-diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
-index caf007ba..5a61cec7 100644
---- a/fs/ext4/ext4.h
-+++ b/fs/ext4/ext4.h
-@@ -1407,6 +1407,9 @@ struct ext4_sb_info {
- unsigned int s_mb_min_to_scan;
- unsigned int s_mb_stats;
- unsigned int s_mb_order2_reqs;
-+ ext4_fsblk_t s_mb_c1_blocks;
-+ ext4_fsblk_t s_mb_c2_blocks;
-+ ext4_fsblk_t s_mb_c3_blocks;
- unsigned long *s_mb_prealloc_table;
- unsigned long s_mb_prealloc_table_size;
- unsigned int s_mb_group_prealloc;
-@@ -1424,6 +1427,9 @@ struct ext4_sb_info {
- atomic_t s_bal_goals; /* goal hits */
- atomic_t s_bal_breaks; /* too long searches */
- atomic_t s_bal_2orders; /* 2^order hits */
-+ /* cX loop didn't find blocks */
-+ atomic64_t s_bal_cX_failed[3];
-+ atomic64_t s_bal_cX_skipped[3];
- spinlock_t s_bal_lock;
- unsigned long s_mb_buddies_generated;
- unsigned long long s_mb_generation_time;
-diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
-index e4a00c78..14a4fb25 100644
---- a/fs/ext4/mballoc.c
-+++ b/fs/ext4/mballoc.c
+Index: linux-stage/fs/ext4/mballoc.c
+===================================================================
+--- linux-stage.orig/fs/ext4/mballoc.c
++++ linux-stage/fs/ext4/mballoc.c
@@ -2098,6 +2098,21 @@ static int ext4_mb_good_group(struct ext4_allocation_context *ac,
return 0;
}
static noinline_for_stack int
ext4_mb_regular_allocator(struct ext4_allocation_context *ac)
{
-@@ -2107,6 +2122,7 @@ ext4_mb_regular_allocator(struct ext4_allocation_context *ac)
+@@ -2087,6 +2102,7 @@ ext4_mb_regular_allocator(struct ext4_al
struct ext4_sb_info *sbi;
struct super_block *sb;
struct ext4_buddy e4b;
sb = ac->ac_sb;
sbi = EXT4_SB(sb);
-@@ -2156,6 +2172,21 @@ ext4_mb_regular_allocator(struct ext4_allocation_context *ac)
+@@ -2136,6 +2152,21 @@ ext4_mb_regular_allocator(struct ext4_al
/* Let's just scan groups to find more-less suitable blocks */
cr = ac->ac_2order ? 0 : 1;
/*
* cr == 0 try to get exact allocation,
* cr == 3 try to get anything
-@@ -2213,6 +2244,9 @@ repeat:
+@@ -2193,6 +2224,9 @@ repeat:
if (ac->ac_status != AC_STATUS_CONTINUE)
break;
}
}
if (ac->ac_b_ex.fe_len > 0 && ac->ac_status != AC_STATUS_FOUND &&
-@@ -2336,6 +2370,93 @@ static const struct seq_operations ext4_mb_seq_groups_ops = {
+@@ -2316,6 +2350,93 @@ static const struct seq_operations ext4_
.show = ext4_mb_seq_groups_show,
};
+
#define EXT4_MB_PREALLOC_TABLE "prealloc_table"
- static ssize_t ext4_mb_prealloc_table_proc_write(struct file *file,
-@@ -2738,6 +2859,7 @@ static int ext4_groupinfo_create_slab(size_t size)
+ static int ext4_mb_check_and_update_prealloc(struct ext4_sb_info *sbi,
+@@ -2730,6 +2851,8 @@ static int ext4_groupinfo_create_slab(si
return 0;
}
-+#define THRESHOLD_BLOCKS(ts) (ext4_blocks_count(sbi->s_es) / 100 * ts)
++#define THRESHOLD_BLOCKS(sbi, percent) \
++ (ext4_blocks_count((sbi)->s_es) / 100 * (percent))
int ext4_mb_init(struct super_block *sb)
{
struct ext4_sb_info *sbi = EXT4_SB(sb);
sbi->s_mb_min_to_scan = MB_DEFAULT_MIN_TO_SCAN;
sbi->s_mb_stats = MB_DEFAULT_STATS;
sbi->s_mb_order2_reqs = MB_DEFAULT_ORDER2_REQS;
-+ sbi->s_mb_c1_blocks =
-+ min_t(u64, THRESHOLD_BLOCKS(MB_DEFAULT_C1_THRESHOLD),
-+ MB_DEFAULT_MAX_C1_BYTES >> sb->s_blocksize_bits);
-+ sbi->s_mb_c2_blocks =
-+ min_t(u64, THRESHOLD_BLOCKS(MB_DEFAULT_C2_THRESHOLD),
-+ MB_DEFAULT_MAX_C2_BYTES >> sb->s_blocksize_bits);
-+ sbi->s_mb_c3_blocks =
-+ min_t(u64, THRESHOLD_BLOCKS(MB_DEFAULT_C3_THRESHOLD),
-+ MB_DEFAULT_MAX_C3_BYTES >> sb->s_blocksize_bits);
++ if (!sbi->s_mb_c1_blocks)
++ sbi->s_mb_c1_blocks =
++ THRESHOLD_BLOCKS(sbi, MB_DEFAULT_C1_THRESHOLD);
++ if (!sbi->s_mb_c2_blocks)
++ sbi->s_mb_c2_blocks =
++ THRESHOLD_BLOCKS(sbi, MB_DEFAULT_C2_THRESHOLD);
++ if (!sbi->s_mb_c3_blocks)
++ sbi->s_mb_c3_blocks =
++ THRESHOLD_BLOCKS(sbi, MB_DEFAULT_C3_THRESHOLD);
/*
* The default group preallocation is 512, which for 4k block
* sizes translates to 2 megabytes. However for bigalloc file
-@@ -2879,6 +3010,8 @@ int ext4_mb_init(struct super_block *sb)
+@@ -2853,6 +2978,8 @@ int ext4_mb_init(struct super_block *sb)
proc_create_data(EXT4_MB_PREALLOC_TABLE, S_IFREG | S_IRUGO |
S_IWUSR, sbi->s_proc,
&ext4_mb_prealloc_seq_fops, sb);
proc_create_data("mb_last_group", S_IFREG | S_IRUGO |
S_IWUSR, sbi->s_proc,
&ext4_mb_seq_last_group_fops, sb);
-@@ -2932,6 +3065,7 @@ int ext4_mb_release(struct super_block *sb)
+@@ -2906,6 +3033,7 @@ int ext4_mb_release(struct super_block *
remove_proc_entry("mb_last_group", sbi->s_proc);
remove_proc_entry("mb_last_start", sbi->s_proc);
remove_proc_entry(EXT4_MB_PREALLOC_TABLE, sbi->s_proc);
}
if (sbi->s_group_info) {
-@@ -2961,6 +3095,16 @@ int ext4_mb_release(struct super_block *sb)
- atomic_read(&sbi->s_bal_allocated),
+@@ -2936,6 +3064,16 @@ int ext4_mb_release(struct super_block *
atomic_read(&sbi->s_bal_reqs),
atomic_read(&sbi->s_bal_success));
-+ ext4_msg(sb, KERN_INFO,
+ ext4_msg(sb, KERN_INFO,
+ "mballoc: (%lu, %lu, %lu) useless c(0,1,2) loops",
+ atomic64_read(&sbi->s_bal_cX_failed[0]),
+ atomic64_read(&sbi->s_bal_cX_failed[1]),
+ atomic64_read(&sbi->s_bal_cX_skipped[0]),
+ atomic64_read(&sbi->s_bal_cX_skipped[1]),
+ atomic64_read(&sbi->s_bal_cX_skipped[2]));
- ext4_msg(sb, KERN_INFO,
++ ext4_msg(sb, KERN_INFO,
"mballoc: %u extents scanned, %u goal hits, "
"%u 2^N hits, %u breaks, %u lost",
-diff --git a/fs/ext4/mballoc.h b/fs/ext4/mballoc.h
-index 81d56d0b..3ee74eac 100644
---- a/fs/ext4/mballoc.h
-+++ b/fs/ext4/mballoc.h
-@@ -84,6 +84,12 @@ extern ushort ext4_mballoc_debug;
- * for which requests use 2^N search using buddies
- */
- #define MB_DEFAULT_ORDER2_REQS 8
-+#define MB_DEFAULT_C1_THRESHOLD 15
-+#define MB_DEFAULT_C2_THRESHOLD 10
-+#define MB_DEFAULT_C3_THRESHOLD 5
-+#define MB_DEFAULT_MAX_C1_BYTES (15ULL << 40) /* 15TB */
-+#define MB_DEFAULT_MAX_C2_BYTES (10ULL << 40)
-+#define MB_DEFAULT_MAX_C3_BYTES (5ULL << 40)
+ atomic_read(&sbi->s_bal_ex_scanned),
+Index: linux-stage/fs/ext4/ext4.h
+===================================================================
+--- linux-stage.orig/fs/ext4/ext4.h
++++ linux-stage/fs/ext4/ext4.h
+@@ -1409,6 +1409,9 @@ struct ext4_sb_info {
+ unsigned int s_mb_min_to_scan;
+ unsigned int s_mb_stats;
+ unsigned int s_mb_order2_reqs;
++ ext4_fsblk_t s_mb_c1_blocks;
++ ext4_fsblk_t s_mb_c2_blocks;
++ ext4_fsblk_t s_mb_c3_blocks;
+ unsigned long *s_mb_prealloc_table;
+ unsigned int s_mb_group_prealloc;
+ unsigned int s_max_dir_size_kb;
+@@ -1425,6 +1428,9 @@ struct ext4_sb_info {
+ atomic_t s_bal_goals; /* goal hits */
+ atomic_t s_bal_breaks; /* too long searches */
+ atomic_t s_bal_2orders; /* 2^order hits */
++ /* cX loop didn't find blocks */
++ atomic64_t s_bal_cX_failed[3];
++ atomic64_t s_bal_cX_skipped[3];
+ spinlock_t s_bal_lock;
+ unsigned long s_mb_buddies_generated;
+ unsigned long long s_mb_generation_time;
+@@ -2115,6 +2121,8 @@ struct ext4_sb_info {
+ extern void ext4_end_bitmap_read(struct buffer_head *bh, int uptodate);
- /*
- * default group prealloc size 512 blocks
-diff --git a/fs/ext4/super.c b/fs/ext4/super.c
-index ca8b50c8..b492245b 100644
---- a/fs/ext4/super.c
-+++ b/fs/ext4/super.c
-@@ -2746,6 +2746,73 @@ static ssize_t sbi_deprecated_show(struct ext4_attr *a,
+ /* mballoc.c */
++extern int save_threshold_percent(struct ext4_sb_info *sbi, const char *buf,
++ ext4_fsblk_t *blocks);
+ extern long ext4_mb_stats;
+ extern long ext4_mb_max_to_scan;
+ extern int ext4_mb_init(struct super_block *);
+Index: linux-stage/fs/ext4/super.c
+===================================================================
+--- linux-stage.orig/fs/ext4/super.c
++++ linux-stage/fs/ext4/super.c
+@@ -1208,6 +1208,7 @@ enum {
+ Opt_nomblk_io_submit, Opt_block_validity, Opt_noblock_validity,
+ Opt_inode_readahead_blks, Opt_journal_ioprio,
+ Opt_dioread_nolock, Opt_dioread_lock,
++ Opt_mb_c1_threshold, Opt_mb_c2_threshold, Opt_mb_c3_threshold,
+ Opt_no_mbcache,
+ Opt_discard, Opt_nodiscard, Opt_init_itable, Opt_noinit_itable,
+ Opt_max_dir_size_kb, Opt_nojournal_checksum,
+@@ -1287,6 +1288,9 @@ static const match_table_t tokens = {
+ {Opt_nodiscard, "nodiscard"},
+ {Opt_init_itable, "init_itable=%u"},
+ {Opt_no_mbcache, "no_mbcache"},
++ {Opt_mb_c1_threshold, "mb_c1_threshold=%s"},
++ {Opt_mb_c2_threshold, "mb_c2_threshold=%s"},
++ {Opt_mb_c3_threshold, "mb_c3_threshold=%s"},
+ {Opt_init_itable, "init_itable"},
+ {Opt_noinit_itable, "noinit_itable"},
+ {Opt_max_dir_size_kb, "max_dir_size_kb=%u"},
+@@ -1449,6 +1453,9 @@ static const struct mount_opts {
+ {Opt_auto_da_alloc, EXT4_MOUNT_NO_AUTO_DA_ALLOC, MOPT_CLEAR},
+ {Opt_noinit_itable, EXT4_MOUNT_INIT_INODE_TABLE, MOPT_CLEAR},
+ {Opt_no_mbcache, EXT4_MOUNT_NO_MBCACHE, MOPT_SET},
++ {Opt_mb_c1_threshold, 0, MOPT_STRING},
++ {Opt_mb_c2_threshold, 0, MOPT_STRING},
++ {Opt_mb_c3_threshold, 0, MOPT_STRING},
+ {Opt_commit, 0, MOPT_GTE0},
+ {Opt_max_batch_time, 0, MOPT_GTE0},
+ {Opt_min_batch_time, 0, MOPT_GTE0},
+@@ -1571,6 +1578,12 @@ static const struct mount_opts {
+ sbi->s_max_dir_size_kb = arg;
+ /* reset s_warning_dir_size and make it re-calculated */
+ sbi->s_warning_dir_size = 0;
++ } else if (token == Opt_mb_c1_threshold) {
++ save_threshold_percent(sbi, args[0].from, &sbi->s_mb_c1_blocks);
++ } else if (token == Opt_mb_c2_threshold) {
++ save_threshold_percent(sbi, args[0].from, &sbi->s_mb_c2_blocks);
++ } else if (token == Opt_mb_c3_threshold) {
++ save_threshold_percent(sbi, args[0].from, &sbi->s_mb_c3_blocks);
+ } else if (token == Opt_stripe) {
+ sbi->s_stripe = arg;
+ } else if (token == Opt_resuid) {
+@@ -2734,6 +2747,74 @@ static ssize_t sbi_deprecated_show(struc
return snprintf(buf, PAGE_SIZE, "%d\n", a->u.deprecated_val);
}
-+static int save_threshold(struct ext4_sb_info *sbi, const char *buf,
-+ ext4_fsblk_t *blocks) {
++int save_threshold_percent(struct ext4_sb_info *sbi, const char *buf,
++ ext4_fsblk_t *blocks) {
+ unsigned long long val;
+
-+ if (!parse_strtoull(buf, 100, &val)) {
++ if (!parse_strtoull(buf, 100, &val) && val <= 100) {
+ *blocks = val * ext4_blocks_count(sbi->s_es) / 100;
+ return 0;
+ }
+ return -EINVAL;
+}
+
-+#define THRESHOLD_PERCENT(ts) (ts * 100 / ext4_blocks_count(sbi->s_es))
++#define THRESHOLD_PERCENT(sbi, blocks) \
++ (((blocks) - 1) * 100 / ext4_blocks_count((sbi)->s_es) + 1)
+static ssize_t mb_c1_threshold_store(struct ext4_attr *a,
+ struct ext4_sb_info *sbi,
+ const char *buf, size_t count)
+{
+ int ret;
+
-+ ret = save_threshold(sbi, buf, &sbi->s_mb_c1_blocks);
++ ret = save_threshold_percent(sbi, buf, &sbi->s_mb_c1_blocks);
+
+ return ret ? ret : count;
+}
+ struct ext4_sb_info *sbi, char *buf)
+{
+ return snprintf(buf, PAGE_SIZE, "%llu\n",
-+ THRESHOLD_PERCENT(sbi->s_mb_c1_blocks));
++ THRESHOLD_PERCENT(sbi, sbi->s_mb_c1_blocks));
+}
+
+static ssize_t mb_c2_threshold_store(struct ext4_attr *a,
+{
+ int ret;
+
-+ ret = save_threshold(sbi, buf, &sbi->s_mb_c2_blocks);
++ ret = save_threshold_percent(sbi, buf, &sbi->s_mb_c2_blocks);
+ return ret ? ret : count;
+}
+
+ struct ext4_sb_info *sbi, char *buf)
+{
+ return snprintf(buf, PAGE_SIZE, "%llu\n",
-+ THRESHOLD_PERCENT(sbi->s_mb_c2_blocks));
++ THRESHOLD_PERCENT(sbi, sbi->s_mb_c2_blocks));
+}
+
+static ssize_t mb_c3_threshold_store(struct ext4_attr *a,
+{
+ int ret;
+
-+ ret = save_threshold(sbi, buf, &sbi->s_mb_c3_blocks);
++ ret = save_threshold_percent(sbi, buf, &sbi->s_mb_c3_blocks);
+
+ return ret ? ret : count;
+}
+ struct ext4_sb_info *sbi, char *buf)
+{
+ return snprintf(buf, PAGE_SIZE, "%llu\n",
-+ THRESHOLD_PERCENT(sbi->s_mb_c3_blocks));
++ THRESHOLD_PERCENT(sbi, sbi->s_mb_c3_blocks));
+}
+
+
#define EXT4_ATTR_OFFSET(_name,_mode,_show,_store,_elname) \
static struct ext4_attr ext4_attr_##_name = { \
.attr = {.name = __stringify(_name), .mode = _mode }, \
-@@ -2802,6 +2869,9 @@ EXT4_RW_ATTR_SBI_UI(mb_stats, s_mb_stats);
+@@ -2790,6 +2857,9 @@ EXT4_RW_ATTR_SBI_UI(mb_stats, s_mb_stats
EXT4_RW_ATTR_SBI_UI(mb_max_to_scan, s_mb_max_to_scan);
EXT4_RW_ATTR_SBI_UI(mb_min_to_scan, s_mb_min_to_scan);
EXT4_RW_ATTR_SBI_UI(mb_order2_req, s_mb_order2_reqs);
EXT4_RW_ATTR_SBI_UI(mb_small_req, s_mb_small_req);
EXT4_RW_ATTR_SBI_UI(mb_large_req, s_mb_large_req);
EXT4_RW_ATTR_SBI_UI(mb_group_prealloc, s_mb_group_prealloc);
-@@ -2833,6 +2903,9 @@ static struct attribute *ext4_attrs[] = {
+@@ -2820,6 +2890,9 @@ static struct attribute *ext4_attrs[] =
ATTR_LIST(mb_max_to_scan),
ATTR_LIST(mb_min_to_scan),
ATTR_LIST(mb_order2_req),
ATTR_LIST(mb_small_req),
ATTR_LIST(mb_large_req),
ATTR_LIST(mb_group_prealloc),
+Index: linux-stage/fs/ext4/mballoc.h
+===================================================================
+--- linux-stage.orig/fs/ext4/mballoc.h
++++ linux-stage/fs/ext4/mballoc.h
+@@ -84,6 +84,12 @@ extern ushort ext4_mballoc_debug;
+ * for which requests use 2^N search using buddies
+ */
+ #define MB_DEFAULT_ORDER2_REQS 8
++#define MB_DEFAULT_C1_THRESHOLD 15
++#define MB_DEFAULT_C2_THRESHOLD 10
++#define MB_DEFAULT_C3_THRESHOLD 5
++#define MB_DEFAULT_MAX_C1_BYTES (15ULL << 40) /* 15TB */
++#define MB_DEFAULT_MAX_C2_BYTES (10ULL << 40)
++#define MB_DEFAULT_MAX_C3_BYTES (5ULL << 40)
+
+ /*
+ * default group prealloc size 512 blocks
git://git.whamcloud.com / fs / lustre-release.git / commitdiff