diff options
| -rw-r--r-- | Documentation/vm/frontswap.txt | 4 | ||||
| -rw-r--r-- | mm/frontswap.c | 150 |
2 files changed, 92 insertions, 62 deletions
diff --git a/Documentation/vm/frontswap.txt b/Documentation/vm/frontswap.txt index 37067cf455f4..5ef2d1366425 100644 --- a/Documentation/vm/frontswap.txt +++ b/Documentation/vm/frontswap.txt @@ -25,7 +25,7 @@ with the specified swap device number (aka "type"). A "store" will copy the page to transcendent memory and associate it with the type and offset associated with the page. A "load" will copy the page, if found, from transcendent memory into kernel memory, but will NOT remove the page -from from transcendent memory. An "invalidate_page" will remove the page +from transcendent memory. An "invalidate_page" will remove the page from transcendent memory and an "invalidate_area" will remove ALL pages associated with the swap type (e.g., like swapoff) and notify the "device" to refuse further stores with that swap type. @@ -99,7 +99,7 @@ server configured with a large amount of RAM... without pre-configuring how much of the RAM is available for each of the clients! In the virtual case, the whole point of virtualization is to statistically -multiplex physical resources acrosst the varying demands of multiple +multiplex physical resources across the varying demands of multiple virtual machines. This is really hard to do with RAM and efforts to do it well with no kernel changes have essentially failed (except in some well-publicized special-case workloads). diff --git a/mm/frontswap.c b/mm/frontswap.c index e25025574a02..6b3e71a2cd48 100644 --- a/mm/frontswap.c +++ b/mm/frontswap.c @@ -11,15 +11,11 @@ * This work is licensed under the terms of the GNU GPL, version 2. */ -#include <linux/mm.h> #include <linux/mman.h> #include <linux/swap.h> #include <linux/swapops.h> -#include <linux/proc_fs.h> #include <linux/security.h> -#include <linux/capability.h> #include <linux/module.h> -#include <linux/uaccess.h> #include <linux/debugfs.h> #include <linux/frontswap.h> #include <linux/swapfile.h> @@ -110,16 +106,21 @@ void __frontswap_init(unsigned type) BUG_ON(sis == NULL); if (sis->frontswap_map == NULL) return; - if (frontswap_enabled) - (*frontswap_ops.init)(type); + frontswap_ops.init(type); } EXPORT_SYMBOL(__frontswap_init); +static inline void __frontswap_clear(struct swap_info_struct *sis, pgoff_t offset) +{ + frontswap_clear(sis, offset); + atomic_dec(&sis->frontswap_pages); +} + /* * "Store" data from a page to frontswap and associate it with the page's * swaptype and offset. Page must be locked and in the swap cache. * If frontswap already contains a page with matching swaptype and - * offset, the frontswap implmentation may either overwrite the data and + * offset, the frontswap implementation may either overwrite the data and * return success or invalidate the page from frontswap and return failure. */ int __frontswap_store(struct page *page) @@ -134,22 +135,21 @@ int __frontswap_store(struct page *page) BUG_ON(sis == NULL); if (frontswap_test(sis, offset)) dup = 1; - ret = (*frontswap_ops.store)(type, offset, page); + ret = frontswap_ops.store(type, offset, page); if (ret == 0) { frontswap_set(sis, offset); inc_frontswap_succ_stores(); if (!dup) atomic_inc(&sis->frontswap_pages); - } else if (dup) { + } else { /* failed dup always results in automatic invalidate of the (older) page from frontswap */ - frontswap_clear(sis, offset); - atomic_dec(&sis->frontswap_pages); - inc_frontswap_failed_stores(); - } else inc_frontswap_failed_stores(); + if (dup) + __frontswap_clear(sis, offset); + } if (frontswap_writethrough_enabled) /* report failure so swap also writes to swap device */ ret = -1; @@ -173,7 +173,7 @@ int __frontswap_load(struct page *page) BUG_ON(!PageLocked(page)); BUG_ON(sis == NULL); if (frontswap_test(sis, offset)) - ret = (*frontswap_ops.load)(type, offset, page); + ret = frontswap_ops.load(type, offset, page); if (ret == 0) inc_frontswap_loads(); return ret; @@ -190,9 +190,8 @@ void __frontswap_invalidate_page(unsigned type, pgoff_t offset) BUG_ON(sis == NULL); if (frontswap_test(sis, offset)) { - (*frontswap_ops.invalidate_page)(type, offset); - atomic_dec(&sis->frontswap_pages); - frontswap_clear(sis, offset); + frontswap_ops.invalidate_page(type, offset); + __frontswap_clear(sis, offset); inc_frontswap_invalidates(); } } @@ -209,67 +208,102 @@ void __frontswap_invalidate_area(unsigned type) BUG_ON(sis == NULL); if (sis->frontswap_map == NULL) return; - (*frontswap_ops.invalidate_area)(type); + frontswap_ops.invalidate_area(type); atomic_set(&sis->frontswap_pages, 0); memset(sis->frontswap_map, 0, sis->max / sizeof(long)); } EXPORT_SYMBOL(__frontswap_invalidate_area); -/* - * Frontswap, like a true swap device, may unnecessarily retain pages - * under certain circumstances; "shrink" frontswap is essentially a - * "partial swapoff" and works by calling try_to_unuse to attempt to - * unuse enough frontswap pages to attempt to -- subject to memory - * constraints -- reduce the number of pages in frontswap to the - * number given in the parameter target_pages. - */ -void frontswap_shrink(unsigned long target_pages) +static unsigned long __frontswap_curr_pages(void) { - struct swap_info_struct *si = NULL; - int si_frontswap_pages; - unsigned long total_pages = 0, total_pages_to_unuse; - unsigned long pages = 0, pages_to_unuse = 0; int type; - bool locked = false; + unsigned long totalpages = 0; + struct swap_info_struct *si = NULL; - /* - * we don't want to hold swap_lock while doing a very - * lengthy try_to_unuse, but swap_list may change - * so restart scan from swap_list.head each time - */ - spin_lock(&swap_lock); - locked = true; - total_pages = 0; + assert_spin_locked(&swap_lock); for (type = swap_list.head; type >= 0; type = si->next) { si = swap_info[type]; - total_pages += atomic_read(&si->frontswap_pages); + totalpages += atomic_read(&si->frontswap_pages); } - if (total_pages <= target_pages) - goto out; - total_pages_to_unuse = total_pages - target_pages; + return totalpages; +} + +static int __frontswap_unuse_pages(unsigned long total, unsigned long *unused, + int *swapid) +{ + int ret = -EINVAL; + struct swap_info_struct *si = NULL; + int si_frontswap_pages; + unsigned long total_pages_to_unuse = total; + unsigned long pages = 0, pages_to_unuse = 0; + int type; + + assert_spin_locked(&swap_lock); for (type = swap_list.head; type >= 0; type = si->next) { si = swap_info[type]; si_frontswap_pages = atomic_read(&si->frontswap_pages); - if (total_pages_to_unuse < si_frontswap_pages) + if (total_pages_to_unuse < si_frontswap_pages) { pages = pages_to_unuse = total_pages_to_unuse; - else { + } else { pages = si_frontswap_pages; pages_to_unuse = 0; /* unuse all */ } /* ensure there is enough RAM to fetch pages from frontswap */ - if (security_vm_enough_memory_mm(current->mm, pages)) + if (security_vm_enough_memory_mm(current->mm, pages)) { + ret = -ENOMEM; continue; + } vm_unacct_memory(pages); + *unused = pages_to_unuse; + *swapid = type; + ret = 0; break; } - if (type < 0) - goto out; - locked = false; + + return ret; +} + +static int __frontswap_shrink(unsigned long target_pages, + unsigned long *pages_to_unuse, + int *type) +{ + unsigned long total_pages = 0, total_pages_to_unuse; + + assert_spin_locked(&swap_lock); + + total_pages = __frontswap_curr_pages(); + if (total_pages <= target_pages) { + /* Nothing to do */ + *pages_to_unuse = 0; + return 0; + } + total_pages_to_unuse = total_pages - target_pages; + return __frontswap_unuse_pages(total_pages_to_unuse, pages_to_unuse, type); +} + +/* + * Frontswap, like a true swap device, may unnecessarily retain pages + * under certain circumstances; "shrink" frontswap is essentially a + * "partial swapoff" and works by calling try_to_unuse to attempt to + * unuse enough frontswap pages to attempt to -- subject to memory + * constraints -- reduce the number of pages in frontswap to the + * number given in the parameter target_pages. + */ +void frontswap_shrink(unsigned long target_pages) +{ + unsigned long pages_to_unuse = 0; + int type, ret; + + /* + * we don't want to hold swap_lock while doing a very + * lengthy try_to_unuse, but swap_list may change + * so restart scan from swap_list.head each time + */ + spin_lock(&swap_lock); + ret = __frontswap_shrink(target_pages, &pages_to_unuse, &type); spin_unlock(&swap_lock); - try_to_unuse(type, true, pages_to_unuse); -out: - if (locked) - spin_unlock(&swap_lock); + if (ret == 0 && pages_to_unuse) + try_to_unuse(type, true, pages_to_unuse); return; } EXPORT_SYMBOL(frontswap_shrink); @@ -281,16 +315,12 @@ EXPORT_SYMBOL(frontswap_shrink); */ unsigned long frontswap_curr_pages(void) { - int type; unsigned long totalpages = 0; - struct swap_info_struct *si = NULL; spin_lock(&swap_lock); - for (type = swap_list.head; type >= 0; type = si->next) { - si = swap_info[type]; - totalpages += atomic_read(&si->frontswap_pages); - } + totalpages = __frontswap_curr_pages(); spin_unlock(&swap_lock); + return totalpages; } EXPORT_SYMBOL(frontswap_curr_pages); |