Re: modules memory (Was: sound volume and kerneld)

Luca Berra (bluca@rettore.rett.polimi.it)
Mon, 15 Apr 1996 11:37:31 +0200 (METDST)

Mike Civil said:
>
>
> In article <96Apr10.082022edt.106763-145+561@vger.rutgers.edu> you write:
> >Actually the problem of keeping settings between modules invocation
> >is not limited to the sound module. Another module which applies is
> >the tape module (i use st.o, but i think also ftape.o applies).
> >the problem is remembering the current file position when you have
> >multiple filemarks.
> >ie: you do
> ># mt fsf 12
> >go drink a coffe while it goes,
> ># mt status
> >and it tells you are at file 0, while the tape IS at file 12 :(
>
> Surely the default action for mt is to use the rewind device (unless you
> override it)? Then I'd expect the tape to be at file 0 after the commands
> above....:)
>
no, the default device for mt is the no-rewind, else mt fsf would not make
sense. anyway if i check the status immediatly after the mt command
it displays the correct result, but if i leave it alone for more than 1 sec
well....
> As regards the floppy tape driver, the zftape author has done this by not
> decrementing the reference count if we use the non-rewind device. This has
> the effect of pinning the module until we use a rewind device. Not ideal in
> the general case, but it works pretty well for tape handling.
this may work, but i still would prefer to handle these issues in a more
general way
>
> Just my $0.02.
>
> Cheers,
>
> Mike
>
thanks for your help

Regards,
Luca

-- 
#include <disclaimer.h>
+----------------------------------------------------------------------------+
|      Luca Berra -- S.I.W.A. Politecnico di Milano                          |
|          Phone: +39-2-23992413  -  e-mail: bluca@www.polimi.it             |
|       PGP public key available via finger                                  |
+----------------------------------------------------------------------------+