sysfs question

sysfs question

[Kurt Van Dijck]

Hi,

Within the socketcan project, we came into a situation that
might benefit with input from the netdev mailing list.

The main issue is the policy to add sysfs properties in
/sys/class/net/canX.

The reason for this is that cards (devices) with multiple network
interfaces may require properties per network.

An obvious property would be the 'channel number of the card'. Other
properties could be things like type of output circuitry, ..., rather
hardware specific.

I see currently 3 options:
1) such properties in /sys/class/net/canX would be allowed.
2) such properties would belong in /sys/class/net/canX/<subdirectory tbd>/
3) such properties would go somewhere else.

Some input with regard to sysfs policies would be helpfull.

Kurt

Re: sysfs question

[Stephen Hemminger]

On Thu, 12 Nov 2009 11:39:41 +0100
Kurt Van Dijck <kurt.van.dijck@eia.be> wrote:

> Hi,
> 
> Within the socketcan project, we came into a situation that
> might benefit with input from the netdev mailing list.
> 
> The main issue is the policy to add sysfs properties in
> /sys/class/net/canX.
> 
> The reason for this is that cards (devices) with multiple network
> interfaces may require properties per network.
> 
> An obvious property would be the 'channel number of the card'. Other
> properties could be things like type of output circuitry, ..., rather
> hardware specific.
> 
> I see currently 3 options:
> 1) such properties in /sys/class/net/canX would be allowed.
> 2) such properties would belong in /sys/class/net/canX/<subdirectory tbd>/
> 3) such properties would go somewhere else.
> 
> Some input with regard to sysfs policies would be helpfull.
> 
> Kurt

It sounds like the property you are proposing is a property of the
upper network layer not the hardware. Putting more properties in sysfs
is good if is hardware related, but awkward if it is really a protocol
attribute.

Re: sysfs question

[Oliver Hartkopp]

Stephen Hemminger wrote:
> On Thu, 12 Nov 2009 11:39:41 +0100
> Kurt Van Dijck <kurt.van.dijck@eia.be> wrote:
> 
>> Hi,
>>
>> Within the socketcan project, we came into a situation that
>> might benefit with input from the netdev mailing list.
>>
>> The main issue is the policy to add sysfs properties in
>> /sys/class/net/canX.
>>
>> The reason for this is that cards (devices) with multiple network
>> interfaces may require properties per network.
>>
>> An obvious property would be the 'channel number of the card'. Other
>> properties could be things like type of output circuitry, ..., rather
>> hardware specific.
>>
>> I see currently 3 options:
>> 1) such properties in /sys/class/net/canX would be allowed.
>> 2) such properties would belong in /sys/class/net/canX/<subdirectory tbd>/
>> 3) such properties would go somewhere else.
>>
>> Some input with regard to sysfs policies would be helpfull.
>>
>> Kurt
> 
> It sounds like the property you are proposing is a property of the
> upper network layer not the hardware. Putting more properties in sysfs
> is good if is hardware related, but awkward if it is really a protocol
> attribute.

Then it would be good here :-)

The problem that can occur here is:

You plug-in a PCMCIA CAN card that has two CAN controllers on-board.

That means, you're getting two CAN independed network interfaces can0 and can1
pointing to ONE pcmcia device. Both of the CAN netdevs may have different
bitrates or other CAN controller specific settings (like the type of output
circuitry Kurt mentioned).

Finally the PCMCIA card itself may contain a serial number and/or a license
information that needs to be made available.

It's definitely no upper network layer stuff.

Regards,
Oliver

Re: sysfs question

[Kurt Van Dijck]

On Thu, Nov 12, 2009 at 06:44:36PM +0100, Oliver Hartkopp wrote:
> Stephen Hemminger wrote:
> > On Thu, 12 Nov 2009 11:39:41 +0100
> > Kurt Van Dijck <kurt.van.dijck@eia.be> wrote:
> > 
[...]
> > 
> > It sounds like the property you are proposing is a property of the
> > upper network layer not the hardware. Putting more properties in sysfs
> > is good if is hardware related, but awkward if it is really a protocol
> > attribute.
> 
> Then it would be good here :-)
> 
[...]
> 
> It's definitely no upper network layer stuff.
> 
Ack. It's definitely datalink or even physical layer we're talking about.
> Regards,
> Oliver

I also encountered an issue with regard to the uevent that is generated
when a netdevice is registered.
If I add sysfs files before register_netdevice, the system complains on
creating the sysfs files.
If I add those after register_netdevice, the uevent is _seems_ to have
triggered userspace already, where the udev (in fact, I use another
home-brew one for boot speed) does not find the extra sysfs file.

Is there a mechanism to hold temporarily the uevent until after the
driver has registered some extra sysfs files?

Kurt

Re: sysfs question

[Daniel Lezcano]

xiaoming.zhang wrote:
> On Thursday 11 September 2008 11:37:33 pm Daniel Lezcano wrote:
>> xiaoming.zhang wrote:
>>> On Thursday 11 September 2008 10:05:43 pm Mark Ryden wrote:
>>>> In fact, after looking more closely, I see that before that message
>>>> appears: Error: unmouting old /sys
>>>> ERROR unmounting old /sys: Invalid argument.
>>>> forcing unmount of /sys
>>>> switchroot...
>>> My experience is you'd better to find a sysfs patch, since the sysfs
>>> filesystem is required by ramdisk and some system utilities, e.g., system
>>> network configuration scripts.
>>>
>>> Patch for mm-tree can be found here:
>>> 	http://lxc.sourceforge.net/patches/2.6.26/
>>>
>>> This is what I used:
>>> http://lxc.sourceforge.net/patches/2.6.26/2.6.26-rc8-mm1-lxc1/broken-out/
>>> sysfs/
>>>
>>> I merged these 11 patches on Linux-2.6.26.1 though I don't understand
>>> these patches, luckily enough it worked for me.
>> As mentioned Benjamin, it is hard to boot a distro without sysfs. I
>> think it is a good idea to use the lxc patchset.
>>
>> You can use the latest version:
>>
>> http://lxc.sourceforge.net/patches/2.6.27/2.6.27-rc6-lxc1/
>>
> Thank you very much. And is there any patchset for Linux-2.6.26.* (not the 
> mm-tree)?

We worked most of the time sticked with the devel version of the kernel 
tree. The kernel which is the nearest of the version you are describing 
is 2.6.27-rc6-lxc1 which is not based on the -mm but on the 2.6.27 
release candidate. IMO, this is the most stable version of the LXC.

Re: sysfs question

[xiaoming.zhang]

On Thursday 11 September 2008 11:37:33 pm Daniel Lezcano wrote:
> xiaoming.zhang wrote:
> > On Thursday 11 September 2008 10:05:43 pm Mark Ryden wrote:
> >> In fact, after looking more closely, I see that before that message
> >> appears: Error: unmouting old /sys
> >> ERROR unmounting old /sys: Invalid argument.
> >> forcing unmount of /sys
> >> switchroot...
> >
> > My experience is you'd better to find a sysfs patch, since the sysfs
> > filesystem is required by ramdisk and some system utilities, e.g., system
> > network configuration scripts.
> >
> > Patch for mm-tree can be found here:
> > 	http://lxc.sourceforge.net/patches/2.6.26/
> >
> > This is what I used:
> > http://lxc.sourceforge.net/patches/2.6.26/2.6.26-rc8-mm1-lxc1/broken-out/
> >sysfs/
> >
> > I merged these 11 patches on Linux-2.6.26.1 though I don't understand
> > these patches, luckily enough it worked for me.
>
> As mentioned Benjamin, it is hard to boot a distro without sysfs. I
> think it is a good idea to use the lxc patchset.
>
> You can use the latest version:
>
> http://lxc.sourceforge.net/patches/2.6.27/2.6.27-rc6-lxc1/
>
Thank you very much. And is there any patchset for Linux-2.6.26.* (not the 
mm-tree)?


> If you want to play with the container, you have the userspace tools at:
>
> http://lxc.sourceforge.net/download/lxc/
>
> Don't forget to check the README in the directory ;)
>
>    -- Daniel



-- 
Have fun,
Xiaoming Zhang

Re: sysfs question

[Daniel Lezcano]

xiaoming.zhang wrote:
> On Thursday 11 September 2008 10:05:43 pm Mark Ryden wrote:
>> In fact, after looking more closely, I see that before that message
>> appears: Error: unmouting old /sys
>> ERROR unmounting old /sys: Invalid argument.
>> forcing unmount of /sys
>> switchroot...
>>
> 
> My experience is you'd better to find a sysfs patch, since the sysfs 
> filesystem is required by ramdisk and some system utilities, e.g., system 
> network configuration scripts.
> 
> Patch for mm-tree can be found here:
> 	http://lxc.sourceforge.net/patches/2.6.26/
> 
> This is what I used:
> http://lxc.sourceforge.net/patches/2.6.26/2.6.26-rc8-mm1-lxc1/broken-out/sysfs/
> 
> I merged these 11 patches on Linux-2.6.26.1 though I don't understand these 
> patches, luckily enough it worked for me.

As mentioned Benjamin, it is hard to boot a distro without sysfs. I 
think it is a good idea to use the lxc patchset.

You can use the latest version:

http://lxc.sourceforge.net/patches/2.6.27/2.6.27-rc6-lxc1/

If you want to play with the container, you have the userspace tools at:

http://lxc.sourceforge.net/download/lxc/

Don't forget to check the README in the directory ;)

   -- Daniel

Re: sysfs question

[Benjamin Thery]

Hi Mark,

Indeed booting a distro without sysfs can be tricky because, for
example, initrd might need it (eg. Fedora).

IIRC, to boot without sysfs, in addition to the "root=" option,  you
have to create some devices nodes in advance in your /dev tree
(mainly the /dev/sda* ones, plus a bunch of others /dev/urandom,
/dev/null, /dev/console, /dev/tty).

mknod /dev/sda b 8 0
mknod /dev/sda1 b 8 1
...

But my advice to test netns is to apply the tagged directories sysfs
patchses. It is a lot easier (and complete).

Gregkh is merging the patchset to his tree today (he has added 6 out of
the 8 remaining patches this morning) :)
Once the patchset is completely merged I'll update the howto.

Also I can send you tarball too if you want. Which kernel version do you 
use?

Regards,
Benjamin


Mark Ryden wrote:
> Hello,
> I am trying to follow the
> Linux Containers - Network Namespace configuration instructions from
> http://lxc.sourceforge.net/network/configuration.php
> 
> 
> I had built a kernel with CONFIG_SYS=n and  CONFIG_NET_NS=y
> and CONFIG_VETH=y , etc, according to the instructions.
> 
> 
> Now, the CONFIG_SYSFS help text says:
> 
> sysfs is currently used by the block subsystem to mount the root
> partition. If sysfs is disabled you must specify the boot device on
> the kernel boot command line via its major and minor numbers. For
> example, "root=03:01" for /dev/hda1.
> 
> 
> So I added root=08,02 to my kernel boot parameter line and booted.
> 
> 
> it starts booting, but then gets:
> switchroot: mount failed: No such file or directory.
> Booting has failed.
> 
> 8 is the majoror number and 2 is the minor number of /dev/sda2, where I have
> my root partition.
> 
> Googling for this found the following, which not helped much:
> http://linux.derkeiler.com/Mailing-Lists/Kernel/2007-11/msg05032.html
> 
> Any ideas?
> Regards,
> Mark
> _______________________________________________
> Containers mailing list
> Containers-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org
> https://lists.linux-foundation.org/mailman/listinfo/containers
> 
> 


-- 
B e n j a m i n   T h e r y  - BULL/DT/Open Software R&D

    http://www.bull.com

Re: sysfs question

[xiaoming.zhang]

On Thursday 11 September 2008 10:05:43 pm Mark Ryden wrote:
> In fact, after looking more closely, I see that before that message
> appears: Error: unmouting old /sys
> ERROR unmounting old /sys: Invalid argument.
> forcing unmount of /sys
> switchroot...
>

My experience is you'd better to find a sysfs patch, since the sysfs 
filesystem is required by ramdisk and some system utilities, e.g., system 
network configuration scripts.

Patch for mm-tree can be found here:
	http://lxc.sourceforge.net/patches/2.6.26/

This is what I used:
http://lxc.sourceforge.net/patches/2.6.26/2.6.26-rc8-mm1-lxc1/broken-out/sysfs/

I merged these 11 patches on Linux-2.6.26.1 though I don't understand these 
patches, luckily enough it worked for me.

> Any ideas?
> Rgs,
> Mark
>
> On Thu, Sep 11, 2008 at 5:01 PM, Mark Ryden <markryde-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
> > Hello,
> > I am trying to follow the
> > Linux Containers - Network Namespace configuration instructions from
> > http://lxc.sourceforge.net/network/configuration.php
> >
> >
> > I had built a kernel with CONFIG_SYS=n and  CONFIG_NET_NS=y
> > and CONFIG_VETH=y , etc, according to the instructions.
> >
> >
> > Now, the CONFIG_SYSFS help text says:
> >
> > sysfs is currently used by the block subsystem to mount the root
> > partition. If sysfs is disabled you must specify the boot device on
> > the kernel boot command line via its major and minor numbers. For
> > example, "root=03:01" for /dev/hda1.
> >
> >
> > So I added root=08,02 to my kernel boot parameter line and booted.
> >
> >
> > it starts booting, but then gets:
> > switchroot: mount failed: No such file or directory.
> > Booting has failed.
> >
> > 8 is the majoror number and 2 is the minor number of /dev/sda2, where I
> > have my root partition.
> >
> > Googling for this found the following, which not helped much:
> > http://linux.derkeiler.com/Mailing-Lists/Kernel/2007-11/msg05032.html
> >
> > Any ideas?
> > Regards,
> > Mark
>
> _______________________________________________
> Containers mailing list
> Containers-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org
> https://lists.linux-foundation.org/mailman/listinfo/containers



-- 
Have fun,
Xiaoming Zhang

Re: sysfs question

[Mark Ryden]

In fact, after looking more closely, I see that before that message appears:
Error: unmouting old /sys
ERROR unmounting old /sys: Invalid argument.
forcing unmount of /sys
switchroot...

Any ideas?
Rgs,
Mark


On Thu, Sep 11, 2008 at 5:01 PM, Mark Ryden <markryde-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
> Hello,
> I am trying to follow the
> Linux Containers - Network Namespace configuration instructions from
> http://lxc.sourceforge.net/network/configuration.php
>
>
> I had built a kernel with CONFIG_SYS=n and  CONFIG_NET_NS=y
> and CONFIG_VETH=y , etc, according to the instructions.
>
>
> Now, the CONFIG_SYSFS help text says:
>
> sysfs is currently used by the block subsystem to mount the root
> partition. If sysfs is disabled you must specify the boot device on
> the kernel boot command line via its major and minor numbers. For
> example, "root=03:01" for /dev/hda1.
>
>
> So I added root=08,02 to my kernel boot parameter line and booted.
>
>
> it starts booting, but then gets:
> switchroot: mount failed: No such file or directory.
> Booting has failed.
>
> 8 is the majoror number and 2 is the minor number of /dev/sda2, where I have
> my root partition.
>
> Googling for this found the following, which not helped much:
> http://linux.derkeiler.com/Mailing-Lists/Kernel/2007-11/msg05032.html
>
> Any ideas?
> Regards,
> Mark
>

sysfs question

[Mark Ryden]

Hello,
I am trying to follow the
Linux Containers - Network Namespace configuration instructions from
http://lxc.sourceforge.net/network/configuration.php


I had built a kernel with CONFIG_SYS=n and  CONFIG_NET_NS=y
and CONFIG_VETH=y , etc, according to the instructions.


Now, the CONFIG_SYSFS help text says:

sysfs is currently used by the block subsystem to mount the root
partition. If sysfs is disabled you must specify the boot device on
the kernel boot command line via its major and minor numbers. For
example, "root=03:01" for /dev/hda1.


So I added root=08,02 to my kernel boot parameter line and booted.


it starts booting, but then gets:
switchroot: mount failed: No such file or directory.
Booting has failed.

8 is the majoror number and 2 is the minor number of /dev/sda2, where I have
my root partition.

Googling for this found the following, which not helped much:
http://linux.derkeiler.com/Mailing-Lists/Kernel/2007-11/msg05032.html

Any ideas?
Regards,
Mark

Re: sysfs question

[Luca]

Il Mon, Jul 24, 2006 at 09:47:59AM -0400, Dennis Munsie ha scritto: 
> quick question regarding sysfs and the framebuffer... I can add my  
> own attributes without any problems, but they always show up with the  
> device (/sys/class/graphics/fb0/device), not the rest of the  
> framebuffer attributes (/sys/class/graphics/fb0). Is there a good  
> way to put them with the fb attributes?

You have to attach the new attributes to the class_device instead of the
device. See class_device_create_file().
I don't know if it's a good idea tough: if the attributes are specific
to only one device it sounds more logical to attach them to the device.

Luca
-- 
Home: http://kronoz.cjb.net
"L'ottimista pensa che questo sia il migliore dei mondi possibili. 
 Il pessimista sa che e` vero" -- J. Robert Oppenheimer

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV

sysfs question

[Dennis Munsie]

quick question regarding sysfs and the framebuffer... I can add my  
own attributes without any problems, but they always show up with the  
device (/sys/class/graphics/fb0/device), not the rest of the  
framebuffer attributes (/sys/class/graphics/fb0).  Is there a good  
way to put them with the fb attributes?  or is this a no-no?

thanks!
dennis

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV

sysfs question

[JaniD++]

Hello list,

I have searching the bottleneck of my system, and this is important to know
what is the 4. and the 8. number from this files exactly:

/sys/block/*/stat

Thanks,

Janos