Re: [PATCH] Device-mapper submission 6/7

From: Jeff Garzik (
Date: Thu Oct 17 2002 - 10:10:19 EST

Andi Kleen wrote:
> Joe Thornber <> writes:
>>Is there anyone out there who is going to argue against using an fs
>>interface when I submit it ? Speak now or forever hold your peace !
>>If dm now misses the feature freeze deadline due to this extra work,
>>is it going to be possible to still place it in 2.5 at a later date ?
>>(dm with an ioctl interface is better than no dm at all).
> How would the fs based interface work ?
> plan9 style echo 'rename foo bla' > /dmfs/command would seem ugly to me
> (just look at the horrible parser code for that in mtrr.c)
> doing it fully as fs objects (mv /dmfs/volume1 /dmfs/volume2 for rename)
> could likely get complicated and it's doubtful that VFS semantics completely
> map to DM volumes.

The simplest interface can be read(2) and write(2) to replace ioctl(2),
but still using a single control node [or whatever granularity currently
exists] I think you are over-complicating a simple issue.

> Unless you have a clear and simple way to handle these issues I would
> suggest to stay with simple ioctls. They look clean enough.

Please go back and read what Linus and Al Viro have repeatedly posted
about ioctl(2)...

Overall, one should consider here
* device mapper has never been in the Linux kernel before, thus we have
a duty to make sure it is clean before it gets into the kernel
* ioctls appear "simple" only at first glance. they require more
maintenance in the long run due to the ioctl32 thunking layers, and are
often riddled with shortsighted 32-bit size limits that reduce their
utility on 64-bit platforms
* ioctls cannot be exported over NFS and similar interfaces
* ioctls are a way to add "do something totally different" functionality
to a file descriptor. IOW you read(2) and write(2) a file, and when you
have other tasks to do to this file, add an "escape hatch"? No, that's
the wrong way to go.

ioctls are analogous to procfs: they are simple, easy, and usually the
wrong thing to do.


To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to
More majordomo info at
Please read the FAQ at

This archive was generated by hypermail 2b29 : Wed Oct 23 2002 - 22:00:35 EST