renegade-project/linux-next
2
0
Fork 0
mirror of https://github.com/edk2-porting/linux-next.git synced 2024-12-21 19:53:59 +08:00
Code
6b692f3b66
linux-next/fs/f2fs/gc.c

730 lines
17 KiB
C
Raw Normal View History

f2fs: adjust kernel coding style As pointed out by Randy Dunlap, this patch removes all usage of "/**" for comment blocks. Instead, just use "/*". Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-29 12:28:09 +08:00
/*
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
* fs/f2fs/gc.c
*
* Copyright (c) 2012 Samsung Electronics Co., Ltd.
* http://www.samsung.com/
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License version 2 as
* published by the Free Software Foundation.
*/
#include <linux/fs.h>
#include <linux/module.h>
#include <linux/backing-dev.h>
#include <linux/init.h>
#include <linux/f2fs_fs.h>
#include <linux/kthread.h>
#include <linux/delay.h>
#include <linux/freezer.h>
#include <linux/blkdev.h>
#include "f2fs.h"
#include "node.h"
#include "segment.h"
#include "gc.h"
f2fs: add tracepoints for GC threads Add tracepoints for tracing the garbage collector threads in f2fs with status of collection & type. Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Pankaj Kumar <pankaj.km@samsung.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> [Jaegeuk: modify slightly to show information] Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-04-23 15:42:53 +08:00
#include <trace/events/f2fs.h>
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
static struct kmem_cache *winode_slab;
static int gc_thread_func(void *data)
{
struct f2fs_sb_info *sbi = data;
wait_queue_head_t *wq = &sbi->gc_thread->gc_wait_queue_head;
long wait_ms;
wait_ms = GC_THREAD_MIN_SLEEP_TIME;
do {
if (try_to_freeze())
continue;
else
wait_event_interruptible_timeout(*wq,
kthread_should_stop(),
msecs_to_jiffies(wait_ms));
if (kthread_should_stop())
break;
f2fs: add un/freeze_fs into super_operations This patch supports ioctl FIFREEZE and FITHAW to snapshot filesystem. Before calling f2fs_freeze, all writers would be suspended and sync_fs would be completed. So no f2fs has to do something. Just background gc operation should be skipped due to generate dirty nodes and data until unfreeze. Signed-off-by: Changman Lee <cm224.lee@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-01-29 17:30:07 +08:00
if (sbi->sb->s_writers.frozen >= SB_FREEZE_WRITE) {
wait_ms = GC_THREAD_MAX_SLEEP_TIME;
continue;
}
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
/*
* [GC triggering condition]
* 0. GC is not conducted currently.
* 1. There are enough dirty segments.
* 2. IO subsystem is idle by checking the # of writeback pages.
* 3. IO subsystem is idle by checking the # of requests in
* bdev's request list.
*
* Note) We have to avoid triggering GCs too much frequently.
* Because it is possible that some segments can be
* invalidated soon after by user update or deletion.
* So, I'd like to wait some time to collect dirty segments.
*/
if (!mutex_trylock(&sbi->gc_mutex))
continue;
if (!is_idle(sbi)) {
wait_ms = increase_sleep_time(wait_ms);
mutex_unlock(&sbi->gc_mutex);
continue;
}
if (has_enough_invalid_blocks(sbi))
wait_ms = decrease_sleep_time(wait_ms);
else
wait_ms = increase_sleep_time(wait_ms);
sbi->bg_gc++;
f2fs: clarify and enhance the f2fs_gc flow This patch makes clearer the ambiguous f2fs_gc flow as follows. 1. Remove intermediate checkpoint condition during f2fs_gc (i.e., should_do_checkpoint() and GC_BLOCKED) 2. Remove unnecessary return values of f2fs_gc because of #1. (i.e., GC_NODE, GC_OK, etc) 3. Simplify write_checkpoint() because of #2. 4. Clarify the main f2fs_gc flow. o monitor how many freed sections during one iteration of do_garbage_collect(). o do GC more without checkpoints if we can't get enough free sections. o do checkpoint once we've got enough free sections through forground GCs. 5. Adopt thread-logging (Slack-Space-Recycle) scheme more aggressively on data log types. See. get_ssr_segement() Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-02-04 14:11:17 +08:00
/* if return value is not zero, no victim was selected */
if (f2fs_gc(sbi))
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
wait_ms = GC_THREAD_NOGC_SLEEP_TIME;
} while (!kthread_should_stop());
return 0;
}
int start_gc_thread(struct f2fs_sb_info *sbi)
{
f2fs: remove unneeded initialization No need to initialize "struct f2fs_gc_kthread *gc_th = NULL", as gc_th = NULL, will be taken care by the return values of kmalloc(). And fix codes in other places. Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Amit Sahrawat <a.sahrawat@samsung.com>
2012-12-01 09:56:13 +08:00
struct f2fs_gc_kthread *gc_th;
f2fs: name gc task as per the block device Currently GC task is started for each f2fs formatted/mounted device. But, when we check the task list, using 'ps', there is no distinguishing factor between the tasks. So, name the task as per the block device just like the flusher threads. Also, remove the macro GC_THREAD_NAME and instead use the name: f2fs_gc to avoid name length truncation, as the command length is 16 -> TASK_COMM_LEN 16 and example name like: f2fs_gc_task:8:16 -> this exceeds name length Before Patch for 2 F2FS formatted partitions: root 28061 0.0 0.0 0 0 ? S 10:31 0:00 [f2fs_gc_task] root 28087 0.0 0.0 0 0 ? S 10:32 0:00 [f2fs_gc_task] After Patch: root 16756 0.0 0.0 0 0 ? S 14:57 0:00 [f2fs_gc-8:18] root 16765 0.0 0.0 0 0 ? S 14:57 0:00 [f2fs_gc-8:19] Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Amit Sahrawat <a.sahrawat@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-02-02 22:52:28 +08:00
dev_t dev = sbi->sb->s_bdev->bd_dev;
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
f2fs: remove unnecessary gc option check and balance_fs 1. If f2fs is mounted with background_gc_off option, checking BG_GC is not redundant. 2. f2fs_balance_fs is checked in f2fs_gc, so this is also redundant. Signed-off-by: Changman Lee <cm224.lee@samsung.com> Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Amit Sahrawat <a.sahrawat@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-02-04 09:05:09 +08:00
if (!test_opt(sbi, BG_GC))
return 0;
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
gc_th = kmalloc(sizeof(struct f2fs_gc_kthread), GFP_KERNEL);
if (!gc_th)
return -ENOMEM;
sbi->gc_thread = gc_th;
init_waitqueue_head(&sbi->gc_thread->gc_wait_queue_head);
sbi->gc_thread->f2fs_gc_task = kthread_run(gc_thread_func, sbi,
f2fs: name gc task as per the block device Currently GC task is started for each f2fs formatted/mounted device. But, when we check the task list, using 'ps', there is no distinguishing factor between the tasks. So, name the task as per the block device just like the flusher threads. Also, remove the macro GC_THREAD_NAME and instead use the name: f2fs_gc to avoid name length truncation, as the command length is 16 -> TASK_COMM_LEN 16 and example name like: f2fs_gc_task:8:16 -> this exceeds name length Before Patch for 2 F2FS formatted partitions: root 28061 0.0 0.0 0 0 ? S 10:31 0:00 [f2fs_gc_task] root 28087 0.0 0.0 0 0 ? S 10:32 0:00 [f2fs_gc_task] After Patch: root 16756 0.0 0.0 0 0 ? S 14:57 0:00 [f2fs_gc-8:18] root 16765 0.0 0.0 0 0 ? S 14:57 0:00 [f2fs_gc-8:19] Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Amit Sahrawat <a.sahrawat@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-02-02 22:52:28 +08:00
"f2fs_gc-%u:%u", MAJOR(dev), MINOR(dev));
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
if (IS_ERR(gc_th->f2fs_gc_task)) {
kfree(gc_th);
f2fs: mark gc_thread as NULL when thread creation is failed When gc thread creation is failed, mark gc_thread as NULL to avoid crash while trying to stop invalid thread in stop_gc_thread->kthread_stop. Instead make it return from: if (!gc_th) return; Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Amit Sahrawat <a.sahrawat@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-02-02 22:52:42 +08:00
sbi->gc_thread = NULL;
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
return -ENOMEM;
}
return 0;
}
void stop_gc_thread(struct f2fs_sb_info *sbi)
{
struct f2fs_gc_kthread *gc_th = sbi->gc_thread;
if (!gc_th)
return;
kthread_stop(gc_th->f2fs_gc_task);
kfree(gc_th);
sbi->gc_thread = NULL;
}
static int select_gc_type(int gc_type)
{
return (gc_type == BG_GC) ? GC_CB : GC_GREEDY;
}
static void select_policy(struct f2fs_sb_info *sbi, int gc_type,
int type, struct victim_sel_policy *p)
{
struct dirty_seglist_info *dirty_i = DIRTY_I(sbi);
f2fs: check completion of foreground GC The foreground GCs are triggered under not enough free sections. So, we should not skip moving valid blocks in the victim segments. Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-03-31 12:49:18 +08:00
if (p->alloc_mode == SSR) {
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
p->gc_mode = GC_GREEDY;
p->dirty_segmap = dirty_i->dirty_segmap[type];
p->ofs_unit = 1;
} else {
p->gc_mode = select_gc_type(gc_type);
p->dirty_segmap = dirty_i->dirty_segmap[DIRTY];
p->ofs_unit = sbi->segs_per_sec;
}
p->offset = sbi->last_victim[p->gc_mode];
}
static unsigned int get_max_cost(struct f2fs_sb_info *sbi,
struct victim_sel_policy *p)
{
f2fs: fix calculation of max. gc cost in the SSR case In the SSR case, the max gc cost should be the number of pages in a segment. Otherwise, f2fs is able to fail getting dirty segments frequently for SSR. In get_victim_by_default() previously, while(1) { ... cost = get_gc_cost(); <- cost is between 0 ~ 512. ... if (cost == get_max_cost(sbi, &p)) <- max cost is UINT_MAX due to GC_CB type continue; if (nsearched++ >= MAX_VICTIM_SEARCH) break; } So, if there are a number of fully valid segments in series, f2fs cannot skip those segments by comparing the cost and max cost of each segment. Note that, the cost is the number of valid blocks at the time of the last checkpoint. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-02-05 12:19:28 +08:00
/* SSR allocates in a segment unit */
if (p->alloc_mode == SSR)
return 1 << sbi->log_blocks_per_seg;
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
if (p->gc_mode == GC_GREEDY)
return (1 << sbi->log_blocks_per_seg) * p->ofs_unit;
else if (p->gc_mode == GC_CB)
return UINT_MAX;
else /* No other gc_mode */
return 0;
}
static unsigned int check_bg_victims(struct f2fs_sb_info *sbi)
{
struct dirty_seglist_info *dirty_i = DIRTY_I(sbi);
f2fs: change GC bitmaps to apply the section granularity This patch removes a bitmap for victim segments selected by foreground GC, and modifies the other bitmap for victim segments selected by background GC. 1) foreground GC bitmap : We don't need to manage this, since we just only one previous victim section number instead of the whole victim history. The f2fs uses the victim section number in order not to allocate currently GC'ed section to current active logs. 2) background GC bitmap : This bitmap is used to avoid selecting victims repeatedly by background GCs. In addition, the victims are able to be selected by foreground GCs, since there is no need to read victim blocks during foreground GCs. By the fact that the foreground GC reclaims segments in a section unit, it'd be better to manage this bitmap based on the section granularity. Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-03-31 12:26:03 +08:00
unsigned int hint = 0;
unsigned int secno;
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
/*
* If the gc_type is FG_GC, we can select victim segments
* selected by background GC before.
* Those segments guarantee they have small valid blocks.
*/
f2fs: change GC bitmaps to apply the section granularity This patch removes a bitmap for victim segments selected by foreground GC, and modifies the other bitmap for victim segments selected by background GC. 1) foreground GC bitmap : We don't need to manage this, since we just only one previous victim section number instead of the whole victim history. The f2fs uses the victim section number in order not to allocate currently GC'ed section to current active logs. 2) background GC bitmap : This bitmap is used to avoid selecting victims repeatedly by background GCs. In addition, the victims are able to be selected by foreground GCs, since there is no need to read victim blocks during foreground GCs. By the fact that the foreground GC reclaims segments in a section unit, it'd be better to manage this bitmap based on the section granularity. Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-03-31 12:26:03 +08:00
next:
secno = find_next_bit(dirty_i->victim_secmap, TOTAL_SECS(sbi), hint++);
if (secno < TOTAL_SECS(sbi)) {
if (sec_usage_check(sbi, secno))
goto next;
clear_bit(secno, dirty_i->victim_secmap);
return secno * sbi->segs_per_sec;
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
}
return NULL_SEGNO;
}
static unsigned int get_cb_cost(struct f2fs_sb_info *sbi, unsigned int segno)
{
struct sit_info *sit_i = SIT_I(sbi);
unsigned int secno = GET_SECNO(sbi, segno);
unsigned int start = secno * sbi->segs_per_sec;
unsigned long long mtime = 0;
unsigned int vblocks;
unsigned char age = 0;
unsigned char u;
unsigned int i;
for (i = 0; i < sbi->segs_per_sec; i++)
mtime += get_seg_entry(sbi, start + i)->mtime;
vblocks = get_valid_blocks(sbi, segno, sbi->segs_per_sec);
mtime = div_u64(mtime, sbi->segs_per_sec);
vblocks = div_u64(vblocks, sbi->segs_per_sec);
u = (vblocks * 100) >> sbi->log_blocks_per_seg;
/* Handle if the system time is changed by user */
if (mtime < sit_i->min_mtime)
sit_i->min_mtime = mtime;
if (mtime > sit_i->max_mtime)
sit_i->max_mtime = mtime;
if (sit_i->max_mtime != sit_i->min_mtime)
age = 100 - div64_u64(100 * (mtime - sit_i->min_mtime),
sit_i->max_mtime - sit_i->min_mtime);
return UINT_MAX - ((100 * (100 - u) * age) / (100 + u));
}
static unsigned int get_gc_cost(struct f2fs_sb_info *sbi, unsigned int segno,
struct victim_sel_policy *p)
{
if (p->alloc_mode == SSR)
return get_seg_entry(sbi, segno)->ckpt_valid_blocks;
/* alloc_mode == LFS */
if (p->gc_mode == GC_GREEDY)
return get_valid_blocks(sbi, segno, sbi->segs_per_sec);
else
return get_cb_cost(sbi, segno);
}
f2fs: adjust kernel coding style As pointed out by Randy Dunlap, this patch removes all usage of "/**" for comment blocks. Instead, just use "/*". Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-29 12:28:09 +08:00
/*
f2fs: fix typo in comments Correct spelling typo in comments Signed-off-by: Masanari Iida <standby24x7@gmail.com> Acked-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-03-19 07:03:35 +08:00
* This function is called from two paths.
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
* One is garbage collection and the other is SSR segment selection.
* When it is called during GC, it just gets a victim segment
* and it does not remove it from dirty seglist.
* When it is called from SSR segment selection, it finds a segment
* which has minimum valid blocks and removes it from dirty seglist.
*/
static int get_victim_by_default(struct f2fs_sb_info *sbi,
unsigned int *result, int gc_type, int type, char alloc_mode)
{
struct dirty_seglist_info *dirty_i = DIRTY_I(sbi);
struct victim_sel_policy p;
f2fs: change GC bitmaps to apply the section granularity This patch removes a bitmap for victim segments selected by foreground GC, and modifies the other bitmap for victim segments selected by background GC. 1) foreground GC bitmap : We don't need to manage this, since we just only one previous victim section number instead of the whole victim history. The f2fs uses the victim section number in order not to allocate currently GC'ed section to current active logs. 2) background GC bitmap : This bitmap is used to avoid selecting victims repeatedly by background GCs. In addition, the victims are able to be selected by foreground GCs, since there is no need to read victim blocks during foreground GCs. By the fact that the foreground GC reclaims segments in a section unit, it'd be better to manage this bitmap based on the section granularity. Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-03-31 12:26:03 +08:00
unsigned int secno;
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
int nsearched = 0;
p.alloc_mode = alloc_mode;
select_policy(sbi, gc_type, type, &p);
p.min_segno = NULL_SEGNO;
p.min_cost = get_max_cost(sbi, &p);
mutex_lock(&dirty_i->seglist_lock);
if (p.alloc_mode == LFS && gc_type == FG_GC) {
p.min_segno = check_bg_victims(sbi);
if (p.min_segno != NULL_SEGNO)
goto got_it;
}
while (1) {
unsigned long cost;
f2fs: change GC bitmaps to apply the section granularity This patch removes a bitmap for victim segments selected by foreground GC, and modifies the other bitmap for victim segments selected by background GC. 1) foreground GC bitmap : We don't need to manage this, since we just only one previous victim section number instead of the whole victim history. The f2fs uses the victim section number in order not to allocate currently GC'ed section to current active logs. 2) background GC bitmap : This bitmap is used to avoid selecting victims repeatedly by background GCs. In addition, the victims are able to be selected by foreground GCs, since there is no need to read victim blocks during foreground GCs. By the fact that the foreground GC reclaims segments in a section unit, it'd be better to manage this bitmap based on the section granularity. Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-03-31 12:26:03 +08:00
unsigned int segno;
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
segno = find_next_bit(p.dirty_segmap,
TOTAL_SEGS(sbi), p.offset);
if (segno >= TOTAL_SEGS(sbi)) {
if (sbi->last_victim[p.gc_mode]) {
sbi->last_victim[p.gc_mode] = 0;
p.offset = 0;
continue;
}
break;
}
p.offset = ((segno / p.ofs_unit) * p.ofs_unit) + p.ofs_unit;
f2fs: change GC bitmaps to apply the section granularity This patch removes a bitmap for victim segments selected by foreground GC, and modifies the other bitmap for victim segments selected by background GC. 1) foreground GC bitmap : We don't need to manage this, since we just only one previous victim section number instead of the whole victim history. The f2fs uses the victim section number in order not to allocate currently GC'ed section to current active logs. 2) background GC bitmap : This bitmap is used to avoid selecting victims repeatedly by background GCs. In addition, the victims are able to be selected by foreground GCs, since there is no need to read victim blocks during foreground GCs. By the fact that the foreground GC reclaims segments in a section unit, it'd be better to manage this bitmap based on the section granularity. Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-03-31 12:26:03 +08:00
secno = GET_SECNO(sbi, segno);
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
f2fs: change GC bitmaps to apply the section granularity This patch removes a bitmap for victim segments selected by foreground GC, and modifies the other bitmap for victim segments selected by background GC. 1) foreground GC bitmap : We don't need to manage this, since we just only one previous victim section number instead of the whole victim history. The f2fs uses the victim section number in order not to allocate currently GC'ed section to current active logs. 2) background GC bitmap : This bitmap is used to avoid selecting victims repeatedly by background GCs. In addition, the victims are able to be selected by foreground GCs, since there is no need to read victim blocks during foreground GCs. By the fact that the foreground GC reclaims segments in a section unit, it'd be better to manage this bitmap based on the section granularity. Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-03-31 12:26:03 +08:00
if (sec_usage_check(sbi, secno))
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
continue;
f2fs: change GC bitmaps to apply the section granularity This patch removes a bitmap for victim segments selected by foreground GC, and modifies the other bitmap for victim segments selected by background GC. 1) foreground GC bitmap : We don't need to manage this, since we just only one previous victim section number instead of the whole victim history. The f2fs uses the victim section number in order not to allocate currently GC'ed section to current active logs. 2) background GC bitmap : This bitmap is used to avoid selecting victims repeatedly by background GCs. In addition, the victims are able to be selected by foreground GCs, since there is no need to read victim blocks during foreground GCs. By the fact that the foreground GC reclaims segments in a section unit, it'd be better to manage this bitmap based on the section granularity. Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-03-31 12:26:03 +08:00
if (gc_type == BG_GC && test_bit(secno, dirty_i->victim_secmap))
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
continue;
cost = get_gc_cost(sbi, segno, &p);
if (p.min_cost > cost) {
p.min_segno = segno;
p.min_cost = cost;
}
if (cost == get_max_cost(sbi, &p))
continue;
if (nsearched++ >= MAX_VICTIM_SEARCH) {
sbi->last_victim[p.gc_mode] = segno;
break;
}
}
got_it:
if (p.min_segno != NULL_SEGNO) {
if (p.alloc_mode == LFS) {
f2fs: change GC bitmaps to apply the section granularity This patch removes a bitmap for victim segments selected by foreground GC, and modifies the other bitmap for victim segments selected by background GC. 1) foreground GC bitmap : We don't need to manage this, since we just only one previous victim section number instead of the whole victim history. The f2fs uses the victim section number in order not to allocate currently GC'ed section to current active logs. 2) background GC bitmap : This bitmap is used to avoid selecting victims repeatedly by background GCs. In addition, the victims are able to be selected by foreground GCs, since there is no need to read victim blocks during foreground GCs. By the fact that the foreground GC reclaims segments in a section unit, it'd be better to manage this bitmap based on the section granularity. Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-03-31 12:26:03 +08:00
secno = GET_SECNO(sbi, p.min_segno);
if (gc_type == FG_GC)
sbi->cur_victim_sec = secno;
else
set_bit(secno, dirty_i->victim_secmap);
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
}
f2fs: change GC bitmaps to apply the section granularity This patch removes a bitmap for victim segments selected by foreground GC, and modifies the other bitmap for victim segments selected by background GC. 1) foreground GC bitmap : We don't need to manage this, since we just only one previous victim section number instead of the whole victim history. The f2fs uses the victim section number in order not to allocate currently GC'ed section to current active logs. 2) background GC bitmap : This bitmap is used to avoid selecting victims repeatedly by background GCs. In addition, the victims are able to be selected by foreground GCs, since there is no need to read victim blocks during foreground GCs. By the fact that the foreground GC reclaims segments in a section unit, it'd be better to manage this bitmap based on the section granularity. Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-03-31 12:26:03 +08:00
*result = (p.min_segno / p.ofs_unit) * p.ofs_unit;
f2fs: add tracepoints for GC threads Add tracepoints for tracing the garbage collector threads in f2fs with status of collection & type. Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Pankaj Kumar <pankaj.km@samsung.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> [Jaegeuk: modify slightly to show information] Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-04-23 15:42:53 +08:00
trace_f2fs_get_victim(sbi->sb, type, gc_type, &p,
sbi->cur_victim_sec,
prefree_segments(sbi), free_segments(sbi));
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
}
mutex_unlock(&dirty_i->seglist_lock);
return (p.min_segno == NULL_SEGNO) ? 0 : 1;
}
static const struct victim_selection default_v_ops = {
.get_victim = get_victim_by_default,
};
static struct inode *find_gc_inode(nid_t ino, struct list_head *ilist)
{
struct list_head *this;
struct inode_entry *ie;
list_for_each(this, ilist) {
ie = list_entry(this, struct inode_entry, list);
if (ie->inode->i_ino == ino)
return ie->inode;
}
return NULL;
}
static void add_gc_inode(struct inode *inode, struct list_head *ilist)
{
struct list_head *this;
struct inode_entry *new_ie, *ie;
list_for_each(this, ilist) {
ie = list_entry(this, struct inode_entry, list);
if (ie->inode == inode) {
iput(inode);
return;
}
}
repeat:
new_ie = kmem_cache_alloc(winode_slab, GFP_NOFS);
if (!new_ie) {
cond_resched();
goto repeat;
}
new_ie->inode = inode;
list_add_tail(&new_ie->list, ilist);
}
static void put_gc_inode(struct list_head *ilist)
{
struct inode_entry *ie, *next_ie;
list_for_each_entry_safe(ie, next_ie, ilist, list) {
iput(ie->inode);
list_del(&ie->list);
kmem_cache_free(winode_slab, ie);
}
}
static int check_valid_map(struct f2fs_sb_info *sbi,
unsigned int segno, int offset)
{
struct sit_info *sit_i = SIT_I(sbi);
struct seg_entry *sentry;
int ret;
mutex_lock(&sit_i->sentry_lock);
sentry = get_seg_entry(sbi, segno);
ret = f2fs_test_bit(offset, sentry->cur_valid_map);
mutex_unlock(&sit_i->sentry_lock);
f2fs: clarify and enhance the f2fs_gc flow This patch makes clearer the ambiguous f2fs_gc flow as follows. 1. Remove intermediate checkpoint condition during f2fs_gc (i.e., should_do_checkpoint() and GC_BLOCKED) 2. Remove unnecessary return values of f2fs_gc because of #1. (i.e., GC_NODE, GC_OK, etc) 3. Simplify write_checkpoint() because of #2. 4. Clarify the main f2fs_gc flow. o monitor how many freed sections during one iteration of do_garbage_collect(). o do GC more without checkpoints if we can't get enough free sections. o do checkpoint once we've got enough free sections through forground GCs. 5. Adopt thread-logging (Slack-Space-Recycle) scheme more aggressively on data log types. See. get_ssr_segement() Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-02-04 14:11:17 +08:00
return ret;
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
}
f2fs: adjust kernel coding style As pointed out by Randy Dunlap, this patch removes all usage of "/**" for comment blocks. Instead, just use "/*". Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-29 12:28:09 +08:00
/*
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
* This function compares node address got in summary with that in NAT.
* On validity, copy that node with cold status, otherwise (invalid node)
* ignore that.
*/
f2fs: clarify and enhance the f2fs_gc flow This patch makes clearer the ambiguous f2fs_gc flow as follows. 1. Remove intermediate checkpoint condition during f2fs_gc (i.e., should_do_checkpoint() and GC_BLOCKED) 2. Remove unnecessary return values of f2fs_gc because of #1. (i.e., GC_NODE, GC_OK, etc) 3. Simplify write_checkpoint() because of #2. 4. Clarify the main f2fs_gc flow. o monitor how many freed sections during one iteration of do_garbage_collect(). o do GC more without checkpoints if we can't get enough free sections. o do checkpoint once we've got enough free sections through forground GCs. 5. Adopt thread-logging (Slack-Space-Recycle) scheme more aggressively on data log types. See. get_ssr_segement() Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-02-04 14:11:17 +08:00
static void gc_node_segment(struct f2fs_sb_info *sbi,
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
struct f2fs_summary *sum, unsigned int segno, int gc_type)
{
bool initial = true;
struct f2fs_summary *entry;
int off;
next_step:
entry = sum;
f2fs: give a chance to merge IOs by IO scheduler Previously, background GC submits many 4KB read requests to load victim blocks and/or its (i)node blocks. ... f2fs_gc : f2fs_readpage: ino = 1, page_index = 0xb61, blkaddr = 0x3b964ed f2fs_gc : block_rq_complete: 8,16 R () 499854968 + 8 [0] f2fs_gc : f2fs_readpage: ino = 1, page_index = 0xb6f, blkaddr = 0x3b964ee f2fs_gc : block_rq_complete: 8,16 R () 499854976 + 8 [0] f2fs_gc : f2fs_readpage: ino = 1, page_index = 0xb79, blkaddr = 0x3b964ef f2fs_gc : block_rq_complete: 8,16 R () 499854984 + 8 [0] ... However, by the fact that many IOs are sequential, we can give a chance to merge the IOs by IO scheduler. In order to do that, let's use blk_plug. ... f2fs_gc : f2fs_iget: ino = 143 f2fs_gc : f2fs_readpage: ino = 143, page_index = 0x1c6, blkaddr = 0x2e6ee f2fs_gc : f2fs_iget: ino = 143 f2fs_gc : f2fs_readpage: ino = 143, page_index = 0x1c7, blkaddr = 0x2e6ef <idle> : block_rq_complete: 8,16 R () 1519616 + 8 [0] <idle> : block_rq_complete: 8,16 R () 1519848 + 8 [0] <idle> : block_rq_complete: 8,16 R () 1520432 + 96 [0] <idle> : block_rq_complete: 8,16 R () 1520536 + 104 [0] <idle> : block_rq_complete: 8,16 R () 1521008 + 112 [0] <idle> : block_rq_complete: 8,16 R () 1521440 + 152 [0] <idle> : block_rq_complete: 8,16 R () 1521688 + 144 [0] <idle> : block_rq_complete: 8,16 R () 1522128 + 192 [0] <idle> : block_rq_complete: 8,16 R () 1523256 + 328 [0] ... Note that this issue should be addressed in checkpoint, and some readahead flows too. Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-04-24 12:19:56 +08:00
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
for (off = 0; off < sbi->blocks_per_seg; off++, entry++) {
nid_t nid = le32_to_cpu(entry->nid);
struct page *node_page;
f2fs: clarify and enhance the f2fs_gc flow This patch makes clearer the ambiguous f2fs_gc flow as follows. 1. Remove intermediate checkpoint condition during f2fs_gc (i.e., should_do_checkpoint() and GC_BLOCKED) 2. Remove unnecessary return values of f2fs_gc because of #1. (i.e., GC_NODE, GC_OK, etc) 3. Simplify write_checkpoint() because of #2. 4. Clarify the main f2fs_gc flow. o monitor how many freed sections during one iteration of do_garbage_collect(). o do GC more without checkpoints if we can't get enough free sections. o do checkpoint once we've got enough free sections through forground GCs. 5. Adopt thread-logging (Slack-Space-Recycle) scheme more aggressively on data log types. See. get_ssr_segement() Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-02-04 14:11:17 +08:00
/* stop BG_GC if there is not enough free sections. */
if (gc_type == BG_GC && has_not_enough_free_secs(sbi, 0))
return;
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
f2fs: clarify and enhance the f2fs_gc flow This patch makes clearer the ambiguous f2fs_gc flow as follows. 1. Remove intermediate checkpoint condition during f2fs_gc (i.e., should_do_checkpoint() and GC_BLOCKED) 2. Remove unnecessary return values of f2fs_gc because of #1. (i.e., GC_NODE, GC_OK, etc) 3. Simplify write_checkpoint() because of #2. 4. Clarify the main f2fs_gc flow. o monitor how many freed sections during one iteration of do_garbage_collect(). o do GC more without checkpoints if we can't get enough free sections. o do checkpoint once we've got enough free sections through forground GCs. 5. Adopt thread-logging (Slack-Space-Recycle) scheme more aggressively on data log types. See. get_ssr_segement() Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-02-04 14:11:17 +08:00
if (check_valid_map(sbi, segno, off) == 0)
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
continue;
if (initial) {
ra_node_page(sbi, nid);
continue;
}
node_page = get_node_page(sbi, nid);
if (IS_ERR(node_page))
continue;
/* set page dirty and write it */
f2fs: check completion of foreground GC The foreground GCs are triggered under not enough free sections. So, we should not skip moving valid blocks in the victim segments. Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-03-31 12:49:18 +08:00
if (gc_type == FG_GC) {
f2fs_submit_bio(sbi, NODE, true);
wait_on_page_writeback(node_page);
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
set_page_dirty(node_page);
f2fs: check completion of foreground GC The foreground GCs are triggered under not enough free sections. So, we should not skip moving valid blocks in the victim segments. Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-03-31 12:49:18 +08:00
} else {
if (!PageWriteback(node_page))
set_page_dirty(node_page);
}
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
f2fs_put_page(node_page, 1);
stat_inc_node_blk_count(sbi, 1);
}
f2fs: give a chance to merge IOs by IO scheduler Previously, background GC submits many 4KB read requests to load victim blocks and/or its (i)node blocks. ... f2fs_gc : f2fs_readpage: ino = 1, page_index = 0xb61, blkaddr = 0x3b964ed f2fs_gc : block_rq_complete: 8,16 R () 499854968 + 8 [0] f2fs_gc : f2fs_readpage: ino = 1, page_index = 0xb6f, blkaddr = 0x3b964ee f2fs_gc : block_rq_complete: 8,16 R () 499854976 + 8 [0] f2fs_gc : f2fs_readpage: ino = 1, page_index = 0xb79, blkaddr = 0x3b964ef f2fs_gc : block_rq_complete: 8,16 R () 499854984 + 8 [0] ... However, by the fact that many IOs are sequential, we can give a chance to merge the IOs by IO scheduler. In order to do that, let's use blk_plug. ... f2fs_gc : f2fs_iget: ino = 143 f2fs_gc : f2fs_readpage: ino = 143, page_index = 0x1c6, blkaddr = 0x2e6ee f2fs_gc : f2fs_iget: ino = 143 f2fs_gc : f2fs_readpage: ino = 143, page_index = 0x1c7, blkaddr = 0x2e6ef <idle> : block_rq_complete: 8,16 R () 1519616 + 8 [0] <idle> : block_rq_complete: 8,16 R () 1519848 + 8 [0] <idle> : block_rq_complete: 8,16 R () 1520432 + 96 [0] <idle> : block_rq_complete: 8,16 R () 1520536 + 104 [0] <idle> : block_rq_complete: 8,16 R () 1521008 + 112 [0] <idle> : block_rq_complete: 8,16 R () 1521440 + 152 [0] <idle> : block_rq_complete: 8,16 R () 1521688 + 144 [0] <idle> : block_rq_complete: 8,16 R () 1522128 + 192 [0] <idle> : block_rq_complete: 8,16 R () 1523256 + 328 [0] ... Note that this issue should be addressed in checkpoint, and some readahead flows too. Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-04-24 12:19:56 +08:00
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
if (initial) {
initial = false;
goto next_step;
}
if (gc_type == FG_GC) {
struct writeback_control wbc = {
.sync_mode = WB_SYNC_ALL,
.nr_to_write = LONG_MAX,
.for_reclaim = 0,
};
sync_node_pages(sbi, 0, &wbc);
f2fs: check completion of foreground GC The foreground GCs are triggered under not enough free sections. So, we should not skip moving valid blocks in the victim segments. Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-03-31 12:49:18 +08:00
/*
* In the case of FG_GC, it'd be better to reclaim this victim
* completely.
*/
if (get_valid_blocks(sbi, segno, 1) != 0)
goto next_step;
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
}
}
f2fs: adjust kernel coding style As pointed out by Randy Dunlap, this patch removes all usage of "/**" for comment blocks. Instead, just use "/*". Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-29 12:28:09 +08:00
/*
f2fs: add comments of start_bidx_of_node The caller of start_bidx_of_node() should give proper node offsets which point only direct node blocks. Otherwise, it is a caller's bug. This patch adds comments to make it clear. Reported-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-01-21 16:34:21 +08:00
* Calculate start block index indicating the given node offset.
* Be careful, caller should give this node offset only indicating direct node
* blocks. If any node offsets, which point the other types of node blocks such
* as indirect or double indirect node blocks, are given, it must be a caller's
* bug.
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
*/
block_t start_bidx_of_node(unsigned int node_ofs)
{
f2fs: clean up the start_bidx_of_node function This patch also resolves the following warning reported by kbuild test robot. fs/f2fs/gc.c: In function 'start_bidx_of_node': fs/f2fs/gc.c:453:21: warning: 'bidx' may be used uninitialized in this function Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-12-26 11:03:22 +08:00
unsigned int indirect_blks = 2 * NIDS_PER_BLOCK + 4;
unsigned int bidx;
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
f2fs: clean up the start_bidx_of_node function This patch also resolves the following warning reported by kbuild test robot. fs/f2fs/gc.c: In function 'start_bidx_of_node': fs/f2fs/gc.c:453:21: warning: 'bidx' may be used uninitialized in this function Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-12-26 11:03:22 +08:00
if (node_ofs == 0)
return 0;
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
f2fs: clean up the start_bidx_of_node function This patch also resolves the following warning reported by kbuild test robot. fs/f2fs/gc.c: In function 'start_bidx_of_node': fs/f2fs/gc.c:453:21: warning: 'bidx' may be used uninitialized in this function Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-12-26 11:03:22 +08:00
if (node_ofs <= 2) {
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
bidx = node_ofs - 1;
} else if (node_ofs <= indirect_blks) {
f2fs: clean up the start_bidx_of_node function This patch also resolves the following warning reported by kbuild test robot. fs/f2fs/gc.c: In function 'start_bidx_of_node': fs/f2fs/gc.c:453:21: warning: 'bidx' may be used uninitialized in this function Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-12-26 11:03:22 +08:00
int dec = (node_ofs - 4) / (NIDS_PER_BLOCK + 1);
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
bidx = node_ofs - 2 - dec;
} else {
f2fs: clean up the start_bidx_of_node function This patch also resolves the following warning reported by kbuild test robot. fs/f2fs/gc.c: In function 'start_bidx_of_node': fs/f2fs/gc.c:453:21: warning: 'bidx' may be used uninitialized in this function Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-12-26 11:03:22 +08:00
int dec = (node_ofs - indirect_blks - 3) / (NIDS_PER_BLOCK + 1);
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
bidx = node_ofs - 5 - dec;
}
f2fs: clean up the start_bidx_of_node function This patch also resolves the following warning reported by kbuild test robot. fs/f2fs/gc.c: In function 'start_bidx_of_node': fs/f2fs/gc.c:453:21: warning: 'bidx' may be used uninitialized in this function Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-12-26 11:03:22 +08:00
return bidx * ADDRS_PER_BLOCK + ADDRS_PER_INODE;
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
}
static int check_dnode(struct f2fs_sb_info *sbi, struct f2fs_summary *sum,
struct node_info *dni, block_t blkaddr, unsigned int *nofs)
{
struct page *node_page;
nid_t nid;
unsigned int ofs_in_node;
block_t source_blkaddr;
nid = le32_to_cpu(sum->nid);
ofs_in_node = le16_to_cpu(sum->ofs_in_node);
node_page = get_node_page(sbi, nid);
if (IS_ERR(node_page))
f2fs: clarify and enhance the f2fs_gc flow This patch makes clearer the ambiguous f2fs_gc flow as follows. 1. Remove intermediate checkpoint condition during f2fs_gc (i.e., should_do_checkpoint() and GC_BLOCKED) 2. Remove unnecessary return values of f2fs_gc because of #1. (i.e., GC_NODE, GC_OK, etc) 3. Simplify write_checkpoint() because of #2. 4. Clarify the main f2fs_gc flow. o monitor how many freed sections during one iteration of do_garbage_collect(). o do GC more without checkpoints if we can't get enough free sections. o do checkpoint once we've got enough free sections through forground GCs. 5. Adopt thread-logging (Slack-Space-Recycle) scheme more aggressively on data log types. See. get_ssr_segement() Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-02-04 14:11:17 +08:00
return 0;
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
get_node_info(sbi, nid, dni);
if (sum->version != dni->version) {
f2fs_put_page(node_page, 1);
f2fs: clarify and enhance the f2fs_gc flow This patch makes clearer the ambiguous f2fs_gc flow as follows. 1. Remove intermediate checkpoint condition during f2fs_gc (i.e., should_do_checkpoint() and GC_BLOCKED) 2. Remove unnecessary return values of f2fs_gc because of #1. (i.e., GC_NODE, GC_OK, etc) 3. Simplify write_checkpoint() because of #2. 4. Clarify the main f2fs_gc flow. o monitor how many freed sections during one iteration of do_garbage_collect(). o do GC more without checkpoints if we can't get enough free sections. o do checkpoint once we've got enough free sections through forground GCs. 5. Adopt thread-logging (Slack-Space-Recycle) scheme more aggressively on data log types. See. get_ssr_segement() Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-02-04 14:11:17 +08:00
return 0;
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
}
*nofs = ofs_of_node(node_page);
source_blkaddr = datablock_addr(node_page, ofs_in_node);
f2fs_put_page(node_page, 1);
if (source_blkaddr != blkaddr)
f2fs: clarify and enhance the f2fs_gc flow This patch makes clearer the ambiguous f2fs_gc flow as follows. 1. Remove intermediate checkpoint condition during f2fs_gc (i.e., should_do_checkpoint() and GC_BLOCKED) 2. Remove unnecessary return values of f2fs_gc because of #1. (i.e., GC_NODE, GC_OK, etc) 3. Simplify write_checkpoint() because of #2. 4. Clarify the main f2fs_gc flow. o monitor how many freed sections during one iteration of do_garbage_collect(). o do GC more without checkpoints if we can't get enough free sections. o do checkpoint once we've got enough free sections through forground GCs. 5. Adopt thread-logging (Slack-Space-Recycle) scheme more aggressively on data log types. See. get_ssr_segement() Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-02-04 14:11:17 +08:00
return 0;
return 1;
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
}
static void move_data_page(struct inode *inode, struct page *page, int gc_type)
{
if (gc_type == BG_GC) {
f2fs: check completion of foreground GC The foreground GCs are triggered under not enough free sections. So, we should not skip moving valid blocks in the victim segments. Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-03-31 12:49:18 +08:00
if (PageWriteback(page))
goto out;
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
set_page_dirty(page);
set_cold_data(page);
} else {
struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
f2fs: check completion of foreground GC The foreground GCs are triggered under not enough free sections. So, we should not skip moving valid blocks in the victim segments. Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-03-31 12:49:18 +08:00
if (PageWriteback(page)) {
f2fs_submit_bio(sbi, DATA, true);
wait_on_page_writeback(page);
}
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
if (clear_page_dirty_for_io(page) &&
S_ISDIR(inode->i_mode)) {
dec_page_count(sbi, F2FS_DIRTY_DENTS);
inode_dec_dirty_dents(inode);
}
set_cold_data(page);
do_write_data_page(page);
clear_cold_data(page);
}
out:
f2fs_put_page(page, 1);
}
f2fs: adjust kernel coding style As pointed out by Randy Dunlap, this patch removes all usage of "/**" for comment blocks. Instead, just use "/*". Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-29 12:28:09 +08:00
/*
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
* This function tries to get parent node of victim data block, and identifies
* data block validity. If the block is valid, copy that with cold status and
* modify parent node.
* If the parent node is not valid or the data block address is different,
* the victim data block is ignored.
*/
f2fs: clarify and enhance the f2fs_gc flow This patch makes clearer the ambiguous f2fs_gc flow as follows. 1. Remove intermediate checkpoint condition during f2fs_gc (i.e., should_do_checkpoint() and GC_BLOCKED) 2. Remove unnecessary return values of f2fs_gc because of #1. (i.e., GC_NODE, GC_OK, etc) 3. Simplify write_checkpoint() because of #2. 4. Clarify the main f2fs_gc flow. o monitor how many freed sections during one iteration of do_garbage_collect(). o do GC more without checkpoints if we can't get enough free sections. o do checkpoint once we've got enough free sections through forground GCs. 5. Adopt thread-logging (Slack-Space-Recycle) scheme more aggressively on data log types. See. get_ssr_segement() Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-02-04 14:11:17 +08:00
static void gc_data_segment(struct f2fs_sb_info *sbi, struct f2fs_summary *sum,
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
struct list_head *ilist, unsigned int segno, int gc_type)
{
struct super_block *sb = sbi->sb;
struct f2fs_summary *entry;
block_t start_addr;
f2fs: clarify and enhance the f2fs_gc flow This patch makes clearer the ambiguous f2fs_gc flow as follows. 1. Remove intermediate checkpoint condition during f2fs_gc (i.e., should_do_checkpoint() and GC_BLOCKED) 2. Remove unnecessary return values of f2fs_gc because of #1. (i.e., GC_NODE, GC_OK, etc) 3. Simplify write_checkpoint() because of #2. 4. Clarify the main f2fs_gc flow. o monitor how many freed sections during one iteration of do_garbage_collect(). o do GC more without checkpoints if we can't get enough free sections. o do checkpoint once we've got enough free sections through forground GCs. 5. Adopt thread-logging (Slack-Space-Recycle) scheme more aggressively on data log types. See. get_ssr_segement() Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-02-04 14:11:17 +08:00
int off;
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
int phase = 0;
start_addr = START_BLOCK(sbi, segno);
next_step:
entry = sum;
f2fs: give a chance to merge IOs by IO scheduler Previously, background GC submits many 4KB read requests to load victim blocks and/or its (i)node blocks. ... f2fs_gc : f2fs_readpage: ino = 1, page_index = 0xb61, blkaddr = 0x3b964ed f2fs_gc : block_rq_complete: 8,16 R () 499854968 + 8 [0] f2fs_gc : f2fs_readpage: ino = 1, page_index = 0xb6f, blkaddr = 0x3b964ee f2fs_gc : block_rq_complete: 8,16 R () 499854976 + 8 [0] f2fs_gc : f2fs_readpage: ino = 1, page_index = 0xb79, blkaddr = 0x3b964ef f2fs_gc : block_rq_complete: 8,16 R () 499854984 + 8 [0] ... However, by the fact that many IOs are sequential, we can give a chance to merge the IOs by IO scheduler. In order to do that, let's use blk_plug. ... f2fs_gc : f2fs_iget: ino = 143 f2fs_gc : f2fs_readpage: ino = 143, page_index = 0x1c6, blkaddr = 0x2e6ee f2fs_gc : f2fs_iget: ino = 143 f2fs_gc : f2fs_readpage: ino = 143, page_index = 0x1c7, blkaddr = 0x2e6ef <idle> : block_rq_complete: 8,16 R () 1519616 + 8 [0] <idle> : block_rq_complete: 8,16 R () 1519848 + 8 [0] <idle> : block_rq_complete: 8,16 R () 1520432 + 96 [0] <idle> : block_rq_complete: 8,16 R () 1520536 + 104 [0] <idle> : block_rq_complete: 8,16 R () 1521008 + 112 [0] <idle> : block_rq_complete: 8,16 R () 1521440 + 152 [0] <idle> : block_rq_complete: 8,16 R () 1521688 + 144 [0] <idle> : block_rq_complete: 8,16 R () 1522128 + 192 [0] <idle> : block_rq_complete: 8,16 R () 1523256 + 328 [0] ... Note that this issue should be addressed in checkpoint, and some readahead flows too. Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-04-24 12:19:56 +08:00
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
for (off = 0; off < sbi->blocks_per_seg; off++, entry++) {
struct page *data_page;
struct inode *inode;
struct node_info dni; /* dnode info for the data */
unsigned int ofs_in_node, nofs;
block_t start_bidx;
f2fs: clarify and enhance the f2fs_gc flow This patch makes clearer the ambiguous f2fs_gc flow as follows. 1. Remove intermediate checkpoint condition during f2fs_gc (i.e., should_do_checkpoint() and GC_BLOCKED) 2. Remove unnecessary return values of f2fs_gc because of #1. (i.e., GC_NODE, GC_OK, etc) 3. Simplify write_checkpoint() because of #2. 4. Clarify the main f2fs_gc flow. o monitor how many freed sections during one iteration of do_garbage_collect(). o do GC more without checkpoints if we can't get enough free sections. o do checkpoint once we've got enough free sections through forground GCs. 5. Adopt thread-logging (Slack-Space-Recycle) scheme more aggressively on data log types. See. get_ssr_segement() Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-02-04 14:11:17 +08:00
/* stop BG_GC if there is not enough free sections. */
if (gc_type == BG_GC && has_not_enough_free_secs(sbi, 0))
return;
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
f2fs: clarify and enhance the f2fs_gc flow This patch makes clearer the ambiguous f2fs_gc flow as follows. 1. Remove intermediate checkpoint condition during f2fs_gc (i.e., should_do_checkpoint() and GC_BLOCKED) 2. Remove unnecessary return values of f2fs_gc because of #1. (i.e., GC_NODE, GC_OK, etc) 3. Simplify write_checkpoint() because of #2. 4. Clarify the main f2fs_gc flow. o monitor how many freed sections during one iteration of do_garbage_collect(). o do GC more without checkpoints if we can't get enough free sections. o do checkpoint once we've got enough free sections through forground GCs. 5. Adopt thread-logging (Slack-Space-Recycle) scheme more aggressively on data log types. See. get_ssr_segement() Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-02-04 14:11:17 +08:00
if (check_valid_map(sbi, segno, off) == 0)
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
continue;
if (phase == 0) {
ra_node_page(sbi, le32_to_cpu(entry->nid));
continue;
}
/* Get an inode by ino with checking validity */
f2fs: clarify and enhance the f2fs_gc flow This patch makes clearer the ambiguous f2fs_gc flow as follows. 1. Remove intermediate checkpoint condition during f2fs_gc (i.e., should_do_checkpoint() and GC_BLOCKED) 2. Remove unnecessary return values of f2fs_gc because of #1. (i.e., GC_NODE, GC_OK, etc) 3. Simplify write_checkpoint() because of #2. 4. Clarify the main f2fs_gc flow. o monitor how many freed sections during one iteration of do_garbage_collect(). o do GC more without checkpoints if we can't get enough free sections. o do checkpoint once we've got enough free sections through forground GCs. 5. Adopt thread-logging (Slack-Space-Recycle) scheme more aggressively on data log types. See. get_ssr_segement() Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-02-04 14:11:17 +08:00
if (check_dnode(sbi, entry, &dni, start_addr + off, &nofs) == 0)
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
continue;
if (phase == 1) {
ra_node_page(sbi, dni.ino);
continue;
}
start_bidx = start_bidx_of_node(nofs);
ofs_in_node = le16_to_cpu(entry->ofs_in_node);
if (phase == 2) {
f2fs: avoid balanc_fs during evict_inode 1. Background Previously, if f2fs tries to move data blocks of an *evicting* inode during the cleaning process, it stops the process incompletely and then restarts the whole process, since it needs a locked inode to grab victim data pages in its address space. In order to get a locked inode, iget_locked() by f2fs_iget() is normally used, but, it waits if the inode is on freeing. So, here is a deadlock scenario. 1. f2fs_evict_inode() <- inode "A" 2. f2fs_balance_fs() 3. f2fs_gc() 4. gc_data_segment() 5. f2fs_iget() <- inode "A" too! If step #1 and #5 treat a same inode "A", step #5 would fall into deadlock since the inode "A" is on freeing. In order to resolve this, f2fs_iget_nowait() which skips __wait_on_freeing_inode() was introduced in step #5, and stops f2fs_gc() to complete f2fs_evict_inode(). 1. f2fs_evict_inode() <- inode "A" 2. f2fs_balance_fs() 3. f2fs_gc() 4. gc_data_segment() 5. f2fs_iget_nowait() <- inode "A", then stop f2fs_gc() w/ -ENOENT 2. Problem and Solution In the above scenario, however, f2fs cannot finish f2fs_evict_inode() only if: o there are not enough free sections, and o f2fs_gc() tries to move data blocks of the *evicting* inode repeatedly. So, the final solution is to use f2fs_iget() and remove f2fs_balance_fs() in f2fs_evict_inode(). The f2fs_evict_inode() actually truncates all the data and node blocks, which means that it doesn't produce any dirty node pages accordingly. So, we don't need to do f2fs_balance_fs() in practical. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-01-31 14:36:04 +08:00
inode = f2fs_iget(sb, dni.ino);
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
if (IS_ERR(inode))
continue;
data_page = find_data_page(inode,
f2fs: give a chance to merge IOs by IO scheduler Previously, background GC submits many 4KB read requests to load victim blocks and/or its (i)node blocks. ... f2fs_gc : f2fs_readpage: ino = 1, page_index = 0xb61, blkaddr = 0x3b964ed f2fs_gc : block_rq_complete: 8,16 R () 499854968 + 8 [0] f2fs_gc : f2fs_readpage: ino = 1, page_index = 0xb6f, blkaddr = 0x3b964ee f2fs_gc : block_rq_complete: 8,16 R () 499854976 + 8 [0] f2fs_gc : f2fs_readpage: ino = 1, page_index = 0xb79, blkaddr = 0x3b964ef f2fs_gc : block_rq_complete: 8,16 R () 499854984 + 8 [0] ... However, by the fact that many IOs are sequential, we can give a chance to merge the IOs by IO scheduler. In order to do that, let's use blk_plug. ... f2fs_gc : f2fs_iget: ino = 143 f2fs_gc : f2fs_readpage: ino = 143, page_index = 0x1c6, blkaddr = 0x2e6ee f2fs_gc : f2fs_iget: ino = 143 f2fs_gc : f2fs_readpage: ino = 143, page_index = 0x1c7, blkaddr = 0x2e6ef <idle> : block_rq_complete: 8,16 R () 1519616 + 8 [0] <idle> : block_rq_complete: 8,16 R () 1519848 + 8 [0] <idle> : block_rq_complete: 8,16 R () 1520432 + 96 [0] <idle> : block_rq_complete: 8,16 R () 1520536 + 104 [0] <idle> : block_rq_complete: 8,16 R () 1521008 + 112 [0] <idle> : block_rq_complete: 8,16 R () 1521440 + 152 [0] <idle> : block_rq_complete: 8,16 R () 1521688 + 144 [0] <idle> : block_rq_complete: 8,16 R () 1522128 + 192 [0] <idle> : block_rq_complete: 8,16 R () 1523256 + 328 [0] ... Note that this issue should be addressed in checkpoint, and some readahead flows too. Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-04-24 12:19:56 +08:00
start_bidx + ofs_in_node, false);
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
if (IS_ERR(data_page))
goto next_iput;
f2fs_put_page(data_page, 0);
add_gc_inode(inode, ilist);
} else {
inode = find_gc_inode(dni.ino, ilist);
if (inode) {
data_page = get_lock_data_page(inode,
start_bidx + ofs_in_node);
if (IS_ERR(data_page))
continue;
move_data_page(inode, data_page, gc_type);
stat_inc_data_blk_count(sbi, 1);
}
}
continue;
next_iput:
iput(inode);
}
f2fs: give a chance to merge IOs by IO scheduler Previously, background GC submits many 4KB read requests to load victim blocks and/or its (i)node blocks. ... f2fs_gc : f2fs_readpage: ino = 1, page_index = 0xb61, blkaddr = 0x3b964ed f2fs_gc : block_rq_complete: 8,16 R () 499854968 + 8 [0] f2fs_gc : f2fs_readpage: ino = 1, page_index = 0xb6f, blkaddr = 0x3b964ee f2fs_gc : block_rq_complete: 8,16 R () 499854976 + 8 [0] f2fs_gc : f2fs_readpage: ino = 1, page_index = 0xb79, blkaddr = 0x3b964ef f2fs_gc : block_rq_complete: 8,16 R () 499854984 + 8 [0] ... However, by the fact that many IOs are sequential, we can give a chance to merge the IOs by IO scheduler. In order to do that, let's use blk_plug. ... f2fs_gc : f2fs_iget: ino = 143 f2fs_gc : f2fs_readpage: ino = 143, page_index = 0x1c6, blkaddr = 0x2e6ee f2fs_gc : f2fs_iget: ino = 143 f2fs_gc : f2fs_readpage: ino = 143, page_index = 0x1c7, blkaddr = 0x2e6ef <idle> : block_rq_complete: 8,16 R () 1519616 + 8 [0] <idle> : block_rq_complete: 8,16 R () 1519848 + 8 [0] <idle> : block_rq_complete: 8,16 R () 1520432 + 96 [0] <idle> : block_rq_complete: 8,16 R () 1520536 + 104 [0] <idle> : block_rq_complete: 8,16 R () 1521008 + 112 [0] <idle> : block_rq_complete: 8,16 R () 1521440 + 152 [0] <idle> : block_rq_complete: 8,16 R () 1521688 + 144 [0] <idle> : block_rq_complete: 8,16 R () 1522128 + 192 [0] <idle> : block_rq_complete: 8,16 R () 1523256 + 328 [0] ... Note that this issue should be addressed in checkpoint, and some readahead flows too. Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-04-24 12:19:56 +08:00
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
if (++phase < 4)
goto next_step;
f2fs: clarify and enhance the f2fs_gc flow This patch makes clearer the ambiguous f2fs_gc flow as follows. 1. Remove intermediate checkpoint condition during f2fs_gc (i.e., should_do_checkpoint() and GC_BLOCKED) 2. Remove unnecessary return values of f2fs_gc because of #1. (i.e., GC_NODE, GC_OK, etc) 3. Simplify write_checkpoint() because of #2. 4. Clarify the main f2fs_gc flow. o monitor how many freed sections during one iteration of do_garbage_collect(). o do GC more without checkpoints if we can't get enough free sections. o do checkpoint once we've got enough free sections through forground GCs. 5. Adopt thread-logging (Slack-Space-Recycle) scheme more aggressively on data log types. See. get_ssr_segement() Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-02-04 14:11:17 +08:00
f2fs: check completion of foreground GC The foreground GCs are triggered under not enough free sections. So, we should not skip moving valid blocks in the victim segments. Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-03-31 12:49:18 +08:00
if (gc_type == FG_GC) {
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
f2fs_submit_bio(sbi, DATA, true);
f2fs: check completion of foreground GC The foreground GCs are triggered under not enough free sections. So, we should not skip moving valid blocks in the victim segments. Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-03-31 12:49:18 +08:00
/*
* In the case of FG_GC, it'd be better to reclaim this victim
* completely.
*/
if (get_valid_blocks(sbi, segno, 1) != 0) {
phase = 2;
goto next_step;
}
}
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
}
static int __get_victim(struct f2fs_sb_info *sbi, unsigned int *victim,
int gc_type, int type)
{
struct sit_info *sit_i = SIT_I(sbi);
int ret;
mutex_lock(&sit_i->sentry_lock);
ret = DIRTY_I(sbi)->v_ops->get_victim(sbi, victim, gc_type, type, LFS);
mutex_unlock(&sit_i->sentry_lock);
return ret;
}
f2fs: clarify and enhance the f2fs_gc flow This patch makes clearer the ambiguous f2fs_gc flow as follows. 1. Remove intermediate checkpoint condition during f2fs_gc (i.e., should_do_checkpoint() and GC_BLOCKED) 2. Remove unnecessary return values of f2fs_gc because of #1. (i.e., GC_NODE, GC_OK, etc) 3. Simplify write_checkpoint() because of #2. 4. Clarify the main f2fs_gc flow. o monitor how many freed sections during one iteration of do_garbage_collect(). o do GC more without checkpoints if we can't get enough free sections. o do checkpoint once we've got enough free sections through forground GCs. 5. Adopt thread-logging (Slack-Space-Recycle) scheme more aggressively on data log types. See. get_ssr_segement() Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-02-04 14:11:17 +08:00
static void do_garbage_collect(struct f2fs_sb_info *sbi, unsigned int segno,
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
struct list_head *ilist, int gc_type)
{
struct page *sum_page;
struct f2fs_summary_block *sum;
f2fs: give a chance to merge IOs by IO scheduler Previously, background GC submits many 4KB read requests to load victim blocks and/or its (i)node blocks. ... f2fs_gc : f2fs_readpage: ino = 1, page_index = 0xb61, blkaddr = 0x3b964ed f2fs_gc : block_rq_complete: 8,16 R () 499854968 + 8 [0] f2fs_gc : f2fs_readpage: ino = 1, page_index = 0xb6f, blkaddr = 0x3b964ee f2fs_gc : block_rq_complete: 8,16 R () 499854976 + 8 [0] f2fs_gc : f2fs_readpage: ino = 1, page_index = 0xb79, blkaddr = 0x3b964ef f2fs_gc : block_rq_complete: 8,16 R () 499854984 + 8 [0] ... However, by the fact that many IOs are sequential, we can give a chance to merge the IOs by IO scheduler. In order to do that, let's use blk_plug. ... f2fs_gc : f2fs_iget: ino = 143 f2fs_gc : f2fs_readpage: ino = 143, page_index = 0x1c6, blkaddr = 0x2e6ee f2fs_gc : f2fs_iget: ino = 143 f2fs_gc : f2fs_readpage: ino = 143, page_index = 0x1c7, blkaddr = 0x2e6ef <idle> : block_rq_complete: 8,16 R () 1519616 + 8 [0] <idle> : block_rq_complete: 8,16 R () 1519848 + 8 [0] <idle> : block_rq_complete: 8,16 R () 1520432 + 96 [0] <idle> : block_rq_complete: 8,16 R () 1520536 + 104 [0] <idle> : block_rq_complete: 8,16 R () 1521008 + 112 [0] <idle> : block_rq_complete: 8,16 R () 1521440 + 152 [0] <idle> : block_rq_complete: 8,16 R () 1521688 + 144 [0] <idle> : block_rq_complete: 8,16 R () 1522128 + 192 [0] <idle> : block_rq_complete: 8,16 R () 1523256 + 328 [0] ... Note that this issue should be addressed in checkpoint, and some readahead flows too. Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-04-24 12:19:56 +08:00
struct blk_plug plug;
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
/* read segment summary of victim */
sum_page = get_sum_page(sbi, segno);
if (IS_ERR(sum_page))
f2fs: clarify and enhance the f2fs_gc flow This patch makes clearer the ambiguous f2fs_gc flow as follows. 1. Remove intermediate checkpoint condition during f2fs_gc (i.e., should_do_checkpoint() and GC_BLOCKED) 2. Remove unnecessary return values of f2fs_gc because of #1. (i.e., GC_NODE, GC_OK, etc) 3. Simplify write_checkpoint() because of #2. 4. Clarify the main f2fs_gc flow. o monitor how many freed sections during one iteration of do_garbage_collect(). o do GC more without checkpoints if we can't get enough free sections. o do checkpoint once we've got enough free sections through forground GCs. 5. Adopt thread-logging (Slack-Space-Recycle) scheme more aggressively on data log types. See. get_ssr_segement() Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-02-04 14:11:17 +08:00
return;
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
f2fs: give a chance to merge IOs by IO scheduler Previously, background GC submits many 4KB read requests to load victim blocks and/or its (i)node blocks. ... f2fs_gc : f2fs_readpage: ino = 1, page_index = 0xb61, blkaddr = 0x3b964ed f2fs_gc : block_rq_complete: 8,16 R () 499854968 + 8 [0] f2fs_gc : f2fs_readpage: ino = 1, page_index = 0xb6f, blkaddr = 0x3b964ee f2fs_gc : block_rq_complete: 8,16 R () 499854976 + 8 [0] f2fs_gc : f2fs_readpage: ino = 1, page_index = 0xb79, blkaddr = 0x3b964ef f2fs_gc : block_rq_complete: 8,16 R () 499854984 + 8 [0] ... However, by the fact that many IOs are sequential, we can give a chance to merge the IOs by IO scheduler. In order to do that, let's use blk_plug. ... f2fs_gc : f2fs_iget: ino = 143 f2fs_gc : f2fs_readpage: ino = 143, page_index = 0x1c6, blkaddr = 0x2e6ee f2fs_gc : f2fs_iget: ino = 143 f2fs_gc : f2fs_readpage: ino = 143, page_index = 0x1c7, blkaddr = 0x2e6ef <idle> : block_rq_complete: 8,16 R () 1519616 + 8 [0] <idle> : block_rq_complete: 8,16 R () 1519848 + 8 [0] <idle> : block_rq_complete: 8,16 R () 1520432 + 96 [0] <idle> : block_rq_complete: 8,16 R () 1520536 + 104 [0] <idle> : block_rq_complete: 8,16 R () 1521008 + 112 [0] <idle> : block_rq_complete: 8,16 R () 1521440 + 152 [0] <idle> : block_rq_complete: 8,16 R () 1521688 + 144 [0] <idle> : block_rq_complete: 8,16 R () 1522128 + 192 [0] <idle> : block_rq_complete: 8,16 R () 1523256 + 328 [0] ... Note that this issue should be addressed in checkpoint, and some readahead flows too. Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-04-24 12:19:56 +08:00
blk_start_plug(&plug);
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
sum = page_address(sum_page);
switch (GET_SUM_TYPE((&sum->footer))) {
case SUM_TYPE_NODE:
f2fs: clarify and enhance the f2fs_gc flow This patch makes clearer the ambiguous f2fs_gc flow as follows. 1. Remove intermediate checkpoint condition during f2fs_gc (i.e., should_do_checkpoint() and GC_BLOCKED) 2. Remove unnecessary return values of f2fs_gc because of #1. (i.e., GC_NODE, GC_OK, etc) 3. Simplify write_checkpoint() because of #2. 4. Clarify the main f2fs_gc flow. o monitor how many freed sections during one iteration of do_garbage_collect(). o do GC more without checkpoints if we can't get enough free sections. o do checkpoint once we've got enough free sections through forground GCs. 5. Adopt thread-logging (Slack-Space-Recycle) scheme more aggressively on data log types. See. get_ssr_segement() Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-02-04 14:11:17 +08:00
gc_node_segment(sbi, sum->entries, segno, gc_type);
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
break;
case SUM_TYPE_DATA:
f2fs: clarify and enhance the f2fs_gc flow This patch makes clearer the ambiguous f2fs_gc flow as follows. 1. Remove intermediate checkpoint condition during f2fs_gc (i.e., should_do_checkpoint() and GC_BLOCKED) 2. Remove unnecessary return values of f2fs_gc because of #1. (i.e., GC_NODE, GC_OK, etc) 3. Simplify write_checkpoint() because of #2. 4. Clarify the main f2fs_gc flow. o monitor how many freed sections during one iteration of do_garbage_collect(). o do GC more without checkpoints if we can't get enough free sections. o do checkpoint once we've got enough free sections through forground GCs. 5. Adopt thread-logging (Slack-Space-Recycle) scheme more aggressively on data log types. See. get_ssr_segement() Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-02-04 14:11:17 +08:00
gc_data_segment(sbi, sum->entries, ilist, segno, gc_type);
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
break;
}
f2fs: give a chance to merge IOs by IO scheduler Previously, background GC submits many 4KB read requests to load victim blocks and/or its (i)node blocks. ... f2fs_gc : f2fs_readpage: ino = 1, page_index = 0xb61, blkaddr = 0x3b964ed f2fs_gc : block_rq_complete: 8,16 R () 499854968 + 8 [0] f2fs_gc : f2fs_readpage: ino = 1, page_index = 0xb6f, blkaddr = 0x3b964ee f2fs_gc : block_rq_complete: 8,16 R () 499854976 + 8 [0] f2fs_gc : f2fs_readpage: ino = 1, page_index = 0xb79, blkaddr = 0x3b964ef f2fs_gc : block_rq_complete: 8,16 R () 499854984 + 8 [0] ... However, by the fact that many IOs are sequential, we can give a chance to merge the IOs by IO scheduler. In order to do that, let's use blk_plug. ... f2fs_gc : f2fs_iget: ino = 143 f2fs_gc : f2fs_readpage: ino = 143, page_index = 0x1c6, blkaddr = 0x2e6ee f2fs_gc : f2fs_iget: ino = 143 f2fs_gc : f2fs_readpage: ino = 143, page_index = 0x1c7, blkaddr = 0x2e6ef <idle> : block_rq_complete: 8,16 R () 1519616 + 8 [0] <idle> : block_rq_complete: 8,16 R () 1519848 + 8 [0] <idle> : block_rq_complete: 8,16 R () 1520432 + 96 [0] <idle> : block_rq_complete: 8,16 R () 1520536 + 104 [0] <idle> : block_rq_complete: 8,16 R () 1521008 + 112 [0] <idle> : block_rq_complete: 8,16 R () 1521440 + 152 [0] <idle> : block_rq_complete: 8,16 R () 1521688 + 144 [0] <idle> : block_rq_complete: 8,16 R () 1522128 + 192 [0] <idle> : block_rq_complete: 8,16 R () 1523256 + 328 [0] ... Note that this issue should be addressed in checkpoint, and some readahead flows too. Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-04-24 12:19:56 +08:00
blk_finish_plug(&plug);
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
stat_inc_seg_count(sbi, GET_SUM_TYPE((&sum->footer)));
stat_inc_call_count(sbi->stat_info);
f2fs: avoid race for summary information In order to do GC more reliably, I'd like to lock the vicitm summary page until its GC is completed, and also prevent any checkpoint process. Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-04-01 07:32:21 +08:00
f2fs_put_page(sum_page, 1);
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
}
f2fs: revisit the f2fs_gc flow I'd like to revisit the f2fs_gc flow and rewrite as follows. 1. In practical, the nGC parameter of f2fs_gc is meaningless. So, let's remove it. 2. Background GC marks victim blocks as dirty one at a time. 3. Foreground GC should do cleaning job until acquiring enough free sections. Afterwards, it needs to do checkpoint. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-01-03 16:55:52 +08:00
int f2fs_gc(struct f2fs_sb_info *sbi)
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
{
struct list_head ilist;
f2fs: revisit the f2fs_gc flow I'd like to revisit the f2fs_gc flow and rewrite as follows. 1. In practical, the nGC parameter of f2fs_gc is meaningless. So, let's remove it. 2. Background GC marks victim blocks as dirty one at a time. 3. Foreground GC should do cleaning job until acquiring enough free sections. Afterwards, it needs to do checkpoint. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-01-03 16:55:52 +08:00
unsigned int segno, i;
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
int gc_type = BG_GC;
f2fs: clarify and enhance the f2fs_gc flow This patch makes clearer the ambiguous f2fs_gc flow as follows. 1. Remove intermediate checkpoint condition during f2fs_gc (i.e., should_do_checkpoint() and GC_BLOCKED) 2. Remove unnecessary return values of f2fs_gc because of #1. (i.e., GC_NODE, GC_OK, etc) 3. Simplify write_checkpoint() because of #2. 4. Clarify the main f2fs_gc flow. o monitor how many freed sections during one iteration of do_garbage_collect(). o do GC more without checkpoints if we can't get enough free sections. o do checkpoint once we've got enough free sections through forground GCs. 5. Adopt thread-logging (Slack-Space-Recycle) scheme more aggressively on data log types. See. get_ssr_segement() Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-02-04 14:11:17 +08:00
int nfree = 0;
int ret = -1;
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
INIT_LIST_HEAD(&ilist);
gc_more:
f2fs: revisit the f2fs_gc flow I'd like to revisit the f2fs_gc flow and rewrite as follows. 1. In practical, the nGC parameter of f2fs_gc is meaningless. So, let's remove it. 2. Background GC marks victim blocks as dirty one at a time. 3. Foreground GC should do cleaning job until acquiring enough free sections. Afterwards, it needs to do checkpoint. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-01-03 16:55:52 +08:00
if (!(sbi->sb->s_flags & MS_ACTIVE))
goto stop;
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
f2fs: write checkpoint before starting FG_GC In order to be aware of prefree and free sections during FG_GC, let's start with write_checkpoint(). Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-04-08 15:01:00 +08:00
if (gc_type == BG_GC && has_not_enough_free_secs(sbi, nfree)) {
f2fs: revisit the f2fs_gc flow I'd like to revisit the f2fs_gc flow and rewrite as follows. 1. In practical, the nGC parameter of f2fs_gc is meaningless. So, let's remove it. 2. Background GC marks victim blocks as dirty one at a time. 3. Foreground GC should do cleaning job until acquiring enough free sections. Afterwards, it needs to do checkpoint. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-01-03 16:55:52 +08:00
gc_type = FG_GC;
f2fs: write checkpoint before starting FG_GC In order to be aware of prefree and free sections during FG_GC, let's start with write_checkpoint(). Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-04-08 15:01:00 +08:00
write_checkpoint(sbi, false);
}
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
f2fs: revisit the f2fs_gc flow I'd like to revisit the f2fs_gc flow and rewrite as follows. 1. In practical, the nGC parameter of f2fs_gc is meaningless. So, let's remove it. 2. Background GC marks victim blocks as dirty one at a time. 3. Foreground GC should do cleaning job until acquiring enough free sections. Afterwards, it needs to do checkpoint. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-01-03 16:55:52 +08:00
if (!__get_victim(sbi, &segno, gc_type, NO_CHECK_TYPE))
goto stop;
f2fs: clarify and enhance the f2fs_gc flow This patch makes clearer the ambiguous f2fs_gc flow as follows. 1. Remove intermediate checkpoint condition during f2fs_gc (i.e., should_do_checkpoint() and GC_BLOCKED) 2. Remove unnecessary return values of f2fs_gc because of #1. (i.e., GC_NODE, GC_OK, etc) 3. Simplify write_checkpoint() because of #2. 4. Clarify the main f2fs_gc flow. o monitor how many freed sections during one iteration of do_garbage_collect(). o do GC more without checkpoints if we can't get enough free sections. o do checkpoint once we've got enough free sections through forground GCs. 5. Adopt thread-logging (Slack-Space-Recycle) scheme more aggressively on data log types. See. get_ssr_segement() Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-02-04 14:11:17 +08:00
ret = 0;
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
f2fs: clarify and enhance the f2fs_gc flow This patch makes clearer the ambiguous f2fs_gc flow as follows. 1. Remove intermediate checkpoint condition during f2fs_gc (i.e., should_do_checkpoint() and GC_BLOCKED) 2. Remove unnecessary return values of f2fs_gc because of #1. (i.e., GC_NODE, GC_OK, etc) 3. Simplify write_checkpoint() because of #2. 4. Clarify the main f2fs_gc flow. o monitor how many freed sections during one iteration of do_garbage_collect(). o do GC more without checkpoints if we can't get enough free sections. o do checkpoint once we've got enough free sections through forground GCs. 5. Adopt thread-logging (Slack-Space-Recycle) scheme more aggressively on data log types. See. get_ssr_segement() Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-02-04 14:11:17 +08:00
for (i = 0; i < sbi->segs_per_sec; i++)
do_garbage_collect(sbi, segno + i, &ilist, gc_type);
f2fs: change GC bitmaps to apply the section granularity This patch removes a bitmap for victim segments selected by foreground GC, and modifies the other bitmap for victim segments selected by background GC. 1) foreground GC bitmap : We don't need to manage this, since we just only one previous victim section number instead of the whole victim history. The f2fs uses the victim section number in order not to allocate currently GC'ed section to current active logs. 2) background GC bitmap : This bitmap is used to avoid selecting victims repeatedly by background GCs. In addition, the victims are able to be selected by foreground GCs, since there is no need to read victim blocks during foreground GCs. By the fact that the foreground GC reclaims segments in a section unit, it'd be better to manage this bitmap based on the section granularity. Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-03-31 12:26:03 +08:00
if (gc_type == FG_GC) {
sbi->cur_victim_sec = NULL_SEGNO;
f2fs: clarify and enhance the f2fs_gc flow This patch makes clearer the ambiguous f2fs_gc flow as follows. 1. Remove intermediate checkpoint condition during f2fs_gc (i.e., should_do_checkpoint() and GC_BLOCKED) 2. Remove unnecessary return values of f2fs_gc because of #1. (i.e., GC_NODE, GC_OK, etc) 3. Simplify write_checkpoint() because of #2. 4. Clarify the main f2fs_gc flow. o monitor how many freed sections during one iteration of do_garbage_collect(). o do GC more without checkpoints if we can't get enough free sections. o do checkpoint once we've got enough free sections through forground GCs. 5. Adopt thread-logging (Slack-Space-Recycle) scheme more aggressively on data log types. See. get_ssr_segement() Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-02-04 14:11:17 +08:00
nfree++;
f2fs: change GC bitmaps to apply the section granularity This patch removes a bitmap for victim segments selected by foreground GC, and modifies the other bitmap for victim segments selected by background GC. 1) foreground GC bitmap : We don't need to manage this, since we just only one previous victim section number instead of the whole victim history. The f2fs uses the victim section number in order not to allocate currently GC'ed section to current active logs. 2) background GC bitmap : This bitmap is used to avoid selecting victims repeatedly by background GCs. In addition, the victims are able to be selected by foreground GCs, since there is no need to read victim blocks during foreground GCs. By the fact that the foreground GC reclaims segments in a section unit, it'd be better to manage this bitmap based on the section granularity. Reviewed-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-03-31 12:26:03 +08:00
WARN_ON(get_valid_blocks(sbi, segno, sbi->segs_per_sec));
}
f2fs: clarify and enhance the f2fs_gc flow This patch makes clearer the ambiguous f2fs_gc flow as follows. 1. Remove intermediate checkpoint condition during f2fs_gc (i.e., should_do_checkpoint() and GC_BLOCKED) 2. Remove unnecessary return values of f2fs_gc because of #1. (i.e., GC_NODE, GC_OK, etc) 3. Simplify write_checkpoint() because of #2. 4. Clarify the main f2fs_gc flow. o monitor how many freed sections during one iteration of do_garbage_collect(). o do GC more without checkpoints if we can't get enough free sections. o do checkpoint once we've got enough free sections through forground GCs. 5. Adopt thread-logging (Slack-Space-Recycle) scheme more aggressively on data log types. See. get_ssr_segement() Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-02-04 14:11:17 +08:00
if (has_not_enough_free_secs(sbi, nfree))
goto gc_more;
if (gc_type == FG_GC)
write_checkpoint(sbi, false);
f2fs: revisit the f2fs_gc flow I'd like to revisit the f2fs_gc flow and rewrite as follows. 1. In practical, the nGC parameter of f2fs_gc is meaningless. So, let's remove it. 2. Background GC marks victim blocks as dirty one at a time. 3. Foreground GC should do cleaning job until acquiring enough free sections. Afterwards, it needs to do checkpoint. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-01-03 16:55:52 +08:00
stop:
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
mutex_unlock(&sbi->gc_mutex);
put_gc_inode(&ilist);
f2fs: clarify and enhance the f2fs_gc flow This patch makes clearer the ambiguous f2fs_gc flow as follows. 1. Remove intermediate checkpoint condition during f2fs_gc (i.e., should_do_checkpoint() and GC_BLOCKED) 2. Remove unnecessary return values of f2fs_gc because of #1. (i.e., GC_NODE, GC_OK, etc) 3. Simplify write_checkpoint() because of #2. 4. Clarify the main f2fs_gc flow. o monitor how many freed sections during one iteration of do_garbage_collect(). o do GC more without checkpoints if we can't get enough free sections. o do checkpoint once we've got enough free sections through forground GCs. 5. Adopt thread-logging (Slack-Space-Recycle) scheme more aggressively on data log types. See. get_ssr_segement() Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-02-04 14:11:17 +08:00
return ret;
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
}
void build_gc_manager(struct f2fs_sb_info *sbi)
{
DIRTY_I(sbi)->v_ops = &default_v_ops;
}
f2fs: add __init to functions in init_f2fs_fs Add __init to functions in init_f2fs_fs for code consistency. Signed-off-by: Namjae Jeon <namjae.jeon@samsung.com> Signed-off-by: Amit Sahrawat <a.sahrawat@samsung.com> Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2013-01-16 23:08:30 +08:00
int __init create_gc_caches(void)
f2fs: add garbage collection functions This adds on-demand and background cleaning functions. - The basic background cleaning policy is trying to do cleaning jobs as much as possible whenever the system is idle. Once the background cleaning is done, the cleaner sleeps an amount of time not to interfere with VFS calls. The time is dynamically adjusted according to the status of whole segments, which is decreased when the following conditions are satisfied. . GC is not conducted currently, and . IO subsystem is idle by checking the number of requets in bdev's request list, and . There are enough dirty segments. Otherwise, the time is increased incrementally until to the maximum time. Note that, min and max times are 10 secs and 30 secs by default. - F2FS adopts a default victim selection policy where background cleaning uses a cost-benefit algorithm, while on-demand cleaning uses a greedy algorithm. - The method of moving data during the cleaning is slightly different between background and on-demand cleaning schemes. In the case of background cleaning, F2FS loads the data, and marks them as dirty. Then, F2FS expects that the data will be moved by flusher or VM. In the case of on-demand cleaning, F2FS should move the data right away. - In order to identify valid blocks in a victim segment, F2FS scans the bitmap of the segment managed as an SIT entry. Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>
2012-11-02 16:13:01 +08:00
{
winode_slab = f2fs_kmem_cache_create("f2fs_gc_inodes",
sizeof(struct inode_entry), NULL);
if (!winode_slab)
return -ENOMEM;
return 0;
}
void destroy_gc_caches(void)
{
kmem_cache_destroy(winode_slab);
}
Copy Permalink