Re: [PATCH] nfs: check nf pointer for validity before use

From: Chuck Lever III <chuck.lever@oracle.com>
To: Muhammad Usama Anjum <usama.anjum@collabora.com>
Cc: Bruce Fields <bfields@fieldses.org>,
	Vasily Averin <vvs@virtuozzo.com>,
	Linux NFS Mailing List <linux-nfs@vger.kernel.org>,
	open list <linux-kernel@vger.kernel.org>,
	"kernel@collabora.com" <kernel@collabora.com>,
	"kernel-janitors@vger.kernel.org"
	<kernel-janitors@vger.kernel.org>
Subject: Re: [PATCH] nfs: check nf pointer for validity before use
Date: Tue, 21 Dec 2021 20:00:01 +0000	[thread overview]
Message-ID: <05877B02-900A-4B22-9460-D2F0D20931DC@oracle.com> (raw)
In-Reply-To: <YcIjJ4jN3ax1rqaE@debian-BULLSEYE-live-builder-AMD64>

Hi-

> On Dec 21, 2021, at 1:55 PM, Muhammad Usama Anjum <usama.anjum@collabora.com> wrote:
> 
> Pointer nf can be NULL. It should be validated before dereferencing it.
> 
> Fixes: 8628027ba8 ("nfs: block notification on fs with its own ->lock")
> Signed-off-by: Muhammad Usama Anjum <usama.anjum@collabora.com>

Thanks. To avoid confusion I've squashed this into 8628027 and
refreshed my for-next branch.

Btw, b4 complained about collabora.com's DKIM:

[cel@bazille linux]$ b4 am https://lore.kernel.org/linux-nfs/YcIjJ4jN3ax1rqaE@debian-BULLSEYE-live-builder-AMD64/raw
Grabbing thread from lore.kernel.org/linux-nfs/YcIjJ4jN3ax1rqaE%40debian-BULLSEYE-live-builder-AMD64/t.mbox.gz
Analyzing 1 messages in the thread
Checking attestation on all messages, may take a moment...
---
  ✗ [PATCH] nfs: check nf pointer for validity before use
  ---
  ✗ BADSIG: DKIM/collabora.com
---
Total patches: 1
---
 Link: https://lore.kernel.org/r/YcIjJ4jN3ax1rqaE@debian-BULLSEYE-live-builder-AMD64
 Base: not specified
       git am ./20211221_usama_anjum_nfs_check_nf_pointer_for_validity_before_use.mbx
[cel@bazille linux]$

The patch is obviously correct, so I applied it despite the
attestation failure.

> ---
> fs/nfsd/nfs4state.c | 10 +++++-----
> 1 file changed, 5 insertions(+), 5 deletions(-)
> 
> diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c
> index a526d4183348..bdd30988e615 100644
> --- a/fs/nfsd/nfs4state.c
> +++ b/fs/nfsd/nfs4state.c
> @@ -6947,6 +6947,11 @@ nfsd4_lock(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
> 		goto out;
> 	}
> 
> +	if (!nf) {
> +		status = nfserr_openmode;
> +		goto out;
> +	}
> +
> 	/*
> 	 * Most filesystems with their own ->lock operations will block
> 	 * the nfsd thread waiting to acquire the lock.  That leads to
> @@ -6957,11 +6962,6 @@ nfsd4_lock(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
> 	if (nf->nf_file->f_op->lock)
> 		fl_flags &= ~FL_SLEEP;
> 
> -	if (!nf) {
> -		status = nfserr_openmode;
> -		goto out;
> -	}
> -
> 	nbl = find_or_allocate_block(lock_sop, &fp->fi_fhandle, nn);
> 	if (!nbl) {
> 		dprintk("NFSD: %s: unable to allocate block!\n", __func__);
> -- 
> 2.30.2
> 

--
Chuck Lever