Commit 1d69a3f8 authored by Minchan Kim's avatar Minchan Kim Committed by Linus Torvalds

zram: idle writeback fixes and cleanup

This patch includes some fixes and cleanup for idle-page writeback.

1. writeback_limit interface

Now writeback_limit interface is rather conusing.  For example, once
writeback limit budget is exausted, admin can see 0 from
/sys/block/zramX/writeback_limit which is same semantic with disable
writeback_limit at this moment.  IOW, admin cannot tell that zero came
from disable writeback limit or exausted writeback limit.

To make the interface clear, let's sepatate enable of writeback limit to
another knob - /sys/block/zram0/writeback_limit_enable

* before:
  while true :
    # to re-enable writeback limit once previous one is used up
    echo 0 > /sys/block/zram0/writeback_limit
    echo $((200<<20)) > /sys/block/zram0/writeback_limit
    ..
    .. # used up the writeback limit budget

* new
  # To enable writeback limit, from the beginning, admin should
  # enable it.
  echo $((200<<20)) > /sys/block/zram0/writeback_limit
  echo 1 > /sys/block/zram/0/writeback_limit_enable
  while true :
    echo $((200<<20)) > /sys/block/zram0/writeback_limit
    ..
    .. # used up the writeback limit budget

It's much strightforward.

2. fix condition check idle/huge writeback mode check

The mode in writeback_store is not bit opeartion any more so no need to
use bit operations.  Furthermore, current condition check is broken in
that it does writeback every pages regardless of huge/idle.

3. clean up idle_store

No need to use goto.

[minchan@kernel.org: missed spin_lock_init]
  Link: http://lkml.kernel.org/r/20190103001601.GA255139@google.com
Link: http://lkml.kernel.org/r/20181224033529.19450-1-minchan@kernel.orgSigned-off-by: default avatarMinchan Kim <minchan@kernel.org>
Suggested-by: default avatarJohn Dias <joaodias@google.com>
Cc: Sergey Senozhatsky <sergey.senozhatsky.work@gmail.com>
Cc: John Dias <joaodias@google.com>
Cc: Srinivas Paladugu <srnvs@google.com>
Signed-off-by: default avatarAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: default avatarLinus Torvalds <torvalds@linux-foundation.org>
parent 3bd6e94b
...@@ -122,11 +122,18 @@ Description: ...@@ -122,11 +122,18 @@ Description:
statistics (bd_count, bd_reads, bd_writes) in a format statistics (bd_count, bd_reads, bd_writes) in a format
similar to block layer statistics file format. similar to block layer statistics file format.
What: /sys/block/zram<id>/writeback_limit_enable
Date: November 2018
Contact: Minchan Kim <minchan@kernel.org>
Description:
The writeback_limit_enable file is read-write and specifies
eanbe of writeback_limit feature. "1" means eable the feature.
No limit "0" is the initial state.
What: /sys/block/zram<id>/writeback_limit What: /sys/block/zram<id>/writeback_limit
Date: November 2018 Date: November 2018
Contact: Minchan Kim <minchan@kernel.org> Contact: Minchan Kim <minchan@kernel.org>
Description: Description:
The writeback_limit file is read-write and specifies the maximum The writeback_limit file is read-write and specifies the maximum
amount of writeback ZRAM can do. The limit could be changed amount of writeback ZRAM can do. The limit could be changed
in run time and "0" means disable the limit. in run time.
No limit is the initial state.
...@@ -156,22 +156,23 @@ Per-device statistics are exported as various nodes under /sys/block/zram<id>/ ...@@ -156,22 +156,23 @@ Per-device statistics are exported as various nodes under /sys/block/zram<id>/
A brief description of exported device attributes. For more details please A brief description of exported device attributes. For more details please
read Documentation/ABI/testing/sysfs-block-zram. read Documentation/ABI/testing/sysfs-block-zram.
Name access description Name access description
---- ------ ----------- ---- ------ -----------
disksize RW show and set the device's disk size disksize RW show and set the device's disk size
initstate RO shows the initialization state of the device initstate RO shows the initialization state of the device
reset WO trigger device reset reset WO trigger device reset
mem_used_max WO reset the `mem_used_max' counter (see later) mem_used_max WO reset the `mem_used_max' counter (see later)
mem_limit WO specifies the maximum amount of memory ZRAM can use mem_limit WO specifies the maximum amount of memory ZRAM can use
to store the compressed data to store the compressed data
writeback_limit WO specifies the maximum amount of write IO zram can writeback_limit WO specifies the maximum amount of write IO zram can
write out to backing device as 4KB unit write out to backing device as 4KB unit
max_comp_streams RW the number of possible concurrent compress operations writeback_limit_enable RW show and set writeback_limit feature
comp_algorithm RW show and change the compression algorithm max_comp_streams RW the number of possible concurrent compress operations
compact WO trigger memory compaction comp_algorithm RW show and change the compression algorithm
debug_stat RO this file is used for zram debugging purposes compact WO trigger memory compaction
backing_dev RW set up backend storage for zram to write out debug_stat RO this file is used for zram debugging purposes
idle WO mark allocated slot as idle backing_dev RW set up backend storage for zram to write out
idle WO mark allocated slot as idle
User space is advised to use the following files to read the device statistics. User space is advised to use the following files to read the device statistics.
...@@ -280,32 +281,51 @@ With the command, zram writeback idle pages from memory to the storage. ...@@ -280,32 +281,51 @@ With the command, zram writeback idle pages from memory to the storage.
If there are lots of write IO with flash device, potentially, it has If there are lots of write IO with flash device, potentially, it has
flash wearout problem so that admin needs to design write limitation flash wearout problem so that admin needs to design write limitation
to guarantee storage health for entire product life. to guarantee storage health for entire product life.
To overcome the concern, zram supports "writeback_limit".
The "writeback_limit"'s default value is 0 so that it doesn't limit To overcome the concern, zram supports "writeback_limit" feature.
any writeback. If admin want to measure writeback count in a certain The "writeback_limit_enable"'s default value is 0 so that it doesn't limit
period, he could know it via /sys/block/zram0/bd_stat's 3rd column. any writeback. IOW, if admin want to apply writeback budget, he should
enable writeback_limit_enable via
$ echo 1 > /sys/block/zramX/writeback_limit_enable
Once writeback_limit_enable is set, zram doesn't allow any writeback
until admin set the budget via /sys/block/zramX/writeback_limit.
(If admin doesn't enable writeback_limit_enable, writeback_limit's value
assigned via /sys/block/zramX/writeback_limit is meaninless.)
If admin want to limit writeback as per-day 400M, he could do it If admin want to limit writeback as per-day 400M, he could do it
like below. like below.
MB_SHIFT=20 $ MB_SHIFT=20
4K_SHIFT=12 $ 4K_SHIFT=12
echo $((400<<MB_SHIFT>>4K_SHIFT)) > \ $ echo $((400<<MB_SHIFT>>4K_SHIFT)) > \
/sys/block/zram0/writeback_limit. /sys/block/zram0/writeback_limit.
$ echo 1 > /sys/block/zram0/writeback_limit_enable
If admin want to allow further write again, he could do it like below If admin want to allow further write again once the bugdet is exausted,
he could do it like below
echo 0 > /sys/block/zram0/writeback_limit $ echo $((400<<MB_SHIFT>>4K_SHIFT)) > \
/sys/block/zram0/writeback_limit
If admin want to see remaining writeback budget since he set, If admin want to see remaining writeback budget since he set,
cat /sys/block/zram0/writeback_limit $ cat /sys/block/zramX/writeback_limit
If admin want to disable writeback limit, he could do
$ echo 0 > /sys/block/zramX/writeback_limit_enable
The writeback_limit count will reset whenever you reset zram(e.g., The writeback_limit count will reset whenever you reset zram(e.g.,
system reboot, echo 1 > /sys/block/zramX/reset) so keeping how many of system reboot, echo 1 > /sys/block/zramX/reset) so keeping how many of
writeback happened until you reset the zram to allocate extra writeback writeback happened until you reset the zram to allocate extra writeback
budget in next setting is user's job. budget in next setting is user's job.
If admin want to measure writeback count in a certain period, he could
know it via /sys/block/zram0/bd_stat's 3rd column.
= memory tracking = memory tracking
With CONFIG_ZRAM_MEMORY_TRACKING, user can know information of the With CONFIG_ZRAM_MEMORY_TRACKING, user can know information of the
......
...@@ -316,11 +316,9 @@ static ssize_t idle_store(struct device *dev, ...@@ -316,11 +316,9 @@ static ssize_t idle_store(struct device *dev,
* See the comment in writeback_store. * See the comment in writeback_store.
*/ */
zram_slot_lock(zram, index); zram_slot_lock(zram, index);
if (!zram_allocated(zram, index) || if (zram_allocated(zram, index) &&
zram_test_flag(zram, index, ZRAM_UNDER_WB)) !zram_test_flag(zram, index, ZRAM_UNDER_WB))
goto next; zram_set_flag(zram, index, ZRAM_IDLE);
zram_set_flag(zram, index, ZRAM_IDLE);
next:
zram_slot_unlock(zram, index); zram_slot_unlock(zram, index);
} }
...@@ -330,6 +328,41 @@ static ssize_t idle_store(struct device *dev, ...@@ -330,6 +328,41 @@ static ssize_t idle_store(struct device *dev,
} }
#ifdef CONFIG_ZRAM_WRITEBACK #ifdef CONFIG_ZRAM_WRITEBACK
static ssize_t writeback_limit_enable_store(struct device *dev,
struct device_attribute *attr, const char *buf, size_t len)
{
struct zram *zram = dev_to_zram(dev);
u64 val;
ssize_t ret = -EINVAL;
if (kstrtoull(buf, 10, &val))
return ret;
down_read(&zram->init_lock);
spin_lock(&zram->wb_limit_lock);
zram->wb_limit_enable = val;
spin_unlock(&zram->wb_limit_lock);
up_read(&zram->init_lock);
ret = len;
return ret;
}
static ssize_t writeback_limit_enable_show(struct device *dev,
struct device_attribute *attr, char *buf)
{
bool val;
struct zram *zram = dev_to_zram(dev);
down_read(&zram->init_lock);
spin_lock(&zram->wb_limit_lock);
val = zram->wb_limit_enable;
spin_unlock(&zram->wb_limit_lock);
up_read(&zram->init_lock);
return scnprintf(buf, PAGE_SIZE, "%d\n", val);
}
static ssize_t writeback_limit_store(struct device *dev, static ssize_t writeback_limit_store(struct device *dev,
struct device_attribute *attr, const char *buf, size_t len) struct device_attribute *attr, const char *buf, size_t len)
{ {
...@@ -341,9 +374,9 @@ static ssize_t writeback_limit_store(struct device *dev, ...@@ -341,9 +374,9 @@ static ssize_t writeback_limit_store(struct device *dev,
return ret; return ret;
down_read(&zram->init_lock); down_read(&zram->init_lock);
atomic64_set(&zram->stats.bd_wb_limit, val); spin_lock(&zram->wb_limit_lock);
if (val == 0) zram->bd_wb_limit = val;
zram->stop_writeback = false; spin_unlock(&zram->wb_limit_lock);
up_read(&zram->init_lock); up_read(&zram->init_lock);
ret = len; ret = len;
...@@ -357,7 +390,9 @@ static ssize_t writeback_limit_show(struct device *dev, ...@@ -357,7 +390,9 @@ static ssize_t writeback_limit_show(struct device *dev,
struct zram *zram = dev_to_zram(dev); struct zram *zram = dev_to_zram(dev);
down_read(&zram->init_lock); down_read(&zram->init_lock);
val = atomic64_read(&zram->stats.bd_wb_limit); spin_lock(&zram->wb_limit_lock);
val = zram->bd_wb_limit;
spin_unlock(&zram->wb_limit_lock);
up_read(&zram->init_lock); up_read(&zram->init_lock);
return scnprintf(buf, PAGE_SIZE, "%llu\n", val); return scnprintf(buf, PAGE_SIZE, "%llu\n", val);
...@@ -588,8 +623,8 @@ static int read_from_bdev_async(struct zram *zram, struct bio_vec *bvec, ...@@ -588,8 +623,8 @@ static int read_from_bdev_async(struct zram *zram, struct bio_vec *bvec,
return 1; return 1;
} }
#define HUGE_WRITEBACK 0x1 #define HUGE_WRITEBACK 1
#define IDLE_WRITEBACK 0x2 #define IDLE_WRITEBACK 2
static ssize_t writeback_store(struct device *dev, static ssize_t writeback_store(struct device *dev,
struct device_attribute *attr, const char *buf, size_t len) struct device_attribute *attr, const char *buf, size_t len)
...@@ -602,7 +637,7 @@ static ssize_t writeback_store(struct device *dev, ...@@ -602,7 +637,7 @@ static ssize_t writeback_store(struct device *dev,
struct page *page; struct page *page;
ssize_t ret, sz; ssize_t ret, sz;
char mode_buf[8]; char mode_buf[8];
unsigned long mode = -1UL; int mode = -1;
unsigned long blk_idx = 0; unsigned long blk_idx = 0;
sz = strscpy(mode_buf, buf, sizeof(mode_buf)); sz = strscpy(mode_buf, buf, sizeof(mode_buf));
...@@ -618,7 +653,7 @@ static ssize_t writeback_store(struct device *dev, ...@@ -618,7 +653,7 @@ static ssize_t writeback_store(struct device *dev,
else if (!strcmp(mode_buf, "huge")) else if (!strcmp(mode_buf, "huge"))
mode = HUGE_WRITEBACK; mode = HUGE_WRITEBACK;
if (mode == -1UL) if (mode == -1)
return -EINVAL; return -EINVAL;
down_read(&zram->init_lock); down_read(&zram->init_lock);
...@@ -645,10 +680,13 @@ static ssize_t writeback_store(struct device *dev, ...@@ -645,10 +680,13 @@ static ssize_t writeback_store(struct device *dev,
bvec.bv_len = PAGE_SIZE; bvec.bv_len = PAGE_SIZE;
bvec.bv_offset = 0; bvec.bv_offset = 0;
if (zram->stop_writeback) { spin_lock(&zram->wb_limit_lock);
if (zram->wb_limit_enable && !zram->bd_wb_limit) {
spin_unlock(&zram->wb_limit_lock);
ret = -EIO; ret = -EIO;
break; break;
} }
spin_unlock(&zram->wb_limit_lock);
if (!blk_idx) { if (!blk_idx) {
blk_idx = alloc_block_bdev(zram); blk_idx = alloc_block_bdev(zram);
...@@ -667,10 +705,11 @@ static ssize_t writeback_store(struct device *dev, ...@@ -667,10 +705,11 @@ static ssize_t writeback_store(struct device *dev,
zram_test_flag(zram, index, ZRAM_UNDER_WB)) zram_test_flag(zram, index, ZRAM_UNDER_WB))
goto next; goto next;
if ((mode & IDLE_WRITEBACK && if (mode == IDLE_WRITEBACK &&
!zram_test_flag(zram, index, ZRAM_IDLE)) && !zram_test_flag(zram, index, ZRAM_IDLE))
(mode & HUGE_WRITEBACK && goto next;
!zram_test_flag(zram, index, ZRAM_HUGE))) if (mode == HUGE_WRITEBACK &&
!zram_test_flag(zram, index, ZRAM_HUGE))
goto next; goto next;
/* /*
* Clearing ZRAM_UNDER_WB is duty of caller. * Clearing ZRAM_UNDER_WB is duty of caller.
...@@ -732,11 +771,10 @@ static ssize_t writeback_store(struct device *dev, ...@@ -732,11 +771,10 @@ static ssize_t writeback_store(struct device *dev,
zram_set_element(zram, index, blk_idx); zram_set_element(zram, index, blk_idx);
blk_idx = 0; blk_idx = 0;
atomic64_inc(&zram->stats.pages_stored); atomic64_inc(&zram->stats.pages_stored);
if (atomic64_add_unless(&zram->stats.bd_wb_limit, spin_lock(&zram->wb_limit_lock);
-1 << (PAGE_SHIFT - 12), 0)) { if (zram->wb_limit_enable && zram->bd_wb_limit > 0)
if (atomic64_read(&zram->stats.bd_wb_limit) == 0) zram->bd_wb_limit -= 1UL << (PAGE_SHIFT - 12);
zram->stop_writeback = true; spin_unlock(&zram->wb_limit_lock);
}
next: next:
zram_slot_unlock(zram, index); zram_slot_unlock(zram, index);
} }
...@@ -1812,6 +1850,7 @@ static DEVICE_ATTR_RW(comp_algorithm); ...@@ -1812,6 +1850,7 @@ static DEVICE_ATTR_RW(comp_algorithm);
static DEVICE_ATTR_RW(backing_dev); static DEVICE_ATTR_RW(backing_dev);
static DEVICE_ATTR_WO(writeback); static DEVICE_ATTR_WO(writeback);
static DEVICE_ATTR_RW(writeback_limit); static DEVICE_ATTR_RW(writeback_limit);
static DEVICE_ATTR_RW(writeback_limit_enable);
#endif #endif
static struct attribute *zram_disk_attrs[] = { static struct attribute *zram_disk_attrs[] = {
...@@ -1828,6 +1867,7 @@ static struct attribute *zram_disk_attrs[] = { ...@@ -1828,6 +1867,7 @@ static struct attribute *zram_disk_attrs[] = {
&dev_attr_backing_dev.attr, &dev_attr_backing_dev.attr,
&dev_attr_writeback.attr, &dev_attr_writeback.attr,
&dev_attr_writeback_limit.attr, &dev_attr_writeback_limit.attr,
&dev_attr_writeback_limit_enable.attr,
#endif #endif
&dev_attr_io_stat.attr, &dev_attr_io_stat.attr,
&dev_attr_mm_stat.attr, &dev_attr_mm_stat.attr,
...@@ -1867,7 +1907,9 @@ static int zram_add(void) ...@@ -1867,7 +1907,9 @@ static int zram_add(void)
device_id = ret; device_id = ret;
init_rwsem(&zram->init_lock); init_rwsem(&zram->init_lock);
#ifdef CONFIG_ZRAM_WRITEBACK
spin_lock_init(&zram->wb_limit_lock);
#endif
queue = blk_alloc_queue(GFP_KERNEL); queue = blk_alloc_queue(GFP_KERNEL);
if (!queue) { if (!queue) {
pr_err("Error allocating disk queue for device %d\n", pr_err("Error allocating disk queue for device %d\n",
......
...@@ -86,7 +86,6 @@ struct zram_stats { ...@@ -86,7 +86,6 @@ struct zram_stats {
atomic64_t bd_count; /* no. of pages in backing device */ atomic64_t bd_count; /* no. of pages in backing device */
atomic64_t bd_reads; /* no. of reads from backing device */ atomic64_t bd_reads; /* no. of reads from backing device */
atomic64_t bd_writes; /* no. of writes from backing device */ atomic64_t bd_writes; /* no. of writes from backing device */
atomic64_t bd_wb_limit; /* writeback limit of backing device */
#endif #endif
}; };
...@@ -114,8 +113,10 @@ struct zram { ...@@ -114,8 +113,10 @@ struct zram {
*/ */
bool claim; /* Protected by bdev->bd_mutex */ bool claim; /* Protected by bdev->bd_mutex */
struct file *backing_dev; struct file *backing_dev;
bool stop_writeback;
#ifdef CONFIG_ZRAM_WRITEBACK #ifdef CONFIG_ZRAM_WRITEBACK
spinlock_t wb_limit_lock;
bool wb_limit_enable;
u64 bd_wb_limit;
struct block_device *bdev; struct block_device *bdev;
unsigned int old_block_size; unsigned int old_block_size;
unsigned long *bitmap; unsigned long *bitmap;
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment