Re: [PATCH] Check validity of cl_rpcclient in nfs_server_list_show

Re: [PATCH] Check validity of cl_rpcclient in nfs_server_list_show

[Myklebust, Trond]

Sorry about that. I'll try to fix it later, but I'm on various planes for most of today so I might not have the ability until tomorrow. 

Malahal Naineni <malahal@us.ibm.com> wrote:

The following patch fixes the panic that we see often. I didn't see it
in the git://git.linux-nfs.org/projects/trondmy/linux-nfs.git git tree
yet as far as I can tell. Any comments on the patch?

Thanks, Malahal.

Malahal Naineni [malahal@us.ibm.com] wrote:
> Any comments on this patch?
> 
> Malahal Naineni [malahal@us.ibm.com] wrote:
> > Malahal Naineni [malahal@us.ibm.com] wrote:
> > > Myklebust, Trond [Trond.Myklebust@netapp.com] wrote:
> > > > Shouldn't we just skip all nfs_client structures that don't have
> > > > cl_cons_state set to NFS_CS_READY? I see no reason to display their
> > > > properties even if they have got far enough as to create an rpc client.
> > > > 
> > > > Cheers
> > > >   Trond
> > > 
> > > Sounds good to me. Here is a tested patch.
> > 
> > Any comments on this one?
> > 
> > Thanks, Malahal.
> > 
> > > 
> > > 
> > > As soon as the nfs_client gets created, its cl_rpcclient is set to
> > > ERR_PTR(-EINVAL). The rpc client structure is allocated later. Check
> > > if the client is ready before using the cl_rpcclient pointer.
> > > 
> > > Signed-off-by: Malahal Naineni <malahal@us.ibm.com>
> > > ---
> > >  fs/nfs/client.c |    4 ++++
> > >  1 files changed, 4 insertions(+), 0 deletions(-)
> > > 
> > > diff --git a/fs/nfs/client.c b/fs/nfs/client.c
> > > index 5833fbb..a54c45e 100644
> > > --- a/fs/nfs/client.c
> > > +++ b/fs/nfs/client.c
> > > @@ -1867,6 +1867,10 @@ static int nfs_server_list_show(struct seq_file *m, void *v)
> > >  	/* display one transport per line on subsequent lines */
> > >  	clp = list_entry(v, struct nfs_client, cl_share_link);
> > > 
> > > +	/* Check if the client is initialized */
> > > +	if (clp->cl_cons_state != NFS_CS_READY)
> > > +		return 0;
> > > +
> > >  	seq_printf(m, "v%u %s %s %3d %s\n",
> > >  		   clp->rpc_ops->version,
> > >  		   rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_HEX_ADDR),
> > > -- 
> > > 1.7.0.4
> > > 
> > > --
> > > To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
> > > the body of a message to majordomo@vger.kernel.org
> > > More majordomo info at  http://vger.kernel.org/majordomo-info.html
> > > 
> > --
> > To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
> > the body of a message to majordomo@vger.kernel.org
> > More majordomo info at  http://vger.kernel.org/majordomo-info.html
> > 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 

[PATCH] Check validity of cl_rpcclient in nfs_server_list_show

[Malahal Naineni]

As soon as the nfs_client gets created, its cl_rpcclient is set to
ERR_PTR(-EINVAL). The rpc client structure is allocated later. Check
for the validity of cl_rpcclient before using it.

Signed-off-by: Malahal Naineni <malahal@us.ibm.com>
---
 fs/nfs/client.c |    4 ++++
 1 files changed, 4 insertions(+), 0 deletions(-)

diff --git a/fs/nfs/client.c b/fs/nfs/client.c
index 5833fbb..ff79e72 100644
--- a/fs/nfs/client.c
+++ b/fs/nfs/client.c
@@ -1867,6 +1867,10 @@ static int nfs_server_list_show(struct seq_file *m, void *v)
 	/* display one transport per line on subsequent lines */
 	clp = list_entry(v, struct nfs_client, cl_share_link);
 
+	/* Check if the rpc client is initialized */
+	if (IS_ERR(clp->cl_rpcclient))
+		return 0;
+
 	seq_printf(m, "v%u %s %s %3d %s\n",
 		   clp->rpc_ops->version,
 		   rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_HEX_ADDR),
-- 
1.7.0.4

Re: [PATCH] Check validity of cl_rpcclient in nfs_server_list_show

[Malahal Naineni]

Any comments on the patch below? It is very easy to reproduce the bug
(at least on SLES11 SP1).

Thanks, Malahal.

Malahal Naineni [malahal@us.ibm.com] wrote:
> As soon as the nfs_client gets created, its cl_rpcclient is set to
> ERR_PTR(-EINVAL). The rpc client structure is allocated later. Check
> for the validity of cl_rpcclient before using it.
> 
> Signed-off-by: Malahal Naineni <malahal@us.ibm.com>
> ---
>  fs/nfs/client.c |    4 ++++
>  1 files changed, 4 insertions(+), 0 deletions(-)
> 
> diff --git a/fs/nfs/client.c b/fs/nfs/client.c
> index 5833fbb..ff79e72 100644
> --- a/fs/nfs/client.c
> +++ b/fs/nfs/client.c
> @@ -1867,6 +1867,10 @@ static int nfs_server_list_show(struct seq_file *m, void *v)
>  	/* display one transport per line on subsequent lines */
>  	clp = list_entry(v, struct nfs_client, cl_share_link);
> 
> +	/* Check if the rpc client is initialized */
> +	if (IS_ERR(clp->cl_rpcclient))
> +		return 0;
> +
>  	seq_printf(m, "v%u %s %s %3d %s\n",
>  		   clp->rpc_ops->version,
>  		   rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_HEX_ADDR),
> -- 
> 1.7.0.4
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

RE: [PATCH] Check validity of cl_rpcclient in nfs_server_list_show

[Myklebust, Trond]

> -----Original Message-----
> From: Malahal Naineni [mailto:malahal@us.ibm.com]
> Sent: Monday, September 19, 2011 11:34 AM
> To: linux-nfs@vger.kernel.org
> Cc: Myklebust, Trond
> Subject: Re: [PATCH] Check validity of cl_rpcclient in
nfs_server_list_show
> 
> Any comments on the patch below? It is very easy to reproduce the bug
(at
> least on SLES11 SP1).

Shouldn't we just skip all nfs_client structures that don't have
cl_cons_state set to NFS_CS_READY? I see no reason to display their
properties even if they have got far enough as to create an rpc client.

Cheers
  Trond


> Thanks, Malahal.
> 
> Malahal Naineni [malahal@us.ibm.com] wrote:
> > As soon as the nfs_client gets created, its cl_rpcclient is set to
> > ERR_PTR(-EINVAL). The rpc client structure is allocated later. Check
> > for the validity of cl_rpcclient before using it.
> >
> > Signed-off-by: Malahal Naineni <malahal@us.ibm.com>
> > ---
> >  fs/nfs/client.c |    4 ++++
> >  1 files changed, 4 insertions(+), 0 deletions(-)
> >
> > diff --git a/fs/nfs/client.c b/fs/nfs/client.c index
5833fbb..ff79e72
> > 100644
> > --- a/fs/nfs/client.c
> > +++ b/fs/nfs/client.c
> > @@ -1867,6 +1867,10 @@ static int nfs_server_list_show(struct
seq_file
> *m, void *v)
> >  	/* display one transport per line on subsequent lines */
> >  	clp = list_entry(v, struct nfs_client, cl_share_link);
> >
> > +	/* Check if the rpc client is initialized */
> > +	if (IS_ERR(clp->cl_rpcclient))
> > +		return 0;
> > +
> >  	seq_printf(m, "v%u %s %s %3d %s\n",
> >  		   clp->rpc_ops->version,
> >  		   rpc_peeraddr2str(clp->cl_rpcclient,
> RPC_DISPLAY_HEX_ADDR),
> > --
> > 1.7.0.4
> >
> > --
> > To unsubscribe from this list: send the line "unsubscribe linux-nfs"
> > in the body of a message to majordomo@vger.kernel.org More
> majordomo
> > info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH] Check validity of cl_rpcclient in nfs_server_list_show

[Malahal Naineni]

Myklebust, Trond [Trond.Myklebust@netapp.com] wrote:
> Shouldn't we just skip all nfs_client structures that don't have
> cl_cons_state set to NFS_CS_READY? I see no reason to display their
> properties even if they have got far enough as to create an rpc client.
> 
> Cheers
>   Trond

Sounds good to me. Here is a tested patch.


As soon as the nfs_client gets created, its cl_rpcclient is set to
ERR_PTR(-EINVAL). The rpc client structure is allocated later. Check
if the client is ready before using the cl_rpcclient pointer.

Signed-off-by: Malahal Naineni <malahal@us.ibm.com>
---
 fs/nfs/client.c |    4 ++++
 1 files changed, 4 insertions(+), 0 deletions(-)

diff --git a/fs/nfs/client.c b/fs/nfs/client.c
index 5833fbb..a54c45e 100644
--- a/fs/nfs/client.c
+++ b/fs/nfs/client.c
@@ -1867,6 +1867,10 @@ static int nfs_server_list_show(struct seq_file *m, void *v)
 	/* display one transport per line on subsequent lines */
 	clp = list_entry(v, struct nfs_client, cl_share_link);
 
+	/* Check if the client is initialized */
+	if (clp->cl_cons_state != NFS_CS_READY)
+		return 0;
+
 	seq_printf(m, "v%u %s %s %3d %s\n",
 		   clp->rpc_ops->version,
 		   rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_HEX_ADDR),
-- 
1.7.0.4

Re: [PATCH] Check validity of cl_rpcclient in nfs_server_list_show

[Malahal Naineni]

Malahal Naineni [malahal@us.ibm.com] wrote:
> Myklebust, Trond [Trond.Myklebust@netapp.com] wrote:
> > Shouldn't we just skip all nfs_client structures that don't have
> > cl_cons_state set to NFS_CS_READY? I see no reason to display their
> > properties even if they have got far enough as to create an rpc client.
> > 
> > Cheers
> >   Trond
> 
> Sounds good to me. Here is a tested patch.

Any comments on this one?

Thanks, Malahal.

> 
> 
> As soon as the nfs_client gets created, its cl_rpcclient is set to
> ERR_PTR(-EINVAL). The rpc client structure is allocated later. Check
> if the client is ready before using the cl_rpcclient pointer.
> 
> Signed-off-by: Malahal Naineni <malahal@us.ibm.com>
> ---
>  fs/nfs/client.c |    4 ++++
>  1 files changed, 4 insertions(+), 0 deletions(-)
> 
> diff --git a/fs/nfs/client.c b/fs/nfs/client.c
> index 5833fbb..a54c45e 100644
> --- a/fs/nfs/client.c
> +++ b/fs/nfs/client.c
> @@ -1867,6 +1867,10 @@ static int nfs_server_list_show(struct seq_file *m, void *v)
>  	/* display one transport per line on subsequent lines */
>  	clp = list_entry(v, struct nfs_client, cl_share_link);
> 
> +	/* Check if the client is initialized */
> +	if (clp->cl_cons_state != NFS_CS_READY)
> +		return 0;
> +
>  	seq_printf(m, "v%u %s %s %3d %s\n",
>  		   clp->rpc_ops->version,
>  		   rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_HEX_ADDR),
> -- 
> 1.7.0.4
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 

Re: [PATCH] Check validity of cl_rpcclient in nfs_server_list_show

[Malahal Naineni]

Any comments on this patch?

Malahal Naineni [malahal@us.ibm.com] wrote:
> Malahal Naineni [malahal@us.ibm.com] wrote:
> > Myklebust, Trond [Trond.Myklebust@netapp.com] wrote:
> > > Shouldn't we just skip all nfs_client structures that don't have
> > > cl_cons_state set to NFS_CS_READY? I see no reason to display their
> > > properties even if they have got far enough as to create an rpc client.
> > > 
> > > Cheers
> > >   Trond
> > 
> > Sounds good to me. Here is a tested patch.
> 
> Any comments on this one?
> 
> Thanks, Malahal.
> 
> > 
> > 
> > As soon as the nfs_client gets created, its cl_rpcclient is set to
> > ERR_PTR(-EINVAL). The rpc client structure is allocated later. Check
> > if the client is ready before using the cl_rpcclient pointer.
> > 
> > Signed-off-by: Malahal Naineni <malahal@us.ibm.com>
> > ---
> >  fs/nfs/client.c |    4 ++++
> >  1 files changed, 4 insertions(+), 0 deletions(-)
> > 
> > diff --git a/fs/nfs/client.c b/fs/nfs/client.c
> > index 5833fbb..a54c45e 100644
> > --- a/fs/nfs/client.c
> > +++ b/fs/nfs/client.c
> > @@ -1867,6 +1867,10 @@ static int nfs_server_list_show(struct seq_file *m, void *v)
> >  	/* display one transport per line on subsequent lines */
> >  	clp = list_entry(v, struct nfs_client, cl_share_link);
> > 
> > +	/* Check if the client is initialized */
> > +	if (clp->cl_cons_state != NFS_CS_READY)
> > +		return 0;
> > +
> >  	seq_printf(m, "v%u %s %s %3d %s\n",
> >  		   clp->rpc_ops->version,
> >  		   rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_HEX_ADDR),
> > -- 
> > 1.7.0.4
> > 
> > --
> > To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
> > the body of a message to majordomo@vger.kernel.org
> > More majordomo info at  http://vger.kernel.org/majordomo-info.html
> > 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 

Re: [PATCH] Check validity of cl_rpcclient in nfs_server_list_show

[Malahal Naineni]

The following patch fixes the panic that we see often. I didn't see it
in the git://git.linux-nfs.org/projects/trondmy/linux-nfs.git git tree
yet as far as I can tell. Any comments on the patch?

Thanks, Malahal.

Malahal Naineni [malahal@us.ibm.com] wrote:
> Any comments on this patch?
> 
> Malahal Naineni [malahal@us.ibm.com] wrote:
> > Malahal Naineni [malahal@us.ibm.com] wrote:
> > > Myklebust, Trond [Trond.Myklebust@netapp.com] wrote:
> > > > Shouldn't we just skip all nfs_client structures that don't have
> > > > cl_cons_state set to NFS_CS_READY? I see no reason to display their
> > > > properties even if they have got far enough as to create an rpc client.
> > > > 
> > > > Cheers
> > > >   Trond
> > > 
> > > Sounds good to me. Here is a tested patch.
> > 
> > Any comments on this one?
> > 
> > Thanks, Malahal.
> > 
> > > 
> > > 
> > > As soon as the nfs_client gets created, its cl_rpcclient is set to
> > > ERR_PTR(-EINVAL). The rpc client structure is allocated later. Check
> > > if the client is ready before using the cl_rpcclient pointer.
> > > 
> > > Signed-off-by: Malahal Naineni <malahal@us.ibm.com>
> > > ---
> > >  fs/nfs/client.c |    4 ++++
> > >  1 files changed, 4 insertions(+), 0 deletions(-)
> > > 
> > > diff --git a/fs/nfs/client.c b/fs/nfs/client.c
> > > index 5833fbb..a54c45e 100644
> > > --- a/fs/nfs/client.c
> > > +++ b/fs/nfs/client.c
> > > @@ -1867,6 +1867,10 @@ static int nfs_server_list_show(struct seq_file *m, void *v)
> > >  	/* display one transport per line on subsequent lines */
> > >  	clp = list_entry(v, struct nfs_client, cl_share_link);
> > > 
> > > +	/* Check if the client is initialized */
> > > +	if (clp->cl_cons_state != NFS_CS_READY)
> > > +		return 0;
> > > +
> > >  	seq_printf(m, "v%u %s %s %3d %s\n",
> > >  		   clp->rpc_ops->version,
> > >  		   rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_HEX_ADDR),
> > > -- 
> > > 1.7.0.4
> > > 
> > > --
> > > To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
> > > the body of a message to majordomo@vger.kernel.org
> > > More majordomo info at  http://vger.kernel.org/majordomo-info.html
> > > 
> > --
> > To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
> > the body of a message to majordomo@vger.kernel.org
> > More majordomo info at  http://vger.kernel.org/majordomo-info.html
> > 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>