Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 29 Jan 2010 21:08:14 +0100
From:      Jens Schweikhardt <schweikh@schweikhardt.net>
To:        questions@freebsd.org
Subject:   Printing via ulpt0 extremely slow
Message-ID:  <20100129200814.GA1919@schweikhardt.net>

next in thread | raw e-mail | index | archive | help
hello, world\n

I have a system with a handful of kernels I chose from with grub.
Recently compiled 8-STABLE systems show strange printing behaviour.
While a one year old 8.0-CURRENT #0 r185532 has no problem printing to
my HP Laserjet 2300d (via cups and USB/ulpt0), newer systems and even
9-CURRENT print extremely slow, on the order of 1 page every 6 minutes.
The printer's "Data" LED blinks sometimes erratically, sometimes is on
for a few seconds, with intermittent periods of 1Hz blinking (which is
the expected normal behavior).

So I'm wondering what causes this oddity. I've ruled out an issue
with hald/dbus which recent systems use for xorg 7.4, by turning them
off, rebooting and printing from the console--same slow printing.

The cups log says it sent the file succesfully (/var/log/cups/access_log):
localhost - - [29/Jan/2010:20:17:11 +0100] "POST /printers/LaserJet_2300d HTTP/1.1" 200 18530 Send-Document successful-ok

I can't find anything obvious in my kernel config that might account
for this behavior.

It does not matter if the printer is on or off when the system
starts.

I've read about interrupt storms (when printing via lpt0), but vmstat -i
looks sane AFAICT:

$ vmstat -i
interrupt                          total       rate
irq16: vgapci0 ahc*               192947         62
irq18: skc0 uhci2++                 3765          1
irq19: fwohci0++                  383725        124
irq23: uhci3 ehci1                  3700          1
cpu0: timer                      6227448       2018
irq256: hdac0                         92          0
cpu1: timer                      6219346       2015
Total                           13031023       4223

Anyone seen something similar? What else can I try to debug this
problem?

Regards,

	Jens
-- 
Jens Schweikhardt http://www.schweikhardt.net/
SIGSIG -- signature too long (core dumped)



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20100129200814.GA1919>