linux/fs/pipe.c, branch v3.0.36

pipes: add a "packetized pipe" mode for writing

2012-05-07T15:56:36Z

commit 9883035ae7edef3ec62ad215611cb8e17d6a1a5d upstream. The actual internal pipe implementation is already really about individual packets (called "pipe buffers"), and this simply exposes that as a special packetized mode. When we are in the packetized mode (marked by O_DIRECT as suggested by Alan Cox), a write() on a pipe will not merge the new data with previous writes, so each write will get a pipe buffer of its own. The pipe buffer is then marked with the PIPE_BUF_FLAG_PACKET flag, which in turn will tell the reader side to break the read at that boundary (and throw away any partial packet contents that do not fit in the read buffer). End result: as long as you do writes less than PIPE_BUF in size (so that the pipe doesn't have to split them up), you can now treat the pipe as a packet interface, where each read() system call will read one packet at a time. You can just use a sufficiently big read buffer (PIPE_BUF is sufficient, since bigger than that doesn't guarantee atomicity anyway), and the return value of the read() will naturally give you the size of the packet. NOTE! We do not support zero-sized packets, and zero-sized reads and writes to a pipe continue to be no-ops. Also note that big packets will currently be split at write time, but that the size at which that happens is not really specified (except that it's bigger than PIPE_BUF). Currently that limit is the system page size, but we might want to explicitly support bigger packets some day. The main user for this is going to be the autofs packet interface, allowing us to stop having to care so deeply about exact packet sizes (which have had bugs with 32/64-bit compatibility modes). But user space can create packetized pipes with "pipe2(fd, O_DIRECT)", which will fail with an EINVAL on kernels that do not support this interface. Tested-by: Michael Tokarev Cc: Alan Cox Cc: David Miller Cc: Ian Kent Cc: Thomas Meyer Signed-off-by: Linus Torvalds Signed-off-by: Greg Kroah-Hartman

Fix broken "pipe: use event aware wakeups" optimization

2011-01-21T00:21:59Z

Commit e462c448fdc8 ("pipe: use event aware wakeups") optimized the pipe event wakeup calls to avoid wakeups if the events do not match the requested set. However, the optimization was buggy, in that it didn't actually use the correct sets for the events: when we make room for more data to be written, the pipe poll() routine will return both the POLLOUT _and_ POLLWRNORM bits. Similarly for read. And most critically, when a pipe is released, that will potentially result in POLLHUP|POLLERR (depending on whether it was the last reader or writer), not just the regular POLLIN|POLLOUT. This bug showed itself as a hung gnome-screensaver-dialog process, stuck forever (or at least until it was poked by a signal or by being traced) in a poll() system call. Cc: Davide Libenzi Cc: David S. Miller Cc: Eric Dumazet Cc: Jens Axboe Cc: Andrew Morton Signed-off-by: Linus Torvalds

sanitize vfsmount refcounting changes

2011-01-16T18:47:07Z

Instead of splitting refcount between (per-cpu) mnt_count and (SMP-only) mnt_longrefs, make all references contribute to mnt_count again and keep track of how many are longterm ones. Accounting rules for longterm count: * 1 for each fs_struct.root.mnt * 1 for each fs_struct.pwd.mnt * 1 for having non-NULL ->mnt_ns * decrement to 0 happens only under vfsmount lock exclusive That allows nice common case for mntput() - since we can't drop the final reference until after mnt_longterm has reached 0 due to the rules above, mntput() can grab vfsmount lock shared and check mnt_longterm. If it turns out to be non-zero (which is the common case), we know that this is not the final mntput() and can just blindly decrement percpu mnt_count. Otherwise we grab vfsmount lock exclusive and do usual decrement-and-check of percpu mnt_count. For fs_struct.c we have mnt_make_longterm() and mnt_make_shortterm(); namespace.c uses the latter in places where we don't already hold vfsmount lock exclusive and opencodes a few remaining spots where we need to manipulate mnt_longterm. Note that we mostly revert the code outside of fs/namespace.c back to what we used to have; in particular, normal code doesn't need to care about two kinds of references, etc. And we get to keep the optimization Nick's variant had bought us... Signed-off-by: Al Viro

Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6

2011-01-13T18:27:28Z

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6: (41 commits) fs: add documentation on fallocate hole punching Gfs2: fail if we try to use hole punch Btrfs: fail if we try to use hole punch Ext4: fail if we try to use hole punch Ocfs2: handle hole punching via fallocate properly XFS: handle hole punching via fallocate properly fs: add hole punching to fallocate vfs: pass struct file to do_truncate on O_TRUNC opens (try #2) fix signedness mess in rw_verify_area() on 64bit architectures fs: fix kernel-doc for dcache::prepend_path fs: fix kernel-doc for dcache::d_validate sanitize ecryptfs ->mount() switch afs move internal-only parts of ncpfs headers to fs/ncpfs switch ncpfs switch 9p pass default dentry_operations to mount_pseudo() switch hostfs switch affs switch configfs ...

pipe: use event aware wakeups

2011-01-13T16:03:15Z

Send the events the wakeup refers to, so that epoll, and even the new poll code in fs/select.c can avoid wakeups if the events do not match the requested set. Signed-off-by: Davide Libenzi Acked-by: David S. Miller Acked-by: Eric Dumazet Cc: Jens Axboe Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds

pass default dentry_operations to mount_pseudo()

2011-01-13T01:03:43Z

Signed-off-by: Al Viro

fs: scale mntget/mntput

2011-01-07T06:50:33Z

The problem that this patch aims to fix is vfsmount refcounting scalability. We need to take a reference on the vfsmount for every successful path lookup, which often go to the same mount point. The fundamental difficulty is that a "simple" reference count can never be made scalable, because any time a reference is dropped, we must check whether that was the last reference. To do that requires communication with all other CPUs that may have taken a reference count. We can make refcounts more scalable in a couple of ways, involving keeping distributed counters, and checking for the global-zero condition less frequently. - check the global sum once every interval (this will delay zero detection for some interval, so it's probably a showstopper for vfsmounts). - keep a local count and only taking the global sum when local reaches 0 (this is difficult for vfsmounts, because we can't hold preempt off for the life of a reference, so a counter would need to be per-thread or tied strongly to a particular CPU which requires more locking). - keep a local difference of increments and decrements, which allows us to sum the total difference and hence find the refcount when summing all CPUs. Then, keep a single integer "long" refcount for slow and long lasting references, and only take the global sum of local counters when the long refcount is 0. This last scheme is what I implemented here. Attached mounts and process root and working directory references are "long" references, and everything else is a short reference. This allows scalable vfsmount references during path walking over mounted subtrees and unattached (lazy umounted) mounts with processes still running in them. This results in one fewer atomic op in the fastpath: mntget is now just a per-CPU inc, rather than an atomic inc; and mntput just requires a spinlock and non-atomic decrement in the common case. However code is otherwise bigger and heavier, so single threaded performance is basically a wash. Signed-off-by: Nick Piggin

fs: improve scalability of pseudo filesystems

2011-01-07T06:50:32Z

Regardless of how much we possibly try to scale dcache, there is likely always going to be some fundamental contention when adding or removing children under the same parent. Pseudo filesystems do not seem need to have connected dentries because by definition they are disconnected. Signed-off-by: Nick Piggin

fs: dcache reduce branches in lookup path

2011-01-07T06:50:28Z

Reduce some branches and memory accesses in dcache lookup by adding dentry flags to indicate common d_ops are set, rather than having to check them. This saves a pointer memory access (dentry->d_op) in common path lookup situations, and saves another pointer load and branch in cases where we have d_op but not the particular operation. Patched with: git grep -E '[.>]([[:space:]])*d_op([[:space:]])*=' | xargs sed -e 's/\([^\t ]*\)->d_op = \(.*\);/d_set_d_op(\1, \2);/' -e 's/\([^\t ]*\)\.d_op = \(.*\);/d_set_d_op(\&\1, \2);/' -i Signed-off-by: Nick Piggin

fs: avoid inode RCU freeing for pseudo fs

2011-01-07T06:50:26Z

Pseudo filesystems that don't put inode on RCU list or reachable by rcu-walk dentries do not need to RCU free their inodes. Signed-off-by: Nick Piggin