From: Goldwyn Rodrigues <rgoldwyn@suse.de>
To: Trond Myklebust <trond.myklebust@primarydata.com>,
Miklos Szeredi <miklos@szeredi.hu>
Cc: "linux-unionfs@vger.kernel.org" <linux-unionfs@vger.kernel.org>,
Linux FS-devel Mailing List <linux-fsdevel@vger.kernel.org>,
Linux NFS Mailing List <linux-nfs@vger.kernel.org>,
David Howells <dhowells@redhat.com>,
Al Viro <viro@zeniv.linux.org.uk>
Subject: Re: [PATCH 3/3] nfs: Store and use inode in nfs_open_context
Date: Fri, 4 Mar 2016 08:52:26 -0600 [thread overview]
Message-ID: <56D9A12A.8050905@suse.de> (raw)
In-Reply-To: <CAHQdGtRzdnGUpvtBnkjWd4NsjAoSTCQULnwXT3mXk+pffFs72g@mail.gmail.com>
On 03/04/2016 07:41 AM, Trond Myklebust wrote:
> On Fri, Mar 4, 2016 at 5:17 AM, Miklos Szeredi <miklos@szeredi.hu> wrote:
>> On Thu, Mar 03, 2016 at 09:16:33AM +0100, Miklos Szeredi wrote:
>>> My plan was to introduce a file_dentry() helper that MUST be used by
>>> filesystems to get the dentry from the file and that makes sure it's
>>> the right one (check against file_inode()). If not, then we could
>>> call into overlayfs to return the right one, similar to
>>> ->d_select_inode(), except we want to have a dentry and we want to
>>> have *a particular dentry* matching file_inode() (the file could have
>>> been copied up in the mean time).
>>
>> Something like the following (untested, against overlayfs-next.git)
>>
>> Thanks,
>> Miklos
>>
>> ---
>> fs/nfs/dir.c | 6 +++---
>> fs/nfs/inode.c | 2 +-
>> fs/nfs/nfs4file.c | 4 ++--
>> fs/open.c | 11 +++++++++++
>> fs/overlayfs/super.c | 16 ++++++++++++++++
>> include/linux/dcache.h | 1 +
>> include/linux/fs.h | 2 ++
>> 7 files changed, 36 insertions(+), 6 deletions(-)
>>
>> --- a/fs/nfs/dir.c
>> +++ b/fs/nfs/dir.c
>> @@ -377,7 +377,7 @@ int nfs_readdir_xdr_filler(struct page *
>> again:
>> timestamp = jiffies;
>> gencount = nfs_inc_attr_generation_counter();
>> - error = NFS_PROTO(inode)->readdir(file->f_path.dentry, cred, entry->cookie, pages,
>> + error = NFS_PROTO(inode)->readdir(file_dentry(file), cred, entry->cookie, pages,
>> NFS_SERVER(inode)->dtsize, desc->plus);
>> if (error < 0) {
>> /* We requested READDIRPLUS, but the server doesn't grok it */
>> @@ -560,7 +560,7 @@ int nfs_readdir_page_filler(nfs_readdir_
>> count++;
>>
>> if (desc->plus != 0)
>> - nfs_prime_dcache(desc->file->f_path.dentry, entry);
>> + nfs_prime_dcache(file_dentry(desc->file), entry);
>>
>> status = nfs_readdir_add_to_array(entry, page);
>> if (status != 0)
>> @@ -864,7 +864,7 @@ static bool nfs_dir_mapping_need_revalid
>> */
>> static int nfs_readdir(struct file *file, struct dir_context *ctx)
>> {
>> - struct dentry *dentry = file->f_path.dentry;
>> + struct dentry *dentry = file_dentry(file);
>> struct inode *inode = d_inode(dentry);
>> nfs_readdir_descriptor_t my_desc,
>> *desc = &my_desc;
>> --- a/fs/nfs/inode.c
>> +++ b/fs/nfs/inode.c
>> @@ -940,7 +940,7 @@ int nfs_open(struct inode *inode, struct
>> {
>> struct nfs_open_context *ctx;
>>
>> - ctx = alloc_nfs_open_context(filp->f_path.dentry, filp->f_mode);
>> + ctx = alloc_nfs_open_context(file_dentry(filp), filp->f_mode);
>> if (IS_ERR(ctx))
>> return PTR_ERR(ctx);
>> nfs_file_set_open_context(filp, ctx);
>> --- a/fs/nfs/nfs4file.c
>> +++ b/fs/nfs/nfs4file.c
>> @@ -26,7 +26,7 @@ static int
>> nfs4_file_open(struct inode *inode, struct file *filp)
>> {
>> struct nfs_open_context *ctx;
>> - struct dentry *dentry = filp->f_path.dentry;
>> + struct dentry *dentry = file_dentry(filp);
>> struct dentry *parent = NULL;
>> struct inode *dir;
>> unsigned openflags = filp->f_flags;
>> @@ -57,7 +57,7 @@ nfs4_file_open(struct inode *inode, stru
>> parent = dget_parent(dentry);
>> dir = d_inode(parent);
>>
>> - ctx = alloc_nfs_open_context(filp->f_path.dentry, filp->f_mode);
>> + ctx = alloc_nfs_open_context(file_dentry(filp), filp->f_mode);
>> err = PTR_ERR(ctx);
>> if (IS_ERR(ctx))
>> goto out;
>> --- a/include/linux/fs.h
>> +++ b/include/linux/fs.h
>> @@ -1234,6 +1234,8 @@ static inline struct inode *file_inode(c
>> return f->f_inode;
>> }
>>
>> +extern struct dentry *file_dentry(const struct file *file);
>> +
>> static inline int locks_lock_file_wait(struct file *filp, struct file_lock *fl)
>> {
>> return locks_lock_inode_wait(file_inode(filp), fl);
>> --- a/fs/open.c
>> +++ b/fs/open.c
>> @@ -831,6 +831,17 @@ char *file_path(struct file *filp, char
>> }
>> EXPORT_SYMBOL(file_path);
>>
>> +struct dentry *file_dentry(const struct file *file)
>> +{
>> + struct dentry *dentry = file->f_path.dentry;
>> +
>> + if (d_inode(dentry) == file_inode(file))
>> + return dentry;
>> + else
>> + return dentry->d_op->d_select_dentry(dentry, file_inode(file));
>> +}
>> +EXPORT_SYMBOL(file_dentry);
>> +
>> /**
>> * vfs_open - open the file at the given path
>> * @path: path to open
>> --- a/fs/overlayfs/super.c
>> +++ b/fs/overlayfs/super.c
>> @@ -336,14 +336,30 @@ static int ovl_dentry_weak_revalidate(st
>> return ret;
>> }
>>
>> +static struct dentry *ovl_d_select_dentry(struct dentry *dentry,
>> + struct inode *inode)
>> +{
>> + struct ovl_entry *oe = dentry->d_fsdata;
>> + struct dentry *realentry = ovl_upperdentry_dereference(oe);
>> +
>> + if (realentry && inode == d_inode(realentry))
>> + return realentry;
>> + realentry = __ovl_dentry_lower(oe);
>> + if (realentry && inode == d_inode(realentry))
>> + return realentry;
>> + BUG();
>> +}
>> +
>> static const struct dentry_operations ovl_dentry_operations = {
>> .d_release = ovl_dentry_release,
>> .d_select_inode = ovl_d_select_inode,
>> + .d_select_dentry = ovl_d_select_dentry,
>> };
>>
>> static const struct dentry_operations ovl_reval_dentry_operations = {
>> .d_release = ovl_dentry_release,
>> .d_select_inode = ovl_d_select_inode,
>> + .d_select_dentry = ovl_d_select_dentry,
>> .d_revalidate = ovl_dentry_revalidate,
>> .d_weak_revalidate = ovl_dentry_weak_revalidate,
>> };
>> --- a/include/linux/dcache.h
>> +++ b/include/linux/dcache.h
>> @@ -161,6 +161,7 @@ struct dentry_operations {
>> struct vfsmount *(*d_automount)(struct path *);
>> int (*d_manage)(struct dentry *, bool);
>> struct inode *(*d_select_inode)(struct dentry *, unsigned);
>> + struct dentry *(*d_select_dentry)(struct dentry *, struct inode *);
>> } ____cacheline_aligned;
>>
>> /*
>
> Thanks Miklos! That looks reasonable, assuming it tests out correctly.
>
Tested and it looks good. Thanks Miklos.
Tested-by: Goldwyn Rodrigues <rgoldwyn@suse.com>
--
Goldwyn
prev parent reply other threads:[~2016-03-04 14:52 UTC|newest]
Thread overview: 14+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-03-01 18:12 [PATCH 0/3] Fix overlayfs with NFS as lowerdir Goldwyn Rodrigues
2016-03-01 18:12 ` [PATCH 1/3] ovl: Add d_select_inode to reval dentry operations Goldwyn Rodrigues
2016-03-01 18:12 ` [PATCH 2/3] vfs: Add d_select_inode for overlayfs translation Goldwyn Rodrigues
2016-03-01 18:12 ` [PATCH 3/3] nfs: Store and use inode in nfs_open_context Goldwyn Rodrigues
2016-03-01 20:46 ` Trond Myklebust
2016-03-02 14:27 ` Goldwyn Rodrigues
2016-03-02 14:31 ` Trond Myklebust
2016-03-02 14:38 ` Trond Myklebust
2016-03-02 14:43 ` Miklos Szeredi
2016-03-02 15:57 ` Goldwyn Rodrigues
2016-03-03 8:16 ` Miklos Szeredi
2016-03-04 10:17 ` Miklos Szeredi
2016-03-04 13:41 ` Trond Myklebust
2016-03-04 14:52 ` Goldwyn Rodrigues [this message]
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=56D9A12A.8050905@suse.de \
--to=rgoldwyn@suse.de \
--cc=dhowells@redhat.com \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-nfs@vger.kernel.org \
--cc=linux-unionfs@vger.kernel.org \
--cc=miklos@szeredi.hu \
--cc=trond.myklebust@primarydata.com \
--cc=viro@zeniv.linux.org.uk \
/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).