xfs
[Top] [All Lists]

Re: [PATCH 6/9] xfs: use vfs inode nlink field everywhere

To: Dave Chinner <david@xxxxxxxxxxxxx>
Subject: Re: [PATCH 6/9] xfs: use vfs inode nlink field everywhere
From: Christoph Hellwig <hch@xxxxxxxxxxxxx>
Date: Mon, 8 Feb 2016 01:40:21 -0800
Cc: xfs@xxxxxxxxxxx
Delivered-to: xfs@xxxxxxxxxxx
In-reply-to: <1454905461-2773-7-git-send-email-david@xxxxxxxxxxxxx>
References: <1454905461-2773-1-git-send-email-david@xxxxxxxxxxxxx> <1454905461-2773-7-git-send-email-david@xxxxxxxxxxxxx>
User-agent: Mutt/1.5.24 (2015-08-30)
On Mon, Feb 08, 2016 at 03:24:18PM +1100, Dave Chinner wrote:
> From: Dave Chinner <dchinner@xxxxxxxxxx>
> 
> The Vfs tracks the inodenlink just like the xfs_icdinode. We can

s/Vfs/VFS/

> +STATIC int
>  xfs_iunlink(
> -     xfs_trans_t     *tp,
> -     xfs_inode_t     *ip)
> +     struct xfs_trans *tp,
> +     struct xfs_inode *ip,
> +     bool            ignore_linkcount)

I'd say kill the ignore_linkcount parameter and the assert depending on it.
There are just two callers: xfs_droplink checks for i_nlink just before
the call, and xfs_create_tmpfile has to set the override flag.

Otherwise looks fine:

Reviewed-by: Christoph Hellwig <hch@xxxxxx>

<Prev in Thread] Current Thread [Next in Thread>