From: "Morten Brørup" <mb@smartsharesystems.com>
To: dev@dpdk.org, Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>,
Bruce Richardson <bruce.richardson@intel.com>,
Stephen Hemminger <stephen@networkplumber.org>
Cc: "Morten Brørup" <mb@smartsharesystems.com>
Subject: [PATCH v5 2/2] mempool: introduce pile driver
Date: Mon, 14 Sep 2026 10:45:13 +0000 [thread overview]
Message-ID: <20260914104513.906736-3-mb@smartsharesystems.com> (raw)
In-Reply-To: <20260914104513.906736-1-mb@smartsharesystems.com>
Added a new "pile" mempool driver, based on the high-performance lock-free
"pile" stack variant.
With this driver added, the max number of mempool drivers
(RTE_MEMPOOL_MAX_OPS_IDX) would be exceeded, so it was increased from 16
to 32.
Changed the other stack drivers to call their specific push/pop functions,
instead of calling the generic stack push/pop API.
Signed-off-by: Morten Brørup <mb@smartsharesystems.com>
---
v5:
* Added release note. (AI advanced model)
* Moved increase of max number of mempool drivers from stack patch to
mempool driver patch, where it belongs.
* Mentioned increase of max number of mempool drivers in patch
description.
* Removed the rte_mempool_register_ops() and rte_mempool_set_ops_byname()
changes.
They are included in the "mempool: various cleanups" patch in the
"mempool: cleanup, fixes, improvements and optimizations" series.
v4:
* No changes.
v3:
* Call other stack types' specific push/pop functions.
v2:
* There is no v2.
---
doc/guides/mempool/stack.rst | 12 ++-
doc/guides/rel_notes/release_26_11.rst | 1 +
drivers/mempool/stack/rte_mempool_stack.c | 90 ++++++++++++++++++++---
lib/mempool/rte_mempool.h | 2 +-
4 files changed, 93 insertions(+), 12 deletions(-)
diff --git a/doc/guides/mempool/stack.rst b/doc/guides/mempool/stack.rst
index 80ea07e65d..c06ab2dc56 100644
--- a/doc/guides/mempool/stack.rst
+++ b/doc/guides/mempool/stack.rst
@@ -1,5 +1,6 @@
.. SPDX-License-Identifier: BSD-3-Clause
Copyright(c) 2020 Intel Corporation.
+ Copyright(c) 2026 SmartShare Systems.
Stack Mempool Driver
====================
@@ -28,6 +29,12 @@ can be selected as described in :ref:`Mempool_Handlers`:
The underlying **rte_stack** operates in lock-free mode. For more
information please refer to :ref:`Stack_Library_LF_Stack`.
+- ``pile``
+
+ The underlying **rte_stack** operates in lock-free mode,
+ and is optimized for bulks of objects.
+ For more information please refer to :ref:`Stack_Library_Pile`.
+
The standard stack outperforms the lock-free stack on average, however the
standard stack is non-preemptive: if a mempool user is preempted while holding
the stack lock, that thread will block all other mempool accesses until it
@@ -35,9 +42,12 @@ returns and releases the lock. As a result, an application using the standard
stack whose threads can be preempted can suffer from brief, infrequent
performance hiccups.
-The lock-free stack, by design, is not susceptible to this problem; one thread can
+The lock-free stack and the pile, by design, are not susceptible to this problem; one thread can
be preempted at any point during a push or pop operation and will not impede
the progress of any other thread.
+The pile is not LIFO per object, but per bulk of objects.
+Although the pile is optimized for bulks of objects, it can handle any request size.
+
For a more detailed description of the stack implementations, please refer to
:doc:`/prog_guide/stack_lib`.
diff --git a/doc/guides/rel_notes/release_26_11.rst b/doc/guides/rel_notes/release_26_11.rst
index 572e579ed5..17a1219222 100644
--- a/doc/guides/rel_notes/release_26_11.rst
+++ b/doc/guides/rel_notes/release_26_11.rst
@@ -61,6 +61,7 @@ New Features
pushed and popped in LIFO manner, but objects within each bulk are not ordered
as expected by a stack.
Furthermore, it is not strictly bounded by its size, but might hold more objects.
+* mempool: Added "pile" driver, using the lock-free "pile" stack-like implementation.
Removed Items
-------------
diff --git a/drivers/mempool/stack/rte_mempool_stack.c b/drivers/mempool/stack/rte_mempool_stack.c
index 1476905227..11c1243bec 100644
--- a/drivers/mempool/stack/rte_mempool_stack.c
+++ b/drivers/mempool/stack/rte_mempool_stack.c
@@ -30,7 +30,7 @@ __stack_alloc(struct rte_mempool *mp, uint32_t flags)
}
static int
-stack_alloc(struct rte_mempool *mp)
+std_stack_alloc(struct rte_mempool *mp)
{
return __stack_alloc(mp, 0);
}
@@ -42,21 +42,81 @@ lf_stack_alloc(struct rte_mempool *mp)
}
static int
-stack_enqueue(struct rte_mempool *mp, void * const *obj_table,
+pile_alloc(struct rte_mempool *mp)
+{
+ return __stack_alloc(mp, RTE_STACK_F_PILE);
+}
+
+static int
+std_stack_enqueue(struct rte_mempool *mp, void * const *obj_table,
+ unsigned int n)
+{
+ struct rte_stack *s = mp->pool_data;
+
+ RTE_ASSERT(s != NULL);
+ RTE_ASSERT(obj_table != NULL);
+
+ return __rte_stack_std_push(s, obj_table, n) == 0 ? -ENOBUFS : 0;
+}
+
+static int
+std_stack_dequeue(struct rte_mempool *mp, void **obj_table,
+ unsigned int n)
+{
+ struct rte_stack *s = mp->pool_data;
+
+ RTE_ASSERT(s != NULL);
+ RTE_ASSERT(obj_table != NULL);
+
+ return __rte_stack_std_pop(s, obj_table, n) == 0 ? -ENOBUFS : 0;
+}
+
+static int
+lf_stack_enqueue(struct rte_mempool *mp, void * const *obj_table,
+ unsigned int n)
+{
+ struct rte_stack *s = mp->pool_data;
+
+ RTE_ASSERT(s != NULL);
+ RTE_ASSERT(obj_table != NULL);
+
+ return __rte_stack_lf_push(s, obj_table, n) == 0 ? -ENOBUFS : 0;
+}
+
+static int
+lf_stack_dequeue(struct rte_mempool *mp, void **obj_table,
unsigned int n)
{
struct rte_stack *s = mp->pool_data;
- return rte_stack_push(s, obj_table, n) == 0 ? -ENOBUFS : 0;
+ RTE_ASSERT(s != NULL);
+ RTE_ASSERT(obj_table != NULL);
+
+ return __rte_stack_lf_pop(s, obj_table, n) == 0 ? -ENOBUFS : 0;
}
static int
-stack_dequeue(struct rte_mempool *mp, void **obj_table,
+pile_enqueue(struct rte_mempool *mp, void * const *obj_table,
unsigned int n)
{
struct rte_stack *s = mp->pool_data;
- return rte_stack_pop(s, obj_table, n) == 0 ? -ENOBUFS : 0;
+ RTE_ASSERT(s != NULL);
+ RTE_ASSERT(obj_table != NULL);
+
+ return __rte_stack_pile_push(s, obj_table, n) == 0 ? -ENOBUFS : 0;
+}
+
+static int
+pile_dequeue(struct rte_mempool *mp, void **obj_table,
+ unsigned int n)
+{
+ struct rte_stack *s = mp->pool_data;
+
+ RTE_ASSERT(s != NULL);
+ RTE_ASSERT(obj_table != NULL);
+
+ return __rte_stack_pile_pop(s, obj_table, n) == 0 ? -ENOBUFS : 0;
}
static unsigned
@@ -77,10 +137,10 @@ stack_free(struct rte_mempool *mp)
static struct rte_mempool_ops ops_stack = {
.name = "stack",
- .alloc = stack_alloc,
+ .alloc = std_stack_alloc,
.free = stack_free,
- .enqueue = stack_enqueue,
- .dequeue = stack_dequeue,
+ .enqueue = std_stack_enqueue,
+ .dequeue = std_stack_dequeue,
.get_count = stack_get_count
};
@@ -88,10 +148,20 @@ static struct rte_mempool_ops ops_lf_stack = {
.name = "lf_stack",
.alloc = lf_stack_alloc,
.free = stack_free,
- .enqueue = stack_enqueue,
- .dequeue = stack_dequeue,
+ .enqueue = lf_stack_enqueue,
+ .dequeue = lf_stack_dequeue,
+ .get_count = stack_get_count
+};
+
+static struct rte_mempool_ops ops_pile = {
+ .name = "pile",
+ .alloc = pile_alloc,
+ .free = stack_free,
+ .enqueue = pile_enqueue,
+ .dequeue = pile_dequeue,
.get_count = stack_get_count
};
RTE_MEMPOOL_REGISTER_OPS(ops_stack);
RTE_MEMPOOL_REGISTER_OPS(ops_lf_stack);
+RTE_MEMPOOL_REGISTER_OPS(ops_pile);
diff --git a/lib/mempool/rte_mempool.h b/lib/mempool/rte_mempool.h
index 50d958c7c6..3e161bfdb9 100644
--- a/lib/mempool/rte_mempool.h
+++ b/lib/mempool/rte_mempool.h
@@ -718,7 +718,7 @@ struct __rte_cache_aligned rte_mempool_ops {
rte_mempool_dequeue_contig_blocks_t dequeue_contig_blocks;
};
-#define RTE_MEMPOOL_MAX_OPS_IDX 16 /**< Max registered ops structs */
+#define RTE_MEMPOOL_MAX_OPS_IDX 32 /**< Max registered ops structs */
/**
* Structure storing the table of registered ops structs, each of which contain
--
2.43.0
next prev parent reply other threads:[~2026-09-14 10:45 UTC|newest]
Thread overview: 38+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-08-12 13:47 [PATCH] stack: introduce pile Morten Brørup
2026-08-12 14:34 ` Bruce Richardson
2026-08-12 16:01 ` Morten Brørup
2026-08-12 16:15 ` Bruce Richardson
2026-08-12 16:28 ` Morten Brørup
2026-08-13 11:50 ` Bruce Richardson
2026-08-17 13:10 ` Bruce Richardson
2026-08-18 8:11 ` Konstantin Ananyev
2026-08-18 8:50 ` Morten Brørup
2026-08-25 7:05 ` Konstantin Ananyev
2026-08-25 9:21 ` Morten Brørup
2026-08-25 11:29 ` Konstantin Ananyev
2026-08-26 8:13 ` Konstantin Ananyev
2026-08-27 13:55 ` [PATCH v3 0/2] introduce pile stack and mempool driver Morten Brørup
2026-08-27 13:55 ` [PATCH v3 1/2] stack: introduce pile Morten Brørup
2026-08-31 8:50 ` Konstantin Ananyev
2026-08-31 9:05 ` Morten Brørup
2026-08-31 9:38 ` Konstantin Ananyev
2026-08-31 10:09 ` Morten Brørup
2026-08-31 16:06 ` Stephen Hemminger
2026-08-31 16:37 ` Morten Brørup
2026-08-27 13:55 ` [PATCH v3 2/2] mempool: introduce pile driver Morten Brørup
2026-09-01 6:43 ` [PATCH v4 0/2] introduce pile stack and mempool driver Morten Brørup
2026-09-01 6:43 ` [PATCH v4 1/2] stack: introduce pile Morten Brørup
2026-09-01 6:43 ` [PATCH v4 2/2] mempool: introduce pile driver Morten Brørup
2026-09-02 16:00 ` Stephen Hemminger
2026-09-14 10:45 ` [PATCH v5 0/2] introduce pile stack and mempool driver Morten Brørup
2026-09-14 10:45 ` [PATCH v5 1/2] stack: introduce pile Morten Brørup
2026-09-14 10:45 ` Morten Brørup [this message]
2026-09-14 14:44 ` [PATCH v6 0/2] introduce pile stack and mempool driver Morten Brørup
2026-09-14 14:44 ` [PATCH v6 1/2] stack: introduce pile Morten Brørup
2026-09-14 14:44 ` [PATCH v6 2/2] mempool: introduce pile driver Morten Brørup
2026-09-14 15:58 ` [PATCH v7 0/2] introduce pile stack and mempool driver Morten Brørup
2026-09-14 15:59 ` [PATCH v7 1/2] stack: introduce pile Morten Brørup
2026-09-14 15:59 ` [PATCH v7 2/2] mempool: introduce pile driver Morten Brørup
2026-10-15 17:22 ` [PATCH v7 0/2] introduce pile stack and mempool driver Thomas Monjalon
2026-10-19 13:56 ` Ali Alnubani
2026-10-19 16:39 ` Morten Brørup
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=20260914104513.906736-3-mb@smartsharesystems.com \
--to=mb@smartsharesystems.com \
--cc=andrew.rybchenko@oktetlabs.ru \
--cc=bruce.richardson@intel.com \
--cc=dev@dpdk.org \
--cc=stephen@networkplumber.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