Re: [basic] nfsd failing to start

["J. Bruce Fields" <bfields@fieldses.org>]

On Mon, Jan 12, 2009 at 04:29:48PM -0500, Chuck Lever wrote:
> On Jan 12, 2009, at 4:09 PM, Harry P wrote:
>> "J. Bruce Fields" <bfields@fieldses.org>
>> writes:
>>
>>>>  CONFIG_NFS_FS=m
>>>>  CONFIG_NFS_V3=y
>>>>  CONFIG_NFS_V3_ACL=y
>>>>  CONFIG_NFS_V4=y
>>>>  CONFIG_NFSD=m
>>>>  CONFIG_NFSD_V2_ACL=y
>>>>  CONFIG_NFSD_V3=y
>>>>  CONFIG_NFSD_V3_ACL=y
>>>>  CONFIG_NFSD_V4=y
>>>>  CONFIG_NFS_ACL_SUPPORT=m
>>>>  CONFIG_NFS_COMMON=y
>>>>  CONFIG_SUNRPC=m
>>>>  CONFIG_SUNRPC_GSS=m
>>>>  CONFIG_SUNRPC_REGISTER_V4=y
>>>
>>> From the config help test: "If you enable this option, you must run a
>>> portmapper daemon that supports rpcbind protocol version 4."
>>
>> Thanks, with your clue I commented all nfs and SunRPC 4 related lines
>> recompiled and rebooted... started up with no problems.
>
> Clearly, setting the Kconfig default to N is not enough to prevent this 
> problem.  I don't understand why this option is getting enabled without 
> any other intervention on the part of kernel builders.  Maybe we need to 
> update arch/defconfig?
>
> It's on my immediate to-do list to see if the compile-time and run-time 
> behavior in this area can be improved.

I'd certainly be happiest if we could just fall back on the earlier
rpcbind protocol version at runtime.

--b.