Compact flash booting problem solved

QNX Newsgroups (Archives) qdn.public.qnxrtp.os
#1

The booting from compact flash problem has been solved… the problem
is that the neutrino dinit program doesn’t write the loader correctly
onto the fash…
the solution is:

obtain a image for a qnx 4 boot disk… (i got mine from support but i
hope they will post it for download somewhere…

once booted in qnx4
no in my case the CF was /dev/hd1 but yous might be different.


Fsys.eide
fdisk /dev/hd1
create your boot partition, don’t forget to make it bootable
reboot into qnx4 again to update the partition table

mount -p /dev/hd1
dinit -h -f mybootimage /dev/hd1t79

i loaded my bootimage when i did the dinit, but you should be able to
boot into neutrino and copy it over that way too.

reboot and if all is well your up and booting off of flash…

vince

#2

I found a similar problem with dinit not writing the boot loader correctly
to a hard-disk.
I found that using the (un-documented) -8 option to write the ‘8G’ boot
loader (whatever that is) wrote a
bootable loader that worked. Maybe the boot loader included with the latest
dinit is corrupted somehow.
(the neutrino 2.0 dinit worked fine, but the rtp dinit didn’t)

“vince” <vgeisler@engineer.com> wrote in message
news:119t2topnt30gdn83l3n35g0fhjme87tcu@4ax.com…

The booting from compact flash problem has been solved… the problem
is that the neutrino dinit program doesn’t write the loader correctly
onto the fash…
the solution is:

obtain a image for a qnx 4 boot disk… (i got mine from support but i
hope they will post it for download somewhere…

once booted in qnx4
no in my case the CF was /dev/hd1 but yous might be different.


Fsys.eide
fdisk /dev/hd1
create your boot partition, don’t forget to make it bootable
reboot into qnx4 again to update the partition table

mount -p /dev/hd1
dinit -h -f mybootimage /dev/hd1t79

i loaded my bootimage when i did the dinit, but you should be able to
boot into neutrino and copy it over that way too.

reboot and if all is well your up and booting off of flash…

vince

\

#3

I’m happy for you a solution was found, big releif I guess :wink:

“vince” <vgeisler@engineer.com> wrote in message
news:119t2topnt30gdn83l3n35g0fhjme87tcu@4ax.com…

The booting from compact flash problem has been solved… the problem
is that the neutrino dinit program doesn’t write the loader correctly
onto the fash…
the solution is:

obtain a image for a qnx 4 boot disk… (i got mine from support but i
hope they will post it for download somewhere…

once booted in qnx4
no in my case the CF was /dev/hd1 but yous might be different.


Fsys.eide
fdisk /dev/hd1
create your boot partition, don’t forget to make it bootable
reboot into qnx4 again to update the partition table

mount -p /dev/hd1
dinit -h -f mybootimage /dev/hd1t79

i loaded my bootimage when i did the dinit, but you should be able to
boot into neutrino and copy it over that way too.

reboot and if all is well your up and booting off of flash…

vince

\

#4

Yep I had the same when trying to boot from DiskOnChip. It looks like the
RTP version of dinit has a new ‘feature’. I had to use the Neutrino 2.0
version…

#5

Dinit wasn’t really corrupt, the problem was that it was seeing CF as a floppy
and not writing the loader correctly. It seems to be fixed now, in the mean time
for those without a QNX system can download a bootable floppy with all the
tools needed (dinit, Fsys.eide, mount) here:

http://www.qnx.com/~emuis/files/nto/cf_dsk.img


Erick.



Kieran Tyrrell <kieran.tyrrell@virgin.net> wrote:

I found a similar problem with dinit not writing the boot loader correctly
to a hard-disk.
I found that using the (un-documented) -8 option to write the ‘8G’ boot
loader (whatever that is) wrote a
bootable loader that worked. Maybe the boot loader included with the latest
dinit is corrupted somehow.
(the neutrino 2.0 dinit worked fine, but the rtp dinit didn’t)

“vince” <> vgeisler@engineer.com> > wrote in message
news:> 119t2topnt30gdn83l3n35g0fhjme87tcu@4ax.com> …

The booting from compact flash problem has been solved… the problem
is that the neutrino dinit program doesn’t write the loader correctly
onto the fash…
the solution is:

obtain a image for a qnx 4 boot disk… (i got mine from support but i
hope they will post it for download somewhere…

once booted in qnx4
no in my case the CF was /dev/hd1 but yous might be different.


Fsys.eide
fdisk /dev/hd1
create your boot partition, don’t forget to make it bootable
reboot into qnx4 again to update the partition table

mount -p /dev/hd1
dinit -h -f mybootimage /dev/hd1t79

i loaded my bootimage when i did the dinit, but you should be able to
boot into neutrino and copy it over that way too.

reboot and if all is well your up and booting off of flash…

vince

\

#6

the aligators have released their grip on my ass slightly :slight_smile:
but i guess thats life working with P.H.B’s … the floggings will
continue untill moral improves.
oh well at least im not hearing anymore “i want daily 10 page reports
on why you’re behind”, from my particular P.H.B

after i got the booting problem fixed i managed to get the system
embedded Neutrino/Photon in a day and in under 10 megs!!!
Very cool…

Vince


On Wed, 6 Dec 2000 22:11:36 -0500, “Mario Charest”
<mcharest@void_zinformatic.com> wrote:

I’m happy for you a solution was found, big releif I guess > :wink:

“vince” <> vgeisler@engineer.com> > wrote in message
news:> 119t2topnt30gdn83l3n35g0fhjme87tcu@4ax.com> …

The booting from compact flash problem has been solved… the problem
is that the neutrino dinit program doesn’t write the loader correctly
onto the fash…
the solution is:

obtain a image for a qnx 4 boot disk… (i got mine from support but i
hope they will post it for download somewhere…

once booted in qnx4
no in my case the CF was /dev/hd1 but yous might be different.


Fsys.eide
fdisk /dev/hd1
create your boot partition, don’t forget to make it bootable
reboot into qnx4 again to update the partition table

mount -p /dev/hd1
dinit -h -f mybootimage /dev/hd1t79

i loaded my bootimage when i did the dinit, but you should be able to
boot into neutrino and copy it over that way too.

reboot and if all is well your up and booting off of flash…

vince


\

#7

ok fine… i give up… its a “Feature” that dinit thinks that CF is a
floppy :wink:





On 7 Dec 2000 16:45:34 GMT, Erick Muis <emuis@qnx.com> wrote:

Dinit wasn’t really corrupt, the problem was that it was seeing CF as a floppy
and not writing the loader correctly. It seems to be fixed now, in the mean time
for those without a QNX system can download a bootable floppy with all the
tools needed (dinit, Fsys.eide, mount) here:

http://www.qnx.com/~emuis/files/nto/cf_dsk.img


Erick.



Kieran Tyrrell <> kieran.tyrrell@virgin.net> > wrote:
I found a similar problem with dinit not writing the boot loader correctly
to a hard-disk.
I found that using the (un-documented) -8 option to write the ‘8G’ boot
loader (whatever that is) wrote a
bootable loader that worked. Maybe the boot loader included with the latest
dinit is corrupted somehow.
(the neutrino 2.0 dinit worked fine, but the rtp dinit didn’t)

“vince” <> vgeisler@engineer.com> > wrote in message
news:> 119t2topnt30gdn83l3n35g0fhjme87tcu@4ax.com> …

The booting from compact flash problem has been solved… the problem
is that the neutrino dinit program doesn’t write the loader correctly
onto the fash…
the solution is:

obtain a image for a qnx 4 boot disk… (i got mine from support but i
hope they will post it for download somewhere…

once booted in qnx4
no in my case the CF was /dev/hd1 but yous might be different.


Fsys.eide
fdisk /dev/hd1
create your boot partition, don’t forget to make it bootable
reboot into qnx4 again to update the partition table

mount -p /dev/hd1
dinit -h -f mybootimage /dev/hd1t79

i loaded my bootimage when i did the dinit, but you should be able to
boot into neutrino and copy it over that way too.

reboot and if all is well your up and booting off of flash…

vince


\

#8

vince <vgeisler@engineer.com> wrote:

ok fine… i give up… its a “Feature” that dinit thinks that CF is a
floppy > :wink:

Erick was responding to someone else who posted that they thought dinit was
corrupt.

He did not say that the bug was a feature.

When someone tells me a program is corrupted, I assume that they mean it
is not an exact replica of what was supposed to be shipped. I do not
consider a program that has a bug in it to be corrupt' any more than I consider a bug in something to be a feature’.

I don’t know the details either, but it could be that dinit is doing
everything correctly, and there is a bug in the driver for the flash
that makes it look like a floppy.

So don’t assume that the entire world is out to get you. In your case,
it is probably true, but as a general assumption it is probably not.

#9

Sorry should mention that the makedisk.exe is required (get this from the
free software section on the QNX website) and a DOS formatted floppy disk.

Erick.


Erick Muis <emuis@qnx.com> wrote:

Dinit wasn’t really corrupt, the problem was that it was seeing CF as a floppy
and not writing the loader correctly. It seems to be fixed now, in the mean time
for those without a QNX system can download a bootable floppy with all the
tools needed (dinit, Fsys.eide, mount) here:

http://www.qnx.com/~emuis/files/nto/cf_dsk.img



Erick.



Kieran Tyrrell <> kieran.tyrrell@virgin.net> > wrote:
I found a similar problem with dinit not writing the boot loader correctly
to a hard-disk.
I found that using the (un-documented) -8 option to write the ‘8G’ boot
loader (whatever that is) wrote a
bootable loader that worked. Maybe the boot loader included with the latest
dinit is corrupted somehow.
(the neutrino 2.0 dinit worked fine, but the rtp dinit didn’t)

“vince” <> vgeisler@engineer.com> > wrote in message
news:> 119t2topnt30gdn83l3n35g0fhjme87tcu@4ax.com> …

The booting from compact flash problem has been solved… the problem
is that the neutrino dinit program doesn’t write the loader correctly
onto the fash…
the solution is:

obtain a image for a qnx 4 boot disk… (i got mine from support but i
hope they will post it for download somewhere…

once booted in qnx4
no in my case the CF was /dev/hd1 but yous might be different.


Fsys.eide
fdisk /dev/hd1
create your boot partition, don’t forget to make it bootable
reboot into qnx4 again to update the partition table

mount -p /dev/hd1
dinit -h -f mybootimage /dev/hd1t79

i loaded my bootimage when i did the dinit, but you should be able to
boot into neutrino and copy it over that way too.

reboot and if all is well your up and booting off of flash…

vince

\

#10

I don’t know the details either, but it could be that dinit is doing
everything correctly, and there is a bug in the driver for the flash
that makes it look like a floppy.

The dinit that shipped with RTP does have a bug. For compact flash it writes
a bootloader which looks for a boot image on the floppy drive. dinit
has apparantly been fixed internally and testing, although limited,
shows that it is working properly now.

#11

Ooops, sorry, looks like I sparked off more than I meant…

My particular problem with the rtp dinit was when writing a bootloader to a
LS120 superdisk (an ide device). The old version worked, the new one didn’t.
I (wrongly) assumed that the actual bootloader had changed, but it seems
that the problem I was seeing may be the same as with CF. ie dinit was
seeing the LS120 as a floppy too…

Sorry, I take back my comment about it being ‘corrupt’.


“Chris Foran” <cforan@qnx.com> wrote in message
news:Voyager.001207165309.16538H@funnel.qnx.com…

I don’t know the details either, but it could be that dinit is doing
everything correctly, and there is a bug in the driver for the flash
that makes it look like a floppy.


The dinit that shipped with RTP does have a bug. For compact flash it
writes
a bootloader which looks for a boot image on the floppy drive. dinit
has apparantly been fixed internally and testing, although limited,
shows that it is working properly now.

#12

And I take back the use of the word ‘feature’ :wink:

#13

Jim Atkins <jamesa@tsd.serco.com> wrote:

And I take back the use of the word ‘feature’ > :wink:

And I take back the use of the word `assume’…