Re: 2.0.35 lp.c very slow

Andrea Arcangeli (arcangeli@mbox.queen.it)
Fri, 24 Jul 1998 01:57:13 +0200 (CEST)


On Thu, 23 Jul 1998, Alan Cox wrote:

>> lp.c in 2.0.35 is very slow, at least under polling. 20 min to
>> print a 40KB monochrome postscript file. The boot message is
>> "lp1 at 0x0378, (polling)".
>>
>> Fixed by using lp.c from 2.0.34. (Prints in about 1 min.)
>
>Ok thanks for the report. So fixing one printer slows others down. Does anyone
^^^^^^^^^^^^^^^^^^

Use `tunelp -w 1000` with 2.0.34 and the one printer will be cured ;-). As
default lp->wait is 0 so lp put the STROBE active for zero loop (I think
that 0 is not the best default setting for LP_INIT_WAIT, the default
should be run everywhere).

I was not aware of the 35 lp patch. Make sense but the problem is the
schedule(). There' s really no time and no need for a schedule there.

2.1.x use the old tunelp way to fix the problem. I am not sure how
checking every time for BUSY could impact performance with very fast
printer and not very fast CPU/BUS clock. The 2.1.x tunelp way _can'_t_ be
slower (even if it' s bit less PnP ;-).

Andrea[s] Arcangeli

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.altern.org/andrebalsa/doc/lkml-faq.html