From owner-freebsd-smp Sun Feb 18 0:15:46 2001 Delivered-To: freebsd-smp@freebsd.org Received: from syncopation-01.iinet.net.au (syncopation-01.iinet.net.au [203.59.24.37]) by hub.freebsd.org (Postfix) with SMTP id 2A25C37B491 for ; Sun, 18 Feb 2001 00:15:39 -0800 (PST) Received: (qmail 20969 invoked by uid 666); 18 Feb 2001 08:27:37 -0000 Received: from i003-020.nv.iinet.net.au (HELO elischer.org) (203.59.3.20) by mail.m.iinet.net.au with SMTP; 18 Feb 2001 08:27:37 -0000 Message-ID: <3A8F8495.F7313ECD@elischer.org> Date: Sun, 18 Feb 2001 00:15:17 -0800 From: Julian Elischer X-Mailer: Mozilla 4.7 [en] (X11; U; FreeBSD 5.0-CURRENT i386) X-Accept-Language: en, hu MIME-Version: 1.0 To: Seth Leigh Cc: freebsd-smp@FreeBSD.ORG Subject: Re: Well I read the stuff, and I get it now. References: <5.0.2.1.0.20010218021929.00aaef98@hobbiton.shire.net> Content-Type: text/plain; charset=iso-8859-15 Content-Transfer-Encoding: 7bit Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Seth Leigh wrote: > > > Actually, I just thought of how it could be done while re-reading my post > prior to sending it off. Is this what you guys have thought of too? The > idea is that since the kernel is already getting interupted 100 times per > second (or however many times FreeBSD does it) anyhow, the running > scheduler activation is *already* going to be preempted off the cpu for the > duration of that tick processing. So, after the tick processing is done > and the kernel dispatcher decides that this particular scheduling > activation may continue running as it was doing before the timer interupt > fired, rather than simply context switch back into that particular > scheduler activation, the kernel would use a *second* scheduler activation > and upcall into the threads library's scheduler. This would basically > allow the threads library's scheduler to "piggyback" onto the kernel's > scheduler, without requiring anymore crossings of the protection boundaries > than were going to be had anyhow. Basically, this scheme would use twice > as many scheduler activations as it wanted to really have be run, basically > using half of them to call up into the threads library after each tick to > decide whether to keep running the preempted thread or scheduler a > different one. > > What do you all think? Or is this already the plan? kind-of.. Basically all returns from the kernel to the user process might go first via the UTS (Userland Thread Scheduler). This includes page faults and interrupts of some types. We have not completely decided how many and which. In addition, timer signals can be delivered after clock interrupts as well. THe current code uses timer signals, but upcalls are almost the same thing. The saved state of the thread (when a timer interrupt completes and retunr is sent back to the UTS) is made to b ecompatible with all stopped threads so that the interrupted thread looks like all the other runnable threads, and the UTS can simply decide which to restart. > > Seth > > To Unsubscribe: send mail to majordomo@FreeBSD.org > with "unsubscribe freebsd-smp" in the body of the message -- __--_|\ Julian Elischer / \ julian@elischer.org ( OZ ) World tour 2000-2001 ---> X_.---._/ v To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Sun Feb 18 6:56:14 2001 Delivered-To: freebsd-smp@freebsd.org Received: from gtei1.bellatlantic.net (gtei1.bellatlantic.net [199.45.40.145]) by hub.freebsd.org (Postfix) with ESMTP id 3E4E237B491 for ; Sun, 18 Feb 2001 06:56:11 -0800 (PST) Received: from me-513q3sc0zun0.pengar.com (adsl-64-223-147-101.mannh.adsl.bellatlantic.net [64.223.147.101]) by gtei1.bellatlantic.net (8.9.1/8.9.1) with ESMTP id JAA01427; Sun, 18 Feb 2001 09:55:50 -0500 (EST) Message-Id: <5.0.2.1.0.20010218094431.00aaf0a8@hobbiton.shire.net> X-Sender: seth-pc@hobbiton.shire.net X-Mailer: QUALCOMM Windows Eudora Version 5.0.2 Date: Sun, 18 Feb 2001 09:52:37 -0500 To: Julian Elischer From: Seth Leigh Subject: Re: Well I read the stuff, and I get it now. Cc: freebsd-smp@FreeBSD.ORG In-Reply-To: <3A8F8495.F7313ECD@elischer.org> References: <5.0.2.1.0.20010218021929.00aaef98@hobbiton.shire.net> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii"; format=flowed Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org At 12:15 AM 2/18/2001 -0800, Julian Elischer wrote: >Seth Leigh wrote: >Basically all returns from the kernel to the user process might go first >via the UTS (Userland Thread Scheduler). This includes page faults and >interrupts of some types. We have not completely decided how many >[snip] >The saved state of the thread (when a timer interrupt completes and >retunr is sent back to the UTS) is made to b ecompatible with all >stopped threads so that the interrupted thread looks like all the >other runnable threads, and the UTS can simply decide which to >restart. Ah, ok, I get it. So you are saying anytime the kernel takes over control on any given processor, the means of the kernel giving control back when it's done doing whatever it was doing would be to upcall back to the process in the context of that scheduler activation, allowing the threads library to be able to make the decision of whether to keep running that thread or not? That's pretty darn cool. Guys, I am very interested in this whole thing. It's making me want to pull out my old Cyrix 166 and install the lastest and greatest FreeBSD on it (it currently isn't running, and has 2.2.1 on it...) so I can try to get up to the code you guys are on and see if I can help out. I am brand spanking new to kernel stuff, so I am not making any promises. I have been paying attention to a lot of kernel talk, and reading some books and such, so I have a reasonable idea of how a lot of stuff works, at least at a high level. It will take me a while after I start delving into code to *really* get it, and be able to be helpful. Unfortunately I don't have a working FreeBSD machine right now. The hardware is there, just the software is way out of date and no longer properly configured. My 1 ghz AMD t-bird is running Win2K and Solaris 8 (well, the machine *was* running Solaris 8 until I upgraded to a new motherboard 3 weeks ago, I haven't booted back up into Solaris 8 yet since the upgrade since I don't know how well it's going to work since the device tree must look different now). Maybe I could get two more hard drives (there are already three in the machine) so I could install Linux and FreeBSD on it too. Yeah, that's the idea. Seth To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Mon Feb 19 11:52:58 2001 Delivered-To: freebsd-smp@freebsd.org Received: from burka.carrier.kiev.ua (burka.carrier.kiev.ua [193.193.193.107]) by hub.freebsd.org (Postfix) with ESMTP id A114B37B491; Mon, 19 Feb 2001 11:52:11 -0800 (PST) Received: from laa@localhost by burka.carrier.kiev.ua id VUB95832; Mon, 19 Feb 2001 21:52:07 +0200 (EET) (envelope-from laa@lucky.net) X-Authentication-Warning: burka.carrier.kiev.ua: laa set sender to laa@lucky.net using -f Date: Mon, 19 Feb 2001 21:52:06 +0200 From: "Alexandr A. Listopad" To: freebsd-smp@freebsd.org, stable@freebsd.org Cc: "Alexandr A. Listopad" Subject: SMP box & trap12 Message-ID: <20010219215206.A95553@burka.carrier.kiev.ua> Reply-To: "Alexandr A. Listopad" , kgp@zgia.zp.ua, foma@zgia.zp.ua, dexim@zgia.zp.ua Mime-Version: 1.0 Content-Type: text/plain; charset=koi8-r Content-Disposition: inline Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org I have 2 SMP boxes with FreeBSD (4.2-ST) -- old and new. both rebooted too often with trap12 and etc... There also message: interrupt mask = net tty <- SMP : XXX Please help me solve my trouble... Thanks! -- Laa To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Wed Feb 21 10:56:16 2001 Delivered-To: freebsd-smp@freebsd.org Received: from mobile.hub.org (mobile.acadiau.ca [131.162.137.70]) by hub.freebsd.org (Postfix) with ESMTP id 75C5337B65D for ; Wed, 21 Feb 2001 10:56:07 -0800 (PST) (envelope-from scrappy@hub.org) Received: from localhost (scrappy@localhost) by mobile.hub.org (8.11.1/8.11.1) with ESMTP id f1LIu3V17809 for ; Wed, 21 Feb 2001 14:56:03 -0400 (AST) (envelope-from scrappy@hub.org) X-Authentication-Warning: mobile.hub.org: scrappy owned process doing -bs Date: Wed, 21 Feb 2001 14:56:03 -0400 (AST) From: The Hermit Hacker To: Subject: IOPs ... what exactly are they? Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Searching google isn't coming up with anything that explains it ... can someone point me to some docs, or provide an explanation? For instance, Tom stated that the 170 does 4000IOPs, vs the 352 at 7000IOPs ... yet, the 352 will do twice the sustained transferof the 170 (200MB/s vs 100MB/s) ... so, IOPs don't == transfer speed ... And, I imagine (can't get to the Mylex web site right now for some reason), the 2000 has an even higher rating then both of those ... So, when evaluating what one needs, how do you determine? are there thresholds one can work with? 4000IOPs will comfortably handle n xgig drives, but if you go above n, then you really need to jump to the 7k IOP level else performance drops substantially? We're looking at this for a database server, so will most likely be going to the 64MB cache level automatically ... but beyond that, I'm virtually uneducated :( Marc G. Fournier ICQ#7615664 IRC Nick: Scrappy Systems Administrator @ hub.org primary: scrappy@hub.org secondary: scrappy@{freebsd|postgresql}.org To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Thu Feb 22 23:12: 9 2001 Delivered-To: freebsd-smp@freebsd.org Received: from obsecurity.dyndns.org (adsl-63-207-60-188.dsl.lsan03.pacbell.net [63.207.60.188]) by hub.freebsd.org (Postfix) with ESMTP id D523737B401; Thu, 22 Feb 2001 23:12:06 -0800 (PST) (envelope-from kris@obsecurity.org) Received: by obsecurity.dyndns.org (Postfix, from userid 1000) id 7F8DB66F38; Thu, 22 Feb 2001 23:11:54 -0800 (PST) Date: Thu, 22 Feb 2001 23:11:54 -0800 From: Kris Kennaway To: smp@freebsd.org, julian@freebsd.org Subject: NETGRAPH broken with witness Message-ID: <20010222231154.A852@mollari.cthul.hu> Mime-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-md5; protocol="application/pgp-signature"; boundary="EeQfGwPcQSOJBaQU" Content-Disposition: inline User-Agent: Mutt/1.2.5i Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org --EeQfGwPcQSOJBaQU Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Apparently this was never tested: early in the boot phase I get the following panic: panic: spin lock netgraph worklist mutex not in order list and indeed the witness code appears to have no knowledge of the netgraph spinlocks. Kris --EeQfGwPcQSOJBaQU Content-Type: application/pgp-signature Content-Disposition: inline -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.0.4 (FreeBSD) Comment: For info see http://www.gnupg.org iD8DBQE6lg06Wry0BWjoQKURAnJ8AJ9R3bcu/PsKmdmH3TkNWq26uF5QDACfWcWj 1oAhp54lX5v32mVtOPrdpc0= =lG1K -----END PGP SIGNATURE----- --EeQfGwPcQSOJBaQU-- To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Fri Feb 23 1:58:10 2001 Delivered-To: freebsd-smp@freebsd.org Received: from syncopation-01.iinet.net.au (syncopation-01.iinet.net.au [203.59.24.37]) by hub.freebsd.org (Postfix) with SMTP id EEDE737B503 for ; Fri, 23 Feb 2001 01:58:07 -0800 (PST) (envelope-from julian@elischer.org) Received: (qmail 32446 invoked by uid 666); 23 Feb 2001 10:09:39 -0000 Received: from i078-087.nv.iinet.net.au (HELO elischer.org) (203.59.78.87) by mail.m.iinet.net.au with SMTP; 23 Feb 2001 10:09:39 -0000 Message-ID: <3A963410.1C656DB3@elischer.org> Date: Fri, 23 Feb 2001 01:57:36 -0800 From: Julian Elischer X-Mailer: Mozilla 4.7 [en] (X11; U; FreeBSD 5.0-CURRENT i386) X-Accept-Language: en, hu MIME-Version: 1.0 To: Kris Kennaway Cc: smp@freebsd.org, julian@freebsd.org Subject: Re: NETGRAPH broken with witness References: <20010222231154.A852@mollari.cthul.hu> Content-Type: text/plain; charset=iso-8859-15 Content-Transfer-Encoding: 7bit Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Kris Kennaway wrote: > > Apparently this was never tested: early in the boot phase I get the > following panic: > > panic: spin lock netgraph worklist mutex not in order list > > and indeed the witness code appears to have no knowledge of the > netgraph spinlocks. so how is it supposed to know? > > Kris > > -------------------------------------------------------------------------------- > Part 1.2Type: application/pgp-signature -- __--_|\ Julian Elischer / \ julian@elischer.org ( OZ ) World tour 2000-2001 ---> X_.---._/ v To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Fri Feb 23 2: 9: 3 2001 Delivered-To: freebsd-smp@freebsd.org Received: from obsecurity.dyndns.org (adsl-63-207-60-57.dsl.lsan03.pacbell.net [63.207.60.57]) by hub.freebsd.org (Postfix) with ESMTP id BE13037B4EC; Fri, 23 Feb 2001 02:09:00 -0800 (PST) (envelope-from kris@obsecurity.org) Received: by obsecurity.dyndns.org (Postfix, from userid 1000) id D09AF66F95; Fri, 23 Feb 2001 02:08:59 -0800 (PST) Date: Fri, 23 Feb 2001 02:08:59 -0800 From: Kris Kennaway To: Julian Elischer Cc: smp@freebsd.org, julian@freebsd.org Subject: Re: NETGRAPH broken with witness Message-ID: <20010223020859.A4771@mollari.cthul.hu> References: <20010222231154.A852@mollari.cthul.hu> <3A963410.1C656DB3@elischer.org> Mime-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-md5; protocol="application/pgp-signature"; boundary="8t9RHnE3ZwKMSgU+" Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: <3A963410.1C656DB3@elischer.org>; from julian@elischer.org on Fri, Feb 23, 2001 at 01:57:36AM -0800 Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org --8t9RHnE3ZwKMSgU+ Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On Fri, Feb 23, 2001 at 01:57:36AM -0800, Julian Elischer wrote: > Kris Kennaway wrote: > >=20 > > Apparently this was never tested: early in the boot phase I get the > > following panic: > >=20 > > panic: spin lock netgraph worklist mutex not in order list > >=20 > > and indeed the witness code appears to have no knowledge of the > > netgraph spinlocks. >=20 > so how is it supposed to know? =46rom what I can gather, you add the locking hierarchy to be enforced to spin_order_list in kern.mutex.c Kris --8t9RHnE3ZwKMSgU+ Content-Type: application/pgp-signature Content-Disposition: inline -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.0.4 (FreeBSD) Comment: For info see http://www.gnupg.org iD8DBQE6lja7Wry0BWjoQKURAt1JAKDOgOLIyjKstBtnO+q7ACGOtkL9zgCeJxJ2 jW12B+9AZpZSz49CYigJkUQ= =obGE -----END PGP SIGNATURE----- --8t9RHnE3ZwKMSgU+-- To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Fri Feb 23 2:48: 2 2001 Delivered-To: freebsd-smp@freebsd.org Received: from urban.iinet.net.au (urban.iinet.net.au [203.59.24.231]) by hub.freebsd.org (Postfix) with ESMTP id 722F537B401; Fri, 23 Feb 2001 02:47:58 -0800 (PST) (envelope-from julian@elischer.org) Received: from elischer.org (i078-087.nv.iinet.net.au [203.59.78.87]) by urban.iinet.net.au (8.8.7/8.8.7) with ESMTP id SAA04408; Fri, 23 Feb 2001 18:47:48 +0800 Message-ID: <3A963FB5.995FEA7D@elischer.org> Date: Fri, 23 Feb 2001 02:47:17 -0800 From: Julian Elischer X-Mailer: Mozilla 4.7 [en] (X11; U; FreeBSD 5.0-CURRENT i386) X-Accept-Language: en, hu MIME-Version: 1.0 To: Kris Kennaway Cc: smp@freebsd.org, julian@freebsd.org Subject: Re: NETGRAPH broken with witness References: <20010222231154.A852@mollari.cthul.hu> <3A963410.1C656DB3@elischer.org> <20010223020859.A4771@mollari.cthul.hu> Content-Type: text/plain; charset=iso-8859-15 Content-Transfer-Encoding: 7bit Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Kris Kennaway wrote: > > On Fri, Feb 23, 2001 at 01:57:36AM -0800, Julian Elischer wrote: > > Kris Kennaway wrote: > > > > > > Apparently this was never tested: early in the boot phase I get the > > > following panic: > > > > > > panic: spin lock netgraph worklist mutex not in order list > > > > > > and indeed the witness code appears to have no knowledge of the > > > netgraph spinlocks. > > > > so how is it supposed to know? > > From what I can gather, you add the locking hierarchy to be enforced > to spin_order_list in kern.mutex.c hmm tricky.. I thought initiating the damned thing would insert it. I have no idea of the correct ordering. > > Kris > > -------------------------------------------------------------------------------- > Part 1.2Type: application/pgp-signature -- __--_|\ Julian Elischer / \ julian@elischer.org ( OZ ) World tour 2000-2001 ---> X_.---._/ v To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Fri Feb 23 6:54:12 2001 Delivered-To: freebsd-smp@freebsd.org Received: from grendel.bsdi.com (grendel.twistedbit.com [199.79.183.5]) by hub.freebsd.org (Postfix) with ESMTP id C9A8B37B491; Fri, 23 Feb 2001 06:54:09 -0800 (PST) (envelope-from cp@grendel.bsdi.com) Received: from grendel.bsdi.com (cp@localhost.bsdi.com [127.0.0.1]) by grendel.bsdi.com (8.11.1/8.9.3) with ESMTP id f1NErsx86721; Fri, 23 Feb 2001 07:53:54 -0700 (MST) (envelope-from cp@grendel.bsdi.com) Message-Id: <200102231453.f1NErsx86721@grendel.bsdi.com> To: Julian Elischer Cc: Kris Kennaway , smp@FreeBSD.ORG, julian@FreeBSD.ORG Subject: Re: NETGRAPH broken with witness In-reply-to: Your message of "Fri, 23 Feb 2001 02:47:17 PST." <3A963FB5.995FEA7D@elischer.org> From: Chuck Paterson Date: Fri, 23 Feb 2001 07:53:54 -0700 Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org } }hmm tricky.. I thought initiating the damned thing would insert it. } }I have no idea of the correct ordering. If this is a leaf lock just put it at the end as there is no ordering issues. If this isn't a leaf lock then you need to figure out where the lock fits in the heirarchy, not just to make the witness code happy. Chuck To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Fri Feb 23 11: 1:14 2001 Delivered-To: freebsd-smp@freebsd.org Received: from pike.osd.bsdi.com (pike.osd.bsdi.com [204.216.28.222]) by hub.freebsd.org (Postfix) with ESMTP id 70E4337B503; Fri, 23 Feb 2001 11:01:11 -0800 (PST) (envelope-from jhb@foo.osd.bsdi.com) Received: from foo.osd.bsdi.com (root@foo.osd.bsdi.com [204.216.28.137]) by pike.osd.bsdi.com (8.11.1/8.9.3) with ESMTP id f1NJ0sK28310; Fri, 23 Feb 2001 11:00:54 -0800 (PST) (envelope-from jhb@foo.osd.bsdi.com) Received: (from jhb@localhost) by foo.osd.bsdi.com (8.11.1/8.11.1) id f1NJ0Ej76425; Fri, 23 Feb 2001 11:00:14 -0800 (PST) (envelope-from jhb) Message-ID: X-Mailer: XFMail 1.4.0 on FreeBSD X-Priority: 3 (Normal) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <3A963FB5.995FEA7D@elischer.org> Date: Fri, 23 Feb 2001 11:00:14 -0800 (PST) Organization: BSD, Inc. From: John Baldwin To: Julian Elischer Subject: Re: NETGRAPH broken with witness Cc: julian@FreeBSD.ORG, smp@FreeBSD.ORG, Kris Kennaway Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On 23-Feb-01 Julian Elischer wrote: > Kris Kennaway wrote: >> >> On Fri, Feb 23, 2001 at 01:57:36AM -0800, Julian Elischer wrote: >> > Kris Kennaway wrote: >> > > >> > > Apparently this was never tested: early in the boot phase I get the >> > > following panic: >> > > >> > > panic: spin lock netgraph worklist mutex not in order list >> > > >> > > and indeed the witness code appears to have no knowledge of the >> > > netgraph spinlocks. >> > >> > so how is it supposed to know? >> >> From what I can gather, you add the locking hierarchy to be enforced >> to spin_order_list in kern.mutex.c > > hmm tricky.. I thought initiating the damned thing would insert it. For sleep mutexes it is automatic, yes. Spin mutexes are intended to only be used in very low level code and thus only in places where the order can be easily laid out. -- John Baldwin -- http://www.FreeBSD.org/~jhb/ PGP Key: http://www.Baldwin.cx/~john/pgpkey.asc "Power Users Use the Power to Serve!" - http://www.FreeBSD.org/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message