linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Hao Li <haoli.tcs@gmail.com>
To: Vlastimil Babka <vbabka@suse.cz>
Cc: Suren Baghdasaryan <surenb@google.com>,
	 "Liam R. Howlett" <Liam.Howlett@oracle.com>,
	Christoph Lameter <cl@gentwo.org>,
	 David Rientjes <rientjes@google.com>,
	Roman Gushchin <roman.gushchin@linux.dev>,
	 Harry Yoo <harry.yoo@oracle.com>,
	Uladzislau Rezki <urezki@gmail.com>,
	 Sidhartha Kumar <sidhartha.kumar@oracle.com>,
	linux-mm@kvack.org, linux-kernel@vger.kernel.org,
	 rcu@vger.kernel.org, maple-tree@lists.infradead.org,
	 Venkat Rao Bagalkote <venkat88@linux.ibm.com>
Subject: [PATCH] slub: add barn_get_full_sheaf() and refine empty-main sheaf
Date: Tue, 2 Dec 2025 16:48:17 +0800	[thread overview]
Message-ID: <fqbhzbjw3nvbylvxayboyug7qayfref4xti7udztygd6zbhcun@b7s2yjihrwos> (raw)
In-Reply-To: <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>

Introduce barn_get_full_sheaf(), a helper that detaches a full sheaf from
the per-node barn without requiring an empty sheaf in exchange.

Use this helper in __pcs_replace_empty_main() to change how an empty main
per-CPU sheaf is handled:

  - If pcs->spare is NULL and pcs->main is empty, first try to obtain a
    full sheaf from the barn via barn_get_full_sheaf(). On success, park
    the empty main sheaf in pcs->spare and install the full sheaf as the
    new pcs->main.

  - If pcs->spare already exists and has objects, keep the existing
    behavior of simply swapping pcs->main and pcs->spare.

  - Only when both pcs->main and pcs->spare are empty do we fall back to
    barn_replace_empty_sheaf() and trade the empty main sheaf into the
    barn in exchange for a full one.

This makes the empty-main path more symmetric with __pcs_replace_full_main(),
which for a full main sheaf parks the full sheaf in pcs->spare and pulls an
empty sheaf from the barn. It also matches the documented design more closely:

  "When both percpu sheaves are found empty during an allocation, an empty
   sheaf may be replaced with a full one from the per-node barn."

Signed-off-by: Hao Li <haoli.tcs@gmail.com>
---

* This patch is based on b4/sheaves-for-all branch

 mm/slub.c | 50 +++++++++++++++++++++++++++++++++++++++++++-------
 1 file changed, 43 insertions(+), 7 deletions(-)

diff --git a/mm/slub.c b/mm/slub.c
index a94c64f56504..1fd28aa204e1 100644
--- a/mm/slub.c
+++ b/mm/slub.c
@@ -2746,6 +2746,32 @@ static void pcs_destroy(struct kmem_cache *s)
 	s->cpu_sheaves = NULL;
 }
 
+static struct slab_sheaf *barn_get_full_sheaf(struct node_barn *barn,
+					       bool allow_spin)
+{
+	struct slab_sheaf *full = NULL;
+	unsigned long flags;
+
+	if (!data_race(barn->nr_full))
+		return NULL;
+
+	if (likely(allow_spin))
+		spin_lock_irqsave(&barn->lock, flags);
+	else if (!spin_trylock_irqsave(&barn->lock, flags))
+		return NULL;
+
+	if (likely(barn->nr_full)) {
+		full = list_first_entry(&barn->sheaves_full,
+					 struct slab_sheaf, barn_list);
+		list_del(&full->barn_list);
+		barn->nr_full--;
+	}
+
+	spin_unlock_irqrestore(&barn->lock, flags);
+
+	return full;
+}
+
 static struct slab_sheaf *barn_get_empty_sheaf(struct node_barn *barn,
 					       bool allow_spin)
 {
@@ -4120,7 +4146,7 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 	struct slab_sheaf *empty = NULL;
 	struct slab_sheaf *full;
 	struct node_barn *barn;
-	bool can_alloc;
+	bool can_alloc, allow_spin;
 
 	lockdep_assert_held(this_cpu_ptr(&s->cpu_sheaves->lock));
 
@@ -4130,10 +4156,7 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 		return NULL;
 	}
 
-	if (pcs->spare && pcs->spare->size > 0) {
-		swap(pcs->main, pcs->spare);
-		return pcs;
-	}
+	allow_spin = gfpflags_allow_spinning(gfp);
 
 	barn = get_barn(s);
 	if (!barn) {
@@ -4141,8 +4164,21 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 		return NULL;
 	}
 
-	full = barn_replace_empty_sheaf(barn, pcs->main,
-					gfpflags_allow_spinning(gfp));
+	if (!pcs->spare) {
+		full = barn_get_full_sheaf(barn, allow_spin);
+		if (full) {
+			pcs->spare = pcs->main;
+			pcs->main = full;
+			return pcs;
+		}
+	} else if (pcs->spare->size > 0) {
+		swap(pcs->main, pcs->spare);
+		return pcs;
+	}
+
+	/* both main and spare are empty */
+
+	full = barn_replace_empty_sheaf(barn, pcs->main, allow_spin);
 
 	if (full) {
 		stat(s, BARN_GET);
-- 
2.50.1


From haoli.tcs@gmail.com Tue Dec  2 16:24:49 2025
Date: Tue, 2 Dec 2025 16:31:49 +0800
From: Hao Li <haoli.tcs@gmail.com>
To: Vlastimil Babka <vbabka@suse.cz>
Cc: Suren Baghdasaryan <surenb@google.com>, 
	"Liam R. Howlett" <Liam.Howlett@oracle.com>, Christoph Lameter <cl@gentwo.org>, 
	David Rientjes <rientjes@google.com>, Roman Gushchin <roman.gushchin@linux.dev>, 
	Harry Yoo <harry.yoo@oracle.com>, Uladzislau Rezki <urezki@gmail.com>, 
	Sidhartha Kumar <sidhartha.kumar@oracle.com>, linux-mm@kvack.org, linux-kernel@vger.kernel.org, 
	rcu@vger.kernel.org, maple-tree@lists.infradead.org, 
	Venkat Rao Bagalkote <venkat88@linux.ibm.com>
Subject: [PATCH] slub: add barn_get_full_sheaf() and refine empty-main sheaf
Message-ID: <fqbhzbjw3nvbylvxayboyug7qayfref4xti7udztygd6zbhcun@b7s2yjihrwos>
Mutt-References: <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
Mutt-Fcc: ~/sent
References: <20250910-slub-percpu-caches-v8-0-ca3099d8352c@suse.cz>
 <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
Mutt-Fcc: ~/sent
Status: RO
Content-Length: 3509
Lines: 119

Introduce barn_get_full_sheaf(), a helper that detaches a full sheaf from
the per-node barn without requiring an empty sheaf in exchange.

Use this helper in __pcs_replace_empty_main() to change how an empty main
per-CPU sheaf is handled:

  - If pcs->spare is NULL and pcs->main is empty, first try to obtain a
    full sheaf from the barn via barn_get_full_sheaf(). On success, park
    the empty main sheaf in pcs->spare and install the full sheaf as the
    new pcs->main.

  - If pcs->spare already exists and has objects, keep the existing
    behavior of simply swapping pcs->main and pcs->spare.

  - Only when both pcs->main and pcs->spare are empty do we fall back to
    barn_replace_empty_sheaf() and trade the empty main sheaf into the
    barn in exchange for a full one.

This makes the empty-main path more symmetric with __pcs_replace_full_main(),
which for a full main sheaf parks the full sheaf in pcs->spare and pulls an
empty sheaf from the barn. It also matches the documented design more closely:

  "When both percpu sheaves are found empty during an allocation, an empty
   sheaf may be replaced with a full one from the per-node barn."

Signed-off-by: Hao Li <haoli.tcs@gmail.com>
---

* This patch is based on b4/sheaves-for-all branch

 mm/slub.c | 50 +++++++++++++++++++++++++++++++++++++++++++-------
 1 file changed, 43 insertions(+), 7 deletions(-)

diff --git a/mm/slub.c b/mm/slub.c
index a94c64f56504..1fd28aa204e1 100644
--- a/mm/slub.c
+++ b/mm/slub.c
@@ -2746,6 +2746,32 @@ static void pcs_destroy(struct kmem_cache *s)
 	s->cpu_sheaves = NULL;
 }
 
+static struct slab_sheaf *barn_get_full_sheaf(struct node_barn *barn,
+					       bool allow_spin)
+{
+	struct slab_sheaf *full = NULL;
+	unsigned long flags;
+
+	if (!data_race(barn->nr_full))
+		return NULL;
+
+	if (likely(allow_spin))
+		spin_lock_irqsave(&barn->lock, flags);
+	else if (!spin_trylock_irqsave(&barn->lock, flags))
+		return NULL;
+
+	if (likely(barn->nr_full)) {
+		full = list_first_entry(&barn->sheaves_full,
+					 struct slab_sheaf, barn_list);
+		list_del(&full->barn_list);
+		barn->nr_full--;
+	}
+
+	spin_unlock_irqrestore(&barn->lock, flags);
+
+	return full;
+}
+
 static struct slab_sheaf *barn_get_empty_sheaf(struct node_barn *barn,
 					       bool allow_spin)
 {
@@ -4120,7 +4146,7 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 	struct slab_sheaf *empty = NULL;
 	struct slab_sheaf *full;
 	struct node_barn *barn;
-	bool can_alloc;
+	bool can_alloc, allow_spin;
 
 	lockdep_assert_held(this_cpu_ptr(&s->cpu_sheaves->lock));
 
@@ -4130,10 +4156,7 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 		return NULL;
 	}
 
-	if (pcs->spare && pcs->spare->size > 0) {
-		swap(pcs->main, pcs->spare);
-		return pcs;
-	}
+	allow_spin = gfpflags_allow_spinning(gfp);
 
 	barn = get_barn(s);
 	if (!barn) {
@@ -4141,8 +4164,21 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 		return NULL;
 	}
 
-	full = barn_replace_empty_sheaf(barn, pcs->main,
-					gfpflags_allow_spinning(gfp));
+	if (!pcs->spare) {
+		full = barn_get_full_sheaf(barn, allow_spin);
+		if (full) {
+			pcs->spare = pcs->main;
+			pcs->main = full;
+			return pcs;
+		}
+	} else if (pcs->spare->size > 0) {
+		swap(pcs->main, pcs->spare);
+		return pcs;
+	}
+
+	/* both main and spare are empty */
+
+	full = barn_replace_empty_sheaf(barn, pcs->main, allow_spin);
 
 	if (full) {
 		stat(s, BARN_GET);
-- 
2.50.1



From haoli.tcs@gmail.com Tue Dec  2 16:24:49 2025
Date: Tue, 2 Dec 2025 16:33:21 +0800
From: Hao Li <haoli.tcs@gmail.com>
To: Vlastimil Babka <vbabka@suse.cz>
Cc: Suren Baghdasaryan <surenb@google.com>, 
	"Liam R. Howlett" <Liam.Howlett@oracle.com>, Christoph Lameter <cl@gentwo.org>, 
	David Rientjes <rientjes@google.com>, Roman Gushchin <roman.gushchin@linux.dev>, 
	Harry Yoo <harry.yoo@oracle.com>, Uladzislau Rezki <urezki@gmail.com>, 
	Sidhartha Kumar <sidhartha.kumar@oracle.com>, linux-mm@kvack.org, linux-kernel@vger.kernel.org, 
	rcu@vger.kernel.org, maple-tree@lists.infradead.org, 
	Venkat Rao Bagalkote <venkat88@linux.ibm.com>
Subject: [PATCH] slub: add barn_get_full_sheaf() and refine empty-main sheaf
Message-ID: <fqbhzbjw3nvbylvxayboyug7qayfref4xti7udztygd6zbhcun@b7s2yjihrwos>
Mutt-References: <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
Mutt-Fcc: ~/sent
References: <20250910-slub-percpu-caches-v8-0-ca3099d8352c@suse.cz>
 <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
Mutt-Fcc: ~/sent
Status: RO
Content-Length: 8250
Lines: 265

Introduce barn_get_full_sheaf(), a helper that detaches a full sheaf from
the per-node barn without requiring an empty sheaf in exchange.

Use this helper in __pcs_replace_empty_main() to change how an empty main
per-CPU sheaf is handled:

  - If pcs->spare is NULL and pcs->main is empty, first try to obtain a
    full sheaf from the barn via barn_get_full_sheaf(). On success, park
    the empty main sheaf in pcs->spare and install the full sheaf as the
    new pcs->main.

  - If pcs->spare already exists and has objects, keep the existing
    behavior of simply swapping pcs->main and pcs->spare.

  - Only when both pcs->main and pcs->spare are empty do we fall back to
    barn_replace_empty_sheaf() and trade the empty main sheaf into the
    barn in exchange for a full one.

This makes the empty-main path more symmetric with __pcs_replace_full_main(),
which for a full main sheaf parks the full sheaf in pcs->spare and pulls an
empty sheaf from the barn. It also matches the documented design more closely:

  "When both percpu sheaves are found empty during an allocation, an empty
   sheaf may be replaced with a full one from the per-node barn."

Signed-off-by: Hao Li <haoli.tcs@gmail.com>
---

* This patch is based on b4/sheaves-for-all branch

 mm/slub.c | 50 +++++++++++++++++++++++++++++++++++++++++++-------
 1 file changed, 43 insertions(+), 7 deletions(-)

diff --git a/mm/slub.c b/mm/slub.c
index a94c64f56504..1fd28aa204e1 100644
--- a/mm/slub.c
+++ b/mm/slub.c
@@ -2746,6 +2746,32 @@ static void pcs_destroy(struct kmem_cache *s)
 	s->cpu_sheaves = NULL;
 }
 
+static struct slab_sheaf *barn_get_full_sheaf(struct node_barn *barn,
+					       bool allow_spin)
+{
+	struct slab_sheaf *full = NULL;
+	unsigned long flags;
+
+	if (!data_race(barn->nr_full))
+		return NULL;
+
+	if (likely(allow_spin))
+		spin_lock_irqsave(&barn->lock, flags);
+	else if (!spin_trylock_irqsave(&barn->lock, flags))
+		return NULL;
+
+	if (likely(barn->nr_full)) {
+		full = list_first_entry(&barn->sheaves_full,
+					 struct slab_sheaf, barn_list);
+		list_del(&full->barn_list);
+		barn->nr_full--;
+	}
+
+	spin_unlock_irqrestore(&barn->lock, flags);
+
+	return full;
+}
+
 static struct slab_sheaf *barn_get_empty_sheaf(struct node_barn *barn,
 					       bool allow_spin)
 {
@@ -4120,7 +4146,7 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 	struct slab_sheaf *empty = NULL;
 	struct slab_sheaf *full;
 	struct node_barn *barn;
-	bool can_alloc;
+	bool can_alloc, allow_spin;
 
 	lockdep_assert_held(this_cpu_ptr(&s->cpu_sheaves->lock));
 
@@ -4130,10 +4156,7 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 		return NULL;
 	}
 
-	if (pcs->spare && pcs->spare->size > 0) {
-		swap(pcs->main, pcs->spare);
-		return pcs;
-	}
+	allow_spin = gfpflags_allow_spinning(gfp);
 
 	barn = get_barn(s);
 	if (!barn) {
@@ -4141,8 +4164,21 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 		return NULL;
 	}
 
-	full = barn_replace_empty_sheaf(barn, pcs->main,
-					gfpflags_allow_spinning(gfp));
+	if (!pcs->spare) {
+		full = barn_get_full_sheaf(barn, allow_spin);
+		if (full) {
+			pcs->spare = pcs->main;
+			pcs->main = full;
+			return pcs;
+		}
+	} else if (pcs->spare->size > 0) {
+		swap(pcs->main, pcs->spare);
+		return pcs;
+	}
+
+	/* both main and spare are empty */
+
+	full = barn_replace_empty_sheaf(barn, pcs->main, allow_spin);
 
 	if (full) {
 		stat(s, BARN_GET);
-- 
2.50.1


From haoli.tcs@gmail.com Tue Dec  2 16:24:49 2025
Date: Tue, 2 Dec 2025 16:31:49 +0800
From: Hao Li <haoli.tcs@gmail.com>
To: Vlastimil Babka <vbabka@suse.cz>
Cc: Suren Baghdasaryan <surenb@google.com>, 
	"Liam R. Howlett" <Liam.Howlett@oracle.com>, Christoph Lameter <cl@gentwo.org>, 
	David Rientjes <rientjes@google.com>, Roman Gushchin <roman.gushchin@linux.dev>, 
	Harry Yoo <harry.yoo@oracle.com>, Uladzislau Rezki <urezki@gmail.com>, 
	Sidhartha Kumar <sidhartha.kumar@oracle.com>, linux-mm@kvack.org, linux-kernel@vger.kernel.org, 
	rcu@vger.kernel.org, maple-tree@lists.infradead.org, 
	Venkat Rao Bagalkote <venkat88@linux.ibm.com>
Subject: [PATCH] slub: add barn_get_full_sheaf() and refine empty-main sheaf
Message-ID: <fqbhzbjw3nvbylvxayboyug7qayfref4xti7udztygd6zbhcun@b7s2yjihrwos>
Mutt-References: <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
Mutt-Fcc: ~/sent
References: <20250910-slub-percpu-caches-v8-0-ca3099d8352c@suse.cz>
 <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
Mutt-Fcc: ~/sent
Status: RO
Content-Length: 3509
Lines: 119

Introduce barn_get_full_sheaf(), a helper that detaches a full sheaf from
the per-node barn without requiring an empty sheaf in exchange.

Use this helper in __pcs_replace_empty_main() to change how an empty main
per-CPU sheaf is handled:

  - If pcs->spare is NULL and pcs->main is empty, first try to obtain a
    full sheaf from the barn via barn_get_full_sheaf(). On success, park
    the empty main sheaf in pcs->spare and install the full sheaf as the
    new pcs->main.

  - If pcs->spare already exists and has objects, keep the existing
    behavior of simply swapping pcs->main and pcs->spare.

  - Only when both pcs->main and pcs->spare are empty do we fall back to
    barn_replace_empty_sheaf() and trade the empty main sheaf into the
    barn in exchange for a full one.

This makes the empty-main path more symmetric with __pcs_replace_full_main(),
which for a full main sheaf parks the full sheaf in pcs->spare and pulls an
empty sheaf from the barn. It also matches the documented design more closely:

  "When both percpu sheaves are found empty during an allocation, an empty
   sheaf may be replaced with a full one from the per-node barn."

Signed-off-by: Hao Li <haoli.tcs@gmail.com>
---

* This patch is based on b4/sheaves-for-all branch

 mm/slub.c | 50 +++++++++++++++++++++++++++++++++++++++++++-------
 1 file changed, 43 insertions(+), 7 deletions(-)

diff --git a/mm/slub.c b/mm/slub.c
index a94c64f56504..1fd28aa204e1 100644
--- a/mm/slub.c
+++ b/mm/slub.c
@@ -2746,6 +2746,32 @@ static void pcs_destroy(struct kmem_cache *s)
 	s->cpu_sheaves = NULL;
 }
 
+static struct slab_sheaf *barn_get_full_sheaf(struct node_barn *barn,
+					       bool allow_spin)
+{
+	struct slab_sheaf *full = NULL;
+	unsigned long flags;
+
+	if (!data_race(barn->nr_full))
+		return NULL;
+
+	if (likely(allow_spin))
+		spin_lock_irqsave(&barn->lock, flags);
+	else if (!spin_trylock_irqsave(&barn->lock, flags))
+		return NULL;
+
+	if (likely(barn->nr_full)) {
+		full = list_first_entry(&barn->sheaves_full,
+					 struct slab_sheaf, barn_list);
+		list_del(&full->barn_list);
+		barn->nr_full--;
+	}
+
+	spin_unlock_irqrestore(&barn->lock, flags);
+
+	return full;
+}
+
 static struct slab_sheaf *barn_get_empty_sheaf(struct node_barn *barn,
 					       bool allow_spin)
 {
@@ -4120,7 +4146,7 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 	struct slab_sheaf *empty = NULL;
 	struct slab_sheaf *full;
 	struct node_barn *barn;
-	bool can_alloc;
+	bool can_alloc, allow_spin;
 
 	lockdep_assert_held(this_cpu_ptr(&s->cpu_sheaves->lock));
 
@@ -4130,10 +4156,7 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 		return NULL;
 	}
 
-	if (pcs->spare && pcs->spare->size > 0) {
-		swap(pcs->main, pcs->spare);
-		return pcs;
-	}
+	allow_spin = gfpflags_allow_spinning(gfp);
 
 	barn = get_barn(s);
 	if (!barn) {
@@ -4141,8 +4164,21 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 		return NULL;
 	}
 
-	full = barn_replace_empty_sheaf(barn, pcs->main,
-					gfpflags_allow_spinning(gfp));
+	if (!pcs->spare) {
+		full = barn_get_full_sheaf(barn, allow_spin);
+		if (full) {
+			pcs->spare = pcs->main;
+			pcs->main = full;
+			return pcs;
+		}
+	} else if (pcs->spare->size > 0) {
+		swap(pcs->main, pcs->spare);
+		return pcs;
+	}
+
+	/* both main and spare are empty */
+
+	full = barn_replace_empty_sheaf(barn, pcs->main, allow_spin);
 
 	if (full) {
 		stat(s, BARN_GET);
-- 
2.50.1




From haoli.tcs@gmail.com Tue Dec  2 16:24:49 2025
Date: Tue, 2 Dec 2025 16:44:16 +0800
From: Hao Li <haoli.tcs@gmail.com>
To: Vlastimil Babka <vbabka@suse.cz>
Cc: Suren Baghdasaryan <surenb@google.com>, 
	"Liam R. Howlett" <Liam.Howlett@oracle.com>, Christoph Lameter <cl@gentwo.org>, 
	David Rientjes <rientjes@google.com>, Roman Gushchin <roman.gushchin@linux.dev>, 
	Harry Yoo <harry.yoo@oracle.com>, Uladzislau Rezki <urezki@gmail.com>, 
	Sidhartha Kumar <sidhartha.kumar@oracle.com>, linux-mm@kvack.org, linux-kernel@vger.kernel.org, 
	rcu@vger.kernel.org, maple-tree@lists.infradead.org, 
	Venkat Rao Bagalkote <venkat88@linux.ibm.com>
Subject: [PATCH] slub: add barn_get_full_sheaf() and refine empty-main sheaf
Message-ID: <fqbhzbjw3nvbylvxayboyug7qayfref4xti7udztygd6zbhcun@b7s2yjihrwos>
Mutt-References: <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
Mutt-Fcc: ~/sent
References: <20250910-slub-percpu-caches-v8-0-ca3099d8352c@suse.cz>
 <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
Mutt-Fcc: ~/sent
Status: RO
Content-Length: 17732
Lines: 557

Introduce barn_get_full_sheaf(), a helper that detaches a full sheaf from
the per-node barn without requiring an empty sheaf in exchange.

Use this helper in __pcs_replace_empty_main() to change how an empty main
per-CPU sheaf is handled:

  - If pcs->spare is NULL and pcs->main is empty, first try to obtain a
    full sheaf from the barn via barn_get_full_sheaf(). On success, park
    the empty main sheaf in pcs->spare and install the full sheaf as the
    new pcs->main.

  - If pcs->spare already exists and has objects, keep the existing
    behavior of simply swapping pcs->main and pcs->spare.

  - Only when both pcs->main and pcs->spare are empty do we fall back to
    barn_replace_empty_sheaf() and trade the empty main sheaf into the
    barn in exchange for a full one.

This makes the empty-main path more symmetric with __pcs_replace_full_main(),
which for a full main sheaf parks the full sheaf in pcs->spare and pulls an
empty sheaf from the barn. It also matches the documented design more closely:

  "When both percpu sheaves are found empty during an allocation, an empty
   sheaf may be replaced with a full one from the per-node barn."

Signed-off-by: Hao Li <haoli.tcs@gmail.com>
---

* This patch is based on b4/sheaves-for-all branch

 mm/slub.c | 50 +++++++++++++++++++++++++++++++++++++++++++-------
 1 file changed, 43 insertions(+), 7 deletions(-)

diff --git a/mm/slub.c b/mm/slub.c
index a94c64f56504..1fd28aa204e1 100644
--- a/mm/slub.c
+++ b/mm/slub.c
@@ -2746,6 +2746,32 @@ static void pcs_destroy(struct kmem_cache *s)
 	s->cpu_sheaves = NULL;
 }
 
+static struct slab_sheaf *barn_get_full_sheaf(struct node_barn *barn,
+					       bool allow_spin)
+{
+	struct slab_sheaf *full = NULL;
+	unsigned long flags;
+
+	if (!data_race(barn->nr_full))
+		return NULL;
+
+	if (likely(allow_spin))
+		spin_lock_irqsave(&barn->lock, flags);
+	else if (!spin_trylock_irqsave(&barn->lock, flags))
+		return NULL;
+
+	if (likely(barn->nr_full)) {
+		full = list_first_entry(&barn->sheaves_full,
+					 struct slab_sheaf, barn_list);
+		list_del(&full->barn_list);
+		barn->nr_full--;
+	}
+
+	spin_unlock_irqrestore(&barn->lock, flags);
+
+	return full;
+}
+
 static struct slab_sheaf *barn_get_empty_sheaf(struct node_barn *barn,
 					       bool allow_spin)
 {
@@ -4120,7 +4146,7 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 	struct slab_sheaf *empty = NULL;
 	struct slab_sheaf *full;
 	struct node_barn *barn;
-	bool can_alloc;
+	bool can_alloc, allow_spin;
 
 	lockdep_assert_held(this_cpu_ptr(&s->cpu_sheaves->lock));
 
@@ -4130,10 +4156,7 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 		return NULL;
 	}
 
-	if (pcs->spare && pcs->spare->size > 0) {
-		swap(pcs->main, pcs->spare);
-		return pcs;
-	}
+	allow_spin = gfpflags_allow_spinning(gfp);
 
 	barn = get_barn(s);
 	if (!barn) {
@@ -4141,8 +4164,21 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 		return NULL;
 	}
 
-	full = barn_replace_empty_sheaf(barn, pcs->main,
-					gfpflags_allow_spinning(gfp));
+	if (!pcs->spare) {
+		full = barn_get_full_sheaf(barn, allow_spin);
+		if (full) {
+			pcs->spare = pcs->main;
+			pcs->main = full;
+			return pcs;
+		}
+	} else if (pcs->spare->size > 0) {
+		swap(pcs->main, pcs->spare);
+		return pcs;
+	}
+
+	/* both main and spare are empty */
+
+	full = barn_replace_empty_sheaf(barn, pcs->main, allow_spin);
 
 	if (full) {
 		stat(s, BARN_GET);
-- 
2.50.1


From haoli.tcs@gmail.com Tue Dec  2 16:24:49 2025
Date: Tue, 2 Dec 2025 16:31:49 +0800
From: Hao Li <haoli.tcs@gmail.com>
To: Vlastimil Babka <vbabka@suse.cz>
Cc: Suren Baghdasaryan <surenb@google.com>, 
	"Liam R. Howlett" <Liam.Howlett@oracle.com>, Christoph Lameter <cl@gentwo.org>, 
	David Rientjes <rientjes@google.com>, Roman Gushchin <roman.gushchin@linux.dev>, 
	Harry Yoo <harry.yoo@oracle.com>, Uladzislau Rezki <urezki@gmail.com>, 
	Sidhartha Kumar <sidhartha.kumar@oracle.com>, linux-mm@kvack.org, linux-kernel@vger.kernel.org, 
	rcu@vger.kernel.org, maple-tree@lists.infradead.org, 
	Venkat Rao Bagalkote <venkat88@linux.ibm.com>
Subject: [PATCH] slub: add barn_get_full_sheaf() and refine empty-main sheaf
Message-ID: <fqbhzbjw3nvbylvxayboyug7qayfref4xti7udztygd6zbhcun@b7s2yjihrwos>
Mutt-References: <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
Mutt-Fcc: ~/sent
References: <20250910-slub-percpu-caches-v8-0-ca3099d8352c@suse.cz>
 <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
Mutt-Fcc: ~/sent
Status: RO
Content-Length: 3509
Lines: 119

Introduce barn_get_full_sheaf(), a helper that detaches a full sheaf from
the per-node barn without requiring an empty sheaf in exchange.

Use this helper in __pcs_replace_empty_main() to change how an empty main
per-CPU sheaf is handled:

  - If pcs->spare is NULL and pcs->main is empty, first try to obtain a
    full sheaf from the barn via barn_get_full_sheaf(). On success, park
    the empty main sheaf in pcs->spare and install the full sheaf as the
    new pcs->main.

  - If pcs->spare already exists and has objects, keep the existing
    behavior of simply swapping pcs->main and pcs->spare.

  - Only when both pcs->main and pcs->spare are empty do we fall back to
    barn_replace_empty_sheaf() and trade the empty main sheaf into the
    barn in exchange for a full one.

This makes the empty-main path more symmetric with __pcs_replace_full_main(),
which for a full main sheaf parks the full sheaf in pcs->spare and pulls an
empty sheaf from the barn. It also matches the documented design more closely:

  "When both percpu sheaves are found empty during an allocation, an empty
   sheaf may be replaced with a full one from the per-node barn."

Signed-off-by: Hao Li <haoli.tcs@gmail.com>
---

* This patch is based on b4/sheaves-for-all branch

 mm/slub.c | 50 +++++++++++++++++++++++++++++++++++++++++++-------
 1 file changed, 43 insertions(+), 7 deletions(-)

diff --git a/mm/slub.c b/mm/slub.c
index a94c64f56504..1fd28aa204e1 100644
--- a/mm/slub.c
+++ b/mm/slub.c
@@ -2746,6 +2746,32 @@ static void pcs_destroy(struct kmem_cache *s)
 	s->cpu_sheaves = NULL;
 }
 
+static struct slab_sheaf *barn_get_full_sheaf(struct node_barn *barn,
+					       bool allow_spin)
+{
+	struct slab_sheaf *full = NULL;
+	unsigned long flags;
+
+	if (!data_race(barn->nr_full))
+		return NULL;
+
+	if (likely(allow_spin))
+		spin_lock_irqsave(&barn->lock, flags);
+	else if (!spin_trylock_irqsave(&barn->lock, flags))
+		return NULL;
+
+	if (likely(barn->nr_full)) {
+		full = list_first_entry(&barn->sheaves_full,
+					 struct slab_sheaf, barn_list);
+		list_del(&full->barn_list);
+		barn->nr_full--;
+	}
+
+	spin_unlock_irqrestore(&barn->lock, flags);
+
+	return full;
+}
+
 static struct slab_sheaf *barn_get_empty_sheaf(struct node_barn *barn,
 					       bool allow_spin)
 {
@@ -4120,7 +4146,7 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 	struct slab_sheaf *empty = NULL;
 	struct slab_sheaf *full;
 	struct node_barn *barn;
-	bool can_alloc;
+	bool can_alloc, allow_spin;
 
 	lockdep_assert_held(this_cpu_ptr(&s->cpu_sheaves->lock));
 
@@ -4130,10 +4156,7 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 		return NULL;
 	}
 
-	if (pcs->spare && pcs->spare->size > 0) {
-		swap(pcs->main, pcs->spare);
-		return pcs;
-	}
+	allow_spin = gfpflags_allow_spinning(gfp);
 
 	barn = get_barn(s);
 	if (!barn) {
@@ -4141,8 +4164,21 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 		return NULL;
 	}
 
-	full = barn_replace_empty_sheaf(barn, pcs->main,
-					gfpflags_allow_spinning(gfp));
+	if (!pcs->spare) {
+		full = barn_get_full_sheaf(barn, allow_spin);
+		if (full) {
+			pcs->spare = pcs->main;
+			pcs->main = full;
+			return pcs;
+		}
+	} else if (pcs->spare->size > 0) {
+		swap(pcs->main, pcs->spare);
+		return pcs;
+	}
+
+	/* both main and spare are empty */
+
+	full = barn_replace_empty_sheaf(barn, pcs->main, allow_spin);
 
 	if (full) {
 		stat(s, BARN_GET);
-- 
2.50.1



From haoli.tcs@gmail.com Tue Dec  2 16:24:49 2025
Date: Tue, 2 Dec 2025 16:33:21 +0800
From: Hao Li <haoli.tcs@gmail.com>
To: Vlastimil Babka <vbabka@suse.cz>
Cc: Suren Baghdasaryan <surenb@google.com>, 
	"Liam R. Howlett" <Liam.Howlett@oracle.com>, Christoph Lameter <cl@gentwo.org>, 
	David Rientjes <rientjes@google.com>, Roman Gushchin <roman.gushchin@linux.dev>, 
	Harry Yoo <harry.yoo@oracle.com>, Uladzislau Rezki <urezki@gmail.com>, 
	Sidhartha Kumar <sidhartha.kumar@oracle.com>, linux-mm@kvack.org, linux-kernel@vger.kernel.org, 
	rcu@vger.kernel.org, maple-tree@lists.infradead.org, 
	Venkat Rao Bagalkote <venkat88@linux.ibm.com>
Subject: [PATCH] slub: add barn_get_full_sheaf() and refine empty-main sheaf
Message-ID: <fqbhzbjw3nvbylvxayboyug7qayfref4xti7udztygd6zbhcun@b7s2yjihrwos>
Mutt-References: <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
Mutt-Fcc: ~/sent
References: <20250910-slub-percpu-caches-v8-0-ca3099d8352c@suse.cz>
 <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
Mutt-Fcc: ~/sent
Status: RO
Content-Length: 8250
Lines: 265

Introduce barn_get_full_sheaf(), a helper that detaches a full sheaf from
the per-node barn without requiring an empty sheaf in exchange.

Use this helper in __pcs_replace_empty_main() to change how an empty main
per-CPU sheaf is handled:

  - If pcs->spare is NULL and pcs->main is empty, first try to obtain a
    full sheaf from the barn via barn_get_full_sheaf(). On success, park
    the empty main sheaf in pcs->spare and install the full sheaf as the
    new pcs->main.

  - If pcs->spare already exists and has objects, keep the existing
    behavior of simply swapping pcs->main and pcs->spare.

  - Only when both pcs->main and pcs->spare are empty do we fall back to
    barn_replace_empty_sheaf() and trade the empty main sheaf into the
    barn in exchange for a full one.

This makes the empty-main path more symmetric with __pcs_replace_full_main(),
which for a full main sheaf parks the full sheaf in pcs->spare and pulls an
empty sheaf from the barn. It also matches the documented design more closely:

  "When both percpu sheaves are found empty during an allocation, an empty
   sheaf may be replaced with a full one from the per-node barn."

Signed-off-by: Hao Li <haoli.tcs@gmail.com>
---

* This patch is based on b4/sheaves-for-all branch

 mm/slub.c | 50 +++++++++++++++++++++++++++++++++++++++++++-------
 1 file changed, 43 insertions(+), 7 deletions(-)

diff --git a/mm/slub.c b/mm/slub.c
index a94c64f56504..1fd28aa204e1 100644
--- a/mm/slub.c
+++ b/mm/slub.c
@@ -2746,6 +2746,32 @@ static void pcs_destroy(struct kmem_cache *s)
 	s->cpu_sheaves = NULL;
 }
 
+static struct slab_sheaf *barn_get_full_sheaf(struct node_barn *barn,
+					       bool allow_spin)
+{
+	struct slab_sheaf *full = NULL;
+	unsigned long flags;
+
+	if (!data_race(barn->nr_full))
+		return NULL;
+
+	if (likely(allow_spin))
+		spin_lock_irqsave(&barn->lock, flags);
+	else if (!spin_trylock_irqsave(&barn->lock, flags))
+		return NULL;
+
+	if (likely(barn->nr_full)) {
+		full = list_first_entry(&barn->sheaves_full,
+					 struct slab_sheaf, barn_list);
+		list_del(&full->barn_list);
+		barn->nr_full--;
+	}
+
+	spin_unlock_irqrestore(&barn->lock, flags);
+
+	return full;
+}
+
 static struct slab_sheaf *barn_get_empty_sheaf(struct node_barn *barn,
 					       bool allow_spin)
 {
@@ -4120,7 +4146,7 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 	struct slab_sheaf *empty = NULL;
 	struct slab_sheaf *full;
 	struct node_barn *barn;
-	bool can_alloc;
+	bool can_alloc, allow_spin;
 
 	lockdep_assert_held(this_cpu_ptr(&s->cpu_sheaves->lock));
 
@@ -4130,10 +4156,7 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 		return NULL;
 	}
 
-	if (pcs->spare && pcs->spare->size > 0) {
-		swap(pcs->main, pcs->spare);
-		return pcs;
-	}
+	allow_spin = gfpflags_allow_spinning(gfp);
 
 	barn = get_barn(s);
 	if (!barn) {
@@ -4141,8 +4164,21 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 		return NULL;
 	}
 
-	full = barn_replace_empty_sheaf(barn, pcs->main,
-					gfpflags_allow_spinning(gfp));
+	if (!pcs->spare) {
+		full = barn_get_full_sheaf(barn, allow_spin);
+		if (full) {
+			pcs->spare = pcs->main;
+			pcs->main = full;
+			return pcs;
+		}
+	} else if (pcs->spare->size > 0) {
+		swap(pcs->main, pcs->spare);
+		return pcs;
+	}
+
+	/* both main and spare are empty */
+
+	full = barn_replace_empty_sheaf(barn, pcs->main, allow_spin);
 
 	if (full) {
 		stat(s, BARN_GET);
-- 
2.50.1


From haoli.tcs@gmail.com Tue Dec  2 16:24:49 2025
Date: Tue, 2 Dec 2025 16:31:49 +0800
From: Hao Li <haoli.tcs@gmail.com>
To: Vlastimil Babka <vbabka@suse.cz>
Cc: Suren Baghdasaryan <surenb@google.com>, 
	"Liam R. Howlett" <Liam.Howlett@oracle.com>, Christoph Lameter <cl@gentwo.org>, 
	David Rientjes <rientjes@google.com>, Roman Gushchin <roman.gushchin@linux.dev>, 
	Harry Yoo <harry.yoo@oracle.com>, Uladzislau Rezki <urezki@gmail.com>, 
	Sidhartha Kumar <sidhartha.kumar@oracle.com>, linux-mm@kvack.org, linux-kernel@vger.kernel.org, 
	rcu@vger.kernel.org, maple-tree@lists.infradead.org, 
	Venkat Rao Bagalkote <venkat88@linux.ibm.com>
Subject: [PATCH] slub: add barn_get_full_sheaf() and refine empty-main sheaf
Message-ID: <fqbhzbjw3nvbylvxayboyug7qayfref4xti7udztygd6zbhcun@b7s2yjihrwos>
Mutt-References: <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
Mutt-Fcc: ~/sent
References: <20250910-slub-percpu-caches-v8-0-ca3099d8352c@suse.cz>
 <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
Mutt-Fcc: ~/sent
Status: RO
Content-Length: 3509
Lines: 119

Introduce barn_get_full_sheaf(), a helper that detaches a full sheaf from
the per-node barn without requiring an empty sheaf in exchange.

Use this helper in __pcs_replace_empty_main() to change how an empty main
per-CPU sheaf is handled:

  - If pcs->spare is NULL and pcs->main is empty, first try to obtain a
    full sheaf from the barn via barn_get_full_sheaf(). On success, park
    the empty main sheaf in pcs->spare and install the full sheaf as the
    new pcs->main.

  - If pcs->spare already exists and has objects, keep the existing
    behavior of simply swapping pcs->main and pcs->spare.

  - Only when both pcs->main and pcs->spare are empty do we fall back to
    barn_replace_empty_sheaf() and trade the empty main sheaf into the
    barn in exchange for a full one.

This makes the empty-main path more symmetric with __pcs_replace_full_main(),
which for a full main sheaf parks the full sheaf in pcs->spare and pulls an
empty sheaf from the barn. It also matches the documented design more closely:

  "When both percpu sheaves are found empty during an allocation, an empty
   sheaf may be replaced with a full one from the per-node barn."

Signed-off-by: Hao Li <haoli.tcs@gmail.com>
---

* This patch is based on b4/sheaves-for-all branch

 mm/slub.c | 50 +++++++++++++++++++++++++++++++++++++++++++-------
 1 file changed, 43 insertions(+), 7 deletions(-)

diff --git a/mm/slub.c b/mm/slub.c
index a94c64f56504..1fd28aa204e1 100644
--- a/mm/slub.c
+++ b/mm/slub.c
@@ -2746,6 +2746,32 @@ static void pcs_destroy(struct kmem_cache *s)
 	s->cpu_sheaves = NULL;
 }
 
+static struct slab_sheaf *barn_get_full_sheaf(struct node_barn *barn,
+					       bool allow_spin)
+{
+	struct slab_sheaf *full = NULL;
+	unsigned long flags;
+
+	if (!data_race(barn->nr_full))
+		return NULL;
+
+	if (likely(allow_spin))
+		spin_lock_irqsave(&barn->lock, flags);
+	else if (!spin_trylock_irqsave(&barn->lock, flags))
+		return NULL;
+
+	if (likely(barn->nr_full)) {
+		full = list_first_entry(&barn->sheaves_full,
+					 struct slab_sheaf, barn_list);
+		list_del(&full->barn_list);
+		barn->nr_full--;
+	}
+
+	spin_unlock_irqrestore(&barn->lock, flags);
+
+	return full;
+}
+
 static struct slab_sheaf *barn_get_empty_sheaf(struct node_barn *barn,
 					       bool allow_spin)
 {
@@ -4120,7 +4146,7 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 	struct slab_sheaf *empty = NULL;
 	struct slab_sheaf *full;
 	struct node_barn *barn;
-	bool can_alloc;
+	bool can_alloc, allow_spin;
 
 	lockdep_assert_held(this_cpu_ptr(&s->cpu_sheaves->lock));
 
@@ -4130,10 +4156,7 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 		return NULL;
 	}
 
-	if (pcs->spare && pcs->spare->size > 0) {
-		swap(pcs->main, pcs->spare);
-		return pcs;
-	}
+	allow_spin = gfpflags_allow_spinning(gfp);
 
 	barn = get_barn(s);
 	if (!barn) {
@@ -4141,8 +4164,21 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 		return NULL;
 	}
 
-	full = barn_replace_empty_sheaf(barn, pcs->main,
-					gfpflags_allow_spinning(gfp));
+	if (!pcs->spare) {
+		full = barn_get_full_sheaf(barn, allow_spin);
+		if (full) {
+			pcs->spare = pcs->main;
+			pcs->main = full;
+			return pcs;
+		}
+	} else if (pcs->spare->size > 0) {
+		swap(pcs->main, pcs->spare);
+		return pcs;
+	}
+
+	/* both main and spare are empty */
+
+	full = barn_replace_empty_sheaf(barn, pcs->main, allow_spin);
 
 	if (full) {
 		stat(s, BARN_GET);
-- 
2.50.1





From haoli.tcs@gmail.com Tue Dec  2 16:24:49 2025
Date: Tue, 2 Dec 2025 16:45:09 +0800
From: Hao Li <haoli.tcs@gmail.com>
To: Vlastimil Babka <vbabka@suse.cz>
Cc: Suren Baghdasaryan <surenb@google.com>, 
	"Liam R. Howlett" <Liam.Howlett@oracle.com>, Christoph Lameter <cl@gentwo.org>, 
	David Rientjes <rientjes@google.com>, Roman Gushchin <roman.gushchin@linux.dev>, 
	Harry Yoo <harry.yoo@oracle.com>, Uladzislau Rezki <urezki@gmail.com>, 
	Sidhartha Kumar <sidhartha.kumar@oracle.com>, linux-mm@kvack.org, linux-kernel@vger.kernel.org, 
	rcu@vger.kernel.org, maple-tree@lists.infradead.org, 
	Venkat Rao Bagalkote <venkat88@linux.ibm.com>
Subject: [PATCH] slub: add barn_get_full_sheaf() and refine empty-main sheaf
Message-ID: <fqbhzbjw3nvbylvxayboyug7qayfref4xti7udztygd6zbhcun@b7s2yjihrwos>
Mutt-References: <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
Mutt-Fcc: ~/sent
References: <20250910-slub-percpu-caches-v8-0-ca3099d8352c@suse.cz>
 <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
Mutt-Fcc: ~/sent
Status: RO
Content-Length: 36697
Lines: 1141

Introduce barn_get_full_sheaf(), a helper that detaches a full sheaf from
the per-node barn without requiring an empty sheaf in exchange.

Use this helper in __pcs_replace_empty_main() to change how an empty main
per-CPU sheaf is handled:

  - If pcs->spare is NULL and pcs->main is empty, first try to obtain a
    full sheaf from the barn via barn_get_full_sheaf(). On success, park
    the empty main sheaf in pcs->spare and install the full sheaf as the
    new pcs->main.

  - If pcs->spare already exists and has objects, keep the existing
    behavior of simply swapping pcs->main and pcs->spare.

  - Only when both pcs->main and pcs->spare are empty do we fall back to
    barn_replace_empty_sheaf() and trade the empty main sheaf into the
    barn in exchange for a full one.

This makes the empty-main path more symmetric with __pcs_replace_full_main(),
which for a full main sheaf parks the full sheaf in pcs->spare and pulls an
empty sheaf from the barn. It also matches the documented design more closely:

  "When both percpu sheaves are found empty during an allocation, an empty
   sheaf may be replaced with a full one from the per-node barn."

Signed-off-by: Hao Li <haoli.tcs@gmail.com>
---

* This patch is based on b4/sheaves-for-all branch

 mm/slub.c | 50 +++++++++++++++++++++++++++++++++++++++++++-------
 1 file changed, 43 insertions(+), 7 deletions(-)

diff --git a/mm/slub.c b/mm/slub.c
index a94c64f56504..1fd28aa204e1 100644
--- a/mm/slub.c
+++ b/mm/slub.c
@@ -2746,6 +2746,32 @@ static void pcs_destroy(struct kmem_cache *s)
 	s->cpu_sheaves = NULL;
 }
 
+static struct slab_sheaf *barn_get_full_sheaf(struct node_barn *barn,
+					       bool allow_spin)
+{
+	struct slab_sheaf *full = NULL;
+	unsigned long flags;
+
+	if (!data_race(barn->nr_full))
+		return NULL;
+
+	if (likely(allow_spin))
+		spin_lock_irqsave(&barn->lock, flags);
+	else if (!spin_trylock_irqsave(&barn->lock, flags))
+		return NULL;
+
+	if (likely(barn->nr_full)) {
+		full = list_first_entry(&barn->sheaves_full,
+					 struct slab_sheaf, barn_list);
+		list_del(&full->barn_list);
+		barn->nr_full--;
+	}
+
+	spin_unlock_irqrestore(&barn->lock, flags);
+
+	return full;
+}
+
 static struct slab_sheaf *barn_get_empty_sheaf(struct node_barn *barn,
 					       bool allow_spin)
 {
@@ -4120,7 +4146,7 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 	struct slab_sheaf *empty = NULL;
 	struct slab_sheaf *full;
 	struct node_barn *barn;
-	bool can_alloc;
+	bool can_alloc, allow_spin;
 
 	lockdep_assert_held(this_cpu_ptr(&s->cpu_sheaves->lock));
 
@@ -4130,10 +4156,7 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 		return NULL;
 	}
 
-	if (pcs->spare && pcs->spare->size > 0) {
-		swap(pcs->main, pcs->spare);
-		return pcs;
-	}
+	allow_spin = gfpflags_allow_spinning(gfp);
 
 	barn = get_barn(s);
 	if (!barn) {
@@ -4141,8 +4164,21 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 		return NULL;
 	}
 
-	full = barn_replace_empty_sheaf(barn, pcs->main,
-					gfpflags_allow_spinning(gfp));
+	if (!pcs->spare) {
+		full = barn_get_full_sheaf(barn, allow_spin);
+		if (full) {
+			pcs->spare = pcs->main;
+			pcs->main = full;
+			return pcs;
+		}
+	} else if (pcs->spare->size > 0) {
+		swap(pcs->main, pcs->spare);
+		return pcs;
+	}
+
+	/* both main and spare are empty */
+
+	full = barn_replace_empty_sheaf(barn, pcs->main, allow_spin);
 
 	if (full) {
 		stat(s, BARN_GET);
-- 
2.50.1


From haoli.tcs@gmail.com Tue Dec  2 16:24:49 2025
Date: Tue, 2 Dec 2025 16:31:49 +0800
From: Hao Li <haoli.tcs@gmail.com>
To: Vlastimil Babka <vbabka@suse.cz>
Cc: Suren Baghdasaryan <surenb@google.com>, 
	"Liam R. Howlett" <Liam.Howlett@oracle.com>, Christoph Lameter <cl@gentwo.org>, 
	David Rientjes <rientjes@google.com>, Roman Gushchin <roman.gushchin@linux.dev>, 
	Harry Yoo <harry.yoo@oracle.com>, Uladzislau Rezki <urezki@gmail.com>, 
	Sidhartha Kumar <sidhartha.kumar@oracle.com>, linux-mm@kvack.org, linux-kernel@vger.kernel.org, 
	rcu@vger.kernel.org, maple-tree@lists.infradead.org, 
	Venkat Rao Bagalkote <venkat88@linux.ibm.com>
Subject: [PATCH] slub: add barn_get_full_sheaf() and refine empty-main sheaf
Message-ID: <fqbhzbjw3nvbylvxayboyug7qayfref4xti7udztygd6zbhcun@b7s2yjihrwos>
Mutt-References: <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
Mutt-Fcc: ~/sent
References: <20250910-slub-percpu-caches-v8-0-ca3099d8352c@suse.cz>
 <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
Mutt-Fcc: ~/sent
Status: RO
Content-Length: 3509
Lines: 119

Introduce barn_get_full_sheaf(), a helper that detaches a full sheaf from
the per-node barn without requiring an empty sheaf in exchange.

Use this helper in __pcs_replace_empty_main() to change how an empty main
per-CPU sheaf is handled:

  - If pcs->spare is NULL and pcs->main is empty, first try to obtain a
    full sheaf from the barn via barn_get_full_sheaf(). On success, park
    the empty main sheaf in pcs->spare and install the full sheaf as the
    new pcs->main.

  - If pcs->spare already exists and has objects, keep the existing
    behavior of simply swapping pcs->main and pcs->spare.

  - Only when both pcs->main and pcs->spare are empty do we fall back to
    barn_replace_empty_sheaf() and trade the empty main sheaf into the
    barn in exchange for a full one.

This makes the empty-main path more symmetric with __pcs_replace_full_main(),
which for a full main sheaf parks the full sheaf in pcs->spare and pulls an
empty sheaf from the barn. It also matches the documented design more closely:

  "When both percpu sheaves are found empty during an allocation, an empty
   sheaf may be replaced with a full one from the per-node barn."

Signed-off-by: Hao Li <haoli.tcs@gmail.com>
---

* This patch is based on b4/sheaves-for-all branch

 mm/slub.c | 50 +++++++++++++++++++++++++++++++++++++++++++-------
 1 file changed, 43 insertions(+), 7 deletions(-)

diff --git a/mm/slub.c b/mm/slub.c
index a94c64f56504..1fd28aa204e1 100644
--- a/mm/slub.c
+++ b/mm/slub.c
@@ -2746,6 +2746,32 @@ static void pcs_destroy(struct kmem_cache *s)
 	s->cpu_sheaves = NULL;
 }
 
+static struct slab_sheaf *barn_get_full_sheaf(struct node_barn *barn,
+					       bool allow_spin)
+{
+	struct slab_sheaf *full = NULL;
+	unsigned long flags;
+
+	if (!data_race(barn->nr_full))
+		return NULL;
+
+	if (likely(allow_spin))
+		spin_lock_irqsave(&barn->lock, flags);
+	else if (!spin_trylock_irqsave(&barn->lock, flags))
+		return NULL;
+
+	if (likely(barn->nr_full)) {
+		full = list_first_entry(&barn->sheaves_full,
+					 struct slab_sheaf, barn_list);
+		list_del(&full->barn_list);
+		barn->nr_full--;
+	}
+
+	spin_unlock_irqrestore(&barn->lock, flags);
+
+	return full;
+}
+
 static struct slab_sheaf *barn_get_empty_sheaf(struct node_barn *barn,
 					       bool allow_spin)
 {
@@ -4120,7 +4146,7 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 	struct slab_sheaf *empty = NULL;
 	struct slab_sheaf *full;
 	struct node_barn *barn;
-	bool can_alloc;
+	bool can_alloc, allow_spin;
 
 	lockdep_assert_held(this_cpu_ptr(&s->cpu_sheaves->lock));
 
@@ -4130,10 +4156,7 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 		return NULL;
 	}
 
-	if (pcs->spare && pcs->spare->size > 0) {
-		swap(pcs->main, pcs->spare);
-		return pcs;
-	}
+	allow_spin = gfpflags_allow_spinning(gfp);
 
 	barn = get_barn(s);
 	if (!barn) {
@@ -4141,8 +4164,21 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 		return NULL;
 	}
 
-	full = barn_replace_empty_sheaf(barn, pcs->main,
-					gfpflags_allow_spinning(gfp));
+	if (!pcs->spare) {
+		full = barn_get_full_sheaf(barn, allow_spin);
+		if (full) {
+			pcs->spare = pcs->main;
+			pcs->main = full;
+			return pcs;
+		}
+	} else if (pcs->spare->size > 0) {
+		swap(pcs->main, pcs->spare);
+		return pcs;
+	}
+
+	/* both main and spare are empty */
+
+	full = barn_replace_empty_sheaf(barn, pcs->main, allow_spin);
 
 	if (full) {
 		stat(s, BARN_GET);
-- 
2.50.1



From haoli.tcs@gmail.com Tue Dec  2 16:24:49 2025
Date: Tue, 2 Dec 2025 16:33:21 +0800
From: Hao Li <haoli.tcs@gmail.com>
To: Vlastimil Babka <vbabka@suse.cz>
Cc: Suren Baghdasaryan <surenb@google.com>, 
	"Liam R. Howlett" <Liam.Howlett@oracle.com>, Christoph Lameter <cl@gentwo.org>, 
	David Rientjes <rientjes@google.com>, Roman Gushchin <roman.gushchin@linux.dev>, 
	Harry Yoo <harry.yoo@oracle.com>, Uladzislau Rezki <urezki@gmail.com>, 
	Sidhartha Kumar <sidhartha.kumar@oracle.com>, linux-mm@kvack.org, linux-kernel@vger.kernel.org, 
	rcu@vger.kernel.org, maple-tree@lists.infradead.org, 
	Venkat Rao Bagalkote <venkat88@linux.ibm.com>
Subject: [PATCH] slub: add barn_get_full_sheaf() and refine empty-main sheaf
Message-ID: <fqbhzbjw3nvbylvxayboyug7qayfref4xti7udztygd6zbhcun@b7s2yjihrwos>
Mutt-References: <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
Mutt-Fcc: ~/sent
References: <20250910-slub-percpu-caches-v8-0-ca3099d8352c@suse.cz>
 <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
Mutt-Fcc: ~/sent
Status: RO
Content-Length: 8250
Lines: 265

Introduce barn_get_full_sheaf(), a helper that detaches a full sheaf from
the per-node barn without requiring an empty sheaf in exchange.

Use this helper in __pcs_replace_empty_main() to change how an empty main
per-CPU sheaf is handled:

  - If pcs->spare is NULL and pcs->main is empty, first try to obtain a
    full sheaf from the barn via barn_get_full_sheaf(). On success, park
    the empty main sheaf in pcs->spare and install the full sheaf as the
    new pcs->main.

  - If pcs->spare already exists and has objects, keep the existing
    behavior of simply swapping pcs->main and pcs->spare.

  - Only when both pcs->main and pcs->spare are empty do we fall back to
    barn_replace_empty_sheaf() and trade the empty main sheaf into the
    barn in exchange for a full one.

This makes the empty-main path more symmetric with __pcs_replace_full_main(),
which for a full main sheaf parks the full sheaf in pcs->spare and pulls an
empty sheaf from the barn. It also matches the documented design more closely:

  "When both percpu sheaves are found empty during an allocation, an empty
   sheaf may be replaced with a full one from the per-node barn."

Signed-off-by: Hao Li <haoli.tcs@gmail.com>
---

* This patch is based on b4/sheaves-for-all branch

 mm/slub.c | 50 +++++++++++++++++++++++++++++++++++++++++++-------
 1 file changed, 43 insertions(+), 7 deletions(-)

diff --git a/mm/slub.c b/mm/slub.c
index a94c64f56504..1fd28aa204e1 100644
--- a/mm/slub.c
+++ b/mm/slub.c
@@ -2746,6 +2746,32 @@ static void pcs_destroy(struct kmem_cache *s)
 	s->cpu_sheaves = NULL;
 }
 
+static struct slab_sheaf *barn_get_full_sheaf(struct node_barn *barn,
+					       bool allow_spin)
+{
+	struct slab_sheaf *full = NULL;
+	unsigned long flags;
+
+	if (!data_race(barn->nr_full))
+		return NULL;
+
+	if (likely(allow_spin))
+		spin_lock_irqsave(&barn->lock, flags);
+	else if (!spin_trylock_irqsave(&barn->lock, flags))
+		return NULL;
+
+	if (likely(barn->nr_full)) {
+		full = list_first_entry(&barn->sheaves_full,
+					 struct slab_sheaf, barn_list);
+		list_del(&full->barn_list);
+		barn->nr_full--;
+	}
+
+	spin_unlock_irqrestore(&barn->lock, flags);
+
+	return full;
+}
+
 static struct slab_sheaf *barn_get_empty_sheaf(struct node_barn *barn,
 					       bool allow_spin)
 {
@@ -4120,7 +4146,7 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 	struct slab_sheaf *empty = NULL;
 	struct slab_sheaf *full;
 	struct node_barn *barn;
-	bool can_alloc;
+	bool can_alloc, allow_spin;
 
 	lockdep_assert_held(this_cpu_ptr(&s->cpu_sheaves->lock));
 
@@ -4130,10 +4156,7 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 		return NULL;
 	}
 
-	if (pcs->spare && pcs->spare->size > 0) {
-		swap(pcs->main, pcs->spare);
-		return pcs;
-	}
+	allow_spin = gfpflags_allow_spinning(gfp);
 
 	barn = get_barn(s);
 	if (!barn) {
@@ -4141,8 +4164,21 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 		return NULL;
 	}
 
-	full = barn_replace_empty_sheaf(barn, pcs->main,
-					gfpflags_allow_spinning(gfp));
+	if (!pcs->spare) {
+		full = barn_get_full_sheaf(barn, allow_spin);
+		if (full) {
+			pcs->spare = pcs->main;
+			pcs->main = full;
+			return pcs;
+		}
+	} else if (pcs->spare->size > 0) {
+		swap(pcs->main, pcs->spare);
+		return pcs;
+	}
+
+	/* both main and spare are empty */
+
+	full = barn_replace_empty_sheaf(barn, pcs->main, allow_spin);
 
 	if (full) {
 		stat(s, BARN_GET);
-- 
2.50.1


From haoli.tcs@gmail.com Tue Dec  2 16:24:49 2025
Date: Tue, 2 Dec 2025 16:31:49 +0800
From: Hao Li <haoli.tcs@gmail.com>
To: Vlastimil Babka <vbabka@suse.cz>
Cc: Suren Baghdasaryan <surenb@google.com>, 
	"Liam R. Howlett" <Liam.Howlett@oracle.com>, Christoph Lameter <cl@gentwo.org>, 
	David Rientjes <rientjes@google.com>, Roman Gushchin <roman.gushchin@linux.dev>, 
	Harry Yoo <harry.yoo@oracle.com>, Uladzislau Rezki <urezki@gmail.com>, 
	Sidhartha Kumar <sidhartha.kumar@oracle.com>, linux-mm@kvack.org, linux-kernel@vger.kernel.org, 
	rcu@vger.kernel.org, maple-tree@lists.infradead.org, 
	Venkat Rao Bagalkote <venkat88@linux.ibm.com>
Subject: [PATCH] slub: add barn_get_full_sheaf() and refine empty-main sheaf
Message-ID: <fqbhzbjw3nvbylvxayboyug7qayfref4xti7udztygd6zbhcun@b7s2yjihrwos>
Mutt-References: <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
Mutt-Fcc: ~/sent
References: <20250910-slub-percpu-caches-v8-0-ca3099d8352c@suse.cz>
 <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
Mutt-Fcc: ~/sent
Status: RO
Content-Length: 3509
Lines: 119

Introduce barn_get_full_sheaf(), a helper that detaches a full sheaf from
the per-node barn without requiring an empty sheaf in exchange.

Use this helper in __pcs_replace_empty_main() to change how an empty main
per-CPU sheaf is handled:

  - If pcs->spare is NULL and pcs->main is empty, first try to obtain a
    full sheaf from the barn via barn_get_full_sheaf(). On success, park
    the empty main sheaf in pcs->spare and install the full sheaf as the
    new pcs->main.

  - If pcs->spare already exists and has objects, keep the existing
    behavior of simply swapping pcs->main and pcs->spare.

  - Only when both pcs->main and pcs->spare are empty do we fall back to
    barn_replace_empty_sheaf() and trade the empty main sheaf into the
    barn in exchange for a full one.

This makes the empty-main path more symmetric with __pcs_replace_full_main(),
which for a full main sheaf parks the full sheaf in pcs->spare and pulls an
empty sheaf from the barn. It also matches the documented design more closely:

  "When both percpu sheaves are found empty during an allocation, an empty
   sheaf may be replaced with a full one from the per-node barn."

Signed-off-by: Hao Li <haoli.tcs@gmail.com>
---

* This patch is based on b4/sheaves-for-all branch

 mm/slub.c | 50 +++++++++++++++++++++++++++++++++++++++++++-------
 1 file changed, 43 insertions(+), 7 deletions(-)

diff --git a/mm/slub.c b/mm/slub.c
index a94c64f56504..1fd28aa204e1 100644
--- a/mm/slub.c
+++ b/mm/slub.c
@@ -2746,6 +2746,32 @@ static void pcs_destroy(struct kmem_cache *s)
 	s->cpu_sheaves = NULL;
 }
 
+static struct slab_sheaf *barn_get_full_sheaf(struct node_barn *barn,
+					       bool allow_spin)
+{
+	struct slab_sheaf *full = NULL;
+	unsigned long flags;
+
+	if (!data_race(barn->nr_full))
+		return NULL;
+
+	if (likely(allow_spin))
+		spin_lock_irqsave(&barn->lock, flags);
+	else if (!spin_trylock_irqsave(&barn->lock, flags))
+		return NULL;
+
+	if (likely(barn->nr_full)) {
+		full = list_first_entry(&barn->sheaves_full,
+					 struct slab_sheaf, barn_list);
+		list_del(&full->barn_list);
+		barn->nr_full--;
+	}
+
+	spin_unlock_irqrestore(&barn->lock, flags);
+
+	return full;
+}
+
 static struct slab_sheaf *barn_get_empty_sheaf(struct node_barn *barn,
 					       bool allow_spin)
 {
@@ -4120,7 +4146,7 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 	struct slab_sheaf *empty = NULL;
 	struct slab_sheaf *full;
 	struct node_barn *barn;
-	bool can_alloc;
+	bool can_alloc, allow_spin;
 
 	lockdep_assert_held(this_cpu_ptr(&s->cpu_sheaves->lock));
 
@@ -4130,10 +4156,7 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 		return NULL;
 	}
 
-	if (pcs->spare && pcs->spare->size > 0) {
-		swap(pcs->main, pcs->spare);
-		return pcs;
-	}
+	allow_spin = gfpflags_allow_spinning(gfp);
 
 	barn = get_barn(s);
 	if (!barn) {
@@ -4141,8 +4164,21 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 		return NULL;
 	}
 
-	full = barn_replace_empty_sheaf(barn, pcs->main,
-					gfpflags_allow_spinning(gfp));
+	if (!pcs->spare) {
+		full = barn_get_full_sheaf(barn, allow_spin);
+		if (full) {
+			pcs->spare = pcs->main;
+			pcs->main = full;
+			return pcs;
+		}
+	} else if (pcs->spare->size > 0) {
+		swap(pcs->main, pcs->spare);
+		return pcs;
+	}
+
+	/* both main and spare are empty */
+
+	full = barn_replace_empty_sheaf(barn, pcs->main, allow_spin);
 
 	if (full) {
 		stat(s, BARN_GET);
-- 
2.50.1




From haoli.tcs@gmail.com Tue Dec  2 16:24:49 2025
Date: Tue, 2 Dec 2025 16:44:16 +0800
From: Hao Li <haoli.tcs@gmail.com>
To: Vlastimil Babka <vbabka@suse.cz>
Cc: Suren Baghdasaryan <surenb@google.com>, 
	"Liam R. Howlett" <Liam.Howlett@oracle.com>, Christoph Lameter <cl@gentwo.org>, 
	David Rientjes <rientjes@google.com>, Roman Gushchin <roman.gushchin@linux.dev>, 
	Harry Yoo <harry.yoo@oracle.com>, Uladzislau Rezki <urezki@gmail.com>, 
	Sidhartha Kumar <sidhartha.kumar@oracle.com>, linux-mm@kvack.org, linux-kernel@vger.kernel.org, 
	rcu@vger.kernel.org, maple-tree@lists.infradead.org, 
	Venkat Rao Bagalkote <venkat88@linux.ibm.com>
Subject: [PATCH] slub: add barn_get_full_sheaf() and refine empty-main sheaf
Message-ID: <fqbhzbjw3nvbylvxayboyug7qayfref4xti7udztygd6zbhcun@b7s2yjihrwos>
Mutt-References: <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
Mutt-Fcc: ~/sent
References: <20250910-slub-percpu-caches-v8-0-ca3099d8352c@suse.cz>
 <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
Mutt-Fcc: ~/sent
Status: RO
Content-Length: 17732
Lines: 557

Introduce barn_get_full_sheaf(), a helper that detaches a full sheaf from
the per-node barn without requiring an empty sheaf in exchange.

Use this helper in __pcs_replace_empty_main() to change how an empty main
per-CPU sheaf is handled:

  - If pcs->spare is NULL and pcs->main is empty, first try to obtain a
    full sheaf from the barn via barn_get_full_sheaf(). On success, park
    the empty main sheaf in pcs->spare and install the full sheaf as the
    new pcs->main.

  - If pcs->spare already exists and has objects, keep the existing
    behavior of simply swapping pcs->main and pcs->spare.

  - Only when both pcs->main and pcs->spare are empty do we fall back to
    barn_replace_empty_sheaf() and trade the empty main sheaf into the
    barn in exchange for a full one.

This makes the empty-main path more symmetric with __pcs_replace_full_main(),
which for a full main sheaf parks the full sheaf in pcs->spare and pulls an
empty sheaf from the barn. It also matches the documented design more closely:

  "When both percpu sheaves are found empty during an allocation, an empty
   sheaf may be replaced with a full one from the per-node barn."

Signed-off-by: Hao Li <haoli.tcs@gmail.com>
---

* This patch is based on b4/sheaves-for-all branch

 mm/slub.c | 50 +++++++++++++++++++++++++++++++++++++++++++-------
 1 file changed, 43 insertions(+), 7 deletions(-)

diff --git a/mm/slub.c b/mm/slub.c
index a94c64f56504..1fd28aa204e1 100644
--- a/mm/slub.c
+++ b/mm/slub.c
@@ -2746,6 +2746,32 @@ static void pcs_destroy(struct kmem_cache *s)
 	s->cpu_sheaves = NULL;
 }
 
+static struct slab_sheaf *barn_get_full_sheaf(struct node_barn *barn,
+					       bool allow_spin)
+{
+	struct slab_sheaf *full = NULL;
+	unsigned long flags;
+
+	if (!data_race(barn->nr_full))
+		return NULL;
+
+	if (likely(allow_spin))
+		spin_lock_irqsave(&barn->lock, flags);
+	else if (!spin_trylock_irqsave(&barn->lock, flags))
+		return NULL;
+
+	if (likely(barn->nr_full)) {
+		full = list_first_entry(&barn->sheaves_full,
+					 struct slab_sheaf, barn_list);
+		list_del(&full->barn_list);
+		barn->nr_full--;
+	}
+
+	spin_unlock_irqrestore(&barn->lock, flags);
+
+	return full;
+}
+
 static struct slab_sheaf *barn_get_empty_sheaf(struct node_barn *barn,
 					       bool allow_spin)
 {
@@ -4120,7 +4146,7 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 	struct slab_sheaf *empty = NULL;
 	struct slab_sheaf *full;
 	struct node_barn *barn;
-	bool can_alloc;
+	bool can_alloc, allow_spin;
 
 	lockdep_assert_held(this_cpu_ptr(&s->cpu_sheaves->lock));
 
@@ -4130,10 +4156,7 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 		return NULL;
 	}
 
-	if (pcs->spare && pcs->spare->size > 0) {
-		swap(pcs->main, pcs->spare);
-		return pcs;
-	}
+	allow_spin = gfpflags_allow_spinning(gfp);
 
 	barn = get_barn(s);
 	if (!barn) {
@@ -4141,8 +4164,21 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 		return NULL;
 	}
 
-	full = barn_replace_empty_sheaf(barn, pcs->main,
-					gfpflags_allow_spinning(gfp));
+	if (!pcs->spare) {
+		full = barn_get_full_sheaf(barn, allow_spin);
+		if (full) {
+			pcs->spare = pcs->main;
+			pcs->main = full;
+			return pcs;
+		}
+	} else if (pcs->spare->size > 0) {
+		swap(pcs->main, pcs->spare);
+		return pcs;
+	}
+
+	/* both main and spare are empty */
+
+	full = barn_replace_empty_sheaf(barn, pcs->main, allow_spin);
 
 	if (full) {
 		stat(s, BARN_GET);
-- 
2.50.1


From haoli.tcs@gmail.com Tue Dec  2 16:24:49 2025
Date: Tue, 2 Dec 2025 16:31:49 +0800
From: Hao Li <haoli.tcs@gmail.com>
To: Vlastimil Babka <vbabka@suse.cz>
Cc: Suren Baghdasaryan <surenb@google.com>, 
	"Liam R. Howlett" <Liam.Howlett@oracle.com>, Christoph Lameter <cl@gentwo.org>, 
	David Rientjes <rientjes@google.com>, Roman Gushchin <roman.gushchin@linux.dev>, 
	Harry Yoo <harry.yoo@oracle.com>, Uladzislau Rezki <urezki@gmail.com>, 
	Sidhartha Kumar <sidhartha.kumar@oracle.com>, linux-mm@kvack.org, linux-kernel@vger.kernel.org, 
	rcu@vger.kernel.org, maple-tree@lists.infradead.org, 
	Venkat Rao Bagalkote <venkat88@linux.ibm.com>
Subject: [PATCH] slub: add barn_get_full_sheaf() and refine empty-main sheaf
Message-ID: <fqbhzbjw3nvbylvxayboyug7qayfref4xti7udztygd6zbhcun@b7s2yjihrwos>
Mutt-References: <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
Mutt-Fcc: ~/sent
References: <20250910-slub-percpu-caches-v8-0-ca3099d8352c@suse.cz>
 <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
Mutt-Fcc: ~/sent
Status: RO
Content-Length: 3509
Lines: 119

Introduce barn_get_full_sheaf(), a helper that detaches a full sheaf from
the per-node barn without requiring an empty sheaf in exchange.

Use this helper in __pcs_replace_empty_main() to change how an empty main
per-CPU sheaf is handled:

  - If pcs->spare is NULL and pcs->main is empty, first try to obtain a
    full sheaf from the barn via barn_get_full_sheaf(). On success, park
    the empty main sheaf in pcs->spare and install the full sheaf as the
    new pcs->main.

  - If pcs->spare already exists and has objects, keep the existing
    behavior of simply swapping pcs->main and pcs->spare.

  - Only when both pcs->main and pcs->spare are empty do we fall back to
    barn_replace_empty_sheaf() and trade the empty main sheaf into the
    barn in exchange for a full one.

This makes the empty-main path more symmetric with __pcs_replace_full_main(),
which for a full main sheaf parks the full sheaf in pcs->spare and pulls an
empty sheaf from the barn. It also matches the documented design more closely:

  "When both percpu sheaves are found empty during an allocation, an empty
   sheaf may be replaced with a full one from the per-node barn."

Signed-off-by: Hao Li <haoli.tcs@gmail.com>
---

* This patch is based on b4/sheaves-for-all branch

 mm/slub.c | 50 +++++++++++++++++++++++++++++++++++++++++++-------
 1 file changed, 43 insertions(+), 7 deletions(-)

diff --git a/mm/slub.c b/mm/slub.c
index a94c64f56504..1fd28aa204e1 100644
--- a/mm/slub.c
+++ b/mm/slub.c
@@ -2746,6 +2746,32 @@ static void pcs_destroy(struct kmem_cache *s)
 	s->cpu_sheaves = NULL;
 }
 
+static struct slab_sheaf *barn_get_full_sheaf(struct node_barn *barn,
+					       bool allow_spin)
+{
+	struct slab_sheaf *full = NULL;
+	unsigned long flags;
+
+	if (!data_race(barn->nr_full))
+		return NULL;
+
+	if (likely(allow_spin))
+		spin_lock_irqsave(&barn->lock, flags);
+	else if (!spin_trylock_irqsave(&barn->lock, flags))
+		return NULL;
+
+	if (likely(barn->nr_full)) {
+		full = list_first_entry(&barn->sheaves_full,
+					 struct slab_sheaf, barn_list);
+		list_del(&full->barn_list);
+		barn->nr_full--;
+	}
+
+	spin_unlock_irqrestore(&barn->lock, flags);
+
+	return full;
+}
+
 static struct slab_sheaf *barn_get_empty_sheaf(struct node_barn *barn,
 					       bool allow_spin)
 {
@@ -4120,7 +4146,7 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 	struct slab_sheaf *empty = NULL;
 	struct slab_sheaf *full;
 	struct node_barn *barn;
-	bool can_alloc;
+	bool can_alloc, allow_spin;
 
 	lockdep_assert_held(this_cpu_ptr(&s->cpu_sheaves->lock));
 
@@ -4130,10 +4156,7 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 		return NULL;
 	}
 
-	if (pcs->spare && pcs->spare->size > 0) {
-		swap(pcs->main, pcs->spare);
-		return pcs;
-	}
+	allow_spin = gfpflags_allow_spinning(gfp);
 
 	barn = get_barn(s);
 	if (!barn) {
@@ -4141,8 +4164,21 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 		return NULL;
 	}
 
-	full = barn_replace_empty_sheaf(barn, pcs->main,
-					gfpflags_allow_spinning(gfp));
+	if (!pcs->spare) {
+		full = barn_get_full_sheaf(barn, allow_spin);
+		if (full) {
+			pcs->spare = pcs->main;
+			pcs->main = full;
+			return pcs;
+		}
+	} else if (pcs->spare->size > 0) {
+		swap(pcs->main, pcs->spare);
+		return pcs;
+	}
+
+	/* both main and spare are empty */
+
+	full = barn_replace_empty_sheaf(barn, pcs->main, allow_spin);
 
 	if (full) {
 		stat(s, BARN_GET);
-- 
2.50.1



From haoli.tcs@gmail.com Tue Dec  2 16:24:49 2025
Date: Tue, 2 Dec 2025 16:33:21 +0800
From: Hao Li <haoli.tcs@gmail.com>
To: Vlastimil Babka <vbabka@suse.cz>
Cc: Suren Baghdasaryan <surenb@google.com>, 
	"Liam R. Howlett" <Liam.Howlett@oracle.com>, Christoph Lameter <cl@gentwo.org>, 
	David Rientjes <rientjes@google.com>, Roman Gushchin <roman.gushchin@linux.dev>, 
	Harry Yoo <harry.yoo@oracle.com>, Uladzislau Rezki <urezki@gmail.com>, 
	Sidhartha Kumar <sidhartha.kumar@oracle.com>, linux-mm@kvack.org, linux-kernel@vger.kernel.org, 
	rcu@vger.kernel.org, maple-tree@lists.infradead.org, 
	Venkat Rao Bagalkote <venkat88@linux.ibm.com>
Subject: [PATCH] slub: add barn_get_full_sheaf() and refine empty-main sheaf
Message-ID: <fqbhzbjw3nvbylvxayboyug7qayfref4xti7udztygd6zbhcun@b7s2yjihrwos>
Mutt-References: <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
Mutt-Fcc: ~/sent
References: <20250910-slub-percpu-caches-v8-0-ca3099d8352c@suse.cz>
 <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
Mutt-Fcc: ~/sent
Status: RO
Content-Length: 8250
Lines: 265

Introduce barn_get_full_sheaf(), a helper that detaches a full sheaf from
the per-node barn without requiring an empty sheaf in exchange.

Use this helper in __pcs_replace_empty_main() to change how an empty main
per-CPU sheaf is handled:

  - If pcs->spare is NULL and pcs->main is empty, first try to obtain a
    full sheaf from the barn via barn_get_full_sheaf(). On success, park
    the empty main sheaf in pcs->spare and install the full sheaf as the
    new pcs->main.

  - If pcs->spare already exists and has objects, keep the existing
    behavior of simply swapping pcs->main and pcs->spare.

  - Only when both pcs->main and pcs->spare are empty do we fall back to
    barn_replace_empty_sheaf() and trade the empty main sheaf into the
    barn in exchange for a full one.

This makes the empty-main path more symmetric with __pcs_replace_full_main(),
which for a full main sheaf parks the full sheaf in pcs->spare and pulls an
empty sheaf from the barn. It also matches the documented design more closely:

  "When both percpu sheaves are found empty during an allocation, an empty
   sheaf may be replaced with a full one from the per-node barn."

Signed-off-by: Hao Li <haoli.tcs@gmail.com>
---

* This patch is based on b4/sheaves-for-all branch

 mm/slub.c | 50 +++++++++++++++++++++++++++++++++++++++++++-------
 1 file changed, 43 insertions(+), 7 deletions(-)

diff --git a/mm/slub.c b/mm/slub.c
index a94c64f56504..1fd28aa204e1 100644
--- a/mm/slub.c
+++ b/mm/slub.c
@@ -2746,6 +2746,32 @@ static void pcs_destroy(struct kmem_cache *s)
 	s->cpu_sheaves = NULL;
 }
 
+static struct slab_sheaf *barn_get_full_sheaf(struct node_barn *barn,
+					       bool allow_spin)
+{
+	struct slab_sheaf *full = NULL;
+	unsigned long flags;
+
+	if (!data_race(barn->nr_full))
+		return NULL;
+
+	if (likely(allow_spin))
+		spin_lock_irqsave(&barn->lock, flags);
+	else if (!spin_trylock_irqsave(&barn->lock, flags))
+		return NULL;
+
+	if (likely(barn->nr_full)) {
+		full = list_first_entry(&barn->sheaves_full,
+					 struct slab_sheaf, barn_list);
+		list_del(&full->barn_list);
+		barn->nr_full--;
+	}
+
+	spin_unlock_irqrestore(&barn->lock, flags);
+
+	return full;
+}
+
 static struct slab_sheaf *barn_get_empty_sheaf(struct node_barn *barn,
 					       bool allow_spin)
 {
@@ -4120,7 +4146,7 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 	struct slab_sheaf *empty = NULL;
 	struct slab_sheaf *full;
 	struct node_barn *barn;
-	bool can_alloc;
+	bool can_alloc, allow_spin;
 
 	lockdep_assert_held(this_cpu_ptr(&s->cpu_sheaves->lock));
 
@@ -4130,10 +4156,7 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 		return NULL;
 	}
 
-	if (pcs->spare && pcs->spare->size > 0) {
-		swap(pcs->main, pcs->spare);
-		return pcs;
-	}
+	allow_spin = gfpflags_allow_spinning(gfp);
 
 	barn = get_barn(s);
 	if (!barn) {
@@ -4141,8 +4164,21 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 		return NULL;
 	}
 
-	full = barn_replace_empty_sheaf(barn, pcs->main,
-					gfpflags_allow_spinning(gfp));
+	if (!pcs->spare) {
+		full = barn_get_full_sheaf(barn, allow_spin);
+		if (full) {
+			pcs->spare = pcs->main;
+			pcs->main = full;
+			return pcs;
+		}
+	} else if (pcs->spare->size > 0) {
+		swap(pcs->main, pcs->spare);
+		return pcs;
+	}
+
+	/* both main and spare are empty */
+
+	full = barn_replace_empty_sheaf(barn, pcs->main, allow_spin);
 
 	if (full) {
 		stat(s, BARN_GET);
-- 
2.50.1


From haoli.tcs@gmail.com Tue Dec  2 16:24:49 2025
Date: Tue, 2 Dec 2025 16:31:49 +0800
From: Hao Li <haoli.tcs@gmail.com>
To: Vlastimil Babka <vbabka@suse.cz>
Cc: Suren Baghdasaryan <surenb@google.com>, 
	"Liam R. Howlett" <Liam.Howlett@oracle.com>, Christoph Lameter <cl@gentwo.org>, 
	David Rientjes <rientjes@google.com>, Roman Gushchin <roman.gushchin@linux.dev>, 
	Harry Yoo <harry.yoo@oracle.com>, Uladzislau Rezki <urezki@gmail.com>, 
	Sidhartha Kumar <sidhartha.kumar@oracle.com>, linux-mm@kvack.org, linux-kernel@vger.kernel.org, 
	rcu@vger.kernel.org, maple-tree@lists.infradead.org, 
	Venkat Rao Bagalkote <venkat88@linux.ibm.com>
Subject: [PATCH] slub: add barn_get_full_sheaf() and refine empty-main sheaf
Message-ID: <fqbhzbjw3nvbylvxayboyug7qayfref4xti7udztygd6zbhcun@b7s2yjihrwos>
Mutt-References: <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
Mutt-Fcc: ~/sent
References: <20250910-slub-percpu-caches-v8-0-ca3099d8352c@suse.cz>
 <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20250910-slub-percpu-caches-v8-3-ca3099d8352c@suse.cz>
Mutt-Fcc: ~/sent
Status: RO
Content-Length: 3509
Lines: 119

Introduce barn_get_full_sheaf(), a helper that detaches a full sheaf from
the per-node barn without requiring an empty sheaf in exchange.

Use this helper in __pcs_replace_empty_main() to change how an empty main
per-CPU sheaf is handled:

  - If pcs->spare is NULL and pcs->main is empty, first try to obtain a
    full sheaf from the barn via barn_get_full_sheaf(). On success, park
    the empty main sheaf in pcs->spare and install the full sheaf as the
    new pcs->main.

  - If pcs->spare already exists and has objects, keep the existing
    behavior of simply swapping pcs->main and pcs->spare.

  - Only when both pcs->main and pcs->spare are empty do we fall back to
    barn_replace_empty_sheaf() and trade the empty main sheaf into the
    barn in exchange for a full one.

This makes the empty-main path more symmetric with __pcs_replace_full_main(),
which for a full main sheaf parks the full sheaf in pcs->spare and pulls an
empty sheaf from the barn. It also matches the documented design more closely:

  "When both percpu sheaves are found empty during an allocation, an empty
   sheaf may be replaced with a full one from the per-node barn."

Signed-off-by: Hao Li <haoli.tcs@gmail.com>
---

* This patch is based on b4/sheaves-for-all branch

 mm/slub.c | 50 +++++++++++++++++++++++++++++++++++++++++++-------
 1 file changed, 43 insertions(+), 7 deletions(-)

diff --git a/mm/slub.c b/mm/slub.c
index a94c64f56504..1fd28aa204e1 100644
--- a/mm/slub.c
+++ b/mm/slub.c
@@ -2746,6 +2746,32 @@ static void pcs_destroy(struct kmem_cache *s)
 	s->cpu_sheaves = NULL;
 }
 
+static struct slab_sheaf *barn_get_full_sheaf(struct node_barn *barn,
+					       bool allow_spin)
+{
+	struct slab_sheaf *full = NULL;
+	unsigned long flags;
+
+	if (!data_race(barn->nr_full))
+		return NULL;
+
+	if (likely(allow_spin))
+		spin_lock_irqsave(&barn->lock, flags);
+	else if (!spin_trylock_irqsave(&barn->lock, flags))
+		return NULL;
+
+	if (likely(barn->nr_full)) {
+		full = list_first_entry(&barn->sheaves_full,
+					 struct slab_sheaf, barn_list);
+		list_del(&full->barn_list);
+		barn->nr_full--;
+	}
+
+	spin_unlock_irqrestore(&barn->lock, flags);
+
+	return full;
+}
+
 static struct slab_sheaf *barn_get_empty_sheaf(struct node_barn *barn,
 					       bool allow_spin)
 {
@@ -4120,7 +4146,7 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 	struct slab_sheaf *empty = NULL;
 	struct slab_sheaf *full;
 	struct node_barn *barn;
-	bool can_alloc;
+	bool can_alloc, allow_spin;
 
 	lockdep_assert_held(this_cpu_ptr(&s->cpu_sheaves->lock));
 
@@ -4130,10 +4156,7 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 		return NULL;
 	}
 
-	if (pcs->spare && pcs->spare->size > 0) {
-		swap(pcs->main, pcs->spare);
-		return pcs;
-	}
+	allow_spin = gfpflags_allow_spinning(gfp);
 
 	barn = get_barn(s);
 	if (!barn) {
@@ -4141,8 +4164,21 @@ __pcs_replace_empty_main(struct kmem_cache *s, struct slub_percpu_sheaves *pcs,
 		return NULL;
 	}
 
-	full = barn_replace_empty_sheaf(barn, pcs->main,
-					gfpflags_allow_spinning(gfp));
+	if (!pcs->spare) {
+		full = barn_get_full_sheaf(barn, allow_spin);
+		if (full) {
+			pcs->spare = pcs->main;
+			pcs->main = full;
+			return pcs;
+		}
+	} else if (pcs->spare->size > 0) {
+		swap(pcs->main, pcs->spare);
+		return pcs;
+	}
+
+	/* both main and spare are empty */
+
+	full = barn_replace_empty_sheaf(barn, pcs->main, allow_spin);
 
 	if (full) {
 		stat(s, BARN_GET);
-- 
2.50.1








  reply	other threads:[~2025-12-02  8:48 UTC|newest]

Thread overview: 95+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-09-10  8:01 [PATCH v8 00/23] SLUB percpu sheaves Vlastimil Babka
2025-09-10  8:01 ` [PATCH v8 01/23] locking/local_lock: Expose dep_map in local_trylock_t Vlastimil Babka
2025-09-24 16:49   ` Suren Baghdasaryan
2025-09-10  8:01 ` [PATCH v8 02/23] slab: simplify init_kmem_cache_nodes() error handling Vlastimil Babka
2025-09-24 16:52   ` Suren Baghdasaryan
2025-09-10  8:01 ` [PATCH v8 03/23] slab: add opt-in caching layer of percpu sheaves Vlastimil Babka
2025-12-02  8:48   ` Hao Li [this message]
2025-12-02  8:55     ` [PATCH] slub: add barn_get_full_sheaf() and refine empty-main sheaf Hao Li
2025-12-02  9:00   ` slub: add barn_get_full_sheaf() and refine empty-main sheaf replacement Hao Li
2025-12-03  5:46     ` Harry Yoo
2025-12-03 11:15       ` Hao Li
2025-09-10  8:01 ` [PATCH v8 04/23] slab: add sheaf support for batching kfree_rcu() operations Vlastimil Babka
2025-09-12  0:38   ` Sergey Senozhatsky
2025-09-12  7:03     ` Vlastimil Babka
2025-09-17  8:30   ` Harry Yoo
2025-09-17  9:55     ` Vlastimil Babka
2025-09-17 11:32       ` Harry Yoo
2025-09-17 12:05         ` Vlastimil Babka
2025-09-17 13:07           ` Harry Yoo
2025-09-17 13:21             ` Vlastimil Babka
2025-09-17 13:34               ` Harry Yoo
2025-09-17 14:14                 ` Vlastimil Babka
2025-09-18  8:09                   ` Vlastimil Babka
2025-09-19  6:47                     ` Harry Yoo
2025-09-19  7:02                       ` Vlastimil Babka
2025-09-19  8:59                         ` Harry Yoo
2025-09-25  4:35                     ` Suren Baghdasaryan
2025-09-25  8:52                       ` Harry Yoo
2025-09-25 13:38                         ` Suren Baghdasaryan
2025-09-26 10:08                       ` Vlastimil Babka
2025-09-26 15:41                         ` Suren Baghdasaryan
2025-09-17 11:36       ` Paul E. McKenney
2025-09-17 12:13         ` Vlastimil Babka
2025-10-31 21:32   ` Daniel Gomez
2025-11-03  3:17     ` Harry Yoo
2025-11-05 11:25       ` Vlastimil Babka
2025-11-27 14:00         ` Daniel Gomez
2025-11-27 19:29           ` Suren Baghdasaryan
2025-11-28 11:37             ` [PATCH V1] mm/slab: introduce kvfree_rcu_barrier_on_cache() for cache destruction Harry Yoo
2025-11-28 12:22               ` Harry Yoo
2025-11-28 12:38               ` Daniel Gomez
2025-12-02  9:29               ` Jon Hunter
2025-12-02 10:18                 ` Harry Yoo
2025-11-27 11:38     ` [PATCH v8 04/23] slab: add sheaf support for batching kfree_rcu() operations Jon Hunter
2025-11-27 11:50       ` Jon Hunter
2025-11-27 12:33       ` Harry Yoo
2025-11-27 12:48         ` Harry Yoo
2025-11-28  8:57           ` Jon Hunter
2025-12-01  6:55             ` Harry Yoo
2025-11-27 13:18       ` Vlastimil Babka
2025-11-28  8:59         ` Jon Hunter
2025-09-10  8:01 ` [PATCH v8 05/23] slab: sheaf prefilling for guaranteed allocations Vlastimil Babka
2025-09-10  8:01 ` [PATCH v8 06/23] slab: determine barn status racily outside of lock Vlastimil Babka
2025-09-10  8:01 ` [PATCH v8 07/23] slab: skip percpu sheaves for remote object freeing Vlastimil Babka
2025-09-25 16:14   ` Suren Baghdasaryan
2025-09-10  8:01 ` [PATCH v8 08/23] slab: allow NUMA restricted allocations to use percpu sheaves Vlastimil Babka
2025-09-25 16:27   ` Suren Baghdasaryan
2025-09-10  8:01 ` [PATCH v8 09/23] maple_tree: remove redundant __GFP_NOWARN Vlastimil Babka
2025-09-10  8:01 ` [PATCH v8 10/23] tools/testing/vma: clean up stubs in vma_internal.h Vlastimil Babka
2025-09-10  8:01 ` [PATCH v8 11/23] maple_tree: Drop bulk insert support Vlastimil Babka
2025-09-25 16:38   ` Suren Baghdasaryan
2025-09-10  8:01 ` [PATCH v8 12/23] tools/testing/vma: Implement vm_refcnt reset Vlastimil Babka
2025-09-25 16:38   ` Suren Baghdasaryan
2025-09-10  8:01 ` [PATCH v8 13/23] tools/testing: Add support for changes to slab for sheaves Vlastimil Babka
2025-09-26 23:28   ` Suren Baghdasaryan
2025-09-10  8:01 ` [PATCH v8 14/23] mm, vma: use percpu sheaves for vm_area_struct cache Vlastimil Babka
2025-09-10  8:01 ` [PATCH v8 15/23] maple_tree: use percpu sheaves for maple_node_cache Vlastimil Babka
2025-09-12  2:20   ` Liam R. Howlett
2025-10-16 15:16   ` D, Suneeth
2025-10-16 16:15     ` Vlastimil Babka
2025-10-17 18:26       ` D, Suneeth
2025-09-10  8:01 ` [PATCH v8 16/23] tools/testing: include maple-shim.c in maple.c Vlastimil Babka
2025-09-26 23:45   ` Suren Baghdasaryan
2025-09-10  8:01 ` [PATCH v8 17/23] testing/radix-tree/maple: Hack around kfree_rcu not existing Vlastimil Babka
2025-09-26 23:53   ` Suren Baghdasaryan
2025-09-10  8:01 ` [PATCH v8 18/23] maple_tree: Use kfree_rcu in ma_free_rcu Vlastimil Babka
2025-09-17 11:46   ` Harry Yoo
2025-09-27  0:05     ` Suren Baghdasaryan
2025-09-10  8:01 ` [PATCH v8 19/23] maple_tree: Replace mt_free_one() with kfree() Vlastimil Babka
2025-09-27  0:06   ` Suren Baghdasaryan
2025-09-10  8:01 ` [PATCH v8 20/23] tools/testing: Add support for prefilled slab sheafs Vlastimil Babka
2025-09-27  0:28   ` Suren Baghdasaryan
2025-09-10  8:01 ` [PATCH v8 21/23] maple_tree: Prefilled sheaf conversion and testing Vlastimil Babka
2025-09-27  1:08   ` Suren Baghdasaryan
2025-09-29  7:30     ` Vlastimil Babka
2025-09-29 16:51       ` Liam R. Howlett
2025-09-10  8:01 ` [PATCH v8 22/23] maple_tree: Add single node allocation support to maple state Vlastimil Babka
2025-09-27  1:17   ` Suren Baghdasaryan
2025-09-29  7:39     ` Vlastimil Babka
2025-09-10  8:01 ` [PATCH v8 23/23] maple_tree: Convert forking to use the sheaf interface Vlastimil Babka
2025-10-07  6:34 ` [PATCH v8 00/23] SLUB percpu sheaves Christoph Hellwig
2025-10-07  8:03   ` Vlastimil Babka
2025-10-08  6:04     ` Christoph Hellwig
2025-10-15  8:32       ` Vlastimil Babka
2025-10-22  6:47         ` Christoph Hellwig

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=fqbhzbjw3nvbylvxayboyug7qayfref4xti7udztygd6zbhcun@b7s2yjihrwos \
    --to=haoli.tcs@gmail.com \
    --cc=Liam.Howlett@oracle.com \
    --cc=cl@gentwo.org \
    --cc=harry.yoo@oracle.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=maple-tree@lists.infradead.org \
    --cc=rcu@vger.kernel.org \
    --cc=rientjes@google.com \
    --cc=roman.gushchin@linux.dev \
    --cc=sidhartha.kumar@oracle.com \
    --cc=surenb@google.com \
    --cc=urezki@gmail.com \
    --cc=vbabka@suse.cz \
    --cc=venkat88@linux.ibm.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox