Date: Tue, 25 May 2010 16:15:35 +0300 From: Alexander Motin <mav@FreeBSD.org> To: John Baldwin <jhb@freebsd.org> Cc: freebsd-arch@freebsd.org Subject: Re: Event timers Message-ID: <4BFBCD77.5030900@FreeBSD.org> In-Reply-To: <201005241530.05116.jhb@freebsd.org> References: <4BFAA1F3.1070206@FreeBSD.org> <201005241530.05116.jhb@freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
John Baldwin wrote: > On Monday 24 May 2010 11:57:39 am Alexander Motin wrote: >> You should agree that event timers code we have now (hardclock, >> statclock and profclock source) is real mess. The code is completely >> machine-dependent, many times duplicated (I have found 4 or 5 equal >> implementations and many more trivial ones), very tangled and >> inflexible. While it is acceptable for some platforms with single >> interrupt source running on hz, x86 with it's many different timers >> definitely doesn't look good. So I've decided to clean it. >> >> I have defined several points on that way: >> 1*. clean low-level timer drivers from unrelated stuff, >> 2*. make some common code machine-independent, >> 3. write common driver API for event timers (alike to one we have now >> for time counters) to make adding more drivers possible, >> 4. add support for HPET as event timer in addition to time counter, >> 5. add support for timers in one-shot mode (LAPIC and HPET). > > Have you looked at the dynamic ticks patches on hackers@? Those already > include changes for 5), although he currently only has support for LAPIC and > is working on adding HPET support. Yes, I've seen it. But without complete timer stuff rewrite it seems hopeless to me. That is one more reason why I want to do it. -- Alexander Motin
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4BFBCD77.5030900>