linux-kernel - [PATCH 1/7] Improve alloc_percpu: make the per cpu reserve configurable and larger.

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>] [day] [month] [year] [list]

Message-Id: <20081117132630.33F09DDDF5@ozlabs.org>
Date:	Mon Nov 17 23:48:26 CST 2008
From:	Rusty Russell <rusty@...tcorp.com.au>
To:	linux-kernel@...r.kernel.org
Cc:	Christoph Lameter <cl@...ux-foundation.org>
Subject: [PATCH 1/7] Improve alloc_percpu: make the per cpu reserve configurable and larger.


This is based on Christoph Lameter's "cpualloc: make the per cpu
reserve configurable" patch, and his "Increase default reserve percpu
area" patch.

Christoph did the hard work of figuring out what the number should be
(based on convering the slub allocator).

Signed-off-by: Rusty Russell <rusty@...tcorp.com.au>
Cc: Christoph Lameter <cl@...ux-foundation.org>
---
 Documentation/kernel-parameters.txt |    7 +++++++
 include/linux/percpu.h              |   10 ++--------
 init/main.c                         |    2 ++
 mm/allocpercpu.c                    |    5 +++++
 4 files changed, 16 insertions(+), 8 deletions(-)

diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt
--- a/Documentation/kernel-parameters.txt
+++ b/Documentation/kernel-parameters.txt
@@ -1704,6 +1704,13 @@
 			Format: { 0 | 1 }
 			See arch/parisc/kernel/pdc_chassis.c
 
+	percpu=		[SMP] Configure the number of percpu bytes that can be
+			dynamically allocated. This is used for per cpu
+			variables of modules and other dynamic per cpu data
+			structures. Creation of per cpu structures after boot
+			may fail if this is set too low.
+			Default is 10k/20k for 32 or 64-bit systems.
+
 	pf.		[PARIDE]
 			See Documentation/blockdev/paride.txt.
 
diff --git a/include/linux/percpu.h b/include/linux/percpu.h
--- a/include/linux/percpu.h
+++ b/include/linux/percpu.h
@@ -41,16 +41,10 @@
 #define EXPORT_PER_CPU_SYMBOL(var) EXPORT_SYMBOL(per_cpu__##var)
 #define EXPORT_PER_CPU_SYMBOL_GPL(var) EXPORT_SYMBOL_GPL(per_cpu__##var)
 
-/* Enough to cover all DEFINE_PER_CPUs in kernel, including modules. */
 #ifndef PERCPU_ENOUGH_ROOM
-#ifdef CONFIG_MODULES
-#define PERCPU_MODULE_RESERVE	8192
-#else
-#define PERCPU_MODULE_RESERVE	0
-#endif
+extern unsigned int percpu_reserve;
 
-#define PERCPU_ENOUGH_ROOM						\
-	(__per_cpu_end - __per_cpu_start + PERCPU_MODULE_RESERVE)
+#define PERCPU_ENOUGH_ROOM (__per_cpu_end - __per_cpu_start + percpu_reserve)
 #endif	/* PERCPU_ENOUGH_ROOM */
 
 /*
diff --git a/init/main.c b/init/main.c
--- a/init/main.c
+++ b/init/main.c
@@ -401,6 +401,8 @@
 	/* Copy section for each CPU (we discard the original) */
 	size = ALIGN(PERCPU_ENOUGH_ROOM, PAGE_SIZE);
 	ptr = alloc_bootmem_pages(size * nr_possible_cpus);
+	printk(KERN_INFO "percpu area: %d bytes total, %d available.\n",
+		size, size - (__per_cpu_end - __per_cpu_start));
 
 	for_each_possible_cpu(i) {
 		__per_cpu_offset[i] = ptr - __per_cpu_start;
diff --git a/mm/allocpercpu.c b/mm/allocpercpu.c
--- a/mm/allocpercpu.c
+++ b/mm/allocpercpu.c
@@ -141,3 +141,8 @@
 	kfree(__percpu_disguise(__pdata));
 }
 EXPORT_SYMBOL_GPL(percpu_free);
+
+/* A heuristic based on observation.  May need to increase. */
+unsigned int percpu_reserve = (sizeof(unsigned long) * 2500);
+
+core_param(percpu, percpu_reserve, uint, 0644);

--
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/