From owner-kdb@oss.sgi.com Sun Oct 1 17:13:54 2000 Received: by oss.sgi.com id ; Sun, 1 Oct 2000 17:13:44 -0700 Received: from pneumatic-tube.sgi.com ([204.94.214.22]:51773 "EHLO pneumatic-tube.sgi.com") by oss.sgi.com with ESMTP id ; Sun, 1 Oct 2000 17:13:16 -0700 Received: from larry.melbourne.sgi.com (larry.melbourne.sgi.com [134.14.52.130]) by pneumatic-tube.sgi.com (980327.SGI.8.8.8-aspam/980310.SGI-aspam) via SMTP id RAA04037 for ; Sun, 1 Oct 2000 17:19:34 -0700 (PDT) mail_from (kaos@melbourne.sgi.com) Received: from kao2.melbourne.sgi.com (kao2.melbourne.sgi.com [134.14.55.180]) by larry.melbourne.sgi.com (950413.SGI.8.6.12/950213.SGI.AUTOCF) via ESMTP id LAA06505 for ; Mon, 2 Oct 2000 11:11:17 +1100 X-Mailer: exmh version 2.1.1 10/15/1999 From: Keith Owens To: kdb@oss.sgi.com Subject: Change of kdb serial attention key Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Mon, 02 Oct 2000 11:11:16 +1100 Message-ID: <1664.970445476@kao2.melbourne.sgi.com> Sender: owner-kdb@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;kdb-outgoing I am thinking of changing the kdb serial attention key from control/A, mainly because ^A is used in emacs, including all utilities that use gnu readline in emacs mode. I have lost count of the number of times I hit ^A to go to start of line in bash and found myself in kdb instead. The keyboard key will still be pause. Any recommendations for a new kdb serial key? The problem is that most single keys are already used so it will probably be a multi key combination. I do not like break because of the problems that some emulators have in sending break. My current idea is ^Q three times within three seconds with no intervening characters. All (?) utilities discard ^Q so it will not cause a problem if kdb is not activated. From owner-kdb@oss.sgi.com Sun Oct 1 17:25:16 2000 Received: by oss.sgi.com id ; Sun, 1 Oct 2000 17:25:06 -0700 Received: from orzan.fi.udc.es ([193.144.60.19]:59559 "EHLO orzan.fi.udc.es") by oss.sgi.com with ESMTP id ; Sun, 1 Oct 2000 17:24:39 -0700 Received: from serpe.mitica (mail@vexeta.dc.fi.udc.es [193.144.51.32]) by orzan.fi.udc.es (8.9.3/8.9.1) with ESMTP id CAA15430; Mon, 2 Oct 2000 02:23:52 +0200 (MET DST) Received: from quintela by serpe.mitica with local (Exim 3.16 #1 (Debian)) id 13ftOa-0006kp-00; Mon, 02 Oct 2000 02:23:52 +0200 To: Keith Owens Cc: kdb@oss.sgi.com Subject: Re: Change of kdb serial attention key References: <1664.970445476@kao2.melbourne.sgi.com> X-Url: http://carpanta.dc.fi.udc.es/~quintela From: "Juan J. Quintela" In-Reply-To: Keith Owens's message of "Mon, 02 Oct 2000 11:11:16 +1100" Date: 02 Oct 2000 02:23:52 +0200 Message-ID: Lines: 20 User-Agent: Gnus/5.0807 (Gnus v5.8.7) Emacs/20.7 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-kdb@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;kdb-outgoing >>>>> "keith" == Keith Owens writes: keith> I am thinking of changing the kdb serial attention key from control/A, keith> mainly because ^A is used in emacs, including all utilities that use keith> gnu readline in emacs mode. I have lost count of the number of times I keith> hit ^A to go to start of line in bash and found myself in kdb instead. keith> The keyboard key will still be pause. Yes, I like that thing really, I use Ctrl-A all the time :))) For me triple ^Q is OK. Later, Juan. PD: Now I am searching for an terminal program that don't use C-a as the request key ... -- In theory, practice and theory are the same, but in practice they are different -- Larry McVoy From owner-kdb@oss.sgi.com Sun Oct 1 17:31:45 2000 Received: by oss.sgi.com id ; Sun, 1 Oct 2000 17:31:36 -0700 Received: from Cantor.suse.de ([194.112.123.193]:36875 "HELO Cantor.suse.de") by oss.sgi.com with SMTP id ; Sun, 1 Oct 2000 17:31:21 -0700 Received: from Hermes.suse.de (Hermes.suse.de [194.112.123.136]) by Cantor.suse.de (Postfix) with ESMTP id 7B9E11E293; Mon, 2 Oct 2000 02:30:37 +0200 (MEST) Received: from gruyere.muc.suse.de (gruyere.muc.suse.de [10.23.1.2]) by Hermes.suse.de (Postfix) with ESMTP id C637A3E456; Mon, 2 Oct 2000 02:30:36 +0200 (MEST) Received: by gruyere.muc.suse.de (Postfix, from userid 14446) id 039EB2F300; Mon, 2 Oct 2000 02:30:27 +0200 (MEST) Date: Mon, 2 Oct 2000 02:30:26 +0200 From: "Andi Kleen" To: Keith Owens Cc: kdb@oss.sgi.com Subject: Re: Change of kdb serial attention key Message-ID: <20001002023026.A31080@gruyere.muc.suse.de> References: <1664.970445476@kao2.melbourne.sgi.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 1.0i In-Reply-To: <1664.970445476@kao2.melbourne.sgi.com>; from kaos@melbourne.sgi.com on Mon, Oct 02, 2000 at 11:11:16AM +1100 Sender: owner-kdb@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;kdb-outgoing On Mon, Oct 02, 2000 at 11:11:16AM +1100, Keith Owens wrote: > I am thinking of changing the kdb serial attention key from control/A, > mainly because ^A is used in emacs, including all utilities that use > gnu readline in emacs mode. I have lost count of the number of times I > hit ^A to go to start of line in bash and found myself in kdb instead. > The keyboard key will still be pause. > > Any recommendations for a new kdb serial key? The problem is that most > single keys are already used so it will probably be a multi key > combination. I do not like break because of the problems that some > emulators have in sending break. My current idea is ^Q three times > within three seconds with no intervening characters. All (?) utilities > discard ^Q so it will not cause a problem if kdb is not activated. I press ^Q in a row regularly because I often have problems with screens that have software flowcontrol enabled (it is XON) How about three ^X ? -Andi From owner-kdb@oss.sgi.com Sun Oct 1 20:05:15 2000 Received: by oss.sgi.com id ; Sun, 1 Oct 2000 20:05:06 -0700 Received: from farmer8.nanobiz.com ([208.176.9.172]:17164 "EHLO nanobiz.com") by oss.sgi.com with ESMTP id ; Sun, 1 Oct 2000 20:04:40 -0700 Received: from pendragon.eng.nanobiz.com (IDENT:root@pendragon [192.168.1.155]) by nanobiz.com (8.9.3/8.9.3) with ESMTP id TAA29148; Sun, 1 Oct 2000 19:59:15 -0700 From: Scott Lurndal Received: by pendragon.eng.nanobiz.com; Sun, 1 Oct 2000 20:04:20 -0700 Message-Id: <200010020304.UAA01697@pendragon.eng.nanobiz.com> Subject: Re: Change of kdb serial attention key To: ak@suse.de (Andi Kleen) Date: Sun, 1 Oct 2000 20:04:20 -0700 (PDT) Cc: kaos@melbourne.sgi.com (Keith Owens), kdb@oss.sgi.com In-Reply-To: <20001002023026.A31080@gruyere.muc.suse.de> from "Andi Kleen" at Oct 02, 2000 02:30:26 AM X-Mailer: ELM [version 2.5 PL1] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-kdb@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;kdb-outgoing > > On Mon, Oct 02, 2000 at 11:11:16AM +1100, Keith Owens wrote: > > I am thinking of changing the kdb serial attention key from control/A, > > mainly because ^A is used in emacs, including all utilities that use > > gnu readline in emacs mode. I have lost count of the number of times I > > hit ^A to go to start of line in bash and found myself in kdb instead. > > The keyboard key will still be pause. > > > > Any recommendations for a new kdb serial key? The problem is that most > > single keys are already used so it will probably be a multi key > > combination. I do not like break because of the problems that some > > emulators have in sending break. My current idea is ^Q three times > > within three seconds with no intervening characters. All (?) utilities > > discard ^Q so it will not cause a problem if kdb is not activated. > > I press ^Q in a row regularly because I often have problems with screens > that have software flowcontrol enabled (it is XON) > > How about three ^X ? > > > -Andi > > > Perhaps ^K? It has mnemonic appeal. If it were as easy to modify utilities as it is to patch the kernel, I'd be tempted to add the kdb hot key to the 'stty' command so that an arbitrary key can be assigned (and pressing it thrice wouldn't be necessary). This would have the additional benefit that it could be usable from all serial ports. speed 9600 baud; rows 65; columns 105; line = 0; intr = ^C; quit = ^\; erase = ^H; kill = ^U; eof = ^D; eol = ; eol2 = ; start = ^Q; kdb = ^A stop = ^S; susp = ^Z; rprnt = ^R; werase = ^W; lnext = ^V; flush = ^O; min = 1; time = 0; -parenb -parodd cs8 -hupcl -cstopb cread -clocal -crtscts ignkdb -ignbrk -brkint -ignpar -parmrk -inpck -istrip -inlcr -igncr icrnl ixon -ixoff -iuclc -ixany -imaxbel opost -olcuc -ocrnl onlcr -onocr -onlret -ofill -ofdel nl0 cr0 tab3 bs0 vt0 ff0 isig icanon iexten echo echoe echok -ec scott From owner-kdb@oss.sgi.com Sun Oct 1 20:47:05 2000 Received: by oss.sgi.com id ; Sun, 1 Oct 2000 20:47:01 -0700 Received: from deliverator.sgi.com ([204.94.214.10]:61190 "EHLO deliverator.sgi.com") by oss.sgi.com with ESMTP id ; Sun, 1 Oct 2000 20:46:29 -0700 Received: from larry.melbourne.sgi.com (larry.melbourne.sgi.com [134.14.52.130]) by deliverator.sgi.com (980309.SGI.8.8.8-aspam-6.2/980310.SGI-aspam) via SMTP id UAA28573 for ; Sun, 1 Oct 2000 20:37:54 -0700 (PDT) mail_from (kaos@melbourne.sgi.com) Received: from kao2.melbourne.sgi.com (kao2.melbourne.sgi.com [134.14.55.180]) by larry.melbourne.sgi.com (950413.SGI.8.6.12/950213.SGI.AUTOCF) via ESMTP id OAA07523 for ; Mon, 2 Oct 2000 14:44:20 +1100 X-Mailer: exmh version 2.1.1 10/15/1999 From: Keith Owens To: kdb@oss.sgi.com Subject: Re: Change of kdb serial attention key In-reply-to: Your message of "Sun, 01 Oct 2000 20:04:20 PDT." <200010020304.UAA01697@pendragon.eng.nanobiz.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Mon, 02 Oct 2000 14:44:17 +1100 Message-ID: <3664.970458257@kao2.melbourne.sgi.com> Sender: owner-kdb@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;kdb-outgoing On Sun, 1 Oct 2000 20:04:20 -0700 (PDT), Scott Lurndal wrote: >> On Mon, Oct 02, 2000 at 11:11:16AM +1100, Keith Owens wrote: >> > I am thinking of changing the kdb serial attention key from control/A, > >Perhaps ^K? It has mnemonic appeal. Kill-line in emacs and gnu readline when in emacs mode. I want to avoid any key that is already used on its own. Andi's suggestion of ^X looks good, it is the emacs escape character but is always used in combination with another key so looking for 3 consecutive ^X will avoid emacs problems. From owner-kdb@oss.sgi.com Sun Oct 1 23:40:06 2000 Received: by oss.sgi.com id ; Sun, 1 Oct 2000 23:39:56 -0700 Received: from deliverator.sgi.com ([204.94.214.10]:41750 "EHLO deliverator.sgi.com") by oss.sgi.com with ESMTP id ; Sun, 1 Oct 2000 23:39:35 -0700 Received: from larry.melbourne.sgi.com (larry.melbourne.sgi.com [134.14.52.130]) by deliverator.sgi.com (980309.SGI.8.8.8-aspam-6.2/980310.SGI-aspam) via SMTP id XAA07661; Sun, 1 Oct 2000 23:31:07 -0700 (PDT) mail_from (kaos@melbourne.sgi.com) Received: from kao2.melbourne.sgi.com (kao2.melbourne.sgi.com [134.14.55.180]) by larry.melbourne.sgi.com (950413.SGI.8.6.12/950213.SGI.AUTOCF) via ESMTP id RAA08310; Mon, 2 Oct 2000 17:36:14 +1100 X-Mailer: exmh version 2.1.1 10/15/1999 From: Keith Owens To: kdb@oss.sgi.com, linux-xfs@oss.sgi.com, linux-kernel@vger.kernel.org, Keir Fraser Subject: [Announce] kdb v1.5 is available Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Mon, 02 Oct 2000 17:36:11 +1100 Message-ID: <6414.970468571@kao2.melbourne.sgi.com> Sender: owner-kdb@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;kdb-outgoing http://oss.sgi.com/projects/kdb/download/ix86/ contains a patch for kdb v1.5 against 2.4.0-test9-pre8. Changes from kdb v1.5-beta2. * Upgrade to 2.4.0-test9-pre8. * Fix premature NMI oops on machines with large numbers of cpus. This version of kdb also includes :- * NMI oopser for uniprocessors. This builds on Keir Fraser's excellent patch for activating the local APIC on P6 systems and (ab)uses the local APIC to generate NMI on systems which have a local APIC but no IO-APIC, typically any P6 and above uniprocessor. So you poor developers with uniprocessors can now get diagnostics when the system gets into a spin loop. On the Processor configuration menu, select 'APIC and IO-APIC support on uniprocessors'. That option will detect the local APIC if it exists. See the help for 'NMI watchdog active for uniprocessors' and Documentation/nmi_watchdog.txt. Expect the configuration method and the use of /proc/sys/kernel/nmi_watchdog to change in future. Ingo Molnar and I are discussing how much of the NMI oopser for uniprocessors should go into the standard kernel. From owner-kdb@oss.sgi.com Mon Oct 2 05:45:50 2000 Received: by oss.sgi.com id ; Mon, 2 Oct 2000 05:45:45 -0700 Received: from orzan.fi.udc.es ([193.144.60.19]:50102 "EHLO orzan.fi.udc.es") by oss.sgi.com with ESMTP id ; Mon, 2 Oct 2000 05:45:16 -0700 Received: from serpe.mitica (mail@vexeta.dc.fi.udc.es [193.144.51.32]) by orzan.fi.udc.es (8.9.3/8.9.1) with ESMTP id OAA23948; Mon, 2 Oct 2000 14:44:14 +0200 (MET DST) Received: from quintela by serpe.mitica with local (Exim 3.16 #1 (Debian)) id 13g4x3-0007o2-00; Mon, 02 Oct 2000 14:44:13 +0200 To: Keith Owens Cc: kdb@oss.sgi.com Subject: Re: Change of kdb serial attention key References: <3664.970458257@kao2.melbourne.sgi.com> X-Url: http://carpanta.dc.fi.udc.es/~quintela From: "Juan J. Quintela" In-Reply-To: Keith Owens's message of "Mon, 02 Oct 2000 14:44:17 +1100" Date: 02 Oct 2000 14:44:13 +0200 Message-ID: Lines: 25 User-Agent: Gnus/5.0807 (Gnus v5.8.7) Emacs/20.7 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-kdb@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;kdb-outgoing >>>>> "keith" == Keith Owens writes: keith> On Sun, 1 Oct 2000 20:04:20 -0700 (PDT), keith> Scott Lurndal wrote: >>> On Mon, Oct 02, 2000 at 11:11:16AM +1100, Keith Owens wrote: >>> > I am thinking of changing the kdb serial attention key from control/A, >> >> Perhaps ^K? It has mnemonic appeal. keith> Kill-line in emacs and gnu readline when in emacs mode. I want to keith> avoid any key that is already used on its own. Andi's suggestion of ^X keith> looks good, it is the emacs escape character but is always used in keith> combination with another key so looking for 3 consecutive ^X will avoid keith> emacs problems. I agree with Andy suggestion of ^X, I tried to use ^Q here and some times it messed up with the Control Flow commands. ^K has the problem that as everybody knows, it is the kill-line comand in emacs (yes I see that there are more emacs users out here :)))) Later, Juan. -- In theory, practice and theory are the same, but in practice they are different -- Larry McVoy From owner-kdb@oss.sgi.com Mon Oct 2 09:26:31 2000 Received: by oss.sgi.com id ; Mon, 2 Oct 2000 09:26:21 -0700 Received: from thalia.fm.intel.com ([132.233.247.11]:11793 "EHLO thalia.fm.intel.com") by oss.sgi.com with ESMTP id ; Mon, 2 Oct 2000 09:26:03 -0700 Received: from SMTP (fmsmsxvs04-1.fm.intel.com [132.233.42.204]) by thalia.fm.intel.com (8.9.1a+p1/8.9.1/d: relay.m4,v 1.31 2000/08/22 00:15:13 dmccart Exp $) with SMTP id QAA15968; Mon, 2 Oct 2000 16:26:23 GMT Received: from fmsmsx26.fm.intel.com ([132.233.48.26]) by 132.233.48.204 (Norton AntiVirus for Internet Email Gateways 1.0) ; Mon, 02 Oct 2000 16:25:09 0000 (GMT) Received: by fmsmsx26.fm.intel.com with Internet Mail Service (5.5.2650.21) id ; Mon, 2 Oct 2000 09:24:50 -0700 Message-ID: <4148FEAAD879D311AC5700A0C969E8905DE388@orsmsx35.jf.intel.com> From: "Grover, Andrew" To: "'Keith Owens'" Cc: "'kdb@oss.sgi.com'" Subject: RE: kdb_read not exported Date: Mon, 2 Oct 2000 09:24:33 -0700 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2650.21) Content-Type: text/plain; charset="iso-8859-1" Sender: owner-kdb@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;kdb-outgoing Hi Keith, I was happy to see the 1.5 release announcement this morning. Wow, that was quick. ;-) Based on that, is it possible to say whether a module's use of kdb_read is okay now? Regards -- Andy -----Original Message----- From: Keith Owens [mailto:kaos@melbourne.sgi.com] Sent: Wednesday, September 20, 2000 11:05 PM To: Grover, Andrew Cc: 'kdb@oss.sgi.com' Subject: Re: kdb_read not exported On Wed, 20 Sep 2000 22:32:01 -0700, "Grover, Andrew" wrote: >I'm writing a module for kdb. > >I would find kdb_read's functionality useful in my module. If I export and >use it, will anything break? kdb v1.5 (out soon) is a major redesign of the cpu state handling, to get finer grained control over single step and recovery from oopsing or hung kdb commands. Part of that redesign affects the way kdb_read() works, especially the longjmp() to break out of a command. Ask me again after v1.5 is out. From owner-kdb@oss.sgi.com Mon Oct 2 10:39:52 2000 Received: by oss.sgi.com id ; Mon, 2 Oct 2000 10:39:41 -0700 Received: from thor.fsc-usa.com ([63.109.18.10]:59405 "EHLO thor.fsc-usa.com") by oss.sgi.com with ESMTP id ; Mon, 2 Oct 2000 10:39:24 -0700 Received: from one.fsc-usa.com (one-internal.eng.pyramid.com [172.25.216.10]) by thor.fsc-usa.com (8.10.1/18SEP00--Fujitsu-Siemens-of-America-Gateway) id e92HbY229594 From: To: <>; Mon, 2 Oct 2000 10:37:34 -0700 Received: from tomb.fsc-usa.com by one.fsc-usa.com (8.8.5/FSC_USA_Internal_Configuration) id KAA06833; Mon, 2 Oct 2000 10:37:07 -0700 (PDT) Received: by tomb.fsc-usa.com (8.8.5/Pyramid_Internal_Configuration) id KAA17083; Mon, 2 Oct 2000 10:37:12 -0700 (PDT) From: shailesh@fsc-usa.com (Shailesh K Basani) Message-Id: <200010021737.KAA17083@tomb.fsc-usa.com> Subject: Re: Change of kdb serial attention key To: kaos@melbourne.sgi.com (Keith Owens) Date: Mon, 2 Oct 2000 10:37:12 -0700 (PDT) Cc: kdb@oss.sgi.com In-Reply-To: <1664.970445476@kao2.melbourne.sgi.com> from "Keith Owens" at Oct 02, 2000 11:11:16 AM Organization: Fujitsu Siemens Computers X-Mailer: ELM [version 2.5 PL2] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-kdb@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;kdb-outgoing Hi Keith, Another place where ^A is used quite often is with the "screen" utility. This utility was mostly used with dumb terminals - and is still a popular utility. (Used to create multiple pseudo terminals). To change to a different window or for most of the commands, ^A is used. 3 times ^X seems good to me. If it is also being used by some other utility, can you pick something like combination of different keys? (Like Solaris uses "~#", our kernel debugger uses CTRL-SHIFT-2 etc..). Could be hard to remember initially, but people will get used to it. regards, shailesh [Keith Owens wrote...] : : I am thinking of changing the kdb serial attention key from control/A, : mainly because ^A is used in emacs, including all utilities that use : gnu readline in emacs mode. I have lost count of the number of times I : hit ^A to go to start of line in bash and found myself in kdb instead. : The keyboard key will still be pause. : : Any recommendations for a new kdb serial key? The problem is that most : single keys are already used so it will probably be a multi key : combination. I do not like break because of the problems that some : emulators have in sending break. My current idea is ^Q three times : within three seconds with no intervening characters. All (?) utilities : discard ^Q so it will not cause a problem if kdb is not activated. : -- From owner-kdb@oss.sgi.com Mon Oct 2 18:22:25 2000 Received: by oss.sgi.com id ; Mon, 2 Oct 2000 18:22:14 -0700 Received: from orzan.fi.udc.es ([193.144.60.19]:25811 "EHLO orzan.fi.udc.es") by oss.sgi.com with ESMTP id ; Mon, 2 Oct 2000 18:21:56 -0700 Received: from serpe.mitica (mail@vexeta.dc.fi.udc.es [193.144.51.32]) by orzan.fi.udc.es (8.9.3/8.9.1) with ESMTP id DAA04177; Tue, 3 Oct 2000 03:21:09 +0200 (MET DST) Received: from quintela by serpe.mitica with local (Exim 3.16 #1 (Debian)) id 13gGlZ-0001Sj-00; Tue, 03 Oct 2000 03:21:09 +0200 To: kdb@oss.sgi.com, Keith Owens Subject: KDB hang X-Url: http://carpanta.dc.fi.udc.es/~quintela From: "Juan J. Quintela" Date: 03 Oct 2000 03:21:09 +0200 Message-ID: Lines: 116 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-kdb@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;kdb-outgoing Hi using kbd-1.5 (final) with test9-pre9 I have got Enter to end, to continue: NMI Watchdog detected LOCKUP on CPU1, regist: CPU: 1 That happened while playing with the debugger in an SMP system. Could be that you forgot to deactivate the NMI watchdog at kdb entry???? Just a guess. THe message after the Oops is: kdb: Debugger re-entered on cpu 1, new reason = 10 Forced recursion is set Attempting recursive mode Notice that if I went to kdb and return to normal mode in less that one minute, I don't get that Oops. Later, Juan. The Ooops message (I have to use ksymoops to do that) is: ksymoops 2.3.4 on i686 2.4.0-test9. Options used -V (default) -K (specified) -L (specified) -O (specified) -m System.map (specified) CPU: 1 EIP: 0010:[] Using defaults from ksymoops -t elf32-i386 -a i386 EFLAGS: 00000002 eax: c01ce260 ebx: c027141c ecx: 000003f8 edx: 000003fd esi: c12f9c9c edi: 0000004e ebp: c12f9c9c esp: c12f9c20 ds: 0018 es: 0018 ss: 0018 Process swapper (pid: 0, stackpage=c12f9000) Stack: c027141c c12f9c9c c01ce392 00000050 c12f9c9c c01f71c5 c12f9f7c c017035f c12f9c9c 00000050 c017038a c12f9c9c 00000050 c01f7200 00000001 c12f9d38 c7e46000 c016f8e9 c12f9c9c 00000050 c01f7200 00000800 c02cfd80 c12f8000 Call Trace: [] [] [] [] [] [< [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] Code: a8 01 74 3a 89 ca ec 88 c3 80 fb 7f 75 02 b3 08 0f b6 f3 80 >>EIP; c01ce2d2 <===== Trace; c01ce392 Trace; c01f71c5 <__nenv+ca1/33bc> Trace; c017035f Trace; c017038a Trace; c01f7200 <__nenv+cdc/33bc> Trace; c01af3aa Trace; c016dd9d Trace; c01f6ee6 <__nenv+9c2/33bc> Trace; c01ce260 Trace; c01ce260 Trace; c017035f Trace; c017038a Trace; c01f6519 <__nkdb_err+81/8c> Trace; c016e0d3 Trace; c01f6519 <__nkdb_err+81/8c> Trace; c01a09ba Trace; c01b83f6 Trace; c01b8200 Trace; c010de70 Trace; c010ac51 Trace; c01cf178 Trace; c0211b9b Trace; c0211b9b Trace; c0211b9b Trace; c01cf5c0 Trace; c0177a06 Trace; c010ac51 Trace; c010ae36 Trace; c0107170 Trace; c0107170 Trace; c0100018 Trace; c010719d Trace; c0107202 Trace; c0119e1c Trace; c010ae75 Code; c01ce2d2 00000000 <_EIP>: Code; c01ce2d2 <===== 0: a8 01 test $0x1,%al <===== Code; c01ce2d4 2: 74 3a je 3e <_EIP+0x3e> c01ce310 Code; c01ce2d6 4: 89 ca mov %ecx,%edx Code; c01ce2d8 6: ec in (%dx),%al Code; c01ce2d9 7: 88 c3 mov %al,%bl Code; c01ce2db 9: 80 fb 7f cmp $0x7f,%bl Code; c01ce2de c: 75 02 jne 10 <_EIP+0x10> c01ce2e2 Code; c01ce2e0 e: b3 08 mov $0x8,%bl Code; c01ce2e2 10: 0f b6 f3 movzbl %bl,%esi Code; c01ce2e5 13: 80 00 00 addb $0x0,(%eax) -- In theory, practice and theory are the same, but in practice they are different -- Larry McVoy From owner-kdb@oss.sgi.com Mon Oct 2 18:56:55 2000 Received: by oss.sgi.com id ; Mon, 2 Oct 2000 18:56:35 -0700 Received: from deliverator.sgi.com ([204.94.214.10]:14599 "EHLO deliverator.sgi.com") by oss.sgi.com with ESMTP id ; Mon, 2 Oct 2000 18:56:11 -0700 Received: from larry.melbourne.sgi.com (larry.melbourne.sgi.com [134.14.52.130]) by deliverator.sgi.com (980309.SGI.8.8.8-aspam-6.2/980310.SGI-aspam) via SMTP id SAA05913 for ; Mon, 2 Oct 2000 18:47:45 -0700 (PDT) mail_from (kaos@melbourne.sgi.com) Received: from kao2.melbourne.sgi.com (kao2.melbourne.sgi.com [134.14.55.180]) by larry.melbourne.sgi.com (950413.SGI.8.6.12/950213.SGI.AUTOCF) via ESMTP id MAA14842; Tue, 3 Oct 2000 12:52:53 +1100 X-Mailer: exmh version 2.1.1 10/15/1999 From: Keith Owens To: "Juan J. Quintela" cc: kdb@oss.sgi.com Subject: Re: KDB hang In-reply-to: Your message of "03 Oct 2000 03:21:09 +0200." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Tue, 03 Oct 2000 12:52:53 +1100 Message-ID: <2558.970537973@kao2.melbourne.sgi.com> Sender: owner-kdb@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;kdb-outgoing On 03 Oct 2000 03:21:09 +0200, "Juan J. Quintela" wrote: > using kbd-1.5 (final) with test9-pre9 I have got > >Enter to end, to continue: >NMI Watchdog detected LOCKUP on CPU1, regist: CPU: 1 > >That happened while playing with the debugger in an SMP system. Could >be that you forgot to deactivate the NMI watchdog at kdb entry???? The NMI watchdog is not deactivated. Instead there is a kdb state flag which says to ignore the NMI watchdog tick. This state bit is twiddled in several places, it is turned on in main kdb, it is turned off during a kdb command (to catch loops in commands), it should be turned on again when waiting for input. Looks like the state is incorrect. Which command did you issue that generated "Enter to end, to continue:"? From owner-kdb@oss.sgi.com Tue Oct 3 04:20:10 2000 Received: by oss.sgi.com id ; Tue, 3 Oct 2000 04:20:00 -0700 Received: from orzan.fi.udc.es ([193.144.60.19]:49380 "EHLO orzan.fi.udc.es") by oss.sgi.com with ESMTP id ; Tue, 3 Oct 2000 04:19:43 -0700 Received: from serpe.mitica (mail@vexeta.dc.fi.udc.es [193.144.51.32]) by orzan.fi.udc.es (8.9.3/8.9.1) with ESMTP id NAA09981; Tue, 3 Oct 2000 13:09:11 +0200 (MET DST) Received: from quintela by serpe.mitica with local (Exim 3.16 #1 (Debian)) id 13gPwc-0001Xz-00; Tue, 03 Oct 2000 13:09:10 +0200 To: Keith Owens Cc: kdb@oss.sgi.com Subject: Re: KDB hang References: <2558.970537973@kao2.melbourne.sgi.com> X-Url: http://carpanta.dc.fi.udc.es/~quintela From: "Juan J. Quintela" In-Reply-To: Keith Owens's message of "Tue, 03 Oct 2000 12:52:53 +1100" Date: 03 Oct 2000 13:09:10 +0200 Message-ID: Lines: 30 User-Agent: Gnus/5.0807 (Gnus v5.8.7) Emacs/20.7 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-kdb@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;kdb-outgoing >>>>> "keith" == Keith Owens writes: keith> On 03 Oct 2000 03:21:09 +0200, keith> "Juan J. Quintela" wrote: >> using kbd-1.5 (final) with test9-pre9 I have got >> >> Enter to end, to continue: >> NMI Watchdog detected LOCKUP on CPU1, regist: CPU: 1 >> >> That happened while playing with the debugger in an SMP system. Could >> be that you forgot to deactivate the NMI watchdog at kdb entry???? keith> The NMI watchdog is not deactivated. Instead there is a kdb state flag keith> which says to ignore the NMI watchdog tick. This state bit is twiddled keith> in several places, it is turned on in main kdb, it is turned off during keith> a kdb command (to catch loops in commands), it should be turned on keith> again when waiting for input. Looks like the state is incorrect. keith> Which command did you issue that generated "Enter to end, to keith> continue:"? bta, and yes I was trying to make sense for a backtrace (i.e. not continuing pressing keys). This has worked for ages (i.e. kdb 1.[34]). Later, Juan. -- In theory, practice and theory are the same, but in practice they are different -- Larry McVoy From owner-kdb@oss.sgi.com Tue Oct 3 06:51:50 2000 Received: by oss.sgi.com id ; Tue, 3 Oct 2000 06:51:41 -0700 Received: from ppp0.ocs.com.au ([203.34.97.3]:23570 "HELO mail.ocs.com.au") by oss.sgi.com with SMTP id ; Tue, 3 Oct 2000 06:51:13 -0700 Received: (qmail 10470 invoked from network); 3 Oct 2000 13:50:13 -0000 Received: from ocs3.ocs-net (192.168.255.3) by mail.ocs.com.au with SMTP; 3 Oct 2000 13:50:13 -0000 X-Mailer: exmh version 2.1.1 10/15/1999 From: Keith Owens To: kdb@oss.sgi.com cc: Mike Galbraith , "Grover, Andrew" Subject: Patch kdb v1.5 for NMI oops in kdb_read Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Wed, 04 Oct 2000 00:50:13 +1100 Message-ID: <4489.970581013@ocs3.ocs-net> Sender: owner-kdb@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;kdb-outgoing This fixes the NMI oops in the bta command. It also exports kdb_read for module use (untested). Index: 0-test9-pre9.2/kdb/kdb_io.c --- 0-test9-pre9.2/kdb/kdb_io.c Mon, 02 Oct 2000 20:59:05 +1100 kaos (linux-2.4/Q/c/32_kdb_io.c 1.6 644) +++ 0-test9-pre9.2(w)/kdb/kdb_io.c Wed, 04 Oct 2000 00:28:11 +1100 kaos (linux-2.4/Q/c/32_kdb_io.c 1.6 644) @@ -33,6 +33,8 @@ * KDB v1.2 */ +#include +#include #include #include #include @@ -93,7 +95,14 @@ static struct console *kdbcons; char * kdb_read(char *buffer, size_t bufsize) { - return kdba_read(buffer, bufsize); + int no_watchdog; + char *ret; + no_watchdog = KDB_STATE(NO_WATCHDOG); + KDB_STATE_SET(NO_WATCHDOG); + ret = kdba_read(buffer, bufsize); + if (!no_watchdog) + KDB_STATE_CLEAR(NO_WATCHDOG); + return(ret); } /* @@ -299,3 +308,5 @@ kdb_io_init(void) } return; } + +EXPORT_SYMBOL(kdb_read); Index: 0-test9-pre9.2/kdb/Makefile --- 0-test9-pre9.2/kdb/Makefile Sun, 24 Sep 2000 22:16:07 +1100 kaos (linux-2.4/Q/c/36_Makefile 1.5 644) +++ 0-test9-pre9.2(w)/kdb/Makefile Tue, 03 Oct 2000 23:06:38 +1100 kaos (linux-2.4/Q/c/36_Makefile 1.5 644) @@ -1,6 +1,6 @@ O_TARGET := kdb.o -O_OBJS = kdb_bt.o kdb_bp.o kdb_id.o kdb_io.o kdbsupport.o gen-kdb_cmds.o -OX_OBJS = kdbmain.o +O_OBJS = kdb_bt.o kdb_bp.o kdb_id.o kdbsupport.o gen-kdb_cmds.o +OX_OBJS = kdbmain.o kdb_io.o MOD_SUB_DIRS += modules From owner-kdb@oss.sgi.com Tue Oct 3 09:58:30 2000 Received: by oss.sgi.com id ; Tue, 3 Oct 2000 09:58:20 -0700 Received: from orzan.fi.udc.es ([193.144.60.19]:10224 "EHLO orzan.fi.udc.es") by oss.sgi.com with ESMTP id ; Tue, 3 Oct 2000 09:57:46 -0700 Received: from serpe.mitica (mail@vexeta.dc.fi.udc.es [193.144.51.32]) by orzan.fi.udc.es (8.9.3/8.9.1) with ESMTP id SAA14245; Tue, 3 Oct 2000 18:28:28 +0200 (MET DST) Received: from quintela by serpe.mitica with local (Exim 3.16 #1 (Debian)) id 13gUvc-0001dn-00; Tue, 03 Oct 2000 18:28:28 +0200 To: Keith Owens Cc: kdb@oss.sgi.com, Mike Galbraith , "Grover, Andrew" Subject: Re: Patch kdb v1.5 for NMI oops in kdb_read References: <4489.970581013@ocs3.ocs-net> X-Url: http://carpanta.dc.fi.udc.es/~quintela From: "Juan J. Quintela" In-Reply-To: Keith Owens's message of "Wed, 04 Oct 2000 00:50:13 +1100" Date: 03 Oct 2000 18:28:28 +0200 Message-ID: Lines: 17 User-Agent: Gnus/5.0807 (Gnus v5.8.7) Emacs/20.7 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-kdb@oss.sgi.com Precedence: bulk Return-Path: X-Orcpt: rfc822;kdb-outgoing >>>>> "keith" == Keith Owens writes: keith> This fixes the NMI oops in the bta command. It also exports kdb_read keith> for module use (untested). The NMI oops fix is working nicely here. Thanks for your fast fix. Later, Juan. -- In theory, practice and theory are the same, but in practice they are different -- Larry McVoy