Re: Ok, are all the panics fixed now?
On Thu, Aug 28, 2003 at 11:34:19PM -0500, Mike Silbersack wrote: On Thu, 28 Aug 2003, Tor Egge wrote: I'm enclosed a patch that introduces PADDR2/PMAP2 which is to be used by pmap_pte() (should not be called from interrupts on noncurrent pmaps) while pmap_pte_quick() (should not be called without splvm() protection) will use PADDR1/PMAP1. The patch also backs out revision 1.250.2.20 of src/sys/i386/i386/pmap.c. - Tor Egge With this patch applied, my Celeron 450 with 12 megs of ram (artifically limited, of course) has been swapping away while running a buildworld for 3+ hours now without a problem. Before your patch, it would die in 15 minutes or so; I think you may have fixed the problem. Good work, as always! Same thing here: With this patch I haven't been able to provoke a panic, while without it I could reliably get one of my machines to panic. -- Insert your favourite quote here. Erik Trulsson [EMAIL PROTECTED] ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Ok, are all the panics fixed now?
At 11:42 PM 27/08/2003 +0400, Maxim Konovalov wrote: It's simple: we need to backout all these untested MFCs. I dont think people throw in untested MFCs into STABLE. They do their best effort and with that, there will still be some bugs. Its that simple. Well, we all have different definitions for 'testing'. To me, 'tesing' is something that takes anywhere from several weeks to several months to do and often involves hundreds of machines, configurations, and load mixes. Perhaps to some others, 'testing' means that the code compiles and the system boots up without panicing. For -current, perhaps the fast-and-loose definition is appropriate. For -stable that is nearing a 'x.9' release, I think my definition (very conservative) should be the standard. By that definition, the PAE changes should be promptly removed (and should have been when the first signs of trouble showed up). -DG David G. Lawrence Download Technologies, Inc. - http://www.downloadtech.com - (866) 399 8500 TeraSolutions, Inc. - http://www.terasolutions.com - (888) 346 7175 The FreeBSD Project - http://www.freebsd.org Pave the road of life with opportunities. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Ok, are all the panics fixed now?
So, I think we'll just include a warning with 4.9: WARNING! Do not attempt to stress a FreeBSD 4.9 machine if you: or Upgrade your FreeBSD to RedHat. s/RedHat/FreeBSD 4.8-RELEASE/ It's simple: we need to backout all these untested MFCs. Or fix the bugs. I don't know anything about the code in question, but now that people are getting repeatable panics, I assume that tracking down the bugs will be rather easier. There was a time when STABLE absolutely needed to be stable, but I'm not sure that's necessarily the case any more; now that we have all the release/security branches, I think it's safe to say that most systems which need absolute stability aren't going to be running STABLE. But the security branches don't get bug fixes, only security fixes. So at the the end we don't have a branch for stability anymore. I think that is a step in the wrong direction. I think by the time we get to x.4 or x.5 of a branch, it should be rock stable and only get bug fixes, with maybe device drivers added. Big changes should be avoided. John -- John Hay -- [EMAIL PROTECTED] / [EMAIL PROTECTED] ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Ok, are all the panics fixed now?
On Wed, 27 Aug 2003, Mike Silbersack wrote: Ah, that's a good idea! I'll try setting hw.physmem down to 12M or so and see if that causes the panics to occur over here. Mike Silby Silbersack Ok, I booted with hw.physmem=16M, and a buildworld paniced with free/cache page dirty. (I was taking a nap at the time the buildworld crashed, we'll talk about that soon.) (Are you guys running with INVARIANTS?) So, I think we'll just include a warning with 4.9: WARNING! Do not attempt to stress a FreeBSD 4.9 machine if you: a) Will be swapping whatsoever or b) Will be taking a nap while it is running. Doing either or both of these may result in panics. Mike Silby Silbersack P.S. - now that we have an easy way to reproduce this, perhaps we can get it fixed. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Ok, are all the panics fixed now?
RAM is cheap, so I don't need swapping. Naps are -not- negotiable. :) - Mike H. X-Original-To: [EMAIL PROTECTED] X-pair-Authenticated: 209.68.2.70 Date: Wed, 27 Aug 2003 13:34:15 -0500 (CDT) From: Mike Silbersack [EMAIL PROTECTED] Cc: [EMAIL PROTECTED] Sender: [EMAIL PROTECTED] X-Spam-Status: No, hits=-5.4 required=5.0 tests=AWL,BAYES_01,EMAIL_ATTRIBUTION,IN_REP_TO,QUOTED_EMAIL_TEXT, RCVD_IN_OSIRUSOFT_COM,REFERENCES,REPLY_WITH_QUOTES version=2.55 X-Spam-Level: X-Spam-Checker-Version: SpamAssassin 2.55 (1.174.2.19-2003-05-19-exp) On Wed, 27 Aug 2003, Mike Silbersack wrote: Ah, that's a good idea! I'll try setting hw.physmem down to 12M or so and see if that causes the panics to occur over here. Mike Silby Silbersack Ok, I booted with hw.physmem=16M, and a buildworld paniced with free/cache page dirty. (I was taking a nap at the time the buildworld crashed, we'll talk about that soon.) (Are you guys running with INVARIANTS?) So, I think we'll just include a warning with 4.9: WARNING! Do not attempt to stress a FreeBSD 4.9 machine if you: a) Will be swapping whatsoever or b) Will be taking a nap while it is running. Doing either or both of these may result in panics. Mike Silby Silbersack P.S. - now that we have an easy way to reproduce this, perhaps we can get it fixed. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to [EMAIL PROTECTED] ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Ok, are all the panics fixed now?
At 01:34 PM 27/08/2003 -0500, Mike Silbersack wrote: Ok, I booted with hw.physmem=16M, and a buildworld paniced with free/cache page dirty. (I was taking a nap at the time the buildworld crashed, we'll talk about that soon.) (Are you guys running with INVARIANTS?) Nope, I am not. Should I be ? ---Mike ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Ok, are all the panics fixed now?
On Wed, 27 Aug 2003, 13:34-0500, Mike Silbersack wrote: On Wed, 27 Aug 2003, Mike Silbersack wrote: Ah, that's a good idea! I'll try setting hw.physmem down to 12M or so and see if that causes the panics to occur over here. Mike Silby Silbersack Ok, I booted with hw.physmem=16M, and a buildworld paniced with free/cache page dirty. (I was taking a nap at the time the buildworld crashed, we'll talk about that soon.) (Are you guys running with INVARIANTS?) So, I think we'll just include a warning with 4.9: WARNING! Do not attempt to stress a FreeBSD 4.9 machine if you: or Upgrade your FreeBSD to RedHat. a) Will be swapping whatsoever or b) Will be taking a nap while it is running. Doing either or both of these may result in panics. It's simple: we need to backout all these untested MFCs. -- Maxim Konovalov, [EMAIL PROTECTED], [EMAIL PROTECTED] ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Ok, are all the panics fixed now?
At 23:42 27/08/2003 +0400, Maxim Konovalov wrote: On Wed, 27 Aug 2003, 13:34-0500, Mike Silbersack wrote: So, I think we'll just include a warning with 4.9: WARNING! Do not attempt to stress a FreeBSD 4.9 machine if you: or Upgrade your FreeBSD to RedHat. s/RedHat/FreeBSD 4.8-RELEASE/ It's simple: we need to backout all these untested MFCs. Or fix the bugs. I don't know anything about the code in question, but now that people are getting repeatable panics, I assume that tracking down the bugs will be rather easier. There was a time when STABLE absolutely needed to be stable, but I'm not sure that's necessarily the case any more; now that we have all the release/security branches, I think it's safe to say that most systems which need absolute stability aren't going to be running STABLE. Colin Percival ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Ok, are all the panics fixed now?
On Wed, 27 Aug 2003, 12:56-0700, Colin Percival wrote: At 23:42 27/08/2003 +0400, Maxim Konovalov wrote: On Wed, 27 Aug 2003, 13:34-0500, Mike Silbersack wrote: So, I think we'll just include a warning with 4.9: WARNING! Do not attempt to stress a FreeBSD 4.9 machine if you: or Upgrade your FreeBSD to RedHat. s/RedHat/FreeBSD 4.8-RELEASE/ No way: SA 03:08 - 03:11, http://www.freebsd.org/security/#adv It's simple: we need to backout all these untested MFCs. Or fix the bugs. I don't know anything about the code in question, but now that people are getting repeatable panics, I assume that tracking down the bugs will be rather easier. There was a time when STABLE absolutely needed to be stable, but I'm not sure that's necessarily the case any more; now that we have all the release/security branches, I think it's safe to say that most systems which need absolute stability aren't going to be running STABLE. We do have -CURRENT already. Look, believe you or not but there are people including me who trying to run -STABLE in a production environment. No sense in tracking RELENG_4_8 because it has some serious bugs, kern/53717 and kern/50803 f.e. No sense in 4.9-REL in such bad quality too. -- Maxim Konovalov, [EMAIL PROTECTED], [EMAIL PROTECTED] ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Ok, are all the panics fixed now?
At 11:42 PM 27/08/2003 +0400, Maxim Konovalov wrote: It's simple: we need to backout all these untested MFCs. I dont think people throw in untested MFCs into STABLE. They do their best effort and with that, there will still be some bugs. Its that simple. ---Mike ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to [EMAIL PROTECTED]
