linux-btrfs.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Miao Xie <miaox@cn.fujitsu.com>
To: Zach Brown <zab@redhat.com>
Cc: linux-btrfs@vger.kernel.org
Subject: Re: [PATCH 1/6] btrfs: set readdir f_pos only after filldir
Date: Wed, 05 Jun 2013 09:19:28 +0800	[thread overview]
Message-ID: <51AE9220.6070102@cn.fujitsu.com> (raw)
In-Reply-To: <1370384280-28652-2-git-send-email-zab@redhat.com>

On 	tue, 4 Jun 2013 15:17:55 -0700, Zach Brown wrote:
> The only time we need to advance f_pos is after we've successfully given
> a result to userspace via filldir.  This simplification gets rid of the
> is_curr variable used to update f_pos for the delayed item readdir
> entries.
> 
> Signed-off-by: Zach Brown <zab@redhat.com>

Reviewed-by: Miao Xie <miaox@cn.fujitsu.com>

> ---
>  fs/btrfs/delayed-inode.c |  5 +++--
>  fs/btrfs/inode.c         | 17 +++++++----------
>  2 files changed, 10 insertions(+), 12 deletions(-)
> 
> diff --git a/fs/btrfs/delayed-inode.c b/fs/btrfs/delayed-inode.c
> index 5615eac..4d846a2 100644
> --- a/fs/btrfs/delayed-inode.c
> +++ b/fs/btrfs/delayed-inode.c
> @@ -1696,8 +1696,6 @@ int btrfs_readdir_delayed_dir_index(struct file *filp, void *dirent,
>  			continue;
>  		}
>  
> -		filp->f_pos = curr->key.offset;
> -
>  		di = (struct btrfs_dir_item *)curr->data;
>  		name = (char *)(di + 1);
>  		name_len = le16_to_cpu(di->name_len);
> @@ -1708,6 +1706,9 @@ int btrfs_readdir_delayed_dir_index(struct file *filp, void *dirent,
>  		over = filldir(dirent, name, name_len, curr->key.offset,
>  			       location.objectid, d_type);
>  
> +		if (!over)
> +			filp->f_pos = curr->key.offset + 1;
> +
>  		if (atomic_dec_and_test(&curr->refs))
>  			kfree(curr);
>  
> diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
> index b11a95e..6a5784b 100644
> --- a/fs/btrfs/inode.c
> +++ b/fs/btrfs/inode.c
> @@ -5007,7 +5007,6 @@ static int btrfs_real_readdir(struct file *filp, void *dirent,
>  	char tmp_name[32];
>  	char *name_ptr;
>  	int name_len;
> -	int is_curr = 0;	/* filp->f_pos points to the current index? */
>  
>  	/* FIXME, use a real flag for deciding about the key type */
>  	if (root->fs_info->tree_root == root)
> @@ -5076,9 +5075,6 @@ static int btrfs_real_readdir(struct file *filp, void *dirent,
>  						  found_key.offset))
>  			goto next;
>  
> -		filp->f_pos = found_key.offset;
> -		is_curr = 1;
> -
>  		di = btrfs_item_ptr(leaf, slot, struct btrfs_dir_item);
>  		di_cur = 0;
>  		di_total = btrfs_item_size(leaf, item);
> @@ -5130,6 +5126,9 @@ skip:
>  
>  			if (over)
>  				goto nopos;
> +
> +			filp->f_pos = found_key.offset + 1;
> +
>  			di_len = btrfs_dir_name_len(leaf, di) +
>  				 btrfs_dir_data_len(leaf, di) + sizeof(*di);
>  			di_cur += di_len;
> @@ -5140,23 +5139,21 @@ next:
>  	}
>  
>  	if (key_type == BTRFS_DIR_INDEX_KEY) {
> -		if (is_curr)
> -			filp->f_pos++;
>  		ret = btrfs_readdir_delayed_dir_index(filp, dirent, filldir,
>  						      &ins_list);
>  		if (ret)
>  			goto nopos;
>  	}
>  
> -	/* Reached end of directory/root. Bump pos past the last item. */
> -	if (key_type == BTRFS_DIR_INDEX_KEY)
> +	/* Reached end of directory/root */
> +	if (key_type == BTRFS_DIR_INDEX_KEY) {
>  		/*
>  		 * 32-bit glibc will use getdents64, but then strtol -
>  		 * so the last number we can serve is this.
>  		 */
>  		filp->f_pos = 0x7fffffff;
> -	else
> -		filp->f_pos++;
> +	}
> +
>  nopos:
>  	ret = 0;
>  err:
> 


  reply	other threads:[~2013-06-05  1:18 UTC|newest]

Thread overview: 22+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-06-04 22:17 [PATCH 0/6] fix INT_MAX readdir hang, plus cleanups Zach Brown
2013-06-04 22:17 ` [PATCH 1/6] btrfs: set readdir f_pos only after filldir Zach Brown
2013-06-05  1:19   ` Miao Xie [this message]
2013-06-04 22:17 ` [PATCH 2/6] btrfs: fix readdir hang with offsets past INT_MAX Zach Brown
2013-06-04 22:17 ` [PATCH 3/6] btrfs: trivial delayed item readdir list cleanups Zach Brown
2013-06-04 22:17 ` [PATCH 4/6] btrfs: simplify finding next/prev delayed items Zach Brown
2013-06-04 22:17 ` [PATCH 5/6] btrfs: add helper to get delayed item root Zach Brown
2013-06-04 22:18 ` [PATCH 6/6] btrfs: get fewer delayed item refs during readdir Zach Brown
2013-06-04 23:16 ` [PATCH 0/6] fix INT_MAX readdir hang, plus cleanups Chris Mason
2013-06-04 23:26   ` Zach Brown
2013-06-05  2:34     ` Miao Xie
2013-06-05 13:36       ` David Sterba
2013-06-06  1:35         ` Miao Xie
2013-06-06 13:55           ` David Sterba
2013-06-06 14:32             ` Chris Mason
2013-06-10 22:39     ` Zach Brown
2013-06-12 12:59       ` Chris Mason
2013-07-01 12:54 ` Josef Bacik
2013-07-01 13:18   ` Chris Mason
2013-07-01 16:10   ` Zach Brown
2013-07-01 17:18     ` Chris Mason
2013-07-11 23:19   ` Zach Brown

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=51AE9220.6070102@cn.fujitsu.com \
    --to=miaox@cn.fujitsu.com \
    --cc=linux-btrfs@vger.kernel.org \
    --cc=zab@redhat.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).