From d9b55eeb1d55ef2dc5a4fdbff9604c2c68cb5649 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Tue, 24 Feb 2009 11:57:21 +0900 Subject: percpu: remove unit_size power-of-2 restriction Impact: allow unit_size to be arbitrary multiple of PAGE_SIZE In dynamic percpu allocator, there is no reason the unit size should be power of two. Remove the restriction. As non-power-of-two unit size means that empty chunks fall into the same slot index as lightly occupied chunks which is bad for reclaming. Reserve an extra slot for empty chunks. Signed-off-by: Tejun Heo --- mm/percpu.c | 33 +++++++++++++++++++-------------- 1 file changed, 19 insertions(+), 14 deletions(-) (limited to 'mm') diff --git a/mm/percpu.c b/mm/percpu.c index 41e7a5f5ab1..d9e6e5d1dbd 100644 --- a/mm/percpu.c +++ b/mm/percpu.c @@ -67,7 +67,7 @@ #include #include -#define PCPU_MIN_UNIT_PAGES_SHIFT 4 /* also max alloc size */ +#define PCPU_MIN_UNIT_PAGES 16 /* max alloc size in pages */ #define PCPU_SLOT_BASE_SHIFT 5 /* 1-31 shares the same slot */ #define PCPU_DFL_MAP_ALLOC 16 /* start a map with 16 ents */ @@ -83,9 +83,7 @@ struct pcpu_chunk { struct page *page[]; /* #cpus * UNIT_PAGES */ }; -static int pcpu_unit_pages_shift; static int pcpu_unit_pages; -static int pcpu_unit_shift; static int pcpu_unit_size; static int pcpu_chunk_size; static int pcpu_nr_slots; @@ -117,12 +115,19 @@ static DEFINE_MUTEX(pcpu_mutex); static struct list_head *pcpu_slot; /* chunk list slots */ static struct rb_root pcpu_addr_root = RB_ROOT; /* chunks by address */ -static int pcpu_size_to_slot(int size) +static int __pcpu_size_to_slot(int size) { int highbit = fls(size); /* size is in bytes */ return max(highbit - PCPU_SLOT_BASE_SHIFT + 2, 1); } +static int pcpu_size_to_slot(int size) +{ + if (size == pcpu_unit_size) + return pcpu_nr_slots - 1; + return __pcpu_size_to_slot(size); +} + static int pcpu_chunk_slot(const struct pcpu_chunk *chunk) { if (chunk->free_size < sizeof(int) || chunk->contig_hint < sizeof(int)) @@ -133,7 +138,7 @@ static int pcpu_chunk_slot(const struct pcpu_chunk *chunk) static int pcpu_page_idx(unsigned int cpu, int page_idx) { - return (cpu << pcpu_unit_pages_shift) + page_idx; + return cpu * pcpu_unit_pages + page_idx; } static struct page **pcpu_chunk_pagep(struct pcpu_chunk *chunk, @@ -659,7 +664,7 @@ static int pcpu_populate_chunk(struct pcpu_chunk *chunk, int off, int size) goto err; for_each_possible_cpu(cpu) - memset(chunk->vm->addr + (cpu << pcpu_unit_shift) + off, 0, + memset(chunk->vm->addr + cpu * pcpu_unit_size + off, 0, size); return 0; @@ -722,7 +727,7 @@ void *__alloc_percpu(size_t size, size_t align) struct pcpu_chunk *chunk; int slot, off; - if (unlikely(!size || size > PAGE_SIZE << PCPU_MIN_UNIT_PAGES_SHIFT || + if (unlikely(!size || size > PCPU_MIN_UNIT_PAGES * PAGE_SIZE || align > PAGE_SIZE)) { WARN(true, "illegal size (%zu) or align (%zu) for " "percpu allocation\n", size, align); @@ -840,19 +845,19 @@ size_t __init pcpu_setup_static(pcpu_populate_pte_fn_t populate_pte_fn, unsigned int cpu; int err, i; - pcpu_unit_pages_shift = max_t(int, PCPU_MIN_UNIT_PAGES_SHIFT, - order_base_2(cpu_size) - PAGE_SHIFT); + pcpu_unit_pages = max_t(int, PCPU_MIN_UNIT_PAGES, PFN_UP(cpu_size)); pcpu_static_size = cpu_size; - pcpu_unit_pages = 1 << pcpu_unit_pages_shift; - pcpu_unit_shift = PAGE_SHIFT + pcpu_unit_pages_shift; - pcpu_unit_size = 1 << pcpu_unit_shift; + pcpu_unit_size = pcpu_unit_pages << PAGE_SHIFT; pcpu_chunk_size = num_possible_cpus() * pcpu_unit_size; - pcpu_nr_slots = pcpu_size_to_slot(pcpu_unit_size) + 1; pcpu_chunk_struct_size = sizeof(struct pcpu_chunk) + num_possible_cpus() * pcpu_unit_pages * sizeof(struct page *); - /* allocate chunk slots */ + /* + * Allocate chunk slots. The additional last slot is for + * empty chunks. + */ + pcpu_nr_slots = __pcpu_size_to_slot(pcpu_unit_size) + 2; pcpu_slot = alloc_bootmem(pcpu_nr_slots * sizeof(pcpu_slot[0])); for (i = 0; i < pcpu_nr_slots; i++) INIT_LIST_HEAD(&pcpu_slot[i]); -- cgit v1.2.3