#!+sb-show *compiler-trace-output*
*last-source-context* *last-original-source*
*last-source-form* *last-format-string* *last-format-args*
- *last-message-count* *lexenv*))
+ *last-message-count* *lexenv* *fun-names-in-this-file*))
+
+;;; Whether call of a function which cannot be defined causes a full
+;;; warning.
+(defvar *flame-on-necessarily-undefined-function* nil)
(defvar *check-consistency* nil)
(defvar *all-components*)
;;; normally causes nested uses to be no-ops).
(defvar *in-compilation-unit* nil)
+;;; This lock is siezed in the same situation: the compiler is not
+;;; presently thread-safe
+(defvar *big-compiler-lock*
+ (sb!thread:make-mutex :name "big compiler lock"))
+
;;; Count of the number of compilation units dynamically enclosed by
;;; the current active WITH-COMPILATION-UNIT that were unwound out of.
(defvar *aborted-compilation-unit-count*)
`(%with-compilation-unit (lambda () ,@body) ,@options))
(defun %with-compilation-unit (fn &key override)
+ (declare (type function fn))
(let ((succeeded-p nil))
(if (and *in-compilation-unit* (not override))
;; Inside another WITH-COMPILATION-UNIT, a WITH-COMPILATION-UNIT is
;; ordinarily (unless OVERRIDE) basically a no-op.
(unwind-protect
- (multiple-value-prog1 (funcall fn) (setf succeeded-p t))
+ (multiple-value-prog1 (funcall fn) (setf succeeded-p t))
(unless succeeded-p
(incf *aborted-compilation-unit-count*)))
- ;; FIXME: Now *COMPILER-FOO-COUNT* stuff is bound in more than
- ;; one place. If we can get rid of the IR1 interpreter, this
- ;; should be easier to clean up.
(let ((*aborted-compilation-unit-count* 0)
(*compiler-error-count* 0)
(*compiler-warning-count* 0)
(*compiler-note-count* 0)
(*undefined-warnings* nil)
(*in-compilation-unit* t))
- (handler-bind ((parse-unknown-type
- (lambda (c)
- (note-undefined-reference
- (parse-unknown-type-specifier c)
- :type))))
- (unwind-protect
- (multiple-value-prog1 (funcall fn) (setf succeeded-p t))
- (unless succeeded-p
- (incf *aborted-compilation-unit-count*))
- (summarize-compilation-unit (not succeeded-p))))))))
+ (sb!thread:with-recursive-lock (*big-compiler-lock*)
+ (handler-bind ((parse-unknown-type
+ (lambda (c)
+ (note-undefined-reference
+ (parse-unknown-type-specifier c)
+ :type))))
+ (unwind-protect
+ (multiple-value-prog1 (funcall fn) (setf succeeded-p t))
+ (unless succeeded-p
+ (incf *aborted-compilation-unit-count*))
+ (summarize-compilation-unit (not succeeded-p)))))))))
+
+;;; Is FUN-NAME something that no conforming program can rely on
+;;; defining as a function?
+(defun fun-name-reserved-by-ansi-p (fun-name)
+ (eq (symbol-package (fun-name-block-name fun-name))
+ *cl-package*))
;;; This is to be called at the end of a compilation unit. It signals
;;; any residual warnings about unknown stuff, then prints the total
(warnings (undefined-warning-warnings undef))
(undefined-warning-count (undefined-warning-count undef)))
(dolist (*compiler-error-context* warnings)
- (compiler-style-warn "undefined ~(~A~): ~S" kind name))
+ (if #-sb-xc-host (and (eq kind :function)
+ (fun-name-reserved-by-ansi-p name)
+ *flame-on-necessarily-undefined-function*)
+ #+sb-xc-host nil
+ (case name
+ ((declare)
+ (compiler-warn
+ "~@<There is no function named ~S. References to ~S in ~
+ some contexts (like starts of blocks) have special ~
+ meaning, but here it would have to be a function, ~
+ and that shouldn't be right.~:@>"
+ name name))
+ (t
+ (compiler-warn
+ "~@<The ~(~A~) ~S is undefined, and its name is ~
+ reserved by ANSI CL so that even if it it were ~
+ defined later, the code doing so would not be ~
+ portable.~:@>"
+ kind name)))
+ (compiler-style-warn "undefined ~(~A~): ~S" kind name)))
(let ((warn-count (length warnings)))
(when (and warnings (> undefined-warning-count warn-count))
(let ((more (- undefined-warning-count warn-count)))
(compiler-style-warn
"~W more use~:P of undefined ~(~A~) ~S"
more kind name))))))
-
+
(dolist (kind '(:variable :function :type))
(let ((summary (mapcar #'undefined-warning-name
(remove kind undefs :test-not #'eq
(multiple-value-bind (code-length trace-table fixups)
(generate-code component)
+ #-sb-xc-host
(when *compiler-trace-output*
(format *compiler-trace-output*
"~|~%disassembly of code for ~S~2%" component)
(defun describe-component (component *standard-output*)
(declare (type component component))
(format t "~|~%;;;; component: ~S~2%" (component-name component))
- (print-blocks component)
+ (print-all-blocks component)
(values))
(defun describe-ir2-component (component *standard-output*)
;;; We parse declarations and then recursively process the body.
(defun process-toplevel-locally (body path compile-time-too &key vars funs)
(declare (list path))
- (multiple-value-bind (forms decls) (parse-body body nil)
- (let* ((*lexenv*
- (process-decls decls vars funs (make-continuation)))
+ (multiple-value-bind (forms decls)
+ (parse-body body :doc-string-allowed nil :toplevel t)
+ (let* ((*lexenv* (process-decls decls vars funs))
+ ;; FIXME: VALUES declaration
+ ;;
;; Binding *POLICY* is pretty much of a hack, since it
;; causes LOCALLY to "capture" enclosed proclamations. It
;; is necessary because CONVERT-AND-MAYBE-COMPILE uses the
(etypecase f
(clambda (list (lambda-component f)))
(optional-dispatch (let ((result nil))
- (labels ((frob (clambda)
- (pushnew (lambda-component clambda)
- result))
- (maybe-frob (maybe-clambda)
- (when maybe-clambda
- (frob maybe-clambda))))
- (mapc #'frob (optional-dispatch-entry-points f))
+ (flet ((maybe-frob (maybe-clambda)
+ (when (and maybe-clambda
+ (promise-ready-p maybe-clambda))
+ (pushnew (lambda-component
+ (force maybe-clambda))
+ result))))
+ (map nil #'maybe-frob (optional-dispatch-entry-points f))
(maybe-frob (optional-dispatch-more-entry f))
- (maybe-frob (optional-dispatch-main-entry f)))))))
+ (maybe-frob (optional-dispatch-main-entry f)))
+ result))))
(defun make-functional-from-toplevel-lambda (definition
&key
(setf (component-name component)
(debug-namify "~S initial component" name))
(setf (component-kind component) :initial)
- (let* ((locall-fun (ir1-convert-lambda
+ (let* ((locall-fun (ir1-convert-lambdalike
definition
:debug-name (debug-namify "top level local call ~S"
- name)))
+ name)
+ ;; KLUDGE: we do this so that we get to have
+ ;; nice debug returnness in functions defined
+ ;; from the REPL
+ :allow-debug-catch-tag t))
(fun (ir1-convert-lambda (make-xep-lambda-expression locall-fun)
:source-name (or name '.anonymous.)
:debug-name (unless name
(compile-component component-from-dfo)
(replace-toplevel-xeps component-from-dfo)))
- (prog1
- (let ((entry-table (etypecase *compile-object*
- (fasl-output (fasl-output-entry-table
- *compile-object*))
- (core-object (core-object-entry-table
- *compile-object*)))))
- (multiple-value-bind (result found-p)
- (gethash (leaf-info fun) entry-table)
- (aver found-p)
- result))
- ;; KLUDGE: This code duplicates some other code in this
- ;; file. In the great reorganzation, the flow of program logic
- ;; changed from the original CMUCL model, and that path (as of
- ;; sbcl-0.7.5 in SUB-COMPILE-FILE) was no longer followed for
- ;; CORE-OBJECTS, leading to BUG 156. This place is
- ;; transparently not the right one for this code, but I don't
- ;; have a clear enough overview of the compiler to know how to
- ;; rearrange it all so that this operation fits in nicely, and
- ;; it was blocking reimplementation of
- ;; (DECLAIM (INLINE FOO)) (MACROLET ((..)) (DEFUN FOO ...))
- ;;
- ;; FIXME: This KLUDGE doesn't solve all the problem in an
- ;; ideal way, as (1) definitions typed in at the REPL without
- ;; an INLINE declaration will give a NULL
- ;; FUNCTION-LAMBDA-EXPRESSION (allowable, but not ideal) and
- ;; (2) INLINE declarations will yield a
- ;; FUNCTION-LAMBDA-EXPRESSION headed by
- ;; SB-C:LAMBDA-WITH-LEXENV, even for null LEXENV.
- ;;
- ;; CSR, 2002-07-02
- (when (core-object-p *compile-object*)
- (fix-core-source-info *source-info* *compile-object*))
-
- (mapc #'clear-ir1-info components-from-dfo)
- (clear-stuff)))))
+ (let ((entry-table (etypecase *compile-object*
+ (fasl-output (fasl-output-entry-table
+ *compile-object*))
+ (core-object (core-object-entry-table
+ *compile-object*)))))
+ (multiple-value-bind (result found-p)
+ (gethash (leaf-info fun) entry-table)
+ (aver found-p)
+ (prog1
+ result
+ ;; KLUDGE: This code duplicates some other code in this
+ ;; file. In the great reorganzation, the flow of program
+ ;; logic changed from the original CMUCL model, and that
+ ;; path (as of sbcl-0.7.5 in SUB-COMPILE-FILE) was no
+ ;; longer followed for CORE-OBJECTS, leading to BUG
+ ;; 156. This place is transparently not the right one for
+ ;; this code, but I don't have a clear enough overview of
+ ;; the compiler to know how to rearrange it all so that
+ ;; this operation fits in nicely, and it was blocking
+ ;; reimplementation of (DECLAIM (INLINE FOO)) (MACROLET
+ ;; ((..)) (DEFUN FOO ...))
+ ;;
+ ;; FIXME: This KLUDGE doesn't solve all the problem in an
+ ;; ideal way, as (1) definitions typed in at the REPL
+ ;; without an INLINE declaration will give a NULL
+ ;; FUNCTION-LAMBDA-EXPRESSION (allowable, but not ideal)
+ ;; and (2) INLINE declarations will yield a
+ ;; FUNCTION-LAMBDA-EXPRESSION headed by
+ ;; SB-C:LAMBDA-WITH-LEXENV, even for null LEXENV. -- CSR,
+ ;; 2002-07-02
+ ;;
+ ;; (2) is probably fairly easy to fix -- it is, after all,
+ ;; a matter of list manipulation (or possibly of teaching
+ ;; CL:FUNCTION about SB-C:LAMBDA-WITH-LEXENV). (1) is
+ ;; significantly harder, as the association between
+ ;; function object and source is a tricky one.
+ ;;
+ ;; FUNCTION-LAMBDA-EXPRESSION "works" (i.e. returns a
+ ;; non-NULL list) when the function in question has been
+ ;; compiled by (COMPILE <x> '(LAMBDA ...)); it does not
+ ;; work when it has been compiled as part of the top-level
+ ;; EVAL strategy of compiling everything inside (LAMBDA ()
+ ;; ...). -- CSR, 2002-11-02
+ (when (core-object-p *compile-object*)
+ (fix-core-source-info *source-info* *compile-object* result))
+
+ (mapc #'clear-ir1-info components-from-dfo)
+ (clear-stuff)))))))
(defun process-toplevel-cold-fset (name lambda-expression path)
(unless (producing-fasl-file)
compile-time-too))))))
(if (atom form)
#+sb-xc-host
- ;; (There are no EVAL-WHEN issues in the ATOM case until
- ;; SBCL gets smart enough to handle global
- ;; DEFINE-SYMBOL-MACRO or SYMBOL-MACROLET.)
+ ;; (There are no xc EVAL-WHEN issues in the ATOM case until
+ ;; (1) SBCL gets smart enough to handle global
+ ;; DEFINE-SYMBOL-MACRO or SYMBOL-MACROLET and (2) SBCL
+ ;; implementors start using symbol macros in a way which
+ ;; interacts with SB-XC/CL distinction.)
(convert-and-maybe-compile form path)
#-sb-xc-host
(default-processor form)
(declare (ignore funs))
(process-toplevel-locally body
path
- compile-time-too))))
+ compile-time-too))
+ :compile))
((symbol-macrolet)
(funcall-in-symbol-macrolet-lexenv
magic
(process-toplevel-locally body
path
compile-time-too
- :vars vars)))))))
+ :vars vars))
+ :compile)))))
((locally)
(process-toplevel-locally (rest form) path compile-time-too))
((progn)
;;; Return (VALUES NIL WARNINGS-P FAILURE-P).
(defun sub-compile-file (info)
(declare (type source-info info))
- (let* ((*block-compile* *block-compile-arg*)
- (*package* (sane-package))
- (*policy* *policy*)
- (*lexenv* (make-null-lexenv))
- (*source-info* info)
- (sb!xc:*compile-file-pathname* nil)
- (sb!xc:*compile-file-truename* nil)
- (*toplevel-lambdas* ())
- (*compiler-error-bailout*
- (lambda ()
- (compiler-mumble "~2&; fatal error, aborting compilation~%")
- (return-from sub-compile-file (values nil t t))))
- (*current-path* nil)
- (*last-source-context* nil)
- (*last-original-source* nil)
- (*last-source-form* nil)
- (*last-format-string* nil)
- (*last-format-args* nil)
- (*last-message-count* 0)
- ;; FIXME: Do we need this rebinding here? It's a literal
- ;; translation of the old CMU CL rebinding to
- ;; (OR *BACKEND-INFO-ENVIRONMENT* *INFO-ENVIRONMENT*),
- ;; and it's not obvious whether the rebinding to itself is
- ;; needed that SBCL doesn't need *BACKEND-INFO-ENVIRONMENT*.
- (*info-environment* *info-environment*)
- (*gensym-counter* 0))
+ (let ((*package* (sane-package))
+ (*readtable* *readtable*)
+ (sb!xc:*compile-file-pathname* nil) ; really bound in
+ (sb!xc:*compile-file-truename* nil) ; SUB-SUB-COMPILE-FILE
+
+ (*policy* *policy*)
+ (*lexenv* (make-null-lexenv))
+ (*block-compile* *block-compile-arg*)
+ (*source-info* info)
+ (*toplevel-lambdas* ())
+ (*fun-names-in-this-file* ())
+ (*compiler-error-bailout*
+ (lambda ()
+ (compiler-mumble "~2&; fatal error, aborting compilation~%")
+ (return-from sub-compile-file (values nil t t))))
+ (*current-path* nil)
+ (*last-source-context* nil)
+ (*last-original-source* nil)
+ (*last-source-form* nil)
+ (*last-format-string* nil)
+ (*last-format-args* nil)
+ (*last-message-count* 0)
+ ;; FIXME: Do we need this rebinding here? It's a literal
+ ;; translation of the old CMU CL rebinding to
+ ;; (OR *BACKEND-INFO-ENVIRONMENT* *INFO-ENVIRONMENT*),
+ ;; and it's not obvious whether the rebinding to itself is
+ ;; needed that SBCL doesn't need *BACKEND-INFO-ENVIRONMENT*.
+ (*info-environment* *info-environment*)
+ (*gensym-counter* 0))
(handler-case
(with-compilation-values
(sb!xc:with-compilation-unit ()
(input-pathname (verify-source-file input-file))
(source-info (make-file-source-info input-pathname))
(*compiler-trace-output* nil)) ; might be modified below
-
+
(unwind-protect
(progn
(when output-file
(namestring input-pathname))))
(when trace-file
(let* ((default-trace-file-pathname
- (make-pathname :type "trace" :defaults input-pathname))
+ (make-pathname :type "ntrace" :defaults input-pathname))
(trace-file-pathname
(if (eql trace-file t)
default-trace-file-pathname
projects / sbcl.git / blobdiff