1 2012-03-22 David Edelsohn <dje.gcc@gmail.com>
3 Backport from mainline:
4 2012-03-09 David Edelsohn <dje.gcc@gmail.com>
6 * src/powerpc/aix_closure.S (ffi_closure_ASM): Adjust for Darwin64
7 change to return value of ffi_closure_helper_DARWIN and load type
10 From Tom Honermann <tom.honermann@oracle.com>:
11 * src/powerpc/aix.S: Declare .ffi_prep_args. Insert nops after
13 * src/powerpc/aix_closure.S: Declare .ffi_closure_helper_DARWIN.
15 2012-03-22 Release Manager
19 2012-02-27 Mikael Pettersson <mikpe@it.uu.se>
22 * Makefile.am (FLAGS_TO_PASS): Define.
23 * Makefile.in: Regenerate.
25 2012-02-23 Kai Tietz <ktietz@redhat.com>
28 * src/x86/ffi.c (ffi_closure_raw_THISCALL): New
30 (ffi_prep_raw_closure_loc): Use ffi_closure_raw_THISCALL for
32 (ffi_raw_call): Use ffi_prep_args_raw.
33 * src/x86/win32.S (ffi_closure_raw_THISCALL): Add
34 implementation for stub.
36 2012-02-13 Kai Tietz <ktietz@redhat.com>
39 * src/x86/ffi.c (ffi_prep_raw_closure_loc): Add thiscall
40 support for X86_WIN32.
41 (FFI_INIT_TRAMPOLINE_THISCALL): Fix displacement.
43 2012-02-11 Eric Botcazou <ebotcazou@adacore.com>
45 * src/sparc/v9.S (STACKFRAME): Bump to 176.
47 2012-02-10 Kai Tietz <ktietz@redhat.com>
49 * configure.ac (AM_LTLDFLAGS): Add -no-undefine for x64
51 * configure: Regenerated.
53 2012-02-08 Kai Tietz <ktietz@redhat.com>
55 * src/prep_cif.c (ffi_prep_cif): Allow for X86_WIN32
57 * src/x86/ffi.c (ffi_closure_THISCALL): Add prototype.
58 (FFI_INIT_TRAMPOLINE_THISCALL): New trampoline code.
59 (ffi_prep_closure_loc): Add FFI_THISCALL support.
60 * src/x86/ffitarget.h (FFI_TRAMPOLINE_SIZE): Adjust size.
61 * src/x86/win32.S (ffi_closure_THISCALL): New closure code
62 for thiscall-calling convention.
63 * testsuite/libffi.call/closure_thiscall.c: New test.
65 2012-01-28 Kai Tietz <ktietz@redhat.com>
67 * src/libffi/src/x86/ffi.c (ffi_call_win32): Add new
68 argument to prototype for specify calling-convention.
69 (ffi_call): Add support for stdcall/thiscall convention.
70 (ffi_prep_args): Likewise.
71 (ffi_raw_call): Likewise.
72 * src/x86/ffitarget.h (ffi_abi): Add FFI_THISCALL and
74 * src/x86/win32.S (_ffi_call_win32): Add support for
75 fastcall/thiscall calling-convention calls.
76 * testsuite/libffi.call/fastthis1_win32.c: New test.
77 * testsuite/libffi.call/fastthis2_win32.c: New test.
78 * testsuite/libffi.call/fastthis3_win32.c: New test.
79 * testsuite/libffi.call/strlen2_win32.c: New test.
80 * testsuite/libffi.call/many2_win32.c: New test.
81 * testsuite/libffi.call/struct1_win32.c: New test.
82 * testsuite/libffi.call/struct2_win32.c: New test.
84 2012-01-23 Andreas Schwab <schwab@linux-m68k.org>
86 * src/m68k/sysv.S (ffi_call_SYSV): Properly test for plain
87 mc68000. Test for __HAVE_68881__ in addition to __MC68881__.
89 2012-01-19 Jakub Jelinek <jakub@redhat.com>
91 PR rtl-optimization/48496
92 * src/ia64/ffi.c (ffi_call): Fix up aliasing violations.
94 2012-01-09 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
96 * configure.ac (i?86-*-*): Set TARGET to X86_64.
97 * configure: Regenerate.
99 2011-12-07 Andrew Pinski <apinski@cavium.com>
102 * src/mips/n32.S: Add ".set mips4".
104 2011-11-21 Andreas Tobler <andreast@fgznet.ch>
106 * configure: Regenerate.
108 2011-11-10 Richard Henderson <rth@redhat.com>
110 * configure.ac (GCC_AS_CFI_PSEUDO_OP): Use it instead of inline check.
111 * configure, aclocal.m4: Rebuild.
113 2011-09-04 Iain Sandoe <iains@gcc.gnu.org>
116 * src/powerpc/darwin_closure.S (stubs): Make the stub binding
117 helper reference track the architecture pointer size.
119 2011-08-25 Andrew Haley <aph@redhat.com>
121 * src/arm/ffi.c (FFI_INIT_TRAMPOLINE): Remove hard-coded assembly
123 * src/arm/sysv.S (ffi_arm_trampoline): Put them here instead.
125 2011-07-11 Andrew Haley <aph@redhat.com>
127 * src/arm/ffi.c (FFI_INIT_TRAMPOLINE): Clear icache.
129 2011-06-29 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
131 * testsuite/libffi.call/cls_double_va.c: Move PR number to comment.
132 * testsuite/libffi.call/cls_longdouble_va.c: Likewise.
134 2011-06-29 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
137 * testsuite/libffi.call/cls_double_va.c: xfail dg-output on
139 * testsuite/libffi.call/cls_longdouble_va.c: Likewise.
141 2011-06-14 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
143 * testsuite/libffi.call/huge_struct.c (test_large_fn): Use PRIu8,
144 PRId8 instead of %hhu, %hhd.
145 * testsuite/libffi.call/ffitest.h [__alpha__ && __osf__] (PRId8,
147 [__sgi__] (PRId8, PRIu8): Define.
149 2011-04-29 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
151 * src/alpha/osf.S (UA_SI, FDE_ENCODING, FDE_ENCODE, FDE_ARANGE):
153 Use them to handle ELF vs. ECOFF differences.
154 [__osf__] (_GLOBAL__F_ffi_call_osf): Define.
156 2011-02-13 Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
158 * configure: Regenerate.
160 2011-02-09 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
163 * testsuite/libffi.call/cls_pointer.c (main): Cast void * to
165 * testsuite/libffi.call/cls_pointer_stack.c (main): Likewise.
167 2011-02-07 Joel Sherrill <joel.sherrill@oarcorp.com>
169 * libffi/src/m68k/ffi.c: Add RTEMS support for cache flushing.
170 Handle case when CPU variant does not have long double support.
171 * libffi/src/m68k/sysv.S: Add support for mc68000, Coldfire,
172 and cores with soft floating point.
174 2011-02-07 Joel Sherrill <joel.sherrill@oarcorp.com>
176 * configure.ac: Add mips*-*-rtems* support.
177 * configure: Regenerate.
178 * src/mips/ffitarget.h: Ensure needed constants are available
179 for targets which do not have sgidefs.h.
181 2011-01-26 Dave Korn <dave.korn.cygwin@gmail.com>
184 * configure.ac (AM_LTLDFLAGS): Add -bindir option for windows DLLs.
185 * configure: Regenerate.
187 2010-12-18 Iain Sandoe <iains@gcc.gnu.org>
191 * src/powerpc/darwin_closure.S: Provide Darwin64 implementation,
193 * src/powerpc/ffitarget.h (POWERPC_DARWIN64): New,
194 (FFI_TRAMPOLINE_SIZE): Update for Darwin64.
195 * src/powerpc/darwin.S: Provide Darwin64 implementation,
197 * src/powerpc/ffi_darwin.c: Likewise.
199 2010-12-06 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
201 * configure.ac (libffi_cv_as_ascii_pseudo_op): Use double
203 (libffi_cv_as_string_pseudo_op): Likewise.
204 * configure: Regenerate.
206 2010-12-03 Chung-Lin Tang <cltang@codesourcery.com>
208 * src/arm/sysv.S (ffi_closure_SYSV): Add UNWIND to .pad directive.
209 (ffi_closure_VFP): Same.
210 (ffi_call_VFP): Move down to before ffi_closure_VFP. Add '.fpu vfp'
213 2010-12-01 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
215 * testsuite/libffi.call/ffitest.h [__sgi] (PRId64, PRIu64): Define.
218 2010-11-29 Richard Henderson <rth@redhat.com>
219 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
221 * src/x86/sysv.S (FDE_ENCODING, FDE_ENCODE): Define.
222 (.eh_frame): Use FDE_ENCODING.
223 (.LASFDE1, .LASFDE2, LASFDE3): Simplify with FDE_ENCODE.
225 2010-11-15 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
227 * testsuite/lib/libffi-dg.exp: Rename ...
228 * testsuite/lib/libffi.exp: ... to this.
229 * libffi/testsuite/libffi.call/call.exp: Don't load libffi-dg.exp.
230 * libffi/testsuite/libffi.special/special.exp: Likewise.
232 2010-10-28 Chung-Lin Tang <cltang@codesourcery.com>
234 * src/arm/ffi.c (ffi_prep_args): Add VFP register argument handling
235 code, new parameter, and return value. Update comments.
236 (ffi_prep_cif_machdep): Add case for VFP struct return values. Add
237 call to layout_vfp_args().
238 (ffi_call_SYSV): Update declaration.
239 (ffi_call_VFP): New declaration.
240 (ffi_call): Add VFP struct return conditions. Call ffi_call_VFP()
242 (ffi_closure_VFP): New declaration.
243 (ffi_closure_SYSV_inner): Add new vfp_args parameter, update call to
244 ffi_prep_incoming_args_SYSV().
245 (ffi_prep_incoming_args_SYSV): Update parameters. Add VFP argument
247 (ffi_prep_closure_loc): Pass ffi_closure_VFP to trampoline
248 construction under VFP hard-float.
249 (rec_vfp_type_p): New function.
251 (place_vfp_arg): Same.
252 (layout_vfp_args): Same.
253 * src/arm/ffitarget.h (ffi_abi): Add FFI_VFP. Define FFI_DEFAULT_ABI
254 based on __ARM_PCS_VFP.
255 (FFI_EXTRA_CIF_FIELDS): Define for adding VFP hard-float specific
257 (FFI_TYPE_STRUCT_VFP_FLOAT): Define internally used type code.
258 (FFI_TYPE_STRUCT_VFP_DOUBLE): Same.
259 * src/arm/sysv.S (ffi_call_SYSV): Change call of ffi_prep_args() to
260 direct call. Move function pointer load upwards.
261 (ffi_call_VFP): New function.
262 (ffi_closure_VFP): Same.
264 * testsuite/lib/libffi-dg.exp (check-flags): New function.
265 (dg-skip-if): New function.
266 * testsuite/libffi.call/cls_double_va.c: Skip if target is arm*-*-*
267 and compiler options include -mfloat-abi=hard.
268 * testsuite/libffi.call/cls_longdouble_va.c: Same.
270 2010-10-01 Jakub Jelinek <jakub@redhat.com>
273 * src/x86/ffi64.c (ffi_prep_cif_machdep): Ensure cif->bytes is
275 * testsuite/libffi.call/many2.c: New test.
277 2010-08-20 Mark Wielaard <mjw@redhat.com>
279 * src/closures.c (open_temp_exec_file_mnt): Check if getmntent_r
282 2010-08-09 Andreas Tobler <andreast@fgznet.ch>
284 * configure.ac: Add target powerpc64-*-freebsd*.
285 * configure: Regenerate.
286 * testsuite/libffi.call/cls_align_longdouble_split.c: Pass
287 -mlong-double-128 only to linux targets.
288 * testsuite/libffi.call/cls_align_longdouble_split2.c: Likewise.
289 * testsuite/libffi.call/cls_longdouble.c: Likewise.
290 * testsuite/libffi.call/huge_struct.c: Likewise.
292 2010-07-10 Evan Phoenix <evan@fallingsnow.net>
294 * src/closures.c (selinux_enabled_check): Fix strncmp usage bug.
296 2010-07-07 Dan Horák <dan@danny.cz>
298 * include/ffi.h.in: Protect #define with #ifndef.
299 * src/powerpc/ffitarget.h: Ditto.
300 * src/s390/ffitarget.h: Ditto.
301 * src/sparc/ffitarget.h: Ditto.
303 2010-07-07 Neil Roberts <neil@linux.intel.com>
305 * src/x86/sysv.S (ffi_call_SYSV): Align the stack pointer to
308 2010-07-02 Jakub Jelinek <jakub@redhat.com>
310 * Makefile.am (AM_MAKEFLAGS): Pass also mandir to submakes.
311 * Makefile.in: Regenerated.
313 2010-05-19 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
315 * configure.ac (libffi_cv_as_x86_pcrel): Check for illegal in as
317 (libffi_cv_as_ascii_pseudo_op): Check for .ascii.
318 (libffi_cv_as_string_pseudo_op): Check for .string.
319 * configure: Regenerate.
320 * fficonfig.h.in: Regenerate.
321 * src/x86/sysv.S (.eh_frame): Use .ascii, .string or error.
323 2010-05-05 Michael Kohler <michaelkohler@live.com>
325 * src/dlmalloc.c (dlfree): Fix spelling.
326 * src/ia64/ffi.c (ffi_prep_cif_machdep): Ditto.
327 * configure.ac: Ditto.
328 * configure: Rebuilt.
330 2010-04-13 Dan Witte <dwitte@mozilla.com>
332 * msvcc.sh: Build with -W3 instead of -Wall.
333 * src/powerpc/ffi_darwin.c: Remove build warnings.
334 * src/x86/ffi.c: Ditto.
335 * src/x86/ffitarget.h: Ditto.
337 2010-04-12 Dan Witte <dwitte@mozilla.com>
338 Walter Meinl <wuno@lsvw.de>
340 * configure.ac: Add OS/2 support.
341 * configure: Rebuilt.
342 * src/closures.c: Ditto.
343 * src/dlmalloc.c: Ditto.
344 * src/x86/win32.S: Ditto.
346 2010-04-07 Jakub Jelinek <jakub@redhat.com>
348 * testsuite/libffi.call/err_bad_abi.c: Remove unused args variable.
350 2010-04-02 Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
352 * Makefile.in: Regenerate.
353 * aclocal.m4: Regenerate.
354 * include/Makefile.in: Regenerate.
355 * man/Makefile.in: Regenerate.
356 * testsuite/Makefile.in: Regenerate.
358 2010-03-15 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
360 * configure.ac (libffi_cv_as_x86_64_unwind_section_type): New test.
361 * configure: Regenerate.
362 * fficonfig.h.in: Regenerate.
363 * libffi/src/x86/unix64.S (.eh_frame)
364 [HAVE_AS_X86_64_UNWIND_SECTION_TYPE]: Use @unwind section type.
366 2010-03-14 Matthias Klose <doko@ubuntu.com>
368 * src/x86/ffi64.c: Fix typo in comment.
369 * src/x86/ffi.c: Use /* ... */ comment style.
371 2010-02-24 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
373 * doc/libffi.texi (The Closure API): Fix typo.
374 * doc/libffi.info: Remove.
376 2010-02-15 Matthias Klose <doko@ubuntu.com>
378 * src/arm/sysv.S (__ARM_ARCH__): Define for processor
381 2010-01-15 Anthony Green <green@redhat.com>
383 * README: Add notes on building with Microsoft Visual C++.
385 2010-01-15 Daniel Witte <dwitte@mozilla.com>
387 * msvcc.sh: New file.
389 * src/x86/win32.S: Port assembly routines to MSVC and #ifdef.
390 * src/x86/ffi.c: Tweak function declaration and remove excess
392 * include/ffi.h.in: Add __declspec(align(8)) to typedef struct
395 * src/x86/ffi.c: Merge ffi_call_SYSV and ffi_call_STDCALL into new
396 function ffi_call_win32 on X86_WIN32.
397 * src/x86/win32.S (ffi_call_SYSV): Rename to ffi_call_win32.
398 (ffi_call_STDCALL): Remove.
400 * src/prep_cif.c (ffi_prep_cif): Move stack space allocation code
401 to ffi_prep_cif_machdep for x86.
402 * src/x86/ffi.c (ffi_prep_cif_machdep): To here.
404 2010-01-15 Oliver Kiddle <okiddle@yahoo.co.uk>
406 * src/x86/ffitarget.h (ffi_abi): Check for __i386 and __amd64 for
407 Sun Studio compiler compatibility.
409 2010-01-12 Conrad Irwin <conrad.irwin@gmail.com>
411 * doc/libffi.texi: Add closure example.
413 2010-01-07 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
416 * testsuite/libffi.call/ffitest.h [__alpha__ && __osf__] (PRIdLL,
417 PRIuLL, PRId64, PRIu64, PRIuPTR): Define.
418 * testsuite/libffi.call/cls_align_sint64.c: Add -Wno-format on
420 * testsuite/libffi.call/cls_align_uint64.c: Likewise.
421 * testsuite/libffi.call/cls_ulonglong.c: Likewise.
422 * testsuite/libffi.call/return_ll1.c: Likewise.
423 * testsuite/libffi.call/stret_medium2.c: Likewise.
424 * testsuite/libffi.special/ffitestcxx.h (allocate_mmap): Cast
425 MAP_FAILED to char *.
427 2010-01-06 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
429 * src/mips/n32.S: Use .abicalls and .eh_frame with __GNUC__.
431 2009-12-31 Anthony Green <green@redhat.com>
433 * README: Update for libffi 3.0.9.
435 2009-12-27 Matthias Klose <doko@ubuntu.com>
437 * configure.ac (HAVE_LONG_DOUBLE): Define for mips when
439 * configure: Rebuilt.
441 2009-12-26 Anthony Green <green@redhat.com>
443 * testsuite/libffi.call/cls_longdouble_va.c: Mark as xfail for
445 * testsuite/libffi.call/cls_double_va.c: Ditto.
447 2009-12-26 Andreas Tobler <a.tobler@schweiz.org>
449 * testsuite/libffi.call/ffitest.h: Conditionally include stdint.h
451 * testsuite/libffi.special/unwindtest.cc: Ditto.
453 2009-12-26 Andreas Tobler <a.tobler@schweiz.org>
455 * configure.ac: Add amd64-*-openbsd*.
456 * configure: Rebuilt.
457 * testsuite/lib/libffi-dg.exp (libffi_target_compile): Link
458 openbsd programs with -lpthread.
460 2009-12-26 Anthony Green <green@redhat.com>
462 * testsuite/libffi.call/cls_double_va.c,
463 testsuite/libffi.call/cls_longdouble.c,
464 testsuite/libffi.call/cls_longdouble_va.c,
465 testsuite/libffi.call/cls_pointer.c,
466 testsuite/libffi.call/cls_pointer_stack.c: Remove xfail for
467 mips*-*-* and arm*-*-*.
468 * testsuite/libffi.call/cls_align_longdouble_split.c,
469 testsuite/libffi.call/cls_align_longdouble_split2.c,
470 testsuite/libffi.call/stret_medium2.c,
471 testsuite/libffi.call/stret_medium.c,
472 testsuite/libffi.call/stret_large.c,
473 testsuite/libffi.call/stret_large2.c: Remove xfail for arm*-*-*.
475 2009-12-31 Kay Tietz <ktietz70@googlemail.com>
477 * testsuite/libffi.call/ffitest.h,
478 testsuite/libffi.special/ffitestcxx.h (PRIdLL, PRuLL): Fix
481 2009-12-31 Carlo Bramini <carlo.bramix@libero.it>
483 * configure.ac (AM_LTLDFLAGS): Define for windows hosts.
484 * Makefile.am (libffi_la_LDFLAGS): Add AM_LTLDFLAGS.
485 * configure: Rebuilt.
486 * Makefile.in: Rebuilt.
488 2009-12-31 Anthony Green <green@redhat.com>
491 * testsuite/libffi.call/huge_struct.c: New test case from Blake
494 2009-12-28 David Edelsohn <edelsohn@gnu.org>
496 * src/powerpc/ffi_darwin.c (ffi_prep_args): Copy abi and nargs to
498 (aix_adjust_aggregate_sizes): New function.
499 (ffi_prep_cif_machdep): Call it.
501 2009-12-26 Andreas Tobler <a.tobler@schweiz.org>
503 * configure.ac: Define FFI_MMAP_EXEC_WRIT for the given targets.
504 * configure: Regenerate.
505 * fficonfig.h.in: Likewise.
506 * src/closures.c: Remove the FFI_MMAP_EXEC_WRIT definition for
509 2009-12-26 Andreas Schwab <schwab@linux-m68k.org>
511 * src/powerpc/ffi.c (ffi_prep_args_SYSV): Advance intarg_count
512 when a float arguments is passed in memory.
513 (ffi_closure_helper_SYSV): Mark general registers as used up when
514 a 64bit or soft-float long double argument is passed in memory.
516 2009-12-25 Matthias Klose <doko@ubuntu.com>
518 * man/ffi_call.3: Fix #include in examples.
519 * doc/libffi.texi: Add dircategory.
521 2009-12-25 Frank Everdij <f.p.x.everdij@tudelft.nl>
523 * include/ffi.h.in: Placed '__GNUC__' ifdef around
524 '__attribute__((aligned(8)))' in ffi_closure, fixes compile for
526 * include/ffi_common.h: Added '__sgi' define to non
527 '__attribute__((__mode__()))' integer typedefs.
528 * src/mips/ffi.c (ffi_call, ffi_closure_mips_inner_O32,
529 ffi_closure_mips_inner_N32): Added 'defined(_MIPSEB)' to BE check.
530 (ffi_closure_mips_inner_O32, ffi_closure_mips_inner_N32): Added
531 FFI_LONGDOUBLE support and alignment(N32 only).
532 * src/mips/ffitarget.h: Corrected '#include <sgidefs.h>' for IRIX and
533 fixed non '__attribute__((__mode__()))' integer typedefs.
534 * src/mips/n32.S: Put '#ifdef linux' around '.abicalls' and '.eh_frame'
535 since they are Linux/GNU Assembler specific.
537 2009-12-25 Bradley Smith <brad@brad-smith.co.uk>
539 * configure.ac, Makefile.am, src/avr32/ffi.c,
540 src/avr32/ffitarget.h,
541 src/avr32/sysv.S: Add AVR32 port.
542 * configure, Makefile.in: Rebuilt.
544 2009-12-21 Andreas Tobler <a.tobler@schweiz.org>
546 * configure.ac: Make i?86 build on FreeBSD and OpenBSD.
547 * configure: Regenerate.
549 2009-12-15 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
551 * testsuite/libffi.call/ffitest.h: Define PRIuPTR on PA HP-UX.
553 2009-12-13 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
555 * src/pa/ffi.c (ffi_closure_inner_pa32): Handle FFI_TYPE_LONGDOUBLE
558 2009-12-11 Eric Botcazou <ebotcazou@adacore.com>
560 * src/sparc/ffi.c (ffi_closure_sparc_inner_v9): Properly align 'long
563 2009-12-11 Eric Botcazou <ebotcazou@adacore.com>
565 * testsuite/libffi.call/ffitest.h: Define PRIuPTR on Solaris < 10.
567 2009-12-10 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
570 * src/closures.c [X86_64 && __sun__ && __svr4__]
571 (FFI_MMAP_EXEC_WRIT): Define.
573 2009-12-08 David Daney <ddaney@caviumnetworks.com>
575 * testsuite/libffi.call/stret_medium.c: Remove xfail for mips*-*-*
576 * testsuite/libffi.call/cls_align_longdouble_split2.c: Same.
577 * testsuite/libffi.call/stret_large.c: Same.
578 * testsuite/libffi.call/cls_align_longdouble_split.c: Same.
579 * testsuite/libffi.call/stret_large2.c: Same.
580 * testsuite/libffi.call/stret_medium2.c: Same.
582 2009-12-07 David Edelsohn <edelsohn@gnu.org>
584 * src/powerpc/aix_closure.S (libffi_closure_ASM): Fix tablejump
587 2009-12-05 David Edelsohn <edelsohn@gnu.org>
589 * src/powerpc/aix.S: Update AIX32 code to be consistent with AIX64
591 * src/powerpc/aix_closure.S: Same.
593 2009-12-05 Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
595 * Makefile.in: Regenerate.
596 * configure: Regenerate.
597 * include/Makefile.in: Regenerate.
598 * man/Makefile.in: Regenerate.
599 * testsuite/Makefile.in: Regenerate.
601 2009-12-04 David Edelsohn <edelsohn@gnu.org>
603 * src/powerpc/aix_closure.S: Reorganize 64-bit code to match
606 2009-12-04 Uros Bizjak <ubizjak@gmail.com>
609 * src/x86/ffi64.c (classify_argument): Update from
610 gcc/config/i386/i386.c.
611 (ffi_closure_unix64_inner): Do not use the address of two consecutive
612 SSE registers directly.
613 * testsuite/libffi.call/cls_dbls_struct.c (main): Remove xfail
614 for x86_64 linux targets.
616 2009-12-04 David Edelsohn <edelsohn@gnu.org>
618 * src/powerpc/ffi_darwin.c (ffi_closure_helper_DARWIN): Increment
619 pfr for long double split between fpr13 and stack.
621 2009-12-03 David Edelsohn <edelsohn@gnu.org>
623 * src/powerpc/ffi_darwin.c (ffi_prep_args): Increment next_arg and
624 fparg_count twice for long double.
626 2009-12-03 David Edelsohn <edelsohn@gnu.org>
629 * src/powerpc/ffi_darwin.c (ffi_prep_args): Remove extra parentheses.
631 2009-12-03 Uros Bizjak <ubizjak@gmail.com>
633 * testsuite/libffi.call/cls_longdouble_va.c (main): Fix format string.
634 Remove xfails for x86 linux targets.
636 2009-12-02 David Edelsohn <edelsohn@gnu.org>
638 * src/powerpc/ffi_darwin.c (ffi_prep_args): Fix typo in INT64
641 2009-12-01 David Edelsohn <edelsohn@gnu.org>
643 * src/powerpc/aix.S (ffi_call_AIX): Convert to more standard
644 register usage. Call ffi_prep_args directly. Add long double
645 return value support.
646 * src/powerpc/ffi_darwin.c (ffi_prep_args): Double arg increment
647 applies to FFI_TYPE_DOUBLE. Correct fpr_base increment typo.
648 Separate FFI_TYPE_SINT32 and FFI_TYPE_UINT32 cases.
649 (ffi_prep_cif_machdep): Only 16 byte stack alignment in 64 bit
651 (ffi_closure_helper_DARWIN): Remove nf and ng counters. Move temp
653 * src/powerpc/aix_closure.S: Maintain 16 byte stack alignment.
654 Allocate result area between params and FPRs.
656 2009-11-30 David Edelsohn <edelsohn@gnu.org>
659 * src/powerpc/ffitarget.h (POWERPC64): Define for PPC64 Linux and
661 * src/powerpc/aix.S: Implement AIX64 version.
662 * src/powerpc/aix_closure.S: Implement AIX64 version.
663 (ffi_closure_ASM): Use extsb, lha and displament addresses.
664 * src/powerpc/ffi_darwin.c (ffi_prep_args): Implement AIX64
666 (ffi_prep_cif_machdep): Same.
668 (ffi_closure_helper_DARWIN): Same.
670 2009-11-02 Andreas Tobler <a.tobler@schweiz.org>
673 * testsuite/libffi.call/testclosure.c: New test.
675 2009-09-28 Kai Tietz <kai.tietz@onevision.com>
677 * src/x86/win64.S (_ffi_call_win64 stack): Remove for gnu
678 assembly version use of ___chkstk.
680 2009-09-23 Matthias Klose <doko@ubuntu.com>
682 PR libffi/40242, PR libffi/41443
683 * src/arm/sysv.S (__ARM_ARCH__): Define for processors
684 __ARM_ARCH_6T2__, __ARM_ARCH_6M__, __ARM_ARCH_7__,
685 __ARM_ARCH_7A__, __ARM_ARCH_7R__, __ARM_ARCH_7M__.
686 Change the conditionals to __SOFTFP__ || __ARM_EABI__
687 for -mfloat-abi=softfp to work.
689 2009-09-17 Loren J. Rittle <ljrittle@acm.org>
691 PR testsuite/32843 (strikes again)
692 * src/x86/ffi.c (ffi_prep_cif_machdep): Add X86_FREEBSD to
693 enable proper extension on char and short.
695 2009-09-15 David Daney <ddaney@caviumnetworks.com>
697 * src/java_raw_api.c (ffi_java_raw_to_rvalue): Remove special
698 handling for FFI_TYPE_POINTER.
699 * src/mips/ffitarget.h (FFI_TYPE_STRUCT_D_SOFT,
700 FFI_TYPE_STRUCT_F_SOFT, FFI_TYPE_STRUCT_DD_SOFT,
701 FFI_TYPE_STRUCT_FF_SOFT, FFI_TYPE_STRUCT_FD_SOFT,
702 FFI_TYPE_STRUCT_DF_SOFT, FFI_TYPE_STRUCT_SOFT): New defines.
703 (FFI_N32_SOFT_FLOAT, FFI_N64_SOFT_FLOAT): New ffi_abi enumerations.
704 (enum ffi_abi): Set FFI_DEFAULT_ABI for soft-float.
705 * src/mips/n32.S (ffi_call_N32): Add handling for soft-float
706 structure and pointer returns.
707 (ffi_closure_N32): Add handling for pointer returns.
708 * src/mips/ffi.c (ffi_prep_args, calc_n32_struct_flags,
709 calc_n32_return_struct_flags): Handle soft-float.
710 (ffi_prep_cif_machdep): Handle soft-float, fix pointer handling.
711 (ffi_call_N32): Declare proper argument types.
712 (ffi_call, copy_struct_N32, ffi_closure_mips_inner_N32): Handle
715 2009-08-24 Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
717 * configure.ac (AC_PREREQ): Bump to 2.64.
719 2009-08-22 Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
721 * Makefile.am (install-html, install-pdf): Remove.
722 * Makefile.in: Regenerate.
724 * Makefile.in: Regenerate.
725 * aclocal.m4: Regenerate.
726 * configure: Regenerate.
727 * fficonfig.h.in: Regenerate.
728 * include/Makefile.in: Regenerate.
729 * man/Makefile.in: Regenerate.
730 * testsuite/Makefile.in: Regenerate.
732 2009-07-30 Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
734 * configure.ac (_AC_ARG_VAR_PRECIOUS): Use m4_rename_force.
736 2009-07-24 Dave Korn <dave.korn.cygwin@gmail.com>
739 * src/x86/ffi.c (ffi_prep_cif_machdep): Also use sign/zero-extending
740 return types for X86_WIN32.
741 * src/x86/win32.S (_ffi_call_SYSV): Handle omitted return types.
742 (_ffi_call_STDCALL, _ffi_closure_SYSV, _ffi_closure_raw_SYSV,
743 _ffi_closure_STDCALL): Likewise.
745 * src/closures.c (is_selinux_enabled): Define to const 0 for Cygwin.
746 (dlmmap, dlmunmap): Also use these functions on Cygwin.
748 2009-07-11 Richard Sandiford <rdsandiford@googlemail.com>
753 * testsuite/lib/libffi-dg.exp: Revert 2009-07-02, 2009-07-01 and
756 2009-07-01 Richard Sandiford <r.sandiford@uk.ibm.com>
758 * testsuite/lib/libffi-dg.exp (libffi-init): Set ld_library_path
759 to "" before adding paths. (This reinstates an assignment that
760 was removed by my 2009-06-30 commit, but changes the initial
761 value from "." to "".)
763 2009-07-01 H.J. Lu <hongjiu.lu@intel.com>
766 * testsuite/lib/libffi-dg.exp (libffi-init): Properly set
767 gccdir. Adjust ld_library_path for gcc only if gccdir isn't
770 2009-06-30 Richard Sandiford <r.sandiford@uk.ibm.com>
772 * testsuite/lib/libffi-dg.exp (libffi-init): Don't add "."
773 to ld_library_path. Use add_path. Add just find_libgcc_s
774 to ld_library_path, not every libgcc multilib directory.
776 2009-06-16 Wim Lewis <wiml@hhhh.org>
778 * src/powerpc/ffi.c: Avoid clobbering cr3 and cr4, which are
779 supposed to be callee-saved.
780 * src/powerpc/sysv.S (small_struct_return_value): Fix overrun of
781 return buffer for odd-size structs.
783 2009-06-16 Andreas Tobler <a.tobler@schweiz.org>
786 * testsuite/lib/libffi-dg.exp (libffi_target_compile): Add
787 allow_stack_execute for Darwin.
789 2009-06-16 Andrew Haley <aph@redhat.com>
791 * configure.ac (TARGETDIR): Add missing blank lines.
792 * configure: Regenerate.
794 2009-06-16 Andrew Haley <aph@redhat.com>
796 * testsuite/libffi.call/cls_align_sint64.c,
797 testsuite/libffi.call/cls_align_uint64.c,
798 testsuite/libffi.call/cls_longdouble_va.c,
799 testsuite/libffi.call/cls_ulonglong.c,
800 testsuite/libffi.call/return_ll1.c,
801 testsuite/libffi.call/stret_medium2.c: Fix printf format
803 * testsuite/libffi.call/ffitest.h,
804 testsuite/libffi.special/ffitestcxx.h (PRIdLL, PRIuLL): Define.
806 2009-06-15 Andrew Haley <aph@redhat.com>
808 * testsuite/libffi.call/err_bad_typedef.c: xfail everywhere.
809 * testsuite/libffi.call/err_bad_abi.c: Likewise.
811 2009-06-12 Andrew Haley <aph@redhat.com>
813 * Makefile.am: Remove info_TEXINFOS.
815 2009-06-12 Andrew Haley <aph@redhat.com>
817 * ChangeLog.libffi: testsuite/libffi.call/cls_align_sint64.c,
818 testsuite/libffi.call/cls_align_uint64.c,
819 testsuite/libffi.call/cls_ulonglong.c,
820 testsuite/libffi.call/return_ll1.c,
821 testsuite/libffi.call/stret_medium2.c: Fix printf format
823 testsuite/libffi.special/unwindtest.cc: include stdint.h.
825 2009-06-11 Timothy Wall <twall@users.sf.net>
830 include/ffi_common.h,
835 src/x86/win64.S (new),
836 README: Added win64 support (mingw or MSVC)
840 testsuite/Makefile.in,
842 aclocal.m4: Regenerated
843 * ltcf-c.sh: properly escape cygwin/w32 path
844 * man/ffi_call.3: Clarify size requirements for return value.
845 * src/x86/ffi64.c: Fix filename in comment.
846 * src/x86/win32.S: Remove unused extern.
848 * testsuite/libffi.call/closure_fn0.c,
849 testsuite/libffi.call/closure_fn1.c,
850 testsuite/libffi.call/closure_fn2.c,
851 testsuite/libffi.call/closure_fn3.c,
852 testsuite/libffi.call/closure_fn4.c,
853 testsuite/libffi.call/closure_fn5.c,
854 testsuite/libffi.call/closure_fn6.c,
855 testsuite/libffi.call/closure_stdcall.c,
856 testsuite/libffi.call/cls_12byte.c,
857 testsuite/libffi.call/cls_16byte.c,
858 testsuite/libffi.call/cls_18byte.c,
859 testsuite/libffi.call/cls_19byte.c,
860 testsuite/libffi.call/cls_1_1byte.c,
861 testsuite/libffi.call/cls_20byte.c,
862 testsuite/libffi.call/cls_20byte1.c,
863 testsuite/libffi.call/cls_24byte.c,
864 testsuite/libffi.call/cls_2byte.c,
865 testsuite/libffi.call/cls_3_1byte.c,
866 testsuite/libffi.call/cls_3byte1.c,
867 testsuite/libffi.call/cls_3byte2.c,
868 testsuite/libffi.call/cls_4_1byte.c,
869 testsuite/libffi.call/cls_4byte.c,
870 testsuite/libffi.call/cls_5_1_byte.c,
871 testsuite/libffi.call/cls_5byte.c,
872 testsuite/libffi.call/cls_64byte.c,
873 testsuite/libffi.call/cls_6_1_byte.c,
874 testsuite/libffi.call/cls_6byte.c,
875 testsuite/libffi.call/cls_7_1_byte.c,
876 testsuite/libffi.call/cls_7byte.c,
877 testsuite/libffi.call/cls_8byte.c,
878 testsuite/libffi.call/cls_9byte1.c,
879 testsuite/libffi.call/cls_9byte2.c,
880 testsuite/libffi.call/cls_align_double.c,
881 testsuite/libffi.call/cls_align_float.c,
882 testsuite/libffi.call/cls_align_longdouble.c,
883 testsuite/libffi.call/cls_align_longdouble_split.c,
884 testsuite/libffi.call/cls_align_longdouble_split2.c,
885 testsuite/libffi.call/cls_align_pointer.c,
886 testsuite/libffi.call/cls_align_sint16.c,
887 testsuite/libffi.call/cls_align_sint32.c,
888 testsuite/libffi.call/cls_align_sint64.c,
889 testsuite/libffi.call/cls_align_uint16.c,
890 testsuite/libffi.call/cls_align_uint32.c,
891 testsuite/libffi.call/cls_align_uint64.c,
892 testsuite/libffi.call/cls_dbls_struct.c,
893 testsuite/libffi.call/cls_double.c,
894 testsuite/libffi.call/cls_double_va.c,
895 testsuite/libffi.call/cls_float.c,
896 testsuite/libffi.call/cls_longdouble.c,
897 testsuite/libffi.call/cls_longdouble_va.c,
898 testsuite/libffi.call/cls_multi_schar.c,
899 testsuite/libffi.call/cls_multi_sshort.c,
900 testsuite/libffi.call/cls_multi_sshortchar.c,
901 testsuite/libffi.call/cls_multi_uchar.c,
902 testsuite/libffi.call/cls_multi_ushort.c,
903 testsuite/libffi.call/cls_multi_ushortchar.c,
904 testsuite/libffi.call/cls_pointer.c,
905 testsuite/libffi.call/cls_pointer_stack.c,
906 testsuite/libffi.call/cls_schar.c,
907 testsuite/libffi.call/cls_sint.c,
908 testsuite/libffi.call/cls_sshort.c,
909 testsuite/libffi.call/cls_uchar.c,
910 testsuite/libffi.call/cls_uint.c,
911 testsuite/libffi.call/cls_ulonglong.c,
912 testsuite/libffi.call/cls_ushort.c,
913 testsuite/libffi.call/err_bad_abi.c,
914 testsuite/libffi.call/err_bad_typedef.c,
915 testsuite/libffi.call/float2.c,
916 testsuite/libffi.call/huge_struct.c,
917 testsuite/libffi.call/nested_struct.c,
918 testsuite/libffi.call/nested_struct1.c,
919 testsuite/libffi.call/nested_struct10.c,
920 testsuite/libffi.call/nested_struct2.c,
921 testsuite/libffi.call/nested_struct3.c,
922 testsuite/libffi.call/nested_struct4.c,
923 testsuite/libffi.call/nested_struct5.c,
924 testsuite/libffi.call/nested_struct6.c,
925 testsuite/libffi.call/nested_struct7.c,
926 testsuite/libffi.call/nested_struct8.c,
927 testsuite/libffi.call/nested_struct9.c,
928 testsuite/libffi.call/problem1.c,
929 testsuite/libffi.call/return_ldl.c,
930 testsuite/libffi.call/return_ll1.c,
931 testsuite/libffi.call/stret_large.c,
932 testsuite/libffi.call/stret_large2.c,
933 testsuite/libffi.call/stret_medium.c,
934 testsuite/libffi.call/stret_medium2.c,
935 testsuite/libffi.special/unwindtest.cc: use ffi_closure_alloc instead
936 of checking for MMAP. Use intptr_t instead of long casts.
938 2009-06-11 Kaz Kojima <kkojima@gcc.gnu.org>
940 * testsuite/libffi.call/cls_longdouble_va.c: Add xfail sh*-*-linux-*.
941 * testsuite/libffi.call/err_bad_abi.c: Add xfail sh*-*-*.
942 * testsuite/libffi.call/err_bad_typedef.c: Likewise.
944 2009-06-09 Andrew Haley <aph@redhat.com>
946 * src/x86/freebsd.S: Add missing file.
948 2009-06-08 Andrew Haley <aph@redhat.com>
950 Import from libffi 3.0.8:
952 * doc/libffi.texi: New file.
953 * doc/libffi.info: Likewise.
954 * doc/stamp-vti: Likewise.
955 * man/Makefile.am: New file.
956 * man/ffi_call.3: New file.
958 * Makefile.am (EXTRA_DIST): Add src/x86/darwin64.S,
960 (nodist_libffi_la_SOURCES): Add X86_FREEBSD.
962 * configure.ac: Bump version to 3.0.8.
963 parisc*-*-linux*: Add.
964 i386-*-freebsd* | i386-*-openbsd*: Add.
965 powerpc-*-beos*: Add.
966 AM_CONDITIONAL X86_FREEBSD: Add.
967 AC_CONFIG_FILES: Add man/Makefile.
969 * include/ffi.h.in (FFI_FN): Change void (*)() to void (*)(void).
971 2009-06-08 Andrew Haley <aph@redhat.com>
973 * README: Import from libffi 3.0.8.
975 2009-06-08 Andrew Haley <aph@redhat.com>
977 * testsuite/libffi.call/err_bad_abi.c: Add xfails.
978 * testsuite/libffi.call/cls_longdouble_va.c: Add xfails.
979 * testsuite/libffi.call/cls_dbls_struct.c: Add xfail x86_64-*-linux-*.
980 * testsuite/libffi.call/err_bad_typedef.c: Add xfails.
982 * testsuite/libffi.call/stret_medium2.c: Add __UNUSED__ to args.
983 * testsuite/libffi.call/stret_medium.c: Likewise.
984 * testsuite/libffi.call/stret_large2.c: Likewise.
985 * testsuite/libffi.call/stret_large.c: Likewise.
987 2008-12-26 Timothy Wall <twall@users.sf.net>
989 * testsuite/libffi.call/cls_longdouble.c,
990 testsuite/libffi.call/cls_longdouble_va.c,
991 testsuite/libffi.call/cls_align_longdouble.c,
992 testsuite/libffi.call/cls_align_longdouble_split.c,
993 testsuite/libffi.call/cls_align_longdouble_split2.c: mark expected
994 failures on x86_64 cygwin/mingw.
996 2008-12-22 Timothy Wall <twall@users.sf.net>
998 * testsuite/libffi.call/closure_fn0.c,
999 testsuite/libffi.call/closure_fn1.c,
1000 testsuite/libffi.call/closure_fn2.c,
1001 testsuite/libffi.call/closure_fn3.c,
1002 testsuite/libffi.call/closure_fn4.c,
1003 testsuite/libffi.call/closure_fn5.c,
1004 testsuite/libffi.call/closure_fn6.c,
1005 testsuite/libffi.call/closure_loc_fn0.c,
1006 testsuite/libffi.call/closure_stdcall.c,
1007 testsuite/libffi.call/cls_align_pointer.c,
1008 testsuite/libffi.call/cls_pointer.c,
1009 testsuite/libffi.call/cls_pointer_stack.c: use portable cast from
1010 pointer to integer (intptr_t).
1011 * testsuite/libffi.call/cls_longdouble.c: disable for win64.
1013 2008-07-24 Anthony Green <green@redhat.com>
1015 * testsuite/libffi.call/cls_dbls_struct.c,
1016 testsuite/libffi.call/cls_double_va.c,
1017 testsuite/libffi.call/cls_longdouble.c,
1018 testsuite/libffi.call/cls_longdouble_va.c,
1019 testsuite/libffi.call/cls_pointer.c,
1020 testsuite/libffi.call/cls_pointer_stack.c,
1021 testsuite/libffi.call/err_bad_abi.c: Clean up failures from
1024 2008-03-04 Anthony Green <green@redhat.com>
1028 * testsuite/libffi.call/cls_align_longdouble_split2.c
1029 testsuite/libffi.call/cls_align_longdouble_split.c
1030 testsuite/libffi.call/cls_dbls_struct.c
1031 testsuite/libffi.call/cls_double_va.c
1032 testsuite/libffi.call/cls_longdouble.c
1033 testsuite/libffi.call/cls_longdouble_va.c
1034 testsuite/libffi.call/cls_pointer.c
1035 testsuite/libffi.call/cls_pointer_stack.c
1036 testsuite/libffi.call/err_bad_abi.c
1037 testsuite/libffi.call/err_bad_typedef.c
1038 testsuite/libffi.call/stret_large2.c
1039 testsuite/libffi.call/stret_large.c
1040 testsuite/libffi.call/stret_medium2.c
1041 testsuite/libffi.call/stret_medium.c: New tests from Apple.
1043 2009-06-05 Andrew Haley <aph@redhat.com>
1045 * src/x86/ffitarget.h, src/x86/ffi.c: Merge stdcall changes from
1048 2009-06-04 Andrew Haley <aph@redhat.com>
1050 * src/x86/ffitarget.h, src/x86/win32.S, src/x86/ffi.c: Back out
1053 2008-02-26 Anthony Green <green@redhat.com>
1054 Thomas Heller <theller@ctypes.org>
1056 * src/x86/ffi.c (ffi_closure_SYSV_inner): Change C++ comment to C
1059 2008-02-03 Timothy Wall <twall@users.sf.net>
1061 * src/x86/ffi.c (FFI_INIT_TRAMPOLINE_STDCALL): Calculate jump return
1062 offset based on code pointer, not data pointer.
1064 2008-01-31 Timothy Wall <twall@users.sf.net>
1066 * testsuite/libffi.call/closure_stdcall.c: Add test for stdcall
1068 * src/x86/ffitarget.h: Increase size of trampoline for stdcall
1070 * src/x86/win32.S: Add assembly for stdcall closure.
1071 * src/x86/ffi.c: Initialize stdcall closure trampoline.
1073 2009-06-04 Andrew Haley <aph@redhat.com>
1075 * include/ffi.h.in: Change void (*)() to void (*)(void).
1076 * src/x86/ffi.c: Likewise.
1078 2009-06-04 Andrew Haley <aph@redhat.com>
1080 * src/powerpc/ppc_closure.S: Insert licence header.
1081 * src/powerpc/linux64_closure.S: Likewise.
1082 * src/m68k/sysv.S: Likewise.
1084 * src/sh64/ffi.c: Change void (*)() to void (*)(void).
1085 * src/powerpc/ffi.c: Likewise.
1086 * src/powerpc/ffi_darwin.c: Likewise.
1087 * src/m32r/ffi.c: Likewise.
1088 * src/sh64/ffi.c: Likewise.
1089 * src/x86/ffi64.c: Likewise.
1090 * src/alpha/ffi.c: Likewise.
1091 * src/alpha/osf.S: Likewise.
1092 * src/frv/ffi.c: Likewise.
1093 * src/s390/ffi.c: Likewise.
1094 * src/pa/ffi.c: Likewise.
1095 * src/pa/hpux32.S: Likewise.
1096 * src/ia64/unix.S: Likewise.
1097 * src/ia64/ffi.c: Likewise.
1098 * src/sparc/ffi.c: Likewise.
1099 * src/mips/ffi.c: Likewise.
1100 * src/sh/ffi.c: Likewise.
1102 2008-02-15 David Daney <ddaney@avtrex.com>
1104 * src/mips/ffi.c (USE__BUILTIN___CLEAR_CACHE):
1105 Define (conditionally), and use it to include cachectl.h.
1106 (ffi_prep_closure_loc): Fix cache flushing.
1107 * src/mips/ffitarget.h (_ABIN32, _ABI64, _ABIO32): Define.
1109 2009-06-04 Andrew Haley <aph@redhat.com>
1112 src/arm/ffitarget.h,
1115 src/powerpc/ffitarget.h,
1117 src/sh64/ffitarget.h,
1122 src/x86/ffitarget.h,
1128 src/alpha/ffitarget.h,
1131 src/m68k/ffitarget.h,
1132 src/frv/ffitarget.h,
1134 src/s390/ffitarget.h,
1136 src/cris/ffitarget.h,
1141 src/ia64/ffitarget.h,
1144 src/ia64/ia64_flags.h,
1148 src/sparc/ffitarget.h,
1151 src/mips/ffitarget.h,
1158 src/sh/sysv.S: Update license text.
1160 2009-05-22 Dave Korn <dave.korn.cygwin@gmail.com>
1162 * src/x86/win32.S (_ffi_closure_STDCALL): New function.
1163 (.eh_frame): Add FDE for it.
1165 2009-05-22 Dave Korn <dave.korn.cygwin@gmail.com>
1167 * configure.ac: Also check if assembler supports pc-relative
1168 relocs on X86_WIN32 targets.
1169 * configure: Regenerate.
1170 * src/x86/win32.S (ffi_prep_args): Declare extern, not global.
1171 (_ffi_call_SYSV): Add missing function type symbol .def and
1172 add EH markup labels.
1173 (_ffi_call_STDCALL): Likewise.
1174 (_ffi_closure_SYSV): Likewise.
1175 (_ffi_closure_raw_SYSV): Likewise.
1176 (.eh_frame): Add hand-crafted EH data.
1178 2009-04-09 Jakub Jelinek <jakub@redhat.com>
1180 * testsuite/lib/libffi-dg.exp: Change copyright header to refer to
1181 version 3 of the GNU General Public License and to point readers
1182 at the COPYING3 file and the FSF's license web page.
1183 * testsuite/libffi.call/call.exp: Likewise.
1184 * testsuite/libffi.special/special.exp: Likewise.
1186 2009-03-01 Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
1188 * configure: Regenerate.
1190 2008-12-18 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
1193 * configure.ac (HAVE_AS_X86_PCREL): New test.
1194 * configure: Regenerate.
1195 * fficonfig.h.in: Regenerate.
1196 * src/x86/sysv.S [!FFI_NO_RAW_API]: Precalculate
1197 RAW_CLOSURE_CIF_OFFSET, RAW_CLOSURE_FUN_OFFSET,
1198 RAW_CLOSURE_USER_DATA_OFFSET for the Solaris 10/x86 assembler.
1199 (.eh_frame): Only use SYMBOL-. iff HAVE_AS_X86_PCREL.
1200 * src/x86/unix64.S (.Lstore_table): Move to .text section.
1201 (.Lload_table): Likewise.
1202 (.eh_frame): Only use SYMBOL-. iff HAVE_AS_X86_PCREL.
1204 2008-12-18 Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
1206 * configure: Regenerate.
1208 2008-11-21 Eric Botcazou <ebotcazou@adacore.com>
1210 * src/sparc/ffi.c (ffi_prep_cif_machdep): Add support for
1211 signed/unsigned int8/16 return values.
1212 * src/sparc/v8.S (ffi_call_v8): Likewise.
1213 (ffi_closure_v8): Likewise.
1215 2008-09-26 Peter O'Gorman <pogma@thewrittenword.com>
1216 Steve Ellcey <sje@cup.hp.com>
1218 * configure: Regenerate for new libtool.
1219 * Makefile.in: Ditto.
1220 * include/Makefile.in: Ditto.
1221 * aclocal.m4: Ditto.
1223 2008-08-25 Andreas Tobler <a.tobler@schweiz.org>
1225 * src/powerpc/ffitarget.h (ffi_abi): Add FFI_LINUX and
1226 FFI_LINUX_SOFT_FLOAT to the POWERPC_FREEBSD enum.
1227 Add note about flag bits used for FFI_SYSV_TYPE_SMALL_STRUCT.
1228 Adjust copyright notice.
1229 * src/powerpc/ffi.c: Add two new flags to indicate if we have one
1230 register or two register to use for FFI_SYSV structs.
1231 (ffi_prep_cif_machdep): Pass the right register flag introduced above.
1232 (ffi_closure_helper_SYSV): Fix the return type for
1233 FFI_SYSV_TYPE_SMALL_STRUCT. Comment.
1234 Adjust copyright notice.
1236 2008-07-16 Kaz Kojima <kkojima@gcc.gnu.org>
1238 * src/sh/ffi.c (ffi_prep_closure_loc): Turn INSN into an unsigned
1241 2008-06-17 Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
1243 * configure: Regenerate.
1244 * include/Makefile.in: Regenerate.
1245 * testsuite/Makefile.in: Regenerate.
1247 2008-06-07 Joseph Myers <joseph@codesourcery.com>
1249 * configure.ac (parisc*-*-linux*, powerpc-*-sysv*,
1250 powerpc-*-beos*): Remove.
1251 * configure: Regenerate.
1253 2008-05-09 Julian Brown <julian@codesourcery.com>
1255 * Makefile.am (LTLDFLAGS): New.
1256 (libffi_la_LDFLAGS): Use above.
1257 * Makefile.in: Regenerate.
1259 2008-04-18 Paolo Bonzini <bonzini@gnu.org>
1262 * aclocal.m4: Regenerate.
1263 * configure: Regenerate.
1265 2008-03-26 Kaz Kojima <kkojima@gcc.gnu.org>
1267 * src/sh/sysv.S: Add .note.GNU-stack on Linux.
1268 * src/sh64/sysv.S: Likewise.
1270 2008-03-26 Daniel Jacobowitz <dan@debian.org>
1272 * src/arm/sysv.S: Fix ARM comment marker.
1274 2008-03-26 Jakub Jelinek <jakub@redhat.com>
1276 * src/alpha/osf.S: Add .note.GNU-stack on Linux.
1277 * src/s390/sysv.S: Likewise.
1278 * src/powerpc/ppc_closure.S: Likewise.
1279 * src/powerpc/sysv.S: Likewise.
1280 * src/x86/unix64.S: Likewise.
1281 * src/x86/sysv.S: Likewise.
1282 * src/sparc/v8.S: Likewise.
1283 * src/sparc/v9.S: Likewise.
1284 * src/m68k/sysv.S: Likewise.
1285 * src/arm/sysv.S: Likewise.
1287 2008-03-16 Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
1289 * aclocal.m4: Regenerate.
1290 * configure: Likewise.
1291 * Makefile.in: Likewise.
1292 * include/Makefile.in: Likewise.
1293 * testsuite/Makefile.in: Likewise.
1295 2008-02-12 Bjoern Koenig <bkoenig@alpha-tierchen.de>
1296 Andreas Tobler <a.tobler@schweiz.org>
1298 * configure.ac: Add amd64-*-freebsd* target.
1299 * configure: Regenerate.
1301 2008-01-30 H.J. Lu <hongjiu.lu@intel.com>
1304 * src/x86/sysv.S (ffi_closure_SYSV): Pop 4 byte from stack when
1307 * testsuite/libffi.call/call.exp: Add "-O2 -fomit-frame-pointer"
1310 2008-01-24 David Edelsohn <edelsohn@gnu.org>
1312 * configure: Regenerate.
1314 2008-01-06 Andreas Tobler <a.tobler@schweiz.org>
1316 * src/x86/ffi.c (ffi_prep_cif_machdep): Fix thinko.
1318 2008-01-05 Andreas Tobler <a.tobler@schweiz.org>
1321 * src/x86/ffi.c (ffi_prep_cif_machdep): Add code for
1322 signed/unsigned int8/16 for X86_DARWIN.
1323 Updated copyright info.
1324 Handle one and two byte structs with special cif->flags.
1325 * src/x86/ffitarget.h: Add special types for one and two byte structs.
1326 Updated copyright info.
1327 * src/x86/darwin.S (ffi_call_SYSV): Rewrite to use a jump table like
1329 Remove code to pop args from the stack after call.
1330 Special-case signed/unsigned for int8/16, one and two byte structs.
1331 (ffi_closure_raw_SYSV): Handle FFI_TYPE_UINT8,
1332 FFI_TYPE_SINT8, FFI_TYPE_UINT16, FFI_TYPE_SINT16, FFI_TYPE_UINT32,
1334 Updated copyright info.
1336 2007-12-08 David Daney <ddaney@avtrex.com>
1338 * src/mips/n32.S (ffi_call_N32): Replace dadd with ADDU, dsub with
1339 SUBU, add with ADDU and use smaller code sequences.
1341 2007-12-07 David Daney <ddaney@avtrex.com>
1343 * src/mips/ffi.c (ffi_prep_cif_machdep): Handle long double return
1346 2007-12-06 David Daney <ddaney@avtrex.com>
1348 * include/ffi.h.in (FFI_SIZEOF_JAVA_RAW): Define if not already
1350 (ffi_java_raw): New typedef.
1351 (ffi_java_raw_call, ffi_java_ptrarray_to_raw,
1352 ffi_java_raw_to_ptrarray): Change parameter types from ffi_raw to
1354 (ffi_java_raw_closure) : Same.
1355 (ffi_prep_java_raw_closure, ffi_prep_java_raw_closure_loc): Change
1357 * src/java_raw_api.c (ffi_java_raw_size): Replace FFI_SIZEOF_ARG with
1358 FFI_SIZEOF_JAVA_RAW.
1359 (ffi_java_raw_to_ptrarray): Change type of raw to ffi_java_raw.
1360 Replace FFI_SIZEOF_ARG with FFI_SIZEOF_JAVA_RAW. Use
1361 sizeof(ffi_java_raw) for alignment calculations.
1362 (ffi_java_ptrarray_to_raw): Same.
1363 (ffi_java_rvalue_to_raw): Add special handling for FFI_TYPE_POINTER
1364 if FFI_SIZEOF_JAVA_RAW == 4.
1365 (ffi_java_raw_to_rvalue): Same.
1366 (ffi_java_raw_call): Change type of raw to ffi_java_raw.
1367 (ffi_java_translate_args): Same.
1368 (ffi_prep_java_raw_closure_loc, ffi_prep_java_raw_closure): Change
1370 * src/mips/ffitarget.h (FFI_SIZEOF_JAVA_RAW): Define for N32 ABI.
1372 2007-12-06 David Daney <ddaney@avtrex.com>
1374 * src/mips/n32.S (ffi_closure_N32): Use 64-bit add instruction on
1377 2007-12-01 Andreas Tobler <a.tobler@schweiz.org>
1380 * src/powerpc/ffitarget.h: Introduce new ABI FFI_LINUX_SOFT_FLOAT.
1381 Add local FFI_TYPE_UINT128 to handle soft-float long-double-128.
1382 * src/powerpc/ffi.c: Distinguish between __NO_FPRS__ and not and
1383 set the NUM_FPR_ARG_REGISTERS according to.
1384 Add support for potential soft-float support under hard-float
1386 (ffi_prep_args_SYSV): Set NUM_FPR_ARG_REGISTERS to 0 in case of
1387 FFI_LINUX_SOFT_FLOAT, handle float, doubles and long-doubles according
1388 to the FFI_LINUX_SOFT_FLOAT ABI.
1389 (ffi_prep_cif_machdep): Likewise.
1390 (ffi_closure_helper_SYSV): Likewise.
1391 * src/powerpc/ppc_closure.S: Make sure not to store float/double
1392 on archs where __NO_FPRS__ is true.
1393 Add FFI_TYPE_UINT128 support.
1394 * src/powerpc/sysv.S: Add support for soft-float long-double-128.
1395 Adjust copyright notice.
1397 2007-11-25 Andreas Tobler <a.tobler@schweiz.org>
1399 * src/closures.c: Move defintion of MAYBE_UNUSED from here to ...
1400 * include/ffi_common.h: ... here.
1403 2007-11-17 Andreas Tobler <a.tobler@schweiz.org>
1405 * src/powerpc/sysv.S: Load correct cr to compare if we have long double.
1406 * src/powerpc/linux64.S: Likewise.
1407 * src/powerpc/ffi.c: Add a comment to show which part goes into cr6.
1408 * testsuite/libffi.call/return_ldl.c: New test.
1410 2007-09-04 <aph@redhat.com>
1412 * src/arm/sysv.S (UNWIND): New.
1413 (Whole file): Conditionally compile unwinder directives.
1414 * src/arm/sysv.S: Add unwinder directives.
1416 * src/arm/ffi.c (ffi_prep_args): Align structs by at least 4 bytes.
1417 Only treat r0 as a struct address if we're actually returning a
1419 Only copy the bytes that are actually within a struct.
1420 (ffi_prep_cif_machdep): A Composite Type not larger than 4 bytes
1421 is returned in r0, not passed by address.
1422 (ffi_call): Allocate a word-sized temporary for the case where
1423 a composite is returned in r0.
1424 (ffi_prep_incoming_args_SYSV): Align as necessary.
1426 2007-08-05 Steven Newbury <s_j_newbury@yahoo.co.uk>
1428 * src/arm/ffi.c (FFI_INIT_TRAMPOLINE): Use __clear_cache instead of
1429 directly using the sys_cacheflush syscall.
1431 2007-07-27 Andrew Haley <aph@redhat.com>
1433 * src/arm/sysv.S (ffi_closure_SYSV): Add soft-float.
1435 2007-09-03 Maciej W. Rozycki <macro@linux-mips.org>
1437 * Makefile.am: Unify MIPS_IRIX and MIPS_LINUX into MIPS.
1438 * configure.ac: Likewise.
1439 * Makefile.in: Regenerate.
1440 * include/Makefile.in: Likewise.
1441 * testsuite/Makefile.in: Likewise.
1442 * configure: Likewise.
1444 2007-08-24 David Daney <ddaney@avtrex.com>
1446 * testsuite/libffi.call/return_sl.c: New test.
1448 2007-08-10 David Daney <ddaney@avtrex.com>
1450 * testsuite/libffi.call/cls_multi_ushort.c,
1451 testsuite/libffi.call/cls_align_uint16.c,
1452 testsuite/libffi.call/nested_struct1.c,
1453 testsuite/libffi.call/nested_struct3.c,
1454 testsuite/libffi.call/cls_7_1_byte.c,
1455 testsuite/libffi.call/nested_struct5.c,
1456 testsuite/libffi.call/cls_double.c,
1457 testsuite/libffi.call/nested_struct7.c,
1458 testsuite/libffi.call/cls_sint.c,
1459 testsuite/libffi.call/nested_struct9.c,
1460 testsuite/libffi.call/cls_20byte1.c,
1461 testsuite/libffi.call/cls_multi_sshortchar.c,
1462 testsuite/libffi.call/cls_align_sint64.c,
1463 testsuite/libffi.call/cls_3byte2.c,
1464 testsuite/libffi.call/cls_multi_schar.c,
1465 testsuite/libffi.call/cls_multi_uchar.c,
1466 testsuite/libffi.call/cls_19byte.c,
1467 testsuite/libffi.call/cls_9byte1.c,
1468 testsuite/libffi.call/cls_align_float.c,
1469 testsuite/libffi.call/closure_fn1.c,
1470 testsuite/libffi.call/problem1.c,
1471 testsuite/libffi.call/closure_fn3.c,
1472 testsuite/libffi.call/cls_sshort.c,
1473 testsuite/libffi.call/closure_fn5.c,
1474 testsuite/libffi.call/cls_align_double.c,
1475 testsuite/libffi.call/nested_struct.c,
1476 testsuite/libffi.call/cls_2byte.c,
1477 testsuite/libffi.call/nested_struct10.c,
1478 testsuite/libffi.call/cls_4byte.c,
1479 testsuite/libffi.call/cls_6byte.c,
1480 testsuite/libffi.call/cls_8byte.c,
1481 testsuite/libffi.call/cls_multi_sshort.c,
1482 testsuite/libffi.call/cls_align_sint16.c,
1483 testsuite/libffi.call/cls_align_uint32.c,
1484 testsuite/libffi.call/cls_20byte.c,
1485 testsuite/libffi.call/cls_float.c,
1486 testsuite/libffi.call/nested_struct2.c,
1487 testsuite/libffi.call/cls_5_1_byte.c,
1488 testsuite/libffi.call/nested_struct4.c,
1489 testsuite/libffi.call/cls_24byte.c,
1490 testsuite/libffi.call/nested_struct6.c,
1491 testsuite/libffi.call/cls_64byte.c,
1492 testsuite/libffi.call/nested_struct8.c,
1493 testsuite/libffi.call/cls_uint.c,
1494 testsuite/libffi.call/cls_multi_ushortchar.c,
1495 testsuite/libffi.call/cls_schar.c,
1496 testsuite/libffi.call/cls_uchar.c,
1497 testsuite/libffi.call/cls_align_uint64.c,
1498 testsuite/libffi.call/cls_ulonglong.c,
1499 testsuite/libffi.call/cls_align_longdouble.c,
1500 testsuite/libffi.call/cls_1_1byte.c,
1501 testsuite/libffi.call/cls_12byte.c,
1502 testsuite/libffi.call/cls_3_1byte.c,
1503 testsuite/libffi.call/cls_3byte1.c,
1504 testsuite/libffi.call/cls_4_1byte.c,
1505 testsuite/libffi.call/cls_6_1_byte.c,
1506 testsuite/libffi.call/cls_16byte.c,
1507 testsuite/libffi.call/cls_18byte.c,
1508 testsuite/libffi.call/closure_fn0.c,
1509 testsuite/libffi.call/cls_9byte2.c,
1510 testsuite/libffi.call/closure_fn2.c,
1511 testsuite/libffi.call/closure_fn4.c,
1512 testsuite/libffi.call/cls_ushort.c,
1513 testsuite/libffi.call/closure_fn6.c,
1514 testsuite/libffi.call/cls_5byte.c,
1515 testsuite/libffi.call/cls_align_pointer.c,
1516 testsuite/libffi.call/cls_7byte.c,
1517 testsuite/libffi.call/cls_align_sint32.c,
1518 testsuite/libffi.special/unwindtest_ffi_call.cc,
1519 testsuite/libffi.special/unwindtest.cc: Remove xfail for mips64*-*-*.
1521 2007-08-10 David Daney <ddaney@avtrex.com>
1524 * configure.ac: Don't treat mips64 as a special case.
1525 * Makefile.am (nodist_libffi_la_SOURCES): Add n32.S.
1526 * configure: Regenerate
1527 * Makefile.in: Ditto.
1528 * fficonfig.h.in: Ditto.
1529 * src/mips/ffitarget.h (REG_L, REG_S, SUBU, ADDU, SRL, LI): Indent.
1530 (LA, EH_FRAME_ALIGN, FDE_ADDR_BYTES): New preprocessor macros.
1531 (FFI_DEFAULT_ABI): Set for n64 case.
1532 (FFI_CLOSURES, FFI_TRAMPOLINE_SIZE): Define for n32 and n64 cases.
1533 * src/mips/n32.S (ffi_call_N32): Add debug macros and labels for FDE.
1534 (ffi_closure_N32): New function.
1535 (.eh_frame): New section
1536 * src/mips/o32.S: Clean up comments.
1537 (ffi_closure_O32): Pass ffi_closure parameter in $12.
1538 * src/mips/ffi.c: Use FFI_MIPS_N32 instead of
1539 _MIPS_SIM == _ABIN32 throughout.
1540 (FFI_MIPS_STOP_HERE): New, use in place of
1542 (ffi_prep_args): Use unsigned long to hold pointer values. Rewrite
1543 to support n32/n64 ABIs.
1544 (calc_n32_struct_flags): Rewrite.
1545 (calc_n32_return_struct_flags): Remove unused variable. Reverse
1546 position of flag bits.
1547 (ffi_prep_cif_machdep): Rewrite n32 portion.
1548 (ffi_call): Enable for n64. Add special handling for small structure
1550 (ffi_prep_closure_loc): Add n32 and n64 support.
1551 (ffi_closure_mips_inner_O32): Add cast to silence warning.
1552 (copy_struct_N32, ffi_closure_mips_inner_N32): New functions.
1554 2007-08-08 David Daney <ddaney@avtrex.com>
1556 * testsuite/libffi.call/ffitest.h (ffi_type_mylong): Remove definition.
1557 * testsuite/libffi.call/cls_align_uint16.c (main): Use correct type
1559 * testsuite/libffi.call/nested_struct1.c (main): Ditto.
1560 * testsuite/libffi.call/cls_sint.c (main): Ditto.
1561 * testsuite/libffi.call/nested_struct9.c (main): Ditto.
1562 * testsuite/libffi.call/cls_20byte1.c (main): Ditto.
1563 * testsuite/libffi.call/cls_9byte1.c (main): Ditto.
1564 * testsuite/libffi.call/closure_fn1.c (main): Ditto.
1565 * testsuite/libffi.call/closure_fn3.c (main): Ditto.
1566 * testsuite/libffi.call/return_dbl2.c (main): Ditto.
1567 * testsuite/libffi.call/cls_sshort.c (main): Ditto.
1568 * testsuite/libffi.call/return_fl3.c (main): Ditto.
1569 * testsuite/libffi.call/closure_fn5.c (main): Ditto.
1570 * testsuite/libffi.call/nested_struct.c (main): Ditto.
1571 * testsuite/libffi.call/nested_struct10.c (main): Ditto.
1572 * testsuite/libffi.call/return_ll1.c (main): Ditto.
1573 * testsuite/libffi.call/cls_8byte.c (main): Ditto.
1574 * testsuite/libffi.call/cls_align_uint32.c (main): Ditto.
1575 * testsuite/libffi.call/cls_align_sint16.c (main): Ditto.
1576 * testsuite/libffi.call/cls_20byte.c (main): Ditto.
1577 * testsuite/libffi.call/nested_struct2.c (main): Ditto.
1578 * testsuite/libffi.call/cls_24byte.c (main): Ditto.
1579 * testsuite/libffi.call/nested_struct6.c (main): Ditto.
1580 * testsuite/libffi.call/cls_uint.c (main): Ditto.
1581 * testsuite/libffi.call/cls_12byte.c (main): Ditto.
1582 * testsuite/libffi.call/cls_16byte.c (main): Ditto.
1583 * testsuite/libffi.call/closure_fn0.c (main): Ditto.
1584 * testsuite/libffi.call/cls_9byte2.c (main): Ditto.
1585 * testsuite/libffi.call/closure_fn2.c (main): Ditto.
1586 * testsuite/libffi.call/return_dbl1.c (main): Ditto.
1587 * testsuite/libffi.call/closure_fn4.c (main): Ditto.
1588 * testsuite/libffi.call/closure_fn6.c (main): Ditto.
1589 * testsuite/libffi.call/cls_align_sint32.c (main): Ditto.
1591 2007-08-07 Andrew Haley <aph@redhat.com>
1593 * src/x86/sysv.S (ffi_closure_raw_SYSV): Fix typo in previous
1596 2007-08-06 Andrew Haley <aph@redhat.com>
1599 * src/x86/sysv.S (ffi_closure_raw_SYSV): Handle FFI_TYPE_UINT8,
1600 FFI_TYPE_SINT8, FFI_TYPE_UINT16, FFI_TYPE_SINT16, FFI_TYPE_UINT32,
1603 2007-08-02 David Daney <ddaney@avtrex.com>
1605 * testsuite/libffi.call/return_ul.c (main): Define return type as
1606 ffi_arg. Use proper printf conversion specifier.
1608 2007-07-30 Andrew Haley <aph@redhat.com>
1611 * src/x86/ffi.c (ffi_prep_cif_machdep): in x86 case, add code for
1612 signed/unsigned int8/16.
1613 * src/x86/sysv.S (ffi_call_SYSV): Rewrite to:
1615 Remove code to pop args from the stack after call.
1616 Special-case signed/unsigned int8/16.
1617 * testsuite/libffi.call/return_sc.c (main): Revert.
1619 2007-07-26 Richard Guenther <rguenther@suse.de>
1622 * testsuite/libffi.call/return_sc.c (main): Verify call
1623 result as signed char, not ffi_arg.
1625 2007-07-16 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
1627 * configure.ac (i?86-*-solaris2.1[0-9]): Set TARGET to X86_64.
1628 * configure: Regenerate.
1630 2007-07-11 David Daney <ddaney@avtrex.com>
1632 * src/mips/ffi.c: Don't include sys/cachectl.h.
1633 (ffi_prep_closure_loc): Use __builtin___clear_cache() instead of
1636 2007-05-18 Aurelien Jarno <aurelien@aurel32.net>
1638 * src/arm/ffi.c (ffi_prep_closure_loc): Renamed and ajusted
1639 from (ffi_prep_closure): ... this.
1640 (FFI_INIT_TRAMPOLINE): Adjust.
1642 2005-12-31 Phil Blundell <pb@reciva.com>
1644 * src/arm/ffi.c (ffi_prep_incoming_args_SYSV,
1645 ffi_closure_SYSV_inner, ffi_prep_closure): New, add closure support.
1646 * src/arm/sysv.S(ffi_closure_SYSV): Likewise.
1647 * src/arm/ffitarget.h (FFI_TRAMPOLINE_SIZE): Likewise.
1648 (FFI_CLOSURES): Enable closure support.
1650 2007-07-03 Andrew Haley <aph@hedges.billgatliff.com>
1652 * testsuite/libffi.call/cls_multi_ushort.c,
1653 testsuite/libffi.call/cls_align_uint16.c,
1654 testsuite/libffi.call/nested_struct1.c,
1655 testsuite/libffi.call/nested_struct3.c,
1656 testsuite/libffi.call/cls_7_1_byte.c,
1657 testsuite/libffi.call/cls_double.c,
1658 testsuite/libffi.call/nested_struct5.c,
1659 testsuite/libffi.call/nested_struct7.c,
1660 testsuite/libffi.call/cls_sint.c,
1661 testsuite/libffi.call/nested_struct9.c,
1662 testsuite/libffi.call/cls_20byte1.c,
1663 testsuite/libffi.call/cls_multi_sshortchar.c,
1664 testsuite/libffi.call/cls_align_sint64.c,
1665 testsuite/libffi.call/cls_3byte2.c,
1666 testsuite/libffi.call/cls_multi_schar.c,
1667 testsuite/libffi.call/cls_multi_uchar.c,
1668 testsuite/libffi.call/cls_19byte.c,
1669 testsuite/libffi.call/cls_9byte1.c,
1670 testsuite/libffi.call/cls_align_float.c,
1671 testsuite/libffi.call/closure_fn1.c,
1672 testsuite/libffi.call/problem1.c,
1673 testsuite/libffi.call/closure_fn3.c,
1674 testsuite/libffi.call/cls_sshort.c,
1675 testsuite/libffi.call/closure_fn5.c,
1676 testsuite/libffi.call/cls_align_double.c,
1677 testsuite/libffi.call/cls_2byte.c,
1678 testsuite/libffi.call/nested_struct.c,
1679 testsuite/libffi.call/nested_struct10.c,
1680 testsuite/libffi.call/cls_4byte.c,
1681 testsuite/libffi.call/cls_6byte.c,
1682 testsuite/libffi.call/cls_8byte.c,
1683 testsuite/libffi.call/cls_multi_sshort.c,
1684 testsuite/libffi.call/cls_align_uint32.c,
1685 testsuite/libffi.call/cls_align_sint16.c,
1686 testsuite/libffi.call/cls_float.c,
1687 testsuite/libffi.call/cls_20byte.c,
1688 testsuite/libffi.call/cls_5_1_byte.c,
1689 testsuite/libffi.call/nested_struct2.c,
1690 testsuite/libffi.call/cls_24byte.c,
1691 testsuite/libffi.call/nested_struct4.c,
1692 testsuite/libffi.call/nested_struct6.c,
1693 testsuite/libffi.call/cls_64byte.c,
1694 testsuite/libffi.call/nested_struct8.c,
1695 testsuite/libffi.call/cls_uint.c,
1696 testsuite/libffi.call/cls_multi_ushortchar.c,
1697 testsuite/libffi.call/cls_schar.c,
1698 testsuite/libffi.call/cls_uchar.c,
1699 testsuite/libffi.call/cls_align_uint64.c,
1700 testsuite/libffi.call/cls_ulonglong.c,
1701 testsuite/libffi.call/cls_align_longdouble.c,
1702 testsuite/libffi.call/cls_1_1byte.c,
1703 testsuite/libffi.call/cls_12byte.c,
1704 testsuite/libffi.call/cls_3_1byte.c,
1705 testsuite/libffi.call/cls_3byte1.c,
1706 testsuite/libffi.call/cls_4_1byte.c,
1707 testsuite/libffi.call/cls_6_1_byte.c,
1708 testsuite/libffi.call/cls_16byte.c,
1709 testsuite/libffi.call/cls_18byte.c,
1710 testsuite/libffi.call/closure_fn0.c,
1711 testsuite/libffi.call/cls_9byte2.c,
1712 testsuite/libffi.call/closure_fn2.c,
1713 testsuite/libffi.call/closure_fn4.c,
1714 testsuite/libffi.call/cls_ushort.c,
1715 testsuite/libffi.call/closure_fn6.c,
1716 testsuite/libffi.call/cls_5byte.c,
1717 testsuite/libffi.call/cls_align_pointer.c,
1718 testsuite/libffi.call/cls_7byte.c,
1719 testsuite/libffi.call/cls_align_sint32.c,
1720 testsuite/libffi.special/unwindtest_ffi_call.cc,
1721 testsuite/libffi.special/unwindtest.cc: Enable for ARM.
1723 2007-07-05 H.J. Lu <hongjiu.lu@intel.com>
1725 * aclocal.m4: Regenerated.
1727 2007-06-02 Paolo Bonzini <bonzini@gnu.org>
1729 * configure: Regenerate.
1731 2007-05-23 Steve Ellcey <sje@cup.hp.com>
1733 * Makefile.in: Regenerate.
1734 * configure: Regenerate.
1735 * aclocal.m4: Regenerate.
1736 * include/Makefile.in: Regenerate.
1737 * testsuite/Makefile.in: Regenerate.
1739 2007-05-10 Roman Zippel <zippel@linux-m68k.org>
1741 * src/m68k/ffi.c (ffi_prep_incoming_args_SYSV,
1742 ffi_closure_SYSV_inner,ffi_prep_closure): New, add closure support.
1743 * src/m68k/sysv.S(ffi_closure_SYSV,ffi_closure_struct_SYSV): Likewise.
1744 * src/m68k/ffitarget.h (FFI_TRAMPOLINE_SIZE): Likewise.
1745 (FFI_CLOSURES): Enable closure support.
1747 2007-05-10 Roman Zippel <zippel@linux-m68k.org>
1749 * configure.ac (HAVE_AS_CFI_PSEUDO_OP): New test.
1750 * configure: Regenerate.
1751 * fficonfig.h.in: Regenerate.
1752 * src/m68k/sysv.S (CFI_STARTPROC,CFI_ENDPROC,
1753 CFI_OFFSET,CFI_DEF_CFA): New macros.
1754 (ffi_call_SYSV): Add callframe annotation.
1756 2007-05-10 Roman Zippel <zippel@linux-m68k.org>
1758 * src/m68k/ffi.c (ffi_prep_args,ffi_prep_cif_machdep): Fix
1759 numerous test suite failures.
1760 * src/m68k/sysv.S (ffi_call_SYSV): Likewise.
1762 2007-04-11 Paolo Bonzini <bonzini@gnu.org>
1764 * Makefile.am (EXTRA_DIST): Bring up to date.
1765 * Makefile.in: Regenerate.
1766 * src/frv/eabi.S: Remove RCS keyword.
1768 2007-04-06 Richard Henderson <rth@redhat.com>
1770 * configure.ac: Tidy target case.
1771 (HAVE_LONG_DOUBLE): Allow the target to override.
1772 * configure: Regenerate.
1773 * include/ffi.h.in: Don't define ffi_type_foo if
1774 LIBFFI_HIDE_BASIC_TYPES is defined.
1775 (ffi_type_longdouble): If not HAVE_LONG_DOUBLE, define
1777 * types.c (LIBFFI_HIDE_BASIC_TYPES): Define.
1778 (FFI_TYPEDEF, ffi_type_void): Mark the data const.
1779 (ffi_type_longdouble): Special case for Alpha. Don't define
1780 if long double == double.
1782 * src/alpha/ffi.c (FFI_TYPE_LONGDOUBLE): Assert unique value.
1783 (ffi_prep_cif_machdep): Handle it as the 128-bit type.
1784 (ffi_call, ffi_closure_osf_inner): Likewise.
1785 (ffi_closure_osf_inner): Likewise. Mark hidden.
1786 (ffi_call_osf, ffi_closure_osf): Mark hidden.
1787 * src/alpha/ffitarget.h (FFI_LAST_ABI): Tidy definition.
1788 * src/alpha/osf.S (ffi_call_osf, ffi_closure_osf): Mark hidden.
1789 (load_table): Handle 128-bit long double.
1791 * testsuite/libffi.call/float4.c: Add -mieee for alpha.
1793 2007-04-06 Tom Tromey <tromey@redhat.com>
1796 * README: Fixed bug in example.
1798 2007-04-03 Jakub Jelinek <jakub@redhat.com>
1800 * src/closures.c: Include sys/statfs.h.
1801 (_GNU_SOURCE): Define on Linux.
1802 (FFI_MMAP_EXEC_SELINUX): Define.
1803 (selinux_enabled): New variable.
1804 (selinux_enabled_check): New function.
1805 (is_selinux_enabled): Define.
1808 2007-03-24 Uros Bizjak <ubizjak@gmail.com>
1810 * testsuite/libffi.call/return_fl2.c (return_fl): Mark as static.
1811 Use 'volatile float sum' to create sum of floats to avoid false
1812 negative due to excess precision on ix86 targets.
1815 2007-03-08 Alexandre Oliva <aoliva@redhat.com>
1817 * src/powerpc/ffi.c (flush_icache): Fix left-over from previous
1819 (ffi_prep_closure_loc): Remove unneeded casts. Add needed ones.
1821 2007-03-07 Alexandre Oliva <aoliva@redhat.com>
1823 * include/ffi.h.in (ffi_closure_alloc, ffi_closure_free): New.
1824 (ffi_prep_closure_loc): New.
1825 (ffi_prep_raw_closure_loc): New.
1826 (ffi_prep_java_raw_closure_loc): New.
1827 * src/closures.c: New file.
1828 * src/dlmalloc.c [FFI_MMAP_EXEC_WRIT] (struct malloc_segment):
1829 Replace sflags with exec_offset.
1830 [FFI_MMAP_EXEC_WRIT] (mmap_exec_offset, add_segment_exec_offset,
1831 sub_segment_exec_offset): New macros.
1832 (get_segment_flags, set_segment_flags, check_segment_merge): New
1834 (is_mmapped_segment, is_extern_segment): Use get_segment_flags.
1835 (add_segment, sys_alloc, create_mspace, create_mspace_with_base,
1836 destroy_mspace): Use new macros.
1837 (sys_alloc): Silence warning.
1838 * Makefile.am (libffi_la_SOURCES): Add src/closures.c.
1839 * Makefile.in: Rebuilt.
1840 * src/prep_cif [FFI_CLOSURES] (ffi_prep_closure): Implement in
1841 terms of ffi_prep_closure_loc.
1842 * src/raw_api.c (ffi_prep_raw_closure_loc): Renamed and adjusted
1844 (ffi_prep_raw_closure): ... this. Re-implement in terms of the
1846 * src/java_raw_api (ffi_prep_java_raw_closure_loc): Renamed and
1848 (ffi_prep_java_raw_closure): ... this. Re-implement in terms of
1849 the renamed version.
1850 * src/alpha/ffi.c (ffi_prep_closure_loc): Renamed from
1851 (ffi_prep_closure): ... this.
1852 * src/pa/ffi.c: Likewise.
1853 * src/cris/ffi.c: Likewise. Adjust.
1854 * src/frv/ffi.c: Likewise.
1855 * src/ia64/ffi.c: Likewise.
1856 * src/mips/ffi.c: Likewise.
1857 * src/powerpc/ffi_darwin.c: Likewise.
1858 * src/s390/ffi.c: Likewise.
1859 * src/sh/ffi.c: Likewise.
1860 * src/sh64/ffi.c: Likewise.
1861 * src/sparc/ffi.c: Likewise.
1862 * src/x86/ffi64.c: Likewise.
1863 * src/x86/ffi.c: Likewise.
1864 (FFI_INIT_TRAMPOLINE): Adjust.
1865 (ffi_prep_raw_closure_loc): Renamed and adjusted from...
1866 (ffi_prep_raw_closure): ... this.
1867 * src/powerpc/ffi.c (ffi_prep_closure_loc): Renamed from
1868 (ffi_prep_closure): ... this.
1869 (flush_icache): Adjust.
1871 2007-03-07 Alexandre Oliva <aoliva@redhat.com>
1873 * src/dlmalloc.c: New file, imported version 2.8.3 of Doug
1876 2007-03-01 Brooks Moses <brooks.moses@codesourcery.com>
1878 * Makefile.am: Add dummy install-pdf target.
1879 * Makefile.in: Regenerate
1881 2007-02-13 Andreas Krebbel <krebbel1@de.ibm.com>
1883 * src/s390/ffi.c (ffi_prep_args, ffi_prep_cif_machdep,
1884 ffi_closure_helper_SYSV): Add long double handling.
1886 2007-02-02 Jakub Jelinek <jakub@redhat.com>
1888 * src/powerpc/linux64.S (ffi_call_LINUX64): Move restore of r2
1889 immediately after bctrl instruction.
1891 2007-01-18 Alexandre Oliva <aoliva@redhat.com>
1893 * Makefile.am (all-recursive, install-recursive,
1894 mostlyclean-recursive, clean-recursive, distclean-recursive,
1895 maintainer-clean-recursive): Add missing targets.
1896 * Makefile.in: Rebuilt.
1898 2006-12-14 Andreas Tobler <a.tobler@schweiz.org>
1900 * configure.ac: Add TARGET for x86_64-*-darwin*.
1901 * Makefile.am (nodist_libffi_la_SOURCES): Add rules for 64-bit sources
1903 * src/x86/ffitarget.h: Set trampoline size for x86_64-*-darwin*.
1904 * src/x86/darwin64.S: New file for x86_64-*-darwin* support.
1905 * configure: Regenerate.
1906 * Makefile.in: Regenerate.
1907 * include/Makefile.in: Regenerate.
1908 * testsuite/Makefile.in: Regenerate.
1909 * testsuite/libffi.special/unwindtest_ffi_call.cc: New test case for
1912 2006-12-13 Andreas Tobler <a.tobler@schweiz.org>
1914 * aclocal.m4: Regenerate with aclocal -I .. as written in the
1917 2006-10-31 Geoffrey Keating <geoffk@apple.com>
1919 * src/powerpc/ffi_darwin.c (darwin_adjust_aggregate_sizes): New.
1920 (ffi_prep_cif_machdep): Call darwin_adjust_aggregate_sizes for
1922 * testsuite/libffi.call/nested_struct4.c: Remove Darwin XFAIL.
1923 * testsuite/libffi.call/nested_struct6.c: Remove Darwin XFAIL.
1925 2006-10-10 Paolo Bonzini <bonzini@gnu.org>
1926 Sandro Tolaini <tolaini@libero.it>
1928 * configure.ac [i*86-*-darwin*]: Set X86_DARWIN symbol and
1930 * configure: Regenerated.
1931 * Makefile.am (nodist_libffi_la_SOURCES) [X86_DARWIN]: New case.
1932 (EXTRA_DIST): Add src/x86/darwin.S.
1933 * Makefile.in: Regenerated.
1934 * include/Makefile.in: Regenerated.
1935 * testsuite/Makefile.in: Regenerated.
1937 * src/x86/ffi.c (ffi_prep_cif_machdep) [X86_DARWIN]: Treat like
1938 X86_WIN32, and additionally align stack to 16 bytes.
1939 * src/x86/darwin.S: New, based on sysv.S.
1940 * src/prep_cif.c (ffi_prep_cif) [X86_DARWIN]: Align > 8-byte structs.
1942 2006-09-12 David Daney <ddaney@avtrex.com>
1945 * include/Makefile.am: Install both ffi.h and ffitarget.h in
1946 $(libdir)/gcc/$(target_alias)/$(gcc_version)/include.
1947 * aclocal.m4: Regenerated for automake 1.9.6.
1948 * Makefile.in: Regenerated.
1949 * include/Makefile.in: Regenerated.
1950 * testsuite/Makefile.in: Regenerated.
1952 2006-08-17 Andreas Tobler <a.tobler@schweiz.ch>
1954 * include/ffi_common.h (struct): Revert accidental commit.
1956 2006-08-15 Andreas Tobler <a.tobler@schweiz.ch>
1958 * include/ffi_common.h: Remove lint directives.
1959 * include/ffi.h.in: Likewise.
1961 2006-07-25 Torsten Schoenfeld <kaffeetisch@gmx.de>
1963 * include/ffi.h.in (ffi_type_ulong, ffi_type_slong): Define correctly
1964 for 32-bit architectures.
1965 * testsuite/libffi.call/return_ul.c: New test case.
1967 2006-07-19 David Daney <ddaney@avtrex.com>
1969 * testsuite/libffi.call/closure_fn6.c: Remove xfail for mips,
1970 xfail remains for mips64.
1972 2006-05-23 Carlos O'Donell <carlos@codesourcery.com>
1974 * Makefile.am: Add install-html target. Add install-html to .PHONY
1975 * Makefile.in: Regenerate.
1976 * aclocal.m4: Regenerate.
1977 * include/Makefile.in: Regenerate.
1978 * testsuite/Makefile.in: Regenerate.
1980 2006-05-18 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
1982 * pa/ffi.c (ffi_prep_args_pa32): Load floating point arguments from
1985 2006-04-22 Andreas Tobler <a.tobler@schweiz.ch>
1987 * README: Remove notice about 'Crazy Comments'.
1988 * src/debug.c: Remove lint directives. Cleanup white spaces.
1989 * src/java_raw_api.c: Likewise.
1990 * src/prep_cif.c: Likewise.
1991 * src/raw_api.c: Likewise.
1992 * src/ffitest.c: Delete. No longer needed, all test cases migrated
1994 * src/arm/ffi.c: Remove lint directives.
1995 * src/m32r/ffi.c: Likewise.
1996 * src/pa/ffi.c: Likewise.
1997 * src/powerpc/ffi.c: Likewise.
1998 * src/powerpc/ffi_darwin.c: Likewise.
1999 * src/sh/ffi.c: Likewise.
2000 * src/sh64/ffi.c: Likewise.
2001 * src/x86/ffi.c: Likewise.
2002 * testsuite/libffi.call/float2.c: Likewise.
2003 * testsuite/libffi.call/promotion.c: Likewise.
2004 * testsuite/libffi.call/struct1.c: Likewise.
2006 2006-04-13 Andreas Tobler <a.tobler@schweiz.ch>
2008 * src/pa/hpux32.S: Correct unwind offset calculation for
2010 * src/pa/linux.S: Likewise.
2012 2006-04-12 James E Wilson <wilson@specifix.com>
2015 * src/ia64/ffi.c (stf_spill, ldf_fill): Rewrite as macros.
2016 (hfa_type_load): Call stf_spill.
2017 (hfa_type_store): Call ldf_fill.
2018 (ffi_call): Adjust calls to above routines. Add local temps for
2021 2006-04-10 Matthias Klose <doko@debian.org>
2023 * testsuite/lib/libffi-dg.exp (libffi-init): Recognize multilib
2024 directory names containing underscores.
2026 2006-04-07 James E Wilson <wilson@specifix.com>
2028 * testsuite/libffi.call/float4.c: New testcase.
2030 2006-04-05 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
2031 Andreas Tobler <a.tobler@schweiz.ch>
2033 * Makefile.am: Add PA_HPUX port.
2034 * Makefile.in: Regenerate.
2035 * include/Makefile.in: Likewise.
2036 * testsuite/Makefile.in: Likewise.
2037 * configure.ac: Add PA_HPUX rules.
2038 * configure: Regenerate.
2039 * src/pa/ffitarget.h: Rename linux target to PA_LINUX.
2040 Add PA_HPUX and PA64_HPUX.
2041 Rename FFI_LINUX ABI to FFI_PA32 ABI.
2042 (FFI_TRAMPOLINE_SIZE): Define for 32-bit HP-UX targets.
2043 (FFI_TYPE_SMALL_STRUCT2): Define.
2044 (FFI_TYPE_SMALL_STRUCT4): Likewise.
2045 (FFI_TYPE_SMALL_STRUCT8): Likewise.
2046 (FFI_TYPE_SMALL_STRUCT3): Redefine.
2047 (FFI_TYPE_SMALL_STRUCT5): Likewise.
2048 (FFI_TYPE_SMALL_STRUCT6): Likewise.
2049 (FFI_TYPE_SMALL_STRUCT7): Likewise.
2050 * src/pa/ffi.c (ROUND_DOWN): Delete.
2051 (fldw, fstw, fldd, fstd): Use '__asm__'.
2052 (ffi_struct_type): Add support for FFI_TYPE_SMALL_STRUCT2,
2053 FFI_TYPE_SMALL_STRUCT4 and FFI_TYPE_SMALL_STRUCT8.
2054 (ffi_prep_args_LINUX): Rename to ffi_prep_args_pa32. Update comment.
2055 Simplify incrementing of stack slot variable. Change type of local
2056 'n' to unsigned int.
2057 (ffi_size_stack_LINUX): Rename to ffi_size_stack_pa32. Handle long
2059 (ffi_prep_cif_machdep): Likewise.
2060 (ffi_call): Likewise.
2061 (ffi_closure_inner_LINUX): Rename to ffi_closure_inner_pa32. Change
2062 return type to ffi_status. Simplify incrementing of stack slot
2063 variable. Only copy floating point argument registers when PA_LINUX
2064 is true. Reformat debug statement.
2065 Add support for FFI_TYPE_SMALL_STRUCT2, FFI_TYPE_SMALL_STRUCT4 and
2066 FFI_TYPE_SMALL_STRUCT8.
2067 (ffi_closure_LINUX): Rename to ffi_closure_pa32. Add 'extern' to
2069 (ffi_prep_closure): Make linux trampoline conditional on PA_LINUX.
2070 Add nops to cache flush. Add trampoline for PA_HPUX.
2071 * src/pa/hpux32.S: New file.
2072 * src/pa/linux.S (ffi_call_LINUX): Rename to ffi_call_pa32. Rename
2073 ffi_prep_args_LINUX to ffi_prep_args_pa32.
2074 Localize labels. Add support for 2, 4 and 8-byte small structs. Handle
2075 unaligned destinations in 3, 5, 6 and 7-byte small structs. Order
2076 argument type checks so that common argument types appear first.
2077 (ffi_closure_LINUX): Rename to ffi_closure_pa32. Rename
2078 ffi_closure_inner_LINUX to ffi_closure_inner_pa32.
2080 2006-03-24 Alan Modra <amodra@bigpond.net.au>
2082 * src/powerpc/ffitarget.h (enum ffi_abi): Add FFI_LINUX. Default
2083 for 32-bit using IBM extended double format. Fix FFI_LAST_ABI.
2084 * src/powerpc/ffi.c (ffi_prep_args_SYSV): Handle linux variant of
2085 FFI_TYPE_LONGDOUBLE.
2086 (ffi_prep_args64): Assert using IBM extended double.
2087 (ffi_prep_cif_machdep): Don't munge FFI_TYPE_LONGDOUBLE type.
2088 Handle FFI_LINUX FFI_TYPE_LONGDOUBLE return and args.
2089 (ffi_call): Handle FFI_LINUX.
2090 (ffi_closure_helper_SYSV): Non FFI_LINUX long double return needs
2091 gpr3 return pointer as for struct return. Handle FFI_LINUX
2092 FFI_TYPE_LONGDOUBLE return and args. Don't increment "nf"
2094 * src/powerpc/ppc_closure.S (ffi_closure_SYSV): Load both f1 and f2
2095 for FFI_TYPE_LONGDOUBLE. Move epilogue insns into case table.
2096 Don't use r6 as pointer to results, instead use sp offset. Don't
2097 make a special call to load lr with case table address, instead
2098 use offset from previous call.
2099 * src/powerpc/sysv.S (ffi_call_SYSV): Save long double return.
2100 * src/powerpc/linux64.S (ffi_call_LINUX64): Simplify long double
2103 2006-03-15 Kaz Kojima <kkojima@gcc.gnu.org>
2105 * src/sh64/ffi.c (ffi_prep_cif_machdep): Handle float arguments
2106 passed with FP registers correctly.
2107 (ffi_closure_helper_SYSV): Likewise.
2108 * src/sh64/sysv.S: Likewise.
2110 2006-03-01 Andreas Tobler <a.tobler@schweiz.ch>
2112 * testsuite/libffi.special/unwindtest.cc (closure_test_fn): Mark cif,
2113 args and userdata unused.
2114 (closure_test_fn1): Mark cif and userdata unused.
2115 (main): Remove unused res.
2117 2006-02-28 Andreas Tobler <a.tobler@schweiz.ch>
2119 * testsuite/libffi.call/call.exp: Adjust FSF address. Add test runs for
2120 -O2, -O3, -Os and the warning flags -W -Wall.
2121 * testsuite/libffi.special/special.exp: Likewise.
2122 * testsuite/libffi.call/ffitest.h: Add an __UNUSED__ macro to mark
2123 unused parameter unused for gcc or else do nothing.
2124 * testsuite/libffi.special/ffitestcxx.h: Likewise.
2125 * testsuite/libffi.call/cls_12byte.c (cls_struct_12byte_gn): Mark cif
2126 and userdata unused.
2127 * testsuite/libffi.call/cls_16byte.c (cls_struct_16byte_gn): Likewise.
2128 * testsuite/libffi.call/cls_18byte.c (cls_struct_18byte_gn): Likewise.
2129 * testsuite/libffi.call/cls_19byte.c (cls_struct_19byte_gn): Likewise.
2130 * testsuite/libffi.call/cls_1_1byte.c (cls_struct_1_1byte_gn): Likewise.
2131 * testsuite/libffi.call/cls_20byte.c (cls_struct_20byte_gn): Likewise.
2132 * testsuite/libffi.call/cls_20byte1.c (cls_struct_20byte_gn): Likewise.
2133 * testsuite/libffi.call/cls_24byte.c (cls_struct_24byte_gn): Likewise.
2134 * testsuite/libffi.call/cls_2byte.c (cls_struct_2byte_gn): Likewise.
2135 * testsuite/libffi.call/cls_3_1byte.c (cls_struct_3_1byte_gn): Likewise.
2136 * testsuite/libffi.call/cls_3byte1.c (cls_struct_3byte_gn): Likewise.
2137 * testsuite/libffi.call/cls_3byte2.c (cls_struct_3byte_gn1): Likewise.
2138 * testsuite/libffi.call/cls_4_1byte.c (cls_struct_4_1byte_gn): Likewise.
2139 * testsuite/libffi.call/cls_4byte.c (cls_struct_4byte_gn): Likewise.
2140 * testsuite/libffi.call/cls_5_1_byte.c (cls_struct_5byte_gn): Likewise.
2141 * testsuite/libffi.call/cls_5byte.c (cls_struct_5byte_gn): Likewise.
2142 * testsuite/libffi.call/cls_64byte.c (cls_struct_64byte_gn): Likewise.
2143 * testsuite/libffi.call/cls_6_1_byte.c (cls_struct_6byte_gn): Likewise.
2144 * testsuite/libffi.call/cls_6byte.c (cls_struct_6byte_gn): Likewise.
2145 * testsuite/libffi.call/cls_7_1_byte.c (cls_struct_7byte_gn): Likewise.
2146 * testsuite/libffi.call/cls_7byte.c (cls_struct_7byte_gn): Likewise.
2147 * testsuite/libffi.call/cls_8byte.c (cls_struct_8byte_gn): Likewise.
2148 * testsuite/libffi.call/cls_9byte1.c (cls_struct_9byte_gn): Likewise.
2149 * testsuite/libffi.call/cls_9byte2.c (cls_struct_9byte_gn): Likewise.
2150 * testsuite/libffi.call/cls_align_double.c (cls_struct_align_gn):
2152 * testsuite/libffi.call/cls_align_float.c (cls_struct_align_gn):
2154 * testsuite/libffi.call/cls_align_longdouble.c (cls_struct_align_gn):
2156 * testsuite/libffi.call/cls_align_pointer.c (cls_struct_align_fn): Cast
2157 void* to avoid compiler warning.
2159 (cls_struct_align_gn): Mark cif and userdata unused.
2160 * testsuite/libffi.call/cls_align_sint16.c (cls_struct_align_gn):
2162 * testsuite/libffi.call/cls_align_sint32.c (cls_struct_align_gn):
2164 * testsuite/libffi.call/cls_align_sint64.c (cls_struct_align_gn):
2166 * testsuite/libffi.call/cls_align_uint16.c (cls_struct_align_gn):
2168 * testsuite/libffi.call/cls_align_uint32.c (cls_struct_align_gn):
2170 * testsuite/libffi.call/cls_double.c (cls_ret_double_fn): Likewise.
2171 * testsuite/libffi.call/cls_float.c (cls_ret_float_fn): Likewise.
2172 * testsuite/libffi.call/cls_multi_schar.c (test_func_gn): Mark cif and
2174 (main): Cast res_call to silence gcc.
2175 * testsuite/libffi.call/cls_multi_sshort.c (test_func_gn): Mark cif and
2177 (main): Cast res_call to silence gcc.
2178 * testsuite/libffi.call/cls_multi_sshortchar.c (test_func_gn): Mark cif
2180 (main): Cast res_call to silence gcc.
2181 * testsuite/libffi.call/cls_multi_uchar.c (test_func_gn): Mark cif and
2183 (main): Cast res_call to silence gcc.
2184 * testsuite/libffi.call/cls_multi_ushort.c (test_func_gn): Mark cif and
2186 (main): Cast res_call to silence gcc.
2187 * testsuite/libffi.call/cls_multi_ushortchar.c (test_func_gn): Mark cif
2189 (main): Cast res_call to silence gcc.
2190 * testsuite/libffi.call/cls_schar.c (cls_ret_schar_fn): Mark cif and
2192 (cls_ret_schar_fn): Cast printf parameter to silence gcc.
2193 * testsuite/libffi.call/cls_sint.c (cls_ret_sint_fn): Mark cif and
2195 (cls_ret_sint_fn): Cast printf parameter to silence gcc.
2196 * testsuite/libffi.call/cls_sshort.c (cls_ret_sshort_fn): Mark cif and
2198 (cls_ret_sshort_fn): Cast printf parameter to silence gcc.
2199 * testsuite/libffi.call/cls_uchar.c (cls_ret_uchar_fn): Mark cif and
2201 (cls_ret_uchar_fn): Cast printf parameter to silence gcc.
2202 * testsuite/libffi.call/cls_uint.c (cls_ret_uint_fn): Mark cif and
2204 (cls_ret_uint_fn): Cast printf parameter to silence gcc.
2205 * testsuite/libffi.call/cls_ulonglong.c (cls_ret_ulonglong_fn): Mark cif
2206 and userdata unused.
2207 * testsuite/libffi.call/cls_ushort.c (cls_ret_ushort_fn): Mark cif and
2209 (cls_ret_ushort_fn): Cast printf parameter to silence gcc.
2210 * testsuite/libffi.call/float.c (floating): Remove unused parameter e.
2211 * testsuite/libffi.call/float1.c (main): Remove unused variable i.
2212 Cleanup white spaces.
2213 * testsuite/libffi.call/negint.c (checking): Remove unused variable i.
2214 * testsuite/libffi.call/nested_struct.c (cls_struct_combined_gn): Mark
2215 cif and userdata unused.
2216 * testsuite/libffi.call/nested_struct1.c (cls_struct_combined_gn):
2218 * testsuite/libffi.call/nested_struct10.c (B_gn): Likewise.
2219 * testsuite/libffi.call/nested_struct2.c (B_fn): Adjust printf
2220 formatters to silence gcc.
2221 (B_gn): Mark cif and userdata unused.
2222 * testsuite/libffi.call/nested_struct3.c (B_gn): Mark cif and userdata
2224 * testsuite/libffi.call/nested_struct4.c: Mention related PR.
2225 (B_gn): Mark cif and userdata unused.
2226 * testsuite/libffi.call/nested_struct5.c (B_gn): Mark cif and userdata
2228 * testsuite/libffi.call/nested_struct6.c: Mention related PR.
2229 (B_gn): Mark cif and userdata unused.
2230 * testsuite/libffi.call/nested_struct7.c (B_gn): Mark cif and userdata
2232 * testsuite/libffi.call/nested_struct8.c (B_gn): Likewise.
2233 * testsuite/libffi.call/nested_struct9.c (B_gn): Likewise.
2234 * testsuite/libffi.call/problem1.c (stub): Likewise.
2235 * testsuite/libffi.call/pyobjc-tc.c (main): Cast the result to silence
2237 * testsuite/libffi.call/return_fl2.c (return_fl): Add the note mentioned
2238 in the last commit for this test case in the test case itself.
2239 * testsuite/libffi.call/closure_fn0.c (closure_test_fn0): Mark cif as
2241 * testsuite/libffi.call/closure_fn1.c (closure_test_fn1): Likewise.
2242 * testsuite/libffi.call/closure_fn2.c (closure_test_fn2): Likewise.
2243 * testsuite/libffi.call/closure_fn3.c (closure_test_fn3): Likewise.
2244 * testsuite/libffi.call/closure_fn4.c (closure_test_fn0): Likewise.
2245 * testsuite/libffi.call/closure_fn5.c (closure_test_fn5): Likewise.
2246 * testsuite/libffi.call/closure_fn6.c (closure_test_fn0): Likewise.
2248 2006-02-22 Kaz Kojima <kkojima@gcc.gnu.org>
2250 * src/sh/sysv.S: Fix register numbers in the FDE for
2253 2006-02-20 Andreas Tobler <a.tobler@schweiz.ch>
2255 * testsuite/libffi.call/return_fl2.c (return_fl): Remove static
2256 declaration to avoid a false negative on ix86. See PR323.
2258 2006-02-18 Kaz Kojima <kkojima@gcc.gnu.org>
2260 * src/sh/ffi.c (ffi_closure_helper_SYSV): Remove unused variable
2261 and cast integer to void * if needed. Update the pointer to
2262 the FP register saved area correctly.
2264 2006-02-17 Andreas Tobler <a.tobler@schweiz.ch>
2266 * testsuite/libffi.call/nested_struct6.c: XFAIL this test until PR25630
2268 * testsuite/libffi.call/nested_struct4.c: Likewise.
2270 2006-02-16 Andreas Tobler <a.tobler@schweiz.ch>
2272 * testsuite/libffi.call/return_dbl.c: New test case.
2273 * testsuite/libffi.call/return_dbl1.c: Likewise.
2274 * testsuite/libffi.call/return_dbl2.c: Likewise.
2275 * testsuite/libffi.call/return_fl.c: Likewise.
2276 * testsuite/libffi.call/return_fl1.c: Likewise.
2277 * testsuite/libffi.call/return_fl2.c: Likewise.
2278 * testsuite/libffi.call/return_fl3.c: Likewise.
2279 * testsuite/libffi.call/closure_fn6.c: Likewise.
2281 * testsuite/libffi.call/nested_struct2.c: Remove ffi_type_mylong
2283 * testsuite/libffi.call/ffitest.h: Add ffi_type_mylong definition
2284 here to be used by other test cases too.
2286 * testsuite/libffi.call/nested_struct10.c: New test case.
2287 * testsuite/libffi.call/nested_struct9.c: Likewise.
2288 * testsuite/libffi.call/nested_struct8.c: Likewise.
2289 * testsuite/libffi.call/nested_struct7.c: Likewise.
2290 * testsuite/libffi.call/nested_struct6.c: Likewise.
2291 * testsuite/libffi.call/nested_struct5.c: Likewise.
2292 * testsuite/libffi.call/nested_struct4.c: Likewise.
2294 2006-01-21 Andreas Tobler <a.tobler@schweiz.ch>
2296 * configure.ac: Enable libffi for sparc64-*-freebsd*.
2297 * configure: Rebuilt.
2299 2006-01-18 Jakub Jelinek <jakub@redhat.com>
2301 * src/powerpc/sysv.S (smst_two_register): Don't call __ashldi3,
2302 instead do the shifting inline.
2303 * src/powerpc/ppc_closure.S (ffi_closure_SYSV): Don't compute %r5
2304 shift count unconditionally. Simplify load sequences for 1, 2, 3, 4
2305 and 8 byte structs, for the remaining struct sizes don't call
2306 __lshrdi3, instead do the shifting inline.
2308 2005-12-07 Thiemo Seufer <ths@networkno.de>
2310 * src/mips/ffitarget.h: Remove obsolete sgidefs.h include. Add
2311 missing parentheses.
2312 * src/mips/o32.S (ffi_call_O32): Code formatting. Define
2313 and use A3_OFF, FP_OFF, RA_OFF. Micro-optimizations.
2314 (ffi_closure_O32): Likewise, but with newly defined A3_OFF2,
2315 A2_OFF2, A1_OFF2, A0_OFF2, RA_OFF2, FP_OFF2, S0_OFF2, GP_OFF2,
2316 V1_OFF2, V0_OFF2, FA_1_1_OFF2, FA_1_0_OFF2, FA_0_1_OFF2,
2318 * src/mips/ffi.c (ffi_prep_args): Code formatting. Fix
2320 (ffi_prep_closure): Improve trampoline instruction scheduling.
2321 (ffi_closure_mips_inner_O32): Fix endianness bugs.
2323 2005-12-03 Alan Modra <amodra@bigpond.net.au>
2325 * src/powerpc/ffi.c: Formatting.
2326 (ffi_prep_args_SYSV): Avoid possible aliasing problems by using unions.
2327 (ffi_prep_args64): Likewise.
2329 2005-09-30 Geoffrey Keating <geoffk@apple.com>
2331 * testsuite/lib/libffi-dg.exp (libffi_target_compile): For
2332 darwin, use -shared-libgcc not -lgcc_s, and explain why.
2334 2005-09-26 Tom Tromey <tromey@redhat.com>
2336 * testsuite/libffi.call/float1.c (value_type): New typedef.
2337 (CANARY): New define.
2338 (main): Check for result buffer overflow.
2339 * src/powerpc/linux64.S: Handle linux64 long double returns.
2340 * src/powerpc/ffi.c (FLAG_RETURNS_128BITS): New constant.
2341 (ffi_prep_cif_machdep): Handle linux64 long double returns.
2343 2005-08-25 Alan Modra <amodra@bigpond.net.au>
2346 * src/powerpc/ffi.c (ffi_prep_args_SYSV): Correct placement of stack
2348 (ffi_status ffi_prep_cif_machdep): Correct stack sizing for same.
2350 2005-08-11 Jakub Jelinek <jakub@redhat.com>
2352 * configure.ac (HAVE_HIDDEN_VISIBILITY_ATTRIBUTE): New test.
2353 (AH_BOTTOM): Add FFI_HIDDEN definition.
2354 * configure: Rebuilt.
2355 * fficonfig.h.in: Rebuilt.
2356 * src/powerpc/ffi.c (hidden): Remove.
2357 (ffi_closure_LINUX64, ffi_prep_args64, ffi_call_LINUX64,
2358 ffi_closure_helper_LINUX64): Use FFI_HIDDEN instead of hidden.
2359 * src/powerpc/linux64_closure.S (ffi_closure_LINUX64,
2360 .ffi_closure_LINUX64): Use FFI_HIDDEN instead of .hidden.
2361 * src/x86/ffi.c (ffi_closure_SYSV, ffi_closure_raw_SYSV): Remove,
2362 add FFI_HIDDEN to its prototype.
2363 (ffi_closure_SYSV_inner): New.
2364 * src/x86/sysv.S (ffi_closure_SYSV, ffi_closure_raw_SYSV): New.
2365 * src/x86/win32.S (ffi_closure_SYSV, ffi_closure_raw_SYSV): New.
2367 2005-08-10 Alfred M. Szmidt <ams@gnu.org>
2370 * configure: Rebuilt.
2371 * configure.ac: Handle i*86-*-gnu*.
2373 2005-08-09 Jakub Jelinek <jakub@redhat.com>
2375 * src/powerpc/ppc_closure.S (ffi_closure_SYSV): Use
2376 DW_CFA_offset_extended_sf rather than
2377 DW_CFA_GNU_negative_offset_extended.
2378 * src/powerpc/sysv.S (ffi_call_SYSV): Likewise.
2380 2005-07-22 SUGIOKA Toshinobu <sugioka@itonet.co.jp>
2382 * src/sh/sysv.S (ffi_call_SYSV): Stop argument popping correctly
2384 (ffi_closure_SYSV): Change the stack layout for sh3 struct argument.
2385 * src/sh/ffi.c (ffi_prep_args): Fix sh3 argument copy, when it is
2386 partially on register.
2387 (ffi_closure_helper_SYSV): Likewise.
2388 (ffi_prep_cif_machdep): Don't set too many cif->flags.
2390 2005-07-20 Kaz Kojima <kkojima@gcc.gnu.org>
2392 * src/sh/ffi.c (ffi_call): Handle small structures correctly.
2394 * src/sh64/ffi.c (simple_type): Remove.
2395 (return_type): Handle small structures correctly.
2396 (ffi_prep_args): Likewise.
2397 (ffi_call): Likewise.
2398 (ffi_closure_helper_SYSV): Likewise.
2399 * src/sh64/sysv.S (ffi_call_SYSV): Handle 1, 2 and 4-byte return.
2400 Emit position independent code if PIC and remove wrong datalabel
2401 prefixes from EH data.
2403 2005-07-19 Andreas Tobler <a.tobler@schweiz.ch>
2405 * Makefile.am (nodist_libffi_la_SOURCES): Add POWERPC_FREEBSD.
2406 * Makefile.in: Regenerate.
2407 * include/Makefile.in: Likewise.
2408 * testsuite/Makefile.in: Likewise.
2409 * configure.ac: Add POWERPC_FREEBSD rules.
2410 * configure: Regenerate.
2411 * src/powerpc/ffitarget.h: Add POWERPC_FREEBSD rules.
2412 (FFI_SYSV_TYPE_SMALL_STRUCT): Define.
2413 * src/powerpc/ffi.c: Add flags to handle small structure returns
2415 (ffi_prep_cif_machdep): Handle small structures for SYSV 4 ABI.
2417 (ffi_closure_helper_SYSV): Likewise.
2418 * src/powerpc/ppc_closure.S: Add return types for small structures.
2419 * src/powerpc/sysv.S: Add bits to handle small structures for
2422 2005-07-10 Andreas Tobler <a.tobler@schweiz.ch>
2424 * testsuite/libffi.call/cls_5_1_byte.c: New test file.
2425 * testsuite/libffi.call/cls_6_1_byte.c: Likewise.
2426 * testsuite/libffi.call/cls_7_1_byte.c: Likewise.
2428 2005-07-05 Randolph Chung <tausq@debian.org>
2430 * src/pa/ffi.c (ffi_struct_type): Rename FFI_TYPE_SMALL_STRUCT1
2431 as FFI_TYPE_SMALL_STRUCT3. Break out handling for 5-7 byte
2432 structures. Kill compilation warnings.
2433 (ffi_closure_inner_LINUX): Print return values as hex in debug
2434 message. Rename FFI_TYPE_SMALL_STRUCT1 as FFI_TYPE_SMALL_STRUCT3.
2435 Properly handle 5-7 byte structure returns.
2436 * src/pa/ffitarget.h (FFI_TYPE_SMALL_STRUCT1)
2437 (FFI_TYPE_SMALL_STRUCT2): Remove.
2438 (FFI_TYPE_SMALL_STRUCT3, FFI_TYPE_SMALL_STRUCT5)
2439 (FFI_TYPE_SMALL_STRUCT6, FFI_TYPE_SMALL_STRUCT7): Define.
2440 * src/pa/linux.S: Mark source file as using PA1.1 assembly.
2441 (checksmst1, checksmst2): Remove.
2442 (checksmst3): Optimize handling of 3-byte struct returns.
2443 (checksmst567): Properly handle 5-7 byte struct returns.
2445 2005-06-15 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
2448 * src/mips/n32.S: Enforce PIC code.
2449 * src/mips/o32.S: Likewise.
2451 2005-06-15 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
2453 * configure.ac: Treat i*86-*-solaris2.10 and up as X86_64.
2454 * configure: Regenerate.
2456 2005-06-01 Alan Modra <amodra@bigpond.net.au>
2458 * src/powerpc/ppc_closure.S (ffi_closure_SYSV): Don't use JUMPTARGET
2459 to call ffi_closure_helper_SYSV. Append @local instead.
2460 * src/powerpc/sysv.S (ffi_call_SYSV): Likewise for ffi_prep_args_SYSV.
2462 2005-05-17 Kelley Cook <kcook@gcc.gnu.org>
2464 * configure.ac: Use AC_C_BIGENDIAN instead of AC_C_BIGENDIAN_CROSS.
2465 Use AC_CHECK_SIZEOF instead of AC_COMPILE_CHECK_SIZEOF.
2466 * Makefile.am (ACLOCAL_AMFLAGS): Remove -I ../config.
2467 * aclocal.m4, configure, fficonfig.h.in, Makefile.in,
2468 include/Makefile.in, testsuite/Makefile.in: Regenerate.
2470 2005-05-09 Mike Stump <mrs@apple.com>
2472 * configure: Regenerate.
2474 2005-05-08 Richard Henderson <rth@redhat.com>
2477 * src/alpha/osf.S: Update unwind into to match code.
2479 2005-05-04 Andreas Degert <ad@papyrus-gmbh.de>
2480 Richard Henderson <rth@redhat.com>
2482 * src/x86/ffi64.c (ffi_prep_cif_machdep): Save sse-used flag in
2484 (ffi_call): Mask return type field. Pass ssecount to ffi_call_unix64.
2485 (ffi_prep_closure): Set carry bit if sse-used flag set.
2486 * src/x86/unix64.S (ffi_call_unix64): Add ssecount argument.
2487 Only load sse registers if ssecount non-zero.
2488 (ffi_closure_unix64): Only save sse registers if carry set on entry.
2490 2005-04-29 Ralf Corsepius <ralf.corsepius@rtems.org>
2492 * configure.ac: Add i*86-*-rtems*, sparc*-*-rtems*,
2493 powerpc-*rtems*, arm*-*-rtems*, sh-*-rtems*.
2494 * configure: Regenerate.
2496 2005-04-20 Hans-Peter Nilsson <hp@axis.com>
2498 * testsuite/lib/libffi-dg.exp (libffi-dg-test-1): In regsub use,
2499 have Tcl8.3-compatible intermediate variable.
2501 2005-04-18 Simon Posnjak <simon.posnjak@siol.net>
2502 Hans-Peter Nilsson <hp@axis.com>
2504 * Makefile.am: Add CRIS support.
2505 * configure.ac: Likewise.
2506 * Makefile.in, configure, testsuite/Makefile.in,
2507 include/Makefile.in: Regenerate.
2508 * src/cris: New directory.
2509 * src/cris/ffi.c, src/cris/sysv.S, src/cris/ffitarget.h: New files.
2510 * src/prep_cif.c (ffi_prep_cif): Wrap in #ifndef __CRIS__.
2512 * testsuite/lib/libffi-dg.exp (libffi-dg-test-1): Replace \n with
2513 \r?\n in output tests.
2515 2005-04-12 Mike Stump <mrs@apple.com>
2517 * configure: Regenerate.
2519 2005-03-30 Hans Boehm <Hans.Boehm@hp.com>
2521 * src/ia64/ffitarget.h (ffi_arg): Use long long instead of DI.
2523 2005-03-30 Steve Ellcey <sje@cup.hp.com>
2525 * src/ia64/ffitarget.h (ffi_arg) ADD DI attribute.
2527 * src/ia64/unix.S (ffi_closure_unix): Extend gp
2528 to 64 bits in ILP32 mode.
2529 Load 64 bits even for short data.
2531 2005-03-23 Mike Stump <mrs@apple.com>
2533 * src/powerpc/darwin.S: Update for -m64 multilib.
2534 * src/powerpc/darwin_closure.S: Likewise.
2536 2005-03-21 Zack Weinberg <zack@codesourcery.com>
2538 * configure.ac: Do not invoke TL_AC_GCC_VERSION.
2539 Do not set tool_include_dir.
2540 * aclocal.m4, configure, Makefile.in, testsuite/Makefile.in:
2542 * include/Makefile.am: Set gcc_version and toollibffidir.
2543 * include/Makefile.in: Regenerate.
2545 2005-02-22 Andrew Haley <aph@redhat.com>
2547 * src/powerpc/ffi.c (ffi_prep_cif_machdep): Bump alignment to
2548 odd-numbered register pairs for 64-bit integer types.
2550 2005-02-23 Andreas Tobler <a.tobler@schweiz.ch>
2553 * testsuite/libffi.call/return_ll1.c: New test case.
2555 2005-02-11 Janis Johnson <janis187@us.ibm.com>
2557 * testsuite/libffi.call/cls_align_longdouble.c: Remove dg-options.
2558 * testsuite/libffi.call/float.c: Ditto.
2559 * testsuite/libffi.call/float2.c: Ditto.
2560 * testsuite/libffi.call/float3.c: Ditto.
2562 2005-02-08 Andreas Tobler <a.tobler@schweiz.ch>
2564 * src/frv/ffitarget.h: Remove PPC stuff which does not belong to frv.
2566 2005-01-12 Eric Botcazou <ebotcazou@libertysurf.fr>
2568 * testsuite/libffi.special/special.exp (cxx_options): Add
2571 2004-12-31 Richard Henderson <rth@redhat.com>
2573 * src/types.c (FFI_AGGREGATE_TYPEDEF): Remove.
2574 (FFI_TYPEDEF): Rename from FFI_INTEGRAL_TYPEDEF. Replace size and
2575 offset parameters with a type parameter; deduce size and structure
2576 alignment. Update all users.
2578 2004-12-31 Richard Henderson <rth@redhat.com>
2580 * src/types.c (FFI_TYPE_POINTER): Define with sizeof.
2581 (FFI_TYPE_LONGDOUBLE): Fix for ia64.
2582 * src/ia64/ffitarget.h (struct ffi_ia64_trampoline_struct): Move
2583 into ffi_prep_closure.
2584 * src/ia64/ia64_flags.h, src/ia64/ffi.c, src/ia64/unix.S: Rewrite
2587 2004-12-27 Richard Henderson <rth@redhat.com>
2589 * src/x86/unix64.S: Fix typo in unwind info.
2591 2004-12-25 Richard Henderson <rth@redhat.com>
2593 * src/x86/ffi64.c (struct register_args): Rename from stackLayout.
2594 (enum x86_64_reg_class): Add X86_64_COMPLEX_X87_CLASS.
2595 (merge_classes): Check for it.
2597 (classify_argument): Pass byte_offset by value; perform all updates
2599 (examine_argument): Add classes argument; handle
2600 X86_64_COMPLEX_X87_CLASS.
2601 (ffi_prep_args): Merge into ...
2602 (ffi_call): ... here. Share stack frame with ffi_call_unix64.
2603 (ffi_prep_cif_machdep): Setup cif->flags for proper structure return.
2604 (ffi_fill_return_value): Remove.
2605 (ffi_prep_closure): Remove dead assert.
2606 (ffi_closure_unix64_inner): Rename from ffi_closure_UNIX64_inner.
2607 Rewrite to use struct register_args instead of va_list. Create
2608 flags for handling structure returns.
2609 * src/x86/unix64.S: Remove dead strings.
2610 (ffi_call_unix64): Rename from ffi_call_UNIX64. Rewrite to share
2611 stack frame with ffi_call. Handle structure returns properly.
2612 (float2sse, floatfloat2sse, double2sse): Remove.
2613 (sse2float, sse2double, sse2floatfloat): Remove.
2614 (ffi_closure_unix64): Rename from ffi_closure_UNIX64. Rewrite
2615 to handle structure returns properly.
2617 2004-12-08 David Edelsohn <edelsohn@gnu.org>
2619 * Makefile.am (AM_MAKEFLAGS): Remove duplicate LIBCFLAGS and
2621 * Makefile.in: Regenerated.
2623 2004-12-02 Richard Sandiford <rsandifo@redhat.com>
2625 * configure.ac: Use TL_AC_GCC_VERSION to set gcc_version.
2626 * configure, aclocal.m4, Makefile.in: Regenerate.
2627 * include/Makefile.in, testsuite/Makefile.in: Regenerate.
2629 2004-11-29 Kelley Cook <kcook@gcc.gnu.org>
2631 * configure: Regenerate for libtool change.
2633 2004-11-25 Kelley Cook <kcook@gcc.gnu.org>
2635 * configure: Regenerate for libtool reversion.
2637 2004-11-24 Kelley Cook <kcook@gcc.gnu.org>
2639 * configure: Regenerate for libtool change.
2641 2004-11-23 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
2643 * testsuite/lib/libffi-dg.exp: Use new procs in target-libpath.exp.
2645 2004-11-23 Richard Sandiford <rsandifo@redhat.com>
2647 * src/mips/o32.S (ffi_call_O32, ffi_closure_O32): Use jalr instead
2648 of jal. Use an absolute encoding for the frame information.
2650 2004-11-23 Kelley Cook <kcook@gcc.gnu.org>
2652 * Makefile.am: Remove no-dependencies. Add ACLOCAL_AMFLAGS.
2653 * acinclude.m4: Delete logic for sincludes.
2654 * aclocal.m4, Makefile.in, configure: Regenerate.
2655 * include/Makefile: Likewise.
2656 * testsuite/Makefile: Likewise.
2658 2004-11-22 Eric Botcazou <ebotcazou@libertysurf.fr>
2660 * src/sparc/ffi.c (ffi_prep_closure): Align doubles and 64-bit integers
2661 on a 8-byte boundary.
2662 * src/sparc/v8.S (ffi_closure_v8): Reserve frame space for arguments.
2664 2004-10-27 Richard Earnshaw <rearnsha@arm.com>
2666 * src/arm/ffi.c (ffi_prep_cif_machdep): Handle functions that return
2667 long long values. Round stack allocation to a multiple of 8 bytes
2668 for ATPCS compatibility.
2669 * src/arm/sysv.S (ffi_call_SYSV): Rework to avoid use of APCS register
2670 names. Handle returning long long types. Add Thumb and interworking
2671 support. Improve soft-float code.
2673 2004-10-27 Richard Earnshaw <rearnsha@arm.com>
2675 * testsuite/lib/libffi-db.exp (load_gcc_lib): New function.
2676 (libffi_exit): New function.
2677 (libffi_init): Build the testglue wrapper if needed.
2679 2004-10-25 Eric Botcazou <ebotcazou@libertysurf.fr>
2682 * testsuite/lib/libffi-dg.exp: Accept more than one multilib libgcc.
2684 2004-10-25 Kazuhiro Inaoka <inaoka.kazuhiro@renesas.com>
2686 * src/m32r/libffitarget.h (FFI_CLOSURES): Set to 0.
2688 2004-10-20 Kaz Kojima <kkojima@gcc.gnu.org>
2690 * src/sh/sysv.S (ffi_call_SYSV): Don't align for double data.
2691 * testsuite/libffi.call/float3.c: New test case.
2693 2004-10-18 Kaz Kojima <kkojima@gcc.gnu.org>
2695 * src/sh/ffi.c (ffi_prep_closure): Set T bit in trampoline for
2696 the function returning a structure pointed with R2.
2697 * src/sh/sysv.S (ffi_closure_SYSV): Use R2 as the pointer to
2698 the structure return value if T bit set. Emit position
2699 independent code and EH data if PIC.
2701 2004-10-13 Kazuhiro Inaoka <inaoka.kazuhiro@renesas.com>
2703 * Makefile.am: Add m32r support.
2704 * configure.ac: Likewise.
2705 * Makefile.in: Regenerate.
2706 * confiugre: Regenerate.
2707 * src/types.c: Add m32r port to FFI_INTERNAL_TYPEDEF
2708 (uint64, sint64, double, longdouble)
2709 * src/m32r: New directory.
2710 * src/m32r/ffi.c: New file.
2711 * src/m32r/sysv.S: Likewise.
2712 * src/m32r/ffitarget.h: Likewise.
2714 2004-10-02 Kaz Kojima <kkojima@gcc.gnu.org>
2716 * testsuite/libffi.call/negint.c: New test case.
2718 2004-09-14 H.J. Lu <hongjiu.lu@intel.com>
2721 * testsuite/lib/libffi-dg.exp: Don't use global ld_library_path.
2722 Set up LD_LIBRARY_PATH, SHLIB_PATH, LD_LIBRARYN32_PATH,
2723 LD_LIBRARY64_PATH, LD_LIBRARY_PATH_32, LD_LIBRARY_PATH_64 and
2726 2004-09-05 Andreas Tobler <a.tobler@schweiz.ch>
2728 * testsuite/libffi.call/many_win32.c: Remove whitespaces.
2729 * testsuite/libffi.call/promotion.c: Likewise.
2730 * testsuite/libffi.call/return_ll.c: Remove unused var. Cleanup
2732 * testsuite/libffi.call/return_sc.c: Likewise.
2733 * testsuite/libffi.call/return_uc.c: Likewise.
2735 2004-09-05 Andreas Tobler <a.tobler@schweiz.ch>
2737 * src/powerpc/darwin.S: Fix comments and identation.
2738 * src/powerpc/darwin_closure.S: Likewise.
2740 2004-09-02 Andreas Tobler <a.tobler@schweiz.ch>
2742 * src/powerpc/ffi_darwin.c: Add flag for longdouble return values.
2743 (ffi_prep_args): Handle longdouble arguments.
2744 (ffi_prep_cif_machdep): Set flags for longdouble. Calculate space for
2746 (ffi_closure_helper_DARWIN): Add closure handling for longdouble.
2747 * src/powerpc/darwin.S (_ffi_call_DARWIN): Add handling of longdouble
2749 * src/powerpc/darwin_closure.S (_ffi_closure_ASM): Likewise.
2750 * src/types.c: Defined longdouble size and alignment for darwin.
2752 2004-09-02 Andreas Tobler <a.tobler@schweiz.ch>
2754 * src/powerpc/aix.S: Remove whitespaces.
2755 * src/powerpc/aix_closure.S: Likewise.
2756 * src/powerpc/asm.h: Likewise.
2757 * src/powerpc/ffi.c: Likewise.
2758 * src/powerpc/ffitarget.h: Likewise.
2759 * src/powerpc/linux64.S: Likewise.
2760 * src/powerpc/linux64_closure.S: Likewise.
2761 * src/powerpc/ppc_closure.S: Likewise.
2762 * src/powerpc/sysv.S: Likewise.
2764 2004-08-30 Anthony Green <green@redhat.com>
2766 * Makefile.am: Add frv support.
2767 * Makefile.in, testsuite/Makefile.in: Rebuilt.
2768 * configure.ac: Read configure.host.
2769 * configure.in: Read configure.host.
2770 * configure.host: New file. frv-elf needs libgloss.
2771 * include/ffi.h.in: Force ffi_closure to have a nice big (8)
2772 alignment. This is needed to frv and shouldn't harm the others.
2773 * include/ffi_common.h (ALIGN_DOWN): New macro.
2774 * src/frv/ffi.c, src/frv/ffitarget.h, src/frv/eabi.S: New files.
2776 2004-08-24 David Daney <daney@avtrex.com>
2778 * testsuite/libffi.call/closure_fn0.c: Xfail mips64* instead of mips*.
2779 * testsuite/libffi.call/closure_fn1.c: Likewise.
2780 * testsuite/libffi.call/closure_fn2.c Likewise.
2781 * testsuite/libffi.call/closure_fn3.c: Likewise.
2782 * testsuite/libffi.call/closure_fn4.c: Likewise.
2783 * testsuite/libffi.call/closure_fn5.c: Likewise.
2784 * testsuite/libffi.call/cls_18byte.c: Likewise.
2785 * testsuite/libffi.call/cls_19byte.c: Likewise.
2786 * testsuite/libffi.call/cls_1_1byte.c: Likewise.
2787 * testsuite/libffi.call/cls_20byte.c: Likewise.
2788 * testsuite/libffi.call/cls_20byte1.c: Likewise.
2789 * testsuite/libffi.call/cls_24byte.c: Likewise.
2790 * testsuite/libffi.call/cls_2byte.c: Likewise.
2791 * testsuite/libffi.call/cls_3_1byte.c: Likewise.
2792 * testsuite/libffi.call/cls_3byte1.c: Likewise.
2793 * testsuite/libffi.call/cls_3byte2.c: Likewise.
2794 * testsuite/libffi.call/cls_4_1byte.c: Likewise.
2795 * testsuite/libffi.call/cls_4byte.c: Likewise.
2796 * testsuite/libffi.call/cls_64byte.c: Likewise.
2797 * testsuite/libffi.call/cls_6byte.c: Likewise.
2798 * testsuite/libffi.call/cls_7byte.c: Likewise.
2799 * testsuite/libffi.call/cls_8byte.c: Likewise.
2800 * testsuite/libffi.call/cls_9byte1.c: Likewise.
2801 * testsuite/libffi.call/cls_9byte2.c: Likewise.
2802 * testsuite/libffi.call/cls_align_double.c: Likewise.
2803 * testsuite/libffi.call/cls_align_float.c: Likewise.
2804 * testsuite/libffi.call/cls_align_longdouble.c: Likewise.
2805 * testsuite/libffi.call/cls_align_pointer.c: Likewise.
2806 * testsuite/libffi.call/cls_align_sint16.c: Likewise.
2807 * testsuite/libffi.call/cls_align_sint32.c: Likewise.
2808 * testsuite/libffi.call/cls_align_sint64.c: Likewise.
2809 * testsuite/libffi.call/cls_align_uint16.c: Likewise.
2810 * testsuite/libffi.call/cls_align_uint32.c: Likewise.
2811 * testsuite/libffi.call/cls_align_uint64.c: Likewise.
2812 * testsuite/libffi.call/cls_double.c: Likewise.
2813 * testsuite/libffi.call/cls_float.c: Likewise.
2814 * testsuite/libffi.call/cls_multi_schar.c: Likewise.
2815 * testsuite/libffi.call/cls_multi_sshort.c: Likewise.
2816 * testsuite/libffi.call/cls_multi_sshortchar.c: Likewise.
2817 * testsuite/libffi.call/cls_multi_uchar.c: Likewise.
2818 * testsuite/libffi.call/cls_multi_ushort.c: Likewise.
2819 * testsuite/libffi.call/cls_multi_ushortchar.c: Likewise.
2820 * testsuite/libffi.call/cls_schar.c: Likewise.
2821 * testsuite/libffi.call/cls_sint.c: Likewise.
2822 * testsuite/libffi.call/cls_sshort.c: Likewise.
2823 * testsuite/libffi.call/cls_uchar.c: Likewise.
2824 * testsuite/libffi.call/cls_uint.c: Likewise.
2825 * testsuite/libffi.call/cls_ulonglong.c: Likewise.
2826 * testsuite/libffi.call/cls_ushort.c: Likewise.
2827 * testsuite/libffi.call/nested_struct.c: Likewise.
2828 * testsuite/libffi.call/nested_struct1.c: Likewise.
2829 * testsuite/libffi.call/nested_struct2.c: Likewise.
2830 * testsuite/libffi.call/nested_struct3.c: Likewise.
2831 * testsuite/libffi.call/problem1.c: Likewise.
2832 * testsuite/libffi.special/unwindtest.cc: Likewise.
2833 * testsuite/libffi.call/cls_12byte.c: Likewise and set return value
2835 * testsuite/libffi.call/cls_16byte.c: Likewise.
2836 * testsuite/libffi.call/cls_5byte.c: Likewise.
2838 2004-08-23 David Daney <daney@avtrex.com>
2841 * src/mips/ffitarget.h (FFI_O32_SOFT_FLOAT): New ABI.
2842 * src/mips/ffi.c (ffi_prep_args): Fix alignment calculation.
2843 (ffi_prep_cif_machdep): Handle FFI_O32_SOFT_FLOAT floating point
2844 parameters and return types.
2845 (ffi_call): Handle FFI_O32_SOFT_FLOAT ABI.
2846 (ffi_prep_closure): Ditto.
2847 (ffi_closure_mips_inner_O32): Handle FFI_O32_SOFT_FLOAT ABI, fix
2848 alignment calculations.
2849 * src/mips/o32.S (ffi_closure_O32): Don't use floating point
2850 instructions if FFI_O32_SOFT_FLOAT, make stack frame ABI compliant.
2852 2004-08-14 Casey Marshall <csm@gnu.org>
2854 * src/mips/ffi.c (ffi_pref_cif_machdep): set `cif->flags' to
2855 contain `FFI_TYPE_UINT64' as return type for any 64-bit
2856 integer (O32 ABI only).
2857 (ffi_prep_closure): new function.
2858 (ffi_closure_mips_inner_O32): new function.
2859 * src/mips/ffitarget.h: Define `FFI_CLOSURES' and
2860 `FFI_TRAMPOLINE_SIZE' appropriately if the ABI is o32.
2861 * src/mips/o32.S (ffi_call_O32): add labels for .eh_frame. Return
2862 64 bit integers correctly.
2863 (ffi_closure_O32): new function.
2864 Added DWARF-2 unwind info for both functions.
2866 2004-08-10 Andrew Haley <aph@redhat.com>
2868 * src/x86/ffi64.c (ffi_prep_args ): 8-align all stack arguments.
2870 2004-08-01 Robert Millan <robertmh@gnu.org>
2872 * configure.ac: Detect knetbsd-gnu and kfreebsd-gnu.
2873 * configure: Regenerate.
2875 2004-07-30 Maciej W. Rozycki <macro@linux-mips.org>
2877 * acinclude.m4 (AC_FUNC_MMAP_BLACKLIST): Check for <sys/mman.h>
2878 and mmap() explicitly instead of relying on preset autoconf cache
2880 * aclocal.m4: Regenerate.
2881 * configure: Regenerate.
2883 2004-07-11 Ulrich Weigand <uweigand@de.ibm.com>
2885 * src/s390/ffi.c (ffi_prep_args): Fix C aliasing violation.
2886 (ffi_check_float_struct): Remove unused prototype.
2888 2004-06-30 Geoffrey Keating <geoffk@apple.com>
2890 * src/powerpc/ffi_darwin.c (flush_icache): ';' is a comment
2891 character on Darwin, use '\n\t' instead.
2893 2004-06-26 Matthias Klose <doko@debian.org>
2895 * libtool-version: Fix typo in revision/age.
2897 2004-06-17 Matthias Klose <doko@debian.org>
2899 * libtool-version: New.
2900 * Makefile.am (libffi_la_LDFLAGS): Use -version-info for soname.
2901 * Makefile.in: Regenerate.
2903 2004-06-15 Paolo Bonzini <bonzini@gnu.org>
2905 * Makefile.am: Remove useless multilib rules.
2906 * Makefile.in: Regenerate.
2907 * aclocal.m4: Regenerate with automake 1.8.5.
2908 * configure.ac: Remove useless multilib configury.
2909 * configure: Regenerate.
2911 2004-06-15 Paolo Bonzini <bonzini@gnu.org>
2913 * .cvsignore: New file.
2915 2004-06-10 Jakub Jelinek <jakub@redhat.com>
2917 * src/ia64/unix.S (ffi_call_unix): Insert group barrier break
2919 (ffi_closure_UNIX): Fix f14/f15 adjustment if FLOAT_SZ is ever
2922 2004-06-06 Sean McNeil <sean@mcneil.com>
2924 * configure.ac: Add x86_64-*-freebsd* support.
2925 * configure: Regenerate.
2927 2004-04-26 Joe Buck <jbuck@welsh-buck.org>
2930 * configure.ac: Test for existence of mmap and sys/mman.h before
2931 checking blacklist. Fix suggested by Jim Wilson.
2932 * configure: Regenerate.
2934 2004-04-26 Matt Austern <austern@apple.com>
2936 * src/powerpc/darwin.S: Go through a non-lazy pointer for initial
2938 * src/powerpc/darwin_closure.S: Likewise.
2940 2004-04-24 Andreas Tobler <a.tobler@schweiz.ch>
2942 * testsuite/libffi.call/cls_multi_schar.c (main): Fix initialization
2943 error. Reported by Thomas Heller <theller@python.net>.
2944 * testsuite/libffi.call/cls_multi_sshort.c (main): Likewise.
2945 * testsuite/libffi.call/cls_multi_ushort.c (main): Likewise.
2947 2004-03-20 Matthias Klose <doko@debian.org>
2949 * src/pa/linux.S: Fix typo.
2951 2004-03-19 Matthias Klose <doko@debian.org>
2953 * Makefile.am: Update.
2954 * Makefile.in: Regenerate.
2955 * src/pa/ffi.h.in: Remove.
2956 * src/pa/ffitarget.h: New file.
2958 2004-02-10 Randolph Chung <tausq@debian.org>
2960 * Makefile.am: Add PA support.
2961 * Makefile.in: Regenerate.
2962 * include/Makefile.in: Regenerate.
2963 * configure.ac: Add PA target.
2964 * configure: Regenerate.
2965 * src/pa/ffi.c: New file.
2966 * src/pa/ffi.h.in: Add PA support.
2967 * src/pa/linux.S: New file.
2968 * prep_cif.c: Add PA support.
2970 2004-03-16 Hosaka Yuji <hos@tamanegi.org>
2972 * src/types.c: Fix alignment size of X86_WIN32 case int64 and
2974 * src/x86/ffi.c (ffi_prep_args): Replace ecif->cif->rtype->type
2975 with ecif->cif->flags.
2976 (ffi_call, ffi_prep_incoming_args_SYSV): Replace cif->rtype->type
2978 (ffi_prep_cif_machdep): Add X86_WIN32 struct case.
2979 (ffi_closure_SYSV): Add 1 or 2-bytes struct case for X86_WIN32.
2980 * src/x86/win32.S (retstruct1b, retstruct2b, sc_retstruct1b,
2981 sc_retstruct2b): Add for 1 or 2-bytes struct case.
2983 2004-03-15 Kelley Cook <kcook@gcc.gnu.org>
2985 * configure.in: Rename file to ...
2986 * configure.ac: ... this.
2987 * fficonfig.h.in: Regenerate.
2988 * Makefile.in: Regenerate.
2989 * include/Makefile.in: Regenerate.
2990 * testsuite/Makefile.in: Regenerate.
2992 2004-03-12 Matt Austern <austern@apple.com>
2994 * src/powerpc/darwin.S: Fix EH information so it corresponds to
2995 changes in EH format resulting from addition of linkonce support.
2996 * src/powerpc/darwin_closure.S: Likewise.
2998 2004-03-11 Andreas Tobler <a.tobler@schweiz.ch>
2999 Paolo Bonzini <bonzini@gnu.org>
3001 * Makefile.am (AUTOMAKE_OPTIONS): Set them.
3002 Remove VPATH. Remove rules for object files. Remove multilib support.
3003 (AM_CCASFLAGS): Add.
3004 * configure.in (AC_CONFIG_HEADERS): Relace AM_CONFIG_HEADER.
3005 (AC_PREREQ): Bump version to 2.59.
3006 (AC_INIT): Fill with version info and bug address.
3007 (ORIGINAL_LD_FOR_MULTILIBS): Remove.
3008 (AM_ENABLE_MULTILIB): Use this instead of AC_ARG_ENABLE.
3009 De-precious CC so that the right flags are passed down to multilibs.
3010 (AC_MSG_ERROR): Replace obsolete macro AC_ERROR.
3011 (AC_CONFIG_FILES): Replace obsolete macro AC_LINK_FILES.
3012 (AC_OUTPUT): Reorganize the output with AC_CONFIG_COMMANDS.
3013 * configure: Rebuilt.
3014 * aclocal.m4: Likewise.
3015 * Makefile.in, include/Makefile.in, testsuite/Makefile.in: Likewise.
3016 * fficonfig.h.in: Likewise.
3018 2004-03-11 Andreas Schwab <schwab@suse.de>
3020 * src/ia64/ffi.c (ffi_prep_incoming_args_UNIX): Get floating point
3021 arguments from fp registers only for the first 8 parameter slots.
3022 Don't convert a float parameter when passed in memory.
3024 2004-03-09 Hans-Peter Nilsson <hp@axis.com>
3026 * configure: Regenerate for config/accross.m4 correction.
3028 2004-02-25 Matt Kraai <kraai@alumni.cmu.edu>
3030 * src/powerpc/ffi.c (ffi_prep_args_SYSV): Change
3031 ecif->cif->bytes to bytes.
3032 (ffi_prep_cif_machdep): Add braces around nested if statement.
3034 2004-02-09 Alan Modra <amodra@bigpond.net.au>
3036 * src/types.c (pointer): POWERPC64 has 8 byte pointers.
3038 * src/powerpc/ffi.c (ffi_prep_args64): Correct long double handling.
3039 (ffi_closure_helper_LINUX64): Fix typo.
3040 * testsuite/libffi.call/cls_align_longdouble.c: Pass -mlong-double-128
3042 * testsuite/libffi.call/float.c: Likewise.
3043 * testsuite/libffi.call/float2.c: Likewise.
3045 2004-02-08 Alan Modra <amodra@bigpond.net.au>
3047 * src/powerpc/ffi.c (ffi_prep_cif_machdep <FFI_LINUX64>): Correct
3048 long double function return and long double arg handling.
3049 (ffi_closure_helper_LINUX64): Formatting. Delete unused "ng" var.
3050 Use "end_pfr" instead of "nf". Correct long double handling.
3052 * src/powerpc/linux64.S (ffi_call_LINUX64): Save f2 long double
3054 * src/powerpc/linux64_closure.S (ffi_closure_LINUX64): Allocate
3055 space for long double return value. Adjust stack frame and offsets.
3056 Load f2 long double return.
3058 2004-02-07 Alan Modra <amodra@bigpond.net.au>
3060 * src/types.c: Use 16 byte long double for POWERPC64.
3062 2004-01-25 Eric Botcazou <ebotcazou@libertysurf.fr>
3064 * src/sparc/ffi.c (ffi_prep_args_v9): Shift the parameter array
3065 when the structure return address is passed in %o0.
3066 (ffi_V9_return_struct): Rename into ffi_v9_layout_struct.
3067 (ffi_v9_layout_struct): Align the field following a nested structure
3068 on a word boundary. Use memmove instead of memcpy.
3069 (ffi_call): Update call to ffi_V9_return_struct.
3070 (ffi_prep_closure): Define 'ctx' only for V8.
3071 (ffi_closure_sparc_inner): Clone into ffi_closure_sparc_inner_v8
3072 and ffi_closure_sparc_inner_v9.
3073 (ffi_closure_sparc_inner_v8): Return long doubles by reference.
3074 Always skip the structure return address. For structures and long
3075 doubles, copy the argument directly.
3076 (ffi_closure_sparc_inner_v9): Skip the structure return address only
3077 if required. Shift the maximum floating-point slot accordingly. For
3078 big structures, copy the argument directly; otherwise, left-justify the
3079 argument and call ffi_v9_layout_struct to lay out the structure on
3081 * src/sparc/v8.S: Undef STACKFRAME before defining it.
3082 (ffi_closure_v8): Pass the structure return address. Update call to
3083 ffi_closure_sparc_inner_v8. Short-circuit FFI_TYPE_INT handling.
3084 Skip the 'unimp' insn when returning long doubles and structures.
3085 * src/sparc/v9.S: Undef STACKFRAME before defining it.
3086 (ffi_closure_v9): Increase the frame size by 2 words. Short-circuit
3087 FFI_TYPE_INT handling. Load structures both in integers and
3088 floating-point registers on return.
3089 * README: Update status of the SPARC port.
3091 2004-01-24 Andreas Tobler <a.tobler@schweiz.ch>
3093 * testsuite/libffi.call/pyobjc-tc.c (main): Treat result value
3095 * testsuite/libffi.call/struct3.c (main): Fix CHECK.
3097 2004-01-22 Ulrich Weigand <uweigand@de.ibm.com>
3099 * testsuite/libffi.call/cls_uint.c (cls_ret_uint_fn): Treat result
3100 value as of type ffi_arg, not unsigned int.
3102 2004-01-21 Michael Ritzert <ritzert@t-online.de>
3104 * ffi64.c (ffi_prep_args): Cast the RHS of an assignment instead
3107 2004-01-12 Andreas Tobler <a.tobler@schweiz.ch>
3109 * testsuite/lib/libffi-dg.exp: Set LD_LIBRARY_PATH_32 for
3112 2004-01-08 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
3114 * testsuite/libffi.call/ffitest.h (allocate_mmap): Cast MAP_FAILED
3117 2003-12-10 Richard Henderson <rth@redhat.com>
3119 * testsuite/libffi.call/cls_align_pointer.c: Cast pointers to
3120 size_t instead of int.
3122 2003-12-04 Hosaka Yuji <hos@tamanegi.org>
3124 * testsuite/libffi.call/many_win32.c: Include <float.h>.
3125 * testsuite/libffi.call/many_win32.c (main): Replace variable
3126 int i with unsigned long ul.
3128 * testsuite/libffi.call/cls_align_uint64.c: New test case.
3129 * testsuite/libffi.call/cls_align_sint64.c: Likewise.
3130 * testsuite/libffi.call/cls_align_uint32.c: Likewise.
3131 * testsuite/libffi.call/cls_align_sint32.c: Likewise.
3132 * testsuite/libffi.call/cls_align_uint16.c: Likewise.
3133 * testsuite/libffi.call/cls_align_sint16.c: Likewise.
3134 * testsuite/libffi.call/cls_align_float.c: Likewise.
3135 * testsuite/libffi.call/cls_align_double.c: Likewise.
3136 * testsuite/libffi.call/cls_align_longdouble.c: Likewise.
3137 * testsuite/libffi.call/cls_align_pointer.c: Likewise.
3139 2003-12-02 Hosaka Yuji <hos@tamanegi.org>
3142 * src/x86/ffi.c (ffi_prep_args, ffi_prep_incoming_args_SYSV):
3143 Align arguments to 32 bits.
3145 2003-12-01 Andreas Tobler <a.tobler@schweiz.ch>
3148 * testsuite/libffi.call/cls_multi_sshort.c: New test case.
3149 * testsuite/libffi.call/cls_multi_sshortchar.c: Likewise.
3150 * testsuite/libffi.call/cls_multi_uchar.c: Likewise.
3151 * testsuite/libffi.call/cls_multi_schar.c: Likewise.
3152 * testsuite/libffi.call/cls_multi_ushortchar.c: Likewise.
3153 * testsuite/libffi.call/cls_multi_ushort.c: Likewise.
3155 * testsuite/libffi.special/unwindtest.cc: Cosmetics.
3157 2003-11-26 Kaveh R. Ghazi <ghazi@caip.rutgers.edu>
3159 * testsuite/libffi.call/ffitest.h: Include <fcntl.h>.
3160 * testsuite/libffi.special/ffitestcxx.h: Likewise.
3162 2003-11-22 Andreas Tobler <a.tobler@schweiz.ch>
3164 * Makefile.in: Rebuilt.
3165 * configure: Likewise.
3166 * testsuite/libffi.special/unwindtest.cc: Convert the mmap to
3169 2003-11-21 Andreas Jaeger <aj@suse.de>
3170 Andreas Tobler <a.tobler@schweiz.ch>
3172 * acinclude.m4: Add AC_FUNC_MMAP_BLACKLIST.
3173 * configure.in: Call AC_FUNC_MMAP_BLACKLIST.
3174 * Makefile.in: Rebuilt.
3175 * aclocal.m4: Likewise.
3176 * configure: Likewise.
3177 * fficonfig.h.in: Likewise.
3178 * testsuite/lib/libffi-dg.exp: Add include dir.
3179 * testsuite/libffi.call/ffitest.h: Add MMAP definitions.
3180 * testsuite/libffi.special/ffitestcxx.h: Likewise.
3181 * testsuite/libffi.call/closure_fn0.c: Use MMAP functionality
3182 for ffi_closure if available.
3183 * testsuite/libffi.call/closure_fn1.c: Likewise.
3184 * testsuite/libffi.call/closure_fn2.c: Likewise.
3185 * testsuite/libffi.call/closure_fn3.c: Likewise.
3186 * testsuite/libffi.call/closure_fn4.c: Likewise.
3187 * testsuite/libffi.call/closure_fn5.c: Likewise.
3188 * testsuite/libffi.call/cls_12byte.c: Likewise.
3189 * testsuite/libffi.call/cls_16byte.c: Likewise.
3190 * testsuite/libffi.call/cls_18byte.c: Likewise.
3191 * testsuite/libffi.call/cls_19byte.c: Likewise.
3192 * testsuite/libffi.call/cls_1_1byte.c: Likewise.
3193 * testsuite/libffi.call/cls_20byte.c: Likewise.
3194 * testsuite/libffi.call/cls_20byte1.c: Likewise.
3195 * testsuite/libffi.call/cls_24byte.c: Likewise.
3196 * testsuite/libffi.call/cls_2byte.c: Likewise.
3197 * testsuite/libffi.call/cls_3_1byte.c: Likewise.
3198 * testsuite/libffi.call/cls_3byte1.c: Likewise.
3199 * testsuite/libffi.call/cls_3byte2.c: Likewise.
3200 * testsuite/libffi.call/cls_4_1byte.c: Likewise.
3201 * testsuite/libffi.call/cls_4byte.c: Likewise.
3202 * testsuite/libffi.call/cls_5byte.c: Likewise.
3203 * testsuite/libffi.call/cls_64byte.c: Likewise.
3204 * testsuite/libffi.call/cls_6byte.c: Likewise.
3205 * testsuite/libffi.call/cls_7byte.c: Likewise.
3206 * testsuite/libffi.call/cls_8byte.c: Likewise.
3207 * testsuite/libffi.call/cls_9byte1.c: Likewise.
3208 * testsuite/libffi.call/cls_9byte2.c: Likewise.
3209 * testsuite/libffi.call/cls_double.c: Likewise.
3210 * testsuite/libffi.call/cls_float.c: Likewise.
3211 * testsuite/libffi.call/cls_schar.c: Likewise.
3212 * testsuite/libffi.call/cls_sint.c: Likewise.
3213 * testsuite/libffi.call/cls_sshort.c: Likewise.
3214 * testsuite/libffi.call/cls_uchar.c: Likewise.
3215 * testsuite/libffi.call/cls_uint.c: Likewise.
3216 * testsuite/libffi.call/cls_ulonglong.c: Likewise.
3217 * testsuite/libffi.call/cls_ushort.c: Likewise.
3218 * testsuite/libffi.call/nested_struct.c: Likewise.
3219 * testsuite/libffi.call/nested_struct1.c: Likewise.
3220 * testsuite/libffi.call/nested_struct2.c: Likewise.
3221 * testsuite/libffi.call/nested_struct3.c: Likewise.
3222 * testsuite/libffi.call/problem1.c: Likewise.
3223 * testsuite/libffi.special/unwindtest.cc: Likewise.
3225 2003-11-20 Andreas Tobler <a.tobler@schweiz.ch>
3227 * testsuite/lib/libffi-dg.exp: Make the -lgcc_s conditional.
3229 2003-11-19 Andreas Tobler <a.tobler@schweiz.ch>
3231 * testsuite/lib/libffi-dg.exp: Add DYLD_LIBRARY_PATH for darwin.
3232 Add -lgcc_s to additional flags.
3234 2003-11-12 Andreas Tobler <a.tobler@schweiz.ch>
3236 * configure.in, include/Makefile.am: PR libgcj/11147, install
3237 the ffitarget.h header file in a gcc versioned and target
3239 * configure: Regenerated.
3240 * Makefile.in, include/Makefile.in: Likewise.
3241 * testsuite/Makefile.in: Likewise.
3243 2003-11-09 Andreas Tobler <a.tobler@schweiz.ch>
3245 * testsuite/libffi.call/closure_fn0.c: Print result and check
3246 with dg-output to make debugging easier.
3247 * testsuite/libffi.call/closure_fn1.c: Likewise.
3248 * testsuite/libffi.call/closure_fn2.c: Likewise.
3249 * testsuite/libffi.call/closure_fn3.c: Likewise.
3250 * testsuite/libffi.call/closure_fn4.c: Likewise.
3251 * testsuite/libffi.call/closure_fn5.c: Likewise.
3252 * testsuite/libffi.call/cls_12byte.c: Likewise.
3253 * testsuite/libffi.call/cls_16byte.c: Likewise.
3254 * testsuite/libffi.call/cls_18byte.c: Likewise.
3255 * testsuite/libffi.call/cls_19byte.c: Likewise.
3256 * testsuite/libffi.call/cls_1_1byte.c: Likewise.
3257 * testsuite/libffi.call/cls_20byte.c: Likewise.
3258 * testsuite/libffi.call/cls_20byte1.c: Likewise.
3259 * testsuite/libffi.call/cls_24byte.c: Likewise.
3260 * testsuite/libffi.call/cls_2byte.c: Likewise.
3261 * testsuite/libffi.call/cls_3_1byte.c: Likewise.
3262 * testsuite/libffi.call/cls_3byte1.c: Likewise.
3263 * testsuite/libffi.call/cls_3byte2.c: Likewise.
3264 * testsuite/libffi.call/cls_4_1byte.c: Likewise.
3265 * testsuite/libffi.call/cls_4byte.c: Likewise.
3266 * testsuite/libffi.call/cls_5byte.c: Likewise.
3267 * testsuite/libffi.call/cls_64byte.c: Likewise.
3268 * testsuite/libffi.call/cls_6byte.c: Likewise.
3269 * testsuite/libffi.call/cls_7byte.c: Likewise.
3270 * testsuite/libffi.call/cls_8byte.c: Likewise.
3271 * testsuite/libffi.call/cls_9byte1.c: Likewise.
3272 * testsuite/libffi.call/cls_9byte2.c: Likewise.
3273 * testsuite/libffi.call/cls_double.c: Likewise.
3274 * testsuite/libffi.call/cls_float.c: Likewise.
3275 * testsuite/libffi.call/cls_schar.c: Likewise.
3276 * testsuite/libffi.call/cls_sint.c: Likewise.
3277 * testsuite/libffi.call/cls_sshort.c: Likewise.
3278 * testsuite/libffi.call/cls_uchar.c: Likewise.
3279 * testsuite/libffi.call/cls_uint.c: Likewise.
3280 * testsuite/libffi.call/cls_ulonglong.c: Likewise.
3281 * testsuite/libffi.call/cls_ushort.c: Likewise.
3282 * testsuite/libffi.call/problem1.c: Likewise.
3284 * testsuite/libffi.special/unwindtest.cc: Make ffi_closure
3287 2003-11-08 Andreas Tobler <a.tobler@schweiz.ch>
3289 * testsuite/libffi.call/cls_9byte2.c: New test case.
3290 * testsuite/libffi.call/cls_9byte1.c: Likewise.
3291 * testsuite/libffi.call/cls_64byte.c: Likewise.
3292 * testsuite/libffi.call/cls_20byte1.c: Likewise.
3293 * testsuite/libffi.call/cls_19byte.c: Likewise.
3294 * testsuite/libffi.call/cls_18byte.c: Likewise.
3295 * testsuite/libffi.call/closure_fn4.c: Likewise.
3296 * testsuite/libffi.call/closure_fn5.c: Likewise.
3297 * testsuite/libffi.call/cls_schar.c: Likewise.
3298 * testsuite/libffi.call/cls_sint.c: Likewise.
3299 * testsuite/libffi.call/cls_sshort.c: Likewise.
3300 * testsuite/libffi.call/nested_struct2.c: Likewise.
3301 * testsuite/libffi.call/nested_struct3.c: Likewise.
3303 2003-11-08 Andreas Tobler <a.tobler@schweiz.ch>
3305 * testsuite/libffi.call/cls_double.c: Do a check on the result.
3306 * testsuite/libffi.call/cls_uchar.c: Likewise.
3307 * testsuite/libffi.call/cls_uint.c: Likewise.
3308 * testsuite/libffi.call/cls_ulonglong.c: Likewise.
3309 * testsuite/libffi.call/cls_ushort.c: Likewise.
3310 * testsuite/libffi.call/return_sc.c: Cleanup whitespaces.
3312 2003-11-06 Andreas Tobler <a.tobler@schweiz.ch>
3314 * src/prep_cif.c (ffi_prep_cif): Move the validity check after
3317 2003-10-23 Andreas Tobler <a.tobler@schweiz.ch>
3319 * src/java_raw_api.c (ffi_java_ptrarray_to_raw): Replace
3320 FFI_ASSERT(FALSE) with FFI_ASSERT(0).
3322 2003-10-22 David Daney <ddaney@avtrex.com>
3324 * src/mips/ffitarget.h: Replace undefined UINT32 and friends with
3325 __attribute__((__mode__(__SI__))) and friends.
3327 2003-10-22 Andreas Schwab <schwab@suse.de>
3329 * src/ia64/ffi.c: Replace FALSE/TRUE with false/true.
3331 2003-10-21 Andreas Tobler <a.tobler@schweiz.ch>
3333 * configure.in: AC_LINK_FILES(ffitarget.h).
3334 * configure: Regenerate.
3335 * Makefile.in: Likewise.
3336 * include/Makefile.in: Likewise.
3337 * testsuite/Makefile.in: Likewise.
3338 * fficonfig.h.in: Likewise.
3340 2003-10-21 Paolo Bonzini <bonzini@gnu.org>
3341 Richard Henderson <rth@redhat.com>
3343 Avoid that ffi.h includes fficonfig.h.
3345 * Makefile.am (EXTRA_DIST): Include ffitarget.h files
3346 (TARGET_SRC_MIPS_GCC): Renamed to TARGET_SRC_MIPS_IRIX.
3347 (TARGET_SRC_MIPS_SGI): Removed.
3348 (MIPS_GCC): Renamed to TARGET_SRC_MIPS_IRIX.
3349 (MIPS_SGI): Removed.
3350 (CLEANFILES): Removed.
3351 (mostlyclean-am, clean-am, mostlyclean-sub, clean-sub): New
3353 * acconfig.h: Removed.
3354 * configure.in: Compute sizeofs only for double and long double.
3355 Use them to define and subst HAVE_LONG_DOUBLE. Include comments
3356 into AC_DEFINE instead of using acconfig.h. Create
3357 include/ffitarget.h instead of include/fficonfig.h. Rename
3358 MIPS_GCC to MIPS_IRIX, drop MIPS_SGI since we are in gcc's tree.
3359 AC_DEFINE EH_FRAME_FLAGS.
3360 * include/Makefile.am (DISTCLEANFILES): New automake macro.
3361 (hack_DATA): Add ffitarget.h.
3362 * include/ffi.h.in: Remove all system specific definitions.
3363 Declare raw API even if it is not installed, why bother?
3364 Use limits.h instead of SIZEOF_* to define ffi_type_*. Do
3365 not define EH_FRAME_FLAGS, it is in fficonfig.h now. Include
3366 ffitarget.h instead of fficonfig.h. Remove ALIGN macro.
3367 (UINT_ARG, INT_ARG): Removed, use ffi_arg and ffi_sarg instead.
3368 * include/ffi_common.h (bool): Do not define.
3369 (ffi_assert): Accept failed assertion.
3370 (ffi_type_test): Return void and accept file/line.
3371 (FFI_ASSERT): Pass stringized failed assertion.
3372 (FFI_ASSERT_AT): New macro.
3373 (FFI_ASSERT_VALID_TYPE): New macro.
3374 (UINT8, SINT8, UINT16, SINT16, UINT32, SINT32,
3375 UINT64, SINT64): Define here with gcc's __attribute__ macro
3377 (FLOAT32, ALIGN): Define here instead of in ffi.h
3378 * include/ffi-mips.h: Removed. Its content moved to
3379 src/mips/ffitarget.h after separating assembly and C sections.
3380 * src/alpha/ffi.c, src/alpha/ffi.c, src/java_raw_api.c
3381 src/prep_cif.c, src/raw_api.c, src/ia64/ffi.c,
3382 src/mips/ffi.c, src/mips/n32.S, src/mips/o32.S,
3383 src/mips/ffitarget.h, src/sparc/ffi.c, src/x86/ffi64.c:
3384 SIZEOF_ARG -> FFI_SIZEOF_ARG.
3385 * src/ia64/ffi.c: Include stdbool.h (provided by GCC 2.95+).
3386 * src/debug.c (ffi_assert): Accept stringized failed assertion.
3387 (ffi_type_test): Rewritten.
3388 * src/prep-cif.c (initialize_aggregate, ffi_prep_cif): Call
3389 FFI_ASSERT_VALID_TYPE.
3390 * src/alpha/ffitarget.h, src/arm/ffitarget.h,
3391 src/ia64/ffitarget.h, src/m68k/ffitarget.h,
3392 src/mips/ffitarget.h, src/powerpc/ffitarget.h,
3393 src/s390/ffitarget.h, src/sh/ffitarget.h,
3394 src/sh64/ffitarget.h, src/sparc/ffitarget.h,
3395 src/x86/ffitarget.h: New files.
3396 * src/alpha/osf.S, src/arm/sysv.S, src/ia64/unix.S,
3397 src/m68k/sysv.S, src/mips/n32.S, src/mips/o32.S,
3398 src/powerpc/aix.S, src/powerpc/darwin.S,
3399 src/powerpc/ffi_darwin.c, src/powerpc/linux64.S,
3400 src/powerpc/linux64_closure.S, src/powerpc/ppc_closure.S,
3401 src/powerpc/sysv.S, src/s390/sysv.S, src/sh/sysv.S,
3402 src/sh64/sysv.S, src/sparc/v8.S, src/sparc/v9.S,
3403 src/x86/sysv.S, src/x86/unix64.S, src/x86/win32.S:
3406 2003-10-20 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
3408 * src/mips/ffi.c: Use _ABIN32, _ABIO32 instead of external
3409 _MIPS_SIM_NABI32, _MIPS_SIM_ABI32.
3411 2003-10-19 Andreas Tobler <a.tobler@schweiz.ch>
3413 * src/powerpc/ffi_darwin.c (ffi_prep_args): Declare bytes again.
3414 Used when FFI_DEBUG = 1.
3416 2003-10-14 Alan Modra <amodra@bigpond.net.au>
3418 * src/types.c (double, longdouble): Default POWERPC64 to 8 byte size
3421 2003-10-06 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
3423 * include/ffi_mips.h: Define FFI_MIPS_N32 for N32/N64 ABIs,
3424 FFI_MIPS_O32 for O32 ABI.
3426 2003-10-01 Andreas Tobler <a.tobler@schweiz.ch>
3428 * testsuite/lib/libffi-dg.exp: Set LD_LIBRARY_PATH_64 for
3429 SPARC64. Cleanup whitespaces.
3431 2003-09-19 Andreas Tobler <a.tobler@schweiz.ch>
3433 * testsuite/libffi.call/closure_fn0.c: Xfail mips, arm,
3434 strongarm, xscale. Cleanup whitespaces.
3435 * testsuite/libffi.call/closure_fn1.c: Likewise.
3436 * testsuite/libffi.call/closure_fn2.c: Likewise.
3437 * testsuite/libffi.call/closure_fn3.c: Likewise.
3438 * testsuite/libffi.call/cls_12byte.c: Likewise.
3439 * testsuite/libffi.call/cls_16byte.c: Likewise.
3440 * testsuite/libffi.call/cls_1_1byte.c: Likewise.
3441 * testsuite/libffi.call/cls_20byte.c: Likewise.
3442 * testsuite/libffi.call/cls_24byte.c: Likewise.
3443 * testsuite/libffi.call/cls_2byte.c: Likewise.
3444 * testsuite/libffi.call/cls_3_1byte.c: Likewise.
3445 * testsuite/libffi.call/cls_3byte1.c: Likewise.
3446 * testsuite/libffi.call/cls_3byte2.c: Likewise.
3447 * testsuite/libffi.call/cls_4_1byte.c: Likewise.
3448 * testsuite/libffi.call/cls_4byte.c: Likewise.
3449 * testsuite/libffi.call/cls_5byte.c: Likewise.
3450 * testsuite/libffi.call/cls_6byte.c: Likewise.
3451 * testsuite/libffi.call/cls_7byte.c: Likewise.
3452 * testsuite/libffi.call/cls_8byte.c: Likewise.
3453 * testsuite/libffi.call/cls_double.c: Likewise.
3454 * testsuite/libffi.call/cls_float.c: Likewise.
3455 * testsuite/libffi.call/cls_uchar.c: Likewise.
3456 * testsuite/libffi.call/cls_uint.c: Likewise.
3457 * testsuite/libffi.call/cls_ulonglong.c: Likewise.
3458 * testsuite/libffi.call/cls_ushort.c: Likewise.
3459 * testsuite/libffi.call/nested_struct.c: Likewise.
3460 * testsuite/libffi.call/nested_struct1.c: Likewise.
3461 * testsuite/libffi.call/problem1.c: Likewise.
3462 * testsuite/libffi.special/unwindtest.cc: Likewise.
3463 * testsuite/libffi.call/pyobjc-tc.c: Cleanup whitespaces.
3465 2003-09-18 David Edelsohn <edelsohn@gnu.org>
3467 * src/powerpc/aix.S: Cleanup whitespaces.
3468 * src/powerpc/aix_closure.S: Likewise.
3470 2003-09-18 Andreas Tobler <a.tobler@schweiz.ch>
3472 * src/powerpc/darwin.S: Cleanup whitespaces, comment formatting.
3473 * src/powerpc/darwin_closure.S: Likewise.
3474 * src/powerpc/ffi_darwin.c: Likewise.
3476 2003-09-18 Andreas Tobler <a.tobler@schweiz.ch>
3477 David Edelsohn <edelsohn@gnu.org>
3479 * src/types.c (double): Add AIX and Darwin to the right TYPEDEF.
3480 * src/powerpc/aix_closure.S: Remove the pointer to the outgoing
3482 * src/powerpc/darwin_closure.S: Likewise.
3483 * src/powerpc/ffi_darwin.c (ffi_prep_args): Handle structures
3484 according to the Darwin/AIX ABI.
3485 (ffi_prep_cif_machdep): Likewise.
3486 (ffi_closure_helper_DARWIN): Likewise.
3487 Remove the outgoing parameter stack logic. Simplify the evaluation
3488 of the different CASE types.
3489 (ffi_prep_clousure): Avoid the casts on lvalues. Change the branch
3490 statement in the trampoline code.
3492 2003-09-18 Kaz Kojima <kkojima@gcc.gnu.org>
3494 * src/sh/ffi.c (ffi_prep_args): Take account into the alignement
3495 for the register size.
3496 (ffi_closure_helper_SYSV): Handle the structure return value
3498 (ffi_closure_helper_SYSV): Return the appropriate type when
3499 the registers are used for the structure return value.
3500 * src/sh/sysv.S (ffi_closure_SYSV): Fix the stack layout for
3501 the 64-bit return value. Update copyright years.
3503 2003-09-17 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
3505 * testsuite/lib/libffi-dg.exp (libffi_target_compile): Search in
3506 srcdir for ffi_mips.h.
3508 2003-09-12 Alan Modra <amodra@bigpond.net.au>
3510 * src/prep_cif.c (initialize_aggregate): Include tail padding in
3512 * src/powerpc/linux64_closure.S (ffi_closure_LINUX64): Correct
3513 placement of float result.
3514 * testsuite/libffi.special/unwindtest.cc (closure_test_fn1): Correct
3515 cast of "resp" for big-endian 64 bit machines.
3517 2003-09-11 Alan Modra <amodra@bigpond.net.au>
3519 * src/types.c (double, longdouble): Merge identical SH and ARM
3520 typedefs, and add POWERPC64.
3521 * src/powerpc/ffi.c (ffi_prep_args64): Correct next_arg calc for
3522 struct split over gpr and rest.
3523 (ffi_prep_cif_machdep): Correct intarg_count for structures.
3524 * src/powerpc/linux64.S (ffi_call_LINUX64): Fix gpr offsets.
3526 2003-09-09 Andreas Tobler <a.tobler@schweiz.ch>
3528 * src/powerpc/ffi.c (ffi_closure_helper_SYSV) Handle struct
3531 2003-09-09 Alan Modra <amodra@bigpond.net.au>
3533 * configure: Regenerate.
3535 2003-09-04 Andreas Tobler <a.tobler@schweiz.ch>
3537 * Makefile.am: Remove build rules for ffitest.
3538 * Makefile.in: Rebuilt.
3540 2003-09-04 Andreas Tobler <a.tobler@schweiz.ch>
3542 * src/java_raw_api.c: Include <stdlib.h> to fix compiler warning
3543 about implicit declaration of abort().
3545 2003-09-04 Andreas Tobler <a.tobler@schweiz.ch>
3547 * Makefile.am: Add dejagnu test framework. Fixes PR other/11411.
3548 * Makefile.in: Rebuilt.
3549 * configure.in: Add dejagnu test framework.
3550 * configure: Rebuilt.
3552 * testsuite/Makefile.am: New file.
3553 * testsuite/Makefile.in: Built
3554 * testsuite/lib/libffi-dg.exp: New file.
3555 * testsuite/config/default.exp: Likewise.
3556 * testsuite/libffi.call/call.exp: Likewise.
3557 * testsuite/libffi.call/ffitest.h: Likewise.
3558 * testsuite/libffi.call/closure_fn0.c: Likewise.
3559 * testsuite/libffi.call/closure_fn1.c: Likewise.
3560 * testsuite/libffi.call/closure_fn2.c: Likewise.
3561 * testsuite/libffi.call/closure_fn3.c: Likewise.
3562 * testsuite/libffi.call/cls_1_1byte.c: Likewise.
3563 * testsuite/libffi.call/cls_3_1byte.c: Likewise.
3564 * testsuite/libffi.call/cls_4_1byte.c: Likewise.
3565 * testsuite/libffi.call/cls_2byte.c: Likewise.
3566 * testsuite/libffi.call/cls_3byte1.c: Likewise.
3567 * testsuite/libffi.call/cls_3byte2.c: Likewise.
3568 * testsuite/libffi.call/cls_4byte.c: Likewise.
3569 * testsuite/libffi.call/cls_5byte.c: Likewise.
3570 * testsuite/libffi.call/cls_6byte.c: Likewise.
3571 * testsuite/libffi.call/cls_7byte.c: Likewise.
3572 * testsuite/libffi.call/cls_8byte.c: Likewise.
3573 * testsuite/libffi.call/cls_12byte.c: Likewise.
3574 * testsuite/libffi.call/cls_16byte.c: Likewise.
3575 * testsuite/libffi.call/cls_20byte.c: Likewise.
3576 * testsuite/libffi.call/cls_24byte.c: Likewise.
3577 * testsuite/libffi.call/cls_double.c: Likewise.
3578 * testsuite/libffi.call/cls_float.c: Likewise.
3579 * testsuite/libffi.call/cls_uchar.c: Likewise.
3580 * testsuite/libffi.call/cls_uint.c: Likewise.
3581 * testsuite/libffi.call/cls_ulonglong.c: Likewise.
3582 * testsuite/libffi.call/cls_ushort.c: Likewise.
3583 * testsuite/libffi.call/float.c: Likewise.
3584 * testsuite/libffi.call/float1.c: Likewise.
3585 * testsuite/libffi.call/float2.c: Likewise.
3586 * testsuite/libffi.call/many.c: Likewise.
3587 * testsuite/libffi.call/many_win32.c: Likewise.
3588 * testsuite/libffi.call/nested_struct.c: Likewise.
3589 * testsuite/libffi.call/nested_struct1.c: Likewise.
3590 * testsuite/libffi.call/pyobjc-tc.c: Likewise.
3591 * testsuite/libffi.call/problem1.c: Likewise.
3592 * testsuite/libffi.call/promotion.c: Likewise.
3593 * testsuite/libffi.call/return_ll.c: Likewise.
3594 * testsuite/libffi.call/return_sc.c: Likewise.
3595 * testsuite/libffi.call/return_uc.c: Likewise.
3596 * testsuite/libffi.call/strlen.c: Likewise.
3597 * testsuite/libffi.call/strlen_win32.c: Likewise.
3598 * testsuite/libffi.call/struct1.c: Likewise.
3599 * testsuite/libffi.call/struct2.c: Likewise.
3600 * testsuite/libffi.call/struct3.c: Likewise.
3601 * testsuite/libffi.call/struct4.c: Likewise.
3602 * testsuite/libffi.call/struct5.c: Likewise.
3603 * testsuite/libffi.call/struct6.c: Likewise.
3604 * testsuite/libffi.call/struct7.c: Likewise.
3605 * testsuite/libffi.call/struct8.c: Likewise.
3606 * testsuite/libffi.call/struct9.c: Likewise.
3607 * testsuite/libffi.special/special.exp: New file.
3608 * testsuite/libffi.special/ffitestcxx.h: Likewise.
3609 * testsuite/libffi.special/unwindtest.cc: Likewise.
3612 2003-08-13 Kaz Kojima <kkojima@gcc.gnu.org>
3614 * src/sh/ffi.c (OFS_INT16): Set 0 for little endian case. Update
3617 2003-08-02 Alan Modra <amodra@bigpond.net.au>
3619 * src/powerpc/ffi.c (ffi_prep_args64): Modify for changed gcc
3621 (ffi_closure_helper_LINUX64): Likewise.
3622 * src/powerpc/linux64.S: Remove code writing to parm save area.
3623 * src/powerpc/linux64_closure.S (ffi_closure_LINUX64): Use return
3624 address in lr from ffi_closure_helper_LINUX64 call to calculate
3625 table address. Optimize function tail.
3627 2003-07-28 Andreas Tobler <a.tobler@schweiz.ch>
3629 * src/sparc/ffi.c: Handle all floating point registers.
3630 * src/sparc/v9.S: Likewise. Fixes second part of PR target/11410.
3632 2003-07-11 Gerald Pfeifer <pfeifer@dbai.tuwien.ac.at>
3634 * README: Note that libffi is not part of GCC. Update the project
3637 2003-06-19 Franz Sirl <Franz.Sirl-kernel@lauterbach.com>
3639 * src/powerpc/ppc_closure.S: Include ffi.h.
3641 2003-06-13 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
3643 * src/x86/sysv.S: Avoid gas-only .uleb128/.sleb128 directives.
3644 Use C style comments.
3646 2003-06-13 Kaz Kojima <kkojima@rr.iij4u.or.jp>
3648 * Makefile.am: Add SHmedia support. Fix a typo of SH support.
3649 * Makefile.in: Regenerate.
3650 * configure.in (sh64-*-linux*, sh5*-*-linux*): Add target.
3651 * configure: Regenerate.
3652 * include/ffi.h.in: Add SHmedia support.
3653 * src/sh64/ffi.c: New file.
3654 * src/sh64/sysv.S: New file.
3656 2003-05-16 Jakub Jelinek <jakub@redhat.com>
3658 * configure.in (HAVE_RO_EH_FRAME): Check whether .eh_frame section
3659 should be read-only.
3660 * configure: Rebuilt.
3661 * fficonfig.h.in: Rebuilt.
3662 * include/ffi.h.in (EH_FRAME_FLAGS): Define.
3663 * src/alpha/osf.S: Use EH_FRAME_FLAGS.
3664 * src/powerpc/linux64.S: Likewise.
3665 * src/powerpc/linux64_closure.S: Likewise. Include ffi.h.
3666 * src/powerpc/sysv.S: Use EH_FRAME_FLAGS. Use pcrel encoding
3667 if -fpic/-fPIC/-mrelocatable.
3668 * src/powerpc/powerpc_closure.S: Likewise.
3669 * src/sparc/v8.S: If HAVE_RO_EH_FRAME is defined, don't include
3670 #write in .eh_frame flags.
3671 * src/sparc/v9.S: Likewise.
3672 * src/x86/unix64.S: Use EH_FRAME_FLAGS.
3673 * src/x86/sysv.S: Likewise. Use pcrel encoding if -fpic/-fPIC.
3674 * src/s390/sysv.S: Use EH_FRAME_FLAGS. Include ffi.h.
3676 2003-05-07 Jeff Sturm <jsturm@one-point.com>
3678 Fixes PR bootstrap/10656
3679 * configure.in (HAVE_AS_REGISTER_PSEUDO_OP): Test assembler
3680 support for .register pseudo-op.
3681 * src/sparc/v8.S: Use it.
3682 * fficonfig.h.in: Rebuilt.
3683 * configure: Rebuilt.
3685 2003-04-18 Jakub Jelinek <jakub@redhat.com>
3687 * include/ffi.h.in (POWERPC64): Define if 64-bit.
3688 (enum ffi_abi): Add FFI_LINUX64 on POWERPC.
3689 Make it the default on POWERPC64.
3690 (FFI_TRAMPOLINE_SIZE): Define to 24 on POWERPC64.
3691 * configure.in: Change powerpc-*-linux* into powerpc*-*-linux*.
3692 * configure: Rebuilt.
3693 * src/powerpc/ffi.c (hidden): Define.
3694 (ffi_prep_args_SYSV): Renamed from
3695 ffi_prep_args. Cast pointers to unsigned long to shut up warnings.
3696 (NUM_GPR_ARG_REGISTERS64, NUM_FPR_ARG_REGISTERS64,
3697 ASM_NEEDS_REGISTERS64): New.
3698 (ffi_prep_args64): New function.
3699 (ffi_prep_cif_machdep): Handle FFI_LINUX64 ABI.
3700 (ffi_call): Likewise.
3701 (ffi_prep_closure): Likewise.
3702 (flush_icache): Surround by #ifndef POWERPC64.
3703 (ffi_dblfl): New union type.
3704 (ffi_closure_helper_SYSV): Use it to avoid aliasing problems.
3705 (ffi_closure_helper_LINUX64): New function.
3706 * src/powerpc/ppc_closure.S: Surround whole file by #ifndef
3708 * src/powerpc/sysv.S: Likewise.
3709 (ffi_call_SYSV): Rename ffi_prep_args to ffi_prep_args_SYSV.
3710 * src/powerpc/linux64.S: New file.
3711 * src/powerpc/linux64_closure.S: New file.
3712 * Makefile.am (EXTRA_DIST): Add src/powerpc/linux64.S and
3713 src/powerpc/linux64_closure.S.
3714 (TARGET_SRC_POWERPC): Likewise.
3716 * src/ffitest.c (closure_test_fn, closure_test_fn1, closure_test_fn2,
3717 closure_test_fn3): Fix result printing on big-endian 64-bit
3719 (main): Print tst2_arg instead of uninitialized tst2_result.
3721 * src/ffitest.c (main): Hide what closure pointer really points to
3724 2003-04-16 Richard Earnshaw <rearnsha@arm.com>
3726 * configure.in (arm-*-netbsdelf*): Add configuration.
3727 (configure): Regenerated.
3729 2003-04-04 Loren J. Rittle <ljrittle@acm.org>
3731 * include/Makefile.in: Regenerate.
3733 2003-03-21 Zdenek Dvorak <rakdver@atrey.karlin.mff.cuni.cz>
3735 * libffi/include/ffi.h.in: Define X86 instead of X86_64 in 32
3737 * libffi/src/x86/ffi.c (ffi_closure_SYSV, ffi_closure_raw_SYSV):
3738 Receive closure pointer through parameter, read args using
3739 __builtin_dwarf_cfa.
3740 (FFI_INIT_TRAMPOLINE): Send closure reference through eax.
3742 2003-03-12 Andreas Schwab <schwab@suse.de>
3744 * configure.in: Avoid trailing /. in toolexeclibdir.
3745 * configure: Rebuilt.
3747 2003-03-03 Andreas Tobler <a.tobler@schweiz.ch>
3749 * src/powerpc/darwin_closure.S: Recode to fit dynamic libraries.
3751 2003-02-06 Andreas Tobler <a.tobler@schweiz.ch>
3753 * libffi/src/powerpc/darwin_closure.S:
3754 Fix alignement bug, allocate 8 bytes for the result.
3755 * libffi/src/powerpc/aix_closure.S:
3757 * libffi/src/powerpc/ffi_darwin.c:
3758 Update stackframe description for aix/darwin_closure.S.
3760 2003-02-06 Jakub Jelinek <jakub@redhat.com>
3762 * src/s390/ffi.c (ffi_closure_helper_SYSV): Add hidden visibility
3765 2003-01-31 Christian Cornelssen <ccorn@cs.tu-berlin.de>,
3766 Andreas Schwab <schwab@suse.de>
3768 * configure.in: Adjust command to source config-ml.in to account
3769 for changes to the libffi_basedir definition.
3770 (libffi_basedir): Remove ${srcdir} from value and include trailing
3773 * configure: Regenerate.
3775 2003-01-29 Franz Sirl <Franz.Sirl-kernel@lauterbach.com>
3777 * src/powerpc/ppc_closure.S: Recode to fit shared libs.
3779 2003-01-28 Andrew Haley <aph@redhat.com>
3781 * include/ffi.h.in: Enable FFI_CLOSURES for x86_64.
3782 * src/x86/ffi64.c (ffi_prep_closure): New.
3783 (ffi_closure_UNIX64_inner): New.
3784 * src/x86/unix64.S (ffi_closure_UNIX64): New.
3786 2003-01-27 Alexandre Oliva <aoliva@redhat.com>
3788 * configure.in (toolexecdir, toolexeclibdir): Set and AC_SUBST.
3789 Remove USE_LIBDIR conditional.
3790 * Makefile.am (toolexecdir, toolexeclibdir): Don't override.
3791 * Makefile.in, configure: Rebuilt.
3793 2003-01027 David Edelsohn <edelsohn@gnu.org>
3795 * Makefile.am (TARGET_SRC_POWERPC_AIX): Fix typo.
3796 * Makefile.in: Regenerate.
3798 2003-01-22 Andrew Haley <aph@redhat.com>
3800 * src/powerpc/darwin.S (_ffi_call_AIX): Add Augmentation size to
3803 2003-01-21 Andreas Tobler <a.tobler@schweiz.ch>
3805 * src/powerpc/darwin.S: Add unwind info.
3806 * src/powerpc/darwin_closure.S: Likewise.
3808 2003-01-14 Andrew Haley <aph@redhat.com>
3810 * src/x86/ffi64.c (ffi_prep_args): Check for void retval.
3811 (ffi_prep_cif_machdep): Likewise.
3812 * src/x86/unix64.S: Add unwind info.
3814 2003-01-14 Andreas Jaeger <aj@suse.de>
3816 * src/ffitest.c (main): Only use ffi_closures if those are
3819 2003-01-13 Andreas Tobler <a.tobler@schweiz.ch>
3821 * libffi/src/ffitest.c
3822 add closure testcases
3824 2003-01-13 Kevin B. Hendricks <khendricks@ivey.uwo.ca>
3826 * libffi/src/powerpc/ffi.c
3827 fix alignment bug for float (4 byte aligned iso 8 byte)
3829 2003-01-09 Geoffrey Keating <geoffk@apple.com>
3831 * src/powerpc/ffi_darwin.c: Remove RCS version string.
3832 * src/powerpc/darwin.S: Remove RCS version string.
3834 2003-01-03 Jeff Sturm <jsturm@one-point.com>
3836 * include/ffi.h.in: Add closure defines for SPARC, SPARC64.
3837 * src/ffitest.c (main): Use static storage for closure.
3838 * src/sparc/ffi.c (ffi_prep_closure, ffi_closure_sparc_inner): New.
3839 * src/sparc/v8.S (ffi_closure_v8): New.
3840 * src/sparc/v9.S (ffi_closure_v9): New.
3842 2002-11-10 Ranjit Mathew <rmathew@hotmail.com>
3844 * include/ffi.h.in: Added FFI_STDCALL ffi_type
3845 enumeration for X86_WIN32.
3846 * src/x86/win32.S: Added ffi_call_STDCALL function
3848 * src/x86/ffi.c (ffi_call/ffi_raw_call): Added
3849 switch cases for recognising FFI_STDCALL and
3850 calling ffi_call_STDCALL if target is X86_WIN32.
3851 * src/ffitest.c (my_stdcall_strlen/stdcall_many):
3852 stdcall versions of the "my_strlen" and "many"
3853 test functions (for X86_WIN32).
3854 Added test cases to test stdcall invocation using
3857 2002-12-02 Kaz Kojima <kkojima@gcc.gnu.org>
3859 * src/sh/sysv.S: Add DWARF2 unwind info.
3861 2002-11-27 Ulrich Weigand <uweigand@de.ibm.com>
3863 * src/s390/sysv.S (.eh_frame section): Make section read-only.
3865 2002-11-26 Jim Wilson <wilson@redhat.com>
3867 * src/types.c (FFI_TYPE_POINTER): Has size 8 on IA64.
3869 2002-11-23 H.J. Lu <hjl@gnu.org>
3871 * acinclude.m4: Add dummy AM_PROG_LIBTOOL.
3872 Include ../config/accross.m4.
3873 * aclocal.m4; Rebuild.
3874 * configure: Likewise.
3876 2002-11-15 Ulrich Weigand <uweigand@de.ibm.com>
3878 * src/s390/sysv.S (.eh_frame section): Adapt to pcrel FDE encoding.
3880 2002-11-11 DJ Delorie <dj@redhat.com>
3882 * configure.in: Look for common files in the right place.
3884 2002-10-08 Ulrich Weigand <uweigand@de.ibm.com>
3886 * src/java_raw_api.c (ffi_java_raw_to_ptrarray): Interpret
3887 raw data as _Jv_word values, not ffi_raw.
3888 (ffi_java_ptrarray_to_raw): Likewise.
3889 (ffi_java_rvalue_to_raw): New function.
3890 (ffi_java_raw_call): Call it.
3891 (ffi_java_raw_to_rvalue): New function.
3892 (ffi_java_translate_args): Call it.
3893 * src/ffitest.c (closure_test_fn): Interpret return value
3894 as ffi_arg, not int.
3895 * src/s390/ffi.c (ffi_prep_cif_machdep): Add missing
3896 FFI_TYPE_POINTER case.
3897 (ffi_closure_helper_SYSV): Likewise. Also, assume return
3898 values extended to word size.
3900 2002-10-02 Andreas Jaeger <aj@suse.de>
3902 * src/x86/ffi64.c (ffi_prep_cif_machdep): Remove debug output.
3904 2002-10-01 Bo Thorsen <bo@smetana.suse.de>
3906 * include/ffi.h.in: Fix i386 win32 compilation.
3908 2002-09-30 Ulrich Weigand <uweigand@de.ibm.com>
3910 * configure.in: Add s390x-*-linux-* target.
3911 * configure: Regenerate.
3912 * include/ffi.h.in: Define S390X for s390x targets.
3913 (FFI_CLOSURES): Define for s390/s390x.
3914 (FFI_TRAMPOLINE_SIZE): Likewise.
3915 (FFI_NATIVE_RAW_API): Likewise.
3916 * src/prep_cif.c (ffi_prep_cif): Do not compute stack space for s390.
3917 * src/types.c (FFI_TYPE_POINTER): Use 8-byte pointers on s390x.
3918 * src/s390/ffi.c: Major rework of existing code. Add support for
3919 s390x targets. Add closure support.
3920 * src/s390/sysv.S: Likewise.
3922 2002-09-29 Richard Earnshaw <rearnsha@arm.com>
3924 * src/arm/sysv.S: Fix typo.
3926 2002-09-28 Richard Earnshaw <rearnsha@arm.com>
3928 * src/arm/sysv.S: If we don't have machine/asm.h and the pre-processor
3929 has defined __USER_LABEL_PREFIX__, then use it in CNAME.
3930 (ffi_call_SYSV): Handle soft-float.
3932 2002-09-27 Bo Thorsen <bo@suse.de>
3934 * include/ffi.h.in: Fix multilib x86-64 support.
3936 2002-09-22 Kaveh R. Ghazi <ghazi@caip.rutgers.edu>
3938 * Makefile.am (all-multi): Fix multilib parallel build.
3940 2002-07-19 Kaz Kojima <kkojima@gcc.gnu.org>
3942 * configure.in (sh[34]*-*-linux*): Add brackets.
3943 * configure: Regenerate.
3945 2002-07-18 Kaz Kojima <kkojima@gcc.gnu.org>
3947 * Makefile.am: Add SH support.
3948 * Makefile.in: Regenerate.
3949 * configure.in (sh-*-linux*, sh[34]*-*-linux*): Add target.
3950 * configure: Regenerate.
3951 * include/ffi.h.in: Add SH support.
3952 * src/sh/ffi.c: New file.
3953 * src/sh/sysv.S: New file.
3954 * src/types.c: Add SH support.
3956 2002-07-16 Bo Thorsen <bo@suse.de>
3958 * src/x86/ffi64.c: New file that adds x86-64 support.
3959 * src/x86/unix64.S: New file that handles argument setup for
3961 * src/x86/sysv.S: Don't use this on x86-64.
3962 * src/x86/ffi.c: Don't use this on x86-64.
3964 * src/prep_cif.c (ffi_prep_cif): Don't do stack size calculation
3966 * src/ffitest.c (struct6): New test that tests a special case in
3968 (struct7): Likewise.
3969 (struct8): Likewise.
3970 (struct9): Likewise.
3971 (closure_test_fn): Silence warning about this when it's not used.
3972 (main): Add the new tests.
3973 (main): Fix a couple of wrong casts and silence some compiler warnings.
3974 * include/ffi.h.in: Add x86-64 ABI definition.
3975 * fficonfig.h.in: Regenerate.
3976 * Makefile.am: Add x86-64 support.
3977 * configure.in: Likewise.
3978 * Makefile.in: Regenerate.
3979 * configure: Likewise.
3981 2002-06-24 Bo Thorsen <bo@suse.de>
3983 * src/types.c: Merge settings for similar architectures.
3984 Add x86-64 sizes and alignments.
3986 2002-06-23 Bo Thorsen <bo@suse.de>
3988 * src/arm/ffi.c (ffi_prep_args): Remove unused vars.
3989 * src/sparc/ffi.c (ffi_prep_args_v8): Likewise.
3990 * src/mips/ffi.c (ffi_prep_args): Likewise.
3991 * src/m68k/ffi.c (ffi_prep_args): Likewise.
3993 2002-07-18 H.J. Lu (hjl@gnu.org)
3995 * Makefile.am (TARGET_SRC_MIPS_LINUX): New.
3996 (libffi_la_SOURCES): Support MIPS_LINUX.
3997 (libffi_convenience_la_SOURCES): Likewise.
3998 * Makefile.in: Regenerated.
4000 * configure.in (mips64*-*): Skip.
4001 (mips*-*-linux*): New.
4002 * configure: Regenerated.
4004 * src/mips/ffi.c: Include <sgidefs.h>.
4006 2002-06-06 Ulrich Weigand <uweigand@de.ibm.com>
4008 * src/s390/sysv.S: Save/restore %r6. Add DWARF-2 unwind info.
4010 2002-05-27 Roger Sayle <roger@eyesopen.com>
4012 * src/x86/ffi.c (ffi_prep_args): Remove reference to avn.
4014 2002-05-27 Bo Thorsen <bo@suse.de>
4016 * src/x86/ffi.c (ffi_prep_args): Remove unused variable and
4019 2002-05-13 Andreas Tobler <a.tobler@schweiz.ch>
4021 * src/powerpc/ffi_darwin.c (ffi_prep_closure): Declare fd at
4022 beginning of function (for older apple cc).
4024 2002-05-08 Alexandre Oliva <aoliva@redhat.com>
4026 * configure.in (ORIGINAL_LD_FOR_MULTILIBS): Preserve LD at
4027 script entry, and set LD to it when configuring multilibs.
4028 * configure: Rebuilt.
4030 2002-05-05 Jason Thorpe <thorpej@wasabisystems.com>
4032 * configure.in (sparc64-*-netbsd*): Add target.
4033 (sparc-*-netbsdelf*): Likewise.
4034 * configure: Regenerate.
4036 2002-04-28 David S. Miller <davem@redhat.com>
4038 * configure.in, configure: Fix SPARC test in previous change.
4040 2002-04-29 Gerhard Tonn <GerhardTonn@swol.de>
4042 * Makefile.am: Add Linux for S/390 support.
4043 * Makefile.in: Regenerate.
4044 * configure.in: Add Linux for S/390 support.
4045 * configure: Regenerate.
4046 * include/ffi.h.in: Add Linux for S/390 support.
4047 * src/s390/ffi.c: New file from libffi CVS tree.
4048 * src/s390/sysv.S: New file from libffi CVS tree.
4050 2002-04-28 Jakub Jelinek <jakub@redhat.com>
4052 * configure.in (HAVE_AS_SPARC_UA_PCREL): Check for working
4054 * src/sparc/v8.S: Use it.
4055 * src/sparc/v9.S: Likewise.
4056 * fficonfig.h.in: Rebuilt.
4057 * configure: Rebuilt.
4059 2002-04-08 Hans Boehm <Hans_Boehm@hp.com>
4061 * src/java_raw_api.c (ffi_java_raw_size): Handle FFI_TYPE_DOUBLE
4063 * src/ia64/unix.S: Add unwind information. Fix comments.
4064 Save sp in a way that's compatible with unwind info.
4065 (ffi_call_unix): Correctly restore sp in all cases.
4066 * src/ia64/ffi.c: Add, fix comments.
4068 2002-04-08 Jakub Jelinek <jakub@redhat.com>
4070 * src/sparc/v8.S: Make .eh_frame dependent on target word size.
4072 2002-04-06 Jason Thorpe <thorpej@wasabisystems.com>
4074 * configure.in (alpha*-*-netbsd*): Add target.
4075 * configure: Regenerate.
4077 2002-04-04 Jeff Sturm <jsturm@one-point.com>
4079 * src/sparc/v8.S: Add unwind info.
4080 * src/sparc/v9.S: Likewise.
4082 2002-03-30 Krister Walfridsson <cato@df.lth.se>
4084 * configure.in: Enable i*86-*-netbsdelf*.
4085 * configure: Rebuilt.
4087 2002-03-29 David Billinghurst <David.Billinghurst@riotinto.com>
4090 * src/mips/n32.s: Delete
4091 * src/mips/o32.s: Delete
4093 2002-03-21 Loren J. Rittle <ljrittle@acm.org>
4095 * configure.in: Enable alpha*-*-freebsd*.
4096 * configure: Rebuilt.
4098 2002-03-17 Bryce McKinlay <bryce@waitaki.otago.ac.nz>
4100 * Makefile.am: libfficonvenience -> libffi_convenience.
4101 * Makefile.in: Rebuilt.
4103 * Makefile.am: Define ffitest_OBJECTS.
4104 * Makefile.in: Rebuilt.
4106 2002-03-07 Andreas Tobler <toa@pop.agri.ch>
4107 David Edelsohn <edelsohn@gnu.org>
4109 * Makefile.am (EXTRA_DIST): Add Darwin and AIX closure files.
4110 (TARGET_SRC_POWERPC_AIX): Add aix_closure.S.
4111 (TARGET_SRC_POWERPC_DARWIN): Add darwin_closure.S.
4112 * Makefile.in: Regenerate.
4113 * include/ffi.h.in: Add AIX and Darwin closure definitions.
4114 * src/powerpc/ffi_darwin.c (ffi_prep_closure): New function.
4115 (flush_icache, flush_range): New functions.
4116 (ffi_closure_helper_DARWIN): New function.
4117 * src/powerpc/aix_closure.S: New file.
4118 * src/powerpc/darwin_closure.S: New file.
4120 2002-02-24 Jeff Sturm <jsturm@one-point.com>
4122 * include/ffi.h.in: Add typedef for ffi_arg.
4123 * src/ffitest.c (main): Declare rint with ffi_arg.
4125 2002-02-21 Andreas Tobler <toa@pop.agri.ch>
4127 * src/powerpc/ffi_darwin.c (ffi_prep_args): Skip appropriate
4128 number of GPRs for floating-point arguments.
4130 2002-01-31 Anthony Green <green@redhat.com>
4132 * configure: Rebuilt.
4133 * configure.in: Replace CHECK_SIZEOF and endian tests with
4134 cross-compiler friendly macros.
4135 * aclocal.m4 (AC_COMPILE_CHECK_SIZEOF, AC_C_BIGENDIAN_CROSS): New
4138 2002-01-18 David Edelsohn <edelsohn@gnu.org>
4140 * src/powerpc/darwin.S (_ffi_call_AIX): New.
4141 * src/powerpc/aix.S (ffi_call_DARWIN): New.
4143 2002-01-17 David Edelsohn <edelsohn@gnu.org>
4145 * Makefile.am (EXTRA_DIST): Add Darwin and AIX files.
4146 (TARGET_SRC_POWERPC_AIX): New.
4147 (POWERPC_AIX): New stanza.
4148 * Makefile.in: Regenerate.
4149 * configure.in: Add AIX case.
4150 * configure: Regenerate.
4151 * include/ffi.h.in (ffi_abi): Add FFI_AIX.
4152 * src/powerpc/ffi_darwin.c (ffi_status): Use "long" to scale frame
4153 size. Fix "long double" support.
4154 (ffi_call): Add FFI_AIX case.
4155 * src/powerpc/aix.S: New.
4157 2001-10-09 John Hornkvist <john@toastedmarshmallow.com>
4159 Implement Darwin PowerPC ABI.
4160 * configure.in: Handle powerpc-*-darwin*.
4161 * Makefile.am: Set source files for POWERPC_DARWIN.
4162 * configure: Rebuilt.
4163 * Makefile.in: Rebuilt.
4164 * include/ffi.h.in: Define FFI_DARWIN and FFI_DEFAULT_ABI for
4166 * src/powerpc/darwin.S: New file.
4167 * src/powerpc/ffi_darwin.c: New file.
4169 2001-10-07 Joseph S. Myers <jsm28@cam.ac.uk>
4171 * src/x86/ffi.c: Fix spelling error of "separate" as "seperate".
4173 2001-07-16 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
4175 * src/x86/sysv.S: Avoid gas-only .balign directive.
4176 Use C style comments.
4178 2001-07-16 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
4180 * src/alpha/ffi.c (ffi_prep_closure): Avoid gas-only mnemonic.
4181 Fixes PR bootstrap/3563.
4183 2001-06-26 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
4185 * src/alpha/osf.S (ffi_closure_osf): Use .rdata for ECOFF.
4187 2001-06-25 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
4189 * configure.in: Recognize sparc*-sun-* host.
4190 * configure: Regenerate.
4192 2001-06-06 Andrew Haley <aph@redhat.com>
4194 * src/alpha/osf.S (__FRAME_BEGIN__): Conditionalize for ELF.
4196 2001-06-03 Andrew Haley <aph@redhat.com>
4198 * src/alpha/osf.S: Add unwind info.
4199 * src/powerpc/sysv.S: Add unwind info.
4200 * src/powerpc/ppc_closure.S: Likewise.
4202 2000-05-31 Jeff Sturm <jsturm@one-point.com>
4204 * configure.in: Fix AC_ARG_ENABLE usage.
4205 * configure: Rebuilt.
4207 2001-05-06 Bryce McKinlay <bryce@waitaki.otago.ac.nz>
4209 * configure.in: Remove warning about beta code.
4210 * configure: Rebuilt.
4212 2001-04-25 Hans Boehm <Hans_Boehm@hp.com>
4214 * src/ia64/unix.S: Restore stack pointer when returning from
4216 * src/ia64/ffi.c: Fix typo in comment.
4218 2001-04-18 Jim Wilson <wilson@redhat.com>
4220 * src/ia64/unix.S: Delete unnecessary increment and decrement of loc2
4221 to eliminate RAW DV.
4223 2001-04-12 Bryce McKinlay <bryce@albatross.co.nz>
4225 * Makefile.am: Make a libtool convenience library.
4226 * Makefile.in: Rebuilt.
4228 2001-03-29 Bryce McKinlay <bryce@albatross.co.nz>
4230 * configure.in: Use different syntax for subdirectory creation.
4231 * configure: Rebuilt.
4233 2001-03-27 Jon Beniston <jon@beniston.com>
4235 * configure.in: Added X86_WIN32 target (Win32, CygWin, MingW).
4236 * configure: Rebuilt.
4237 * Makefile.am: Added X86_WIN32 target support.
4238 * Makefile.in: Rebuilt.
4240 * include/ffi.h.in: Added X86_WIN32 target support.
4242 * src/ffitest.c: Doesn't run structure tests for X86_WIN32 targets.
4243 * src/types.c: Added X86_WIN32 target support.
4245 * src/x86/win32.S: New file. Based on sysv.S, but with EH
4246 stuff removed and made to work with CygWin's gas.
4248 2001-03-26 Bryce McKinlay <bryce@albatross.co.nz>
4250 * configure.in: Make target subdirectory in build dir.
4251 * Makefile.am: Override suffix based rules to specify correct output
4253 * Makefile.in: Rebuilt.
4254 * configure: Rebuilt.
4256 2001-03-23 Kevin B Hendricks <khendricks@ivey.uwo.ca>
4258 * src/powerpc/ppc_closure.S: New file.
4259 * src/powerpc/ffi.c (ffi_prep_args): Fixed ABI compatibility bug
4260 involving long long and register pairs.
4261 (ffi_prep_closure): New function.
4262 (flush_icache): Likewise.
4263 (ffi_closure_helper_SYSV): Likewise.
4264 * include/ffi.h.in (FFI_CLOSURES): Define on PPC.
4265 (FFI_TRAMPOLINE_SIZE): Likewise.
4266 (FFI_NATIVE_RAW_API): Likewise.
4267 * Makefile.in: Rebuilt.
4268 * Makefile.am (EXTRA_DIST): Added src/powerpc/ppc_closure.S.
4269 (TARGET_SRC_POWERPC): Likewise.
4271 2001-03-19 Tom Tromey <tromey@redhat.com>
4273 * Makefile.in: Rebuilt.
4274 * Makefile.am (ffitest_LDFLAGS): New macro.
4276 2001-03-02 Nick Clifton <nickc@redhat.com>
4278 * include/ffi.h.in: Remove RCS ident string.
4279 * include/ffi_mips.h: Remove RCS ident string.
4280 * src/debug.c: Remove RCS ident string.
4281 * src/ffitest.c: Remove RCS ident string.
4282 * src/prep_cif.c: Remove RCS ident string.
4283 * src/types.c: Remove RCS ident string.
4284 * src/alpha/ffi.c: Remove RCS ident string.
4285 * src/alpha/osf.S: Remove RCS ident string.
4286 * src/arm/ffi.c: Remove RCS ident string.
4287 * src/arm/sysv.S: Remove RCS ident string.
4288 * src/mips/ffi.c: Remove RCS ident string.
4289 * src/mips/n32.S: Remove RCS ident string.
4290 * src/mips/o32.S: Remove RCS ident string.
4291 * src/sparc/ffi.c: Remove RCS ident string.
4292 * src/sparc/v8.S: Remove RCS ident string.
4293 * src/sparc/v9.S: Remove RCS ident string.
4294 * src/x86/ffi.c: Remove RCS ident string.
4295 * src/x86/sysv.S: Remove RCS ident string.
4297 2001-02-08 Joseph S. Myers <jsm28@cam.ac.uk>
4299 * include/ffi.h.in: Change sourceware.cygnus.com references to
4302 2000-12-09 Richard Henderson <rth@redhat.com>
4304 * src/alpha/ffi.c (ffi_call): Simplify struct return test.
4305 (ffi_closure_osf_inner): Index rather than increment avalue
4306 and arg_types. Give ffi_closure_osf the raw return value type.
4307 * src/alpha/osf.S (ffi_closure_osf): Handle return value type
4310 2000-12-07 Richard Henderson <rth@redhat.com>
4312 * src/raw_api.c (ffi_translate_args): Fix typo.
4313 (ffi_prep_closure): Likewise.
4315 * include/ffi.h.in [ALPHA]: Define FFI_CLOSURES and
4316 FFI_TRAMPOLINE_SIZE.
4317 * src/alpha/ffi.c (ffi_prep_cif_machdep): Adjust minimal
4318 cif->bytes for new ffi_call_osf implementation.
4319 (ffi_prep_args): Absorb into ...
4320 (ffi_call): ... here. Do all stack allocation here and
4321 avoid a callback function.
4322 (ffi_prep_closure, ffi_closure_osf_inner): New.
4323 * src/alpha/osf.S (ffi_call_osf): Reimplement with no callback.
4324 (ffi_closure_osf): New.
4326 2000-09-10 Alexandre Oliva <aoliva@redhat.com>
4328 * config.guess, config.sub, install-sh: Removed.
4329 * ltconfig, ltmain.sh, missing, mkinstalldirs: Likewise.
4330 * Makefile.in: Rebuilt.
4332 * acinclude.m4: Include libtool macros from the top level.
4333 * aclocal.m4, configure: Rebuilt.
4335 2000-08-22 Alexandre Oliva <aoliva@redhat.com>
4337 * configure.in [i*86-*-freebsd*] (TARGET, TARGETDIR): Set.
4338 * configure: Rebuilt.
4340 2000-05-11 Scott Bambrough <scottb@netwinder.org>
4342 * libffi/src/arm/sysv.S (ffi_call_SYSV): Doubles are not saved to
4343 memory correctly. Use conditional instructions, not branches where
4346 2000-05-04 Tom Tromey <tromey@cygnus.com>
4348 * configure: Rebuilt.
4349 * configure.in: Match `arm*-*-linux-*'.
4350 From Chris Dornan <cdornan@arm.com>.
4352 2000-04-28 Jakub Jelinek <jakub@redhat.com>
4354 * Makefile.am (SUBDIRS): Define.
4355 (AM_MAKEFLAGS): Likewise.
4356 (Multilib support.): Add section.
4357 * Makefile.in: Rebuilt.
4358 * ltconfig (extra_compiler_flags, extra_compiler_flags_value):
4359 New variables. Set for gcc using -print-multi-lib. Export them
4361 (sparc64-*-linux-gnu*): Use libsuff 64 for search paths.
4362 * ltmain.sh (B|b|V): Don't throw away gcc's -B, -b and -V options
4364 (extra_compiler_flags_value, extra_compiler_flags): Check these
4365 for extra compiler options which need to be passed down in
4368 2000-04-16 Anthony Green <green@redhat.com>
4370 * configure: Rebuilt.
4371 * configure.in: Change i*86-pc-linux* to i*86-*-linux*.
4373 2000-04-14 Jakub Jelinek <jakub@redhat.com>
4375 * include/ffi.h.in (SPARC64): Define for 64bit SPARC builds.
4376 Set SPARC FFI_DEFAULT_ABI based on SPARC64 define.
4377 * src/sparc/ffi.c (ffi_prep_args_v8): Renamed from ffi_prep_args.
4378 Replace all void * sizeofs with sizeof(int).
4379 Only compare type with FFI_TYPE_LONGDOUBLE if LONGDOUBLE is
4380 different than DOUBLE.
4381 Remove FFI_TYPE_SINT32 and FFI_TYPE_UINT32 cases (handled elsewhere).
4382 (ffi_prep_args_v9): New function.
4383 (ffi_prep_cif_machdep): Handle V9 ABI and long long on V8.
4384 (ffi_V9_return_struct): New function.
4385 (ffi_call): Handle FFI_V9 ABI from 64bit code and FFI_V8 ABI from
4386 32bit code (not yet cross-arch calls).
4387 * src/sparc/v8.S: Add struct return delay nop.
4389 * src/sparc/v9.S: New file.
4390 * src/prep_cif.c (ffi_prep_cif): Return structure pointer
4391 is used on sparc64 only for structures larger than 32 bytes.
4392 Pass by reference for structures is done for structure arguments
4393 larger than 16 bytes.
4394 * src/ffitest.c (main): Use 64bit rint on sparc64.
4395 Run long long tests on sparc.
4396 * src/types.c (FFI_TYPE_POINTER): Pointer is 64bit on alpha and
4398 (FFI_TYPE_LONGDOUBLE): long double is 128 bit aligned to 128 bits
4400 * configure.in (sparc-*-linux*): New supported target.
4401 (sparc64-*-linux*): Likewise.
4402 * configure: Rebuilt.
4403 * Makefile.am: Add v9.S to SPARC files.
4404 * Makefile.in: Likewise.
4405 (LINK): Surround $(CCLD) into double quotes, so that multilib
4406 compiles work correctly.
4408 2000-04-04 Alexandre Petit-Bianco <apbianco@cygnus.com>
4410 * configure: Rebuilt.
4411 * configure.in: (i*86-*-solaris*): New libffi target. Patch
4412 proposed by Bryce McKinlay.
4414 2000-03-20 Tom Tromey <tromey@cygnus.com>
4416 * Makefile.in: Hand edit for java_raw_api.lo.
4418 2000-03-08 Bryce McKinlay <bryce@albatross.co.nz>
4420 * config.guess, config.sub: Update from the gcc tree.
4421 Fix for PR libgcj/168.
4423 2000-03-03 Tom Tromey <tromey@cygnus.com>
4425 * Makefile.in: Fixed ia64 by hand.
4427 * configure: Rebuilt.
4428 * configure.in (--enable-multilib): New option.
4429 (libffi_basedir): New subst.
4430 (AC_OUTPUT): Added multilib code.
4432 2000-03-02 Tom Tromey <tromey@cygnus.com>
4434 * Makefile.in: Rebuilt.
4435 * Makefile.am (TARGET_SRC_IA64): Use `ia64', not `alpha', as
4438 2000-02-25 Hans Boehm <boehm@acm.org>
4440 * src/ia64/ffi.c, src/ia64/ia64_flags.h, src/ia64/unix.S: New
4442 * src/raw_api.c (ffi_translate_args): Fixed typo in argument
4444 (ffi_prep_raw_closure): Use ffi_translate_args, not
4445 ffi_closure_translate.
4446 * src/java_raw_api.c: New file.
4447 * src/ffitest.c (closure_test_fn): New function.
4448 (main): Define `rint' as long long on IA64. Added new test when
4449 FFI_CLOSURES is defined.
4450 * include/ffi.h.in (ALIGN): Use size_t, not unsigned.
4451 (ffi_abi): Recognize IA64.
4452 (ffi_raw): Added `flt' field.
4453 Added "Java raw API" code.
4454 * configure.in: Recognize ia64.
4455 * Makefile.am (TARGET_SRC_IA64): New macro.
4456 (libffi_la_common_SOURCES): Added java_raw_api.c.
4457 (libffi_la_SOURCES): Define in IA64 case.
4459 2000-01-04 Tom Tromey <tromey@cygnus.com>
4461 * Makefile.in: Rebuilt with newer automake.
4463 1999-12-31 Tom Tromey <tromey@cygnus.com>
4465 * Makefile.am (INCLUDES): Added -I$(top_srcdir)/src.
4467 1999-09-01 Tom Tromey <tromey@cygnus.com>
4469 * include/ffi.h.in: Removed PACKAGE and VERSION defines and
4471 * fficonfig.h.in: Rebuilt.
4472 * configure: Rebuilt.
4473 * configure.in: Pass 3rd argument to AM_INIT_AUTOMAKE.
4474 Use AM_PROG_LIBTOOL (automake 1.4 compatibility).
4475 * acconfig.h: Don't #undef PACKAGE or VERSION.
4477 1999-08-09 Anthony Green <green@cygnus.com>
4479 * include/ffi.h.in: Try to work around messy header problem
4480 with PACKAGE and VERSION.
4482 * configure: Rebuilt.
4483 * configure.in: Change version to 2.00-beta.
4485 * fficonfig.h.in: Rebuilt.
4486 * acconfig.h (FFI_NO_STRUCTS, FFI_NO_RAW_API): Define.
4488 * src/x86/ffi.c (ffi_raw_call): Rename.
4490 1999-08-02 Kresten Krab Thorup <krab@dominiq.is.s.u-tokyo.ac.jp>
4492 * src/x86/ffi.c (ffi_closure_SYSV): New function.
4493 (ffi_prep_incoming_args_SYSV): Ditto.
4494 (ffi_prep_closure): Ditto.
4495 (ffi_closure_raw_SYSV): Ditto.
4496 (ffi_prep_raw_closure): More ditto.
4497 (ffi_call_raw): Final ditto.
4499 * include/ffi.h.in: Add definitions for closure and raw API.
4501 * src/x86/ffi.c (ffi_prep_cif_machdep): Added case for
4504 * Makefile.am (libffi_la_common_SOURCES): Added raw_api.c
4506 * src/raw_api.c: New file.
4508 * include/ffi.h.in (ffi_raw): New type.
4509 (UINT_ARG, SINT_ARG): New defines.
4510 (ffi_closure, ffi_raw_closure): New types.
4511 (ffi_prep_closure, ffi_prep_raw_closure): New declarations.
4513 * configure.in: Add check for endianness and sizeof void*.
4515 * src/x86/sysv.S (ffi_call_SYSV): Call fixup routine via argument,
4516 instead of directly.
4518 * configure: Rebuilt.
4520 Thu Jul 8 14:28:42 1999 Anthony Green <green@cygnus.com>
4522 * configure.in: Add x86 and powerpc BeOS configurations.
4523 From Makoto Kato <m_kato@ga2.so-net.ne.jp>.
4525 1999-05-09 Anthony Green <green@cygnus.com>
4527 * configure.in: Add warning about this being beta code.
4528 Remove src/Makefile.am from the picture.
4529 * configure: Rebuilt.
4531 * Makefile.am: Move logic from src/Makefile.am. Add changes
4532 to support libffi as a target library.
4533 * Makefile.in: Rebuilt.
4535 * aclocal.m4, config.guess, config.sub, ltconfig, ltmain.sh:
4536 Upgraded to new autoconf, automake, libtool.
4540 * LICENSE: Update copyright date.
4542 * src/Makefile.am, src/Makefile.in: Removed.
4544 1998-11-29 Anthony Green <green@cygnus.com>
4546 * include/ChangeLog: Removed.
4547 * src/ChangeLog: Removed.
4548 * src/mips/ChangeLog: Removed.
4549 * src/sparc/ChangeLog: Remboved.
4550 * src/x86/ChangeLog: Removed.
4552 * ChangeLog.v1: Created.