Very Computer

Something happened with the floppies; I don't know what.  First I
tried to write to B: with no disk in there, but I managed to kill the
offending process (why is so hard to kill processes that are stuck
dealing will ill-prepared floppies?).  But then when I tried to write
to A: (which >did< have a disk in it) I got an endless stream of
``reset-floppy called'' messages, which followed me around from
VC to VC.

First, there's a bug somewhere (I know that's not real helpful)
because accessing A: should have worked.

And second, it seems to me that after a certain amount of floppy
resets the kernel should just give up, or at least let the offending
process die a peaceful death.

Finally, why do these messages appear on every VC?  In
/etc/syslog.conf, I (thought I had) instructed the messages only to
appear on tty8.

-Joel

--
-----------------------------------------------------------------------------
|_|~~ Germany, 1943.   ``A little garden, fragrant and full of roses.
__|~| 16 Million DEAD.   The path is narrow, and a little boy walks along it.
                         A little boy, a sweet boy, like that growing blossom,
 cnc  Bosnia, 1993.      When the blossom comes to bloom,
 cnc  HOW MANY MORE?     The little boy will be no more.''

                                        - Franta Bass,
                                          killed at age 14 by the Nazis
-----------------------------------------------------------------------------

|> Something happened with the floppies; I don't know what.  First I
|> tried to write to B: with no disk in there, but I managed to kill the
|> offending process (why is so hard to kill processes that are stuck
|> dealing will ill-prepared floppies?).  But then when I tried to write
|> to A: (which >did< have a disk in it) I got an endless stream of
|> ``reset-floppy called'' messages, which followed me around from
|> VC to VC.
|>
|> First, there's a bug somewhere (I know that's not real helpful)
|> because accessing A: should have worked.
|>
|> And second, it seems to me that after a certain amount of floppy
|> resets the kernel should just give up, or at least let the offending
|> process die a peaceful death.
|>
|> Finally, why do these messages appear on every VC?  In
|> /etc/syslog.conf, I (thought I had) instructed the messages only to
|> appear on tty8.
|>
|> -Joel

|> --

Sorry, what is A: and B: under Linux ?????
Do you still think in messy-DOS ?

Or do you talk about mtools ?

However:

1) What you call A: is /dev/fd0, B: is /dev/fd1.
2) Disks have to be formatted:
    fdformat /dev/fd0H1440 (for a 3.5" HD Drive) will do the job (must be root)
3) If you want to have msdos-formatted disk, type
    mformat a: (or b:)
4) If you want to have a Minix filesystem on these disks, type
    mkfs /dev/fd0 (or /dev/fd1) <number of blocks>,
   where number of blocks is 1440 for 3.5" HD.
5) If you want to tar to or from the floppy, type
    tar xvf /dev/fd0   or   tar cvf /dev/fd0 <files>

6) If still something's going wrong, just pull out the floppy. The error
   messages will stop after a while !

Everything clear ?

Andy

|> 1) What you call A: is /dev/fd0, B: is /dev/fd1.
|> 2) Disks have to be formatted:
|>     fdformat /dev/fd0H1440 (for a 3.5" HD Drive) will do the job (must be root)
|> 3) If you want to have msdos-formatted disk, type
|>     mformat a: (or b:)
|> 4) If you want to have a Minix filesystem on these disks, type
|>     mkfs /dev/fd0 (or /dev/fd1) <number of blocks>,
|>    where number of blocks is 1440 for 3.5" HD.
|> 5) If you want to tar to or from the floppy, type
|>     tar xvf /dev/fd0   or   tar cvf /dev/fd0 <files>
|>
|> 6) If still something's going wrong, just pull out the floppy. The error
|>    messages will stop after a while !
|>
|> Everything clear ?

Not so darn *fast* Andy :-)

I've got two machines running Linux 0.99pl9 from SLS 1.02. One, a no-name VESA
486DX33 never has problems with the floppy. But the other, a Compaq Contura
3/25, often gets in a twist with the floppy drive. Hmmmmmm, maybe there's a
lesson in that :-)

As Joel says, on accasion (but not always) when I try to cpio (or tar) from (or
to) the floppy I get the dreaded 'reset-floppy called' message and the whole
machine jumps about like it's going to blow up (slight exageration there ;-).

The only way to shut the damm thing up is to switch it off :-( Your claim that
ejecting the floppy cures the problem is, unfortunatly, incorrect in my case.

So, questions remain:

        What causes this to happen?
        Why can't the faulty process be interupted?

Actualy I think I may know the answer to the second one - could be that I have
the console device linked to /dev/tty0 so the error messages appear on all
vc's which makes it a * to log in again and type in a ps/kill sequence.

BTW. The floppies in question were formated on a variety of UNIX's: AIX,
Interactive & Xenix in particular.

-- David
________________________________________________________________________________
David Harrison                              Phone  +44 (0)565 621000 Ext 4197
UNIX Systems Support Team                   Fax    +44 (0)565 654473
Barclays Computer Operations                Mobile 0836 583085 (UK only)

Cheshire WA16 9EU                           #include <disclaimer.h>

|> |> 1) What you call A: is /dev/fd0, B: is /dev/fd1.
|> |> 2) Disks have to be formatted:
|> |>     fdformat /dev/fd0H1440 (for a 3.5" HD Drive) will do the job (must be root)
|> |> 3) If you want to have msdos-formatted disk, type
|> |>     mformat a: (or b:)
|> |> 4) If you want to have a Minix filesystem on these disks, type
|> |>     mkfs /dev/fd0 (or /dev/fd1) <number of blocks>,
|> |>    where number of blocks is 1440 for 3.5" HD.
|> |> 5) If you want to tar to or from the floppy, type
|> |>     tar xvf /dev/fd0   or   tar cvf /dev/fd0 <files>
|> |>
|> |> 6) If still something's going wrong, just pull out the floppy. The error
|> |>    messages will stop after a while !
|> |>
|> |> Everything clear ?
|>
|> Not so darn *fast* Andy :-)

Sorry. Next time I'll try to explain more slowly. :-)

|>
|> I've got two machines running Linux 0.99pl9 from SLS 1.02. One, a no-name VESA
|> 486DX33 never has problems with the floppy. But the other, a Compaq Contura
|> 3/25, often gets in a twist with the floppy drive. Hmmmmmm, maybe there's a
|> lesson in that :-)
|>
|> As Joel says, on accasion (but not always) when I try to cpio (or tar) from (or
|> to) the floppy I get the dreaded 'reset-floppy called' message and the whole
|> machine jumps about like it's going to blow up (slight exageration there ;-).
|>
|> The only way to shut the damm thing up is to switch it off :-( Your claim that
|> ejecting the floppy cures the problem is, unfortunatly, incorrect in my case.
|>
|> So, questions remain:
|>
|>   What causes this to happen?
|>   Why can't the faulty process be interupted?
|>
|> Actualy I think I may know the answer to the second one - could be that I have
|> the console device linked to /dev/tty0 so the error messages appear on all
|> vc's which makes it a * to log in again and type in a ps/kill sequence.
|>
|> BTW. The floppies in question were formated on a variety of UNIX's: AIX,
|> Interactive & Xenix in particular.
|>
|> -- David

If those error message won't stop after pulling out the disk you might have to kill
the process still trying or reading from the devide. Perhaps a ^C will work, but
probably you'll have to do a #ps -ax to see what's going on.
One problem might be the different floppy drives. Although this isn't a very
good solution, you might have to throw out the Compaq drive and buy a new
one. I never had trouble with TEAC drives, but SONY drives are *always* calibrated
a bit different.

Andy

[Who wrote what deleted]

Quote:>If those error message won't stop after pulling out the disk you might have to kill
>the process still trying or reading from the devide. Perhaps a ^C will work, but
>probably you'll have to do a #ps -ax to see what's going on.
>One problem might be the different floppy drives. Although this isn't a very
>good solution, you might have to throw out the Compaq drive and buy a new
>one. I never had trouble with TEAC drives, but SONY drives are *always* calibrated
>a bit different.

The problem is that kill doesn't work, at least not for me.  I usually
try just "kill" first, then a few "kill -9"'s and nothing helps.  I
understand that this is common, standard behavior for processes doing
what these are doing (I forget the details -- I think it has to do
with the system calls or something), but at any rate the processes are
un-killable.  

-Joel

--
-----------------------------------------------------------------------------
|_|~~ Germany, 1943.   ``A little garden, fragrant and full of roses.
__|~| 16 Million DEAD.   The path is narrow, and a little boy walks along it.
                         A little boy, a sweet boy, like that growing blossom,
 cnc  Bosnia, 1993.      When the blossom comes to bloom,
 cnc  HOW MANY MORE?     The little boy will be no more.''

                                        - Franta Bass,
                                          killed at age 14 by the Nazis
-----------------------------------------------------------------------------