mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: "Liam R. Howlett (Oracle)" <liam@infradead.org>
To: Andrew Morton <akpm@linux-foundation.org>
Cc: linux-mm@kvack.org, linux-kernel@vger.kernel.org,
	maple-tree@lists.infradead.org,
	"Liam R. Howlett (Oracle)" <liam@infradead.org>
Subject: [PATCH 09/18] maple_tree: Optimise mas_wr_node_store() when not in rcu mode
Date: Mon, 29 Jun 2026 10:41:36 -0400	[thread overview]
Message-ID: <20260629144145.1572283-10-liam@infradead.org> (raw)
In-Reply-To: <20260629144145.1572283-1-liam@infradead.org>

Clearing the entire node on the stack is unnecessary since most of the
node will be overwritten anyways.  Just clear what isn't used after the
data is in place.

Benchmarking shows a speedup of 0.67% on a height 4 tree with 2048
entries.

Signed-off-by: Liam R. Howlett (Oracle) <liam@infradead.org>
---
 lib/maple_tree.c | 16 +++++++++++++---
 1 file changed, 13 insertions(+), 3 deletions(-)

diff --git a/lib/maple_tree.c b/lib/maple_tree.c
index fe10ff83e4937..a2b96ef05cb32 100644
--- a/lib/maple_tree.c
+++ b/lib/maple_tree.c
@@ -3183,7 +3183,7 @@ static void mas_wr_spanning_store(struct ma_wr_state *wr_mas)
 static inline void mas_wr_node_store(struct ma_wr_state *wr_mas)
 {
 	unsigned char dst_offset, offset_end;
-	unsigned char copy_size, node_pivots;
+	unsigned char copy_size, node_pivots, node_slots;
 	struct maple_node reuse, *newnode;
 	unsigned long *dst_pivots;
 	void __rcu **dst_slots;
@@ -3196,6 +3196,7 @@ static inline void mas_wr_node_store(struct ma_wr_state *wr_mas)
 	in_rcu = mt_in_rcu(mas->tree);
 	offset_end = wr_mas->offset_end;
 	node_pivots = mt_pivots[wr_mas->type];
+	node_slots = mt_slots[wr_mas->type];
 	/* Assume last adds an entry */
 	new_end = mas->end + 1 - offset_end + mas->offset;
 	if (mas->last == wr_mas->end_piv) {
@@ -3207,7 +3208,6 @@ static inline void mas_wr_node_store(struct ma_wr_state *wr_mas)
 	if (in_rcu) {
 		newnode = mas_pop_node(mas);
 	} else {
-		memset(&reuse, 0, sizeof(struct maple_node));
 		newnode = &reuse;
 	}
 
@@ -3251,7 +3251,17 @@ static inline void mas_wr_node_store(struct ma_wr_state *wr_mas)
 		dst_pivots[new_end] = mas->max;
 
 done:
-	mas_leaf_set_meta(newnode, maple_leaf_64, new_end);
+	if (!in_rcu && new_end + 2 < node_slots) {
+		unsigned char clear_from = new_end + 1;
+
+		memset(dst_slots + clear_from, 0,
+		       sizeof(void __rcu *) * (node_slots - clear_from));
+		if (clear_from < node_pivots)
+			memset(dst_pivots + clear_from, 0,
+			       sizeof(unsigned long) * (node_pivots - clear_from));
+	}
+
+	mas_leaf_set_meta(newnode, wr_mas->type, new_end);
 	if (in_rcu) {
 		struct maple_enode *old_enode = mas->node;
 
-- 
2.47.3


  parent reply	other threads:[~2026-06-29 14:42 UTC|newest]

Thread overview: 21+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-06-29 14:41 [PATCH 00/18] maple_tree: lock checking and clean ups Liam R. Howlett (Oracle)
2026-06-29 14:41 ` [PATCH 01/18] maple_tree: Add rcu locking check when LOCKDEP is enabled Liam R. Howlett (Oracle)
2026-06-29 14:41 ` [PATCH 02/18] locking/lockdep: Add sequence counter to held_lock Liam R. Howlett (Oracle)
2026-06-29 14:41 ` [PATCH 03/18] maple_tree: Add write lock checking with lockdep sequence numbers Liam R. Howlett (Oracle)
2026-06-29 15:03   ` [PATCH fixup] maple_tree: Static define mas_lockdep_map() Liam R. Howlett (Oracle)
2026-06-29 14:41 ` [PATCH 04/18] maple_tree: Documentation fix Liam R. Howlett (Oracle)
2026-06-29 14:41 ` [PATCH 05/18] maple_tree: Drop dead code from mas_extend_spanning_null() Liam R. Howlett (Oracle)
2026-06-29 14:41 ` [PATCH 06/18] maple_tree: Drop MAPLE_ALLOC_SLOTS Liam R. Howlett (Oracle)
2026-06-29 14:41 ` [PATCH 07/18] maple_tree: Clarify comments on mas_nomem() Liam R. Howlett (Oracle)
2026-06-29 14:41 ` [PATCH 08/18] maple_tree: Use prefetched value in mas_wr_store_type() Liam R. Howlett (Oracle)
2026-06-29 14:41 ` Liam R. Howlett (Oracle) [this message]
2026-06-29 14:41 ` [PATCH 10/18] maple_tree: micro optimisation of mas_wr_store_type() Liam R. Howlett (Oracle)
2026-06-29 14:41 ` [PATCH 11/18] maple_tree: Add bulk parent set helper Liam R. Howlett (Oracle)
2026-06-29 15:22   ` [PATCH fixup] maple_tree: Fix dereference of rcu type Liam R. Howlett (Oracle)
2026-06-29 14:41 ` [PATCH 12/18] maple_tree: Catch race in mas_alloc_cyclic() Liam R. Howlett (Oracle)
2026-06-29 14:41 ` [PATCH 13/18] maple_tree: Document that erase may use GFP_KERNEL for allocations Liam R. Howlett (Oracle)
2026-06-29 14:41 ` [PATCH 14/18] maple_tree: WARN_ON_ONCE when allocations fail Liam R. Howlett (Oracle)
2026-06-29 14:41 ` [PATCH 15/18] maple_tree: Document erase and allocations better Liam R. Howlett (Oracle)
2026-06-29 14:41 ` [PATCH 16/18] maple_tree: Change two GFP flags in tests Liam R. Howlett (Oracle)
2026-06-29 14:41 ` [PATCH 17/18] maple_tree: Fix argument name in header Liam R. Howlett (Oracle)
2026-06-29 14:41 ` [PATCH 18/18] maple_tree: Avoid extra gap calculation Liam R. Howlett (Oracle)

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=20260629144145.1572283-10-liam@infradead.org \
    --to=liam@infradead.org \
    --cc=akpm@linux-foundation.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=maple-tree@lists.infradead.org \
    /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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®