From katia_dilkina at verizon.net Mon Apr 11 23:41:58 2005 From: katia_dilkina at verizon.net (Shawn Betts) Date: Mon, 11 Apr 2005 19:41:58 -0400 Subject: [movitz-devel] test check? Message-ID: <86br8kq3sp.fsf@monk.lamenet> cheeeeck test! is this mailing list alive? -Shawn From sabetts at vcn.bc.ca Tue Apr 12 06:01:38 2005 From: sabetts at vcn.bc.ca (Shawn Betts) Date: Tue, 12 Apr 2005 02:01:38 -0400 Subject: [movitz-devel] compile error in CVS Message-ID: <86zmw4jzy5.fsf@monk.lamenet> Hi folks, I'm getting a compile error in the latest CVS: [1]> (load "load") ... [2]> (movitz:create-image) ;; NIL value: #x65. WARNING: Compiler-macro <= defined after being called as function (first in "losp/muerte/typep.lisp") ;; While Movitz compiling LDB%BYTE in losp/muerte/integers.lisp: ** - Continuable Error Label NIL multiply defined. ... Am I doing something wrong or is this an error in CVS? I've tried clisp, cmucl, and sbcl all with the same result. -Shawn From spambait at CloudDancer.com Tue Apr 12 05:35:12 2005 From: spambait at CloudDancer.com (GP lisper) Date: Mon, 11 Apr 2005 22:35:12 -0700 Subject: [movitz-devel] Re: test check? References: <86br8kq3sp.fsf@monk.lamenet> Message-ID: On Mon, 11 Apr 2005 19:41:58 -0400, wrote: > cheeeeck test! > > is this mailing list alive? > A good question. I have a VMware setup with DRDos set aside to try Movitz out. I was expecting the mailing list to prod me into finishing that project. -- Everyman has three hearts; one to show the world, one to show friends, and one only he knows. From jdz at dir.lv Tue Apr 12 06:16:32 2005 From: jdz at dir.lv (Janis Dzerins) Date: Tue, 12 Apr 2005 09:16:32 +0300 Subject: [movitz-devel] test check? In-Reply-To: <86br8kq3sp.fsf@monk.lamenet> References: <86br8kq3sp.fsf@monk.lamenet> Message-ID: <1113286592.8405.0.camel@localhost> On Mon, 2005-04-11 at 19:41 -0400, Shawn Betts wrote: > cheeeeck test! > > is this mailing list alive? I received the message successfully... -- Janis Dzerins If million people say a stupid thing, it's still a stupid thing. From ffjeld at common-lisp.net Tue Apr 12 09:34:10 2005 From: ffjeld at common-lisp.net (Frode Vatvedt Fjeld) Date: Tue, 12 Apr 2005 11:34:10 +0200 Subject: [movitz-devel] Re: compile error in CVS References: <86zmw4jzy5.fsf@monk.lamenet> Message-ID: <2hbr8kbap9.fsf@vserver.cs.uit.no> Shawn Betts writes: > Am I doing something wrong or is this an error in CVS? I've tried > clisp, cmucl, and sbcl all with the same result. Hi, I'm actually unable to reproduce the error your report. I've tried cmucl on a fresh check-out from CVS, and it seems to work fine. The error might suggest a problem with the assembler, are you sure your ia-x86 directory is up to date? -- Frode Vatvedt Fjeld From larsnostdal at gmail.com Tue Apr 12 10:03:05 2005 From: larsnostdal at gmail.com (=?ISO-8859-1?Q?Lars_Rune_N=F8stdal?=) Date: Tue, 12 Apr 2005 12:03:05 +0200 Subject: [movitz-devel] test check? In-Reply-To: <1113286592.8405.0.camel@localhost> References: <86br8kq3sp.fsf@monk.lamenet> <1113286592.8405.0.camel@localhost> Message-ID: i'm lurking here .. :] On Apr 12, 2005 8:16 AM, Janis Dzerins wrote: > On Mon, 2005-04-11 at 19:41 -0400, Shawn Betts wrote: > > cheeeeck test! > > > > is this mailing list alive? > > I received the message successfully... > > -- > Janis Dzerins > > If million people say a stupid thing, it's still a stupid thing. > > > _______________________________________________ > movitz-devel site list > movitz-devel at common-lisp.net > http://common-lisp.net/mailman/listinfo/movitz-devel > -- Mvh, Lars Rune N?stdal http://lars.nostdal.org/ From sabetts at vcn.bc.ca Wed Apr 13 01:46:51 2005 From: sabetts at vcn.bc.ca (Shawn Betts) Date: Tue, 12 Apr 2005 21:46:51 -0400 Subject: [movitz-devel] Re: compile error in CVS In-Reply-To: <2hbr8kbap9.fsf@vserver.cs.uit.no> References: <86zmw4jzy5.fsf@monk.lamenet> <2hbr8kbap9.fsf@vserver.cs.uit.no> Message-ID: <868y3njvn8.fsf@monk.lamenet> Frode Vatvedt Fjeld writes: > Shawn Betts writes: > > > Am I doing something wrong or is this an error in CVS? I've tried > > clisp, cmucl, and sbcl all with the same result. > > Hi, I'm actually unable to reproduce the error your report. I've tried > cmucl on a fresh check-out from CVS, and it seems to work fine. The > error might suggest a problem with the assembler, are you sure your > ia-x86 directory is up to date? Hi Frode, Yeah, I updated an old cvs copy I had from ages ago and didn't realized there were 3 repositories to update. After updating it worked! What is the reason for splitting the project like this? Wouldn't it be just as easy to have the 3 projects as directories in the movitz repository? It would have helped me anyway :)...though I'm not suggesting you change your setup because of incompetent users. -Shawn From ffjeld at common-lisp.net Wed Apr 13 06:41:40 2005 From: ffjeld at common-lisp.net (Frode Vatvedt Fjeld) Date: Wed, 13 Apr 2005 08:41:40 +0200 Subject: [movitz-devel] Re: compile error in CVS References: <86zmw4jzy5.fsf@monk.lamenet> <2hbr8kbap9.fsf@vserver.cs.uit.no> <868y3njvn8.fsf@monk.lamenet> Message-ID: <2hy8bn9o0r.fsf@vserver.cs.uit.no> Shawn Betts writes: > What is the reason for splitting the project like this? The ia-x86 package was started before movitz, and was really supposed mostly to be primarily a disassembler etc. for dynamic patching and re-writing of machine code. This is partly why it's quite inefficient when it comes to assembling output from the movitz compiler (I suspect the assembler consumes a considerable portion of the CPU power during movitz compiles--if anyone would be interested in a relatively simple project that'd speed up and simplify movitz compilation significantly, a new assembler would be the way to go). The binary-types package was also developed independently of movitz, and in fact for the same original project as ia-x86---it's no accident that binary-types/example.lisp is an ELF parser and as such fits neatly together with a disassembler. -- Frode Vatvedt Fjeld From luke at synap.se Wed Apr 13 10:37:36 2005 From: luke at synap.se (Luke Gorrie) Date: 13 Apr 2005 12:37:36 +0200 Subject: [movitz-devel] Re: compile error in CVS References: <86zmw4jzy5.fsf@monk.lamenet> <2hbr8kbap9.fsf@vserver.cs.uit.no> <868y3njvn8.fsf@monk.lamenet> Message-ID: Shawn Betts writes: > What is the reason for splitting the project like this? Wouldn't it be > just as easy to have the 3 projects as directories in the movitz > repository? It would have helped me anyway :)...though I'm not > suggesting you change your setup because of incompetent users. They are actually in one repository, you can grab the whole lot like this: mkdir movitz cd movitz export CVSROOT=:pserver:anonymous at common-lisp.net:/project/movitz/cvsroot cvs -d $CVSROOT co . From mulix at mulix.org Sat Apr 16 14:48:58 2005 From: mulix at mulix.org (Muli Ben-Yehuda) Date: Sat, 16 Apr 2005 17:48:58 +0300 Subject: [movitz-devel] [PATCH] should ll-testing.lisp provide ll-testing? Message-ID: <20050416144858.GA12567@granada.merseine.nu> Hi, I need this simple patch to compile CVS with cmucl 18e. Not sure it's the right thing to do, but it works for me... Error in function MOVITZ::SPECIAL-OPERATOR-REQUIRE: Compiling file "losp/ll-testing.lisp" didn't provide module :LL-TESTING. [Condition of type SIMPLE-ERROR] Index: losp/ll-testing.lisp =================================================================== RCS file: /project/movitz/cvsroot/movitz/losp/ll-testing.lisp,v retrieving revision 1.1 diff -u -r1.1 ll-testing.lisp --- losp/ll-testing.lisp 15 Apr 2005 07:03:09 -0000 1.1 +++ losp/ll-testing.lisp 16 Apr 2005 14:42:37 -0000 @@ -14,6 +14,8 @@ ;;;; ;;;;------------------------------------------------------------------ +(provide :ll-testing) + (in-package muerte) (defun dump-global-segment-table (&key table entries nofill) Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ From ffjeld at common-lisp.net Sat Apr 16 16:21:09 2005 From: ffjeld at common-lisp.net (Frode Vatvedt Fjeld) Date: Sat, 16 Apr 2005 18:21:09 +0200 Subject: [movitz-devel] Re: [PATCH] should ll-testing.lisp provide ll-testing? References: <20050416144858.GA12567@granada.merseine.nu> Message-ID: <2hfyxq8zgq.fsf@vserver.cs.uit.no> Muli Ben-Yehuda writes: > I need this simple patch to compile CVS with cmucl 18e. Not sure > it's the right thing to do, but it works for me... Right, I forgot that. I guess this also shows how poor the current file/module system is, but that how it is. I'll check in a patch shortly. -- Frode Vatvedt Fjeld From sabetts at vcn.bc.ca Sun Apr 17 15:23:48 2005 From: sabetts at vcn.bc.ca (Shawn Betts) Date: Sun, 17 Apr 2005 11:23:48 -0400 Subject: [movitz-devel] list-all-packages Message-ID: <86pswte8aj.fsf@monk.lamenet> I noticed movitz was missing list-all-packages. *** packages.lisp 28 Feb 2005 23:36:08 -0000 1.7 --- packages.lisp 17 Apr 2005 15:21:13 -0000 *************** *** 41,46 **** --- 41,53 ---- name (find-package-string (string name)))) + (defun list-all-packages () + (let (pkgs) + (maphash (lambda (k v) + (pushnew v pkgs)) + (get-global-property :packages)) + pkgs)) + (defun find-package-string (name &optional (start 0) (end (length name)) (key 'identity)) (values (gethash-string name start end (get-global-property :packages) From sabetts at vcn.bc.ca Sun Apr 17 15:57:25 2005 From: sabetts at vcn.bc.ca (Shawn Betts) Date: Sun, 17 Apr 2005 11:57:25 -0400 Subject: [movitz-devel] &REST variable OBJECTS must be dynamic-extent. Message-ID: <86ll7he6qi.fsf@monk.lamenet> Hello, It seems sometimes having a &rest variable in a function causes a strange error. Is this a known issue? It's kind of irritating :). I added this silly function: *** los0.lisp 15 Apr 2005 07:04:24 -0000 1.38 --- los0.lisp 17 Apr 2005 15:51:59 -0000 *************** *** 16,21 **** --- 16,25 ---- (provide :los0 :load-priority 0) (require :common-lisp) + + (defun silly-fn (&rest objects) + (mapcar #'+ objects)) + (require :x86-pc/all) (require :x86-pc/io-space) (require :x86-pc/ne2k) and I get this error: Error in function LISP::ASSERT-ERROR: &REST variable OBJECTS must be dynamic-extent. [Condition of type COMMON-LISP:SIMPLE-ERROR] Restarts: 0: [CONTINUE ] Retry assertion. 1: [RETRY-FUNOBJ ] Retry compilation of SILLY-FN. 2: [SKIP-TOPLEVEL-FORM] Skip the compilation of top-level form DEFUN. 3: [RETRY ] Restart Movitz compilation of #p"losp/los0.lisp". 4: Skip Movitz compilation of #p"losp/los0.lisp". 5: [ABORT ] Return to Top-Level. Debug (type H for help) (LISP::ASSERT-ERROR (COMMON-LISP:OR (COMMON-LISP:TYPEP MOVITZ::BINDING 'MOVITZ::HIDDEN-REST-FUNCTION-ARGUMENT) (MOVITZ::MOVITZ-ENV-GET # 'COMMON-LISP:DYNAMIC-EXTENT NIL #)) NIL "&REST variable ~S must be dynamic-extent." OBJECTS) From ffjeld at common-lisp.net Sun Apr 17 18:21:51 2005 From: ffjeld at common-lisp.net (Frode Vatvedt Fjeld) Date: Sun, 17 Apr 2005 20:21:51 +0200 Subject: [movitz-devel] Re: list-all-packages References: <86pswte8aj.fsf@monk.lamenet> Message-ID: <2hbr8d8ds0.fsf@vserver.cs.uit.no> Shawn Betts writes: > I noticed movitz was missing list-all-packages. I checked it in, thanks. -- Frode Vatvedt Fjeld From ffjeld at common-lisp.net Sun Apr 17 18:45:42 2005 From: ffjeld at common-lisp.net (Frode Vatvedt Fjeld) Date: Sun, 17 Apr 2005 20:45:42 +0200 Subject: [movitz-devel] Re: &REST variable OBJECTS must be dynamic-extent. References: <86ll7he6qi.fsf@monk.lamenet> Message-ID: <2h7jj18co9.fsf@vserver.cs.uit.no> Shawn Betts writes: > It seems sometimes having a &rest variable in a function causes a > strange error. Is this a known issue? It's kind of irritating :). Yes.. it's a remainder from before there was any dynamic memory allocation, and which I haven't come around to fix yet. Please just put in a declaration (and obey it..) until I can get it fixed. (defun silly-fn (&rest objects) (declare (dynamic-extent objects)) (mapcar #'+ objects)) should work. -- Frode Vatvedt Fjeld From ffjeld at common-lisp.net Sun Apr 17 22:31:34 2005 From: ffjeld at common-lisp.net (Frode Vatvedt Fjeld) Date: Mon, 18 Apr 2005 00:31:34 +0200 Subject: [movitz-devel] Re: &REST variable OBJECTS must be dynamic-extent. References: <86ll7he6qi.fsf@monk.lamenet> Message-ID: <2h3btp827t.fsf@vserver.cs.uit.no> Shawn Betts writes: > It seems sometimes having a &rest variable in a function causes a > strange error. Is this a known issue? It's kind of irritating :). Turned out this could be fixed without too much effort. What I've done is to have e.g. (defun silly-fn (&rest objects) (mapcar #'+ objects)) compile as (defun silly-fn (&rest objects) (declare (dynamic-extent objects)) (let ((objects (copy-list objects))) (mapcar #'+ objects))) I've not tested it extensively, but I hope it works for you. -- Frode Vatvedt Fjeld From ffjeld at common-lisp.net Mon Apr 18 06:57:10 2005 From: ffjeld at common-lisp.net (Frode Vatvedt Fjeld) Date: Mon, 18 Apr 2005 08:57:10 +0200 Subject: [movitz-devel] los0-gc.lisp is heavily debug-instrumented Message-ID: <2hy8bg7et5.fsf@vserver.cs.uit.no> I just thought I'd mention that for a couple of months now, the stop-and-copy in los0-gc.lisp has been instrumented with debugging code that is quite unreasonable for any sort of normal use. It's randomly migrating code-vectors to newspace, to mention one thing. While I do expect GC/memory-management eventually to be "plug-and-play" by default, that's not quite the case at the moment (although in principle just removing all that debugging stuff from the current stop-and-copy should work well. (defun simple-stop-and-copy) is an approximation to that). -- Frode Vatvedt Fjeld From mulix at mulix.org Sun Apr 24 09:05:54 2005 From: mulix at mulix.org (Muli Ben-Yehuda) Date: Sun, 24 Apr 2005 12:05:54 +0300 Subject: [movitz-devel] initial binary and memory layout Message-ID: <20050424090554.GC1325@granada.merseine.nu> Hi, Is there any documentation (except from the code...) on the los0-image layout and how and where it gets loaded into memory? Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: Digital signature URL: From ffjeld at common-lisp.net Sun Apr 24 10:34:06 2005 From: ffjeld at common-lisp.net (Frode Vatvedt Fjeld) Date: Sun, 24 Apr 2005 12:34:06 +0200 Subject: [movitz-devel] Re: initial binary and memory layout References: <20050424090554.GC1325@granada.merseine.nu> Message-ID: <2h64yc79b5.fsf@vserver.cs.uit.no> Muli Ben-Yehuda writes: > Is there any documentation (except from the code...) on the > los0-image layout and how and where it gets loaded into memory? Unfortunately I don't think it's much documented. However, there isn't too much to say about this. Most of the details are embedded in what happens in dump-image. Briefly, everything gets loaded (by the bootloader, i.e. that in bootloader.lisp or GRUB) at 1MB (physical) and up. Logical adresses start at 0, because the segment descriptors are given a 1MB base-address. (This segmentation scheme is completely optional btw, it's not any intrinsic property of movitz.) The first object (i.e whose logical location is 0) is the "global" run-time-context that is activated at boot-up (this happens in the code generated at dump-time by make-toplevel-funobj), which is called by the bootloader. That's it, really. After the global run-time-context, the memory is just a heap of (8-byte aligned) objects, in no particular order. -- Frode Vatvedt Fjeld