linux/lib/idr.c
Mel Gorman c7df8ad291 mm, truncate: do not check mapping for every page being truncated
During truncation, the mapping has already been checked for shmem and
dax so it's known that workingset_update_node is required.

This patch avoids the checks on mapping for each page being truncated.
In all other cases, a lookup helper is used to determine if
workingset_update_node() needs to be called.  The one danger is that the
API is slightly harder to use as calling workingset_update_node directly
without checking for dax or shmem mappings could lead to surprises.
However, the API rarely needs to be used and hopefully the comment is
enough to give people the hint.

sparsetruncate (tiny)
                              4.14.0-rc4             4.14.0-rc4
                             oneirq-v1r1        pickhelper-v1r1
Min          Time      141.00 (   0.00%)      140.00 (   0.71%)
1st-qrtle    Time      142.00 (   0.00%)      141.00 (   0.70%)
2nd-qrtle    Time      142.00 (   0.00%)      142.00 (   0.00%)
3rd-qrtle    Time      143.00 (   0.00%)      143.00 (   0.00%)
Max-90%      Time      144.00 (   0.00%)      144.00 (   0.00%)
Max-95%      Time      147.00 (   0.00%)      145.00 (   1.36%)
Max-99%      Time      195.00 (   0.00%)      191.00 (   2.05%)
Max          Time      230.00 (   0.00%)      205.00 (  10.87%)
Amean        Time      144.37 (   0.00%)      143.82 (   0.38%)
Stddev       Time       10.44 (   0.00%)        9.00 (  13.74%)
Coeff        Time        7.23 (   0.00%)        6.26 (  13.41%)
Best99%Amean Time      143.72 (   0.00%)      143.34 (   0.26%)
Best95%Amean Time      142.37 (   0.00%)      142.00 (   0.26%)
Best90%Amean Time      142.19 (   0.00%)      141.85 (   0.24%)
Best75%Amean Time      141.92 (   0.00%)      141.58 (   0.24%)
Best50%Amean Time      141.69 (   0.00%)      141.31 (   0.27%)
Best25%Amean Time      141.38 (   0.00%)      140.97 (   0.29%)

As you'd expect, the gain is marginal but it can be detected.  The
differences in bonnie are all within the noise which is not surprising
given the impact on the microbenchmark.

radix_tree_update_node_t is a callback for some radix operations that
optionally passes in a private field.  The only user of the callback is
workingset_update_node and as it no longer requires a mapping, the
private field is removed.

Link: http://lkml.kernel.org/r/20171018075952.10627-3-mgorman@techsingularity.net
Signed-off-by: Mel Gorman <mgorman@techsingularity.net>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2017-11-15 18:21:06 -08:00

493 lines
14 KiB
C

#include <linux/bitmap.h>
#include <linux/export.h>
#include <linux/idr.h>
#include <linux/slab.h>
#include <linux/spinlock.h>
DEFINE_PER_CPU(struct ida_bitmap *, ida_bitmap);
static DEFINE_SPINLOCK(simple_ida_lock);
int idr_alloc_cmn(struct idr *idr, void *ptr, unsigned long *index,
unsigned long start, unsigned long end, gfp_t gfp,
bool ext)
{
struct radix_tree_iter iter;
void __rcu **slot;
if (WARN_ON_ONCE(radix_tree_is_internal_node(ptr)))
return -EINVAL;
radix_tree_iter_init(&iter, start);
if (ext)
slot = idr_get_free_ext(&idr->idr_rt, &iter, gfp, end);
else
slot = idr_get_free(&idr->idr_rt, &iter, gfp, end);
if (IS_ERR(slot))
return PTR_ERR(slot);
radix_tree_iter_replace(&idr->idr_rt, &iter, slot, ptr);
radix_tree_iter_tag_clear(&idr->idr_rt, &iter, IDR_FREE);
if (index)
*index = iter.index;
return 0;
}
EXPORT_SYMBOL_GPL(idr_alloc_cmn);
/**
* idr_alloc_cyclic - allocate new idr entry in a cyclical fashion
* @idr: idr handle
* @ptr: pointer to be associated with the new id
* @start: the minimum id (inclusive)
* @end: the maximum id (exclusive)
* @gfp: memory allocation flags
*
* Allocates an ID larger than the last ID allocated if one is available.
* If not, it will attempt to allocate the smallest ID that is larger or
* equal to @start.
*/
int idr_alloc_cyclic(struct idr *idr, void *ptr, int start, int end, gfp_t gfp)
{
int id, curr = idr->idr_next;
if (curr < start)
curr = start;
id = idr_alloc(idr, ptr, curr, end, gfp);
if ((id == -ENOSPC) && (curr > start))
id = idr_alloc(idr, ptr, start, curr, gfp);
if (id >= 0)
idr->idr_next = id + 1U;
return id;
}
EXPORT_SYMBOL(idr_alloc_cyclic);
/**
* idr_for_each - iterate through all stored pointers
* @idr: idr handle
* @fn: function to be called for each pointer
* @data: data passed to callback function
*
* The callback function will be called for each entry in @idr, passing
* the id, the pointer and the data pointer passed to this function.
*
* If @fn returns anything other than %0, the iteration stops and that
* value is returned from this function.
*
* idr_for_each() can be called concurrently with idr_alloc() and
* idr_remove() if protected by RCU. Newly added entries may not be
* seen and deleted entries may be seen, but adding and removing entries
* will not cause other entries to be skipped, nor spurious ones to be seen.
*/
int idr_for_each(const struct idr *idr,
int (*fn)(int id, void *p, void *data), void *data)
{
struct radix_tree_iter iter;
void __rcu **slot;
radix_tree_for_each_slot(slot, &idr->idr_rt, &iter, 0) {
int ret = fn(iter.index, rcu_dereference_raw(*slot), data);
if (ret)
return ret;
}
return 0;
}
EXPORT_SYMBOL(idr_for_each);
/**
* idr_get_next - Find next populated entry
* @idr: idr handle
* @nextid: Pointer to lowest possible ID to return
*
* Returns the next populated entry in the tree with an ID greater than
* or equal to the value pointed to by @nextid. On exit, @nextid is updated
* to the ID of the found value. To use in a loop, the value pointed to by
* nextid must be incremented by the user.
*/
void *idr_get_next(struct idr *idr, int *nextid)
{
struct radix_tree_iter iter;
void __rcu **slot;
slot = radix_tree_iter_find(&idr->idr_rt, &iter, *nextid);
if (!slot)
return NULL;
*nextid = iter.index;
return rcu_dereference_raw(*slot);
}
EXPORT_SYMBOL(idr_get_next);
void *idr_get_next_ext(struct idr *idr, unsigned long *nextid)
{
struct radix_tree_iter iter;
void __rcu **slot;
slot = radix_tree_iter_find(&idr->idr_rt, &iter, *nextid);
if (!slot)
return NULL;
*nextid = iter.index;
return rcu_dereference_raw(*slot);
}
EXPORT_SYMBOL(idr_get_next_ext);
/**
* idr_replace - replace pointer for given id
* @idr: idr handle
* @ptr: New pointer to associate with the ID
* @id: Lookup key
*
* Replace the pointer registered with an ID and return the old value.
* This function can be called under the RCU read lock concurrently with
* idr_alloc() and idr_remove() (as long as the ID being removed is not
* the one being replaced!).
*
* Returns: the old value on success. %-ENOENT indicates that @id was not
* found. %-EINVAL indicates that @id or @ptr were not valid.
*/
void *idr_replace(struct idr *idr, void *ptr, int id)
{
if (id < 0)
return ERR_PTR(-EINVAL);
return idr_replace_ext(idr, ptr, id);
}
EXPORT_SYMBOL(idr_replace);
void *idr_replace_ext(struct idr *idr, void *ptr, unsigned long id)
{
struct radix_tree_node *node;
void __rcu **slot = NULL;
void *entry;
if (WARN_ON_ONCE(radix_tree_is_internal_node(ptr)))
return ERR_PTR(-EINVAL);
entry = __radix_tree_lookup(&idr->idr_rt, id, &node, &slot);
if (!slot || radix_tree_tag_get(&idr->idr_rt, id, IDR_FREE))
return ERR_PTR(-ENOENT);
__radix_tree_replace(&idr->idr_rt, node, slot, ptr, NULL);
return entry;
}
EXPORT_SYMBOL(idr_replace_ext);
/**
* DOC: IDA description
*
* The IDA is an ID allocator which does not provide the ability to
* associate an ID with a pointer. As such, it only needs to store one
* bit per ID, and so is more space efficient than an IDR. To use an IDA,
* define it using DEFINE_IDA() (or embed a &struct ida in a data structure,
* then initialise it using ida_init()). To allocate a new ID, call
* ida_simple_get(). To free an ID, call ida_simple_remove().
*
* If you have more complex locking requirements, use a loop around
* ida_pre_get() and ida_get_new() to allocate a new ID. Then use
* ida_remove() to free an ID. You must make sure that ida_get_new() and
* ida_remove() cannot be called at the same time as each other for the
* same IDA.
*
* You can also use ida_get_new_above() if you need an ID to be allocated
* above a particular number. ida_destroy() can be used to dispose of an
* IDA without needing to free the individual IDs in it. You can use
* ida_is_empty() to find out whether the IDA has any IDs currently allocated.
*
* IDs are currently limited to the range [0-INT_MAX]. If this is an awkward
* limitation, it should be quite straightforward to raise the maximum.
*/
/*
* Developer's notes:
*
* The IDA uses the functionality provided by the IDR & radix tree to store
* bitmaps in each entry. The IDR_FREE tag means there is at least one bit
* free, unlike the IDR where it means at least one entry is free.
*
* I considered telling the radix tree that each slot is an order-10 node
* and storing the bit numbers in the radix tree, but the radix tree can't
* allow a single multiorder entry at index 0, which would significantly
* increase memory consumption for the IDA. So instead we divide the index
* by the number of bits in the leaf bitmap before doing a radix tree lookup.
*
* As an optimisation, if there are only a few low bits set in any given
* leaf, instead of allocating a 128-byte bitmap, we use the 'exceptional
* entry' functionality of the radix tree to store BITS_PER_LONG - 2 bits
* directly in the entry. By being really tricksy, we could store
* BITS_PER_LONG - 1 bits, but there're diminishing returns after optimising
* for 0-3 allocated IDs.
*
* We allow the radix tree 'exceptional' count to get out of date. Nothing
* in the IDA nor the radix tree code checks it. If it becomes important
* to maintain an accurate exceptional count, switch the rcu_assign_pointer()
* calls to radix_tree_iter_replace() which will correct the exceptional
* count.
*
* The IDA always requires a lock to alloc/free. If we add a 'test_bit'
* equivalent, it will still need locking. Going to RCU lookup would require
* using RCU to free bitmaps, and that's not trivial without embedding an
* RCU head in the bitmap, which adds a 2-pointer overhead to each 128-byte
* bitmap, which is excessive.
*/
#define IDA_MAX (0x80000000U / IDA_BITMAP_BITS)
/**
* ida_get_new_above - allocate new ID above or equal to a start id
* @ida: ida handle
* @start: id to start search at
* @id: pointer to the allocated handle
*
* Allocate new ID above or equal to @start. It should be called
* with any required locks to ensure that concurrent calls to
* ida_get_new_above() / ida_get_new() / ida_remove() are not allowed.
* Consider using ida_simple_get() if you do not have complex locking
* requirements.
*
* If memory is required, it will return %-EAGAIN, you should unlock
* and go back to the ida_pre_get() call. If the ida is full, it will
* return %-ENOSPC. On success, it will return 0.
*
* @id returns a value in the range @start ... %0x7fffffff.
*/
int ida_get_new_above(struct ida *ida, int start, int *id)
{
struct radix_tree_root *root = &ida->ida_rt;
void __rcu **slot;
struct radix_tree_iter iter;
struct ida_bitmap *bitmap;
unsigned long index;
unsigned bit, ebit;
int new;
index = start / IDA_BITMAP_BITS;
bit = start % IDA_BITMAP_BITS;
ebit = bit + RADIX_TREE_EXCEPTIONAL_SHIFT;
slot = radix_tree_iter_init(&iter, index);
for (;;) {
if (slot)
slot = radix_tree_next_slot(slot, &iter,
RADIX_TREE_ITER_TAGGED);
if (!slot) {
slot = idr_get_free(root, &iter, GFP_NOWAIT, IDA_MAX);
if (IS_ERR(slot)) {
if (slot == ERR_PTR(-ENOMEM))
return -EAGAIN;
return PTR_ERR(slot);
}
}
if (iter.index > index) {
bit = 0;
ebit = RADIX_TREE_EXCEPTIONAL_SHIFT;
}
new = iter.index * IDA_BITMAP_BITS;
bitmap = rcu_dereference_raw(*slot);
if (radix_tree_exception(bitmap)) {
unsigned long tmp = (unsigned long)bitmap;
ebit = find_next_zero_bit(&tmp, BITS_PER_LONG, ebit);
if (ebit < BITS_PER_LONG) {
tmp |= 1UL << ebit;
rcu_assign_pointer(*slot, (void *)tmp);
*id = new + ebit - RADIX_TREE_EXCEPTIONAL_SHIFT;
return 0;
}
bitmap = this_cpu_xchg(ida_bitmap, NULL);
if (!bitmap)
return -EAGAIN;
memset(bitmap, 0, sizeof(*bitmap));
bitmap->bitmap[0] = tmp >> RADIX_TREE_EXCEPTIONAL_SHIFT;
rcu_assign_pointer(*slot, bitmap);
}
if (bitmap) {
bit = find_next_zero_bit(bitmap->bitmap,
IDA_BITMAP_BITS, bit);
new += bit;
if (new < 0)
return -ENOSPC;
if (bit == IDA_BITMAP_BITS)
continue;
__set_bit(bit, bitmap->bitmap);
if (bitmap_full(bitmap->bitmap, IDA_BITMAP_BITS))
radix_tree_iter_tag_clear(root, &iter,
IDR_FREE);
} else {
new += bit;
if (new < 0)
return -ENOSPC;
if (ebit < BITS_PER_LONG) {
bitmap = (void *)((1UL << ebit) |
RADIX_TREE_EXCEPTIONAL_ENTRY);
radix_tree_iter_replace(root, &iter, slot,
bitmap);
*id = new;
return 0;
}
bitmap = this_cpu_xchg(ida_bitmap, NULL);
if (!bitmap)
return -EAGAIN;
memset(bitmap, 0, sizeof(*bitmap));
__set_bit(bit, bitmap->bitmap);
radix_tree_iter_replace(root, &iter, slot, bitmap);
}
*id = new;
return 0;
}
}
EXPORT_SYMBOL(ida_get_new_above);
/**
* ida_remove - Free the given ID
* @ida: ida handle
* @id: ID to free
*
* This function should not be called at the same time as ida_get_new_above().
*/
void ida_remove(struct ida *ida, int id)
{
unsigned long index = id / IDA_BITMAP_BITS;
unsigned offset = id % IDA_BITMAP_BITS;
struct ida_bitmap *bitmap;
unsigned long *btmp;
struct radix_tree_iter iter;
void __rcu **slot;
slot = radix_tree_iter_lookup(&ida->ida_rt, &iter, index);
if (!slot)
goto err;
bitmap = rcu_dereference_raw(*slot);
if (radix_tree_exception(bitmap)) {
btmp = (unsigned long *)slot;
offset += RADIX_TREE_EXCEPTIONAL_SHIFT;
if (offset >= BITS_PER_LONG)
goto err;
} else {
btmp = bitmap->bitmap;
}
if (!test_bit(offset, btmp))
goto err;
__clear_bit(offset, btmp);
radix_tree_iter_tag_set(&ida->ida_rt, &iter, IDR_FREE);
if (radix_tree_exception(bitmap)) {
if (rcu_dereference_raw(*slot) ==
(void *)RADIX_TREE_EXCEPTIONAL_ENTRY)
radix_tree_iter_delete(&ida->ida_rt, &iter, slot);
} else if (bitmap_empty(btmp, IDA_BITMAP_BITS)) {
kfree(bitmap);
radix_tree_iter_delete(&ida->ida_rt, &iter, slot);
}
return;
err:
WARN(1, "ida_remove called for id=%d which is not allocated.\n", id);
}
EXPORT_SYMBOL(ida_remove);
/**
* ida_destroy - Free the contents of an ida
* @ida: ida handle
*
* Calling this function releases all resources associated with an IDA. When
* this call returns, the IDA is empty and can be reused or freed. The caller
* should not allow ida_remove() or ida_get_new_above() to be called at the
* same time.
*/
void ida_destroy(struct ida *ida)
{
struct radix_tree_iter iter;
void __rcu **slot;
radix_tree_for_each_slot(slot, &ida->ida_rt, &iter, 0) {
struct ida_bitmap *bitmap = rcu_dereference_raw(*slot);
if (!radix_tree_exception(bitmap))
kfree(bitmap);
radix_tree_iter_delete(&ida->ida_rt, &iter, slot);
}
}
EXPORT_SYMBOL(ida_destroy);
/**
* ida_simple_get - get a new id.
* @ida: the (initialized) ida.
* @start: the minimum id (inclusive, < 0x8000000)
* @end: the maximum id (exclusive, < 0x8000000 or 0)
* @gfp_mask: memory allocation flags
*
* Allocates an id in the range start <= id < end, or returns -ENOSPC.
* On memory allocation failure, returns -ENOMEM.
*
* Compared to ida_get_new_above() this function does its own locking, and
* should be used unless there are special requirements.
*
* Use ida_simple_remove() to get rid of an id.
*/
int ida_simple_get(struct ida *ida, unsigned int start, unsigned int end,
gfp_t gfp_mask)
{
int ret, id;
unsigned int max;
unsigned long flags;
BUG_ON((int)start < 0);
BUG_ON((int)end < 0);
if (end == 0)
max = 0x80000000;
else {
BUG_ON(end < start);
max = end - 1;
}
again:
if (!ida_pre_get(ida, gfp_mask))
return -ENOMEM;
spin_lock_irqsave(&simple_ida_lock, flags);
ret = ida_get_new_above(ida, start, &id);
if (!ret) {
if (id > max) {
ida_remove(ida, id);
ret = -ENOSPC;
} else {
ret = id;
}
}
spin_unlock_irqrestore(&simple_ida_lock, flags);
if (unlikely(ret == -EAGAIN))
goto again;
return ret;
}
EXPORT_SYMBOL(ida_simple_get);
/**
* ida_simple_remove - remove an allocated id.
* @ida: the (initialized) ida.
* @id: the id returned by ida_simple_get.
*
* Use to release an id allocated with ida_simple_get().
*
* Compared to ida_remove() this function does its own locking, and should be
* used unless there are special requirements.
*/
void ida_simple_remove(struct ida *ida, unsigned int id)
{
unsigned long flags;
BUG_ON((int)id < 0);
spin_lock_irqsave(&simple_ida_lock, flags);
ida_remove(ida, id);
spin_unlock_irqrestore(&simple_ida_lock, flags);
}
EXPORT_SYMBOL(ida_simple_remove);