2009-03-16 12:51:52

by Alexander Beregalov

[permalink] [raw]
Subject: [PATCH next-20090316] sunrpc: fix build breakage

net/sunrpc/clnt.c: In function 'call_bind_status':
net/sunrpc/clnt.c:994: error: 'struct rpc_task' has no member named
'tk_pid'
net/sunrpc/clnt.c: In function 'call_connect_status':
net/sunrpc/clnt.c:1051: error: 'struct rpc_task' has no member named
'tk_pid'

rpc_task has tk_pid if RPC_DEBUG is defined.

Signed-off-by: Alexander Beregalov <[email protected]>
---

net/sunrpc/clnt.c | 4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c
index efd0c7b..68cb91a 100644
--- a/net/sunrpc/clnt.c
+++ b/net/sunrpc/clnt.c
@@ -990,7 +990,7 @@ call_bind_status(struct rpc_task *task)
return;
default:
if (net_ratelimit())
- printk("RPC: %5u unrecognized rpcbind error (%d)\n",
+ dprintk("RPC: %5u unrecognized rpcbind error (%d)\n",
task->tk_pid, -task->tk_status);
}

@@ -1047,7 +1047,7 @@ call_connect_status(struct rpc_task *task)
break;
default:
if (net_ratelimit())
- printk("RPC: %5u unrecognized connect error (%d)\n",
+ dprintk("RPC: %5u unrecognized connect error (%d)\n",
task->tk_pid, status);
rpc_exit(task, -EIO);
}


2009-03-16 13:04:40

by Myklebust, Trond

[permalink] [raw]
Subject: Re: [PATCH next-20090316] sunrpc: fix build breakage

On Mon, 2009-03-16 at 15:51 +0300, Alexander Beregalov wrote:
> net/sunrpc/clnt.c: In function 'call_bind_status':
> net/sunrpc/clnt.c:994: error: 'struct rpc_task' has no member named
> 'tk_pid'
> net/sunrpc/clnt.c: In function 'call_connect_status':
> net/sunrpc/clnt.c:1051: error: 'struct rpc_task' has no member named
> 'tk_pid'
>
> rpc_task has tk_pid if RPC_DEBUG is defined.
>
> Signed-off-by: Alexander Beregalov <[email protected]>
> ---
>
> net/sunrpc/clnt.c | 4 ++--
> 1 files changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c
> index efd0c7b..68cb91a 100644
> --- a/net/sunrpc/clnt.c
> +++ b/net/sunrpc/clnt.c
> @@ -990,7 +990,7 @@ call_bind_status(struct rpc_task *task)
> return;
> default:
> if (net_ratelimit())
> - printk("RPC: %5u unrecognized rpcbind error (%d)\n",
> + dprintk("RPC: %5u unrecognized rpcbind error (%d)\n",
> task->tk_pid, -task->tk_status);
> }
>
> @@ -1047,7 +1047,7 @@ call_connect_status(struct rpc_task *task)
> break;
> default:
> if (net_ratelimit())
> - printk("RPC: %5u unrecognized connect error (%d)\n",
> + dprintk("RPC: %5u unrecognized connect error (%d)\n",
> task->tk_pid, status);
> rpc_exit(task, -EIO);
> }

No, that's not right. The intention is to always report errors that are
fatal and are due to unforeseen circumstances. Your patch means that
those reports are suppressed entirely when RPC_DEBUG is undefined.

Trond
--
Trond Myklebust
Linux NFS client maintainer

NetApp
[email protected]
http://www.netapp.com