Received-SPF: pass (google.com: domain of linux-nfs-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <d33278347061e7276a4aef427687d46cc4f69736.camel@kernel.org>
Subject: Re: [PATCH v7 03/14] NFSD: Add an NFSD_FILE_GC flag to enable
 nfsd_file garbage collection
From:   Jeff Layton <jlayton@kernel.org>
To:     Chuck Lever <chuck.lever@oracle.com>, linux-nfs@vger.kernel.org
Cc:     neilb@suse.de
Date:   Mon, 31 Oct 2022 12:32:05 -0400
In-Reply-To: <166696841100.106044.8347912040040290155.stgit@klimt.1015granger.net>
References: <166696812922.106044.679812521105874329.stgit@klimt.1015granger.net>
         <166696841100.106044.8347912040040290155.stgit@klimt.1015granger.net>
Content-Type: text/plain; charset="ISO-8859-15"
Content-Transfer-Encoding: quoted-printable
User-Agent: Evolution 3.44.4 (3.44.4-2.fc36) 
MIME-Version: 1.0
Precedence: bulk

On Fri, 2022-10-28 at 10:46 -0400, Chuck Lever wrote:
> NFSv4 operations manage the lifetime of nfsd_file items they use by
> means of NFSv4 OPEN and CLOSE. Hence there's no need for them to be
> garbage collected.
>=20
> Introduce a mechanism to enable garbage collection for nfsd_file
> items used only by NFSv2/3 callers.
>=20
> Note that the change in nfsd_file_put() ensures that both CLOSE and
> DELEGRETURN will actually close out and free an nfsd_file on last
> reference of a non-garbage-collected file.
>=20
> Link: https://bugzilla.linux-nfs.org/show_bug.cgi?id=3D394
> Suggested-by: Trond Myklebust <trond.myklebust@hammerspace.com>
> Tested-by: Jeff Layton <jlayton@kernel.org>
> Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
> Reviewed-by: NeilBrown <neilb@suse.de>
> ---
>  fs/nfsd/filecache.c |   63 +++++++++++++++++++++++++++++++++++++++++++++=
------
>  fs/nfsd/filecache.h |    3 ++
>  fs/nfsd/nfs3proc.c  |    4 ++-
>  fs/nfsd/trace.h     |    3 ++
>  fs/nfsd/vfs.c       |    4 ++-
>  5 files changed, 64 insertions(+), 13 deletions(-)
>=20
> diff --git a/fs/nfsd/filecache.c b/fs/nfsd/filecache.c
> index beb41a507623..965c7b13fddc 100644
> --- a/fs/nfsd/filecache.c
> +++ b/fs/nfsd/filecache.c
> @@ -63,6 +63,7 @@ struct nfsd_file_lookup_key {
>  	struct net			*net;
>  	const struct cred		*cred;
>  	unsigned char			need;
> +	bool				gc;
>  	enum nfsd_file_lookup_type	type;
>  };
> =20
> @@ -162,6 +163,8 @@ static int nfsd_file_obj_cmpfn(struct rhashtable_comp=
are_arg *arg,
>  			return 1;
>  		if (!nfsd_match_cred(nf->nf_cred, key->cred))
>  			return 1;
> +		if (!!test_bit(NFSD_FILE_GC, &nf->nf_flags) !=3D key->gc)
> +			return 1;
>  		if (test_bit(NFSD_FILE_HASHED, &nf->nf_flags) =3D=3D 0)
>  			return 1;
>  		break;
> @@ -297,6 +300,8 @@ nfsd_file_alloc(struct nfsd_file_lookup_key *key, uns=
igned int may)
>  		nf->nf_flags =3D 0;
>  		__set_bit(NFSD_FILE_HASHED, &nf->nf_flags);
>  		__set_bit(NFSD_FILE_PENDING, &nf->nf_flags);
> +		if (key->gc)
> +			__set_bit(NFSD_FILE_GC, &nf->nf_flags);
>  		nf->nf_inode =3D key->inode;
>  		/* nf_ref is pre-incremented for hash table */
>  		refcount_set(&nf->nf_ref, 2);
> @@ -428,16 +433,27 @@ nfsd_file_put_noref(struct nfsd_file *nf)
>  	}
>  }
> =20
> +static void
> +nfsd_file_unhash_and_put(struct nfsd_file *nf)
> +{
> +	if (nfsd_file_unhash(nf))
> +		nfsd_file_put_noref(nf);
> +}
> +
>  void
>  nfsd_file_put(struct nfsd_file *nf)
>  {
>  	might_sleep();
> =20
> -	nfsd_file_lru_add(nf);
> -	if (test_bit(NFSD_FILE_HASHED, &nf->nf_flags) =3D=3D 0) {
> +	if (test_bit(NFSD_FILE_GC, &nf->nf_flags))
> +		nfsd_file_lru_add(nf);
> +	else if (refcount_read(&nf->nf_ref) =3D=3D 2)
> +		nfsd_file_unhash_and_put(nf);
> +
> +	if (!test_bit(NFSD_FILE_HASHED, &nf->nf_flags)) {
>  		nfsd_file_flush(nf);
>  		nfsd_file_put_noref(nf);
> -	} else if (nf->nf_file) {
> +	} else if (nf->nf_file && test_bit(NFSD_FILE_GC, &nf->nf_flags)) {
>  		nfsd_file_put_noref(nf);
>  		nfsd_file_schedule_laundrette();
>  	} else
> @@ -1017,12 +1033,14 @@ nfsd_file_is_cached(struct inode *inode)
> =20
>  static __be32
>  nfsd_file_do_acquire(struct svc_rqst *rqstp, struct svc_fh *fhp,
> -		     unsigned int may_flags, struct nfsd_file **pnf, bool open)
> +		     unsigned int may_flags, struct nfsd_file **pnf,
> +		     bool open, bool want_gc)
>  {
>  	struct nfsd_file_lookup_key key =3D {
>  		.type	=3D NFSD_FILE_KEY_FULL,
>  		.need	=3D may_flags & NFSD_FILE_MAY_MASK,
>  		.net	=3D SVC_NET(rqstp),
> +		.gc	=3D want_gc,
>  	};
>  	bool open_retry =3D true;
>  	struct nfsd_file *nf;
> @@ -1117,14 +1135,35 @@ nfsd_file_do_acquire(struct svc_rqst *rqstp, stru=
ct svc_fh *fhp,
>  	 * then unhash.
>  	 */
>  	if (status !=3D nfs_ok || key.inode->i_nlink =3D=3D 0)
> -		if (nfsd_file_unhash(nf))
> -			nfsd_file_put_noref(nf);
> +		nfsd_file_unhash_and_put(nf);
>  	clear_bit_unlock(NFSD_FILE_PENDING, &nf->nf_flags);
>  	smp_mb__after_atomic();
>  	wake_up_bit(&nf->nf_flags, NFSD_FILE_PENDING);
>  	goto out;
>  }
> =20
> +/**
> + * nfsd_file_acquire_gc - Get a struct nfsd_file with an open file
> + * @rqstp: the RPC transaction being executed
> + * @fhp: the NFS filehandle of the file to be opened
> + * @may_flags: NFSD_MAY_ settings for the file
> + * @pnf: OUT: new or found "struct nfsd_file" object
> + *
> + * The nfsd_file object returned by this API is reference-counted
> + * and garbage-collected. The object is retained for a few
> + * seconds after the final nfsd_file_put() in case the caller
> + * wants to re-use it.
> + *
> + * Returns nfs_ok and sets @pnf on success; otherwise an nfsstat in
> + * network byte order is returned.
> + */
> +__be32
> +nfsd_file_acquire_gc(struct svc_rqst *rqstp, struct svc_fh *fhp,
> +		     unsigned int may_flags, struct nfsd_file **pnf)
> +{
> +	return nfsd_file_do_acquire(rqstp, fhp, may_flags, pnf, true, true);
> +}
> +
>  /**
>   * nfsd_file_acquire - Get a struct nfsd_file with an open file
>   * @rqstp: the RPC transaction being executed
> @@ -1132,6 +1171,10 @@ nfsd_file_do_acquire(struct svc_rqst *rqstp, struc=
t svc_fh *fhp,
>   * @may_flags: NFSD_MAY_ settings for the file
>   * @pnf: OUT: new or found "struct nfsd_file" object
>   *
> + * The nfsd_file_object returned by this API is reference-counted
> + * but not garbage-collected. The object is unhashed after the
> + * final nfsd_file_put().
> + *
>   * Returns nfs_ok and sets @pnf on success; otherwise an nfsstat in
>   * network byte order is returned.
>   */
> @@ -1139,7 +1182,7 @@ __be32
>  nfsd_file_acquire(struct svc_rqst *rqstp, struct svc_fh *fhp,
>  		  unsigned int may_flags, struct nfsd_file **pnf)
>  {
> -	return nfsd_file_do_acquire(rqstp, fhp, may_flags, pnf, true);
> +	return nfsd_file_do_acquire(rqstp, fhp, may_flags, pnf, true, false);
>  }
> =20
>  /**
> @@ -1149,6 +1192,10 @@ nfsd_file_acquire(struct svc_rqst *rqstp, struct s=
vc_fh *fhp,
>   * @may_flags: NFSD_MAY_ settings for the file
>   * @pnf: OUT: new or found "struct nfsd_file" object
>   *
> + * The nfsd_file_object returned by this API is reference-counted
> + * but not garbage-collected. The object is released immediately
> + * one RCU grace period after the final nfsd_file_put().
> + *
>   * Returns nfs_ok and sets @pnf on success; otherwise an nfsstat in
>   * network byte order is returned.
>   */
> @@ -1156,7 +1203,7 @@ __be32
>  nfsd_file_create(struct svc_rqst *rqstp, struct svc_fh *fhp,
>  		 unsigned int may_flags, struct nfsd_file **pnf)
>  {
> -	return nfsd_file_do_acquire(rqstp, fhp, may_flags, pnf, false);
> +	return nfsd_file_do_acquire(rqstp, fhp, may_flags, pnf, false, false);
>  }
> =20
>  /*
> diff --git a/fs/nfsd/filecache.h b/fs/nfsd/filecache.h
> index 6b012ea4bd9d..b7efb2c3ddb1 100644
> --- a/fs/nfsd/filecache.h
> +++ b/fs/nfsd/filecache.h
> @@ -38,6 +38,7 @@ struct nfsd_file {
>  #define NFSD_FILE_HASHED	(0)
>  #define NFSD_FILE_PENDING	(1)
>  #define NFSD_FILE_REFERENCED	(2)
> +#define NFSD_FILE_GC		(3)
>  	unsigned long		nf_flags;
>  	struct inode		*nf_inode;	/* don't deref */
>  	refcount_t		nf_ref;
> @@ -55,6 +56,8 @@ void nfsd_file_put(struct nfsd_file *nf);
>  struct nfsd_file *nfsd_file_get(struct nfsd_file *nf);
>  void nfsd_file_close_inode_sync(struct inode *inode);
>  bool nfsd_file_is_cached(struct inode *inode);
> +__be32 nfsd_file_acquire_gc(struct svc_rqst *rqstp, struct svc_fh *fhp,
> +		  unsigned int may_flags, struct nfsd_file **nfp);
>  __be32 nfsd_file_acquire(struct svc_rqst *rqstp, struct svc_fh *fhp,
>  		  unsigned int may_flags, struct nfsd_file **nfp);
>  __be32 nfsd_file_create(struct svc_rqst *rqstp, struct svc_fh *fhp,
> diff --git a/fs/nfsd/nfs3proc.c b/fs/nfsd/nfs3proc.c
> index ff2920546333..d01b29aba662 100644
> --- a/fs/nfsd/nfs3proc.c
> +++ b/fs/nfsd/nfs3proc.c
> @@ -772,8 +772,8 @@ nfsd3_proc_commit(struct svc_rqst *rqstp)
>  				(unsigned long long) argp->offset);
> =20
>  	fh_copy(&resp->fh, &argp->fh);
> -	resp->status =3D nfsd_file_acquire(rqstp, &resp->fh, NFSD_MAY_WRITE |
> -					 NFSD_MAY_NOT_BREAK_LEASE, &nf);
> +	resp->status =3D nfsd_file_acquire_gc(rqstp, &resp->fh, NFSD_MAY_WRITE =
|
> +					    NFSD_MAY_NOT_BREAK_LEASE, &nf);
>  	if (resp->status)
>  		goto out;
>  	resp->status =3D nfsd_commit(rqstp, &resp->fh, nf, argp->offset,
> diff --git a/fs/nfsd/trace.h b/fs/nfsd/trace.h
> index 06a96e955bd0..b065a4b1e0dc 100644
> --- a/fs/nfsd/trace.h
> +++ b/fs/nfsd/trace.h
> @@ -814,7 +814,8 @@ DEFINE_CLID_EVENT(confirmed_r);
>  	__print_flags(val, "|",						\
>  		{ 1 << NFSD_FILE_HASHED,	"HASHED" },		\
>  		{ 1 << NFSD_FILE_PENDING,	"PENDING" },		\
> -		{ 1 << NFSD_FILE_REFERENCED,	"REFERENCED"})
> +		{ 1 << NFSD_FILE_REFERENCED,	"REFERENCED"},		\
> +		{ 1 << NFSD_FILE_GC,		"GC"})
> =20
>  DECLARE_EVENT_CLASS(nfsd_file_class,
>  	TP_PROTO(struct nfsd_file *nf),
> diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c
> index 91c34cef11d8..df1830a1a295 100644
> --- a/fs/nfsd/vfs.c
> +++ b/fs/nfsd/vfs.c
> @@ -1084,7 +1084,7 @@ __be32 nfsd_read(struct svc_rqst *rqstp, struct svc=
_fh *fhp,
>  	__be32 err;
> =20
>  	trace_nfsd_read_start(rqstp, fhp, offset, *count);
> -	err =3D nfsd_file_acquire(rqstp, fhp, NFSD_MAY_READ, &nf);
> +	err =3D nfsd_file_acquire_gc(rqstp, fhp, NFSD_MAY_READ, &nf);
>  	if (err)
>  		return err;
> =20
> @@ -1116,7 +1116,7 @@ nfsd_write(struct svc_rqst *rqstp, struct svc_fh *f=
hp, loff_t offset,
> =20
>  	trace_nfsd_write_start(rqstp, fhp, offset, *cnt);
> =20
> -	err =3D nfsd_file_acquire(rqstp, fhp, NFSD_MAY_WRITE, &nf);
> +	err =3D nfsd_file_acquire_gc(rqstp, fhp, NFSD_MAY_WRITE, &nf);
>  	if (err)
>  		goto out;
> =20
>=20
>=20

Reviewed-by: Jeff Layton <jlayton@kernel.org>