devfs

devfs

[Ian Molton]

On 11 Nov 2002 20:50:39 -0500
Robert Love <rml@tech9.net> wrote:

> On Mon, 2002-11-11 at 20:32, Ryan Anderson wrote:
> 
> > From an outsider point of view (and because nobody else responded),
> > I think the big question here would be: Would you use it after this
> > cleanup?
> > 
> > If you say yes, I'd say that's a good sign in its favor.
> 
> Good heuristic, except Al would not use devfs in either case :)

Personally I love devfs. I've not looked at its internals, but its never
failed me yet, and I like the way /dev only contains stuff that actually
exists.

And if Al doesnt like the code... hes free to reimplement it...

Re: devfs

[Xavier Bestel]

Le mar 12/11/2002 à 10:32, Ian Molton a écrit :
> On 11 Nov 2002 20:50:39 -0500
> Robert Love <rml@tech9.net> wrote:
> 
> > On Mon, 2002-11-11 at 20:32, Ryan Anderson wrote:
> > 
> > > From an outsider point of view (and because nobody else responded),
> > > I think the big question here would be: Would you use it after this
> > > cleanup?
> > > 
> > > If you say yes, I'd say that's a good sign in its favor.
> > 
> > Good heuristic, except Al would not use devfs in either case :)
> 
> Personally I love devfs. I've not looked at its internals, but its never
> failed me yet, and I like the way /dev only contains stuff that actually
> exists.

I'm wondering if a totally userspace solution could replace devs ?
Something using hotplug + sysfs and creating directories/nodes as they
appear on the system. This way, the policy (how do I name what) could be
moved out of the kernel.

	Xav

Re: devfs

[john slee]

On Tue, Nov 12, 2002 at 10:43:41AM +0100, Xavier Bestel wrote:
> I'm wondering if a totally userspace solution could replace devs ?
> Something using hotplug + sysfs and creating directories/nodes as they
> appear on the system. This way, the policy (how do I name what) could be
> moved out of the kernel.

curious!  you mean similar to (and a logical extension of) the 'disks'
command in solaris?  at least i think thats what its called...

j.

-- 
toyota power: http://indigoid.net/

Re: devfs

[Sean Neakums]

commence  john slee quotation:

> On Tue, Nov 12, 2002 at 10:43:41AM +0100, Xavier Bestel wrote:
>> I'm wondering if a totally userspace solution could replace devs ?
>> Something using hotplug + sysfs and creating directories/nodes as
>> they appear on the system. This way, the policy (how do I name
>> what) could be moved out of the kernel.
>
> curious!  you mean similar to (and a logical extension of) the
> 'disks' command in solaris?  at least i think thats what its
> called...

Except that a /sbin/hotplug er, "solution", would be dynamic.  I had
always assumed in the back of my mind that such a solution would make
devfs go away some day, although I don't think I actually read about
it anywhere.

Actually, here's a question: are /sbin/hotplug upcalls serialized in
some fashion?  I'd hate to online a thousand devices in my disk array
and have the machine forkbomb itself.

-- 
 /                          |
[|] Sean Neakums            |  Questions are a burden to others;
[|] <sneakums@zork.net>     |      answers a prison for oneself.
 \                          |

Re: devfs

[Alexander Viro]

On 12 Nov 2002, Xavier Bestel wrote:

> I'm wondering if a totally userspace solution could replace devs ?
> Something using hotplug + sysfs and creating directories/nodes as they
> appear on the system. This way, the policy (how do I name what) could be
> moved out of the kernel.

	Guys, may I remind you that Oct 31 had been more than a week ago?
Devfs *is* a race-ridden pile of crap, but we are in a goddamn feature
freeze, so let's get real.

	Interfaces can and should be cleaned up.  Ditto for semantics of
registering/unregistering - that allows to make glue in drivers more
straightforward.  Majestic flamewars about removing the thing completely/
moving it to userland/etc. are exercises in masturbation by that point.

	Again, WE ARE IN FEATURE FREEZE.

	Now, does somebody have technical comments on the proposed changes?

Re: devfs

[Xavier Bestel]

Le mar 12/11/2002 à 10:49, john slee a écrit :
> On Tue, Nov 12, 2002 at 10:43:41AM +0100, Xavier Bestel wrote:
> > I'm wondering if a totally userspace solution could replace devs ?
> > Something using hotplug + sysfs and creating directories/nodes as they
> > appear on the system. This way, the policy (how do I name what) could be
> > moved out of the kernel.
> 
> curious!  you mean similar to (and a logical extension of) the 'disks'
> command in solaris?  at least i think thats what its called...

I don't know solaris (from an admin POV), but I meant something like,
well, devfs. On top of which we could add feature like device naming
stability across system upgrades (I know solaris does that).

Xav

Re: devfs

[Ian Molton]

On Tue, 12 Nov 2002 05:04:41 -0500 (EST)
Alexander Viro <viro@math.psu.edu> wrote:

> 	Again, WE ARE IN FEATURE FREEZE.
> 
> 	Now, does somebody have technical comments on the proposed
> 	changes?

And since when did feature freeze affect, as the guy said, *purely*
userspace implementations?

(can it be done with current hotplug/sysfs?)

Re: devfs

[Dave Jones]

On Tue, Nov 12, 2002 at 10:25:35AM +0000, Ian Molton wrote:
 > > 	Again, WE ARE IN FEATURE FREEZE.
 > And since when did feature freeze affect, as the guy said, *purely*
 > userspace implementations?

 Since it would a *feature* to move it out of kernel space.
 To reiterate : _FEATURE_ _FREEZE_. Nothing[1] new[2]
 should be going into mainline at this point.

 We should now be in the stabilising period, where we don't require
 testers to have to upgrade their userspace every fortnight.

		Dave

[1] Rusty's modules rewrite and whatever else Linus queed up
    seems to have exemption from this rule. Hopefully there
    isn't too much of this stuff.
[2] Where 'new' can also mean 'bloody large reimplementation'
 
-- 
| Dave Jones.        http://www.codemonkey.org.uk
| SuSE Labs

Re: devfs

[Ian Molton]

On Tue, 12 Nov 2002 10:46:50 +0000
Dave Jones <davej@codemonkey.org.uk> wrote:

> > And since when did feature freeze affect, as the guy said, *purely*
>  > userspace implementations?
> 
>  Since it would a *feature* to move it out of kernel space.
>  To reiterate : _FEATURE_ _FREEZE_. Nothing[1] new[2]
>  should be going into mainline at this point.

nothing new WOULD be going in. nor would devfs need to move out - just
disable it. It *is* still an option, is it not?

>  We should now be in the stabilising period, where we don't require
>  testers to have to upgrade their userspace every fortnight.

its not a kernel issue anymore, if its totally in userspace, is it?

Re: devfs

[Rando Christensen]

Tue, 12 Nov 2002 10:46:50 +0000: Dave Jones (Dave Jones
<davej@codemonkey.org.uk>):

> On Tue, Nov 12, 2002 at 10:25:35AM +0000, Ian Molton wrote:
>  > > 	Again, WE ARE IN FEATURE FREEZE.
>  > And since when did feature freeze affect, as the guy said, *purely*
>  > userspace implementations?
>  Since it would a *feature* to move it out of kernel space.
>  To reiterate : _FEATURE_ _FREEZE_. Nothing[1] new[2]
>  should be going into mainline at this point.

Well, it does make sense to start planning out what might be done to
devfs in the /next/ devel kernel, then. If one were so inclined, they
could have a stable enough system for it ready for inclusion next time
around, depending on who they get to test it.

Clean up the devfs API for now, and start working on what might
eventually replace it- Whether that's devfs2 or a purely userspace
implementation or something new and wacky is up to whoever writes it.

Rather than saying "Devfs sucks, and we can't do anything about it other
than fix it's more minor problems because we're in feature freeze", we
should be saying "devfs sucks; we're a little late for feature freeze,
so let's clean up what we can and work on something much better for the
next time around."

Devfs as it is now is a nice idea, to me. It's an extremely organized
/dev which is also an accurate portrayal of what's available to use on a
particular machine. I'd hate to see it just ripped out entirely, and
while they've never affected me, I've heard of many issues with it. I'd
like to hear what ideas people may have /for/ a future replacement.

-- 
< There is a light that shines on the frontier >
<   And maybe someday, We're gonna be there.   >
<    Rando Christensen / rando@babblica.net    >

Re: devfs

[Helge Hafting]

Alexander Viro wrote:
> 
> On 12 Nov 2002, Xavier Bestel wrote:
> 
> > I'm wondering if a totally userspace solution could replace devs ?
> > Something using hotplug + sysfs and creating directories/nodes as they
> > appear on the system. This way, the policy (how do I name what) could be
> > moved out of the kernel.
> 
>         Guys, may I remind you that Oct 31 had been more than a week ago?
> Devfs *is* a race-ridden pile of crap, but we are in a goddamn feature
> freeze, so let's get real.
> 
A kernel feature freeze don't apply to a pure userspace implemetation,
so let him try.  :-)  It won't affect the current devfs code, it'll
simply be configured out.
/dev on tmpfs, populated by hotplug. Could be interesting to see.

[...]
>         Now, does somebody have technical comments on the proposed changes?

Only the obvious - a cleanup is generally good and so is getting rid
of unused parameters.

Helge Hafting

Re: devfs

[Oliver Neukum]

> Actually, here's a question: are /sbin/hotplug upcalls serialized in
> some fashion?  I'd hate to online a thousand devices in my disk array
> and have the machine forkbomb itself.

Nope, no serialisation. You don't have any guarantee even that
addition will be delivered before removal.

	Regards
		Oliver

Re: devfs

[Alexander Viro]

On Tue, 12 Nov 2002, Rando Christensen wrote:

> Rather than saying "Devfs sucks, and we can't do anything about it other
> than fix it's more minor problems because we're in feature freeze", we
> should be saying "devfs sucks; we're a little late for feature freeze,
> so let's clean up what we can and work on something much better for the
> next time around."

Whatever is going to happen with devfs, believe me, the first thing
you'll need is stable glue in drivers - as simple and natural from the
driver POV as possible.  Complexity of doing development in 2.6 will
directly depend on the amount of code in drivers touched by patches.
BTDT - one can carry (and gradually merge) deep rewrites of core code
during -STABLE if it's done carefully.  But as soon as your patchset
hits the drivers - you are in for a world of pain just porting it to
next versions.

_That_ is critical - get interfaces right in -CURRENT, so that further
work would not cross these boundaries; then work in the resulting areas
becomes independent.  

And in situations like that of devfs, simple rules for callers are pretty
much the main criteria - if users of the interface have to jump through
some hoops, it's a sign that interface needs changes...

Re: devfs

[Alan Cox]

On Tue, 2002-11-12 at 10:46, Dave Jones wrote:
> On Tue, Nov 12, 2002 at 10:25:35AM +0000, Ian Molton wrote:
>  > > 	Again, WE ARE IN FEATURE FREEZE.
>  > And since when did feature freeze affect, as the guy said, *purely*
>  > userspace implementations?
> 
>  Since it would a *feature* to move it out of kernel space.
>  To reiterate : _FEATURE_ _FREEZE_. Nothing[1] new[2]
>  should be going into mainline at this point.

Who cares. You can do most of it already with hotplug and the remaining
bits are very much "oops should tell hotplug" bug fixes nothing more.

Alan

Re: devfs

[Ian Molton]

On 12 Nov 2002 14:53:38 +0000
Alan Cox <alan@lxorguk.ukuu.org.uk> wrote:

> >  Since it would a *feature* to move it out of kernel space.
> >  To reiterate : _FEATURE_ _FREEZE_. Nothing[1] new[2]
> >  should be going into mainline at this point.
> 
> Who cares. You can do most of it already with hotplug and the
> remaining bits are very much "oops should tell hotplug" bug fixes
> nothing more.

The more I think about this userspace devfs, the more I like it. simple,
clean, AND doesnt affect the kernel.

Pure coding beauty.

Re: devfs

[Greg KH]

On Tue, Nov 12, 2002 at 10:43:41AM +0100, Xavier Bestel wrote:
> 
> I'm wondering if a totally userspace solution could replace devs ?
> Something using hotplug + sysfs and creating directories/nodes as they
> appear on the system. This way, the policy (how do I name what) could be
> moved out of the kernel.

Yes, that is _exactly_ what I am working on doing, and have stated as
such on this list a number of times in the past.

As for it being after the feature freeze comment, like Alan noted,
everything is already present to do this, it's just going to take some
more notifiers being added and some userspace code written.

And no, even if all of this gets done, I would not want to remove devfs
from the kernel, yet :)

thanks,

greg k-h

Re: hotplug (was devfs)

[Nick Craig-Wood]

On Tue, Nov 12, 2002 at 01:51:08PM +0100, Oliver Neukum wrote:
> > Actually, here's a question: are /sbin/hotplug upcalls serialized in
> > some fashion?  I'd hate to online a thousand devices in my disk array
> > and have the machine forkbomb itself.
> 
> Nope, no serialisation. You don't have any guarantee even that
> addition will be delivered before removal.

And that is why (we finally discovered) we were getting
non-deterministic device numbering of USB nodes.

We have machines with 6 x 4 port USB <-> serial converters attached.
These would get randomly assigned usb device ids and hence random
/dev/ttyUSB nodes.  Not very useful when there is a load of different
things attached to the 24 serial ports!

Sometimes we also found that one of the devices wouldn't get
initialised properly.

We fixed these problems by removing hotplug and loading the relevant
kernel modules in the correct order and voila a perfectly
deterministic order for the /dev/ttyUSBs with all devices initialised.

Plugging in our USB bus with 24 devices on it does indeed produce a
mini-forkbomb effect ;-) (Especially since these Keyspan devices are
initialised twice - once without firmware and once with firmware.)

So - perhaps hotplug ought to be serialised?

-- 
Nick Craig-Wood
ncw1@axis.demon.co.uk

Re: hotplug (was devfs)

[Greg KH]

On Wed, Nov 13, 2002 at 10:48:09AM +0000, Nick Craig-Wood wrote:
> 
> We fixed these problems by removing hotplug and loading the relevant
> kernel modules in the correct order and voila a perfectly
> deterministic order for the /dev/ttyUSBs with all devices initialised.

deterministic for you :)

What hotplug will do is allow you to assign a /dev entry to a specific
device, so that you can go off of the topology, and not just the order
in which the devices are found.  That is how this problem will be
solved properly.

> Plugging in our USB bus with 24 devices on it does indeed produce a
> mini-forkbomb effect ;-) (Especially since these Keyspan devices are
> initialised twice - once without firmware and once with firmware.)
> 
> So - perhaps hotplug ought to be serialised?

No, it's not needed for this problem.  There has been talk of
serializing stuff in userspace, which is the proper way to handle some
of the remove before add was seen problems.

thanks,

greg k-h

Re: hotplug (was devfs)

[Greg KH]

On Wed, Nov 13, 2002 at 06:06:06PM +0000, Nick Craig-Wood wrote:
> 
> So I'll be able to say usb bus1/1/4/1 port 3 should be /dev/ttyUSB15
> and it will always be that port?  That would be perfect.

Yes, that is the goal.

thanks,

greg k-h

Re: hotplug (was devfs)

[Nick Craig-Wood]

On Wed, Nov 13, 2002 at 09:02:04AM -0800, Greg KH wrote:
> On Wed, Nov 13, 2002 at 10:48:09AM +0000, Nick Craig-Wood wrote:
> > 
> > We fixed these problems by removing hotplug and loading the relevant
> > kernel modules in the correct order and voila a perfectly
> > deterministic order for the /dev/ttyUSBs with all devices initialised.
> 
> deterministic for you :)

Indeed!  It was deterministic in the sense of

1) I booted the machine 10 times and the devices came up in the same
order ;-)

2) the order of the devices was related to the usb topology, like this
(these are usb bus positions as noted by hub.c)

  1/1     <- hub1
  1/1/1   <- keyspan 1 /dev/ttyUSB0..3
  1/1/2   <- keyspan 2 /dev/ttyUSB4..7
  1/1/3   <- keyspan 3 /dev/ttyUSB8..11
  1/1/4   <- hub2
  1/1/4/1 <- keyspan 4 /dev/ttyUSB12..15
  1/1/4/2 <- keyspan 5 /dev/ttyUSB16..19
  1/1/4/3 <- keyspan 6 /dev/ttyUSB20..23

That seemed like a sensible order to me!

> What hotplug will do is allow you to assign a /dev entry to a specific
> device, so that you can go off of the topology, and not just the order
> in which the devices are found.  That is how this problem will be
> solved properly.

So I'll be able to say usb bus1/1/4/1 port 3 should be /dev/ttyUSB15
and it will always be that port?  That would be perfect.

> > Plugging in our USB bus with 24 devices on it does indeed produce a
> > mini-forkbomb effect ;-) (Especially since these Keyspan devices are
> > initialised twice - once without firmware and once with firmware.)
> > 
> > So - perhaps hotplug ought to be serialised?
> 
> No, it's not needed for this problem.  There has been talk of
> serializing stuff in userspace, which is the proper way to handle some
> of the remove before add was seen problems.

Userspace serialisation would have solved this problem for us too I
think without the extra mapping mechanism.

-- 
Nick Craig-Wood
ncw1@axis.demon.co.uk

Re: hotplug (was devfs)

[Oliver Neukum]

> > So - perhaps hotplug ought to be serialised?
>
> No, it's not needed for this problem.  There has been talk of
> serializing stuff in userspace, which is the proper way to handle some
> of the remove before add was seen problems.

We may need some kind of load control.
The thought of firing up hundreds of hotplug scripts
simultaneously is not pretty.

	Regards
		Oliver

Re: devfs

[Ed Sweetman]

On Wed, 2002-08-21 at 00:49, Richard Gooch wrote:
> It seems the reason you removed devfs is that you followed Al's bad
> advice:
> Alexander Viro writes:
> > Don't be silly - if you want to test anything, devfs is the last
> > thing you want on the system.
> 
> In fact, devfs works quite robustly for many people, and wasn't
> involved in the IDE problems you were having. Al is an absolutist: if
> it's not 100% provably correct, it falls into his other category,
> "spawn of satan".
> 
> So next time someone claims devfs is causing you problems, treat it
> with the skepticism it deserves.
> 
> 				Regards,
> 
> 					Richard....


Ok, well the more happy the people who have the ability to know what the
problem is the better so despite a simple usb issue, moving out of devfs
wasn't a hassle.  I'm willing to move back to vanilla if it helps in
chasing a problem down.  I'm just kind of disappointed that the
discussion of me playing around with devfs got more attention from the
ide guys than the actual problem since it seems to be strictly DMA
related and not a drive problem so if it's chipset conflicts it should
be documented either in the good-bad firmware so the kernel disables dma
on boot of this particular promise chipset when this particular via
chipset is present or in the HELP for the config option so at least
users are aware that drive corruption has been known to happen on the
linux promise driver when dma is enabled on some via chipsets.  That is
unless it's a fixable problem.  Which i rather hope it is.  neither the
abit motherboard or promise controller card are uncommon.   But since
nobody i know has via + promise setups I cant ask them to run the
cerberus test to see the same behavior to see the extent of the promise
+ via problems or if it's even got anything to do with via.  (shrugs)

Re: devfs

[Richard Gooch]

Ed Sweetman writes:
> Ok, that all makes sense.  I removed the dev entries because they
> weren't needed by me anymore but I suppose it doesn't hurt anything
> to just keep them there anyways so I've fixed all that.  Either way,
> removing devfs did nothing but apparently it was asked to be done to
> allow better testing and/or debugging to be done.  But i've yet to
> get any reason why I removed devfs to investigate promise ide
> controller's dma related memory failures.  I've removed devfs and
> replaced the old /dev entries, no problem.  I'm not getting off
> topic about that.  It's all done so i'm waiting for the next step
> here.

It seems the reason you removed devfs is that you followed Al's bad
advice:
Alexander Viro writes:
> Don't be silly - if you want to test anything, devfs is the last
> thing you want on the system.

In fact, devfs works quite robustly for many people, and wasn't
involved in the IDE problems you were having. Al is an absolutist: if
it's not 100% provably correct, it falls into his other category,
"spawn of satan".

So next time someone claims devfs is causing you problems, treat it
with the skepticism it deserves.

				Regards,

					Richard....
Permanent: rgooch@atnf.csiro.au
Current:   rgooch@ras.ucalgary.ca

Re: devfs

[Greg KH]

On Sun, Aug 18, 2002 at 09:06:18PM -0400, Olivier Galibert wrote:
> 
> I've been wondering, imagine that in the future we have a working
> dynamic device filesystem (be it devfs, driverfs, whatever) nice
> enough that we don't want a disk-based /dev anymore.  How are we
> supposed to mount it so that the kernel's open("/dev/console")
> succeeds?

initramfs might already contain a minimial /dev that has those kinds of
entries in it.

thanks,

greg k-h

Re: devfs

[Olivier Galibert]

On Sun, Aug 18, 2002 at 07:03:42PM -0400, Alexander Viro wrote:
> devfs "mount" option is an idiotic kludge that makes _kernel_ mount
> it on /dev after the root fs had been mounted.  Why it had been
> introduced is a great mistery, since the normal way is to have a
> corresponding line in /etc/fstab and have userland mount whatever
> it needs.

I've been wondering, imagine that in the future we have a working
dynamic device filesystem (be it devfs, driverfs, whatever) nice
enough that we don't want a disk-based /dev anymore.  How are we
supposed to mount it so that the kernel's open("/dev/console")
succeeds?

  OG.

Re: devfs

[Barry K. Nathan]

On Sun, Aug 18, 2002 at 06:26:35PM -0400, Ed Sweetman wrote:
> On Sun, 2002-08-18 at 17:53, Barry K. Nathan wrote:
[snip]
> > Starting at line 722 of
> > linux-2.4.19/Documentation/filesystems/devfs/README:
> > 
> > > In general, a kernel built with CONFIG_DEVFS_FS=y but without mounting
> > > devfs onto /dev is completely safe, and requires no
> > > configuration changes.
> > 
> > I skimmed through the documentation and it appears to assume that you're
> > not deleting all the stuff in /dev before switching over to devfs.
> 
> This has nothing to do with not mounting devfs and still using devfs to
> work with devices.   If devfs is not mounted but you're still using
> devfs, you shouldn't need anything in /dev.

IMO the combination of common sense and a little Linux/Unix knowledge
would suggest that you can't use a filesystem if it's not mounted.
(Also, see my next paragraph in this message.)

> The documentation says you
> can use devfs without mounting

No, that's not what it says. It says that you can run a kernel with
devfs enabled but not mounted. This does not imply that devfs is in use
and providing the device nodes, just that it is enabled and present in
the event that it should be mounted.

> and This is what i'm saying is
> problematic and doesn't seem possible in normal usage.   It's an
> optional config so are we using devfs when we dont mount it or not?  
> and if not, then why make not mounting it an option ? 

Why make not mounting it an option? There's more than one reason. You
might want to wait until well into the boot process before mounting it
(I think this one's mentioned in the documentation but I'm not 100%
sure). You might also want to temporarily disable devfs mounting to
avoid a security hole in the event that one is found in devfs, until an
updated kernel is available (this actually happened earlier this year
with recent Mandrake Linux releases that use devfs by default).

> If it's using the old device files in /dev then how can it be using
> devfs and how can accessing physical inodes on the disk be intentional
> to devfs? 

If it's using the old /dev nodes then it's not using devfs -- but you
can switch over later after or during boot.

> > In any event, it might be a good idea to make the documentation a bit
> > more explicit about this, and I might send a patch to the mailing
> > list later today.
> 
> I'm not talking about booting without devfs enabled being the problem, i
> know booting without devfs enabled I'll have issues booting the system
> without physical /dev entries, i was referring to having devfs enabled
> and not mounting it.  Which according to the documentation should be
> perfectly functional and valid.  This is not the case though.   devfs
> should not require the old /dev entries at all since it doesn't use them
> so why would keeping them be required at all when using it (not counting
> the "if i want to not use devfs" argument).  This is what should be
> cleared up in the documentation.  

I believe you misunderstood the existing documentation. Nonetheless it
probably should be clarified, and I'm about to send a patch to the
mailing list.

-Barry K. Nathan <barryn@pobox.com>

Re: devfs

[Ed Sweetman]

On Sun, 2002-08-18 at 19:03, Alexander Viro wrote:
> 
> 
> On 18 Aug 2002, Ed Sweetman wrote:
> 
> > This has nothing to do with not mounting devfs and still using devfs to
> > work with devices.   If devfs is not mounted but you're still using
> > devfs, you shouldn't need anything in /dev.   The documentation says you
> > can use devfs without mounting and This is what i'm saying is
> > problematic and doesn't seem possible in normal usage.   It's an
> > optional config so are we using devfs when we dont mount it or not?  
> > and if not, then why make not mounting it an option ? 
> 
> What?  If program calls open("/dev/zero",...) and there's no such file,
> how the fuck would having devfs enabled help you?
> 
> Come on, use common sense - devfs provides a tree with some device nodes.
> You can mount it wherever you want (or not mount it anywhere).  Just as
> with any other filesystem.
> 
> If you mount it on /dev - well, duh, you see that tree on /dev.  If you
> do not - you see whatever is in /dev on underlying fs.
> 
> If program wants to access a device, it opens that device.  Just as any
> other file.  By name.  There is nothing magical about names that begin
> with /dev/ - it's just a conventional place for device nodes.
> 
> devfs "mount" option is an idiotic kludge that makes _kernel_ mount
> it on /dev after the root fs had been mounted.  Why it had been
> introduced is a great mistery, since the normal way is to have a
> corresponding line in /etc/fstab and have userland mount whatever
> it needs.
> 
> Said option is, indeed, not required for anything - in a sense that
> it does nothing that system wouldn't be perfectly capable of in regular
> ways.
> 
> But you _do_ need stuff in /dev, no matter what filesystem it comes
> from.  Kernel doesn't need it, but userland programs expect to find
> it there.  If you had deleted device nodes from underlying /dev and
> do not care to mount something on top of it - well, there won't be
> anything in that directory.
> 

Ok, that all makes sense.  I removed the dev entries because they
weren't needed by me anymore but I suppose it doesn't hurt anything to
just keep them there anyways so I've fixed all that.   Either way,
removing devfs did nothing but apparently it was asked to be done to
allow better testing and/or debugging to be done.  But i've yet to get
any reason why I removed devfs to investigate promise ide controller's
dma related memory failures.  I've removed devfs and replaced the old
/dev entries, no problem.  I'm not getting off topic about that.  It's
all done so i'm waiting for the next step here.  

Re: devfs

[Alexander Viro]

On 18 Aug 2002, Ed Sweetman wrote:

> This has nothing to do with not mounting devfs and still using devfs to
> work with devices.   If devfs is not mounted but you're still using
> devfs, you shouldn't need anything in /dev.   The documentation says you
> can use devfs without mounting and This is what i'm saying is
> problematic and doesn't seem possible in normal usage.   It's an
> optional config so are we using devfs when we dont mount it or not?  
> and if not, then why make not mounting it an option ? 

What?  If program calls open("/dev/zero",...) and there's no such file,
how the fuck would having devfs enabled help you?

Come on, use common sense - devfs provides a tree with some device nodes.
You can mount it wherever you want (or not mount it anywhere).  Just as
with any other filesystem.

If you mount it on /dev - well, duh, you see that tree on /dev.  If you
do not - you see whatever is in /dev on underlying fs.

If program wants to access a device, it opens that device.  Just as any
other file.  By name.  There is nothing magical about names that begin
with /dev/ - it's just a conventional place for device nodes.

devfs "mount" option is an idiotic kludge that makes _kernel_ mount
it on /dev after the root fs had been mounted.  Why it had been
introduced is a great mistery, since the normal way is to have a
corresponding line in /etc/fstab and have userland mount whatever
it needs.

Said option is, indeed, not required for anything - in a sense that
it does nothing that system wouldn't be perfectly capable of in regular
ways.

But you _do_ need stuff in /dev, no matter what filesystem it comes
from.  Kernel doesn't need it, but userland programs expect to find
it there.  If you had deleted device nodes from underlying /dev and
do not care to mount something on top of it - well, there won't be
anything in that directory.

Re: devfs

[Sean Neakums]

commence  Ed Sweetman quotation:

> On Sun, 2002-08-18 at 17:53, Barry K. Nathan wrote:
>> On Sun, Aug 18, 2002 at 07:36:47PM +0100, Sean Neakums wrote:
>> > commence  Ed Sweetman quotation:
>> [snip]
>> > > the devfs documentation says it doesn't need to have devfs mounted
>> > > to work, but this doesn't seem to be true at all.
>> > 
>> > If it does say exactly that, then it is outrageously wrong.
>> 
>> Starting at line 722 of
>> linux-2.4.19/Documentation/filesystems/devfs/README:
>> 
>> > In general, a kernel built with CONFIG_DEVFS_FS=y but without mounting
>> > devfs onto /dev is completely safe, and requires no
>> > configuration changes.
>> 
>> I skimmed through the documentation and it appears to assume that you're
>> not deleting all the stuff in /dev before switching over to devfs.
>
> This has nothing to do with not mounting devfs and still using devfs to
> work with devices.

If you don't mount devfs, you are not using it to "work with devices".
You use your existing device nodes, unless you deleted them, in which
case you are in trouble, as you discovered.

> If devfs is not mounted but you're still using devfs, you shouldn't
> need anything in /dev.

If devfs is not mounted, you are not "using it".

> The documentation says you can use devfs without mounting [...]

It does not.  It says that if devfs is built as part of your kernel's
configuration and you do not mount it, everything works as before.
For everythign to work as before, your device nodes need to be
present.

> and This is what i'm saying is problematic and doesn't seem possible
> in normal usage.  It's an optional config so are we using devfs when
> we dont mount it or not?  and if not, then why make not mounting it
> an option ?

I can imagine it being useful for for vendors.  Makes it easy to make
devfs usage a simple selection a install time without needsing to ship
two different kernels.

> If it's using the old device files in /dev then how can it be using
> devfs and how can accessing physical inodes on the disk be
> intentional to devfs?

Dunno what you mean here.

>> Right, there's no way around that. If you deleted everything in /dev --
>> which you're not supposed to do -- then there's no way for anything to
>> find any devices if devfs isn't enabled. (And you should have a rescue
>> CD around anyway -- you never know when you might need it! BTW, what
>> distribution are you (Ed) using? Some distributions have special boot
>> options you can use when booting their install CDs to get into a rescue
>> mode.)
>> 
>> In any event, it might be a good idea to make the documentation a bit
>> more explicit about this, and I might send a patch to the mailing
>> list later today.
>
> I'm not talking about booting without devfs enabled being the problem, i
> know booting without devfs enabled I'll have issues booting the system
> without physical /dev entries, i was referring to having devfs enabled
> and not mounting it.  Which according to the documentation should be
> perfectly functional and valid.  This is not the case though.

BECAUSE YOU DELETED YOUR DEVICE NODES.

> devfs should not require the old /dev entries at all since it
> doesn't use them so why would keeping them be required at all when
> using it (not counting the "if i want to not use devfs" argument).
> This is what should be cleared up in the documentation.

devfs does not require your old devices nodes.  Nowhere does the
documentation say that it does.  HOWEVER, if you want to boot a
devfs-capable kernel and not mount devfs (or simply boot a kernel with
no devfs capability at all), you have to have SOMETHING in /dev,
i.e. a set of standard device nodes.

-- 
 /                          |
[|] Sean Neakums            |  Questions are a burden to others;
[|] <sneakums@zork.net>     |      answers a prison for oneself.
 \                          |

devfs

[Ed Sweetman]

On Sun, 2002-08-18 at 17:53, Barry K. Nathan wrote:
> On Sun, Aug 18, 2002 at 07:36:47PM +0100, Sean Neakums wrote:
> > commence  Ed Sweetman quotation:
> [snip]
> > > the devfs documentation says it doesn't need to have devfs mounted
> > > to work, but this doesn't seem to be true at all.
> > 
> > If it does say exactly that, then it is outrageously wrong.
> 
> Starting at line 722 of
> linux-2.4.19/Documentation/filesystems/devfs/README:
> 
> > In general, a kernel built with CONFIG_DEVFS_FS=y but without mounting
> > devfs onto /dev is completely safe, and requires no
> > configuration changes.
> 
> I skimmed through the documentation and it appears to assume that you're
> not deleting all the stuff in /dev before switching over to devfs.

This has nothing to do with not mounting devfs and still using devfs to
work with devices.   If devfs is not mounted but you're still using
devfs, you shouldn't need anything in /dev.   The documentation says you
can use devfs without mounting and This is what i'm saying is
problematic and doesn't seem possible in normal usage.   It's an
optional config so are we using devfs when we dont mount it or not?  
and if not, then why make not mounting it an option ? 

If it's using the old device files in /dev then how can it be using
devfs and how can accessing physical inodes on the disk be intentional
to devfs? 


> Right, there's no way around that. If you deleted everything in /dev --
> which you're not supposed to do -- then there's no way for anything to
> find any devices if devfs isn't enabled. (And you should have a rescue
> CD around anyway -- you never know when you might need it! BTW, what
> distribution are you (Ed) using? Some distributions have special boot
> options you can use when booting their install CDs to get into a rescue
> mode.)
> 
> In any event, it might be a good idea to make the documentation a bit
> more explicit about this, and I might send a patch to the mailing
> list later today.

I'm not talking about booting without devfs enabled being the problem, i
know booting without devfs enabled I'll have issues booting the system
without physical /dev entries, i was referring to having devfs enabled
and not mounting it.  Which according to the documentation should be
perfectly functional and valid.  This is not the case though.   devfs
should not require the old /dev entries at all since it doesn't use them
so why would keeping them be required at all when using it (not counting
the "if i want to not use devfs" argument).  This is what should be
cleared up in the documentation.