Date: Tue, 11 May 1999 19:20:17 +1000 From: Bruce Evans <bde@zeta.org.au> To: bde@freebsd.org, dg@freebsd.org, ru@ucb.crimea.ua Cc: hackers@freebsd.org Subject: Re: CY_PCI_FASTINTR (Was: Re: Cyclom-Y driver for FreeBSD - help!!!) Message-ID: <199905110920.TAA02979@godzilla.zeta.org.au>
next in thread | raw e-mail | index | archive | help
>Am I understand this right: if I configure my PCI card to an >exclusive interrupt, and then turn on CY_PCI_FASTINTR, I'll get >rid of ``silo overflows'' on it. Except in -current, where CY_PCI_FASTINTR is broken (it has no effect). >Just out of curiosity. >What is the ``interrupt latency in the FreeBSD kernel'' problem >and what are the ``Bruce's "fast" interrupt hacks''? The problem is that some mouldy old drivers and subsystems disable interrupts for too long (e.g. 2-5msec for programming keyboard LEDs; this breaks unbuffered serial input at speeds larger than 200-500 bps, and it breaks the Cyclades' 12-character fifos at speeds larger than 600-2400 bps). The "fast" interrupt hacks consist of ignoring the normal interrupt masking system so that drivers using "fast" interrupts are not affected by the problem unless there are enough of them (or enough activity on them) to create their own latency problems. Bruce To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199905110920.TAA02979>