Re: [PATCH 09/10] nfsd: expose some more information about NFSv4 opens

["Benjamin Coddington" <bcodding@redhat.com>]

On 25 Apr 2019, at 10:04, J. Bruce Fields wrote:

> From: "J. Bruce Fields" <bfields@redhat.com>
>
> Add open modes, device numbers, inode numbers, and open owners to each
> line of nfsd/clients/#/opens.
>
> Open owners are totally opaque but seem to sometimes have some useful
> ascii strings included, so passing through printable ascii characters
> and escaping the rest seems useful while still being machine-readable.
> ---
>  fs/nfsd/nfs4state.c            | 40 
> ++++++++++++++++++++++++++++------
>  fs/nfsd/state.h                |  2 +-
>  fs/seq_file.c                  | 17 +++++++++++++++
>  include/linux/seq_file.h       |  2 ++
>  include/linux/string_helpers.h |  1 +
>  lib/string_helpers.c           |  5 +++--
>  6 files changed, 57 insertions(+), 10 deletions(-)
>
> diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c
> index 829d1e5440d3..f53621a65e60 100644
> --- a/fs/nfsd/nfs4state.c
> +++ b/fs/nfsd/nfs4state.c
> @@ -42,6 +42,7 @@
>  #include <linux/sunrpc/svcauth_gss.h>
>  #include <linux/sunrpc/addr.h>
>  #include <linux/jhash.h>
> +#include <linux/string_helpers.h>
>  #include "xdr4.h"
>  #include "xdr4cb.h"
>  #include "vfs.h"
> @@ -2261,16 +2262,41 @@ static void opens_stop(struct seq_file *s, 
> void *v)
>  static int opens_show(struct seq_file *s, void *v)
>  {
>  	struct nfs4_stid *st = v;
> -	struct nfs4_ol_stateid *os;
> -	u64 stateid;
> +	struct nfs4_ol_stateid *ols;
> +	struct nfs4_file *nf;
> +	struct file *file;
> +	struct inode *inode;
> +	struct nfs4_stateowner *oo;
> +	unsigned int access, deny;
>
>  	if (st->sc_type != NFS4_OPEN_STID)
>  		return 0; /* XXX: or SEQ_SKIP? */
> -	os = openlockstateid(st);
> -	/* XXX: get info about file, etc., here */
> +	ols = openlockstateid(st);
> +	oo = ols->st_stateowner;
> +	nf = st->sc_file;
> +	file = find_any_file(nf);
> +	inode = d_inode(file->f_path.dentry);
> +
> +	seq_printf(s, STATEID_FMT, STATEID_VAL(&st->sc_stateid));

Should we match the byte order printed with what wireshark/tshark sees?

For example, this will print:

5ccb016e/6d028c97/00000002/00000002 -w  --  fd:00:9163439   'open 
id:\x00\x00\x00.\x00\x00\x00\x00\x00\x00\x021\x8dp\xbe\xc7'

But, tshark -V prints:

         Opcode: OPEN (18)
             Status: NFS4_OK (0)
             stateid
                 [StateID Hash: 0x8298]
                 seqid: 0x00000002
                 Data: 6e01cb5c978c026d02000000
                 [Data hash (CRC-32): 0x8391069f]

I think this is the first time we've exposed state ids to users from 
knfsd,
I wonder if we should make it easier for everyone that might want to try 
to
correlate this information with what they can see in a packet capture.

Ben