Skip to Content.
Sympa Menu

sm-discuss - Re: [SM-Discuss] eudev

sm-discuss AT lists.ibiblio.org

Subject: Public SourceMage Discussion List

List archive

Chronological Thread  
  • From: Treeve Jelbert <treeve AT scarlet.be>
  • To: "sm, discuss" <sm-discuss AT lists.ibiblio.org>
  • Subject: Re: [SM-Discuss] eudev
  • Date: Wed, 31 Jul 2013 12:13 +0200

On Tuesday 30 July 2013 17:38:00 you wrote:
> On Mon, Jul 29, 2013 at 6:30 AM, Treeve Jelbert <treeve AT scarlet.be> wrote:
> > ...
> >
> > The only strange thifindm ng is that /dev/pts and /dev/shm did not get
created.
>
> Would this require some additional init scripts to mount those devs?

maybe
>
> > ...
> > I will then add this spell as an alternative to udev.
>
> It reached test already, :-)
>
> I believe udev 182 conflicts with module-init-tools. And according to
> the eudev spell, it's a fork of udev 195. That leads me to believe
> eudev also conflicts with module-init-tools as well. Therefore it
> also requires basesystem 0.9.6, whereas the current test one is 0.9.5.
>
> Is that correct?
I don't think so, as eudev has its own internal version of kmod and does not
need a separate module-init-tools or kmod in order to load modules.
In this case module-init-tools or kmod merely provide lsmod, depmod and
friends. modprobe is not needed by eudev itself, although it can be used
manually.
Linking to an external kmod was dropped by the eudev team as a bad idea,
although it can optionally be enabled during configure.

# ldd /sbin/udevd
linux-vdso.so.1 (0x00007fff372e0000)
libudev.so.1 => /lib/libudev.so.1 (0x00007fcf798f8000)
libblkid.so.1 => /lib/libblkid.so.1 (0x00007fcf796b0000)
libc.so.6 => /lib/libc.so.6 (0x00007fcf792f0000)
libm.so.6 => /lib/libm.so.6 (0x00007fcf79068000)
/lib/ld-linux-x86-64.so.2 (0x00007fcf79b20000)
libuuid.so.1 => /lib/libuuid.so.1 (0x00007fcf78e60000)

In fact I have disabled kmod on one of my systems and all my devices appear,
except for /dev/pts, even though /etc/init.d/runlevels/%DEV/devices (from
init.d) runs before udevd. I suspect some sort of race condition related to
the kernel constructing an internal device tree, which is later copied to the
/dev devtmpfs

# findmnt /dev
TARGET SOURCE FSTYPE OPTIONS

/dev devtmpfs devtmpfs rw,relatime,size=3786908k,nr_inodes=946727,mode=755


# cat /etc/fstab
proc /proc proc defaults 0 0
devpts /dev/pts devpts defaults 0 0
sysfs /sys sysfs defaults 0 0


Perhaps the /dev/pts is overwritten during the startup phase?
I am trying to get some more debug info on this.


One of my machines has kmod-8, the other kmod-14
>
> Thanks,
>
> Javier.





Archive powered by MHonArc 2.6.24.

Top of Page