1 2007-12-06 David Daney <ddaney@avtrex.com>
3 * src/mips/n32.S (ffi_closure_N32): Use 64-bit add instruction on
6 2007-12-01 Andreas Tobler <a.tobler@schweiz.org>
9 * src/powerpc/ffitarget.h: Introduce new ABI FFI_LINUX_SOFT_FLOAT.
10 Add local FFI_TYPE_UINT128 to handle soft-float long-double-128.
11 * src/powerpc/ffi.c: Distinguish between __NO_FPRS__ and not and
12 set the NUM_FPR_ARG_REGISTERS according to.
13 Add support for potential soft-float support under hard-float
15 (ffi_prep_args_SYSV): Set NUM_FPR_ARG_REGISTERS to 0 in case of
16 FFI_LINUX_SOFT_FLOAT, handle float, doubles and long-doubles according
17 to the FFI_LINUX_SOFT_FLOAT ABI.
18 (ffi_prep_cif_machdep): Likewise.
19 (ffi_closure_helper_SYSV): Likewise.
20 * src/powerpc/ppc_closure.S: Make sure not to store float/double
21 on archs where __NO_FPRS__ is true.
22 Add FFI_TYPE_UINT128 support.
23 * src/powerpc/sysv.S: Add support for soft-float long-double-128.
24 Adjust copyright notice.
26 2007-11-25 Andreas Tobler <a.tobler@schweiz.org>
28 * src/closures.c: Move defintion of MAYBE_UNUSED from here to ...
29 * include/ffi_common.h: ... here.
32 2007-11-17 Andreas Tobler <a.tobler@schweiz.org>
34 * src/powerpc/sysv.S: Load correct cr to compare if we have long double.
35 * src/powerpc/linux64.S: Likewise.
36 * src/powerpc/ffi.c: Add a comment to show which part goes into cr6.
37 * testsuite/libffi.call/return_ldl.c: New test.
39 2007-09-04 <aph@redhat.com>
41 * src/arm/sysv.S (UNWIND): New.
42 (Whole file): Conditionally compile unwinder directives.
43 * src/arm/sysv.S: Add unwinder directives.
45 * src/arm/ffi.c (ffi_prep_args): Align structs by at least 4 bytes.
46 Only treat r0 as a struct address if we're actually returning a
48 Only copy the bytes that are actually within a struct.
49 (ffi_prep_cif_machdep): A Composite Type not larger than 4 bytes
50 is returned in r0, not passed by address.
51 (ffi_call): Allocate a word-sized temporary for the case where
52 a composite is returned in r0.
53 (ffi_prep_incoming_args_SYSV): Align as necessary.
55 2007-08-05 Steven Newbury <s_j_newbury@yahoo.co.uk>
57 * src/arm/ffi.c (FFI_INIT_TRAMPOLINE): Use __clear_cache instead of
58 directly using the sys_cacheflush syscall.
60 2007-07-27 Andrew Haley <aph@redhat.com>
62 * src/arm/sysv.S (ffi_closure_SYSV): Add soft-float.
64 2007-09-03 Maciej W. Rozycki <macro@linux-mips.org>
66 * Makefile.am: Unify MIPS_IRIX and MIPS_LINUX into MIPS.
67 * configure.ac: Likewise.
68 * Makefile.in: Regenerate.
69 * include/Makefile.in: Likewise.
70 * testsuite/Makefile.in: Likewise.
71 * configure: Likewise.
73 2007-08-24 David Daney <ddaney@avtrex.com>
75 * testsuite/libffi.call/return_sl.c: New test.
77 2007-08-10 David Daney <ddaney@avtrex.com>
79 * testsuite/libffi.call/cls_multi_ushort.c,
80 testsuite/libffi.call/cls_align_uint16.c,
81 testsuite/libffi.call/nested_struct1.c,
82 testsuite/libffi.call/nested_struct3.c,
83 testsuite/libffi.call/cls_7_1_byte.c,
84 testsuite/libffi.call/nested_struct5.c,
85 testsuite/libffi.call/cls_double.c,
86 testsuite/libffi.call/nested_struct7.c,
87 testsuite/libffi.call/cls_sint.c,
88 testsuite/libffi.call/nested_struct9.c,
89 testsuite/libffi.call/cls_20byte1.c,
90 testsuite/libffi.call/cls_multi_sshortchar.c,
91 testsuite/libffi.call/cls_align_sint64.c,
92 testsuite/libffi.call/cls_3byte2.c,
93 testsuite/libffi.call/cls_multi_schar.c,
94 testsuite/libffi.call/cls_multi_uchar.c,
95 testsuite/libffi.call/cls_19byte.c,
96 testsuite/libffi.call/cls_9byte1.c,
97 testsuite/libffi.call/cls_align_float.c,
98 testsuite/libffi.call/closure_fn1.c,
99 testsuite/libffi.call/problem1.c,
100 testsuite/libffi.call/closure_fn3.c,
101 testsuite/libffi.call/cls_sshort.c,
102 testsuite/libffi.call/closure_fn5.c,
103 testsuite/libffi.call/cls_align_double.c,
104 testsuite/libffi.call/nested_struct.c,
105 testsuite/libffi.call/cls_2byte.c,
106 testsuite/libffi.call/nested_struct10.c,
107 testsuite/libffi.call/cls_4byte.c,
108 testsuite/libffi.call/cls_6byte.c,
109 testsuite/libffi.call/cls_8byte.c,
110 testsuite/libffi.call/cls_multi_sshort.c,
111 testsuite/libffi.call/cls_align_sint16.c,
112 testsuite/libffi.call/cls_align_uint32.c,
113 testsuite/libffi.call/cls_20byte.c,
114 testsuite/libffi.call/cls_float.c,
115 testsuite/libffi.call/nested_struct2.c,
116 testsuite/libffi.call/cls_5_1_byte.c,
117 testsuite/libffi.call/nested_struct4.c,
118 testsuite/libffi.call/cls_24byte.c,
119 testsuite/libffi.call/nested_struct6.c,
120 testsuite/libffi.call/cls_64byte.c,
121 testsuite/libffi.call/nested_struct8.c,
122 testsuite/libffi.call/cls_uint.c,
123 testsuite/libffi.call/cls_multi_ushortchar.c,
124 testsuite/libffi.call/cls_schar.c,
125 testsuite/libffi.call/cls_uchar.c,
126 testsuite/libffi.call/cls_align_uint64.c,
127 testsuite/libffi.call/cls_ulonglong.c,
128 testsuite/libffi.call/cls_align_longdouble.c,
129 testsuite/libffi.call/cls_1_1byte.c,
130 testsuite/libffi.call/cls_12byte.c,
131 testsuite/libffi.call/cls_3_1byte.c,
132 testsuite/libffi.call/cls_3byte1.c,
133 testsuite/libffi.call/cls_4_1byte.c,
134 testsuite/libffi.call/cls_6_1_byte.c,
135 testsuite/libffi.call/cls_16byte.c,
136 testsuite/libffi.call/cls_18byte.c,
137 testsuite/libffi.call/closure_fn0.c,
138 testsuite/libffi.call/cls_9byte2.c,
139 testsuite/libffi.call/closure_fn2.c,
140 testsuite/libffi.call/closure_fn4.c,
141 testsuite/libffi.call/cls_ushort.c,
142 testsuite/libffi.call/closure_fn6.c,
143 testsuite/libffi.call/cls_5byte.c,
144 testsuite/libffi.call/cls_align_pointer.c,
145 testsuite/libffi.call/cls_7byte.c,
146 testsuite/libffi.call/cls_align_sint32.c,
147 testsuite/libffi.special/unwindtest_ffi_call.cc,
148 testsuite/libffi.special/unwindtest.cc: Remove xfail for mips64*-*-*.
150 2007-08-10 David Daney <ddaney@avtrex.com>
153 * configure.ac: Don't treat mips64 as a special case.
154 * Makefile.am (nodist_libffi_la_SOURCES): Add n32.S.
155 * configure: Regenerate
156 * Makefile.in: Ditto.
157 * fficonfig.h.in: Ditto.
158 * src/mips/ffitarget.h (REG_L, REG_S, SUBU, ADDU, SRL, LI): Indent.
159 (LA, EH_FRAME_ALIGN, FDE_ADDR_BYTES): New preprocessor macros.
160 (FFI_DEFAULT_ABI): Set for n64 case.
161 (FFI_CLOSURES, FFI_TRAMPOLINE_SIZE): Define for n32 and n64 cases.
162 * src/mips/n32.S (ffi_call_N32): Add debug macros and labels for FDE.
163 (ffi_closure_N32): New function.
164 (.eh_frame): New section
165 * src/mips/o32.S: Clean up comments.
166 (ffi_closure_O32): Pass ffi_closure parameter in $12.
167 * src/mips/ffi.c: Use FFI_MIPS_N32 instead of
168 _MIPS_SIM == _ABIN32 throughout.
169 (FFI_MIPS_STOP_HERE): New, use in place of
171 (ffi_prep_args): Use unsigned long to hold pointer values. Rewrite
172 to support n32/n64 ABIs.
173 (calc_n32_struct_flags): Rewrite.
174 (calc_n32_return_struct_flags): Remove unused variable. Reverse
175 position of flag bits.
176 (ffi_prep_cif_machdep): Rewrite n32 portion.
177 (ffi_call): Enable for n64. Add special handling for small structure
179 (ffi_prep_closure_loc): Add n32 and n64 support.
180 (ffi_closure_mips_inner_O32): Add cast to silence warning.
181 (copy_struct_N32, ffi_closure_mips_inner_N32): New functions.
183 2007-08-08 David Daney <ddaney@avtrex.com>
185 * testsuite/libffi.call/ffitest.h (ffi_type_mylong): Remove definition.
186 * testsuite/libffi.call/cls_align_uint16.c (main): Use correct type
188 * testsuite/libffi.call/nested_struct1.c (main): Ditto.
189 * testsuite/libffi.call/cls_sint.c (main): Ditto.
190 * testsuite/libffi.call/nested_struct9.c (main): Ditto.
191 * testsuite/libffi.call/cls_20byte1.c (main): Ditto.
192 * testsuite/libffi.call/cls_9byte1.c (main): Ditto.
193 * testsuite/libffi.call/closure_fn1.c (main): Ditto.
194 * testsuite/libffi.call/closure_fn3.c (main): Ditto.
195 * testsuite/libffi.call/return_dbl2.c (main): Ditto.
196 * testsuite/libffi.call/cls_sshort.c (main): Ditto.
197 * testsuite/libffi.call/return_fl3.c (main): Ditto.
198 * testsuite/libffi.call/closure_fn5.c (main): Ditto.
199 * testsuite/libffi.call/nested_struct.c (main): Ditto.
200 * testsuite/libffi.call/nested_struct10.c (main): Ditto.
201 * testsuite/libffi.call/return_ll1.c (main): Ditto.
202 * testsuite/libffi.call/cls_8byte.c (main): Ditto.
203 * testsuite/libffi.call/cls_align_uint32.c (main): Ditto.
204 * testsuite/libffi.call/cls_align_sint16.c (main): Ditto.
205 * testsuite/libffi.call/cls_20byte.c (main): Ditto.
206 * testsuite/libffi.call/nested_struct2.c (main): Ditto.
207 * testsuite/libffi.call/cls_24byte.c (main): Ditto.
208 * testsuite/libffi.call/nested_struct6.c (main): Ditto.
209 * testsuite/libffi.call/cls_uint.c (main): Ditto.
210 * testsuite/libffi.call/cls_12byte.c (main): Ditto.
211 * testsuite/libffi.call/cls_16byte.c (main): Ditto.
212 * testsuite/libffi.call/closure_fn0.c (main): Ditto.
213 * testsuite/libffi.call/cls_9byte2.c (main): Ditto.
214 * testsuite/libffi.call/closure_fn2.c (main): Ditto.
215 * testsuite/libffi.call/return_dbl1.c (main): Ditto.
216 * testsuite/libffi.call/closure_fn4.c (main): Ditto.
217 * testsuite/libffi.call/closure_fn6.c (main): Ditto.
218 * testsuite/libffi.call/cls_align_sint32.c (main): Ditto.
220 2007-08-07 Andrew Haley <aph@redhat.com>
222 * src/x86/sysv.S (ffi_closure_raw_SYSV): Fix typo in previous
225 2007-08-06 Andrew Haley <aph@redhat.com>
228 * src/x86/sysv.S (ffi_closure_raw_SYSV): Handle FFI_TYPE_UINT8,
229 FFI_TYPE_SINT8, FFI_TYPE_UINT16, FFI_TYPE_SINT16, FFI_TYPE_UINT32,
232 2007-08-02 David Daney <ddaney@avtrex.com>
234 * testsuite/libffi.call/return_ul.c (main): Define return type as
235 ffi_arg. Use proper printf conversion specifier.
237 2007-07-30 Andrew Haley <aph@redhat.com>
240 * src/x86/ffi.c (ffi_prep_cif_machdep): in x86 case, add code for
241 signed/unsigned int8/16.
242 * src/x86/sysv.S (ffi_call_SYSV): Rewrite to:
244 Remove code to pop args from the stack after call.
245 Special-case signed/unsigned int8/16.
246 * testsuite/libffi.call/return_sc.c (main): Revert.
248 2007-07-26 Richard Guenther <rguenther@suse.de>
251 * testsuite/libffi.call/return_sc.c (main): Verify call
252 result as signed char, not ffi_arg.
254 2007-07-16 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
256 * configure.ac (i?86-*-solaris2.1[0-9]): Set TARGET to X86_64.
257 * configure: Regenerate.
259 2007-07-11 David Daney <ddaney@avtrex.com>
261 * src/mips/ffi.c: Don't include sys/cachectl.h.
262 (ffi_prep_closure_loc): Use __builtin___clear_cache() instead of
265 2007-05-18 Aurelien Jarno <aurelien@aurel32.net>
267 * src/arm/ffi.c (ffi_prep_closure_loc): Renamed and ajusted
268 from (ffi_prep_closure): ... this.
269 (FFI_INIT_TRAMPOLINE): Adjust.
271 2005-12-31 Phil Blundell <pb@reciva.com>
273 * src/arm/ffi.c (ffi_prep_incoming_args_SYSV,
274 ffi_closure_SYSV_inner, ffi_prep_closure): New, add closure support.
275 * src/arm/sysv.S(ffi_closure_SYSV): Likewise.
276 * src/arm/ffitarget.h (FFI_TRAMPOLINE_SIZE): Likewise.
277 (FFI_CLOSURES): Enable closure support.
279 2007-07-03 Andrew Haley <aph@hedges.billgatliff.com>
281 * testsuite/libffi.call/cls_multi_ushort.c,
282 testsuite/libffi.call/cls_align_uint16.c,
283 testsuite/libffi.call/nested_struct1.c,
284 testsuite/libffi.call/nested_struct3.c,
285 testsuite/libffi.call/cls_7_1_byte.c,
286 testsuite/libffi.call/cls_double.c,
287 testsuite/libffi.call/nested_struct5.c,
288 testsuite/libffi.call/nested_struct7.c,
289 testsuite/libffi.call/cls_sint.c,
290 testsuite/libffi.call/nested_struct9.c,
291 testsuite/libffi.call/cls_20byte1.c,
292 testsuite/libffi.call/cls_multi_sshortchar.c,
293 testsuite/libffi.call/cls_align_sint64.c,
294 testsuite/libffi.call/cls_3byte2.c,
295 testsuite/libffi.call/cls_multi_schar.c,
296 testsuite/libffi.call/cls_multi_uchar.c,
297 testsuite/libffi.call/cls_19byte.c,
298 testsuite/libffi.call/cls_9byte1.c,
299 testsuite/libffi.call/cls_align_float.c,
300 testsuite/libffi.call/closure_fn1.c,
301 testsuite/libffi.call/problem1.c,
302 testsuite/libffi.call/closure_fn3.c,
303 testsuite/libffi.call/cls_sshort.c,
304 testsuite/libffi.call/closure_fn5.c,
305 testsuite/libffi.call/cls_align_double.c,
306 testsuite/libffi.call/cls_2byte.c,
307 testsuite/libffi.call/nested_struct.c,
308 testsuite/libffi.call/nested_struct10.c,
309 testsuite/libffi.call/cls_4byte.c,
310 testsuite/libffi.call/cls_6byte.c,
311 testsuite/libffi.call/cls_8byte.c,
312 testsuite/libffi.call/cls_multi_sshort.c,
313 testsuite/libffi.call/cls_align_uint32.c,
314 testsuite/libffi.call/cls_align_sint16.c,
315 testsuite/libffi.call/cls_float.c,
316 testsuite/libffi.call/cls_20byte.c,
317 testsuite/libffi.call/cls_5_1_byte.c,
318 testsuite/libffi.call/nested_struct2.c,
319 testsuite/libffi.call/cls_24byte.c,
320 testsuite/libffi.call/nested_struct4.c,
321 testsuite/libffi.call/nested_struct6.c,
322 testsuite/libffi.call/cls_64byte.c,
323 testsuite/libffi.call/nested_struct8.c,
324 testsuite/libffi.call/cls_uint.c,
325 testsuite/libffi.call/cls_multi_ushortchar.c,
326 testsuite/libffi.call/cls_schar.c,
327 testsuite/libffi.call/cls_uchar.c,
328 testsuite/libffi.call/cls_align_uint64.c,
329 testsuite/libffi.call/cls_ulonglong.c,
330 testsuite/libffi.call/cls_align_longdouble.c,
331 testsuite/libffi.call/cls_1_1byte.c,
332 testsuite/libffi.call/cls_12byte.c,
333 testsuite/libffi.call/cls_3_1byte.c,
334 testsuite/libffi.call/cls_3byte1.c,
335 testsuite/libffi.call/cls_4_1byte.c,
336 testsuite/libffi.call/cls_6_1_byte.c,
337 testsuite/libffi.call/cls_16byte.c,
338 testsuite/libffi.call/cls_18byte.c,
339 testsuite/libffi.call/closure_fn0.c,
340 testsuite/libffi.call/cls_9byte2.c,
341 testsuite/libffi.call/closure_fn2.c,
342 testsuite/libffi.call/closure_fn4.c,
343 testsuite/libffi.call/cls_ushort.c,
344 testsuite/libffi.call/closure_fn6.c,
345 testsuite/libffi.call/cls_5byte.c,
346 testsuite/libffi.call/cls_align_pointer.c,
347 testsuite/libffi.call/cls_7byte.c,
348 testsuite/libffi.call/cls_align_sint32.c,
349 testsuite/libffi.special/unwindtest_ffi_call.cc,
350 testsuite/libffi.special/unwindtest.cc: Enable for ARM.
352 2007-07-05 H.J. Lu <hongjiu.lu@intel.com>
354 * aclocal.m4: Regenerated.
356 2007-06-02 Paolo Bonzini <bonzini@gnu.org>
358 * configure: Regenerate.
360 2007-05-23 Steve Ellcey <sje@cup.hp.com>
362 * Makefile.in: Regenerate.
363 * configure: Regenerate.
364 * aclocal.m4: Regenerate.
365 * include/Makefile.in: Regenerate.
366 * testsuite/Makefile.in: Regenerate.
368 2007-05-10 Roman Zippel <zippel@linux-m68k.org>
370 * src/m68k/ffi.c (ffi_prep_incoming_args_SYSV,
371 ffi_closure_SYSV_inner,ffi_prep_closure): New, add closure support.
372 * src/m68k/sysv.S(ffi_closure_SYSV,ffi_closure_struct_SYSV): Likewise.
373 * src/m68k/ffitarget.h (FFI_TRAMPOLINE_SIZE): Likewise.
374 (FFI_CLOSURES): Enable closure support.
376 2007-05-10 Roman Zippel <zippel@linux-m68k.org>
378 * configure.ac (HAVE_AS_CFI_PSEUDO_OP): New test.
379 * configure: Regenerate.
380 * fficonfig.h.in: Regenerate.
381 * src/m68k/sysv.S (CFI_STARTPROC,CFI_ENDPROC,
382 CFI_OFFSET,CFI_DEF_CFA): New macros.
383 (ffi_call_SYSV): Add callframe annotation.
385 2007-05-10 Roman Zippel <zippel@linux-m68k.org>
387 * src/m68k/ffi.c (ffi_prep_args,ffi_prep_cif_machdep): Fix
388 numerous test suite failures.
389 * src/m68k/sysv.S (ffi_call_SYSV): Likewise.
391 2007-04-11 Paolo Bonzini <bonzini@gnu.org>
393 * Makefile.am (EXTRA_DIST): Bring up to date.
394 * Makefile.in: Regenerate.
395 * src/frv/eabi.S: Remove RCS keyword.
397 2007-04-06 Richard Henderson <rth@redhat.com>
399 * configure.ac: Tidy target case.
400 (HAVE_LONG_DOUBLE): Allow the target to override.
401 * configure: Regenerate.
402 * include/ffi.h.in: Don't define ffi_type_foo if
403 LIBFFI_HIDE_BASIC_TYPES is defined.
404 (ffi_type_longdouble): If not HAVE_LONG_DOUBLE, define
406 * types.c (LIBFFI_HIDE_BASIC_TYPES): Define.
407 (FFI_TYPEDEF, ffi_type_void): Mark the data const.
408 (ffi_type_longdouble): Special case for Alpha. Don't define
409 if long double == double.
411 * src/alpha/ffi.c (FFI_TYPE_LONGDOUBLE): Assert unique value.
412 (ffi_prep_cif_machdep): Handle it as the 128-bit type.
413 (ffi_call, ffi_closure_osf_inner): Likewise.
414 (ffi_closure_osf_inner): Likewise. Mark hidden.
415 (ffi_call_osf, ffi_closure_osf): Mark hidden.
416 * src/alpha/ffitarget.h (FFI_LAST_ABI): Tidy definition.
417 * src/alpha/osf.S (ffi_call_osf, ffi_closure_osf): Mark hidden.
418 (load_table): Handle 128-bit long double.
420 * testsuite/libffi.call/float4.c: Add -mieee for alpha.
422 2007-04-06 Tom Tromey <tromey@redhat.com>
425 * README: Fixed bug in example.
427 2007-04-03 Jakub Jelinek <jakub@redhat.com>
429 * src/closures.c: Include sys/statfs.h.
430 (_GNU_SOURCE): Define on Linux.
431 (FFI_MMAP_EXEC_SELINUX): Define.
432 (selinux_enabled): New variable.
433 (selinux_enabled_check): New function.
434 (is_selinux_enabled): Define.
437 2007-03-24 Uros Bizjak <ubizjak@gmail.com>
439 * testsuite/libffi.call/return_fl2.c (return_fl): Mark as static.
440 Use 'volatile float sum' to create sum of floats to avoid false
441 negative due to excess precision on ix86 targets.
444 2007-03-08 Alexandre Oliva <aoliva@redhat.com>
446 * src/powerpc/ffi.c (flush_icache): Fix left-over from previous
448 (ffi_prep_closure_loc): Remove unneeded casts. Add needed ones.
450 2007-03-07 Alexandre Oliva <aoliva@redhat.com>
452 * include/ffi.h.in (ffi_closure_alloc, ffi_closure_free): New.
453 (ffi_prep_closure_loc): New.
454 (ffi_prep_raw_closure_loc): New.
455 (ffi_prep_java_raw_closure_loc): New.
456 * src/closures.c: New file.
457 * src/dlmalloc.c [FFI_MMAP_EXEC_WRIT] (struct malloc_segment):
458 Replace sflags with exec_offset.
459 [FFI_MMAP_EXEC_WRIT] (mmap_exec_offset, add_segment_exec_offset,
460 sub_segment_exec_offset): New macros.
461 (get_segment_flags, set_segment_flags, check_segment_merge): New
463 (is_mmapped_segment, is_extern_segment): Use get_segment_flags.
464 (add_segment, sys_alloc, create_mspace, create_mspace_with_base,
465 destroy_mspace): Use new macros.
466 (sys_alloc): Silence warning.
467 * Makefile.am (libffi_la_SOURCES): Add src/closures.c.
468 * Makefile.in: Rebuilt.
469 * src/prep_cif [FFI_CLOSURES] (ffi_prep_closure): Implement in
470 terms of ffi_prep_closure_loc.
471 * src/raw_api.c (ffi_prep_raw_closure_loc): Renamed and adjusted
473 (ffi_prep_raw_closure): ... this. Re-implement in terms of the
475 * src/java_raw_api (ffi_prep_java_raw_closure_loc): Renamed and
477 (ffi_prep_java_raw_closure): ... this. Re-implement in terms of
479 * src/alpha/ffi.c (ffi_prep_closure_loc): Renamed from
480 (ffi_prep_closure): ... this.
481 * src/pa/ffi.c: Likewise.
482 * src/cris/ffi.c: Likewise. Adjust.
483 * src/frv/ffi.c: Likewise.
484 * src/ia64/ffi.c: Likewise.
485 * src/mips/ffi.c: Likewise.
486 * src/powerpc/ffi_darwin.c: Likewise.
487 * src/s390/ffi.c: Likewise.
488 * src/sh/ffi.c: Likewise.
489 * src/sh64/ffi.c: Likewise.
490 * src/sparc/ffi.c: Likewise.
491 * src/x86/ffi64.c: Likewise.
492 * src/x86/ffi.c: Likewise.
493 (FFI_INIT_TRAMPOLINE): Adjust.
494 (ffi_prep_raw_closure_loc): Renamed and adjusted from...
495 (ffi_prep_raw_closure): ... this.
496 * src/powerpc/ffi.c (ffi_prep_closure_loc): Renamed from
497 (ffi_prep_closure): ... this.
498 (flush_icache): Adjust.
500 2007-03-07 Alexandre Oliva <aoliva@redhat.com>
502 * src/dlmalloc.c: New file, imported version 2.8.3 of Doug
505 2007-03-01 Brooks Moses <brooks.moses@codesourcery.com>
507 * Makefile.am: Add dummy install-pdf target.
508 * Makefile.in: Regenerate
510 2007-02-13 Andreas Krebbel <krebbel1@de.ibm.com>
512 * src/s390/ffi.c (ffi_prep_args, ffi_prep_cif_machdep,
513 ffi_closure_helper_SYSV): Add long double handling.
515 2007-02-02 Jakub Jelinek <jakub@redhat.com>
517 * src/powerpc/linux64.S (ffi_call_LINUX64): Move restore of r2
518 immediately after bctrl instruction.
520 2007-01-18 Alexandre Oliva <aoliva@redhat.com>
522 * Makefile.am (all-recursive, install-recursive,
523 mostlyclean-recursive, clean-recursive, distclean-recursive,
524 maintainer-clean-recursive): Add missing targets.
525 * Makefile.in: Rebuilt.
527 2006-12-14 Andreas Tobler <a.tobler@schweiz.org>
529 * configure.ac: Add TARGET for x86_64-*-darwin*.
530 * Makefile.am (nodist_libffi_la_SOURCES): Add rules for 64-bit sources
532 * src/x86/ffitarget.h: Set trampoline size for x86_64-*-darwin*.
533 * src/x86/darwin64.S: New file for x86_64-*-darwin* support.
534 * configure: Regenerate.
535 * Makefile.in: Regenerate.
536 * include/Makefile.in: Regenerate.
537 * testsuite/Makefile.in: Regenerate.
538 * testsuite/libffi.special/unwindtest_ffi_call.cc: New test case for
541 2006-12-13 Andreas Tobler <a.tobler@schweiz.org>
543 * aclocal.m4: Regenerate with aclocal -I .. as written in the
546 2006-10-31 Geoffrey Keating <geoffk@apple.com>
548 * src/powerpc/ffi_darwin.c (darwin_adjust_aggregate_sizes): New.
549 (ffi_prep_cif_machdep): Call darwin_adjust_aggregate_sizes for
551 * testsuite/libffi.call/nested_struct4.c: Remove Darwin XFAIL.
552 * testsuite/libffi.call/nested_struct6.c: Remove Darwin XFAIL.
554 2006-10-10 Paolo Bonzini <bonzini@gnu.org>
555 Sandro Tolaini <tolaini@libero.it>
557 * configure.ac [i*86-*-darwin*]: Set X86_DARWIN symbol and
559 * configure: Regenerated.
560 * Makefile.am (nodist_libffi_la_SOURCES) [X86_DARWIN]: New case.
561 (EXTRA_DIST): Add src/x86/darwin.S.
562 * Makefile.in: Regenerated.
563 * include/Makefile.in: Regenerated.
564 * testsuite/Makefile.in: Regenerated.
566 * src/x86/ffi.c (ffi_prep_cif_machdep) [X86_DARWIN]: Treat like
567 X86_WIN32, and additionally align stack to 16 bytes.
568 * src/x86/darwin.S: New, based on sysv.S.
569 * src/prep_cif.c (ffi_prep_cif) [X86_DARWIN]: Align > 8-byte structs.
571 2006-09-12 David Daney <ddaney@avtrex.com>
574 * include/Makefile.am: Install both ffi.h and ffitarget.h in
575 $(libdir)/gcc/$(target_alias)/$(gcc_version)/include.
576 * aclocal.m4: Regenerated for automake 1.9.6.
577 * Makefile.in: Regenerated.
578 * include/Makefile.in: Regenerated.
579 * testsuite/Makefile.in: Regenerated.
581 2006-08-17 Andreas Tobler <a.tobler@schweiz.ch>
583 * include/ffi_common.h (struct): Revert accidental commit.
585 2006-08-15 Andreas Tobler <a.tobler@schweiz.ch>
587 * include/ffi_common.h: Remove lint directives.
588 * include/ffi.h.in: Likewise.
590 2006-07-25 Torsten Schoenfeld <kaffeetisch@gmx.de>
592 * include/ffi.h.in (ffi_type_ulong, ffi_type_slong): Define correctly
593 for 32-bit architectures.
594 * testsuite/libffi.call/return_ul.c: New test case.
596 2006-07-19 David Daney <ddaney@avtrex.com>
598 * testsuite/libffi.call/closure_fn6.c: Remove xfail for mips,
599 xfail remains for mips64.
601 2006-05-23 Carlos O'Donell <carlos@codesourcery.com>
603 * Makefile.am: Add install-html target. Add install-html to .PHONY
604 * Makefile.in: Regenerate.
605 * aclocal.m4: Regenerate.
606 * include/Makefile.in: Regenerate.
607 * testsuite/Makefile.in: Regenerate.
609 2006-05-18 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
611 * pa/ffi.c (ffi_prep_args_pa32): Load floating point arguments from
614 2006-04-22 Andreas Tobler <a.tobler@schweiz.ch>
616 * README: Remove notice about 'Crazy Comments'.
617 * src/debug.c: Remove lint directives. Cleanup white spaces.
618 * src/java_raw_api.c: Likewise.
619 * src/prep_cif.c: Likewise.
620 * src/raw_api.c: Likewise.
621 * src/ffitest.c: Delete. No longer needed, all test cases migrated
623 * src/arm/ffi.c: Remove lint directives.
624 * src/m32r/ffi.c: Likewise.
625 * src/pa/ffi.c: Likewise.
626 * src/powerpc/ffi.c: Likewise.
627 * src/powerpc/ffi_darwin.c: Likewise.
628 * src/sh/ffi.c: Likewise.
629 * src/sh64/ffi.c: Likewise.
630 * src/x86/ffi.c: Likewise.
631 * testsuite/libffi.call/float2.c: Likewise.
632 * testsuite/libffi.call/promotion.c: Likewise.
633 * testsuite/libffi.call/struct1.c: Likewise.
635 2006-04-13 Andreas Tobler <a.tobler@schweiz.ch>
637 * src/pa/hpux32.S: Correct unwind offset calculation for
639 * src/pa/linux.S: Likewise.
641 2006-04-12 James E Wilson <wilson@specifix.com>
644 * src/ia64/ffi.c (stf_spill, ldf_fill): Rewrite as macros.
645 (hfa_type_load): Call stf_spill.
646 (hfa_type_store): Call ldf_fill.
647 (ffi_call): Adjust calls to above routines. Add local temps for
650 2006-04-10 Matthias Klose <doko@debian.org>
652 * testsuite/lib/libffi-dg.exp (libffi-init): Recognize multilib
653 directory names containing underscores.
655 2006-04-07 James E Wilson <wilson@specifix.com>
657 * testsuite/libffi.call/float4.c: New testcase.
659 2006-04-05 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
660 Andreas Tobler <a.tobler@schweiz.ch>
662 * Makefile.am: Add PA_HPUX port.
663 * Makefile.in: Regenerate.
664 * include/Makefile.in: Likewise.
665 * testsuite/Makefile.in: Likewise.
666 * configure.ac: Add PA_HPUX rules.
667 * configure: Regenerate.
668 * src/pa/ffitarget.h: Rename linux target to PA_LINUX.
669 Add PA_HPUX and PA64_HPUX.
670 Rename FFI_LINUX ABI to FFI_PA32 ABI.
671 (FFI_TRAMPOLINE_SIZE): Define for 32-bit HP-UX targets.
672 (FFI_TYPE_SMALL_STRUCT2): Define.
673 (FFI_TYPE_SMALL_STRUCT4): Likewise.
674 (FFI_TYPE_SMALL_STRUCT8): Likewise.
675 (FFI_TYPE_SMALL_STRUCT3): Redefine.
676 (FFI_TYPE_SMALL_STRUCT5): Likewise.
677 (FFI_TYPE_SMALL_STRUCT6): Likewise.
678 (FFI_TYPE_SMALL_STRUCT7): Likewise.
679 * src/pa/ffi.c (ROUND_DOWN): Delete.
680 (fldw, fstw, fldd, fstd): Use '__asm__'.
681 (ffi_struct_type): Add support for FFI_TYPE_SMALL_STRUCT2,
682 FFI_TYPE_SMALL_STRUCT4 and FFI_TYPE_SMALL_STRUCT8.
683 (ffi_prep_args_LINUX): Rename to ffi_prep_args_pa32. Update comment.
684 Simplify incrementing of stack slot variable. Change type of local
686 (ffi_size_stack_LINUX): Rename to ffi_size_stack_pa32. Handle long
688 (ffi_prep_cif_machdep): Likewise.
689 (ffi_call): Likewise.
690 (ffi_closure_inner_LINUX): Rename to ffi_closure_inner_pa32. Change
691 return type to ffi_status. Simplify incrementing of stack slot
692 variable. Only copy floating point argument registers when PA_LINUX
693 is true. Reformat debug statement.
694 Add support for FFI_TYPE_SMALL_STRUCT2, FFI_TYPE_SMALL_STRUCT4 and
695 FFI_TYPE_SMALL_STRUCT8.
696 (ffi_closure_LINUX): Rename to ffi_closure_pa32. Add 'extern' to
698 (ffi_prep_closure): Make linux trampoline conditional on PA_LINUX.
699 Add nops to cache flush. Add trampoline for PA_HPUX.
700 * src/pa/hpux32.S: New file.
701 * src/pa/linux.S (ffi_call_LINUX): Rename to ffi_call_pa32. Rename
702 ffi_prep_args_LINUX to ffi_prep_args_pa32.
703 Localize labels. Add support for 2, 4 and 8-byte small structs. Handle
704 unaligned destinations in 3, 5, 6 and 7-byte small structs. Order
705 argument type checks so that common argument types appear first.
706 (ffi_closure_LINUX): Rename to ffi_closure_pa32. Rename
707 ffi_closure_inner_LINUX to ffi_closure_inner_pa32.
709 2006-03-24 Alan Modra <amodra@bigpond.net.au>
711 * src/powerpc/ffitarget.h (enum ffi_abi): Add FFI_LINUX. Default
712 for 32-bit using IBM extended double format. Fix FFI_LAST_ABI.
713 * src/powerpc/ffi.c (ffi_prep_args_SYSV): Handle linux variant of
715 (ffi_prep_args64): Assert using IBM extended double.
716 (ffi_prep_cif_machdep): Don't munge FFI_TYPE_LONGDOUBLE type.
717 Handle FFI_LINUX FFI_TYPE_LONGDOUBLE return and args.
718 (ffi_call): Handle FFI_LINUX.
719 (ffi_closure_helper_SYSV): Non FFI_LINUX long double return needs
720 gpr3 return pointer as for struct return. Handle FFI_LINUX
721 FFI_TYPE_LONGDOUBLE return and args. Don't increment "nf"
723 * src/powerpc/ppc_closure.S (ffi_closure_SYSV): Load both f1 and f2
724 for FFI_TYPE_LONGDOUBLE. Move epilogue insns into case table.
725 Don't use r6 as pointer to results, instead use sp offset. Don't
726 make a special call to load lr with case table address, instead
727 use offset from previous call.
728 * src/powerpc/sysv.S (ffi_call_SYSV): Save long double return.
729 * src/powerpc/linux64.S (ffi_call_LINUX64): Simplify long double
732 2006-03-15 Kaz Kojima <kkojima@gcc.gnu.org>
734 * src/sh64/ffi.c (ffi_prep_cif_machdep): Handle float arguments
735 passed with FP registers correctly.
736 (ffi_closure_helper_SYSV): Likewise.
737 * src/sh64/sysv.S: Likewise.
739 2006-03-01 Andreas Tobler <a.tobler@schweiz.ch>
741 * testsuite/libffi.special/unwindtest.cc (closure_test_fn): Mark cif,
742 args and userdata unused.
743 (closure_test_fn1): Mark cif and userdata unused.
744 (main): Remove unused res.
746 2006-02-28 Andreas Tobler <a.tobler@schweiz.ch>
748 * testsuite/libffi.call/call.exp: Adjust FSF address. Add test runs for
749 -O2, -O3, -Os and the warning flags -W -Wall.
750 * testsuite/libffi.special/special.exp: Likewise.
751 * testsuite/libffi.call/ffitest.h: Add an __UNUSED__ macro to mark
752 unused parameter unused for gcc or else do nothing.
753 * testsuite/libffi.special/ffitestcxx.h: Likewise.
754 * testsuite/libffi.call/cls_12byte.c (cls_struct_12byte_gn): Mark cif
756 * testsuite/libffi.call/cls_16byte.c (cls_struct_16byte_gn): Likewise.
757 * testsuite/libffi.call/cls_18byte.c (cls_struct_18byte_gn): Likewise.
758 * testsuite/libffi.call/cls_19byte.c (cls_struct_19byte_gn): Likewise.
759 * testsuite/libffi.call/cls_1_1byte.c (cls_struct_1_1byte_gn): Likewise.
760 * testsuite/libffi.call/cls_20byte.c (cls_struct_20byte_gn): Likewise.
761 * testsuite/libffi.call/cls_20byte1.c (cls_struct_20byte_gn): Likewise.
762 * testsuite/libffi.call/cls_24byte.c (cls_struct_24byte_gn): Likewise.
763 * testsuite/libffi.call/cls_2byte.c (cls_struct_2byte_gn): Likewise.
764 * testsuite/libffi.call/cls_3_1byte.c (cls_struct_3_1byte_gn): Likewise.
765 * testsuite/libffi.call/cls_3byte1.c (cls_struct_3byte_gn): Likewise.
766 * testsuite/libffi.call/cls_3byte2.c (cls_struct_3byte_gn1): Likewise.
767 * testsuite/libffi.call/cls_4_1byte.c (cls_struct_4_1byte_gn): Likewise.
768 * testsuite/libffi.call/cls_4byte.c (cls_struct_4byte_gn): Likewise.
769 * testsuite/libffi.call/cls_5_1_byte.c (cls_struct_5byte_gn): Likewise.
770 * testsuite/libffi.call/cls_5byte.c (cls_struct_5byte_gn): Likewise.
771 * testsuite/libffi.call/cls_64byte.c (cls_struct_64byte_gn): Likewise.
772 * testsuite/libffi.call/cls_6_1_byte.c (cls_struct_6byte_gn): Likewise.
773 * testsuite/libffi.call/cls_6byte.c (cls_struct_6byte_gn): Likewise.
774 * testsuite/libffi.call/cls_7_1_byte.c (cls_struct_7byte_gn): Likewise.
775 * testsuite/libffi.call/cls_7byte.c (cls_struct_7byte_gn): Likewise.
776 * testsuite/libffi.call/cls_8byte.c (cls_struct_8byte_gn): Likewise.
777 * testsuite/libffi.call/cls_9byte1.c (cls_struct_9byte_gn): Likewise.
778 * testsuite/libffi.call/cls_9byte2.c (cls_struct_9byte_gn): Likewise.
779 * testsuite/libffi.call/cls_align_double.c (cls_struct_align_gn):
781 * testsuite/libffi.call/cls_align_float.c (cls_struct_align_gn):
783 * testsuite/libffi.call/cls_align_longdouble.c (cls_struct_align_gn):
785 * testsuite/libffi.call/cls_align_pointer.c (cls_struct_align_fn): Cast
786 void* to avoid compiler warning.
788 (cls_struct_align_gn): Mark cif and userdata unused.
789 * testsuite/libffi.call/cls_align_sint16.c (cls_struct_align_gn):
791 * testsuite/libffi.call/cls_align_sint32.c (cls_struct_align_gn):
793 * testsuite/libffi.call/cls_align_sint64.c (cls_struct_align_gn):
795 * testsuite/libffi.call/cls_align_uint16.c (cls_struct_align_gn):
797 * testsuite/libffi.call/cls_align_uint32.c (cls_struct_align_gn):
799 * testsuite/libffi.call/cls_double.c (cls_ret_double_fn): Likewise.
800 * testsuite/libffi.call/cls_float.c (cls_ret_float_fn): Likewise.
801 * testsuite/libffi.call/cls_multi_schar.c (test_func_gn): Mark cif and
803 (main): Cast res_call to silence gcc.
804 * testsuite/libffi.call/cls_multi_sshort.c (test_func_gn): Mark cif and
806 (main): Cast res_call to silence gcc.
807 * testsuite/libffi.call/cls_multi_sshortchar.c (test_func_gn): Mark cif
809 (main): Cast res_call to silence gcc.
810 * testsuite/libffi.call/cls_multi_uchar.c (test_func_gn): Mark cif and
812 (main): Cast res_call to silence gcc.
813 * testsuite/libffi.call/cls_multi_ushort.c (test_func_gn): Mark cif and
815 (main): Cast res_call to silence gcc.
816 * testsuite/libffi.call/cls_multi_ushortchar.c (test_func_gn): Mark cif
818 (main): Cast res_call to silence gcc.
819 * testsuite/libffi.call/cls_schar.c (cls_ret_schar_fn): Mark cif and
821 (cls_ret_schar_fn): Cast printf parameter to silence gcc.
822 * testsuite/libffi.call/cls_sint.c (cls_ret_sint_fn): Mark cif and
824 (cls_ret_sint_fn): Cast printf parameter to silence gcc.
825 * testsuite/libffi.call/cls_sshort.c (cls_ret_sshort_fn): Mark cif and
827 (cls_ret_sshort_fn): Cast printf parameter to silence gcc.
828 * testsuite/libffi.call/cls_uchar.c (cls_ret_uchar_fn): Mark cif and
830 (cls_ret_uchar_fn): Cast printf parameter to silence gcc.
831 * testsuite/libffi.call/cls_uint.c (cls_ret_uint_fn): Mark cif and
833 (cls_ret_uint_fn): Cast printf parameter to silence gcc.
834 * testsuite/libffi.call/cls_ulonglong.c (cls_ret_ulonglong_fn): Mark cif
836 * testsuite/libffi.call/cls_ushort.c (cls_ret_ushort_fn): Mark cif and
838 (cls_ret_ushort_fn): Cast printf parameter to silence gcc.
839 * testsuite/libffi.call/float.c (floating): Remove unused parameter e.
840 * testsuite/libffi.call/float1.c (main): Remove unused variable i.
841 Cleanup white spaces.
842 * testsuite/libffi.call/negint.c (checking): Remove unused variable i.
843 * testsuite/libffi.call/nested_struct.c (cls_struct_combined_gn): Mark
844 cif and userdata unused.
845 * testsuite/libffi.call/nested_struct1.c (cls_struct_combined_gn):
847 * testsuite/libffi.call/nested_struct10.c (B_gn): Likewise.
848 * testsuite/libffi.call/nested_struct2.c (B_fn): Adjust printf
849 formatters to silence gcc.
850 (B_gn): Mark cif and userdata unused.
851 * testsuite/libffi.call/nested_struct3.c (B_gn): Mark cif and userdata
853 * testsuite/libffi.call/nested_struct4.c: Mention related PR.
854 (B_gn): Mark cif and userdata unused.
855 * testsuite/libffi.call/nested_struct5.c (B_gn): Mark cif and userdata
857 * testsuite/libffi.call/nested_struct6.c: Mention related PR.
858 (B_gn): Mark cif and userdata unused.
859 * testsuite/libffi.call/nested_struct7.c (B_gn): Mark cif and userdata
861 * testsuite/libffi.call/nested_struct8.c (B_gn): Likewise.
862 * testsuite/libffi.call/nested_struct9.c (B_gn): Likewise.
863 * testsuite/libffi.call/problem1.c (stub): Likewise.
864 * testsuite/libffi.call/pyobjc-tc.c (main): Cast the result to silence
866 * testsuite/libffi.call/return_fl2.c (return_fl): Add the note mentioned
867 in the last commit for this test case in the test case itself.
868 * testsuite/libffi.call/closure_fn0.c (closure_test_fn0): Mark cif as
870 * testsuite/libffi.call/closure_fn1.c (closure_test_fn1): Likewise.
871 * testsuite/libffi.call/closure_fn2.c (closure_test_fn2): Likewise.
872 * testsuite/libffi.call/closure_fn3.c (closure_test_fn3): Likewise.
873 * testsuite/libffi.call/closure_fn4.c (closure_test_fn0): Likewise.
874 * testsuite/libffi.call/closure_fn5.c (closure_test_fn5): Likewise.
875 * testsuite/libffi.call/closure_fn6.c (closure_test_fn0): Likewise.
877 2006-02-22 Kaz Kojima <kkojima@gcc.gnu.org>
879 * src/sh/sysv.S: Fix register numbers in the FDE for
882 2006-02-20 Andreas Tobler <a.tobler@schweiz.ch>
884 * testsuite/libffi.call/return_fl2.c (return_fl): Remove static
885 declaration to avoid a false negative on ix86. See PR323.
887 2006-02-18 Kaz Kojima <kkojima@gcc.gnu.org>
889 * src/sh/ffi.c (ffi_closure_helper_SYSV): Remove unused variable
890 and cast integer to void * if needed. Update the pointer to
891 the FP register saved area correctly.
893 2006-02-17 Andreas Tobler <a.tobler@schweiz.ch>
895 * testsuite/libffi.call/nested_struct6.c: XFAIL this test until PR25630
897 * testsuite/libffi.call/nested_struct4.c: Likewise.
899 2006-02-16 Andreas Tobler <a.tobler@schweiz.ch>
901 * testsuite/libffi.call/return_dbl.c: New test case.
902 * testsuite/libffi.call/return_dbl1.c: Likewise.
903 * testsuite/libffi.call/return_dbl2.c: Likewise.
904 * testsuite/libffi.call/return_fl.c: Likewise.
905 * testsuite/libffi.call/return_fl1.c: Likewise.
906 * testsuite/libffi.call/return_fl2.c: Likewise.
907 * testsuite/libffi.call/return_fl3.c: Likewise.
908 * testsuite/libffi.call/closure_fn6.c: Likewise.
910 * testsuite/libffi.call/nested_struct2.c: Remove ffi_type_mylong
912 * testsuite/libffi.call/ffitest.h: Add ffi_type_mylong definition
913 here to be used by other test cases too.
915 * testsuite/libffi.call/nested_struct10.c: New test case.
916 * testsuite/libffi.call/nested_struct9.c: Likewise.
917 * testsuite/libffi.call/nested_struct8.c: Likewise.
918 * testsuite/libffi.call/nested_struct7.c: Likewise.
919 * testsuite/libffi.call/nested_struct6.c: Likewise.
920 * testsuite/libffi.call/nested_struct5.c: Likewise.
921 * testsuite/libffi.call/nested_struct4.c: Likewise.
923 2006-01-21 Andreas Tobler <a.tobler@schweiz.ch>
925 * configure.ac: Enable libffi for sparc64-*-freebsd*.
926 * configure: Rebuilt.
928 2006-01-18 Jakub Jelinek <jakub@redhat.com>
930 * src/powerpc/sysv.S (smst_two_register): Don't call __ashldi3,
931 instead do the shifting inline.
932 * src/powerpc/ppc_closure.S (ffi_closure_SYSV): Don't compute %r5
933 shift count unconditionally. Simplify load sequences for 1, 2, 3, 4
934 and 8 byte structs, for the remaining struct sizes don't call
935 __lshrdi3, instead do the shifting inline.
937 2005-12-07 Thiemo Seufer <ths@networkno.de>
939 * src/mips/ffitarget.h: Remove obsolete sgidefs.h include. Add
941 * src/mips/o32.S (ffi_call_O32): Code formatting. Define
942 and use A3_OFF, FP_OFF, RA_OFF. Micro-optimizations.
943 (ffi_closure_O32): Likewise, but with newly defined A3_OFF2,
944 A2_OFF2, A1_OFF2, A0_OFF2, RA_OFF2, FP_OFF2, S0_OFF2, GP_OFF2,
945 V1_OFF2, V0_OFF2, FA_1_1_OFF2, FA_1_0_OFF2, FA_0_1_OFF2,
947 * src/mips/ffi.c (ffi_prep_args): Code formatting. Fix
949 (ffi_prep_closure): Improve trampoline instruction scheduling.
950 (ffi_closure_mips_inner_O32): Fix endianness bugs.
952 2005-12-03 Alan Modra <amodra@bigpond.net.au>
954 * src/powerpc/ffi.c: Formatting.
955 (ffi_prep_args_SYSV): Avoid possible aliasing problems by using unions.
956 (ffi_prep_args64): Likewise.
958 2005-09-30 Geoffrey Keating <geoffk@apple.com>
960 * testsuite/lib/libffi-dg.exp (libffi_target_compile): For
961 darwin, use -shared-libgcc not -lgcc_s, and explain why.
963 2005-09-26 Tom Tromey <tromey@redhat.com>
965 * testsuite/libffi.call/float1.c (value_type): New typedef.
966 (CANARY): New define.
967 (main): Check for result buffer overflow.
968 * src/powerpc/linux64.S: Handle linux64 long double returns.
969 * src/powerpc/ffi.c (FLAG_RETURNS_128BITS): New constant.
970 (ffi_prep_cif_machdep): Handle linux64 long double returns.
972 2005-08-25 Alan Modra <amodra@bigpond.net.au>
975 * src/powerpc/ffi.c (ffi_prep_args_SYSV): Correct placement of stack
977 (ffi_status ffi_prep_cif_machdep): Correct stack sizing for same.
979 2005-08-11 Jakub Jelinek <jakub@redhat.com>
981 * configure.ac (HAVE_HIDDEN_VISIBILITY_ATTRIBUTE): New test.
982 (AH_BOTTOM): Add FFI_HIDDEN definition.
983 * configure: Rebuilt.
984 * fficonfig.h.in: Rebuilt.
985 * src/powerpc/ffi.c (hidden): Remove.
986 (ffi_closure_LINUX64, ffi_prep_args64, ffi_call_LINUX64,
987 ffi_closure_helper_LINUX64): Use FFI_HIDDEN instead of hidden.
988 * src/powerpc/linux64_closure.S (ffi_closure_LINUX64,
989 .ffi_closure_LINUX64): Use FFI_HIDDEN instead of .hidden.
990 * src/x86/ffi.c (ffi_closure_SYSV, ffi_closure_raw_SYSV): Remove,
991 add FFI_HIDDEN to its prototype.
992 (ffi_closure_SYSV_inner): New.
993 * src/x86/sysv.S (ffi_closure_SYSV, ffi_closure_raw_SYSV): New.
994 * src/x86/win32.S (ffi_closure_SYSV, ffi_closure_raw_SYSV): New.
996 2005-08-10 Alfred M. Szmidt <ams@gnu.org>
999 * configure: Rebuilt.
1000 * configure.ac: Handle i*86-*-gnu*.
1002 2005-08-09 Jakub Jelinek <jakub@redhat.com>
1004 * src/powerpc/ppc_closure.S (ffi_closure_SYSV): Use
1005 DW_CFA_offset_extended_sf rather than
1006 DW_CFA_GNU_negative_offset_extended.
1007 * src/powerpc/sysv.S (ffi_call_SYSV): Likewise.
1009 2005-07-22 SUGIOKA Toshinobu <sugioka@itonet.co.jp>
1011 * src/sh/sysv.S (ffi_call_SYSV): Stop argument popping correctly
1013 (ffi_closure_SYSV): Change the stack layout for sh3 struct argument.
1014 * src/sh/ffi.c (ffi_prep_args): Fix sh3 argument copy, when it is
1015 partially on register.
1016 (ffi_closure_helper_SYSV): Likewise.
1017 (ffi_prep_cif_machdep): Don't set too many cif->flags.
1019 2005-07-20 Kaz Kojima <kkojima@gcc.gnu.org>
1021 * src/sh/ffi.c (ffi_call): Handle small structures correctly.
1023 * src/sh64/ffi.c (simple_type): Remove.
1024 (return_type): Handle small structures correctly.
1025 (ffi_prep_args): Likewise.
1026 (ffi_call): Likewise.
1027 (ffi_closure_helper_SYSV): Likewise.
1028 * src/sh64/sysv.S (ffi_call_SYSV): Handle 1, 2 and 4-byte return.
1029 Emit position independent code if PIC and remove wrong datalabel
1030 prefixes from EH data.
1032 2005-07-19 Andreas Tobler <a.tobler@schweiz.ch>
1034 * Makefile.am (nodist_libffi_la_SOURCES): Add POWERPC_FREEBSD.
1035 * Makefile.in: Regenerate.
1036 * include/Makefile.in: Likewise.
1037 * testsuite/Makefile.in: Likewise.
1038 * configure.ac: Add POWERPC_FREEBSD rules.
1039 * configure: Regenerate.
1040 * src/powerpc/ffitarget.h: Add POWERPC_FREEBSD rules.
1041 (FFI_SYSV_TYPE_SMALL_STRUCT): Define.
1042 * src/powerpc/ffi.c: Add flags to handle small structure returns
1044 (ffi_prep_cif_machdep): Handle small structures for SYSV 4 ABI.
1046 (ffi_closure_helper_SYSV): Likewise.
1047 * src/powerpc/ppc_closure.S: Add return types for small structures.
1048 * src/powerpc/sysv.S: Add bits to handle small structures for
1051 2005-07-10 Andreas Tobler <a.tobler@schweiz.ch>
1053 * testsuite/libffi.call/cls_5_1_byte.c: New test file.
1054 * testsuite/libffi.call/cls_6_1_byte.c: Likewise.
1055 * testsuite/libffi.call/cls_7_1_byte.c: Likewise.
1057 2005-07-05 Randolph Chung <tausq@debian.org>
1059 * src/pa/ffi.c (ffi_struct_type): Rename FFI_TYPE_SMALL_STRUCT1
1060 as FFI_TYPE_SMALL_STRUCT3. Break out handling for 5-7 byte
1061 structures. Kill compilation warnings.
1062 (ffi_closure_inner_LINUX): Print return values as hex in debug
1063 message. Rename FFI_TYPE_SMALL_STRUCT1 as FFI_TYPE_SMALL_STRUCT3.
1064 Properly handle 5-7 byte structure returns.
1065 * src/pa/ffitarget.h (FFI_TYPE_SMALL_STRUCT1)
1066 (FFI_TYPE_SMALL_STRUCT2): Remove.
1067 (FFI_TYPE_SMALL_STRUCT3, FFI_TYPE_SMALL_STRUCT5)
1068 (FFI_TYPE_SMALL_STRUCT6, FFI_TYPE_SMALL_STRUCT7): Define.
1069 * src/pa/linux.S: Mark source file as using PA1.1 assembly.
1070 (checksmst1, checksmst2): Remove.
1071 (checksmst3): Optimize handling of 3-byte struct returns.
1072 (checksmst567): Properly handle 5-7 byte struct returns.
1074 2005-06-15 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
1077 * src/mips/n32.S: Enforce PIC code.
1078 * src/mips/o32.S: Likewise.
1080 2005-06-15 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
1082 * configure.ac: Treat i*86-*-solaris2.10 and up as X86_64.
1083 * configure: Regenerate.
1085 2005-06-01 Alan Modra <amodra@bigpond.net.au>
1087 * src/powerpc/ppc_closure.S (ffi_closure_SYSV): Don't use JUMPTARGET
1088 to call ffi_closure_helper_SYSV. Append @local instead.
1089 * src/powerpc/sysv.S (ffi_call_SYSV): Likewise for ffi_prep_args_SYSV.
1091 2005-05-17 Kelley Cook <kcook@gcc.gnu.org>
1093 * configure.ac: Use AC_C_BIGENDIAN instead of AC_C_BIGENDIAN_CROSS.
1094 Use AC_CHECK_SIZEOF instead of AC_COMPILE_CHECK_SIZEOF.
1095 * Makefile.am (ACLOCAL_AMFLAGS): Remove -I ../config.
1096 * aclocal.m4, configure, fficonfig.h.in, Makefile.in,
1097 include/Makefile.in, testsuite/Makefile.in: Regenerate.
1099 2005-05-09 Mike Stump <mrs@apple.com>
1101 * configure: Regenerate.
1103 2005-05-08 Richard Henderson <rth@redhat.com>
1106 * src/alpha/osf.S: Update unwind into to match code.
1108 2005-05-04 Andreas Degert <ad@papyrus-gmbh.de>
1109 Richard Henderson <rth@redhat.com>
1111 * src/x86/ffi64.c (ffi_prep_cif_machdep): Save sse-used flag in
1113 (ffi_call): Mask return type field. Pass ssecount to ffi_call_unix64.
1114 (ffi_prep_closure): Set carry bit if sse-used flag set.
1115 * src/x86/unix64.S (ffi_call_unix64): Add ssecount argument.
1116 Only load sse registers if ssecount non-zero.
1117 (ffi_closure_unix64): Only save sse registers if carry set on entry.
1119 2005-04-29 Ralf Corsepius <ralf.corsepius@rtems.org>
1121 * configure.ac: Add i*86-*-rtems*, sparc*-*-rtems*,
1122 powerpc-*rtems*, arm*-*-rtems*, sh-*-rtems*.
1123 * configure: Regenerate.
1125 2005-04-20 Hans-Peter Nilsson <hp@axis.com>
1127 * testsuite/lib/libffi-dg.exp (libffi-dg-test-1): In regsub use,
1128 have Tcl8.3-compatible intermediate variable.
1130 2005-04-18 Simon Posnjak <simon.posnjak@siol.net>
1131 Hans-Peter Nilsson <hp@axis.com>
1133 * Makefile.am: Add CRIS support.
1134 * configure.ac: Likewise.
1135 * Makefile.in, configure, testsuite/Makefile.in,
1136 include/Makefile.in: Regenerate.
1137 * src/cris: New directory.
1138 * src/cris/ffi.c, src/cris/sysv.S, src/cris/ffitarget.h: New files.
1139 * src/prep_cif.c (ffi_prep_cif): Wrap in #ifndef __CRIS__.
1141 * testsuite/lib/libffi-dg.exp (libffi-dg-test-1): Replace \n with
1142 \r?\n in output tests.
1144 2005-04-12 Mike Stump <mrs@apple.com>
1146 * configure: Regenerate.
1148 2005-03-30 Hans Boehm <Hans.Boehm@hp.com>
1150 * src/ia64/ffitarget.h (ffi_arg): Use long long instead of DI.
1152 2005-03-30 Steve Ellcey <sje@cup.hp.com>
1154 * src/ia64/ffitarget.h (ffi_arg) ADD DI attribute.
1156 * src/ia64/unix.S (ffi_closure_unix): Extend gp
1157 to 64 bits in ILP32 mode.
1158 Load 64 bits even for short data.
1160 2005-03-23 Mike Stump <mrs@apple.com>
1162 * src/powerpc/darwin.S: Update for -m64 multilib.
1163 * src/powerpc/darwin_closure.S: Likewise.
1165 2005-03-21 Zack Weinberg <zack@codesourcery.com>
1167 * configure.ac: Do not invoke TL_AC_GCC_VERSION.
1168 Do not set tool_include_dir.
1169 * aclocal.m4, configure, Makefile.in, testsuite/Makefile.in:
1171 * include/Makefile.am: Set gcc_version and toollibffidir.
1172 * include/Makefile.in: Regenerate.
1174 2005-02-22 Andrew Haley <aph@redhat.com>
1176 * src/powerpc/ffi.c (ffi_prep_cif_machdep): Bump alignment to
1177 odd-numbered register pairs for 64-bit integer types.
1179 2005-02-23 Andreas Tobler <a.tobler@schweiz.ch>
1182 * testsuite/libffi.call/return_ll1.c: New test case.
1184 2005-02-11 Janis Johnson <janis187@us.ibm.com>
1186 * testsuite/libffi.call/cls_align_longdouble.c: Remove dg-options.
1187 * testsuite/libffi.call/float.c: Ditto.
1188 * testsuite/libffi.call/float2.c: Ditto.
1189 * testsuite/libffi.call/float3.c: Ditto.
1191 2005-02-08 Andreas Tobler <a.tobler@schweiz.ch>
1193 * src/frv/ffitarget.h: Remove PPC stuff which does not belong to frv.
1195 2005-01-12 Eric Botcazou <ebotcazou@libertysurf.fr>
1197 * testsuite/libffi.special/special.exp (cxx_options): Add
1200 2004-12-31 Richard Henderson <rth@redhat.com>
1202 * src/types.c (FFI_AGGREGATE_TYPEDEF): Remove.
1203 (FFI_TYPEDEF): Rename from FFI_INTEGRAL_TYPEDEF. Replace size and
1204 offset parameters with a type parameter; deduce size and structure
1205 alignment. Update all users.
1207 2004-12-31 Richard Henderson <rth@redhat.com>
1209 * src/types.c (FFI_TYPE_POINTER): Define with sizeof.
1210 (FFI_TYPE_LONGDOUBLE): Fix for ia64.
1211 * src/ia64/ffitarget.h (struct ffi_ia64_trampoline_struct): Move
1212 into ffi_prep_closure.
1213 * src/ia64/ia64_flags.h, src/ia64/ffi.c, src/ia64/unix.S: Rewrite
1216 2004-12-27 Richard Henderson <rth@redhat.com>
1218 * src/x86/unix64.S: Fix typo in unwind info.
1220 2004-12-25 Richard Henderson <rth@redhat.com>
1222 * src/x86/ffi64.c (struct register_args): Rename from stackLayout.
1223 (enum x86_64_reg_class): Add X86_64_COMPLEX_X87_CLASS.
1224 (merge_classes): Check for it.
1226 (classify_argument): Pass byte_offset by value; perform all updates
1228 (examine_argument): Add classes argument; handle
1229 X86_64_COMPLEX_X87_CLASS.
1230 (ffi_prep_args): Merge into ...
1231 (ffi_call): ... here. Share stack frame with ffi_call_unix64.
1232 (ffi_prep_cif_machdep): Setup cif->flags for proper structure return.
1233 (ffi_fill_return_value): Remove.
1234 (ffi_prep_closure): Remove dead assert.
1235 (ffi_closure_unix64_inner): Rename from ffi_closure_UNIX64_inner.
1236 Rewrite to use struct register_args instead of va_list. Create
1237 flags for handling structure returns.
1238 * src/x86/unix64.S: Remove dead strings.
1239 (ffi_call_unix64): Rename from ffi_call_UNIX64. Rewrite to share
1240 stack frame with ffi_call. Handle structure returns properly.
1241 (float2sse, floatfloat2sse, double2sse): Remove.
1242 (sse2float, sse2double, sse2floatfloat): Remove.
1243 (ffi_closure_unix64): Rename from ffi_closure_UNIX64. Rewrite
1244 to handle structure returns properly.
1246 2004-12-08 David Edelsohn <edelsohn@gnu.org>
1248 * Makefile.am (AM_MAKEFLAGS): Remove duplicate LIBCFLAGS and
1250 * Makefile.in: Regenerated.
1252 2004-12-02 Richard Sandiford <rsandifo@redhat.com>
1254 * configure.ac: Use TL_AC_GCC_VERSION to set gcc_version.
1255 * configure, aclocal.m4, Makefile.in: Regenerate.
1256 * include/Makefile.in, testsuite/Makefile.in: Regenerate.
1258 2004-11-29 Kelley Cook <kcook@gcc.gnu.org>
1260 * configure: Regenerate for libtool change.
1262 2004-11-25 Kelley Cook <kcook@gcc.gnu.org>
1264 * configure: Regenerate for libtool reversion.
1266 2004-11-24 Kelley Cook <kcook@gcc.gnu.org>
1268 * configure: Regenerate for libtool change.
1270 2004-11-23 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
1272 * testsuite/lib/libffi-dg.exp: Use new procs in target-libpath.exp.
1274 2004-11-23 Richard Sandiford <rsandifo@redhat.com>
1276 * src/mips/o32.S (ffi_call_O32, ffi_closure_O32): Use jalr instead
1277 of jal. Use an absolute encoding for the frame information.
1279 2004-11-23 Kelley Cook <kcook@gcc.gnu.org>
1281 * Makefile.am: Remove no-dependencies. Add ACLOCAL_AMFLAGS.
1282 * acinclude.m4: Delete logic for sincludes.
1283 * aclocal.m4, Makefile.in, configure: Regenerate.
1284 * include/Makefile: Likewise.
1285 * testsuite/Makefile: Likewise.
1287 2004-11-22 Eric Botcazou <ebotcazou@libertysurf.fr>
1289 * src/sparc/ffi.c (ffi_prep_closure): Align doubles and 64-bit integers
1290 on a 8-byte boundary.
1291 * src/sparc/v8.S (ffi_closure_v8): Reserve frame space for arguments.
1293 2004-10-27 Richard Earnshaw <rearnsha@arm.com>
1295 * src/arm/ffi.c (ffi_prep_cif_machdep): Handle functions that return
1296 long long values. Round stack allocation to a multiple of 8 bytes
1297 for ATPCS compatibility.
1298 * src/arm/sysv.S (ffi_call_SYSV): Rework to avoid use of APCS register
1299 names. Handle returning long long types. Add Thumb and interworking
1300 support. Improve soft-float code.
1302 2004-10-27 Richard Earnshaw <rearnsha@arm.com>
1304 * testsuite/lib/libffi-db.exp (load_gcc_lib): New function.
1305 (libffi_exit): New function.
1306 (libffi_init): Build the testglue wrapper if needed.
1308 2004-10-25 Eric Botcazou <ebotcazou@libertysurf.fr>
1311 * testsuite/lib/libffi-dg.exp: Accept more than one multilib libgcc.
1313 2004-10-25 Kazuhiro Inaoka <inaoka.kazuhiro@renesas.com>
1315 * src/m32r/libffitarget.h (FFI_CLOSURES): Set to 0.
1317 2004-10-20 Kaz Kojima <kkojima@gcc.gnu.org>
1319 * src/sh/sysv.S (ffi_call_SYSV): Don't align for double data.
1320 * testsuite/libffi.call/float3.c: New test case.
1322 2004-10-18 Kaz Kojima <kkojima@gcc.gnu.org>
1324 * src/sh/ffi.c (ffi_prep_closure): Set T bit in trampoline for
1325 the function returning a structure pointed with R2.
1326 * src/sh/sysv.S (ffi_closure_SYSV): Use R2 as the pointer to
1327 the structure return value if T bit set. Emit position
1328 independent code and EH data if PIC.
1330 2004-10-13 Kazuhiro Inaoka <inaoka.kazuhiro@renesas.com>
1332 * Makefile.am: Add m32r support.
1333 * configure.ac: Likewise.
1334 * Makefile.in: Regenerate.
1335 * confiugre: Regenerate.
1336 * src/types.c: Add m32r port to FFI_INTERNAL_TYPEDEF
1337 (uint64, sint64, double, longdouble)
1338 * src/m32r: New directory.
1339 * src/m32r/ffi.c: New file.
1340 * src/m32r/sysv.S: Likewise.
1341 * src/m32r/ffitarget.h: Likewise.
1343 2004-10-02 Kaz Kojima <kkojima@gcc.gnu.org>
1345 * testsuite/libffi.call/negint.c: New test case.
1347 2004-09-14 H.J. Lu <hongjiu.lu@intel.com>
1350 * testsuite/lib/libffi-dg.exp: Don't use global ld_library_path.
1351 Set up LD_LIBRARY_PATH, SHLIB_PATH, LD_LIBRARYN32_PATH,
1352 LD_LIBRARY64_PATH, LD_LIBRARY_PATH_32, LD_LIBRARY_PATH_64 and
1355 2004-09-05 Andreas Tobler <a.tobler@schweiz.ch>
1357 * testsuite/libffi.call/many_win32.c: Remove whitespaces.
1358 * testsuite/libffi.call/promotion.c: Likewise.
1359 * testsuite/libffi.call/return_ll.c: Remove unused var. Cleanup
1361 * testsuite/libffi.call/return_sc.c: Likewise.
1362 * testsuite/libffi.call/return_uc.c: Likewise.
1364 2004-09-05 Andreas Tobler <a.tobler@schweiz.ch>
1366 * src/powerpc/darwin.S: Fix comments and identation.
1367 * src/powerpc/darwin_closure.S: Likewise.
1369 2004-09-02 Andreas Tobler <a.tobler@schweiz.ch>
1371 * src/powerpc/ffi_darwin.c: Add flag for longdouble return values.
1372 (ffi_prep_args): Handle longdouble arguments.
1373 (ffi_prep_cif_machdep): Set flags for longdouble. Calculate space for
1375 (ffi_closure_helper_DARWIN): Add closure handling for longdouble.
1376 * src/powerpc/darwin.S (_ffi_call_DARWIN): Add handling of longdouble
1378 * src/powerpc/darwin_closure.S (_ffi_closure_ASM): Likewise.
1379 * src/types.c: Defined longdouble size and alignment for darwin.
1381 2004-09-02 Andreas Tobler <a.tobler@schweiz.ch>
1383 * src/powerpc/aix.S: Remove whitespaces.
1384 * src/powerpc/aix_closure.S: Likewise.
1385 * src/powerpc/asm.h: Likewise.
1386 * src/powerpc/ffi.c: Likewise.
1387 * src/powerpc/ffitarget.h: Likewise.
1388 * src/powerpc/linux64.S: Likewise.
1389 * src/powerpc/linux64_closure.S: Likewise.
1390 * src/powerpc/ppc_closure.S: Likewise.
1391 * src/powerpc/sysv.S: Likewise.
1393 2004-08-30 Anthony Green <green@redhat.com>
1395 * Makefile.am: Add frv support.
1396 * Makefile.in, testsuite/Makefile.in: Rebuilt.
1397 * configure.ac: Read configure.host.
1398 * configure.in: Read configure.host.
1399 * configure.host: New file. frv-elf needs libgloss.
1400 * include/ffi.h.in: Force ffi_closure to have a nice big (8)
1401 alignment. This is needed to frv and shouldn't harm the others.
1402 * include/ffi_common.h (ALIGN_DOWN): New macro.
1403 * src/frv/ffi.c, src/frv/ffitarget.h, src/frv/eabi.S: New files.
1405 2004-08-24 David Daney <daney@avtrex.com>
1407 * testsuite/libffi.call/closure_fn0.c: Xfail mips64* instead of mips*.
1408 * testsuite/libffi.call/closure_fn1.c: Likewise.
1409 * testsuite/libffi.call/closure_fn2.c Likewise.
1410 * testsuite/libffi.call/closure_fn3.c: Likewise.
1411 * testsuite/libffi.call/closure_fn4.c: Likewise.
1412 * testsuite/libffi.call/closure_fn5.c: Likewise.
1413 * testsuite/libffi.call/cls_18byte.c: Likewise.
1414 * testsuite/libffi.call/cls_19byte.c: Likewise.
1415 * testsuite/libffi.call/cls_1_1byte.c: Likewise.
1416 * testsuite/libffi.call/cls_20byte.c: Likewise.
1417 * testsuite/libffi.call/cls_20byte1.c: Likewise.
1418 * testsuite/libffi.call/cls_24byte.c: Likewise.
1419 * testsuite/libffi.call/cls_2byte.c: Likewise.
1420 * testsuite/libffi.call/cls_3_1byte.c: Likewise.
1421 * testsuite/libffi.call/cls_3byte1.c: Likewise.
1422 * testsuite/libffi.call/cls_3byte2.c: Likewise.
1423 * testsuite/libffi.call/cls_4_1byte.c: Likewise.
1424 * testsuite/libffi.call/cls_4byte.c: Likewise.
1425 * testsuite/libffi.call/cls_64byte.c: Likewise.
1426 * testsuite/libffi.call/cls_6byte.c: Likewise.
1427 * testsuite/libffi.call/cls_7byte.c: Likewise.
1428 * testsuite/libffi.call/cls_8byte.c: Likewise.
1429 * testsuite/libffi.call/cls_9byte1.c: Likewise.
1430 * testsuite/libffi.call/cls_9byte2.c: Likewise.
1431 * testsuite/libffi.call/cls_align_double.c: Likewise.
1432 * testsuite/libffi.call/cls_align_float.c: Likewise.
1433 * testsuite/libffi.call/cls_align_longdouble.c: Likewise.
1434 * testsuite/libffi.call/cls_align_pointer.c: Likewise.
1435 * testsuite/libffi.call/cls_align_sint16.c: Likewise.
1436 * testsuite/libffi.call/cls_align_sint32.c: Likewise.
1437 * testsuite/libffi.call/cls_align_sint64.c: Likewise.
1438 * testsuite/libffi.call/cls_align_uint16.c: Likewise.
1439 * testsuite/libffi.call/cls_align_uint32.c: Likewise.
1440 * testsuite/libffi.call/cls_align_uint64.c: Likewise.
1441 * testsuite/libffi.call/cls_double.c: Likewise.
1442 * testsuite/libffi.call/cls_float.c: Likewise.
1443 * testsuite/libffi.call/cls_multi_schar.c: Likewise.
1444 * testsuite/libffi.call/cls_multi_sshort.c: Likewise.
1445 * testsuite/libffi.call/cls_multi_sshortchar.c: Likewise.
1446 * testsuite/libffi.call/cls_multi_uchar.c: Likewise.
1447 * testsuite/libffi.call/cls_multi_ushort.c: Likewise.
1448 * testsuite/libffi.call/cls_multi_ushortchar.c: Likewise.
1449 * testsuite/libffi.call/cls_schar.c: Likewise.
1450 * testsuite/libffi.call/cls_sint.c: Likewise.
1451 * testsuite/libffi.call/cls_sshort.c: Likewise.
1452 * testsuite/libffi.call/cls_uchar.c: Likewise.
1453 * testsuite/libffi.call/cls_uint.c: Likewise.
1454 * testsuite/libffi.call/cls_ulonglong.c: Likewise.
1455 * testsuite/libffi.call/cls_ushort.c: Likewise.
1456 * testsuite/libffi.call/nested_struct.c: Likewise.
1457 * testsuite/libffi.call/nested_struct1.c: Likewise.
1458 * testsuite/libffi.call/nested_struct2.c: Likewise.
1459 * testsuite/libffi.call/nested_struct3.c: Likewise.
1460 * testsuite/libffi.call/problem1.c: Likewise.
1461 * testsuite/libffi.special/unwindtest.cc: Likewise.
1462 * testsuite/libffi.call/cls_12byte.c: Likewise and set return value
1464 * testsuite/libffi.call/cls_16byte.c: Likewise.
1465 * testsuite/libffi.call/cls_5byte.c: Likewise.
1467 2004-08-23 David Daney <daney@avtrex.com>
1470 * src/mips/ffitarget.h (FFI_O32_SOFT_FLOAT): New ABI.
1471 * src/mips/ffi.c (ffi_prep_args): Fix alignment calculation.
1472 (ffi_prep_cif_machdep): Handle FFI_O32_SOFT_FLOAT floating point
1473 parameters and return types.
1474 (ffi_call): Handle FFI_O32_SOFT_FLOAT ABI.
1475 (ffi_prep_closure): Ditto.
1476 (ffi_closure_mips_inner_O32): Handle FFI_O32_SOFT_FLOAT ABI, fix
1477 alignment calculations.
1478 * src/mips/o32.S (ffi_closure_O32): Don't use floating point
1479 instructions if FFI_O32_SOFT_FLOAT, make stack frame ABI compliant.
1481 2004-08-14 Casey Marshall <csm@gnu.org>
1483 * src/mips/ffi.c (ffi_pref_cif_machdep): set `cif->flags' to
1484 contain `FFI_TYPE_UINT64' as return type for any 64-bit
1485 integer (O32 ABI only).
1486 (ffi_prep_closure): new function.
1487 (ffi_closure_mips_inner_O32): new function.
1488 * src/mips/ffitarget.h: Define `FFI_CLOSURES' and
1489 `FFI_TRAMPOLINE_SIZE' appropriately if the ABI is o32.
1490 * src/mips/o32.S (ffi_call_O32): add labels for .eh_frame. Return
1491 64 bit integers correctly.
1492 (ffi_closure_O32): new function.
1493 Added DWARF-2 unwind info for both functions.
1495 2004-08-10 Andrew Haley <aph@redhat.com>
1497 * src/x86/ffi64.c (ffi_prep_args ): 8-align all stack arguments.
1499 2004-08-01 Robert Millan <robertmh@gnu.org>
1501 * configure.ac: Detect knetbsd-gnu and kfreebsd-gnu.
1502 * configure: Regenerate.
1504 2004-07-30 Maciej W. Rozycki <macro@linux-mips.org>
1506 * acinclude.m4 (AC_FUNC_MMAP_BLACKLIST): Check for <sys/mman.h>
1507 and mmap() explicitly instead of relying on preset autoconf cache
1509 * aclocal.m4: Regenerate.
1510 * configure: Regenerate.
1512 2004-07-11 Ulrich Weigand <uweigand@de.ibm.com>
1514 * src/s390/ffi.c (ffi_prep_args): Fix C aliasing violation.
1515 (ffi_check_float_struct): Remove unused prototype.
1517 2004-06-30 Geoffrey Keating <geoffk@apple.com>
1519 * src/powerpc/ffi_darwin.c (flush_icache): ';' is a comment
1520 character on Darwin, use '\n\t' instead.
1522 2004-06-26 Matthias Klose <doko@debian.org>
1524 * libtool-version: Fix typo in revision/age.
1526 2004-06-17 Matthias Klose <doko@debian.org>
1528 * libtool-version: New.
1529 * Makefile.am (libffi_la_LDFLAGS): Use -version-info for soname.
1530 * Makefile.in: Regenerate.
1532 2004-06-15 Paolo Bonzini <bonzini@gnu.org>
1534 * Makefile.am: Remove useless multilib rules.
1535 * Makefile.in: Regenerate.
1536 * aclocal.m4: Regenerate with automake 1.8.5.
1537 * configure.ac: Remove useless multilib configury.
1538 * configure: Regenerate.
1540 2004-06-15 Paolo Bonzini <bonzini@gnu.org>
1542 * .cvsignore: New file.
1544 2004-06-10 Jakub Jelinek <jakub@redhat.com>
1546 * src/ia64/unix.S (ffi_call_unix): Insert group barrier break
1548 (ffi_closure_UNIX): Fix f14/f15 adjustment if FLOAT_SZ is ever
1551 2004-06-06 Sean McNeil <sean@mcneil.com>
1553 * configure.ac: Add x86_64-*-freebsd* support.
1554 * configure: Regenerate.
1556 2004-04-26 Joe Buck <jbuck@welsh-buck.org>
1559 * configure.ac: Test for existence of mmap and sys/mman.h before
1560 checking blacklist. Fix suggested by Jim Wilson.
1561 * configure: Regenerate.
1563 2004-04-26 Matt Austern <austern@apple.com>
1565 * src/powerpc/darwin.S: Go through a non-lazy pointer for initial
1567 * src/powerpc/darwin_closure.S: Likewise.
1569 2004-04-24 Andreas Tobler <a.tobler@schweiz.ch>
1571 * testsuite/libffi.call/cls_multi_schar.c (main): Fix initialization
1572 error. Reported by Thomas Heller <theller@python.net>.
1573 * testsuite/libffi.call/cls_multi_sshort.c (main): Likewise.
1574 * testsuite/libffi.call/cls_multi_ushort.c (main): Likewise.
1576 2004-03-20 Matthias Klose <doko@debian.org>
1578 * src/pa/linux.S: Fix typo.
1580 2004-03-19 Matthias Klose <doko@debian.org>
1582 * Makefile.am: Update.
1583 * Makefile.in: Regenerate.
1584 * src/pa/ffi.h.in: Remove.
1585 * src/pa/ffitarget.h: New file.
1587 2004-02-10 Randolph Chung <tausq@debian.org>
1589 * Makefile.am: Add PA support.
1590 * Makefile.in: Regenerate.
1591 * include/Makefile.in: Regenerate.
1592 * configure.ac: Add PA target.
1593 * configure: Regenerate.
1594 * src/pa/ffi.c: New file.
1595 * src/pa/ffi.h.in: Add PA support.
1596 * src/pa/linux.S: New file.
1597 * prep_cif.c: Add PA support.
1599 2004-03-16 Hosaka Yuji <hos@tamanegi.org>
1601 * src/types.c: Fix alignment size of X86_WIN32 case int64 and
1603 * src/x86/ffi.c (ffi_prep_args): Replace ecif->cif->rtype->type
1604 with ecif->cif->flags.
1605 (ffi_call, ffi_prep_incoming_args_SYSV): Replace cif->rtype->type
1607 (ffi_prep_cif_machdep): Add X86_WIN32 struct case.
1608 (ffi_closure_SYSV): Add 1 or 2-bytes struct case for X86_WIN32.
1609 * src/x86/win32.S (retstruct1b, retstruct2b, sc_retstruct1b,
1610 sc_retstruct2b): Add for 1 or 2-bytes struct case.
1612 2004-03-15 Kelley Cook <kcook@gcc.gnu.org>
1614 * configure.in: Rename file to ...
1615 * configure.ac: ... this.
1616 * fficonfig.h.in: Regenerate.
1617 * Makefile.in: Regenerate.
1618 * include/Makefile.in: Regenerate.
1619 * testsuite/Makefile.in: Regenerate.
1621 2004-03-12 Matt Austern <austern@apple.com>
1623 * src/powerpc/darwin.S: Fix EH information so it corresponds to
1624 changes in EH format resulting from addition of linkonce support.
1625 * src/powerpc/darwin_closure.S: Likewise.
1627 2004-03-11 Andreas Tobler <a.tobler@schweiz.ch>
1628 Paolo Bonzini <bonzini@gnu.org>
1630 * Makefile.am (AUTOMAKE_OPTIONS): Set them.
1631 Remove VPATH. Remove rules for object files. Remove multilib support.
1632 (AM_CCASFLAGS): Add.
1633 * configure.in (AC_CONFIG_HEADERS): Relace AM_CONFIG_HEADER.
1634 (AC_PREREQ): Bump version to 2.59.
1635 (AC_INIT): Fill with version info and bug address.
1636 (ORIGINAL_LD_FOR_MULTILIBS): Remove.
1637 (AM_ENABLE_MULTILIB): Use this instead of AC_ARG_ENABLE.
1638 De-precious CC so that the right flags are passed down to multilibs.
1639 (AC_MSG_ERROR): Replace obsolete macro AC_ERROR.
1640 (AC_CONFIG_FILES): Replace obsolete macro AC_LINK_FILES.
1641 (AC_OUTPUT): Reorganize the output with AC_CONFIG_COMMANDS.
1642 * configure: Rebuilt.
1643 * aclocal.m4: Likewise.
1644 * Makefile.in, include/Makefile.in, testsuite/Makefile.in: Likewise.
1645 * fficonfig.h.in: Likewise.
1647 2004-03-11 Andreas Schwab <schwab@suse.de>
1649 * src/ia64/ffi.c (ffi_prep_incoming_args_UNIX): Get floating point
1650 arguments from fp registers only for the first 8 parameter slots.
1651 Don't convert a float parameter when passed in memory.
1653 2004-03-09 Hans-Peter Nilsson <hp@axis.com>
1655 * configure: Regenerate for config/accross.m4 correction.
1657 2004-02-25 Matt Kraai <kraai@alumni.cmu.edu>
1659 * src/powerpc/ffi.c (ffi_prep_args_SYSV): Change
1660 ecif->cif->bytes to bytes.
1661 (ffi_prep_cif_machdep): Add braces around nested if statement.
1663 2004-02-09 Alan Modra <amodra@bigpond.net.au>
1665 * src/types.c (pointer): POWERPC64 has 8 byte pointers.
1667 * src/powerpc/ffi.c (ffi_prep_args64): Correct long double handling.
1668 (ffi_closure_helper_LINUX64): Fix typo.
1669 * testsuite/libffi.call/cls_align_longdouble.c: Pass -mlong-double-128
1671 * testsuite/libffi.call/float.c: Likewise.
1672 * testsuite/libffi.call/float2.c: Likewise.
1674 2004-02-08 Alan Modra <amodra@bigpond.net.au>
1676 * src/powerpc/ffi.c (ffi_prep_cif_machdep <FFI_LINUX64>): Correct
1677 long double function return and long double arg handling.
1678 (ffi_closure_helper_LINUX64): Formatting. Delete unused "ng" var.
1679 Use "end_pfr" instead of "nf". Correct long double handling.
1681 * src/powerpc/linux64.S (ffi_call_LINUX64): Save f2 long double
1683 * src/powerpc/linux64_closure.S (ffi_closure_LINUX64): Allocate
1684 space for long double return value. Adjust stack frame and offsets.
1685 Load f2 long double return.
1687 2004-02-07 Alan Modra <amodra@bigpond.net.au>
1689 * src/types.c: Use 16 byte long double for POWERPC64.
1691 2004-01-25 Eric Botcazou <ebotcazou@libertysurf.fr>
1693 * src/sparc/ffi.c (ffi_prep_args_v9): Shift the parameter array
1694 when the structure return address is passed in %o0.
1695 (ffi_V9_return_struct): Rename into ffi_v9_layout_struct.
1696 (ffi_v9_layout_struct): Align the field following a nested structure
1697 on a word boundary. Use memmove instead of memcpy.
1698 (ffi_call): Update call to ffi_V9_return_struct.
1699 (ffi_prep_closure): Define 'ctx' only for V8.
1700 (ffi_closure_sparc_inner): Clone into ffi_closure_sparc_inner_v8
1701 and ffi_closure_sparc_inner_v9.
1702 (ffi_closure_sparc_inner_v8): Return long doubles by reference.
1703 Always skip the structure return address. For structures and long
1704 doubles, copy the argument directly.
1705 (ffi_closure_sparc_inner_v9): Skip the structure return address only
1706 if required. Shift the maximum floating-point slot accordingly. For
1707 big structures, copy the argument directly; otherwise, left-justify the
1708 argument and call ffi_v9_layout_struct to lay out the structure on
1710 * src/sparc/v8.S: Undef STACKFRAME before defining it.
1711 (ffi_closure_v8): Pass the structure return address. Update call to
1712 ffi_closure_sparc_inner_v8. Short-circuit FFI_TYPE_INT handling.
1713 Skip the 'unimp' insn when returning long doubles and structures.
1714 * src/sparc/v9.S: Undef STACKFRAME before defining it.
1715 (ffi_closure_v9): Increase the frame size by 2 words. Short-circuit
1716 FFI_TYPE_INT handling. Load structures both in integers and
1717 floating-point registers on return.
1718 * README: Update status of the SPARC port.
1720 2004-01-24 Andreas Tobler <a.tobler@schweiz.ch>
1722 * testsuite/libffi.call/pyobjc-tc.c (main): Treat result value
1724 * testsuite/libffi.call/struct3.c (main): Fix CHECK.
1726 2004-01-22 Ulrich Weigand <uweigand@de.ibm.com>
1728 * testsuite/libffi.call/cls_uint.c (cls_ret_uint_fn): Treat result
1729 value as of type ffi_arg, not unsigned int.
1731 2004-01-21 Michael Ritzert <ritzert@t-online.de>
1733 * ffi64.c (ffi_prep_args): Cast the RHS of an assignment instead
1736 2004-01-12 Andreas Tobler <a.tobler@schweiz.ch>
1738 * testsuite/lib/libffi-dg.exp: Set LD_LIBRARY_PATH_32 for
1741 2004-01-08 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
1743 * testsuite/libffi.call/ffitest.h (allocate_mmap): Cast MAP_FAILED
1746 2003-12-10 Richard Henderson <rth@redhat.com>
1748 * testsuite/libffi.call/cls_align_pointer.c: Cast pointers to
1749 size_t instead of int.
1751 2003-12-04 Hosaka Yuji <hos@tamanegi.org>
1753 * testsuite/libffi.call/many_win32.c: Include <float.h>.
1754 * testsuite/libffi.call/many_win32.c (main): Replace variable
1755 int i with unsigned long ul.
1757 * testsuite/libffi.call/cls_align_uint64.c: New test case.
1758 * testsuite/libffi.call/cls_align_sint64.c: Likewise.
1759 * testsuite/libffi.call/cls_align_uint32.c: Likewise.
1760 * testsuite/libffi.call/cls_align_sint32.c: Likewise.
1761 * testsuite/libffi.call/cls_align_uint16.c: Likewise.
1762 * testsuite/libffi.call/cls_align_sint16.c: Likewise.
1763 * testsuite/libffi.call/cls_align_float.c: Likewise.
1764 * testsuite/libffi.call/cls_align_double.c: Likewise.
1765 * testsuite/libffi.call/cls_align_longdouble.c: Likewise.
1766 * testsuite/libffi.call/cls_align_pointer.c: Likewise.
1768 2003-12-02 Hosaka Yuji <hos@tamanegi.org>
1771 * src/x86/ffi.c (ffi_prep_args, ffi_prep_incoming_args_SYSV):
1772 Align arguments to 32 bits.
1774 2003-12-01 Andreas Tobler <a.tobler@schweiz.ch>
1777 * testsuite/libffi.call/cls_multi_sshort.c: New test case.
1778 * testsuite/libffi.call/cls_multi_sshortchar.c: Likewise.
1779 * testsuite/libffi.call/cls_multi_uchar.c: Likewise.
1780 * testsuite/libffi.call/cls_multi_schar.c: Likewise.
1781 * testsuite/libffi.call/cls_multi_ushortchar.c: Likewise.
1782 * testsuite/libffi.call/cls_multi_ushort.c: Likewise.
1784 * testsuite/libffi.special/unwindtest.cc: Cosmetics.
1786 2003-11-26 Kaveh R. Ghazi <ghazi@caip.rutgers.edu>
1788 * testsuite/libffi.call/ffitest.h: Include <fcntl.h>.
1789 * testsuite/libffi.special/ffitestcxx.h: Likewise.
1791 2003-11-22 Andreas Tobler <a.tobler@schweiz.ch>
1793 * Makefile.in: Rebuilt.
1794 * configure: Likewise.
1795 * testsuite/libffi.special/unwindtest.cc: Convert the mmap to
1798 2003-11-21 Andreas Jaeger <aj@suse.de>
1799 Andreas Tobler <a.tobler@schweiz.ch>
1801 * acinclude.m4: Add AC_FUNC_MMAP_BLACKLIST.
1802 * configure.in: Call AC_FUNC_MMAP_BLACKLIST.
1803 * Makefile.in: Rebuilt.
1804 * aclocal.m4: Likewise.
1805 * configure: Likewise.
1806 * fficonfig.h.in: Likewise.
1807 * testsuite/lib/libffi-dg.exp: Add include dir.
1808 * testsuite/libffi.call/ffitest.h: Add MMAP definitions.
1809 * testsuite/libffi.special/ffitestcxx.h: Likewise.
1810 * testsuite/libffi.call/closure_fn0.c: Use MMAP functionality
1811 for ffi_closure if available.
1812 * testsuite/libffi.call/closure_fn1.c: Likewise.
1813 * testsuite/libffi.call/closure_fn2.c: Likewise.
1814 * testsuite/libffi.call/closure_fn3.c: Likewise.
1815 * testsuite/libffi.call/closure_fn4.c: Likewise.
1816 * testsuite/libffi.call/closure_fn5.c: Likewise.
1817 * testsuite/libffi.call/cls_12byte.c: Likewise.
1818 * testsuite/libffi.call/cls_16byte.c: Likewise.
1819 * testsuite/libffi.call/cls_18byte.c: Likewise.
1820 * testsuite/libffi.call/cls_19byte.c: Likewise.
1821 * testsuite/libffi.call/cls_1_1byte.c: Likewise.
1822 * testsuite/libffi.call/cls_20byte.c: Likewise.
1823 * testsuite/libffi.call/cls_20byte1.c: Likewise.
1824 * testsuite/libffi.call/cls_24byte.c: Likewise.
1825 * testsuite/libffi.call/cls_2byte.c: Likewise.
1826 * testsuite/libffi.call/cls_3_1byte.c: Likewise.
1827 * testsuite/libffi.call/cls_3byte1.c: Likewise.
1828 * testsuite/libffi.call/cls_3byte2.c: Likewise.
1829 * testsuite/libffi.call/cls_4_1byte.c: Likewise.
1830 * testsuite/libffi.call/cls_4byte.c: Likewise.
1831 * testsuite/libffi.call/cls_5byte.c: Likewise.
1832 * testsuite/libffi.call/cls_64byte.c: Likewise.
1833 * testsuite/libffi.call/cls_6byte.c: Likewise.
1834 * testsuite/libffi.call/cls_7byte.c: Likewise.
1835 * testsuite/libffi.call/cls_8byte.c: Likewise.
1836 * testsuite/libffi.call/cls_9byte1.c: Likewise.
1837 * testsuite/libffi.call/cls_9byte2.c: Likewise.
1838 * testsuite/libffi.call/cls_double.c: Likewise.
1839 * testsuite/libffi.call/cls_float.c: Likewise.
1840 * testsuite/libffi.call/cls_schar.c: Likewise.
1841 * testsuite/libffi.call/cls_sint.c: Likewise.
1842 * testsuite/libffi.call/cls_sshort.c: Likewise.
1843 * testsuite/libffi.call/cls_uchar.c: Likewise.
1844 * testsuite/libffi.call/cls_uint.c: Likewise.
1845 * testsuite/libffi.call/cls_ulonglong.c: Likewise.
1846 * testsuite/libffi.call/cls_ushort.c: Likewise.
1847 * testsuite/libffi.call/nested_struct.c: Likewise.
1848 * testsuite/libffi.call/nested_struct1.c: Likewise.
1849 * testsuite/libffi.call/nested_struct2.c: Likewise.
1850 * testsuite/libffi.call/nested_struct3.c: Likewise.
1851 * testsuite/libffi.call/problem1.c: Likewise.
1852 * testsuite/libffi.special/unwindtest.cc: Likewise.
1854 2003-11-20 Andreas Tobler <a.tobler@schweiz.ch>
1856 * testsuite/lib/libffi-dg.exp: Make the -lgcc_s conditional.
1858 2003-11-19 Andreas Tobler <a.tobler@schweiz.ch>
1860 * testsuite/lib/libffi-dg.exp: Add DYLD_LIBRARY_PATH for darwin.
1861 Add -lgcc_s to additional flags.
1863 2003-11-12 Andreas Tobler <a.tobler@schweiz.ch>
1865 * configure.in, include/Makefile.am: PR libgcj/11147, install
1866 the ffitarget.h header file in a gcc versioned and target
1868 * configure: Regenerated.
1869 * Makefile.in, include/Makefile.in: Likewise.
1870 * testsuite/Makefile.in: Likewise.
1872 2003-11-09 Andreas Tobler <a.tobler@schweiz.ch>
1874 * testsuite/libffi.call/closure_fn0.c: Print result and check
1875 with dg-output to make debugging easier.
1876 * testsuite/libffi.call/closure_fn1.c: Likewise.
1877 * testsuite/libffi.call/closure_fn2.c: Likewise.
1878 * testsuite/libffi.call/closure_fn3.c: Likewise.
1879 * testsuite/libffi.call/closure_fn4.c: Likewise.
1880 * testsuite/libffi.call/closure_fn5.c: Likewise.
1881 * testsuite/libffi.call/cls_12byte.c: Likewise.
1882 * testsuite/libffi.call/cls_16byte.c: Likewise.
1883 * testsuite/libffi.call/cls_18byte.c: Likewise.
1884 * testsuite/libffi.call/cls_19byte.c: Likewise.
1885 * testsuite/libffi.call/cls_1_1byte.c: Likewise.
1886 * testsuite/libffi.call/cls_20byte.c: Likewise.
1887 * testsuite/libffi.call/cls_20byte1.c: Likewise.
1888 * testsuite/libffi.call/cls_24byte.c: Likewise.
1889 * testsuite/libffi.call/cls_2byte.c: Likewise.
1890 * testsuite/libffi.call/cls_3_1byte.c: Likewise.
1891 * testsuite/libffi.call/cls_3byte1.c: Likewise.
1892 * testsuite/libffi.call/cls_3byte2.c: Likewise.
1893 * testsuite/libffi.call/cls_4_1byte.c: Likewise.
1894 * testsuite/libffi.call/cls_4byte.c: Likewise.
1895 * testsuite/libffi.call/cls_5byte.c: Likewise.
1896 * testsuite/libffi.call/cls_64byte.c: Likewise.
1897 * testsuite/libffi.call/cls_6byte.c: Likewise.
1898 * testsuite/libffi.call/cls_7byte.c: Likewise.
1899 * testsuite/libffi.call/cls_8byte.c: Likewise.
1900 * testsuite/libffi.call/cls_9byte1.c: Likewise.
1901 * testsuite/libffi.call/cls_9byte2.c: Likewise.
1902 * testsuite/libffi.call/cls_double.c: Likewise.
1903 * testsuite/libffi.call/cls_float.c: Likewise.
1904 * testsuite/libffi.call/cls_schar.c: Likewise.
1905 * testsuite/libffi.call/cls_sint.c: Likewise.
1906 * testsuite/libffi.call/cls_sshort.c: Likewise.
1907 * testsuite/libffi.call/cls_uchar.c: Likewise.
1908 * testsuite/libffi.call/cls_uint.c: Likewise.
1909 * testsuite/libffi.call/cls_ulonglong.c: Likewise.
1910 * testsuite/libffi.call/cls_ushort.c: Likewise.
1911 * testsuite/libffi.call/problem1.c: Likewise.
1913 * testsuite/libffi.special/unwindtest.cc: Make ffi_closure
1916 2003-11-08 Andreas Tobler <a.tobler@schweiz.ch>
1918 * testsuite/libffi.call/cls_9byte2.c: New test case.
1919 * testsuite/libffi.call/cls_9byte1.c: Likewise.
1920 * testsuite/libffi.call/cls_64byte.c: Likewise.
1921 * testsuite/libffi.call/cls_20byte1.c: Likewise.
1922 * testsuite/libffi.call/cls_19byte.c: Likewise.
1923 * testsuite/libffi.call/cls_18byte.c: Likewise.
1924 * testsuite/libffi.call/closure_fn4.c: Likewise.
1925 * testsuite/libffi.call/closure_fn5.c: Likewise.
1926 * testsuite/libffi.call/cls_schar.c: Likewise.
1927 * testsuite/libffi.call/cls_sint.c: Likewise.
1928 * testsuite/libffi.call/cls_sshort.c: Likewise.
1929 * testsuite/libffi.call/nested_struct2.c: Likewise.
1930 * testsuite/libffi.call/nested_struct3.c: Likewise.
1932 2003-11-08 Andreas Tobler <a.tobler@schweiz.ch>
1934 * testsuite/libffi.call/cls_double.c: Do a check on the result.
1935 * testsuite/libffi.call/cls_uchar.c: Likewise.
1936 * testsuite/libffi.call/cls_uint.c: Likewise.
1937 * testsuite/libffi.call/cls_ulonglong.c: Likewise.
1938 * testsuite/libffi.call/cls_ushort.c: Likewise.
1939 * testsuite/libffi.call/return_sc.c: Cleanup whitespaces.
1941 2003-11-06 Andreas Tobler <a.tobler@schweiz.ch>
1943 * src/prep_cif.c (ffi_prep_cif): Move the validity check after
1946 2003-10-23 Andreas Tobler <a.tobler@schweiz.ch>
1948 * src/java_raw_api.c (ffi_java_ptrarray_to_raw): Replace
1949 FFI_ASSERT(FALSE) with FFI_ASSERT(0).
1951 2003-10-22 David Daney <ddaney@avtrex.com>
1953 * src/mips/ffitarget.h: Replace undefined UINT32 and friends with
1954 __attribute__((__mode__(__SI__))) and friends.
1956 2003-10-22 Andreas Schwab <schwab@suse.de>
1958 * src/ia64/ffi.c: Replace FALSE/TRUE with false/true.
1960 2003-10-21 Andreas Tobler <a.tobler@schweiz.ch>
1962 * configure.in: AC_LINK_FILES(ffitarget.h).
1963 * configure: Regenerate.
1964 * Makefile.in: Likewise.
1965 * include/Makefile.in: Likewise.
1966 * testsuite/Makefile.in: Likewise.
1967 * fficonfig.h.in: Likewise.
1969 2003-10-21 Paolo Bonzini <bonzini@gnu.org>
1970 Richard Henderson <rth@redhat.com>
1972 Avoid that ffi.h includes fficonfig.h.
1974 * Makefile.am (EXTRA_DIST): Include ffitarget.h files
1975 (TARGET_SRC_MIPS_GCC): Renamed to TARGET_SRC_MIPS_IRIX.
1976 (TARGET_SRC_MIPS_SGI): Removed.
1977 (MIPS_GCC): Renamed to TARGET_SRC_MIPS_IRIX.
1978 (MIPS_SGI): Removed.
1979 (CLEANFILES): Removed.
1980 (mostlyclean-am, clean-am, mostlyclean-sub, clean-sub): New
1982 * acconfig.h: Removed.
1983 * configure.in: Compute sizeofs only for double and long double.
1984 Use them to define and subst HAVE_LONG_DOUBLE. Include comments
1985 into AC_DEFINE instead of using acconfig.h. Create
1986 include/ffitarget.h instead of include/fficonfig.h. Rename
1987 MIPS_GCC to MIPS_IRIX, drop MIPS_SGI since we are in gcc's tree.
1988 AC_DEFINE EH_FRAME_FLAGS.
1989 * include/Makefile.am (DISTCLEANFILES): New automake macro.
1990 (hack_DATA): Add ffitarget.h.
1991 * include/ffi.h.in: Remove all system specific definitions.
1992 Declare raw API even if it is not installed, why bother?
1993 Use limits.h instead of SIZEOF_* to define ffi_type_*. Do
1994 not define EH_FRAME_FLAGS, it is in fficonfig.h now. Include
1995 ffitarget.h instead of fficonfig.h. Remove ALIGN macro.
1996 (UINT_ARG, INT_ARG): Removed, use ffi_arg and ffi_sarg instead.
1997 * include/ffi_common.h (bool): Do not define.
1998 (ffi_assert): Accept failed assertion.
1999 (ffi_type_test): Return void and accept file/line.
2000 (FFI_ASSERT): Pass stringized failed assertion.
2001 (FFI_ASSERT_AT): New macro.
2002 (FFI_ASSERT_VALID_TYPE): New macro.
2003 (UINT8, SINT8, UINT16, SINT16, UINT32, SINT32,
2004 UINT64, SINT64): Define here with gcc's __attribute__ macro
2006 (FLOAT32, ALIGN): Define here instead of in ffi.h
2007 * include/ffi-mips.h: Removed. Its content moved to
2008 src/mips/ffitarget.h after separating assembly and C sections.
2009 * src/alpha/ffi.c, src/alpha/ffi.c, src/java_raw_api.c
2010 src/prep_cif.c, src/raw_api.c, src/ia64/ffi.c,
2011 src/mips/ffi.c, src/mips/n32.S, src/mips/o32.S,
2012 src/mips/ffitarget.h, src/sparc/ffi.c, src/x86/ffi64.c:
2013 SIZEOF_ARG -> FFI_SIZEOF_ARG.
2014 * src/ia64/ffi.c: Include stdbool.h (provided by GCC 2.95+).
2015 * src/debug.c (ffi_assert): Accept stringized failed assertion.
2016 (ffi_type_test): Rewritten.
2017 * src/prep-cif.c (initialize_aggregate, ffi_prep_cif): Call
2018 FFI_ASSERT_VALID_TYPE.
2019 * src/alpha/ffitarget.h, src/arm/ffitarget.h,
2020 src/ia64/ffitarget.h, src/m68k/ffitarget.h,
2021 src/mips/ffitarget.h, src/powerpc/ffitarget.h,
2022 src/s390/ffitarget.h, src/sh/ffitarget.h,
2023 src/sh64/ffitarget.h, src/sparc/ffitarget.h,
2024 src/x86/ffitarget.h: New files.
2025 * src/alpha/osf.S, src/arm/sysv.S, src/ia64/unix.S,
2026 src/m68k/sysv.S, src/mips/n32.S, src/mips/o32.S,
2027 src/powerpc/aix.S, src/powerpc/darwin.S,
2028 src/powerpc/ffi_darwin.c, src/powerpc/linux64.S,
2029 src/powerpc/linux64_closure.S, src/powerpc/ppc_closure.S,
2030 src/powerpc/sysv.S, src/s390/sysv.S, src/sh/sysv.S,
2031 src/sh64/sysv.S, src/sparc/v8.S, src/sparc/v9.S,
2032 src/x86/sysv.S, src/x86/unix64.S, src/x86/win32.S:
2035 2003-10-20 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
2037 * src/mips/ffi.c: Use _ABIN32, _ABIO32 instead of external
2038 _MIPS_SIM_NABI32, _MIPS_SIM_ABI32.
2040 2003-10-19 Andreas Tobler <a.tobler@schweiz.ch>
2042 * src/powerpc/ffi_darwin.c (ffi_prep_args): Declare bytes again.
2043 Used when FFI_DEBUG = 1.
2045 2003-10-14 Alan Modra <amodra@bigpond.net.au>
2047 * src/types.c (double, longdouble): Default POWERPC64 to 8 byte size
2050 2003-10-06 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
2052 * include/ffi_mips.h: Define FFI_MIPS_N32 for N32/N64 ABIs,
2053 FFI_MIPS_O32 for O32 ABI.
2055 2003-10-01 Andreas Tobler <a.tobler@schweiz.ch>
2057 * testsuite/lib/libffi-dg.exp: Set LD_LIBRARY_PATH_64 for
2058 SPARC64. Cleanup whitespaces.
2060 2003-09-19 Andreas Tobler <a.tobler@schweiz.ch>
2062 * testsuite/libffi.call/closure_fn0.c: Xfail mips, arm,
2063 strongarm, xscale. Cleanup whitespaces.
2064 * testsuite/libffi.call/closure_fn1.c: Likewise.
2065 * testsuite/libffi.call/closure_fn2.c: Likewise.
2066 * testsuite/libffi.call/closure_fn3.c: Likewise.
2067 * testsuite/libffi.call/cls_12byte.c: Likewise.
2068 * testsuite/libffi.call/cls_16byte.c: Likewise.
2069 * testsuite/libffi.call/cls_1_1byte.c: Likewise.
2070 * testsuite/libffi.call/cls_20byte.c: Likewise.
2071 * testsuite/libffi.call/cls_24byte.c: Likewise.
2072 * testsuite/libffi.call/cls_2byte.c: Likewise.
2073 * testsuite/libffi.call/cls_3_1byte.c: Likewise.
2074 * testsuite/libffi.call/cls_3byte1.c: Likewise.
2075 * testsuite/libffi.call/cls_3byte2.c: Likewise.
2076 * testsuite/libffi.call/cls_4_1byte.c: Likewise.
2077 * testsuite/libffi.call/cls_4byte.c: Likewise.
2078 * testsuite/libffi.call/cls_5byte.c: Likewise.
2079 * testsuite/libffi.call/cls_6byte.c: Likewise.
2080 * testsuite/libffi.call/cls_7byte.c: Likewise.
2081 * testsuite/libffi.call/cls_8byte.c: Likewise.
2082 * testsuite/libffi.call/cls_double.c: Likewise.
2083 * testsuite/libffi.call/cls_float.c: Likewise.
2084 * testsuite/libffi.call/cls_uchar.c: Likewise.
2085 * testsuite/libffi.call/cls_uint.c: Likewise.
2086 * testsuite/libffi.call/cls_ulonglong.c: Likewise.
2087 * testsuite/libffi.call/cls_ushort.c: Likewise.
2088 * testsuite/libffi.call/nested_struct.c: Likewise.
2089 * testsuite/libffi.call/nested_struct1.c: Likewise.
2090 * testsuite/libffi.call/problem1.c: Likewise.
2091 * testsuite/libffi.special/unwindtest.cc: Likewise.
2092 * testsuite/libffi.call/pyobjc-tc.c: Cleanup whitespaces.
2094 2003-09-18 David Edelsohn <edelsohn@gnu.org>
2096 * src/powerpc/aix.S: Cleanup whitespaces.
2097 * src/powerpc/aix_closure.S: Likewise.
2099 2003-09-18 Andreas Tobler <a.tobler@schweiz.ch>
2101 * src/powerpc/darwin.S: Cleanup whitespaces, comment formatting.
2102 * src/powerpc/darwin_closure.S: Likewise.
2103 * src/powerpc/ffi_darwin.c: Likewise.
2105 2003-09-18 Andreas Tobler <a.tobler@schweiz.ch>
2106 David Edelsohn <edelsohn@gnu.org>
2108 * src/types.c (double): Add AIX and Darwin to the right TYPEDEF.
2109 * src/powerpc/aix_closure.S: Remove the pointer to the outgoing
2111 * src/powerpc/darwin_closure.S: Likewise.
2112 * src/powerpc/ffi_darwin.c (ffi_prep_args): Handle structures
2113 according to the Darwin/AIX ABI.
2114 (ffi_prep_cif_machdep): Likewise.
2115 (ffi_closure_helper_DARWIN): Likewise.
2116 Remove the outgoing parameter stack logic. Simplify the evaluation
2117 of the different CASE types.
2118 (ffi_prep_clousure): Avoid the casts on lvalues. Change the branch
2119 statement in the trampoline code.
2121 2003-09-18 Kaz Kojima <kkojima@gcc.gnu.org>
2123 * src/sh/ffi.c (ffi_prep_args): Take account into the alignement
2124 for the register size.
2125 (ffi_closure_helper_SYSV): Handle the structure return value
2127 (ffi_closure_helper_SYSV): Return the appropriate type when
2128 the registers are used for the structure return value.
2129 * src/sh/sysv.S (ffi_closure_SYSV): Fix the stack layout for
2130 the 64-bit return value. Update copyright years.
2132 2003-09-17 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
2134 * testsuite/lib/libffi-dg.exp (libffi_target_compile): Search in
2135 srcdir for ffi_mips.h.
2137 2003-09-12 Alan Modra <amodra@bigpond.net.au>
2139 * src/prep_cif.c (initialize_aggregate): Include tail padding in
2141 * src/powerpc/linux64_closure.S (ffi_closure_LINUX64): Correct
2142 placement of float result.
2143 * testsuite/libffi.special/unwindtest.cc (closure_test_fn1): Correct
2144 cast of "resp" for big-endian 64 bit machines.
2146 2003-09-11 Alan Modra <amodra@bigpond.net.au>
2148 * src/types.c (double, longdouble): Merge identical SH and ARM
2149 typedefs, and add POWERPC64.
2150 * src/powerpc/ffi.c (ffi_prep_args64): Correct next_arg calc for
2151 struct split over gpr and rest.
2152 (ffi_prep_cif_machdep): Correct intarg_count for structures.
2153 * src/powerpc/linux64.S (ffi_call_LINUX64): Fix gpr offsets.
2155 2003-09-09 Andreas Tobler <a.tobler@schweiz.ch>
2157 * src/powerpc/ffi.c (ffi_closure_helper_SYSV) Handle struct
2160 2003-09-09 Alan Modra <amodra@bigpond.net.au>
2162 * configure: Regenerate.
2164 2003-09-04 Andreas Tobler <a.tobler@schweiz.ch>
2166 * Makefile.am: Remove build rules for ffitest.
2167 * Makefile.in: Rebuilt.
2169 2003-09-04 Andreas Tobler <a.tobler@schweiz.ch>
2171 * src/java_raw_api.c: Include <stdlib.h> to fix compiler warning
2172 about implicit declaration of abort().
2174 2003-09-04 Andreas Tobler <a.tobler@schweiz.ch>
2176 * Makefile.am: Add dejagnu test framework. Fixes PR other/11411.
2177 * Makefile.in: Rebuilt.
2178 * configure.in: Add dejagnu test framework.
2179 * configure: Rebuilt.
2181 * testsuite/Makefile.am: New file.
2182 * testsuite/Makefile.in: Built
2183 * testsuite/lib/libffi-dg.exp: New file.
2184 * testsuite/config/default.exp: Likewise.
2185 * testsuite/libffi.call/call.exp: Likewise.
2186 * testsuite/libffi.call/ffitest.h: Likewise.
2187 * testsuite/libffi.call/closure_fn0.c: Likewise.
2188 * testsuite/libffi.call/closure_fn1.c: Likewise.
2189 * testsuite/libffi.call/closure_fn2.c: Likewise.
2190 * testsuite/libffi.call/closure_fn3.c: Likewise.
2191 * testsuite/libffi.call/cls_1_1byte.c: Likewise.
2192 * testsuite/libffi.call/cls_3_1byte.c: Likewise.
2193 * testsuite/libffi.call/cls_4_1byte.c: Likewise.
2194 * testsuite/libffi.call/cls_2byte.c: Likewise.
2195 * testsuite/libffi.call/cls_3byte1.c: Likewise.
2196 * testsuite/libffi.call/cls_3byte2.c: Likewise.
2197 * testsuite/libffi.call/cls_4byte.c: Likewise.
2198 * testsuite/libffi.call/cls_5byte.c: Likewise.
2199 * testsuite/libffi.call/cls_6byte.c: Likewise.
2200 * testsuite/libffi.call/cls_7byte.c: Likewise.
2201 * testsuite/libffi.call/cls_8byte.c: Likewise.
2202 * testsuite/libffi.call/cls_12byte.c: Likewise.
2203 * testsuite/libffi.call/cls_16byte.c: Likewise.
2204 * testsuite/libffi.call/cls_20byte.c: Likewise.
2205 * testsuite/libffi.call/cls_24byte.c: Likewise.
2206 * testsuite/libffi.call/cls_double.c: Likewise.
2207 * testsuite/libffi.call/cls_float.c: Likewise.
2208 * testsuite/libffi.call/cls_uchar.c: Likewise.
2209 * testsuite/libffi.call/cls_uint.c: Likewise.
2210 * testsuite/libffi.call/cls_ulonglong.c: Likewise.
2211 * testsuite/libffi.call/cls_ushort.c: Likewise.
2212 * testsuite/libffi.call/float.c: Likewise.
2213 * testsuite/libffi.call/float1.c: Likewise.
2214 * testsuite/libffi.call/float2.c: Likewise.
2215 * testsuite/libffi.call/many.c: Likewise.
2216 * testsuite/libffi.call/many_win32.c: Likewise.
2217 * testsuite/libffi.call/nested_struct.c: Likewise.
2218 * testsuite/libffi.call/nested_struct1.c: Likewise.
2219 * testsuite/libffi.call/pyobjc-tc.c: Likewise.
2220 * testsuite/libffi.call/problem1.c: Likewise.
2221 * testsuite/libffi.call/promotion.c: Likewise.
2222 * testsuite/libffi.call/return_ll.c: Likewise.
2223 * testsuite/libffi.call/return_sc.c: Likewise.
2224 * testsuite/libffi.call/return_uc.c: Likewise.
2225 * testsuite/libffi.call/strlen.c: Likewise.
2226 * testsuite/libffi.call/strlen_win32.c: Likewise.
2227 * testsuite/libffi.call/struct1.c: Likewise.
2228 * testsuite/libffi.call/struct2.c: Likewise.
2229 * testsuite/libffi.call/struct3.c: Likewise.
2230 * testsuite/libffi.call/struct4.c: Likewise.
2231 * testsuite/libffi.call/struct5.c: Likewise.
2232 * testsuite/libffi.call/struct6.c: Likewise.
2233 * testsuite/libffi.call/struct7.c: Likewise.
2234 * testsuite/libffi.call/struct8.c: Likewise.
2235 * testsuite/libffi.call/struct9.c: Likewise.
2236 * testsuite/libffi.special/special.exp: New file.
2237 * testsuite/libffi.special/ffitestcxx.h: Likewise.
2238 * testsuite/libffi.special/unwindtest.cc: Likewise.
2241 2003-08-13 Kaz Kojima <kkojima@gcc.gnu.org>
2243 * src/sh/ffi.c (OFS_INT16): Set 0 for little endian case. Update
2246 2003-08-02 Alan Modra <amodra@bigpond.net.au>
2248 * src/powerpc/ffi.c (ffi_prep_args64): Modify for changed gcc
2250 (ffi_closure_helper_LINUX64): Likewise.
2251 * src/powerpc/linux64.S: Remove code writing to parm save area.
2252 * src/powerpc/linux64_closure.S (ffi_closure_LINUX64): Use return
2253 address in lr from ffi_closure_helper_LINUX64 call to calculate
2254 table address. Optimize function tail.
2256 2003-07-28 Andreas Tobler <a.tobler@schweiz.ch>
2258 * src/sparc/ffi.c: Handle all floating point registers.
2259 * src/sparc/v9.S: Likewise. Fixes second part of PR target/11410.
2261 2003-07-11 Gerald Pfeifer <pfeifer@dbai.tuwien.ac.at>
2263 * README: Note that libffi is not part of GCC. Update the project
2266 2003-06-19 Franz Sirl <Franz.Sirl-kernel@lauterbach.com>
2268 * src/powerpc/ppc_closure.S: Include ffi.h.
2270 2003-06-13 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
2272 * src/x86/sysv.S: Avoid gas-only .uleb128/.sleb128 directives.
2273 Use C style comments.
2275 2003-06-13 Kaz Kojima <kkojima@rr.iij4u.or.jp>
2277 * Makefile.am: Add SHmedia support. Fix a typo of SH support.
2278 * Makefile.in: Regenerate.
2279 * configure.in (sh64-*-linux*, sh5*-*-linux*): Add target.
2280 * configure: Regenerate.
2281 * include/ffi.h.in: Add SHmedia support.
2282 * src/sh64/ffi.c: New file.
2283 * src/sh64/sysv.S: New file.
2285 2003-05-16 Jakub Jelinek <jakub@redhat.com>
2287 * configure.in (HAVE_RO_EH_FRAME): Check whether .eh_frame section
2288 should be read-only.
2289 * configure: Rebuilt.
2290 * fficonfig.h.in: Rebuilt.
2291 * include/ffi.h.in (EH_FRAME_FLAGS): Define.
2292 * src/alpha/osf.S: Use EH_FRAME_FLAGS.
2293 * src/powerpc/linux64.S: Likewise.
2294 * src/powerpc/linux64_closure.S: Likewise. Include ffi.h.
2295 * src/powerpc/sysv.S: Use EH_FRAME_FLAGS. Use pcrel encoding
2296 if -fpic/-fPIC/-mrelocatable.
2297 * src/powerpc/powerpc_closure.S: Likewise.
2298 * src/sparc/v8.S: If HAVE_RO_EH_FRAME is defined, don't include
2299 #write in .eh_frame flags.
2300 * src/sparc/v9.S: Likewise.
2301 * src/x86/unix64.S: Use EH_FRAME_FLAGS.
2302 * src/x86/sysv.S: Likewise. Use pcrel encoding if -fpic/-fPIC.
2303 * src/s390/sysv.S: Use EH_FRAME_FLAGS. Include ffi.h.
2305 2003-05-07 Jeff Sturm <jsturm@one-point.com>
2307 Fixes PR bootstrap/10656
2308 * configure.in (HAVE_AS_REGISTER_PSEUDO_OP): Test assembler
2309 support for .register pseudo-op.
2310 * src/sparc/v8.S: Use it.
2311 * fficonfig.h.in: Rebuilt.
2312 * configure: Rebuilt.
2314 2003-04-18 Jakub Jelinek <jakub@redhat.com>
2316 * include/ffi.h.in (POWERPC64): Define if 64-bit.
2317 (enum ffi_abi): Add FFI_LINUX64 on POWERPC.
2318 Make it the default on POWERPC64.
2319 (FFI_TRAMPOLINE_SIZE): Define to 24 on POWERPC64.
2320 * configure.in: Change powerpc-*-linux* into powerpc*-*-linux*.
2321 * configure: Rebuilt.
2322 * src/powerpc/ffi.c (hidden): Define.
2323 (ffi_prep_args_SYSV): Renamed from
2324 ffi_prep_args. Cast pointers to unsigned long to shut up warnings.
2325 (NUM_GPR_ARG_REGISTERS64, NUM_FPR_ARG_REGISTERS64,
2326 ASM_NEEDS_REGISTERS64): New.
2327 (ffi_prep_args64): New function.
2328 (ffi_prep_cif_machdep): Handle FFI_LINUX64 ABI.
2329 (ffi_call): Likewise.
2330 (ffi_prep_closure): Likewise.
2331 (flush_icache): Surround by #ifndef POWERPC64.
2332 (ffi_dblfl): New union type.
2333 (ffi_closure_helper_SYSV): Use it to avoid aliasing problems.
2334 (ffi_closure_helper_LINUX64): New function.
2335 * src/powerpc/ppc_closure.S: Surround whole file by #ifndef
2337 * src/powerpc/sysv.S: Likewise.
2338 (ffi_call_SYSV): Rename ffi_prep_args to ffi_prep_args_SYSV.
2339 * src/powerpc/linux64.S: New file.
2340 * src/powerpc/linux64_closure.S: New file.
2341 * Makefile.am (EXTRA_DIST): Add src/powerpc/linux64.S and
2342 src/powerpc/linux64_closure.S.
2343 (TARGET_SRC_POWERPC): Likewise.
2345 * src/ffitest.c (closure_test_fn, closure_test_fn1, closure_test_fn2,
2346 closure_test_fn3): Fix result printing on big-endian 64-bit
2348 (main): Print tst2_arg instead of uninitialized tst2_result.
2350 * src/ffitest.c (main): Hide what closure pointer really points to
2353 2003-04-16 Richard Earnshaw <rearnsha@arm.com>
2355 * configure.in (arm-*-netbsdelf*): Add configuration.
2356 (configure): Regenerated.
2358 2003-04-04 Loren J. Rittle <ljrittle@acm.org>
2360 * include/Makefile.in: Regenerate.
2362 2003-03-21 Zdenek Dvorak <rakdver@atrey.karlin.mff.cuni.cz>
2364 * libffi/include/ffi.h.in: Define X86 instead of X86_64 in 32
2366 * libffi/src/x86/ffi.c (ffi_closure_SYSV, ffi_closure_raw_SYSV):
2367 Receive closure pointer through parameter, read args using
2368 __builtin_dwarf_cfa.
2369 (FFI_INIT_TRAMPOLINE): Send closure reference through eax.
2371 2003-03-12 Andreas Schwab <schwab@suse.de>
2373 * configure.in: Avoid trailing /. in toolexeclibdir.
2374 * configure: Rebuilt.
2376 2003-03-03 Andreas Tobler <a.tobler@schweiz.ch>
2378 * src/powerpc/darwin_closure.S: Recode to fit dynamic libraries.
2380 2003-02-06 Andreas Tobler <a.tobler@schweiz.ch>
2382 * libffi/src/powerpc/darwin_closure.S:
2383 Fix alignement bug, allocate 8 bytes for the result.
2384 * libffi/src/powerpc/aix_closure.S:
2386 * libffi/src/powerpc/ffi_darwin.c:
2387 Update stackframe description for aix/darwin_closure.S.
2389 2003-02-06 Jakub Jelinek <jakub@redhat.com>
2391 * src/s390/ffi.c (ffi_closure_helper_SYSV): Add hidden visibility
2394 2003-01-31 Christian Cornelssen <ccorn@cs.tu-berlin.de>,
2395 Andreas Schwab <schwab@suse.de>
2397 * configure.in: Adjust command to source config-ml.in to account
2398 for changes to the libffi_basedir definition.
2399 (libffi_basedir): Remove ${srcdir} from value and include trailing
2402 * configure: Regenerate.
2404 2003-01-29 Franz Sirl <Franz.Sirl-kernel@lauterbach.com>
2406 * src/powerpc/ppc_closure.S: Recode to fit shared libs.
2408 2003-01-28 Andrew Haley <aph@redhat.com>
2410 * include/ffi.h.in: Enable FFI_CLOSURES for x86_64.
2411 * src/x86/ffi64.c (ffi_prep_closure): New.
2412 (ffi_closure_UNIX64_inner): New.
2413 * src/x86/unix64.S (ffi_closure_UNIX64): New.
2415 2003-01-27 Alexandre Oliva <aoliva@redhat.com>
2417 * configure.in (toolexecdir, toolexeclibdir): Set and AC_SUBST.
2418 Remove USE_LIBDIR conditional.
2419 * Makefile.am (toolexecdir, toolexeclibdir): Don't override.
2420 * Makefile.in, configure: Rebuilt.
2422 2003-01027 David Edelsohn <edelsohn@gnu.org>
2424 * Makefile.am (TARGET_SRC_POWERPC_AIX): Fix typo.
2425 * Makefile.in: Regenerate.
2427 2003-01-22 Andrew Haley <aph@redhat.com>
2429 * src/powerpc/darwin.S (_ffi_call_AIX): Add Augmentation size to
2432 2003-01-21 Andreas Tobler <a.tobler@schweiz.ch>
2434 * src/powerpc/darwin.S: Add unwind info.
2435 * src/powerpc/darwin_closure.S: Likewise.
2437 2003-01-14 Andrew Haley <aph@redhat.com>
2439 * src/x86/ffi64.c (ffi_prep_args): Check for void retval.
2440 (ffi_prep_cif_machdep): Likewise.
2441 * src/x86/unix64.S: Add unwind info.
2443 2003-01-14 Andreas Jaeger <aj@suse.de>
2445 * src/ffitest.c (main): Only use ffi_closures if those are
2448 2003-01-13 Andreas Tobler <a.tobler@schweiz.ch>
2450 * libffi/src/ffitest.c
2451 add closure testcases
2453 2003-01-13 Kevin B. Hendricks <khendricks@ivey.uwo.ca>
2455 * libffi/src/powerpc/ffi.c
2456 fix alignment bug for float (4 byte aligned iso 8 byte)
2458 2003-01-09 Geoffrey Keating <geoffk@apple.com>
2460 * src/powerpc/ffi_darwin.c: Remove RCS version string.
2461 * src/powerpc/darwin.S: Remove RCS version string.
2463 2003-01-03 Jeff Sturm <jsturm@one-point.com>
2465 * include/ffi.h.in: Add closure defines for SPARC, SPARC64.
2466 * src/ffitest.c (main): Use static storage for closure.
2467 * src/sparc/ffi.c (ffi_prep_closure, ffi_closure_sparc_inner): New.
2468 * src/sparc/v8.S (ffi_closure_v8): New.
2469 * src/sparc/v9.S (ffi_closure_v9): New.
2471 2002-11-10 Ranjit Mathew <rmathew@hotmail.com>
2473 * include/ffi.h.in: Added FFI_STDCALL ffi_type
2474 enumeration for X86_WIN32.
2475 * src/x86/win32.S: Added ffi_call_STDCALL function
2477 * src/x86/ffi.c (ffi_call/ffi_raw_call): Added
2478 switch cases for recognising FFI_STDCALL and
2479 calling ffi_call_STDCALL if target is X86_WIN32.
2480 * src/ffitest.c (my_stdcall_strlen/stdcall_many):
2481 stdcall versions of the "my_strlen" and "many"
2482 test functions (for X86_WIN32).
2483 Added test cases to test stdcall invocation using
2486 2002-12-02 Kaz Kojima <kkojima@gcc.gnu.org>
2488 * src/sh/sysv.S: Add DWARF2 unwind info.
2490 2002-11-27 Ulrich Weigand <uweigand@de.ibm.com>
2492 * src/s390/sysv.S (.eh_frame section): Make section read-only.
2494 2002-11-26 Jim Wilson <wilson@redhat.com>
2496 * src/types.c (FFI_TYPE_POINTER): Has size 8 on IA64.
2498 2002-11-23 H.J. Lu <hjl@gnu.org>
2500 * acinclude.m4: Add dummy AM_PROG_LIBTOOL.
2501 Include ../config/accross.m4.
2502 * aclocal.m4; Rebuild.
2503 * configure: Likewise.
2505 2002-11-15 Ulrich Weigand <uweigand@de.ibm.com>
2507 * src/s390/sysv.S (.eh_frame section): Adapt to pcrel FDE encoding.
2509 2002-11-11 DJ Delorie <dj@redhat.com>
2511 * configure.in: Look for common files in the right place.
2513 2002-10-08 Ulrich Weigand <uweigand@de.ibm.com>
2515 * src/java_raw_api.c (ffi_java_raw_to_ptrarray): Interpret
2516 raw data as _Jv_word values, not ffi_raw.
2517 (ffi_java_ptrarray_to_raw): Likewise.
2518 (ffi_java_rvalue_to_raw): New function.
2519 (ffi_java_raw_call): Call it.
2520 (ffi_java_raw_to_rvalue): New function.
2521 (ffi_java_translate_args): Call it.
2522 * src/ffitest.c (closure_test_fn): Interpret return value
2523 as ffi_arg, not int.
2524 * src/s390/ffi.c (ffi_prep_cif_machdep): Add missing
2525 FFI_TYPE_POINTER case.
2526 (ffi_closure_helper_SYSV): Likewise. Also, assume return
2527 values extended to word size.
2529 2002-10-02 Andreas Jaeger <aj@suse.de>
2531 * src/x86/ffi64.c (ffi_prep_cif_machdep): Remove debug output.
2533 2002-10-01 Bo Thorsen <bo@smetana.suse.de>
2535 * include/ffi.h.in: Fix i386 win32 compilation.
2537 2002-09-30 Ulrich Weigand <uweigand@de.ibm.com>
2539 * configure.in: Add s390x-*-linux-* target.
2540 * configure: Regenerate.
2541 * include/ffi.h.in: Define S390X for s390x targets.
2542 (FFI_CLOSURES): Define for s390/s390x.
2543 (FFI_TRAMPOLINE_SIZE): Likewise.
2544 (FFI_NATIVE_RAW_API): Likewise.
2545 * src/prep_cif.c (ffi_prep_cif): Do not compute stack space for s390.
2546 * src/types.c (FFI_TYPE_POINTER): Use 8-byte pointers on s390x.
2547 * src/s390/ffi.c: Major rework of existing code. Add support for
2548 s390x targets. Add closure support.
2549 * src/s390/sysv.S: Likewise.
2551 2002-09-29 Richard Earnshaw <rearnsha@arm.com>
2553 * src/arm/sysv.S: Fix typo.
2555 2002-09-28 Richard Earnshaw <rearnsha@arm.com>
2557 * src/arm/sysv.S: If we don't have machine/asm.h and the pre-processor
2558 has defined __USER_LABEL_PREFIX__, then use it in CNAME.
2559 (ffi_call_SYSV): Handle soft-float.
2561 2002-09-27 Bo Thorsen <bo@suse.de>
2563 * include/ffi.h.in: Fix multilib x86-64 support.
2565 2002-09-22 Kaveh R. Ghazi <ghazi@caip.rutgers.edu>
2567 * Makefile.am (all-multi): Fix multilib parallel build.
2569 2002-07-19 Kaz Kojima <kkojima@gcc.gnu.org>
2571 * configure.in (sh[34]*-*-linux*): Add brackets.
2572 * configure: Regenerate.
2574 2002-07-18 Kaz Kojima <kkojima@gcc.gnu.org>
2576 * Makefile.am: Add SH support.
2577 * Makefile.in: Regenerate.
2578 * configure.in (sh-*-linux*, sh[34]*-*-linux*): Add target.
2579 * configure: Regenerate.
2580 * include/ffi.h.in: Add SH support.
2581 * src/sh/ffi.c: New file.
2582 * src/sh/sysv.S: New file.
2583 * src/types.c: Add SH support.
2585 2002-07-16 Bo Thorsen <bo@suse.de>
2587 * src/x86/ffi64.c: New file that adds x86-64 support.
2588 * src/x86/unix64.S: New file that handles argument setup for
2590 * src/x86/sysv.S: Don't use this on x86-64.
2591 * src/x86/ffi.c: Don't use this on x86-64.
2593 * src/prep_cif.c (ffi_prep_cif): Don't do stack size calculation
2595 * src/ffitest.c (struct6): New test that tests a special case in
2597 (struct7): Likewise.
2598 (struct8): Likewise.
2599 (struct9): Likewise.
2600 (closure_test_fn): Silence warning about this when it's not used.
2601 (main): Add the new tests.
2602 (main): Fix a couple of wrong casts and silence some compiler warnings.
2603 * include/ffi.h.in: Add x86-64 ABI definition.
2604 * fficonfig.h.in: Regenerate.
2605 * Makefile.am: Add x86-64 support.
2606 * configure.in: Likewise.
2607 * Makefile.in: Regenerate.
2608 * configure: Likewise.
2610 2002-06-24 Bo Thorsen <bo@suse.de>
2612 * src/types.c: Merge settings for similar architectures.
2613 Add x86-64 sizes and alignments.
2615 2002-06-23 Bo Thorsen <bo@suse.de>
2617 * src/arm/ffi.c (ffi_prep_args): Remove unused vars.
2618 * src/sparc/ffi.c (ffi_prep_args_v8): Likewise.
2619 * src/mips/ffi.c (ffi_prep_args): Likewise.
2620 * src/m68k/ffi.c (ffi_prep_args): Likewise.
2622 2002-07-18 H.J. Lu (hjl@gnu.org)
2624 * Makefile.am (TARGET_SRC_MIPS_LINUX): New.
2625 (libffi_la_SOURCES): Support MIPS_LINUX.
2626 (libffi_convenience_la_SOURCES): Likewise.
2627 * Makefile.in: Regenerated.
2629 * configure.in (mips64*-*): Skip.
2630 (mips*-*-linux*): New.
2631 * configure: Regenerated.
2633 * src/mips/ffi.c: Include <sgidefs.h>.
2635 2002-06-06 Ulrich Weigand <uweigand@de.ibm.com>
2637 * src/s390/sysv.S: Save/restore %r6. Add DWARF-2 unwind info.
2639 2002-05-27 Roger Sayle <roger@eyesopen.com>
2641 * src/x86/ffi.c (ffi_prep_args): Remove reference to avn.
2643 2002-05-27 Bo Thorsen <bo@suse.de>
2645 * src/x86/ffi.c (ffi_prep_args): Remove unused variable and
2648 2002-05-13 Andreas Tobler <a.tobler@schweiz.ch>
2650 * src/powerpc/ffi_darwin.c (ffi_prep_closure): Declare fd at
2651 beginning of function (for older apple cc).
2653 2002-05-08 Alexandre Oliva <aoliva@redhat.com>
2655 * configure.in (ORIGINAL_LD_FOR_MULTILIBS): Preserve LD at
2656 script entry, and set LD to it when configuring multilibs.
2657 * configure: Rebuilt.
2659 2002-05-05 Jason Thorpe <thorpej@wasabisystems.com>
2661 * configure.in (sparc64-*-netbsd*): Add target.
2662 (sparc-*-netbsdelf*): Likewise.
2663 * configure: Regenerate.
2665 2002-04-28 David S. Miller <davem@redhat.com>
2667 * configure.in, configure: Fix SPARC test in previous change.
2669 2002-04-29 Gerhard Tonn <GerhardTonn@swol.de>
2671 * Makefile.am: Add Linux for S/390 support.
2672 * Makefile.in: Regenerate.
2673 * configure.in: Add Linux for S/390 support.
2674 * configure: Regenerate.
2675 * include/ffi.h.in: Add Linux for S/390 support.
2676 * src/s390/ffi.c: New file from libffi CVS tree.
2677 * src/s390/sysv.S: New file from libffi CVS tree.
2679 2002-04-28 Jakub Jelinek <jakub@redhat.com>
2681 * configure.in (HAVE_AS_SPARC_UA_PCREL): Check for working
2683 * src/sparc/v8.S: Use it.
2684 * src/sparc/v9.S: Likewise.
2685 * fficonfig.h.in: Rebuilt.
2686 * configure: Rebuilt.
2688 2002-04-08 Hans Boehm <Hans_Boehm@hp.com>
2690 * src/java_raw_api.c (ffi_java_raw_size): Handle FFI_TYPE_DOUBLE
2692 * src/ia64/unix.S: Add unwind information. Fix comments.
2693 Save sp in a way that's compatible with unwind info.
2694 (ffi_call_unix): Correctly restore sp in all cases.
2695 * src/ia64/ffi.c: Add, fix comments.
2697 2002-04-08 Jakub Jelinek <jakub@redhat.com>
2699 * src/sparc/v8.S: Make .eh_frame dependent on target word size.
2701 2002-04-06 Jason Thorpe <thorpej@wasabisystems.com>
2703 * configure.in (alpha*-*-netbsd*): Add target.
2704 * configure: Regenerate.
2706 2002-04-04 Jeff Sturm <jsturm@one-point.com>
2708 * src/sparc/v8.S: Add unwind info.
2709 * src/sparc/v9.S: Likewise.
2711 2002-03-30 Krister Walfridsson <cato@df.lth.se>
2713 * configure.in: Enable i*86-*-netbsdelf*.
2714 * configure: Rebuilt.
2716 2002-03-29 David Billinghurst <David.Billinghurst@riotinto.com>
2719 * src/mips/n32.s: Delete
2720 * src/mips/o32.s: Delete
2722 2002-03-21 Loren J. Rittle <ljrittle@acm.org>
2724 * configure.in: Enable alpha*-*-freebsd*.
2725 * configure: Rebuilt.
2727 2002-03-17 Bryce McKinlay <bryce@waitaki.otago.ac.nz>
2729 * Makefile.am: libfficonvenience -> libffi_convenience.
2730 * Makefile.in: Rebuilt.
2732 * Makefile.am: Define ffitest_OBJECTS.
2733 * Makefile.in: Rebuilt.
2735 2002-03-07 Andreas Tobler <toa@pop.agri.ch>
2736 David Edelsohn <edelsohn@gnu.org>
2738 * Makefile.am (EXTRA_DIST): Add Darwin and AIX closure files.
2739 (TARGET_SRC_POWERPC_AIX): Add aix_closure.S.
2740 (TARGET_SRC_POWERPC_DARWIN): Add darwin_closure.S.
2741 * Makefile.in: Regenerate.
2742 * include/ffi.h.in: Add AIX and Darwin closure definitions.
2743 * src/powerpc/ffi_darwin.c (ffi_prep_closure): New function.
2744 (flush_icache, flush_range): New functions.
2745 (ffi_closure_helper_DARWIN): New function.
2746 * src/powerpc/aix_closure.S: New file.
2747 * src/powerpc/darwin_closure.S: New file.
2749 2002-02-24 Jeff Sturm <jsturm@one-point.com>
2751 * include/ffi.h.in: Add typedef for ffi_arg.
2752 * src/ffitest.c (main): Declare rint with ffi_arg.
2754 2002-02-21 Andreas Tobler <toa@pop.agri.ch>
2756 * src/powerpc/ffi_darwin.c (ffi_prep_args): Skip appropriate
2757 number of GPRs for floating-point arguments.
2759 2002-01-31 Anthony Green <green@redhat.com>
2761 * configure: Rebuilt.
2762 * configure.in: Replace CHECK_SIZEOF and endian tests with
2763 cross-compiler friendly macros.
2764 * aclocal.m4 (AC_COMPILE_CHECK_SIZEOF, AC_C_BIGENDIAN_CROSS): New
2767 2002-01-18 David Edelsohn <edelsohn@gnu.org>
2769 * src/powerpc/darwin.S (_ffi_call_AIX): New.
2770 * src/powerpc/aix.S (ffi_call_DARWIN): New.
2772 2002-01-17 David Edelsohn <edelsohn@gnu.org>
2774 * Makefile.am (EXTRA_DIST): Add Darwin and AIX files.
2775 (TARGET_SRC_POWERPC_AIX): New.
2776 (POWERPC_AIX): New stanza.
2777 * Makefile.in: Regenerate.
2778 * configure.in: Add AIX case.
2779 * configure: Regenerate.
2780 * include/ffi.h.in (ffi_abi): Add FFI_AIX.
2781 * src/powerpc/ffi_darwin.c (ffi_status): Use "long" to scale frame
2782 size. Fix "long double" support.
2783 (ffi_call): Add FFI_AIX case.
2784 * src/powerpc/aix.S: New.
2786 2001-10-09 John Hornkvist <john@toastedmarshmallow.com>
2788 Implement Darwin PowerPC ABI.
2789 * configure.in: Handle powerpc-*-darwin*.
2790 * Makefile.am: Set source files for POWERPC_DARWIN.
2791 * configure: Rebuilt.
2792 * Makefile.in: Rebuilt.
2793 * include/ffi.h.in: Define FFI_DARWIN and FFI_DEFAULT_ABI for
2795 * src/powerpc/darwin.S: New file.
2796 * src/powerpc/ffi_darwin.c: New file.
2798 2001-10-07 Joseph S. Myers <jsm28@cam.ac.uk>
2800 * src/x86/ffi.c: Fix spelling error of "separate" as "seperate".
2802 2001-07-16 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
2804 * src/x86/sysv.S: Avoid gas-only .balign directive.
2805 Use C style comments.
2807 2001-07-16 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
2809 * src/alpha/ffi.c (ffi_prep_closure): Avoid gas-only mnemonic.
2810 Fixes PR bootstrap/3563.
2812 2001-06-26 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
2814 * src/alpha/osf.S (ffi_closure_osf): Use .rdata for ECOFF.
2816 2001-06-25 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
2818 * configure.in: Recognize sparc*-sun-* host.
2819 * configure: Regenerate.
2821 2001-06-06 Andrew Haley <aph@redhat.com>
2823 * src/alpha/osf.S (__FRAME_BEGIN__): Conditionalize for ELF.
2825 2001-06-03 Andrew Haley <aph@redhat.com>
2827 * src/alpha/osf.S: Add unwind info.
2828 * src/powerpc/sysv.S: Add unwind info.
2829 * src/powerpc/ppc_closure.S: Likewise.
2831 2000-05-31 Jeff Sturm <jsturm@one-point.com>
2833 * configure.in: Fix AC_ARG_ENABLE usage.
2834 * configure: Rebuilt.
2836 2001-05-06 Bryce McKinlay <bryce@waitaki.otago.ac.nz>
2838 * configure.in: Remove warning about beta code.
2839 * configure: Rebuilt.
2841 2001-04-25 Hans Boehm <Hans_Boehm@hp.com>
2843 * src/ia64/unix.S: Restore stack pointer when returning from
2845 * src/ia64/ffi.c: Fix typo in comment.
2847 2001-04-18 Jim Wilson <wilson@redhat.com>
2849 * src/ia64/unix.S: Delete unnecessary increment and decrement of loc2
2850 to eliminate RAW DV.
2852 2001-04-12 Bryce McKinlay <bryce@albatross.co.nz>
2854 * Makefile.am: Make a libtool convenience library.
2855 * Makefile.in: Rebuilt.
2857 2001-03-29 Bryce McKinlay <bryce@albatross.co.nz>
2859 * configure.in: Use different syntax for subdirectory creation.
2860 * configure: Rebuilt.
2862 2001-03-27 Jon Beniston <jon@beniston.com>
2864 * configure.in: Added X86_WIN32 target (Win32, CygWin, MingW).
2865 * configure: Rebuilt.
2866 * Makefile.am: Added X86_WIN32 target support.
2867 * Makefile.in: Rebuilt.
2869 * include/ffi.h.in: Added X86_WIN32 target support.
2871 * src/ffitest.c: Doesn't run structure tests for X86_WIN32 targets.
2872 * src/types.c: Added X86_WIN32 target support.
2874 * src/x86/win32.S: New file. Based on sysv.S, but with EH
2875 stuff removed and made to work with CygWin's gas.
2877 2001-03-26 Bryce McKinlay <bryce@albatross.co.nz>
2879 * configure.in: Make target subdirectory in build dir.
2880 * Makefile.am: Override suffix based rules to specify correct output
2882 * Makefile.in: Rebuilt.
2883 * configure: Rebuilt.
2885 2001-03-23 Kevin B Hendricks <khendricks@ivey.uwo.ca>
2887 * src/powerpc/ppc_closure.S: New file.
2888 * src/powerpc/ffi.c (ffi_prep_args): Fixed ABI compatibility bug
2889 involving long long and register pairs.
2890 (ffi_prep_closure): New function.
2891 (flush_icache): Likewise.
2892 (ffi_closure_helper_SYSV): Likewise.
2893 * include/ffi.h.in (FFI_CLOSURES): Define on PPC.
2894 (FFI_TRAMPOLINE_SIZE): Likewise.
2895 (FFI_NATIVE_RAW_API): Likewise.
2896 * Makefile.in: Rebuilt.
2897 * Makefile.am (EXTRA_DIST): Added src/powerpc/ppc_closure.S.
2898 (TARGET_SRC_POWERPC): Likewise.
2900 2001-03-19 Tom Tromey <tromey@redhat.com>
2902 * Makefile.in: Rebuilt.
2903 * Makefile.am (ffitest_LDFLAGS): New macro.
2905 2001-03-02 Nick Clifton <nickc@redhat.com>
2907 * include/ffi.h.in: Remove RCS ident string.
2908 * include/ffi_mips.h: Remove RCS ident string.
2909 * src/debug.c: Remove RCS ident string.
2910 * src/ffitest.c: Remove RCS ident string.
2911 * src/prep_cif.c: Remove RCS ident string.
2912 * src/types.c: Remove RCS ident string.
2913 * src/alpha/ffi.c: Remove RCS ident string.
2914 * src/alpha/osf.S: Remove RCS ident string.
2915 * src/arm/ffi.c: Remove RCS ident string.
2916 * src/arm/sysv.S: Remove RCS ident string.
2917 * src/mips/ffi.c: Remove RCS ident string.
2918 * src/mips/n32.S: Remove RCS ident string.
2919 * src/mips/o32.S: Remove RCS ident string.
2920 * src/sparc/ffi.c: Remove RCS ident string.
2921 * src/sparc/v8.S: Remove RCS ident string.
2922 * src/sparc/v9.S: Remove RCS ident string.
2923 * src/x86/ffi.c: Remove RCS ident string.
2924 * src/x86/sysv.S: Remove RCS ident string.
2926 2001-02-08 Joseph S. Myers <jsm28@cam.ac.uk>
2928 * include/ffi.h.in: Change sourceware.cygnus.com references to
2931 2000-12-09 Richard Henderson <rth@redhat.com>
2933 * src/alpha/ffi.c (ffi_call): Simplify struct return test.
2934 (ffi_closure_osf_inner): Index rather than increment avalue
2935 and arg_types. Give ffi_closure_osf the raw return value type.
2936 * src/alpha/osf.S (ffi_closure_osf): Handle return value type
2939 2000-12-07 Richard Henderson <rth@redhat.com>
2941 * src/raw_api.c (ffi_translate_args): Fix typo.
2942 (ffi_prep_closure): Likewise.
2944 * include/ffi.h.in [ALPHA]: Define FFI_CLOSURES and
2945 FFI_TRAMPOLINE_SIZE.
2946 * src/alpha/ffi.c (ffi_prep_cif_machdep): Adjust minimal
2947 cif->bytes for new ffi_call_osf implementation.
2948 (ffi_prep_args): Absorb into ...
2949 (ffi_call): ... here. Do all stack allocation here and
2950 avoid a callback function.
2951 (ffi_prep_closure, ffi_closure_osf_inner): New.
2952 * src/alpha/osf.S (ffi_call_osf): Reimplement with no callback.
2953 (ffi_closure_osf): New.
2955 2000-09-10 Alexandre Oliva <aoliva@redhat.com>
2957 * config.guess, config.sub, install-sh: Removed.
2958 * ltconfig, ltmain.sh, missing, mkinstalldirs: Likewise.
2959 * Makefile.in: Rebuilt.
2961 * acinclude.m4: Include libtool macros from the top level.
2962 * aclocal.m4, configure: Rebuilt.
2964 2000-08-22 Alexandre Oliva <aoliva@redhat.com>
2966 * configure.in [i*86-*-freebsd*] (TARGET, TARGETDIR): Set.
2967 * configure: Rebuilt.
2969 2000-05-11 Scott Bambrough <scottb@netwinder.org>
2971 * libffi/src/arm/sysv.S (ffi_call_SYSV): Doubles are not saved to
2972 memory correctly. Use conditional instructions, not branches where
2975 2000-05-04 Tom Tromey <tromey@cygnus.com>
2977 * configure: Rebuilt.
2978 * configure.in: Match `arm*-*-linux-*'.
2979 From Chris Dornan <cdornan@arm.com>.
2981 2000-04-28 Jakub Jelinek <jakub@redhat.com>
2983 * Makefile.am (SUBDIRS): Define.
2984 (AM_MAKEFLAGS): Likewise.
2985 (Multilib support.): Add section.
2986 * Makefile.in: Rebuilt.
2987 * ltconfig (extra_compiler_flags, extra_compiler_flags_value):
2988 New variables. Set for gcc using -print-multi-lib. Export them
2990 (sparc64-*-linux-gnu*): Use libsuff 64 for search paths.
2991 * ltmain.sh (B|b|V): Don't throw away gcc's -B, -b and -V options
2993 (extra_compiler_flags_value, extra_compiler_flags): Check these
2994 for extra compiler options which need to be passed down in
2997 2000-04-16 Anthony Green <green@redhat.com>
2999 * configure: Rebuilt.
3000 * configure.in: Change i*86-pc-linux* to i*86-*-linux*.
3002 2000-04-14 Jakub Jelinek <jakub@redhat.com>
3004 * include/ffi.h.in (SPARC64): Define for 64bit SPARC builds.
3005 Set SPARC FFI_DEFAULT_ABI based on SPARC64 define.
3006 * src/sparc/ffi.c (ffi_prep_args_v8): Renamed from ffi_prep_args.
3007 Replace all void * sizeofs with sizeof(int).
3008 Only compare type with FFI_TYPE_LONGDOUBLE if LONGDOUBLE is
3009 different than DOUBLE.
3010 Remove FFI_TYPE_SINT32 and FFI_TYPE_UINT32 cases (handled elsewhere).
3011 (ffi_prep_args_v9): New function.
3012 (ffi_prep_cif_machdep): Handle V9 ABI and long long on V8.
3013 (ffi_V9_return_struct): New function.
3014 (ffi_call): Handle FFI_V9 ABI from 64bit code and FFI_V8 ABI from
3015 32bit code (not yet cross-arch calls).
3016 * src/sparc/v8.S: Add struct return delay nop.
3018 * src/sparc/v9.S: New file.
3019 * src/prep_cif.c (ffi_prep_cif): Return structure pointer
3020 is used on sparc64 only for structures larger than 32 bytes.
3021 Pass by reference for structures is done for structure arguments
3022 larger than 16 bytes.
3023 * src/ffitest.c (main): Use 64bit rint on sparc64.
3024 Run long long tests on sparc.
3025 * src/types.c (FFI_TYPE_POINTER): Pointer is 64bit on alpha and
3027 (FFI_TYPE_LONGDOUBLE): long double is 128 bit aligned to 128 bits
3029 * configure.in (sparc-*-linux*): New supported target.
3030 (sparc64-*-linux*): Likewise.
3031 * configure: Rebuilt.
3032 * Makefile.am: Add v9.S to SPARC files.
3033 * Makefile.in: Likewise.
3034 (LINK): Surround $(CCLD) into double quotes, so that multilib
3035 compiles work correctly.
3037 2000-04-04 Alexandre Petit-Bianco <apbianco@cygnus.com>
3039 * configure: Rebuilt.
3040 * configure.in: (i*86-*-solaris*): New libffi target. Patch
3041 proposed by Bryce McKinlay.
3043 2000-03-20 Tom Tromey <tromey@cygnus.com>
3045 * Makefile.in: Hand edit for java_raw_api.lo.
3047 2000-03-08 Bryce McKinlay <bryce@albatross.co.nz>
3049 * config.guess, config.sub: Update from the gcc tree.
3050 Fix for PR libgcj/168.
3052 2000-03-03 Tom Tromey <tromey@cygnus.com>
3054 * Makefile.in: Fixed ia64 by hand.
3056 * configure: Rebuilt.
3057 * configure.in (--enable-multilib): New option.
3058 (libffi_basedir): New subst.
3059 (AC_OUTPUT): Added multilib code.
3061 2000-03-02 Tom Tromey <tromey@cygnus.com>
3063 * Makefile.in: Rebuilt.
3064 * Makefile.am (TARGET_SRC_IA64): Use `ia64', not `alpha', as
3067 2000-02-25 Hans Boehm <boehm@acm.org>
3069 * src/ia64/ffi.c, src/ia64/ia64_flags.h, src/ia64/unix.S: New
3071 * src/raw_api.c (ffi_translate_args): Fixed typo in argument
3073 (ffi_prep_raw_closure): Use ffi_translate_args, not
3074 ffi_closure_translate.
3075 * src/java_raw_api.c: New file.
3076 * src/ffitest.c (closure_test_fn): New function.
3077 (main): Define `rint' as long long on IA64. Added new test when
3078 FFI_CLOSURES is defined.
3079 * include/ffi.h.in (ALIGN): Use size_t, not unsigned.
3080 (ffi_abi): Recognize IA64.
3081 (ffi_raw): Added `flt' field.
3082 Added "Java raw API" code.
3083 * configure.in: Recognize ia64.
3084 * Makefile.am (TARGET_SRC_IA64): New macro.
3085 (libffi_la_common_SOURCES): Added java_raw_api.c.
3086 (libffi_la_SOURCES): Define in IA64 case.
3088 2000-01-04 Tom Tromey <tromey@cygnus.com>
3090 * Makefile.in: Rebuilt with newer automake.
3092 1999-12-31 Tom Tromey <tromey@cygnus.com>
3094 * Makefile.am (INCLUDES): Added -I$(top_srcdir)/src.
3096 1999-09-01 Tom Tromey <tromey@cygnus.com>
3098 * include/ffi.h.in: Removed PACKAGE and VERSION defines and
3100 * fficonfig.h.in: Rebuilt.
3101 * configure: Rebuilt.
3102 * configure.in: Pass 3rd argument to AM_INIT_AUTOMAKE.
3103 Use AM_PROG_LIBTOOL (automake 1.4 compatibility).
3104 * acconfig.h: Don't #undef PACKAGE or VERSION.
3106 1999-08-09 Anthony Green <green@cygnus.com>
3108 * include/ffi.h.in: Try to work around messy header problem
3109 with PACKAGE and VERSION.
3111 * configure: Rebuilt.
3112 * configure.in: Change version to 2.00-beta.
3114 * fficonfig.h.in: Rebuilt.
3115 * acconfig.h (FFI_NO_STRUCTS, FFI_NO_RAW_API): Define.
3117 * src/x86/ffi.c (ffi_raw_call): Rename.
3119 1999-08-02 Kresten Krab Thorup <krab@dominiq.is.s.u-tokyo.ac.jp>
3121 * src/x86/ffi.c (ffi_closure_SYSV): New function.
3122 (ffi_prep_incoming_args_SYSV): Ditto.
3123 (ffi_prep_closure): Ditto.
3124 (ffi_closure_raw_SYSV): Ditto.
3125 (ffi_prep_raw_closure): More ditto.
3126 (ffi_call_raw): Final ditto.
3128 * include/ffi.h.in: Add definitions for closure and raw API.
3130 * src/x86/ffi.c (ffi_prep_cif_machdep): Added case for
3133 * Makefile.am (libffi_la_common_SOURCES): Added raw_api.c
3135 * src/raw_api.c: New file.
3137 * include/ffi.h.in (ffi_raw): New type.
3138 (UINT_ARG, SINT_ARG): New defines.
3139 (ffi_closure, ffi_raw_closure): New types.
3140 (ffi_prep_closure, ffi_prep_raw_closure): New declarations.
3142 * configure.in: Add check for endianness and sizeof void*.
3144 * src/x86/sysv.S (ffi_call_SYSV): Call fixup routine via argument,
3145 instead of directly.
3147 * configure: Rebuilt.
3149 Thu Jul 8 14:28:42 1999 Anthony Green <green@cygnus.com>
3151 * configure.in: Add x86 and powerpc BeOS configurations.
3152 From Makoto Kato <m_kato@ga2.so-net.ne.jp>.
3154 1999-05-09 Anthony Green <green@cygnus.com>
3156 * configure.in: Add warning about this being beta code.
3157 Remove src/Makefile.am from the picture.
3158 * configure: Rebuilt.
3160 * Makefile.am: Move logic from src/Makefile.am. Add changes
3161 to support libffi as a target library.
3162 * Makefile.in: Rebuilt.
3164 * aclocal.m4, config.guess, config.sub, ltconfig, ltmain.sh:
3165 Upgraded to new autoconf, automake, libtool.
3169 * LICENSE: Update copyright date.
3171 * src/Makefile.am, src/Makefile.in: Removed.
3173 1998-11-29 Anthony Green <green@cygnus.com>
3175 * include/ChangeLog: Removed.
3176 * src/ChangeLog: Removed.
3177 * src/mips/ChangeLog: Removed.
3178 * src/sparc/ChangeLog: Remboved.
3179 * src/x86/ChangeLog: Removed.
3181 * ChangeLog.v1: Created.