Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755608AbZFXGsZ (ORCPT ); Wed, 24 Jun 2009 02:48:25 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1754412AbZFXGqc (ORCPT ); Wed, 24 Jun 2009 02:46:32 -0400 Received: from hera.kernel.org ([140.211.167.34]:57107 "EHLO hera.kernel.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754228AbZFXGq2 (ORCPT ); Wed, 24 Jun 2009 02:46:28 -0400 From: Tejun Heo To: linux-kernel@vger.kernel.org, x86@kernel.org, linux-arch@vger.kernel.org, mingo@elte.hu, kyle@mcmartin.ca, cl@linux-foundation.org, Jesper.Nilsson@axis.com, benh@kernel.crashing.org, paulmck@linux.vnet.ibm.com, rusty@rustcorp.com.au, torvalds@linux-foundation.org, akpm@linux-foundation.org Cc: Tejun Heo , Paul Mundt , Russell King , Mikael Starvik , Ralf Baechle , Bryan Wu , Matthew Wilcox , Grant Grundler , Hirokazu Takata , Richard Henderson , Ivan Kokshaysky , Heiko Carstens Subject: [PATCH 01/10] percpu: use dynamic percpu allocator as the default percpu allocator Date: Wed, 24 Jun 2009 15:45:15 +0900 Message-Id: <1245825924-30412-2-git-send-email-tj@kernel.org> X-Mailer: git-send-email 1.6.0.2 In-Reply-To: <1245825924-30412-1-git-send-email-tj@kernel.org> References: <1245825924-30412-1-git-send-email-tj@kernel.org> X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.0 (hera.kernel.org [127.0.0.1]); Wed, 24 Jun 2009 06:45:34 +0000 (UTC) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 12136 Lines: 405 This patch makes most !CONFIG_HAVE_SETUP_PER_CPU_AREA archs use dynamic percpu allocator. The first chunk is allocated using embedding helper and 8k is reserved for modules. This ensures that the new allocator behaves almost identically to the original allocator as long as static percpu variables are concerned, so it shouldn't introduce much breakage. s390 and alpha use custom SHIFT_PERCPU_PTR() to work around addressing range limit the addressing model imposes. Unfortunately, this breaks if the address is specified using a variable, so for now, the two archs aren't converted. The following architectures are affected by this change. * sh * arm * cris * mips * sparc(32) * blackfin * avr32 * parisc (broken, under investigation) * m32r * powerpc(32) As this change makes the dynamic allocator the default one, CONFIG_HAVE_DYNAMIC_PER_CPU_AREA is replaced with its invert - CONFIG_HAVE_LEGACY_PER_CPU_AREA, which is added to yet-to-be converted archs. These archs implement their own setup_per_cpu_areas() and the conversion is not trivial. * powerpc(64) * sparc(64) * ia64 * alpha * s390 Boot and batch alloc/free tests on x86_32 with debug code (x86_32 doesn't use default first chunk initialization). Compile tested on sparc(32), powerpc(32), arm and alpha. Kyle McMartin reported that this change breaks parisc. The problem is still under investigation and he is okay with pushing this patch forward and fixing parisc later. [ Impact: use dynamic allocator for most archs w/o custom percpu setup ] Signed-off-by: Tejun Heo Acked-by: Rusty Russell Acked-by: David S. Miller Acked-by: Benjamin Herrenschmidt Acked-by: Martin Schwidefsky Reviewed-by: Christoph Lameter Cc: Paul Mundt Cc: Russell King Cc: Mikael Starvik Cc: Ralf Baechle Cc: Bryan Wu Cc: Kyle McMartin Cc: Matthew Wilcox Cc: Grant Grundler Cc: Hirokazu Takata Cc: Richard Henderson Cc: Ivan Kokshaysky Cc: Heiko Carstens Cc: Ingo Molnar --- arch/alpha/Kconfig | 3 +++ arch/ia64/Kconfig | 3 +++ arch/powerpc/Kconfig | 3 +++ arch/s390/Kconfig | 3 +++ arch/sparc/Kconfig | 3 +++ arch/x86/Kconfig | 3 --- include/linux/percpu.h | 12 +++++++++--- init/main.c | 24 ------------------------ kernel/module.c | 6 +++--- mm/Makefile | 2 +- mm/allocpercpu.c | 28 ++++++++++++++++++++++++++++ mm/percpu.c | 40 +++++++++++++++++++++++++++++++++++++++- 12 files changed, 95 insertions(+), 35 deletions(-) diff --git a/arch/alpha/Kconfig b/arch/alpha/Kconfig index 9fb8aae..05d8640 100644 --- a/arch/alpha/Kconfig +++ b/arch/alpha/Kconfig @@ -70,6 +70,9 @@ config AUTO_IRQ_AFFINITY depends on SMP default y +config HAVE_LEGACY_PER_CPU_AREA + def_bool y + source "init/Kconfig" source "kernel/Kconfig.freezer" diff --git a/arch/ia64/Kconfig b/arch/ia64/Kconfig index 170042b..328d2f8 100644 --- a/arch/ia64/Kconfig +++ b/arch/ia64/Kconfig @@ -89,6 +89,9 @@ config GENERIC_TIME_VSYSCALL bool default y +config HAVE_LEGACY_PER_CPU_AREA + def_bool y + config HAVE_SETUP_PER_CPU_AREA def_bool y diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig index bf6cedf..a774c2a 100644 --- a/arch/powerpc/Kconfig +++ b/arch/powerpc/Kconfig @@ -46,6 +46,9 @@ config GENERIC_HARDIRQS_NO__DO_IRQ bool default y +config HAVE_LEGACY_PER_CPU_AREA + def_bool PPC64 + config HAVE_SETUP_PER_CPU_AREA def_bool PPC64 diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig index a14dba0..f4a3cc6 100644 --- a/arch/s390/Kconfig +++ b/arch/s390/Kconfig @@ -75,6 +75,9 @@ config VIRT_CPU_ACCOUNTING config ARCH_SUPPORTS_DEBUG_PAGEALLOC def_bool y +config HAVE_LEGACY_PER_CPU_AREA + def_bool y + mainmenu "Linux Kernel Configuration" config S390 diff --git a/arch/sparc/Kconfig b/arch/sparc/Kconfig index 3f8b6a9..7a8698b 100644 --- a/arch/sparc/Kconfig +++ b/arch/sparc/Kconfig @@ -92,6 +92,9 @@ config AUDIT_ARCH bool default y +config HAVE_LEGACY_PER_CPU_AREA + def_bool y if SPARC64 + config HAVE_SETUP_PER_CPU_AREA def_bool y if SPARC64 diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index d1430ef..a48a900 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -149,9 +149,6 @@ config ARCH_HAS_CACHE_LINE_SIZE config HAVE_SETUP_PER_CPU_AREA def_bool y -config HAVE_DYNAMIC_PER_CPU_AREA - def_bool y - config HAVE_CPUMASK_OF_CPU_MAP def_bool X86_64_SMP diff --git a/include/linux/percpu.h b/include/linux/percpu.h index 26fd9d1..e500034 100644 --- a/include/linux/percpu.h +++ b/include/linux/percpu.h @@ -34,7 +34,7 @@ #ifdef CONFIG_SMP -#ifdef CONFIG_HAVE_DYNAMIC_PER_CPU_AREA +#ifndef CONFIG_HAVE_LEGACY_PER_CPU_AREA /* minimum unit size, also is the maximum supported allocation size */ #define PCPU_MIN_UNIT_SIZE PFN_ALIGN(64 << 10) @@ -80,7 +80,7 @@ extern ssize_t __init pcpu_embed_first_chunk( extern void *__alloc_reserved_percpu(size_t size, size_t align); -#else /* CONFIG_HAVE_DYNAMIC_PER_CPU_AREA */ +#else /* CONFIG_HAVE_LEGACY_PER_CPU_AREA */ struct percpu_data { void *ptrs[1]; @@ -99,11 +99,15 @@ struct percpu_data { (__typeof__(ptr))__p->ptrs[(cpu)]; \ }) -#endif /* CONFIG_HAVE_DYNAMIC_PER_CPU_AREA */ +#endif /* CONFIG_HAVE_LEGACY_PER_CPU_AREA */ extern void *__alloc_percpu(size_t size, size_t align); extern void free_percpu(void *__pdata); +#ifndef CONFIG_HAVE_SETUP_PER_CPU_AREA +extern void __init setup_per_cpu_areas(void); +#endif + #else /* CONFIG_SMP */ #define per_cpu_ptr(ptr, cpu) ({ (void)(cpu); (ptr); }) @@ -124,6 +128,8 @@ static inline void free_percpu(void *p) kfree(p); } +static inline void __init setup_per_cpu_areas(void) { } + #endif /* CONFIG_SMP */ #define alloc_percpu(type) (type *)__alloc_percpu(sizeof(type), \ diff --git a/init/main.c b/init/main.c index 09131ec..602d724 100644 --- a/init/main.c +++ b/init/main.c @@ -357,7 +357,6 @@ static void __init smp_init(void) #define smp_init() do { } while (0) #endif -static inline void setup_per_cpu_areas(void) { } static inline void setup_nr_cpu_ids(void) { } static inline void smp_prepare_cpus(unsigned int maxcpus) { } @@ -378,29 +377,6 @@ static void __init setup_nr_cpu_ids(void) nr_cpu_ids = find_last_bit(cpumask_bits(cpu_possible_mask),NR_CPUS) + 1; } -#ifndef CONFIG_HAVE_SETUP_PER_CPU_AREA -unsigned long __per_cpu_offset[NR_CPUS] __read_mostly; - -EXPORT_SYMBOL(__per_cpu_offset); - -static void __init setup_per_cpu_areas(void) -{ - unsigned long size, i; - char *ptr; - unsigned long nr_possible_cpus = num_possible_cpus(); - - /* Copy section for each CPU (we discard the original) */ - size = ALIGN(PERCPU_ENOUGH_ROOM, PAGE_SIZE); - ptr = alloc_bootmem_pages(size * nr_possible_cpus); - - for_each_possible_cpu(i) { - __per_cpu_offset[i] = ptr - __per_cpu_start; - memcpy(ptr, __per_cpu_start, __per_cpu_end - __per_cpu_start); - ptr += size; - } -} -#endif /* CONFIG_HAVE_SETUP_PER_CPU_AREA */ - /* Called by boot processor to activate the rest. */ static void __init smp_init(void) { diff --git a/kernel/module.c b/kernel/module.c index 38928fc..f593495 100644 --- a/kernel/module.c +++ b/kernel/module.c @@ -364,7 +364,7 @@ EXPORT_SYMBOL_GPL(find_module); #ifdef CONFIG_SMP -#ifdef CONFIG_HAVE_DYNAMIC_PER_CPU_AREA +#ifndef CONFIG_HAVE_LEGACY_PER_CPU_AREA static void *percpu_modalloc(unsigned long size, unsigned long align, const char *name) @@ -389,7 +389,7 @@ static void percpu_modfree(void *freeme) free_percpu(freeme); } -#else /* ... !CONFIG_HAVE_DYNAMIC_PER_CPU_AREA */ +#else /* ... CONFIG_HAVE_LEGACY_PER_CPU_AREA */ /* Number of blocks used and allocated. */ static unsigned int pcpu_num_used, pcpu_num_allocated; @@ -535,7 +535,7 @@ static int percpu_modinit(void) } __initcall(percpu_modinit); -#endif /* CONFIG_HAVE_DYNAMIC_PER_CPU_AREA */ +#endif /* CONFIG_HAVE_LEGACY_PER_CPU_AREA */ static unsigned int find_pcpusec(Elf_Ehdr *hdr, Elf_Shdr *sechdrs, diff --git a/mm/Makefile b/mm/Makefile index 5e0bd64..c77c648 100644 --- a/mm/Makefile +++ b/mm/Makefile @@ -33,7 +33,7 @@ obj-$(CONFIG_FAILSLAB) += failslab.o obj-$(CONFIG_MEMORY_HOTPLUG) += memory_hotplug.o obj-$(CONFIG_FS_XIP) += filemap_xip.o obj-$(CONFIG_MIGRATION) += migrate.o -ifdef CONFIG_HAVE_DYNAMIC_PER_CPU_AREA +ifndef CONFIG_HAVE_LEGACY_PER_CPU_AREA obj-$(CONFIG_SMP) += percpu.o else obj-$(CONFIG_SMP) += allocpercpu.o diff --git a/mm/allocpercpu.c b/mm/allocpercpu.c index dfdee6a..df34cea 100644 --- a/mm/allocpercpu.c +++ b/mm/allocpercpu.c @@ -5,6 +5,8 @@ */ #include #include +#include +#include #ifndef cache_line_size #define cache_line_size() L1_CACHE_BYTES @@ -147,3 +149,29 @@ void free_percpu(void *__pdata) kfree(__percpu_disguise(__pdata)); } EXPORT_SYMBOL_GPL(free_percpu); + +/* + * Generic percpu area setup. + */ +#ifndef CONFIG_HAVE_SETUP_PER_CPU_AREA +unsigned long __per_cpu_offset[NR_CPUS] __read_mostly; + +EXPORT_SYMBOL(__per_cpu_offset); + +void __init setup_per_cpu_areas(void) +{ + unsigned long size, i; + char *ptr; + unsigned long nr_possible_cpus = num_possible_cpus(); + + /* Copy section for each CPU (we discard the original) */ + size = ALIGN(PERCPU_ENOUGH_ROOM, PAGE_SIZE); + ptr = alloc_bootmem_pages(size * nr_possible_cpus); + + for_each_possible_cpu(i) { + __per_cpu_offset[i] = ptr - __per_cpu_start; + memcpy(ptr, __per_cpu_start, __per_cpu_end - __per_cpu_start); + ptr += size; + } +} +#endif /* CONFIG_HAVE_SETUP_PER_CPU_AREA */ diff --git a/mm/percpu.c b/mm/percpu.c index b70f2ac..b149845 100644 --- a/mm/percpu.c +++ b/mm/percpu.c @@ -43,7 +43,7 @@ * * To use this allocator, arch code should do the followings. * - * - define CONFIG_HAVE_DYNAMIC_PER_CPU_AREA + * - drop CONFIG_HAVE_LEGACY_PER_CPU_AREA * * - define __addr_to_pcpu_ptr() and __pcpu_ptr_to_addr() to translate * regular address to percpu pointer and back if they need to be @@ -1275,3 +1275,41 @@ ssize_t __init pcpu_embed_first_chunk(size_t static_size, size_t reserved_size, reserved_size, dyn_size, pcpue_unit_size, pcpue_ptr, NULL); } + +/* + * Generic percpu area setup. + * + * The embedding helper is used because its behavior closely resembles + * the original non-dynamic generic percpu area setup. This is + * important because many archs have addressing restrictions and might + * fail if the percpu area is located far away from the previous + * location. As an added bonus, in non-NUMA cases, embedding is + * generally a good idea TLB-wise because percpu area can piggy back + * on the physical linear memory mapping which uses large page + * mappings on applicable archs. + */ +#ifndef CONFIG_HAVE_SETUP_PER_CPU_AREA +unsigned long __per_cpu_offset[NR_CPUS] __read_mostly; +EXPORT_SYMBOL(__per_cpu_offset); + +void __init setup_per_cpu_areas(void) +{ + size_t static_size = __per_cpu_end - __per_cpu_start; + ssize_t unit_size; + unsigned long delta; + unsigned int cpu; + + /* + * Always reserve area for module percpu variables. That's + * what the legacy allocator did. + */ + unit_size = pcpu_embed_first_chunk(static_size, PERCPU_MODULE_RESERVE, + PERCPU_DYNAMIC_RESERVE, -1); + if (unit_size < 0) + panic("Failed to initialized percpu areas."); + + delta = (unsigned long)pcpu_base_addr - (unsigned long)__per_cpu_start; + for_each_possible_cpu(cpu) + __per_cpu_offset[cpu] = delta + cpu * unit_size; +} +#endif /* CONFIG_HAVE_SETUP_PER_CPU_AREA */ -- 1.6.0.2 -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/