Re: [Cooker] Re: A couple of kernel issues

2000-02-11 Thread Pixel

"Brian J. Murrell" [EMAIL PROTECTED] writes:

  
  this is not true. if you give hdX=ide-scsi to the kernel, it works
 
 I just tried this and you are right it does work.  Not that I doubted it
 you but I am/was sure I tried that at one time.  Guess I must be
 recalling incorrectly.
 
  Yep, the problem with that is the kernel is not really made for this,
 
 In what way?  I have been doing this since RH 5.0 at least.  Not that
 it's RH specific of course, just used as a time reference.

what i'm meaning is that few people uses ide as module. Either they don't use it
or have it builtin the kernel.

but as you're saying, i must be wrong...



Re: [Cooker] Re: A couple of kernel issues

2000-02-11 Thread Brian J. Murrell

from the quill of Chmouel Boudjnah [EMAIL PROTECTED] on scroll
[EMAIL PROTECTED]
 
 Yes, and this is a pain to generate with the normal kenrel rpm
 (because i
 will have to list each modules).

I don't think I follow.  In my kernel rpm I have a "detour" in the
kernel build to go build freeswan after which it goes back to the kernel
building.  When the build for a kernel type X (where X can be fb, smp,
secure) is done, why not just "move" the resulting ipsec.o to ipsec.o-X
and have ipsec.o-fb and ipsec.o-smp, etc. in the freeswan package and
link or move (or whatever) it into the modules directory on installation
depeneding on what kernel is running?

 on the boot you have a block of memory reserved for the bios and after
 you store the whole unziped initrd with unzipped kernel in memory.

Yes, but what I am saying is that I was under the impression that once
the kernel is booted and running, that block of memory that the initrd
disk image was on is freed back to the kernel.  From the initrd.txt in
linux/Documentation/initrd.txt:

Operation
-

When using initrd, the system boots as follows:

  1) the boot loader loads the kernel and the initial RAM disk
  2) the kernel converts initrd into a "normal" RAM disk and
 frees the memory used by initrd
  3) initrd is mounted read-write as root
  4) /linuxrc is executed (this can be any valid executable, including
 shell scripts; it is run with uid 0 and can do basically everything
 init can do)
  5) when linuxrc terminates, the "real" root file system is mounted
  6) if a directory /initrd exists, the initrd is moved there
 otherwise, initrd is unmounted
  7) the usual boot sequence (e.g. invocation of /sbin/init) is
 performed on the root file system

Further thots?

b.


--
Brian J. Murrell  InterLinx Support Services, Inc.
North Vancouver, B.C. 604 983 UNIX
Platform and Brand Independent UNIX Support - R3.2 - R4 - BSD



Re: [Cooker] Re: A couple of kernel issues

2000-02-10 Thread Chmouel Boudjnah

"Brian J. Murrell" [EMAIL PROTECTED] writes:

 Different?  As in different spec file/SRPM?  How come?  I have it merged
 into the standard Mandrake spec file and it works perfectamundo.  The
 only thing I have not done is isolate the ipsec.o module to the freeswan
 package and not include it with the kernels.  That would be required to
 keep the resulting kernels free from crypto and be able to put the
 freeswan completely on the crypto site.

Ok, but how you do for different kernel ? the ipsec.o should be
compiled for smp/fb/secure/standar or maybe it's only for secure
kernel, right ?

  you want always to use initrd ?
 Yeah, essentially.  :-)

I think it's expensive for low memory machine.

-- 
MandrakeSoft  http://www.mandrakesoft.com/
 --Chmouel



[Cooker] Re: A couple of kernel issues

2000-02-10 Thread Brian J. Murrell

from the quill of Pixel [EMAIL PROTECTED] on scroll
[EMAIL PROTECTED]
 
 this is not true. if you give hdX=ide-scsi to the kernel, it works

I just tried this and you are right it does work.  Not that I doubted it
you but I am/was sure I tried that at one time.  Guess I must be
recalling incorrectly.

 Yep, the problem with that is the kernel is not really made for this,

In what way?  I have been doing this since RH 5.0 at least.  Not that
it's RH specific of course, just used as a time reference.

 and less
 tested that way.

Really?  Meaning by Mandrake or in general?  Everybody who does SCSI for
their boot disk probably uses it, no?

 This is what i'm trying to do for the kernel boot and i don't get
 access to
 /proc/ide/ideX for example

I am not too familiar with the IDE driver and don't know what's supposed
to be in /proc/ide/ideX, but with my "ide as a module" kernel I am
running right now, I don't have it either.

b.


--
Brian J. Murrell  InterLinx Support Services, Inc.
North Vancouver, B.C. 604 983 UNIX
Platform and Brand Independent UNIX Support - R3.2 - R4 - BSD



[Cooker] Re: A couple of kernel issues

2000-02-09 Thread Brian J. Murrell

from the quill of Chmouel Boudjnah [EMAIL PROTECTED] on scroll
[EMAIL PROTECTED]
 
 Yes i would like to do this but we need a different kernel package for
 ipsec, and i don't have the time to maintain it.

Different?  As in different spec file/SRPM?  How come?  I have it merged
into the standard Mandrake spec file and it works perfectamundo.  The
only thing I have not done is isolate the ipsec.o module to the freeswan
package and not include it with the kernels.  That would be required to
keep the resulting kernels free from crypto and be able to put the
freeswan completely on the crypto site.

 you want always to use initrd ?

Yeah, essentially.  :-)

b.


--
Brian J. Murrell  InterLinx Support Services, Inc.
North Vancouver, B.C. 604 983 UNIX
Platform and Brand Independent UNIX Support - R3.2 - R4 - BSD