From: Zi Yan <ziy@nvidia.com>
To: linux-mm@kvack.org, Andrew Morton <akpm@linux-foundation.org>,
Hugh Dickins <hughd@google.com>,
"Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Ryan Roberts <ryan.roberts@arm.com>,
"Kirill A . Shutemov" <kirill.shutemov@linux.intel.com>,
David Hildenbrand <david@redhat.com>,
Yang Shi <yang@os.amperecomputing.com>,
Miaohe Lin <linmiaohe@huawei.com>,
Kefeng Wang <wangkefeng.wang@huawei.com>,
Yu Zhao <yuzhao@google.com>, John Hubbard <jhubbard@nvidia.com>,
Baolin Wang <baolin.wang@linux.alibaba.com>,
linux-kselftest@vger.kernel.org, linux-kernel@vger.kernel.org,
Zi Yan <ziy@nvidia.com>, Kairui Song <kasong@tencent.com>
Subject: [PATCH v10 1/8] xarray: add xas_try_split() to split a multi-index entry
Date: Fri, 7 Mar 2025 12:39:54 -0500 [thread overview]
Message-ID: <20250307174001.242794-2-ziy@nvidia.com> (raw)
In-Reply-To: <20250307174001.242794-1-ziy@nvidia.com>
A preparation patch for non-uniform folio split, which always split a
folio into half iteratively, and minimal xarray entry split.
Currently, xas_split_alloc() and xas_split() always split all slots from a
multi-index entry. They cost the same number of xa_node as the
to-be-split slots. For example, to split an order-9 entry, which takes
2^(9-6)=8 slots, assuming XA_CHUNK_SHIFT is 6 (!CONFIG_BASE_SMALL), 8
xa_node are needed. Instead xas_try_split() is intended to be used
iteratively to split the order-9 entry into 2 order-8 entries, then split
one order-8 entry, based on the given index, to 2 order-7 entries, ...,
and split one order-1 entry to 2 order-0 entries. When splitting the
order-6 entry and a new xa_node is needed, xas_try_split() will try to
allocate one if possible. As a result, xas_try_split() would only need
1 xa_node instead of 8.
When a new xa_node is needed during the split, xas_try_split() can try to
allocate one but no more. -ENOMEM will be return if a node cannot be
allocated. -EINVAL will be return if a sibling node is split or cascade
split happens, where two or more new nodes are needed, and these are not
supported by xas_try_split().
xas_split_alloc() and xas_split() split an order-9 to order-0:
---------------------------------
| | | | | | | | |
| 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 |
| | | | | | | | |
---------------------------------
| | | |
------- --- --- -------
| | ... | |
V V V V
----------- ----------- ----------- -----------
| xa_node | | xa_node | ... | xa_node | | xa_node |
----------- ----------- ----------- -----------
xas_try_split() splits an order-9 to order-0:
---------------------------------
| | | | | | | | |
| 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 |
| | | | | | | | |
---------------------------------
|
|
V
-----------
| xa_node |
-----------
Signed-off-by: Zi Yan <ziy@nvidia.com>
Cc: Baolin Wang <baolin.wang@linux.alibaba.com>
Cc: David Hildenbrand <david@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Kefeng Wang <wangkefeng.wang@huawei.com>
Cc: Kirill A. Shuemov <kirill.shutemov@linux.intel.com>
Cc: Miaohe Lin <linmiaohe@huawei.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Ryan Roberts <ryan.roberts@arm.com>
Cc: Yang Shi <yang@os.amperecomputing.com>
Cc: Yu Zhao <yuzhao@google.com>
Cc: Zi Yan <ziy@nvidia.com>
Cc: Kairui Song <kasong@tencent.com>
---
Documentation/core-api/xarray.rst | 14 +++-
include/linux/xarray.h | 6 ++
lib/test_xarray.c | 52 ++++++++++++
lib/xarray.c | 132 +++++++++++++++++++++++++++---
tools/testing/radix-tree/Makefile | 1 +
5 files changed, 192 insertions(+), 13 deletions(-)
diff --git a/Documentation/core-api/xarray.rst b/Documentation/core-api/xarray.rst
index f6a3eef4fe7f..c6c91cbd0c3c 100644
--- a/Documentation/core-api/xarray.rst
+++ b/Documentation/core-api/xarray.rst
@@ -489,7 +489,19 @@ Storing ``NULL`` into any index of a multi-index entry will set the
entry at every index to ``NULL`` and dissolve the tie. A multi-index
entry can be split into entries occupying smaller ranges by calling
xas_split_alloc() without the xa_lock held, followed by taking the lock
-and calling xas_split().
+and calling xas_split() or calling xas_try_split() with xa_lock. The
+difference between xas_split_alloc()+xas_split() and xas_try_alloc() is
+that xas_split_alloc() + xas_split() split the entry from the original
+order to the new order in one shot uniformly, whereas xas_try_split()
+iteratively splits the entry containing the index non-uniformly.
+For example, to split an order-9 entry, which takes 2^(9-6)=8 slots,
+assuming ``XA_CHUNK_SHIFT`` is 6, xas_split_alloc() + xas_split() need
+8 xa_node. xas_try_split() splits the order-9 entry into
+2 order-8 entries, then split one order-8 entry, based on the given index,
+to 2 order-7 entries, ..., and split one order-1 entry to 2 order-0 entries.
+When splitting the order-6 entry and a new xa_node is needed, xas_try_split()
+will try to allocate one if possible. As a result, xas_try_split() would only
+need 1 xa_node instead of 8.
Functions and structures
========================
diff --git a/include/linux/xarray.h b/include/linux/xarray.h
index 0b618ec04115..4010195201c9 100644
--- a/include/linux/xarray.h
+++ b/include/linux/xarray.h
@@ -1555,6 +1555,7 @@ int xa_get_order(struct xarray *, unsigned long index);
int xas_get_order(struct xa_state *xas);
void xas_split(struct xa_state *, void *entry, unsigned int order);
void xas_split_alloc(struct xa_state *, void *entry, unsigned int order, gfp_t);
+void xas_try_split(struct xa_state *xas, void *entry, unsigned int order);
#else
static inline int xa_get_order(struct xarray *xa, unsigned long index)
{
@@ -1576,6 +1577,11 @@ static inline void xas_split_alloc(struct xa_state *xas, void *entry,
unsigned int order, gfp_t gfp)
{
}
+
+static inline void xas_try_split(struct xa_state *xas, void *entry,
+ unsigned int order)
+{
+}
#endif
/**
diff --git a/lib/test_xarray.c b/lib/test_xarray.c
index 0e865bab4a10..080a39d22e73 100644
--- a/lib/test_xarray.c
+++ b/lib/test_xarray.c
@@ -1858,6 +1858,54 @@ static void check_split_1(struct xarray *xa, unsigned long index,
xa_destroy(xa);
}
+static void check_split_2(struct xarray *xa, unsigned long index,
+ unsigned int order, unsigned int new_order)
+{
+ XA_STATE_ORDER(xas, xa, index, new_order);
+ unsigned int i, found;
+ void *entry;
+
+ xa_store_order(xa, index, order, xa, GFP_KERNEL);
+ xa_set_mark(xa, index, XA_MARK_1);
+
+ /* allocate a node for xas_try_split() */
+ xas_set_err(&xas, -ENOMEM);
+ XA_BUG_ON(xa, !xas_nomem(&xas, GFP_KERNEL));
+
+ xas_lock(&xas);
+ xas_try_split(&xas, xa, order);
+ if (((new_order / XA_CHUNK_SHIFT) < (order / XA_CHUNK_SHIFT)) &&
+ new_order < order - 1) {
+ XA_BUG_ON(xa, !xas_error(&xas) || xas_error(&xas) != -EINVAL);
+ xas_unlock(&xas);
+ goto out;
+ }
+ for (i = 0; i < (1 << order); i += (1 << new_order))
+ __xa_store(xa, index + i, xa_mk_index(index + i), 0);
+ xas_unlock(&xas);
+
+ for (i = 0; i < (1 << order); i++) {
+ unsigned int val = index + (i & ~((1 << new_order) - 1));
+ XA_BUG_ON(xa, xa_load(xa, index + i) != xa_mk_index(val));
+ }
+
+ xa_set_mark(xa, index, XA_MARK_0);
+ XA_BUG_ON(xa, !xa_get_mark(xa, index, XA_MARK_0));
+
+ xas_set_order(&xas, index, 0);
+ found = 0;
+ rcu_read_lock();
+ xas_for_each_marked(&xas, entry, ULONG_MAX, XA_MARK_1) {
+ found++;
+ XA_BUG_ON(xa, xa_is_internal(entry));
+ }
+ rcu_read_unlock();
+ XA_BUG_ON(xa, found != 1 << (order - new_order));
+out:
+ xas_destroy(&xas);
+ xa_destroy(xa);
+}
+
static noinline void check_split(struct xarray *xa)
{
unsigned int order, new_order;
@@ -1869,6 +1917,10 @@ static noinline void check_split(struct xarray *xa)
check_split_1(xa, 0, order, new_order);
check_split_1(xa, 1UL << order, order, new_order);
check_split_1(xa, 3UL << order, order, new_order);
+
+ check_split_2(xa, 0, order, new_order);
+ check_split_2(xa, 1UL << order, order, new_order);
+ check_split_2(xa, 3UL << order, order, new_order);
}
}
}
diff --git a/lib/xarray.c b/lib/xarray.c
index 116e9286c64e..3bae48558e21 100644
--- a/lib/xarray.c
+++ b/lib/xarray.c
@@ -278,6 +278,7 @@ void xas_destroy(struct xa_state *xas)
xas->xa_alloc = node = next;
}
}
+EXPORT_SYMBOL_GPL(xas_destroy);
/**
* xas_nomem() - Allocate memory if needed.
@@ -1007,6 +1008,26 @@ static void node_set_marks(struct xa_node *node, unsigned int offset,
}
}
+static void __xas_init_node_for_split(struct xa_state *xas,
+ struct xa_node *node, void *entry)
+{
+ unsigned int i;
+ void *sibling = NULL;
+ unsigned int mask = xas->xa_sibs;
+
+ if (!node)
+ return;
+ node->array = xas->xa;
+ for (i = 0; i < XA_CHUNK_SIZE; i++) {
+ if ((i & mask) == 0) {
+ RCU_INIT_POINTER(node->slots[i], entry);
+ sibling = xa_mk_sibling(i);
+ } else {
+ RCU_INIT_POINTER(node->slots[i], sibling);
+ }
+ }
+}
+
/**
* xas_split_alloc() - Allocate memory for splitting an entry.
* @xas: XArray operation state.
@@ -1025,7 +1046,6 @@ void xas_split_alloc(struct xa_state *xas, void *entry, unsigned int order,
gfp_t gfp)
{
unsigned int sibs = (1 << (order % XA_CHUNK_SHIFT)) - 1;
- unsigned int mask = xas->xa_sibs;
/* XXX: no support for splitting really large entries yet */
if (WARN_ON(xas->xa_shift + 2 * XA_CHUNK_SHIFT <= order))
@@ -1034,22 +1054,13 @@ void xas_split_alloc(struct xa_state *xas, void *entry, unsigned int order,
return;
do {
- unsigned int i;
- void *sibling = NULL;
struct xa_node *node;
node = kmem_cache_alloc_lru(radix_tree_node_cachep, xas->xa_lru, gfp);
if (!node)
goto nomem;
- node->array = xas->xa;
- for (i = 0; i < XA_CHUNK_SIZE; i++) {
- if ((i & mask) == 0) {
- RCU_INIT_POINTER(node->slots[i], entry);
- sibling = xa_mk_sibling(i);
- } else {
- RCU_INIT_POINTER(node->slots[i], sibling);
- }
- }
+
+ __xas_init_node_for_split(xas, node, entry);
RCU_INIT_POINTER(node->parent, xas->xa_alloc);
xas->xa_alloc = node;
} while (sibs-- > 0);
@@ -1122,6 +1133,103 @@ void xas_split(struct xa_state *xas, void *entry, unsigned int order)
xas_update(xas, node);
}
EXPORT_SYMBOL_GPL(xas_split);
+
+/**
+ * xas_try_split() - Try to split a multi-index entry.
+ * @xas: XArray operation state.
+ * @entry: New entry to store in the array.
+ * @order: Current entry order.
+ *
+ * The size of the new entries is set in @xas. The value in @entry is
+ * copied to all the replacement entries. If and only if one new xa_node is
+ * needed, the function will use GFP_NOWAIT to get one if xas->xa_alloc is
+ * NULL. If more new xa_node are needed, the function gives EINVAL error.
+ *
+ * Context: Any context. The caller should hold the xa_lock.
+ */
+void xas_try_split(struct xa_state *xas, void *entry, unsigned int order)
+{
+ unsigned int sibs = (1 << (order % XA_CHUNK_SHIFT)) - 1;
+ unsigned int offset, marks;
+ struct xa_node *node;
+ void *curr = xas_load(xas);
+ int values = 0;
+ gfp_t gfp = GFP_NOWAIT;
+
+ node = xas->xa_node;
+ if (xas_top(node))
+ return;
+
+ if (xas->xa->xa_flags & XA_FLAGS_ACCOUNT)
+ gfp |= __GFP_ACCOUNT;
+
+ marks = node_get_marks(node, xas->xa_offset);
+
+ offset = xas->xa_offset + sibs;
+
+ if (xas->xa_shift < node->shift) {
+ struct xa_node *child = xas->xa_alloc;
+ unsigned int expected_sibs =
+ (1 << ((order - 1) % XA_CHUNK_SHIFT)) - 1;
+
+ /*
+ * No support for splitting sibling entries
+ * (horizontally) or cascade split (vertically), which
+ * requires two or more new xa_nodes.
+ * Since if one xa_node allocation fails,
+ * it is hard to free the prior allocations.
+ */
+ if (sibs || xas->xa_sibs != expected_sibs) {
+ xas_destroy(xas);
+ xas_set_err(xas, -EINVAL);
+ return;
+ }
+
+ if (!child) {
+ child = kmem_cache_alloc_lru(radix_tree_node_cachep,
+ xas->xa_lru, gfp);
+ if (!child) {
+ xas_destroy(xas);
+ xas_set_err(xas, -ENOMEM);
+ return;
+ }
+ RCU_INIT_POINTER(child->parent, xas->xa_alloc);
+ }
+ __xas_init_node_for_split(xas, child, entry);
+
+ xas->xa_alloc = rcu_dereference_raw(child->parent);
+ child->shift = node->shift - XA_CHUNK_SHIFT;
+ child->offset = offset;
+ child->count = XA_CHUNK_SIZE;
+ child->nr_values = xa_is_value(entry) ?
+ XA_CHUNK_SIZE : 0;
+ RCU_INIT_POINTER(child->parent, node);
+ node_set_marks(node, offset, child, xas->xa_sibs,
+ marks);
+ rcu_assign_pointer(node->slots[offset],
+ xa_mk_node(child));
+ if (xa_is_value(curr))
+ values--;
+ xas_update(xas, child);
+
+ } else {
+ do {
+ unsigned int canon = offset - xas->xa_sibs;
+
+ node_set_marks(node, canon, NULL, 0, marks);
+ rcu_assign_pointer(node->slots[canon], entry);
+ while (offset > canon)
+ rcu_assign_pointer(node->slots[offset--],
+ xa_mk_sibling(canon));
+ values += (xa_is_value(entry) - xa_is_value(curr)) *
+ (xas->xa_sibs + 1);
+ } while (offset-- > xas->xa_offset);
+ }
+
+ node->nr_values += values;
+ xas_update(xas, node);
+}
+EXPORT_SYMBOL_GPL(xas_try_split);
#endif
/**
diff --git a/tools/testing/radix-tree/Makefile b/tools/testing/radix-tree/Makefile
index 8b3591a51e1f..b2a6660bbd92 100644
--- a/tools/testing/radix-tree/Makefile
+++ b/tools/testing/radix-tree/Makefile
@@ -14,6 +14,7 @@ include ../shared/shared.mk
main: $(OFILES)
+xarray.o: ../../../lib/test_xarray.c
idr-test.o: ../../../lib/test_ida.c
idr-test: idr-test.o $(CORE_OFILES)
--
2.47.2
next prev parent reply other threads:[~2025-03-07 17:40 UTC|newest]
Thread overview: 17+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-03-07 17:39 [PATCH v10 0/8] Buddy allocator like (or non-uniform) folio split Zi Yan
2025-03-07 17:39 ` Zi Yan [this message]
2025-03-07 17:39 ` [PATCH v10 2/8] mm/huge_memory: add two new (not yet used) functions for folio_split() Zi Yan
2025-03-10 16:14 ` Zi Yan
2025-03-10 17:32 ` Zi Yan
2025-03-13 14:08 ` Zi Yan
2025-03-10 16:30 ` Matthew Wilcox
2025-03-10 16:39 ` Zi Yan
2025-03-10 16:42 ` Zi Yan
2025-03-10 17:00 ` Matthew Wilcox
2025-03-10 17:05 ` Zi Yan
2025-03-07 17:39 ` [PATCH v10 3/8] mm/huge_memory: move folio split common code to __folio_split() Zi Yan
2025-03-07 17:39 ` [PATCH v10 4/8] mm/huge_memory: add buddy allocator like (non-uniform) folio_split() Zi Yan
2025-03-07 17:39 ` [PATCH v10 5/8] mm/huge_memory: remove the old, unused __split_huge_page() Zi Yan
2025-03-07 17:39 ` [PATCH v10 6/8] mm/huge_memory: add folio_split() to debugfs testing interface Zi Yan
2025-03-07 17:40 ` [PATCH v10 7/8] mm/truncate: use folio_split() in truncate operation Zi Yan
2025-03-07 17:40 ` [PATCH v10 8/8] selftests/mm: add tests for folio_split(), buddy allocator like split Zi Yan
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20250307174001.242794-2-ziy@nvidia.com \
--to=ziy@nvidia.com \
--cc=akpm@linux-foundation.org \
--cc=baolin.wang@linux.alibaba.com \
--cc=david@redhat.com \
--cc=hughd@google.com \
--cc=jhubbard@nvidia.com \
--cc=kasong@tencent.com \
--cc=kirill.shutemov@linux.intel.com \
--cc=linmiaohe@huawei.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-kselftest@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=ryan.roberts@arm.com \
--cc=wangkefeng.wang@huawei.com \
--cc=willy@infradead.org \
--cc=yang@os.amperecomputing.com \
--cc=yuzhao@google.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.