See upstream commit:
commit
0ede61d8589cc2d93aa78230d74ac58b5b8d0244
Author: Christian Brauner <brauner@kernel.org>
Date: Fri Sep 29 08:45:59 2023 +0200
file: convert to SLAB_TYPESAFE_BY_RCU
In recent discussions around some performance improvements in the file
handling area we discussed switching the file cache to rely on
SLAB_TYPESAFE_BY_RCU which allows us to get rid of call_rcu() based
freeing for files completely. This is a pretty sensitive change overall
but it might actually be worth doing.
The main downside is the subtlety. The other one is that we should
really wait for Jann's patch to land that enables KASAN to handle
SLAB_TYPESAFE_BY_RCU UAFs. Currently it doesn't but a patch for this
exists.
With SLAB_TYPESAFE_BY_RCU objects may be freed and reused multiple times
which requires a few changes. So it isn't sufficient anymore to just
acquire a reference to the file in question under rcu using
atomic_long_inc_not_zero() since the file might have already been
recycled and someone else might have bumped the reference.
In other words, callers might see reference count bumps from newer
users. For this reason it is necessary to verify that the pointer is the
same before and after the reference count increment. This pattern can be
seen in get_file_rcu() and __files_get_rcu().
In addition, it isn't possible to access or check fields in struct file
without first aqcuiring a reference on it. Not doing that was always
very dodgy and it was only usable for non-pointer data in struct file.
With SLAB_TYPESAFE_BY_RCU it is necessary that callers first acquire a
reference under rcu or they must hold the files_lock of the fdtable.
Failing to do either one of this is a bug.
Thanks to Jann for pointing out that we need to ensure memory ordering
between reallocations and pointer check by ensuring that all subsequent
loads have a dependency on the second load in get_file_rcu() and
providing a fixup that was folded into this patch.
Signed-off-by: Kienan Stewart <kstewart@efficios.com>
Change-Id: Iba3663f19a54820afd31a8eeec24b3b5d4b06589
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
#include <linux/fdtable.h>
#include <linux/sched.h>
-#if (LTTNG_LINUX_VERSION_CODE >= LTTNG_KERNEL_VERSION(5,11,0))
+#if (LTTNG_LINUX_VERSION_CODE >= LTTNG_KERNEL_VERSION(6,7,0))
static inline
-struct file *lttng_lookup_fd_rcu(unsigned int fd)
+struct file *lttng_lookup_fdget_rcu(unsigned int fd)
{
- return lookup_fd_rcu(fd);
+ return lookup_fdget_rcu(fd);
+}
+
+#elif (LTTNG_LINUX_VERSION_CODE >= LTTNG_KERNEL_VERSION(5,11,0))
+static inline
+struct file *lttng_lookup_fdget_rcu(unsigned int fd)
+{
+ struct file* file = lookup_fd_rcu(fd);
+
+ if (unlikely(!file || !get_file_rcu(file)))
+ return NULL;
+ return file;
}
#else
static inline
-struct file *lttng_lookup_fd_rcu(unsigned int fd)
+struct file *lttng_lookup_fdget_rcu(unsigned int fd)
{
- return fcheck(fd);
+ struct file* file = fcheck(fd);
+
+ if (unlikely(!file || !get_file_rcu(file)))
+ return NULL;
+ return file;
}
#endif
*/
#include <wrapper/fdtable.h>
+#include <linux/file.h>
#include <linux/list.h>
#include <linux/module.h>
#include <linux/namei.h>
* Returns the file backing the given fd. Needs to be done inside an RCU
* critical section.
*/
- file = lttng_lookup_fd_rcu(fd);
+ file = lttng_lookup_fdget_rcu(fd);
if (file == NULL) {
printk(KERN_WARNING "LTTng: Cannot access file backing the fd(%d)\n", fd);
inode = NULL;
inode = igrab(file->f_path.dentry->d_inode);
if (inode == NULL)
printk(KERN_WARNING "LTTng: Cannot grab a reference on the inode.\n");
+
error:
rcu_read_unlock();
+ if (file)
+ fput(file);
return inode;
}