[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <239E2F87-E595-4132-B133-504EFF9103B8@oracle.com>
Date: Wed, 15 Jul 2020 12:05:19 -0400
From: Chuck Lever <chuck.lever@...cle.com>
To: Colin Ian King <colin.king@...onical.com>
Cc: Trond Myklebust <trond.myklebust@...merspace.com>,
Anna Schumaker <anna.schumaker@...app.com>,
Bruce Fields <bfields@...ldses.org>,
"David S. Miller" <davem@...emloft.net>,
Jakub Kicinski <kuba@...nel.org>,
Linux NFS Mailing List <linux-nfs@...r.kernel.org>,
"netdev@...r.kernel.org" <netdev@...r.kernel.org>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: xprtrdma: Prevent inline overflow
> On Jul 15, 2020, at 11:56 AM, Colin Ian King <colin.king@...onical.com> wrote:
>
> Hi,
>
> Static analysis with Coverity has found a potential issue with the
> header size calculations in source net/sunrpc/xprtrdma/rpc_rdma.c in
> functions rpcrdma_max_call_header_size and rpcrdma_max_reply_header_size.
>
> The commit in question is relatively old:
>
> commit 302d3deb20682a076e1ab551821cacfdc81c5e4f
> Author: Chuck Lever <chuck.lever@...cle.com>
> Date: Mon May 2 14:41:05 2016 -0400
>
> xprtrdma: Prevent inline overflow
>
> The two issues are as follows:
>
> Issue #1:
>
> 66 static unsigned int rpcrdma_max_call_header_size(unsigned int maxsegs)
> 67 {
> 68 unsigned int size;
> 69
> 70 /* Fixed header fields and list discriminators */
>
> Unused value (UNUSED_VALUE)
>
> 71 size = RPCRDMA_HDRLEN_MIN;
> 72
> 73 /* Maximum Read list size */
> 74 size = maxsegs * rpcrdma_readchunk_maxsz * sizeof(__be32);
> 75
>
> should the size assignment on line 74 be instead:
>
> size += maxsegs * rpcrdma_readchunk_maxsz * sizeof(__be32);
>
>
> Issue #2:
>
> 89 static unsigned int rpcrdma_max_reply_header_size(unsigned int maxsegs)
> 90 {
> 91 unsigned int size;
> 92
> 93 /* Fixed header fields and list discriminators */
>
> Unused value (UNUSED_VALUE)
>
> 94 size = RPCRDMA_HDRLEN_MIN;
> 95
> 96 /* Maximum Write list size */
> 97 size = sizeof(__be32); /* segment count */
>
> should the size assignment in line 97 be instead:
>
> size += sizeof(__be32)?
Colin, Yes to both questions. Can you send a fix to Anna?
--
Chuck Lever
Powered by blists - more mailing lists