- cffi-devel - mailman3.common-lisp.net

[cffi-devel] define-foreign-library behavior
by hbabcockos1＠mac.com 26 May '06

26 May '06

Hello, On OS-X, this works: (define-foreign-library libc (t (:default "libc"))) (use-foreign-library libc) But this doesn't: (define-foreign-library libc (t (:or "libc" "libc.so.6"))) (use-foreign-library libc) This seems to be because load-foreign-library doesn't attempt to guess the appropriate suffix for the library name when used with :or. This is the documented behavior but it could be a little confusing since the second example appears to the naive (like me) to be a super- set of the first example. Also: (define-foreign-library libc (:unix (:or "libc.so.6")) (t (:default "libc"))) (use-foreign-library libc) Doesn't work since OS-X is unix, but not linux. Maybe after failing on the specific case (i.e. unix), it could fall back to the general case? I could try and provide patches for one or both of these behaviors if people think that this would be an improvement. best, -Hazen

4 4

[cffi-devel] Re: [Sbcl-devel] encoding of alien c-string once again
by Juho Snellman 22 May '06

22 May '06

[ CC:d to cffi-devel, in case the FFI experts have any opinions on external-format handling. ] Yaroslav Kavenchuk <kavenchuk(a)jenty.by> writes: > I have impudence to offer dear community patch for non-ascii encoding > of alien c-string once again. > > (Maybe me have forgotten? :) Thanks. There are some issues with this patch. * On non-windows platforms we need to naturalize a string to find out the default external format, and after this change we need to know the default external format to naturalize a string. But that's easy to fix. * Also, the patch can't be used stand-alone, since other SBCL internals (e.g. pathname handling) assume that a c-string alien type will be naturalized to a simple-base-string. What's the right thing to do here? Add a new alien-type that does the conversion, and leave c-string as-is, or change to change the behaviour of c-string? * I'm not sure that the interface is quite right. It seems probable that at one point or another somebody will need to use multiple external formats at once (ebcdic for pathnames and latin-1 for a database connection, or something). So we might need to be able to parametrize the external format to be used when defining the types. As a silly example: (define-alien-routine strdup (c-string :external-format :latin-1) (str (c-string :external-format :utf-8))) (Hmm... maybe using a keyword parameter for this is excessive, and an optional would do). James, Luis, would something like this be useful for CFFI? Would the proposed interface work for you? Any opinions on the c-string issue mentioned above? A patch implementing this functionality (including Yaroslav's work): http://jsnell.iki.fi/tmp/alien-again.diff -- Juho Snellman

4 4

[cffi-devel] CFFI::CANONICALIZE with arguments (NIL)
by Lou Vanek 20 May '06

20 May '06

I'm a glutton for punishment: . latest darcs cffi source code (about an hour old) . latest stable clisp (January 2006, cygwin version) . cygwin (latest version) hosted on windows xp sp2 . mysql 4.1.12 . latest clsql 3.6.1 (a few hours old) I'm trying to get all of these to play nice but the clsql tutorial example (clsql-tutorial-mysql.lisp) dies during compilation of clsql/db-mysql/mysql-api.lisp. The trace is below. I'm using uffi-compat.lisp. I'm trying to load "ctm" which is just a link to the clsql example file clsql-tutorial-mysql.lisp. *compile-print* and *load-print* are turned on for greater viewing pleasure. I'm aware clsql doesn't officially support clisp (especially on windows) but I'm hoping somebody got this to work. All the DLLs load. Can anybody tell me whether this is a cffi or a clsql or a platform issue? Please holler if there's more info I can provide. Thanks P.S. I tried SBCL but it's still too unstable on windows. I believe the first struct (in mysql-api.lisp) is being compiled when the error occurs. This is what the def-struct macro / NOTICE-FOREIGN-STRUCT-DEFINITION function is choking on: (uffi:def-struct mysql-net (vio :pointer-void) (fd mysql-socket) (fcntl :int) (buff (* :unsigned-char)) (buff-end (* :unsigned-char)) (write-pos (* :unsigned-char)) (read-pos (* :unsigned-char)) (last-error (:array :char 200)) (last-errno :unsigned-int) (max-packet :unsigned-int) (timeout :unsigned-int) (pkt-nr :unsigned-int) (error mysql-bool) (return-errno mysql-bool) (compress mysql-bool) (no-send-ok mysql-bool) (remain-in-buf :unsigned-long) (length :unsigned-long) (buf-length :unsigned-long) (where-b :unsigned-long) (return-status (* :unsigned-int)) (reading-or-writing :unsigned-char) (save-char :char)) ... ;; Loaded file /usr/local/asdf-install/site/clsql-3.6.1/db-mysql/mysql-client-info.fas ;; Compiling file /usr/local/asdf-install/site/clsql-3.6.1/db-mysql/mysql-api.lisp ... ; (IN-PACKAGE #:MYSQL) ; (CFFI-UFFI-COMPAT:DEF-FOREIGN-TYPE MYSQL-SOCKET :INT) ; (CFFI-UFFI-COMPAT:DEF-FOREIGN-TYPE MYSQL-BOOL :BYTE) ; (CFFI-UFFI-COMPAT:DEF-FOREIGN-TYPE MYSQL-BYTE :UNSIGNED-CHAR) ; (CFFI-UFFI-COMPAT:DEF-ENUM MYSQL-NET-TYPE (:TCP-IP :SOCKET :NAMED-PIPE)) ; (CFFI-UFFI-COMPAT:DEF-STRUCT MYSQL-NET (VIO :POINTER-VOID) ...) *** - NO-APPLICABLE-METHOD: When calling #<STANDARD-GENERIC-FUNCTION CFFI::CANONICALIZE> with arguments (NIL), no method is applicable. The following restarts are available: RETRY :R1 try calling CFFI::CANONICALIZE again RETURN :R2 specify return values RETRY :R3 Retry performing #<ASDF:COMPILE-OP NIL #x007A79CD> on #<ASDF:CL-SOURCE-FILE "mysql-api" #x008BA485>. ACCEPT :R4 Continue, treating #<ASDF:COMPILE-OP NIL #x007A79CD> on #<ASDF:CL-SOURCE-FILE "mysql-api" #x008BA485> as having been successful. SKIP :R5 skip (CASE *TUTORIAL-DATABASE-TYPE* # ...) STOP :R6 stop loading file /usr/local/asdf-install/site/clsql-3.6.1/examples/clsql-tutorial-mysql.lisp SKIP :R7 skip (LOAD ctm) STOP :R8 stop loading file /home/Lou_Vanek/.clisprc.lisp ABORT :R9 ABORT Break 1 MYSQL[2]> :bt1 <1> #<SYSTEM-FUNCTION EXT:SHOW-STACK> 3 <2> #<COMPILED-FUNCTION SYSTEM::PRINT-BACKTRACE> <3> #<COMPILED-FUNCTION SYSTEM::DEBUG-BACKTRACE> <4> #<COMPILED-FUNCTION SYSTEM::DEBUG-BACKTRACE-1> <5> #<SYSTEM-FUNCTION SYSTEM::READ-EVAL-PRINT> 2 <6> #<COMPILED-FUNCTION SYSTEM::BREAK-LOOP-2-2> <7> #<SYSTEM-FUNCTION SYSTEM::SAME-ENV-AS> 2 <8> #<COMPILED-FUNCTION SYSTEM::BREAK-LOOP-2> <9> #<SYSTEM-FUNCTION SYSTEM::DRIVER> - #<CLOS:METHOD-CALL-TYPE-ERROR #x006BFBC9> - NIL <10> #<COMPILED-FUNCTION SYSTEM::BREAK-LOOP> - #<CLOS:METHOD-CALL-TYPE-ERROR #x006BFBC9> <11> #<SYSTEM-FUNCTION INVOKE-DEBUGGER> 1 frame binding variables (~ = dynamically): | ~ SYSTEM::*PRIN-STREAM* <--> #<UNBOUND> frame binding variables (~ = dynamically): | ~ *PRINT-READABLY* <--> NIL frame binding variables (~ = dynamically): | ~ *PRINT-ESCAPE* <--> T - #<CLOS:METHOD-CALL-TYPE-ERROR #x006BFBC9> <12> #<SYSTEM-FUNCTION ERROR> frame binding variables (~ = dynamically): | ~ SYSTEM::*CONDITION-RESTARTS* <--> NIL frame binding variables (~ = dynamically): | ~ SYSTEM::*ACTIVE-RESTARTS* <--> (#S(RESTART :NAME ASDF:RETRY :TEST #<COMPILED-FUNCTION SYSTEM::DEFAULT-RESTART-TEST> :INVOKE-TAG NIL :INVOKE-FUNCTION #<FUNCTION :LAMBDA (&REST SYSTEM::ARGUMENTS) (SETQ #:G3623 SYSTEM::ARGUMENTS) (GO #:G3620)> :REPORT #<FUNCTION :LAMBDA (ASDF::S) (FORMAT ASDF::S "~@<Retry performing ~S on ~S.~@:>" ASDF::OP ASDF:COMPONENT)> :INTERACTIVE #<COMPILED-FUNCTION SYSTEM::DEFAULT-RESTART-INTERACTIVE> :MEANINGFULP T) #S(RESTART :NAME ASDF:ACCEPT :TEST #<COMPILED-FUNCTION SYSTEM::DEFAULT-RESTART-TEST> :INVOKE-TAG NIL :INVOKE-FUNCTION #<FUNCTION :LAMBDA (&REST SYSTEM::ARGUMENTS) (SETQ #:G3623 SYSTEM::ARGUMENTS) (GO #:G3621)> :REPORT #<FUNCTION :LAMBDA (ASDF::S) (FORMAT ASDF::S "~@<Continue, treating ~S on ~S as ~ having been successful.~@:>" ASDF::OP ASDF:COMPONENT)> :INTERACTIVE #<COMPILED-FUNCTION SYSTEM::DEFAULT-RESTART-INTERACTIVE> :MEANINGFULP T) #S(RESTART :NAME SYSTEM::SKIP :TEST #<COMPILED-FUNCTION SYSTEM::DEFAULT-RESTART-TEST> :INVOKE-TAG NIL :INVOKE-FUNCTION #<COMPILED-FUNCTION SYSTEM::EVAL-LOADED-FORM-1> :REPORT #<COMPILED-FUNCTION SYSTEM::EVAL-LOADED-FORM-2> :INTERACTIVE #<COMPILED-FUNCTION SYSTEM::DEFAULT-RESTART-INTERACTIVE> :MEANINGFULP T) #S(RESTART :NAME SYSTEM::STOP :TEST #<COMPILED-FUNCTION SYSTEM::DEFAULT-RESTART-TEST> :INVOKE-TAG NIL :INVOKE-FUNCTION #<COMPILED-FUNCTION SYSTEM::EVAL-LOADED-FORM-3> :REPORT #<COMPILED-FUNCTION SYSTEM::EVAL-LOADED-FORM-4> :INTERACTIVE #<COMPILED-FUNCTION SYSTEM::DEFAULT-RESTART-INTERACTIVE> :MEANINGFULP T) #S(RESTART :NAME SYSTEM::SKIP :TEST #<COMPILED-FUNCTION SYSTEM::DEFAULT-RESTART-TEST> :INVOKE-TAG NIL :INVOKE-FUNCTION #<COMPILED-FUNCTION SYSTEM::EVAL-LOADED-FORM-1> :REPORT #<COMPILED-FUNCTION SYSTEM::EVAL-LOADED-FORM-2> :INTERACTIVE #<COMPILED-FUNCTION SYSTEM::DEFAULT-RESTART-INTERACTIVE> :MEANINGFULP T) #S(RESTART :NAME SYSTEM::STOP :TEST #<COMPILED-FUNCTION SYSTEM::DEFAULT-RESTART-TEST> :INVOKE-TAG NIL :INVOKE-FUNCTION #<COMPILED-FUNCTION SYSTEM::EVAL-LOADED-FORM-3> :REPORT #<COMPILED-FUNCTION SYSTEM::EVAL-LOADED-FORM-4> :INTERACTIVE #<COMPILED-FUNCTION SYSTEM::DEFAULT-RESTART-INTERACTIVE> :MEANINGFULP T) #S(RESTART :NAME ABORT :TEST #<COMPILED-FUNCTION SYSTEM::DEFAULT-RESTART-TEST> :INVOKE-TAG NIL :INVOKE-FUNCTION #<COMPILED-FUNCTION SYSTEM::MAIN-LOOP-1-1> :REPORT NIL :INTERACTIVE #<COMPILED-FUNCTION SYSTEM::DEFAULT-RESTART-INTERACTIVE> :MEANINGFULP T)) compiled tagbody frame for #(NIL NIL) - 93 - 107 - #(#(NIL #<STANDARD-GENERIC-FUNCTION CFFI::CANONICALIZE>) NIL) - #(NIL #<STANDARD-GENERIC-FUNCTION CFFI::CANONICALIZE>) - #<COMPILED-FUNCTION SYSTEM::RETRY-FUNCTION-CALL> - (NIL) - #<STANDARD-GENERIC-FUNCTION CFFI::CANONICALIZE> - #<CLOS:METHOD-CALL-TYPE-ERROR #x006BFBC9> <13> #<COMPILED-FUNCTION SYSTEM::RETRY-FUNCTION-CALL> - 0 - (#<STANDARD-METHOD (#<STANDARD-CLASS CFFI::FOREIGN-STRUCT-TYPE>)> #<STANDARD-METHOD (#<STANDARD-CLASS CFFI::FOREIGN-TYPE-ALIAS>)> #<STANDARD-METHOD (#<STANDARD-CLASS CFFI::FOREIGN-POINTER-TYPE>)> #<STANDARD-METHOD (#<STANDARD-CLASS CFFI::FOREIGN-BUILT-IN-TYPE>)>) - #<COMPILED-FUNCTION #:|87 107 (DEFGENERIC NO-APPLICABLE-METHOD (GF &REST ARGS) ...)-9-1|> - (NIL) - #<STANDARD-GENERIC-FUNCTION CFFI::CANONICALIZE> <14> #<COMPILED-FUNCTION #:|87 107 (DEFGENERIC NO-APPLICABLE-METHOD (GF &REST ARGS) ...)-9-1|> <15> #<STANDARD-GENERIC-FUNCTION NO-APPLICABLE-METHOD> <16> #<STANDARD-GENERIC-FUNCTION NO-APPLICABLE-METHOD> - NO-APPLICABLE-METHOD - #<COMPILED-FUNCTION CLOS::NO-METHOD-CALLER-1> - (NIL) <17> #<COMPILED-FUNCTION CLOS::NO-METHOD-CALLER-1> <18> #<STANDARD-GENERIC-FUNCTION CFFI::CANONICALIZE> - #<COMPILED-FUNCTION #:|267 269 (DEFMETHOD CANONICALIZE (#) ...)-41-1-1|> - #<CFFI-UFFI-COMPAT::UFFI-ARRAY-TYPE #:ANONYMOUS-CFFI-TYPE7352> <19> #<COMPILED-FUNCTION #:|267 269 (DEFMETHOD CANONICALIZE (#) ...)-41-1-1|> <20> #<STANDARD-GENERIC-FUNCTION CFFI::CANONICALIZE> - #<COMPILED-FUNCTION CFFI::CANONICALIZE-FOREIGN-TYPE> - (CFFI-UFFI-COMPAT::UFFI-ARRAY (CFFI-UFFI-COMPAT::UFFI-CHAR :CHAR) 200) <21> #<COMPILED-FUNCTION CFFI::CANONICALIZE-FOREIGN-TYPE> - NIL - 1 - NIL - (CFFI-UFFI-COMPAT::UFFI-ARRAY (CFFI-UFFI-COMPAT::UFFI-CHAR :CHAR) 200) - LAST-ERROR - (LAST-ERROR (CFFI-UFFI-COMPAT::UFFI-ARRAY (CFFI-UFFI-COMPAT::UFFI-CHAR :CHAR) 200)) - ((LAST-ERROR (CFFI-UFFI-COMPAT::UFFI-ARRAY (CFFI-UFFI-COMPAT::UFFI-CHAR :CHAR) 200)) (LAST-ERRNO :UNSIGNED-INT) (MAX-PACKET :UNSIGNED-INT) (TIMEOUT :UNSIGNED-INT) (PKT-NR :UNSIGNED-INT) (ERROR MYSQL-BOOL) (RETURN-ERRNO MYSQL-BOOL) (COMPRESS MYSQL-BOOL) (NO-SEND-OK MYSQL-BOOL) (REMAIN-IN-BUF :UNSIGNED-LONG) (LENGTH :UNSIGNED-LONG) (BUF-LENGTH :UNSIGNED-LONG) (WHERE-B :UNSIGNED-LONG) (RETURN-STATUS :POINTER) (READING-OR-WRITING (CFFI-UFFI-COMPAT::UFFI-CHAR :UNSIGNED-CHAR)) (SAVE-CHAR (CFFI-UFFI-COMPAT::UFFI-CHAR :CHAR))) - NIL - 4 - 28 - #<CFFI::FOREIGN-STRUCT-TYPE MYSQL-NET> - NIL - NIL - MYSQL-NET - (MYSQL-NET) - #<COMPILED-FUNCTION CFFI::NOTICE-FOREIGN-STRUCT-DEFINITION> - ((VIO :POINTER) (FD MYSQL-SOCKET) (FCNTL :INT) (BUFF :POINTER) (BUFF-END :POINTER) (WRITE-POS :POINTER) (READ-POS :POINTER) (LAST-ERROR (CFFI-UFFI-COMPAT::UFFI-ARRAY (CFFI-UFFI-COMPAT::UFFI-CHAR :CHAR) 200)) (LAST-ERRNO :UNSIGNED-INT) (MAX-PACKET :UNSIGNED-INT) (TIMEOUT :UNSIGNED-INT) (PKT-NR :UNSIGNED-INT) (ERROR MYSQL-BOOL) (RETURN-ERRNO MYSQL-BOOL) (COMPRESS MYSQL-BOOL) (NO-SEND-OK MYSQL-BOOL) (REMAIN-IN-BUF :UNSIGNED-LONG) (LENGTH :UNSIGNED-LONG) (BUF-LENGTH :UNSIGNED-LONG) (WHERE-B :UNSIGNED-LONG) (RETURN-STATUS :POINTER) (READING-OR-WRITING (CFFI-UFFI-COMPAT::UFFI-CHAR :UNSIGNED-CHAR)) (SAVE-CHAR (CFFI-UFFI-COMPAT::UFFI-CHAR :CHAR))) - MYSQL-NET <22> #<COMPILED-FUNCTION CFFI::NOTICE-FOREIGN-STRUCT-DEFINITION> - #<COMPILED-FUNCTION CFFI::NOTICE-FOREIGN-STRUCT-DEFINITION> EVAL frame for form (CFFI::NOTICE-FOREIGN-STRUCT-DEFINITION 'MYSQL-NET '((VIO :POINTER) (FD MYSQL-SOCKET) (FCNTL :INT) (BUFF :POINTER) (BUFF-END :POINTER) (WRITE-POS :POINTER) (READ-POS :POINTER) (LAST-ERROR (CFFI-UFFI-COMPAT::UFFI-ARRAY (CFFI-UFFI-COMPAT::UFFI-CHAR :CHAR) 200)) (LAST-ERRNO :UNSIGNED-INT) (MAX-PACKET :UNSIGNED-INT) (TIMEOUT :UNSIGNED-INT) (PKT-NR :UNSIGNED-INT) (ERROR MYSQL-BOOL) (RETURN-ERRNO MYSQL-BOOL) (COMPRESS MYSQL-BOOL) (NO-SEND-OK MYSQL-BOOL) (REMAIN-IN-BUF :UNSIGNED-LONG) (LENGTH :UNSIGNED-LONG) (BUF-LENGTH :UNSIGNED-LONG) (WHERE-B :UNSIGNED-LONG) (RETURN-STATUS :POINTER) (READING-OR-WRITING (CFFI-UFFI-COMPAT::UFFI-CHAR :UNSIGNED-CHAR)) (SAVE-CHAR (CFFI-UFFI-COMPAT::UFFI-CHAR :CHAR)))) - NIL <23> #<SPECIAL-OPERATOR PROGN> EVAL frame for form (PROGN (CFFI::NOTICE-FOREIGN-STRUCT-DEFINITION 'MYSQL-NET '((VIO :POINTER) (FD MYSQL-SOCKET) (FCNTL :INT) (BUFF :POINTER) (BUFF-END :POINTER) (WRITE-POS :POINTER) (READ-POS :POINTER) (LAST-ERROR (CFFI-UFFI-COMPAT::UFFI-ARRAY (CFFI-UFFI-COMPAT::UFFI-CHAR :CHAR) 200)) (LAST-ERRNO :UNSIGNED-INT) (MAX-PACKET :UNSIGNED-INT) (TIMEOUT :UNSIGNED-INT) (PKT-NR :UNSIGNED-INT) (ERROR MYSQL-BOOL) (RETURN-ERRNO MYSQL-BOOL) (COMPRESS MYSQL-BOOL) (NO-SEND-OK MYSQL-BOOL) (REMAIN-IN-BUF :UNSIGNED-LONG) (LENGTH :UNSIGNED-LONG) (BUF-LENGTH :UNSIGNED-LONG) (WHERE-B :UNSIGNED-LONG) (RETURN-STATUS :POINTER) (READING-OR-WRITING (CFFI-UFFI-COMPAT::UFFI-CHAR :UNSIGNED-CHAR)) (SAVE-CHAR (CFFI-UFFI-COMPAT::UFFI-CHAR :CHAR))))) frame binding environments VAR_ENV <--> #<ADDRESS #x1918084C> FUN_ENV <--> #(NEXT-METHOD-P #<FUNCTION-MACRO #<FUNCTION NEXT-METHOD-P (SYSTEM::<MACRO-FORM> SYSTEM::<ENV-ARG>) (DECLARE (CONS SYSTEM::<MACRO-FORM>)) (DECLARE (IGNORE SYSTEM::<ENV-ARG>)) (IF (> (EXT:LIST-LENGTH-DOTTED SYSTEM::<MACRO-FORM>) 1) (SYSTEM::MACRO-CALL-ERROR SYSTEM::<MACRO-FORM>) (LET* NIL (BLOCK NEXT-METHOD-P '#:G3549)))> #<FUNCTION NEXT-METHOD-P NIL #:G3549>> CALL-NEXT-METHOD #<FUNCTION-MACRO #<FUNCTION CALL-NEXT-METHOD (SYSTEM::<MACRO-FORM> SYSTEM::<ENV-ARG>) (DECLARE (CONS SYSTEM::<MACRO-FORM>)) (DECLARE (IGNORE SYSTEM::<ENV-ARG>)) (LET* ((CLOS::NEW-ARG-EXPRS (CDR SYSTEM::<MACRO-FORM>))) (BLOCK CALL-NEXT-METHOD (IF CLOS::NEW-ARG-EXPRS (LIST 'IF '(EVAL-WHEN (EVAL) T) (LIST 'CLOS::%CALL-NEXT-METHOD '#:G3548 '#:G3549 (LIST 'LIST '#:G3550 '#:G3551) (CONS 'LIST CLOS::NEW-ARG-EXPRS)) (LIST 'IF '#:G3549 (LIST* 'FUNCALL '#:G3549 CLOS::NEW-ARG-EXPRS) (LIST* 'CLOS::%NO-NEXT-METHOD '#:G3548 CLOS::NEW-ARG-EXPRS))) (LIST 'IF '#:G3549 (LIST 'FUNCALL '#:G3549 '#:G3550 '#:G3551) (LIST 'CLOS::%NO-NEXT-METHOD '#:G3548 '#:G3550 '#:G3551)))))> #<FUNCTION CALL-NEXT-METHOD (&REST CLOS::NEW-ARGS) (IF CLOS::NEW-ARGS (IF (EVAL-WHEN (EVAL) T) (CLOS::%CALL-NEXT-METHOD #:G3548 #:G3549 (LIST #:G3550 #:G3551) CLOS::NEW-ARGS) (IF #:G3549 (APPLY #:G3549 CLOS::NEW-ARGS) (APPLY #'CLOS::%NO-NEXT-METHOD #:G3548 CLOS::NEW-ARGS))) (IF #:G3549 (FUNCALL #:G3549 #:G3550 #:G3551) (CLOS::%NO-NEXT-METHOD #:G3548 #:G3550 #:G3551)))>> NIL) BLOCK_ENV <--> NIL GO_ENV <--> NIL DECL_ENV <--> ((DECLARATION OPTIMIZE DECLARATION)) <24> #<SYSTEM-FUNCTION EVAL> - (PROGN (CFFI::NOTICE-FOREIGN-STRUCT-DEFINITION 'MYSQL-NET ...

2 6

[cffi-devel] callbacks with variable arguments
by Anselm Helbig 20 May '06

20 May '06

Hi list! Background: I'm trying to write a CFFI interface to the silc client library (see silcnet.org) for use with my silcbot. This is on SBCL 0.9.11, Debian sarge, CFFI-tarball from 2006-04-24, silc toolkit 1.0.2. Problem: the silc client library needs callbacks with variable argument lists. Is there a portable way to access these arguments? Possible solutions: - I could define additional arguments to fit the maximal size of the argument list, and do some byte-shifting if necessary. Yuck! - This should be possible to do in ECL. Cons: I lose portability, and ECL does not seem to be as widely supported on Linux as SBCL (libraries might break), and since this is my first CL-project I don't feel like patching libraries myself yet. Has anyone a better idea how to accomplish this? Regards, Anselm

2 1

[cffi-devel] cffi-uffi-compat vs :float
by Ken Tilton 20 May '06

20 May '06

I think this kind of code does not like cffi-uffi-compat: (defun gl-apply-matrix (a b c d x y) (let ((m (allocate-foreign-object :float 16)) (l (list a c 0 0 b d 0 0 0 0 1 0 x y 0 1))) (loop for i from 0 to 15 and j in l do (setf (deref-array m :float i) (coerce j 'float))) (gl-mult-matrixf m) (free-foreign-object m))) The allocate ended up with CFFI being passed ':float, which really made CFFI unhappy. I beat on things a little and got past that with the dubious: (defmacro allocate-foreign-object (type &optional (size 1)) "Allocate one or more instance of a foreign type." `(cffi:foreign-alloc ,(cond ((keywordp type) `,(convert-uffi-type type)) ((constantp type) `',(convert-uffi-type (eval type))) (t `(convert-uffi-type ,type))) :count ,size)) But a similar hack on deref-array did not fly and when I saw why I decided to port the code to native CFFI since there is not that much of it. Are those easily fixable? There is a lot of UFFI code out there that might like CFFI's portability. (I know you know that. <g>) kenny

2 2

[cffi-devel] defcallback has to be invoked on every run? yes? no?
by Ken Tilton 20 May '06

20 May '06

Found this comment in some code I am porting to CFFI: ; Seems like ff-defun-callable should be executed every time. ; In OpenMCL, loading .dfsl file is not enough. ff-defun-callable was my creation when UFFI did not handle callbacks. It was like UFFI in that it Just Wrapped native Lisp implementation FFI callback syntax. Uhhhh... if I am using CFFI, can I just code defcallback at the toplevel and invoke get-callback at runtime when passing the callback to the library? Existing code calls a defun which executes ff-defun-callable each time the app runs. kt

2 1

[cffi-devel] :ensure-float? Was I dreaming?
by Ken Tilton 17 May '06

17 May '06

Yo, dawgs! Lissen up... sorry, too much American Idol. Here's the deal. One of my bindings was not working because the parameter was float and I was passing a fixnum. Coerced to a float and all was groovy. Two things: 1. Minor question: I coulda swore the first time I fixed this I dug around and discovered :ensure-float and used that in my DEFCFUN. The second time I ran into it I went looking for :ensure-float and could not find it. Found :ensure-XXX for a couple other things in the UFFI-compat bit. Maybe it was one of the XXXs. No big. 2. Am I missing something? Is there some reason we cannot have an ensure-float? Is it one of those deals where for some obscure reason unbeknownst to me there is no reliable way to guess what is needed and programmers need to Just Provide the Right Type? thx, kenny

2 1

Re: [cffi-devel] :ensure-float? Was I dreaming?
by Ken Tilton 17 May '06

17 May '06

> Am 17.05.2006 um 21:24 schrieb Ken Tilton: > > > Yo, dawgs! Lissen up... sorry, too much American Idol. > > > > > > Here's the deal. One of my bindings was not working because the > > parameter was float and I was passing a fixnum. Coerced to a float > > and all was groovy. Two things: > > > > 1. Minor question: I coulda swore the first time I fixed this I dug > > around and discovered :ensure-float and used that in my DEFCFUN. > > The second time I ran into it I went looking for :ensure-float and > > could not find it. Found :ensure-XXX for a couple other things in > > the UFFI-compat bit. Maybe it was one of the XXXs. No big. Thx, Frank. But I try to resolve such tears in the space-time continuum before continuing because it often means I missed Something Important. Not this time: found ensure-float as part of the renegade cl-opengl package, which I did use on the Celtk Gears demo. whew! Meanwhile, OT, cl-openal is alive and well and working atop CFFI, via my Hello-CFFI package. I do believe it is Snack time. :) kenny

1 0

[cffi-devel] cffi_0.9.1 vs cffi-060417
by Rick Taube 16 May '06

16 May '06

I have a program that runs well in cffi-060417 (SBCL 0.9.10/linux and openmcl/darwin), but crashes in the 0.9.1 release in SBCL 0.9.10/linux The error message is: vicious metacircle: The computation of an effective method of #<STANDARD-GENERIC-FUNCTION CFFI:FREE-TRANSLATED-OBJECT (3)> for arguments of types (#<BUILT-IN-CLASS T> #<BUILT-IN-CLASS SYMBOL> #<BUILT-IN-CLASS T>) uses the effective method being computed. unfortunately I have no idea what is causing this error other than its happening under my scheduler's callback into lisp. any help trying to figure this out would be appreciated! --rick * (rts) ; RTS running * (sprout (process repeat 5 output (new midi :keynum (between 60 80)) wait .5)) debugger invoked on a SIMPLE-ERROR in thread #<THREAD "initial thread" {A68D371}>: vicious metacircle: The computation of an effective method of #<STANDARD-GENERIC-FUNCTION CFFI:FREE-TRANSLATED-OBJECT (3)> for arguments of types (#<BUILT-IN-CLASS T> #<BUILT-IN-CLASS SYMBOL> #<BUILT-IN-CLASS T>) uses the effective method being computed. Type HELP for debugger help, or (SB-EXT:QUIT) to exit from SBCL. restarts (invokable by number or by possibly-abbreviated name): 0: [ABORT-RTS ] Abort RTS, stopping scheduler. 1: [CONTINUE-RTS ] Delete current entry and continue. 2: [FLUSH-QUEUE ] Flush pending events from scheduler. 3: [REMOVE-CALLBACK] Remove callback function. 4: [ABORT ] Exit debugger, returning to top level. (SB-PCL::BREAK-VICIOUS-METACIRCLE #<STANDARD-GENERIC-FUNCTION CFFI:FREE-TRANSLATED-OBJECT (3)> (#<BUILT-IN-CLASS T> #<BUILT-IN-CLASS SYMBOL> #<BUILT-IN-CLASS T>) #S(SB-PCL::ARG-INFO :ARG-INFO-LAMBDA-LIST (CFFI::VALUE CFFI::NAME CFFI::PARAM) :ARG-INFO-PRECEDENCE (0 1 2) :ARG-INFO-METATYPES (T CLASS T) :ARG-INFO-NUMBER-OPTIONAL 0 :ARG-INFO-KEY/REST-P NIL :ARG-INFO-KEYS NIL :GF-INFO-SIMPLE-ACCESSOR-TYPE NIL :GF-PRECOMPUTE-DFUN-AND-EMF-P NIL :GF-INFO-STATIC-C-A-M-EMF #S(SB-PCL::FAST-METHOD-CALL :FUNCTION # :PV-CELL NIL :NEXT-METHOD-CALL NIL :ARG-INFO (2)) :GF-INFO-C-A-M-EMF-STD-P T :GF-INFO-FAST-MF-P T)) 0] debugger invoked on a SB-KERNEL::MEMORY-FAULT-ERROR in thread #<THREAD "initial thread" {A68D371}>: memory fault Segmentation fault -bash-3.00$

1 0

[cffi-devel] defcallback ignores ignorable <g>
by Ken Tilton 15 May '06

15 May '06

Whoa, CFFI rocks, having a lot of fun creating an FFI-based version of Celtk. As the subject suggests, methinks the defcallback macro knows about IGNORE but not IGNORABLE. I use the latter a lot because of debug statements that come and go. best, kenny

3 4