From sacadmin Fri Jul 21 17:01:08 2006 Received: from zion.eng.sun.com (zion.SFBay.Sun.COM [129.146.17.75]) by sac.sfbay.sun.com (8.13.6+Sun/8.13.6) with ESMTP id k6M018L3017771 for ; Fri, 21 Jul 2006 17:01:08 -0700 (PDT) Received: from zion.eng.sun.com (localhost [127.0.0.1]) by zion.eng.sun.com (8.13.7+Sun/8.13.7) with ESMTP id k6M017bP012551; Fri, 21 Jul 2006 17:01:07 -0700 (PDT) Received: (from ahl@localhost) by zion.eng.sun.com (8.13.7+Sun/8.13.7/Submit) id k6M017tn012550; Fri, 21 Jul 2006 17:01:07 -0700 (PDT) Date: Fri, 21 Jul 2006 17:01:07 -0700 From: Adam Leventhal To: psarc@sac.sfbay.sun.com Cc: Adam Leventhal , Vitezslav.Batrla@sun.com Subject: PSARC 2006/447 LWP support for prun(1) and pstop(1) Message-ID: <20060722000107.GR3313@eng.sun.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.5.10i Status: RO Content-Length: 2728 I'm sponsoring the following fasttrack on behalf of Vita Batrla. This case introduces new ways of invoking the prun(1) and pstop(1) commands so that they apply only to particular LWPs. This uses the same methodology that was introduced in PSARC 2004/258 for utilities such as pstack(1). The stability it Evolving and the binding is Patch. The timer is set for 7/28/2006. Adam ---8<--- A. Introduction Our prun(1) and pstop(1) tools are used to stop and run processes. These tools currently operate on whole process. In some situations one may need to to stop and run only particular thread(s)/lwp(s) in the process. Similar enhancement was done in the past for truss(1), pstack(1) and pflags(1) tools, see PSARC 2004/258. B. Proposed Changes This case proposes per-thread modes for prun(1) and pstop(1) to allow users to stop selected threads using pstop(1) and set them running back using prun(1). The syntax for selecting threads is the same as in PSARC 2004/258: pstop 100123/2 # stop thread 2 in process 100123 pstop 100123/2,3-5 # stop threads 2, 3, 4, and 5 pstop 100123/-7 # stop threads whose IDs are 7 or below pstop 100123/8- # stop threads whose IDs are 8 or above pstop 100123/8- 321/3 # stop threads whose IDs are 8 or above in # process 100123 and stop thread 3 in # process 321 The thread specification is considered as a mask - prun(1)/pstop(1) will iterate over all threads in the process and stop/set running only threads whose ID will match the mask. For example: if the user does "prun 100123/2,3-5" and thread with ID 2 and 4 will not exist in the process, then no error will be printed, and threads 3 and 5 will be set running. If the user has provided thread specification, which did not match any thread, an error will be printed. This behaviour is consistent with pstack(1). C. Man page changes proc(1) --- /tmp/proc.1.orig Fri Jun 30 13:24:23 2006 +++ /tmp/proc.1 Fri Jun 30 13:26:56 2006 @@ -29,9 +29,9 @@ /usr/bin/pwdx pid... - /usr/bin/pstop pid... + /usr/bin/pstop pid[/lwps] ... - /usr/bin/prun pid... + /usr/bin/prun pid[/lwps] ... /usr/bin/pwait [-v] pid... YY. Stability _____________________________________________________________________ | Interface | Classification | Binding | |_____________________________________|___________________|_________| | prun(1) per-thread mode | Evolving | Minor | | pstop(1) per-thread mode | Evolving | Minor | |_____________________________________|___________________|_________| -- Adam Leventhal, Solaris Kernel Development http://blogs.sun.com/ahl From sacadmin Fri Jul 21 17:24:03 2006 Received: from jurassic.eng.sun.com (jurassic.SFBay.Sun.COM [129.146.224.130]) by sac.sfbay.sun.com (8.13.6+Sun/8.13.6) with ESMTP id k6M0O3FN018231 for ; Fri, 21 Jul 2006 17:24:03 -0700 (PDT) Received: from [129.146.11.163] (sr1-umpk-05.SFBay.Sun.COM [129.146.11.163]) by jurassic.eng.sun.com (8.13.7+Sun/8.13.6) with ESMTP id k6M0O2nX111051; Fri, 21 Jul 2006 17:24:02 -0700 (PDT) Message-ID: <44C17022.50707@Sun.COM> Date: Fri, 21 Jul 2006 17:24:02 -0700 From: Kais Belgaied User-Agent: Mozilla/5.0 (X11; U; SunOS sun4u; en-US; rv:1.7) Gecko/20050530 X-Accept-Language: ar-eg, en-us, en, ar, ar-dz, ar-bh, ar-iq, ar-jo, ar-kw, ar-lb, ar-ly, ar-ma, ar-om, ar-qa, ar-sa, ar-sy, ar-tn, ar-ae, ar-ye MIME-Version: 1.0 To: Adam Leventhal CC: psarc@sac.sfbay.sun.com, Vitezslav.Batrla@Sun.COM Subject: Re: PSARC 2006/447 LWP support for prun(1) and pstop(1) References: <20060722000107.GR3313@eng.sun.com> In-Reply-To: <20060722000107.GR3313@eng.sun.com> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Status: RO Content-Length: 729 Adam Leventhal wrote On 07/21/06 17:01,: > >The thread specification is considered as a mask - prun(1)/pstop(1) will iterate >over all threads in the process and stop/set running only threads whose ID will >match the mask. For example: if the user does "prun 100123/2,3-5" and thread >with ID 2 and 4 will not exist in the process, then no error will be printed, >and threads 3 and 5 will be set running. If the user has provided thread >specification, which did not match any thread, an error will be printed. This >behaviour is consistent with pstack(1). > > what will ps -l show for the state of a process with some of its threads being stopped and some being runnable? 'T' (for being "partially" traced?) Kais From sacadmin Tue Jul 25 03:06:31 2006 Received: from sfbaymail1sca.SFBay.Sun.COM (sfbaymail1sca.SFBay.Sun.COM [129.145.154.35]) by sac.sfbay.sun.com (8.13.6+Sun/8.13.6) with ESMTP id k6PA6Vqw015320 for ; Tue, 25 Jul 2006 03:06:31 -0700 (PDT) Received: from gmpea-pix-1.sun.com (gmpes-gis-mail-2.UK.Sun.COM [129.156.42.6]) by sfbaymail1sca.SFBay.Sun.COM (8.13.6+Sun/8.13.6/ENSMAIL,v2.2) with ESMTP id k6PA6Uok003666 for ; Tue, 25 Jul 2006 03:06:31 -0700 (PDT) Received: from d1-emea-10.sun.com ([192.18.2.120]) by gmpea-pix-1.sun.com (8.13.6+Sun/8.12.9) with ESMTP id k6PA6Osj015528 for ; Tue, 25 Jul 2006 11:06:25 +0100 (BST) Received: from conversion-daemon.d1-emea-10.sun.com by d1-emea-10.sun.com (Sun Java System Messaging Server 6.2-4.02 (built Sep 9 2005)) id <0J2Y00G01ENQAW00@d1-emea-10.sun.com> (original mail from Vitezslav.Batrla@Sun.COM) for psarc@sac.sfbay.sun.com; Tue, 25 Jul 2006 11:06:24 +0100 (BST) Received: from 129.157.18.72 ([129.157.18.72]) by d1-emea-10.sun.com (Sun Java System Messaging Server 6.2-4.02 (built Sep 9 2005)) with ESMTPSA id <0J2Y00456EQIV4ZS@d1-emea-10.sun.com>; Tue, 25 Jul 2006 11:06:24 +0100 (BST) Date: Tue, 25 Jul 2006 12:06:01 +0200 From: Vita Batrla Subject: Re: PSARC 2006/447 LWP support for prun(1) and pstop(1) In-reply-to: <44C17022.50707@Sun.COM> Sender: Vitezslav.Batrla@Sun.COM To: Kais Belgaied Cc: Adam Leventhal , psarc@sac.sfbay.sun.com Message-id: <1153821961.4500.14.camel@cml> MIME-version: 1.0 X-Mailer: Ximian Evolution 1.4.6.331 Content-type: text/plain Content-transfer-encoding: 7BIT References: <20060722000107.GR3313@eng.sun.com> <44C17022.50707@Sun.COM> Status: RO Content-Length: 1340 Hi Kais, The current behavior is: the output of *ps -l* doesn't show state 'T' until all threads are stopped. If some threads are stopped and some are running/sleeping, then 'R'/'S' is displayed, etc. The command *ps -lL* displays status per lwp and shows correctly 'T' for stopped threads. This behavior is not new. The ability to run/stop the individual LWP already exists in /proc filesystem, see man proc(4). You can run/stop the particular LWP in the process by writing to /proc//lwp//lwpctl file. Vita On Sat, 2006-07-22 at 02:24, Kais Belgaied wrote: > Adam Leventhal wrote On 07/21/06 17:01,: > > > > >The thread specification is considered as a mask - prun(1)/pstop(1) will iterate > >over all threads in the process and stop/set running only threads whose ID will > >match the mask. For example: if the user does "prun 100123/2,3-5" and thread > >with ID 2 and 4 will not exist in the process, then no error will be printed, > >and threads 3 and 5 will be set running. If the user has provided thread > >specification, which did not match any thread, an error will be printed. This > >behaviour is consistent with pstack(1). > > > > > > what will ps -l show for the state of a process with some of its threads > being stopped and some > being runnable? 'T' (for being "partially" traced?) > > > Kais From sacadmin Tue Jul 25 06:09:17 2006 Received: from localhost.east.sun.com (punchin-sommerfeld.East.Sun.COM [129.148.19.3]) by sac.sfbay.sun.com (8.13.6+Sun/8.13.6) with ESMTP id k6PD9Gr7019027 for ; Tue, 25 Jul 2006 06:09:17 -0700 (PDT) Received: from localhost.east.sun.com (localhost [127.0.0.1]) by localhost.east.sun.com (8.13.6+Sun/8.13.6) with ESMTP id k6PD95hK003696; Tue, 25 Jul 2006 13:09:06 GMT Received: (from sommerfeld@localhost) by localhost.east.sun.com (8.13.6+Sun/8.13.6/Submit) id k6PD950V003695; Tue, 25 Jul 2006 09:09:05 -0400 (EDT) X-Authentication-Warning: localhost.east.sun.com: sommerfeld set sender to sommerfeld@sun.com using -f Subject: Re: PSARC 2006/447 LWP support for prun(1) and pstop(1) From: Bill Sommerfeld To: Adam Leventhal Cc: psarc@sac.sfbay.sun.com, Vitezslav.Batrla@sun.com In-Reply-To: <20060722000107.GR3313@eng.sun.com> References: <20060722000107.GR3313@eng.sun.com> Content-Type: text/plain Content-Transfer-Encoding: 7bit Message-Id: <1153832945.2948.28.camel@localhost> Mime-Version: 1.0 X-Mailer: Ximian Evolution 1.4.6.340 Date: Tue, 25 Jul 2006 09:09:05 -0400 Status: RO Content-Length: 1332 On Fri, 2006-07-21 at 20:01, Adam Leventhal wrote: > I'm sponsoring the following fasttrack on behalf of Vita Batrla. This case > introduces new ways of invoking the prun(1) and pstop(1) commands so that > they apply only to particular LWPs. This uses the same methodology that was > introduced in PSARC 2004/258 for utilities such as pstack(1). The stability > it Evolving and the binding is Patch. The timer is set for 7/28/2006. Adding this functionality is great but it seems to me that it would be worthwhile to enhance the existing cautions in the proc(1) man page. For instance, in the "WARNINGS" section, you could change most instances of "process" to "process or thread"... nit: at the top you say: "The stability i[s] Evolving and the binding is Patch." while at the end you say: _____________________________________________________________________ | Interface | Classification | Binding | |_____________________________________|___________________|_________| | prun(1) per-thread mode | Evolving | Minor | | pstop(1) per-thread mode | Evolving | Minor | |_____________________________________|___________________|_________| This looks Patch-eligible to me. - Bill From sacadmin Tue Jul 25 09:34:10 2006 Received: from zion.eng.sun.com (zion.SFBay.Sun.COM [129.146.17.75]) by sac.sfbay.sun.com (8.13.6+Sun/8.13.6) with ESMTP id k6PGY9ET025293 for ; Tue, 25 Jul 2006 09:34:09 -0700 (PDT) Received: from zion.eng.sun.com (localhost [127.0.0.1]) by zion.eng.sun.com (8.13.7+Sun/8.13.7) with ESMTP id k6PGY91W024307; Tue, 25 Jul 2006 09:34:09 -0700 (PDT) Received: (from ahl@localhost) by zion.eng.sun.com (8.13.7+Sun/8.13.7/Submit) id k6PGY9Fj024306; Tue, 25 Jul 2006 09:34:09 -0700 (PDT) Date: Tue, 25 Jul 2006 09:34:09 -0700 From: Adam Leventhal To: Bill Sommerfeld Cc: psarc@sac.sfbay.sun.com, Vitezslav.Batrla@sun.com Subject: Re: PSARC 2006/447 LWP support for prun(1) and pstop(1) Message-ID: <20060725163409.GE18073@eng.sun.com> References: <20060722000107.GR3313@eng.sun.com> <1153832945.2948.28.camel@localhost> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1153832945.2948.28.camel@localhost> User-Agent: Mutt/1.5.10i Status: RO Content-Length: 840 On Tue, Jul 25, 2006 at 09:09:05AM -0400, Bill Sommerfeld wrote: > nit: > at the top you say: > > "The stability i[s] Evolving and the binding is Patch." > > while at the end you say: > > _____________________________________________________________________ > | Interface | Classification | Binding | > |_____________________________________|___________________|_________| > | prun(1) per-thread mode | Evolving | Minor | > | pstop(1) per-thread mode | Evolving | Minor | > |_____________________________________|___________________|_________| > > This looks Patch-eligible to me. Correct; the binding is Patch. Adam -- Adam Leventhal, Solaris Kernel Development http://blogs.sun.com/ahl From sacadmin Mon Jul 31 11:03:29 2006 Received: from zion.eng.sun.com (zion.SFBay.Sun.COM [129.146.17.75]) by sac.sfbay.sun.com (8.13.6+Sun/8.13.6) with ESMTP id k6VI3T0X013494 for ; Mon, 31 Jul 2006 11:03:29 -0700 (PDT) Received: from zion.eng.sun.com (localhost [127.0.0.1]) by zion.eng.sun.com (8.13.7+Sun/8.13.7) with ESMTP id k6VI3SiS018701; Mon, 31 Jul 2006 11:03:29 -0700 (PDT) Received: (from ahl@localhost) by zion.eng.sun.com (8.13.7+Sun/8.13.7/Submit) id k6VI3SNa018700; Mon, 31 Jul 2006 11:03:28 -0700 (PDT) Date: Mon, 31 Jul 2006 11:03:28 -0700 From: Adam Leventhal To: psarc@sac.sfbay.sun.com Cc: Vitezslav.Batrla@sun.com, Adam Leventhal Subject: Re: PSARC 2006/447 LWP support for prun(1) and pstop(1) Message-ID: <20060731180328.GG12272@eng.sun.com> References: <20060722000107.GR3313@eng.sun.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20060722000107.GR3313@eng.sun.com> User-Agent: Mutt/1.5.10i Status: RO Content-Length: 117 This case was approved on 7/26. Adam -- Adam Leventhal, Solaris Kernel Development http://blogs.sun.com/ahl