linux-stable.git/fs/proc/generic.c, branch v3.19

fs/proc.c: use rb_entry_safe() instead of rb_entry()

2014-12-11T01:41:09+00:00

Better to use existing macro that rewriting them.

Signed-off-by: Nicolas Dichtel 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

procfs: fix error handling of proc_register()

2014-12-11T01:41:09+00:00

proc_register() error paths are leaking inodes and directory refcounts.

Signed-off-by: Debabrata Banerjee 
Cc: Alexander Viro 
Acked-by: Nicolas Dichtel 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

fs/proc: use a rb tree for the directory entries

2014-12-11T01:41:09+00:00

When a lot of netdevices are created, one of the bottleneck is the
creation of proc entries.  This serie aims to accelerate this part.

The current implementation for the directories in /proc is using a single
linked list.  This is slow when handling directories with large numbers of
entries (eg netdevice-related entries when lots of tunnels are opened).

This patch replaces this linked list by a red-black tree.

Here are some numbers:

dummy30000.batch contains 30 000 times 'link add type dummy'.

Before the patch:
  $ time ip -b dummy30000.batch
  real    2m31.950s
  user    0m0.440s
  sys     2m21.440s
  $ time rmmod dummy
  real    1m35.764s
  user    0m0.000s
  sys     1m24.088s

After the patch:
  $ time ip -b dummy30000.batch
  real    2m0.874s
  user    0m0.448s
  sys     1m49.720s
  $ time rmmod dummy
  real    1m13.988s
  user    0m0.000s
  sys     1m1.008s

The idea of improving this part was suggested by Thierry Herbelot.

[akpm@linux-foundation.org: initialise proc_root.subdir at compile time]
Signed-off-by: Nicolas Dichtel 
Acked-by: David S. Miller 
Cc: Thierry Herbelot .
Acked-by: "Eric W. Biederman" 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

proc: make proc_subdir_lock static

2014-08-08T22:57:22+00:00

Signed-off-by: Alexey Dobriyan 
Cc: Al Viro 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

proc: add and remove /proc entry create checks

2014-08-08T22:57:22+00:00

* remove proc_create(NULL, ...) check, let it oops

* warn about proc_create("", ...) and proc_create("very very long name", ...)
  proc code keeps length as u8, no 256+ name length possible

* warn about proc_create("123", ...)
  /proc/$PID and /proc/misc namespaces are separate things,
  but dumb module might create funky a-la $PID entry.

* remove post mortem strchr('/') check
  Triggering it implies either strchr() is buggy or memory corruption.
  It should be VFS check anyway.

In reality, none of these checks will ever trigger,
it is preparation for the next patch.

Based on patch from Al Viro.

Signed-off-by: Alexey Dobriyan 
Cc: Al Viro 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

proc: set attributes of pde using accessor functions

2014-01-24T00:37:01+00:00

Use existing accessors proc_set_user() and proc_set_size() to set
attributes.  Just a cleanup.

Signed-off-by: Rui Xiang 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

consolidate simple ->d_delete() instances

2013-11-16T03:04:17+00:00

Rename simple_delete_dentry() to always_delete_dentry() and export it.
Export simple_dentry_operations, while we are at it, and get rid of
their duplicates

Signed-off-by: Al Viro

proc: more readdir conversion bug-fixes

2013-08-19T23:26:12+00:00

In the previous commit, Richard Genoud fixed proc_root_readdir(), which
had lost the check for whether all of the non-process /proc entries had
been returned or not.

But that in turn exposed _another_ bug, namely that the original readdir
conversion patch had yet another problem: it had lost the return value
of proc_readdir_de(), so now checking whether it had completed
successfully or not didn't actually work right anyway.

This reinstates the non-zero return for the "end of base entries" that
had also gotten lost in commit f0c3b5093add ("[readdir] convert
procfs").  So now you get all the base entries *and* you get all the
process entries, regardless of getdents buffer size.

(Side note: the Linux "getdents" manual page actually has a nice example
application for testing getdents, which can be easily modified to use
different buffers.  Who knew? Man-pages can be useful)

Reported-by: Emmanuel Benisty 
Reported-by: Marc Dionne 
Cc: Richard Genoud 
Cc: Al Viro 
Signed-off-by: Linus Torvalds

[readdir] convert procfs

2013-06-29T08:56:32+00:00

Signed-off-by: Al Viro

proc: Make the PROC_I() and PDE() macros internal to procfs

2013-05-01T21:29:47+00:00

Make the PROC_I() and PDE() macros internal to procfs.  This means making
PDE_DATA() out of line.  This could be made more optimal by storing
PDE()->data into inode->i_private.

Also provide a __PDE_DATA() that is inline and internal to procfs.

Signed-off-by: David Howells 
Signed-off-by: Al Viro