Date: Thu, 20 May 1999 19:30:13 -0500 From: Constantine Shkolnyy <stan@osgroup.com> To: "freebsd-hackers@FreeBSD.ORG" <freebsd-hackers@FreeBSD.ORG> Subject: RE: Lazy SPLs Message-ID: <01BEA2F7.2F422B40.stan@osgroup.com>
next in thread | raw e-mail | index | archive | help
On Thursday, May 20, 1999 7:18 PM, Joel Ray Holveck [SMTP:joelh@gnu.org] wrote: > >>> "Lazy SPLs - The kernel no longer masks hardware events unless a > >>> hardware event actually occurs, avoiding many expensive > >>> operations." > >> We've been doing it for as long as I can remember, at least as far > >> back as 2.0.5, probably as far back as 1.x. > > My earliest memory of it was as "Bruce's new interrupt code" for 386bsd. > > > > It was part of the 386bsd patchkit I think. > > Why mask out the interrupts at all, instead of queuing them in handler > level? Because only the device's driver knows how to stop the device from interrupting again and again, but calling its handler is prohibited. Lazy SPLs is an optimization. Drivers play with SPLs very often, so it would be unefficient to program the interrupt controller each time when somebody wants to increment a counter. However, if the device has indeed interrupted, there is no choice left except disabling it in the interrupt controller. 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?01BEA2F7.2F422B40.stan>