On Tue, Jan 5, 2010 at 2:38 PM, Steve Dickson <SteveD@xxxxxxxxxx> wrote:
> On 01/05/2010 02:23 PM, Chuck Lever wrote:
>>>
>>> Author: Steve Dickson <steved@xxxxxxxxxx>
>>> Date: Tue Jan 5 13:29:07 2010 -0500
>>>
>>> showmount: Try the highest mount version then fall back to lower ones
>>>
>>> Showmount should try the highest mount version first then fall
>>> back to the lower ones when the server returns a RPC_PROGVERSMISMATCH
>>> error. The idea being not using the lower mount versions will begin
>>> the process of moving away from NFSv2 support.
>>>
>>> Signed-off-by: Steve Dickson <steved@xxxxxxxxxx>
>>>
>>> diff --git a/utils/showmount/showmount.c b/utils/showmount/showmount.c
>>> index 418e8b9..17224a6 100644
>>> --- a/utils/showmount/showmount.c
>>> +++ b/utils/showmount/showmount.c
>>> @@ -85,22 +85,29 @@ static const char *nfs_sm_pgmtbl[] = {
>>> NULL,
>>> };
>>>
>>> +static const int mount_vers[] = {
>>
>> RPC version numbers are rpcvers_t, not int.
> True..
>
>>
>> The array above this one is called "nfs_sm_pgmtbl"... to be consistent
>> with existing code you should name this one "nfs_sm_verstbl".
>>
>>> + MOUNTVERS_NFSV3,
>>> + MOUNTVERS_POSIX,
>>> + MOUNTVERS,
>>> +};
>>> +static const int max_vers = (sizeof(mount_vers)/sizeof(mount_vers[0]));
>>
>> Array indices are unsigned.
> Sure...
>
>>
>> Calling this "max_vers" suggests it's actually a version number, and not
>> an index into the version array; that's confusing.
> Its pretty simple code.. don't see this confusion.... but..
> It changed to mount_vers_tbl
>>
>> The array above this one has a NULL entry as it's final element. To be
>> consistent with existing code, you should copy that logic for this array
>> instead of using a separate "max_vers" variable.
> Having the NULL messes up the sizeof calculations
>
>>
>>> +
>>> /*
>>> * Generate an RPC client handle connected to the mountd service
>>> * at @hostname, or die trying.
>>> *
>>> * Supports both AF_INET and AF_INET6 server addresses.
>>> */
>>> -static CLIENT *nfs_get_mount_client(const char *hostname)
>>> +static CLIENT *nfs_get_mount_client(const char *hostname, int vers)
>>
>> RPC version numbers are rpcvers_t, not int. At the very least,
>> clnt_create(3t) takes an unsigned version number argument.
>>
>>> {
>>> rpcprog_t program = nfs_getrpcbyname(MOUNTPROG, nfs_sm_pgmtbl);
>>> CLIENT *client;
>>>
>>> - client = clnt_create(hostname, program, MOUNTVERS, "tcp");
>>> + client = clnt_create(hostname, program, vers, "tcp");
>>> if (client)
>>> return client;
>>>
>>> - client = clnt_create(hostname, program, MOUNTVERS, "udp");
>>> + client = clnt_create(hostname, program, vers, "udp");
>>> if (client)
>>> return client;
>>>
>>> @@ -122,7 +129,7 @@ int main(int argc, char **argv)
>>> mountlist list;
>>> int i;
>>> int n;
>>> - int maxlen;
>>> + int maxlen, vers=0;
>>
>> Array indices are unsigned integers.
>>
>> Calling this "vers" suggests it's actually a version number, and not an
>> index into the version array; that's confusing.
> and an index call 'j' or 'i' is descriptive?? vers is clear enough imho...
>
>>
>>> char **dumpv;
>>>
>>> program_name = argv[0];
>>> @@ -185,7 +192,8 @@ int main(int argc, char **argv)
>>> break;
>>> }
>>>
>>> - mclient = nfs_get_mount_client(hostname);
>>> +again:
>>> + mclient = nfs_get_mount_client(hostname, mount_vers[vers]);
>>> mclient->cl_auth = authunix_create_default();
>>> total_timeout.tv_sec = TOTAL_TIMEOUT;
>>> total_timeout.tv_usec = 0;
>>> @@ -197,6 +205,10 @@ int main(int argc, char **argv)
>>> (xdrproc_t) xdr_void, NULL,
>>> (xdrproc_t) xdr_exports, (caddr_t) &exportlist,
>>> total_timeout);
>>> + if (clnt_stat == RPC_PROGVERSMISMATCH) {
>>> + if (++vers < max_vers)
>>> + goto again;
>>> + }
>>
>> clnt_create(3t) should tell you if the requested version number is
>> working; libtirpc's version does a NULLPROC call, for example. It would
>> be simpler to move this check into nfs_get_mount_client(), then you
>> wouldn't need the retry loop here.
> clnt_create() does not fail.. the clnt_call() does..
>
>>
>> But nfs_get_mount_client() already doesn't care why the clnt_create(3t)
>> call fails, it just retries the creation with different parameters if it
>> didn't succeed. Do we really care enough to retry _only_ if the
>> specific RPC version isn't available? We use the same procedure number
>> with the same arguments and the same results for all three protocol
>> versions, yes?
> Let not make a mountain out of mole hill... the goal of this patch is stop using the lower mount protocol version so server can only support v3 and v4... nothing
> more...
As a side note, I don't care what gets accepted as long as it works
with modern versions of NFS. Let me know if you need anything more
from me, but I'm glad I got the discussion going.
-Dan
--
To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
