January 2005 - movitz-cvs@common-lisp.net

[movitz-cvs] CVS update: ia-x86/README
by ffjeld＠common-lisp.net 31 Jan '05

31 Jan '05

Update of /project/movitz/cvsroot/ia-x86 In directory common-lisp.net:/tmp/cvs-serv10289 Modified Files: README Log Message: *** empty log message *** Date: Mon Jan 31 14:58:48 2005 Author: ffjeld Index: ia-x86/README diff -u ia-x86/README:1.4 ia-x86/README:1.5 --- ia-x86/README:1.4 Sun Jan 30 10:59:58 2005 +++ ia-x86/README Mon Jan 31 14:58:48 2005 @@ -10,7 +10,7 @@ ## Author: Frode Vatvedt Fjeld <frodef(a)acm.org> ## Created at: Sat Jan 29 16:48:42 2005 ## -## $Id: README,v 1.4 2005/01/30 18:59:58 ffjeld Exp $ +## $Id: README,v 1.5 2005/01/31 22:58:48 ffjeld Exp $ ## ###################################################################### @@ -100,6 +100,13 @@ --------------------------------------------------------- +One of the original intentions of having proglists as the intermediate +representation of assembly programs, was to facilitate computational +reasoning about the program (hence the class hierarchy rooted at +"instruction"). I don't really know how well this idea worked out, I +suspect most people will just pass the proglists directly to +proglist-encode. + The function proglist-encode takes a proglist and produces machine-code in the form of e.g. a vector of 8-bit bytes. A typical @@ -137,3 +144,8 @@ (big-endian) binary encoding of the instruction, and the imaginary component is the number of bytes. + + + Disassembly + +[To be written.]

1 0

[movitz-cvs] CVS update: movitz/losp/muerte/scavenge.lisp
by ffjeld＠common-lisp.net 31 Jan '05

31 Jan '05

Update of /project/movitz/cvsroot/movitz/losp/muerte In directory common-lisp.net:/tmp/cvs-serv26913 Modified Files: scavenge.lisp Log Message: Do RET roll-forward in map-stack-dit. Date: Mon Jan 31 09:54:04 2005 Author: ffjeld Index: movitz/losp/muerte/scavenge.lisp diff -u movitz/losp/muerte/scavenge.lisp:1.44 movitz/losp/muerte/scavenge.lisp:1.45 --- movitz/losp/muerte/scavenge.lisp:1.44 Fri Jan 28 00:47:18 2005 +++ movitz/losp/muerte/scavenge.lisp Mon Jan 31 09:54:03 2005 @@ -10,7 +10,7 @@ ;;;; Author: Frode Vatvedt Fjeld <frodef(a)acm.org> ;;;; Created at: Mon Mar 29 14:54:08 2004 ;;;; -;;;; $Id: scavenge.lisp,v 1.44 2005/01/28 08:47:18 ffjeld Exp $ +;;;; $Id: scavenge.lisp,v 1.45 2005/01/31 17:54:03 ffjeld Exp $ ;;;; ;;;;------------------------------------------------------------------ @@ -168,7 +168,14 @@ (+ start-frame 1) map-region)) -(defun scavenge-find-code-vector (location casf-funobj esi &optional searchp) +(defun scavenge-find-pf (location) + (loop for (slot-name type) in (slot-value (class-of (current-run-time-context)) 'slot-map) + do (when (eq type 'code-vector-word) + (let ((code-vector (%run-time-context-slot slot-name))) + (when (location-in-object-p code-vector location) + (return code-vector)))))) + +(defun scavenge-find-code-vector (location casf-funobj esi &optional primitive-function-p) (flet ((match-funobj (funobj location) (cond ((not (typep funobj 'function)) @@ -201,7 +208,9 @@ (break "Unknown funobj/frame-type: ~S" casf-funobj)) ((location-in-object-p (%run-time-context-slot 'dynamic-jump-next) location) (%run-time-context-slot 'dynamic-jump-next)) - ((when searchp + ((when primitive-function-p + (scavenge-find-pf location) + #+ignore (%find-code-vector location))) (t (with-simple-restart (continue "Try to perform a code-vector-search.") (error "Unable to decode EIP #x~X funobj ~S, ESI ~S." @@ -216,6 +225,8 @@ (funcall function value frame))) (defun map-stack (function frame frame-bottom eip-index map-region) + "Scavenge the stack starting at location <frame> which ends at <frame-bottom> +and whose return instruction-pointer is at location eip-index." (with-funcallable (map-region) (loop ;; for frame = frame then (stack-frame-uplink frame) @@ -252,6 +263,7 @@ #'map-header-vals))) (defun map-stack-dit (function dit-frame frame-bottom eip-index map-region) + "Scavenge the stack, starting at a DIT stack-frame." (with-funcallable (map-region) (let* ((atomically (dit-frame-ref nil dit-frame :atomically-continuation :unsigned-byte32)) @@ -271,11 +283,10 @@ (= 0 (ldb (byte 2 0) atomically))) ;; Interrupt occurred inside an (non-pf) atomically, so none of the ;; GC-root registers are active. - #+ignore - (setf (dit-frame-ref nil dit-frame :eax) nil - (dit-frame-ref nil dit-frame :ebx) nil - (dit-frame-ref nil dit-frame :edx) nil - (dit-frame-ref nil dit-frame :esi) nil) + #+ignore (setf (dit-frame-ref nil dit-frame :eax) nil + (dit-frame-ref nil dit-frame :ebx) nil + (dit-frame-ref nil dit-frame :edx) nil + (dit-frame-ref nil dit-frame :esi) nil) (map-region function frame-bottom (+ dit-frame 1 (dit-frame-index :scratch1)))) (secondary-register-mode-p ;; EBX is also active @@ -294,12 +305,15 @@ ;; (multiple-value-bind (x0-location x0-tag) (stack-frame-ref nil next-frame-bottom 0 :signed-byte30+2) - ;; (warn "X0: ~S ~S" x0-location x0-tag) (cond ((and (or (eq x0-tag 1) ; 1 or 5? (eq x0-tag 3) ; 3 or 7? (and (oddp x0-location) (eq x0-tag 2))) ; 6? (location-in-object-p casf-code-vector x0-location)) + (when (= #xc3 (memref-int (stack-frame-ref nil next-eip-index 0 :unsigned-byte32) + :physicalp nil :type :unsigned-byte8)) + (setf (stack-frame-ref nil next-eip-index 0 :code-vector) + (symbol-value 'ret-trampoline))) (let* ((old-x0-code-vector (scavenge-find-code-vector (stack-frame-ref nil next-eip-index 0 :location) casf-funobj interrupted-esi t))) @@ -329,15 +343,5 @@ (when (not (eq old-code-vector new-code-vector)) (break "Code-vector for stack instruction-pointer moved. [index: ~S]" index)) new-code-vector)) - -(defun map-stack-flaccid-pointer (function index) - "If the pointed-to object is moved, reset pointer to NIL." - (let ((old (stack-frame-ref nil index 0))) - (cond - ((not (typep old 'pointer)) - old) - ((eq old (funcall function old index)) - old) - (t (setf (stack-frame-ref nil index 0) nil)))))

1 0

[movitz-cvs] CVS update: movitz/losp/muerte/functions.lisp
by ffjeld＠common-lisp.net 31 Jan '05

31 Jan '05

Update of /project/movitz/cvsroot/movitz/losp/muerte In directory common-lisp.net:/tmp/cvs-serv20276 Modified Files: functions.lisp Log Message: Fixed nasty bug in funobj-code-vector%{1,2,3} which caused completely bogus (integer) values to be returned for many funobjs. Date: Mon Jan 31 07:47:58 2005 Author: ffjeld Index: movitz/losp/muerte/functions.lisp diff -u movitz/losp/muerte/functions.lisp:1.26 movitz/losp/muerte/functions.lisp:1.27 --- movitz/losp/muerte/functions.lisp:1.26 Tue Jan 25 05:46:54 2005 +++ movitz/losp/muerte/functions.lisp Mon Jan 31 07:47:57 2005 @@ -10,7 +10,7 @@ ;;;; Author: Frode Vatvedt Fjeld <frodef(a)acm.org> ;;;; Created at: Tue Mar 12 22:58:54 2002 ;;;; -;;;; $Id: functions.lisp,v 1.26 2005/01/25 13:46:54 ffjeld Exp $ +;;;; $Id: functions.lisp,v 1.27 2005/01/31 15:47:57 ffjeld Exp $ ;;;; ;;;;------------------------------------------------------------------ @@ -133,20 +133,19 @@ retry (:movl (:esp) :ebp) - (:compile-form (:result-mode :ebx) funobj) - (:movl (:ebx (:offset movitz-funobj code-vector)) :eax) ; EAX = code-vector - (:locally (:movl :esp (:edi (:edi-offset :atomically-continuation)))) ;; Now inside atomically section. + (:compile-form (:result-mode :ebx) funobj) + (:movl (:ebx (:offset movitz-funobj code-vector)) :eax) ; EAX = code-vector (:movl (:ebx (:offset movitz-funobj code-vector%1op)) :ecx) - ;; determine if ECX is a pointer into EBX + ;; determine if ECX is a pointer into EAX (:subl :eax :ecx) (:jl 'return-vector) (:leal ((:ecx #.movitz:+movitz-fixnum-factor+)) :ecx) - (:cmpl (:ebx (:offset movitz-basic-vector num-elements)) :ecx) - (:jg 'return-vector) - ;; return the integer offset EAX-EBX + (:cmpl (:eax (:offset movitz-basic-vector num-elements -2)) :ecx) + (:jnc 'return-vector) + ;; return the integer offset (:movl :ecx :eax) (:jmp 'done) return-vector @@ -191,21 +190,19 @@ (:locally (:pushl (:edi (:edi-offset :atomically-continuation)))) (:pushl :ebp) retry - (:movl (:esp) :ebp) - (:compile-form (:result-mode :ebx) funobj) - (:movl (:ebx (:offset movitz-funobj code-vector)) :eax) ; EAX = code-vector - (:locally (:movl :esp (:edi (:edi-offset :atomically-continuation)))) ;; Now inside atomically section. - + + (:compile-form (:result-mode :ebx) funobj) + (:movl (:ebx (:offset movitz-funobj code-vector)) :eax) ; EAX = code-vector (:movl (:ebx (:offset movitz-funobj code-vector%2op)) :ecx) - ;; determine if ECX is a pointer into EBX + ;; determine if ECX is a pointer into EAX (:subl :eax :ecx) (:jl 'return-vector) (:leal ((:ecx #.movitz:+movitz-fixnum-factor+)) :ecx) - (:cmpl (:ebx (:offset movitz-basic-vector num-elements)) :ecx) - (:jg 'return-vector) + (:cmpl (:eax (:offset movitz-basic-vector num-elements -2)) :ecx) + (:jnc 'return-vector) ;; return the integer offset EAX-EBX (:movl :ecx :eax) (:jmp 'done) @@ -251,21 +248,19 @@ (:locally (:pushl (:edi (:edi-offset :atomically-continuation)))) (:pushl :ebp) retry - (:movl (:esp) :ebp) - (:compile-form (:result-mode :ebx) funobj) - (:movl (:ebx (:offset movitz-funobj code-vector)) :eax) ; EAX = code-vector - (:locally (:movl :esp (:edi (:edi-offset :atomically-continuation)))) ;; Now inside atomically section. + (:compile-form (:result-mode :ebx) funobj) + (:movl (:ebx (:offset movitz-funobj code-vector)) :eax) ; EAX = code-vector (:movl (:ebx (:offset movitz-funobj code-vector%3op)) :ecx) - ;; determine if ECX is a pointer into EBX + ;; determine if ECX is a pointer into EAX (:subl :eax :ecx) (:jl 'return-vector) (:leal ((:ecx #.movitz:+movitz-fixnum-factor+)) :ecx) - (:cmpl (:ebx (:offset movitz-basic-vector num-elements)) :ecx) - (:jg 'return-vector) + (:cmpl (:eax (:offset movitz-basic-vector num-elements -2)) :ecx) + (:jnc 'return-vector) ;; return the integer offset EAX-EBX (:movl :ecx :eax) (:jmp 'done)

1 0

[movitz-cvs] CVS update: movitz/compiler.lisp
by ffjeld＠common-lisp.net 31 Jan '05

31 Jan '05

Update of /project/movitz/cvsroot/movitz In directory common-lisp.net:/tmp/cvs-serv14515 Modified Files: compiler.lisp Log Message: Make the descision of whether a function should have a proper stack-frame be in accord with the stack discipline. That is, if ESI is written to, a stack-frame must be installed first. Date: Mon Jan 31 06:11:14 2005 Author: ffjeld Index: movitz/compiler.lisp diff -u movitz/compiler.lisp:1.132 movitz/compiler.lisp:1.133 --- movitz/compiler.lisp:1.132 Mon Jan 31 03:19:23 2005 +++ movitz/compiler.lisp Mon Jan 31 06:11:14 2005 @@ -8,7 +8,7 @@ ;;;; Created at: Wed Oct 25 12:30:49 2000 ;;;; Distribution: See the accompanying file COPYING. ;;;; -;;;; $Id: compiler.lisp,v 1.132 2005/01/31 11:19:23 ffjeld Exp $ +;;;; $Id: compiler.lisp,v 1.133 2005/01/31 14:11:14 ffjeld Exp $ ;;;; ;;;;------------------------------------------------------------------ @@ -34,7 +34,7 @@ "Make every compiled function check upon entry that the stack-pointer is within bounds. Costs 3 code-bytes and a few cycles.") -(defvar *compiler-allow-transients* nil +(defvar *compiler-allow-transients* t "Allow the compiler to keep function arguments solely in registers. Hurst debugging, improves performance.") @@ -796,7 +796,11 @@ (stack-frame-size (frame-map-size (frame-map function-env))) (use-stack-frame-p (or (plusp stack-frame-size) (tree-search resolved-code - '(:pushl :popl :ebp :esp :call :leave))))) + '(:pushl :popl :ebp :esp :call :leave)) + (some (lambda (x) + (and (not (equal '(:movl (:ebp -4) :esi) x)) + (tree-search x ':esi))) + resolved-code)))) (let* ((function-code (let* ((req-binding (movitz-binding (first (required-vars function-env)) function-env nil)) @@ -884,7 +888,12 @@ (resolved-code (finalize-code (extended-code function-env) funobj frame-map)) (stack-frame-size (frame-map-size (frame-map function-env))) (use-stack-frame-p (or (plusp stack-frame-size) - (tree-search resolved-code '(:ebp :esp :call :leave))))) + (tree-search resolved-code + '(:push :pop :ebp :esp :call :leave)) + (some (lambda (x) + (and (not (equal '(:movl (:ebp -4) :esi) x)) + (tree-search x ':esi))) + resolved-code)))) (multiple-value-bind (prelude-code have-normalized-ecx-p) (make-compiled-function-prelude stack-frame-size function-env use-stack-frame-p (need-normalized-ecx-p function-env) frame-map

1 0

[movitz-cvs] CVS update: movitz/compiler.lisp
by ffjeld＠common-lisp.net 31 Jan '05

31 Jan '05

Update of /project/movitz/cvsroot/movitz In directory common-lisp.net:/tmp/cvs-serv6094 Modified Files: compiler.lisp Log Message: Use muerte.cl:lambda, not cl:lambda. Date: Mon Jan 31 03:19:24 2005 Author: ffjeld Index: movitz/compiler.lisp diff -u movitz/compiler.lisp:1.131 movitz/compiler.lisp:1.132 --- movitz/compiler.lisp:1.131 Thu Jan 27 01:00:25 2005 +++ movitz/compiler.lisp Mon Jan 31 03:19:23 2005 @@ -8,7 +8,7 @@ ;;;; Created at: Wed Oct 25 12:30:49 2000 ;;;; Distribution: See the accompanying file COPYING. ;;;; -;;;; $Id: compiler.lisp,v 1.131 2005/01/27 09:00:25 ffjeld Exp $ +;;;; $Id: compiler.lisp,v 1.132 2005/01/31 11:19:23 ffjeld Exp $ ;;;; ;;;;------------------------------------------------------------------ @@ -696,8 +696,6 @@ function-binding-usage)))) (defun resolve-sub-functions (toplevel-funobj function-binding-usage) -;;; (multiple-value-bind (toplevel-funobj function-binding-usage) -;;; (resolve-borrowed-bindings toplevel-funobj) (assert (null (borrowed-bindings toplevel-funobj)) () "Can't deal with toplevel closures yet. Borrowed: ~S" (borrowed-bindings toplevel-funobj)) @@ -709,7 +707,7 @@ (case (car (movitz-funobj-name sub-funobj)) ((muerte.cl:lambda) (setf (movitz-funobj-name sub-funobj) - (list 'lambda + (list 'muerte.cl:lambda (movitz-funobj-name toplevel-funobj) (post-incf sub-funobj-index))))) (loop for borrowed-binding in (borrowed-bindings sub-funobj) @@ -733,12 +731,7 @@ (change-class function-binding 'closure-binding)) (t (change-class function-binding 'closure-binding) (setf (movitz-funobj-extent sub-funobj) - :indefinite-extent))) ; XXX - #+ignore - (warn "extent usage ~S: ~S => ~S" - usage - sub-funobj - (movitz-funobj-extent sub-funobj))))) + :indefinite-extent)))))) (loop for function-binding in function-binding-usage by #'cddr do (finalize-funobj (function-binding-funobj function-binding))) (finalize-funobj toplevel-funobj)) @@ -1017,12 +1010,7 @@ (make-movitz-vector (length code-vector) :fill-pointer code-length :element-type 'code - :initial-contents code-vector) - #+ignore - (make-movitz-code-vector code-vector - (slot-value funobj 'code-vector%1op) - (slot-value funobj 'code-vector%2op) - (slot-value funobj 'code-vector%3op))))) + :initial-contents code-vector)))) funobj) (defun check-locate-concistency (code-vector)

1 0

[movitz-cvs] CVS update: ia-x86/README
by ffjeld＠common-lisp.net 30 Jan '05

30 Jan '05

Update of /project/movitz/cvsroot/ia-x86 In directory common-lisp.net:/tmp/cvs-serv19851 Modified Files: README Log Message: *** empty log message *** Date: Sun Jan 30 10:59:58 2005 Author: ffjeld Index: ia-x86/README diff -u ia-x86/README:1.3 ia-x86/README:1.4 --- ia-x86/README:1.3 Sun Jan 30 06:54:33 2005 +++ ia-x86/README Sun Jan 30 10:59:58 2005 @@ -10,7 +10,7 @@ ## Author: Frode Vatvedt Fjeld <frodef(a)acm.org> ## Created at: Sat Jan 29 16:48:42 2005 ## -## $Id: README,v 1.3 2005/01/30 14:54:33 ffjeld Exp $ +## $Id: README,v 1.4 2005/01/30 18:59:58 ffjeld Exp $ ## ###################################################################### @@ -115,8 +115,10 @@ just pass zero). Finally, there is the proglist to be encoded to machine code. Additionally, the keyword argument :symtab-lookup can provide a function that proglist-encode will use to map (otherwise -undefined) labels to integers. The primary return value is the byte -vector, while the secondary return value is an (assoc) symbol-table: +undefined) labels to integers: The function should expect one symbol +argument and return a suitable integer. The primary return value is +the byte vector, while the secondary return value is an (assoc) +symbol-table: IA-X86(7): (proglist-encode :octet-vector :16-bit #x7c00 (read-proglist (mkasm16-bios-print))) @@ -131,7 +133,7 @@ IA-X86(8): (asm :movl :eax (:ebx 1)) => #c(#x894301 #x3) -The return value is a complex, whose real component is the binary -encoding of the instruction, and the imaginary component is the number -of bytes. +The return value is a complex, whose real component is the +(big-endian) binary encoding of the instruction, and the imaginary +component is the number of bytes.

1 0

[movitz-cvs] CVS update: ia-x86/README
by ffjeld＠common-lisp.net 30 Jan '05

30 Jan '05

Update of /project/movitz/cvsroot/ia-x86 In directory common-lisp.net:/tmp/cvs-serv6929 Modified Files: README Log Message: *** empty log message *** Date: Sun Jan 30 06:54:33 2005 Author: ffjeld Index: ia-x86/README diff -u ia-x86/README:1.2 ia-x86/README:1.3 --- ia-x86/README:1.2 Sun Jan 30 03:35:16 2005 +++ ia-x86/README Sun Jan 30 06:54:33 2005 @@ -10,7 +10,7 @@ ## Author: Frode Vatvedt Fjeld <frodef(a)acm.org> ## Created at: Sat Jan 29 16:48:42 2005 ## -## $Id: README,v 1.2 2005/01/30 11:35:16 ffjeld Exp $ +## $Id: README,v 1.3 2005/01/30 14:54:33 ffjeld Exp $ ## ###################################################################### @@ -67,6 +67,8 @@ This is a more-or-less exact description of the syntax: +--------------------------------------------------------- + program ::= (<sexpr>*) sexpr ::= ( <expr> ) | <label> | (% <inline-data> %) | (:include . <program>) @@ -96,5 +98,40 @@ prefix ::= <segment-override> | (:size <size>) +--------------------------------------------------------- + + The function proglist-encode takes a proglist and produces -machine-code in the form of e.g. a vector of 8-bit bytes. +machine-code in the form of e.g. a vector of 8-bit bytes. A typical +usage would be this: + + (proglist-encode :octet-vector :16-bit #x7c00 + (read-proglist (mkasm16-bios-print))) + +The first agument, :octet-vector, requests that the result be in the +form of a vector of octets. The second argument says to prodce machine +code for the 16-bit mode of x86. Next is the address to assume the +code is located in memory (if the program is position-independent, +just pass zero). Finally, there is the proglist to be encoded to +machine code. Additionally, the keyword argument :symtab-lookup can +provide a function that proglist-encode will use to map (otherwise +undefined) labels to integers. The primary return value is the byte +vector, while the secondary return value is an (assoc) symbol-table: + + IA-X86(7): (proglist-encode :octet-vector :16-bit #x7c00 + (read-proglist (mkasm16-bios-print))) + => #(#xf #xb6 #xc #x46 #xb4 #xe #xbb #x7 #x0 #xac ...) + => ((PRINT-LOOP . #x7c09)) + + +Another operator that can be useful for interactive and experimental +use, is the macro "asm" which encodes a single instruction. This can +be used e.g. to verify syntax or the encoded size of an instruction: + + IA-X86(8): (asm :movl :eax (:ebx 1)) + => #c(#x894301 #x3) + +The return value is a complex, whose real component is the binary +encoding of the instruction, and the imaginary component is the number +of bytes. +

1 0

[movitz-cvs] CVS update: ia-x86/README
by ffjeld＠common-lisp.net 30 Jan '05

30 Jan '05

Update of /project/movitz/cvsroot/ia-x86 In directory common-lisp.net:/tmp/cvs-serv29444 Modified Files: README Log Message: *** empty log message *** Date: Sun Jan 30 03:35:16 2005 Author: ffjeld Index: ia-x86/README diff -u ia-x86/README:1.1 ia-x86/README:1.2 --- ia-x86/README:1.1 Sat Jan 29 08:33:31 2005 +++ ia-x86/README Sun Jan 30 03:35:16 2005 @@ -10,7 +10,7 @@ ## Author: Frode Vatvedt Fjeld <frodef(a)acm.org> ## Created at: Sat Jan 29 16:48:42 2005 ## -## $Id: README,v 1.1 2005/01/29 16:33:31 ffjeld Exp $ +## $Id: README,v 1.2 2005/01/30 11:35:16 ffjeld Exp $ ## ###################################################################### @@ -22,7 +22,7 @@ the Movitz compiler. The following documents the most interesting API operators for this package. -A few terms requires explanation. A "proglist" is a representation of +A few terms requires explanation. A "program" is a representation of an assembly program that is reasonably convenient to work with for humans. This is a list, whose elements are either a symbol that represents a label, or a list that represents an instruction. The @@ -31,10 +31,70 @@ by 1" (more on this syntax below). However, instructions are also represented internally ia-x86 by instances of the various subclasses of the "instruction" standard-class, and lists of such objects are -also someplaces referred to as proglists. +referred to as "proglists". Assembly +The function read-proglist reads a program into proglist form, which +is typically the first step in producing machine code. This function +(and its helper functions, in read.lisp) defines the human-readable +syntax for assembly programs. This is an example program (or rather, a +lisp function that produces an assembly program): + + (defun mkasm16-bios-print () + "Print something to the terminal. [es:si] points to the text" + `((:movzxb (:si) :cx) + (:incw :si) + (:movb #xe :ah) + (:movw 7 :bx) + print-loop + (:lodsb) + (:int #x10) + (:loop 'print-loop) + (:ret))) + +I personally tend to use keywords for instruction names, although the +instructions are recognized by name rather than identity, so any +package will do. Labels, however, are recognized by identity. Label +references are on the form (quote <label>), so that the loop +instruction above will transfer control to the print-loop label two +instructions before it. Indirect memory references are writen by +placing the pointer inside parens, such as in the first movzxb above. +If the first element of an instructon is a list (rather than an +instruction name), this is interpreted as a list of instruction +prefixes (such as REP, LOCK, etc.). + +This is a more-or-less exact description of the syntax: + + program ::= (<sexpr>*) + + sexpr ::= ( <expr> ) | <label> | (% <inline-data> %) | (:include . <program>) + expr ::= <instro> | ( { <prefix> } ) <instro> + instro ::= <instruction> { <operand> } + + operand ::= <concrete> | <abstract> + + concrete ::= <immediate> | <register> | <indirect> + immediate ::= <number> + register ::= eax | ecx | ... + indirect ::= ( iexpr ) + iexpr ::= <address> + | (quote <label>) + | <segment> <address> + | pc+ <offset> + | pc <address> + | { (quote <label>) } <offset> <scaled-register> <register> + scaled-register ::= <register> | ( <register> <scale> ) + scale ::= 1 | 2 | 4 | 8 + address ::= <number> + offset ::= <signed-number> + + abstract ::= (quote <absexpr>) + absexpr ::= <label> | <number> | append-prg + append-prg ::= program + prefix ::= <segment-override> | (:size <size>) +The function proglist-encode takes a proglist and produces +machine-code in the form of e.g. a vector of 8-bit bytes.

1 0

[movitz-cvs] CVS update: ia-x86/README
by ffjeld＠common-lisp.net 29 Jan '05

29 Jan '05

Update of /project/movitz/cvsroot/ia-x86 In directory common-lisp.net:/tmp/cvs-serv3027 Added Files: README Log Message: First few lines. Date: Sat Jan 29 08:33:31 2005 Author: ffjeld

1 0

[movitz-cvs] CVS update: public_html/ChangeLog public_html/index.html
by ffjeld＠common-lisp.net 29 Jan '05

29 Jan '05

Update of /project/movitz/cvsroot/public_html In directory common-lisp.net:/tmp/cvs-serv17004 Modified Files: index.html Added Files: ChangeLog Log Message: moved this here. Date: Sat Jan 29 02:42:28 2005 Author: ffjeld Index: public_html/index.html diff -u public_html/index.html:1.23 public_html/index.html:1.24 --- public_html/index.html:1.23 Sat Jan 29 02:38:47 2005 +++ public_html/index.html Sat Jan 29 02:42:28 2005 @@ -39,7 +39,7 @@ with Movitz, and which is now fixed in QEMU's CVS. <p> For more news, see the <a - href="files/ChangeLog"><tt>ChangeLog</tt></a>. + href="ChangeLog"><tt>ChangeLog</tt></a>. <h3>Introduction</h3>

1 0