Date: Sun, 23 Dec 2001 14:00:56 -0500 From: "Louis A. Mamakos" <louie@TransSys.COM> To: Robert Watson <rwatson@FreeBSD.ORG> Cc: Greg Lehey <grog@FreeBSD.ORG>, Bruce Evans <bde@FreeBSD.ORG>, Matt Dillon <dillon@FreeBSD.ORG>, cvs-committers@FreeBSD.ORG, cvs-all@FreeBSD.ORG, FreeBSD Core Team <core@FreeBSD.ORG> Subject: Re: cvs commit: src/sys/dev/sio sio.c Message-ID: <200112231900.fBNJ0u745326@whizzo.transsys.com> In-Reply-To: Your message of "Sun, 23 Dec 2001 12:33:12 EST." <Pine.NEB.3.96L.1011223120751.8511h-100000@fledge.watson.org> References: <Pine.NEB.3.96L.1011223120751.8511h-100000@fledge.watson.org>
next in thread | previous in thread | raw e-mail | index | archive | help
It's been a couple of years since I did an experiment to measure interrupt latency, and that on FreeBSD 3.2-CURRENT base. Here we are two years later, and it's unclear that end-users are likely to see any significant improvement in their end-user experience. Having the end-user suffer from continuing serial I/O input silo errors hardly seems like a useful result. Surely this email thread will be a much more effective reminder of the interrupt latency issue than the silo overflow printf's from the sio driver. So, speaking as a FreeBSD end-user - please fix the behavior that I see by the unclean, un-holy, and architecturally bankrupt expediant hack of changing the input silo theshold. This is seems to be morally offensive to some; perhaps that will serve as motivation to fix the underlying cause rather than just ignoring it. I'm all for curing the disease rather than just treating the symptoms, but this aching pain has gone on for quite a while when a simple work-around is readily available. There must be more offensive crocks on the FreeBSD source tree to get wound-up over than this. louie To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe cvs-all" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200112231900.fBNJ0u745326>