From shartley@king.mcs.drexel.edu Tue Jan 10 12:56:51 1995 Received: from mcs.drexel.edu (king.mcs.drexel.edu) by optima.CS.Arizona.EDU (5.65c/15) via SMTP id AA13681; Tue, 10 Jan 1995 10:56:05 MST Received: by mcs.drexel.edu (4.1/SMI-4.0) id AA27244; Tue, 10 Jan 95 12:56:51 EST Date: Tue, 10 Jan 95 12:56:51 EST From: shartley@king.mcs.drexel.edu (Stephen J. Hartley) Message-Id: <9501101756.AA27244@mcs.drexel.edu> To: info-sr@cs.arizona.edu Subject: client/server (RPC) programming It is possible to do client/server programming using SR, as shown in the attached example. The process in the server resource waits in an `in' statement for incoming requests from clients for service. Using the `forward' statement, the server creates a new process to handle the request. While the request is being handled, the server can field new requests. By adding code to create virtual machines on different physical machines, the clients can be on remote machines and send their requests to the server process on the server machine. My question is this. Is is possible to write the client code and the server code as separate programs? Can the server process be started on the server machine from a stand-alone program and then later can the clients be created and send their requests to the server? Or must the whole thing be started in the context of one SR program? Thank you for your help. Steve Hartley shartley@mcs.drexel.edu =============================== rpc.sr =================================== resource server optype serve = (client_id, parameters: int) op service1, service2, service3 : serve body server() op do_service1, do_service2, do_service3 : serve process serving do true -> in service1(client_id, parameters) -> forward do_service1(client_id, parameters) [] service2(client_id, parameters) -> forward do_service2(client_id, parameters) [] service3(client_id, parameters) -> forward do_service3(client_id, parameters) ni od end serving proc do_service1(client_id, parameters) nap(parameters) end do_service1 proc do_service2(client_id, parameters) nap(parameters) end do_service2 proc do_service3(client_id, parameters) nap(parameters) end do_service3 end server resource client import server body client(client_id : int; server_cap : cap server) process go do true -> nap(10000) call server_cap.service1(client_id, age()) nap(10000) call server_cap.service2(client_id, age()) nap(10000) call server_cap.service3(client_id, age()) od end go end client resource client_server() import server, client var server_cap : cap server var num_clients : int := 5 getarg(1, num_clients) server_cap := create server() fa i := 1 to num_clients -> create client(i, server_cap) af end client_server ========================================================================== From owner-info-sr Tue Jan 10 13:12:43 1995 Received: from baskerville.CS.Arizona.EDU by optima.CS.Arizona.EDU (5.65c/15) via SMTP id AA22987; Tue, 10 Jan 1995 13:12:44 MST Received: (from gmt@localhost) by baskerville.CS.Arizona.EDU (8.6.9/8.6.6) id NAA10861 for info-sr; Tue, 10 Jan 1995 13:12:43 -0700 Date: Tue, 10 Jan 1995 13:12:43 -0700 From: Gregg Townsend Message-Id: <199501102012.NAA10861@baskerville.CS.Arizona.EDU> To: info-sr@baskerville Subject: Re: client/server (RPC) programming Content-Length: 688 From: shartley@king.mcs.drexel.edu (Stephen J. Hartley) ... My question is this. Is is possible to write the client code and the server code as separate programs? Can the server process be started on the server machine from a stand-alone program and then later can the clients be created and send their requests to the server? Or must the whole thing be started in the context of one SR program? It all has to be one program; SR has no facilities for estabilishing communications with another program. Gregg Townsend / Computer Science Dept / Univ of Arizona / Tucson, AZ 85721 +1 602 621 4325 gmt@cs.arizona.edu 110 57 16 W / 32 13 45 N / +758m From marcos@inf.puc-rio.br Thu Jan 12 12:10:55 1995 Received: from exu.inf.puc-rio.br by optima.cs.arizona.edu (5.65c/15) via SMTP id AA11536; Thu, 12 Jan 1995 07:12:12 MST Received: from omolu (omolu.inf.puc-rio.br) by exu.inf.puc-rio.br (4.1/SMI-4.1) id AA17847; Thu, 12 Jan 95 12:10:55 EDT Date: Thu, 12 Jan 95 12:10:55 EDT From: marcos@inf.puc-rio.br (Marcos Rodrigues de Lyra) Message-Id: <9501121410.AA17847@exu.inf.puc-rio.br> To: info-sr@cs.arizona.edu Subject: SrWin Cc: marcos@inf.puc-rio.br I have tryed to use the SrWin functions but I had some problems. The srl can't find the directory or files ,i don't know, of X11. I want to know what is necessary for srl to create the executable file ? , what srl needs of X11 directory. Please, if anyone has answers contact me. marcos@inf.puc-rio.br From owner-info-sr Thu Jan 12 13:10:38 1995 Received: from baskerville.CS.Arizona.EDU by optima.cs.arizona.edu (5.65c/15) via SMTP id AA06303; Thu, 12 Jan 1995 13:10:39 MST Received: (from gmt@localhost) by baskerville.CS.Arizona.EDU (8.6.9/8.6.6) id NAA22200; Thu, 12 Jan 1995 13:10:38 -0700 Date: Thu, 12 Jan 1995 13:10:38 -0700 From: Gregg Townsend Message-Id: <199501122010.NAA22200@baskerville.CS.Arizona.EDU> To: info-sr@baskerville Cc: marcos@inf.puc-rio.br Subject: Re: SrWin Content-Length: 933 Date: Thu, 12 Jan 95 12:10:55 EDT From: marcos@inf.puc-rio.br (Marcos Rodrigues de Lyra) I have tryed to use the SrWin functions but I had some problems. The srl can't find the directory or files ,i don't know, of X11. I want to know what is necessary for srl to create the executable file ? , what srl needs of X11 directory. Please, if anyone has answers contact me. The usual way to build an SRWin program "foo.sr" is to use the command sr foo.sr srwin.o -lX11 (thats a lowercase L, an X, a number 11) If that doesn't work, please send the exact error message to sr-project@cs.arizona.edu along with information specifying your computer model, operating system, and SR version, and we'll see if we can figure out what's gone wrong. Gregg Townsend / Computer Science Dept / Univ of Arizona / Tucson, AZ 85721 +1 602 621 4325 gmt@cs.arizona.edu 110 57 16 W / 32 13 45 N / +758m From canedo@taec.enet.dec.com Fri Jan 13 03:35:24 1995 Received: from inet-gw-1.pa.dec.com by optima.cs.arizona.edu (5.65c/15) via SMTP id AA02611; Fri, 13 Jan 1995 03:35:24 MST Received: from vbormc.vbo.dec.com by inet-gw-1.pa.dec.com (5.65/10Aug94) id AA20104; Fri, 13 Jan 95 02:30:46 -0800 Received: by vbormc.vbo.dec.com; id AA13080; Fri, 13 Jan 95 11:07:29 +0100 From: canedo@taec.enet.dec.com Message-Id: <9501131007.AA13080@vbormc.vbo.dec.com> Received: from taec.enet; by vbormc.enet; Fri, 13 Jan 95 11:07:29 MET Date: Fri, 13 Jan 95 11:07:29 MET To: info-sr@cs.arizona.edu Apparently-To: info-sr@cs.arizona.edu Subject: Tracing a variable in sr ? Hello every one, I wish to implement a tracing mecanism to log variable modification. More precisely, I have a variable "toto" and I wish to call a command every time this variable is changed. Is it possible to do that ? Thanks in advance Best regards Joseph Canedo From owner-info-sr Fri Jan 13 09:38:21 1995 Received: from baskerville.CS.Arizona.EDU by optima.cs.arizona.edu (5.65c/15) via SMTP id AA18387; Fri, 13 Jan 1995 09:38:22 MST Received: (from gmt@localhost) by baskerville.CS.Arizona.EDU (8.6.9/8.6.6) id JAA09001 for info-sr; Fri, 13 Jan 1995 09:38:21 -0700 Date: Fri, 13 Jan 1995 09:38:21 -0700 From: Gregg Townsend Message-Id: <199501131638.JAA09001@baskerville.CS.Arizona.EDU> To: info-sr@baskerville Subject: Re: Tracing a variable in sr ? Content-Length: 482 From: canedo@taec.enet.dec.com I wish to implement a tracing mecanism to log variable modification. More precisely, I have a variable "toto" and I wish to call a command every time this variable is changed. Is it possible to do that ? Sorry, SR doesn't have anything like that available. Gregg Townsend / Computer Science Dept / Univ of Arizona / Tucson, AZ 85721 +1 602 621 4325 gmt@cs.arizona.edu 110 57 16 W / 32 13 45 N / +758m From john_townsend@taligent.com Fri Jan 13 18:46:25 1995 Received: from mailserv.taligent.com by optima.cs.arizona.edu (5.65c/15) via SMTP id AA11233; Fri, 13 Jan 1995 19:49:28 MST Received: from qm.taligent.com by mailserv.taligent.com (AIX 3.2/UCB 5.64/4.03) id AA46548; Fri, 13 Jan 1995 18:49:31 -0800 Message-Id: Date: 13 Jan 1995 18:46:25 -0800 From: "John Townsend" Subject: Question regarding syntax To: "SR Mailing List" X-Mailer: Mail*Link SMTP-QM 3.0.1 I am new to sr and I am still adjusting to the syntax. (This doesn't mean that I don't like it. In fact, there are things that I like about it alot). I did have a question. Consider the following code fragment: resource main() type TName = string[40]; type TStreet = string(40); # Note the difference in syntax end I have sr 2.3 running on an RS/6000 and this program will compile and run but yields a warning regarding the usage of older string syntax. Will code that uses the newer syntax (the brackets) work on older versions of sr? I am concerned because I am using 2.3 and the computers at school are using an older version. Do I have to worry about source compatibility between these two versions? Am I worrying for nothing? :-) -- John Townsend, John_Townsend@taligent.com From greg Sat Jan 14 10:07:07 1995 Received: from paloverde.CS.Arizona.EDU by optima.cs.arizona.edu (5.65c/15) via SMTP id AA15332; Sat, 14 Jan 1995 10:05:54 MST Date: Sat, 14 Jan 1995 10:07:07 MST From: "Greg Andrews" Message-Id: <199501141707.AA19742@paloverde.cs.arizona.edu> Received: by paloverde.cs.arizona.edu; Sat, 14 Jan 1995 10:07:07 MST To: info-sr@cs.arizona.edu, john_townsend@taligent.com Subject: Re: Question regarding syntax Date: 13 Jan 1995 18:46:25 -0800 From: "John Townsend" Subject: Question regarding syntax I am new to sr and I am still adjusting to the syntax. ... resource main() type TName = string[40]; type TStreet = string(40); # Note the difference in syntax end I have sr 2.3 running on an RS/6000 and this program will compile and run but yields a warning regarding the usage of older string syntax. Will code that uses the newer syntax (the brackets) work on older versions of sr? I am concerned because I am using 2.3 and the computers at school are using an older version. Do I have to worry about source compatibility between these two versions? Am I worrying for nothing? :-) The new syntax for string size has been around for ages, so it is unlikely that your school is using such an old version of SR that it won't work. I'd suggest just using the new syntax, and then trying your program on the older system. It will complain if it expects parentheses instead of brackets. Greg Andrews From canedo@taec.enet.dec.com Wed Jan 18 03:21:36 1995 Received: from inet-gw-2.pa.dec.com by optima.cs.arizona.edu (5.65c/15) via SMTP id AA26938; Wed, 18 Jan 1995 03:21:36 MST Received: from vbormc.vbo.dec.com by inet-gw-2.pa.dec.com (5.65/10Aug94) id AA07529; Wed, 18 Jan 95 02:19:31 -0800 Received: by vbormc.vbo.dec.com; id AA01135; Wed, 18 Jan 95 10:55:42 +0100 From: canedo@taec.enet.dec.com Message-Id: <9501180955.AA01135@vbormc.vbo.dec.com> Received: from taec.enet; by vbormc.enet; Wed, 18 Jan 95 10:55:43 MET Date: Wed, 18 Jan 95 10:55:43 MET To: info-sr@cs.arizona.edu Apparently-To: info-sr@cs.arizona.edu Subject: Trying to interface sr and tcl/tk Hello every one, For one month or so, I'm trying to interface sr and tcl/tk. I think I found a fairly good solution but I don't know exactly if it is possible to implement it... I wish to have some opinions on that. The first step is to invoke tcl scripts in sr code. I do that use C external interface of sr and calling Tcl_Eval and Tcl_evalFile. The only problem is the script must not block (If I remember well). This works well (communication sr -> tcl). Tcl_LinkVar allows link variables between sr and tcl (tested also with ints, doubles...) The second step is to call sr code in tcl script (to make callbacks for example). This can be done using C interface of sr and implementing new tcl commands in C. For example I could make a rendez-vous or setting a semaphore calling sr C interface corresponding procedure in a tcl command (I don't know if it is possible, I 've just begun to read runtime.ps file in sr distribution). For example I could define a tcl command V or P which acts on a semaphore identified by its name (shared between tcl and sr)... Thank you for your attention and your remarks. Best regards Joseph Canedo From greg Wed Jan 18 08:48:22 1995 Received: from paloverde.CS.Arizona.EDU by optima.cs.arizona.edu (5.65c/15) via SMTP id AA10521; Wed, 18 Jan 1995 08:47:03 MST Date: Wed, 18 Jan 1995 08:48:22 MST From: "Greg Andrews" Message-Id: <199501181548.AA25784@paloverde.cs.arizona.edu> Received: by paloverde.cs.arizona.edu; Wed, 18 Jan 1995 08:48:22 MST To: canedo@taec.enet.dec.com, info-sr@cs.arizona.edu Subject: Re: Trying to interface sr and tcl/tk It's fairly easy to call *out* of SR to C, in ways such as you describe. Calling in is much trickier, as it requires a detailed understanding of the SR run-time. You have to worry not only about the names and parameters of the appropriate SR runtime routines, but also have to be very careful about blocking/context switching. Greg Andrews From nlam@eecs.wsu.edu Fri Jan 27 10:51:47 1995 Received: from dns1.eecs.wsu.edu by optima.cs.arizona.edu (5.65c/15) via SMTP id AA27986; Fri, 27 Jan 1995 11:51:51 MST Received: from dryden.eecs.wsu.edu by dns1.eecs.wsu.edu (8.6.8.1/8.940801) id KAA23422; Fri, 27 Jan 1995 10:51:48 -0800 From: Norman Lam - EECS Received: by dryden.eecs.wsu.edu (1.37.109.4) id AA11613; Fri, 27 Jan 95 10:51:47 -0800 Message-Id: <9501271851.AA11613@dryden.eecs.wsu.edu> Subject: age() function To: info-sr@cs.arizona.edu Date: Fri, 27 Jan 1995 10:51:47 -0800 (PST) X-Mailer: ELM [version 2.4 PL23] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 249 Hi, I've used the age() function to time the execution of the matrix multiplication program in chapter 1 of the book "The SR Programming Language". I found that the time would change from time to time. Is there any explanation to that? Norman Lam From gmt@baskerville Fri Jan 27 15:31:12 1995 Received: from baskerville.CS.Arizona.EDU by optima.cs.arizona.edu (5.65c/15) via SMTP id AA14145; Fri, 27 Jan 1995 15:31:13 MST Received: (from gmt@localhost) by baskerville.CS.Arizona.EDU (8.6.9/8.6.6) id PAA29918; Fri, 27 Jan 1995 15:31:12 -0700 Date: Fri, 27 Jan 1995 15:31:12 -0700 From: Gregg Townsend Message-Id: <199501272231.PAA29918@baskerville.CS.Arizona.EDU> To: info-sr@cs.arizona.edu, nlam@eecs.wsu.edu Subject: Re: age() function Content-Length: 525 Many things can cause variations in age() measurements. Perhaps the most obvious is other processes running on the machine -- if they are using much of the CPU, there will be less available to your SR program, so it will take longer to finish. There can be other variables such as the cache hit rate. Even an idle system usually shows some amount of variation. Gregg Townsend / Computer Science Dept / Univ of Arizona / Tucson, AZ 85721 +1 602 621 4325 gmt@cs.arizona.edu 110 57 16 W / 32 13 45 N / +758m From @exeter.ac.uk:steve@dcs.exeter.ac.uk Mon Feb 6 12:25:45 1995 Received: from sun2.nsfnet-relay.ac.uk by optima.cs.arizona.edu (5.65c/15) via SMTP id AA09398; Mon, 6 Feb 1995 07:37:12 MST Via: uk.ac.exeter; Mon, 6 Feb 1995 12:30:12 +0000 Received: from atlas.exeter.ac.uk by exub.exeter.ac.uk with SMTP (PP) id <27627-0@exub.exeter.ac.uk>; Mon, 6 Feb 1995 12:25:56 +0000 Received: from sirius.dcs.exeter.ac.uk by atlas.dcs.exeter.ac.uk; Mon, 6 Feb 95 12:25:46 GMT From: steve@atlas.ex.ac.uk Date: Mon, 6 Feb 95 12:25:45 GMT Message-Id: <3116.9502061225@sirius.dcs.exeter.ac.uk> To: greg@cs.arizona.edu Cc: info-sr@cs.arizona.edu Subject: Operation capabilities We have noticed what appear to be inconsistencies in the mechanism for changing an operation capability using a resource capability variable (as described on p94 of the SR book). I guess this is all to do with the way calls of procs are optimized, but my question is whether the language is meant to behave in this way. resource Test op a() # operation capability to be changed op b() # calls a() from within resource op c(x: cap Test) # calls a() using resource capability x op d() # calls a() using myresource() capability body Test() var y: cap Test := myresource() proc a() write("old version of a") end proc b() writes("b: "); a() end proc c(x) writes("c: "); x.a() end proc d() writes("d: "); y.a() end end resource Main import Test op newa() body Main() proc newa() write("new version of a") end var t: cap Test := create Test() t.a(); t.b(); t.c(t); t.d() # all use old version t.a := newa # assign to operation capability t.a() # uses new version of a t.b() # uses old version of a t.c(t) # uses new version of a t.d() # uses old version of a end 64 % sr test.sr 65 % a.out old version of a b: old version of a c: old version of a d: old version of a new version of a b: old version of a c: new version of a d: old version of a 66 % We are currently using SR to explore some issues in computational reflection [1], where we need to use this or some similar technique, so it would be helpful for us to have a clear understanding of the semantics of this mechanism. [1] T. Watanabe and A. Yonezawa: "Reflection in an Object-Oriented Concurrent Language", in A. Yonezawa "ABCL: An Object-Oriented Concurrent System" (MIT Press, 1990). Regards, Steve ------------------------------------------------------------------------------- Parallel Systems Research Laboratory | Stephen Turner Email: steve@dcs.exeter.ac.uk | Department of Computer Science WWW: http://www.dcs.exeter.ac.uk | University of Exeter Fax: +44 392 264067 | Prince of Wales Road Tel: +44 392 264048 | Exeter EX4 4PT England ------------------------------------------------------------------------------- From pmcgonig@CS.Trinity.Edu Mon Feb 6 13:19:43 1995 Message-Id: <199502061916.AA24162@optima.cs.arizona.edu> Received: from sol.CS.Trinity.Edu ([131.194.131.158]) by optima.cs.arizona.edu (5.65c/15) via SMTP id AA24162; Mon, 6 Feb 1995 12:16:07 MST Received: by sol.CS.Trinity.Edu (1.37.109.11/16.2) id AA150068388; Mon, 6 Feb 1995 13:19:48 -0600 From: patrick mcgonigle Subject: SR system call To: info-sr@cs.arizona.edu Date: Mon, 6 Feb 1995 13:19:43 -0600 (CST) X-Mailer: ELM [version 2.4 PL23] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 286 Hi! How can I call the UNIX command "date" from an SR program? Basically, I just want to be able to print the date to stdout by simply calling a procedure, as in: PrintDate() It seems simple enough, but I'm not having any luck. Any help would be much appreciated. - Pat McGonigle From gmt@baskerville Mon Feb 6 14:35:14 1995 Received: from baskerville.CS.Arizona.EDU by optima.cs.arizona.edu (5.65c/15) via SMTP id AA01812; Mon, 6 Feb 1995 14:35:21 MST Received: (from gmt@localhost) by baskerville.CS.Arizona.EDU (8.6.9/8.6.6) id OAA18348; Mon, 6 Feb 1995 14:35:14 -0700 Date: Mon, 6 Feb 1995 14:35:14 -0700 From: Gregg Townsend Message-Id: <199502062135.OAA18348@baskerville.CS.Arizona.EDU> To: info-sr@cs.arizona.edu, pmcgonig@CS.Trinity.Edu Subject: Re: SR system call Content-Length: 455 From: patrick mcgonigle How can I call the UNIX command "date" from an SR program? Basically, I just want to be able to print the date to stdout by simply calling a procedure.... Try this: external system (cmd: string[*]) system("date") Gregg Townsend / Computer Science Dept / Univ of Arizona / Tucson, AZ 85721 +1 602 621 4325 gmt@cs.arizona.edu 110 57 16 W / 32 13 45 N / +758m From owner-info-sr Mon Feb 6 15:59:12 1995 Received: from baskerville.CS.Arizona.EDU by optima.cs.arizona.edu (5.65c/15) via SMTP id AA07268; Mon, 6 Feb 1995 15:59:13 MST Received: (from gmt@localhost) by baskerville.CS.Arizona.EDU (8.6.9/8.6.6) id PAA22533; Mon, 6 Feb 1995 15:59:12 -0700 Date: Mon, 6 Feb 1995 15:59:12 -0700 From: Gregg Townsend Message-Id: <199502062259.PAA22533@baskerville.CS.Arizona.EDU> To: steve@atlas.ex.ac.uk Cc: info-sr@baskerville Subject: Re: Operation capabilities Content-Length: 913 From: steve@atlas.ex.ac.uk We have noticed what appear to be inconsistencies in the mechanism for changing an operation capability using a resource capability variable (as described on p94 of the SR book). I guess this is all to do with the way calls of procs are optimized, but my question is whether the language is meant to behave in this way.... I believe it's behaving correctly. Assignment to an operation capability contained within a resource capability changes only that copy of the resource capability; it does not alter anything in the underlying resource. You can think of a resource capability as a record containing opcaps as fields; in fact, that's how it's implemented in the underlying C code. Gregg Townsend / Computer Science Dept / Univ of Arizona / Tucson, AZ 85721 +1 602 621 4325 gmt@cs.arizona.edu 110 57 16 W / 32 13 45 N / +758m From marcos@inf.puc-rio.br Thu Feb 9 15:40:16 1995 Received: from exu.inf.puc-rio.br by optima.cs.arizona.edu (5.65c/15) via SMTP id AA04391; Thu, 9 Feb 1995 10:40:36 MST Received: from oxala.puc-rio.br (oxala.inf.puc-rio.br) by exu.inf.puc-rio.br (4.1/SMI-4.1) id AA02696; Thu, 9 Feb 95 15:40:16 EDT Date: Thu, 9 Feb 95 15:40:16 EDT From: marcos@inf.puc-rio.br (Marcos Rodrigues de Lyra) Message-Id: <9502091740.AA02696@exu.inf.puc-rio.br> To: info-sr@cs.arizona.edu Subject: locate I need more information about the command "locate". I've tried to use it but it didn't work. Below is the way that i used it: var NumMachine : int . . . write ("My machine is:",NumMachine) locate (NumMaq, "oxala") write ("My machine is:",NumMachine) . . . - Before and after the "locate" i get -808464433 as my machine number. Is it anything wrong with my enviroment? Thanks. marcos@inf.puc-rio.br From gmt Fri Feb 10 08:52:22 1995 Received: from hawk.CS.Arizona.EDU by optima.cs.arizona.edu (5.65c/15) via SMTP id AA13559; Fri, 10 Feb 1995 08:52:28 MST Received: by hawk.CS.Arizona.EDU; (5.65/1.1.8.2/08Nov94-0446PM) id AA04580; Fri, 10 Feb 1995 08:52:22 -0700 Date: Fri, 10 Feb 1995 08:52:22 -0700 From: Gregg Townsend Message-Id: <9502101552.AA04580@hawk.CS.Arizona.EDU> To: info-sr, marcos@inf.puc-rio.br Subject: Re: locate From: marcos@inf.puc-rio.br (Marcos Rodrigues de Lyra) var NumMachine : int ... write ("My machine is:",NumMachine) locate (NumMaq, "oxala") write ("My machine is:",NumMachine) - Before and after the "locate" i get -808464433 as my machine number. Is it anything wrong with my enviroment? It appears that NumMachine is never initialized, and that it has nothing to do with the locate() call. Note that the purpose of locate() is to define the meaning of an integer value x when used in "create vm() on x". It has no effect on the current machine. Gregg Townsend / Computer Science Dept / Univ of Arizona / Tucson, AZ 85721 +1 602 621 4325 gmt@cs.arizona.edu 110 57 16 W / 32 13 45 N / +758m From marcos@inf.puc-rio.br Fri Feb 10 16:53:29 1995 Received: from exu.inf.puc-rio.br by optima.cs.arizona.edu (5.65c/15) via SMTP id AA24180; Fri, 10 Feb 1995 11:53:58 MST Received: from brahma (brahma.inf.puc-rio.br) by exu.inf.puc-rio.br (4.1/SMI-4.1) id AA11099; Fri, 10 Feb 95 16:53:29 EDT Date: Fri, 10 Feb 95 16:53:29 EDT From: marcos@inf.puc-rio.br (Marcos Rodrigues de Lyra) Message-Id: <9502101853.AA11099@exu.inf.puc-rio.br> To: info-sr@cs.arizona.edu Subject: srwin I am trying to use srwin in a distribute program. My program has many resources, each of them manages one window. The problem happens in execution time. The message error is : "ld.so: libX11.so.4: not found ". I just need to know on what directory SRX looks after this file. Thanks! marcos@inf.puc-rio.br From azhao@cc.gatech.edu Fri Feb 10 14:56:38 1995 Received: from burdell.cc.gatech.edu by optima.cs.arizona.edu (5.65c/15) via SMTP id AA28392; Fri, 10 Feb 1995 12:56:43 MST Received: from oakmont.cc.gatech.edu (azhao@oakmont.cc.gatech.edu [130.207.119.218]) by burdell.cc.gatech.edu (8.6.9/8.6.9) with ESMTP id OAA08459; Fri, 10 Feb 1995 14:56:41 -0500 Received: (from azhao@localhost) by oakmont.cc.gatech.edu (8.6.9/8.6.9) id OAA01394; Fri, 10 Feb 1995 14:56:40 -0500 From: azhao@cc.gatech.edu (Q. Alex Zhao) Message-Id: <199502101956.OAA01394@oakmont.cc.gatech.edu> Subject: Re: srwin To: marcos@inf.puc-rio.br (Marcos Rodrigues de Lyra) Date: Fri, 10 Feb 1995 14:56:38 -0500 (EST) Cc: info-sr@cs.arizona.edu In-Reply-To: <9502101853.AA11099@exu.inf.puc-rio.br> from "Marcos Rodrigues de Lyra" at Feb 10, 95 04:53:29 pm X-Mailer: ELM [version 2.4 PL23] Content-Type: text Content-Length: 855 According to Marcos Rodrigues de Lyra in a previous message: ] ] I am trying to use srwin in a distribute program. My program has many resources, each of them manages one window. The problem happens in execution time. The message error is : "ld.so: libX11.so.4: not found ". I just need to know on what directory SRX looks after this file. ] ] Thanks! ] ] marcos@inf.puc-rio.br It was the UNIX runtime library linker which could not find the library. I'm guessing that if you setenv LD_LIBRARY_PATH in your .profile or .cshrc etc. to the X lib directory "ld.so" should find it... Cheers, = Q. Alex Zhao ~{0"UT~} ! Graphics, Visualization & Usability Center Email: azhao@cc.gatech.edu ! College of Computing FAX: 404-853-9378 ! Georgia Institute of Technology Lab: 404-894-9636 ! Atlanta, Georgia 30332-0280 From gmt Fri Feb 10 13:04:44 1995 Received: from hawk.CS.Arizona.EDU by optima.cs.arizona.edu (5.65c/15) via SMTP id AA28967; Fri, 10 Feb 1995 13:04:47 MST Received: by hawk.CS.Arizona.EDU; (5.65/1.1.8.2/08Nov94-0446PM) id AA05101; Fri, 10 Feb 1995 13:04:44 -0700 Date: Fri, 10 Feb 1995 13:04:44 -0700 From: Gregg Townsend Message-Id: <9502102004.AA05101@hawk.CS.Arizona.EDU> To: marcos@inf.puc-rio.br Subject: Re: srwin Cc: info-sr From: marcos@inf.puc-rio.br (Marcos Rodrigues de Lyra) I am trying to use srwin in a distribute program. My program has many resources, each of them manages one window. The problem happens in execution time. The message error is : "ld.so: libX11.so.4: not found ". I just need to know on what directory SRX looks after this file. That's a system library that's part of the X window system. If you're running on a Sun Solaris system, you might try setting LD_RUN_PATH=/usr/openwin/lib in your environment at the time you build the program. Gregg Townsend / Computer Science Dept / Univ of Arizona / Tucson, AZ 85721 +1 602 621 4325 gmt@cs.arizona.edu 110 57 16 W / 32 13 45 N / +758m From john_townsend@taligent.com Thu Feb 16 17:47:31 1995 Received: from mailserv.taligent.com by optima.cs.arizona.edu (5.65c/15) via SMTP id AA19131; Thu, 16 Feb 1995 18:50:59 MST Received: from qm.taligent.com by mailserv.taligent.com (AIX 3.2/UCB 5.64/4.03) id AA213702; Thu, 16 Feb 1995 17:51:40 -0800 Message-Id: Date: 16 Feb 1995 17:47:31 -0800 From: "John Townsend" Subject: Problems To: "info-sr" X-Mailer: Mail*Link SMTP-QM 3.0.2 I am trying to run the distributed file system example from the book on my RS/6000. I am running into problems: prompt> a.out jaguar rshd: 0826-813 permission is denied. [vm 1] RTS abort: attedmpting to create resource on null machine. Now.. Jaguar is my own machine. This it the only machine I have access to! It looks like rshd is owned by root and I can't run it. Can someone tell me what I need to have done to my system to allow me to run this example? -- John From john_townsend@taligent.com Thu Feb 16 23:28:06 1995 Received: from mailserv.taligent.com by optima.cs.arizona.edu (5.65c/15) via SMTP id AA08897; Fri, 17 Feb 1995 00:32:46 MST Received: from qm.taligent.com by mailserv.taligent.com (AIX 3.2/UCB 5.64/4.03) id AA105804; Thu, 16 Feb 1995 23:33:28 -0800 Message-Id: Date: 16 Feb 1995 23:28:06 -0800 From: "John Townsend" Subject: More problems with dfs To: "info-sr Mailing List" X-Mailer: Mail*Link SMTP-QM 3.0.2 Well.. I made progress. I have the program running.. sort of. I found the first problem. It wasn't working because I didn't have an .rhosts file and when I finally figured that out, I discovered that rsh won't recognize your rhosts file unless the permissions are set to 600. Once I did that, I was able to get things to work. Unfortunately.. I am still confused. I run the program and enter some terminals. (pts/3 and pts/4) This seems to work okay except that their are shells attached to these terminals and I get contention for the terminal (one character goes to the shell, two character to the terminal, etc.) Am I doing this wrong? I would like to figure this stuff out because I would like to use the server design in this program as the basis for a project. Any help would be greatly appreciated. -- John From greg Fri Feb 17 08:41:15 1995 Received: from paloverde.CS.Arizona.EDU by optima.cs.arizona.edu (5.65c/15) via SMTP id AA26831; Fri, 17 Feb 1995 08:41:17 MST Date: Fri, 17 Feb 1995 08:41:15 MST From: "Greg Andrews" Message-Id: <199502171541.AA26877@paloverde.cs.arizona.edu> Received: by paloverde.cs.arizona.edu; Fri, 17 Feb 1995 08:41:15 MST To: info-sr@cs.arizona.edu, john_townsend@taligent.com Subject: Re: More problems with dfs Date: 16 Feb 1995 23:28:06 -0800 From: "John Townsend" Subject: More problems with dfs ... Unfortunately.. I am still confused. I run the program and enter some terminals. (pts/3 and pts/4) This seems to work okay except that their are shells attached to these terminals and I get contention for the terminal (one character goes to the shell, two character to the terminal, etc.) Read the discussion on page 263 of the SR book, just below the program for the Login resource. It describes how to solve your problem. Greg Andrews From john_townsend@taligent.com Sat Feb 18 00:52:31 1995 Received: from mailserv.taligent.com by optima.cs.arizona.edu (5.65c/15) via SMTP id AA26174; Sat, 18 Feb 1995 01:56:32 MST Received: from qm.taligent.com by mailserv.taligent.com (AIX 3.2/UCB 5.64/4.03) id AA64859; Sat, 18 Feb 1995 00:55:29 -0800 Message-Id: Date: 18 Feb 1995 00:52:31 -0800 Priority: Urgent From: "John Townsend" Subject: AIX 4.1? To: "info-sr Mailing List" X-Mailer: Mail*Link SMTP-QM 3.0.2 AIX 4.1? Aaaaggghhhhhh I did something that appears to be very bad. Today my RS/6000 was upgraded from AIX 3.2.5 to AIX 4.1. Unfortunately.. my sr setup is now badly hosted. I can't link the assembly cstest code in the /csw directory. I get three link errors on the stack underflow, overflow, etc. externs. Does anyone know anything about this? Has anyone ported SR to AIX 4.1? Am I completely hosed? I hope not.. I really need to get some coding done in SR! Please help me if you can. -- John From admin@ibmserv.edu.tw Mon Feb 20 14:21:43 1995 Received: from ibmserv.edu.tw by optima.cs.arizona.edu (5.65c/15) via SMTP id AA11997; Tue, 21 Feb 1995 22:59:14 MST Received: by ibmserv.edu.tw (AIX 3.2/UCB 5.64/4.03) id AA14745; Mon, 20 Feb 1995 14:21:43 -1600 From: admin@ibmserv.edu.tw Message-Id: <9502210621.AA14745@ibmserv.edu.tw> Subject: AIX 4.1? (fwd) To: info-sr@cs.arizona.edu Date: Mon, 20 Feb 1995 14:21:43 -1600 () X-Mailer: ELM [version 2.4 PL23alpha2] Content-Type: text Content-Length: 1945 Forwarded message: > From root Thu Feb 16 17:21:28 1995 > Message-Id: > Date: 18 Feb 1995 00:52:31 -0800 > Priority: Urgent > From: "John Townsend" > Subject: AIX 4.1? > To: "info-sr Mailing List" > X-Mailer: Mail*Link SMTP-QM 3.0.2 > Errors-To: info-sr-errors@cs.arizona.edu > > AIX 4.1? > Aaaaggghhhhhh > > I did something that appears to be very bad. Today my RS/6000 was upgraded > from AIX 3.2.5 to AIX 4.1. > > Unfortunately.. my sr setup is now badly hosted. I can't link the assembly > cstest code in the /csw directory. I get three link errors on the stack > underflow, overflow, etc. externs. > > Does anyone know anything about this? Has anyone ported SR to AIX 4.1? Am I > completely hosed? I hope not.. I really need to get some coding done in SR! > > Please help me if you can. > > -- John > > > I have the same problem when linking the cstest in ./csw. Here is the error message: as -o asm.o asm.s . ../paths.sh; cc -o cstest cstest.o asm.o $LIBM $LIBC ld: 0711-780 SEVERE ERROR: Symbol .contextdat (entry 18) in object asm.o: Relocation overflow in reference to: .sr_stk_overflow (entry 2) Address: 0x00000154; RLD type: R_RBR; RLD length: 26 ld: 0711-780 SEVERE ERROR: Symbol .contextdat (entry 18) in object asm.o: Relocation overflow in reference to: .sr_stk_underflow (entry 4) Address: 0x00000158; RLD type: R_RBR; RLD length: 26 ld: 0711-780 SEVERE ERROR: Symbol .contextdat (entry 18) in object asm.o: Relocation overflow in reference to: .sr_stk_corrupted (entry 6) Address: 0x0000015c; RLD type: R_RBR; RLD length: 26 make: 1254-004 The error code from the last command is 12. The c compiler V3.1.2 is used for building sr in AIX4.1.1. Is there anything to do with the cstest.c, asm.c, or rs6000.s? Richard From john_townsend@taligent.com Wed Feb 22 11:22:26 1995 Received: from mailserv.taligent.com by optima.cs.arizona.edu (5.65c/15) via SMTP id AA22415; Wed, 22 Feb 1995 12:30:14 MST Received: from qm.taligent.com by mailserv.taligent.com (AIX 3.2/UCB 5.64/4.03) id AA25028; Wed, 22 Feb 1995 11:30:41 -0800 Message-Id: Date: 22 Feb 1995 11:22:26 -0800 From: "John Townsend" Subject: Re: AIX 4.1? (fwd) To: admin@ibmserv.edu.tw, "info-sr" X-Mailer: Mail*Link SMTP-QM 3.0.2 I solved this problem. Turns out that there are some things in the xlC defs file that weren't there on AIX 3.2.5 that are they are AIX 4.1. They have to do with base addresses of the text and data segments. Unfortunately, I am at home at the moment and all of the info is at work. Tommorrow I will send out a message describing the problem to the info-sr mailing list. One note of warning.. to change the xlc defs file, you will need root access on your machine (at least I did). -------------------------------------- Date: 2/22/95 10:32 AM To: John Townsend From: admin@ibmserv.edu.tw Forwarded message: > From root Thu Feb 16 17:21:28 1995 > Message-Id: > Date: 18 Feb 1995 00:52:31 -0800 > Priority: Urgent > From: "John Townsend" > Subject: AIX 4.1? > To: "info-sr Mailing List" > X-Mailer: Mail*Link SMTP-QM 3.0.2 > Errors-To: info-sr-errors@cs.arizona.edu > > AIX 4.1? > Aaaaggghhhhhh > > I did something that appears to be very bad. Today my RS/6000 was upgraded > from AIX 3.2.5 to AIX 4.1. > > Unfortunately.. my sr setup is now badly hosted. I can't link the assembly > cstest code in the /csw directory. I get three link errors on the stack > underflow, overflow, etc. externs. > > Does anyone know anything about this? Has anyone ported SR to AIX 4.1? Am I > completely hosed? I hope not.. I really need to get some coding done in SR! > > Please help me if you can. > > -- John > > > I have the same problem when linking the cstest in ./csw. Here is the error message: as -o asm.o asm.s . ../paths.sh; cc -o cstest cstest.o asm.o $LIBM $LIBC ld: 0711-780 SEVERE ERROR: Symbol .contextdat (entry 18) in object asm.o: Relocation overflow in reference to: .sr_stk_overflow (entry 2) Address: 0x00000154; RLD type: R_RBR; RLD length: 26 ld: 0711-780 SEVERE ERROR: Symbol .contextdat (entry 18) in object asm.o: Relocation overflow in reference to: .sr_stk_underflow (entry 4) Address: 0x00000158; RLD type: R_RBR; RLD length: 26 ld: 0711-780 SEVERE ERROR: Symbol .contextdat (entry 18) in object asm.o: Relocation overflow in reference to: .sr_stk_corrupted (entry 6) Address: 0x0000015c; RLD type: R_RBR; RLD length: 26 make: 1254-004 The error code from the last command is 12. The c compiler V3.1.2 is used for building sr in AIX4.1.1. Is there anything to do with the cstest.c, asm.c, or rs6000.s? Richard ------------------ RFC822 Header Follows ------------------ Received: by qm.taligent.com with SMTP;21 Feb 1995 22:21:05 -0800 Received: from taligent.com by mailserv.taligent.com (AIX 3.2/UCB 5.64/4.03) id AA256491; Tue, 21 Feb 1995 22:19:58 -0800 Received: from optima.CS.Arizona.EDU by taligent.com with SMTP (5.67/23-Oct-1991-eef) id AA11931; Tue, 21 Feb 95 22:14:31 -0800 for Received: by optima.cs.arizona.edu (5.65c/15) id AA12009; Tue, 21 Feb 1995 22:59:33 MST From: admin@ibmserv.edu.tw Message-Id: <9502210621.AA14745@ibmserv.edu.tw> Subject: AIX 4.1? (fwd) To: info-sr@cs.arizona.edu Date: Mon, 20 Feb 1995 14:21:43 -1600 () X-Mailer: ELM [version 2.4 PL23alpha2] Content-Type: text Content-Length: 1945 Errors-To: info-sr-errors@cs.arizona.edu From shartley@king.mcs.drexel.edu Fri Feb 24 12:06:55 1995 Received: from mcs.drexel.edu (king.mcs.drexel.edu) by optima.cs.arizona.edu (5.65c/15) via SMTP id AA16117; Fri, 24 Feb 1995 10:06:11 MST Received: by mcs.drexel.edu (4.1/SMI-4.0) id AA20722; Fri, 24 Feb 95 12:06:55 EST Date: Fri, 24 Feb 95 12:06:55 EST From: shartley@king.mcs.drexel.edu (Stephen J. Hartley) Message-Id: <9502241706.AA20722@mcs.drexel.edu> To: info-sr@cs.arizona.edu Subject: SR on IBM SP2 I am pleased to report that SR compiles and runs "out of the box" on IBM's SP2 distributed memory parallel machine. I am not sure what versions of AIX and the C compiler (cc command) are running at the Maui High Performance Computing Center where I tested SR. The SP2 at MHPCC has 400 nodes, each with its own CPU and private memory, connected to each other by Ethernet and by a special high-speed switch. Not all nodes are always accessible by all users. Compiling SR "out of the box" allows virtual machines on different nodes to communicate over the Ethernet connections but not the switch connections. Implementing the latter will take some work digging through the SP2 documentation. The program attached at the end starts worker processes on a number of nodes to solve the N queens problem using the master/slave technique. SP2 users must have an .rhosts file in their home directories containing all the Ethernet node names for rsh commands to work. Program sample output is at the end. The only problem I have encountered so far is that srv/srv -v quick reports a problem with the optimized jumble.sr test program. 0 sr -O -T -c jumble.sr 0 srl ggg rrr jumble 0 a.out 01992 161.803e-2 refrigerator >withopt.out srv/srv: 19748 Memory fault: A memory image file is created as "core". expected 0, got 139 (segmentation violation) from a.out Steve Hartley Drexel Univ shartley@mcs.drexel.edu =============== queens.sr ============================ global rhosts const NUM_RHOSTS : int := 400 var rhost[1:NUM_RHOSTS] : string[64] body rhosts() var rhosts_file : file var name : string[64] rhosts_file := open("/u/shartley/.rhosts", READ) if rhosts_file = null -> write("cannot open .rhosts"); stop(1) fi fa i := 1 to NUM_RHOSTS -> read(rhosts_file, name) rhost[i] := name af write("rhosts:", NUM_RHOSTS, "rhosts read, rhost[1]=", rhost[1]) end rhosts resource worker import Master op my_assignment(r: int) body worker(id : int; n : int; mcap : cap Master) write("Worker", id, "starting up with board size", n) var solutions : int var board[1:n]: int := ([n] 0) procedure safe(row, column: int; board[1:*]: int) returns answer: bool # Check whether it is safe to place a queen at (row,column); # i.e., is board[column]=row a safe configuration? fa j := 1 to column-1 -> if board[column-j] = row or board[column-j] = row-j or board[column-j] = row+j -> answer := false return fi af answer := true end safe procedure place(column: int; board[1:*]: int) # Place a queen in all safe positions of column c, # then try placing a queen in the next column. # If a position in column n is safe, print the board. fa row := 1 to n -> board[column] := row # try placing a queen in (row,column) if safe(row, column, board) -> if column = n -> solutions++ # have a solution [] column place(column+1, board) # try next column fi fi board[column] := 0 # unrecord that a queen was placed af end place process working var r : int send mcap.give_assignment(id, 0) # tell master this worker is alive solutions := 0 do true -> receive my_assignment(r) solutions := 0 board[1] := r write("Worker", id, "starts queen in column 1, row", r) place(2, board) send mcap.give_assignment(id, solutions) od end working end worker resource Master import rhosts, worker op give_assignment(id : int; solutions : int) body Master() var board_size : int := 0 var num_workers : int := 0 var first_rhost : int := 1 var id : int getarg(1, board_size) # First argument on command line is board size. if board_size <= 1 -> write("board_size =", board_size, "is too small"); stop fi getarg(2, num_workers) # Second argument on command line is number workers. if num_workers <= 0 -> write("num_workers =", num_workers, "is too small"); stop fi getarg(3, first_rhost) # Third argument on command line is first rhost. if first_rhost <= 0 or first_rhost + num_workers - 1 > NUM_RHOSTS -> write("first_rhost =", first_rhost, "is too big or small"); stop fi var machine_cap[1:num_workers] : cap vm fa i := 1 to num_workers -> machine_cap[i] := create vm() on rhost[first_rhost+i-1] if machine_cap[i] = null -> write("Cannot create vm on", rhost[first_rhost+i-1], "so abort"); stop fi write("age=", age(), "Virtual machine up on", rhost[first_rhost+i-1]) af var worker_cap[1:num_workers] : cap worker var my_cap : cap Master var total_solutions : int var solutions : int my_cap := myresource() fa i := 1 to num_workers -> worker_cap[i] := create worker(i, board_size, my_cap) on machine_cap[i] af write("age=", age(), "Solutions to the", board_size, "queens problem will be farmed out to", num_workers, "workers") total_solutions := 0 # Send all the different initial board positions (queen in some row of column # one) to any worker waiting for an assignment. The parameter solutions in # give_assignment will be zero the first time it is received from a worker. # This serves as a signal to the master that the worker is alive. fa r := 1 to board_size -> receive give_assignment(id, solutions) if solutions > 0 -> total_solutions := total_solutions + solutions write("age=", age(), "Worker", id, "found", solutions, "solutions") fi send worker_cap[id].my_assignment(r) af # All initial positions have been passed out. Wait for workers to finish. fa i := 1 to num_workers -> receive give_assignment(id, solutions) total_solutions := total_solutions + solutions write("age=", age(), "Worker", id, "done, found", solutions, "solutions") af write("age=", age(), "Total number of solutions found was", total_solutions) stop end Master /* ............... Example compile and run(s) % sr -o queens queens.sr % ./queens 9 4 rhosts: 400 rhosts read, rhost[1]= fr1n01.mhpcc.edu Your account is not authorized to login to this machine, please login in to another machine for all uses. Cannot create vm on fr1n01.mhpcc.edu so abort % ./queens 9 4 48 rhosts: 400 rhosts read, rhost[1]= fr1n01.mhpcc.edu age= 880 Virtual machine up on fr3n16.mhpcc.edu age= 1525 Virtual machine up on fr4n01.mhpcc.edu fr4n02.mhpcc.edu: A remote host did not respond within the timeout period. Cannot create vm on fr4n02.mhpcc.edu so abort % ./queens 9 4 53 rhosts: 400 rhosts read, rhost[1]= fr1n01.mhpcc.edu age= 670 Virtual machine up on fr4n05.mhpcc.edu age= 1287 Virtual machine up on fr4n06.mhpcc.edu age= 2089 Virtual machine up on fr4n07.mhpcc.edu age= 2706 Virtual machine up on fr4n08.mhpcc.edu rhosts: 400 rhosts read, rhost[1]= fr1n01.mhpcc.edu Worker 1 starting up with board size 9 rhosts: 400 rhosts read, rhost[1]= fr1n01.mhpcc.edu Worker 2 starting up with board size 9 rhosts: 400 rhosts read, rhost[1]= fr1n01.mhpcc.edu Worker 3 starting up with board size 9 rhosts: 400 rhosts read, rhost[1]= fr1n01.mhpcc.edu age= 4124 Solutions to the 9 queens problem will be farmed out to 4 workers Worker 1 starts queen in column 1, row 1 Worker 2 starts queen in column 1, row 2 Worker 3 starts queen in column 1, row 3 Worker 4 starting up with board size 9 age= 4303 Worker 1 found 28 solutions Worker 4 starts queen in column 1, row 4 Worker 1 starts queen in column 1, row 5 age= 4328 Worker 2 found 30 solutions Worker 2 starts queen in column 1, row 6 age= 4335 Worker 3 found 47 solutions age= 4336 Worker 4 found 44 solutions Worker 3 starts queen in column 1, row 7 age= 4518 Worker 1 found 54 solutions Worker 4 starts queen in column 1, row 8 Worker 1 starts queen in column 1, row 9 age= 4530 Worker 2 done, found 44 solutions age= 4537 Worker 4 done, found 30 solutions age= 4542 Worker 3 done, found 47 solutions age= 4707 Worker 1 done, found 28 solutions age= 4708 Total number of solutions found was 352 */ From grindsta@cs.ucdavis.edu Mon Feb 27 18:21:52 1995 Received: from toadflax.cs.ucdavis.edu by optima.cs.arizona.edu (5.65c/15) via SMTP id AA11659; Mon, 27 Feb 1995 19:22:02 MST Received: from burks.cs.ucdavis.edu by toadflax.cs.ucdavis.edu (4.1/UCD.CS.2.6) id AA21121; Mon, 27 Feb 95 18:21:58 PST Received: by burks.cs.ucdavis.edu (5.65/UCD.CS.2.6) id AA17288; Mon, 27 Feb 1995 18:21:53 -0800 Date: Mon, 27 Feb 1995 18:21:52 -0800 (PST) From: Marcus Andrew Grindstaff To: info-sr@cs.arizona.edu Subject: Placing BMP or GIF in window in SRWin (fwd) Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII ---------- Forwarded message ---------- Date: Mon, 27 Feb 1995 18:10:35 -0800 From: Ron Olsson To: grindsta@burks.cs.ucdavis.edu Subject: Placing BMP or GIF in window in SRWin > I don't know. I presume you've checked the man page. If so, post a > message to info-sr@cs.arizona.edu. Please follow up anything you find > out to the newsgroup. Thanks. Is it possible to display BMPs or GIFs (or any other graphics format) in SRWin? Thanks, Marcus ---------------------------- Marcus Grindstaff Technical Marketing Engineer PCI Components (916)948-8493 From azhao@cc.gatech.edu Tue Feb 28 00:45:18 1995 Received: from burdell.cc.gatech.edu by optima.cs.arizona.edu (5.65c/15) via SMTP id AA22307; Mon, 27 Feb 1995 22:45:21 MST Received: from oakmont.cc.gatech.edu (azhao@oakmont.cc.gatech.edu [130.207.119.218]) by burdell.cc.gatech.edu (8.6.10/8.6.9) with ESMTP id AAA18016; Tue, 28 Feb 1995 00:45:19 -0500 Received: (from azhao@localhost) by oakmont.cc.gatech.edu (8.6.10/8.6.9) id AAA01382; Tue, 28 Feb 1995 00:45:18 -0500 From: azhao@cc.gatech.edu (Q. Alex Zhao) Message-Id: <199502280545.AAA01382@oakmont.cc.gatech.edu> Subject: Re: Placing BMP or GIF in window in SRWin To: grindsta@cs.ucdavis.edu (Marcus Andrew Grindstaff) Date: Tue, 28 Feb 1995 00:45:18 -0500 (EST) Cc: info-sr@cs.arizona.edu In-Reply-To: from "Marcus Andrew Grindstaff" at Feb 27, 95 06:21:52 pm X-Mailer: ELM [version 2.4 PL23] Content-Type: text Content-Length: 1681 According to Marcus Andrew Grindstaff in a previous message: ] ] From: Ron Olsson ] To: grindsta@burks.cs.ucdavis.edu ] Subject: Placing BMP or GIF in window in SRWin ] ] > I don't know. I presume you've checked the man page. If so, post a ] > message to info-sr@cs.arizona.edu. Please follow up anything you find ] > out to the newsgroup. Thanks. ] ] Is it possible to display BMPs or GIFs (or any other graphics format) in ] SRWin? ] ] Thanks, ] Marcus ] ] ---------------------------- ] Marcus Grindstaff ] Technical Marketing Engineer ] PCI Components ] (916)948-8493 SRWin doesn't directly support image file I/O. If you just want to display an image you could write the image to a file (in the correct format) and spawn an "xloadimage" or "xv" to display it. SRWin uses the default colormap which is shared among all X clients on the same display. An SRWin program could easily run out of color if it displays color images. I guess SRWin need to be extended to include better support of colormap manipulation to be flexible in handling images... In the short term you could try building a C function and an SR interface (similar to how SRWin is implemented); using existing data structures, they could read in the image, find the off-screen pixmap associated with a "winWindow" structure and put the image (dithered to the available colorspace) on the pixmap, then call "WinUpdate"... Cheers, = Q. Alex Zhao ~{0"UT~} ! Graphics, Visualization & Usability Center Email: azhao@cc.gatech.edu ! College of Computing FAX: 404-853-9378 ! Georgia Institute of Technology Lab: 404-894-9636 ! Atlanta, Georgia 30332-0280 From admin@ibmserv.edu.tw Mon Feb 27 18:00:25 1995 Received: from ibmserv.edu.tw by optima.cs.arizona.edu (5.65c/15) via SMTP id AA15663; Wed, 1 Mar 1995 02:39:51 MST Received: by ibmserv.edu.tw (AIX 3.2/UCB 5.64/4.03) id AA12199; Mon, 27 Feb 1995 18:00:26 -1600 From: admin@ibmserv.edu.tw Message-Id: <9502281000.AA12199@ibmserv.edu.tw> Subject: Re: AIX 4.1? (fwd) To: info-sr@cs.arizona.edu Date: Mon, 27 Feb 1995 18:00:25 -1600 () X-Mailer: ELM [version 2.4 PL23alpha2] Content-Type: text Content-Length: 2067 What are the xlC defs needed to be changed? could you show me the details? -- Richard Forwarded message: > From root Tue Feb 21 03:57:34 1995 > Message-Id: > Date: 22 Feb 1995 11:22:26 -0800 > From: "John Townsend" > Subject: Re: AIX 4.1? (fwd) > To: admin@ibmserv.edu.tw, "info-sr" > X-Mailer: Mail*Link SMTP-QM 3.0.2 > Errors-To: info-sr-errors@cs.arizona.edu > > I solved this problem. Turns out that there are some things in the xlC defs > file that weren't there on AIX 3.2.5 that are they are AIX 4.1. They have to > do with base addresses of the text and data segments. > > Unfortunately, I am at home at the moment and all of the info is at work. > Tommorrow I will send out a message describing the problem to the info-sr > mailing list. > > One note of warning.. to change the xlc defs file, you will need root access > on your machine (at least I did). > -------------------------------------- > > > > I have the same problem when linking the cstest in ./csw. Here is the error > message: > > as -o asm.o asm.s > . ../paths.sh; cc -o cstest cstest.o asm.o $LIBM $LIBC > ld: 0711-780 SEVERE ERROR: Symbol .contextdat (entry 18) in object asm.o: > Relocation overflow in reference to: .sr_stk_overflow (entry 2) > Address: 0x00000154; RLD type: R_RBR; RLD length: 26 > ld: 0711-780 SEVERE ERROR: Symbol .contextdat (entry 18) in object asm.o: > Relocation overflow in reference to: .sr_stk_underflow (entry 4) > Address: 0x00000158; RLD type: R_RBR; RLD length: 26 > ld: 0711-780 SEVERE ERROR: Symbol .contextdat (entry 18) in object asm.o: > Relocation overflow in reference to: .sr_stk_corrupted (entry 6) > Address: 0x0000015c; RLD type: R_RBR; RLD length: 26 > make: 1254-004 The error code from the last command is 12. > > The c compiler V3.1.2 is used for building sr in AIX4.1.1. Is there > anything to do with the cstest.c, asm.c, or rs6000.s? > > > Richard > From shartley@king.mcs.drexel.edu Mon Mar 6 07:17:55 1995 Received: from mcs.drexel.edu (king.mcs.drexel.edu) by optima.cs.arizona.edu (5.65c/15) via SMTP id AA02789; Mon, 6 Mar 1995 05:17:09 MST Received: by mcs.drexel.edu (4.1/SMI-4.0) id AA10914; Mon, 6 Mar 95 07:17:55 EST Date: Mon, 6 Mar 95 07:17:55 EST From: shartley@king.mcs.drexel.edu (Stephen J. Hartley) Message-Id: <9503061217.AA10914@mcs.drexel.edu> To: info-sr@cs.arizona.edu Subject: finally The following is finally (whew!) available from Oxford University Press: "Operating Systems Programming: The SR Programming Language". The ISBN is 0-19-5095790. Contact them at (800) 451-7556 for a copy. Suggested list price is $25.00. About 236 pages. The example SR programs used in the book are avaliable by anonymous ftp from site mcs.drexel.edu in directory pub/shartley; retrieve file book_SR_examples.tar.gz (remember to use binary mode) to get a tar archive compressed with the GNU compression program gzip. The SR language is available by anonymous ftp from site cs.arizona.edu in directory sr. Steve Hartley Drexel University shartley@mcs.drexel.edu From the Preface: SR is a language for concurrent programming. This book describes the SR language, presents some examples of SR programs in the context of an operating systems or concurrent programming course, and provides some programming as- signments in the form of open Student Laboratories. The SR language can be used by instructors of concurrent programming or operating systems courses to give students experience in writing concurrent programs that use multiple processes, semaphores, monitors, message passing, remote procedure calls, and the rendezvous. The language can also be used for parallel computing in a shared-memory multiprocessor or a distributed memory cluster environment. From gln@cs.simpson.edu Tue Mar 7 15:27:24 1995 Received: from cs.simpson.edu by optima.cs.arizona.edu (5.65c/15) via SMTP id AA28366; Tue, 7 Mar 1995 14:27:28 MST Received: by cs.simpson.edu; id AA08256; Tue, 7 Mar 1995 15:27:24 -0600 Date: Tue, 7 Mar 1995 15:27:24 -0600 From: Gary L. Newell Message-Id: <9503072127.AA08256@cs.simpson.edu> To: info-sr@cs.arizona.edu Subject: hello Hi folks, hows it goin? I recently (actually as I write this) decided to download SR for our departmental machine. It is a DEC Alpha running OSF and as I know that you have at least one such beast in the department now I was wondering if there is a clean port of SR to OSF or if I'm on my own. This question might be answered once I take a look at the downloaded material and if so let me apologize in advance. Either way, could you please add me to the mailing list? Thanks. Hope things are going well in sunny Tucson - I'm not sure who is officially part of the SR group now a days so I'm afraid I will have to avoid any personal hellos... later Gary Newell gln@cs.simpson.edu From shartley@king.mcs.drexel.edu Thu Mar 9 17:50:37 1995 Received: from mcs.drexel.edu (king.mcs.drexel.edu) by optima.cs.arizona.edu (5.65c/15) via SMTP id AA10634; Thu, 9 Mar 1995 15:49:56 MST Received: by mcs.drexel.edu (4.1/SMI-4.0) id AA26731; Thu, 9 Mar 95 17:50:37 EST Date: Thu, 9 Mar 95 17:50:37 EST From: shartley@king.mcs.drexel.edu (Stephen J. Hartley) Message-Id: <9503092250.AA26731@mcs.drexel.edu> To: info-sr@cs.arizona.edu Subject: performance question The attached program was run on a Sparc multiprocessor with 4 CPUs. It was compiled with the MultiSR version of SR. It runs markedly slower when SR_PARALLEL is set bigger than 1. What is slowing it down? Sample output is at the end of the program in the form of a big comment. ================ queens.sr ==================== resource worker import Master op my_assignment(r: int) body worker(id : int; n : int; mcap : cap Master) write("Worker", id, "starting up with board size", n) var solutions : int var board[1:n]: int := ([n] 0) procedure safe(row, column: int; board[1:*]: int) returns answer: bool # Check whether it is safe to place a queen at (row,column); # i.e., is board[column]=row a safe configuration? fa j := 1 to column-1 -> if board[column-j] = row or board[column-j] = row-j or board[column-j] = row+j -> answer := false return fi af answer := true end safe procedure place(column: int; board[1:*]: int) # Place a queen in all safe positions of column c, # then try placing a queen in the next column. # If a position in column n is safe, print the board. fa row := 1 to n -> board[column] := row # try placing a queen in (row,column) if safe(row, column, board) -> if column = n -> solutions++ # have a solution [] column place(column+1, board) # try next column fi fi board[column] := 0 # unrecord that a queen was placed af end place process working var r : int send mcap.give_assignment(id, 0) # tell master this worker is alive solutions := 0 do true -> receive my_assignment(r) solutions := 0 board[1] := r write("Worker", id, "starts queen in column 1, row", r) place(2, board) send mcap.give_assignment(id, solutions) od end working end worker resource Master import worker op give_assignment(id : int; solutions : int) body Master() var board_size : int := 0 var num_workers : int := 0 var id : int getarg(1, board_size) if board_size <= 1 -> write("board_size =", board_size, "is too small"); stop fi getarg(2, num_workers) if num_workers <= 0 -> write("num_workers =", num_workers, "is too small"); stop fi var worker_cap[1:num_workers] : cap worker var my_cap : cap Master var total_solutions : int var solutions : int my_cap := myresource() fa i := 1 to num_workers -> worker_cap[i] := create worker(i, board_size, my_cap) af write("age=", age(), "Solutions to the", board_size, "queens problem will be farmed out to", num_workers, "workers") total_solutions := 0 # Send all the different initial board positions (queen in some row of column # one) to any worker waiting for an assignment. The parameter solutions in # give_assignment will be zero the first time it is received from a worker. # This serves as a signal to the master that the worker is alive. fa r := 1 to board_size -> receive give_assignment(id, solutions) if solutions > 0 -> total_solutions := total_solutions + solutions write("age=", age(), "Worker", id, "found", solutions, "solutions") fi send worker_cap[id].my_assignment(r) af # All initial positions have been passed out. Wait for workers to finish. fa i := 1 to num_workers -> receive give_assignment(id, solutions) total_solutions := total_solutions + solutions write("age=", age(), "Worker", id, "done, found", solutions, "solutions") af write("age=", age(), "Total number of solutions found was", total_solutions) stop end Master /* ............... Example compile and run(s) % uname -a SunOS bandersnatch 5.3 Generic sun4m sparc % sr -v MultiSR version 2.3, October 1994 % sr -o queens queens.sr % time ./queens 10 4 Worker 1 starting up with board size 10 Worker 2 starting up with board size 10 Worker 3 starting up with board size 10 Worker 4 starting up with board size 10 age= 42 Solutions to the 10 queens problem will be farmed out to 4 workers Worker 1 starts queen in column 1, row 1 Worker 2 starts queen in column 1, row 2 Worker 3 starts queen in column 1, row 3 Worker 4 starts queen in column 1, row 4 age= 5149 Worker 1 found 64 solutions Worker 1 starts queen in column 1, row 5 age= 5915 Worker 4 found 93 solutions age= 5916 Worker 2 found 48 solutions age= 5917 Worker 3 found 65 solutions Worker 4 starts queen in column 1, row 6 Worker 2 starts queen in column 1, row 7 Worker 3 starts queen in column 1, row 8 age= 11589 Worker 1 found 92 solutions age= 11590 Worker 2 found 93 solutions Worker 1 starts queen in column 1, row 9 Worker 2 starts queen in column 1, row 10 age= 11793 Worker 4 done, found 92 solutions age= 11794 Worker 3 done, found 65 solutions age= 14163 Worker 2 done, found 64 solutions age= 14363 Worker 1 done, found 48 solutions age= 14364 Total number of solutions found was 724 14.0u 0.0s 0:14 96% 0+0k 0+0io 0pf+0w % setenv SR_PARALLEL 1 % time ./queens 10 4 Worker 1 starting up with board size 10 Worker 2 starting up with board size 10 Worker 3 starting up with board size 10 Worker 4 starting up with board size 10 age= 42 Solutions to the 10 queens problem will be farmed out to 4 workers Worker 1 starts queen in column 1, row 1 Worker 2 starts queen in column 1, row 2 Worker 3 starts queen in column 1, row 3 Worker 4 starts queen in column 1, row 4 age= 5633 Worker 1 found 64 solutions Worker 1 starts queen in column 1, row 5 age= 6432 Worker 4 found 93 solutions age= 6433 Worker 2 found 48 solutions age= 6434 Worker 3 found 65 solutions Worker 4 starts queen in column 1, row 6 Worker 2 starts queen in column 1, row 7 Worker 3 starts queen in column 1, row 8 age= 12807 Worker 1 found 92 solutions age= 12808 Worker 2 found 93 solutions Worker 1 starts queen in column 1, row 9 Worker 2 starts queen in column 1, row 10 age= 13030 Worker 4 done, found 92 solutions age= 13031 Worker 3 done, found 65 solutions age= 15332 Worker 2 done, found 64 solutions age= 15523 Worker 1 done, found 48 solutions age= 15524 Total number of solutions found was 724 15.0u 0.0s 0:15 95% 0+0k 0+0io 0pf+0w % setenv SR_PARALLEL 2 % time ./queens 10 4 Worker 1 starting up with board size 10 Worker 2 starting up with board size 10 Worker 3 starting up with board size 10 Worker 4 starting up with board size 10 age= 47 Solutions to the 10 queens problem will be farmed out to 4 workers Worker 1 starts queen in column 1, row 1 Worker 2 starts queen in column 1, row 2 Worker 3 starts queen in column 1, row 3 Worker 4 starts queen in column 1, row 4 age= 17378 Worker 1 found 64 solutions Worker 1 starts queen in column 1, row 5 age= 19392 Worker 2 found 48 solutions Worker 2 starts queen in column 1, row 6 age= 21436 Worker 4 found 93 solutions Worker 4 starts queen in column 1, row 7 age= 24186 Worker 3 found 65 solutions Worker 3 starts queen in column 1, row 8 age= 37968 Worker 1 found 92 solutions age= 37970 Worker 2 found 92 solutions Worker 1 starts queen in column 1, row 9 Worker 2 starts queen in column 1, row 10 age= 41285 Worker 4 done, found 93 solutions age= 47439 Worker 3 done, found 65 solutions age= 49919 Worker 2 done, found 64 solutions age= 50093 Worker 1 done, found 48 solutions age= 50094 Total number of solutions found was 724 51.0u 37.0s 0:50 175% 0+0k 0+0io 0pf+0w % setenv SR_PARALLEL 3 % time ./queens 10 4 Worker 1 starting up with board size 10 Worker 2 starting up with board size 10 Worker 3 starting up with board size 10 Worker 4 starting up with board size 10 age= 53 Solutions to the 10 queens problem will be farmed out to 4 workers Worker 1 starts queen in column 1, row 1 Worker 3 starts queen in column 1, row 3 Worker 2 starts queen in column 1, row 2 Worker 4 starts queen in column 1, row 4 age= 17910 Worker 1 found 64 solutions Worker 1 starts queen in column 1, row 5 age= 19555 Worker 4 found 93 solutions age= 19557 Worker 2 found 48 solutions Worker 4 starts queen in column 1, row 6 Worker 2 starts queen in column 1, row 7 age= 21026 Worker 3 found 65 solutions Worker 3 starts queen in column 1, row 8 age= 38338 Worker 2 found 93 solutions age= 38339 Worker 1 found 92 solutions Worker 2 starts queen in column 1, row 9 Worker 1 starts queen in column 1, row 10 age= 40594 Worker 4 done, found 92 solutions age= 41867 Worker 3 done, found 65 solutions age= 48549 Worker 2 done, found 48 solutions age= 48553 Worker 1 done, found 64 solutions age= 48554 Total number of solutions found was 724 51.0u 58.0s 0:48 223% 0+0k 0+0io 0pf+0w % setenv SR_PARALLEL 4 % time ./queens 10 4 Worker 1 starting up with board size 10 Worker 2 starting up with board size 10 Worker 3 starting up with board size 10 Worker 4 starting up with board size 10 age= 55 Solutions to the 10 queens problem will be farmed out to 4 workers Worker 1 starts queen in column 1, row 1 Worker 2 starts queen in column 1, row 2 Worker 3 starts queen in column 1, row 3 Worker 4 starts queen in column 1, row 4 age= 14024 Worker 2 found 48 solutions Worker 2 starts queen in column 1, row 5 age= 15591 Worker 1 found 64 solutions Worker 1 starts queen in column 1, row 6 age= 16641 Worker 4 found 93 solutions Worker 4 starts queen in column 1, row 7 age= 20334 Worker 3 found 65 solutions Worker 3 starts queen in column 1, row 8 age= 30768 Worker 2 found 92 solutions Worker 2 starts queen in column 1, row 9 age= 31640 Worker 1 found 92 solutions Worker 1 starts queen in column 1, row 10 age= 32458 Worker 4 done, found 93 solutions age= 39402 Worker 3 done, found 65 solutions age= 39476 Worker 2 done, found 48 solutions age= 39699 Worker 1 done, found 64 solutions age= 39700 Total number of solutions found was 724 41.0u 49.0s 0:39 226% 0+0k 0+0io 0pf+0w % time ./queens 10 3 Worker 1 starting up with board size 10 Worker 2 starting up with board size 10 Worker 3 starting up with board size 10 age= 46 Solutions to the 10 queens problem will be farmed out to 3 workers Worker 1 starts queen in column 1, row 1 Worker 2 starts queen in column 1, row 2 Worker 3 starts queen in column 1, row 3 age= 12149 Worker 1 found 64 solutions Worker 1 starts queen in column 1, row 4 age= 14508 Worker 3 found 65 solutions Worker 3 starts queen in column 1, row 5 age= 14829 Worker 2 found 48 solutions Worker 2 starts queen in column 1, row 6 age= 25927 Worker 1 found 93 solutions Worker 1 starts queen in column 1, row 7 age= 27898 Worker 2 found 92 solutions Worker 2 starts queen in column 1, row 8 age= 29851 Worker 3 found 92 solutions Worker 3 starts queen in column 1, row 9 age= 38592 Worker 1 found 93 solutions Worker 1 starts queen in column 1, row 10 age= 42044 Worker 2 done, found 65 solutions age= 43960 Worker 3 done, found 48 solutions age= 44311 Worker 1 done, found 64 solutions age= 44312 Total number of solutions found was 724 47.0u 47.0s 0:44 211% 0+0k 0+0io 0pf+0w % unsetenv SR_PARALLEL % time ./queens 10 3 Worker 1 starting up with board size 10 Worker 2 starting up with board size 10 Worker 3 starting up with board size 10 age= 35 Solutions to the 10 queens problem will be farmed out to 3 workers Worker 1 starts queen in column 1, row 1 Worker 2 starts queen in column 1, row 2 Worker 3 starts queen in column 1, row 3 age= 3660 Worker 1 found 64 solutions Worker 1 starts queen in column 1, row 4 age= 4329 Worker 2 found 48 solutions age= 4330 Worker 3 found 65 solutions Worker 2 starts queen in column 1, row 5 Worker 3 starts queen in column 1, row 6 age= 7662 Worker 1 found 93 solutions Worker 1 starts queen in column 1, row 7 age= 8713 Worker 3 found 92 solutions age= 8714 Worker 2 found 92 solutions Worker 3 starts queen in column 1, row 8 Worker 2 starts queen in column 1, row 9 age= 11753 Worker 1 found 93 solutions Worker 1 starts queen in column 1, row 10 age= 12860 Worker 2 done, found 48 solutions age= 12861 Worker 3 done, found 65 solutions age= 13773 Worker 1 done, found 64 solutions age= 13774 Total number of solutions found was 724 13.0u 0.0s 0:13 93% 0+0k 0+0io 0pf+0w */ From nlam@eecs.wsu.edu Fri Mar 10 09:40:01 1995 Received: from dns1.eecs.wsu.edu by optima.cs.arizona.edu (5.65c/15) via SMTP id AA11929; Fri, 10 Mar 1995 10:40:04 MST Received: from dryden.eecs.wsu.edu by dns1.eecs.wsu.edu (8.6.10/8.940801) id JAA25551; Fri, 10 Mar 1995 09:40:02 -0800 From: Norman Lam - EECS Received: by dryden.eecs.wsu.edu (1.37.109.4) id AA10259; Fri, 10 Mar 95 09:40:01 -0800 Message-Id: <9503101740.AA10259@dryden.eecs.wsu.edu> Subject: Stack overflow To: info-sr@cs.arizona.edu (SR User Group) Date: Fri, 10 Mar 1995 09:40:01 -0800 (PST) X-Mailer: ELM [version 2.4 PL23] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 378 Hi, What does the error message "RTS: stack overflow" mean typically? Is there any document on what the error messages mean? My sr program compiled correctly but gave the above error message during execution. The error occurred where a procedure is called and none of the statements inside the procedure got executed as a result. Your help is greatly appreciated. Norman Lam From gmt Fri Mar 10 13:32:09 1995 Received: from hawk.CS.Arizona.EDU by optima.cs.arizona.edu (5.65c/15) via SMTP id AA23195; Fri, 10 Mar 1995 13:32:14 MST Received: by hawk.CS.Arizona.EDU; (5.65/1.1.8.2/08Nov94-0446PM) id AA22447; Fri, 10 Mar 1995 13:32:09 -0700 Date: Fri, 10 Mar 1995 13:32:09 -0700 From: Gregg Townsend Message-Id: <9503102032.AA22447@hawk.CS.Arizona.EDU> To: info-sr@cs.arizona.edu, shartley@king.mcs.drexel.edu Subject: Re: performance question From: shartley@king.mcs.drexel.edu (Stephen J. Hartley) The attached program was run on a Sparc multiprocessor with 4 CPUs. It was compiled with the MultiSR version of SR. It runs markedly slower when SR_PARALLEL is set bigger than 1. What is slowing it down? % setenv SR_PARALLEL 1 % time ./queens 10 4 15.0u 0.0s 0:15 95% 0+0k 0+0io 0pf+0w % setenv SR_PARALLEL 2 % time ./queens 10 4 51.0u 37.0s 0:50 175% 0+0k 0+0io 0pf+0w % setenv SR_PARALLEL 3 % time ./queens 10 4 51.0u 58.0s 0:48 223% 0+0k 0+0io 0pf+0w % setenv SR_PARALLEL 4 % time ./queens 10 4 41.0u 49.0s 0:39 226% 0+0k 0+0io 0pf+0w Of course, there is some cost to running with parallelism enabled, as the various parts of the runtime system need to coordinate with each other. truss(1) shows a huge number of system lock/unlock/p/v calls with SR_PARALLEL>1 that aren't there with SR_PARALLEL=1. Still, I wouldn't expect to see a slowdown like you reported. Without trying to analyze this particular program, I ran it several times on our multiprocessor Sparcstation. This system is running the same versions of Solaris and SR as you reported. Here are the real (elapsed) times I recorded for "queens 10 4": SR_PARALLEL=1 12.6, 12.4, 12.9 SR_PARALLEL=2 12.8, 13.8, 14.2 SR_PARALLEL=3 29.3, 29.2, 30.9 SR_PARALLEL=4 30.2, 21.7, 33.1 In my experiments, the big increase came with the attempt to run with parallelism of 3. That makes some sense here, because our Sparcstation has only two processors. The user and system times you reported, though, do seem to indicate more than two processors available, so that doesn't explain your figures. I did not have a dedicated machine available for these experiments, so I watched /usr/openwin/bin/perfmeter to verify that the system was idle. I noticed that when I asked for more parallelism than really available, the number of system context switches on the perfmeter went way up -- whereas there were none earlier. On the on the other hand, with SR_PARALLEL=2 the program did not appear to actually make use of the second CPU. One really interesting discovery was that the SR_PARALLEL=3 case runs more quickly when the system is NOT idle. In fact, if I run two of those at once, each of them finishes in under 15 seconds, while alone a single one runs for about 30 seconds! I'm not sure what conclusions to draw from this other than that the Solaris scheduling sometimes seems strange. Gregg Townsend / Computer Science Dept / Univ of Arizona / Tucson, AZ 85721 +1 602 621 4325 gmt@cs.arizona.edu 110 57 16 W / 32 13 45 N / +758m From gmt Fri Mar 10 14:03:22 1995 Received: from hawk.CS.Arizona.EDU by optima.cs.arizona.edu (5.65c/15) via SMTP id AA25389; Fri, 10 Mar 1995 14:03:27 MST Received: by hawk.CS.Arizona.EDU; (5.65/1.1.8.2/08Nov94-0446PM) id AA22583; Fri, 10 Mar 1995 14:03:22 -0700 Date: Fri, 10 Mar 1995 14:03:22 -0700 From: Gregg Townsend Message-Id: <9503102103.AA22583@hawk.CS.Arizona.EDU> To: info-sr@cs.arizona.edu, nlam@eecs.wsu.edu Subject: Re: Stack overflow From: Norman Lam - EECS What does the error message "RTS: stack overflow" mean typically? Is there any document on what the error messages mean? My sr program compiled correctly but gave the above error message during execution. The error occurred where a procedure is called and none of the statements inside the procedure got executed as a result. There's no document describing SR error messages. Infinite recursion will produce a "stack overflow" (or worse) but the message can also occur in a correct program that just needs a larger stack. The stack size can be set using the "-S" option of srl. Unfortunately, we've found nothing better than trial and error for predicting stack sizes. The amount needed depends on lots of system- dependent things beyond the user's control, such as the requirements of the C runtime system. The default of 40000 is sufficient to run the programs in the SR test suite on all the machines we normally use. Gregg Townsend / Computer Science Dept / Univ of Arizona / Tucson, AZ 85721 +1 602 621 4325 gmt@cs.arizona.edu 110 57 16 W / 32 13 45 N / +758m From john_townsend@taligent.com Tue Mar 14 10:51:04 1995 Received: from mailserv.taligent.com by optima.cs.arizona.edu (5.65c/15) via SMTP id AA07072; Tue, 14 Mar 1995 11:52:09 MST Received: from qm.taligent.com by mailserv.taligent.com (AIX 3.2/UCB 5.64/4.03) id AA26827; Tue, 14 Mar 1995 10:53:23 -0800 Message-Id: Date: 14 Mar 1995 10:51:04 -0800 From: "John Townsend" Subject: SR on AIX 4.1 To: "info-sr Mailing List" X-Mailer: Mail*Link SMTP-QM 3.0.2 I apologize for not posting this sooner. I have been extrordinarily busy with projects, finals, job interviews, etc. and haven't had the time to remember what we did to solve this problem. Anyway.. here it is: Making SR run on AIX 4.1 1. You need root access on your RS/6000 box. 2. You need to make the following change in /etc/xlC.cfg * common definitions DEFLT: ... ... ... * this is the old line. The new line is the second line. * options = -D_AIX,-D_AIX32,-D_IBMR2,-D_POWER,-bpT:0x10000000,-bpD:0x20000000 options = -D_AIX,-D_AIX32,-D_IBMR2,-D_POWER Change the options line to NOT include the -bpT and -bpD linker options. These are options that specify the base address of text and data for the AIX linker. Near as we can tell, these are default settings and aren't neccessary. In fact, they mess up SR. 3. Please consider recompiling the entire SR package. 4. Now things should be working for you. That's all I did. Your mileage may vary. I may be using a different version of AIX 4.1 than you are using and I don't know how this change will effect non-SR work. (I build LOTS of stuff with xlC and everything is working just fine for me. Your experience could be different). And that's it. If you have any questions, please let me know... -- John Townsend From bergenhe@db.erau.edu Sat Mar 18 19:44:57 1995 Received: from db.erau.edu by optima.cs.arizona.edu (5.65c/15) via SMTP id AA13501; Sat, 18 Mar 1995 17:44:57 MST Received: from zeta.sun by db.erau.edu with smtp (Smail3.1.28.1 #2) id m0rq95q-0007zhC; Sat, 18 Mar 95 19:43 EST Message-Id: From: bergenhe@db.erau.edu (Erik Bergenholtz) Subject: virtual machine communication To: info-sr@cs.arizona.edu Date: Sat, 18 Mar 95 19:44:57 EST X-Mailer: ELM [version 2.3 PL11] Hi! I am trying to implement the Byzantine Generals Problem using SR. I am having difficulty getting virtual machines to communicate with each other. Is this possible to do? The "SR Programming Language" book does not explicitly mention this. Any information or help is greatly appreciated. Best Regards, Erik Bergenholtz e-mail: bergenhe@db.erau.edu From greg Mon Mar 20 08:24:07 1995 Received: from paloverde.CS.Arizona.EDU by optima.cs.arizona.edu (5.65c/15) via SMTP id AA20259; Mon, 20 Mar 1995 08:24:10 MST Received: by paloverde.CS.Arizona.EDU; (5.65/1.1.8.2/08Nov94-0446PM) id AA03425; Mon, 20 Mar 1995 08:24:07 -0700 Date: Mon, 20 Mar 1995 08:24:07 -0700 From: Greg Andrews Message-Id: <9503201524.AA03425@paloverde.CS.Arizona.EDU> To: bergenhe@db.erau.edu, info-sr@cs.arizona.edu Subject: Re: virtual machine communication From: bergenhe@db.erau.edu (Erik Bergenholtz) Subject: virtual machine communication To: info-sr@cs.arizona.edu Date: Sat, 18 Mar 95 19:44:57 EST Hi! I am trying to implement the Byzantine Generals Problem using SR. I am having difficulty getting virtual machines to communicate with each other. Is this possible to do? The "SR Programming Language" book does not explicitly mention this. Any information or help is greatly appreciated. Best Regards, Erik Bergenholtz e-mail: bergenhe@db.erau.edu Virtual machines are merely address spaces. Resources are the "active" unit in SR. Resources communicate with each other, and they can be placed on different virtual machines. There are examples of this in the SR book. Look at the Distributed File System chapter, for example. Greg Andrews From sirirut@csgrad.cs.vt.edu Fri Mar 31 10:05:08 1995 Received: from csgrad.cs.vt.edu by optima.cs.arizona.edu (5.65c/15) via SMTP id AA03384; Fri, 31 Mar 1995 08:05:10 MST Received: by csgrad.cs.vt.edu; (5.65/1.1.8.2/19Sep94-1023AM) id AA18111; Fri, 31 Mar 1995 10:05:09 -0500 Message-Id: <9503311505.AA18111@csgrad.cs.vt.edu> Subject: Want to be in mailing list To: info-sr@cs.arizona.edu Date: Fri, 31 Mar 1995 10:05:08 -0500 (EST) From: "Sirirut Vanichayobon" X-Mailer: ELM [version 2.4 PL23] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 73 I am interested in SR. My e-mail address is sirirut@csgrad.cs.vt.edu. From jperez@lensis3.lsi.us.es Wed Apr 12 18:49:00 1995 Received: from lensis3.lsi.us.es by optima.cs.arizona.edu (5.65c/15) via SMTP id AA20903; Wed, 12 Apr 1995 08:50:21 MST Received: by lensis3.lsi.us.es (AIX 3.2/UCB 5.64/4.03) id AA14161; Wed, 12 Apr 1995 18:49:00 -0500 Date: Wed, 12 Apr 1995 18:49:00 -0500 From: jperez@lensis3.lsi.us.es (Jos Antonio Prez Castellanos) Message-Id: <9504122349.AA14161@lensis3.lsi.us.es> To: info-sr@cs.arizona.edu INFO-SR Mailing List Hello. Last week I mailed you for being included in the INFO-SR Mailing List. My e-mail has just been changed, so please, you could delete my old e-mail and insert my new one: my old e-mail: jperez@lensis1.fie.us.es my new e-mail: jperez@lensis1.lsi.us.e Sincerelly Yours, Jose Antonio Perez Castellanos Facultad Informatica y Estadistica Sevilla - ESPA%A jperez@lensis1.lsi.us.e From mcnamee@dromio.ecs.csus.edu Mon Apr 17 20:07:42 1995 Received: from dromio.ecs.csus.edu by optima.cs.arizona.edu (5.65c/15) via SMTP id AA06653; Mon, 17 Apr 1995 20:05:21 MST Received: (from mcnamee@localhost) by dromio.ecs.csus.edu (8.6.10/8.6.9) id UAA10246; Mon, 17 Apr 1995 20:07:43 -0700 Date: Mon, 17 Apr 1995 20:07:42 -0700 (PDT) From: Carole McNamee To: info-sr@cs.arizona.edu Cc: crowa@dromio.ecs.csus.edu Subject: timing sr progs Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII We are timing some sr programs using /bin/time and csh time on both sparcs and HP machines. We are getting some results that are contrary to results obtained five years ago with a much older version of sr. The results for elapsed time seem quite reasonable but the user time figures seem to give the inverse of the expected result. Have you or anyone else observed this phenomenon and might you have an explanation for it??? We are quite puzzled! tks, Carole McNamee mcnamee@dromio.csus.edu From Tsomber@aol.com Tue Apr 25 12:37:49 1995 Received: from mail06.mail.aol.com by optima.cs.arizona.edu (5.65c/15) via SMTP id AA10523; Tue, 25 Apr 1995 09:37:51 MST Received: by mail06.mail.aol.com (1.37.109.11/16.2) id AA275797869; Tue, 25 Apr 1995 12:37:49 -0400 Date: Tue, 25 Apr 1995 12:37:49 -0400 From: Tsomber@aol.com Message-Id: <950425123745_95668901@aol.com> To: info-sr@cs.arizona.edu Subject: installing on linux Hi! I'd appreciate some feedback on the following problem I am having trying to install SR on a PC running Slackware linux. I am a student at California State Univ. Hayward and SR is being used in an Operating Systems class there taught by Profession G. Lippman. I installed the latest version of SR on a Pent. 16meg machine running Slackware. I performed the Configuration change mentioned in the file Systems, and performed the make, and make install proecess. No errors were seen at this time. I then ran the tester program -p -v quick /vm . This seemd to run fine but at the end the following message was seen before termination : 0 $RSHPath $HOST -n date >date.out ./sev: /usr/bin/rsh: no such directory exists expected 0, got 126 from $RSHPATH Now, I tried running some example programs from Chap01 examples. The 'Hello.SR' program worked fine, as did the 'Factorial.SR' . However, other programs in Chap01 and Chap02 would generate errors like "array bounds overfow.." AND more alarmingly "Parser error at ' Procedcure ' " . I know you are busy, but *any* help on this matter is greatly appreciated. I am supposed to demonstrate the SR installation process soon so it can be used in classes at CSUH. thomas smith From gmt Tue Apr 25 11:15:37 1995 Received: from hawk.CS.Arizona.EDU by optima.cs.arizona.edu (5.65c/15) via SMTP id AA17650; Tue, 25 Apr 1995 11:15:39 MST Received: by hawk.CS.Arizona.EDU; (5.65/1.1.8.2/08Nov94-0446PM) id AA12066; Tue, 25 Apr 1995 11:15:37 -0700 Date: Tue, 25 Apr 1995 11:15:37 -0700 From: Gregg Townsend Message-Id: <9504251815.AA12066@hawk.CS.Arizona.EDU> To: Tsomber@aol.com Subject: Re: installing on linux Cc: info-sr The first problem sounds like SR can't find the rsh (remote shell) program. Check your setting of RSHPATH in the Configuration file. I don't know about "Parser error" and "bounds overflow" messages. Those messages aren't part of SR. Perhaps another of our Linux users can shed some light there. Gregg Townsend / Computer Science Dept / Univ of Arizona / Tucson, AZ 85721 +1 520 621 4325 gmt@CS.Arizona.EDU 110 57 16 W / 32 13 45 N / +758m From pingu!edgar@tmpuhf.saar.de Tue May 2 12:10:23 1995 Received: from shug-internet.saar.de by optima.cs.arizona.edu (5.65c/15) via SMTP id AA10625; Tue, 2 May 1995 12:10:23 MST Received: from TMPuhf.Saar.DE (tmpuhf.saar.de [192.109.53.3]) by Shug-Internet.Saar.DE (8.6.8.1/8.5) with SMTP id VAA09702; Tue, 2 May 1995 21:10:07 +0200 Received: from pingu by TMPuhf.Saar.DE with uucp (Smail3.1.28.1 #1) id m0s6NKf-00020zC; Tue, 2 May 95 21:10 WET DST Received: by pingu.saar.de (Smail3.1.28.1 #4) id m0s6LOA-0002KoC; Tue, 2 May 95 19:05 MET DST Message-Id: Date: Tue, 2 May 95 19:05 MET DST From: edgar@pingu.saar.de (Edgar Greuter) To: info-sr@cs.arizona.edu Subject: Patch for .../srlatex/print.c Hi! The SR 2.3 version of srlatex has a small error. When a @-sign is output, a break was missing in the C-source, and so another ^-char is echoed behind it in the output. The following patch corrects the problem: --- srlatex/print.c.orig Sat Mar 6 00:57:37 1993 +++ srlatex/print.c Tue May 2 18:47:53 1995 @@ -275,6 +275,7 @@ case '@': mmode (); out ("\\char`\\@ "); + break; case '^': mmode (); out ("\\char`\\^ "); Thanks for your attention! - Edgar From gmt Tue May 2 12:31:37 1995 Received: from hawk.CS.Arizona.EDU by optima.cs.arizona.edu (5.65c/15) via SMTP id AA12095; Tue, 2 May 1995 12:31:40 MST Received: by hawk.CS.Arizona.EDU; (5.65/1.1.8.2/08Nov94-0446PM) id AA20945; Tue, 2 May 1995 12:31:37 -0700 Date: Tue, 2 May 1995 12:31:37 -0700 From: Gregg Townsend Message-Id: <9505021931.AA20945@hawk.CS.Arizona.EDU> To: edgar@pingu.saar.de Subject: Re: Patch for .../srlatex/print.c Cc: info-sr Thanks for the fix. I'll see that it's included in the next version. Gregg Townsend / Computer Science Dept / Univ of Arizona / Tucson, AZ 85721 +1 520 621 4325 gmt@CS.Arizona.EDU 110 57 16 W / 32 13 45 N / +758m From colin@dcs.st-and.ac.uk Mon Jun 5 10:20:44 1995 Received: from tamdhu.dcs.st-andrews.ac.uk (tamdhu.dcs.st-and.ac.uk) by optima.cs.arizona.edu (5.65c/15) via SMTP id AA26606; Mon, 5 Jun 1995 10:20:44 MST Received: by tamdhu.dcs.st-andrews.ac.uk (4.1/SMI-4.1) id AA16427; Mon, 5 Jun 95 18:21:37 BST Date: Mon, 5 Jun 95 18:21:37 BST From: colin@dcs.st-and.ac.uk (Colin Allison) Message-Id: <9506051721.AA16427@tamdhu.dcs.st-andrews.ac.uk> To: info-sr@cs.arizona.edu Subject: non-blocking read() ? I'm seeing some unexpected behaviour using the read() command in resource instances created on a different vm from the "main" part of a program - basically the read() never waits and in the example below the loop goes on indefinitely outputting the prompt. If the example is modified so the "remote" resource is created on the same vm() it behaves as expected. can anyone explain this? thanks, Colin Allison St Andrews ------------------------------------------------- resource remote() var answer: string[8] do true -> writes("Enter (o for OK, q for quit):") read (answer) if answer = "o" -> write("OK") [] answer = "q" -> write("quitting"); exit fi od end remote resource main() import remote var remote_resource: cap remote var host1: cap vm host1 := create vm() on "glenkinchie" remote_resource := create remote() on host1 end main From gmt Tue Jun 6 09:13:26 1995 Received: from hawk.CS.Arizona.EDU by optima.cs.arizona.edu (5.65c/15) via SMTP id AA00987; Tue, 6 Jun 1995 09:13:44 MST Received: by hawk.CS.Arizona.EDU; (5.65/1.1.8.2/08Nov94-0446PM) id AA29216; Tue, 6 Jun 1995 09:13:26 -0700 Date: Tue, 6 Jun 1995 09:13:26 -0700 From: Gregg Townsend Message-Id: <9506061613.AA29216@hawk.CS.Arizona.EDU> To: colin@dcs.st-and.ac.uk, info-sr@cs.arizona.edu Subject: Re: non-blocking read() ? Date: Mon, 5 Jun 95 18:21:37 BST From: colin@dcs.st-and.ac.uk (Colin Allison) I'm seeing some unexpected behaviour using the read() command in resource instances created on a different vm from the "main" part of a program - basically the read() never waits and in the example below the loop goes on indefinitely outputting the prompt. New virtual machines inherit only standard output and standard error from the original VM. In particular, standard input is connected to /dev/null. The read() is getting an immediate EOF, resulting in a loop. Gregg Townsend / Computer Science Dept / Univ of Arizona / Tucson, AZ 85721 +1 520 621 4325 gmt@CS.Arizona.EDU 110 57 16 W / 32 13 45 N / +758m From vanmaren@fast.cs.utah.edu Wed Jul 12 17:43:26 1995 Received: from cs.utah.edu by optima.cs.arizona.edu (5.65c/15) via SMTP id AA00122; Wed, 12 Jul 1995 16:43:34 MST Received: from fast.cs.utah.edu by cs.utah.edu (8.6.12/utah-2.21-cs) id RAA01004; Wed, 12 Jul 1995 17:43:26 -0600 Received: by fast.cs.utah.edu (8.6.10/utah-2.15-leaf) id RAA25778; Wed, 12 Jul 1995 17:43:26 -0600 Date: Wed, 12 Jul 1995 17:43:26 -0600 From: vanmaren@fast.cs.utah.edu (Kevin Van maren) Message-Id: <199507122343.RAA25778@fast.cs.utah.edu> To: info-sr@cs.arizona.edu I used SR last semester for a class at the University of Wyoming. Now that I am at Utah, I thought I'd install SR 2.3 on the HP RISC (800 or 735) machines that I have access to. Unfortunatly, it doesn't seem to like the installation process. It chokes on building the asm.o file. I am using gcc 2.5.8 (cc & as 2.5.2). The error message is asm.s:26: Fatal error:Not in a space. when I build that one file manually using make asm.o in the csw directory. Is this a problem with g-cc/g-as? I am wondering if someone has already figured out the fix for this. There are some other machines around here I could try, but they all use the gcc compiler. I also get *** Error code 1 (ignored) right after the cc of util.c and before the cc of main.c. (twice) I haven't subscribed tothe list yet, so I would appreciate it if you cound send a copy of the response directly to me. Thanks, Kevin Van Maren From gmt Thu Jul 13 10:35:29 1995 Received: from hawk.CS.Arizona.EDU by optima.cs.arizona.edu (5.65c/15) via SMTP id AA22185; Thu, 13 Jul 1995 10:35:30 MST Received: by hawk.CS.Arizona.EDU; (5.65/1.1.8.2/08Nov94-0446PM) id AA23701; Thu, 13 Jul 1995 10:35:29 -0700 Date: Thu, 13 Jul 1995 10:35:29 -0700 From: Gregg Townsend Message-Id: <9507131735.AA23701@hawk.CS.Arizona.EDU> To: vanmaren@fast.cs.utah.edu Subject: Re: errors building SR on HP using gcc Cc: info-sr Date: Wed, 12 Jul 1995 17:43:26 -0600 From: vanmaren@fast.cs.utah.edu (Kevin Van maren) ... I thought I'd install SR 2.3 on the HP RISC (800 or 735) machines that I have access to. Unfortunatly, it doesn't seem to like the installation process. It chokes on building the asm.o file. I am using gcc 2.5.8 (cc & as 2.5.2).... This is just a guess, but does gcc predefine the same symbols as the HP C compiler? In particular, SR (in the top-level arch.h file) depends on the predefined symbol "hp9000s800" to identify any HP-RISC architecture. Gregg Townsend / Computer Science Dept / Univ of Arizona / Tucson, AZ 85721 +1 520 621 4325 gmt@CS.Arizona.EDU 110 57 16 W / 32 13 45 N / +758m From shartley@king.mcs.drexel.edu Mon Jul 17 07:10:28 1995 Received: from mcs.drexel.edu (king.mcs.drexel.edu) by optima.cs.arizona.edu (5.65c/15) via SMTP id AA18341; Mon, 17 Jul 1995 04:10:16 MST Received: by mcs.drexel.edu (4.1/SMI-4.0) id AA08313; Mon, 17 Jul 95 07:10:28 EDT Date: Mon, 17 Jul 95 07:10:28 EDT From: shartley@king.mcs.drexel.edu (Stephen J. Hartley) Message-Id: <9507171110.AA08313@mcs.drexel.edu> To: info-sr@cs.arizona.edu Subject: The future of SR I am wondering what plans there are for SR's future by the developers, such as new features contemplated. Or maybe SR has stabilized and future releases will be of the bug fix and maintenance kind. Thanks. Steve Hartley Math and Computer Science Drexel University From shartley@king.mcs.drexel.edu Thu Jul 27 12:08:10 1995 Received: from mcs.drexel.edu (king.mcs.drexel.edu) by optima.cs.arizona.edu (5.65c/15) via SMTP id AA13350; Thu, 27 Jul 1995 09:07:52 MST Received: by mcs.drexel.edu (4.1/SMI-4.0) id AA13499; Thu, 27 Jul 95 12:08:10 EDT Date: Thu, 27 Jul 95 12:08:10 EDT From: shartley@king.mcs.drexel.edu (Stephen J. Hartley) Message-Id: <9507271608.AA13499@mcs.drexel.edu> To: info-sr@cs.arizona.edu Subject: book errata An errata Web page has been set up for the book "Operating Systems Programming: The SR Programming Language", Oxford University Press, 1995, ISBN 0-19-5095790. See http://www.mcs.drexel.edu/~shartley/BookOSusingSR/errata.html From s2157613@cse.unsw.edu.au Sun Aug 20 12:44:40 1995 Received: from albeniz.orchestra.cse.unsw.EDU.AU by optima.CS.Arizona.EDU (5.65c/15) via SMTP id AA02803; Sat, 19 Aug 1995 19:49:26 MST Return-Path: s2157613@cse.unsw.edu.au Received: From bribe With LocalMail ; Sun, 20 Aug 95 12:46:32 +1000 From: s2157613@cse.unsw.edu.au (Tuan Van Nguyen) To: info-sr@cs.arizona.edu Date: Sun, 20 Aug 95 12:44:40 +1000 Message-Id: <950820024440.23913@cse.unsw.edu.au> Subject: hello My name is TUAN , I am from N.S.W university . I wuold like to say hello to all of you . I have just learned on Concurrent about few months ago , but i am really interesting on that . when you receive this mail, please mail back to me . Thank you. TUAN. From farraj@sybase.com Fri Sep 22 09:49:30 1995 Resent-Date: Fri, 22 Sep 1995 09:49:30 -0700 (PDT) Resent-From: farraj@sybase.com Received: from Arizona.EDU (Penny.Telcom.Arizona.EDU) by optima.cs.arizona.edu (5.65c/15) via SMTP id AA11421; Fri, 22 Sep 1995 09:48:17 MST Received: from Arizona.EDU by Arizona.EDU (PMDF V5.0-5 #2381) id <01HVKM8SQWGY9D4EV9@Arizona.EDU> for info-sr@CS.Arizona.EDU; Fri, 22 Sep 1995 09:48:11 -0700 (MST) Received: from halon.sybase.com (192.138.151.33) by Arizona.EDU (PMDF V5.0-5 #2381) id <01HVKM8EIRRKBRB0NX@Arizona.EDU> for info-sr@Arizona.EDU; Fri, 22 Sep 1995 09:47:55 -0700 (MST) Received: from sybase.com (sybgate.sybase.com) by halon.sybase.com (5.x/SMI-SVR4/SybFW4.0) id AA16297; Fri, 22 Sep 1995 09:47:17 -0700 Received: from surf.sybgate.sybase.com by sybase.com (4.1/SMI-4.1/SybH3.4) id AA18313; Fri, 22 Sep 1995 09:49:31 -0700 (PDT) Received: by surf.sybgate.sybase.com (4.1/SMI-4.1/SybEC3.2) id AA08188; Fri, 22 Sep 1995 09:49:30 -0700 (PDT) Date: Fri, 22 Sep 1995 09:49:30 -0700 (PDT) From: farraj@sybase.com (Jack Farraj) Subject: Resent-To: info-sr@cs.arizona.edu To: info-sr@arizona.edu Resent-Message-Id: <01HVKM8SXLLW9D4EV9@Arizona.EDU> Message-Id: <9509221649.AA08188@surf.sybgate.sybase.com> X-Vms-To: IN%"info-sr@Arizona.EDU" Mime-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII Content-Transfer-Encoding: 7BIT From 75162.1362@compuserve.com Sat Sep 23 02:08:25 1995 Received: from arl-img-5.compuserve.com by optima.cs.arizona.edu (5.65c/15) via SMTP id AA26785; Fri, 22 Sep 1995 23:10:19 MST Received: by arl-img-5.compuserve.com (8.6.10/5.950515) id CAA02683; Sat, 23 Sep 1995 02:10:17 -0400 Date: 23 Sep 95 02:08:25 EDT From: Julie <75162.1362@compuserve.com> To: Info-SR Mailing List Subject: SR Mailing List Message-Id: <950923060824_75162.1362_FHI5-1@CompuServe.COM> Please put me on your mailing list for SR issues. If you have additional information on the SR Programming Language that will be helpful for a graduate course project or ideas for an internet project, please e-mail the information. My e-mail addresses: 75162.1362@compuserve.com dujulie@orion.etsu.edu Thank you very much. Julie Du From shartley@king.mcs.drexel.edu Fri Sep 29 13:24:25 1995 Received: from mcs.drexel.edu (king.mcs.drexel.edu) by optima.cs.arizona.edu (5.65c/15) via SMTP id AA07940; Fri, 29 Sep 1995 10:24:26 MST Received: by mcs.drexel.edu (4.1/SMI-4.0) id AA18787; Fri, 29 Sep 95 13:24:25 EDT Date: Fri, 29 Sep 95 13:24:25 EDT From: shartley@king.mcs.drexel.edu (Stephen J. Hartley) Message-Id: <9509291724.AA18787@mcs.drexel.edu> To: info-sr@cs.arizona.edu Subject: Using SR in OS classes Concurrent Programming for Operating Systems Classes Using SR For Web-browsable text, example programs, animations, and laboratory exercises about using the SR programming language in operating systems classes, see http://www.mcs.drexel.edu/~shartley/OSusingSR/index.html Table of Contents Introduction Processes The SR Programming Language Laboratory Race Conditions and Process Synchronization Laboratory Semaphores Laboratory Monitors Laboratory Message Passing Laboratory Rendezvous Laboratory This work is supported by the National Science Foundation, Division of Undergraduate Education, Course and Curriculum Development Program, DUE-9455307. Stephen J. Hartley, Assistant Professor Math and Computer Science, Drexel University, Philadelphia, PA 19104 office (215) 895-2678, fax (215) 895-1582/4999, e-mail shartley@mcs.drexel.edu WWW URL http://www.mcs.drexel.edu/~shartley From gmt Mon Oct 16 11:31:36 1995 Received: from hawk.CS.Arizona.EDU by optima.cs.arizona.edu (5.65c/15) via SMTP id AA13051; Mon, 16 Oct 1995 11:31:37 MST Received: by hawk.CS.Arizona.EDU; (5.65/1.1.8.2/08Nov94-0446PM) id AA23549; Mon, 16 Oct 1995 11:31:36 -0700 Date: Mon, 16 Oct 1995 11:31:36 -0700 From: Gregg Townsend Message-Id: <9510161831.AA23549@hawk.CS.Arizona.EDU> To: info-sr Subject: Building SR on Linux From benson@cs.ucdavis.edu Sun Oct 15 03:53:00 1995 [fowarded with permission] In recent versions of Slackware (A popular Linux distribution) SR fails to build with the following link error: ld: Output file requires shared library `libc.so.4' gcc: Internal compiler error: program ld got fatal signal 6 A fix I have used in the past is to remove "-g" from the definition of CFLAGS in the main Makefile. This allows SR to build, but without debugging support. For most users this is fine. A cleaner fix is to change the LIBM line in Configuration to: LIBM = -lm -lc This allows SR to be compiled on Linux with debugging support. Note that the debugging info is removed (stripped) from the executables when they are installed. It seems that when gcc is told to generate debugging information a dependency is created by the math library. I am guessing libm depends on something in libc which is needed by gcc to produce debugging info. This might not be a Slackware problem, but rather a gcc+linux problem. Sorry for the lack of a better explanation. I know this is a problem for Slackware 2.2 and 2.3. It might apply to other distributions as well. Perhaps this should be mentioned in the Systems file. I have run into at least two other people that have had this problem, I am sure there are others. Greg From steve@atlas.ex.ac.uk Tue Oct 31 14:37:29 1995 Received: from hermes (hermes.ex.ac.uk) by optima.cs.arizona.edu (5.65c/15) via SMTP id AA04668; Tue, 31 Oct 1995 07:38:40 MST Received: from atlas.dcs.exeter.ac.uk by hermes via SMTP (OAA19388); Tue, 31 Oct 1995 14:38:51 GMT From: steve@atlas.ex.ac.uk Received: from vega.dcs.exeter.ac.uk by atlas.dcs.exeter.ac.uk; Tue, 31 Oct 95 14:37:26 GMT Date: Tue, 31 Oct 95 14:37:29 GMT Message-Id: <15170.9510311437@vega.dcs.exeter.ac.uk> To: info-sr@cs.arizona.edu Subject: Signaling Disciplines Re: Signal and Exit in Monitors I wrote a small program in an attempt to illustrate the difference between the alternative signaling disciplines in monitors: file sigdis.m: _monitor(test) op req(id: int), wake() _body(test) _condvar(cond); _proc( req(id) ) printf("Process Pr%d enters at time %d\n", id, age()) nap(1000) printf("Process Pr%d queued at time %d\n", id, age()) _wait(cond) printf("Process Pr%d awakes at time %d\n", id, age()) _proc_end _proc( wake() ) printf("Process Sig enters at time %d\n", age()) _signal(cond) printf("Process Sig continues at time %d\n", age()) _proc_end _monitor_end file sigmain.sr: resource SigDis() import test process Pr (i := 1 to 3) nap(i * 1000) test.req(i) end process Sig nap(2500) test.wake() end end Using the m2sr preprocessor, I get the behaviour I would expect with signal and continue, signal and wait, and signal and urgent wait. However, the signal and exit option does not appear to be preemptive: 20 % m2sr -sx sigdis.m 21 % sr sigdis.sr sigmain.sr sigdis.sr: sigmain.sr: sigdis.sr: sigmain.sr: linking: 22 % a.out Process Pr1 enters at time 1042 Process Pr1 queued at time 2111 Process Pr2 enters at time 2168 Process Pr2 queued at time 3242 Process Sig enters at time 3297 Process Pr3 enters at time 3300 Process Pr3 queued at time 4312 Process Pr1 awakes at time 4317 i.e. Pr3 is allowed to enter the monitor even though Pr1 was the process awakened by the signal. Looking at the macro definitions, it seems that what has been implemented is the equivalent of: sem e := 1 sem c := 0 var nc:int := 0 monitor entry: P(e) wait(cond): nc++; V(e); P(c); P(e) signal(cond): if nc > 0 -> nc--; V(c) fi; V(e); return monitor exit: V(e) whereas for preemptive behaviour I would expect something like: sem e := 1 sem c := 0 var nc:int := 0 monitor entry: P(e) wait(cond): nc++; V(e); P(c) signal(cond): if nc > 0 -> nc--; V(c) [] nc = 0 -> V(e) fi; return monitor exit: V(e) Do you agree with this? If so, I think the following modification to m2sr.h is required: *** m2sr.old Mon Oct 30 09:47:53 1995 --- m2sr.h Mon Oct 30 09:52:04 1995 *************** *** 132,138 **** send x.m_wait_ranks(y); \ send m_mutex(); \ receive m_return_from_wait(); \ - receive m_mutex(); /* enclose in if statement in attempt to avoid unreachable code * messages that come in case _signal comes right before _proc_end. --- 132,137 ---- *************** *** 140,147 **** #define _signal(x) \ SLINE \ if true -> \ ! m_bozo := x.m_signal(); \ ! send m_mutex(); \ return; \ fi; --- 139,147 ---- #define _signal(x) \ SLINE \ if true -> \ ! if not x.m_signal() -> \ ! send m_mutex(); \ ! fi; \ return; \ fi; This gives the preemptive behaviour I would expect: 29 % m2sr -sx sigdis.m 30 % sr sigdis.sr sigmain.sr sigdis.sr: sigmain.sr: sigdis.sr: sigmain.sr: linking: 31 % a.out Process Pr1 enters at time 1035 Process Pr1 queued at time 2105 Process Pr2 enters at time 2163 Process Pr2 queued at time 3226 Process Sig enters at time 3231 Process Pr1 awakes at time 3234 Process Pr3 enters at time 3235 Process Pr3 queued at time 4246 Steve ------------------------------------------------------------------------------- Parallel Systems Research Laboratory | Stephen Turner Email: steve@dcs.exeter.ac.uk | Department of Computer Science WWW: http://www.dcs.exeter.ac.uk | University of Exeter Fax: +44 1392 264067 | Prince of Wales Road Tel: +44 1392 264048 | Exeter EX4 4PT England ------------------------------------------------------------------------------- From olsson@cs.ucdavis.edu Tue Oct 31 12:51:32 1995 Received: from toadflax.cs.ucdavis.edu by optima.cs.arizona.edu (5.65c/15) via SMTP id AA27320; Tue, 31 Oct 1995 13:51:36 MST Received: from ivy.cs.ucdavis.edu by toadflax.cs.ucdavis.edu (4.1/UCD.CS.2.6) id AA14156; Tue, 31 Oct 95 12:51:33 PST Received: by ivy.cs.ucdavis.edu (5.65/UCD.CS.2.6) id AA02865; Tue, 31 Oct 1995 12:51:32 -0800 Date: Tue, 31 Oct 1995 12:51:32 -0800 From: olsson@cs.ucdavis.edu (Ron Olsson) Message-Id: <9510312051.AA02865@ivy.cs.ucdavis.edu> To: steve@atlas.ex.ac.uk Cc: info-sr@cs.arizona.edu In-Reply-To: <15170.9510311437@vega.dcs.exeter.ac.uk> (steve@atlas.ex.ac.uk) Subject: Signaling Disciplines I agree with your analysis and your patch looks good. Thanks. Sorry for the no-brainer. From bartis@cnu.edu Thu Nov 2 12:14:19 1995 Received: from drake.cnu.edu by optima.cs.arizona.edu (5.65c/15) via SMTP id AA24074; Thu, 2 Nov 1995 10:15:54 MST Received: from drake.cnu.edu (localhost [127.0.0.1]) by drake.cnu.edu (8.6.12/8.6.9) with SMTP id MAA05605 for ; Thu, 2 Nov 1995 12:14:19 -0500 Message-Id: <199511021714.MAA05605@drake.cnu.edu> Date: Thu, 02 Nov 95 12:14:19 -0500 Sender: bartis@cnu.edu From: Brian Artis X-Mailer: Mozilla 1.12 (X11; I; SunOS 5.4 sun4m) Mime-Version: 1.0 To: info-sr@cs.arizona.edu Subject: Buddy System X-Url: http://www.mcs.drexel.edu/~shartley/OSusingSR/index.html Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=us-ascii Are there any examples of the Buddy System available in SR code? This seems like it would be a good topic for an Operating Systems class. bla From greg Thu Nov 2 13:27:28 1995 Received: from paloverde.CS.Arizona.EDU by optima.cs.arizona.edu (5.65c/15) via SMTP id AA03175; Thu, 2 Nov 1995 13:27:34 MST Received: by paloverde.CS.Arizona.EDU; (5.65/1.1.8.2/08Nov94-0446PM) id AA27452; Thu, 2 Nov 1995 13:27:28 -0700 Date: Thu, 2 Nov 1995 13:27:28 -0700 From: Greg Andrews Message-Id: <9511022027.AA27452@paloverde.CS.Arizona.EDU> To: bartis@cnu.edu, info-sr@cs.arizona.edu Subject: Re: Buddy System Sorry, but there are no examples of which I am aware. Perhaps someone will write one? It would be a good topic. Greg Andrews From lschwe@ns Fri Nov 3 21:01:02 1995 Received: from tiny.sprintlink.net by optima.cs.arizona.edu (5.65c/15) via SMTP id AA01915; Fri, 3 Nov 1995 20:02:28 MST Received: from ns.cencom.net (ns.cencom.net [198.69.20.2]) by tiny.sprintlink.net (8.6.9/8.6.9) with SMTP id WAA16512 for ; Fri, 3 Nov 1995 22:02:03 -0500 Received: by ns.cencom.net (5.x/SMI-SVR4) id AA02173; Fri, 3 Nov 1995 21:01:02 -0600 Date: Fri, 3 Nov 1995 21:01:02 -0600 (CST) From: Larry Schweizer DIALUP Subject: Installing SR To: info-sr@cs.arizona.edu Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII I am running Linux 1.2.8 on a Pentium 60 PC. I get the following error(s) when I run 'make all': make[1]: Entering directory /sr/sr .... expect 4 shift/reduction conflicts : yacc -d grammar.y make: /usr/bin/make: cannot execute binary file make[1]:***[grammar.c] Error 126 make[1]: Leaving directory /sr/sr make:***[all] Error 2 Please help! Also, any suggestions on the proper values for the Configuration file would be gratefully appreciated. I wasn't at all clear on some of the necessary values for my Linux/Pentium box. Thanks-a-many! Larry Schweizer lschwe@ns.cencom.net From gmt Mon Nov 6 09:20:07 1995 Received: from hawk.CS.Arizona.EDU by optima.cs.arizona.edu (5.65c/15) via SMTP id AA09210; Mon, 6 Nov 1995 09:20:07 MST Received: by hawk.CS.Arizona.EDU; (5.65/1.1.8.2/08Nov94-0446PM) id AA24047; Mon, 6 Nov 1995 09:20:07 -0700 Date: Mon, 6 Nov 1995 09:20:07 -0700 From: Gregg Townsend Message-Id: <9511061620.AA24047@hawk.CS.Arizona.EDU> To: lschwe@ns.cencom.net Subject: Re: Installing SR Cc: info-sr From: Larry Schweizer DIALUP I am running Linux 1.2.8 on a Pentium 60 PC. I get the following error(s) when I run 'make all': make[1]: Entering directory /sr/sr .... expect 4 shift/reduction conflicts : yacc -d grammar.y make: /usr/bin/make: cannot execute binary file make[1]:***[grammar.c] Error 126 make[1]: Leaving directory /sr/sr make:***[all] Error 2 My guess is that for some reason yacc isn't executable. It appears to have something do to with how your particular system is configured / installed. Also, any suggestions on the proper values for the Configuration file would be gratefully appreciated. I wasn't at all clear on some of the necessary values for my Linux/Pentium box. Our latest info is that you should leave everything at the settings distributed except for three items in the Configuration file: CCPATH = /usr/bin/cc LIBL = LIBM = -lm -lc Gregg Townsend / gmt@CS.Arizona.EDU / +1 520 621 4325 / 32 13 45N 110 57 16W Computer Science / Univ of Arizona / 1040 E 4th St / Tucson AZ 85721-0077 From lschwe@ns Tue Nov 7 23:36:22 1995 Received: from tiny.sprintlink.net by optima.cs.arizona.edu (5.65c/15) via SMTP id AA16342; Tue, 7 Nov 1995 22:37:22 MST Received: from ns.cencom.net (ns.cencom.net [198.69.20.2]) by tiny.sprintlink.net (8.6.9/8.6.9) with SMTP id AAA16180 for ; Wed, 8 Nov 1995 00:37:21 -0500 Received: by ns.cencom.net (5.x/SMI-SVR4) id AA09903; Tue, 7 Nov 1995 23:36:22 -0600 Date: Tue, 7 Nov 1995 23:36:22 -0600 (CST) From: Larry Schweizer DIALUP Subject: SR install To: info-sr@cs.arizona.edu Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Many thanks to William R. Patters, Jose A Perez Castellone and Greg Townsend for responding to my request for help. I hope I can return the favor someday. However, that may be a while because ...... I have a new problem! But first, the good news: I found out that I'm actually running GNU Bison instead of Yacc, and that Bison has different output filenames than Yacc. So 'Make' was looking for a different filename. Setting YACC=bison and YFLAGS=-y in the /sr/sr Makefile makes Bison's output filenames the same as Yacc's. I also set my Configuration file to the values suggested by all of you. So now when I run 'make all', it gets past the original problem to this new one: gcc -g -c errors.c -o errors.o errors.c: In function 'yyerror': errors.c:58: conflicting types for 'yylval' tokens.h:137: previous declaration of 'yylval" make[1]:***{errors.o} Error 1 make[1]: Leaving directory '/usr/sr/sr' make: ***{all} Error 2 Once again, any help would be truly appreciated! Thanx! Larry Schweizer lschwe@ns.cencom.net From lschwe@ns Sat Nov 11 19:08:15 1995 Received: from tiny.sprintlink.net by optima.cs.arizona.edu (5.65c/15) via SMTP id AA12572; Sat, 11 Nov 1995 18:09:14 MST Received: from ns.cencom.net (ns.cencom.net [198.69.20.2]) by tiny.sprintlink.net (8.6.9/8.6.9) with SMTP id UAA00526 for ; Sat, 11 Nov 1995 20:09:13 -0500 Received: by ns.cencom.net (5.x/SMI-SVR4) id AA20571; Sat, 11 Nov 1995 19:08:16 -0600 Date: Sat, 11 Nov 1995 19:08:15 -0600 (CST) From: Larry Schweizer DIALUP Subject: SR install To: info-sr@cs.arizona.edu Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Thanks once more to Jose A Perez Castellone and Greg Townsend for your help. I installed Yacc, ran 'make', and everything went fine. Obviously, Bison and Yacc are NOT compatible in this circumstance! One more question, though. I requested the following in the Configuration file: SRSRC = /usr/local/src/sr SRCMD = /usr/local/bin SRLIB = /usr/local/lib/sr MAN1 = /usr/local/man/man1 MAN3 = /usr/local/man/man3 MAN5 = /sur/local/man/man5 When 'make' completed, I went to those directories and they were empty. Am I supposed to move the various files to their respective directories manually? Otherwise everything stays in the /usr/local/src/sr directory and various subdirectories. Gratefully, Larry Schweizer lschwe@ns.cencom.net From jperez@arturo.lsi.us.es Mon Nov 13 13:09:19 1995 Received: from opalo.us.es by optima.cs.arizona.edu (5.65c/15) via SMTP id AA15795; Mon, 13 Nov 1995 05:09:18 MST Received: from arturo.lsi.us.es by opalo.us.es (AIX 3.2/UCB 5.64/4.03) id AA21076; Mon, 13 Nov 1995 13:07:02 +0100 Received: from selene.lsi.us.es by lsi.us.es (5.x/SMI-SVR4) id AA01831; Mon, 13 Nov 1995 13:07:33 +0100 Received: by selene.lsi.us.es (5.x/SMI-SVR4) id AA21631; Mon, 13 Nov 1995 13:09:20 +0100 From: jperez@arturo.lsi.us.es (JOSE A. PEREZ CASTELLANOS) Message-Id: <9511131209.AA21631@selene.lsi.us.es> Subject: INSTALLING SR To: info-sr@cs.arizona.edu Date: Mon, 13 Nov 1995 13:09:19 +0100 (MET) X-Mailer: ELM [version 2.4 PL23] Content-Type: text No, of course you haven't to copy files manually. Once SR is built with 'make all', you must install each file to its location by typing 'make install'. When this is done, and if the location where SR is placed is within the PATH environment, SR is ready to run. Good luck. Jose Antonio Perez Castellanos Dep. Lenguajes y Sistemas Informaticos Universidad de Sevilla Spain jperez@lsi.us.es From lschwe@ns Wed Nov 15 21:32:55 1995 Received: from tiny.sprintlink.net by optima.cs.arizona.edu (5.65c/15) via SMTP id AA00350; Wed, 15 Nov 1995 20:33:52 MST Received: from ns.cencom.net (ns.cencom.net [198.69.20.2]) by tiny.sprintlink.net (8.6.9/8.6.9) with SMTP id WAA18935 for ; Wed, 15 Nov 1995 22:33:49 -0500 Received: by ns.cencom.net (5.x/SMI-SVR4) id AA29543; Wed, 15 Nov 1995 21:32:56 -0600 Date: Wed, 15 Nov 1995 21:32:55 -0600 (CST) From: Larry Schweizer DIALUP Subject: SR install To: info-sr@cs.arizona.edu Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Just wanted to say thanks to Ron Olsson and Jose Antonio Perez Castellanos for their helpful responses. Yes, I forgot to run 'make install'! A big DUH for me. However, I did successfully install gcc-2.7.0 tonight, and Linux-1.2.13 last night, so I'm on a roll! Thanks again! Yours truly, Larry Schweizer lschwe@ns.cencom.net From cliff@convx1.ccit.arizona.edu Mon Dec 11 14:51:19 1995 Received: from optima.CS.Arizona.EDU by cheltenham.CS.Arizona.EDU; Mon, 11 Dec 1995 14:52:08 MST Received: from Arizona.EDU (Penny.Telcom.Arizona.EDU) by optima.cs.arizona.edu (5.65c/15) via SMTP id AA11674; Mon, 11 Dec 1995 14:52:08 MST Return-Path: cliff@Arizona.EDU Received: from convx1.ccit.arizona.edu by Arizona.EDU (PMDF V5.0-5 #2381) id <01HYOO67K5HS95MUA5@Arizona.EDU> for info-sr@cs.arizona.edu; Mon, 11 Dec 1995 14:52:05 -0700 (MST) Received: from localhost by convx1.ccit.arizona.edu; Mon, 11 Dec 1995 14:51:19 -0700 Date: Mon, 11 Dec 1995 14:51:19 -0700 From: cliff@convx1.ccit.arizona.edu (Clifford Hathaway) Subject: test from convx1 through che To: info-sr@cs.arizona.edu Message-Id: <199512112151.OAA21287@convx1.ccit.arizona.edu> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII Content-Transfer-Encoding: 7BIT From cliff Mon Dec 11 15:07:01 1995 Date: Mon, 11 Dec 1995 15:07:01 MST From: "Cliff Hathaway" Message-Id: <199512112207.AA02435@cheltenham.cs.arizona.edu> Received: by cheltenham.cs.arizona.edu; Mon, 11 Dec 1995 15:07:01 MST To: info-sr Subject: test from che through che From gmt Fri Dec 15 16:35:51 1995 Received: from hawk.CS.Arizona.EDU by cheltenham.CS.Arizona.EDU; Fri, 15 Dec 1995 16:35:52 MST Received: by hawk.CS.Arizona.EDU; (5.65/1.1.8.2/08Nov94-0446PM) id AA10886; Fri, 15 Dec 1995 16:35:51 -0700 Date: Fri, 15 Dec 1995 16:35:51 -0700 From: Gregg Townsend Message-Id: <9512152335.AA10886@hawk.CS.Arizona.EDU> To: info-sr Subject: Maintenance release of SR SR version 2.3.1 is now available for FTP from the /sr directory on ftp.cs.arizona.edu. This version contains an accumulation of portability fixes, notably for AIX 4.1, Linux, and FreeBSD. If you had trouble building SR before, this release may help. The number of bug fixes is small. Unless you have a specific reason, there's no need to upgrade a copy of SR that's currently working. Gregg Townsend / gmt@CS.Arizona.EDU / +1 520 621 4325 / 32 13 45N 110 57 16W Computer Science / Univ of Arizona / 1040 E 4th St / Tucson AZ 85721-0077