linux-kernel - [PATCH 09/10] x86: add embedding percpu first chunk allocator

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-Id: <1235445101-7882-10-git-send-email-tj@kernel.org>
Date:	Tue, 24 Feb 2009 12:11:40 +0900
From:	Tejun Heo <tj@...nel.org>
To:	mingo@...e.hu, rusty@...tcorp.com.au, tglx@...utronix.de,
	x86@...nel.org, linux-kernel@...r.kernel.org, hpa@...or.com,
	jeremy@...p.org, cpw@....com, nickpiggin@...oo.com.au,
	ink@...assic.park.msu.ru
Cc:	Tejun Heo <tj@...nel.org>
Subject: [PATCH 09/10] x86: add embedding percpu first chunk allocator

Impact: add better first percpu allocation for !NUMA

On !NUMA, we can simply allocate contiguous memory and use it for the
first chunk without mapping it into vmalloc area.  As the memory area
is covered by the large page physical memory mapping, it allows the
dynamic perpcu allocator to not add any TLB overhead for the static
percpu area and whatever falls into the first chunk and the
implementation is very simple too.

Signed-off-by: Tejun Heo <tj@...nel.org>
---
 arch/x86/kernel/setup_percpu.c |   86 +++++++++++++++++++++++++++++++++++++++-
 1 files changed, 85 insertions(+), 1 deletions(-)

diff --git a/arch/x86/kernel/setup_percpu.c b/arch/x86/kernel/setup_percpu.c
index 4a17c96..fd4c399 100644
--- a/arch/x86/kernel/setup_percpu.c
+++ b/arch/x86/kernel/setup_percpu.c
@@ -43,6 +43,35 @@ unsigned long __per_cpu_offset[NR_CPUS] __read_mostly = {
 EXPORT_SYMBOL(__per_cpu_offset);
 
 /**
+ * pcpu_need_numa - determine percpu allocation needs to consider NUMA
+ *
+ * If NUMA is not configured or there is only one NUMA node available,
+ * there is no reason to consider NUMA.  This function determines
+ * whether percpu allocation should consider NUMA or not.
+ *
+ * RETURNS:
+ * true if NUMA should be considered; otherwise, false.
+ */
+static bool __init pcpu_need_numa(void)
+{
+#ifdef CONFIG_NEED_MULTIPLE_NODES
+	pg_data_t *last = NULL;
+	unsigned int cpu;
+
+	for_each_possible_cpu(cpu) {
+		int node = early_cpu_to_node(cpu);
+
+		if (node_online(node) && NODE_DATA(node) &&
+		    last && last != NODE_DATA(node))
+			return true;
+
+		last = NODE_DATA(node);
+	}
+#endif
+	return false;
+}
+
+/**
  * pcpu_alloc_bootmem - NUMA friendly alloc_bootmem wrapper for percpu
  * @cpu: cpu to allocate for
  * @size: size allocation in bytes
@@ -82,6 +111,59 @@ static void * __init pcpu_alloc_bootmem(unsigned int cpu, unsigned long size,
 }
 
 /*
+ * Embedding allocator
+ *
+ * The first chunk is sized to just contain the static area plus
+ * PERCPU_DYNAMIC_RESERVE and allocated as a contiguous area using
+ * bootmem allocator and used as-is without being mapped into vmalloc
+ * area.  This enables the first chunk to piggy back on the linear
+ * physical PMD mapping and doesn't add any additional pressure to
+ * TLB.
+ */
+static void *pcpue_ptr __initdata;
+static size_t pcpue_unit_size __initdata;
+
+static struct page * __init pcpue_get_page(unsigned int cpu, int pageno)
+{
+	return virt_to_page(pcpue_ptr + cpu * pcpue_unit_size
+			    + ((size_t)pageno << PAGE_SHIFT));
+}
+
+static ssize_t __init setup_pcpu_embed(size_t static_size)
+{
+	unsigned int cpu;
+
+	/*
+	 * If large page isn't supported, there's no benefit in doing
+	 * this.  Also, embedding allocation doesn't play well with
+	 * NUMA.
+	 */
+	if (!cpu_has_pse || pcpu_need_numa())
+		return -EINVAL;
+
+	/* allocate and copy */
+	pcpue_unit_size = PFN_ALIGN(static_size + PERCPU_DYNAMIC_RESERVE);
+	pcpue_unit_size = max(pcpue_unit_size, PCPU_MIN_UNIT_SIZE);
+	pcpue_ptr = pcpu_alloc_bootmem(0, num_possible_cpus() * pcpue_unit_size,
+				       PAGE_SIZE);
+	if (!pcpue_ptr)
+		return -ENOMEM;
+
+	for_each_possible_cpu(cpu)
+		memcpy(pcpue_ptr + cpu * pcpue_unit_size, __per_cpu_load,
+		       static_size);
+
+	/* we're ready, commit */
+	pr_info("PERCPU: Embedded %zu pages at %p, static data %zu bytes\n",
+		pcpue_unit_size >> PAGE_SHIFT, pcpue_ptr, static_size);
+
+	return pcpu_setup_first_chunk(pcpue_get_page, static_size,
+				      pcpue_unit_size,
+				      pcpue_unit_size - static_size, pcpue_ptr,
+				      NULL);
+}
+
+/*
  * 4k page allocator
  *
  * This is the basic allocator.  Static percpu area is allocated
@@ -178,7 +260,9 @@ void __init setup_per_cpu_areas(void)
 		NR_CPUS, nr_cpumask_bits, nr_cpu_ids, nr_node_ids);
 
 	/* allocate percpu area */
-	ret = setup_pcpu_4k(static_size);
+	ret = setup_pcpu_embed(static_size);
+	if (ret < 0)
+		ret = setup_pcpu_4k(static_size);
 	if (ret < 0)
 		panic("cannot allocate static percpu area (%zu bytes, err=%zd)",
 		      static_size, ret);
-- 
1.6.0.2

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/