[PATCH v6 0/8] fix debugfs file removal races

[PATCH v6 0/8] fix debugfs file removal races

[Nicolai Stange]

Hi Greg,

this series' v6 is meant to replace the v4 currently sitting in your
driver-core tree's driver-core-testing branch.

All the three fixup patches to v4 (not applied anywhere) have been
incorporated into this v6. I'll explicitly mark them as "please drop".


Former v5 thread can be found here:

  http://lkml.kernel.org/g/1457267103-28910-1-git-send-email-nicstange@gmail.com


Applicable to current mainline.


Thanks,

Nicolai


Changes v5 -> v6:
  [1/8] ("debugfs: prevent access to possibly dead file_operations at file open")
   - Fix the double unlock issue found by Rasmus Villemoes and reported
     in reply to v5.


Changes v4 -> v5:
  [1/8] ("debugfs: prevent access to possibly dead file_operations at file open")
   - In the DocBook comment for debugfs_use_file_start(), rename the
     "@file" parameter into the now correct "@dentry".

  [2/8] ("debugfs: prevent access to removed files' private data)"
   - In include/linux/debugfs.h, add the inline keyword to the
     CONFIG_DEBUG_FS=n dummy implementations of
     debugfs_use_file_start()/_finish()

  [3-8/8] unchanged


Changes v3 -> v4:
  [4/8] ("debugfs, coccinelle: check for obsolete DEFINE_SIMPLE_ATTRIBUTE() usage")
   - This one is new, the Coccinelle related changes have been split
     off from former
       [3/7] ("debugfs: add support for self-protecting attribute file fops")
     into this patch.
   - Style fixes as suggested by Julia Lawall have been applied to the
     contained cocci script's comment header.

  [5/8] ("debugfs: unproxify integer attribute files")
   - The commit messages has been reworded in order to get rid of the
     unfortunate triple-X in former [4/7].

  [6-8/8]
   - Former [5-7/8], only the numbering has changed.


Changes v2 -> v3:
  [1/7] ("debugfs: prevent access to possibly dead file_operations at file open")
   - move the definition of the debugfs_use_file_start() and _end() from former
     [2/2] to [1/7]. Also, they've been renamed from debugfs_file_use_data*().
   - Make the ->open() proxy use the debugfs_use_file_*() helpers.
   - In debugfs_use_file_start(), use d_unlinked() rather than
     (->d_fsdata == NULL) as a flag whether the dentry is dead.
   - Make the ->open() proxy include the forwarded call to the original fops' ->open
     within the SRCU read side critical section.
   - debugfs_proxy_file_operations has been renamed to
     "debugfs_open_proxy_file_operations"  to distinguish it from the full proxy
     introduced in [2/7].

  [2/7] ("debugfs: prevent access to removed files' private data")
   - This one has changed completely: instead of providing file
     removal-safe fops helpers to opt-into at the debugfs users, the
     original struct file_operations get completely and
     unconditionally proxied now.

  [3-7/7]
   New. Opt-out from the full proxying introduced in [2/7] for some
   special case struct file_operations provided by debugfs itself.


Changes v1 -> v2:
  [1/2] ("debugfs: prevent access to possibly dead file_operations at file open")
   - Resolve trivial diff conflict in debugfs_remove_recursive():
     in the meanwhile, an unrelated 'mutex_unlock(...)' had been rewritten to
     'inode_unlock(...)' which broke the diff's context.
   - Introduce the fs/debugfs/internal.h header and move the declarations of
     debugfs_noop_file_operations, debugfs_proxy_file_operations and
     debugfs_rcu from include/linux/debugfs.h thereinto. Include this header
     from file.c and inode.c.
   - Add a word about the new internal header to the commit message.
   - Move the inclusion of linux/srcu.h from include/linux/debugfs.h
     into file.c and inode.c respectively.

  [2/2] ("debugfs: prevent access to removed files' private data")
   - Move the definitions of debugfs_file_use_data_start() and
     debugfs_file_use_data_finish() from include/linux/debugfs.h to
     file.c. Export them and keep their declarations in debugfs.h,
   - In order to be able to attach proper __acquires() and __releases() tags
     to the decalarations of debugfs_file_use_data_*() in debugfs.h,
     move the debugfs_srcu declaration from internal.h into debugfs.h.
   - Since the definitions as well as the docstrings of
     debugfs_file_use_data_*() have been moved into file.c,
     there is no need to run DocBook on debugfs.h: do not modify
     Documentation/DocBook/filesystems.tmpl anymore.
   - In the commit message, encourage new users of debugfs to prefer
     DEFINE_DEBUGFS_ATTRIBUTE() and friends over DEFINE_SIMPLE_ATTRIBUTE().



Nicolai Stange (8):
  debugfs: prevent access to possibly dead file_operations at file open
  debugfs: prevent access to removed files' private data
  debugfs: add support for self-protecting attribute file fops
  debugfs, coccinelle: check for obsolete DEFINE_SIMPLE_ATTRIBUTE()
    usage
  debugfs: unproxify integer attribute files
  debugfs: unproxify files created through debugfs_create_bool()
  debugfs: unproxify files created through debugfs_create_blob()
  debugfs: unproxify files created through debugfs_create_u32_array()

 fs/debugfs/file.c                                  | 436 +++++++++++++++++----
 fs/debugfs/inode.c                                 | 101 ++++-
 fs/debugfs/internal.h                              |  26 ++
 include/linux/debugfs.h                            |  49 ++-
 lib/Kconfig.debug                                  |   1 +
 .../api/debugfs/debugfs_simple_attr.cocci          |  67 ++++
 6 files changed, 592 insertions(+), 88 deletions(-)
 create mode 100644 fs/debugfs/internal.h
 create mode 100644 scripts/coccinelle/api/debugfs/debugfs_simple_attr.cocci

-- 
2.7.4

[Cocci] [PATCH v6 0/8] fix debugfs file removal races

[Nicolai Stange]

Hi Greg,

this series' v6 is meant to replace the v4 currently sitting in your
driver-core tree's driver-core-testing branch.

All the three fixup patches to v4 (not applied anywhere) have been
incorporated into this v6. I'll explicitly mark them as "please drop".


Former v5 thread can be found here:

  http://lkml.kernel.org/g/1457267103-28910-1-git-send-email-nicstange at gmail.com


Applicable to current mainline.


Thanks,

Nicolai


Changes v5 -> v6:
  [1/8] ("debugfs: prevent access to possibly dead file_operations at file open")
   - Fix the double unlock issue found by Rasmus Villemoes and reported
     in reply to v5.


Changes v4 -> v5:
  [1/8] ("debugfs: prevent access to possibly dead file_operations at file open")
   - In the DocBook comment for debugfs_use_file_start(), rename the
     "@file" parameter into the now correct "@dentry".

  [2/8] ("debugfs: prevent access to removed files' private data)"
   - In include/linux/debugfs.h, add the inline keyword to the
     CONFIG_DEBUG_FS=n dummy implementations of
     debugfs_use_file_start()/_finish()

  [3-8/8] unchanged


Changes v3 -> v4:
  [4/8] ("debugfs, coccinelle: check for obsolete DEFINE_SIMPLE_ATTRIBUTE() usage")
   - This one is new, the Coccinelle related changes have been split
     off from former
       [3/7] ("debugfs: add support for self-protecting attribute file fops")
     into this patch.
   - Style fixes as suggested by Julia Lawall have been applied to the
     contained cocci script's comment header.

  [5/8] ("debugfs: unproxify integer attribute files")
   - The commit messages has been reworded in order to get rid of the
     unfortunate triple-X in former [4/7].

  [6-8/8]
   - Former [5-7/8], only the numbering has changed.


Changes v2 -> v3:
  [1/7] ("debugfs: prevent access to possibly dead file_operations at file open")
   - move the definition of the debugfs_use_file_start() and _end() from former
     [2/2] to [1/7]. Also, they've been renamed from debugfs_file_use_data*().
   - Make the ->open() proxy use the debugfs_use_file_*() helpers.
   - In debugfs_use_file_start(), use d_unlinked() rather than
     (->d_fsdata == NULL) as a flag whether the dentry is dead.
   - Make the ->open() proxy include the forwarded call to the original fops' ->open
     within the SRCU read side critical section.
   - debugfs_proxy_file_operations has been renamed to
     "debugfs_open_proxy_file_operations"  to distinguish it from the full proxy
     introduced in [2/7].

  [2/7] ("debugfs: prevent access to removed files' private data")
   - This one has changed completely: instead of providing file
     removal-safe fops helpers to opt-into at the debugfs users, the
     original struct file_operations get completely and
     unconditionally proxied now.

  [3-7/7]
   New. Opt-out from the full proxying introduced in [2/7] for some
   special case struct file_operations provided by debugfs itself.


Changes v1 -> v2:
  [1/2] ("debugfs: prevent access to possibly dead file_operations at file open")
   - Resolve trivial diff conflict in debugfs_remove_recursive():
     in the meanwhile, an unrelated 'mutex_unlock(...)' had been rewritten to
     'inode_unlock(...)' which broke the diff's context.
   - Introduce the fs/debugfs/internal.h header and move the declarations of
     debugfs_noop_file_operations, debugfs_proxy_file_operations and
     debugfs_rcu from include/linux/debugfs.h thereinto. Include this header
     from file.c and inode.c.
   - Add a word about the new internal header to the commit message.
   - Move the inclusion of linux/srcu.h from include/linux/debugfs.h
     into file.c and inode.c respectively.

  [2/2] ("debugfs: prevent access to removed files' private data")
   - Move the definitions of debugfs_file_use_data_start() and
     debugfs_file_use_data_finish() from include/linux/debugfs.h to
     file.c. Export them and keep their declarations in debugfs.h,
   - In order to be able to attach proper __acquires() and __releases() tags
     to the decalarations of debugfs_file_use_data_*() in debugfs.h,
     move the debugfs_srcu declaration from internal.h into debugfs.h.
   - Since the definitions as well as the docstrings of
     debugfs_file_use_data_*() have been moved into file.c,
     there is no need to run DocBook on debugfs.h: do not modify
     Documentation/DocBook/filesystems.tmpl anymore.
   - In the commit message, encourage new users of debugfs to prefer
     DEFINE_DEBUGFS_ATTRIBUTE() and friends over DEFINE_SIMPLE_ATTRIBUTE().



Nicolai Stange (8):
  debugfs: prevent access to possibly dead file_operations at file open
  debugfs: prevent access to removed files' private data
  debugfs: add support for self-protecting attribute file fops
  debugfs, coccinelle: check for obsolete DEFINE_SIMPLE_ATTRIBUTE()
    usage
  debugfs: unproxify integer attribute files
  debugfs: unproxify files created through debugfs_create_bool()
  debugfs: unproxify files created through debugfs_create_blob()
  debugfs: unproxify files created through debugfs_create_u32_array()

 fs/debugfs/file.c                                  | 436 +++++++++++++++++----
 fs/debugfs/inode.c                                 | 101 ++++-
 fs/debugfs/internal.h                              |  26 ++
 include/linux/debugfs.h                            |  49 ++-
 lib/Kconfig.debug                                  |   1 +
 .../api/debugfs/debugfs_simple_attr.cocci          |  67 ++++
 6 files changed, 592 insertions(+), 88 deletions(-)
 create mode 100644 fs/debugfs/internal.h
 create mode 100644 scripts/coccinelle/api/debugfs/debugfs_simple_attr.cocci

-- 
2.7.4

[PATCH v6 1/8] debugfs: prevent access to possibly dead file_operations at file open

[Nicolai Stange]

Nothing prevents a dentry found by path lookup before a return of
__debugfs_remove() to actually get opened after that return. Now, after
the return of __debugfs_remove(), there are no guarantees whatsoever
regarding the memory the corresponding inode's file_operations object
had been kept in.

Since __debugfs_remove() is seldomly invoked, usually from module exit
handlers only, the race is hard to trigger and the impact is very low.

A discussion of the problem outlined above as well as a suggested
solution can be found in the (sub-)thread rooted at

  http://lkml.kernel.org/g/20130401203445.GA20862@ZenIV.linux.org.uk
  ("Yet another pipe related oops.")

Basically, Greg KH suggests to introduce an intermediate fops and
Al Viro points out that a pointer to the original ones may be stored in
->d_fsdata.

Follow this line of reasoning:
- Add SRCU as a reverse dependency of DEBUG_FS.
- Introduce a srcu_struct object for the debugfs subsystem.
- In debugfs_create_file(), store a pointer to the original
  file_operations object in ->d_fsdata.
- Make debugfs_remove() and debugfs_remove_recursive() wait for a
  SRCU grace period after the dentry has been delete()'d and before they
  return to their callers.
- Introduce an intermediate file_operations object named
  "debugfs_open_proxy_file_operations". It's ->open() functions checks,
  under the protection of a SRCU read lock, whether the dentry is still
  alive, i.e. has not been d_delete()'d and if so, tries to acquire a
  reference on the owning module.
  On success, it sets the file object's ->f_op to the original
  file_operations and forwards the ongoing open() call to the original
  ->open().
- For clarity, rename the former debugfs_file_operations to
  debugfs_noop_file_operations -- they are in no way canonical.

The choice of SRCU over "normal" RCU is justified by the fact, that the
former may also be used to protect ->i_private data from going away
during the execution of a file's readers and writers which may (and do)
sleep.

Finally, introduce the fs/debugfs/internal.h header containing some
declarations internal to the debugfs implementation.

Signed-off-by: Nicolai Stange <nicstange@gmail.com>
---
 fs/debugfs/file.c       | 91 ++++++++++++++++++++++++++++++++++++++++++++++++-
 fs/debugfs/inode.c      | 13 ++++++-
 fs/debugfs/internal.h   | 24 +++++++++++++
 include/linux/debugfs.h |  3 --
 lib/Kconfig.debug       |  1 +
 5 files changed, 127 insertions(+), 5 deletions(-)
 create mode 100644 fs/debugfs/internal.h

diff --git a/fs/debugfs/file.c b/fs/debugfs/file.c
index d2ba12e..736ab3c 100644
--- a/fs/debugfs/file.c
+++ b/fs/debugfs/file.c
@@ -22,6 +22,9 @@
 #include <linux/slab.h>
 #include <linux/atomic.h>
 #include <linux/device.h>
+#include <linux/srcu.h>
+
+#include "internal.h"
 
 static ssize_t default_read_file(struct file *file, char __user *buf,
 				 size_t count, loff_t *ppos)
@@ -35,13 +38,99 @@ static ssize_t default_write_file(struct file *file, const char __user *buf,
 	return count;
 }
 
-const struct file_operations debugfs_file_operations = {
+const struct file_operations debugfs_noop_file_operations = {
 	.read =		default_read_file,
 	.write =	default_write_file,
 	.open =		simple_open,
 	.llseek =	noop_llseek,
 };
 
+/**
+ * debugfs_use_file_start - mark the beginning of file data access
+ * @dentry: the dentry object whose data is being accessed.
+ * @srcu_idx: a pointer to some memory to store a SRCU index in.
+ *
+ * Up to a matching call to debugfs_use_file_finish(), any
+ * successive call into the file removing functions debugfs_remove()
+ * and debugfs_remove_recursive() will block. Since associated private
+ * file data may only get freed after a successful return of any of
+ * the removal functions, you may safely access it after a successful
+ * call to debugfs_use_file_start() without worrying about
+ * lifetime issues.
+ *
+ * If -%EIO is returned, the file has already been removed and thus,
+ * it is not safe to access any of its data. If, on the other hand,
+ * it is allowed to access the file data, zero is returned.
+ *
+ * Regardless of the return code, any call to
+ * debugfs_use_file_start() must be followed by a matching call
+ * to debugfs_use_file_finish().
+ */
+static int debugfs_use_file_start(const struct dentry *dentry, int *srcu_idx)
+	__acquires(&debugfs_srcu)
+{
+	*srcu_idx = srcu_read_lock(&debugfs_srcu);
+	barrier();
+	if (d_unlinked(dentry))
+		return -EIO;
+	return 0;
+}
+
+/**
+ * debugfs_use_file_finish - mark the end of file data access
+ * @srcu_idx: the SRCU index "created" by a former call to
+ *            debugfs_use_file_start().
+ *
+ * Allow any ongoing concurrent call into debugfs_remove() or
+ * debugfs_remove_recursive() blocked by a former call to
+ * debugfs_use_file_start() to proceed and return to its caller.
+ */
+static void debugfs_use_file_finish(int srcu_idx) __releases(&debugfs_srcu)
+{
+	srcu_read_unlock(&debugfs_srcu, srcu_idx);
+}
+
+#define F_DENTRY(filp) ((filp)->f_path.dentry)
+
+#define REAL_FOPS_DEREF(dentry)					\
+	((const struct file_operations *)(dentry)->d_fsdata)
+
+static int open_proxy_open(struct inode *inode, struct file *filp)
+{
+	const struct dentry *dentry = F_DENTRY(filp);
+	const struct file_operations *real_fops = NULL;
+	int srcu_idx, r;
+
+	r = debugfs_use_file_start(dentry, &srcu_idx);
+	if (r) {
+		r = -ENOENT;
+		goto out;
+	}
+
+	real_fops = REAL_FOPS_DEREF(dentry);
+	real_fops = fops_get(real_fops);
+	if (!real_fops) {
+		/* Huh? Module did not clean up after itself at exit? */
+		WARN(1, "debugfs file owner did not clean up at exit: %pd",
+			dentry);
+		r = -ENXIO;
+		goto out;
+	}
+	replace_fops(filp, real_fops);
+
+	if (real_fops->open)
+		r = real_fops->open(inode, filp);
+
+out:
+	fops_put(real_fops);
+	debugfs_use_file_finish(srcu_idx);
+	return r;
+}
+
+const struct file_operations debugfs_open_proxy_file_operations = {
+	.open = open_proxy_open,
+};
+
 static struct dentry *debugfs_create_mode(const char *name, umode_t mode,
 					  struct dentry *parent, void *value,
 				          const struct file_operations *fops,
diff --git a/fs/debugfs/inode.c b/fs/debugfs/inode.c
index bece948..26880ad 100644
--- a/fs/debugfs/inode.c
+++ b/fs/debugfs/inode.c
@@ -27,9 +27,14 @@
 #include <linux/parser.h>
 #include <linux/magic.h>
 #include <linux/slab.h>
+#include <linux/srcu.h>
+
+#include "internal.h"
 
 #define DEBUGFS_DEFAULT_MODE	0700
 
+DEFINE_SRCU(debugfs_srcu);
+
 static struct vfsmount *debugfs_mount;
 static int debugfs_mount_count;
 static bool debugfs_registered;
@@ -340,8 +345,12 @@ struct dentry *debugfs_create_file(const char *name, umode_t mode,
 		return failed_creating(dentry);
 
 	inode->i_mode = mode;
-	inode->i_fop = fops ? fops : &debugfs_file_operations;
 	inode->i_private = data;
+
+	inode->i_fop = fops ? &debugfs_open_proxy_file_operations
+		: &debugfs_noop_file_operations;
+	dentry->d_fsdata = (void *)fops;
+
 	d_instantiate(dentry, inode);
 	fsnotify_create(d_inode(dentry->d_parent), dentry);
 	return end_creating(dentry);
@@ -565,6 +574,7 @@ void debugfs_remove(struct dentry *dentry)
 	inode_unlock(d_inode(parent));
 	if (!ret)
 		simple_release_fs(&debugfs_mount, &debugfs_mount_count);
+	synchronize_srcu(&debugfs_srcu);
 }
 EXPORT_SYMBOL_GPL(debugfs_remove);
 
@@ -642,6 +652,7 @@ void debugfs_remove_recursive(struct dentry *dentry)
 	if (!__debugfs_remove(child, parent))
 		simple_release_fs(&debugfs_mount, &debugfs_mount_count);
 	inode_unlock(d_inode(parent));
+	synchronize_srcu(&debugfs_srcu);
 }
 EXPORT_SYMBOL_GPL(debugfs_remove_recursive);
 
diff --git a/fs/debugfs/internal.h b/fs/debugfs/internal.h
new file mode 100644
index 0000000..c7aaa5c
--- /dev/null
+++ b/fs/debugfs/internal.h
@@ -0,0 +1,24 @@
+/*
+ *  internal.h - declarations internal to debugfs
+ *
+ *  Copyright (C) 2016 Nicolai Stange <nicstange@gmail.com>
+ *
+ *	This program is free software; you can redistribute it and/or
+ *	modify it under the terms of the GNU General Public License version
+ *	2 as published by the Free Software Foundation.
+ *
+ */
+
+#ifndef _DEBUGFS_INTERNAL_H_
+#define _DEBUGFS_INTERNAL_H_
+
+struct file_operations;
+struct srcu_struct;
+
+/* declared over in file.c */
+extern const struct file_operations debugfs_noop_file_operations;
+extern const struct file_operations debugfs_open_proxy_file_operations;
+
+extern struct srcu_struct debugfs_srcu;
+
+#endif /* _DEBUGFS_INTERNAL_H_ */
diff --git a/include/linux/debugfs.h b/include/linux/debugfs.h
index 981e53a..fcafe2d 100644
--- a/include/linux/debugfs.h
+++ b/include/linux/debugfs.h
@@ -43,9 +43,6 @@ extern struct dentry *arch_debugfs_dir;
 
 #if defined(CONFIG_DEBUG_FS)
 
-/* declared over in file.c */
-extern const struct file_operations debugfs_file_operations;
-
 struct dentry *debugfs_create_file(const char *name, umode_t mode,
 				   struct dentry *parent, void *data,
 				   const struct file_operations *fops);
diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug
index 5a60f45..0d5177f 100644
--- a/lib/Kconfig.debug
+++ b/lib/Kconfig.debug
@@ -257,6 +257,7 @@ config PAGE_OWNER
 
 config DEBUG_FS
 	bool "Debug Filesystem"
+	select SRCU
 	help
 	  debugfs is a virtual file system that kernel developers use to put
 	  debugging files into.  Enable this option to be able to read and
-- 
2.7.4

[Cocci] [PATCH v6 1/8] debugfs: prevent access to possibly dead file_operations at file open

[Nicolai Stange]

Nothing prevents a dentry found by path lookup before a return of
__debugfs_remove() to actually get opened after that return. Now, after
the return of __debugfs_remove(), there are no guarantees whatsoever
regarding the memory the corresponding inode's file_operations object
had been kept in.

Since __debugfs_remove() is seldomly invoked, usually from module exit
handlers only, the race is hard to trigger and the impact is very low.

A discussion of the problem outlined above as well as a suggested
solution can be found in the (sub-)thread rooted at

  http://lkml.kernel.org/g/20130401203445.GA20862 at ZenIV.linux.org.uk
  ("Yet another pipe related oops.")

Basically, Greg KH suggests to introduce an intermediate fops and
Al Viro points out that a pointer to the original ones may be stored in
->d_fsdata.

Follow this line of reasoning:
- Add SRCU as a reverse dependency of DEBUG_FS.
- Introduce a srcu_struct object for the debugfs subsystem.
- In debugfs_create_file(), store a pointer to the original
  file_operations object in ->d_fsdata.
- Make debugfs_remove() and debugfs_remove_recursive() wait for a
  SRCU grace period after the dentry has been delete()'d and before they
  return to their callers.
- Introduce an intermediate file_operations object named
  "debugfs_open_proxy_file_operations". It's ->open() functions checks,
  under the protection of a SRCU read lock, whether the dentry is still
  alive, i.e. has not been d_delete()'d and if so, tries to acquire a
  reference on the owning module.
  On success, it sets the file object's ->f_op to the original
  file_operations and forwards the ongoing open() call to the original
  ->open().
- For clarity, rename the former debugfs_file_operations to
  debugfs_noop_file_operations -- they are in no way canonical.

The choice of SRCU over "normal" RCU is justified by the fact, that the
former may also be used to protect ->i_private data from going away
during the execution of a file's readers and writers which may (and do)
sleep.

Finally, introduce the fs/debugfs/internal.h header containing some
declarations internal to the debugfs implementation.

Signed-off-by: Nicolai Stange <nicstange@gmail.com>
---
 fs/debugfs/file.c       | 91 ++++++++++++++++++++++++++++++++++++++++++++++++-
 fs/debugfs/inode.c      | 13 ++++++-
 fs/debugfs/internal.h   | 24 +++++++++++++
 include/linux/debugfs.h |  3 --
 lib/Kconfig.debug       |  1 +
 5 files changed, 127 insertions(+), 5 deletions(-)
 create mode 100644 fs/debugfs/internal.h

diff --git a/fs/debugfs/file.c b/fs/debugfs/file.c
index d2ba12e..736ab3c 100644
--- a/fs/debugfs/file.c
+++ b/fs/debugfs/file.c
@@ -22,6 +22,9 @@
 #include <linux/slab.h>
 #include <linux/atomic.h>
 #include <linux/device.h>
+#include <linux/srcu.h>
+
+#include "internal.h"
 
 static ssize_t default_read_file(struct file *file, char __user *buf,
 				 size_t count, loff_t *ppos)
@@ -35,13 +38,99 @@ static ssize_t default_write_file(struct file *file, const char __user *buf,
 	return count;
 }
 
-const struct file_operations debugfs_file_operations = {
+const struct file_operations debugfs_noop_file_operations = {
 	.read =		default_read_file,
 	.write =	default_write_file,
 	.open =		simple_open,
 	.llseek =	noop_llseek,
 };
 
+/**
+ * debugfs_use_file_start - mark the beginning of file data access
+ * @dentry: the dentry object whose data is being accessed.
+ * @srcu_idx: a pointer to some memory to store a SRCU index in.
+ *
+ * Up to a matching call to debugfs_use_file_finish(), any
+ * successive call into the file removing functions debugfs_remove()
+ * and debugfs_remove_recursive() will block. Since associated private
+ * file data may only get freed after a successful return of any of
+ * the removal functions, you may safely access it after a successful
+ * call to debugfs_use_file_start() without worrying about
+ * lifetime issues.
+ *
+ * If -%EIO is returned, the file has already been removed and thus,
+ * it is not safe to access any of its data. If, on the other hand,
+ * it is allowed to access the file data, zero is returned.
+ *
+ * Regardless of the return code, any call to
+ * debugfs_use_file_start() must be followed by a matching call
+ * to debugfs_use_file_finish().
+ */
+static int debugfs_use_file_start(const struct dentry *dentry, int *srcu_idx)
+	__acquires(&debugfs_srcu)
+{
+	*srcu_idx = srcu_read_lock(&debugfs_srcu);
+	barrier();
+	if (d_unlinked(dentry))
+		return -EIO;
+	return 0;
+}
+
+/**
+ * debugfs_use_file_finish - mark the end of file data access
+ * @srcu_idx: the SRCU index "created" by a former call to
+ *            debugfs_use_file_start().
+ *
+ * Allow any ongoing concurrent call into debugfs_remove() or
+ * debugfs_remove_recursive() blocked by a former call to
+ * debugfs_use_file_start() to proceed and return to its caller.
+ */
+static void debugfs_use_file_finish(int srcu_idx) __releases(&debugfs_srcu)
+{
+	srcu_read_unlock(&debugfs_srcu, srcu_idx);
+}
+
+#define F_DENTRY(filp) ((filp)->f_path.dentry)
+
+#define REAL_FOPS_DEREF(dentry)					\
+	((const struct file_operations *)(dentry)->d_fsdata)
+
+static int open_proxy_open(struct inode *inode, struct file *filp)
+{
+	const struct dentry *dentry = F_DENTRY(filp);
+	const struct file_operations *real_fops = NULL;
+	int srcu_idx, r;
+
+	r = debugfs_use_file_start(dentry, &srcu_idx);
+	if (r) {
+		r = -ENOENT;
+		goto out;
+	}
+
+	real_fops = REAL_FOPS_DEREF(dentry);
+	real_fops = fops_get(real_fops);
+	if (!real_fops) {
+		/* Huh? Module did not clean up after itself at exit? */
+		WARN(1, "debugfs file owner did not clean up at exit: %pd",
+			dentry);
+		r = -ENXIO;
+		goto out;
+	}
+	replace_fops(filp, real_fops);
+
+	if (real_fops->open)
+		r = real_fops->open(inode, filp);
+
+out:
+	fops_put(real_fops);
+	debugfs_use_file_finish(srcu_idx);
+	return r;
+}
+
+const struct file_operations debugfs_open_proxy_file_operations = {
+	.open = open_proxy_open,
+};
+
 static struct dentry *debugfs_create_mode(const char *name, umode_t mode,
 					  struct dentry *parent, void *value,
 				          const struct file_operations *fops,
diff --git a/fs/debugfs/inode.c b/fs/debugfs/inode.c
index bece948..26880ad 100644
--- a/fs/debugfs/inode.c
+++ b/fs/debugfs/inode.c
@@ -27,9 +27,14 @@
 #include <linux/parser.h>
 #include <linux/magic.h>
 #include <linux/slab.h>
+#include <linux/srcu.h>
+
+#include "internal.h"
 
 #define DEBUGFS_DEFAULT_MODE	0700
 
+DEFINE_SRCU(debugfs_srcu);
+
 static struct vfsmount *debugfs_mount;
 static int debugfs_mount_count;
 static bool debugfs_registered;
@@ -340,8 +345,12 @@ struct dentry *debugfs_create_file(const char *name, umode_t mode,
 		return failed_creating(dentry);
 
 	inode->i_mode = mode;
-	inode->i_fop = fops ? fops : &debugfs_file_operations;
 	inode->i_private = data;
+
+	inode->i_fop = fops ? &debugfs_open_proxy_file_operations
+		: &debugfs_noop_file_operations;
+	dentry->d_fsdata = (void *)fops;
+
 	d_instantiate(dentry, inode);
 	fsnotify_create(d_inode(dentry->d_parent), dentry);
 	return end_creating(dentry);
@@ -565,6 +574,7 @@ void debugfs_remove(struct dentry *dentry)
 	inode_unlock(d_inode(parent));
 	if (!ret)
 		simple_release_fs(&debugfs_mount, &debugfs_mount_count);
+	synchronize_srcu(&debugfs_srcu);
 }
 EXPORT_SYMBOL_GPL(debugfs_remove);
 
@@ -642,6 +652,7 @@ void debugfs_remove_recursive(struct dentry *dentry)
 	if (!__debugfs_remove(child, parent))
 		simple_release_fs(&debugfs_mount, &debugfs_mount_count);
 	inode_unlock(d_inode(parent));
+	synchronize_srcu(&debugfs_srcu);
 }
 EXPORT_SYMBOL_GPL(debugfs_remove_recursive);
 
diff --git a/fs/debugfs/internal.h b/fs/debugfs/internal.h
new file mode 100644
index 0000000..c7aaa5c
--- /dev/null
+++ b/fs/debugfs/internal.h
@@ -0,0 +1,24 @@
+/*
+ *  internal.h - declarations internal to debugfs
+ *
+ *  Copyright (C) 2016 Nicolai Stange <nicstange@gmail.com>
+ *
+ *	This program is free software; you can redistribute it and/or
+ *	modify it under the terms of the GNU General Public License version
+ *	2 as published by the Free Software Foundation.
+ *
+ */
+
+#ifndef _DEBUGFS_INTERNAL_H_
+#define _DEBUGFS_INTERNAL_H_
+
+struct file_operations;
+struct srcu_struct;
+
+/* declared over in file.c */
+extern const struct file_operations debugfs_noop_file_operations;
+extern const struct file_operations debugfs_open_proxy_file_operations;
+
+extern struct srcu_struct debugfs_srcu;
+
+#endif /* _DEBUGFS_INTERNAL_H_ */
diff --git a/include/linux/debugfs.h b/include/linux/debugfs.h
index 981e53a..fcafe2d 100644
--- a/include/linux/debugfs.h
+++ b/include/linux/debugfs.h
@@ -43,9 +43,6 @@ extern struct dentry *arch_debugfs_dir;
 
 #if defined(CONFIG_DEBUG_FS)
 
-/* declared over in file.c */
-extern const struct file_operations debugfs_file_operations;
-
 struct dentry *debugfs_create_file(const char *name, umode_t mode,
 				   struct dentry *parent, void *data,
 				   const struct file_operations *fops);
diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug
index 5a60f45..0d5177f 100644
--- a/lib/Kconfig.debug
+++ b/lib/Kconfig.debug
@@ -257,6 +257,7 @@ config PAGE_OWNER
 
 config DEBUG_FS
 	bool "Debug Filesystem"
+	select SRCU
 	help
 	  debugfs is a virtual file system that kernel developers use to put
 	  debugging files into.  Enable this option to be able to read and
-- 
2.7.4

[PATCH v6 2/8] debugfs: prevent access to removed files' private data

[Nicolai Stange]

Upon return of debugfs_remove()/debugfs_remove_recursive(), it might
still be attempted to access associated private file data through
previously opened struct file objects. If that data has been freed by
the caller of debugfs_remove*() in the meanwhile, the reading/writing
process would either encounter a fault or, if the memory address in
question has been reassigned again, unrelated data structures could get
overwritten.

However, since debugfs files are seldomly removed, usually from module
exit handlers only, the impact is very low.

Currently, there are ~1000 call sites of debugfs_create_file() spread
throughout the whole tree and touching all of those struct file_operations
in order to make them file removal aware by means of checking the result of
debugfs_use_file_start() from within their methods is unfeasible.

Instead, wrap the struct file_operations by a lifetime managing proxy at
file open:
- In debugfs_create_file(), the original fops handed in has got stashed
  away in ->d_fsdata already.
- In debugfs_create_file(), install a proxy file_operations factory,
  debugfs_full_proxy_file_operations, at ->i_fop.

This proxy factory has got an ->open() method only. It carries out some
lifetime checks and if successful, dynamically allocates and sets up a new
struct file_operations proxy at ->f_op. Afterwards, it forwards to the
->open() of the original struct file_operations in ->d_fsdata, if any.

The dynamically set up proxy at ->f_op has got a lifetime managing wrapper
set for each of the methods defined in the original struct file_operations
in ->d_fsdata.

Its ->release()er frees the proxy again and forwards to the original
->release(), if any.

In order not to mislead the VFS layer, it is strictly necessary to leave
those fields blank in the proxy that have been NULL in the original
struct file_operations also, i.e. aren't supported. This is why there is a
need for dynamically allocated proxies. The choice made not to allocate a
proxy instance for every dentry at file creation, but for every
struct file object instantiated thereof is justified by the expected usage
pattern of debugfs, namely that in general very few files get opened more
than once at a time.

The wrapper methods set in the struct file_operations implement lifetime
managing by means of the SRCU protection facilities already in place for
debugfs:
They set up a SRCU read side critical section and check whether the dentry
is still alive by means of debugfs_use_file_start(). If so, they forward
the call to the original struct file_operation stored in ->d_fsdata, still
under the protection of the SRCU read side critical section.
This SRCU read side critical section prevents any pending debugfs_remove()
and friends to return to their callers. Since a file's private data must
only be freed after the return of debugfs_remove(), the ongoing proxied
call is guarded against any file removal race.

If, on the other hand, the initial call to debugfs_use_file_start() detects
that the dentry is dead, the wrapper simply returns -EIO and does not
forward the call. Note that the ->poll() wrapper is special in that its
signature does not allow for the return of arbitrary -EXXX values and thus,
POLLHUP is returned here.

In order not to pollute debugfs with wrapper definitions that aren't ever
needed, I chose not to define a wrapper for every struct file_operations
method possible. Instead, a wrapper is defined only for the subset of
methods which are actually set by any debugfs users.
Currently, these are:

  ->llseek()
  ->read()
  ->write()
  ->unlocked_ioctl()
  ->poll()

The ->release() wrapper is special in that it does not protect the original
->release() in any way from dead files in order not to leak resources.
Thus, any ->release() handed to debugfs must implement file lifetime
management manually, if needed.
For only 33 out of a total of 434 releasers handed in to debugfs, it could
not be verified immediately whether they access data structures that might
have been freed upon a debugfs_remove() return in the meanwhile.

Export debugfs_use_file_start() and debugfs_use_file_finish() in order to
allow any ->release() to manually implement file lifetime management.

For a set of common cases of struct file_operations implemented by the
debugfs_core itself, future patches will incorporate file lifetime
management directly within those in order to allow for their unproxied
operation. Rename the original, non-proxying "debugfs_create_file()" to
"debugfs_create_file_unsafe()" and keep it for future internal use by
debugfs itself. Factor out code common to both into the new
__debugfs_create_file().

Signed-off-by: Nicolai Stange <nicstange@gmail.com>
---
 fs/debugfs/file.c       | 157 +++++++++++++++++++++++++++++++++++++++++++++++-
 fs/debugfs/inode.c      |  70 ++++++++++++++-------
 fs/debugfs/internal.h   |   6 +-
 include/linux/debugfs.h |  20 ++++++
 4 files changed, 226 insertions(+), 27 deletions(-)

diff --git a/fs/debugfs/file.c b/fs/debugfs/file.c
index 736ab3c..6eb58a8 100644
--- a/fs/debugfs/file.c
+++ b/fs/debugfs/file.c
@@ -23,9 +23,12 @@
 #include <linux/atomic.h>
 #include <linux/device.h>
 #include <linux/srcu.h>
+#include <asm/poll.h>
 
 #include "internal.h"
 
+struct poll_table_struct;
+
 static ssize_t default_read_file(struct file *file, char __user *buf,
 				 size_t count, loff_t *ppos)
 {
@@ -66,7 +69,7 @@ const struct file_operations debugfs_noop_file_operations = {
  * debugfs_use_file_start() must be followed by a matching call
  * to debugfs_use_file_finish().
  */
-static int debugfs_use_file_start(const struct dentry *dentry, int *srcu_idx)
+int debugfs_use_file_start(const struct dentry *dentry, int *srcu_idx)
 	__acquires(&debugfs_srcu)
 {
 	*srcu_idx = srcu_read_lock(&debugfs_srcu);
@@ -75,6 +78,7 @@ static int debugfs_use_file_start(const struct dentry *dentry, int *srcu_idx)
 		return -EIO;
 	return 0;
 }
+EXPORT_SYMBOL_GPL(debugfs_use_file_start);
 
 /**
  * debugfs_use_file_finish - mark the end of file data access
@@ -85,10 +89,11 @@ static int debugfs_use_file_start(const struct dentry *dentry, int *srcu_idx)
  * debugfs_remove_recursive() blocked by a former call to
  * debugfs_use_file_start() to proceed and return to its caller.
  */
-static void debugfs_use_file_finish(int srcu_idx) __releases(&debugfs_srcu)
+void debugfs_use_file_finish(int srcu_idx) __releases(&debugfs_srcu)
 {
 	srcu_read_unlock(&debugfs_srcu, srcu_idx);
 }
+EXPORT_SYMBOL_GPL(debugfs_use_file_finish);
 
 #define F_DENTRY(filp) ((filp)->f_path.dentry)
 
@@ -131,6 +136,154 @@ const struct file_operations debugfs_open_proxy_file_operations = {
 	.open = open_proxy_open,
 };
 
+#define PROTO(args...) args
+#define ARGS(args...) args
+
+#define FULL_PROXY_FUNC(name, ret_type, filp, proto, args)		\
+static ret_type full_proxy_ ## name(proto)				\
+{									\
+	const struct dentry *dentry = F_DENTRY(filp);			\
+	const struct file_operations *real_fops =			\
+		REAL_FOPS_DEREF(dentry);				\
+	int srcu_idx;							\
+	ret_type r;							\
+									\
+	r = debugfs_use_file_start(dentry, &srcu_idx);			\
+	if (likely(!r))						\
+		r = real_fops->name(args);				\
+	debugfs_use_file_finish(srcu_idx);				\
+	return r;							\
+}
+
+FULL_PROXY_FUNC(llseek, loff_t, filp,
+		PROTO(struct file *filp, loff_t offset, int whence),
+		ARGS(filp, offset, whence));
+
+FULL_PROXY_FUNC(read, ssize_t, filp,
+		PROTO(struct file *filp, char __user *buf, size_t size,
+			loff_t *ppos),
+		ARGS(filp, buf, size, ppos));
+
+FULL_PROXY_FUNC(write, ssize_t, filp,
+		PROTO(struct file *filp, const char __user *buf, size_t size,
+			loff_t *ppos),
+		ARGS(filp, buf, size, ppos));
+
+FULL_PROXY_FUNC(unlocked_ioctl, long, filp,
+		PROTO(struct file *filp, unsigned int cmd, unsigned long arg),
+		ARGS(filp, cmd, arg));
+
+static unsigned int full_proxy_poll(struct file *filp,
+				struct poll_table_struct *wait)
+{
+	const struct dentry *dentry = F_DENTRY(filp);
+	const struct file_operations *real_fops = REAL_FOPS_DEREF(dentry);
+	int srcu_idx;
+	unsigned int r = 0;
+
+	if (debugfs_use_file_start(dentry, &srcu_idx)) {
+		debugfs_use_file_finish(srcu_idx);
+		return POLLHUP;
+	}
+
+	r = real_fops->poll(filp, wait);
+	debugfs_use_file_finish(srcu_idx);
+	return r;
+}
+
+static int full_proxy_release(struct inode *inode, struct file *filp)
+{
+	const struct dentry *dentry = F_DENTRY(filp);
+	const struct file_operations *real_fops = REAL_FOPS_DEREF(dentry);
+	const struct file_operations *proxy_fops = filp->f_op;
+	int r = 0;
+
+	/*
+	 * We must not protect this against removal races here: the
+	 * original releaser should be called unconditionally in order
+	 * not to leak any resources. Releasers must not assume that
+	 * ->i_private is still being meaningful here.
+	 */
+	if (real_fops->release)
+		r = real_fops->release(inode, filp);
+
+	replace_fops(filp, d_inode(dentry)->i_fop);
+	kfree((void *)proxy_fops);
+	fops_put(real_fops);
+	return 0;
+}
+
+static void __full_proxy_fops_init(struct file_operations *proxy_fops,
+				const struct file_operations *real_fops)
+{
+	proxy_fops->release = full_proxy_release;
+	if (real_fops->llseek)
+		proxy_fops->llseek = full_proxy_llseek;
+	if (real_fops->read)
+		proxy_fops->read = full_proxy_read;
+	if (real_fops->write)
+		proxy_fops->write = full_proxy_write;
+	if (real_fops->poll)
+		proxy_fops->poll = full_proxy_poll;
+	if (real_fops->unlocked_ioctl)
+		proxy_fops->unlocked_ioctl = full_proxy_unlocked_ioctl;
+}
+
+static int full_proxy_open(struct inode *inode, struct file *filp)
+{
+	const struct dentry *dentry = F_DENTRY(filp);
+	const struct file_operations *real_fops = NULL;
+	struct file_operations *proxy_fops = NULL;
+	int srcu_idx, r;
+
+	r = debugfs_use_file_start(dentry, &srcu_idx);
+	if (r) {
+		r = -ENOENT;
+		goto out;
+	}
+
+	real_fops = REAL_FOPS_DEREF(dentry);
+	real_fops = fops_get(real_fops);
+	if (!real_fops) {
+		/* Huh? Module did not cleanup after itself at exit? */
+		WARN(1, "debugfs file owner did not clean up at exit: %pd",
+			dentry);
+		r = -ENXIO;
+		goto out;
+	}
+
+	proxy_fops = kzalloc(sizeof(*proxy_fops), GFP_KERNEL);
+	if (!proxy_fops) {
+		r = -ENOMEM;
+		goto free_proxy;
+	}
+	__full_proxy_fops_init(proxy_fops, real_fops);
+	replace_fops(filp, proxy_fops);
+
+	if (real_fops->open) {
+		r = real_fops->open(inode, filp);
+
+		if (filp->f_op != proxy_fops) {
+			/* No protection against file removal anymore. */
+			WARN(1, "debugfs file owner replaced proxy fops: %pd",
+				dentry);
+			goto free_proxy;
+		}
+	}
+
+	goto out;
+free_proxy:
+	kfree(proxy_fops);
+	fops_put(real_fops);
+out:
+	debugfs_use_file_finish(srcu_idx);
+	return r;
+}
+
+const struct file_operations debugfs_full_proxy_file_operations = {
+	.open = full_proxy_open,
+};
+
 static struct dentry *debugfs_create_mode(const char *name, umode_t mode,
 					  struct dentry *parent, void *value,
 				          const struct file_operations *fops,
diff --git a/fs/debugfs/inode.c b/fs/debugfs/inode.c
index 26880ad..42a9b34 100644
--- a/fs/debugfs/inode.c
+++ b/fs/debugfs/inode.c
@@ -299,6 +299,37 @@ static struct dentry *end_creating(struct dentry *dentry)
 	return dentry;
 }
 
+static struct dentry *__debugfs_create_file(const char *name, umode_t mode,
+				struct dentry *parent, void *data,
+				const struct file_operations *proxy_fops,
+				const struct file_operations *real_fops)
+{
+	struct dentry *dentry;
+	struct inode *inode;
+
+	if (!(mode & S_IFMT))
+		mode |= S_IFREG;
+	BUG_ON(!S_ISREG(mode));
+	dentry = start_creating(name, parent);
+
+	if (IS_ERR(dentry))
+		return NULL;
+
+	inode = debugfs_get_inode(dentry->d_sb);
+	if (unlikely(!inode))
+		return failed_creating(dentry);
+
+	inode->i_mode = mode;
+	inode->i_private = data;
+
+	inode->i_fop = proxy_fops;
+	dentry->d_fsdata = (void *)real_fops;
+
+	d_instantiate(dentry, inode);
+	fsnotify_create(d_inode(dentry->d_parent), dentry);
+	return end_creating(dentry);
+}
+
 /**
  * debugfs_create_file - create a file in the debugfs filesystem
  * @name: a pointer to a string containing the name of the file to create.
@@ -329,33 +360,24 @@ struct dentry *debugfs_create_file(const char *name, umode_t mode,
 				   struct dentry *parent, void *data,
 				   const struct file_operations *fops)
 {
-	struct dentry *dentry;
-	struct inode *inode;
-
-	if (!(mode & S_IFMT))
-		mode |= S_IFREG;
-	BUG_ON(!S_ISREG(mode));
-	dentry = start_creating(name, parent);
-
-	if (IS_ERR(dentry))
-		return NULL;
-
-	inode = debugfs_get_inode(dentry->d_sb);
-	if (unlikely(!inode))
-		return failed_creating(dentry);
 
-	inode->i_mode = mode;
-	inode->i_private = data;
+	return __debugfs_create_file(name, mode, parent, data,
+				fops ? &debugfs_full_proxy_file_operations :
+					&debugfs_noop_file_operations,
+				fops);
+}
+EXPORT_SYMBOL_GPL(debugfs_create_file);
 
-	inode->i_fop = fops ? &debugfs_open_proxy_file_operations
-		: &debugfs_noop_file_operations;
-	dentry->d_fsdata = (void *)fops;
+struct dentry *debugfs_create_file_unsafe(const char *name, umode_t mode,
+				   struct dentry *parent, void *data,
+				   const struct file_operations *fops)
+{
 
-	d_instantiate(dentry, inode);
-	fsnotify_create(d_inode(dentry->d_parent), dentry);
-	return end_creating(dentry);
+	return __debugfs_create_file(name, mode, parent, data,
+				fops ? &debugfs_open_proxy_file_operations :
+					&debugfs_noop_file_operations,
+				fops);
 }
-EXPORT_SYMBOL_GPL(debugfs_create_file);
 
 /**
  * debugfs_create_file_size - create a file in the debugfs filesystem
@@ -574,6 +596,7 @@ void debugfs_remove(struct dentry *dentry)
 	inode_unlock(d_inode(parent));
 	if (!ret)
 		simple_release_fs(&debugfs_mount, &debugfs_mount_count);
+
 	synchronize_srcu(&debugfs_srcu);
 }
 EXPORT_SYMBOL_GPL(debugfs_remove);
@@ -652,6 +675,7 @@ void debugfs_remove_recursive(struct dentry *dentry)
 	if (!__debugfs_remove(child, parent))
 		simple_release_fs(&debugfs_mount, &debugfs_mount_count);
 	inode_unlock(d_inode(parent));
+
 	synchronize_srcu(&debugfs_srcu);
 }
 EXPORT_SYMBOL_GPL(debugfs_remove_recursive);
diff --git a/fs/debugfs/internal.h b/fs/debugfs/internal.h
index c7aaa5c..bba5263 100644
--- a/fs/debugfs/internal.h
+++ b/fs/debugfs/internal.h
@@ -13,12 +13,14 @@
 #define _DEBUGFS_INTERNAL_H_
 
 struct file_operations;
-struct srcu_struct;
 
 /* declared over in file.c */
 extern const struct file_operations debugfs_noop_file_operations;
 extern const struct file_operations debugfs_open_proxy_file_operations;
+extern const struct file_operations debugfs_full_proxy_file_operations;
 
-extern struct srcu_struct debugfs_srcu;
+struct dentry *debugfs_create_file_unsafe(const char *name, umode_t mode,
+					struct dentry *parent, void *data,
+					const struct file_operations *fops);
 
 #endif /* _DEBUGFS_INTERNAL_H_ */
diff --git a/include/linux/debugfs.h b/include/linux/debugfs.h
index fcafe2d..a63e6ea 100644
--- a/include/linux/debugfs.h
+++ b/include/linux/debugfs.h
@@ -19,9 +19,11 @@
 #include <linux/seq_file.h>
 
 #include <linux/types.h>
+#include <linux/compiler.h>
 
 struct device;
 struct file_operations;
+struct srcu_struct;
 
 struct debugfs_blob_wrapper {
 	void *data;
@@ -41,6 +43,8 @@ struct debugfs_regset32 {
 
 extern struct dentry *arch_debugfs_dir;
 
+extern struct srcu_struct debugfs_srcu;
+
 #if defined(CONFIG_DEBUG_FS)
 
 struct dentry *debugfs_create_file(const char *name, umode_t mode,
@@ -65,6 +69,11 @@ struct dentry *debugfs_create_automount(const char *name,
 void debugfs_remove(struct dentry *dentry);
 void debugfs_remove_recursive(struct dentry *dentry);
 
+int debugfs_use_file_start(const struct dentry *dentry, int *srcu_idx)
+	__acquires(&debugfs_srcu);
+
+void debugfs_use_file_finish(int srcu_idx) __releases(&debugfs_srcu);
+
 struct dentry *debugfs_rename(struct dentry *old_dir, struct dentry *old_dentry,
                 struct dentry *new_dir, const char *new_name);
 
@@ -173,6 +182,17 @@ static inline void debugfs_remove(struct dentry *dentry)
 static inline void debugfs_remove_recursive(struct dentry *dentry)
 { }
 
+static inline int debugfs_use_file_start(const struct dentry *dentry,
+					int *srcu_idx)
+	__acquires(&debugfs_srcu)
+{
+	return 0;
+}
+
+static inline void debugfs_use_file_finish(int srcu_idx)
+	__releases(&debugfs_srcu)
+{ }
+
 static inline struct dentry *debugfs_rename(struct dentry *old_dir, struct dentry *old_dentry,
                 struct dentry *new_dir, char *new_name)
 {
-- 
2.7.4

[Cocci] [PATCH v6 2/8] debugfs: prevent access to removed files' private data

[Nicolai Stange]

Upon return of debugfs_remove()/debugfs_remove_recursive(), it might
still be attempted to access associated private file data through
previously opened struct file objects. If that data has been freed by
the caller of debugfs_remove*() in the meanwhile, the reading/writing
process would either encounter a fault or, if the memory address in
question has been reassigned again, unrelated data structures could get
overwritten.

However, since debugfs files are seldomly removed, usually from module
exit handlers only, the impact is very low.

Currently, there are ~1000 call sites of debugfs_create_file() spread
throughout the whole tree and touching all of those struct file_operations
in order to make them file removal aware by means of checking the result of
debugfs_use_file_start() from within their methods is unfeasible.

Instead, wrap the struct file_operations by a lifetime managing proxy at
file open:
- In debugfs_create_file(), the original fops handed in has got stashed
  away in ->d_fsdata already.
- In debugfs_create_file(), install a proxy file_operations factory,
  debugfs_full_proxy_file_operations, at ->i_fop.

This proxy factory has got an ->open() method only. It carries out some
lifetime checks and if successful, dynamically allocates and sets up a new
struct file_operations proxy at ->f_op. Afterwards, it forwards to the
->open() of the original struct file_operations in ->d_fsdata, if any.

The dynamically set up proxy at ->f_op has got a lifetime managing wrapper
set for each of the methods defined in the original struct file_operations
in ->d_fsdata.

Its ->release()er frees the proxy again and forwards to the original
->release(), if any.

In order not to mislead the VFS layer, it is strictly necessary to leave
those fields blank in the proxy that have been NULL in the original
struct file_operations also, i.e. aren't supported. This is why there is a
need for dynamically allocated proxies. The choice made not to allocate a
proxy instance for every dentry at file creation, but for every
struct file object instantiated thereof is justified by the expected usage
pattern of debugfs, namely that in general very few files get opened more
than once at a time.

The wrapper methods set in the struct file_operations implement lifetime
managing by means of the SRCU protection facilities already in place for
debugfs:
They set up a SRCU read side critical section and check whether the dentry
is still alive by means of debugfs_use_file_start(). If so, they forward
the call to the original struct file_operation stored in ->d_fsdata, still
under the protection of the SRCU read side critical section.
This SRCU read side critical section prevents any pending debugfs_remove()
and friends to return to their callers. Since a file's private data must
only be freed after the return of debugfs_remove(), the ongoing proxied
call is guarded against any file removal race.

If, on the other hand, the initial call to debugfs_use_file_start() detects
that the dentry is dead, the wrapper simply returns -EIO and does not
forward the call. Note that the ->poll() wrapper is special in that its
signature does not allow for the return of arbitrary -EXXX values and thus,
POLLHUP is returned here.

In order not to pollute debugfs with wrapper definitions that aren't ever
needed, I chose not to define a wrapper for every struct file_operations
method possible. Instead, a wrapper is defined only for the subset of
methods which are actually set by any debugfs users.
Currently, these are:

  ->llseek()
  ->read()
  ->write()
  ->unlocked_ioctl()
  ->poll()

The ->release() wrapper is special in that it does not protect the original
->release() in any way from dead files in order not to leak resources.
Thus, any ->release() handed to debugfs must implement file lifetime
management manually, if needed.
For only 33 out of a total of 434 releasers handed in to debugfs, it could
not be verified immediately whether they access data structures that might
have been freed upon a debugfs_remove() return in the meanwhile.

Export debugfs_use_file_start() and debugfs_use_file_finish() in order to
allow any ->release() to manually implement file lifetime management.

For a set of common cases of struct file_operations implemented by the
debugfs_core itself, future patches will incorporate file lifetime
management directly within those in order to allow for their unproxied
operation. Rename the original, non-proxying "debugfs_create_file()" to
"debugfs_create_file_unsafe()" and keep it for future internal use by
debugfs itself. Factor out code common to both into the new
__debugfs_create_file().

Signed-off-by: Nicolai Stange <nicstange@gmail.com>
---
 fs/debugfs/file.c       | 157 +++++++++++++++++++++++++++++++++++++++++++++++-
 fs/debugfs/inode.c      |  70 ++++++++++++++-------
 fs/debugfs/internal.h   |   6 +-
 include/linux/debugfs.h |  20 ++++++
 4 files changed, 226 insertions(+), 27 deletions(-)

diff --git a/fs/debugfs/file.c b/fs/debugfs/file.c
index 736ab3c..6eb58a8 100644
--- a/fs/debugfs/file.c
+++ b/fs/debugfs/file.c
@@ -23,9 +23,12 @@
 #include <linux/atomic.h>
 #include <linux/device.h>
 #include <linux/srcu.h>
+#include <asm/poll.h>
 
 #include "internal.h"
 
+struct poll_table_struct;
+
 static ssize_t default_read_file(struct file *file, char __user *buf,
 				 size_t count, loff_t *ppos)
 {
@@ -66,7 +69,7 @@ const struct file_operations debugfs_noop_file_operations = {
  * debugfs_use_file_start() must be followed by a matching call
  * to debugfs_use_file_finish().
  */
-static int debugfs_use_file_start(const struct dentry *dentry, int *srcu_idx)
+int debugfs_use_file_start(const struct dentry *dentry, int *srcu_idx)
 	__acquires(&debugfs_srcu)
 {
 	*srcu_idx = srcu_read_lock(&debugfs_srcu);
@@ -75,6 +78,7 @@ static int debugfs_use_file_start(const struct dentry *dentry, int *srcu_idx)
 		return -EIO;
 	return 0;
 }
+EXPORT_SYMBOL_GPL(debugfs_use_file_start);
 
 /**
  * debugfs_use_file_finish - mark the end of file data access
@@ -85,10 +89,11 @@ static int debugfs_use_file_start(const struct dentry *dentry, int *srcu_idx)
  * debugfs_remove_recursive() blocked by a former call to
  * debugfs_use_file_start() to proceed and return to its caller.
  */
-static void debugfs_use_file_finish(int srcu_idx) __releases(&debugfs_srcu)
+void debugfs_use_file_finish(int srcu_idx) __releases(&debugfs_srcu)
 {
 	srcu_read_unlock(&debugfs_srcu, srcu_idx);
 }
+EXPORT_SYMBOL_GPL(debugfs_use_file_finish);
 
 #define F_DENTRY(filp) ((filp)->f_path.dentry)
 
@@ -131,6 +136,154 @@ const struct file_operations debugfs_open_proxy_file_operations = {
 	.open = open_proxy_open,
 };
 
+#define PROTO(args...) args
+#define ARGS(args...) args
+
+#define FULL_PROXY_FUNC(name, ret_type, filp, proto, args)		\
+static ret_type full_proxy_ ## name(proto)				\
+{									\
+	const struct dentry *dentry = F_DENTRY(filp);			\
+	const struct file_operations *real_fops =			\
+		REAL_FOPS_DEREF(dentry);				\
+	int srcu_idx;							\
+	ret_type r;							\
+									\
+	r = debugfs_use_file_start(dentry, &srcu_idx);			\
+	if (likely(!r))						\
+		r = real_fops->name(args);				\
+	debugfs_use_file_finish(srcu_idx);				\
+	return r;							\
+}
+
+FULL_PROXY_FUNC(llseek, loff_t, filp,
+		PROTO(struct file *filp, loff_t offset, int whence),
+		ARGS(filp, offset, whence));
+
+FULL_PROXY_FUNC(read, ssize_t, filp,
+		PROTO(struct file *filp, char __user *buf, size_t size,
+			loff_t *ppos),
+		ARGS(filp, buf, size, ppos));
+
+FULL_PROXY_FUNC(write, ssize_t, filp,
+		PROTO(struct file *filp, const char __user *buf, size_t size,
+			loff_t *ppos),
+		ARGS(filp, buf, size, ppos));
+
+FULL_PROXY_FUNC(unlocked_ioctl, long, filp,
+		PROTO(struct file *filp, unsigned int cmd, unsigned long arg),
+		ARGS(filp, cmd, arg));
+
+static unsigned int full_proxy_poll(struct file *filp,
+				struct poll_table_struct *wait)
+{
+	const struct dentry *dentry = F_DENTRY(filp);
+	const struct file_operations *real_fops = REAL_FOPS_DEREF(dentry);
+	int srcu_idx;
+	unsigned int r = 0;
+
+	if (debugfs_use_file_start(dentry, &srcu_idx)) {
+		debugfs_use_file_finish(srcu_idx);
+		return POLLHUP;
+	}
+
+	r = real_fops->poll(filp, wait);
+	debugfs_use_file_finish(srcu_idx);
+	return r;
+}
+
+static int full_proxy_release(struct inode *inode, struct file *filp)
+{
+	const struct dentry *dentry = F_DENTRY(filp);
+	const struct file_operations *real_fops = REAL_FOPS_DEREF(dentry);
+	const struct file_operations *proxy_fops = filp->f_op;
+	int r = 0;
+
+	/*
+	 * We must not protect this against removal races here: the
+	 * original releaser should be called unconditionally in order
+	 * not to leak any resources. Releasers must not assume that
+	 * ->i_private is still being meaningful here.
+	 */
+	if (real_fops->release)
+		r = real_fops->release(inode, filp);
+
+	replace_fops(filp, d_inode(dentry)->i_fop);
+	kfree((void *)proxy_fops);
+	fops_put(real_fops);
+	return 0;
+}
+
+static void __full_proxy_fops_init(struct file_operations *proxy_fops,
+				const struct file_operations *real_fops)
+{
+	proxy_fops->release = full_proxy_release;
+	if (real_fops->llseek)
+		proxy_fops->llseek = full_proxy_llseek;
+	if (real_fops->read)
+		proxy_fops->read = full_proxy_read;
+	if (real_fops->write)
+		proxy_fops->write = full_proxy_write;
+	if (real_fops->poll)
+		proxy_fops->poll = full_proxy_poll;
+	if (real_fops->unlocked_ioctl)
+		proxy_fops->unlocked_ioctl = full_proxy_unlocked_ioctl;
+}
+
+static int full_proxy_open(struct inode *inode, struct file *filp)
+{
+	const struct dentry *dentry = F_DENTRY(filp);
+	const struct file_operations *real_fops = NULL;
+	struct file_operations *proxy_fops = NULL;
+	int srcu_idx, r;
+
+	r = debugfs_use_file_start(dentry, &srcu_idx);
+	if (r) {
+		r = -ENOENT;
+		goto out;
+	}
+
+	real_fops = REAL_FOPS_DEREF(dentry);
+	real_fops = fops_get(real_fops);
+	if (!real_fops) {
+		/* Huh? Module did not cleanup after itself at exit? */
+		WARN(1, "debugfs file owner did not clean up at exit: %pd",
+			dentry);
+		r = -ENXIO;
+		goto out;
+	}
+
+	proxy_fops = kzalloc(sizeof(*proxy_fops), GFP_KERNEL);
+	if (!proxy_fops) {
+		r = -ENOMEM;
+		goto free_proxy;
+	}
+	__full_proxy_fops_init(proxy_fops, real_fops);
+	replace_fops(filp, proxy_fops);
+
+	if (real_fops->open) {
+		r = real_fops->open(inode, filp);
+
+		if (filp->f_op != proxy_fops) {
+			/* No protection against file removal anymore. */
+			WARN(1, "debugfs file owner replaced proxy fops: %pd",
+				dentry);
+			goto free_proxy;
+		}
+	}
+
+	goto out;
+free_proxy:
+	kfree(proxy_fops);
+	fops_put(real_fops);
+out:
+	debugfs_use_file_finish(srcu_idx);
+	return r;
+}
+
+const struct file_operations debugfs_full_proxy_file_operations = {
+	.open = full_proxy_open,
+};
+
 static struct dentry *debugfs_create_mode(const char *name, umode_t mode,
 					  struct dentry *parent, void *value,
 				          const struct file_operations *fops,
diff --git a/fs/debugfs/inode.c b/fs/debugfs/inode.c
index 26880ad..42a9b34 100644
--- a/fs/debugfs/inode.c
+++ b/fs/debugfs/inode.c
@@ -299,6 +299,37 @@ static struct dentry *end_creating(struct dentry *dentry)
 	return dentry;
 }
 
+static struct dentry *__debugfs_create_file(const char *name, umode_t mode,
+				struct dentry *parent, void *data,
+				const struct file_operations *proxy_fops,
+				const struct file_operations *real_fops)
+{
+	struct dentry *dentry;
+	struct inode *inode;
+
+	if (!(mode & S_IFMT))
+		mode |= S_IFREG;
+	BUG_ON(!S_ISREG(mode));
+	dentry = start_creating(name, parent);
+
+	if (IS_ERR(dentry))
+		return NULL;
+
+	inode = debugfs_get_inode(dentry->d_sb);
+	if (unlikely(!inode))
+		return failed_creating(dentry);
+
+	inode->i_mode = mode;
+	inode->i_private = data;
+
+	inode->i_fop = proxy_fops;
+	dentry->d_fsdata = (void *)real_fops;
+
+	d_instantiate(dentry, inode);
+	fsnotify_create(d_inode(dentry->d_parent), dentry);
+	return end_creating(dentry);
+}
+
 /**
  * debugfs_create_file - create a file in the debugfs filesystem
  * @name: a pointer to a string containing the name of the file to create.
@@ -329,33 +360,24 @@ struct dentry *debugfs_create_file(const char *name, umode_t mode,
 				   struct dentry *parent, void *data,
 				   const struct file_operations *fops)
 {
-	struct dentry *dentry;
-	struct inode *inode;
-
-	if (!(mode & S_IFMT))
-		mode |= S_IFREG;
-	BUG_ON(!S_ISREG(mode));
-	dentry = start_creating(name, parent);
-
-	if (IS_ERR(dentry))
-		return NULL;
-
-	inode = debugfs_get_inode(dentry->d_sb);
-	if (unlikely(!inode))
-		return failed_creating(dentry);
 
-	inode->i_mode = mode;
-	inode->i_private = data;
+	return __debugfs_create_file(name, mode, parent, data,
+				fops ? &debugfs_full_proxy_file_operations :
+					&debugfs_noop_file_operations,
+				fops);
+}
+EXPORT_SYMBOL_GPL(debugfs_create_file);
 
-	inode->i_fop = fops ? &debugfs_open_proxy_file_operations
-		: &debugfs_noop_file_operations;
-	dentry->d_fsdata = (void *)fops;
+struct dentry *debugfs_create_file_unsafe(const char *name, umode_t mode,
+				   struct dentry *parent, void *data,
+				   const struct file_operations *fops)
+{
 
-	d_instantiate(dentry, inode);
-	fsnotify_create(d_inode(dentry->d_parent), dentry);
-	return end_creating(dentry);
+	return __debugfs_create_file(name, mode, parent, data,
+				fops ? &debugfs_open_proxy_file_operations :
+					&debugfs_noop_file_operations,
+				fops);
 }
-EXPORT_SYMBOL_GPL(debugfs_create_file);
 
 /**
  * debugfs_create_file_size - create a file in the debugfs filesystem
@@ -574,6 +596,7 @@ void debugfs_remove(struct dentry *dentry)
 	inode_unlock(d_inode(parent));
 	if (!ret)
 		simple_release_fs(&debugfs_mount, &debugfs_mount_count);
+
 	synchronize_srcu(&debugfs_srcu);
 }
 EXPORT_SYMBOL_GPL(debugfs_remove);
@@ -652,6 +675,7 @@ void debugfs_remove_recursive(struct dentry *dentry)
 	if (!__debugfs_remove(child, parent))
 		simple_release_fs(&debugfs_mount, &debugfs_mount_count);
 	inode_unlock(d_inode(parent));
+
 	synchronize_srcu(&debugfs_srcu);
 }
 EXPORT_SYMBOL_GPL(debugfs_remove_recursive);
diff --git a/fs/debugfs/internal.h b/fs/debugfs/internal.h
index c7aaa5c..bba5263 100644
--- a/fs/debugfs/internal.h
+++ b/fs/debugfs/internal.h
@@ -13,12 +13,14 @@
 #define _DEBUGFS_INTERNAL_H_
 
 struct file_operations;
-struct srcu_struct;
 
 /* declared over in file.c */
 extern const struct file_operations debugfs_noop_file_operations;
 extern const struct file_operations debugfs_open_proxy_file_operations;
+extern const struct file_operations debugfs_full_proxy_file_operations;
 
-extern struct srcu_struct debugfs_srcu;
+struct dentry *debugfs_create_file_unsafe(const char *name, umode_t mode,
+					struct dentry *parent, void *data,
+					const struct file_operations *fops);
 
 #endif /* _DEBUGFS_INTERNAL_H_ */
diff --git a/include/linux/debugfs.h b/include/linux/debugfs.h
index fcafe2d..a63e6ea 100644
--- a/include/linux/debugfs.h
+++ b/include/linux/debugfs.h
@@ -19,9 +19,11 @@
 #include <linux/seq_file.h>
 
 #include <linux/types.h>
+#include <linux/compiler.h>
 
 struct device;
 struct file_operations;
+struct srcu_struct;
 
 struct debugfs_blob_wrapper {
 	void *data;
@@ -41,6 +43,8 @@ struct debugfs_regset32 {
 
 extern struct dentry *arch_debugfs_dir;
 
+extern struct srcu_struct debugfs_srcu;
+
 #if defined(CONFIG_DEBUG_FS)
 
 struct dentry *debugfs_create_file(const char *name, umode_t mode,
@@ -65,6 +69,11 @@ struct dentry *debugfs_create_automount(const char *name,
 void debugfs_remove(struct dentry *dentry);
 void debugfs_remove_recursive(struct dentry *dentry);
 
+int debugfs_use_file_start(const struct dentry *dentry, int *srcu_idx)
+	__acquires(&debugfs_srcu);
+
+void debugfs_use_file_finish(int srcu_idx) __releases(&debugfs_srcu);
+
 struct dentry *debugfs_rename(struct dentry *old_dir, struct dentry *old_dentry,
                 struct dentry *new_dir, const char *new_name);
 
@@ -173,6 +182,17 @@ static inline void debugfs_remove(struct dentry *dentry)
 static inline void debugfs_remove_recursive(struct dentry *dentry)
 { }
 
+static inline int debugfs_use_file_start(const struct dentry *dentry,
+					int *srcu_idx)
+	__acquires(&debugfs_srcu)
+{
+	return 0;
+}
+
+static inline void debugfs_use_file_finish(int srcu_idx)
+	__releases(&debugfs_srcu)
+{ }
+
 static inline struct dentry *debugfs_rename(struct dentry *old_dir, struct dentry *old_dentry,
                 struct dentry *new_dir, char *new_name)
 {
-- 
2.7.4

Re: [PATCH v6 2/8] debugfs: prevent access to removed files' private data

[Sasha Levin]

On 03/22/2016 09:11 AM, Nicolai Stange wrote:
> Upon return of debugfs_remove()/debugfs_remove_recursive(), it might
> still be attempted to access associated private file data through
> previously opened struct file objects. If that data has been freed by
> the caller of debugfs_remove*() in the meanwhile, the reading/writing
> process would either encounter a fault or, if the memory address in
> question has been reassigned again, unrelated data structures could get
> overwritten.
> 
> However, since debugfs files are seldomly removed, usually from module
> exit handlers only, the impact is very low.
> 
> Currently, there are ~1000 call sites of debugfs_create_file() spread
> throughout the whole tree and touching all of those struct file_operations
> in order to make them file removal aware by means of checking the result of
> debugfs_use_file_start() from within their methods is unfeasible.
> 
> Instead, wrap the struct file_operations by a lifetime managing proxy at
> file open:
> - In debugfs_create_file(), the original fops handed in has got stashed
>   away in ->d_fsdata already.
> - In debugfs_create_file(), install a proxy file_operations factory,
>   debugfs_full_proxy_file_operations, at ->i_fop.
> 
> This proxy factory has got an ->open() method only. It carries out some
> lifetime checks and if successful, dynamically allocates and sets up a new
> struct file_operations proxy at ->f_op. Afterwards, it forwards to the
> ->open() of the original struct file_operations in ->d_fsdata, if any.
> 
> The dynamically set up proxy at ->f_op has got a lifetime managing wrapper
> set for each of the methods defined in the original struct file_operations
> in ->d_fsdata.
> 
> Its ->release()er frees the proxy again and forwards to the original
> ->release(), if any.
> 
> In order not to mislead the VFS layer, it is strictly necessary to leave
> those fields blank in the proxy that have been NULL in the original
> struct file_operations also, i.e. aren't supported. This is why there is a
> need for dynamically allocated proxies. The choice made not to allocate a
> proxy instance for every dentry at file creation, but for every
> struct file object instantiated thereof is justified by the expected usage
> pattern of debugfs, namely that in general very few files get opened more
> than once at a time.
> 
> The wrapper methods set in the struct file_operations implement lifetime
> managing by means of the SRCU protection facilities already in place for
> debugfs:
> They set up a SRCU read side critical section and check whether the dentry
> is still alive by means of debugfs_use_file_start(). If so, they forward
> the call to the original struct file_operation stored in ->d_fsdata, still
> under the protection of the SRCU read side critical section.
> This SRCU read side critical section prevents any pending debugfs_remove()
> and friends to return to their callers. Since a file's private data must
> only be freed after the return of debugfs_remove(), the ongoing proxied
> call is guarded against any file removal race.
> 
> If, on the other hand, the initial call to debugfs_use_file_start() detects
> that the dentry is dead, the wrapper simply returns -EIO and does not
> forward the call. Note that the ->poll() wrapper is special in that its
> signature does not allow for the return of arbitrary -EXXX values and thus,
> POLLHUP is returned here.
> 
> In order not to pollute debugfs with wrapper definitions that aren't ever
> needed, I chose not to define a wrapper for every struct file_operations
> method possible. Instead, a wrapper is defined only for the subset of
> methods which are actually set by any debugfs users.
> Currently, these are:
> 
>   ->llseek()
>   ->read()
>   ->write()
>   ->unlocked_ioctl()
>   ->poll()
> 
> The ->release() wrapper is special in that it does not protect the original
> ->release() in any way from dead files in order not to leak resources.
> Thus, any ->release() handed to debugfs must implement file lifetime
> management manually, if needed.
> For only 33 out of a total of 434 releasers handed in to debugfs, it could
> not be verified immediately whether they access data structures that might
> have been freed upon a debugfs_remove() return in the meanwhile.
> 
> Export debugfs_use_file_start() and debugfs_use_file_finish() in order to
> allow any ->release() to manually implement file lifetime management.
> 
> For a set of common cases of struct file_operations implemented by the
> debugfs_core itself, future patches will incorporate file lifetime
> management directly within those in order to allow for their unproxied
> operation. Rename the original, non-proxying "debugfs_create_file()" to
> "debugfs_create_file_unsafe()" and keep it for future internal use by
> debugfs itself. Factor out code common to both into the new
> __debugfs_create_file().
> 
> Signed-off-by: Nicolai Stange <nicstange@gmail.com>

Hey,

I'm seeing a hang on boot which was bisected (twice) to this commit. I don't
see any lockup messages, but everything just seems to stop.


Thanks,
Sasha

Re: [PATCH v6 2/8] debugfs: prevent access to removed files' private data

[Nicolai Stange]

Hi Sasha,

Sasha Levin <sasha.levin@oracle.com> writes:

> On 03/22/2016 09:11 AM, Nicolai Stange wrote:
>> Upon return of debugfs_remove()/debugfs_remove_recursive(), it might
>> still be attempted to access associated private file data through
>> previously opened struct file objects. If that data has been freed by
>> the caller of debugfs_remove*() in the meanwhile, the reading/writing
>> process would either encounter a fault or, if the memory address in
>> question has been reassigned again, unrelated data structures could get
>> overwritten.
>> 
>> However, since debugfs files are seldomly removed, usually from module
>> exit handlers only, the impact is very low.
>> 
>> Currently, there are ~1000 call sites of debugfs_create_file() spread
>> throughout the whole tree and touching all of those struct file_operations
>> in order to make them file removal aware by means of checking the result of
>> debugfs_use_file_start() from within their methods is unfeasible.
>> 
>> Instead, wrap the struct file_operations by a lifetime managing proxy at
>> file open:
>> - In debugfs_create_file(), the original fops handed in has got stashed
>>   away in ->d_fsdata already.
>> - In debugfs_create_file(), install a proxy file_operations factory,
>>   debugfs_full_proxy_file_operations, at ->i_fop.
>> 
>> This proxy factory has got an ->open() method only. It carries out some
>> lifetime checks and if successful, dynamically allocates and sets up a new
>> struct file_operations proxy at ->f_op. Afterwards, it forwards to the
>> ->open() of the original struct file_operations in ->d_fsdata, if any.
>> 
>> The dynamically set up proxy at ->f_op has got a lifetime managing wrapper
>> set for each of the methods defined in the original struct file_operations
>> in ->d_fsdata.
>> 
>> Its ->release()er frees the proxy again and forwards to the original
>> ->release(), if any.
>> 
>> In order not to mislead the VFS layer, it is strictly necessary to leave
>> those fields blank in the proxy that have been NULL in the original
>> struct file_operations also, i.e. aren't supported. This is why there is a
>> need for dynamically allocated proxies. The choice made not to allocate a
>> proxy instance for every dentry at file creation, but for every
>> struct file object instantiated thereof is justified by the expected usage
>> pattern of debugfs, namely that in general very few files get opened more
>> than once at a time.
>> 
>> The wrapper methods set in the struct file_operations implement lifetime
>> managing by means of the SRCU protection facilities already in place for
>> debugfs:
>> They set up a SRCU read side critical section and check whether the dentry
>> is still alive by means of debugfs_use_file_start(). If so, they forward
>> the call to the original struct file_operation stored in ->d_fsdata, still
>> under the protection of the SRCU read side critical section.
>> This SRCU read side critical section prevents any pending debugfs_remove()
>> and friends to return to their callers. Since a file's private data must
>> only be freed after the return of debugfs_remove(), the ongoing proxied
>> call is guarded against any file removal race.
>> 
>> If, on the other hand, the initial call to debugfs_use_file_start() detects
>> that the dentry is dead, the wrapper simply returns -EIO and does not
>> forward the call. Note that the ->poll() wrapper is special in that its
>> signature does not allow for the return of arbitrary -EXXX values and thus,
>> POLLHUP is returned here.
>> 
>> In order not to pollute debugfs with wrapper definitions that aren't ever
>> needed, I chose not to define a wrapper for every struct file_operations
>> method possible. Instead, a wrapper is defined only for the subset of
>> methods which are actually set by any debugfs users.
>> Currently, these are:
>> 
>>   ->llseek()
>>   ->read()
>>   ->write()
>>   ->unlocked_ioctl()
>>   ->poll()
>> 
>> The ->release() wrapper is special in that it does not protect the original
>> ->release() in any way from dead files in order not to leak resources.
>> Thus, any ->release() handed to debugfs must implement file lifetime
>> management manually, if needed.
>> For only 33 out of a total of 434 releasers handed in to debugfs, it could
>> not be verified immediately whether they access data structures that might
>> have been freed upon a debugfs_remove() return in the meanwhile.
>> 
>> Export debugfs_use_file_start() and debugfs_use_file_finish() in order to
>> allow any ->release() to manually implement file lifetime management.
>> 
>> For a set of common cases of struct file_operations implemented by the
>> debugfs_core itself, future patches will incorporate file lifetime
>> management directly within those in order to allow for their unproxied
>> operation. Rename the original, non-proxying "debugfs_create_file()" to
>> "debugfs_create_file_unsafe()" and keep it for future internal use by
>> debugfs itself. Factor out code common to both into the new
>> __debugfs_create_file().
>> 
>> Signed-off-by: Nicolai Stange <nicstange@gmail.com>

Thanks a million for reporting!

1.) Do you have lockdep enabled?

2.) Does this happen before or after userspace init has been spawned,
    i.e. does the lockup happen at debugfs file creation time or
    possibly at usage time?

Thank you,

Nicolai

[Cocci] [PATCH v6 2/8] debugfs: prevent access to removed files' private data

[Nicolai Stange]

Hi Sasha,

Sasha Levin <sasha.levin@oracle.com> writes:

> On 03/22/2016 09:11 AM, Nicolai Stange wrote:
>> Upon return of debugfs_remove()/debugfs_remove_recursive(), it might
>> still be attempted to access associated private file data through
>> previously opened struct file objects. If that data has been freed by
>> the caller of debugfs_remove*() in the meanwhile, the reading/writing
>> process would either encounter a fault or, if the memory address in
>> question has been reassigned again, unrelated data structures could get
>> overwritten.
>> 
>> However, since debugfs files are seldomly removed, usually from module
>> exit handlers only, the impact is very low.
>> 
>> Currently, there are ~1000 call sites of debugfs_create_file() spread
>> throughout the whole tree and touching all of those struct file_operations
>> in order to make them file removal aware by means of checking the result of
>> debugfs_use_file_start() from within their methods is unfeasible.
>> 
>> Instead, wrap the struct file_operations by a lifetime managing proxy at
>> file open:
>> - In debugfs_create_file(), the original fops handed in has got stashed
>>   away in ->d_fsdata already.
>> - In debugfs_create_file(), install a proxy file_operations factory,
>>   debugfs_full_proxy_file_operations, at ->i_fop.
>> 
>> This proxy factory has got an ->open() method only. It carries out some
>> lifetime checks and if successful, dynamically allocates and sets up a new
>> struct file_operations proxy at ->f_op. Afterwards, it forwards to the
>> ->open() of the original struct file_operations in ->d_fsdata, if any.
>> 
>> The dynamically set up proxy at ->f_op has got a lifetime managing wrapper
>> set for each of the methods defined in the original struct file_operations
>> in ->d_fsdata.
>> 
>> Its ->release()er frees the proxy again and forwards to the original
>> ->release(), if any.
>> 
>> In order not to mislead the VFS layer, it is strictly necessary to leave
>> those fields blank in the proxy that have been NULL in the original
>> struct file_operations also, i.e. aren't supported. This is why there is a
>> need for dynamically allocated proxies. The choice made not to allocate a
>> proxy instance for every dentry at file creation, but for every
>> struct file object instantiated thereof is justified by the expected usage
>> pattern of debugfs, namely that in general very few files get opened more
>> than once at a time.
>> 
>> The wrapper methods set in the struct file_operations implement lifetime
>> managing by means of the SRCU protection facilities already in place for
>> debugfs:
>> They set up a SRCU read side critical section and check whether the dentry
>> is still alive by means of debugfs_use_file_start(). If so, they forward
>> the call to the original struct file_operation stored in ->d_fsdata, still
>> under the protection of the SRCU read side critical section.
>> This SRCU read side critical section prevents any pending debugfs_remove()
>> and friends to return to their callers. Since a file's private data must
>> only be freed after the return of debugfs_remove(), the ongoing proxied
>> call is guarded against any file removal race.
>> 
>> If, on the other hand, the initial call to debugfs_use_file_start() detects
>> that the dentry is dead, the wrapper simply returns -EIO and does not
>> forward the call. Note that the ->poll() wrapper is special in that its
>> signature does not allow for the return of arbitrary -EXXX values and thus,
>> POLLHUP is returned here.
>> 
>> In order not to pollute debugfs with wrapper definitions that aren't ever
>> needed, I chose not to define a wrapper for every struct file_operations
>> method possible. Instead, a wrapper is defined only for the subset of
>> methods which are actually set by any debugfs users.
>> Currently, these are:
>> 
>>   ->llseek()
>>   ->read()
>>   ->write()
>>   ->unlocked_ioctl()
>>   ->poll()
>> 
>> The ->release() wrapper is special in that it does not protect the original
>> ->release() in any way from dead files in order not to leak resources.
>> Thus, any ->release() handed to debugfs must implement file lifetime
>> management manually, if needed.
>> For only 33 out of a total of 434 releasers handed in to debugfs, it could
>> not be verified immediately whether they access data structures that might
>> have been freed upon a debugfs_remove() return in the meanwhile.
>> 
>> Export debugfs_use_file_start() and debugfs_use_file_finish() in order to
>> allow any ->release() to manually implement file lifetime management.
>> 
>> For a set of common cases of struct file_operations implemented by the
>> debugfs_core itself, future patches will incorporate file lifetime
>> management directly within those in order to allow for their unproxied
>> operation. Rename the original, non-proxying "debugfs_create_file()" to
>> "debugfs_create_file_unsafe()" and keep it for future internal use by
>> debugfs itself. Factor out code common to both into the new
>> __debugfs_create_file().
>> 
>> Signed-off-by: Nicolai Stange <nicstange@gmail.com>

Thanks a million for reporting!

1.) Do you have lockdep enabled?

2.) Does this happen before or after userspace init has been spawned,
    i.e. does the lockup happen at debugfs file creation time or
    possibly at usage time?

Thank you,

Nicolai

Re: [PATCH v6 2/8] debugfs: prevent access to removed files' private data

[Sasha Levin]

On 05/18/2016 11:01 AM, Nicolai Stange wrote:
> Thanks a million for reporting!
> 
> 1.) Do you have lockdep enabled?

Yup, nothing there.

> 2.) Does this happen before or after userspace init has been spawned,
>     i.e. does the lockup happen at debugfs file creation time or
>     possibly at usage time?

So I looked closer, and it seems to happen after starting syzkaller, which
as far as I know tries to open many different debugfs files.

Is there debug code I can add it that'll help us figure out what's up?


Thanks,
Sasha

Re: [PATCH v6 2/8] debugfs: prevent access to removed files' private data

[Greg Kroah-Hartman]

On Wed, May 18, 2016 at 11:18:16AM -0400, Sasha Levin wrote:
> On 05/18/2016 11:01 AM, Nicolai Stange wrote:
> > Thanks a million for reporting!
> > 
> > 1.) Do you have lockdep enabled?
> 
> Yup, nothing there.
> 
> > 2.) Does this happen before or after userspace init has been spawned,
> >     i.e. does the lockup happen at debugfs file creation time or
> >     possibly at usage time?
> 
> So I looked closer, and it seems to happen after starting syzkaller, which
> as far as I know tries to open many different debugfs files.
> 
> Is there debug code I can add it that'll help us figure out what's up?

Trying to figure out _which_ debugfs file is causing this would be
great, if at all possible.  strace?

Re: [PATCH v6 2/8] debugfs: prevent access to removed files' private data

[Sasha Levin]

On 05/18/2016 12:05 PM, Greg Kroah-Hartman wrote:
> On Wed, May 18, 2016 at 11:18:16AM -0400, Sasha Levin wrote:
>> On 05/18/2016 11:01 AM, Nicolai Stange wrote:
>>> Thanks a million for reporting!
>>>
>>> 1.) Do you have lockdep enabled?
>>
>> Yup, nothing there.
>>
>>> 2.) Does this happen before or after userspace init has been spawned,
>>>     i.e. does the lockup happen at debugfs file creation time or
>>>     possibly at usage time?
>>
>> So I looked closer, and it seems to happen after starting syzkaller, which
>> as far as I know tries to open many different debugfs files.
>>
>> Is there debug code I can add it that'll help us figure out what's up?
> 
> Trying to figure out _which_ debugfs file is causing this would be
> great, if at all possible.  strace?

What seems to be failing is syzkaller's attempt to mmap the coverage
debugfs file. So this isn't actually a kernel deadlock but syzkaller
misbehaves when that scenario happens.

Either way, it only fails to mmap with that commit that I've pointed
out.

	th->cover_fd = open("/sys/kernel/debug/kcov", O_RDWR);
	if (th->cover_fd == -1)
		fail("open of /sys/kernel/debug/kcov failed");
	if (ioctl(th->cover_fd, KCOV_INIT_TRACE, kCoverSize))
		fail("cover enable write failed");
	th->cover_data = (uintptr_t*)mmap(NULL, kCoverSize * sizeof(th->cover_data[0]), PROT_READ | PROT_WRITE, MAP_SHARED, th->cover_fd, 0);
	if ((void*)th->cover_data == MAP_FAILED)
		fail("cover mmap failed");

And it's the mmap() that fails with -ENODEV.


Thanks,
Sasha

Re: [PATCH v6 2/8] debugfs: prevent access to removed files' private data

[Nicolai Stange]

Sasha Levin <sasha.levin@oracle.com> writes:

> On 05/18/2016 12:05 PM, Greg Kroah-Hartman wrote:
>> On Wed, May 18, 2016 at 11:18:16AM -0400, Sasha Levin wrote:
>>> On 05/18/2016 11:01 AM, Nicolai Stange wrote:
>>>> Thanks a million for reporting!
>>>>
>>>> 1.) Do you have lockdep enabled?
>>>
>>> Yup, nothing there.
>>>
>>>> 2.) Does this happen before or after userspace init has been spawned,
>>>>     i.e. does the lockup happen at debugfs file creation time or
>>>>     possibly at usage time?
>>>
>>> So I looked closer, and it seems to happen after starting syzkaller, which
>>> as far as I know tries to open many different debugfs files.
>>>
>>> Is there debug code I can add it that'll help us figure out what's up?
>> 
>> Trying to figure out _which_ debugfs file is causing this would be
>> great, if at all possible.  strace?
>
> What seems to be failing is syzkaller's attempt to mmap the coverage
> debugfs file. So this isn't actually a kernel deadlock but syzkaller
> misbehaves when that scenario happens.
>
> Either way, it only fails to mmap with that commit that I've pointed
> out.

That info is really helpful here: the proxy file_operations introduced by
this commit doesn't have a ->mmap() defined, i.e. it is NULL from the
VFS layer's point of view.

The simple reason is that at the time I submitted this series, my
Coccinelle script didn't find any debugfs user with a ->mmap()
defined. Thus either that script was broken or things have changed in
the meanwhile.

I'll look into this tomorrow.

Thank you very much for the effort you put into this!

>
> 	th->cover_fd = open("/sys/kernel/debug/kcov", O_RDWR);
> 	if (th->cover_fd == -1)
> 		fail("open of /sys/kernel/debug/kcov failed");
> 	if (ioctl(th->cover_fd, KCOV_INIT_TRACE, kCoverSize))
> 		fail("cover enable write failed");
> 	th->cover_data = (uintptr_t*)mmap(NULL, kCoverSize * sizeof(th->cover_data[0]), PROT_READ | PROT_WRITE, MAP_SHARED, th->cover_fd, 0);
> 	if ((void*)th->cover_data == MAP_FAILED)
> 		fail("cover mmap failed");
>
> And it's the mmap() that fails with -ENODEV.

[Cocci] [PATCH v6 2/8] debugfs: prevent access to removed files' private data

[Nicolai Stange]

Sasha Levin <sasha.levin@oracle.com> writes:

> On 05/18/2016 12:05 PM, Greg Kroah-Hartman wrote:
>> On Wed, May 18, 2016 at 11:18:16AM -0400, Sasha Levin wrote:
>>> On 05/18/2016 11:01 AM, Nicolai Stange wrote:
>>>> Thanks a million for reporting!
>>>>
>>>> 1.) Do you have lockdep enabled?
>>>
>>> Yup, nothing there.
>>>
>>>> 2.) Does this happen before or after userspace init has been spawned,
>>>>     i.e. does the lockup happen at debugfs file creation time or
>>>>     possibly at usage time?
>>>
>>> So I looked closer, and it seems to happen after starting syzkaller, which
>>> as far as I know tries to open many different debugfs files.
>>>
>>> Is there debug code I can add it that'll help us figure out what's up?
>> 
>> Trying to figure out _which_ debugfs file is causing this would be
>> great, if at all possible.  strace?
>
> What seems to be failing is syzkaller's attempt to mmap the coverage
> debugfs file. So this isn't actually a kernel deadlock but syzkaller
> misbehaves when that scenario happens.
>
> Either way, it only fails to mmap with that commit that I've pointed
> out.

That info is really helpful here: the proxy file_operations introduced by
this commit doesn't have a ->mmap() defined, i.e. it is NULL from the
VFS layer's point of view.

The simple reason is that at the time I submitted this series, my
Coccinelle script didn't find any debugfs user with a ->mmap()
defined. Thus either that script was broken or things have changed in
the meanwhile.

I'll look into this tomorrow.

Thank you very much for the effort you put into this!

>
> 	th->cover_fd = open("/sys/kernel/debug/kcov", O_RDWR);
> 	if (th->cover_fd == -1)
> 		fail("open of /sys/kernel/debug/kcov failed");
> 	if (ioctl(th->cover_fd, KCOV_INIT_TRACE, kCoverSize))
> 		fail("cover enable write failed");
> 	th->cover_data = (uintptr_t*)mmap(NULL, kCoverSize * sizeof(th->cover_data[0]), PROT_READ | PROT_WRITE, MAP_SHARED, th->cover_fd, 0);
> 	if ((void*)th->cover_data == MAP_FAILED)
> 		fail("cover mmap failed");
>
> And it's the mmap() that fails with -ENODEV.

Re: [PATCH v6 2/8] debugfs: prevent access to removed files' private data

[Nicolai Stange]

[-- Attachment #1: Type: text/plain, Size: 2719 bytes --]

Nicolai Stange <nicstange@gmail.com> writes:

> Sasha Levin <sasha.levin@oracle.com> writes:
>
>> On 05/18/2016 12:05 PM, Greg Kroah-Hartman wrote:
>>> On Wed, May 18, 2016 at 11:18:16AM -0400, Sasha Levin wrote:
>>>> On 05/18/2016 11:01 AM, Nicolai Stange wrote:
>>>>> Thanks a million for reporting!
>>>>>
>>>>> 1.) Do you have lockdep enabled?
>>>>
>>>> Yup, nothing there.
>>>>
>>>>> 2.) Does this happen before or after userspace init has been spawned,
>>>>>     i.e. does the lockup happen at debugfs file creation time or
>>>>>     possibly at usage time?
>>>>
>>>> So I looked closer, and it seems to happen after starting syzkaller, which
>>>> as far as I know tries to open many different debugfs files.
>>>>
>>>> Is there debug code I can add it that'll help us figure out what's up?
>>> 
>>> Trying to figure out _which_ debugfs file is causing this would be
>>> great, if at all possible.  strace?
>>
>> What seems to be failing is syzkaller's attempt to mmap the coverage
>> debugfs file. So this isn't actually a kernel deadlock but syzkaller
>> misbehaves when that scenario happens.
>>
>> Either way, it only fails to mmap with that commit that I've pointed
>> out.
>
> That info is really helpful here: the proxy file_operations introduced by
> this commit doesn't have a ->mmap() defined, i.e. it is NULL from the
> VFS layer's point of view.
>
> The simple reason is that at the time I submitted this series, my
> Coccinelle script didn't find any debugfs user with a ->mmap()
> defined. Thus either that script was broken or things have changed in
> the meanwhile.

Thankfully, it's the latter :)
See the attached cocci script I used back then.

It now reports:

  ./drivers/staging/android/sync_debug.c:330:1-20: unsupported file_operations given to debugfs
  ./kernel/kcov.c:267:6-25: unsupported file_operations given to debugfs

The kcov's ->mmap() has been introduced by

  5c9a8750a640 ("kernel: add kcov code coverage")

dated from March this year.

Since that kcov debugfs file is never removed, it needs no protecting
proxy and thus, a replacement of debugfs_create_file() by
debugfs_create_file_unsafe() will do the trick here.

I'll send patches addressing the above two issues.

>>
>> 	th->cover_fd = open("/sys/kernel/debug/kcov", O_RDWR);
>> 	if (th->cover_fd == -1)
>> 		fail("open of /sys/kernel/debug/kcov failed");
>> 	if (ioctl(th->cover_fd, KCOV_INIT_TRACE, kCoverSize))
>> 		fail("cover enable write failed");
>> 	th->cover_data = (uintptr_t*)mmap(NULL, kCoverSize * sizeof(th->cover_data[0]), PROT_READ | PROT_WRITE, MAP_SHARED, th->cover_fd, 0);
>> 	if ((void*)th->cover_data == MAP_FAILED)
>> 		fail("cover mmap failed");
>>
>> And it's the mmap() that fails with -ENODEV.


[-- Attachment #2: debugfs_unsupp_fops.cocci --]
[-- Type: text/plain, Size: 644 bytes --]

virtual report
virtual org

@unsupp_fops@
identifier fops;
expression e;
identifier m != {owner, open, release, llseek, read, write, poll, unlocked_ioctl};
@@
struct file_operations fops = {
 .m = e,
};

@unsupp_debugfs_fops@
expression name, mode, parent, data;
identifier unsupp_fops.fops;
position p;
@@
debugfs_create_file@p(name, mode, parent, data, &fops)



@script:python depends on report@
p << unsupp_debugfs_fops.p;
@@
coccilib.report.print_report(p[0], "unsupported file_operations given to debugfs")

@script:python depends on org@
p << unsupp_debugfs_fops.p;
@@
cocci.print_main("unsupported file_operations given to debugfs", p)

[Cocci] [PATCH v6 2/8] debugfs: prevent access to removed files' private data

[Nicolai Stange]

Nicolai Stange <nicstange@gmail.com> writes:

> Sasha Levin <sasha.levin@oracle.com> writes:
>
>> On 05/18/2016 12:05 PM, Greg Kroah-Hartman wrote:
>>> On Wed, May 18, 2016 at 11:18:16AM -0400, Sasha Levin wrote:
>>>> On 05/18/2016 11:01 AM, Nicolai Stange wrote:
>>>>> Thanks a million for reporting!
>>>>>
>>>>> 1.) Do you have lockdep enabled?
>>>>
>>>> Yup, nothing there.
>>>>
>>>>> 2.) Does this happen before or after userspace init has been spawned,
>>>>>     i.e. does the lockup happen at debugfs file creation time or
>>>>>     possibly at usage time?
>>>>
>>>> So I looked closer, and it seems to happen after starting syzkaller, which
>>>> as far as I know tries to open many different debugfs files.
>>>>
>>>> Is there debug code I can add it that'll help us figure out what's up?
>>> 
>>> Trying to figure out _which_ debugfs file is causing this would be
>>> great, if at all possible.  strace?
>>
>> What seems to be failing is syzkaller's attempt to mmap the coverage
>> debugfs file. So this isn't actually a kernel deadlock but syzkaller
>> misbehaves when that scenario happens.
>>
>> Either way, it only fails to mmap with that commit that I've pointed
>> out.
>
> That info is really helpful here: the proxy file_operations introduced by
> this commit doesn't have a ->mmap() defined, i.e. it is NULL from the
> VFS layer's point of view.
>
> The simple reason is that at the time I submitted this series, my
> Coccinelle script didn't find any debugfs user with a ->mmap()
> defined. Thus either that script was broken or things have changed in
> the meanwhile.

Thankfully, it's the latter :)
See the attached cocci script I used back then.

It now reports:

  ./drivers/staging/android/sync_debug.c:330:1-20: unsupported file_operations given to debugfs
  ./kernel/kcov.c:267:6-25: unsupported file_operations given to debugfs

The kcov's ->mmap() has been introduced by

  5c9a8750a640 ("kernel: add kcov code coverage")

dated from March this year.

Since that kcov debugfs file is never removed, it needs no protecting
proxy and thus, a replacement of debugfs_create_file() by
debugfs_create_file_unsafe() will do the trick here.

I'll send patches addressing the above two issues.

>>
>> 	th->cover_fd = open("/sys/kernel/debug/kcov", O_RDWR);
>> 	if (th->cover_fd == -1)
>> 		fail("open of /sys/kernel/debug/kcov failed");
>> 	if (ioctl(th->cover_fd, KCOV_INIT_TRACE, kCoverSize))
>> 		fail("cover enable write failed");
>> 	th->cover_data = (uintptr_t*)mmap(NULL, kCoverSize * sizeof(th->cover_data[0]), PROT_READ | PROT_WRITE, MAP_SHARED, th->cover_fd, 0);
>> 	if ((void*)th->cover_data == MAP_FAILED)
>> 		fail("cover mmap failed");
>>
>> And it's the mmap() that fails with -ENODEV.

-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: debugfs_unsupp_fops.cocci
URL: <https://systeme.lip6.fr/pipermail/cocci/attachments/20160522/23b3ce79/attachment.ksh>

Re: [PATCH v6 2/8] debugfs: prevent access to removed files' private data

[Nicolai Stange]

Sasha Levin <sasha.levin@oracle.com> writes:

> On 05/18/2016 11:01 AM, Nicolai Stange wrote:
>> Thanks a million for reporting!
>> 
>> 1.) Do you have lockdep enabled?
>
> Yup, nothing there.
>
>> 2.) Does this happen before or after userspace init has been spawned,
>>     i.e. does the lockup happen at debugfs file creation time or
>>     possibly at usage time?
>
> So I looked closer, and it seems to happen after starting syzkaller, which
> as far as I know tries to open many different debugfs files.
>
> Is there debug code I can add it that'll help us figure out what's up?

Could you try the patch below? I stared at the new full_proxy_open() for
a while now and had to recognize the fact that if the original real_fops'
->open() fails, then its owning module's reference won't ever get
dropped :(

diff --git a/fs/debugfs/file.c b/fs/debugfs/file.c
index 6eb58a8..2e663d4 100644
--- a/fs/debugfs/file.c
+++ b/fs/debugfs/file.c
@@ -263,10 +263,14 @@ static int full_proxy_open(struct inode *inode, struct file *filp)
        if (real_fops->open) {
                r = real_fops->open(inode, filp);

-               if (filp->f_op != proxy_fops) {
+               if (r) {
+                       replace_fops(filp, d_inode(dentry)->i_fop);
+                       goto free_proxy;
+               } else if (filp->f_op != proxy_fops) {
                        /* No protection against file removal anymore. */
                        WARN(1, "debugfs file owner replaced proxy fops: %pd",
                                dentry);
+                       replace_fops(filp, d_inode(dentry)->i_fop);
                        goto free_proxy;
                }
        }


I don't see directly how this could lead to lockups, but I think it's
better to rule out the obvious before inserting more or less random
printks...

Thank you very much again,

Nicolai

[Cocci] [PATCH v6 2/8] debugfs: prevent access to removed files' private data

[Nicolai Stange]

Sasha Levin <sasha.levin@oracle.com> writes:

> On 05/18/2016 11:01 AM, Nicolai Stange wrote:
>> Thanks a million for reporting!
>> 
>> 1.) Do you have lockdep enabled?
>
> Yup, nothing there.
>
>> 2.) Does this happen before or after userspace init has been spawned,
>>     i.e. does the lockup happen at debugfs file creation time or
>>     possibly at usage time?
>
> So I looked closer, and it seems to happen after starting syzkaller, which
> as far as I know tries to open many different debugfs files.
>
> Is there debug code I can add it that'll help us figure out what's up?

Could you try the patch below? I stared at the new full_proxy_open() for
a while now and had to recognize the fact that if the original real_fops'
->open() fails, then its owning module's reference won't ever get
dropped :(

diff --git a/fs/debugfs/file.c b/fs/debugfs/file.c
index 6eb58a8..2e663d4 100644
--- a/fs/debugfs/file.c
+++ b/fs/debugfs/file.c
@@ -263,10 +263,14 @@ static int full_proxy_open(struct inode *inode, struct file *filp)
        if (real_fops->open) {
                r = real_fops->open(inode, filp);

-               if (filp->f_op != proxy_fops) {
+               if (r) {
+                       replace_fops(filp, d_inode(dentry)->i_fop);
+                       goto free_proxy;
+               } else if (filp->f_op != proxy_fops) {
                        /* No protection against file removal anymore. */
                        WARN(1, "debugfs file owner replaced proxy fops: %pd",
                                dentry);
+                       replace_fops(filp, d_inode(dentry)->i_fop);
                        goto free_proxy;
                }
        }


I don't see directly how this could lead to lockups, but I think it's
better to rule out the obvious before inserting more or less random
printks...

Thank you very much again,

Nicolai

Re: [PATCH v6 2/8] debugfs: prevent access to removed files' private data

[Sasha Levin]

On 05/18/2016 12:32 PM, Nicolai Stange wrote:
> Sasha Levin <sasha.levin@oracle.com> writes:
> 
>> On 05/18/2016 11:01 AM, Nicolai Stange wrote:
>>> Thanks a million for reporting!
>>>
>>> 1.) Do you have lockdep enabled?
>>
>> Yup, nothing there.
>>
>>> 2.) Does this happen before or after userspace init has been spawned,
>>>     i.e. does the lockup happen at debugfs file creation time or
>>>     possibly at usage time?
>>
>> So I looked closer, and it seems to happen after starting syzkaller, which
>> as far as I know tries to open many different debugfs files.
>>
>> Is there debug code I can add it that'll help us figure out what's up?
> 
> Could you try the patch below? I stared at the new full_proxy_open() for
> a while now and had to recognize the fact that if the original real_fops'
> ->open() fails, then its owning module's reference won't ever get
> dropped :(
> 
> diff --git a/fs/debugfs/file.c b/fs/debugfs/file.c
> index 6eb58a8..2e663d4 100644
> --- a/fs/debugfs/file.c
> +++ b/fs/debugfs/file.c
> @@ -263,10 +263,14 @@ static int full_proxy_open(struct inode *inode, struct file *filp)
>         if (real_fops->open) {
>                 r = real_fops->open(inode, filp);
> 
> -               if (filp->f_op != proxy_fops) {
> +               if (r) {
> +                       replace_fops(filp, d_inode(dentry)->i_fop);
> +                       goto free_proxy;
> +               } else if (filp->f_op != proxy_fops) {
>                         /* No protection against file removal anymore. */
>                         WARN(1, "debugfs file owner replaced proxy fops: %pd",
>                                 dentry);
> +                       replace_fops(filp, d_inode(dentry)->i_fop);
>                         goto free_proxy;
>                 }
>         }
> 
> 
> I don't see directly how this could lead to lockups, but I think it's
> better to rule out the obvious before inserting more or less random
> printks...
> 
> Thank you very much again,

Nope, that didn't do the trick.


Thanks,
Sasha

[PATCH v6 3/8] debugfs: add support for self-protecting attribute file fops

[Nicolai Stange]

In order to protect them against file removal issues, debugfs_create_file()
creates a lifetime managing proxy around each struct file_operations
handed in.

In cases where this struct file_operations is able to manage file lifetime
by itself already, the proxy created by debugfs is a waste of resources.

The most common class of struct file_operations given to debugfs are those
defined by means of the DEFINE_SIMPLE_ATTRIBUTE() macro.

Introduce a DEFINE_DEBUGFS_ATTRIBUTE() macro to allow any
struct file_operations of this class to be easily made file lifetime aware
and thus, to be operated unproxied.

Specifically, introduce debugfs_attr_read() and debugfs_attr_write()
which wrap simple_attr_read() and simple_attr_write() under the protection
of a debugfs_use_file_start()/debugfs_use_file_finish() pair.

Make DEFINE_DEBUGFS_ATTRIBUTE() set the defined struct file_operations'
->read() and ->write() members to these wrappers.

Export debugfs_create_file_unsafe() in order to allow debugfs users to
create their files in non-proxying operation mode.

Signed-off-by: Nicolai Stange <nicstange@gmail.com>
---
 fs/debugfs/file.c       | 28 ++++++++++++++++++++++++++++
 fs/debugfs/inode.c      | 28 ++++++++++++++++++++++++++++
 include/linux/debugfs.h | 26 ++++++++++++++++++++++++++
 3 files changed, 82 insertions(+)

diff --git a/fs/debugfs/file.c b/fs/debugfs/file.c
index 6eb58a8..8ef56d9 100644
--- a/fs/debugfs/file.c
+++ b/fs/debugfs/file.c
@@ -284,6 +284,34 @@ const struct file_operations debugfs_full_proxy_file_operations = {
 	.open = full_proxy_open,
 };
 
+ssize_t debugfs_attr_read(struct file *file, char __user *buf,
+			size_t len, loff_t *ppos)
+{
+	ssize_t ret;
+	int srcu_idx;
+
+	ret = debugfs_use_file_start(F_DENTRY(file), &srcu_idx);
+	if (likely(!ret))
+		ret = simple_attr_read(file, buf, len, ppos);
+	debugfs_use_file_finish(srcu_idx);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(debugfs_attr_read);
+
+ssize_t debugfs_attr_write(struct file *file, const char __user *buf,
+			 size_t len, loff_t *ppos)
+{
+	ssize_t ret;
+	int srcu_idx;
+
+	ret = debugfs_use_file_start(F_DENTRY(file), &srcu_idx);
+	if (likely(!ret))
+		ret = simple_attr_write(file, buf, len, ppos);
+	debugfs_use_file_finish(srcu_idx);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(debugfs_attr_write);
+
 static struct dentry *debugfs_create_mode(const char *name, umode_t mode,
 					  struct dentry *parent, void *value,
 				          const struct file_operations *fops,
diff --git a/fs/debugfs/inode.c b/fs/debugfs/inode.c
index 42a9b34..f95e355 100644
--- a/fs/debugfs/inode.c
+++ b/fs/debugfs/inode.c
@@ -368,6 +368,33 @@ struct dentry *debugfs_create_file(const char *name, umode_t mode,
 }
 EXPORT_SYMBOL_GPL(debugfs_create_file);
 
+/**
+ * debugfs_create_file_unsafe - create a file in the debugfs filesystem
+ * @name: a pointer to a string containing the name of the file to create.
+ * @mode: the permission that the file should have.
+ * @parent: a pointer to the parent dentry for this file.  This should be a
+ *          directory dentry if set.  If this parameter is NULL, then the
+ *          file will be created in the root of the debugfs filesystem.
+ * @data: a pointer to something that the caller will want to get to later
+ *        on.  The inode.i_private pointer will point to this value on
+ *        the open() call.
+ * @fops: a pointer to a struct file_operations that should be used for
+ *        this file.
+ *
+ * debugfs_create_file_unsafe() is completely analogous to
+ * debugfs_create_file(), the only difference being that the fops
+ * handed it will not get protected against file removals by the
+ * debugfs core.
+ *
+ * It is your responsibility to protect your struct file_operation
+ * methods against file removals by means of debugfs_use_file_start()
+ * and debugfs_use_file_finish(). ->open() is still protected by
+ * debugfs though.
+ *
+ * Any struct file_operations defined by means of
+ * DEFINE_DEBUGFS_ATTRIBUTE() is protected against file removals and
+ * thus, may be used here.
+ */
 struct dentry *debugfs_create_file_unsafe(const char *name, umode_t mode,
 				   struct dentry *parent, void *data,
 				   const struct file_operations *fops)
@@ -378,6 +405,7 @@ struct dentry *debugfs_create_file_unsafe(const char *name, umode_t mode,
 					&debugfs_noop_file_operations,
 				fops);
 }
+EXPORT_SYMBOL_GPL(debugfs_create_file_unsafe);
 
 /**
  * debugfs_create_file_size - create a file in the debugfs filesystem
diff --git a/include/linux/debugfs.h b/include/linux/debugfs.h
index a63e6ea..1438e23 100644
--- a/include/linux/debugfs.h
+++ b/include/linux/debugfs.h
@@ -50,6 +50,9 @@ extern struct srcu_struct debugfs_srcu;
 struct dentry *debugfs_create_file(const char *name, umode_t mode,
 				   struct dentry *parent, void *data,
 				   const struct file_operations *fops);
+struct dentry *debugfs_create_file_unsafe(const char *name, umode_t mode,
+				   struct dentry *parent, void *data,
+				   const struct file_operations *fops);
 
 struct dentry *debugfs_create_file_size(const char *name, umode_t mode,
 					struct dentry *parent, void *data,
@@ -74,6 +77,26 @@ int debugfs_use_file_start(const struct dentry *dentry, int *srcu_idx)
 
 void debugfs_use_file_finish(int srcu_idx) __releases(&debugfs_srcu);
 
+ssize_t debugfs_attr_read(struct file *file, char __user *buf,
+			size_t len, loff_t *ppos);
+ssize_t debugfs_attr_write(struct file *file, const char __user *buf,
+			size_t len, loff_t *ppos);
+
+#define DEFINE_DEBUGFS_ATTRIBUTE(__fops, __get, __set, __fmt)		\
+static int __fops ## _open(struct inode *inode, struct file *file)	\
+{									\
+	__simple_attr_check_format(__fmt, 0ull);			\
+	return simple_attr_open(inode, file, __get, __set, __fmt);	\
+}									\
+static const struct file_operations __fops = {				\
+	.owner	 = THIS_MODULE,					\
+	.open	 = __fops ## _open,					\
+	.release = simple_attr_release,				\
+	.read	 = debugfs_attr_read,					\
+	.write	 = debugfs_attr_write,					\
+	.llseek  = generic_file_llseek,				\
+}
+
 struct dentry *debugfs_rename(struct dentry *old_dir, struct dentry *old_dentry,
                 struct dentry *new_dir, const char *new_name);
 
@@ -193,6 +216,9 @@ static inline void debugfs_use_file_finish(int srcu_idx)
 	__releases(&debugfs_srcu)
 { }
 
+#define DEFINE_DEBUGFS_ATTRIBUTE(__fops, __get, __set, __fmt)	\
+	static const struct file_operations __fops = { 0 }
+
 static inline struct dentry *debugfs_rename(struct dentry *old_dir, struct dentry *old_dentry,
                 struct dentry *new_dir, char *new_name)
 {
-- 
2.7.4

[Cocci] [PATCH v6 3/8] debugfs: add support for self-protecting attribute file fops

[Nicolai Stange]

In order to protect them against file removal issues, debugfs_create_file()
creates a lifetime managing proxy around each struct file_operations
handed in.

In cases where this struct file_operations is able to manage file lifetime
by itself already, the proxy created by debugfs is a waste of resources.

The most common class of struct file_operations given to debugfs are those
defined by means of the DEFINE_SIMPLE_ATTRIBUTE() macro.

Introduce a DEFINE_DEBUGFS_ATTRIBUTE() macro to allow any
struct file_operations of this class to be easily made file lifetime aware
and thus, to be operated unproxied.

Specifically, introduce debugfs_attr_read() and debugfs_attr_write()
which wrap simple_attr_read() and simple_attr_write() under the protection
of a debugfs_use_file_start()/debugfs_use_file_finish() pair.

Make DEFINE_DEBUGFS_ATTRIBUTE() set the defined struct file_operations'
->read() and ->write() members to these wrappers.

Export debugfs_create_file_unsafe() in order to allow debugfs users to
create their files in non-proxying operation mode.

Signed-off-by: Nicolai Stange <nicstange@gmail.com>
---
 fs/debugfs/file.c       | 28 ++++++++++++++++++++++++++++
 fs/debugfs/inode.c      | 28 ++++++++++++++++++++++++++++
 include/linux/debugfs.h | 26 ++++++++++++++++++++++++++
 3 files changed, 82 insertions(+)

diff --git a/fs/debugfs/file.c b/fs/debugfs/file.c
index 6eb58a8..8ef56d9 100644
--- a/fs/debugfs/file.c
+++ b/fs/debugfs/file.c
@@ -284,6 +284,34 @@ const struct file_operations debugfs_full_proxy_file_operations = {
 	.open = full_proxy_open,
 };
 
+ssize_t debugfs_attr_read(struct file *file, char __user *buf,
+			size_t len, loff_t *ppos)
+{
+	ssize_t ret;
+	int srcu_idx;
+
+	ret = debugfs_use_file_start(F_DENTRY(file), &srcu_idx);
+	if (likely(!ret))
+		ret = simple_attr_read(file, buf, len, ppos);
+	debugfs_use_file_finish(srcu_idx);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(debugfs_attr_read);
+
+ssize_t debugfs_attr_write(struct file *file, const char __user *buf,
+			 size_t len, loff_t *ppos)
+{
+	ssize_t ret;
+	int srcu_idx;
+
+	ret = debugfs_use_file_start(F_DENTRY(file), &srcu_idx);
+	if (likely(!ret))
+		ret = simple_attr_write(file, buf, len, ppos);
+	debugfs_use_file_finish(srcu_idx);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(debugfs_attr_write);
+
 static struct dentry *debugfs_create_mode(const char *name, umode_t mode,
 					  struct dentry *parent, void *value,
 				          const struct file_operations *fops,
diff --git a/fs/debugfs/inode.c b/fs/debugfs/inode.c
index 42a9b34..f95e355 100644
--- a/fs/debugfs/inode.c
+++ b/fs/debugfs/inode.c
@@ -368,6 +368,33 @@ struct dentry *debugfs_create_file(const char *name, umode_t mode,
 }
 EXPORT_SYMBOL_GPL(debugfs_create_file);
 
+/**
+ * debugfs_create_file_unsafe - create a file in the debugfs filesystem
+ * @name: a pointer to a string containing the name of the file to create.
+ * @mode: the permission that the file should have.
+ * @parent: a pointer to the parent dentry for this file.  This should be a
+ *          directory dentry if set.  If this parameter is NULL, then the
+ *          file will be created in the root of the debugfs filesystem.
+ * @data: a pointer to something that the caller will want to get to later
+ *        on.  The inode.i_private pointer will point to this value on
+ *        the open() call.
+ * @fops: a pointer to a struct file_operations that should be used for
+ *        this file.
+ *
+ * debugfs_create_file_unsafe() is completely analogous to
+ * debugfs_create_file(), the only difference being that the fops
+ * handed it will not get protected against file removals by the
+ * debugfs core.
+ *
+ * It is your responsibility to protect your struct file_operation
+ * methods against file removals by means of debugfs_use_file_start()
+ * and debugfs_use_file_finish(). ->open() is still protected by
+ * debugfs though.
+ *
+ * Any struct file_operations defined by means of
+ * DEFINE_DEBUGFS_ATTRIBUTE() is protected against file removals and
+ * thus, may be used here.
+ */
 struct dentry *debugfs_create_file_unsafe(const char *name, umode_t mode,
 				   struct dentry *parent, void *data,
 				   const struct file_operations *fops)
@@ -378,6 +405,7 @@ struct dentry *debugfs_create_file_unsafe(const char *name, umode_t mode,
 					&debugfs_noop_file_operations,
 				fops);
 }
+EXPORT_SYMBOL_GPL(debugfs_create_file_unsafe);
 
 /**
  * debugfs_create_file_size - create a file in the debugfs filesystem
diff --git a/include/linux/debugfs.h b/include/linux/debugfs.h
index a63e6ea..1438e23 100644
--- a/include/linux/debugfs.h
+++ b/include/linux/debugfs.h
@@ -50,6 +50,9 @@ extern struct srcu_struct debugfs_srcu;
 struct dentry *debugfs_create_file(const char *name, umode_t mode,
 				   struct dentry *parent, void *data,
 				   const struct file_operations *fops);
+struct dentry *debugfs_create_file_unsafe(const char *name, umode_t mode,
+				   struct dentry *parent, void *data,
+				   const struct file_operations *fops);
 
 struct dentry *debugfs_create_file_size(const char *name, umode_t mode,
 					struct dentry *parent, void *data,
@@ -74,6 +77,26 @@ int debugfs_use_file_start(const struct dentry *dentry, int *srcu_idx)
 
 void debugfs_use_file_finish(int srcu_idx) __releases(&debugfs_srcu);
 
+ssize_t debugfs_attr_read(struct file *file, char __user *buf,
+			size_t len, loff_t *ppos);
+ssize_t debugfs_attr_write(struct file *file, const char __user *buf,
+			size_t len, loff_t *ppos);
+
+#define DEFINE_DEBUGFS_ATTRIBUTE(__fops, __get, __set, __fmt)		\
+static int __fops ## _open(struct inode *inode, struct file *file)	\
+{									\
+	__simple_attr_check_format(__fmt, 0ull);			\
+	return simple_attr_open(inode, file, __get, __set, __fmt);	\
+}									\
+static const struct file_operations __fops = {				\
+	.owner	 = THIS_MODULE,					\
+	.open	 = __fops ## _open,					\
+	.release = simple_attr_release,				\
+	.read	 = debugfs_attr_read,					\
+	.write	 = debugfs_attr_write,					\
+	.llseek  = generic_file_llseek,				\
+}
+
 struct dentry *debugfs_rename(struct dentry *old_dir, struct dentry *old_dentry,
                 struct dentry *new_dir, const char *new_name);
 
@@ -193,6 +216,9 @@ static inline void debugfs_use_file_finish(int srcu_idx)
 	__releases(&debugfs_srcu)
 { }
 
+#define DEFINE_DEBUGFS_ATTRIBUTE(__fops, __get, __set, __fmt)	\
+	static const struct file_operations __fops = { 0 }
+
 static inline struct dentry *debugfs_rename(struct dentry *old_dir, struct dentry *old_dentry,
                 struct dentry *new_dir, char *new_name)
 {
-- 
2.7.4

[PATCH v6 4/8] debugfs, coccinelle: check for obsolete DEFINE_SIMPLE_ATTRIBUTE() usage

[Nicolai Stange]

In order to protect against file removal races, debugfs files created via
debugfs_create_file() now get wrapped by a struct file_operations at their
opening.

If the original struct file_operations are known to be safe against removal
races by themselves already, the proxy creation may be bypassed by creating
the files through debugfs_create_file_unsafe().

In order to help debugfs users who use the common
  DEFINE_SIMPLE_ATTRIBUTE() + debugfs_create_file()
idiom to transition to removal safe struct file_operations, the helper
macro DEFINE_DEBUGFS_ATTRIBUTE() has been introduced.

Thus, the preferred strategy is to use
  DEFINE_DEBUGFS_ATTRIBUTE() + debugfs_create_file_unsafe()
now.

Introduce a Coccinelle script that searches for
DEFINE_SIMPLE_ATTRIBUTE()-defined struct file_operations handed into
debugfs_create_file(). Suggest to turn these usages into the
  DEFINE_DEBUGFS_ATTRIBUTE() + debugfs_create_file_unsafe()
pattern.

Signed-off-by: Nicolai Stange <nicstange@gmail.com>
---
 .../api/debugfs/debugfs_simple_attr.cocci          | 67 ++++++++++++++++++++++
 1 file changed, 67 insertions(+)
 create mode 100644 scripts/coccinelle/api/debugfs/debugfs_simple_attr.cocci

diff --git a/scripts/coccinelle/api/debugfs/debugfs_simple_attr.cocci b/scripts/coccinelle/api/debugfs/debugfs_simple_attr.cocci
new file mode 100644
index 0000000..85cf540
--- /dev/null
+++ b/scripts/coccinelle/api/debugfs/debugfs_simple_attr.cocci
@@ -0,0 +1,67 @@
+/// Use DEFINE_DEBUGFS_ATTRIBUTE rather than DEFINE_SIMPLE_ATTRIBUTE
+/// for debugfs files.
+///
+//# Rationale: DEFINE_SIMPLE_ATTRIBUTE + debugfs_create_file()
+//# imposes some significant overhead as compared to
+//# DEFINE_DEBUGFS_ATTRIBUTE + debugfs_create_file_unsafe().
+//
+// Copyright (C): 2016 Nicolai Stange
+// Options: --no-includes
+//
+
+virtual context
+virtual patch
+virtual org
+virtual report
+
+@dsa@
+declarer name DEFINE_SIMPLE_ATTRIBUTE;
+identifier dsa_fops;
+expression dsa_get, dsa_set, dsa_fmt;
+position p;
+@@
+DEFINE_SIMPLE_ATTRIBUTE@p(dsa_fops, dsa_get, dsa_set, dsa_fmt);
+
+@dcf@
+expression name, mode, parent, data;
+identifier dsa.dsa_fops;
+@@
+debugfs_create_file(name, mode, parent, data, &dsa_fops)
+
+
+@context_dsa depends on context && dcf@
+declarer name DEFINE_DEBUGFS_ATTRIBUTE;
+identifier dsa.dsa_fops;
+expression dsa.dsa_get, dsa.dsa_set, dsa.dsa_fmt;
+@@
+* DEFINE_SIMPLE_ATTRIBUTE(dsa_fops, dsa_get, dsa_set, dsa_fmt);
+
+
+@patch_dcf depends on patch expression@
+expression name, mode, parent, data;
+identifier dsa.dsa_fops;
+@@
+- debugfs_create_file(name, mode, parent, data, &dsa_fops)
++ debugfs_create_file_unsafe(name, mode, parent, data, &dsa_fops)
+
+@patch_dsa depends on patch_dcf && patch@
+identifier dsa.dsa_fops;
+expression dsa.dsa_get, dsa.dsa_set, dsa.dsa_fmt;
+@@
+- DEFINE_SIMPLE_ATTRIBUTE(dsa_fops, dsa_get, dsa_set, dsa_fmt);
++ DEFINE_DEBUGFS_ATTRIBUTE(dsa_fops, dsa_get, dsa_set, dsa_fmt);
+
+
+@script:python depends on org && dcf@
+fops << dsa.dsa_fops;
+p << dsa.p;
+@@
+msg="%s should be defined with DEFINE_DEBUGFS_ATTRIBUTE" % (fops)
+coccilib.org.print_todo(p[0], msg)
+
+@script:python depends on report && dcf@
+fops << dsa.dsa_fops;
+p << dsa.p;
+@@
+msg="WARNING: %s should be defined with DEFINE_DEBUGFS_ATTRIBUTE" % (fops)
+coccilib.report.print_report(p[0], msg)
-- 
2.7.4

[Cocci] [PATCH v6 4/8] debugfs, coccinelle: check for obsolete DEFINE_SIMPLE_ATTRIBUTE() usage

[Nicolai Stange]

In order to protect against file removal races, debugfs files created via
debugfs_create_file() now get wrapped by a struct file_operations at their
opening.

If the original struct file_operations are known to be safe against removal
races by themselves already, the proxy creation may be bypassed by creating
the files through debugfs_create_file_unsafe().

In order to help debugfs users who use the common
  DEFINE_SIMPLE_ATTRIBUTE() + debugfs_create_file()
idiom to transition to removal safe struct file_operations, the helper
macro DEFINE_DEBUGFS_ATTRIBUTE() has been introduced.

Thus, the preferred strategy is to use
  DEFINE_DEBUGFS_ATTRIBUTE() + debugfs_create_file_unsafe()
now.

Introduce a Coccinelle script that searches for
DEFINE_SIMPLE_ATTRIBUTE()-defined struct file_operations handed into
debugfs_create_file(). Suggest to turn these usages into the
  DEFINE_DEBUGFS_ATTRIBUTE() + debugfs_create_file_unsafe()
pattern.

Signed-off-by: Nicolai Stange <nicstange@gmail.com>
---
 .../api/debugfs/debugfs_simple_attr.cocci          | 67 ++++++++++++++++++++++
 1 file changed, 67 insertions(+)
 create mode 100644 scripts/coccinelle/api/debugfs/debugfs_simple_attr.cocci

diff --git a/scripts/coccinelle/api/debugfs/debugfs_simple_attr.cocci b/scripts/coccinelle/api/debugfs/debugfs_simple_attr.cocci
new file mode 100644
index 0000000..85cf540
--- /dev/null
+++ b/scripts/coccinelle/api/debugfs/debugfs_simple_attr.cocci
@@ -0,0 +1,67 @@
+/// Use DEFINE_DEBUGFS_ATTRIBUTE rather than DEFINE_SIMPLE_ATTRIBUTE
+/// for debugfs files.
+///
+//# Rationale: DEFINE_SIMPLE_ATTRIBUTE + debugfs_create_file()
+//# imposes some significant overhead as compared to
+//# DEFINE_DEBUGFS_ATTRIBUTE + debugfs_create_file_unsafe().
+//
+// Copyright (C): 2016 Nicolai Stange
+// Options: --no-includes
+//
+
+virtual context
+virtual patch
+virtual org
+virtual report
+
+ at dsa@
+declarer name DEFINE_SIMPLE_ATTRIBUTE;
+identifier dsa_fops;
+expression dsa_get, dsa_set, dsa_fmt;
+position p;
+@@
+DEFINE_SIMPLE_ATTRIBUTE at p(dsa_fops, dsa_get, dsa_set, dsa_fmt);
+
+ at dcf@
+expression name, mode, parent, data;
+identifier dsa.dsa_fops;
+@@
+debugfs_create_file(name, mode, parent, data, &dsa_fops)
+
+
+ at context_dsa depends on context && dcf@
+declarer name DEFINE_DEBUGFS_ATTRIBUTE;
+identifier dsa.dsa_fops;
+expression dsa.dsa_get, dsa.dsa_set, dsa.dsa_fmt;
+@@
+* DEFINE_SIMPLE_ATTRIBUTE(dsa_fops, dsa_get, dsa_set, dsa_fmt);
+
+
+ at patch_dcf depends on patch expression@
+expression name, mode, parent, data;
+identifier dsa.dsa_fops;
+@@
+- debugfs_create_file(name, mode, parent, data, &dsa_fops)
++ debugfs_create_file_unsafe(name, mode, parent, data, &dsa_fops)
+
+ at patch_dsa depends on patch_dcf && patch@
+identifier dsa.dsa_fops;
+expression dsa.dsa_get, dsa.dsa_set, dsa.dsa_fmt;
+@@
+- DEFINE_SIMPLE_ATTRIBUTE(dsa_fops, dsa_get, dsa_set, dsa_fmt);
++ DEFINE_DEBUGFS_ATTRIBUTE(dsa_fops, dsa_get, dsa_set, dsa_fmt);
+
+
+ at script:python depends on org && dcf@
+fops << dsa.dsa_fops;
+p << dsa.p;
+@@
+msg="%s should be defined with DEFINE_DEBUGFS_ATTRIBUTE" % (fops)
+coccilib.org.print_todo(p[0], msg)
+
+@script:python depends on report && dcf@
+fops << dsa.dsa_fops;
+p << dsa.p;
+@@
+msg="WARNING: %s should be defined with DEFINE_DEBUGFS_ATTRIBUTE" % (fops)
+coccilib.report.print_report(p[0], msg)
-- 
2.7.4

Re: [PATCH v6 4/8] debugfs, coccinelle: check for obsolete DEFINE_SIMPLE_ATTRIBUTE() usage

[Julia Lawall]

On Tue, 22 Mar 2016, Nicolai Stange wrote:

> In order to protect against file removal races, debugfs files created via
> debugfs_create_file() now get wrapped by a struct file_operations at their
> opening.
>
> If the original struct file_operations are known to be safe against removal
> races by themselves already, the proxy creation may be bypassed by creating
> the files through debugfs_create_file_unsafe().
>
> In order to help debugfs users who use the common
>   DEFINE_SIMPLE_ATTRIBUTE() + debugfs_create_file()
> idiom to transition to removal safe struct file_operations, the helper
> macro DEFINE_DEBUGFS_ATTRIBUTE() has been introduced.
>
> Thus, the preferred strategy is to use
>   DEFINE_DEBUGFS_ATTRIBUTE() + debugfs_create_file_unsafe()
> now.
>
> Introduce a Coccinelle script that searches for
> DEFINE_SIMPLE_ATTRIBUTE()-defined struct file_operations handed into
> debugfs_create_file(). Suggest to turn these usages into the
>   DEFINE_DEBUGFS_ATTRIBUTE() + debugfs_create_file_unsafe()
> pattern.
>
> Signed-off-by: Nicolai Stange <nicstange@gmail.com>

In terms of the structure of the semantic patch:
Acked-by: Julia Lawall <Julia.Lawall@lip6.fr>

> ---
>  .../api/debugfs/debugfs_simple_attr.cocci          | 67 ++++++++++++++++++++++
>  1 file changed, 67 insertions(+)
>  create mode 100644 scripts/coccinelle/api/debugfs/debugfs_simple_attr.cocci
>
> diff --git a/scripts/coccinelle/api/debugfs/debugfs_simple_attr.cocci b/scripts/coccinelle/api/debugfs/debugfs_simple_attr.cocci
> new file mode 100644
> index 0000000..85cf540
> --- /dev/null
> +++ b/scripts/coccinelle/api/debugfs/debugfs_simple_attr.cocci
> @@ -0,0 +1,67 @@
> +/// Use DEFINE_DEBUGFS_ATTRIBUTE rather than DEFINE_SIMPLE_ATTRIBUTE
> +/// for debugfs files.
> +///
> +//# Rationale: DEFINE_SIMPLE_ATTRIBUTE + debugfs_create_file()
> +//# imposes some significant overhead as compared to
> +//# DEFINE_DEBUGFS_ATTRIBUTE + debugfs_create_file_unsafe().
> +//
> +// Copyright (C): 2016 Nicolai Stange
> +// Options: --no-includes
> +//
> +
> +virtual context
> +virtual patch
> +virtual org
> +virtual report
> +
> +@dsa@
> +declarer name DEFINE_SIMPLE_ATTRIBUTE;
> +identifier dsa_fops;
> +expression dsa_get, dsa_set, dsa_fmt;
> +position p;
> +@@
> +DEFINE_SIMPLE_ATTRIBUTE@p(dsa_fops, dsa_get, dsa_set, dsa_fmt);
> +
> +@dcf@
> +expression name, mode, parent, data;
> +identifier dsa.dsa_fops;
> +@@
> +debugfs_create_file(name, mode, parent, data, &dsa_fops)
> +
> +
> +@context_dsa depends on context && dcf@
> +declarer name DEFINE_DEBUGFS_ATTRIBUTE;
> +identifier dsa.dsa_fops;
> +expression dsa.dsa_get, dsa.dsa_set, dsa.dsa_fmt;
> +@@
> +* DEFINE_SIMPLE_ATTRIBUTE(dsa_fops, dsa_get, dsa_set, dsa_fmt);
> +
> +
> +@patch_dcf depends on patch expression@
> +expression name, mode, parent, data;
> +identifier dsa.dsa_fops;
> +@@
> +- debugfs_create_file(name, mode, parent, data, &dsa_fops)
> ++ debugfs_create_file_unsafe(name, mode, parent, data, &dsa_fops)
> +
> +@patch_dsa depends on patch_dcf && patch@
> +identifier dsa.dsa_fops;
> +expression dsa.dsa_get, dsa.dsa_set, dsa.dsa_fmt;
> +@@
> +- DEFINE_SIMPLE_ATTRIBUTE(dsa_fops, dsa_get, dsa_set, dsa_fmt);
> ++ DEFINE_DEBUGFS_ATTRIBUTE(dsa_fops, dsa_get, dsa_set, dsa_fmt);
> +
> +
> +@script:python depends on org && dcf@
> +fops << dsa.dsa_fops;
> +p << dsa.p;
> +@@
> +msg="%s should be defined with DEFINE_DEBUGFS_ATTRIBUTE" % (fops)
> +coccilib.org.print_todo(p[0], msg)
> +
> +@script:python depends on report && dcf@
> +fops << dsa.dsa_fops;
> +p << dsa.p;
> +@@
> +msg="WARNING: %s should be defined with DEFINE_DEBUGFS_ATTRIBUTE" % (fops)
> +coccilib.report.print_report(p[0], msg)
> --
> 2.7.4
>
>

[Cocci] [PATCH v6 4/8] debugfs, coccinelle: check for obsolete DEFINE_SIMPLE_ATTRIBUTE() usage

[Julia Lawall]

On Tue, 22 Mar 2016, Nicolai Stange wrote:

> In order to protect against file removal races, debugfs files created via
> debugfs_create_file() now get wrapped by a struct file_operations at their
> opening.
>
> If the original struct file_operations are known to be safe against removal
> races by themselves already, the proxy creation may be bypassed by creating
> the files through debugfs_create_file_unsafe().
>
> In order to help debugfs users who use the common
>   DEFINE_SIMPLE_ATTRIBUTE() + debugfs_create_file()
> idiom to transition to removal safe struct file_operations, the helper
> macro DEFINE_DEBUGFS_ATTRIBUTE() has been introduced.
>
> Thus, the preferred strategy is to use
>   DEFINE_DEBUGFS_ATTRIBUTE() + debugfs_create_file_unsafe()
> now.
>
> Introduce a Coccinelle script that searches for
> DEFINE_SIMPLE_ATTRIBUTE()-defined struct file_operations handed into
> debugfs_create_file(). Suggest to turn these usages into the
>   DEFINE_DEBUGFS_ATTRIBUTE() + debugfs_create_file_unsafe()
> pattern.
>
> Signed-off-by: Nicolai Stange <nicstange@gmail.com>

In terms of the structure of the semantic patch:
Acked-by: Julia Lawall <Julia.Lawall@lip6.fr>

> ---
>  .../api/debugfs/debugfs_simple_attr.cocci          | 67 ++++++++++++++++++++++
>  1 file changed, 67 insertions(+)
>  create mode 100644 scripts/coccinelle/api/debugfs/debugfs_simple_attr.cocci
>
> diff --git a/scripts/coccinelle/api/debugfs/debugfs_simple_attr.cocci b/scripts/coccinelle/api/debugfs/debugfs_simple_attr.cocci
> new file mode 100644
> index 0000000..85cf540
> --- /dev/null
> +++ b/scripts/coccinelle/api/debugfs/debugfs_simple_attr.cocci
> @@ -0,0 +1,67 @@
> +/// Use DEFINE_DEBUGFS_ATTRIBUTE rather than DEFINE_SIMPLE_ATTRIBUTE
> +/// for debugfs files.
> +///
> +//# Rationale: DEFINE_SIMPLE_ATTRIBUTE + debugfs_create_file()
> +//# imposes some significant overhead as compared to
> +//# DEFINE_DEBUGFS_ATTRIBUTE + debugfs_create_file_unsafe().
> +//
> +// Copyright (C): 2016 Nicolai Stange
> +// Options: --no-includes
> +//
> +
> +virtual context
> +virtual patch
> +virtual org
> +virtual report
> +
> + at dsa@
> +declarer name DEFINE_SIMPLE_ATTRIBUTE;
> +identifier dsa_fops;
> +expression dsa_get, dsa_set, dsa_fmt;
> +position p;
> +@@
> +DEFINE_SIMPLE_ATTRIBUTE at p(dsa_fops, dsa_get, dsa_set, dsa_fmt);
> +
> + at dcf@
> +expression name, mode, parent, data;
> +identifier dsa.dsa_fops;
> +@@
> +debugfs_create_file(name, mode, parent, data, &dsa_fops)
> +
> +
> + at context_dsa depends on context && dcf@
> +declarer name DEFINE_DEBUGFS_ATTRIBUTE;
> +identifier dsa.dsa_fops;
> +expression dsa.dsa_get, dsa.dsa_set, dsa.dsa_fmt;
> +@@
> +* DEFINE_SIMPLE_ATTRIBUTE(dsa_fops, dsa_get, dsa_set, dsa_fmt);
> +
> +
> + at patch_dcf depends on patch expression@
> +expression name, mode, parent, data;
> +identifier dsa.dsa_fops;
> +@@
> +- debugfs_create_file(name, mode, parent, data, &dsa_fops)
> ++ debugfs_create_file_unsafe(name, mode, parent, data, &dsa_fops)
> +
> + at patch_dsa depends on patch_dcf && patch@
> +identifier dsa.dsa_fops;
> +expression dsa.dsa_get, dsa.dsa_set, dsa.dsa_fmt;
> +@@
> +- DEFINE_SIMPLE_ATTRIBUTE(dsa_fops, dsa_get, dsa_set, dsa_fmt);
> ++ DEFINE_DEBUGFS_ATTRIBUTE(dsa_fops, dsa_get, dsa_set, dsa_fmt);
> +
> +
> + at script:python depends on org && dcf@
> +fops << dsa.dsa_fops;
> +p << dsa.p;
> +@@
> +msg="%s should be defined with DEFINE_DEBUGFS_ATTRIBUTE" % (fops)
> +coccilib.org.print_todo(p[0], msg)
> +
> + at script:python depends on report && dcf@
> +fops << dsa.dsa_fops;
> +p << dsa.p;
> +@@
> +msg="WARNING: %s should be defined with DEFINE_DEBUGFS_ATTRIBUTE" % (fops)
> +coccilib.report.print_report(p[0], msg)
> --
> 2.7.4
>
>

[PATCH v6 5/8] debugfs: unproxify integer attribute files

[Nicolai Stange]

Currently, the struct file_operations associated with the integer attribute
style files created through the debugfs_create_*() helpers are not file
lifetime aware as they are defined by means of DEFINE_SIMPLE_ATTRIBUTE().

Thus, a lifetime managing proxy is created around the original fops each
time such a file is opened which is an unnecessary waste of resources.

Migrate all usages of DEFINE_SIMPLE_ATTRIBUTE() within debugfs itself
to DEFINE_DEBUGFS_ATTRIBUTE() in order to implement file lifetime managing
within the struct file_operations thus defined.

Introduce the debugfs_create_mode_unsafe() helper, analogous to
debugfs_create_mode(), but distinct in that it creates the files in
non-proxying operation mode through debugfs_create_file_unsafe().

Feed all struct file_operations migrated to DEFINE_DEBUGFS_ATTRIBUTE()
into debugfs_create_mode_unsafe() instead of former debugfs_create_mode().

Signed-off-by: Nicolai Stange <nicstange@gmail.com>
---
 fs/debugfs/file.c | 123 +++++++++++++++++++++++++++++++++---------------------
 1 file changed, 75 insertions(+), 48 deletions(-)

diff --git a/fs/debugfs/file.c b/fs/debugfs/file.c
index 8ef56d9..4b3967e 100644
--- a/fs/debugfs/file.c
+++ b/fs/debugfs/file.c
@@ -328,6 +328,24 @@ static struct dentry *debugfs_create_mode(const char *name, umode_t mode,
 	return debugfs_create_file(name, mode, parent, value, fops);
 }
 
+static struct dentry *debugfs_create_mode_unsafe(const char *name, umode_t mode,
+					struct dentry *parent, void *value,
+					const struct file_operations *fops,
+					const struct file_operations *fops_ro,
+					const struct file_operations *fops_wo)
+{
+	/* if there are no write bits set, make read only */
+	if (!(mode & S_IWUGO))
+		return debugfs_create_file_unsafe(name, mode, parent, value,
+						fops_ro);
+	/* if there are no read bits set, make write only */
+	if (!(mode & S_IRUGO))
+		return debugfs_create_file_unsafe(name, mode, parent, value,
+						fops_wo);
+
+	return debugfs_create_file_unsafe(name, mode, parent, value, fops);
+}
+
 static int debugfs_u8_set(void *data, u64 val)
 {
 	*(u8 *)data = val;
@@ -338,9 +356,9 @@ static int debugfs_u8_get(void *data, u64 *val)
 	*val = *(u8 *)data;
 	return 0;
 }
-DEFINE_SIMPLE_ATTRIBUTE(fops_u8, debugfs_u8_get, debugfs_u8_set, "%llu\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_u8_ro, debugfs_u8_get, NULL, "%llu\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_u8_wo, NULL, debugfs_u8_set, "%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_u8, debugfs_u8_get, debugfs_u8_set, "%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_u8_ro, debugfs_u8_get, NULL, "%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_u8_wo, NULL, debugfs_u8_set, "%llu\n");
 
 /**
  * debugfs_create_u8 - create a debugfs file that is used to read and write an unsigned 8-bit value
@@ -369,7 +387,7 @@ DEFINE_SIMPLE_ATTRIBUTE(fops_u8_wo, NULL, debugfs_u8_set, "%llu\n");
 struct dentry *debugfs_create_u8(const char *name, umode_t mode,
 				 struct dentry *parent, u8 *value)
 {
-	return debugfs_create_mode(name, mode, parent, value, &fops_u8,
+	return debugfs_create_mode_unsafe(name, mode, parent, value, &fops_u8,
 				   &fops_u8_ro, &fops_u8_wo);
 }
 EXPORT_SYMBOL_GPL(debugfs_create_u8);
@@ -384,9 +402,9 @@ static int debugfs_u16_get(void *data, u64 *val)
 	*val = *(u16 *)data;
 	return 0;
 }
-DEFINE_SIMPLE_ATTRIBUTE(fops_u16, debugfs_u16_get, debugfs_u16_set, "%llu\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_u16_ro, debugfs_u16_get, NULL, "%llu\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_u16_wo, NULL, debugfs_u16_set, "%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_u16, debugfs_u16_get, debugfs_u16_set, "%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_u16_ro, debugfs_u16_get, NULL, "%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_u16_wo, NULL, debugfs_u16_set, "%llu\n");
 
 /**
  * debugfs_create_u16 - create a debugfs file that is used to read and write an unsigned 16-bit value
@@ -415,7 +433,7 @@ DEFINE_SIMPLE_ATTRIBUTE(fops_u16_wo, NULL, debugfs_u16_set, "%llu\n");
 struct dentry *debugfs_create_u16(const char *name, umode_t mode,
 				  struct dentry *parent, u16 *value)
 {
-	return debugfs_create_mode(name, mode, parent, value, &fops_u16,
+	return debugfs_create_mode_unsafe(name, mode, parent, value, &fops_u16,
 				   &fops_u16_ro, &fops_u16_wo);
 }
 EXPORT_SYMBOL_GPL(debugfs_create_u16);
@@ -430,9 +448,9 @@ static int debugfs_u32_get(void *data, u64 *val)
 	*val = *(u32 *)data;
 	return 0;
 }
-DEFINE_SIMPLE_ATTRIBUTE(fops_u32, debugfs_u32_get, debugfs_u32_set, "%llu\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_u32_ro, debugfs_u32_get, NULL, "%llu\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_u32_wo, NULL, debugfs_u32_set, "%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_u32, debugfs_u32_get, debugfs_u32_set, "%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_u32_ro, debugfs_u32_get, NULL, "%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_u32_wo, NULL, debugfs_u32_set, "%llu\n");
 
 /**
  * debugfs_create_u32 - create a debugfs file that is used to read and write an unsigned 32-bit value
@@ -461,7 +479,7 @@ DEFINE_SIMPLE_ATTRIBUTE(fops_u32_wo, NULL, debugfs_u32_set, "%llu\n");
 struct dentry *debugfs_create_u32(const char *name, umode_t mode,
 				 struct dentry *parent, u32 *value)
 {
-	return debugfs_create_mode(name, mode, parent, value, &fops_u32,
+	return debugfs_create_mode_unsafe(name, mode, parent, value, &fops_u32,
 				   &fops_u32_ro, &fops_u32_wo);
 }
 EXPORT_SYMBOL_GPL(debugfs_create_u32);
@@ -477,9 +495,9 @@ static int debugfs_u64_get(void *data, u64 *val)
 	*val = *(u64 *)data;
 	return 0;
 }
-DEFINE_SIMPLE_ATTRIBUTE(fops_u64, debugfs_u64_get, debugfs_u64_set, "%llu\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_u64_ro, debugfs_u64_get, NULL, "%llu\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_u64_wo, NULL, debugfs_u64_set, "%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_u64, debugfs_u64_get, debugfs_u64_set, "%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_u64_ro, debugfs_u64_get, NULL, "%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_u64_wo, NULL, debugfs_u64_set, "%llu\n");
 
 /**
  * debugfs_create_u64 - create a debugfs file that is used to read and write an unsigned 64-bit value
@@ -508,7 +526,7 @@ DEFINE_SIMPLE_ATTRIBUTE(fops_u64_wo, NULL, debugfs_u64_set, "%llu\n");
 struct dentry *debugfs_create_u64(const char *name, umode_t mode,
 				 struct dentry *parent, u64 *value)
 {
-	return debugfs_create_mode(name, mode, parent, value, &fops_u64,
+	return debugfs_create_mode_unsafe(name, mode, parent, value, &fops_u64,
 				   &fops_u64_ro, &fops_u64_wo);
 }
 EXPORT_SYMBOL_GPL(debugfs_create_u64);
@@ -524,9 +542,10 @@ static int debugfs_ulong_get(void *data, u64 *val)
 	*val = *(unsigned long *)data;
 	return 0;
 }
-DEFINE_SIMPLE_ATTRIBUTE(fops_ulong, debugfs_ulong_get, debugfs_ulong_set, "%llu\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_ulong_ro, debugfs_ulong_get, NULL, "%llu\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_ulong_wo, NULL, debugfs_ulong_set, "%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_ulong, debugfs_ulong_get, debugfs_ulong_set,
+			"%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_ulong_ro, debugfs_ulong_get, NULL, "%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_ulong_wo, NULL, debugfs_ulong_set, "%llu\n");
 
 /**
  * debugfs_create_ulong - create a debugfs file that is used to read and write
@@ -556,26 +575,30 @@ DEFINE_SIMPLE_ATTRIBUTE(fops_ulong_wo, NULL, debugfs_ulong_set, "%llu\n");
 struct dentry *debugfs_create_ulong(const char *name, umode_t mode,
 				    struct dentry *parent, unsigned long *value)
 {
-	return debugfs_create_mode(name, mode, parent, value, &fops_ulong,
-				   &fops_ulong_ro, &fops_ulong_wo);
+	return debugfs_create_mode_unsafe(name, mode, parent, value,
+					&fops_ulong, &fops_ulong_ro,
+					&fops_ulong_wo);
 }
 EXPORT_SYMBOL_GPL(debugfs_create_ulong);
 
-DEFINE_SIMPLE_ATTRIBUTE(fops_x8, debugfs_u8_get, debugfs_u8_set, "0x%02llx\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_x8_ro, debugfs_u8_get, NULL, "0x%02llx\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_x8_wo, NULL, debugfs_u8_set, "0x%02llx\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_x8, debugfs_u8_get, debugfs_u8_set, "0x%02llx\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_x8_ro, debugfs_u8_get, NULL, "0x%02llx\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_x8_wo, NULL, debugfs_u8_set, "0x%02llx\n");
 
-DEFINE_SIMPLE_ATTRIBUTE(fops_x16, debugfs_u16_get, debugfs_u16_set, "0x%04llx\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_x16_ro, debugfs_u16_get, NULL, "0x%04llx\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_x16_wo, NULL, debugfs_u16_set, "0x%04llx\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_x16, debugfs_u16_get, debugfs_u16_set,
+			"0x%04llx\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_x16_ro, debugfs_u16_get, NULL, "0x%04llx\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_x16_wo, NULL, debugfs_u16_set, "0x%04llx\n");
 
-DEFINE_SIMPLE_ATTRIBUTE(fops_x32, debugfs_u32_get, debugfs_u32_set, "0x%08llx\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_x32_ro, debugfs_u32_get, NULL, "0x%08llx\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_x32_wo, NULL, debugfs_u32_set, "0x%08llx\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_x32, debugfs_u32_get, debugfs_u32_set,
+			"0x%08llx\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_x32_ro, debugfs_u32_get, NULL, "0x%08llx\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_x32_wo, NULL, debugfs_u32_set, "0x%08llx\n");
 
-DEFINE_SIMPLE_ATTRIBUTE(fops_x64, debugfs_u64_get, debugfs_u64_set, "0x%016llx\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_x64_ro, debugfs_u64_get, NULL, "0x%016llx\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_x64_wo, NULL, debugfs_u64_set, "0x%016llx\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_x64, debugfs_u64_get, debugfs_u64_set,
+			"0x%016llx\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_x64_ro, debugfs_u64_get, NULL, "0x%016llx\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_x64_wo, NULL, debugfs_u64_set, "0x%016llx\n");
 
 /*
  * debugfs_create_x{8,16,32,64} - create a debugfs file that is used to read and write an unsigned {8,16,32,64}-bit value
@@ -598,7 +621,7 @@ DEFINE_SIMPLE_ATTRIBUTE(fops_x64_wo, NULL, debugfs_u64_set, "0x%016llx\n");
 struct dentry *debugfs_create_x8(const char *name, umode_t mode,
 				 struct dentry *parent, u8 *value)
 {
-	return debugfs_create_mode(name, mode, parent, value, &fops_x8,
+	return debugfs_create_mode_unsafe(name, mode, parent, value, &fops_x8,
 				   &fops_x8_ro, &fops_x8_wo);
 }
 EXPORT_SYMBOL_GPL(debugfs_create_x8);
@@ -616,7 +639,7 @@ EXPORT_SYMBOL_GPL(debugfs_create_x8);
 struct dentry *debugfs_create_x16(const char *name, umode_t mode,
 				 struct dentry *parent, u16 *value)
 {
-	return debugfs_create_mode(name, mode, parent, value, &fops_x16,
+	return debugfs_create_mode_unsafe(name, mode, parent, value, &fops_x16,
 				   &fops_x16_ro, &fops_x16_wo);
 }
 EXPORT_SYMBOL_GPL(debugfs_create_x16);
@@ -634,7 +657,7 @@ EXPORT_SYMBOL_GPL(debugfs_create_x16);
 struct dentry *debugfs_create_x32(const char *name, umode_t mode,
 				 struct dentry *parent, u32 *value)
 {
-	return debugfs_create_mode(name, mode, parent, value, &fops_x32,
+	return debugfs_create_mode_unsafe(name, mode, parent, value, &fops_x32,
 				   &fops_x32_ro, &fops_x32_wo);
 }
 EXPORT_SYMBOL_GPL(debugfs_create_x32);
@@ -652,7 +675,7 @@ EXPORT_SYMBOL_GPL(debugfs_create_x32);
 struct dentry *debugfs_create_x64(const char *name, umode_t mode,
 				 struct dentry *parent, u64 *value)
 {
-	return debugfs_create_mode(name, mode, parent, value, &fops_x64,
+	return debugfs_create_mode_unsafe(name, mode, parent, value, &fops_x64,
 				   &fops_x64_ro, &fops_x64_wo);
 }
 EXPORT_SYMBOL_GPL(debugfs_create_x64);
@@ -668,10 +691,10 @@ static int debugfs_size_t_get(void *data, u64 *val)
 	*val = *(size_t *)data;
 	return 0;
 }
-DEFINE_SIMPLE_ATTRIBUTE(fops_size_t, debugfs_size_t_get, debugfs_size_t_set,
-			"%llu\n");	/* %llu and %zu are more or less the same */
-DEFINE_SIMPLE_ATTRIBUTE(fops_size_t_ro, debugfs_size_t_get, NULL, "%llu\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_size_t_wo, NULL, debugfs_size_t_set, "%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_size_t, debugfs_size_t_get, debugfs_size_t_set,
+			"%llu\n"); /* %llu and %zu are more or less the same */
+DEFINE_DEBUGFS_ATTRIBUTE(fops_size_t_ro, debugfs_size_t_get, NULL, "%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_size_t_wo, NULL, debugfs_size_t_set, "%llu\n");
 
 /**
  * debugfs_create_size_t - create a debugfs file that is used to read and write an size_t value
@@ -686,8 +709,9 @@ DEFINE_SIMPLE_ATTRIBUTE(fops_size_t_wo, NULL, debugfs_size_t_set, "%llu\n");
 struct dentry *debugfs_create_size_t(const char *name, umode_t mode,
 				     struct dentry *parent, size_t *value)
 {
-	return debugfs_create_mode(name, mode, parent, value, &fops_size_t,
-				   &fops_size_t_ro, &fops_size_t_wo);
+	return debugfs_create_mode_unsafe(name, mode, parent, value,
+					&fops_size_t, &fops_size_t_ro,
+					&fops_size_t_wo);
 }
 EXPORT_SYMBOL_GPL(debugfs_create_size_t);
 
@@ -701,10 +725,12 @@ static int debugfs_atomic_t_get(void *data, u64 *val)
 	*val = atomic_read((atomic_t *)data);
 	return 0;
 }
-DEFINE_SIMPLE_ATTRIBUTE(fops_atomic_t, debugfs_atomic_t_get,
+DEFINE_DEBUGFS_ATTRIBUTE(fops_atomic_t, debugfs_atomic_t_get,
 			debugfs_atomic_t_set, "%lld\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_atomic_t_ro, debugfs_atomic_t_get, NULL, "%lld\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_atomic_t_wo, NULL, debugfs_atomic_t_set, "%lld\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_atomic_t_ro, debugfs_atomic_t_get, NULL,
+			"%lld\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_atomic_t_wo, NULL, debugfs_atomic_t_set,
+			"%lld\n");
 
 /**
  * debugfs_create_atomic_t - create a debugfs file that is used to read and
@@ -720,8 +746,9 @@ DEFINE_SIMPLE_ATTRIBUTE(fops_atomic_t_wo, NULL, debugfs_atomic_t_set, "%lld\n");
 struct dentry *debugfs_create_atomic_t(const char *name, umode_t mode,
 				 struct dentry *parent, atomic_t *value)
 {
-	return debugfs_create_mode(name, mode, parent, value, &fops_atomic_t,
-				   &fops_atomic_t_ro, &fops_atomic_t_wo);
+	return debugfs_create_mode_unsafe(name, mode, parent, value,
+					&fops_atomic_t, &fops_atomic_t_ro,
+					&fops_atomic_t_wo);
 }
 EXPORT_SYMBOL_GPL(debugfs_create_atomic_t);
 
-- 
2.7.4

[Cocci] [PATCH v6 5/8] debugfs: unproxify integer attribute files

[Nicolai Stange]

Currently, the struct file_operations associated with the integer attribute
style files created through the debugfs_create_*() helpers are not file
lifetime aware as they are defined by means of DEFINE_SIMPLE_ATTRIBUTE().

Thus, a lifetime managing proxy is created around the original fops each
time such a file is opened which is an unnecessary waste of resources.

Migrate all usages of DEFINE_SIMPLE_ATTRIBUTE() within debugfs itself
to DEFINE_DEBUGFS_ATTRIBUTE() in order to implement file lifetime managing
within the struct file_operations thus defined.

Introduce the debugfs_create_mode_unsafe() helper, analogous to
debugfs_create_mode(), but distinct in that it creates the files in
non-proxying operation mode through debugfs_create_file_unsafe().

Feed all struct file_operations migrated to DEFINE_DEBUGFS_ATTRIBUTE()
into debugfs_create_mode_unsafe() instead of former debugfs_create_mode().

Signed-off-by: Nicolai Stange <nicstange@gmail.com>
---
 fs/debugfs/file.c | 123 +++++++++++++++++++++++++++++++++---------------------
 1 file changed, 75 insertions(+), 48 deletions(-)

diff --git a/fs/debugfs/file.c b/fs/debugfs/file.c
index 8ef56d9..4b3967e 100644
--- a/fs/debugfs/file.c
+++ b/fs/debugfs/file.c
@@ -328,6 +328,24 @@ static struct dentry *debugfs_create_mode(const char *name, umode_t mode,
 	return debugfs_create_file(name, mode, parent, value, fops);
 }
 
+static struct dentry *debugfs_create_mode_unsafe(const char *name, umode_t mode,
+					struct dentry *parent, void *value,
+					const struct file_operations *fops,
+					const struct file_operations *fops_ro,
+					const struct file_operations *fops_wo)
+{
+	/* if there are no write bits set, make read only */
+	if (!(mode & S_IWUGO))
+		return debugfs_create_file_unsafe(name, mode, parent, value,
+						fops_ro);
+	/* if there are no read bits set, make write only */
+	if (!(mode & S_IRUGO))
+		return debugfs_create_file_unsafe(name, mode, parent, value,
+						fops_wo);
+
+	return debugfs_create_file_unsafe(name, mode, parent, value, fops);
+}
+
 static int debugfs_u8_set(void *data, u64 val)
 {
 	*(u8 *)data = val;
@@ -338,9 +356,9 @@ static int debugfs_u8_get(void *data, u64 *val)
 	*val = *(u8 *)data;
 	return 0;
 }
-DEFINE_SIMPLE_ATTRIBUTE(fops_u8, debugfs_u8_get, debugfs_u8_set, "%llu\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_u8_ro, debugfs_u8_get, NULL, "%llu\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_u8_wo, NULL, debugfs_u8_set, "%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_u8, debugfs_u8_get, debugfs_u8_set, "%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_u8_ro, debugfs_u8_get, NULL, "%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_u8_wo, NULL, debugfs_u8_set, "%llu\n");
 
 /**
  * debugfs_create_u8 - create a debugfs file that is used to read and write an unsigned 8-bit value
@@ -369,7 +387,7 @@ DEFINE_SIMPLE_ATTRIBUTE(fops_u8_wo, NULL, debugfs_u8_set, "%llu\n");
 struct dentry *debugfs_create_u8(const char *name, umode_t mode,
 				 struct dentry *parent, u8 *value)
 {
-	return debugfs_create_mode(name, mode, parent, value, &fops_u8,
+	return debugfs_create_mode_unsafe(name, mode, parent, value, &fops_u8,
 				   &fops_u8_ro, &fops_u8_wo);
 }
 EXPORT_SYMBOL_GPL(debugfs_create_u8);
@@ -384,9 +402,9 @@ static int debugfs_u16_get(void *data, u64 *val)
 	*val = *(u16 *)data;
 	return 0;
 }
-DEFINE_SIMPLE_ATTRIBUTE(fops_u16, debugfs_u16_get, debugfs_u16_set, "%llu\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_u16_ro, debugfs_u16_get, NULL, "%llu\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_u16_wo, NULL, debugfs_u16_set, "%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_u16, debugfs_u16_get, debugfs_u16_set, "%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_u16_ro, debugfs_u16_get, NULL, "%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_u16_wo, NULL, debugfs_u16_set, "%llu\n");
 
 /**
  * debugfs_create_u16 - create a debugfs file that is used to read and write an unsigned 16-bit value
@@ -415,7 +433,7 @@ DEFINE_SIMPLE_ATTRIBUTE(fops_u16_wo, NULL, debugfs_u16_set, "%llu\n");
 struct dentry *debugfs_create_u16(const char *name, umode_t mode,
 				  struct dentry *parent, u16 *value)
 {
-	return debugfs_create_mode(name, mode, parent, value, &fops_u16,
+	return debugfs_create_mode_unsafe(name, mode, parent, value, &fops_u16,
 				   &fops_u16_ro, &fops_u16_wo);
 }
 EXPORT_SYMBOL_GPL(debugfs_create_u16);
@@ -430,9 +448,9 @@ static int debugfs_u32_get(void *data, u64 *val)
 	*val = *(u32 *)data;
 	return 0;
 }
-DEFINE_SIMPLE_ATTRIBUTE(fops_u32, debugfs_u32_get, debugfs_u32_set, "%llu\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_u32_ro, debugfs_u32_get, NULL, "%llu\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_u32_wo, NULL, debugfs_u32_set, "%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_u32, debugfs_u32_get, debugfs_u32_set, "%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_u32_ro, debugfs_u32_get, NULL, "%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_u32_wo, NULL, debugfs_u32_set, "%llu\n");
 
 /**
  * debugfs_create_u32 - create a debugfs file that is used to read and write an unsigned 32-bit value
@@ -461,7 +479,7 @@ DEFINE_SIMPLE_ATTRIBUTE(fops_u32_wo, NULL, debugfs_u32_set, "%llu\n");
 struct dentry *debugfs_create_u32(const char *name, umode_t mode,
 				 struct dentry *parent, u32 *value)
 {
-	return debugfs_create_mode(name, mode, parent, value, &fops_u32,
+	return debugfs_create_mode_unsafe(name, mode, parent, value, &fops_u32,
 				   &fops_u32_ro, &fops_u32_wo);
 }
 EXPORT_SYMBOL_GPL(debugfs_create_u32);
@@ -477,9 +495,9 @@ static int debugfs_u64_get(void *data, u64 *val)
 	*val = *(u64 *)data;
 	return 0;
 }
-DEFINE_SIMPLE_ATTRIBUTE(fops_u64, debugfs_u64_get, debugfs_u64_set, "%llu\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_u64_ro, debugfs_u64_get, NULL, "%llu\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_u64_wo, NULL, debugfs_u64_set, "%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_u64, debugfs_u64_get, debugfs_u64_set, "%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_u64_ro, debugfs_u64_get, NULL, "%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_u64_wo, NULL, debugfs_u64_set, "%llu\n");
 
 /**
  * debugfs_create_u64 - create a debugfs file that is used to read and write an unsigned 64-bit value
@@ -508,7 +526,7 @@ DEFINE_SIMPLE_ATTRIBUTE(fops_u64_wo, NULL, debugfs_u64_set, "%llu\n");
 struct dentry *debugfs_create_u64(const char *name, umode_t mode,
 				 struct dentry *parent, u64 *value)
 {
-	return debugfs_create_mode(name, mode, parent, value, &fops_u64,
+	return debugfs_create_mode_unsafe(name, mode, parent, value, &fops_u64,
 				   &fops_u64_ro, &fops_u64_wo);
 }
 EXPORT_SYMBOL_GPL(debugfs_create_u64);
@@ -524,9 +542,10 @@ static int debugfs_ulong_get(void *data, u64 *val)
 	*val = *(unsigned long *)data;
 	return 0;
 }
-DEFINE_SIMPLE_ATTRIBUTE(fops_ulong, debugfs_ulong_get, debugfs_ulong_set, "%llu\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_ulong_ro, debugfs_ulong_get, NULL, "%llu\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_ulong_wo, NULL, debugfs_ulong_set, "%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_ulong, debugfs_ulong_get, debugfs_ulong_set,
+			"%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_ulong_ro, debugfs_ulong_get, NULL, "%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_ulong_wo, NULL, debugfs_ulong_set, "%llu\n");
 
 /**
  * debugfs_create_ulong - create a debugfs file that is used to read and write
@@ -556,26 +575,30 @@ DEFINE_SIMPLE_ATTRIBUTE(fops_ulong_wo, NULL, debugfs_ulong_set, "%llu\n");
 struct dentry *debugfs_create_ulong(const char *name, umode_t mode,
 				    struct dentry *parent, unsigned long *value)
 {
-	return debugfs_create_mode(name, mode, parent, value, &fops_ulong,
-				   &fops_ulong_ro, &fops_ulong_wo);
+	return debugfs_create_mode_unsafe(name, mode, parent, value,
+					&fops_ulong, &fops_ulong_ro,
+					&fops_ulong_wo);
 }
 EXPORT_SYMBOL_GPL(debugfs_create_ulong);
 
-DEFINE_SIMPLE_ATTRIBUTE(fops_x8, debugfs_u8_get, debugfs_u8_set, "0x%02llx\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_x8_ro, debugfs_u8_get, NULL, "0x%02llx\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_x8_wo, NULL, debugfs_u8_set, "0x%02llx\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_x8, debugfs_u8_get, debugfs_u8_set, "0x%02llx\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_x8_ro, debugfs_u8_get, NULL, "0x%02llx\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_x8_wo, NULL, debugfs_u8_set, "0x%02llx\n");
 
-DEFINE_SIMPLE_ATTRIBUTE(fops_x16, debugfs_u16_get, debugfs_u16_set, "0x%04llx\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_x16_ro, debugfs_u16_get, NULL, "0x%04llx\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_x16_wo, NULL, debugfs_u16_set, "0x%04llx\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_x16, debugfs_u16_get, debugfs_u16_set,
+			"0x%04llx\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_x16_ro, debugfs_u16_get, NULL, "0x%04llx\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_x16_wo, NULL, debugfs_u16_set, "0x%04llx\n");
 
-DEFINE_SIMPLE_ATTRIBUTE(fops_x32, debugfs_u32_get, debugfs_u32_set, "0x%08llx\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_x32_ro, debugfs_u32_get, NULL, "0x%08llx\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_x32_wo, NULL, debugfs_u32_set, "0x%08llx\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_x32, debugfs_u32_get, debugfs_u32_set,
+			"0x%08llx\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_x32_ro, debugfs_u32_get, NULL, "0x%08llx\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_x32_wo, NULL, debugfs_u32_set, "0x%08llx\n");
 
-DEFINE_SIMPLE_ATTRIBUTE(fops_x64, debugfs_u64_get, debugfs_u64_set, "0x%016llx\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_x64_ro, debugfs_u64_get, NULL, "0x%016llx\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_x64_wo, NULL, debugfs_u64_set, "0x%016llx\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_x64, debugfs_u64_get, debugfs_u64_set,
+			"0x%016llx\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_x64_ro, debugfs_u64_get, NULL, "0x%016llx\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_x64_wo, NULL, debugfs_u64_set, "0x%016llx\n");
 
 /*
  * debugfs_create_x{8,16,32,64} - create a debugfs file that is used to read and write an unsigned {8,16,32,64}-bit value
@@ -598,7 +621,7 @@ DEFINE_SIMPLE_ATTRIBUTE(fops_x64_wo, NULL, debugfs_u64_set, "0x%016llx\n");
 struct dentry *debugfs_create_x8(const char *name, umode_t mode,
 				 struct dentry *parent, u8 *value)
 {
-	return debugfs_create_mode(name, mode, parent, value, &fops_x8,
+	return debugfs_create_mode_unsafe(name, mode, parent, value, &fops_x8,
 				   &fops_x8_ro, &fops_x8_wo);
 }
 EXPORT_SYMBOL_GPL(debugfs_create_x8);
@@ -616,7 +639,7 @@ EXPORT_SYMBOL_GPL(debugfs_create_x8);
 struct dentry *debugfs_create_x16(const char *name, umode_t mode,
 				 struct dentry *parent, u16 *value)
 {
-	return debugfs_create_mode(name, mode, parent, value, &fops_x16,
+	return debugfs_create_mode_unsafe(name, mode, parent, value, &fops_x16,
 				   &fops_x16_ro, &fops_x16_wo);
 }
 EXPORT_SYMBOL_GPL(debugfs_create_x16);
@@ -634,7 +657,7 @@ EXPORT_SYMBOL_GPL(debugfs_create_x16);
 struct dentry *debugfs_create_x32(const char *name, umode_t mode,
 				 struct dentry *parent, u32 *value)
 {
-	return debugfs_create_mode(name, mode, parent, value, &fops_x32,
+	return debugfs_create_mode_unsafe(name, mode, parent, value, &fops_x32,
 				   &fops_x32_ro, &fops_x32_wo);
 }
 EXPORT_SYMBOL_GPL(debugfs_create_x32);
@@ -652,7 +675,7 @@ EXPORT_SYMBOL_GPL(debugfs_create_x32);
 struct dentry *debugfs_create_x64(const char *name, umode_t mode,
 				 struct dentry *parent, u64 *value)
 {
-	return debugfs_create_mode(name, mode, parent, value, &fops_x64,
+	return debugfs_create_mode_unsafe(name, mode, parent, value, &fops_x64,
 				   &fops_x64_ro, &fops_x64_wo);
 }
 EXPORT_SYMBOL_GPL(debugfs_create_x64);
@@ -668,10 +691,10 @@ static int debugfs_size_t_get(void *data, u64 *val)
 	*val = *(size_t *)data;
 	return 0;
 }
-DEFINE_SIMPLE_ATTRIBUTE(fops_size_t, debugfs_size_t_get, debugfs_size_t_set,
-			"%llu\n");	/* %llu and %zu are more or less the same */
-DEFINE_SIMPLE_ATTRIBUTE(fops_size_t_ro, debugfs_size_t_get, NULL, "%llu\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_size_t_wo, NULL, debugfs_size_t_set, "%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_size_t, debugfs_size_t_get, debugfs_size_t_set,
+			"%llu\n"); /* %llu and %zu are more or less the same */
+DEFINE_DEBUGFS_ATTRIBUTE(fops_size_t_ro, debugfs_size_t_get, NULL, "%llu\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_size_t_wo, NULL, debugfs_size_t_set, "%llu\n");
 
 /**
  * debugfs_create_size_t - create a debugfs file that is used to read and write an size_t value
@@ -686,8 +709,9 @@ DEFINE_SIMPLE_ATTRIBUTE(fops_size_t_wo, NULL, debugfs_size_t_set, "%llu\n");
 struct dentry *debugfs_create_size_t(const char *name, umode_t mode,
 				     struct dentry *parent, size_t *value)
 {
-	return debugfs_create_mode(name, mode, parent, value, &fops_size_t,
-				   &fops_size_t_ro, &fops_size_t_wo);
+	return debugfs_create_mode_unsafe(name, mode, parent, value,
+					&fops_size_t, &fops_size_t_ro,
+					&fops_size_t_wo);
 }
 EXPORT_SYMBOL_GPL(debugfs_create_size_t);
 
@@ -701,10 +725,12 @@ static int debugfs_atomic_t_get(void *data, u64 *val)
 	*val = atomic_read((atomic_t *)data);
 	return 0;
 }
-DEFINE_SIMPLE_ATTRIBUTE(fops_atomic_t, debugfs_atomic_t_get,
+DEFINE_DEBUGFS_ATTRIBUTE(fops_atomic_t, debugfs_atomic_t_get,
 			debugfs_atomic_t_set, "%lld\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_atomic_t_ro, debugfs_atomic_t_get, NULL, "%lld\n");
-DEFINE_SIMPLE_ATTRIBUTE(fops_atomic_t_wo, NULL, debugfs_atomic_t_set, "%lld\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_atomic_t_ro, debugfs_atomic_t_get, NULL,
+			"%lld\n");
+DEFINE_DEBUGFS_ATTRIBUTE(fops_atomic_t_wo, NULL, debugfs_atomic_t_set,
+			"%lld\n");
 
 /**
  * debugfs_create_atomic_t - create a debugfs file that is used to read and
@@ -720,8 +746,9 @@ DEFINE_SIMPLE_ATTRIBUTE(fops_atomic_t_wo, NULL, debugfs_atomic_t_set, "%lld\n");
 struct dentry *debugfs_create_atomic_t(const char *name, umode_t mode,
 				 struct dentry *parent, atomic_t *value)
 {
-	return debugfs_create_mode(name, mode, parent, value, &fops_atomic_t,
-				   &fops_atomic_t_ro, &fops_atomic_t_wo);
+	return debugfs_create_mode_unsafe(name, mode, parent, value,
+					&fops_atomic_t, &fops_atomic_t_ro,
+					&fops_atomic_t_wo);
 }
 EXPORT_SYMBOL_GPL(debugfs_create_atomic_t);
 
-- 
2.7.4

[PATCH v6 6/8] debugfs: unproxify files created through debugfs_create_bool()

[Nicolai Stange]

Currently, the struct file_operations fops_bool associated with files
created through the debugfs_create_bool() helpers are not file
lifetime aware.

Thus, a lifetime managing proxy is created around fops_bool each time such
a file is opened which is an unnecessary waste of resources.

Implement file lifetime management for the fops_bool file_operations.
Namely, make debugfs_read_file_bool() and debugfs_write_file_bool() safe
against file removals by means of debugfs_use_file_start() and
debugfs_use_file_finish().

Make debugfs_create_bool() create its files in non-proxying operation mode
through debugfs_create_mode_unsafe().

Finally, purge debugfs_create_mode() as debugfs_create_bool() had been its
last user.

Signed-off-by: Nicolai Stange <nicstange@gmail.com>
---
 fs/debugfs/file.c | 41 ++++++++++++++++++++---------------------
 1 file changed, 20 insertions(+), 21 deletions(-)

diff --git a/fs/debugfs/file.c b/fs/debugfs/file.c
index 4b3967e..8a548be 100644
--- a/fs/debugfs/file.c
+++ b/fs/debugfs/file.c
@@ -312,22 +312,6 @@ ssize_t debugfs_attr_write(struct file *file, const char __user *buf,
 }
 EXPORT_SYMBOL_GPL(debugfs_attr_write);
 
-static struct dentry *debugfs_create_mode(const char *name, umode_t mode,
-					  struct dentry *parent, void *value,
-				          const struct file_operations *fops,
-				          const struct file_operations *fops_ro,
-				          const struct file_operations *fops_wo)
-{
-	/* if there are no write bits set, make read only */
-	if (!(mode & S_IWUGO))
-		return debugfs_create_file(name, mode, parent, value, fops_ro);
-	/* if there are no read bits set, make write only */
-	if (!(mode & S_IRUGO))
-		return debugfs_create_file(name, mode, parent, value, fops_wo);
-
-	return debugfs_create_file(name, mode, parent, value, fops);
-}
-
 static struct dentry *debugfs_create_mode_unsafe(const char *name, umode_t mode,
 					struct dentry *parent, void *value,
 					const struct file_operations *fops,
@@ -756,9 +740,17 @@ ssize_t debugfs_read_file_bool(struct file *file, char __user *user_buf,
 			       size_t count, loff_t *ppos)
 {
 	char buf[3];
-	bool *val = file->private_data;
+	bool val;
+	int r, srcu_idx;
+
+	r = debugfs_use_file_start(F_DENTRY(file), &srcu_idx);
+	if (likely(!r))
+		val = *(bool *)file->private_data;
+	debugfs_use_file_finish(srcu_idx);
+	if (r)
+		return r;
 
-	if (*val)
+	if (val)
 		buf[0] = 'Y';
 	else
 		buf[0] = 'N';
@@ -774,6 +766,7 @@ ssize_t debugfs_write_file_bool(struct file *file, const char __user *user_buf,
 	char buf[32];
 	size_t buf_size;
 	bool bv;
+	int r, srcu_idx;
 	bool *val = file->private_data;
 
 	buf_size = min(count, (sizeof(buf)-1));
@@ -781,8 +774,14 @@ ssize_t debugfs_write_file_bool(struct file *file, const char __user *user_buf,
 		return -EFAULT;
 
 	buf[buf_size] = '\0';
-	if (strtobool(buf, &bv) == 0)
-		*val = bv;
+	if (strtobool(buf, &bv) == 0) {
+		r = debugfs_use_file_start(F_DENTRY(file), &srcu_idx);
+		if (likely(!r))
+			*val = bv;
+		debugfs_use_file_finish(srcu_idx);
+		if (r)
+			return r;
+	}
 
 	return count;
 }
@@ -834,7 +833,7 @@ static const struct file_operations fops_bool_wo = {
 struct dentry *debugfs_create_bool(const char *name, umode_t mode,
 				   struct dentry *parent, bool *value)
 {
-	return debugfs_create_mode(name, mode, parent, value, &fops_bool,
+	return debugfs_create_mode_unsafe(name, mode, parent, value, &fops_bool,
 				   &fops_bool_ro, &fops_bool_wo);
 }
 EXPORT_SYMBOL_GPL(debugfs_create_bool);
-- 
2.7.4

[Cocci] [PATCH v6 6/8] debugfs: unproxify files created through debugfs_create_bool()

[Nicolai Stange]

Currently, the struct file_operations fops_bool associated with files
created through the debugfs_create_bool() helpers are not file
lifetime aware.

Thus, a lifetime managing proxy is created around fops_bool each time such
a file is opened which is an unnecessary waste of resources.

Implement file lifetime management for the fops_bool file_operations.
Namely, make debugfs_read_file_bool() and debugfs_write_file_bool() safe
against file removals by means of debugfs_use_file_start() and
debugfs_use_file_finish().

Make debugfs_create_bool() create its files in non-proxying operation mode
through debugfs_create_mode_unsafe().

Finally, purge debugfs_create_mode() as debugfs_create_bool() had been its
last user.

Signed-off-by: Nicolai Stange <nicstange@gmail.com>
---
 fs/debugfs/file.c | 41 ++++++++++++++++++++---------------------
 1 file changed, 20 insertions(+), 21 deletions(-)

diff --git a/fs/debugfs/file.c b/fs/debugfs/file.c
index 4b3967e..8a548be 100644
--- a/fs/debugfs/file.c
+++ b/fs/debugfs/file.c
@@ -312,22 +312,6 @@ ssize_t debugfs_attr_write(struct file *file, const char __user *buf,
 }
 EXPORT_SYMBOL_GPL(debugfs_attr_write);
 
-static struct dentry *debugfs_create_mode(const char *name, umode_t mode,
-					  struct dentry *parent, void *value,
-				          const struct file_operations *fops,
-				          const struct file_operations *fops_ro,
-				          const struct file_operations *fops_wo)
-{
-	/* if there are no write bits set, make read only */
-	if (!(mode & S_IWUGO))
-		return debugfs_create_file(name, mode, parent, value, fops_ro);
-	/* if there are no read bits set, make write only */
-	if (!(mode & S_IRUGO))
-		return debugfs_create_file(name, mode, parent, value, fops_wo);
-
-	return debugfs_create_file(name, mode, parent, value, fops);
-}
-
 static struct dentry *debugfs_create_mode_unsafe(const char *name, umode_t mode,
 					struct dentry *parent, void *value,
 					const struct file_operations *fops,
@@ -756,9 +740,17 @@ ssize_t debugfs_read_file_bool(struct file *file, char __user *user_buf,
 			       size_t count, loff_t *ppos)
 {
 	char buf[3];
-	bool *val = file->private_data;
+	bool val;
+	int r, srcu_idx;
+
+	r = debugfs_use_file_start(F_DENTRY(file), &srcu_idx);
+	if (likely(!r))
+		val = *(bool *)file->private_data;
+	debugfs_use_file_finish(srcu_idx);
+	if (r)
+		return r;
 
-	if (*val)
+	if (val)
 		buf[0] = 'Y';
 	else
 		buf[0] = 'N';
@@ -774,6 +766,7 @@ ssize_t debugfs_write_file_bool(struct file *file, const char __user *user_buf,
 	char buf[32];
 	size_t buf_size;
 	bool bv;
+	int r, srcu_idx;
 	bool *val = file->private_data;
 
 	buf_size = min(count, (sizeof(buf)-1));
@@ -781,8 +774,14 @@ ssize_t debugfs_write_file_bool(struct file *file, const char __user *user_buf,
 		return -EFAULT;
 
 	buf[buf_size] = '\0';
-	if (strtobool(buf, &bv) == 0)
-		*val = bv;
+	if (strtobool(buf, &bv) == 0) {
+		r = debugfs_use_file_start(F_DENTRY(file), &srcu_idx);
+		if (likely(!r))
+			*val = bv;
+		debugfs_use_file_finish(srcu_idx);
+		if (r)
+			return r;
+	}
 
 	return count;
 }
@@ -834,7 +833,7 @@ static const struct file_operations fops_bool_wo = {
 struct dentry *debugfs_create_bool(const char *name, umode_t mode,
 				   struct dentry *parent, bool *value)
 {
-	return debugfs_create_mode(name, mode, parent, value, &fops_bool,
+	return debugfs_create_mode_unsafe(name, mode, parent, value, &fops_bool,
 				   &fops_bool_ro, &fops_bool_wo);
 }
 EXPORT_SYMBOL_GPL(debugfs_create_bool);
-- 
2.7.4

[PATCH v6 7/8] debugfs: unproxify files created through debugfs_create_blob()

[Nicolai Stange]

Currently, the struct file_operations fops_blob associated with files
created through the debugfs_create_blob() helpers are not file
lifetime aware.

Thus, a lifetime managing proxy is created around fops_blob each time such
a file is opened which is an unnecessary waste of resources.

Implement file lifetime management for the fops_bool file_operations.
Namely, make read_file_blob() safe gainst file removals by means of
debugfs_use_file_start() and debugfs_use_file_finish().

Make debugfs_create_blob() create its files in non-proxying operation mode
by means of debugfs_create_file_unsafe().

Signed-off-by: Nicolai Stange <nicstange@gmail.com>
---
 fs/debugfs/file.c | 13 ++++++++++---
 1 file changed, 10 insertions(+), 3 deletions(-)

diff --git a/fs/debugfs/file.c b/fs/debugfs/file.c
index 8a548be..2e86d66 100644
--- a/fs/debugfs/file.c
+++ b/fs/debugfs/file.c
@@ -842,8 +842,15 @@ static ssize_t read_file_blob(struct file *file, char __user *user_buf,
 			      size_t count, loff_t *ppos)
 {
 	struct debugfs_blob_wrapper *blob = file->private_data;
-	return simple_read_from_buffer(user_buf, count, ppos, blob->data,
-			blob->size);
+	ssize_t r;
+	int srcu_idx;
+
+	r = debugfs_use_file_start(F_DENTRY(file), &srcu_idx);
+	if (likely(!r))
+		r = simple_read_from_buffer(user_buf, count, ppos, blob->data,
+					blob->size);
+	debugfs_use_file_finish(srcu_idx);
+	return r;
 }
 
 static const struct file_operations fops_blob = {
@@ -880,7 +887,7 @@ struct dentry *debugfs_create_blob(const char *name, umode_t mode,
 				   struct dentry *parent,
 				   struct debugfs_blob_wrapper *blob)
 {
-	return debugfs_create_file(name, mode, parent, blob, &fops_blob);
+	return debugfs_create_file_unsafe(name, mode, parent, blob, &fops_blob);
 }
 EXPORT_SYMBOL_GPL(debugfs_create_blob);
 
-- 
2.7.4

[Cocci] [PATCH v6 7/8] debugfs: unproxify files created through debugfs_create_blob()

[Nicolai Stange]

Currently, the struct file_operations fops_blob associated with files
created through the debugfs_create_blob() helpers are not file
lifetime aware.

Thus, a lifetime managing proxy is created around fops_blob each time such
a file is opened which is an unnecessary waste of resources.

Implement file lifetime management for the fops_bool file_operations.
Namely, make read_file_blob() safe gainst file removals by means of
debugfs_use_file_start() and debugfs_use_file_finish().

Make debugfs_create_blob() create its files in non-proxying operation mode
by means of debugfs_create_file_unsafe().

Signed-off-by: Nicolai Stange <nicstange@gmail.com>
---
 fs/debugfs/file.c | 13 ++++++++++---
 1 file changed, 10 insertions(+), 3 deletions(-)

diff --git a/fs/debugfs/file.c b/fs/debugfs/file.c
index 8a548be..2e86d66 100644
--- a/fs/debugfs/file.c
+++ b/fs/debugfs/file.c
@@ -842,8 +842,15 @@ static ssize_t read_file_blob(struct file *file, char __user *user_buf,
 			      size_t count, loff_t *ppos)
 {
 	struct debugfs_blob_wrapper *blob = file->private_data;
-	return simple_read_from_buffer(user_buf, count, ppos, blob->data,
-			blob->size);
+	ssize_t r;
+	int srcu_idx;
+
+	r = debugfs_use_file_start(F_DENTRY(file), &srcu_idx);
+	if (likely(!r))
+		r = simple_read_from_buffer(user_buf, count, ppos, blob->data,
+					blob->size);
+	debugfs_use_file_finish(srcu_idx);
+	return r;
 }
 
 static const struct file_operations fops_blob = {
@@ -880,7 +887,7 @@ struct dentry *debugfs_create_blob(const char *name, umode_t mode,
 				   struct dentry *parent,
 				   struct debugfs_blob_wrapper *blob)
 {
-	return debugfs_create_file(name, mode, parent, blob, &fops_blob);
+	return debugfs_create_file_unsafe(name, mode, parent, blob, &fops_blob);
 }
 EXPORT_SYMBOL_GPL(debugfs_create_blob);
 
-- 
2.7.4

[PATCH v6 8/8] debugfs: unproxify files created through debugfs_create_u32_array()

[Nicolai Stange]

The struct file_operations u32_array_fops associated with files created
through debugfs_create_u32_array() has been lifetime aware already:
everything needed for subsequent operation is copied to a ->f_private
buffer at file opening time in u32_array_open(). Now, ->open() is always
protected against file removal issues by the debugfs core.

There is no need for the debugfs core to wrap the u32_array_fops
with a file lifetime managing proxy.

Make debugfs_create_u32_array() create its files in non-proxying operation
mode by means of debugfs_create_file_unsafe().

Signed-off-by: Nicolai Stange <nicstange@gmail.com>
---
 fs/debugfs/file.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/fs/debugfs/file.c b/fs/debugfs/file.c
index 2e86d66..9c1c9a0 100644
--- a/fs/debugfs/file.c
+++ b/fs/debugfs/file.c
@@ -992,7 +992,8 @@ struct dentry *debugfs_create_u32_array(const char *name, umode_t mode,
 	data->array = array;
 	data->elements = elements;
 
-	return debugfs_create_file(name, mode, parent, data, &u32_array_fops);
+	return debugfs_create_file_unsafe(name, mode, parent, data,
+					&u32_array_fops);
 }
 EXPORT_SYMBOL_GPL(debugfs_create_u32_array);
 
-- 
2.7.4

[Cocci] [PATCH v6 8/8] debugfs: unproxify files created through debugfs_create_u32_array()

[Nicolai Stange]

The struct file_operations u32_array_fops associated with files created
through debugfs_create_u32_array() has been lifetime aware already:
everything needed for subsequent operation is copied to a ->f_private
buffer at file opening time in u32_array_open(). Now, ->open() is always
protected against file removal issues by the debugfs core.

There is no need for the debugfs core to wrap the u32_array_fops
with a file lifetime managing proxy.

Make debugfs_create_u32_array() create its files in non-proxying operation
mode by means of debugfs_create_file_unsafe().

Signed-off-by: Nicolai Stange <nicstange@gmail.com>
---
 fs/debugfs/file.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/fs/debugfs/file.c b/fs/debugfs/file.c
index 2e86d66..9c1c9a0 100644
--- a/fs/debugfs/file.c
+++ b/fs/debugfs/file.c
@@ -992,7 +992,8 @@ struct dentry *debugfs_create_u32_array(const char *name, umode_t mode,
 	data->array = array;
 	data->elements = elements;
 
-	return debugfs_create_file(name, mode, parent, data, &u32_array_fops);
+	return debugfs_create_file_unsafe(name, mode, parent, data,
+					&u32_array_fops);
 }
 EXPORT_SYMBOL_GPL(debugfs_create_u32_array);
 
-- 
2.7.4

Re: [PATCH v6 2/8] debugfs: prevent access to removed files' private data

[Arkady]

Nicolai,

Regarding the debugfs_create_file() which ignores .mmap field in the
struct file_operations. The patch is a part of the kernel 4.8 in the
Ubuntu 16.10 and breaks quite a few online code examples of mmap
drivers.

Is is acceptable to replace the debugfs_create_file() by a macro and
fail the compilation if the .mmap field is not initialized.
Alternatively debugfs_create_file() can return an error if .mmap is
not NULL.

Thank you, Arkady.