[PATCH] ext4: fall back to vmalloc() for large allocations

[PATCH] ext4: fall back to vmalloc() for large allocations

[Andreas Dilger]

For very large ext4 filesystems (128TB and larger) kmalloc() of
some per-group structures can fail at mount time due to memory
fragmentation.  If kmalloc() fails, fall back to vmalloc() for
the s_group_info and s_group_desc arrays.

Signed-off-by: Yu Jian <yujian@whamcloud.com>
Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
---
 fs/ext4/mballoc.c |   49 +++++++++++++++++++++++++++++++++----------------
 fs/ext4/super.c   |   29 +++++++++++++++++++++++------
 2 files changed, 56 insertions(+), 22 deletions(-)

diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
index 6ed859d..72c5796 100644
--- a/fs/ext4/mballoc.c
+++ b/fs/ext4/mballoc.c
@@ -2325,25 +2325,37 @@ static int ext4_mb_init_backend(struct super_block *sb)
 	while (array_size < sizeof(*sbi->s_group_info) *
 	       num_meta_group_infos_max)
 		array_size = array_size << 1;
-	/* An 8TB filesystem with 64-bit pointers requires a 4096 byte
-	 * kmalloc. A 128kb malloc should suffice for a 256TB filesystem.
-	 * So a two level scheme suffices for now. */
+	/* A 16TB filesystem with 64-bit pointers requires an 8192 byte
+	 * kmalloc(). Filesystems larger than 2^32 blocks (16TB normally)
+	 * have group descriptors at least twice as large (64 bytes or
+	 * more vs. 32 bytes for traditional ext3 filesystems), so a 128TB
+	 * filesystem needs a 128kB allocation, which may need vmalloc(). */
 	sbi->s_group_info = kzalloc(array_size, GFP_KERNEL);
 	if (sbi->s_group_info == NULL) {
-		printk(KERN_ERR "EXT4-fs: can't allocate buddy meta group\n");
-		return -ENOMEM;
+		sbi->s_group_info = vmalloc(array_size);
+		if (sbi->s_group_info != NULL) {
+			memset(sbi->s_group_info, 0, array_size);
+		} else {
+			ext4_msg(sb, KERN_ERR, "no memory for groupinfo (%u)\n",
+				 array_size);
+			return -ENOMEM;
+		}
 	}
 	sbi->s_buddy_cache = new_inode(sb);
 	if (sbi->s_buddy_cache == NULL) {
-		printk(KERN_ERR "EXT4-fs: can't get new inode\n");
+		ext4_msg(sb, KERN_ERR, "can't get new inode\n");
 		goto err_freesgi;
 	}
-	sbi->s_buddy_cache->i_ino = get_next_ino();
+	/* To avoid potentially colliding with an valid on-disk inode number,
+	 * use EXT4_BAD_INO for the buddy cache inode number.  This inode is
+	 * not in the inode hash, so it should never be found by iget(), but
+	 * this will avoid confusion if it ever shows up during debugging. */
+	sbi->s_buddy_cache->i_ino = EXT4_BAD_INO;
 	EXT4_I(sbi->s_buddy_cache)->i_disksize = 0;
 	for (i = 0; i < ngroups; i++) {
 		desc = ext4_get_group_desc(sb, i, NULL);
 		if (desc == NULL) {
-			printk(KERN_ERR
+			ext4_msg(sb, KERN_ERR,
 				"EXT4-fs: can't read descriptor %u\n", i);
 			goto err_freebuddy;
 		}
@@ -2362,7 +2374,10 @@ err_freebuddy:
 		kfree(sbi->s_group_info[i]);
 	iput(sbi->s_buddy_cache);
 err_freesgi:
-	kfree(sbi->s_group_info);
+	if (is_vmalloc_addr(sbi->s_group_info))
+		vfree(sbi->s_group_info);
+	else
+		kfree(sbi->s_group_info);
 	return -ENOMEM;
 }
 
@@ -2457,12 +2472,6 @@ int ext4_mb_init(struct super_block *sb, int needs_recovery)
 		i++;
 	} while (i <= sb->s_blocksize_bits + 1);
 
-	/* init file for buddy data */
-	ret = ext4_mb_init_backend(sb);
-	if (ret != 0) {
-		goto out;
-	}
-
 	spin_lock_init(&sbi->s_md_lock);
 	spin_lock_init(&sbi->s_bal_lock);
 
@@ -2487,6 +2496,11 @@ int ext4_mb_init(struct super_block *sb, int needs_recovery)
 		spin_lock_init(&lg->lg_prealloc_lock);
 	}
 
+	/* init file for buddy data */
+	ret = ext4_mb_init_backend(sb);
+	if (ret != 0)
+		goto out;
+
 	if (sbi->s_proc)
 		proc_create_data("mb_groups", S_IRUGO, sbi->s_proc,
 				 &ext4_mb_seq_groups_fops, sb);
@@ -2544,7 +2558,10 @@ int ext4_mb_release(struct super_block *sb)
 			EXT4_DESC_PER_BLOCK_BITS(sb);
 		for (i = 0; i < num_meta_group_infos; i++)
 			kfree(sbi->s_group_info[i]);
-		kfree(sbi->s_group_info);
+		if (is_vmalloc_addr(sbi->s_group_info))
+			vfree(sbi->s_group_info);
+		else
+			kfree(sbi->s_group_info);
 	}
 	kfree(sbi->s_mb_offsets);
 	kfree(sbi->s_mb_maxs);
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index 9ea71aa..556084b 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -789,7 +789,12 @@ static void ext4_put_super(struct super_block *sb)
 
 	for (i = 0; i < sbi->s_gdb_count; i++)
 		brelse(sbi->s_group_desc[i]);
-	kfree(sbi->s_group_desc);
+
+	if (is_vmalloc_addr(sbi->s_group_desc))
+		vfree(sbi->s_group_desc);
+	else
+		kfree(sbi->s_group_desc);
+
 	if (is_vmalloc_addr(sbi->s_flex_groups))
 		vfree(sbi->s_flex_groups);
 	else
@@ -3059,6 +3064,7 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
 	int ret = -ENOMEM;
 	int blocksize;
 	unsigned int db_count;
+	size_t size;
 	unsigned int i;
 	int needs_recovery, has_huge_files;
 	__u64 blocks_count;
@@ -3408,11 +3414,18 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
 			(EXT4_MAX_BLOCK_FILE_PHYS / EXT4_BLOCKS_PER_GROUP(sb)));
 	db_count = (sbi->s_groups_count + EXT4_DESC_PER_BLOCK(sb) - 1) /
 		   EXT4_DESC_PER_BLOCK(sb);
-	sbi->s_group_desc = kmalloc(db_count * sizeof(struct buffer_head *),
-				    GFP_KERNEL);
+	size = (size_t)db_count * sizeof(struct buffer_head *);
+	sbi->s_group_desc = kzalloc(size, GFP_KERNEL);
 	if (sbi->s_group_desc == NULL) {
-		ext4_msg(sb, KERN_ERR, "not enough memory");
-		goto failed_mount;
+		sbi->s_group_desc = vmalloc(size);
+		if (sbi->s_group_desc != NULL) {
+			memset(sbi->s_group_desc, 0, size);
+		} else {
+			ext4_msg(sb, KERN_ERR, "no memory for %u groups (%u)\n",
+				 sbi->s_groups_count, (unsigned int)size);
+			ret = -ENOMEM;
+			goto failed_mount;
+	 	}
 	}
 
 #ifdef CONFIG_PROC_FS
@@ -3756,7 +3769,11 @@ failed_mount3:
 failed_mount2:
 	for (i = 0; i < db_count; i++)
 		brelse(sbi->s_group_desc[i]);
-	kfree(sbi->s_group_desc);
+
+	if (is_vmalloc_addr(sbi->s_group_desc))
+		vfree(sbi->s_group_desc);
+	else
+		kfree(sbi->s_group_desc);
 failed_mount:
 	if (sbi->s_proc) {
 		remove_proc_entry(sb->s_id, ext4_proc_root);
-- 
1.7.3.4

Re: [PATCH] ext4: fall back to vmalloc() for large allocations

[Ted Ts'o]

On Tue, Jul 12, 2011 at 03:40:46PM -0600, Andreas Dilger wrote:
> For very large ext4 filesystems (128TB and larger) kmalloc() of
> some per-group structures can fail at mount time due to memory
> fragmentation.  If kmalloc() fails, fall back to vmalloc() for
> the s_group_info and s_group_desc arrays.
> 
> Signed-off-by: Yu Jian <yujian@whamcloud.com>
> Signed-off-by: Andreas Dilger <adilger@whamcloud.com>

Andras, was this patch authored by Yu Jian or by you?

>  	sbi->s_buddy_cache = new_inode(sb);
>  	if (sbi->s_buddy_cache == NULL) {
> -		printk(KERN_ERR "EXT4-fs: can't get new inode\n");
> +		ext4_msg(sb, KERN_ERR, "can't get new inode\n");
>  		goto err_freesgi;
>  	}

Using ext4_msg instead of printk is good, but that really should be a
separate patch.


> -	sbi->s_buddy_cache->i_ino = get_next_ino();
> +	/* To avoid potentially colliding with an valid on-disk inode number,
> +	 * use EXT4_BAD_INO for the buddy cache inode number.  This inode is
> +	 * not in the inode hash, so it should never be found by iget(), but
> +	 * this will avoid confusion if it ever shows up during debugging. */
> +	sbi->s_buddy_cache->i_ino = EXT4_BAD_INO;

This should be a separate patch.

> @@ -2457,12 +2472,6 @@ int ext4_mb_init(struct super_block *sb, int needs_recovery)
>  		i++;
>  	} while (i <= sb->s_blocksize_bits + 1);
>  
> -	/* init file for buddy data */
> -	ret = ext4_mb_init_backend(sb);
> -	if (ret != 0) {
> -		goto out;
> -	}
> -
>  	spin_lock_init(&sbi->s_md_lock);
>  	spin_lock_init(&sbi->s_bal_lock);
>  
> @@ -2487,6 +2496,11 @@ int ext4_mb_init(struct super_block *sb, int needs_recovery)
>  		spin_lock_init(&lg->lg_prealloc_lock);
>  	}
>  
> +	/* init file for buddy data */
> +	ret = ext4_mb_init_backend(sb);
> +	if (ret != 0)
> +		goto out;
> +
>  	if (sbi->s_proc)
>  		proc_create_data("mb_groups", S_IRUGO, sbi->s_proc,
>  				 &ext4_mb_seq_groups_fops, sb);

Why are you moving ext4_mb_init_backend()?  This should be a separate
patch, with an explanation of what is going on....

       	       		      	      	    - Ted

Re: [PATCH] ext4: fall back to vmalloc() for large allocations

[Andreas Dilger]

On 2011-07-17, at 7:25 PM, Ted Ts'o wrote:
> On Tue, Jul 12, 2011 at 03:40:46PM -0600, Andreas Dilger wrote:
>> For very large ext4 filesystems (128TB and larger) kmalloc() of
>> some per-group structures can fail at mount time due to memory
>> fragmentation.  If kmalloc() fails, fall back to vmalloc() for
>> the s_group_info and s_group_desc arrays.
>> 
>> Signed-off-by: Yu Jian <yujian@whamcloud.com>
>> Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
> 
> Andreas, was this patch authored by Yu Jian or by you?

Yu Jian wrote it with help from me, and I made some modifications afterward,
so I put both of our names down.

>> 	sbi->s_buddy_cache = new_inode(sb);
>> 	if (sbi->s_buddy_cache == NULL) {
>> -		printk(KERN_ERR "EXT4-fs: can't get new inode\n");
>> +		ext4_msg(sb, KERN_ERR, "can't get new inode\n");
>> 		goto err_freesgi;
>> 	}
> 
> Using ext4_msg instead of printk is good, but that really should be a
> separate patch.

> 
>> -	sbi->s_buddy_cache->i_ino = get_next_ino();
>> +	/* To avoid potentially colliding with an valid on-disk inode number,
>> +	 * use EXT4_BAD_INO for the buddy cache inode number.  This inode is
>> +	 * not in the inode hash, so it should never be found by iget(), but
>> +	 * this will avoid confusion if it ever shows up during debugging. */
>> +	sbi->s_buddy_cache->i_ino = EXT4_BAD_INO;
> 
> This should be a separate patch.

Sure.

>> @@ -2457,12 +2472,6 @@ int ext4_mb_init(struct super_block *sb, int needs_recovery)
>> 		i++;
>> 	} while (i <= sb->s_blocksize_bits + 1);
>> 
>> -	/* init file for buddy data */
>> -	ret = ext4_mb_init_backend(sb);
>> -	if (ret != 0) {
>> -		goto out;
>> -	}
>> -
>> 	spin_lock_init(&sbi->s_md_lock);
>> 	spin_lock_init(&sbi->s_bal_lock);
>> 
>> @@ -2487,6 +2496,11 @@ int ext4_mb_init(struct super_block *sb, int needs_recovery)
>> 		spin_lock_init(&lg->lg_prealloc_lock);
>> 	}
>> 
>> +	/* init file for buddy data */
>> +	ret = ext4_mb_init_backend(sb);
>> +	if (ret != 0)
>> +		goto out;
>> +
>> 	if (sbi->s_proc)
>> 		proc_create_data("mb_groups", S_IRUGO, sbi->s_proc,
>> 				 &ext4_mb_seq_groups_fops, sb);
> 
> Why are you moving ext4_mb_init_backend()?  This should be a separate
> patch, with an explanation of what is going on....

In ext4_mb_init(), if the s_locality_group allocation fails it will
currently cause the allocations made in ext4_mb_init_backend() to
be leaked.  Moving the ext4_mb_init_backend() allocation after the
s_locality_group allocation avoids that problem.

Cheers, Andreas
--
Andreas Dilger 
Principal Engineer
Whamcloud, Inc.

Re: [PATCH] ext4: fall back to vmalloc() for large allocations

[Ted Ts'o]

On Sun, Jul 17, 2011 at 11:12:34PM -0600, Andreas Dilger wrote:
> 
> Yu Jian wrote it with help from me, and I made some modifications afterward,
> so I put both of our names down.

OK, I just wanted to know who I should credit as the author in Git.
If you think it should be Yu Jian, could you put in a From: Yu Jian
line?  If you think you made sufficient modifications that you think
it's mainly your code, that's fine.  I just wasn't sure which would
have been more correct, and whether you might have just forgotten an
explicit From: header or not.  Just want to give credit where credit
is due!  :-)

I'm going to mark this patch as "revisions requested" in patchworkand
want for you to send out a broken out patch series.

Many thanks!!

					- Ted

[PATCH 0/5] Break up "fall back to vmalloc() for large allocations"

[Theodore Ts'o]

This patch series breaks up Andreas and Yu Jian's patch "fall back to
vmalloc() for large allocations" into 5 patches, with the following
changes:

1) I created new helper functions, ext4_kvmalloc(), ext4_kzalloc, and
	ext4_kvfree() to simplify the code.

2) Fixed the patch so that online resize would work correctly if
	s_group_info and s_group_desc were allocated using vmalloc().
	(Yeah, there are other reasons why online resize won't work for
	such large allocations, but let's not add to the problems; also,
	there's always the possibility that kmalloc might fail for even
	for a small allocation.)

3) Fixed many more places in mballoc.c where ext4_msg() would be better
	than just printk().

I was waiting for Andreas and/or Yu Jian to resubmit, but since they
didn't, and I wanted these fixes, I decided to break up the patch
myself.

						- Ted

Theodore Ts'o (3):
  ext4: introduce ext4_kvmalloc(), ext4_kzalloc(), and ext4_kvfree()
  ext4: use ext4_kvzalloc()/ext4_kvmalloc() for s_group_desc and
    s_group_info
  ext4: use ext4_msg() instead of printk in mballoc

Yu Jian (2):
  ext4: use EXT4_BAD_INO for buddy cache to avoid colliding with valid
    inode #
  ext4: prevent memory leaks from ext4_mb_init_backend() on error path

 fs/ext4/ext4.h    |    3 ++
 fs/ext4/mballoc.c |  103 +++++++++++++++++++++++++++++------------------------
 fs/ext4/resize.c  |   13 ++++---
 fs/ext4/super.c   |   63 +++++++++++++++++++++-----------
 4 files changed, 107 insertions(+), 75 deletions(-)

-- 
1.7.4.1.22.gec8e1.dirty

[PATCH 1/5] ext4: introduce ext4_kvmalloc(), ext4_kzalloc(), and ext4_kvfree()

[Theodore Ts'o]

Introduce new helper functions which try kmalloc, and then fall back
to vmalloc if necessary, and use them for allocating and deallocating
s_flex_groups.

Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
 fs/ext4/ext4.h  |    3 +++
 fs/ext4/super.c |   54 ++++++++++++++++++++++++++++++++++++------------------
 2 files changed, 39 insertions(+), 18 deletions(-)

diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index ba2009b..db9fead 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -1874,6 +1874,9 @@ extern int ext4_group_extend(struct super_block *sb,
 				ext4_fsblk_t n_blocks_count);
 
 /* super.c */
+extern void *ext4_kvmalloc(size_t size, gfp_t flags);
+extern void *ext4_kvzalloc(size_t size, gfp_t flags);
+extern void ext4_kvfree(void *ptr);
 extern void __ext4_error(struct super_block *, const char *, unsigned int,
 			 const char *, ...)
 	__attribute__ ((format (printf, 4, 5)));
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index cfe9f39..658f586 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -110,6 +110,35 @@ static struct file_system_type ext3_fs_type = {
 #define IS_EXT3_SB(sb) (0)
 #endif
 
+void *ext4_kvmalloc(size_t size, gfp_t flags)
+{
+	void *ret;
+
+	ret = kmalloc(size, flags);
+	if (!ret)
+		ret = __vmalloc(size, flags, PAGE_KERNEL);
+	return ret;
+}
+
+void *ext4_kvzalloc(size_t size, gfp_t flags)
+{
+	void *ret;
+
+	ret = kmalloc(size, flags);
+	if (!ret)
+		ret = __vmalloc(size, flags | __GFP_ZERO, PAGE_KERNEL);
+	return ret;
+}
+
+void ext4_kvfree(void *ptr)
+{
+	if (is_vmalloc_addr(ptr))
+		vfree(ptr);
+	else
+		kfree(ptr);
+
+}
+
 ext4_fsblk_t ext4_block_bitmap(struct super_block *sb,
 			       struct ext4_group_desc *bg)
 {
@@ -791,10 +820,7 @@ static void ext4_put_super(struct super_block *sb)
 	for (i = 0; i < sbi->s_gdb_count; i++)
 		brelse(sbi->s_group_desc[i]);
 	kfree(sbi->s_group_desc);
-	if (is_vmalloc_addr(sbi->s_flex_groups))
-		vfree(sbi->s_flex_groups);
-	else
-		kfree(sbi->s_flex_groups);
+	ext4_kvfree(sbi->s_flex_groups);
 	percpu_counter_destroy(&sbi->s_freeblocks_counter);
 	percpu_counter_destroy(&sbi->s_freeinodes_counter);
 	percpu_counter_destroy(&sbi->s_dirs_counter);
@@ -1977,15 +2003,11 @@ static int ext4_fill_flex_info(struct super_block *sb)
 			((le16_to_cpu(sbi->s_es->s_reserved_gdt_blocks) + 1) <<
 			      EXT4_DESC_PER_BLOCK_BITS(sb))) / groups_per_flex;
 	size = flex_group_count * sizeof(struct flex_groups);
-	sbi->s_flex_groups = kzalloc(size, GFP_KERNEL);
+	sbi->s_flex_groups = ext4_kvzalloc(size, GFP_KERNEL);
 	if (sbi->s_flex_groups == NULL) {
-		sbi->s_flex_groups = vzalloc(size);
-		if (sbi->s_flex_groups == NULL) {
-			ext4_msg(sb, KERN_ERR,
-				 "not enough memory for %u flex groups",
-				 flex_group_count);
-			goto failed;
-		}
+		ext4_msg(sb, KERN_ERR, "not enough memory for %u flex groups",
+			 flex_group_count);
+		goto failed;
 	}
 
 	for (i = 0; i < sbi->s_groups_count; i++) {
@@ -3750,12 +3772,8 @@ failed_mount_wq:
 	}
 failed_mount3:
 	del_timer(&sbi->s_err_report);
-	if (sbi->s_flex_groups) {
-		if (is_vmalloc_addr(sbi->s_flex_groups))
-			vfree(sbi->s_flex_groups);
-		else
-			kfree(sbi->s_flex_groups);
-	}
+	if (sbi->s_flex_groups)
+		ext4_kvfree(sbi->s_flex_groups);
 	percpu_counter_destroy(&sbi->s_freeblocks_counter);
 	percpu_counter_destroy(&sbi->s_freeinodes_counter);
 	percpu_counter_destroy(&sbi->s_dirs_counter);
-- 
1.7.4.1.22.gec8e1.dirty

[PATCH] ext4: use kzalloc in ext4_kzalloc()

[Mathias Krause]

Commit 9933fc0i (ext4: introduce ext4_kvmalloc(), ext4_kzalloc(), and
ext4_kvfree()) intruduced wrappers around k*alloc/vmalloc but introduced
a typo for ext4_kzalloc() by not using kzalloc() but kmalloc().

Signed-off-by: Mathias Krause <minipli@googlemail.com>
---
 fs/ext4/super.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index e2d88ba..4687fea 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -124,7 +124,7 @@ void *ext4_kvzalloc(size_t size, gfp_t flags)
 {
 	void *ret;
 
-	ret = kmalloc(size, flags);
+	ret = kzalloc(size, flags);
 	if (!ret)
 		ret = __vmalloc(size, flags | __GFP_ZERO, PAGE_KERNEL);
 	return ret;
-- 
1.5.6.5

Re: [PATCH] ext4: use kzalloc in ext4_kzalloc()

[Ted Ts'o]

On Tue, Aug 02, 2011 at 08:07:15AM +0200, Mathias Krause wrote:
> Commit 9933fc0i (ext4: introduce ext4_kvmalloc(), ext4_kzalloc(), and
> ext4_kvfree()) intruduced wrappers around k*alloc/vmalloc but introduced
> a typo for ext4_kzalloc() by not using kzalloc() but kmalloc().
> 
> Signed-off-by: Mathias Krause <minipli@googlemail.com>

Oops.  Thanks for catching this.  I'll make sure this gets pushed to
Linus ASAP.

					- Ted

[PATCH 2/5] ext4: use ext4_kvzalloc()/ext4_kvmalloc() for s_group_desc and s_group_info

[Theodore Ts'o]

Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
 fs/ext4/mballoc.c |    6 +++---
 fs/ext4/resize.c  |   13 +++++++------
 fs/ext4/super.c   |    9 +++++----
 3 files changed, 15 insertions(+), 13 deletions(-)

diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
index fa716c9..d5021e8 100644
--- a/fs/ext4/mballoc.c
+++ b/fs/ext4/mballoc.c
@@ -2331,7 +2331,7 @@ static int ext4_mb_init_backend(struct super_block *sb)
 	/* An 8TB filesystem with 64-bit pointers requires a 4096 byte
 	 * kmalloc. A 128kb malloc should suffice for a 256TB filesystem.
 	 * So a two level scheme suffices for now. */
-	sbi->s_group_info = kzalloc(array_size, GFP_KERNEL);
+	sbi->s_group_info = ext4_kvzalloc(array_size, GFP_KERNEL);
 	if (sbi->s_group_info == NULL) {
 		printk(KERN_ERR "EXT4-fs: can't allocate buddy meta group\n");
 		return -ENOMEM;
@@ -2365,7 +2365,7 @@ err_freebuddy:
 		kfree(sbi->s_group_info[i]);
 	iput(sbi->s_buddy_cache);
 err_freesgi:
-	kfree(sbi->s_group_info);
+	ext4_kvfree(sbi->s_group_info);
 	return -ENOMEM;
 }
 
@@ -2559,7 +2559,7 @@ int ext4_mb_release(struct super_block *sb)
 			EXT4_DESC_PER_BLOCK_BITS(sb);
 		for (i = 0; i < num_meta_group_infos; i++)
 			kfree(sbi->s_group_info[i]);
-		kfree(sbi->s_group_info);
+		ext4_kvfree(sbi->s_group_info);
 	}
 	kfree(sbi->s_mb_offsets);
 	kfree(sbi->s_mb_maxs);
diff --git a/fs/ext4/resize.c b/fs/ext4/resize.c
index 71085df..707d3f1 100644
--- a/fs/ext4/resize.c
+++ b/fs/ext4/resize.c
@@ -467,12 +467,13 @@ static int add_new_gdb(handle_t *handle, struct inode *inode,
 	if (unlikely(err))
 		goto exit_dindj;
 
-	n_group_desc = kmalloc((gdb_num + 1) * sizeof(struct buffer_head *),
-			GFP_NOFS);
+	n_group_desc = ext4_kvmalloc((gdb_num + 1) *
+				     sizeof(struct buffer_head *),
+				     GFP_NOFS);
 	if (!n_group_desc) {
 		err = -ENOMEM;
-		ext4_warning(sb,
-			      "not enough memory for %lu groups", gdb_num + 1);
+		ext4_warning(sb, "not enough memory for %lu groups",
+			     gdb_num + 1);
 		goto exit_inode;
 	}
 
@@ -507,7 +508,7 @@ static int add_new_gdb(handle_t *handle, struct inode *inode,
 	n_group_desc[gdb_num] = gdb_bh;
 	EXT4_SB(sb)->s_group_desc = n_group_desc;
 	EXT4_SB(sb)->s_gdb_count++;
-	kfree(o_group_desc);
+	ext4_kvfree(o_group_desc);
 
 	le16_add_cpu(&es->s_reserved_gdt_blocks, -1);
 	err = ext4_handle_dirty_metadata(handle, NULL, EXT4_SB(sb)->s_sbh);
@@ -517,7 +518,7 @@ static int add_new_gdb(handle_t *handle, struct inode *inode,
 	return err;
 
 exit_inode:
-	kfree(n_group_desc);
+	ext4_kvfree(n_group_desc);
 	/* ext4_handle_release_buffer(handle, iloc.bh); */
 	brelse(iloc.bh);
 exit_dindj:
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index 658f586..e2d88ba 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -819,7 +819,7 @@ static void ext4_put_super(struct super_block *sb)
 
 	for (i = 0; i < sbi->s_gdb_count; i++)
 		brelse(sbi->s_group_desc[i]);
-	kfree(sbi->s_group_desc);
+	ext4_kvfree(sbi->s_group_desc);
 	ext4_kvfree(sbi->s_flex_groups);
 	percpu_counter_destroy(&sbi->s_freeblocks_counter);
 	percpu_counter_destroy(&sbi->s_freeinodes_counter);
@@ -3439,8 +3439,9 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
 			(EXT4_MAX_BLOCK_FILE_PHYS / EXT4_BLOCKS_PER_GROUP(sb)));
 	db_count = (sbi->s_groups_count + EXT4_DESC_PER_BLOCK(sb) - 1) /
 		   EXT4_DESC_PER_BLOCK(sb);
-	sbi->s_group_desc = kmalloc(db_count * sizeof(struct buffer_head *),
-				    GFP_KERNEL);
+	sbi->s_group_desc = ext4_kvmalloc(db_count *
+					  sizeof(struct buffer_head *),
+					  GFP_KERNEL);
 	if (sbi->s_group_desc == NULL) {
 		ext4_msg(sb, KERN_ERR, "not enough memory");
 		goto failed_mount;
@@ -3783,7 +3784,7 @@ failed_mount3:
 failed_mount2:
 	for (i = 0; i < db_count; i++)
 		brelse(sbi->s_group_desc[i]);
-	kfree(sbi->s_group_desc);
+	ext4_kvfree(sbi->s_group_desc);
 failed_mount:
 	if (sbi->s_proc) {
 		remove_proc_entry(sb->s_id, ext4_proc_root);
-- 
1.7.4.1.22.gec8e1.dirty

[PATCH 3/5] ext4: use ext4_msg() instead of printk in mballoc

[Theodore Ts'o]

Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
 fs/ext4/mballoc.c |   79 ++++++++++++++++++++++++++++-------------------------
 1 files changed, 42 insertions(+), 37 deletions(-)

diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
index d5021e8..262fa47 100644
--- a/fs/ext4/mballoc.c
+++ b/fs/ext4/mballoc.c
@@ -493,10 +493,11 @@ static void mb_cmp_bitmaps(struct ext4_buddy *e4b, void *bitmap)
 		b2 = (unsigned char *) bitmap;
 		for (i = 0; i < e4b->bd_sb->s_blocksize; i++) {
 			if (b1[i] != b2[i]) {
-				printk(KERN_ERR "corruption in group %u "
-				       "at byte %u(%u): %x in copy != %x "
-				       "on disk/prealloc\n",
-				       e4b->bd_group, i, i * 8, b1[i], b2[i]);
+				ext4_msg(e4b->bd_sb, KERN_ERR,
+					 "corruption in group %u "
+					 "at byte %u(%u): %x in copy != %x "
+					 "on disk/prealloc",
+					 e4b->bd_group, i, i * 8, b1[i], b2[i]);
 				BUG();
 			}
 		}
@@ -2224,8 +2225,8 @@ int ext4_mb_add_groupinfo(struct super_block *sb, ext4_group_t group,
 			EXT4_DESC_PER_BLOCK_BITS(sb);
 		meta_group_info = kmalloc(metalen, GFP_KERNEL);
 		if (meta_group_info == NULL) {
-			printk(KERN_ERR "EXT4-fs: can't allocate mem for a "
-			       "buddy group\n");
+			ext4_msg(sb, KERN_ERR, "EXT4-fs: can't allocate mem "
+				 "for a buddy group");
 			goto exit_meta_group_info;
 		}
 		sbi->s_group_info[group >> EXT4_DESC_PER_BLOCK_BITS(sb)] =
@@ -2238,7 +2239,7 @@ int ext4_mb_add_groupinfo(struct super_block *sb, ext4_group_t group,
 
 	meta_group_info[i] = kmem_cache_alloc(cachep, GFP_KERNEL);
 	if (meta_group_info[i] == NULL) {
-		printk(KERN_ERR "EXT4-fs: can't allocate buddy mem\n");
+		ext4_msg(sb, KERN_ERR, "EXT4-fs: can't allocate buddy mem");
 		goto exit_group_info;
 	}
 	memset(meta_group_info[i], 0, kmem_cache_size(cachep));
@@ -2333,12 +2334,12 @@ static int ext4_mb_init_backend(struct super_block *sb)
 	 * So a two level scheme suffices for now. */
 	sbi->s_group_info = ext4_kvzalloc(array_size, GFP_KERNEL);
 	if (sbi->s_group_info == NULL) {
-		printk(KERN_ERR "EXT4-fs: can't allocate buddy meta group\n");
+		ext4_msg(sb, KERN_ERR, "can't allocate buddy meta group");
 		return -ENOMEM;
 	}
 	sbi->s_buddy_cache = new_inode(sb);
 	if (sbi->s_buddy_cache == NULL) {
-		printk(KERN_ERR "EXT4-fs: can't get new inode\n");
+		ext4_msg(sb, KERN_ERR, "can't get new inode");
 		goto err_freesgi;
 	}
 	sbi->s_buddy_cache->i_ino = get_next_ino();
@@ -2346,8 +2347,7 @@ static int ext4_mb_init_backend(struct super_block *sb)
 	for (i = 0; i < ngroups; i++) {
 		desc = ext4_get_group_desc(sb, i, NULL);
 		if (desc == NULL) {
-			printk(KERN_ERR
-				"EXT4-fs: can't read descriptor %u\n", i);
+			ext4_msg(sb, KERN_ERR, "can't read descriptor %u", i);
 			goto err_freebuddy;
 		}
 		if (ext4_mb_add_groupinfo(sb, i, desc) != 0)
@@ -2411,7 +2411,8 @@ static int ext4_groupinfo_create_slab(size_t size)
 
 	mutex_unlock(&ext4_grpinfo_slab_create_mutex);
 	if (!cachep) {
-		printk(KERN_EMERG "EXT4: no memory for groupinfo slab cache\n");
+		printk(KERN_EMERG,
+		       "EXT4-fs: no memory for groupinfo slab cache\n");
 		return -ENOMEM;
 	}
 
@@ -2566,25 +2567,25 @@ int ext4_mb_release(struct super_block *sb)
 	if (sbi->s_buddy_cache)
 		iput(sbi->s_buddy_cache);
 	if (sbi->s_mb_stats) {
-		printk(KERN_INFO
-		       "EXT4-fs: mballoc: %u blocks %u reqs (%u success)\n",
+		ext4_msg(sb, KERN_INFO,
+		       "mballoc: %u blocks %u reqs (%u success)",
 				atomic_read(&sbi->s_bal_allocated),
 				atomic_read(&sbi->s_bal_reqs),
 				atomic_read(&sbi->s_bal_success));
-		printk(KERN_INFO
-		      "EXT4-fs: mballoc: %u extents scanned, %u goal hits, "
-				"%u 2^N hits, %u breaks, %u lost\n",
+		ext4_msg(sb, KERN_INFO,
+		      "mballoc: %u extents scanned, %u goal hits, "
+				"%u 2^N hits, %u breaks, %u lost",
 				atomic_read(&sbi->s_bal_ex_scanned),
 				atomic_read(&sbi->s_bal_goals),
 				atomic_read(&sbi->s_bal_2orders),
 				atomic_read(&sbi->s_bal_breaks),
 				atomic_read(&sbi->s_mb_lost_chunks));
-		printk(KERN_INFO
-		       "EXT4-fs: mballoc: %lu generated and it took %Lu\n",
+		ext4_msg(sb, KERN_INFO,
+		       "mballoc: %lu generated and it took %Lu",
 				sbi->s_mb_buddies_generated,
 				sbi->s_mb_generation_time);
-		printk(KERN_INFO
-		       "EXT4-fs: mballoc: %u preallocated, %u discarded\n",
+		ext4_msg(sb, KERN_INFO,
+		       "mballoc: %u preallocated, %u discarded",
 				atomic_read(&sbi->s_mb_preallocated),
 				atomic_read(&sbi->s_mb_discarded));
 	}
@@ -3024,9 +3025,10 @@ ext4_mb_normalize_request(struct ext4_allocation_context *ac,
 
 	if (start + size <= ac->ac_o_ex.fe_logical &&
 			start > ac->ac_o_ex.fe_logical) {
-		printk(KERN_ERR "start %lu, size %lu, fe_logical %lu\n",
-			(unsigned long) start, (unsigned long) size,
-			(unsigned long) ac->ac_o_ex.fe_logical);
+		ext4_msg(ac->ac_sb, KERN_ERR,
+			 "start %lu, size %lu, fe_logical %lu",
+			 (unsigned long) start, (unsigned long) size,
+			 (unsigned long) ac->ac_o_ex.fe_logical);
 	}
 	BUG_ON(start + size <= ac->ac_o_ex.fe_logical &&
 			start > ac->ac_o_ex.fe_logical);
@@ -3607,10 +3609,11 @@ ext4_mb_release_inode_pa(struct ext4_buddy *e4b, struct buffer_head *bitmap_bh,
 		bit = next + 1;
 	}
 	if (free != pa->pa_free) {
-		printk(KERN_CRIT "pa %p: logic %lu, phys. %lu, len %lu\n",
-			pa, (unsigned long) pa->pa_lstart,
-			(unsigned long) pa->pa_pstart,
-			(unsigned long) pa->pa_len);
+		ext4_msg(e4b->bd_sb, KERN_CRIT,
+			 "pa %p: logic %lu, phys. %lu, len %lu",
+			 pa, (unsigned long) pa->pa_lstart,
+			 (unsigned long) pa->pa_pstart,
+			 (unsigned long) pa->pa_len);
 		ext4_grp_locked_error(sb, group, 0, 0, "free %u, pa_free %u",
 					free, pa->pa_free);
 		/*
@@ -3798,7 +3801,8 @@ repeat:
 			 * use preallocation while we're discarding it */
 			spin_unlock(&pa->pa_lock);
 			spin_unlock(&ei->i_prealloc_lock);
-			printk(KERN_ERR "uh-oh! used pa while discarding\n");
+			ext4_msg(sb, KERN_ERR,
+				 "uh-oh! used pa while discarding");
 			WARN_ON(1);
 			schedule_timeout_uninterruptible(HZ);
 			goto repeat;
@@ -3875,12 +3879,13 @@ static void ext4_mb_show_ac(struct ext4_allocation_context *ac)
 	    (EXT4_SB(sb)->s_mount_flags & EXT4_MF_FS_ABORTED))
 		return;
 
-	printk(KERN_ERR "EXT4-fs: Can't allocate:"
-			" Allocation context details:\n");
-	printk(KERN_ERR "EXT4-fs: status %d flags %d\n",
+	ext4_msg(ac->ac_sb, KERN_ERR, "EXT4-fs: Can't allocate:"
+			" Allocation context details:");
+	ext4_msg(ac->ac_sb, KERN_ERR, "EXT4-fs: status %d flags %d",
 			ac->ac_status, ac->ac_flags);
-	printk(KERN_ERR "EXT4-fs: orig %lu/%lu/%lu@%lu, goal %lu/%lu/%lu@%lu, "
-			"best %lu/%lu/%lu@%lu cr %d\n",
+	ext4_msg(ac->ac_sb, KERN_ERR, "EXT4-fs: orig %lu/%lu/%lu@%lu, "
+		 	"goal %lu/%lu/%lu@%lu, "
+			"best %lu/%lu/%lu@%lu cr %d",
 			(unsigned long)ac->ac_o_ex.fe_group,
 			(unsigned long)ac->ac_o_ex.fe_start,
 			(unsigned long)ac->ac_o_ex.fe_len,
@@ -3894,9 +3899,9 @@ static void ext4_mb_show_ac(struct ext4_allocation_context *ac)
 			(unsigned long)ac->ac_b_ex.fe_len,
 			(unsigned long)ac->ac_b_ex.fe_logical,
 			(int)ac->ac_criteria);
-	printk(KERN_ERR "EXT4-fs: %lu scanned, %d found\n", ac->ac_ex_scanned,
-		ac->ac_found);
-	printk(KERN_ERR "EXT4-fs: groups: \n");
+	ext4_msg(ac->ac_sb, KERN_ERR, "EXT4-fs: %lu scanned, %d found",
+		 ac->ac_ex_scanned, ac->ac_found);
+	ext4_msg(ac->ac_sb, KERN_ERR, "EXT4-fs: groups: ");
 	ngroups = ext4_get_groups_count(sb);
 	for (i = 0; i < ngroups; i++) {
 		struct ext4_group_info *grp = ext4_get_group_info(sb, i);
-- 
1.7.4.1.22.gec8e1.dirty

Re: [PATCH 3/5] ext4: use ext4_msg() instead of printk in mballoc

[Andreas Dilger]

On 2011-08-01, at 7:13 AM, Theodore Ts'o <tytso@mit.edu> wrote:
> Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
> ---
> fs/ext4/mballoc.c |   79 ++++++++++++++++++++++++++++-------------------------
> 1 files changed, 42 insertions(+), 37 deletions(-)

Hi Ted, thanks for updating this patch set.  One minor issue below. 

> diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
> index d5021e8..262fa47 100644
> --- a/fs/ext4/mballoc.c
> +++ b/fs/ext4/mballoc.c
> @@ -493,10 +493,11 @@ static void mb_cmp_bitmaps(struct ext4_buddy *e4b, void *bitmap)
>        b2 = (unsigned char *) bitmap;
>        for (i = 0; i < e4b->bd_sb->s_blocksize; i++) {
>            if (b1[i] != b2[i]) {
> -                printk(KERN_ERR "corruption in group %u "
> -                       "at byte %u(%u): %x in copy != %x "
> -                       "on disk/prealloc\n",
> -                       e4b->bd_group, i, i * 8, b1[i], b2[i]);
> +                ext4_msg(e4b->bd_sb, KERN_ERR,
> +                     "corruption in group %u "
> +                     "at byte %u(%u): %x in copy != %x "
> +                     "on disk/prealloc",
> +                     e4b->bd_group, i, i * 8, b1[i], b2[i]);
>                BUG();
>            }
>        }
> @@ -2224,8 +2225,8 @@ int ext4_mb_add_groupinfo(struct super_block *sb, ext4_group_t group,
>            EXT4_DESC_PER_BLOCK_BITS(sb);
>        meta_group_info = kmalloc(metalen, GFP_KERNEL);
>        if (meta_group_info == NULL) {
> -            printk(KERN_ERR "EXT4-fs: can't allocate mem for a "
> -                   "buddy group\n");
> +            ext4_msg(sb, KERN_ERR, "EXT4-fs: can't allocate mem "
> +                 "for a buddy group");

I don't think ext4_msg() format strings should still have "Ext4-fs:" at the beginning. Isn't that the whole point of using ext4_msg()?

>            goto exit_meta_group_info;
>        }
>        sbi->s_group_info[group >> EXT4_DESC_PER_BLOCK_BITS(sb)] =
> @@ -2238,7 +2239,7 @@ int ext4_mb_add_groupinfo(struct super_block *sb, ext4_group_t group,
> 
>    meta_group_info[i] = kmem_cache_alloc(cachep, GFP_KERNEL);
>    if (meta_group_info[i] == NULL) {
> -        printk(KERN_ERR "EXT4-fs: can't allocate buddy mem\n");
> +        ext4_msg(sb, KERN_ERR, "EXT4-fs: can't allocate buddy mem");

Ditto. 

>        goto exit_group_info;
>    }
>    memset(meta_group_info[i], 0, kmem_cache_size(cachep));
> @@ -2333,12 +2334,12 @@ static int ext4_mb_init_backend(struct super_block *sb)
>     * So a two level scheme suffices for now. */
>    sbi->s_group_info = ext4_kvzalloc(array_size, GFP_KERNEL);
>    if (sbi->s_group_info == NULL) {
> -        printk(KERN_ERR "EXT4-fs: can't allocate buddy meta group\n");
> +        ext4_msg(sb, KERN_ERR, "can't allocate buddy meta group");
>        return -ENOMEM;
>    }
>    sbi->s_buddy_cache = new_inode(sb);
>    if (sbi->s_buddy_cache == NULL) {
> -        printk(KERN_ERR "EXT4-fs: can't get new inode\n");
> +        ext4_msg(sb, KERN_ERR, "can't get new inode");
>        goto err_freesgi;
>    }
>    sbi->s_buddy_cache->i_ino = get_next_ino();
> @@ -2346,8 +2347,7 @@ static int ext4_mb_init_backend(struct super_block *sb)
>    for (i = 0; i < ngroups; i++) {
>        desc = ext4_get_group_desc(sb, i, NULL);
>        if (desc == NULL) {
> -            printk(KERN_ERR
> -                "EXT4-fs: can't read descriptor %u\n", i);
> +            ext4_msg(sb, KERN_ERR, "can't read descriptor %u", i);
>            goto err_freebuddy;
>        }
>        if (ext4_mb_add_groupinfo(sb, i, desc) != 0)
> @@ -2411,7 +2411,8 @@ static int ext4_groupinfo_create_slab(size_t size)
> 
>    mutex_unlock(&ext4_grpinfo_slab_create_mutex);
>    if (!cachep) {
> -        printk(KERN_EMERG "EXT4: no memory for groupinfo slab cache\n");
> +        printk(KERN_EMERG,
> +               "EXT4-fs: no memory for groupinfo slab cache\n");

This one didn't actually get changed to ext4_msg(). 

>        return -ENOMEM;
>    }
> 
> @@ -2566,25 +2567,25 @@ int ext4_mb_release(struct super_block *sb)
>    if (sbi->s_buddy_cache)
>        iput(sbi->s_buddy_cache);
>    if (sbi->s_mb_stats) {
> -        printk(KERN_INFO
> -               "EXT4-fs: mballoc: %u blocks %u reqs (%u success)\n",
> +        ext4_msg(sb, KERN_INFO,
> +               "mballoc: %u blocks %u reqs (%u success)",
>                atomic_read(&sbi->s_bal_allocated),
>                atomic_read(&sbi->s_bal_reqs),
>                atomic_read(&sbi->s_bal_success));
> -        printk(KERN_INFO
> -              "EXT4-fs: mballoc: %u extents scanned, %u goal hits, "
> -                "%u 2^N hits, %u breaks, %u lost\n",
> +        ext4_msg(sb, KERN_INFO,
> +              "mballoc: %u extents scanned, %u goal hits, "
> +                "%u 2^N hits, %u breaks, %u lost",
>                atomic_read(&sbi->s_bal_ex_scanned),
>                atomic_read(&sbi->s_bal_goals),
>                atomic_read(&sbi->s_bal_2orders),
>                atomic_read(&sbi->s_bal_breaks),
>                atomic_read(&sbi->s_mb_lost_chunks));
> -        printk(KERN_INFO
> -               "EXT4-fs: mballoc: %lu generated and it took %Lu\n",
> +        ext4_msg(sb, KERN_INFO,
> +               "mballoc: %lu generated and it took %Lu",
>                sbi->s_mb_buddies_generated,
>                sbi->s_mb_generation_time);
> -        printk(KERN_INFO
> -               "EXT4-fs: mballoc: %u preallocated, %u discarded\n",
> +        ext4_msg(sb, KERN_INFO,
> +               "mballoc: %u preallocated, %u discarded",
>                atomic_read(&sbi->s_mb_preallocated),
>                atomic_read(&sbi->s_mb_discarded));
>    }
> @@ -3024,9 +3025,10 @@ ext4_mb_normalize_request(struct ext4_allocation_context *ac,
> 
>    if (start + size <= ac->ac_o_ex.fe_logical &&
>            start > ac->ac_o_ex.fe_logical) {
> -        printk(KERN_ERR "start %lu, size %lu, fe_logical %lu\n",
> -            (unsigned long) start, (unsigned long) size,
> -            (unsigned long) ac->ac_o_ex.fe_logical);
> +        ext4_msg(ac->ac_sb, KERN_ERR,
> +             "start %lu, size %lu, fe_logical %lu",
> +             (unsigned long) start, (unsigned long) size,
> +             (unsigned long) ac->ac_o_ex.fe_logical);
>    }
>    BUG_ON(start + size <= ac->ac_o_ex.fe_logical &&
>            start > ac->ac_o_ex.fe_logical);
> @@ -3607,10 +3609,11 @@ ext4_mb_release_inode_pa(struct ext4_buddy *e4b, struct buffer_head *bitmap_bh,
>        bit = next + 1;
>    }
>    if (free != pa->pa_free) {
> -        printk(KERN_CRIT "pa %p: logic %lu, phys. %lu, len %lu\n",
> -            pa, (unsigned long) pa->pa_lstart,
> -            (unsigned long) pa->pa_pstart,
> -            (unsigned long) pa->pa_len);
> +        ext4_msg(e4b->bd_sb, KERN_CRIT,
> +             "pa %p: logic %lu, phys. %lu, len %lu",
> +             pa, (unsigned long) pa->pa_lstart,
> +             (unsigned long) pa->pa_pstart,
> +             (unsigned long) pa->pa_len);
>        ext4_grp_locked_error(sb, group, 0, 0, "free %u, pa_free %u",
>                    free, pa->pa_free);
>        /*
> @@ -3798,7 +3801,8 @@ repeat:
>             * use preallocation while we're discarding it */
>            spin_unlock(&pa->pa_lock);
>            spin_unlock(&ei->i_prealloc_lock);
> -            printk(KERN_ERR "uh-oh! used pa while discarding\n");
> +            ext4_msg(sb, KERN_ERR,
> +                 "uh-oh! used pa while discarding");
>            WARN_ON(1);
>            schedule_timeout_uninterruptible(HZ);
>            goto repeat;
> @@ -3875,12 +3879,13 @@ static void ext4_mb_show_ac(struct ext4_allocation_context *ac)
>        (EXT4_SB(sb)->s_mount_flags & EXT4_MF_FS_ABORTED))
>        return;
> 
> -    printk(KERN_ERR "EXT4-fs: Can't allocate:"
> -            " Allocation context details:\n");
> -    printk(KERN_ERR "EXT4-fs: status %d flags %d\n",
> +    ext4_msg(ac->ac_sb, KERN_ERR, "EXT4-fs: Can't allocate:"

Ditto. 

> +            " Allocation context details:");
> +    ext4_msg(ac->ac_sb, KERN_ERR, "EXT4-fs: status %d flags %d",
>            ac->ac_status, ac->ac_flags);
> -    printk(KERN_ERR "EXT4-fs: orig %lu/%lu/%lu@%lu, goal %lu/%lu/%lu@%lu, "
> -            "best %lu/%lu/%lu@%lu cr %d\n",
> +    ext4_msg(ac->ac_sb, KERN_ERR, "EXT4-fs: orig %lu/%lu/%lu@%lu, "

Here too. 

> +            "goal %lu/%lu/%lu@%lu, "
> +            "best %lu/%lu/%lu@%lu cr %d",
>            (unsigned long)ac->ac_o_ex.fe_group,
>            (unsigned long)ac->ac_o_ex.fe_start,
>            (unsigned long)ac->ac_o_ex.fe_len,
> @@ -3894,9 +3899,9 @@ static void ext4_mb_show_ac(struct ext4_allocation_context *ac)
>            (unsigned long)ac->ac_b_ex.fe_len,
>            (unsigned long)ac->ac_b_ex.fe_logical,
>            (int)ac->ac_criteria);
> -    printk(KERN_ERR "EXT4-fs: %lu scanned, %d found\n", ac->ac_ex_scanned,
> -        ac->ac_found);
> -    printk(KERN_ERR "EXT4-fs: groups: \n");
> +    ext4_msg(ac->ac_sb, KERN_ERR, "EXT4-fs: %lu scanned, %d found",

...

> +         ac->ac_ex_scanned, ac->ac_found);
> +    ext4_msg(ac->ac_sb, KERN_ERR, "EXT4-fs: groups: ");
>    ngroups = ext4_get_groups_count(sb);
>    for (i = 0; i < ngroups; i++) {
>        struct ext4_group_info *grp = ext4_get_group_info(sb, i);
> -- 
> 1.7.4.1.22.gec8e1.dirty
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 4/5] ext4: use EXT4_BAD_INO for buddy cache to avoid colliding with valid inode #

[Theodore Ts'o]

From: Yu Jian <yujian@whamcloud.com>

Signed-off-by: Yu Jian <yujian@whamcloud.com>
Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
 fs/ext4/mballoc.c |    6 +++++-
 1 files changed, 5 insertions(+), 1 deletions(-)

diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
index 262fa47..8beccd8 100644
--- a/fs/ext4/mballoc.c
+++ b/fs/ext4/mballoc.c
@@ -2342,7 +2342,11 @@ static int ext4_mb_init_backend(struct super_block *sb)
 		ext4_msg(sb, KERN_ERR, "can't get new inode");
 		goto err_freesgi;
 	}
-	sbi->s_buddy_cache->i_ino = get_next_ino();
+	/* To avoid potentially colliding with an valid on-disk inode number,
+	 * use EXT4_BAD_INO for the buddy cache inode number.  This inode is
+	 * not in the inode hash, so it should never be found by iget(), but
+	 * this will avoid confusion if it ever shows up during debugging. */
+	sbi->s_buddy_cache->i_ino = EXT4_BAD_INO;
 	EXT4_I(sbi->s_buddy_cache)->i_disksize = 0;
 	for (i = 0; i < ngroups; i++) {
 		desc = ext4_get_group_desc(sb, i, NULL);
-- 
1.7.4.1.22.gec8e1.dirty

[PATCH 5/5] ext4: prevent memory leaks from ext4_mb_init_backend() on error path

[Theodore Ts'o]

From: Yu Jian <yujian@whamcloud.com>

In ext4_mb_init(), if the s_locality_group allocation fails it will
currently cause the allocations made in ext4_mb_init_backend() to
be leaked.  Moving the ext4_mb_init_backend() allocation after the
s_locality_group allocation avoids that problem.

Signed-off-by: Yu Jian <yujian@whamcloud.com>
Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
 fs/ext4/mballoc.c |   12 ++++++------
 1 files changed, 6 insertions(+), 6 deletions(-)

diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
index 8beccd8..c6c71af 100644
--- a/fs/ext4/mballoc.c
+++ b/fs/ext4/mballoc.c
@@ -2465,12 +2465,6 @@ int ext4_mb_init(struct super_block *sb, int needs_recovery)
 		i++;
 	} while (i <= sb->s_blocksize_bits + 1);
 
-	/* init file for buddy data */
-	ret = ext4_mb_init_backend(sb);
-	if (ret != 0) {
-		goto out;
-	}