Re: [ 042/108] NFS: Alias the nfs module to nfs4

[Josh Boyer]

On Wed, Sep 12, 2012 at 9:24 PM, Myklebust, Trond
<Trond.Myklebust@xxxxxxxxxx> wrote:
> On Wed, 2012-09-12 at 21:08 -0400, Josh Boyer wrote:
>> On Wed, Sep 12, 2012 at 7:28 PM, Greg Kroah-Hartman
>> <gregkh@xxxxxxxxxxxxxxxxxxx> wrote:
>> > From: Greg KH <gregkh@xxxxxxxxxxxxxxxxxxx>
>> >
>> > 3.5-stable review patch.  If anyone has any objections, please let me know.
>> >
>> > ------------------
>> >
>> > commit 425e776d93a7a5070b77d4f458a5bab0f924652c upstream.
>> >
>> > This allows distros to remove the line from their modprobe
>> > configuration.
>> >
>> > Signed-off-by: Bryan Schumaker <bjschuma@xxxxxxxxxx>
>> > Signed-off-by: Trond Myklebust <Trond.Myklebust@xxxxxxxxxx>
>> > [bwh: Backported to 3.2: adjust context]
>> > Signed-off-by: Ben Hutchings <ben@xxxxxxxxxxxxxxx>
>> > Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
>>
>> Could someone elaborate why this is acceptable now, while it wasn't
>> during the timeframe I sent an identical patch months ago?
>>
>> http://lkml.indiana.edu/hypermail/linux/kernel/1204.3/02064.html
>>
>> We had some trouble in Fedora 17 because of this and would up with
>> patches to nfs-utils because the modalias was NAKed.  If we're bringing
>> it into stable (and upstream) at this point, it would be good to know
>> what changed so I can go and undo what we did because I was told no.
>>
>> Please don't get me wrong, I'm all for simplicity.  I'm just perhaps
>> so simple that the brief changelog confuses me because it doesn't
>> actually describe why the original solution wasn't followed up on.
>
> Linux-3.6 converts NFSv2/v3/v4 into modules, and so modprobe.conf
> entries of the form
>
>   alias nfs4 nfs
>
> will now fail to ensure that 'mount -t nfs4 ...' works. In order to

Ah, ok.  That makes sense.

> facilitate the task for the distros when dealing with kernels >= 3.5.x,
> we've added the automatic module alias so that they can remove the
> hard-coded aliases in modprobe.conf while remaining backward-compatible.

This is queued for 3.5.x, 3.4.x, 3.2.x, and 3.0.x.  That's probably OK,
but it's a slightly broader scope that what you mentioned above.

At any rate, I'm glad to see this go in.  It will make things easier
going forward.  Thanks for the reply.

josh
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/