[PATCH] Btrfs: fix extent_from_logical to deal with skinny metadata

[PATCH] Btrfs: fix extent_from_logical to deal with skinny metadata

[Josef Bacik]

I don't think this is an issue and I've not seen it in practice but
extent_from_logical will fail to find a skinny extent because it uses
btrfs_previous_item and gives it the normal extent item type.  This is just not
a place to use btrfs_previous_item since we care about either normal extents or
skinny extents, so open code btrfs_previous_item to properly check.  This would
only affect metadata and the only place this is used for metadata is scrub and
I'm pretty sure it's just for printing stuff out, not actually doing any work so
hopefully it was never a problem other than a cosmetic one.  Thanks,

Signed-off-by: Josef Bacik <jbacik@fb.com>
---
 fs/btrfs/backref.c | 41 +++++++++++++++++++++++++++++++++--------
 1 file changed, 33 insertions(+), 8 deletions(-)

diff --git a/fs/btrfs/backref.c b/fs/btrfs/backref.c
index 34a8952..dcf2448 100644
--- a/fs/btrfs/backref.c
+++ b/fs/btrfs/backref.c
@@ -1302,20 +1302,45 @@ int extent_from_logical(struct btrfs_fs_info *fs_info, u64 logical,
 	ret = btrfs_search_slot(NULL, fs_info->extent_root, &key, path, 0, 0);
 	if (ret < 0)
 		return ret;
-	ret = btrfs_previous_item(fs_info->extent_root, path,
-					0, BTRFS_EXTENT_ITEM_KEY);
-	if (ret < 0)
-		return ret;
 
-	btrfs_item_key_to_cpu(path->nodes[0], found_key, path->slots[0]);
+	while (1) {
+		u32 nritems;
+		if (path->slots[0] == 0) {
+			btrfs_set_path_blocking(path);
+			ret = btrfs_prev_leaf(fs_info->extent_root, path);
+			if (ret != 0) {
+				if (ret > 0) {
+					pr_debug("logical %llu is not within "
+						 "any extent\n", logical);
+					ret = -ENOENT;
+				}
+				return ret;
+			}
+		} else {
+			path->slots[0]--;
+		}
+		nritems = btrfs_header_nritems(path->nodes[0]);
+		if (nritems == 0) {
+			pr_debug("logical %llu is not within any extent\n",
+				 logical);
+			return -ENOENT;
+		}
+		if (path->slots[0] == nritems)
+			path->slots[0]--;
+
+		btrfs_item_key_to_cpu(path->nodes[0], found_key,
+				      path->slots[0]);
+		if (found_key->type == BTRFS_EXTENT_ITEM_KEY ||
+		    found_key->type == BTRFS_METADATA_ITEM_KEY)
+			break;
+	}
+
 	if (found_key->type == BTRFS_METADATA_ITEM_KEY)
 		size = fs_info->extent_root->leafsize;
 	else if (found_key->type == BTRFS_EXTENT_ITEM_KEY)
 		size = found_key->offset;
 
-	if ((found_key->type != BTRFS_EXTENT_ITEM_KEY &&
-	     found_key->type != BTRFS_METADATA_ITEM_KEY) ||
-	    found_key->objectid > logical ||
+	if (found_key->objectid > logical ||
 	    found_key->objectid + size <= logical) {
 		pr_debug("logical %llu is not within any extent\n", logical);
 		return -ENOENT;
-- 
1.8.3.1

Re: [PATCH] Btrfs: fix extent_from_logical to deal with skinny metadata

[Wang Shilong]

Hello Josef,

I have sent a patch before to address this issue:
       https://patchwork.kernel.org/patch/3472211/

I introduced another function btrfs_previous_extent_item() to
search previous extent item, I think you must miss my path before
giving this patch ^_^.

Thanks,
Wang

On 01/24/2014 05:06 AM, Josef Bacik wrote:
> I don't think this is an issue and I've not seen it in practice but
> extent_from_logical will fail to find a skinny extent because it uses
> btrfs_previous_item and gives it the normal extent item type.  This is just not
> a place to use btrfs_previous_item since we care about either normal extents or
> skinny extents, so open code btrfs_previous_item to properly check.  This would
> only affect metadata and the only place this is used for metadata is scrub and
> I'm pretty sure it's just for printing stuff out, not actually doing any work so
> hopefully it was never a problem other than a cosmetic one.  Thanks,
>
> Signed-off-by: Josef Bacik <jbacik@fb.com>
> ---
>   fs/btrfs/backref.c | 41 +++++++++++++++++++++++++++++++++--------
>   1 file changed, 33 insertions(+), 8 deletions(-)
>
> diff --git a/fs/btrfs/backref.c b/fs/btrfs/backref.c
> index 34a8952..dcf2448 100644
> --- a/fs/btrfs/backref.c
> +++ b/fs/btrfs/backref.c
> @@ -1302,20 +1302,45 @@ int extent_from_logical(struct btrfs_fs_info *fs_info, u64 logical,
>   	ret = btrfs_search_slot(NULL, fs_info->extent_root, &key, path, 0, 0);
>   	if (ret < 0)
>   		return ret;
> -	ret = btrfs_previous_item(fs_info->extent_root, path,
> -					0, BTRFS_EXTENT_ITEM_KEY);
> -	if (ret < 0)
> -		return ret;
>   
> -	btrfs_item_key_to_cpu(path->nodes[0], found_key, path->slots[0]);
> +	while (1) {
> +		u32 nritems;
> +		if (path->slots[0] == 0) {
> +			btrfs_set_path_blocking(path);
> +			ret = btrfs_prev_leaf(fs_info->extent_root, path);
> +			if (ret != 0) {
> +				if (ret > 0) {
> +					pr_debug("logical %llu is not within "
> +						 "any extent\n", logical);
> +					ret = -ENOENT;
> +				}
> +				return ret;
> +			}
> +		} else {
> +			path->slots[0]--;
> +		}
> +		nritems = btrfs_header_nritems(path->nodes[0]);
> +		if (nritems == 0) {
> +			pr_debug("logical %llu is not within any extent\n",
> +				 logical);
> +			return -ENOENT;
> +		}
> +		if (path->slots[0] == nritems)
> +			path->slots[0]--;
> +
> +		btrfs_item_key_to_cpu(path->nodes[0], found_key,
> +				      path->slots[0]);
> +		if (found_key->type == BTRFS_EXTENT_ITEM_KEY ||
> +		    found_key->type == BTRFS_METADATA_ITEM_KEY)
> +			break;
> +	}
> +
>   	if (found_key->type == BTRFS_METADATA_ITEM_KEY)
>   		size = fs_info->extent_root->leafsize;
>   	else if (found_key->type == BTRFS_EXTENT_ITEM_KEY)
>   		size = found_key->offset;
>   
> -	if ((found_key->type != BTRFS_EXTENT_ITEM_KEY &&
> -	     found_key->type != BTRFS_METADATA_ITEM_KEY) ||
> -	    found_key->objectid > logical ||
> +	if (found_key->objectid > logical ||
>   	    found_key->objectid + size <= logical) {
>   		pr_debug("logical %llu is not within any extent\n", logical);
>   		return -ENOENT;

Re: [PATCH] Btrfs: fix extent_from_logical to deal with skinny metadata

[Josef Bacik]

On 01/23/2014 08:07 PM, Wang Shilong wrote:
> Hello Josef,
>
> I have sent a patch before to address this issue:
>       https://patchwork.kernel.org/patch/3472211/
>
> I introduced another function btrfs_previous_extent_item() to
> search previous extent item, I think you must miss my path before
> giving this patch ^_^.

I sure did, we'll take yours since it was first.  Thanks,

Josef