Subject: Re: unexpected halt of system while ... problem solved !
To: Wilfried Heller <Ing.-Buero_Heller@t-online.de>
From: Chris G. Demetriou <email@example.com>
Date: 03/07/1998 11:31:23
> I probably found out, what went wrong:
> -The statically allocated printer-buffer size was too small to
> handle a brute force 'cat anytext > /dev/lpt'. Probably the
> printer-buffer was overun and caused an illegal memory access.
> The workaround is , to increase the buffer-size.
> the file in question is '/sys/dev/ic/lpt.c'.=20
> I increased '#define LPT_BSIZE 1024' to 4098.
> Perhaps a 'LPT_BSIZE =3D pagesize used' would be better.
> The best solution would be, if this buffer could dynamically
> resize itself. Is this possible ???
"I suggest you read the driver again."
If making the change that you suggest does in fact fix the problem for
you, that's interesting, because it indicates a set of possible places
where the timing problem could be.
But your suggestion of why the cat fails is ... silly.