![]() |
Thanks much for the response.. Ill look into reporting the issue to my kernel maintainer. You mention below mainpine is trying to create a work-around? Im curious do you have any idea what form that work-around will take? An update to the fax boards firmware or some such thing or some alternative serial driver for linux? Thanks again -Mike On Fri, Sep 12, 2008 at 2:35 PM, Lee Howard <lee.howard@xxxxxxxxxxxx> wrote: > I've had numerous conversations with a few people regarding this topic... > Furthermore, I've done a lot of research on it... And I still don't know > for-certain what the problem is or how, exactly, to fix it. > > What seems clear, however, is that there is a bug in the kernel serial > driver that is, for some reason, more prominent on newer kernel versions and > is, for some reason, more prominent when using high port-count IQ Express > fax cards (although there have been reports of this kind of thing using > other serial hardware and software). > > There was a confirmed bug report with two patches from Robert Evans of > Stratus Technologies that circulated on the linux-serial mailing list on 13 > August that dealt specifically with SMP systems. (The systems on which I've > reproduced this problem have all been SMP.) I do not know if the patches > were ever committed to the Linux kernel source repository - although one is > apparently in the RedHat 5.2 version - or if they will be making it in to a > subsequent kernel release. Furthermore, I do not know if the patches will > resolve the matter. (I have not tested them.) > > http://article.gmane.org/gmane.linux.serial/2633 > > Mainpine has chosen to develop a work-around to the problem which won't be > available for another few weeks. Please check back with me for updates on > that. Meanwhile, I would suggest reporting a bug to the kernel maintainer > for your distribution. Please do point them at the above-mentioned URL. > Hopefully they can provide you with a patched kernel which you could test. > > Thanks, > > Lee. > > Lee Howard > Mainpine, Inc. Software Development Lead > Tel: +1 866 363 6680 ext 5 | Fax: +1 360 462 8160 > lee.howard@xxxxxxxxxxxx | www.mainpine.com > > -----Original Message----- > From: msbrentlinger@xxxxxxxxx [mailto:msbrentlinger@xxxxxxxxx] > Sent: Friday, September 12, 2008 11:04 AM > To: Lee Howard > Cc: hylafax-users@xxxxxxxxxxx > Subject: Re: IRQ Overload > > > I have the same issue as well.. Did anyone ever come to some resolution to > this issue? > > I have HylaFAX (tm) Version 5.2.7 from hylafax-5.2.7-1.fc9.i386.rpm > installed on Fedora release 9 (Sulphur) 2.6.25-14.fc9.i686 with a Mainpine > IQ Express 8-port (Low Profile) RF2123 > > > at night we send a batch of anywhere from a few dozen to a few hundred > faxes, things start off fine and then. > > /var/log/messages starts showing > Sep 11 20:33:10 hylafax kernel: serial8250: too much work for irq16 Sep 11 > 20:33:10 hylafax kernel: serial8250: too much work for irq16 Sep 11 20:33:10 > hylafax kernel: serial8250: too much work for irq16 Sep 11 20:33:10 hylafax > kernel: serial8250: too much work for irq16 Sep 11 20:33:10 hylafax kernel: > serial8250: too much work for irq16 Sep 11 20:33:10 hylafax kernel: > serial8250: too much work for irq16 Sep 11 20:33:10 hylafax kernel: > serial8250: too much work for irq16 ... > ... > > shortly there after the remaining faxes that are queued up go out start > failing with' too many attempts to dial'. > > Eventually things seem to resolve itself because in the morning things are > going fine again without any reboots or anything. Throughout the day we do > send occasional faxes, but maybe 2-3 at a time every 10 minutes or so... > Nothing like our larger night batches of a couple hundered that get queued > up all at once. > > > > > ____________________ HylaFAX(tm) Users Mailing List _______________________ To subscribe/unsubscribe, click http://lists.hylafax.org/cgi-bin/lsg2.cgi On UNIX: mail -s unsubscribe hylafax-users-request@xxxxxxxxxxx < /dev/null *To learn about commercial HylaFAX(tm) support, mail sales@xxxxxxxxx*