Require filesystems be aware of .d_revalidate being called in rcu-walk
mode (nd->flags & LOOKUP_RCU). For now do a simple push down, returning
-ECHILD from all implementations.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

Put dentry and inode fields into top of data structure.  This allows RCU path
traversal to perform an RCU dentry lookup in a path walk by touching only the
first 56 bytes of the dentry.

We also fit in 8 bytes of inline name in the first 64 bytes, so for short
names, only 64 bytes needs to be touched to perform the lookup. We should
get rid of the hash->prev pointer from the first 64 bytes, and fit 16 bytes
of name in there, which will take care of 81% rather than 32% of the kernel
tree.

inode is also rearranged so that RCU lookup will only touch a single cacheline
in the inode, plus one in the i_ops structure.

This is important for directory component lookups in RCU path walking. In the
kernel source, directory names average is around 6 chars, so this works.

When we reach the last element of the lookup, we need to lock it and take its
refcount which requires another cacheline access.

Align dentry and inode operations structs, so members will be at predictable
offsets and we can group common operations into head of structure.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

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 <npiggin@kernel.dk>

Rather than keep a d_mounted count in the dentry, set a dentry flag instead.
The flag can be cleared by checking the hash table to see if there are any
mounts left, which is not time critical because it is performed at detach time.

The mounted state of a dentry is only used to speculatively take a look in the
mount hash table if it is set -- before following the mount, vfsmount lock is
taken and mount re-checked without races.

This saves 4 bytes on 32-bit, nothing on 64-bit but it does provide a hole I
might use later (and some configs have larger than 32-bit spinlocks which might
make use of the hole).

Autofs4 conversion and changelog by Ian Kent <raven@themaw.net>:
In autofs4, when expring direct (or offset) mounts we need to ensure that we
block user path walks into the autofs mount, which is covered by another mount.
To do this we clear the mounted status so that follows stop before walking into
the mount and are essentially blocked until the expire is completed. The
automount daemon still finds the correct dentry for the umount due to the
follow mount logic in fs/autofs4/root.c:autofs4_follow_link(), which is set as
an inode operation for direct and offset mounts only and is called following
the lookup that stopped at the covered mount.

At the end of the expire the covering mount probably has gone away so the
mounted status need not be restored. But we need to check this and only restore
the mounted status if the expire failed.

XXX: autofs may not work right if we have other mounts go over the top of it?
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

Use a seqlock in the fs_struct to enable us to take an atomic copy of the
complete cwd and root paths. Use this in the RCU lookup path to avoid a
thread-shared spinlock in RCU lookup operations.

Multi-threaded apps may now perform path lookups with scalability matching
multi-process apps. Operations such as stat(2) become very scalable for
multi-threaded workload.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

Perform common cases of path lookups without any stores or locking in the
ancestor dentry elements. This is called rcu-walk, as opposed to the current
algorithm which is a refcount based walk, or ref-walk.

This results in far fewer atomic operations on every path element,
significantly improving path lookup performance. It also avoids cacheline
bouncing on common dentries, significantly improving scalability.

The overall design is like this:
* LOOKUP_RCU is set in nd->flags, which distinguishes rcu-walk from ref-walk.
* Take the RCU lock for the entire path walk, starting with the acquiring
  of the starting path (eg. root/cwd/fd-path). So now dentry refcounts are
  not required for dentry persistence.
* synchronize_rcu is called when unregistering a filesystem, so we can
  access d_ops and i_ops during rcu-walk.
* Similarly take the vfsmount lock for the entire path walk. So now mnt
  refcounts are not required for persistence. Also we are free to perform mount
  lookups, and to assume dentry mount points and mount roots are stable up and
  down the path.
* Have a per-dentry seqlock to protect the dentry name, parent, and inode,
  so we can load this tuple atomically, and also check whether any of its
  members have changed.
* Dentry lookups (based on parent, candidate string tuple) recheck the parent
  sequence after the child is found in case anything changed in the parent
  during the path walk.
* inode is also RCU protected so we can load d_inode and use the inode for
  limited things.
* i_mode, i_uid, i_gid can be tested for exec permissions during path walk.
* i_op can be loaded.

When we reach the destination dentry, we lock it, recheck lookup sequence,
and increment its refcount and mountpoint refcount. RCU and vfsmount locks
are dropped. This is termed "dropping rcu-walk". If the dentry refcount does
not match, we can not drop rcu-walk gracefully at the current point in the
lokup, so instead return -ECHILD (for want of a better errno). This signals the
path walking code to re-do the entire lookup with a ref-walk.

Aside from the final dentry, there are other situations that may be encounted
where we cannot continue rcu-walk. In that case, we drop rcu-walk (ie. take
a reference on the last good dentry) and continue with a ref-walk. Again, if
we can drop rcu-walk gracefully, we return -ECHILD and do the whole lookup
using ref-walk. But it is very important that we can continue with ref-walk
for most cases, particularly to avoid the overhead of double lookups, and to
gain the scalability advantages on common path elements (like cwd and root).

The cases where rcu-walk cannot continue are:
* NULL dentry (ie. any uncached path element)
* parent with d_inode->i_op->permission or ACLs
* dentries with d_revalidate
* Following links

In future patches, permission checks and d_revalidate become rcu-walk aware. It
may be possible eventually to make following links rcu-walk aware.

Uncached path elements will always require dropping to ref-walk mode, at the
very least because i_mutex needs to be grabbed, and objects allocated.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

Add branch annotations for seqlock read fastpath, and introduce
__read_seqcount_begin and __read_seqcount_end functions, that can avoid the
smp_rmb() if used carefully. These will be used by store-free path walking
algorithm performance is critical and seqlocks are in use.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

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 <npiggin@kernel.dk>

RCU free the struct inode. This will allow:

- Subsequent store-free path walking patch. The inode must be consulted for
  permissions when walking, so an RCU inode reference is a must.
- sb_inode_list_lock to be moved inside i_lock because sb list walkers who want
  to take i_lock no longer need to take sb_inode_list_lock to walk the list in
  the first place. This will simplify and optimize locking.
- Could remove some nested trylock loops in dcache code
- Could potentially simplify things a bit in VM land. Do not need to take the
  page lock to follow page->mapping.

The downsides of this is the performance cost of using RCU. In a simple
creat/unlink microbenchmark, performance drops by about 10% due to inability to
reuse cache-hot slab objects. As iterations increase and RCU freeing starts
kicking over, this increases to about 20%.

In cases where inode lifetimes are longer (ie. many inodes may be allocated
during the average life span of a single inode), a lot of this cache reuse is
not applicable, so the regression caused by this patch is smaller.

The cache-hot regression could largely be avoided by using SLAB_DESTROY_BY_RCU,
however this adds some complexity to list walking and store-free path walking,
so I prefer to implement this at a later date, if it is shown to be a win in
real situations. I haven't found a regression in any non-micro benchmark so I
doubt it will be a problem.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

The tricky locking for disposing of a dentry is duplicated 3 times in the
dcache (dput, pruning a dentry from the LRU, and pruning its ancestors).
Consolidate them all into a single function dentry_kill.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

Signed-off-by: Nick Piggin <npiggin@kernel.dk>

Signed-off-by: Nick Piggin <npiggin@kernel.dk>

prune_one_dentry can avoid quite a bit of locking in the common case where
ancestors have an elevated refcount. Alternatively, we could have gone the
other way and made fewer trylocks in the case where d_count goes to zero, but
is probably less common.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

Use RCU to simplify locking in dget_parent.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

dget_locked was a shortcut to avoid the lazy lru manipulation when we already
held dcache_lock (lru manipulation was relatively cheap at that point).
However, how that the lru lock is an innermost one, we never hold it at any
caller, so the lock cost can now be avoided. We already have well working lazy
dcache LRU, so it should be fine to defer LRU manipulations to scan time.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

dcache_inode_lock can be avoided in d_delete() and d_materialise_unique()
in cases where it is not required.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

Signed-off-by: Nick Piggin <npiggin@kernel.dk>

It is possible to run dput without taking data structure locks up-front. In
many cases where we don't kill the dentry anyway, these locks are not required.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

Long lived dcache "multi-step" operations which retry on rename seq can
be starved with a lot of rename activity. If they fail after the 1st pass,
take the rename_lock for writing to avoid further starvation.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

dcache_lock no longer protects anything. remove it.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

The remaining usages for dcache_lock is to allow atomic, multi-step read-side
operations over the directory tree by excluding modifications to the tree.
Also, to walk in the leaf->root direction in the tree where we don't have
a natural d_lock ordering.

This could be accomplished by taking every d_lock, but this would mean a
huge number of locks and actually gets very tricky.

Solve this instead by using the rename seqlock for multi-step read-side
operations, retry in case of a rename so we don't walk up the wrong parent.
Concurrent dentry insertions are not serialised against.  Concurrent deletes
are tricky when walking up the directory: our parent might have been deleted
when dropping locks so also need to check and retry for that.

We can also use the rename lock in cases where livelock is a worry (and it
is introduced in subsequent patch).
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

Cover d_name with d_lock in more cases, where there may be concurrent
modification to it.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

Add a new lock, dcache_inode_lock, to protect the inode's i_dentry list
from concurrent modification. d_alias is also protected by d_lock.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

Protect d_subdirs and d_child with d_lock, except in filesystems that aren't
using dcache_lock for these anyway (eg. using i_mutex).

Note: if we change the locking rule in future so that ->d_child protection is
provided only with ->d_parent->d_lock, it may allow us to reduce some locking.
But it would be an exception to an otherwise regular locking scheme, so we'd
have to see some good results. Probably not worthwhile.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

Protect d_unhashed(dentry) condition with d_lock. This means keeping
DCACHE_UNHASHED bit in synch with hash manipulations.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

Make d_count non-atomic and protect it with d_lock. This allows us to ensure a
0 refcount dentry remains 0 without dcache_lock. It is also fairly natural when
we start protecting many other dentry members with d_lock.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

Add a new lock, dcache_lru_lock, to protect the dcache LRU list from concurrent
modification. d_lru is also protected by d_lock, which allows LRU lists to be
accessed without the lru lock, using RCU in future patches.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

Add a new lock, dcache_hash_lock, to protect the dcache hash table from
concurrent modification. d_hash is also protected by d_lock.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

Remove dcache_lock locking from hostfs filesystem, and move it into dcache
helpers. All that is required is a coherent path name. Protection from
concurrent modification of the namespace after path name generation is not
provided in current code, because dcache_lock is dropped before the path is
used.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

Change d_hash so it may be called from lock-free RCU lookups. See similar
patch for d_compare for details.

For in-tree filesystems, this is just a mechanical change.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

Change d_compare so it may be called from lock-free RCU lookups. This
does put significant restrictions on what may be done from the callback,
however there don't seem to have been any problems with in-tree fses.
If some strange use case pops up that _really_ cannot cope with the
rcu-walk rules, we can just add new rcu-unaware callbacks, which would
cause name lookup to drop out of rcu-walk mode.

For in-tree filesystems, this is just a mechanical change.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

smpfs and ncpfs want to update a live dentry name in-place. Rather than
have them open code the locking, provide a documented dcache API.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

Use vfat's method for dealing with negative dentries to preserve case,
rather than overwrite dentry name in d_revalidate, which is a bit ugly
and also gets in the way of doing lock-free path walking.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

Use vfat's method for dealing with negative dentries to preserve case,
rather than overwrite dentry name in d_revalidate, which is a bit ugly
and also gets in the way of doing lock-free path walking.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

Change d_delete from a dentry deletion notification to a dentry caching
advise, more like ->drop_inode. Require it to be constant and idempotent,
and not take d_lock. This is how all existing filesystems use the callback
anyway.

This makes fine grained dentry locking of dput and dentry lru scanning
much simpler.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

Remove redundant (and incorrect, since dcache RCU lookup) dentry locking
documentation and point to the canonical document.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

Switching d_op on a live dentry is racy in general, so avoid it. In this case
it is a negative dentry, which is safer, but there are still concurrent ops
which may be called on d_op in that case (eg. d_revalidate). So in general
a filesystem may not do this. Fix configfs so as not to do this.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

Switching d_op on a live dentry is racy in general, so avoid it. In this case
it is a negative dentry, which is safer, but there are still concurrent ops
which may be called on d_op in that case (eg. d_revalidate). So in general
a filesystem may not do this. Fix cgroupfs so as not to do this.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

percpu_counter library generates quite nasty code, so unless you need
to dynamically allocate counters or take fast approximate value, a
simple per cpu set of counters is much better.

The percpu_counter can never be made to work as well, because it has an
indirection from pointer to percpu memory, and it can't use direct
this_cpu_inc interfaces because it doesn't use static PER_CPU data, so
code will always be worse.

In the fastpath, it is the difference between this:

        incl %gs:nr_dentry      # nr_dentry

and this:

        movl    percpu_counter_batch(%rip), %edx        # percpu_counter_batch,
        movl    $1, %esi        #,
        movq    $nr_dentry, %rdi        #,
        call    __percpu_counter_add    # (plus I clobber registers)

__percpu_counter_add:
        pushq   %rbp    #
        movq    %rsp, %rbp      #,
        subq    $32, %rsp       #,
        movq    %rbx, -24(%rbp) #,
        movq    %r12, -16(%rbp) #,
        movq    %r13, -8(%rbp)  #,
        movq    %rdi, %rbx      # fbc, fbc
#APP
# 216 "/home/npiggin/usr/src/linux-2.6/arch/x86/include/asm/thread_info.h" 1
        movq %gs:kernel_stack,%rax      #, pfo_ret__
# 0 "" 2
#NO_APP
        incl    -8124(%rax)     # <variable>.preempt_count
        movq    32(%rdi), %r12  # <variable>.counters, tcp_ptr__
#APP
# 78 "lib/percpu_counter.c" 1
        add %gs:this_cpu_off, %r12      # this_cpu_off, tcp_ptr__
# 0 "" 2
#NO_APP
        movslq  (%r12),%r13     #* tcp_ptr__, tmp73
        movslq  %edx,%rax       # batch, batch
        addq    %rsi, %r13      # amount, count
        cmpq    %rax, %r13      # batch, count
        jge     .L27    #,
        negl    %edx    # tmp76
        movslq  %edx,%rdx       # tmp76, tmp77
        cmpq    %rdx, %r13      # tmp77, count
        jg      .L28    #,
.L27:
        movq    %rbx, %rdi      # fbc,
        call    _raw_spin_lock  #
        addq    %r13, 8(%rbx)   # count, <variable>.count
        movq    %rbx, %rdi      # fbc,
        movl    $0, (%r12)      #,* tcp_ptr__
        call    _raw_spin_unlock        #
.L29:
#APP
# 216 "/home/npiggin/usr/src/linux-2.6/arch/x86/include/asm/thread_info.h" 1
        movq %gs:kernel_stack,%rax      #, pfo_ret__
# 0 "" 2
#NO_APP
        decl    -8124(%rax)     # <variable>.preempt_count
        movq    -8136(%rax), %rax       #, D.14625
        testb   $8, %al #, D.14625
        jne     .L32    #,
.L31:
        movq    -24(%rbp), %rbx #,
        movq    -16(%rbp), %r12 #,
        movq    -8(%rbp), %r13  #,
        leave
        ret
        .p2align 4,,10
        .p2align 3
.L28:
        movl    %r13d, (%r12)   # count,*
        jmp     .L29    #
.L32:
        call    preempt_schedule        #
        .p2align 4,,6
        jmp     .L31    #
        .size   __percpu_counter_add, .-__percpu_counter_add
        .p2align 4,,15
Signed-off-by: Nick Piggin <npiggin@kernel.dk>

The nr_unused counters count the number of objects on an LRU, and as such they
are synchronized with LRU object insertion and removal and scanning, and
protected under the LRU lock.

Making it per-cpu does not actually get any concurrency improvements because of
this lock, and summing the counter is much slower, and
incrementing/decrementing it costs more code size and is slower too.

These counters should stay per-LRU, which currently means global.
Signed-off-by: Nick Piggin <npiggin@kernel.dk>