QNX 4.25E
Dell Latitude Notebook
Toshiba 30G Hard drive
Fsys.eide (also tried Fsys.ide with same result)
I am able to boot QNX from a floppy, access the network, load
drivers, etc. The partition table looks reasonable,
Head,Sectors,Cylinders = 255,63,1386 (or so, I’m working from
memory here), existing windows partitions appear to be in the
right places. QNX partition is in cylinders 650-702 (or so),
so it isn’t a 1024 issue.
Boot image is built and written to /.boot on the hard disk.
On reboot, it gets two dots into the boot, then hangs (and
sometimes reboots spontaneously). If I put the same boot
image on a floppy, it works fine.
I have seen mention of a -H option to Fsys that may be required
for larger disks, but all the comments mention that it is
undocumented, and - did I mention? - it’s undocumented.
I have tried, and my Fsys accepts the -H without complaint,
but lacking documentation, I don’t know what kind of argument
to give it, or how to interpret the results.
Is this relevant to the problem at hand? If so, what version
of Fsys and/or Fsys.eide is required? (I haven’t advanced to
G because I still lack a CD installer - we’ve been running
on downloaded updates for quite awhile now!)
Thanks, -Norton Allen
Norton Allen <allen@huarp.harvard.edu> wrote:
I have seen mention of a -H option to Fsys that may be required
for larger disks, but all the comments mention that it is
This has nothing to do with booting; it may be relevant later with
spurious ENFILE errors; usage has been discussed many times in this NG.
Is this relevant to the problem at hand? If so, what version
of Fsys and/or Fsys.eide is required? (I haven’t advanced to
8.1GB is a definite EIDE limit, so to use your disk may require a
newer version of Fsys.eide (symptoms may include not booting or
only seeing the first 8G of the disk). You can just update the
Fsys.eide driver without needing a full CD install. You may also
want to try a newer boot loader (for instance the QNX/NTO one, or
the one built-in to newer dinit versions) for larger disks.
John Garvey wrote:
Norton Allen <> allen@huarp.harvard.edu> > wrote:
I have seen mention of a -H option to Fsys that may be required
for larger disks, but all the comments mention that it is
This has nothing to do with booting; it may be relevant later with
spurious ENFILE errors; usage has been discussed many times in this NG.
Is this relevant to the problem at hand? If so, what version
of Fsys and/or Fsys.eide is required? (I haven’t advanced to
8.1GB is a definite EIDE limit, so to use your disk may require a
newer version of Fsys.eide (symptoms may include not booting or
only seeing the first 8G of the disk). You can just update the
Fsys.eide driver without needing a full CD install. You may also
want to try a newer boot loader (for instance the QNX/NTO one, or
the one built-in to newer dinit versions) for larger disks.
As I said, I’m using 4.25E, and Fsys.eide certainly claims to be
able to see 30GB (the C,H,S I stated before were badly recollected.
In particular, it reports >3000 cylinders. This partition is all
below 1024). I am able to dinit and copy lots of stuff onto the
disk, and it verifies, so it certainly looks like my version of
Fsys.eide is late enough, but do you know what version I would
need (and how I can get ahold of it?)
-Norton Allen
It seems to be a similar problem as “install problem” from 21.03.2003"
Have you tried using the old loader? ‘dinit -b -O /dev/hd0t77’
Previously, Norton Allen wrote in qdn.public.qnx4:
QNX 4.25E
Dell Latitude Notebook
Toshiba 30G Hard drive
Fsys.eide (also tried Fsys.ide with same result)
I am able to boot QNX from a floppy, access the network, load
drivers, etc. The partition table looks reasonable,
Head,Sectors,Cylinders = 255,63,1386 (or so, I’m working from
memory here), existing windows partitions appear to be in the
right places. QNX partition is in cylinders 650-702 (or so),
so it isn’t a 1024 issue.
Boot image is built and written to /.boot on the hard disk.
On reboot, it gets two dots into the boot, then hangs (and
sometimes reboots spontaneously). If I put the same boot
image on a floppy, it works fine.
I have seen mention of a -H option to Fsys that may be required
for larger disks, but all the comments mention that it is
undocumented, and - did I mention? - it’s undocumented.
I have tried, and my Fsys accepts the -H without complaint,
but lacking documentation, I don’t know what kind of argument
to give it, or how to interpret the results.
Is this relevant to the problem at hand? If so, what version
of Fsys and/or Fsys.eide is required? (I haven’t advanced to
G because I still lack a CD installer - we’ve been running
on downloaded updates for quite awhile now!)
Thanks, -Norton Allen
Hugh Brown wrote:
Have you tried using the old loader? ‘dinit -b -O /dev/hd0t77’
No, I hadn’t tried that. Just did, and it boots!
But now I get the old spinning arrow instead of dots.
That’s a problem 
Thanks, Hugh.
Anyone care to explain why one works and the other doesn’t?
-Norton
Previously, Norton Allen wrote in qdn.public.qnx4:
Hugh Brown wrote:
Have you tried using the old loader? ‘dinit -b -O /dev/hd0t77’
No, I hadn’t tried that. Just did, and it boots!
But now I get the old spinning arrow instead of dots.
That’s a problem > > Thanks, Hugh.
Anyone care to explain why one works and the other doesn’t?
I have no idea! All the new Dell laptops seem to exhibit the same problem,
so I guess something has changed in their BIOS.
-Norton
Not only the laptops; also some of the newer 2G desktops. (Also had this
problem and also discovered that the old boot loader worked.) I speculated
that it might have had to do with the new “ultra 100 eide” disks (or
whatever they’re called) with which the new bootloader couldn’t cope
with(?); and that the old bootloader worked because it would turn off
multi-sector transfer during booting(??) [as I said: speculation]
BTW, I found that the I/O performance of these new disks is worse than
older EIDE disks. (Around 1 to 2 M/s) I did simple tests (timing a cp
from one file to another or to /dev/null). I’d expect that this new “ultra
100” stuff would perform better. Are there any Fsys.eide option that I
can play with to squeeze more through-put out of it?
regards,
rick
Hugh Brown <hsbrown@qnx.com> wrote:
Previously, Norton Allen wrote in qdn.public.qnx4:
Hugh Brown wrote:
Have you tried using the old loader? ‘dinit -b -O /dev/hd0t77’
No, I hadn’t tried that. Just did, and it boots!
But now I get the old spinning arrow instead of dots.
That’s a problem > > Thanks, Hugh.
Anyone care to explain why one works and the other doesn’t?
I have no idea! All the new Dell laptops seem to exhibit the same problem,
so I guess something has changed in their BIOS.
-Norton
Rick I have some question concerning you hardware, would you be so kind as
to send me your email address at mcharest at zinformatic dot com.
“Rick Lake” <rwlake@spamfree.domain.invalid> wrote in message
news:bbg8vt$bsh$1@inn.qnx.com…
Not only the laptops; also some of the newer 2G desktops. (Also had this
problem and also discovered that the old boot loader worked.) I speculated
that it might have had to do with the new “ultra 100 eide” disks (or
whatever they’re called) with which the new bootloader couldn’t cope
with(?); and that the old bootloader worked because it would turn off
multi-sector transfer during booting(??) [as I said: speculation]
BTW, I found that the I/O performance of these new disks is worse than
older EIDE disks. (Around 1 to 2 M/s) I did simple tests (timing a cp
from one file to another or to /dev/null). I’d expect that this new “ultra
100” stuff would perform better. Are there any Fsys.eide option that I
can play with to squeeze more through-put out of it?
regards,
rick
Hugh Brown <> hsbrown@qnx.com> > wrote:
Previously, Norton Allen wrote in qdn.public.qnx4:
Hugh Brown wrote:
Have you tried using the old loader? ‘dinit -b -O /dev/hd0t77’
No, I hadn’t tried that. Just did, and it boots!
But now I get the old spinning arrow instead of dots.
That’s a problem > > Thanks, Hugh.
Anyone care to explain why one works and the other doesn’t?
I have no idea! All the new Dell laptops seem to exhibit the same
problem,
so I guess something has changed in their BIOS.
-Norton
Hi Mario,
I’m not sure I can reach you at your email address at zinformatic. (I get
connection refused by newmarket dot qenesis dot com) Anyway, you can email
me at my address: rwlake at anp dot nl. (Which I thought you already knew,
since you sent me a fake spam once as a joke
regards,
rick
Mario Charest postmaster@127.0.0.1 wrote:
Rick I have some question concerning you hardware, would you be so kind as
to send me your email address at mcharest at zinformatic dot com.
“Rick Lake” <> rwlake@spamfree.domain.invalid> > wrote in message
news:bbg8vt$bsh$> 1@inn.qnx.com> …
Not only the laptops; also some of the newer 2G desktops. (Also had this
problem and also discovered that the old boot loader worked.) I speculated
that it might have had to do with the new “ultra 100 eide” disks (or
whatever they’re called) with which the new bootloader couldn’t cope
with(?); and that the old bootloader worked because it would turn off
multi-sector transfer during booting(??) [as I said: speculation]
BTW, I found that the I/O performance of these new disks is worse than
older EIDE disks. (Around 1 to 2 M/s) I did simple tests (timing a cp
from one file to another or to /dev/null). I’d expect that this new “ultra
100” stuff would perform better. Are there any Fsys.eide option that I
can play with to squeeze more through-put out of it?
regards,
rick
Hugh Brown <> hsbrown@qnx.com> > wrote:
Previously, Norton Allen wrote in qdn.public.qnx4:
Hugh Brown wrote:
Have you tried using the old loader? ‘dinit -b -O /dev/hd0t77’
No, I hadn’t tried that. Just did, and it boots!
But now I get the old spinning arrow instead of dots.
That’s a problem > > Thanks, Hugh.
Anyone care to explain why one works and the other doesn’t?
I have no idea! All the new Dell laptops seem to exhibit the same
problem,
so I guess something has changed in their BIOS.
-Norton
Helped also me.
Thanks Dieter
Dieter Schemmelmann schrieb:
It seems to be a similar problem as “install problem” from 21.03.2003"