Red Hat Bugzilla – Bug 29573
erroneous IRQ conflict message
Last modified: 2005-10-31 17:00:50 EST
When lpd starts it scans all ports and then reports irq conflict. Have
added irq=none to modules.conf with no change on startup. Fisher beta did
not have this problem although lpd didn't start on boot.
One other note: there is no installed /proc/parport directory.
This looks more like a kernel issue rather than an LPRng issue.
Yes, the message is incorrect and should be ignored.
/proc/parport has moved to /proc/sys/dev/parport, and the ports have names
'parport0', 'parport1', instead of '0', '1', etc.
I would like to ignore the error message but printing a simple page takes
anywhere from 1 to 2 minutes
If you print a page of text by printing directly to the port, is that slow too?
cat textfile > /dev/lp0
Also, is it consistently slow, or does print normally and then stall?
See also bug #26996, which might or might not be what you're seeing.
cat textfile > /dev/lp0 produces the same result. Have looked at Bug 26996 which
appears to have a similar stalling problem except this printer stalls from the
get go. Appears that much troubleshooting has occurred on 26996 with no solution
Will try other options if requested.
Please try this patch, which I think is the resolution for bug 26996.
Created attachment 11360 [details]
Ed, are you still getting slow printing?
Yes, because I was unable to successfully apply the linux-sema.patch and
rebuild the kernel.
Could you try a newer kernel? -0.1.19 is in rawhide now.
Printing is working using the -0.1.19 kernel!
Some other possibly irrelevent info:
1. Still receive the irq error message on boot.
2. Installed -i the -0.1.19 kernel and -U the headers.
3. Had to hand edit lilo.conf to add -0.1.19.
Thanks for your help.
Okay, great. I've made a patch to get rid of the warning message since it's
triggered by unreliable information, and it's in my queue to send to Linus.
I assume this patch is in our latest kernels and will close this bug.
It isn't, although I've sent it to Linus twice now.
FWIW, this cosmetic bug is fixed in 2.4.3.