[gentoo-user] Making a init thingy. Step two I guess.

[gentoo-user] Making a init thingy. Step two I guess.

[Dale]

OK.  Sort of dealt with the LVM part so now to step two.  I'm trying to 
follow this:

http://en.gentoo-wiki.com/wiki/Initramfs

Then I get confused.  I get to Applications and I'm sort of lost here.  
In there it talks about copying nano and its friends over to the init 
directory.  Then below that it says to use busybox.  Well, which is it?  
Do I do both of those or just one?

I also tried to figure out Dracut but it seems like a work in progress 
or something.  I can't seem to find any Gentoo docs for it.  I found a 
Redhat doc but that's not exactly Gentoo.

Am I starting to work on this to soon?  Maybe wait a while longer until 
some more stuff gets sorted out?

Dang, I'm full of questions today.  ;-)

Dale

:-)  :-)

Re: [gentoo-user] Making a init thingy. Step two I guess.

[Mark Knecht]

On Sat, Sep 17, 2011 at 1:02 PM, Dale <rdalek1967@gmail.com> wrote:
> OK.  Sort of dealt with the LVM part so now to step two.  I'm trying to
> follow this:
>
> http://en.gentoo-wiki.com/wiki/Initramfs
>
> Then I get confused.  I get to Applications and I'm sort of lost here.  In
> there it talks about copying nano and its friends over to the init
> directory.  Then below that it says to use busybox.  Well, which is it?  Do
> I do both of those or just one?
>

It's been a while for me but I believe it's both. I think busybox is
the thing that gives you command line tools like cd, ls, pwd, etc.
However you also can include applications in your initramfs that give
you more access to the hardware or the net.


> I also tried to figure out Dracut but it seems like a work in progress or
> something.  I can't seem to find any Gentoo docs for it.  I found a Redhat
> doc but that's not exactly Gentoo.
>
> Am I starting to work on this to soon?  Maybe wait a while longer until some
> more stuff gets sorted out?
>
> Dang, I'm full of questions today.  ;-)
>
> Dale

I'm just reading some Dracut stuff myself. I figured I might as well
get educated. I likely won't actually build it and try it for a few
days at the soonest, but it's good to get prepared. Cheers to you for
doing that.

- Mark

Re: [gentoo-user] Making a init thingy. Step two I guess.

[Dale]

Mark Knecht wrote:
> On Sat, Sep 17, 2011 at 1:02 PM, Dale<rdalek1967@gmail.com>  wrote:
>> OK.  Sort of dealt with the LVM part so now to step two.  I'm trying to
>> follow this:
>>
>> http://en.gentoo-wiki.com/wiki/Initramfs
>>
>> Then I get confused.  I get to Applications and I'm sort of lost here.  In
>> there it talks about copying nano and its friends over to the init
>> directory.  Then below that it says to use busybox.  Well, which is it?  Do
>> I do both of those or just one?
>>
> It's been a while for me but I believe it's both. I think busybox is
> the thing that gives you command line tools like cd, ls, pwd, etc.
> However you also can include applications in your initramfs that give
> you more access to the hardware or the net.

That's what I was thinking to.  I know busybox does a LOT of things but 
I didn't think it was a nano replacement either.  OK.  I got both of 
those in there at least.  First problem dealt with I guess.  Wonder what 
else I should put in there too. < scratches head >

>
>> I also tried to figure out Dracut but it seems like a work in progress or
>> something.  I can't seem to find any Gentoo docs for it.  I found a Redhat
>> doc but that's not exactly Gentoo.
>>
>> Am I starting to work on this to soon?  Maybe wait a while longer until some
>> more stuff gets sorted out?
>>
>> Dang, I'm full of questions today.  ;-)
>>
>> Dale
> I'm just reading some Dracut stuff myself. I figured I might as well
> get educated. I likely won't actually build it and try it for a few
> days at the soonest, but it's good to get prepared. Cheers to you for
> doing that.
>
> - Mark
>

Yea, it's either learn this or find a new distro.  Still haven't ruled 
out the later tho.  Just going to try at least.

Dale

:-)  :-)

Re: [gentoo-user] Making a init thingy. Step two I guess.

[Michael Mol]

On Sat, Sep 17, 2011 at 4:44 PM, Dale <rdalek1967@gmail.com> wrote:
> Mark Knecht wrote:
>> On Sat, Sep 17, 2011 at 1:02 PM, Dale<rdalek1967@gmail.com>  wrote:
> That's what I was thinking to.  I know busybox does a LOT of things but I
> didn't think it was a nano replacement either.  OK.  I got both of those in
> there at least.  First problem dealt with I guess.  Wonder what else I
> should put in there too. < scratches head >

Anything that makes life convenient, I'd guess.

I'd look at stuffing these in there:
* screen
* file
* ssh
* links/lynx/w3m

You'd also need the .so files those depend on. I don't know how to
identify those.

-- 
:wq

Re: [gentoo-user] Making a init thingy. Step two I guess.

[Mark Knecht]

On Sat, Sep 17, 2011 at 1:51 PM, Michael Mol <mikemol@gmail.com> wrote:
> On Sat, Sep 17, 2011 at 4:44 PM, Dale <rdalek1967@gmail.com> wrote:
>> Mark Knecht wrote:
>>> On Sat, Sep 17, 2011 at 1:02 PM, Dale<rdalek1967@gmail.com>  wrote:
>> That's what I was thinking to.  I know busybox does a LOT of things but I
>> didn't think it was a nano replacement either.  OK.  I got both of those in
>> there at least.  First problem dealt with I guess.  Wonder what else I
>> should put in there too. < scratches head >
>
> Anything that makes life convenient, I'd guess.
>
> I'd look at stuffing these in there:
> * screen
> * file
> * ssh
> * links/lynx/w3m
>
> You'd also need the .so files those depend on. I don't know how to
> identify those.
>
> --
> :wq
>
>

Supposedly

ldd /usr/bin/screen
ldd /usr/bin/file
ldd /usr/bin/ssh

etc.

I think that's what I did when I built one a while back...

- Mark

Re: [gentoo-user] Making a init thingy. Step two I guess.

[Dale]

Mark Knecht wrote:
> On Sat, Sep 17, 2011 at 1:51 PM, Michael Mol<mikemol@gmail.com>  wrote:
>> On Sat, Sep 17, 2011 at 4:44 PM, Dale<rdalek1967@gmail.com>  wrote:
>>> Mark Knecht wrote:
>>>> On Sat, Sep 17, 2011 at 1:02 PM, Dale<rdalek1967@gmail.com>    wrote:
>>> That's what I was thinking to.  I know busybox does a LOT of things but I
>>> didn't think it was a nano replacement either.  OK.  I got both of those in
>>> there at least.  First problem dealt with I guess.  Wonder what else I
>>> should put in there too.<  scratches head>
>> Anything that makes life convenient, I'd guess.
>>
>> I'd look at stuffing these in there:
>> * screen
>> * file
>> * ssh
>> * links/lynx/w3m
>>
>> You'd also need the .so files those depend on. I don't know how to
>> identify those.
>>
>> --
>> :wq
>>
>>
> Supposedly
>
> ldd /usr/bin/screen
> ldd /usr/bin/file
> ldd /usr/bin/ssh
>
> etc.
>
> I think that's what I did when I built one a while back...
>
> - Mark
>
>

Yep, that's how the howto says it too.  It seems to work.

root@fireball / # ldd /bin/nano
         linux-vdso.so.1 =>  (0x00007fffc25ff000)
         libncursesw.so.5 => /lib64/libncursesw.so.5 (0x00007f25fb500000)
         libc.so.6 => /lib64/libc.so.6 (0x00007f25fb171000)
         libdl.so.2 => /lib64/libdl.so.2 (0x00007f25faf6d000)
         /lib64/ld-linux-x86-64.so.2 (0x00007f25fb75f000)
root@fireball / #

So, I copied those over too.

Adding ssh and screen are good ideas.  I very rarely use file.  Still 
might tho. Interesting:

root@fireball / # ldd /usr/bin/screen
         linux-vdso.so.1 =>  (0x00007fffe1573000)
         libcrypt.so.1 => /lib64/libcrypt.so.1 (0x00007f921dfca000)
         libncursesw.so.5 => /lib64/libncursesw.so.5 (0x00007f921dd6b000)
         libpam.so.0 => /lib64/libpam.so.0 (0x00007f921db5c000)
         libc.so.6 => /lib64/libc.so.6 (0x00007f921d7cd000)
         libdl.so.2 => /lib64/libdl.so.2 (0x00007f921d5c9000)
         /lib64/ld-linux-x86-64.so.2 (0x00007f921e230000)
root@fireball / # ldd /usr/bin/ssh
         linux-vdso.so.1 =>  (0x00007fffc6bff000)
         libcrypto.so.1.0.0 => /usr/lib64/libcrypto.so.1.0.0 
(0x00007ff6aa7b9000)
         libdl.so.2 => /lib64/libdl.so.2 (0x00007ff6aa5b5000)
         libz.so.1 => /lib64/libz.so.1 (0x00007ff6aa39d000)
         libresolv.so.2 => /lib64/libresolv.so.2 (0x00007ff6aa186000)
         libpthread.so.0 => /lib64/libpthread.so.0 (0x00007ff6a9f69000)
         libc.so.6 => /lib64/libc.so.6 (0x00007ff6a9bda000)
         /lib64/ld-linux-x86-64.so.2 (0x00007ff6aab73000)
root@fireball / #

Those seem to have a bit in common and no static flag for either.

Dale

:-)  :-)

Re: [gentoo-user] Making a init thingy. Step two I guess.

[Mark Knecht]

On Sat, Sep 17, 2011 at 2:17 PM, Dale <rdalek1967@gmail.com> wrote:
> Mark Knecht wrote:
>>
>> On Sat, Sep 17, 2011 at 1:51 PM, Michael Mol<mikemol@gmail.com>  wrote:
>>>
>>> On Sat, Sep 17, 2011 at 4:44 PM, Dale<rdalek1967@gmail.com>  wrote:
>>>>
>>>> Mark Knecht wrote:
>>>>>
>>>>> On Sat, Sep 17, 2011 at 1:02 PM, Dale<rdalek1967@gmail.com>    wrote:
>>>>
>>>> That's what I was thinking to.  I know busybox does a LOT of things but
>>>> I
>>>> didn't think it was a nano replacement either.  OK.  I got both of those
>>>> in
>>>> there at least.  First problem dealt with I guess.  Wonder what else I
>>>> should put in there too.<  scratches head>
>>>
>>> Anything that makes life convenient, I'd guess.
>>>
>>> I'd look at stuffing these in there:
>>> * screen
>>> * file
>>> * ssh
>>> * links/lynx/w3m
>>>
>>> You'd also need the .so files those depend on. I don't know how to
>>> identify those.
>>>
>>> --
>>> :wq
>>>
>>>
>> Supposedly
>>
>> ldd /usr/bin/screen
>> ldd /usr/bin/file
>> ldd /usr/bin/ssh
>>
>> etc.
>>
>> I think that's what I did when I built one a while back...
>>
>> - Mark
>>
>>
>
> Yep, that's how the howto says it too.  It seems to work.
>
> root@fireball / # ldd /bin/nano
>        linux-vdso.so.1 =>  (0x00007fffc25ff000)
>        libncursesw.so.5 => /lib64/libncursesw.so.5 (0x00007f25fb500000)
>        libc.so.6 => /lib64/libc.so.6 (0x00007f25fb171000)
>        libdl.so.2 => /lib64/libdl.so.2 (0x00007f25faf6d000)
>        /lib64/ld-linux-x86-64.so.2 (0x00007f25fb75f000)
> root@fireball / #
>
> So, I copied those over too.
>
> Adding ssh and screen are good ideas.  I very rarely use file.  Still might
> tho. Interesting:
>
> root@fireball / # ldd /usr/bin/screen
>        linux-vdso.so.1 =>  (0x00007fffe1573000)
>        libcrypt.so.1 => /lib64/libcrypt.so.1 (0x00007f921dfca000)
>        libncursesw.so.5 => /lib64/libncursesw.so.5 (0x00007f921dd6b000)
>        libpam.so.0 => /lib64/libpam.so.0 (0x00007f921db5c000)
>        libc.so.6 => /lib64/libc.so.6 (0x00007f921d7cd000)
>        libdl.so.2 => /lib64/libdl.so.2 (0x00007f921d5c9000)
>        /lib64/ld-linux-x86-64.so.2 (0x00007f921e230000)
> root@fireball / # ldd /usr/bin/ssh
>        linux-vdso.so.1 =>  (0x00007fffc6bff000)
>        libcrypto.so.1.0.0 => /usr/lib64/libcrypto.so.1.0.0
> (0x00007ff6aa7b9000)
>        libdl.so.2 => /lib64/libdl.so.2 (0x00007ff6aa5b5000)
>        libz.so.1 => /lib64/libz.so.1 (0x00007ff6aa39d000)
>        libresolv.so.2 => /lib64/libresolv.so.2 (0x00007ff6aa186000)
>        libpthread.so.0 => /lib64/libpthread.so.0 (0x00007ff6a9f69000)
>        libc.so.6 => /lib64/libc.so.6 (0x00007ff6a9bda000)
>        /lib64/ld-linux-x86-64.so.2 (0x00007ff6aab73000)
> root@fireball / #
>
> Those seem to have a bit in common and no static flag for either.
>
> Dale

While I was out walking my dog I sort of remembered that there are
just a few apps I had to build static, or at least that I build
static. One, I think was grub. The first and only time I did it I had
to do it 3 or 4 times before I got everything I needed working
correctly. This page which is like the first one Google comes up with
talks about that stuff, although you have to sort of dig it out and
read between the lines.

http://en.gentoo-wiki.com/wiki/Initramfs

And, I think with the ldd command as long as you include the libraries
ldd points at in your equivalent /lib directory within the initramfs
then that works for most apps and keeps the initramfs smaller.

HTH,
Mark

Re: [gentoo-user] Making a init thingy. Step two I guess.

[Dale]

Mark Knecht wrote:
> While I was out walking my dog I sort of remembered that there are
> just a few apps I had to build static, or at least that I build
> static. One, I think was grub. The first and only time I did it I had
> to do it 3 or 4 times before I got everything I needed working
> correctly. This page which is like the first one Google comes up with
> talks about that stuff, although you have to sort of dig it out and
> read between the lines.
>
> http://en.gentoo-wiki.com/wiki/Initramfs
>
> And, I think with the ldd command as long as you include the libraries
> ldd points at in your equivalent /lib directory within the initramfs
> then that works for most apps and keeps the initramfs smaller.
>
> HTH,
> Mark
>
>

I think there are only a few that has that flag, at least that I would 
put in the init thingy anyway.  Maybe this is something that the devs 
will work on if it can be done.  May be a big if there.

That is the guide I am trying to go by but I think I am missing 
something.  This is the script they have posted:

#!/bin/busybox sh

# Mount the /proc and /sys filesystems.
mount -t proc none /proc
mount -t sysfs none /sys

# Do your stuff here.
echo "This script mounts rootfs and boots it up, nothing more!"

# Mount the root filesystem.
mount -o ro /dev/sda1 /mnt/root

# Clean up.
umount /proc
umount /sys

# Boot the real thing.
exec switch_root /mnt/root /sbin/init

That doesn't really make much sense to me.  First it mounts the stuff then umounts it right after that.  Huh?  Is the relevant part the "mount -o ro /dev/sda1 /mnt/root" ?  Then the exec switch_root part after that?  The rest seems to cancel each other out.

Looking forward to that light bulb moment here.  ;-)

Dale

:-)  :-)

Re: [gentoo-user] Making a init thingy. Step two I guess.

[Michael Mol]

On Sat, Sep 17, 2011 at 6:27 PM, Dale <rdalek1967@gmail.com> wrote:
> Mark Knecht wrote:
> I think there are only a few that has that flag, at least that I would put
> in the init thingy anyway.  Maybe this is something that the devs will work
> on if it can be done.  May be a big if there.
>
> That is the guide I am trying to go by but I think I am missing something.
>  This is the script they have posted:
>
> #!/bin/busybox sh
>
> # Mount the /proc and /sys filesystems.
> mount -t proc none /proc
> mount -t sysfs none /sys
>
> # Do your stuff here.
> echo "This script mounts rootfs and boots it up, nothing more!"
>
> # Mount the root filesystem.
> mount -o ro /dev/sda1 /mnt/root
>
> # Clean up.
> umount /proc
> umount /sys
>
> # Boot the real thing.
> exec switch_root /mnt/root /sbin/init
>
> That doesn't really make much sense to me.  First it mounts the stuff then
> umounts it right after that.  Huh?  Is the relevant part the "mount -o ro
> /dev/sda1 /mnt/root" ?  Then the exec switch_root part after that?  The rest
> seems to cancel each other out.
>
> Looking forward to that light bulb moment here.  ;-)

Here's how I read it.

First, it mounts /proc and /sys, since just about anything is going to
need at least one of those.

Second, it mounts your (desired) / filesystem at /mnt/root.

Third, it unmounts /proc and /sys.

Fourth, it switches out / with the filesystem it already mounted at
/mnt/root. This is similar (idential) to chroot. At the _same_ time,
it launches your init script.

Your init script launches and sees a / without a /proc or a /sys. The
/ it sees is what _was_ /mnt/root only moments before. The stuff that
was originally at / is no longer accessible. (Which, incidentally, is
why you unmount /proc and /sys; nothing would be able to get to those
particular mounted filesystems, since everything else gets to see the
world with /mnt/root/ as the /.

Your init script (the one at /sbin/init), seeing itself in a fresh,
needs-to-be-booted system, mounts /proc, /sys, etc...everything the
init script is configured to do.

-- 
:wq

Re: [gentoo-user] Making a init thingy. Step two I guess.

[Dale]

Michael Mol wrote:
> On Sat, Sep 17, 2011 at 6:27 PM, Dale<rdalek1967@gmail.com>  wrote:
>> Mark Knecht wrote:
>> I think there are only a few that has that flag, at least that I would put
>> in the init thingy anyway.  Maybe this is something that the devs will work
>> on if it can be done.  May be a big if there.
>>
>> That is the guide I am trying to go by but I think I am missing something.
>>   This is the script they have posted:
>>
>> #!/bin/busybox sh
>>
>> # Mount the /proc and /sys filesystems.
>> mount -t proc none /proc
>> mount -t sysfs none /sys
>>
>> # Do your stuff here.
>> echo "This script mounts rootfs and boots it up, nothing more!"
>>
>> # Mount the root filesystem.
>> mount -o ro /dev/sda1 /mnt/root
>>
>> # Clean up.
>> umount /proc
>> umount /sys
>>
>> # Boot the real thing.
>> exec switch_root /mnt/root /sbin/init
>>
>> That doesn't really make much sense to me.  First it mounts the stuff then
>> umounts it right after that.  Huh?  Is the relevant part the "mount -o ro
>> /dev/sda1 /mnt/root" ?  Then the exec switch_root part after that?  The rest
>> seems to cancel each other out.
>>
>> Looking forward to that light bulb moment here.  ;-)
> Here's how I read it.
>
> First, it mounts /proc and /sys, since just about anything is going to
> need at least one of those.
>
> Second, it mounts your (desired) / filesystem at /mnt/root.
>
> Third, it unmounts /proc and /sys.
>
> Fourth, it switches out / with the filesystem it already mounted at
> /mnt/root. This is similar (idential) to chroot. At the _same_ time,
> it launches your init script.
>
> Your init script launches and sees a / without a /proc or a /sys. The
> / it sees is what _was_ /mnt/root only moments before. The stuff that
> was originally at / is no longer accessible. (Which, incidentally, is
> why you unmount /proc and /sys; nothing would be able to get to those
> particular mounted filesystems, since everything else gets to see the
> world with /mnt/root/ as the /.
>
> Your init script (the one at /sbin/init), seeing itself in a fresh,
> needs-to-be-booted system, mounts /proc, /sys, etc...everything the
> init script is configured to do.
>

Ahhhhh.  I see now.  So, it mounts proc and sys but that is in the init 
then it mounts the real root outside the init.  Then it umounts the proc 
and sys under the init and then switches to the real root and starts 
init there.

Where does /usr and /var come in here?  Isn't the init supposed to mount 
that too?  Do I add that myself?  If so, when to fsck get ran?

Dale

:-)  :-)

Re: [gentoo-user] Making a init thingy. Step two I guess.

[Michael Mol]

On Sat, Sep 17, 2011 at 10:13 PM, Dale <rdalek1967@gmail.com> wrote:
> Michael Mol wrote:

[snippage]

> Ahhhhh.  I see now.  So, it mounts proc and sys but that is in the init then
> it mounts the real root outside the init.  Then it umounts the proc and sys
> under the init and then switches to the real root and starts init there.
>
> Where does /usr and /var come in here?  Isn't the init supposed to mount
> that too?  Do I add that myself?  If so, when to fsck get ran?

From the look of it, yeah, it'd be your responsibility to do that in
the "# Do your stuff here." area.

Probably why the script has a line that says: echo "This script mounts
rootfs and boots it up, nothing more!"

-- 
:wq

Re: [gentoo-user] Making a init thingy. Step two I guess.

[Pandu Poluan]

[-- Attachment #1: Type: text/plain, Size: 909 bytes --]

On Sep 18, 2011 9:22 AM, "Michael Mol" <mikemol@gmail.com> wrote:
>
> On Sat, Sep 17, 2011 at 10:13 PM, Dale <rdalek1967@gmail.com> wrote:
> > Michael Mol wrote:
>
> [snippage]
>
> > Ahhhhh.  I see now.  So, it mounts proc and sys but that is in the init
then
> > it mounts the real root outside the init.  Then it umounts the proc and
sys
> > under the init and then switches to the real root and starts init there.
> >
> > Where does /usr and /var come in here?  Isn't the init supposed to mount
> > that too?  Do I add that myself?  If so, when to fsck get ran?
>
> From the look of it, yeah, it'd be your responsibility to do that in
> the "# Do your stuff here." area.
>
> Probably why the script has a line that says: echo "This script mounts
> rootfs and boots it up, nothing more!"
>

Of course, you have to remember to mount usr in /mnt/root instead of in /

That switch_root thing, you know.

Rgds,

[-- Attachment #2: Type: text/html, Size: 1260 bytes --]

Re: [gentoo-user] Making a init thingy. Step two I guess.

[Mark Knecht]

On Sat, Sep 17, 2011 at 7:13 PM, Dale <rdalek1967@gmail.com> wrote:
<SNIP>
>
> Ahhhhh.  I see now.  So, it mounts proc and sys but that is in the init then
> it mounts the real root outside the init.  Then it umounts the proc and sys
> under the init and then switches to the real root and starts init there.
>
> Where does /usr and /var come in here?  Isn't the init supposed to mount
> that too?  Do I add that myself?  If so, when to fsck get ran?
>
> Dale

Hi Dale,
   Let's take a look at the script I sent you yesterday, copied here
again. This system uses a RAID6 for root. This RAID6 is built using
metadata type 1.2. (The newest at the time) Since the kernel only
assembles metadata type 0.9 automatically I had to use the initramfs
to do the assembly and mount.


c2stable / # cat /usr/src/initramfs/init
#!/bin/busybox sh

rescue_shell() {
   echo "Something went wrong. Dropping you to a shell."
   busybox --install -s
   exec /bin/sh
}


# Mount the /proc and /sys filesystems.
mount -t proc none /proc
mount -t sysfs none /sys

# Do your stuff here.
echo "This script mounts rootfs and boots it up, nothing more!"

mdadm --assemble --name=c2stable:3 /dev/md3

# Mount the root filesystem.
mount -o ro /dev/md3 /mnt/root  || rescue_shell

# Clean up.
umount /proc
umount /sys

# Boot the real thing.
exec switch_root /mnt/root /sbin/init
c2stable / #


There are really just 3 interesting parts to this:

1) The rescue_shell function
2) The assembly command
3) The mount command

When you read through the script you see I do the assembly and then
the mount. If the mount command fails then I drop into the rescue
shell which is a simple Linux environment where I can run mdadm by
hand to figure out what went wrong. If the mount succeeds then the
script just continues and Linux boots.

The only difference concerning your /usr question would be change the
mount command to something like:

mount -o ro /dev/sdaX /usr || rescue_shell

Does this make basic sense?



One point about static flag. Normally I don't build mdadm with static enabled:

c2stable ~ # emerge -pv mdadm

These are the packages that would be merged, in order:

Calculating dependencies... done!
[ebuild   R    ] sys-fs/mdadm-3.1.4  USE="-static" 282 kB

Total: 1 package (1 reinstall), Size of downloads: 282 kB
c2stable ~ #

but for the version I put in the initramfs I did enable it.

Cheers,
Mark

Re: [gentoo-user] Making a init thingy. Step two I guess.

[Mark Knecht]

On Sat, Sep 17, 2011 at 3:27 PM, Dale <rdalek1967@gmail.com> wrote:
> Mark Knecht wrote:
>>
>> While I was out walking my dog I sort of remembered that there are
>> just a few apps I had to build static, or at least that I build
>> static. One, I think was grub. The first and only time I did it I had
>> to do it 3 or 4 times before I got everything I needed working
>> correctly. This page which is like the first one Google comes up with
>> talks about that stuff, although you have to sort of dig it out and
>> read between the lines.
>>
>> http://en.gentoo-wiki.com/wiki/Initramfs
>>
>> And, I think with the ldd command as long as you include the libraries
>> ldd points at in your equivalent /lib directory within the initramfs
>> then that works for most apps and keeps the initramfs smaller.
>>
>> HTH,
>> Mark
>>
>>
>
> I think there are only a few that has that flag, at least that I would put
> in the init thingy anyway.  Maybe this is something that the devs will work
> on if it can be done.  May be a big if there.
>
> That is the guide I am trying to go by but I think I am missing something.
>  This is the script they have posted:
>
> #!/bin/busybox sh
>
> # Mount the /proc and /sys filesystems.
> mount -t proc none /proc
> mount -t sysfs none /sys
>
> # Do your stuff here.
> echo "This script mounts rootfs and boots it up, nothing more!"
>
> # Mount the root filesystem.
> mount -o ro /dev/sda1 /mnt/root
>
> # Clean up.
> umount /proc
> umount /sys
>
> # Boot the real thing.
> exec switch_root /mnt/root /sbin/init
>
> That doesn't really make much sense to me.  First it mounts the stuff then
> umounts it right after that.  Huh?  Is the relevant part the "mount -o ro
> /dev/sda1 /mnt/root" ?  Then the exec switch_root part after that?  The rest
> seems to cancel each other out.
>
> Looking forward to that light bulb moment here.  ;-)
>
> Dale

Yup, that's what it does... ;-)

As you know, I'm a user type, not a sys admin or anything like that,
so I have no real training in how the Linux initialization process
really works. It's my uninformed opinion that the kernel in the early
stages needs /proc and /sys to basically run. I don't know what it
does with them but they need to be mounted so that the kernel can see
them. Remember, this isn't necessarily the same as what we see within
those directories once we really boot, it's just something for the
kernel to use for a while. The kernel does what it needs to do, then
unmounts them and lets things proceed.

If it helps here's an init file from the one I did to figure my mdadm
RAID boot problems here:

c2stable / # cat /usr/src/initramfs/init
#!/bin/busybox sh

rescue_shell() {
    echo "Something went wrong. Dropping you to a shell."
    busybox --install -s
    exec /bin/sh
}


# Mount the /proc and /sys filesystems.
mount -t proc none /proc
mount -t sysfs none /sys

# Do your stuff here.
echo "This script mounts rootfs and boots it up, nothing more!"

mdadm --assemble --name=c2stable:3 /dev/md3

# Mount the root filesystem.
mount -o ro /dev/md3 /mnt/root  || rescue_shell

# Clean up.
umount /proc
umount /sys

# Boot the real thing.
exec switch_root /mnt/root /sbin/init
c2stable / #

If I remember correctly I couldn't figure out what the right dev
number was at the time, so I built the initramfs without the mdadm
command, then used mdadm commands within the initramfs process using
the busybox shell. When I figured it out I used the updated init file
to prove it worked.

That could all be wrong - it was over a year ago - but that's what I
remember. Adding the busybox part at the top - the rescue shell part -
was REALLY useful.

HTH,
Mark

Re: [gentoo-user] Making a init thingy. Step two I guess.

[Mark Knecht]

On Sat, Sep 17, 2011 at 1:44 PM, Dale <rdalek1967@gmail.com> wrote:
> Mark Knecht wrote:
>>
>> On Sat, Sep 17, 2011 at 1:02 PM, Dale<rdalek1967@gmail.com>  wrote:
>>>
>>> OK.  Sort of dealt with the LVM part so now to step two.  I'm trying to
>>> follow this:
>>>
>>> http://en.gentoo-wiki.com/wiki/Initramfs
>>>
>>> Then I get confused.  I get to Applications and I'm sort of lost here.
>>>  In
>>> there it talks about copying nano and its friends over to the init
>>> directory.  Then below that it says to use busybox.  Well, which is it?
>>>  Do
>>> I do both of those or just one?
>>>
>> It's been a while for me but I believe it's both. I think busybox is
>> the thing that gives you command line tools like cd, ls, pwd, etc.
>> However you also can include applications in your initramfs that give
>> you more access to the hardware or the net.
>
> That's what I was thinking to.  I know busybox does a LOT of things but I
> didn't think it was a nano replacement either.  OK.  I got both of those in
> there at least.  First problem dealt with I guess.  Wonder what else I
> should put in there too. < scratches head >
>

Right. I don't think it replaces nano. If you want to edit in that
environment than I think you need to include an editor.

Note that one of the initramfs requirements for applications is that
they need to be built as 'static'. (I think) so that they don't
require libraries or other things that aren't yet mounted.

>>
>>> I also tried to figure out Dracut but it seems like a work in progress or
>>> something.  I can't seem to find any Gentoo docs for it.  I found a
>>> Redhat
>>> doc but that's not exactly Gentoo.
>>>
>>> Am I starting to work on this to soon?  Maybe wait a while longer until
>>> some
>>> more stuff gets sorted out?
>>>
>>> Dang, I'm full of questions today.  ;-)
>>>
>>> Dale
>>
>> I'm just reading some Dracut stuff myself. I figured I might as well
>> get educated. I likely won't actually build it and try it for a few
>> days at the soonest, but it's good to get prepared. Cheers to you for
>> doing that.
>>
>> - Mark
>>
>
> Yea, it's either learn this or find a new distro.  Still haven't ruled out
> the later tho.  Just going to try at least.
>
> Dale

Nahh...you're smarted than becoming an Ubuntu guy. Of course, if I'm
wrong maybe you'll become an Ubuntu guy... ;-)

- Mark

Re: [gentoo-user] Making a init thingy. Step two I guess.

[Dale]

Mark Knecht wrote:
> On Sat, Sep 17, 2011 at 1:44 PM, Dale<rdalek1967@gmail.com>  wrote:
>
>> That's what I was thinking to.  I know busybox does a LOT of things but I
>> didn't think it was a nano replacement either.  OK.  I got both of those in
>> there at least.  First problem dealt with I guess.  Wonder what else I
>> should put in there too.<  scratches head>
>>
> Right. I don't think it replaces nano. If you want to edit in that
> environment than I think you need to include an editor.
>
> Note that one of the initramfs requirements for applications is that
> they need to be built as 'static'. (I think) so that they don't
> require libraries or other things that aren't yet mounted.

I noticed that in the command for busybox.  Trick is remembering it.  
:/   Oh, looky here:

[ebuild   R    ] app-editors/nano-2.2.5  USE="justify ncurses nls spell 
unicode -debug -minimal -slang"

No static build for nano.  Bummer.  It couldn't be that easy huh?  :-(

>> Yea, it's either learn this or find a new distro.  Still haven't ruled out
>> the later tho.  Just going to try at least.
>>
>> Dale
> Nahh...you're smarted than becoming an Ubuntu guy. Of course, if I'm
> wrong maybe you'll become an Ubuntu guy... ;-)
>
> - Mark
>

I'm not sure at times.  It seems as times goes by, the reasons I left 
Mandrake are coming to Gentoo.  The only thing is, I only updated 
Mandrake once every few months and VERY little to compile either.  Plus, 
I had powstatd for my UPS as well.  I like it a LOT more than upsmon and 
such.

I made it to the init script part and I'm clueless.  Let me see if I got 
this right.  I copy the script but I don't see where it mounts /usr and 
/var anywhere.  Do I need to add that to the script or does something 
figure this out later?  Isn't that the point of the script, to mount 
/usr and /var?

Progress, ever so sloooooooowly.  o_O

Dale

:-)  :-)

Re: [gentoo-user] Making a init thingy. Step two I guess.

[pk]

On 2011-09-17 22:54, Mark Knecht wrote:

>> That's what I was thinking to.  I know busybox does a LOT of things but I
>> didn't think it was a nano replacement either.  OK.  I got both of those in
>> there at least.  First problem dealt with I guess.  Wonder what else I
>> should put in there too. < scratches head >

http://linux.die.net/man/1/busybox

> Right. I don't think it replaces nano. If you want to edit in that
> environment than I think you need to include an editor.

cat & (s)ed (included in busybox) works fine if you want to edit text
files...
https://www.ibm.com/developerworks/aix/library/au-textedit.html

Best regards

Peter K

Re: [gentoo-user] Making a init thingy. Step two I guess.

[Alan McKinnon]

On Sat, 17 Sep 2011 13:27:45 -0700
Mark Knecht <markknecht@gmail.com> wrote:

> > Then I get confused.  I get to Applications and I'm sort of lost
> > here.  In there it talks about copying nano and its friends over to
> > the init directory.  Then below that it says to use busybox.  Well,
> > which is it?  Do I do both of those or just one?
> >  
> 
> It's been a while for me but I believe it's both. I think busybox is
> the thing that gives you command line tools like cd, ls, pwd, etc.
> However you also can include applications in your initramfs that give
> you more access to the hardware or the net.

True.

Busybox is a tiny userland implementing most of the common options for
most of the common Unix commands. When you log into your ADSL
router/modem and get a shell, it's probably busybox running there,
not GNU util-linux stuff.

Binary distros often put busybox in their initrds as it doubles up as a
rescue environment and busybox is many times smaller than the full GNU
stuff. It's up to you if you want to do that or not; if all you use an
initrd for is a convenient place to store drivers to be able to
mount /usr, then you will have no need for busybox in it.

-- 
Alan McKinnnon
alan.mckinnon@gmail.com

Re: [gentoo-user] Making a init thingy. Step two I guess.

[Dale]

Alan McKinnon wrote:
> On Sat, 17 Sep 2011 13:27:45 -0700
> Mark Knecht<markknecht@gmail.com>  wrote:
>
>>> Then I get confused.  I get to Applications and I'm sort of lost
>>> here.  In there it talks about copying nano and its friends over to
>>> the init directory.  Then below that it says to use busybox.  Well,
>>> which is it?  Do I do both of those or just one?
>>>
>> It's been a while for me but I believe it's both. I think busybox is
>> the thing that gives you command line tools like cd, ls, pwd, etc.
>> However you also can include applications in your initramfs that give
>> you more access to the hardware or the net.
> True.
>
> Busybox is a tiny userland implementing most of the common options for
> most of the common Unix commands. When you log into your ADSL
> router/modem and get a shell, it's probably busybox running there,
> not GNU util-linux stuff.
>
> Binary distros often put busybox in their initrds as it doubles up as a
> rescue environment and busybox is many times smaller than the full GNU
> stuff. It's up to you if you want to do that or not; if all you use an
> initrd for is a convenient place to store drivers to be able to
> mount /usr, then you will have no need for busybox in it.
>

Unless something goes wrong.  You know how I am.  Murphy's law and all.  
lol

Dale

:-)  :-)