i tried to get a new hd bootable and copied an existing qnx-partition to
that hd.
Now i try to boot it, but it seems i have to do a little fix to that
“copied” bootpart. At time qnx seems to find the bootsector and starts
booting:
Searching for Boot-Record from IDE-1…OK
HitESC for .altboot … (many rows of points)
…and then it stops.
Where can i find information what to do next to get it booting?
As far as I remember, you can have such problem if your image is not
below the 1024th cylinder and you don’t use the good OS loader. Look at
fdisk.
cheers,
Alain.
CG a écrit:
Hi,
i tried to get a new hd bootable and copied an existing qnx-partition to
that hd.
Now i try to boot it, but it seems i have to do a little fix to that
“copied” bootpart. At time qnx seems to find the bootsector and starts
booting:
Searching for Boot-Record from IDE-1…OK
HitESC for .altboot … (many rows of points)
…and then it stops.
Where can i find information what to do next to get it booting?
i tried to get a new hd bootable and copied an existing qnx-partition to
that hd.
What did you use for copying?
Now i try to boot it, but it seems i have to do a little fix to that
“copied” bootpart. At time qnx seems to find the bootsector and starts
booting:
Searching for Boot-Record from IDE-1…OK
HitESC for .altboot … (many rows of points)
…and then it stops.
Try
dinit -r /dev/hd0t78
chkfsys -u /dev/hd0t78
I suppose hd0t78 is your “copied” partition. Also I suppose it is not big
deal if you loss “copied” partition. Do you have original one? But I hope
commands above will useful
Best regards.
Eduard.
Where can i find information what to do next to get it booting?
I had copied my hd (complete root fs - without /proc and /net and … of
course). This was ok,
BUT the .boot copied from /.boot was a zero byte file.
After a new copy command for this specific file it was copied correctly.
AND the .diskroot was not copied, too.
This file seems to have to be present in root directory. It’s empty. It only
has to be there to show which partition to mount to root …
Is this correct?
After this two fixes my new hd is a fine working copy of my old one!
AND the .diskroot was not copied, too.
This file seems to have to be present in root directory. It’s empty. It only
has to be there to show which partition to mount to root …
Is this correct?
Yes.
(Bonus point: If you had two partitions containing /.diskroot,
the “diskboot” lets you select which one to use)
The problem is these various “diskboot” features aren’t documented.
I had copied my hd (complete root fs - without /proc and /net and … of
course). This was ok,
BUT the .boot copied from /.boot was a zero byte file.
After a new copy command for this specific file it was copied correctly.
AND the .diskroot was not copied, too.
This file seems to have to be present in root directory. It’s empty. It
only
has to be there to show which partition to mount to root …
Is this correct?
Mainly, yes. And I saw sometimes, this file wasn’t empty. It contained some
path. I’ll very thankful if someone would explain this feature.
Eduard.
After this two fixes my new hd is a fine working copy of my old one!
AND the .diskroot was not copied, too.
This file seems to have to be present in root directory. It’s empty. It
Mainly, yes. And I saw sometimes, this file wasn’t empty. It contained some
path. I’ll very thankful if someone would explain this feature.
Well, I don’t use diskboot, but from looking at the source …
It can contain lines of the format “mountpt=XXX” or “options=XXX”,
which can be used to control the pathname to mount at or any per-fsys
options. An empty file is equivalent to a file containing the line
“mountpt=/”. Any filesystems that have specific requirements are
unmounted following the initial probe and remounted at their final
place (note that this can upset the default mount naming scheme).
So presumably you could edit a .diskroot file on all your partitions
and have them mount at non-standard ("/fs/hdN-XXX-N") locations …
It can contain lines of the format “mountpt=XXX” or “options=XXX”,
which can be used to control the pathname to mount at or any per-fsys
options. An empty file is equivalent to a file containing the line
“mountpt=/”. Any filesystems that have specific requirements are
unmounted following the initial probe and remounted at their final
place (note that this can upset the default mount naming scheme).
So presumably you could edit a .diskroot file on all your partitions
and have them mount at non-standard ("/fs/hdN-XXX-N") locations …
“Bill Caroselli (Q-TPS)” <> QTPS@earthlink.net> > wrote:
: This is big news. Is this documented anywhere?
: STEVE ! ! !
It doesn’t seem to be. I’ll ask the OS group about it.
While I don’t have a current use for this, years ago I would have loved to
have had this feature under QNX4. We had 4 SCSI Tahiti MO drives and a
different MO disk that was created once a week. We could have any 4 weeks
of data on line at any one time that we wanted. When a disk was mounted I
had to read what was on it and then mount if under a specific name in the
pathspace. I would have liked this feature.
First of all, huge thanks to John. It’s not a hacker’s observation anymore
and you told us very useful information (I didn’t know about fsys options
)).
Also I’m surprised the sys admin guide is behind the train. Actually, it’s
QNX 6.1 sys admin guide now. Congratulations. Now the guide consists of
unactully links and some notes. Also, there is no any word about .diskroot
in the guide. Are you going to rid of the guide?
Eduard.
Steve Reid > stever@qnx.com
TechPubs (Technical Publications)
QNX Software Systems
ed1k <ed1k@nobody.fools.ca> wrote:
: Also I’m surprised the sys admin guide is behind the train. Actually, it’s
: QNX 6.1 sys admin guide now. Congratulations. Now the guide consists of
: unactully links and some notes. Also, there is no any word about .diskroot
: in the guide. Are you going to rid of the guide?
Unfortunately, the writer who was working on that book is seriously ill. We
hope to do more work on that book soon.
Steve Reid stever@qnx.com
TechPubs (Technical Publications)
QNX Software Systems