walk" is also sometimes