Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

here's how to use sol9 ISOs without burning

2 views
Skip to first unread message

Philip Brown

unread,
Jun 7, 2002, 1:58:23 AM6/7/02
to
I saw a request or two a while back about what to do with the durn
sol9 iso's from sun, if you want to make a jumpstart server, but dont want
to actually burn CDs from the images.
The trouble being that 'lofiadm' is not multi-slice aware.

I saw an "interesting" way to do this on an old website, involving use of
od , and other "interesting" things. But here's a more straightforward
method:


lofiadm -a cd1.iso
mount -o ro -F hsfs /dev/lofi/1 /mnt1
df -k /mnt1
dd if=cd1.iso of=/tmp/cd1.s1.img bs=1k iseek=(number of blocks from df)

You can now do : mount -o ro `lofiadm -a /tmp/cd1.s1.img` /mnt2

This information is recorded on my main solaris page, along with other good
stuff:

http://www.bolthole.com/solaris/#quickies


--
[Trim the no-bots from my address to reply to me by email!]
[ Do NOT email-CC me on posts. Pick one or the other.]
S.1618 http://thomas.loc.gov/cgi-bin/bdquery/z?d105:SN01618:@@@D
http://www.spamlaws.com/state/ca1.html

Joerg Schilling

unread,
Jun 7, 2002, 4:36:19 AM6/7/02
to
In article <slrnag0j2u....@bolthole.com>,

Philip Brown <phi...@bolthole.no-bots.com> wrote:
>I saw a request or two a while back about what to do with the durn
>sol9 iso's from sun, if you want to make a jumpstart server, but dont want
>to actually burn CDs from the images.
>The trouble being that 'lofiadm' is not multi-slice aware.
>
>I saw an "interesting" way to do this on an old website, involving use of
>od , and other "interesting" things. But here's a more straightforward
>method:
>
>
> lofiadm -a cd1.iso
> mount -o ro -F hsfs /dev/lofi/1 /mnt1
> df -k /mnt1
> dd if=cd1.iso of=/tmp/cd1.s1.img bs=1k iseek=(number of blocks from df)
>
> You can now do : mount -o ro `lofiadm -a /tmp/cd1.s1.img` /mnt2

This would not work ad the ISO FS needs internal space.
It would make more sende to call:

isoinfo -i cd1.iso -d

and look for the number os 2048 byte sectors - but even this is wrong because
of padding.

YOu need to print the label information from sector #0 and get
the right offset.

--
EMail:jo...@schily.isdn.cs.tu-berlin.de (home) Jörg Schilling D-13353 Berlin
j...@cs.tu-berlin.de (uni) If you don't have iso-8859-1
schi...@fokus.gmd.de (work) chars I am J"org Schilling
URL: http://www.fokus.gmd.de/usr/schilling ftp://ftp.fokus.gmd.de/pub/unix

Sven-Olaf Hilmer

unread,
Jun 7, 2002, 6:37:19 AM6/7/02
to
Joerg Schilling wrote:
...

> This would not work ad the ISO FS needs internal space.
> It would make more sende to call:
>
> isoinfo -i cd1.iso -d
>
> and look for the number os 2048 byte sectors - but even this is wrong because
> of padding.

where can I get an 'isoinfo' binary for Solaris 8
or a source package ?
I've searched the GNU site and Google and nothing appapriate found.

Sven

Richard L. Hamilton

unread,
Jun 7, 2002, 8:41:26 AM6/7/02
to
In article <slrnag0j2u....@bolthole.com>,

phi...@bolthole.no-bots.com (Philip Brown) writes:
> I saw a request or two a while back about what to do with the durn
> sol9 iso's from sun, if you want to make a jumpstart server, but dont want
> to actually burn CDs from the images.
> The trouble being that 'lofiadm' is not multi-slice aware.
>
> I saw an "interesting" way to do this on an old website, involving use of
> od , and other "interesting" things. But here's a more straightforward
> method:
>
>
> lofiadm -a cd1.iso
> mount -o ro -F hsfs /dev/lofi/1 /mnt1
> df -k /mnt1
> dd if=cd1.iso of=/tmp/cd1.s1.img bs=1k iseek=(number of blocks from df)
>
> You can now do : mount -o ro `lofiadm -a /tmp/cd1.s1.img` /mnt2
>
> This information is recorded on my main solaris page, along with other good
> stuff:
>
> http://www.bolthole.com/solaris/#quickies

In another thread, I (re)posted a C program that I wrote to extract
all but the slice at offset 0 of a multi-slice ISO image such as is common
for SPARC Solaris boot CDs. Search groups.google.com for message-id
<ufv7776...@corp.supernews.com>

Usage is

imgslice [-c] iso_file

Without -c, it will just list the slices; with -c, it will extract
all that are not at offset 0, provided they are all validly within
the image file. (A slice at offset 0 can be accessed directly
via the ISO image file and doesn't need to be copied to a separate file.)

No commands, no scripts, no foolin' around, just compile it and go.
A truly crazed person could perhaps even rewrite it in perl (at
the expense of using read/write rather than mmap(), so it might be a
fair bit slower) and eliminate the need to compile it at all.

--
mailto:rlh...@mindwarp.smart.net http://www.smart.net/~rlhamil

Andrew Gabriel

unread,
Jun 7, 2002, 10:43:51 AM6/7/02
to
In article <ug1afm5...@corp.supernews.com>,

rlh...@smart.net (Richard L. Hamilton) writes:
> In another thread, I (re)posted a C program that I wrote to extract
> all but the slice at offset 0 of a multi-slice ISO image such as is common
> for SPARC Solaris boot CDs. Search groups.google.com for message-id
> <ufv7776...@corp.supernews.com>
>
> Usage is
>
> imgslice [-c] iso_file
>
> Without -c, it will just list the slices; with -c, it will extract
> all that are not at offset 0, provided they are all validly within
> the image file. (A slice at offset 0 can be accessed directly
> via the ISO image file and doesn't need to be copied to a separate file.)
>
> No commands, no scripts, no foolin' around, just compile it and go.
> A truly crazed person could perhaps even rewrite it in perl (at
> the expense of using read/write rather than mmap(), so it might be a
> fair bit slower) and eliminate the need to compile it at all.

Somewhere, I have one which does something similar, but puts
the disk back together from the individual files, constructing
appropriately sized entries in the VToC (so you can change them),
and it works on x86 too (which has a differently structured VToC,
looking more like the SVR4 one than the SunOS one).

I think mkisofs now has some/all of this functionality, although
I've never used it in mkisofs.

--
Andrew Gabriel

Philip Brown

unread,
Jun 7, 2002, 12:07:08 PM6/7/02
to
On 7 Jun 2002 08:36:19 GMT, j...@cs.tu-berlin.de wrote:
>In article <slrnag0j2u....@bolthole.com>,
>Philip Brown <phi...@bolthole.no-bots.com> wrote:
>> lofiadm -a cd1.iso
>> mount -o ro -F hsfs /dev/lofi/1 /mnt1
>> df -k /mnt1
>> dd if=cd1.iso of=/tmp/cd1.s1.img bs=1k iseek=(number of blocks from df)
>>
>> You can now do : mount -o ro `lofiadm -a /tmp/cd1.s1.img` /mnt2
>
>This would not work [if?] the ISO FS needs internal space.

>It would make more sende to call:
> isoinfo -i cd1.iso -d ...

>
>YOu need to print the label information from sector #0 and get
>the right offset.

eh.. the proceedure I posted, works for me.
I'm upgrading a sparc from 9beta to 9fcs as I write this :-)

Philip Brown

unread,
Jun 7, 2002, 12:09:55 PM6/7/02
to
On Fri, 07 Jun 2002 12:41:26 -0000, rlh...@smart.net wrote:
>
>In another thread, I (re)posted a C program that I wrote to extract
>all but the slice at offset 0 of a multi-slice ISO image such as is common
>for SPARC Solaris boot CDs.

cool.
found it.

Unfortunately, according to your comments, it only works for sparc.
The proceedure I wrote, works on solx86 as well.

proved by the fact that I'm upgrading a sparc box, from a sol8x86 jumpstart
server ;-)

Joerg Schilling

unread,
Jun 7, 2002, 2:47:36 PM6/7/02
to
In article <3D008CDF...@post.ch>,

It is part of the mkisofs part of the cdrtools package.

I have no idea why Sun did not include it into Solaris. Looks as if somebody
at Sun would decide to remove 'fstyp'.

Richard L. Hamilton

unread,
Jun 7, 2002, 3:57:59 PM6/7/02
to
In article <slrnag1mti....@bolthole.com>,

phi...@bolthole.no-bots.com (Philip Brown) writes:
> On Fri, 07 Jun 2002 12:41:26 -0000, rlh...@smart.net wrote:
>>
>>In another thread, I (re)posted a C program that I wrote to extract
>>all but the slice at offset 0 of a multi-slice ISO image such as is common
>>for SPARC Solaris boot CDs.
>
> cool.
> found it.
>
> Unfortunately, according to your comments, it only works for sparc.
> The proceedure I wrote, works on solx86 as well.
>
> proved by the fact that I'm upgrading a sparc box, from a sol8x86 jumpstart
> server ;-)


Hey, I wouldn't mind trying to make it work on x86, but I don't have (or
otherwise need) an x86 box to test it on.

0 new messages