Re: [Ext2-devel] [PATCH 9/9]ext4 super block changes for >32 bitblocks numbers
From: Avantika Mathur LTC
Date: Wed Aug 23 2006 - 20:26:00 EST
On Wed, 2006-08-09 at 23:41 -0700, Andrew Morton wrote:
On Wed, 09 Aug 2006 18:22:25 -0700
Mingming Cao <cmm@xxxxxxxxxx <mailto:cmm@xxxxxxxxxx>> wrote:
> In-kernel and on-disk super block changes to support >32 bit blocks numbers.
>
> +static inline u32 EXT4_RELATIVE_ENCODE(ext4_fsblk_t group_base,
> + ext4_fsblk_t fs_block)
> +{
> + s32 gdp_block;
> +
> + if (fs_block < (1ULL<<32) && group_base < (1ULL<<32))
> + return fs_block;
> +
> + gdp_block = (fs_block - group_base);
> + BUG_ON ((group_base + gdp_block) != fs_block);
> +
> + return gdp_block;
> +}
> +
> +static inline ext4_fsblk_t EXT4_RELATIVE_DECODE(ext4_fsblk_t group_base,
> + u32 gdp_block)
> +{
> + if (group_base >= (1ULL<<32))
> + return group_base + (s32) gdp_block;
> +
> + if ((s32) gdp_block >= 0 && gdp_block < group_base &&
> + group_base + gdp_block >= (1ULL<<32))
> + return group_base + gdp_block;
> +
> + return gdp_block;
> +}
These seem far too large and far too commonly called to be inlined.
>
> +
> +#define EXT4_BLOCKS_COUNT(s) \
> + (ext4_fsblk_t)(((__u64)le32_to_cpu((s)->s_blocks_count_hi) << 32) | \
> + (__u64)le32_to_cpu((s)->s_blocks_count))
> +#define EXT4_BLOCKS_COUNT_SET(s,v) do { \
> + (s)->s_blocks_count = cpu_to_le32((v)); \
> + (s)->s_blocks_count_hi = cpu_to_le32(((__u64)(v)) >> 32); \
> +} while (0)
> +
> +#define EXT4_R_BLOCKS_COUNT(s) \
> + (ext4_fsblk_t)(((__u64)le32_to_cpu((s)->s_r_blocks_count_hi) << 32) | \
> + (__u64)le32_to_cpu((s)->s_r_blocks_count))
> +#define EXT4_R_BLOCKS_COUNT_SET(s,v) do { \
> + (s)->s_r_blocks_count = cpu_to_le32((v)); \
> + (s)->s_r_blocks_count_hi = cpu_to_le32(((__u64)(v)) >> 32); \
> +} while (0)
> +
> +#define EXT4_FREE_BLOCKS_COUNT(s) \
> + (ext4_fsblk_t)(((__u64)le32_to_cpu((s)->s_free_blocks_count_hi) << 32) | \
> + (__u64)le32_to_cpu((s)->s_free_blocks_count))
> +#define EXT4_FREE_BLOCKS_COUNT_SET(s,v) do { \
> + (s)->s_free_blocks_count = cpu_to_le32((v)); \
> + (s)->s_free_blocks_count_hi = cpu_to_le32(((__u64)(v)) >> 32); \
> +} while (0)
Can these not be implemented as C functions?
Here is a patch to implement these macros as C functions
Thanks,
Avantika
---
Avantika Mathur
IBM LTC
---------------------------
diff -uprN linux-2.6.18-rc4/fs/ext4/balloc.c linux-2.6.18-rc4-patched/fs/ext4/balloc.c
--- linux-2.6.18-rc4/fs/ext4/balloc.c 2006-08-17 09:31:13.000000000 -0700
+++ linux-2.6.18-rc4-patched/fs/ext4/balloc.c 2006-08-14 15:33:30.000000000 -0700
@@ -331,7 +331,7 @@ void ext4_free_blocks_sb(handle_t *handl
es = sbi->s_es;
if (block < le32_to_cpu(es->s_first_data_block) ||
block + count < block ||
- block + count > EXT4_BLOCKS_COUNT(es)) {
+ block + count > ext4_blocks_count(es)) {
ext4_error (sb, "ext4_free_blocks",
"Freeing blocks not in datazone - "
"block = "E3FSBLK", count = %lu", block, count);
@@ -1174,7 +1174,7 @@ static int ext4_has_free_blocks(struct e
ext4_fsblk_t free_blocks, root_blocks;
free_blocks = percpu_counter_read_positive(&sbi->s_freeblocks_counter);
- root_blocks = EXT4_R_BLOCKS_COUNT(sbi->s_es);
+ root_blocks = ext4_r_blocks_count(sbi->s_es);
if (free_blocks < root_blocks + 1 && !capable(CAP_SYS_RESOURCE) &&
sbi->s_resuid != current->fsuid &&
(sbi->s_resgid == 0 || !in_group_p (sbi->s_resgid))) {
@@ -1273,7 +1273,7 @@ ext4_fsblk_t ext4_new_blocks(handle_t *h
* First, test whether the goal block is free.
*/
if (goal < le32_to_cpu(es->s_first_data_block) ||
- goal >= EXT4_BLOCKS_COUNT(es))
+ goal >= ext4_blocks_count(es))
goal = le32_to_cpu(es->s_first_data_block);
ext4_get_group_no_and_offset(sb, goal, &group_no, &grp_target_blk);
gdp = ext4_get_group_desc(sb, group_no, &gdp_bh);
@@ -1419,11 +1419,11 @@ allocated:
jbd_unlock_bh_state(bitmap_bh);
#endif
- if (ret_block + num - 1 >= EXT4_BLOCKS_COUNT(es)) {
+ if (ret_block + num - 1 >= ext4_blocks_count(es)) {
ext4_error(sb, "ext4_new_block",
"block("E3FSBLK") >= blocks count("E3FSBLK") - "
"block_group = %lu, es == %p ", ret_block,
- EXT4_BLOCKS_COUNT(es), group_no, es);
+ ext4_blocks_count(es), group_no, es);
goto out;
}
diff -uprN linux-2.6.18-rc4/fs/ext4/ialloc.c linux-2.6.18-rc4-patched/fs/ext4/ialloc.c
--- linux-2.6.18-rc4/fs/ext4/ialloc.c 2006-08-17 09:31:13.000000000 -0700
+++ linux-2.6.18-rc4-patched/fs/ext4/ialloc.c 2006-08-14 11:17:51.000000000 -0700
@@ -306,7 +306,7 @@ static int find_group_orlov(struct super
goto fallback;
}
- blocks_per_dir = EXT4_BLOCKS_COUNT(es) - freeb;
+ blocks_per_dir = ext4_blocks_count(es) - freeb;
sector_div(blocks_per_dir, ndirs);
max_dirs = ndirs / ngroups + inodes_per_group / 16;
diff -uprN linux-2.6.18-rc4/fs/ext4/resize.c linux-2.6.18-rc4-patched/fs/ext4/resize.c
--- linux-2.6.18-rc4/fs/ext4/resize.c 2006-08-17 09:31:13.000000000 -0700
+++ linux-2.6.18-rc4-patched/fs/ext4/resize.c 2006-08-14 11:00:25.000000000 -0700
@@ -27,7 +27,7 @@ static int verify_group_input(struct sup
{
struct ext4_sb_info *sbi = EXT4_SB(sb);
struct ext4_super_block *es = sbi->s_es;
- ext4_fsblk_t start = EXT4_BLOCKS_COUNT(es);
+ ext4_fsblk_t start = ext4_blocks_count(es);
ext4_fsblk_t end = start + input->blocks_count;
unsigned group = input->group;
ext4_fsblk_t itend = input->inode_table + sbi->s_itb_per_group;
@@ -286,6 +286,18 @@ exit_journal:
return err;
}
+static void ext4_blocks_count_set(struct ext4_super_block *es, __u32 v)
+{
+ es->s_blocks_count = cpu_to_le32(v);
+ es->s_blocks_count_hi = cpu_to_le32(((__u64) v) >> 32);
+}
+
+static void ext4_r_blocks_count_set(struct ext4_super_block *es, __u32 v)
+{
+ es->s_r_blocks_count = cpu_to_le32(v);
+ es->s_r_blocks_count_hi = cpu_to_le32(((__u64) v) >> 32);
+}
+
/*
* Iterate through the groups which hold BACKUP superblock/GDT copies in an
* ext4 filesystem. The counters should be initialized to 1, 5, and 7 before
@@ -836,7 +848,7 @@ int ext4_group_add(struct super_block *s
* blocks/inodes before the group is live won't actually let us
* allocate the new space yet.
*/
- EXT4_BLOCKS_COUNT_SET(es, EXT4_BLOCKS_COUNT(es) +
+ ext4_blocks_count_set(es, ext4_blocks_count(es) +
input->blocks_count);
es->s_inodes_count = cpu_to_le32(le32_to_cpu(es->s_inodes_count) +
EXT4_INODES_PER_GROUP(sb));
@@ -872,7 +884,7 @@ int ext4_group_add(struct super_block *s
/* Update the reserved block counts only once the new group is
* active. */
- EXT4_R_BLOCKS_COUNT_SET(es, EXT4_R_BLOCKS_COUNT(es) +
+ ext4_r_blocks_count_set(es, ext4_r_blocks_count(es) +
input->reserved_blocks);
/* Update the free space counts */
@@ -923,7 +935,7 @@ int ext4_group_extend(struct super_block
/* We don't need to worry about locking wrt other resizers just
* yet: we're going to revalidate es->s_blocks_count after
* taking lock_super() below. */
- o_blocks_count = EXT4_BLOCKS_COUNT(es);
+ o_blocks_count = ext4_blocks_count(es);
o_groups_count = EXT4_SB(sb)->s_groups_count;
if (test_opt(sb, DEBUG))
@@ -989,7 +1001,7 @@ int ext4_group_extend(struct super_block
}
lock_super(sb);
- if (o_blocks_count != EXT4_BLOCKS_COUNT(es)) {
+ if (o_blocks_count != ext4_blocks_count(es)) {
ext4_warning(sb, __FUNCTION__,
"multiple resizers run on filesystem!");
unlock_super(sb);
@@ -1005,7 +1017,7 @@ int ext4_group_extend(struct super_block
ext4_journal_stop(handle);
goto exit_put;
}
- EXT4_BLOCKS_COUNT_SET(es, o_blocks_count + add);
+ ext4_blocks_count_set(es, o_blocks_count + add);
ext4_journal_dirty_metadata(handle, EXT4_SB(sb)->s_sbh);
sb->s_dirt = 1;
unlock_super(sb);
@@ -1018,7 +1030,7 @@ int ext4_group_extend(struct super_block
goto exit_put;
if (test_opt(sb, DEBUG))
printk(KERN_DEBUG "EXT4-fs: extended group to %llu blocks\n",
- EXT4_BLOCKS_COUNT(es));
+ ext4_blocks_count(es));
update_backups(sb, EXT4_SB(sb)->s_sbh->b_blocknr, (char *)es,
sizeof(struct ext4_super_block));
exit_put:
diff -uprN linux-2.6.18-rc4/fs/ext4/super.c linux-2.6.18-rc4-patched/fs/ext4/super.c
--- linux-2.6.18-rc4/fs/ext4/super.c 2006-08-17 09:31:13.000000000 -0700
+++ linux-2.6.18-rc4-patched/fs/ext4/super.c 2006-08-17 14:46:51.000000000 -0700
@@ -62,6 +62,50 @@ static void ext4_unlockfs(struct super_b
static void ext4_write_super (struct super_block * sb);
static void ext4_write_super_lockfs(struct super_block *sb);
+ext4_fsblk_t ext4_blocks_count(struct ext4_super_block *es)
+{
+ return (ext4_fsblk_t) ((__u64)le32_to_cpu(es->s_blocks_count_hi) << 32 |
+ (__u64)le32_to_cpu(es->s_blocks_count));
+}
+
+ext4_fsblk_t ext4_r_blocks_count(struct ext4_super_block *es)
+{
+ return (ext4_fsblk_t) ((__u64)le32_to_cpu(es->s_r_blocks_count_hi) << 32 |
+ (__u64)le32_to_cpu(es->s_r_blocks_count));
+}
+
+u32 ext4_relative_encode(ext4_fsblk_t group_base, ext4_fsblk_t fs_block)
+{
+ s32 gdp_block;
+
+ if (fs_block < (1ULL<<32) && group_base < (1ULL<<32))
+ return fs_block;
+
+ gdp_block = (fs_block - group_base);
+ BUG_ON ((group_base + gdp_block) != fs_block);
+
+ return gdp_block;
+}
+
+ext4_fsblk_t ext4_relative_decode(ext4_fsblk_t group_base, u32 gdp_block)
+{
+ if (group_base >= (1ULL<<32))
+ return group_base + (s32) gdp_block;
+
+ if ((s32) gdp_block >= 0 && gdp_block < group_base &&
+ group_base + gdp_block >= (1ULL<<32))
+ return group_base + gdp_block;
+
+ return gdp_block;
+}
+
+
+static void ext4_free_blocks_count_set(struct ext4_super_block *es, __u32 v)
+{
+ es->s_free_blocks_count = cpu_to_le32(v);
+ es->s_free_blocks_count_hi = cpu_to_le32(((__u64) v) >> 32);
+}
+
/*
* Wrappers for jbd2_journal_start/end.
*
@@ -1191,7 +1235,7 @@ static int ext4_check_descriptors (struc
gdp++;
}
- EXT4_FREE_BLOCKS_COUNT_SET(sbi->s_es, ext4_count_free_blocks(sb));
+ ext4_free_blocks_count_set(sbi->s_es, ext4_count_free_blocks(sb));
sbi->s_es->s_free_inodes_count=cpu_to_le32(ext4_count_free_inodes(sb));
return 1;
}
@@ -1579,7 +1623,7 @@ static int ext4_fill_super (struct super
goto failed_mount;
}
- if (EXT4_BLOCKS_COUNT(es) >
+ if (ext4_blocks_count(es) >
(sector_t)(~0ULL) >> (sb->s_blocksize_bits - 9)) {
printk(KERN_ERR "EXT4-fs: filesystem on %s:"
" too large to mount safely\n", sb->s_id);
@@ -1591,7 +1635,7 @@ static int ext4_fill_super (struct super
if (EXT4_BLOCKS_PER_GROUP(sb) == 0)
goto cantfind_ext4;
- blocks_count = (EXT4_BLOCKS_COUNT(es) -
+ blocks_count = (ext4_blocks_count(es) -
le32_to_cpu(es->s_first_data_block) +
EXT4_BLOCKS_PER_GROUP(sb) - 1);
do_div(blocks_count, EXT4_BLOCKS_PER_GROUP(sb));
@@ -1909,7 +1953,7 @@ static journal_t *ext4_get_dev_journal(s
goto out_bdev;
}
- len = EXT4_BLOCKS_COUNT(es);
+ len = ext4_blocks_count(es);
start = sb_block + 1;
brelse(bh); /* we're done with the superblock */
@@ -2079,7 +2123,7 @@ static void ext4_commit_super (struct su
if (!sbh)
return;
es->s_wtime = cpu_to_le32(get_seconds());
- EXT4_FREE_BLOCKS_COUNT_SET(es, ext4_count_free_blocks(sb));
+ ext4_free_blocks_count_set(es, ext4_count_free_blocks(sb));
es->s_free_inodes_count = cpu_to_le32(ext4_count_free_inodes(sb));
BUFFER_TRACE(sbh, "marking dirty");
mark_buffer_dirty(sbh);
@@ -2272,7 +2316,7 @@ static int ext4_remount (struct super_bl
ext4_init_journal_params(sb, sbi->s_journal);
if ((*flags & MS_RDONLY) != (sb->s_flags & MS_RDONLY) ||
- n_blocks_count > EXT4_BLOCKS_COUNT(es)) {
+ n_blocks_count > ext4_blocks_count(es)) {
if (sbi->s_mount_opt & EXT4_MOUNT_ABORT) {
err = -EROFS;
goto restore_opts;
@@ -2393,10 +2437,10 @@ static int ext4_statfs (struct dentry *
buf->f_type = EXT4_SUPER_MAGIC;
buf->f_bsize = sb->s_blocksize;
- buf->f_blocks = EXT4_BLOCKS_COUNT(es) - overhead;
+ buf->f_blocks = ext4_blocks_count(es) - overhead;
buf->f_bfree = percpu_counter_sum(&sbi->s_freeblocks_counter);
- buf->f_bavail = buf->f_bfree - EXT4_R_BLOCKS_COUNT(es);
- if (buf->f_bfree < EXT4_R_BLOCKS_COUNT(es))
+ buf->f_bavail = buf->f_bfree - ext4_r_blocks_count(es);
+ if (buf->f_bfree < ext4_r_blocks_count(es))
buf->f_bavail = 0;
buf->f_files = le32_to_cpu(es->s_inodes_count);
buf->f_ffree = percpu_counter_sum(&sbi->s_freeinodes_counter);
diff -uprN linux-2.6.18-rc4/include/linux/ext4_fs.h linux-2.6.18-rc4-patched/include/linux/ext4_fs.h
--- linux-2.6.18-rc4/include/linux/ext4_fs.h 2006-08-17 09:31:13.000000000 -0700
+++ linux-2.6.18-rc4-patched/include/linux/ext4_fs.h 2006-08-17 14:39:22.000000000 -0700
@@ -139,46 +139,20 @@ struct ext4_group_desc
#ifdef __KERNEL__
#include <linux/ext4_fs_i.h>
#include <linux/ext4_fs_sb.h>
-static inline u32 EXT4_RELATIVE_ENCODE(ext4_fsblk_t group_base,
- ext4_fsblk_t fs_block)
-{
- s32 gdp_block;
-
- if (fs_block < (1ULL<<32) && group_base < (1ULL<<32))
- return fs_block;
-
- gdp_block = (fs_block - group_base);
- BUG_ON ((group_base + gdp_block) != fs_block);
-
- return gdp_block;
-}
-
-static inline ext4_fsblk_t EXT4_RELATIVE_DECODE(ext4_fsblk_t group_base,
- u32 gdp_block)
-{
- if (group_base >= (1ULL<<32))
- return group_base + (s32) gdp_block;
-
- if ((s32) gdp_block >= 0 && gdp_block < group_base &&
- group_base + gdp_block >= (1ULL<<32))
- return group_base + gdp_block;
-
- return gdp_block;
-}
#define EXT4_BLOCK_BITMAP(bg, group_base) \
- EXT4_RELATIVE_DECODE(group_base, le32_to_cpu((bg)->bg_block_bitmap))
+ ext4_relative_decode(group_base, le32_to_cpu((bg)->bg_block_bitmap))
#define EXT4_INODE_BITMAP(bg, group_base) \
- EXT4_RELATIVE_DECODE(group_base, le32_to_cpu((bg)->bg_inode_bitmap))
+ ext4_relative_decode(group_base, le32_to_cpu((bg)->bg_inode_bitmap))
#define EXT4_INODE_TABLE(bg, group_base) \
- EXT4_RELATIVE_DECODE(group_base, le32_to_cpu((bg)->bg_inode_table))
+ ext4_relative_decode(group_base, le32_to_cpu((bg)->bg_inode_table))
#define EXT4_BLOCK_BITMAP_SET(bg, group_base, value) \
- do {(bg)->bg_block_bitmap = EXT4_RELATIVE_ENCODE(group_base, value);} while(0)
+ do {(bg)->bg_block_bitmap = ext4_relative_encode(group_base, value);} while(0)
#define EXT4_INODE_BITMAP_SET(bg, group_base, value) \
- do {(bg)->bg_inode_bitmap = EXT4_RELATIVE_ENCODE(group_base, value);} while(0)
+ do {(bg)->bg_inode_bitmap = ext4_relative_encode(group_base, value);} while(0)
#define EXT4_INODE_TABLE_SET(bg, group_base, value) \
- do {(bg)->bg_inode_table = EXT4_RELATIVE_ENCODE(group_base, value);} while(0)
+ do {(bg)->bg_inode_table = ext4_relative_encode(group_base, value);} while(0)
#define EXT4_IS_USED_BLOCK_BITMAP(bg) \
((bg)->bg_block_bitmap != 0)
@@ -543,30 +517,9 @@ struct ext4_super_block {
__u32 s_reserved[169]; /* Padding to the end of the block */
};
-
-#define EXT4_BLOCKS_COUNT(s) \
- (ext4_fsblk_t)(((__u64)le32_to_cpu((s)->s_blocks_count_hi) << 32) | \
- (__u64)le32_to_cpu((s)->s_blocks_count))
-#define EXT4_BLOCKS_COUNT_SET(s,v) do { \
- (s)->s_blocks_count = cpu_to_le32((v)); \
- (s)->s_blocks_count_hi = cpu_to_le32(((__u64)(v)) >> 32); \
-} while (0)
-
-#define EXT4_R_BLOCKS_COUNT(s) \
- (ext4_fsblk_t)(((__u64)le32_to_cpu((s)->s_r_blocks_count_hi) << 32) | \
- (__u64)le32_to_cpu((s)->s_r_blocks_count))
-#define EXT4_R_BLOCKS_COUNT_SET(s,v) do { \
- (s)->s_r_blocks_count = cpu_to_le32((v)); \
- (s)->s_r_blocks_count_hi = cpu_to_le32(((__u64)(v)) >> 32); \
-} while (0)
-
#define EXT4_FREE_BLOCKS_COUNT(s) \
(ext4_fsblk_t)(((__u64)le32_to_cpu((s)->s_free_blocks_count_hi) << 32) | \
(__u64)le32_to_cpu((s)->s_free_blocks_count))
-#define EXT4_FREE_BLOCKS_COUNT_SET(s,v) do { \
- (s)->s_free_blocks_count = cpu_to_le32((v)); \
- (s)->s_free_blocks_count_hi = cpu_to_le32(((__u64)(v)) >> 32); \
-} while (0)
#ifdef __KERNEL__
static inline struct ext4_sb_info * EXT4_SB(struct super_block *sb)
@@ -947,6 +900,10 @@ extern void ext4_abort (struct super_blo
extern void ext4_warning (struct super_block *, const char *, const char *, ...)
__attribute__ ((format (printf, 3, 4)));
extern void ext4_update_dynamic_rev (struct super_block *sb);
+extern ext4_fsblk_t ext4_blocks_count(struct ext4_super_block *es);
+extern ext4_fsblk_t ext4_r_blocks_count(struct ext4_super_block *es);
+extern u32 ext4_relative_encode(ext4_fsblk_t group_base, ext4_fsblk_t fs_block);
+extern ext4_fsblk_t ext4_relative_decode(ext4_fsblk_t group_base, u32 gdp_block);
#define ext4_std_error(sb, errno) \
do { \
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/