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

[cffi-devel] Status of the ECL Backend
by James Bielman 19 Feb '06

19 Feb '06

Hi, I tried to compile and test the CFFI backend for ECL using the latest CVS and ran into problems related to type translators and the ECL fasl dumper apparently not respecting MAKE-LOAD-FORM. I don't know if anyone is using the ECL backend for CFFI, but I'll try to prepare a proper bug report for the ECL folks... James

1 0

[cffi-devel] New patches: 17-Feb-2006
by loliveira＠common-lisp.net 18 Feb '06

18 Feb '06

Fri Feb 17 14:44:57 CST 2006 James Bielman <jamesjb(a)jamesjb.com> * Fix LOOP indentation in backends. - Thanks for Luís for showing me how to configure Emacs to get this right. M ./src/cffi-allegro.lisp -3 +3 M ./src/cffi-clisp.lisp -1 +1 M ./src/cffi-cmucl.lisp -15 +12 M ./src/cffi-ecl.lisp -1 +1 M ./src/cffi-lispworks.lisp -3 +3 M ./src/cffi-openmcl.lisp -12 +9 M ./src/cffi-sbcl.lisp -15 +12 An updated tarball of CFFI's source can be downloaded here: http://common-lisp.net/project/cffi/tarballs/cffi-060217.tar.gz Darcsweb URL: http://common-lisp.net/cgi-bin/darcsweb/darcsweb.cgi?r=cffi;a=summary

1 0

Re: [cffi-devel] New patches: 3-Feb-2006: x86-64
by Hoehle, Joerg-Cyril 17 Feb '06

17 Feb '06

James, >Sorry... here's my train of thought: I wanted to support long long on >CLISP either based on the version or whether it is built for a 64-bit >target (amd64). I'm not sure this makes sense. I'd try this: o if it's a 64bit machine, assume longlong has always worked (since it's equal to long). o otherwise, check it's >2.38 (you cannot AFAIK detect the CVS number, so this check would currently fail...) Maybe this would work for the first test? (= (ffi:sizeof 'ffi:long) (sizeof 'ffi:int64)) In fact, that should be the precise test to detect whether longlong worked in the past: when long was used instead, there was no problem. Regards, Jorg Hohle

1 0

[cffi-devel] [ANN] SWIG Release - 1.3.28 - with CFFI
by efuzzyone＠netscape.net 17 Feb '06

17 Feb '06

The SWIG Development team has released SWIG 1.3.28, which now supports CFFI. http://www.swig.org/download.html Full Announcement: *** ANNOUNCE: SWIG 1.3.28 (February 12, 2006) *** http://www.swig.org We're pleased to announce SWIG-1.3.28, the latest installment in the SWIG development effort. SWIG-1.3.28 includes a number of bug fixes and large number of enhancements throughout. What is SWIG? ------------- SWIG is a software development tool that reads C/C++ header files and generates the wrapper code needed to make C and C++ code accessible from other languages including Perl, Python, Tcl, Ruby, PHP, Java, Scheme (Guile, MzScheme, CHICKEN), Ocaml, Lua, Pike, C#, Modula-3, and Common Lisp (CLISP, Allegro CL, CFFI, UFFI). SWIG can also export its parse tree in the form of XML and Lisp s-expressions. Major applications of SWIG include generation of scripting language extension modules, rapid prototyping, testing, and user interface development for large C/C++ systems. Availability: ------------- The release is available for download on Sourceforge at http://prdownloads.sourceforge.net/swig/swig-1.3.28.tar.gz A Windows version should also be available at http://prdownloads.sourceforge.net/swig/swigwin-1.3.28.zip What's New? =========== SWIG-1.3.28 summary: - More powerful renaming (%rename) capability. - More user friendly warning handling. - Add finer control for default constructors and destructors. We discourage the use of the 'nodefault' option, which disables both constructors and destructors, leading to possible memory leaks. Use instead 'nodefaultctor' and/or 'nodefaultdtor'. - Automatic copy constructor wrapper generation via the 'copyctor' option/feature. - Better handling of Windows extensions and types. - Better runtime error reporting. - Add the %catches directive to catch and dispatch exceptions. - Add the %naturalvar directive for more 'natural' variable wrapping. - Better default handling of std::string variables using the %naturalvar directive. - Add the %allowexcept and %exceptionvar directives to handle exceptions when accessing a variable. - Add the %delobject directive to mark methods that act like destructors. - Add the -fastdispatch option to enable smaller and faster overload dispatch mechanism. - Template support for %rename, %feature and %typemap improved. - Add/doc more debug options, such as -dump_module, -debug_typemaps, etc. - Unified typemap library (UTL) potentially providing core typemaps for all scripting languages based on the recently evolving Python typemaps. - New language module: Common Lisp with CFFI. - Python, Ruby, Perl and Tcl use the new UTL, many old reported and hidden errors with typemaps are now fixed. - Initial Java support for languages using the UTL via GCJ, you can now use Java libraries in your favorite script language using gcj + swig. - PHP4 module update, many error fixes and actively maintained again. - Allegrocl support for C++, also enhanced C support. - Ruby support for bang methods. - Ruby support for user classes as native exceptions. - Perl improved dispatching in overloaded functions via the new cast and rank mechanism. - Perl improved backward compatibility, 5.004 and later tested and working. - Python improved backward compatibility, 1.5.2 and later tested and working. - Python can use the same cast/rank mechanism via the -castmode option. - Python implicit conversion mechanism similar to C++, via the %implicitconv directive (replaces and improves the implicit.i library). - Python threading support added. - Python STL support improved, iterators are supported and STL containers can use now the native PyObject type. - Python many performance options and improvements, try the -O option to test all of them. Python runtime benchmarks show up to 20 times better performance compared to 1.3.27 and older versions. - Python support for 'multi-inheritance' on the python side. - Python simplified proxy classes, now swig doesn't need to generate the additional 'ClassPtr' classes. - Python extended support for smart pointers. - Python better support for static member variables. - Python backward compatibility improved, many projects that used to work only with swig-1.3.21 to swig-1.3.24 are working again with swig-1.3.28 - Python test-suite is now 'valgrinded' before release, and swig also reports memory leaks due to missing destructors. - Minor bug fixes and improvements to the Lua, Ruby, Java, C#, Python, Guile, Chicken, Tcl and Perl modules. Release numbers --------------- With SWIG-1.3, we are adopting an odd/even version numbering scheme for SWIG. Odd version numbers (1.3, 1.5, 1.7, etc...) are considered to be development releases. Even numbers (1.4,1.6,1.8) are stable releases. The current 1.3 effort is working to produce a stable 2.0 release. A stable 2.0 release will not be made until it can accompanied by fully updated documentation. In the meantime, we will continue to make periodic 1.3.x releases. We need your help! ------------------ Even if you are perfectly happy with SWIG-1.1, we can still use your feedback. First, we like to know about compilation problems and other issues concerning the building of SWIG. Second, if SWIG-1.3 is unable to compile your old interface files, we would like to get information about the features you are using. This information will help us find bugs in the SWIG-1.3 release, develop techniques for supporting backwards compatibility, and write documentation that addresses specific issues related to migrating from SWIG-1.1 to SWIG-1.3. We are also looking for volunteers who would like to work on various aspects of SWIG development. SWIG is an unfunded project that would not exist without volunteers. We are also looking for the developers of other SWIG language modules. If you have developed a SWIG module and would like to see it incorporated into the new release, please contact us to obtain SWIG-CVS access. We are also more than willing to help port your module from SWIG-1.1 to SWIG-1.3. Please send email to beazley(a)cs.uchicago.edu for further information. Please report problems with this release to the swig-dev mailing list, details at http://www.swig.org/mail.html. --- The SWIG Developers

1 0

[cffi-devel] New patches: 16-Feb-2006
by loliveira＠common-lisp.net 17 Feb '06

17 Feb '06

Thu Feb 16 02:34:48 CST 2006 James Bielman <jamesjb(a)jamesjb.com> * Fix and add tests for %MEM-REF and %MEM-SET evaluation order. - Some minor reformatting of LOOP forms to pacify cl-indent. - Add new regression tests to check the evaluation order of %MEM-REF and %MEM-SET. - Add the necessary ONCE-ONLY forms to the compiler macros. - Have the CFFI-TESTS package use CFFI-SYS for testing primitives. M ./src/cffi-allegro.lisp -3 +4 M ./src/cffi-cmucl.lisp -10 +14 M ./src/cffi-lispworks.lisp -14 +16 M ./src/cffi-openmcl.lisp -10 +14 M ./src/cffi-sbcl.lisp -23 +27 M ./tests/memory.lisp +34 M ./tests/package.lisp -1 +1 An updated tarball of CFFI's source can be downloaded here: http://common-lisp.net/project/cffi/tarballs/cffi-060216.tar.gz Darcsweb URL: http://common-lisp.net/cgi-bin/darcsweb/darcsweb.cgi?r=cffi;a=summary

1 0

Re: [cffi-devel] New patches: 3-Feb-2006: x86-64
by Hoehle, Joerg-Cyril 16 Feb '06

16 Feb '06

James Bielman writes: >Hmm, I've tried to install an x86 CLISP in a 32-bit chroot to prepare >to implement and test this, but I'm a little concerned about the fact >that MACHINE-TYPE still returns X86-64 on a 32-bit x86 build. Is >there a better way to tell whether we are running an x86 or x86-64 >CLISP? I'm sorry, I don't understand what you observed, what you want or what you need. Feel free to elaborate. Regards, Jorg Hohle.

2 1

[cffi-devel] New patches: 15-Feb-2006
by loliveira＠common-lisp.net 16 Feb '06

16 Feb '06

Wed Feb 15 23:15:44 CST 2006 Luis Oliveira <loliveira(a)common-lisp.net> * Fix typo. extend -> extent M ./doc/cffi-manual.texinfo -1 +1 Wed Feb 15 23:07:35 CST 2006 Luis Oliveira <loliveira(a)common-lisp.net> * Minor change to tests/defcfun.lisp - Remove unnecessary (setf (mem-ref s :char) 0) forms. M ./tests/defcfun.lisp -7 Wed Feb 15 22:55:47 CST 2006 Luis Oliveira <loliveira(a)common-lisp.net> * CFFI Manual update - Update known issues (callback.bff.[12] failures). - defbitfield, foreign-bitfield-symbols and foreign-bitfield-value. - Fix typo in tutorial. defenum -> defcenum. - New section "Other Types" documenting :string, :boolean and :wrapper. - Add note about the translate-* methods not being meant to be called directly and suggesting convert-* instead. - Document the macroexpansion-time translators in a new "Optimizing Type Translators" sub-section. - Document defcenum's base-type option. - Document defcfun's varargs support. M ./doc/cffi-manual.texinfo -15 +371 Wed Feb 15 22:53:35 CST 2006 Luis Oliveira <loliveira(a)common-lisp.net> * New functions convert-to/from-foreign and free-converted-object - These functions basically export the functionality of translate-type-to-foreign, translate-type-from-foreign and free-translated-object. - TODO: document these. M ./TODO -3 +2 M ./src/package.lisp +3 M ./src/types.lisp +23 Wed Feb 15 22:52:40 CST 2006 Luis Oliveira <loliveira(a)common-lisp.net> * New TODO item: defcfun compiler macros M ./TODO +3 Wed Feb 15 16:25:59 CST 2006 James Bielman <jamesjb(a)jamesjb.com> * Implement compiler macros for %MEM-REF and %MEM-SET in LispWorks. - Use FLI:FOREIGN-TYPED-AREF if available and dereferencing a number. - Avoid calling INC-POINTER when the offset is a multiple of the type size allowing direct use of the INDEX argument to FOREIGN-TYPED-AREF. - Fall back to open-coding the call to FLI:DEREFERENCE otherwise. M ./src/cffi-lispworks.lisp +84 Wed Feb 15 15:34:42 CST 2006 James Bielman <jamesjb(a)jamesjb.com> * Delete LispWorks/Linux .ufsl files on make clean. M ./Makefile -1 +1 Wed Feb 15 14:15:14 CST 2006 James Bielman <jamesjb(a)jamesjb.com> * Use C limits for foreign floats and doubles instead of Lisp's. - Export 'float_min', 'float_max', 'double_min', and 'double_max' from the libtest shared library with the values of FLT_MIN, FLT_MAX, DBL_MIN, and DBL_MAX, respectively. - Use *FLOAT-MIN*, *FLOAT-MAX*, *DOUBLE-MIN*, and *DOUBLE-MAX* as test values for foreign floats/doubles instead of the Lisp constants <foo>-POSITIVE-<bar>-FLOAT. M ./tests/bindings.lisp -1 +9 M ./tests/libtest.c -1 +7 M ./tests/memory.lisp -16 +16 Wed Feb 15 11:43:16 CST 2006 Luis Oliveira <loliveira(a)common-lisp.net> * Oops, fix tests. - MISC-TYPES.BOOLEAN.2 was using the wrong ff name. - MISC-TYPES.EXPAND.* need the expand-* methods defined at macroexpansion-time. M ./tests/misc-types.lisp -5 +6 Wed Feb 15 11:16:31 CST 2006 James Bielman <jamesjb(a)jamesjb.com> * Change +REQUIRED-DLL-VERSION+ to *REQUIRED-DLL-VERSION*. M ./tests/bindings.lisp -3 +3 Wed Feb 15 03:13:37 CST 2006 James Bielman <jamesjb(a)jamesjb.com> * Remove completed TODO item about building 32 and 64 bit libtest. M ./TODO -2 Wed Feb 15 10:35:51 CST 2006 Luis Oliveira <loliveira(a)common-lisp.net> * A few more tests - Tests for the new macroexpansion-time type translator interface: MISC-TYPES.EXPAND.[1234] - New tests: DEREF.LONG-LONG and DEREF.UNSIGNED-LONG-LONG. - Tests for mem-ref with non-constant type arguments: DEREF.NONCONST.* - Fix some comments in tests/memory.lisp. - New test: MISC-TYPES.BOOLEAN.2 (accepting typedefs to integer types) M ./tests/libtest.c +3 M ./tests/memory.lisp -16 +140 M ./tests/misc-types.lisp -3 +66 An updated tarball of CFFI's source can be downloaded here: http://common-lisp.net/project/cffi/tarballs/cffi-060215.tar.gz Darcsweb URL: http://common-lisp.net/cgi-bin/darcsweb/darcsweb.cgi?r=cffi;a=summary

1 0

[cffi-devel] New patches: 14-Feb-2006
by loliveira＠common-lisp.net 15 Feb '06

15 Feb '06

Tue Feb 14 21:01:52 CST 2006 Luis Oliveira <loliveira(a)common-lisp.net> * Minor test changes - Use long instead of int in the big C functions in libtest.c. - Mark callbacks.bff.[12] as expect failures for a couple of lisps. M ./tests/callbacks.lisp -2 +12 M ./tests/libtest.c -31 +31 Tue Feb 14 20:59:06 CST 2006 Luis Oliveira <loliveira(a)common-lisp.net> * Make (%callback 'non-existant-callback) signal an error - Make %callback signal an error for non-existing callbacks. (allegro, sbcl and openmcl) - Regression test: callbacks.non-existant M ./src/cffi-allegro.lisp -1 +2 M ./src/cffi-openmcl.lisp -1 +2 M ./src/cffi-sbcl.lisp -1 +2 M ./tests/callbacks.lisp +6 Tue Feb 14 20:52:57 CST 2006 James Bielman <jamesjb(a)jamesjb.com> * Implement %MEM-REF and %MEM-SET compiler macros for Allegro CL. M ./src/cffi-allegro.lisp +20 Tue Feb 14 20:26:51 CST 2006 James Bielman <jamesjb(a)jamesjb.com> * Ignore RETTYPE in Allegro CL %DEFCALLBACK. M ./src/cffi-allegro.lisp +1 An updated tarball of CFFI's source can be downloaded here: http://common-lisp.net/project/cffi/tarballs/cffi-060214.tar.gz Darcsweb URL: http://common-lisp.net/cgi-bin/darcsweb/darcsweb.cgi?r=cffi;a=summary

1 0

[cffi-devel] New patches: 13-Feb-2006
by loliveira＠common-lisp.net 14 Feb '06

14 Feb '06

Mon Feb 13 22:29:28 CST 2006 Luis Oliveira <loliveira(a)common-lisp.net> * Fix float varargs promotion. - Convert floats to doubles in foreign-funcall-varargs. Test DEFCFUN.VARARGS.FLOAT now passes. M ./src/functions.lisp -4 +9 Mon Feb 13 22:27:33 CST 2006 Luis Oliveira <loliveira(a)common-lisp.net> * Preliminary random tester. - Push random-tester.lisp. This was used to generate the BFF tests, but is still not automated. - Update respective TODO item. M ./TODO -6 +1 A ./tests/random-tester.lisp Mon Feb 13 21:28:30 CST 2006 Luis Oliveira <loliveira(a)common-lisp.net> * MORE TESTS - New tests: callbacks.funcall.2, callbacks.bff.[12], defcfun.bff.[12]. - Re-enable defcfun.varargs.double. - Use #'float instead of #'coerce in the defcfun.varargs.* tests. - Fix dll version in libtest.c. M ./tests/callbacks.lisp -2 +136 M ./tests/defcfun.lisp -5 +125 M ./tests/libtest.c -1 +222 Mon Feb 13 18:28:39 CST 2006 Luis Oliveira <loliveira(a)common-lisp.net> * Change defconstant to defparameter in bindings.lisp M ./tests/bindings.lisp -1 +3 Mon Feb 13 18:19:31 CST 2006 Luis Oliveira <loliveira(a)common-lisp.net> * Macroexpansion-time translators - New interface. expand-to-foreign-dyn, expand-to-foreign and expand-from-foreign. - defcfun/foreign-funcall/defcallback/defcvar use this as well as the compiler macros and setf-expanders for mem-ref, mem-aref, and foreign-slot-value. - Also, parse ignore declarations and use them to avoid translating such types (thus avoiding breaking the ignore declaration). M ./src/foreign-vars.lisp -13 +7 M ./src/functions.lisp -46 +44 M ./src/package.lisp -1 +4 M ./src/types.lisp -40 +118 Sun Feb 12 12:27:09 CST 2006 Jack Unrue <jdunrue(a)gmail.com> * Implement simple version check for libtest DLL M ./tests/bindings.lisp +9 M ./tests/libtest.c +2 An updated tarball of CFFI's source can be downloaded here: http://common-lisp.net/project/cffi/tarballs/cffi-060213.tar.gz Darcsweb URL: http://common-lisp.net/cgi-bin/darcsweb/darcsweb.cgi?r=cffi;a=summary

1 0

Re: [cffi-devel] wanted: testsuite .dll
by Hoehle, Joerg-Cyril 14 Feb '06

14 Feb '06

jamesjb(a)jamesjb.com writes: >It would be nice to figure out how to get the ASDF system to compile >the DLL on Windows. That would be a bad idea. Presumably most users of CFFI on MS-Windows don't have MS-VC (or Mingw). But IMHO all users should be entitled to run a testsuite. > I'm not sure I'd want to include the DLL in the >repository as the risk of it not getting updated when the testsuite is >updated is somewhat high [...] IMHO not really a problem, because the .dll will stabilize, while the testsuite will continue to grow. Strict solution: add a version number (foreign-var) to the .dll and check it when loading the testsuite. Then people will know when it is outdated and hope for a new one from the CFFI site. As I said, presumably the .dll will receive seldom updates. Regards, Jorg Hohle.

4 11