]> git.ipfire.org Git - thirdparty/gcc.git/blame - libffi/ChangeLog
projects / thirdparty / gcc.git / blame
? search:
summary | shortlog | log | commit | commitdiff | tree
blob | blame (incremental) | history | HEAD
PR libffi/28313
[thirdparty/gcc.git] / libffi / ChangeLog
CommitLineData
edd57a0a 12007-08-10 David Daney <ddaney@avtrex.com>
2
3 PR libffi/28313
4 * configure.ac: Don't treat mips64 as a special case.
5 * Makefile.am (nodist_libffi_la_SOURCES): Add n32.S.
6 * configure: Regenerate
7 * Makefile.in: Ditto.
8 * fficonfig.h.in: Ditto.
9 * src/mips/ffitarget.h (REG_L, REG_S, SUBU, ADDU, SRL, LI): Indent.
10 (LA, EH_FRAME_ALIGN, FDE_ADDR_BYTES): New preprocessor macros.
11 (FFI_DEFAULT_ABI): Set for n64 case.
12 (FFI_CLOSURES, FFI_TRAMPOLINE_SIZE): Define for n32 and n64 cases.
13 * src/mips/n32.S (ffi_call_N32): Add debug macros and labels for FDE.
14 (ffi_closure_N32): New function.
15 (.eh_frame): New section
16 * src/mips/o32.S: Clean up comments.
17 (ffi_closure_O32): Pass ffi_closure parameter in $12.
18 * src/mips/ffi.c: Use FFI_MIPS_N32 instead of
19 _MIPS_SIM == _ABIN32 throughout.
20 (FFI_MIPS_STOP_HERE): New, use in place of
21 ffi_stop_here.
22 (ffi_prep_args): Use unsigned long to hold pointer values. Rewrite
23 to support n32/n64 ABIs.
24 (calc_n32_struct_flags): Rewrite.
25 (calc_n32_return_struct_flags): Remove unused variable. Reverse
26 position of flag bits.
27 (ffi_prep_cif_machdep): Rewrite n32 portion.
28 (ffi_call): Enable for n64. Add special handling for small structure
29 return values.
30 (ffi_prep_closure_loc): Add n32 and n64 support.
31 (ffi_closure_mips_inner_O32): Add cast to silence warning.
32 (copy_struct_N32, ffi_closure_mips_inner_N32): New functions.
33
3471c5eb 342007-08-08 David Daney <ddaney@avtrex.com>
35
36 * testsuite/libffi.call/ffitest.h (ffi_type_mylong): Remove definition.
37 * testsuite/libffi.call/cls_align_uint16.c (main): Use correct type
38 specifiers.
39 * testsuite/libffi.call/nested_struct1.c (main): Ditto.
40 * testsuite/libffi.call/cls_sint.c (main): Ditto.
41 * testsuite/libffi.call/nested_struct9.c (main): Ditto.
42 * testsuite/libffi.call/cls_20byte1.c (main): Ditto.
43 * testsuite/libffi.call/cls_9byte1.c (main): Ditto.
44 * testsuite/libffi.call/closure_fn1.c (main): Ditto.
45 * testsuite/libffi.call/closure_fn3.c (main): Ditto.
46 * testsuite/libffi.call/return_dbl2.c (main): Ditto.
47 * testsuite/libffi.call/cls_sshort.c (main): Ditto.
48 * testsuite/libffi.call/return_fl3.c (main): Ditto.
49 * testsuite/libffi.call/closure_fn5.c (main): Ditto.
50 * testsuite/libffi.call/nested_struct.c (main): Ditto.
51 * testsuite/libffi.call/nested_struct10.c (main): Ditto.
52 * testsuite/libffi.call/return_ll1.c (main): Ditto.
53 * testsuite/libffi.call/cls_8byte.c (main): Ditto.
54 * testsuite/libffi.call/cls_align_uint32.c (main): Ditto.
55 * testsuite/libffi.call/cls_align_sint16.c (main): Ditto.
56 * testsuite/libffi.call/cls_20byte.c (main): Ditto.
57 * testsuite/libffi.call/nested_struct2.c (main): Ditto.
58 * testsuite/libffi.call/cls_24byte.c (main): Ditto.
59 * testsuite/libffi.call/nested_struct6.c (main): Ditto.
60 * testsuite/libffi.call/cls_uint.c (main): Ditto.
61 * testsuite/libffi.call/cls_12byte.c (main): Ditto.
62 * testsuite/libffi.call/cls_16byte.c (main): Ditto.
63 * testsuite/libffi.call/closure_fn0.c (main): Ditto.
64 * testsuite/libffi.call/cls_9byte2.c (main): Ditto.
65 * testsuite/libffi.call/closure_fn2.c (main): Ditto.
66 * testsuite/libffi.call/return_dbl1.c (main): Ditto.
67 * testsuite/libffi.call/closure_fn4.c (main): Ditto.
68 * testsuite/libffi.call/closure_fn6.c (main): Ditto.
69 * testsuite/libffi.call/cls_align_sint32.c (main): Ditto.
70
ad7bf8e7 712007-08-07 Andrew Haley <aph@redhat.com>
72
73 * src/x86/sysv.S (ffi_closure_raw_SYSV): Fix typo in previous
74 checkin.
75
6cc1286e 762007-08-06 Andrew Haley <aph@redhat.com>
77
78 PR testsuite/32843
79 * src/x86/sysv.S (ffi_closure_raw_SYSV): Handle FFI_TYPE_UINT8,
80 FFI_TYPE_SINT8, FFI_TYPE_UINT16, FFI_TYPE_SINT16, FFI_TYPE_UINT32,
81 FFI_TYPE_SINT32.
82
c6f5483d 832007-08-02 David Daney <ddaney@avtrex.com>
84
85 * testsuite/libffi.call/return_ul.c (main): Define return type as
86 ffi_arg. Use proper printf conversion specifier.
87
f6179f0e 882007-07-30 Andrew Haley <aph@redhat.com>
89
90 PR testsuite/32843
91 * src/x86/ffi.c (ffi_prep_cif_machdep): in x86 case, add code for
92 signed/unsigned int8/16.
93 * src/x86/sysv.S (ffi_call_SYSV): Rewrite to:
94 Use a jump table.
95 Remove code to pop args from the stack after call.
96 Special-case signed/unsigned int8/16.
97 * testsuite/libffi.call/return_sc.c (main): Revert.
98
7babed79 992007-07-26 Richard Guenther <rguenther@suse.de>
100
101 PR testsuite/32843
102 * testsuite/libffi.call/return_sc.c (main): Verify call
103 result as signed char, not ffi_arg.
104
f54a4242 1052007-07-16 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
106
107 * configure.ac (i?86-*-solaris2.1[0-9]): Set TARGET to X86_64.
108 * configure: Regenerate.
109
e96a69da 1102007-07-11 David Daney <ddaney@avtrex.com>
111
112 * src/mips/ffi.c: Don't include sys/cachectl.h.
113 (ffi_prep_closure_loc): Use __builtin___clear_cache() instead of
114 cacheflush().
115
187f6543 1162007-05-18 Aurelien Jarno <aurelien@aurel32.net>
117
118 * src/arm/ffi.c (ffi_prep_closure_loc): Renamed and ajusted
119 from (ffi_prep_closure): ... this.
120 (FFI_INIT_TRAMPOLINE): Adjust.
121
1222005-12-31 Phil Blundell <pb@reciva.com>
123
124 * src/arm/ffi.c (ffi_prep_incoming_args_SYSV,
125 ffi_closure_SYSV_inner, ffi_prep_closure): New, add closure support.
126 * src/arm/sysv.S(ffi_closure_SYSV): Likewise.
127 * src/arm/ffitarget.h (FFI_TRAMPOLINE_SIZE): Likewise.
128 (FFI_CLOSURES): Enable closure support.
129
1302007-07-03 Andrew Haley <aph@hedges.billgatliff.com>
131
132 * testsuite/libffi.call/cls_multi_ushort.c,
133 testsuite/libffi.call/cls_align_uint16.c,
134 testsuite/libffi.call/nested_struct1.c,
135 testsuite/libffi.call/nested_struct3.c,
136 testsuite/libffi.call/cls_7_1_byte.c,
137 testsuite/libffi.call/cls_double.c,
138 testsuite/libffi.call/nested_struct5.c,
139 testsuite/libffi.call/nested_struct7.c,
140 testsuite/libffi.call/cls_sint.c,
141 testsuite/libffi.call/nested_struct9.c,
142 testsuite/libffi.call/cls_20byte1.c,
143 testsuite/libffi.call/cls_multi_sshortchar.c,
144 testsuite/libffi.call/cls_align_sint64.c,
145 testsuite/libffi.call/cls_3byte2.c,
146 testsuite/libffi.call/cls_multi_schar.c,
147 testsuite/libffi.call/cls_multi_uchar.c,
148 testsuite/libffi.call/cls_19byte.c,
149 testsuite/libffi.call/cls_9byte1.c,
150 testsuite/libffi.call/cls_align_float.c,
151 testsuite/libffi.call/closure_fn1.c,
152 testsuite/libffi.call/problem1.c,
153 testsuite/libffi.call/closure_fn3.c,
154 testsuite/libffi.call/cls_sshort.c,
155 testsuite/libffi.call/closure_fn5.c,
156 testsuite/libffi.call/cls_align_double.c,
157 testsuite/libffi.call/cls_2byte.c,
158 testsuite/libffi.call/nested_struct.c,
159 testsuite/libffi.call/nested_struct10.c,
160 testsuite/libffi.call/cls_4byte.c,
161 testsuite/libffi.call/cls_6byte.c,
162 testsuite/libffi.call/cls_8byte.c,
163 testsuite/libffi.call/cls_multi_sshort.c,
164 testsuite/libffi.call/cls_align_uint32.c,
165 testsuite/libffi.call/cls_align_sint16.c,
166 testsuite/libffi.call/cls_float.c,
167 testsuite/libffi.call/cls_20byte.c,
168 testsuite/libffi.call/cls_5_1_byte.c,
169 testsuite/libffi.call/nested_struct2.c,
170 testsuite/libffi.call/cls_24byte.c,
171 testsuite/libffi.call/nested_struct4.c,
172 testsuite/libffi.call/nested_struct6.c,
173 testsuite/libffi.call/cls_64byte.c,
174 testsuite/libffi.call/nested_struct8.c,
175 testsuite/libffi.call/cls_uint.c,
176 testsuite/libffi.call/cls_multi_ushortchar.c,
177 testsuite/libffi.call/cls_schar.c,
178 testsuite/libffi.call/cls_uchar.c,
179 testsuite/libffi.call/cls_align_uint64.c,
180 testsuite/libffi.call/cls_ulonglong.c,
181 testsuite/libffi.call/cls_align_longdouble.c,
182 testsuite/libffi.call/cls_1_1byte.c,
183 testsuite/libffi.call/cls_12byte.c,
184 testsuite/libffi.call/cls_3_1byte.c,
185 testsuite/libffi.call/cls_3byte1.c,
186 testsuite/libffi.call/cls_4_1byte.c,
187 testsuite/libffi.call/cls_6_1_byte.c,
188 testsuite/libffi.call/cls_16byte.c,
189 testsuite/libffi.call/cls_18byte.c,
190 testsuite/libffi.call/closure_fn0.c,
191 testsuite/libffi.call/cls_9byte2.c,
192 testsuite/libffi.call/closure_fn2.c,
193 testsuite/libffi.call/closure_fn4.c,
194 testsuite/libffi.call/cls_ushort.c,
195 testsuite/libffi.call/closure_fn6.c,
196 testsuite/libffi.call/cls_5byte.c,
197 testsuite/libffi.call/cls_align_pointer.c,
198 testsuite/libffi.call/cls_7byte.c,
199 testsuite/libffi.call/cls_align_sint32.c,
200 testsuite/libffi.special/unwindtest_ffi_call.cc,
201 testsuite/libffi.special/unwindtest.cc: Enable for ARM.
202
cecc75a9 2032007-07-05 H.J. Lu <hongjiu.lu@intel.com>
204
205 * aclocal.m4: Regenerated.
206
07c46636 2072007-06-02 Paolo Bonzini <bonzini@gnu.org>
208
209 * configure: Regenerate.
210
542f9784 2112007-05-23 Steve Ellcey <sje@cup.hp.com>
212
213 * Makefile.in: Regenerate.
214 * configure: Regenerate.
215 * aclocal.m4: Regenerate.
216 * include/Makefile.in: Regenerate.
217 * testsuite/Makefile.in: Regenerate.
218
0e3a3bdb 2192007-05-10 Roman Zippel <zippel@linux-m68k.org>
220
221 * src/m68k/ffi.c (ffi_prep_incoming_args_SYSV,
222 ffi_closure_SYSV_inner,ffi_prep_closure): New, add closure support.
223 * src/m68k/sysv.S(ffi_closure_SYSV,ffi_closure_struct_SYSV): Likewise.
224 * src/m68k/ffitarget.h (FFI_TRAMPOLINE_SIZE): Likewise.
225 (FFI_CLOSURES): Enable closure support.
226
a53add80 2272007-05-10 Roman Zippel <zippel@linux-m68k.org>
228
229 * configure.ac (HAVE_AS_CFI_PSEUDO_OP): New test.
230 * configure: Regenerate.
231 * fficonfig.h.in: Regenerate.
232 * src/m68k/sysv.S (CFI_STARTPROC,CFI_ENDPROC,
233 CFI_OFFSET,CFI_DEF_CFA): New macros.
234 (ffi_call_SYSV): Add callframe annotation.
235
bb3218e9 2362007-05-10 Roman Zippel <zippel@linux-m68k.org>
237
238 * src/m68k/ffi.c (ffi_prep_args,ffi_prep_cif_machdep): Fix
239 numerous test suite failures.
240 * src/m68k/sysv.S (ffi_call_SYSV): Likewise.
241
5d0df87f 2422007-04-11 Paolo Bonzini <bonzini@gnu.org>
243
244 * Makefile.am (EXTRA_DIST): Bring up to date.
245 * Makefile.in: Regenerate.
246 * src/frv/eabi.S: Remove RCS keyword.
247
ea4d299a 2482007-04-06 Richard Henderson <rth@redhat.com>
249
250 * configure.ac: Tidy target case.
251 (HAVE_LONG_DOUBLE): Allow the target to override.
252 * configure: Regenerate.
253 * include/ffi.h.in: Don't define ffi_type_foo if
254 LIBFFI_HIDE_BASIC_TYPES is defined.
255 (ffi_type_longdouble): If not HAVE_LONG_DOUBLE, define
256 to ffi_type_double.
257 * types.c (LIBFFI_HIDE_BASIC_TYPES): Define.
258 (FFI_TYPEDEF, ffi_type_void): Mark the data const.
259 (ffi_type_longdouble): Special case for Alpha. Don't define
260 if long double == double.
261
262 * src/alpha/ffi.c (FFI_TYPE_LONGDOUBLE): Assert unique value.
263 (ffi_prep_cif_machdep): Handle it as the 128-bit type.
264 (ffi_call, ffi_closure_osf_inner): Likewise.
265 (ffi_closure_osf_inner): Likewise. Mark hidden.
266 (ffi_call_osf, ffi_closure_osf): Mark hidden.
267 * src/alpha/ffitarget.h (FFI_LAST_ABI): Tidy definition.
268 * src/alpha/osf.S (ffi_call_osf, ffi_closure_osf): Mark hidden.
269 (load_table): Handle 128-bit long double.
270
271 * testsuite/libffi.call/float4.c: Add -mieee for alpha.
272
770e494d 2732007-04-06 Tom Tromey <tromey@redhat.com>
274
275 PR libffi/31491:
276 * README: Fixed bug in example.
277
58606959 2782007-04-03 Jakub Jelinek <jakub@redhat.com>
279
280 * src/closures.c: Include sys/statfs.h.
281 (_GNU_SOURCE): Define on Linux.
282 (FFI_MMAP_EXEC_SELINUX): Define.
283 (selinux_enabled): New variable.
284 (selinux_enabled_check): New function.
285 (is_selinux_enabled): Define.
286 (dlmmap): Use it.
287
7fed151e 2882007-03-24 Uros Bizjak <ubizjak@gmail.com>
289
290 * testsuite/libffi.call/return_fl2.c (return_fl): Mark as static.
291 Use 'volatile float sum' to create sum of floats to avoid false
292 negative due to excess precision on ix86 targets.
293 (main): Ditto.
294
8fb4628d 2952007-03-08 Alexandre Oliva <aoliva@redhat.com>
296
297 * src/powerpc/ffi.c (flush_icache): Fix left-over from previous
298 patch.
299 (ffi_prep_closure_loc): Remove unneeded casts. Add needed ones.
300
f20fe604 3012007-03-07 Alexandre Oliva <aoliva@redhat.com>
302
303 * include/ffi.h.in (ffi_closure_alloc, ffi_closure_free): New.
304 (ffi_prep_closure_loc): New.
305 (ffi_prep_raw_closure_loc): New.
306 (ffi_prep_java_raw_closure_loc): New.
307 * src/closures.c: New file.
308 * src/dlmalloc.c [FFI_MMAP_EXEC_WRIT] (struct malloc_segment):
309 Replace sflags with exec_offset.
310 [FFI_MMAP_EXEC_WRIT] (mmap_exec_offset, add_segment_exec_offset,
311 sub_segment_exec_offset): New macros.
312 (get_segment_flags, set_segment_flags, check_segment_merge): New
313 macros.
314 (is_mmapped_segment, is_extern_segment): Use get_segment_flags.
315 (add_segment, sys_alloc, create_mspace, create_mspace_with_base,
316 destroy_mspace): Use new macros.
317 (sys_alloc): Silence warning.
318 * Makefile.am (libffi_la_SOURCES): Add src/closures.c.
319 * Makefile.in: Rebuilt.
320 * src/prep_cif [FFI_CLOSURES] (ffi_prep_closure): Implement in
321 terms of ffi_prep_closure_loc.
322 * src/raw_api.c (ffi_prep_raw_closure_loc): Renamed and adjusted
323 from...
324 (ffi_prep_raw_closure): ... this. Re-implement in terms of the
325 renamed version.
326 * src/java_raw_api (ffi_prep_java_raw_closure_loc): Renamed and
327 adjusted from...
328 (ffi_prep_java_raw_closure): ... this. Re-implement in terms of
329 the renamed version.
330 * src/alpha/ffi.c (ffi_prep_closure_loc): Renamed from
331 (ffi_prep_closure): ... this.
332 * src/pa/ffi.c: Likewise.
333 * src/cris/ffi.c: Likewise. Adjust.
334 * src/frv/ffi.c: Likewise.
335 * src/ia64/ffi.c: Likewise.
336 * src/mips/ffi.c: Likewise.
337 * src/powerpc/ffi_darwin.c: Likewise.
338 * src/s390/ffi.c: Likewise.
339 * src/sh/ffi.c: Likewise.
340 * src/sh64/ffi.c: Likewise.
341 * src/sparc/ffi.c: Likewise.
342 * src/x86/ffi64.c: Likewise.
343 * src/x86/ffi.c: Likewise.
344 (FFI_INIT_TRAMPOLINE): Adjust.
345 (ffi_prep_raw_closure_loc): Renamed and adjusted from...
346 (ffi_prep_raw_closure): ... this.
347 * src/powerpc/ffi.c (ffi_prep_closure_loc): Renamed from
348 (ffi_prep_closure): ... this.
349 (flush_icache): Adjust.
350
c3265eac 3512007-03-07 Alexandre Oliva <aoliva@redhat.com>
352
353 * src/dlmalloc.c: New file, imported version 2.8.3 of Doug
354 Lea's malloc.
355
0ce4a5ce 3562007-03-01 Brooks Moses <brooks.moses@codesourcery.com>
357
358 * Makefile.am: Add dummy install-pdf target.
359 * Makefile.in: Regenerate
360
45444175 3612007-02-13 Andreas Krebbel <krebbel1@de.ibm.com>
362
363 * src/s390/ffi.c (ffi_prep_args, ffi_prep_cif_machdep,
364 ffi_closure_helper_SYSV): Add long double handling.
365
1df875cc 3662007-02-02 Jakub Jelinek <jakub@redhat.com>
367
368 * src/powerpc/linux64.S (ffi_call_LINUX64): Move restore of r2
369 immediately after bctrl instruction.
370
27c4194f 3712007-01-18 Alexandre Oliva <aoliva@redhat.com>
372
373 * Makefile.am (all-recursive, install-recursive,
374 mostlyclean-recursive, clean-recursive, distclean-recursive,
375 maintainer-clean-recursive): Add missing targets.
376 * Makefile.in: Rebuilt.
377
eb6316a2 3782006-12-14 Andreas Tobler <a.tobler@schweiz.org>
379
380 * configure.ac: Add TARGET for x86_64-*-darwin*.
381 * Makefile.am (nodist_libffi_la_SOURCES): Add rules for 64-bit sources
382 for X86_DARWIN.
383 * src/x86/ffitarget.h: Set trampoline size for x86_64-*-darwin*.
384 * src/x86/darwin64.S: New file for x86_64-*-darwin* support.
385 * configure: Regenerate.
386 * Makefile.in: Regenerate.
387 * include/Makefile.in: Regenerate.
388 * testsuite/Makefile.in: Regenerate.
389 * testsuite/libffi.special/unwindtest_ffi_call.cc: New test case for
390 ffi_call only.
391
75a99c5e 3922006-12-13 Andreas Tobler <a.tobler@schweiz.org>
393
394 * aclocal.m4: Regenerate with aclocal -I .. as written in the
395 Makefile.am.
396
a39e926c 3972006-10-31 Geoffrey Keating <geoffk@apple.com>
c8dfb8ef 398
399 * src/powerpc/ffi_darwin.c (darwin_adjust_aggregate_sizes): New.
400 (ffi_prep_cif_machdep): Call darwin_adjust_aggregate_sizes for
401 Darwin.
402 * testsuite/libffi.call/nested_struct4.c: Remove Darwin XFAIL.
403 * testsuite/libffi.call/nested_struct6.c: Remove Darwin XFAIL.
404
eb819ecd 4052006-10-10 Paolo Bonzini <bonzini@gnu.org>
406 Sandro Tolaini <tolaini@libero.it>
407
408 * configure.ac [i*86-*-darwin*]: Set X86_DARWIN symbol and
409 conditional.
410 * configure: Regenerated.
411 * Makefile.am (nodist_libffi_la_SOURCES) [X86_DARWIN]: New case.
412 (EXTRA_DIST): Add src/x86/darwin.S.
413 * Makefile.in: Regenerated.
414 * include/Makefile.in: Regenerated.
415 * testsuite/Makefile.in: Regenerated.
416
417 * src/x86/ffi.c (ffi_prep_cif_machdep) [X86_DARWIN]: Treat like
418 X86_WIN32, and additionally align stack to 16 bytes.
419 * src/x86/darwin.S: New, based on sysv.S.
420 * src/prep_cif.c (ffi_prep_cif) [X86_DARWIN]: Align > 8-byte structs.
421
23d9d417 4222006-09-12 David Daney <ddaney@avtrex.com>
423
424 PR libffi/23935
425 * include/Makefile.am: Install both ffi.h and ffitarget.h in
426 $(libdir)/gcc/$(target_alias)/$(gcc_version)/include.
427 * aclocal.m4: Regenerated for automake 1.9.6.
428 * Makefile.in: Regenerated.
429 * include/Makefile.in: Regenerated.
430 * testsuite/Makefile.in: Regenerated.
431
ec045626 4322006-08-17 Andreas Tobler <a.tobler@schweiz.ch>
433
434 * include/ffi_common.h (struct): Revert accidental commit.
435
b9463909 4362006-08-15 Andreas Tobler <a.tobler@schweiz.ch>
437
438 * include/ffi_common.h: Remove lint directives.
439 * include/ffi.h.in: Likewise.
440
c695c024 4412006-07-25 Torsten Schoenfeld <kaffeetisch@gmx.de>
442
443 * include/ffi.h.in (ffi_type_ulong, ffi_type_slong): Define correctly
444 for 32-bit architectures.
445 * testsuite/libffi.call/return_ul.c: New test case.
446
3a9d2a82 4472006-07-19 David Daney <ddaney@avtrex.com>
448
449 * testsuite/libffi.call/closure_fn6.c: Remove xfail for mips,
450 xfail remains for mips64.
451
8f16e3ad 4522006-05-23 Carlos O'Donell <carlos@codesourcery.com>
453
454 * Makefile.am: Add install-html target. Add install-html to .PHONY
455 * Makefile.in: Regenerate.
456 * aclocal.m4: Regenerate.
457 * include/Makefile.in: Regenerate.
458 * testsuite/Makefile.in: Regenerate.
459
db951627 4602006-05-18 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
461
462 * pa/ffi.c (ffi_prep_args_pa32): Load floating point arguments from
463 stack slot.
464
9a333c86 4652006-04-22 Andreas Tobler <a.tobler@schweiz.ch>
466
467 * README: Remove notice about 'Crazy Comments'.
468 * src/debug.c: Remove lint directives. Cleanup white spaces.
469 * src/java_raw_api.c: Likewise.
470 * src/prep_cif.c: Likewise.
471 * src/raw_api.c: Likewise.
472 * src/ffitest.c: Delete. No longer needed, all test cases migrated
473 to the testsuite.
474 * src/arm/ffi.c: Remove lint directives.
475 * src/m32r/ffi.c: Likewise.
476 * src/pa/ffi.c: Likewise.
477 * src/powerpc/ffi.c: Likewise.
478 * src/powerpc/ffi_darwin.c: Likewise.
479 * src/sh/ffi.c: Likewise.
480 * src/sh64/ffi.c: Likewise.
481 * src/x86/ffi.c: Likewise.
482 * testsuite/libffi.call/float2.c: Likewise.
483 * testsuite/libffi.call/promotion.c: Likewise.
484 * testsuite/libffi.call/struct1.c: Likewise.
485
7610fe2b 4862006-04-13 Andreas Tobler <a.tobler@schweiz.ch>
487
488 * src/pa/hpux32.S: Correct unwind offset calculation for
489 ffi_closure_pa32.
490 * src/pa/linux.S: Likewise.
491
5adc6320 4922006-04-12 James E Wilson <wilson@specifix.com>
493
494 PR libgcj/26483
495 * src/ia64/ffi.c (stf_spill, ldf_fill): Rewrite as macros.
496 (hfa_type_load): Call stf_spill.
497 (hfa_type_store): Call ldf_fill.
498 (ffi_call): Adjust calls to above routines. Add local temps for
499 macro result.
7610fe2b 500
d35f63a9 5012006-04-10 Matthias Klose <doko@debian.org>
502
503 * testsuite/lib/libffi-dg.exp (libffi-init): Recognize multilib
504 directory names containing underscores.
505
2d3dcfaa 5062006-04-07 James E Wilson <wilson@specifix.com>
507
508 * testsuite/libffi.call/float4.c: New testcase.
509
1a82d6f6 5102006-04-05 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
511 Andreas Tobler <a.tobler@schweiz.ch>
512
513 * Makefile.am: Add PA_HPUX port.
514 * Makefile.in: Regenerate.
515 * include/Makefile.in: Likewise.
516 * testsuite/Makefile.in: Likewise.
517 * configure.ac: Add PA_HPUX rules.
518 * configure: Regenerate.
519 * src/pa/ffitarget.h: Rename linux target to PA_LINUX.
520 Add PA_HPUX and PA64_HPUX.
521 Rename FFI_LINUX ABI to FFI_PA32 ABI.
522 (FFI_TRAMPOLINE_SIZE): Define for 32-bit HP-UX targets.
523 (FFI_TYPE_SMALL_STRUCT2): Define.
524 (FFI_TYPE_SMALL_STRUCT4): Likewise.
525 (FFI_TYPE_SMALL_STRUCT8): Likewise.
526 (FFI_TYPE_SMALL_STRUCT3): Redefine.
527 (FFI_TYPE_SMALL_STRUCT5): Likewise.
528 (FFI_TYPE_SMALL_STRUCT6): Likewise.
529 (FFI_TYPE_SMALL_STRUCT7): Likewise.
530 * src/pa/ffi.c (ROUND_DOWN): Delete.
531 (fldw, fstw, fldd, fstd): Use '__asm__'.
532 (ffi_struct_type): Add support for FFI_TYPE_SMALL_STRUCT2,
533 FFI_TYPE_SMALL_STRUCT4 and FFI_TYPE_SMALL_STRUCT8.
534 (ffi_prep_args_LINUX): Rename to ffi_prep_args_pa32. Update comment.
535 Simplify incrementing of stack slot variable. Change type of local
536 'n' to unsigned int.
537 (ffi_size_stack_LINUX): Rename to ffi_size_stack_pa32. Handle long
538 double on PA_HPUX.
539 (ffi_prep_cif_machdep): Likewise.
540 (ffi_call): Likewise.
541 (ffi_closure_inner_LINUX): Rename to ffi_closure_inner_pa32. Change
542 return type to ffi_status. Simplify incrementing of stack slot
543 variable. Only copy floating point argument registers when PA_LINUX
544 is true. Reformat debug statement.
545 Add support for FFI_TYPE_SMALL_STRUCT2, FFI_TYPE_SMALL_STRUCT4 and
546 FFI_TYPE_SMALL_STRUCT8.
547 (ffi_closure_LINUX): Rename to ffi_closure_pa32. Add 'extern' to
548 declaration.
549 (ffi_prep_closure): Make linux trampoline conditional on PA_LINUX.
550 Add nops to cache flush. Add trampoline for PA_HPUX.
551 * src/pa/hpux32.S: New file.
552 * src/pa/linux.S (ffi_call_LINUX): Rename to ffi_call_pa32. Rename
553 ffi_prep_args_LINUX to ffi_prep_args_pa32.
554 Localize labels. Add support for 2, 4 and 8-byte small structs. Handle
555 unaligned destinations in 3, 5, 6 and 7-byte small structs. Order
556 argument type checks so that common argument types appear first.
557 (ffi_closure_LINUX): Rename to ffi_closure_pa32. Rename
558 ffi_closure_inner_LINUX to ffi_closure_inner_pa32.
559
7f5ddc97 5602006-03-24 Alan Modra <amodra@bigpond.net.au>
561
562 * src/powerpc/ffitarget.h (enum ffi_abi): Add FFI_LINUX. Default
563 for 32-bit using IBM extended double format. Fix FFI_LAST_ABI.
564 * src/powerpc/ffi.c (ffi_prep_args_SYSV): Handle linux variant of
565 FFI_TYPE_LONGDOUBLE.
566 (ffi_prep_args64): Assert using IBM extended double.
567 (ffi_prep_cif_machdep): Don't munge FFI_TYPE_LONGDOUBLE type.
568 Handle FFI_LINUX FFI_TYPE_LONGDOUBLE return and args.
569 (ffi_call): Handle FFI_LINUX.
570 (ffi_closure_helper_SYSV): Non FFI_LINUX long double return needs
571 gpr3 return pointer as for struct return. Handle FFI_LINUX
572 FFI_TYPE_LONGDOUBLE return and args. Don't increment "nf"
573 unnecessarily.
574 * src/powerpc/ppc_closure.S (ffi_closure_SYSV): Load both f1 and f2
575 for FFI_TYPE_LONGDOUBLE. Move epilogue insns into case table.
576 Don't use r6 as pointer to results, instead use sp offset. Don't
577 make a special call to load lr with case table address, instead
578 use offset from previous call.
579 * src/powerpc/sysv.S (ffi_call_SYSV): Save long double return.
580 * src/powerpc/linux64.S (ffi_call_LINUX64): Simplify long double
581 return.
582
cd10a5a7 5832006-03-15 Kaz Kojima <kkojima@gcc.gnu.org>
584
585 * src/sh64/ffi.c (ffi_prep_cif_machdep): Handle float arguments
586 passed with FP registers correctly.
587 (ffi_closure_helper_SYSV): Likewise.
588 * src/sh64/sysv.S: Likewise.
589
a07533b6 5902006-03-01 Andreas Tobler <a.tobler@schweiz.ch>
591
592 * testsuite/libffi.special/unwindtest.cc (closure_test_fn): Mark cif,
593 args and userdata unused.
594 (closure_test_fn1): Mark cif and userdata unused.
595 (main): Remove unused res.
596
bdc6f488 5972006-02-28 Andreas Tobler <a.tobler@schweiz.ch>
598
599 * testsuite/libffi.call/call.exp: Adjust FSF address. Add test runs for
600 -O2, -O3, -Os and the warning flags -W -Wall.
601 * testsuite/libffi.special/special.exp: Likewise.
602 * testsuite/libffi.call/ffitest.h: Add an __UNUSED__ macro to mark
603 unused parameter unused for gcc or else do nothing.
604 * testsuite/libffi.special/ffitestcxx.h: Likewise.
605 * testsuite/libffi.call/cls_12byte.c (cls_struct_12byte_gn): Mark cif
606 and userdata unused.
607 * testsuite/libffi.call/cls_16byte.c (cls_struct_16byte_gn): Likewise.
608 * testsuite/libffi.call/cls_18byte.c (cls_struct_18byte_gn): Likewise.
609 * testsuite/libffi.call/cls_19byte.c (cls_struct_19byte_gn): Likewise.
610 * testsuite/libffi.call/cls_1_1byte.c (cls_struct_1_1byte_gn): Likewise.
611 * testsuite/libffi.call/cls_20byte.c (cls_struct_20byte_gn): Likewise.
612 * testsuite/libffi.call/cls_20byte1.c (cls_struct_20byte_gn): Likewise.
613 * testsuite/libffi.call/cls_24byte.c (cls_struct_24byte_gn): Likewise.
614 * testsuite/libffi.call/cls_2byte.c (cls_struct_2byte_gn): Likewise.
615 * testsuite/libffi.call/cls_3_1byte.c (cls_struct_3_1byte_gn): Likewise.
616 * testsuite/libffi.call/cls_3byte1.c (cls_struct_3byte_gn): Likewise.
617 * testsuite/libffi.call/cls_3byte2.c (cls_struct_3byte_gn1): Likewise.
618 * testsuite/libffi.call/cls_4_1byte.c (cls_struct_4_1byte_gn): Likewise.
619 * testsuite/libffi.call/cls_4byte.c (cls_struct_4byte_gn): Likewise.
620 * testsuite/libffi.call/cls_5_1_byte.c (cls_struct_5byte_gn): Likewise.
621 * testsuite/libffi.call/cls_5byte.c (cls_struct_5byte_gn): Likewise.
622 * testsuite/libffi.call/cls_64byte.c (cls_struct_64byte_gn): Likewise.
623 * testsuite/libffi.call/cls_6_1_byte.c (cls_struct_6byte_gn): Likewise.
624 * testsuite/libffi.call/cls_6byte.c (cls_struct_6byte_gn): Likewise.
625 * testsuite/libffi.call/cls_7_1_byte.c (cls_struct_7byte_gn): Likewise.
626 * testsuite/libffi.call/cls_7byte.c (cls_struct_7byte_gn): Likewise.
627 * testsuite/libffi.call/cls_8byte.c (cls_struct_8byte_gn): Likewise.
628 * testsuite/libffi.call/cls_9byte1.c (cls_struct_9byte_gn): Likewise.
629 * testsuite/libffi.call/cls_9byte2.c (cls_struct_9byte_gn): Likewise.
630 * testsuite/libffi.call/cls_align_double.c (cls_struct_align_gn):
631 Likewise.
632 * testsuite/libffi.call/cls_align_float.c (cls_struct_align_gn):
633 Likewise.
634 * testsuite/libffi.call/cls_align_longdouble.c (cls_struct_align_gn):
635 Likewise.
636 * testsuite/libffi.call/cls_align_pointer.c (cls_struct_align_fn): Cast
637 void* to avoid compiler warning.
638 (main): Likewise.
639 (cls_struct_align_gn): Mark cif and userdata unused.
640 * testsuite/libffi.call/cls_align_sint16.c (cls_struct_align_gn):
641 Likewise.
642 * testsuite/libffi.call/cls_align_sint32.c (cls_struct_align_gn):
643 Likewise.
644 * testsuite/libffi.call/cls_align_sint64.c (cls_struct_align_gn):
645 Likewise.
646 * testsuite/libffi.call/cls_align_uint16.c (cls_struct_align_gn):
647 Likewise.
648 * testsuite/libffi.call/cls_align_uint32.c (cls_struct_align_gn):
649 Likewise.
650 * testsuite/libffi.call/cls_double.c (cls_ret_double_fn): Likewise.
651 * testsuite/libffi.call/cls_float.c (cls_ret_float_fn): Likewise.
652 * testsuite/libffi.call/cls_multi_schar.c (test_func_gn): Mark cif and
653 data unused.
654 (main): Cast res_call to silence gcc.
655 * testsuite/libffi.call/cls_multi_sshort.c (test_func_gn): Mark cif and
656 data unused.
657 (main): Cast res_call to silence gcc.
658 * testsuite/libffi.call/cls_multi_sshortchar.c (test_func_gn): Mark cif
659 and data unused.
660 (main): Cast res_call to silence gcc.
661 * testsuite/libffi.call/cls_multi_uchar.c (test_func_gn): Mark cif and
662 data unused.
663 (main): Cast res_call to silence gcc.
664 * testsuite/libffi.call/cls_multi_ushort.c (test_func_gn): Mark cif and
665 data unused.
666 (main): Cast res_call to silence gcc.
667 * testsuite/libffi.call/cls_multi_ushortchar.c (test_func_gn): Mark cif
668 and data unused.
669 (main): Cast res_call to silence gcc.
670 * testsuite/libffi.call/cls_schar.c (cls_ret_schar_fn): Mark cif and
671 userdata unused.
672 (cls_ret_schar_fn): Cast printf parameter to silence gcc.
673 * testsuite/libffi.call/cls_sint.c (cls_ret_sint_fn): Mark cif and
674 userdata unused.
675 (cls_ret_sint_fn): Cast printf parameter to silence gcc.
676 * testsuite/libffi.call/cls_sshort.c (cls_ret_sshort_fn): Mark cif and
677 userdata unused.
678 (cls_ret_sshort_fn): Cast printf parameter to silence gcc.
679 * testsuite/libffi.call/cls_uchar.c (cls_ret_uchar_fn): Mark cif and
680 userdata unused.
681 (cls_ret_uchar_fn): Cast printf parameter to silence gcc.
682 * testsuite/libffi.call/cls_uint.c (cls_ret_uint_fn): Mark cif and
683 userdata unused.
684 (cls_ret_uint_fn): Cast printf parameter to silence gcc.
685 * testsuite/libffi.call/cls_ulonglong.c (cls_ret_ulonglong_fn): Mark cif
686 and userdata unused.
687 * testsuite/libffi.call/cls_ushort.c (cls_ret_ushort_fn): Mark cif and
688 userdata unused.
689 (cls_ret_ushort_fn): Cast printf parameter to silence gcc.
690 * testsuite/libffi.call/float.c (floating): Remove unused parameter e.
691 * testsuite/libffi.call/float1.c (main): Remove unused variable i.
692 Cleanup white spaces.
693 * testsuite/libffi.call/negint.c (checking): Remove unused variable i.
694 * testsuite/libffi.call/nested_struct.c (cls_struct_combined_gn): Mark
695 cif and userdata unused.
696 * testsuite/libffi.call/nested_struct1.c (cls_struct_combined_gn):
697 Likewise.
698 * testsuite/libffi.call/nested_struct10.c (B_gn): Likewise.
699 * testsuite/libffi.call/nested_struct2.c (B_fn): Adjust printf
700 formatters to silence gcc.
701 (B_gn): Mark cif and userdata unused.
702 * testsuite/libffi.call/nested_struct3.c (B_gn): Mark cif and userdata
703 unused.
704 * testsuite/libffi.call/nested_struct4.c: Mention related PR.
705 (B_gn): Mark cif and userdata unused.
706 * testsuite/libffi.call/nested_struct5.c (B_gn): Mark cif and userdata
707 unused.
708 * testsuite/libffi.call/nested_struct6.c: Mention related PR.
709 (B_gn): Mark cif and userdata unused.
710 * testsuite/libffi.call/nested_struct7.c (B_gn): Mark cif and userdata
711 unused.
712 * testsuite/libffi.call/nested_struct8.c (B_gn): Likewise.
713 * testsuite/libffi.call/nested_struct9.c (B_gn): Likewise.
714 * testsuite/libffi.call/problem1.c (stub): Likewise.
715 * testsuite/libffi.call/pyobjc-tc.c (main): Cast the result to silence
716 gcc.
717 * testsuite/libffi.call/return_fl2.c (return_fl): Add the note mentioned
718 in the last commit for this test case in the test case itself.
719 * testsuite/libffi.call/closure_fn0.c (closure_test_fn0): Mark cif as
720 unused.
721 * testsuite/libffi.call/closure_fn1.c (closure_test_fn1): Likewise.
722 * testsuite/libffi.call/closure_fn2.c (closure_test_fn2): Likewise.
723 * testsuite/libffi.call/closure_fn3.c (closure_test_fn3): Likewise.
724 * testsuite/libffi.call/closure_fn4.c (closure_test_fn0): Likewise.
725 * testsuite/libffi.call/closure_fn5.c (closure_test_fn5): Likewise.
726 * testsuite/libffi.call/closure_fn6.c (closure_test_fn0): Likewise.
727
5ab2e6e4 7282006-02-22 Kaz Kojima <kkojima@gcc.gnu.org>
729
730 * src/sh/sysv.S: Fix register numbers in the FDE for
731 ffi_closure_SYSV.
732
849b9f91 7332006-02-20 Andreas Tobler <a.tobler@schweiz.ch>
734
735 * testsuite/libffi.call/return_fl2.c (return_fl): Remove static
736 declaration to avoid a false negative on ix86. See PR323.
737
003f258f 7382006-02-18 Kaz Kojima <kkojima@gcc.gnu.org>
739
740 * src/sh/ffi.c (ffi_closure_helper_SYSV): Remove unused variable
741 and cast integer to void * if needed. Update the pointer to
742 the FP register saved area correctly.
743
9121151b 7442006-02-17 Andreas Tobler <a.tobler@schweiz.ch>
745
746 * testsuite/libffi.call/nested_struct6.c: XFAIL this test until PR25630
747 is fixed.
748 * testsuite/libffi.call/nested_struct4.c: Likewise.
749
019a3b64 7502006-02-16 Andreas Tobler <a.tobler@schweiz.ch>
751
6cbfba8c 752 * testsuite/libffi.call/return_dbl.c: New test case.
753 * testsuite/libffi.call/return_dbl1.c: Likewise.
754 * testsuite/libffi.call/return_dbl2.c: Likewise.
755 * testsuite/libffi.call/return_fl.c: Likewise.
756 * testsuite/libffi.call/return_fl1.c: Likewise.
757 * testsuite/libffi.call/return_fl2.c: Likewise.
758 * testsuite/libffi.call/return_fl3.c: Likewise.
759 * testsuite/libffi.call/closure_fn6.c: Likewise.
760
019a3b64 761 * testsuite/libffi.call/nested_struct2.c: Remove ffi_type_mylong
762 definition.
763 * testsuite/libffi.call/ffitest.h: Add ffi_type_mylong definition
764 here to be used by other test cases too.
765
766 * testsuite/libffi.call/nested_struct10.c: New test case.
767 * testsuite/libffi.call/nested_struct9.c: Likewise.
768 * testsuite/libffi.call/nested_struct8.c: Likewise.
769 * testsuite/libffi.call/nested_struct7.c: Likewise.
770 * testsuite/libffi.call/nested_struct6.c: Likewise.
771 * testsuite/libffi.call/nested_struct5.c: Likewise.
772 * testsuite/libffi.call/nested_struct4.c: Likewise.
773
946a8d62 7742006-01-21 Andreas Tobler <a.tobler@schweiz.ch>
775
776 * configure.ac: Enable libffi for sparc64-*-freebsd*.
777 * configure: Rebuilt.
778
cb99d24e 7792006-01-18 Jakub Jelinek <jakub@redhat.com>
780
781 * src/powerpc/sysv.S (smst_two_register): Don't call __ashldi3,
782 instead do the shifting inline.
783 * src/powerpc/ppc_closure.S (ffi_closure_SYSV): Don't compute %r5
784 shift count unconditionally. Simplify load sequences for 1, 2, 3, 4
785 and 8 byte structs, for the remaining struct sizes don't call
786 __lshrdi3, instead do the shifting inline.
787
06bf2366 7882005-12-07 Thiemo Seufer <ths@networkno.de>
789
790 * src/mips/ffitarget.h: Remove obsolete sgidefs.h include. Add
791 missing parentheses.
792 * src/mips/o32.S (ffi_call_O32): Code formatting. Define
793 and use A3_OFF, FP_OFF, RA_OFF. Micro-optimizations.
794 (ffi_closure_O32): Likewise, but with newly defined A3_OFF2,
795 A2_OFF2, A1_OFF2, A0_OFF2, RA_OFF2, FP_OFF2, S0_OFF2, GP_OFF2,
796 V1_OFF2, V0_OFF2, FA_1_1_OFF2, FA_1_0_OFF2, FA_0_1_OFF2,
797 FA_0_0_OFF2.
798 * src/mips/ffi.c (ffi_prep_args): Code formatting. Fix
799 endianness bugs.
800 (ffi_prep_closure): Improve trampoline instruction scheduling.
801 (ffi_closure_mips_inner_O32): Fix endianness bugs.
802
9a385484 8032005-12-03 Alan Modra <amodra@bigpond.net.au>
804
805 * src/powerpc/ffi.c: Formatting.
806 (ffi_prep_args_SYSV): Avoid possible aliasing problems by using unions.
807 (ffi_prep_args64): Likewise.
808
193c2942 8092005-09-30 Geoffrey Keating <geoffk@apple.com>
810
811 * testsuite/lib/libffi-dg.exp (libffi_target_compile): For
812 darwin, use -shared-libgcc not -lgcc_s, and explain why.
813
d3eb23fb 8142005-09-26 Tom Tromey <tromey@redhat.com>
815
816 * testsuite/libffi.call/float1.c (value_type): New typedef.
817 (CANARY): New define.
818 (main): Check for result buffer overflow.
819 * src/powerpc/linux64.S: Handle linux64 long double returns.
820 * src/powerpc/ffi.c (FLAG_RETURNS_128BITS): New constant.
821 (ffi_prep_cif_machdep): Handle linux64 long double returns.
822
5497bbdf 8232005-08-25 Alan Modra <amodra@bigpond.net.au>
824
825 PR target/23404
826 * src/powerpc/ffi.c (ffi_prep_args_SYSV): Correct placement of stack
827 homed fp args.
828 (ffi_status ffi_prep_cif_machdep): Correct stack sizing for same.
829
33b4ca90 8302005-08-11 Jakub Jelinek <jakub@redhat.com>
831
832 * configure.ac (HAVE_HIDDEN_VISIBILITY_ATTRIBUTE): New test.
833 (AH_BOTTOM): Add FFI_HIDDEN definition.
834 * configure: Rebuilt.
835 * fficonfig.h.in: Rebuilt.
836 * src/powerpc/ffi.c (hidden): Remove.
837 (ffi_closure_LINUX64, ffi_prep_args64, ffi_call_LINUX64,
838 ffi_closure_helper_LINUX64): Use FFI_HIDDEN instead of hidden.
839 * src/powerpc/linux64_closure.S (ffi_closure_LINUX64,
840 .ffi_closure_LINUX64): Use FFI_HIDDEN instead of .hidden.
841 * src/x86/ffi.c (ffi_closure_SYSV, ffi_closure_raw_SYSV): Remove,
842 add FFI_HIDDEN to its prototype.
843 (ffi_closure_SYSV_inner): New.
844 * src/x86/sysv.S (ffi_closure_SYSV, ffi_closure_raw_SYSV): New.
845 * src/x86/win32.S (ffi_closure_SYSV, ffi_closure_raw_SYSV): New.
846
967ec51b 8472005-08-10 Alfred M. Szmidt <ams@gnu.org>
848
849 PR libffi/21819:
850 * configure: Rebuilt.
851 * configure.ac: Handle i*86-*-gnu*.
852
80207f81 8532005-08-09 Jakub Jelinek <jakub@redhat.com>
854
855 * src/powerpc/ppc_closure.S (ffi_closure_SYSV): Use
856 DW_CFA_offset_extended_sf rather than
857 DW_CFA_GNU_negative_offset_extended.
858 * src/powerpc/sysv.S (ffi_call_SYSV): Likewise.
859
6a0e74dc 8602005-07-22 SUGIOKA Toshinobu <sugioka@itonet.co.jp>
861
862 * src/sh/sysv.S (ffi_call_SYSV): Stop argument popping correctly
863 on sh3.
864 (ffi_closure_SYSV): Change the stack layout for sh3 struct argument.
865 * src/sh/ffi.c (ffi_prep_args): Fix sh3 argument copy, when it is
866 partially on register.
867 (ffi_closure_helper_SYSV): Likewise.
868 (ffi_prep_cif_machdep): Don't set too many cif->flags.
869
cef3f52b 8702005-07-20 Kaz Kojima <kkojima@gcc.gnu.org>
871
872 * src/sh/ffi.c (ffi_call): Handle small structures correctly.
873 Remove empty line.
874 * src/sh64/ffi.c (simple_type): Remove.
875 (return_type): Handle small structures correctly.
876 (ffi_prep_args): Likewise.
877 (ffi_call): Likewise.
878 (ffi_closure_helper_SYSV): Likewise.
879 * src/sh64/sysv.S (ffi_call_SYSV): Handle 1, 2 and 4-byte return.
880 Emit position independent code if PIC and remove wrong datalabel
881 prefixes from EH data.
882
66c44239 8832005-07-19 Andreas Tobler <a.tobler@schweiz.ch>
884
885 * Makefile.am (nodist_libffi_la_SOURCES): Add POWERPC_FREEBSD.
886 * Makefile.in: Regenerate.
887 * include/Makefile.in: Likewise.
888 * testsuite/Makefile.in: Likewise.
889 * configure.ac: Add POWERPC_FREEBSD rules.
890 * configure: Regenerate.
891 * src/powerpc/ffitarget.h: Add POWERPC_FREEBSD rules.
892 (FFI_SYSV_TYPE_SMALL_STRUCT): Define.
893 * src/powerpc/ffi.c: Add flags to handle small structure returns
894 in ffi_call_SYSV.
895 (ffi_prep_cif_machdep): Handle small structures for SYSV 4 ABI.
896 Aka FFI_SYSV.
897 (ffi_closure_helper_SYSV): Likewise.
898 * src/powerpc/ppc_closure.S: Add return types for small structures.
899 * src/powerpc/sysv.S: Add bits to handle small structures for
900 final SYSV 4 ABI.
901
a21bd97f 9022005-07-10 Andreas Tobler <a.tobler@schweiz.ch>
903
904 * testsuite/libffi.call/cls_5_1_byte.c: New test file.
905 * testsuite/libffi.call/cls_6_1_byte.c: Likewise.
906 * testsuite/libffi.call/cls_7_1_byte.c: Likewise.
907
c350327d 9082005-07-05 Randolph Chung <tausq@debian.org>
909
910 * src/pa/ffi.c (ffi_struct_type): Rename FFI_TYPE_SMALL_STRUCT1
911 as FFI_TYPE_SMALL_STRUCT3. Break out handling for 5-7 byte
912 structures. Kill compilation warnings.
913 (ffi_closure_inner_LINUX): Print return values as hex in debug
914 message. Rename FFI_TYPE_SMALL_STRUCT1 as FFI_TYPE_SMALL_STRUCT3.
915 Properly handle 5-7 byte structure returns.
916 * src/pa/ffitarget.h (FFI_TYPE_SMALL_STRUCT1)
917 (FFI_TYPE_SMALL_STRUCT2): Remove.
918 (FFI_TYPE_SMALL_STRUCT3, FFI_TYPE_SMALL_STRUCT5)
919 (FFI_TYPE_SMALL_STRUCT6, FFI_TYPE_SMALL_STRUCT7): Define.
920 * src/pa/linux.S: Mark source file as using PA1.1 assembly.
921 (checksmst1, checksmst2): Remove.
922 (checksmst3): Optimize handling of 3-byte struct returns.
923 (checksmst567): Properly handle 5-7 byte struct returns.
924
20750e69 9252005-06-15 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
926
927 PR libgcj/21943
928 * src/mips/n32.S: Enforce PIC code.
929 * src/mips/o32.S: Likewise.
930
58841df8 9312005-06-15 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
932
933 * configure.ac: Treat i*86-*-solaris2.10 and up as X86_64.
934 * configure: Regenerate.
935
cc46fb30 9362005-06-01 Alan Modra <amodra@bigpond.net.au>
937
938 * src/powerpc/ppc_closure.S (ffi_closure_SYSV): Don't use JUMPTARGET
939 to call ffi_closure_helper_SYSV. Append @local instead.
940 * src/powerpc/sysv.S (ffi_call_SYSV): Likewise for ffi_prep_args_SYSV.
941
6e912c1e 9422005-05-17 Kelley Cook <kcook@gcc.gnu.org>
69738b31 943
944 * configure.ac: Use AC_C_BIGENDIAN instead of AC_C_BIGENDIAN_CROSS.
945 Use AC_CHECK_SIZEOF instead of AC_COMPILE_CHECK_SIZEOF.
946 * Makefile.am (ACLOCAL_AMFLAGS): Remove -I ../config.
947 * aclocal.m4, configure, fficonfig.h.in, Makefile.in,
948 include/Makefile.in, testsuite/Makefile.in: Regenerate.
949
5f0a087e 9502005-05-09 Mike Stump <mrs@apple.com>
951
952 * configure: Regenerate.
953
81734468 9542005-05-08 Richard Henderson <rth@redhat.com>
955
956 PR libffi/21285
957 * src/alpha/osf.S: Update unwind into to match code.
958
4e9df534 9592005-05-04 Andreas Degert <ad@papyrus-gmbh.de>
960 Richard Henderson <rth@redhat.com>
961
962 * src/x86/ffi64.c (ffi_prep_cif_machdep): Save sse-used flag in
963 bit 11 of flags.
964 (ffi_call): Mask return type field. Pass ssecount to ffi_call_unix64.
965 (ffi_prep_closure): Set carry bit if sse-used flag set.
966 * src/x86/unix64.S (ffi_call_unix64): Add ssecount argument.
967 Only load sse registers if ssecount non-zero.
968 (ffi_closure_unix64): Only save sse registers if carry set on entry.
969
9702005-04-29 Ralf Corsepius <ralf.corsepius@rtems.org>
6155bf60 971
972 * configure.ac: Add i*86-*-rtems*, sparc*-*-rtems*,
973 powerpc-*rtems*, arm*-*-rtems*, sh-*-rtems*.
974 * configure: Regenerate.
975
643e4a19 9762005-04-20 Hans-Peter Nilsson <hp@axis.com>
977
978 * testsuite/lib/libffi-dg.exp (libffi-dg-test-1): In regsub use,
979 have Tcl8.3-compatible intermediate variable.
980
7e5823cf 9812005-04-18 Simon Posnjak <simon.posnjak@siol.net>
982 Hans-Peter Nilsson <hp@axis.com>
983
984 * Makefile.am: Add CRIS support.
985 * configure.ac: Likewise.
942132a7 986 * Makefile.in, configure, testsuite/Makefile.in,
987 include/Makefile.in: Regenerate.
7e5823cf 988 * src/cris: New directory.
989 * src/cris/ffi.c, src/cris/sysv.S, src/cris/ffitarget.h: New files.
990 * src/prep_cif.c (ffi_prep_cif): Wrap in #ifndef __CRIS__.
f0a2eecf 991
992 * testsuite/lib/libffi-dg.exp (libffi-dg-test-1): Replace \n with
993 \r?\n in output tests.
994
3bcf2cd5 9952005-04-12 Mike Stump <mrs@apple.com>
996
997 * configure: Regenerate.
998
b7292ee2 9992005-03-30 Hans Boehm <Hans.Boehm@hp.com>
1000
1001 * src/ia64/ffitarget.h (ffi_arg): Use long long instead of DI.
1002
3bdc9b0e 10032005-03-30 Steve Ellcey <sje@cup.hp.com>
1004
1005 * src/ia64/ffitarget.h (ffi_arg) ADD DI attribute.
1006 (ffi_sarg) Ditto.
1007 * src/ia64/unix.S (ffi_closure_unix): Extend gp
1008 to 64 bits in ILP32 mode.
1009 Load 64 bits even for short data.
1010
7abd8367 10112005-03-23 Mike Stump <mrs@apple.com>
1012
1013 * src/powerpc/darwin.S: Update for -m64 multilib.
1014 * src/powerpc/darwin_closure.S: Likewise.
1015
bbb07d35 10162005-03-21 Zack Weinberg <zack@codesourcery.com>
1017
1018 * configure.ac: Do not invoke TL_AC_GCC_VERSION.
1019 Do not set tool_include_dir.
1020 * aclocal.m4, configure, Makefile.in, testsuite/Makefile.in:
1021 Regenerate.
1022 * include/Makefile.am: Set gcc_version and toollibffidir.
1023 * include/Makefile.in: Regenerate.
1024
e6fd470b 10252005-02-22 Andrew Haley <aph@redhat.com>
1026
1027 * src/powerpc/ffi.c (ffi_prep_cif_machdep): Bump alignment to
1028 odd-numbered register pairs for 64-bit integer types.
1029
4f373876 10302005-02-23 Andreas Tobler <a.tobler@schweiz.ch>
1031
1032 PR libffi/20104
1033 * testsuite/libffi.call/return_ll1.c: New test case.
1034
c2cf854f 10352005-02-11 Janis Johnson <janis187@us.ibm.com>
1036
1037 * testsuite/libffi.call/cls_align_longdouble.c: Remove dg-options.
1038 * testsuite/libffi.call/float.c: Ditto.
1039 * testsuite/libffi.call/float2.c: Ditto.
1040 * testsuite/libffi.call/float3.c: Ditto.
1041
9e71f2c8 10422005-02-08 Andreas Tobler <a.tobler@schweiz.ch>
1043
1044 * src/frv/ffitarget.h: Remove PPC stuff which does not belong to frv.
1045
45e075b4 10462005-01-12 Eric Botcazou <ebotcazou@libertysurf.fr>
1047
1048 * testsuite/libffi.special/special.exp (cxx_options): Add
1049 -shared-libgcc.
1050
f587fca2 10512004-12-31 Richard Henderson <rth@redhat.com>
1052
1053 * src/types.c (FFI_AGGREGATE_TYPEDEF): Remove.
1054 (FFI_TYPEDEF): Rename from FFI_INTEGRAL_TYPEDEF. Replace size and
1055 offset parameters with a type parameter; deduce size and structure
1056 alignment. Update all users.
1057
717c502e 10582004-12-31 Richard Henderson <rth@redhat.com>
1059
1060 * src/types.c (FFI_TYPE_POINTER): Define with sizeof.
1061 (FFI_TYPE_LONGDOUBLE): Fix for ia64.
bbb07d35 1062 * src/ia64/ffitarget.h (struct ffi_ia64_trampoline_struct): Move
717c502e 1063 into ffi_prep_closure.
1064 * src/ia64/ia64_flags.h, src/ia64/ffi.c, src/ia64/unix.S: Rewrite
1065 from scratch.
1066
4a60643d 10672004-12-27 Richard Henderson <rth@redhat.com>
1068
1069 * src/x86/unix64.S: Fix typo in unwind info.
1070
66d912b3 10712004-12-25 Richard Henderson <rth@redhat.com>
1072
1073 * src/x86/ffi64.c (struct register_args): Rename from stackLayout.
1074 (enum x86_64_reg_class): Add X86_64_COMPLEX_X87_CLASS.
1075 (merge_classes): Check for it.
1076 (SSE_CLASS_P): New.
1077 (classify_argument): Pass byte_offset by value; perform all updates
1078 inside struct case.
1079 (examine_argument): Add classes argument; handle
1080 X86_64_COMPLEX_X87_CLASS.
1081 (ffi_prep_args): Merge into ...
1082 (ffi_call): ... here. Share stack frame with ffi_call_unix64.
1083 (ffi_prep_cif_machdep): Setup cif->flags for proper structure return.
1084 (ffi_fill_return_value): Remove.
1085 (ffi_prep_closure): Remove dead assert.
1086 (ffi_closure_unix64_inner): Rename from ffi_closure_UNIX64_inner.
1087 Rewrite to use struct register_args instead of va_list. Create
1088 flags for handling structure returns.
1089 * src/x86/unix64.S: Remove dead strings.
1090 (ffi_call_unix64): Rename from ffi_call_UNIX64. Rewrite to share
1091 stack frame with ffi_call. Handle structure returns properly.
1092 (float2sse, floatfloat2sse, double2sse): Remove.
1093 (sse2float, sse2double, sse2floatfloat): Remove.
1094 (ffi_closure_unix64): Rename from ffi_closure_UNIX64. Rewrite
1095 to handle structure returns properly.
1096
a8594486 10972004-12-08 David Edelsohn <edelsohn@gnu.org>
1098
1099 * Makefile.am (AM_MAKEFLAGS): Remove duplicate LIBCFLAGS and
1100 PICFLAG.
1101 * Makefile.in: Regenerated.
1102
14ac9144 11032004-12-02 Richard Sandiford <rsandifo@redhat.com>
1104
1105 * configure.ac: Use TL_AC_GCC_VERSION to set gcc_version.
1106 * configure, aclocal.m4, Makefile.in: Regenerate.
1107 * include/Makefile.in, testsuite/Makefile.in: Regenerate.
1108
d081dd29 11092004-11-29 Kelley Cook <kcook@gcc.gnu.org>
1110
1111 * configure: Regenerate for libtool change.
1112
a234d196 11132004-11-25 Kelley Cook <kcook@gcc.gnu.org>
1114
1115 * configure: Regenerate for libtool reversion.
1116
7f93fe7a 11172004-11-24 Kelley Cook <kcook@gcc.gnu.org>
1118
1119 * configure: Regenerate for libtool change.
1120
c0e31427 11212004-11-23 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
1122
1123 * testsuite/lib/libffi-dg.exp: Use new procs in target-libpath.exp.
1124
11252004-11-23 Richard Sandiford <rsandifo@redhat.com>
1126
1127 * src/mips/o32.S (ffi_call_O32, ffi_closure_O32): Use jalr instead
1128 of jal. Use an absolute encoding for the frame information.
1129
11302004-11-23 Kelley Cook <kcook@gcc.gnu.org>
1131
1132 * Makefile.am: Remove no-dependencies. Add ACLOCAL_AMFLAGS.
1133 * acinclude.m4: Delete logic for sincludes.
1134 * aclocal.m4, Makefile.in, configure: Regenerate.
1135 * include/Makefile: Likewise.
1136 * testsuite/Makefile: Likewise.
1137
68fa8e89 11382004-11-22 Eric Botcazou <ebotcazou@libertysurf.fr>
1139
1140 * src/sparc/ffi.c (ffi_prep_closure): Align doubles and 64-bit integers
1141 on a 8-byte boundary.
1142 * src/sparc/v8.S (ffi_closure_v8): Reserve frame space for arguments.
1143
c11820d9 11442004-10-27 Richard Earnshaw <rearnsha@arm.com>
1145
1146 * src/arm/ffi.c (ffi_prep_cif_machdep): Handle functions that return
1147 long long values. Round stack allocation to a multiple of 8 bytes
1148 for ATPCS compatibility.
1149 * src/arm/sysv.S (ffi_call_SYSV): Rework to avoid use of APCS register
1150 names. Handle returning long long types. Add Thumb and interworking
1151 support. Improve soft-float code.
1152
03b2b968 11532004-10-27 Richard Earnshaw <rearnsha@arm.com>
1154
1155 * testsuite/lib/libffi-db.exp (load_gcc_lib): New function.
1156 (libffi_exit): New function.
1157 (libffi_init): Build the testglue wrapper if needed.
1158
9f234538 11592004-10-25 Eric Botcazou <ebotcazou@libertysurf.fr>
1160
1161 PR other/18138
1162 * testsuite/lib/libffi-dg.exp: Accept more than one multilib libgcc.
1163
61dc6540 11642004-10-25 Kazuhiro Inaoka <inaoka.kazuhiro@renesas.com>
1165
1166 * src/m32r/libffitarget.h (FFI_CLOSURES): Set to 0.
1167
3b77fb00 11682004-10-20 Kaz Kojima <kkojima@gcc.gnu.org>
1169
1170 * src/sh/sysv.S (ffi_call_SYSV): Don't align for double data.
1171 * testsuite/libffi.call/float3.c: New test case.
1172
7c2fd652 11732004-10-18 Kaz Kojima <kkojima@gcc.gnu.org>
1174
1175 * src/sh/ffi.c (ffi_prep_closure): Set T bit in trampoline for
1176 the function returning a structure pointed with R2.
1177 * src/sh/sysv.S (ffi_closure_SYSV): Use R2 as the pointer to
1178 the structure return value if T bit set. Emit position
1179 independent code and EH data if PIC.
1180
7e74b6d4 11812004-10-13 Kazuhiro Inaoka <inaoka.kazuhiro@renesas.com>
1182
1183 * Makefile.am: Add m32r support.
1184 * configure.ac: Likewise.
1185 * Makefile.in: Regenerate.
1186 * confiugre: Regenerate.
1187 * src/types.c: Add m32r port to FFI_INTERNAL_TYPEDEF
1188 (uint64, sint64, double, longdouble)
1189 * src/m32r: New directory.
1190 * src/m32r/ffi.c: New file.
1191 * src/m32r/sysv.S: Likewise.
1192 * src/m32r/ffitarget.h: Likewise.
1193
d601026d 11942004-10-02 Kaz Kojima <kkojima@gcc.gnu.org>
1195
1196 * testsuite/libffi.call/negint.c: New test case.
1197
ec1bc4ab 11982004-09-14 H.J. Lu <hongjiu.lu@intel.com>
1199
1200 PR libgcj/17465
1201 * testsuite/lib/libffi-dg.exp: Don't use global ld_library_path.
1202 Set up LD_LIBRARY_PATH, SHLIB_PATH, LD_LIBRARYN32_PATH,
1203 LD_LIBRARY64_PATH, LD_LIBRARY_PATH_32, LD_LIBRARY_PATH_64 and
1204 DYLD_LIBRARY_PATH.
1205
41067a17 12062004-09-05 Andreas Tobler <a.tobler@schweiz.ch>
1207
1208 * testsuite/libffi.call/many_win32.c: Remove whitespaces.
1209 * testsuite/libffi.call/promotion.c: Likewise.
1210 * testsuite/libffi.call/return_ll.c: Remove unused var. Cleanup
1211 whitespaces.
1212 * testsuite/libffi.call/return_sc.c: Likewise.
1213 * testsuite/libffi.call/return_uc.c: Likewise.
1214
acbdb39a 12152004-09-05 Andreas Tobler <a.tobler@schweiz.ch>
1216
1217 * src/powerpc/darwin.S: Fix comments and identation.
1218 * src/powerpc/darwin_closure.S: Likewise.
1219
de7fad36 12202004-09-02 Andreas Tobler <a.tobler@schweiz.ch>
1221
1222 * src/powerpc/ffi_darwin.c: Add flag for longdouble return values.
1223 (ffi_prep_args): Handle longdouble arguments.
1224 (ffi_prep_cif_machdep): Set flags for longdouble. Calculate space for
1225 longdouble.
1226 (ffi_closure_helper_DARWIN): Add closure handling for longdouble.
1227 * src/powerpc/darwin.S (_ffi_call_DARWIN): Add handling of longdouble
1228 values.
1229 * src/powerpc/darwin_closure.S (_ffi_closure_ASM): Likewise.
1230 * src/types.c: Defined longdouble size and alignment for darwin.
1231
8bf7bfd1 12322004-09-02 Andreas Tobler <a.tobler@schweiz.ch>
1233
1234 * src/powerpc/aix.S: Remove whitespaces.
1235 * src/powerpc/aix_closure.S: Likewise.
1236 * src/powerpc/asm.h: Likewise.
1237 * src/powerpc/ffi.c: Likewise.
1238 * src/powerpc/ffitarget.h: Likewise.
1239 * src/powerpc/linux64.S: Likewise.
1240 * src/powerpc/linux64_closure.S: Likewise.
1241 * src/powerpc/ppc_closure.S: Likewise.
1242 * src/powerpc/sysv.S: Likewise.
1243
2e818f50 12442004-08-30 Anthony Green <green@redhat.com>
1245
1246 * Makefile.am: Add frv support.
1247 * Makefile.in, testsuite/Makefile.in: Rebuilt.
1248 * configure.ac: Read configure.host.
1249 * configure.in: Read configure.host.
1250 * configure.host: New file. frv-elf needs libgloss.
1251 * include/ffi.h.in: Force ffi_closure to have a nice big (8)
752da755 1252 alignment. This is needed to frv and shouldn't harm the others.
2e818f50 1253 * include/ffi_common.h (ALIGN_DOWN): New macro.
1254 * src/frv/ffi.c, src/frv/ffitarget.h, src/frv/eabi.S: New files.
1255
d36998ee 12562004-08-24 David Daney <daney@avtrex.com>
1257
1258 * testsuite/libffi.call/closure_fn0.c: Xfail mips64* instead of mips*.
1259 * testsuite/libffi.call/closure_fn1.c: Likewise.
1260 * testsuite/libffi.call/closure_fn2.c Likewise.
1261 * testsuite/libffi.call/closure_fn3.c: Likewise.
1262 * testsuite/libffi.call/closure_fn4.c: Likewise.
1263 * testsuite/libffi.call/closure_fn5.c: Likewise.
1264 * testsuite/libffi.call/cls_18byte.c: Likewise.
1265 * testsuite/libffi.call/cls_19byte.c: Likewise.
1266 * testsuite/libffi.call/cls_1_1byte.c: Likewise.
1267 * testsuite/libffi.call/cls_20byte.c: Likewise.
1268 * testsuite/libffi.call/cls_20byte1.c: Likewise.
1269 * testsuite/libffi.call/cls_24byte.c: Likewise.
1270 * testsuite/libffi.call/cls_2byte.c: Likewise.
1271 * testsuite/libffi.call/cls_3_1byte.c: Likewise.
1272 * testsuite/libffi.call/cls_3byte1.c: Likewise.
1273 * testsuite/libffi.call/cls_3byte2.c: Likewise.
1274 * testsuite/libffi.call/cls_4_1byte.c: Likewise.
1275 * testsuite/libffi.call/cls_4byte.c: Likewise.
1276 * testsuite/libffi.call/cls_64byte.c: Likewise.
1277 * testsuite/libffi.call/cls_6byte.c: Likewise.
1278 * testsuite/libffi.call/cls_7byte.c: Likewise.
1279 * testsuite/libffi.call/cls_8byte.c: Likewise.
1280 * testsuite/libffi.call/cls_9byte1.c: Likewise.
1281 * testsuite/libffi.call/cls_9byte2.c: Likewise.
1282 * testsuite/libffi.call/cls_align_double.c: Likewise.
1283 * testsuite/libffi.call/cls_align_float.c: Likewise.
1284 * testsuite/libffi.call/cls_align_longdouble.c: Likewise.
1285 * testsuite/libffi.call/cls_align_pointer.c: Likewise.
1286 * testsuite/libffi.call/cls_align_sint16.c: Likewise.
1287 * testsuite/libffi.call/cls_align_sint32.c: Likewise.
1288 * testsuite/libffi.call/cls_align_sint64.c: Likewise.
1289 * testsuite/libffi.call/cls_align_uint16.c: Likewise.
1290 * testsuite/libffi.call/cls_align_uint32.c: Likewise.
1291 * testsuite/libffi.call/cls_align_uint64.c: Likewise.
1292 * testsuite/libffi.call/cls_double.c: Likewise.
1293 * testsuite/libffi.call/cls_float.c: Likewise.
1294 * testsuite/libffi.call/cls_multi_schar.c: Likewise.
1295 * testsuite/libffi.call/cls_multi_sshort.c: Likewise.
1296 * testsuite/libffi.call/cls_multi_sshortchar.c: Likewise.
1297 * testsuite/libffi.call/cls_multi_uchar.c: Likewise.
1298 * testsuite/libffi.call/cls_multi_ushort.c: Likewise.
1299 * testsuite/libffi.call/cls_multi_ushortchar.c: Likewise.
1300 * testsuite/libffi.call/cls_schar.c: Likewise.
1301 * testsuite/libffi.call/cls_sint.c: Likewise.
1302 * testsuite/libffi.call/cls_sshort.c: Likewise.
1303 * testsuite/libffi.call/cls_uchar.c: Likewise.
1304 * testsuite/libffi.call/cls_uint.c: Likewise.
1305 * testsuite/libffi.call/cls_ulonglong.c: Likewise.
1306 * testsuite/libffi.call/cls_ushort.c: Likewise.
1307 * testsuite/libffi.call/nested_struct.c: Likewise.
1308 * testsuite/libffi.call/nested_struct1.c: Likewise.
1309 * testsuite/libffi.call/nested_struct2.c: Likewise.
1310 * testsuite/libffi.call/nested_struct3.c: Likewise.
1311 * testsuite/libffi.call/problem1.c: Likewise.
1312 * testsuite/libffi.special/unwindtest.cc: Likewise.
1313 * testsuite/libffi.call/cls_12byte.c: Likewise and set return value
1314 to zero.
1315 * testsuite/libffi.call/cls_16byte.c: Likewise.
1316 * testsuite/libffi.call/cls_5byte.c: Likewise.
1317
29ac6c83 13182004-08-23 David Daney <daney@avtrex.com>
1319
1320 PR libgcj/13141
1321 * src/mips/ffitarget.h (FFI_O32_SOFT_FLOAT): New ABI.
1322 * src/mips/ffi.c (ffi_prep_args): Fix alignment calculation.
1323 (ffi_prep_cif_machdep): Handle FFI_O32_SOFT_FLOAT floating point
1324 parameters and return types.
1325 (ffi_call): Handle FFI_O32_SOFT_FLOAT ABI.
1326 (ffi_prep_closure): Ditto.
1327 (ffi_closure_mips_inner_O32): Handle FFI_O32_SOFT_FLOAT ABI, fix
1328 alignment calculations.
1329 * src/mips/o32.S (ffi_closure_O32): Don't use floating point
1330 instructions if FFI_O32_SOFT_FLOAT, make stack frame ABI compliant.
1331
4487a7e8 13322004-08-14 Casey Marshall <csm@gnu.org>
1333
1334 * src/mips/ffi.c (ffi_pref_cif_machdep): set `cif->flags' to
1335 contain `FFI_TYPE_UINT64' as return type for any 64-bit
1336 integer (O32 ABI only).
1337 (ffi_prep_closure): new function.
1338 (ffi_closure_mips_inner_O32): new function.
bbb07d35 1339 * src/mips/ffitarget.h: Define `FFI_CLOSURES' and
4487a7e8 1340 `FFI_TRAMPOLINE_SIZE' appropriately if the ABI is o32.
1341 * src/mips/o32.S (ffi_call_O32): add labels for .eh_frame. Return
1342 64 bit integers correctly.
1343 (ffi_closure_O32): new function.
1344 Added DWARF-2 unwind info for both functions.
1345
bb2d2363 13462004-08-10 Andrew Haley <aph@redhat.com>
1347
1348 * src/x86/ffi64.c (ffi_prep_args ): 8-align all stack arguments.
1349
cb8a7007 13502004-08-01 Robert Millan <robertmh@gnu.org>
1351
1352 * configure.ac: Detect knetbsd-gnu and kfreebsd-gnu.
1353 * configure: Regenerate.
1354
e2f66690 13552004-07-30 Maciej W. Rozycki <macro@linux-mips.org>
1356
1357 * acinclude.m4 (AC_FUNC_MMAP_BLACKLIST): Check for <sys/mman.h>
1358 and mmap() explicitly instead of relying on preset autoconf cache
1359 variables.
1360 * aclocal.m4: Regenerate.
1361 * configure: Regenerate.
1362
e395522d 13632004-07-11 Ulrich Weigand <uweigand@de.ibm.com>
1364
1365 * src/s390/ffi.c (ffi_prep_args): Fix C aliasing violation.
1366 (ffi_check_float_struct): Remove unused prototype.
1367
b589ba75 13682004-06-30 Geoffrey Keating <geoffk@apple.com>
1369
1370 * src/powerpc/ffi_darwin.c (flush_icache): ';' is a comment
1371 character on Darwin, use '\n\t' instead.
1372
7b709c61 13732004-06-26 Matthias Klose <doko@debian.org>
1374
1375 * libtool-version: Fix typo in revision/age.
1376
06e2077b 13772004-06-17 Matthias Klose <doko@debian.org>
1378
1379 * libtool-version: New.
1380 * Makefile.am (libffi_la_LDFLAGS): Use -version-info for soname.
1381 * Makefile.in: Regenerate.
1382
4889e098 13832004-06-15 Paolo Bonzini <bonzini@gnu.org>
d947df39 1384
1385 * Makefile.am: Remove useless multilib rules.
1386 * Makefile.in: Regenerate.
1387 * aclocal.m4: Regenerate with automake 1.8.5.
1388 * configure.ac: Remove useless multilib configury.
1389 * configure: Regenerate.
1390
4889e098 13912004-06-15 Paolo Bonzini <bonzini@gnu.org>
b2191698 1392
1393 * .cvsignore: New file.
1394
6bc491fa 13952004-06-10 Jakub Jelinek <jakub@redhat.com>
1396
1397 * src/ia64/unix.S (ffi_call_unix): Insert group barrier break
1398 fp_done.
1399 (ffi_closure_UNIX): Fix f14/f15 adjustment if FLOAT_SZ is ever
1400 changed from 8.
1401
e1c1d65f 14022004-06-06 Sean McNeil <sean@mcneil.com>
1403
1404 * configure.ac: Add x86_64-*-freebsd* support.
1405 * configure: Regenerate.
1406
2be99030 14072004-04-26 Joe Buck <jbuck@welsh-buck.org>
1408
1409 Bug 15093
1410 * configure.ac: Test for existence of mmap and sys/mman.h before
1411 checking blacklist. Fix suggested by Jim Wilson.
1412 * configure: Regenerate.
1413
efefde65 14142004-04-26 Matt Austern <austern@apple.com>
1415
1416 * src/powerpc/darwin.S: Go through a non-lazy pointer for initial
1417 FDE location.
1418 * src/powerpc/darwin_closure.S: Likewise.
bbb07d35 1419
5766842b 14202004-04-24 Andreas Tobler <a.tobler@schweiz.ch>
1421
1422 * testsuite/libffi.call/cls_multi_schar.c (main): Fix initialization
1423 error. Reported by Thomas Heller <theller@python.net>.
1424 * testsuite/libffi.call/cls_multi_sshort.c (main): Likewise.
1425 * testsuite/libffi.call/cls_multi_ushort.c (main): Likewise.
1426
43935228 14272004-03-20 Matthias Klose <doko@debian.org>
1428
1429 * src/pa/linux.S: Fix typo.
1430
f35c338f 14312004-03-19 Matthias Klose <doko@debian.org>
1432
5766842b 1433 * Makefile.am: Update.
f35c338f 1434 * Makefile.in: Regenerate.
1435 * src/pa/ffi.h.in: Remove.
1436 * src/pa/ffitarget.h: New file.
1437
14382004-02-10 Randolph Chung <tausq@debian.org>
1439
1440 * Makefile.am: Add PA support.
1441 * Makefile.in: Regenerate.
1442 * include/Makefile.in: Regenerate.
1443 * configure.ac: Add PA target.
1444 * configure: Regenerate.
1445 * src/pa/ffi.c: New file.
1446 * src/pa/ffi.h.in: Add PA support.
1447 * src/pa/linux.S: New file.
1448 * prep_cif.c: Add PA support.
1449
021bf237 14502004-03-16 Hosaka Yuji <hos@tamanegi.org>
1451
1452 * src/types.c: Fix alignment size of X86_WIN32 case int64 and
1453 double.
1454 * src/x86/ffi.c (ffi_prep_args): Replace ecif->cif->rtype->type
1455 with ecif->cif->flags.
1456 (ffi_call, ffi_prep_incoming_args_SYSV): Replace cif->rtype->type
1457 with cif->flags.
1458 (ffi_prep_cif_machdep): Add X86_WIN32 struct case.
1459 (ffi_closure_SYSV): Add 1 or 2-bytes struct case for X86_WIN32.
1460 * src/x86/win32.S (retstruct1b, retstruct2b, sc_retstruct1b,
1461 sc_retstruct2b): Add for 1 or 2-bytes struct case.
1462
4a900809 14632004-03-15 Kelley Cook <kcook@gcc.gnu.org>
1464
1465 * configure.in: Rename file to ...
1466 * configure.ac: ... this.
1467 * fficonfig.h.in: Regenerate.
1468 * Makefile.in: Regenerate.
1469 * include/Makefile.in: Regenerate.
1470 * testsuite/Makefile.in: Regenerate.
1471
d128584a 14722004-03-12 Matt Austern <austern@apple.com>
1473
1474 * src/powerpc/darwin.S: Fix EH information so it corresponds to
1475 changes in EH format resulting from addition of linkonce support.
1476 * src/powerpc/darwin_closure.S: Likewise.
021bf237 1477
b375837d 14782004-03-11 Andreas Tobler <a.tobler@schweiz.ch>
1479 Paolo Bonzini <bonzini@gnu.org>
1480
1481 * Makefile.am (AUTOMAKE_OPTIONS): Set them.
1482 Remove VPATH. Remove rules for object files. Remove multilib support.
1483 (AM_CCASFLAGS): Add.
1484 * configure.in (AC_CONFIG_HEADERS): Relace AM_CONFIG_HEADER.
1485 (AC_PREREQ): Bump version to 2.59.
1486 (AC_INIT): Fill with version info and bug address.
1487 (ORIGINAL_LD_FOR_MULTILIBS): Remove.
1488 (AM_ENABLE_MULTILIB): Use this instead of AC_ARG_ENABLE.
1489 De-precious CC so that the right flags are passed down to multilibs.
1490 (AC_MSG_ERROR): Replace obsolete macro AC_ERROR.
1491 (AC_CONFIG_FILES): Replace obsolete macro AC_LINK_FILES.
1492 (AC_OUTPUT): Reorganize the output with AC_CONFIG_COMMANDS.
1493 * configure: Rebuilt.
1494 * aclocal.m4: Likewise.
1495 * Makefile.in, include/Makefile.in, testsuite/Makefile.in: Likewise.
1496 * fficonfig.h.in: Likewise.
1497
52cee448 14982004-03-11 Andreas Schwab <schwab@suse.de>
1499
1500 * src/ia64/ffi.c (ffi_prep_incoming_args_UNIX): Get floating point
1501 arguments from fp registers only for the first 8 parameter slots.
1502 Don't convert a float parameter when passed in memory.
1503
ac09b42b 15042004-03-09 Hans-Peter Nilsson <hp@axis.com>
1505
1506 * configure: Regenerate for config/accross.m4 correction.
1507
497c7020 15082004-02-25 Matt Kraai <kraai@alumni.cmu.edu>
1509
1510 * src/powerpc/ffi.c (ffi_prep_args_SYSV): Change
1511 ecif->cif->bytes to bytes.
1512 (ffi_prep_cif_machdep): Add braces around nested if statement.
1513
0ad895fb 15142004-02-09 Alan Modra <amodra@bigpond.net.au>
1515
ad98ab60 1516 * src/types.c (pointer): POWERPC64 has 8 byte pointers.
1517
0ad895fb 1518 * src/powerpc/ffi.c (ffi_prep_args64): Correct long double handling.
1519 (ffi_closure_helper_LINUX64): Fix typo.
1520 * testsuite/libffi.call/cls_align_longdouble.c: Pass -mlong-double-128
1521 for powerpc64-*-*.
1522 * testsuite/libffi.call/float.c: Likewise.
1523 * testsuite/libffi.call/float2.c: Likewise.
1524
de9c7178 15252004-02-08 Alan Modra <amodra@bigpond.net.au>
1526
1527 * src/powerpc/ffi.c (ffi_prep_cif_machdep <FFI_LINUX64>): Correct
1528 long double function return and long double arg handling.
1529 (ffi_closure_helper_LINUX64): Formatting. Delete unused "ng" var.
1530 Use "end_pfr" instead of "nf". Correct long double handling.
1531 Localise "temp".
1532 * src/powerpc/linux64.S (ffi_call_LINUX64): Save f2 long double
1533 return value.
1534 * src/powerpc/linux64_closure.S (ffi_closure_LINUX64): Allocate
1535 space for long double return value. Adjust stack frame and offsets.
1536 Load f2 long double return.
1537
3474a085 15382004-02-07 Alan Modra <amodra@bigpond.net.au>
1539
1540 * src/types.c: Use 16 byte long double for POWERPC64.
1541
071ff5d6 15422004-01-25 Eric Botcazou <ebotcazou@libertysurf.fr>
1543
1544 * src/sparc/ffi.c (ffi_prep_args_v9): Shift the parameter array
1545 when the structure return address is passed in %o0.
1546 (ffi_V9_return_struct): Rename into ffi_v9_layout_struct.
1547 (ffi_v9_layout_struct): Align the field following a nested structure
1548 on a word boundary. Use memmove instead of memcpy.
1549 (ffi_call): Update call to ffi_V9_return_struct.
1550 (ffi_prep_closure): Define 'ctx' only for V8.
1551 (ffi_closure_sparc_inner): Clone into ffi_closure_sparc_inner_v8
1552 and ffi_closure_sparc_inner_v9.
1553 (ffi_closure_sparc_inner_v8): Return long doubles by reference.
1554 Always skip the structure return address. For structures and long
1555 doubles, copy the argument directly.
1556 (ffi_closure_sparc_inner_v9): Skip the structure return address only
1557 if required. Shift the maximum floating-point slot accordingly. For
1558 big structures, copy the argument directly; otherwise, left-justify the
1559 argument and call ffi_v9_layout_struct to lay out the structure on
1560 the stack.
1561 * src/sparc/v8.S: Undef STACKFRAME before defining it.
1562 (ffi_closure_v8): Pass the structure return address. Update call to
1563 ffi_closure_sparc_inner_v8. Short-circuit FFI_TYPE_INT handling.
1564 Skip the 'unimp' insn when returning long doubles and structures.
1565 * src/sparc/v9.S: Undef STACKFRAME before defining it.
1566 (ffi_closure_v9): Increase the frame size by 2 words. Short-circuit
1567 FFI_TYPE_INT handling. Load structures both in integers and
1568 floating-point registers on return.
1569 * README: Update status of the SPARC port.
1570
24ee5f14 15712004-01-24 Andreas Tobler <a.tobler@schweiz.ch>
1572
1573 * testsuite/libffi.call/pyobjc-tc.c (main): Treat result value
1574 as of type ffi_arg.
1575 * testsuite/libffi.call/struct3.c (main): Fix CHECK.
1576
1e0ee4d8 15772004-01-22 Ulrich Weigand <uweigand@de.ibm.com>
1578
1579 * testsuite/libffi.call/cls_uint.c (cls_ret_uint_fn): Treat result
1580 value as of type ffi_arg, not unsigned int.
1581
c886ae3f 15822004-01-21 Michael Ritzert <ritzert@t-online.de>
1583
1584 * ffi64.c (ffi_prep_args): Cast the RHS of an assignment instead
1585 of the LHS.
1586
fcd85db3 15872004-01-12 Andreas Tobler <a.tobler@schweiz.ch>
1588
1589 * testsuite/lib/libffi-dg.exp: Set LD_LIBRARY_PATH_32 for
1590 Solaris.
1591
dcf910e2 15922004-01-08 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
1593
1594 * testsuite/libffi.call/ffitest.h (allocate_mmap): Cast MAP_FAILED
1595 to void *.
1596
3d3042b8 15972003-12-10 Richard Henderson <rth@redhat.com>
1598
1599 * testsuite/libffi.call/cls_align_pointer.c: Cast pointers to
1600 size_t instead of int.
1601
dc36d4d1 16022003-12-04 Hosaka Yuji <hos@tamanegi.org>
1603
1604 * testsuite/libffi.call/many_win32.c: Include <float.h>.
1605 * testsuite/libffi.call/many_win32.c (main): Replace variable
1606 int i with unsigned long ul.
1607
1608 * testsuite/libffi.call/cls_align_uint64.c: New test case.
1609 * testsuite/libffi.call/cls_align_sint64.c: Likewise.
1610 * testsuite/libffi.call/cls_align_uint32.c: Likewise.
1611 * testsuite/libffi.call/cls_align_sint32.c: Likewise.
1612 * testsuite/libffi.call/cls_align_uint16.c: Likewise.
1613 * testsuite/libffi.call/cls_align_sint16.c: Likewise.
1614 * testsuite/libffi.call/cls_align_float.c: Likewise.
1615 * testsuite/libffi.call/cls_align_double.c: Likewise.
1616 * testsuite/libffi.call/cls_align_longdouble.c: Likewise.
1617 * testsuite/libffi.call/cls_align_pointer.c: Likewise.
1618
f90351a5 16192003-12-02 Hosaka Yuji <hos@tamanegi.org>
1620
1621 PR other/13221
1622 * src/x86/ffi.c (ffi_prep_args, ffi_prep_incoming_args_SYSV):
1623 Align arguments to 32 bits.
1624
1c4a2a48 16252003-12-01 Andreas Tobler <a.tobler@schweiz.ch>
1626
1627 PR other/13221
1628 * testsuite/libffi.call/cls_multi_sshort.c: New test case.
1629 * testsuite/libffi.call/cls_multi_sshortchar.c: Likewise.
1630 * testsuite/libffi.call/cls_multi_uchar.c: Likewise.
1631 * testsuite/libffi.call/cls_multi_schar.c: Likewise.
1632 * testsuite/libffi.call/cls_multi_ushortchar.c: Likewise.
1633 * testsuite/libffi.call/cls_multi_ushort.c: Likewise.
1634
1635 * testsuite/libffi.special/unwindtest.cc: Cosmetics.
1636
8c899267 16372003-11-26 Kaveh R. Ghazi <ghazi@caip.rutgers.edu>
1638
1639 * testsuite/libffi.call/ffitest.h: Include <fcntl.h>.
1640 * testsuite/libffi.special/ffitestcxx.h: Likewise.
1641
857794c9 16422003-11-22 Andreas Tobler <a.tobler@schweiz.ch>
1643
1644 * Makefile.in: Rebuilt.
1645 * configure: Likewise.
1646 * testsuite/libffi.special/unwindtest.cc: Convert the mmap to
1647 the right type.
1648
e7f352cc 16492003-11-21 Andreas Jaeger <aj@suse.de>
1650 Andreas Tobler <a.tobler@schweiz.ch>
1651
1652 * acinclude.m4: Add AC_FUNC_MMAP_BLACKLIST.
1653 * configure.in: Call AC_FUNC_MMAP_BLACKLIST.
1654 * Makefile.in: Rebuilt.
1655 * aclocal.m4: Likewise.
1656 * configure: Likewise.
1657 * fficonfig.h.in: Likewise.
1658 * testsuite/lib/libffi-dg.exp: Add include dir.
1659 * testsuite/libffi.call/ffitest.h: Add MMAP definitions.
1660 * testsuite/libffi.special/ffitestcxx.h: Likewise.
1661 * testsuite/libffi.call/closure_fn0.c: Use MMAP functionality
1662 for ffi_closure if available.
1663 * testsuite/libffi.call/closure_fn1.c: Likewise.
1664 * testsuite/libffi.call/closure_fn2.c: Likewise.
1665 * testsuite/libffi.call/closure_fn3.c: Likewise.
1666 * testsuite/libffi.call/closure_fn4.c: Likewise.
1667 * testsuite/libffi.call/closure_fn5.c: Likewise.
1668 * testsuite/libffi.call/cls_12byte.c: Likewise.
1669 * testsuite/libffi.call/cls_16byte.c: Likewise.
1670 * testsuite/libffi.call/cls_18byte.c: Likewise.
1671 * testsuite/libffi.call/cls_19byte.c: Likewise.
1672 * testsuite/libffi.call/cls_1_1byte.c: Likewise.
1673 * testsuite/libffi.call/cls_20byte.c: Likewise.
1674 * testsuite/libffi.call/cls_20byte1.c: Likewise.
1675 * testsuite/libffi.call/cls_24byte.c: Likewise.
1676 * testsuite/libffi.call/cls_2byte.c: Likewise.
1677 * testsuite/libffi.call/cls_3_1byte.c: Likewise.
1678 * testsuite/libffi.call/cls_3byte1.c: Likewise.
1679 * testsuite/libffi.call/cls_3byte2.c: Likewise.
1680 * testsuite/libffi.call/cls_4_1byte.c: Likewise.
1681 * testsuite/libffi.call/cls_4byte.c: Likewise.
1682 * testsuite/libffi.call/cls_5byte.c: Likewise.
1683 * testsuite/libffi.call/cls_64byte.c: Likewise.
1684 * testsuite/libffi.call/cls_6byte.c: Likewise.
1685 * testsuite/libffi.call/cls_7byte.c: Likewise.
1686 * testsuite/libffi.call/cls_8byte.c: Likewise.
1687 * testsuite/libffi.call/cls_9byte1.c: Likewise.
1688 * testsuite/libffi.call/cls_9byte2.c: Likewise.
1689 * testsuite/libffi.call/cls_double.c: Likewise.
1690 * testsuite/libffi.call/cls_float.c: Likewise.
1691 * testsuite/libffi.call/cls_schar.c: Likewise.
1692 * testsuite/libffi.call/cls_sint.c: Likewise.
1693 * testsuite/libffi.call/cls_sshort.c: Likewise.
1694 * testsuite/libffi.call/cls_uchar.c: Likewise.
1695 * testsuite/libffi.call/cls_uint.c: Likewise.
1696 * testsuite/libffi.call/cls_ulonglong.c: Likewise.
1697 * testsuite/libffi.call/cls_ushort.c: Likewise.
1698 * testsuite/libffi.call/nested_struct.c: Likewise.
1699 * testsuite/libffi.call/nested_struct1.c: Likewise.
1700 * testsuite/libffi.call/nested_struct2.c: Likewise.
1701 * testsuite/libffi.call/nested_struct3.c: Likewise.
1702 * testsuite/libffi.call/problem1.c: Likewise.
1703 * testsuite/libffi.special/unwindtest.cc: Likewise.
1704
71e96ac5 17052003-11-20 Andreas Tobler <a.tobler@schweiz.ch>
1706
1707 * testsuite/lib/libffi-dg.exp: Make the -lgcc_s conditional.
1708
fd0a337b 17092003-11-19 Andreas Tobler <a.tobler@schweiz.ch>
1710
1711 * testsuite/lib/libffi-dg.exp: Add DYLD_LIBRARY_PATH for darwin.
1712 Add -lgcc_s to additional flags.
1713
c2dfd8f6 17142003-11-12 Andreas Tobler <a.tobler@schweiz.ch>
1715
1716 * configure.in, include/Makefile.am: PR libgcj/11147, install
1717 the ffitarget.h header file in a gcc versioned and target
1718 dependent place.
1719 * configure: Regenerated.
1720 * Makefile.in, include/Makefile.in: Likewise.
1721 * testsuite/Makefile.in: Likewise.
1722
17c9ab63 17232003-11-09 Andreas Tobler <a.tobler@schweiz.ch>
1724
1725 * testsuite/libffi.call/closure_fn0.c: Print result and check
1726 with dg-output to make debugging easier.
1727 * testsuite/libffi.call/closure_fn1.c: Likewise.
1728 * testsuite/libffi.call/closure_fn2.c: Likewise.
1729 * testsuite/libffi.call/closure_fn3.c: Likewise.
1730 * testsuite/libffi.call/closure_fn4.c: Likewise.
1731 * testsuite/libffi.call/closure_fn5.c: Likewise.
1732 * testsuite/libffi.call/cls_12byte.c: Likewise.
1733 * testsuite/libffi.call/cls_16byte.c: Likewise.
1734 * testsuite/libffi.call/cls_18byte.c: Likewise.
1735 * testsuite/libffi.call/cls_19byte.c: Likewise.
1736 * testsuite/libffi.call/cls_1_1byte.c: Likewise.
1737 * testsuite/libffi.call/cls_20byte.c: Likewise.
1738 * testsuite/libffi.call/cls_20byte1.c: Likewise.
1739 * testsuite/libffi.call/cls_24byte.c: Likewise.
1740 * testsuite/libffi.call/cls_2byte.c: Likewise.
1741 * testsuite/libffi.call/cls_3_1byte.c: Likewise.
1742 * testsuite/libffi.call/cls_3byte1.c: Likewise.
1743 * testsuite/libffi.call/cls_3byte2.c: Likewise.
1744 * testsuite/libffi.call/cls_4_1byte.c: Likewise.
1745 * testsuite/libffi.call/cls_4byte.c: Likewise.
1746 * testsuite/libffi.call/cls_5byte.c: Likewise.
1747 * testsuite/libffi.call/cls_64byte.c: Likewise.
1748 * testsuite/libffi.call/cls_6byte.c: Likewise.
1749 * testsuite/libffi.call/cls_7byte.c: Likewise.
1750 * testsuite/libffi.call/cls_8byte.c: Likewise.
1751 * testsuite/libffi.call/cls_9byte1.c: Likewise.
1752 * testsuite/libffi.call/cls_9byte2.c: Likewise.
1753 * testsuite/libffi.call/cls_double.c: Likewise.
1754 * testsuite/libffi.call/cls_float.c: Likewise.
1755 * testsuite/libffi.call/cls_schar.c: Likewise.
1756 * testsuite/libffi.call/cls_sint.c: Likewise.
1757 * testsuite/libffi.call/cls_sshort.c: Likewise.
1758 * testsuite/libffi.call/cls_uchar.c: Likewise.
1759 * testsuite/libffi.call/cls_uint.c: Likewise.
1760 * testsuite/libffi.call/cls_ulonglong.c: Likewise.
1761 * testsuite/libffi.call/cls_ushort.c: Likewise.
1762 * testsuite/libffi.call/problem1.c: Likewise.
1763
1764 * testsuite/libffi.special/unwindtest.cc: Make ffi_closure
1765 static.
1766
a0b53ce0 17672003-11-08 Andreas Tobler <a.tobler@schweiz.ch>
1768
1769 * testsuite/libffi.call/cls_9byte2.c: New test case.
1770 * testsuite/libffi.call/cls_9byte1.c: Likewise.
1771 * testsuite/libffi.call/cls_64byte.c: Likewise.
1772 * testsuite/libffi.call/cls_20byte1.c: Likewise.
1773 * testsuite/libffi.call/cls_19byte.c: Likewise.
1774 * testsuite/libffi.call/cls_18byte.c: Likewise.
1775 * testsuite/libffi.call/closure_fn4.c: Likewise.
1776 * testsuite/libffi.call/closure_fn5.c: Likewise.
1777 * testsuite/libffi.call/cls_schar.c: Likewise.
1778 * testsuite/libffi.call/cls_sint.c: Likewise.
1779 * testsuite/libffi.call/cls_sshort.c: Likewise.
1780 * testsuite/libffi.call/nested_struct2.c: Likewise.
1781 * testsuite/libffi.call/nested_struct3.c: Likewise.
1782
ed0ed2e9 17832003-11-08 Andreas Tobler <a.tobler@schweiz.ch>
1784
1785 * testsuite/libffi.call/cls_double.c: Do a check on the result.
1786 * testsuite/libffi.call/cls_uchar.c: Likewise.
1787 * testsuite/libffi.call/cls_uint.c: Likewise.
1788 * testsuite/libffi.call/cls_ulonglong.c: Likewise.
1789 * testsuite/libffi.call/cls_ushort.c: Likewise.
1790 * testsuite/libffi.call/return_sc.c: Cleanup whitespaces.
1791
61641cc0 17922003-11-06 Andreas Tobler <a.tobler@schweiz.ch>
1793
1794 * src/prep_cif.c (ffi_prep_cif): Move the validity check after
1795 the initialization.
1796
6a759884 17972003-10-23 Andreas Tobler <a.tobler@schweiz.ch>
1798
61641cc0 1799 * src/java_raw_api.c (ffi_java_ptrarray_to_raw): Replace
6a759884 1800 FFI_ASSERT(FALSE) with FFI_ASSERT(0).
1801
4f388831 18022003-10-22 David Daney <ddaney@avtrex.com>
1803
1804 * src/mips/ffitarget.h: Replace undefined UINT32 and friends with
1805 __attribute__((__mode__(__SI__))) and friends.
1806
b4891aa6 18072003-10-22 Andreas Schwab <schwab@suse.de>
1808
1809 * src/ia64/ffi.c: Replace FALSE/TRUE with false/true.
1810
3d37e05c 18112003-10-21 Andreas Tobler <a.tobler@schweiz.ch>
1812
1813 * configure.in: AC_LINK_FILES(ffitarget.h).
1814 * configure: Regenerate.
1815 * Makefile.in: Likewise.
1816 * include/Makefile.in: Likewise.
1817 * testsuite/Makefile.in: Likewise.
1818 * fficonfig.h.in: Likewise.
1819
18202003-10-21 Paolo Bonzini <bonzini@gnu.org>
1821 Richard Henderson <rth@redhat.com>
1822
1823 Avoid that ffi.h includes fficonfig.h.
1824
1825 * Makefile.am (EXTRA_DIST): Include ffitarget.h files
1826 (TARGET_SRC_MIPS_GCC): Renamed to TARGET_SRC_MIPS_IRIX.
1827 (TARGET_SRC_MIPS_SGI): Removed.
1828 (MIPS_GCC): Renamed to TARGET_SRC_MIPS_IRIX.
1829 (MIPS_SGI): Removed.
1830 (CLEANFILES): Removed.
1831 (mostlyclean-am, clean-am, mostlyclean-sub, clean-sub): New
1832 targets.
1833 * acconfig.h: Removed.
1834 * configure.in: Compute sizeofs only for double and long double.
1835 Use them to define and subst HAVE_LONG_DOUBLE. Include comments
1836 into AC_DEFINE instead of using acconfig.h. Create
1837 include/ffitarget.h instead of include/fficonfig.h. Rename
1838 MIPS_GCC to MIPS_IRIX, drop MIPS_SGI since we are in gcc's tree.
1839 AC_DEFINE EH_FRAME_FLAGS.
1840 * include/Makefile.am (DISTCLEANFILES): New automake macro.
1841 (hack_DATA): Add ffitarget.h.
1842 * include/ffi.h.in: Remove all system specific definitions.
1843 Declare raw API even if it is not installed, why bother?
1844 Use limits.h instead of SIZEOF_* to define ffi_type_*. Do
1845 not define EH_FRAME_FLAGS, it is in fficonfig.h now. Include
1846 ffitarget.h instead of fficonfig.h. Remove ALIGN macro.
1847 (UINT_ARG, INT_ARG): Removed, use ffi_arg and ffi_sarg instead.
1848 * include/ffi_common.h (bool): Do not define.
1849 (ffi_assert): Accept failed assertion.
1850 (ffi_type_test): Return void and accept file/line.
1851 (FFI_ASSERT): Pass stringized failed assertion.
1852 (FFI_ASSERT_AT): New macro.
1853 (FFI_ASSERT_VALID_TYPE): New macro.
1854 (UINT8, SINT8, UINT16, SINT16, UINT32, SINT32,
1855 UINT64, SINT64): Define here with gcc's __attribute__ macro
1856 instead of in ffi.h
1857 (FLOAT32, ALIGN): Define here instead of in ffi.h
1858 * include/ffi-mips.h: Removed. Its content moved to
1859 src/mips/ffitarget.h after separating assembly and C sections.
1860 * src/alpha/ffi.c, src/alpha/ffi.c, src/java_raw_api.c
1861 src/prep_cif.c, src/raw_api.c, src/ia64/ffi.c,
1862 src/mips/ffi.c, src/mips/n32.S, src/mips/o32.S,
1863 src/mips/ffitarget.h, src/sparc/ffi.c, src/x86/ffi64.c:
1864 SIZEOF_ARG -> FFI_SIZEOF_ARG.
1865 * src/ia64/ffi.c: Include stdbool.h (provided by GCC 2.95+).
1866 * src/debug.c (ffi_assert): Accept stringized failed assertion.
1867 (ffi_type_test): Rewritten.
1868 * src/prep-cif.c (initialize_aggregate, ffi_prep_cif): Call
1869 FFI_ASSERT_VALID_TYPE.
1870 * src/alpha/ffitarget.h, src/arm/ffitarget.h,
1871 src/ia64/ffitarget.h, src/m68k/ffitarget.h,
1872 src/mips/ffitarget.h, src/powerpc/ffitarget.h,
1873 src/s390/ffitarget.h, src/sh/ffitarget.h,
1874 src/sh64/ffitarget.h, src/sparc/ffitarget.h,
1875 src/x86/ffitarget.h: New files.
1876 * src/alpha/osf.S, src/arm/sysv.S, src/ia64/unix.S,
1877 src/m68k/sysv.S, src/mips/n32.S, src/mips/o32.S,
1878 src/powerpc/aix.S, src/powerpc/darwin.S,
1879 src/powerpc/ffi_darwin.c, src/powerpc/linux64.S,
1880 src/powerpc/linux64_closure.S, src/powerpc/ppc_closure.S,
1881 src/powerpc/sysv.S, src/s390/sysv.S, src/sh/sysv.S,
1882 src/sh64/sysv.S, src/sparc/v8.S, src/sparc/v9.S,
1883 src/x86/sysv.S, src/x86/unix64.S, src/x86/win32.S:
1884 include fficonfig.h
1885
2a5f9ffc 18862003-10-20 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
1887
1888 * src/mips/ffi.c: Use _ABIN32, _ABIO32 instead of external
1889 _MIPS_SIM_NABI32, _MIPS_SIM_ABI32.
1890
51c5f96c 18912003-10-19 Andreas Tobler <a.tobler@schweiz.ch>
1892
1893 * src/powerpc/ffi_darwin.c (ffi_prep_args): Declare bytes again.
1894 Used when FFI_DEBUG = 1.
1895
ee46e78f 18962003-10-14 Alan Modra <amodra@bigpond.net.au>
1897
1898 * src/types.c (double, longdouble): Default POWERPC64 to 8 byte size
1899 and align.
1900
a3abe7fe 19012003-10-06 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
1902
1903 * include/ffi_mips.h: Define FFI_MIPS_N32 for N32/N64 ABIs,
1904 FFI_MIPS_O32 for O32 ABI.
1905
1ee4561e 19062003-10-01 Andreas Tobler <a.tobler@schweiz.ch>
1907
1908 * testsuite/lib/libffi-dg.exp: Set LD_LIBRARY_PATH_64 for
1909 SPARC64. Cleanup whitespaces.
1910
93421fe6 19112003-09-19 Andreas Tobler <a.tobler@schweiz.ch>
1912
1913 * testsuite/libffi.call/closure_fn0.c: Xfail mips, arm,
1914 strongarm, xscale. Cleanup whitespaces.
1915 * testsuite/libffi.call/closure_fn1.c: Likewise.
1916 * testsuite/libffi.call/closure_fn2.c: Likewise.
1917 * testsuite/libffi.call/closure_fn3.c: Likewise.
1918 * testsuite/libffi.call/cls_12byte.c: Likewise.
1919 * testsuite/libffi.call/cls_16byte.c: Likewise.
1920 * testsuite/libffi.call/cls_1_1byte.c: Likewise.
1921 * testsuite/libffi.call/cls_20byte.c: Likewise.
1922 * testsuite/libffi.call/cls_24byte.c: Likewise.
1923 * testsuite/libffi.call/cls_2byte.c: Likewise.
1924 * testsuite/libffi.call/cls_3_1byte.c: Likewise.
1925 * testsuite/libffi.call/cls_3byte1.c: Likewise.
1926 * testsuite/libffi.call/cls_3byte2.c: Likewise.
1927 * testsuite/libffi.call/cls_4_1byte.c: Likewise.
1928 * testsuite/libffi.call/cls_4byte.c: Likewise.
1929 * testsuite/libffi.call/cls_5byte.c: Likewise.
1930 * testsuite/libffi.call/cls_6byte.c: Likewise.
1931 * testsuite/libffi.call/cls_7byte.c: Likewise.
1932 * testsuite/libffi.call/cls_8byte.c: Likewise.
1933 * testsuite/libffi.call/cls_double.c: Likewise.
1934 * testsuite/libffi.call/cls_float.c: Likewise.
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/nested_struct.c: Likewise.
1940 * testsuite/libffi.call/nested_struct1.c: Likewise.
1941 * testsuite/libffi.call/problem1.c: Likewise.
1942 * testsuite/libffi.special/unwindtest.cc: Likewise.
1943 * testsuite/libffi.call/pyobjc-tc.c: Cleanup whitespaces.
1944
8a94dd75 19452003-09-18 David Edelsohn <edelsohn@gnu.org>
1946
1947 * src/powerpc/aix.S: Cleanup whitespaces.
1948 * src/powerpc/aix_closure.S: Likewise.
1949
bd1f3533 19502003-09-18 Andreas Tobler <a.tobler@schweiz.ch>
1951
1952 * src/powerpc/darwin.S: Cleanup whitespaces, comment formatting.
1953 * src/powerpc/darwin_closure.S: Likewise.
1954 * src/powerpc/ffi_darwin.c: Likewise.
1955
7a142533 19562003-09-18 Andreas Tobler <a.tobler@schweiz.ch>
1957 David Edelsohn <edelsohn@gnu.org>
1958
1959 * src/types.c (double): Add AIX and Darwin to the right TYPEDEF.
1960 * src/powerpc/aix_closure.S: Remove the pointer to the outgoing
1961 parameter stack.
1962 * src/powerpc/darwin_closure.S: Likewise.
1963 * src/powerpc/ffi_darwin.c (ffi_prep_args): Handle structures
1964 according to the Darwin/AIX ABI.
1965 (ffi_prep_cif_machdep): Likewise.
1966 (ffi_closure_helper_DARWIN): Likewise.
1967 Remove the outgoing parameter stack logic. Simplify the evaluation
1968 of the different CASE types.
1969 (ffi_prep_clousure): Avoid the casts on lvalues. Change the branch
1970 statement in the trampoline code.
1971
d80727ae 19722003-09-18 Kaz Kojima <kkojima@gcc.gnu.org>
1973
1974 * src/sh/ffi.c (ffi_prep_args): Take account into the alignement
1975 for the register size.
1976 (ffi_closure_helper_SYSV): Handle the structure return value
1977 address correctly.
1978 (ffi_closure_helper_SYSV): Return the appropriate type when
1979 the registers are used for the structure return value.
1980 * src/sh/sysv.S (ffi_closure_SYSV): Fix the stack layout for
1981 the 64-bit return value. Update copyright years.
1982
47fab6c3 19832003-09-17 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
1984
1985 * testsuite/lib/libffi-dg.exp (libffi_target_compile): Search in
1986 srcdir for ffi_mips.h.
1987
f397026a 19882003-09-12 Alan Modra <amodra@bigpond.net.au>
1989
1990 * src/prep_cif.c (initialize_aggregate): Include tail padding in
1991 structure size.
1992 * src/powerpc/linux64_closure.S (ffi_closure_LINUX64): Correct
1993 placement of float result.
1994 * testsuite/libffi.special/unwindtest.cc (closure_test_fn1): Correct
1995 cast of "resp" for big-endian 64 bit machines.
1996
341b7dcf 19972003-09-11 Alan Modra <amodra@bigpond.net.au>
1998
1999 * src/types.c (double, longdouble): Merge identical SH and ARM
2000 typedefs, and add POWERPC64.
2001 * src/powerpc/ffi.c (ffi_prep_args64): Correct next_arg calc for
2002 struct split over gpr and rest.
2003 (ffi_prep_cif_machdep): Correct intarg_count for structures.
2004 * src/powerpc/linux64.S (ffi_call_LINUX64): Fix gpr offsets.
2005
06855f3a 20062003-09-09 Andreas Tobler <a.tobler@schweiz.ch>
2007
2008 * src/powerpc/ffi.c (ffi_closure_helper_SYSV) Handle struct
2009 passing correctly.
c886ae3f 2010
9165b0af 20112003-09-09 Alan Modra <amodra@bigpond.net.au>
2012
2013 * configure: Regenerate.
2014
98d58335 20152003-09-04 Andreas Tobler <a.tobler@schweiz.ch>
2016
2017 * Makefile.am: Remove build rules for ffitest.
2018 * Makefile.in: Rebuilt.
c886ae3f 2019
98d58335 20202003-09-04 Andreas Tobler <a.tobler@schweiz.ch>
0262ee1f 2021
2022 * src/java_raw_api.c: Include <stdlib.h> to fix compiler warning
2023 about implicit declaration of abort().
2024
3743c332 20252003-09-04 Andreas Tobler <a.tobler@schweiz.ch>
2026
2027 * Makefile.am: Add dejagnu test framework. Fixes PR other/11411.
2028 * Makefile.in: Rebuilt.
2029 * configure.in: Add dejagnu test framework.
2030 * configure: Rebuilt.
2031
2032 * testsuite/Makefile.am: New file.
2033 * testsuite/Makefile.in: Built
c886ae3f 2034 * testsuite/lib/libffi-dg.exp: New file.
3743c332 2035 * testsuite/config/default.exp: Likewise.
2036 * testsuite/libffi.call/call.exp: Likewise.
2037 * testsuite/libffi.call/ffitest.h: Likewise.
2038 * testsuite/libffi.call/closure_fn0.c: Likewise.
2039 * testsuite/libffi.call/closure_fn1.c: Likewise.
2040 * testsuite/libffi.call/closure_fn2.c: Likewise.
2041 * testsuite/libffi.call/closure_fn3.c: Likewise.
2042 * testsuite/libffi.call/cls_1_1byte.c: Likewise.
2043 * testsuite/libffi.call/cls_3_1byte.c: Likewise.
2044 * testsuite/libffi.call/cls_4_1byte.c: Likewise.
2045 * testsuite/libffi.call/cls_2byte.c: Likewise.
2046 * testsuite/libffi.call/cls_3byte1.c: Likewise.
2047 * testsuite/libffi.call/cls_3byte2.c: Likewise.
2048 * testsuite/libffi.call/cls_4byte.c: Likewise.
2049 * testsuite/libffi.call/cls_5byte.c: Likewise.
2050 * testsuite/libffi.call/cls_6byte.c: Likewise.
2051 * testsuite/libffi.call/cls_7byte.c: Likewise.
2052 * testsuite/libffi.call/cls_8byte.c: Likewise.
2053 * testsuite/libffi.call/cls_12byte.c: Likewise.
2054 * testsuite/libffi.call/cls_16byte.c: Likewise.
2055 * testsuite/libffi.call/cls_20byte.c: Likewise.
2056 * testsuite/libffi.call/cls_24byte.c: Likewise.
2057 * testsuite/libffi.call/cls_double.c: Likewise.
2058 * testsuite/libffi.call/cls_float.c: Likewise.
2059 * testsuite/libffi.call/cls_uchar.c: Likewise.
2060 * testsuite/libffi.call/cls_uint.c: Likewise.
2061 * testsuite/libffi.call/cls_ulonglong.c: Likewise.
2062 * testsuite/libffi.call/cls_ushort.c: Likewise.
2063 * testsuite/libffi.call/float.c: Likewise.
2064 * testsuite/libffi.call/float1.c: Likewise.
2065 * testsuite/libffi.call/float2.c: Likewise.
2066 * testsuite/libffi.call/many.c: Likewise.
2067 * testsuite/libffi.call/many_win32.c: Likewise.
2068 * testsuite/libffi.call/nested_struct.c: Likewise.
2069 * testsuite/libffi.call/nested_struct1.c: Likewise.
2070 * testsuite/libffi.call/pyobjc-tc.c: Likewise.
2071 * testsuite/libffi.call/problem1.c: Likewise.
2072 * testsuite/libffi.call/promotion.c: Likewise.
2073 * testsuite/libffi.call/return_ll.c: Likewise.
2074 * testsuite/libffi.call/return_sc.c: Likewise.
2075 * testsuite/libffi.call/return_uc.c: Likewise.
2076 * testsuite/libffi.call/strlen.c: Likewise.
2077 * testsuite/libffi.call/strlen_win32.c: Likewise.
2078 * testsuite/libffi.call/struct1.c: Likewise.
2079 * testsuite/libffi.call/struct2.c: Likewise.
2080 * testsuite/libffi.call/struct3.c: Likewise.
2081 * testsuite/libffi.call/struct4.c: Likewise.
2082 * testsuite/libffi.call/struct5.c: Likewise.
2083 * testsuite/libffi.call/struct6.c: Likewise.
2084 * testsuite/libffi.call/struct7.c: Likewise.
2085 * testsuite/libffi.call/struct8.c: Likewise.
c886ae3f 2086 * testsuite/libffi.call/struct9.c: Likewise.
3743c332 2087 * testsuite/libffi.special/special.exp: New file.
2088 * testsuite/libffi.special/ffitestcxx.h: Likewise.
2089 * testsuite/libffi.special/unwindtest.cc: Likewise.
2090
c886ae3f 2091
3a845f75 20922003-08-13 Kaz Kojima <kkojima@gcc.gnu.org>
2093
2094 * src/sh/ffi.c (OFS_INT16): Set 0 for little endian case. Update
2095 copyright years.
2096
4637dfa6 20972003-08-02 Alan Modra <amodra@bigpond.net.au>
2098
2099 * src/powerpc/ffi.c (ffi_prep_args64): Modify for changed gcc
2100 structure passing.
2101 (ffi_closure_helper_LINUX64): Likewise.
2102 * src/powerpc/linux64.S: Remove code writing to parm save area.
2103 * src/powerpc/linux64_closure.S (ffi_closure_LINUX64): Use return
2104 address in lr from ffi_closure_helper_LINUX64 call to calculate
2105 table address. Optimize function tail.
2106
8bcdc7f0 21072003-07-28 Andreas Tobler <a.tobler@schweiz.ch>
2108
2109 * src/sparc/ffi.c: Handle all floating point registers.
2110 * src/sparc/v9.S: Likewise. Fixes second part of PR target/11410.
2111
3d6cc549 21122003-07-11 Gerald Pfeifer <pfeifer@dbai.tuwien.ac.at>
2113
2114 * README: Note that libffi is not part of GCC. Update the project
2115 URL and status.
2116
9b770b5a 21172003-06-19 Franz Sirl <Franz.Sirl-kernel@lauterbach.com>
2118
2119 * src/powerpc/ppc_closure.S: Include ffi.h.
2120
d46ab9f3 21212003-06-13 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
2122
2123 * src/x86/sysv.S: Avoid gas-only .uleb128/.sleb128 directives.
2124 Use C style comments.
2125
81f536a3 21262003-06-13 Kaz Kojima <kkojima@rr.iij4u.or.jp>
2127
2128 * Makefile.am: Add SHmedia support. Fix a typo of SH support.
2129 * Makefile.in: Regenerate.
2130 * configure.in (sh64-*-linux*, sh5*-*-linux*): Add target.
2131 * configure: Regenerate.
2132 * include/ffi.h.in: Add SHmedia support.
2133 * src/sh64/ffi.c: New file.
2134 * src/sh64/sysv.S: New file.
2135
1da888ec 21362003-05-16 Jakub Jelinek <jakub@redhat.com>
2137
2138 * configure.in (HAVE_RO_EH_FRAME): Check whether .eh_frame section
2139 should be read-only.
2140 * configure: Rebuilt.
2141 * fficonfig.h.in: Rebuilt.
2142 * include/ffi.h.in (EH_FRAME_FLAGS): Define.
2143 * src/alpha/osf.S: Use EH_FRAME_FLAGS.
2144 * src/powerpc/linux64.S: Likewise.
2145 * src/powerpc/linux64_closure.S: Likewise. Include ffi.h.
2146 * src/powerpc/sysv.S: Use EH_FRAME_FLAGS. Use pcrel encoding
2147 if -fpic/-fPIC/-mrelocatable.
2148 * src/powerpc/powerpc_closure.S: Likewise.
2149 * src/sparc/v8.S: If HAVE_RO_EH_FRAME is defined, don't include
2150 #write in .eh_frame flags.
2151 * src/sparc/v9.S: Likewise.
2152 * src/x86/unix64.S: Use EH_FRAME_FLAGS.
2153 * src/x86/sysv.S: Likewise. Use pcrel encoding if -fpic/-fPIC.
2154 * src/s390/sysv.S: Use EH_FRAME_FLAGS. Include ffi.h.
2155
f3bd7a2c 21562003-05-07 Jeff Sturm <jsturm@one-point.com>
2157
2158 Fixes PR bootstrap/10656
2159 * configure.in (HAVE_AS_REGISTER_PSEUDO_OP): Test assembler
2160 support for .register pseudo-op.
2161 * src/sparc/v8.S: Use it.
2162 * fficonfig.h.in: Rebuilt.
2163 * configure: Rebuilt.
2164
3647950d 21652003-04-18 Jakub Jelinek <jakub@redhat.com>
2166
2167 * include/ffi.h.in (POWERPC64): Define if 64-bit.
2168 (enum ffi_abi): Add FFI_LINUX64 on POWERPC.
2169 Make it the default on POWERPC64.
2170 (FFI_TRAMPOLINE_SIZE): Define to 24 on POWERPC64.
2171 * configure.in: Change powerpc-*-linux* into powerpc*-*-linux*.
2172 * configure: Rebuilt.
2173 * src/powerpc/ffi.c (hidden): Define.
2174 (ffi_prep_args_SYSV): Renamed from
2175 ffi_prep_args. Cast pointers to unsigned long to shut up warnings.
2176 (NUM_GPR_ARG_REGISTERS64, NUM_FPR_ARG_REGISTERS64,
2177 ASM_NEEDS_REGISTERS64): New.
2178 (ffi_prep_args64): New function.
2179 (ffi_prep_cif_machdep): Handle FFI_LINUX64 ABI.
2180 (ffi_call): Likewise.
2181 (ffi_prep_closure): Likewise.
2182 (flush_icache): Surround by #ifndef POWERPC64.
2183 (ffi_dblfl): New union type.
2184 (ffi_closure_helper_SYSV): Use it to avoid aliasing problems.
2185 (ffi_closure_helper_LINUX64): New function.
2186 * src/powerpc/ppc_closure.S: Surround whole file by #ifndef
2187 __powerpc64__.
2188 * src/powerpc/sysv.S: Likewise.
2189 (ffi_call_SYSV): Rename ffi_prep_args to ffi_prep_args_SYSV.
2190 * src/powerpc/linux64.S: New file.
2191 * src/powerpc/linux64_closure.S: New file.
2192 * Makefile.am (EXTRA_DIST): Add src/powerpc/linux64.S and
2193 src/powerpc/linux64_closure.S.
2194 (TARGET_SRC_POWERPC): Likewise.
2195
2196 * src/ffitest.c (closure_test_fn, closure_test_fn1, closure_test_fn2,
2197 closure_test_fn3): Fix result printing on big-endian 64-bit
2198 machines.
2199 (main): Print tst2_arg instead of uninitialized tst2_result.
2200
2201 * src/ffitest.c (main): Hide what closure pointer really points to
2202 from the compiler.
2203
1636e920 22042003-04-16 Richard Earnshaw <rearnsha@arm.com>
2205
2206 * configure.in (arm-*-netbsdelf*): Add configuration.
2207 (configure): Regenerated.
2208
e663564a 22092003-04-04 Loren J. Rittle <ljrittle@acm.org>
2210
2211 * include/Makefile.in: Regenerate.
2212
02f5d4c9 22132003-03-21 Zdenek Dvorak <rakdver@atrey.karlin.mff.cuni.cz>
2214
2215 * libffi/include/ffi.h.in: Define X86 instead of X86_64 in 32
2216 bit mode.
2217 * libffi/src/x86/ffi.c (ffi_closure_SYSV, ffi_closure_raw_SYSV):
2218 Receive closure pointer through parameter, read args using
2219 __builtin_dwarf_cfa.
2220 (FFI_INIT_TRAMPOLINE): Send closure reference through eax.
2221
b8e27232 22222003-03-12 Andreas Schwab <schwab@suse.de>
2223
2224 * configure.in: Avoid trailing /. in toolexeclibdir.
2225 * configure: Rebuilt.
2226
5fee68c6 22272003-03-03 Andreas Tobler <a.tobler@schweiz.ch>
2228
bbb07d35 2229 * src/powerpc/darwin_closure.S: Recode to fit dynamic libraries.
5fee68c6 2230
ecadab3f 22312003-02-06 Andreas Tobler <a.tobler@schweiz.ch>
2232
2233 * libffi/src/powerpc/darwin_closure.S:
2234 Fix alignement bug, allocate 8 bytes for the result.
c886ae3f 2235 * libffi/src/powerpc/aix_closure.S:
ecadab3f 2236 Likewise.
2237 * libffi/src/powerpc/ffi_darwin.c:
2238 Update stackframe description for aix/darwin_closure.S.
c886ae3f 2239
a0294641 22402003-02-06 Jakub Jelinek <jakub@redhat.com>
2241
2242 * src/s390/ffi.c (ffi_closure_helper_SYSV): Add hidden visibility
2243 attribute.
2244
624476ed 22452003-01-31 Christian Cornelssen <ccorn@cs.tu-berlin.de>,
2246 Andreas Schwab <schwab@suse.de>
2247
bbb07d35 2248 * configure.in: Adjust command to source config-ml.in to account
624476ed 2249 for changes to the libffi_basedir definition.
2250 (libffi_basedir): Remove ${srcdir} from value and include trailing
2251 slash if nonempty.
c886ae3f 2252
bbb07d35 2253 * configure: Regenerate.
624476ed 2254
1143cc9b 22552003-01-29 Franz Sirl <Franz.Sirl-kernel@lauterbach.com>
2256
2257 * src/powerpc/ppc_closure.S: Recode to fit shared libs.
2258
967ae8fb 22592003-01-28 Andrew Haley <aph@redhat.com>
2260
2261 * include/ffi.h.in: Enable FFI_CLOSURES for x86_64.
2262 * src/x86/ffi64.c (ffi_prep_closure): New.
2263 (ffi_closure_UNIX64_inner): New.
2264 * src/x86/unix64.S (ffi_closure_UNIX64): New.
2265
c8d4fa8c 22662003-01-27 Alexandre Oliva <aoliva@redhat.com>
2267
2268 * configure.in (toolexecdir, toolexeclibdir): Set and AC_SUBST.
2269 Remove USE_LIBDIR conditional.
2270 * Makefile.am (toolexecdir, toolexeclibdir): Don't override.
2271 * Makefile.in, configure: Rebuilt.
2272
2dd7852c 22732003-01027 David Edelsohn <edelsohn@gnu.org>
2274
2275 * Makefile.am (TARGET_SRC_POWERPC_AIX): Fix typo.
2276 * Makefile.in: Regenerate.
2277
008eadcd 22782003-01-22 Andrew Haley <aph@redhat.com>
2279
2280 * src/powerpc/darwin.S (_ffi_call_AIX): Add Augmentation size to
2281 unwind info.
2282
cb440058 22832003-01-21 Andreas Tobler <a.tobler@schweiz.ch>
2284
2285 * src/powerpc/darwin.S: Add unwind info.
2286 * src/powerpc/darwin_closure.S: Likewise.
2287
149417bc 22882003-01-14 Andrew Haley <aph@redhat.com>
2289
2290 * src/x86/ffi64.c (ffi_prep_args): Check for void retval.
2291 (ffi_prep_cif_machdep): Likewise.
2292 * src/x86/unix64.S: Add unwind info.
2293
d4dcd973 22942003-01-14 Andreas Jaeger <aj@suse.de>
2295
2296 * src/ffitest.c (main): Only use ffi_closures if those are
2297 supported.
2298
a7898d0c 22992003-01-13 Andreas Tobler <a.tobler@schweiz.ch>
2300
2301 * libffi/src/ffitest.c
bbb07d35 2302 add closure testcases
a7898d0c 2303
23042003-01-13 Kevin B. Hendricks <khendricks@ivey.uwo.ca>
2305
2306 * libffi/src/powerpc/ffi.c
bbb07d35 2307 fix alignment bug for float (4 byte aligned iso 8 byte)
a7898d0c 2308
4e844289 23092003-01-09 Geoffrey Keating <geoffk@apple.com>
2310
2311 * src/powerpc/ffi_darwin.c: Remove RCS version string.
2312 * src/powerpc/darwin.S: Remove RCS version string.
2313
0ef3e9c5 23142003-01-03 Jeff Sturm <jsturm@one-point.com>
2315
2316 * include/ffi.h.in: Add closure defines for SPARC, SPARC64.
2317 * src/ffitest.c (main): Use static storage for closure.
2318 * src/sparc/ffi.c (ffi_prep_closure, ffi_closure_sparc_inner): New.
2319 * src/sparc/v8.S (ffi_closure_v8): New.
2320 * src/sparc/v9.S (ffi_closure_v9): New.
2321
acdc3fc3 23222002-11-10 Ranjit Mathew <rmathew@hotmail.com>
2323
2324 * include/ffi.h.in: Added FFI_STDCALL ffi_type
2325 enumeration for X86_WIN32.
2326 * src/x86/win32.S: Added ffi_call_STDCALL function
2327 definition.
2328 * src/x86/ffi.c (ffi_call/ffi_raw_call): Added
2329 switch cases for recognising FFI_STDCALL and
2330 calling ffi_call_STDCALL if target is X86_WIN32.
2331 * src/ffitest.c (my_stdcall_strlen/stdcall_many):
2332 stdcall versions of the "my_strlen" and "many"
2333 test functions (for X86_WIN32).
2334 Added test cases to test stdcall invocation using
2335 these functions.
2336
805e22b2 23372002-12-02 Kaz Kojima <kkojima@gcc.gnu.org>
2338
2339 * src/sh/sysv.S: Add DWARF2 unwind info.
2340
6d219176 23412002-11-27 Ulrich Weigand <uweigand@de.ibm.com>
2342
2343 * src/s390/sysv.S (.eh_frame section): Make section read-only.
2344
4827afc9 23452002-11-26 Jim Wilson <wilson@redhat.com>
2346
2347 * src/types.c (FFI_TYPE_POINTER): Has size 8 on IA64.
2348
805e22b2 23492002-11-23 H.J. Lu <hjl@gnu.org>
2350
2351 * acinclude.m4: Add dummy AM_PROG_LIBTOOL.
2352 Include ../config/accross.m4.
2353 * aclocal.m4; Rebuild.
2354 * configure: Likewise.
2355
e3a04290 23562002-11-15 Ulrich Weigand <uweigand@de.ibm.com>
2357
2358 * src/s390/sysv.S (.eh_frame section): Adapt to pcrel FDE encoding.
2359
be36f77d 23602002-11-11 DJ Delorie <dj@redhat.com>
2361
2362 * configure.in: Look for common files in the right place.
2363
dd985e54 23642002-10-08 Ulrich Weigand <uweigand@de.ibm.com>
2365
2366 * src/java_raw_api.c (ffi_java_raw_to_ptrarray): Interpret
2367 raw data as _Jv_word values, not ffi_raw.
2368 (ffi_java_ptrarray_to_raw): Likewise.
2369 (ffi_java_rvalue_to_raw): New function.
2370 (ffi_java_raw_call): Call it.
2371 (ffi_java_raw_to_rvalue): New function.
2372 (ffi_java_translate_args): Call it.
2373 * src/ffitest.c (closure_test_fn): Interpret return value
2374 as ffi_arg, not int.
2375 * src/s390/ffi.c (ffi_prep_cif_machdep): Add missing
2376 FFI_TYPE_POINTER case.
2377 (ffi_closure_helper_SYSV): Likewise. Also, assume return
2378 values extended to word size.
2379
4a18e5fa 23802002-10-02 Andreas Jaeger <aj@suse.de>
2381
2382 * src/x86/ffi64.c (ffi_prep_cif_machdep): Remove debug output.
2383
3d8ac088 23842002-10-01 Bo Thorsen <bo@smetana.suse.de>
2385
2386 * include/ffi.h.in: Fix i386 win32 compilation.
2387
e0aaecf3 23882002-09-30 Ulrich Weigand <uweigand@de.ibm.com>
2389
2390 * configure.in: Add s390x-*-linux-* target.
2391 * configure: Regenerate.
2392 * include/ffi.h.in: Define S390X for s390x targets.
2393 (FFI_CLOSURES): Define for s390/s390x.
2394 (FFI_TRAMPOLINE_SIZE): Likewise.
2395 (FFI_NATIVE_RAW_API): Likewise.
2396 * src/prep_cif.c (ffi_prep_cif): Do not compute stack space for s390.
2397 * src/types.c (FFI_TYPE_POINTER): Use 8-byte pointers on s390x.
805e22b2 2398 * src/s390/ffi.c: Major rework of existing code. Add support for
e0aaecf3 2399 s390x targets. Add closure support.
2400 * src/s390/sysv.S: Likewise.
2401
c3e9c301 24022002-09-29 Richard Earnshaw <rearnsha@arm.com>
2403
2404 * src/arm/sysv.S: Fix typo.
2405
d13886be 24062002-09-28 Richard Earnshaw <rearnsha@arm.com>
2407
2408 * src/arm/sysv.S: If we don't have machine/asm.h and the pre-processor
2409 has defined __USER_LABEL_PREFIX__, then use it in CNAME.
2410 (ffi_call_SYSV): Handle soft-float.
2411
aaff811e 24122002-09-27 Bo Thorsen <bo@suse.de>
2413
2414 * include/ffi.h.in: Fix multilib x86-64 support.
2415
5f7912d8 24162002-09-22 Kaveh R. Ghazi <ghazi@caip.rutgers.edu>
2417
2418 * Makefile.am (all-multi): Fix multilib parallel build.
2419
ccda5e55 24202002-07-19 Kaz Kojima <kkojima@gcc.gnu.org>
2421
2422 * configure.in (sh[34]*-*-linux*): Add brackets.
2423 * configure: Regenerate.
2424
b1e5d71d 24252002-07-18 Kaz Kojima <kkojima@gcc.gnu.org>
2426
2427 * Makefile.am: Add SH support.
2428 * Makefile.in: Regenerate.
2429 * configure.in (sh-*-linux*, sh[34]*-*-linux*): Add target.
2430 * configure: Regenerate.
2431 * include/ffi.h.in: Add SH support.
2432 * src/sh/ffi.c: New file.
2433 * src/sh/sysv.S: New file.
2434 * src/types.c: Add SH support.
2435
42c71854 24362002-07-16 Bo Thorsen <bo@suse.de>
2437
2438 * src/x86/ffi64.c: New file that adds x86-64 support.
2439 * src/x86/unix64.S: New file that handles argument setup for
2440 x86-64.
2441 * src/x86/sysv.S: Don't use this on x86-64.
2442 * src/x86/ffi.c: Don't use this on x86-64.
2443 Remove unused vars.
2444 * src/prep_cif.c (ffi_prep_cif): Don't do stack size calculation
2445 for x86-64.
2446 * src/ffitest.c (struct6): New test that tests a special case in
2447 the x86-64 ABI.
2448 (struct7): Likewise.
2449 (struct8): Likewise.
2450 (struct9): Likewise.
2451 (closure_test_fn): Silence warning about this when it's not used.
2452 (main): Add the new tests.
2453 (main): Fix a couple of wrong casts and silence some compiler warnings.
2454 * include/ffi.h.in: Add x86-64 ABI definition.
2455 * fficonfig.h.in: Regenerate.
2456 * Makefile.am: Add x86-64 support.
2457 * configure.in: Likewise.
2458 * Makefile.in: Regenerate.
2459 * configure: Likewise.
2460
24612002-06-24 Bo Thorsen <bo@suse.de>
2462
2463 * src/types.c: Merge settings for similar architectures.
2464 Add x86-64 sizes and alignments.
2465
24662002-06-23 Bo Thorsen <bo@suse.de>
2467
2468 * src/arm/ffi.c (ffi_prep_args): Remove unused vars.
2469 * src/sparc/ffi.c (ffi_prep_args_v8): Likewise.
2470 * src/mips/ffi.c (ffi_prep_args): Likewise.
2471 * src/m68k/ffi.c (ffi_prep_args): Likewise.
2472
72726889 24732002-07-18 H.J. Lu (hjl@gnu.org)
2474
2475 * Makefile.am (TARGET_SRC_MIPS_LINUX): New.
2476 (libffi_la_SOURCES): Support MIPS_LINUX.
2477 (libffi_convenience_la_SOURCES): Likewise.
2478 * Makefile.in: Regenerated.
2479
2480 * configure.in (mips64*-*): Skip.
2481 (mips*-*-linux*): New.
2482 * configure: Regenerated.
2483
2484 * src/mips/ffi.c: Include <sgidefs.h>.
2485
e17e640b 24862002-06-06 Ulrich Weigand <uweigand@de.ibm.com>
2487
2488 * src/s390/sysv.S: Save/restore %r6. Add DWARF-2 unwind info.
2489
b5d69d52 24902002-05-27 Roger Sayle <roger@eyesopen.com>
2491
2492 * src/x86/ffi.c (ffi_prep_args): Remove reference to avn.
2493
eadca1d5 24942002-05-27 Bo Thorsen <bo@suse.de>
2495
2496 * src/x86/ffi.c (ffi_prep_args): Remove unused variable and
2497 fix formatting.
2498
2abdadc9 24992002-05-13 Andreas Tobler <a.tobler@schweiz.ch>
2500
2501 * src/powerpc/ffi_darwin.c (ffi_prep_closure): Declare fd at
2502 beginning of function (for older apple cc).
2503
0b045cf7 25042002-05-08 Alexandre Oliva <aoliva@redhat.com>
2505
2506 * configure.in (ORIGINAL_LD_FOR_MULTILIBS): Preserve LD at
2507 script entry, and set LD to it when configuring multilibs.
2508 * configure: Rebuilt.
2509
7f0e056d 25102002-05-05 Jason Thorpe <thorpej@wasabisystems.com>
2511
2512 * configure.in (sparc64-*-netbsd*): Add target.
2513 (sparc-*-netbsdelf*): Likewise.
2514 * configure: Regenerate.
2515
b19313f4 25162002-04-28 David S. Miller <davem@redhat.com>
2517
2518 * configure.in, configure: Fix SPARC test in previous change.
2519
ea1f83c5 25202002-04-29 Gerhard Tonn <GerhardTonn@swol.de>
2521
2522 * Makefile.am: Add Linux for S/390 support.
2523 * Makefile.in: Regenerate.
2524 * configure.in: Add Linux for S/390 support.
2525 * configure: Regenerate.
2526 * include/ffi.h.in: Add Linux for S/390 support.
2527 * src/s390/ffi.c: New file from libffi CVS tree.
2528 * src/s390/sysv.S: New file from libffi CVS tree.
2529
b40c762b 25302002-04-28 Jakub Jelinek <jakub@redhat.com>
2531
2532 * configure.in (HAVE_AS_SPARC_UA_PCREL): Check for working
2533 %r_disp32().
2534 * src/sparc/v8.S: Use it.
2535 * src/sparc/v9.S: Likewise.
2536 * fficonfig.h.in: Rebuilt.
2537 * configure: Rebuilt.
2538
00f87a50 25392002-04-08 Hans Boehm <Hans_Boehm@hp.com>
2540
2541 * src/java_raw_api.c (ffi_java_raw_size): Handle FFI_TYPE_DOUBLE
2542 correctly.
2543 * src/ia64/unix.S: Add unwind information. Fix comments.
2544 Save sp in a way that's compatible with unwind info.
2545 (ffi_call_unix): Correctly restore sp in all cases.
2546 * src/ia64/ffi.c: Add, fix comments.
2547
a622ae9d 25482002-04-08 Jakub Jelinek <jakub@redhat.com>
2549
2550 * src/sparc/v8.S: Make .eh_frame dependent on target word size.
2551
c1d86bcd 25522002-04-06 Jason Thorpe <thorpej@wasabisystems.com>
2553
2554 * configure.in (alpha*-*-netbsd*): Add target.
2555 * configure: Regenerate.
2556
3faa4340 25572002-04-04 Jeff Sturm <jsturm@one-point.com>
2558
2559 * src/sparc/v8.S: Add unwind info.
2560 * src/sparc/v9.S: Likewise.
2561
ec66e65b 25622002-03-30 Krister Walfridsson <cato@df.lth.se>
2563
2564 * configure.in: Enable i*86-*-netbsdelf*.
2565 * configure: Rebuilt.
805e22b2 2566
9ee9948b 25672002-03-29 David Billinghurst <David.Billinghurst@riotinto.com>
2568
2569 PR other/2620
2570 * src/mips/n32.s: Delete
2571 * src/mips/o32.s: Delete
2572
6c74b4f2 25732002-03-21 Loren J. Rittle <ljrittle@acm.org>
2574
2575 * configure.in: Enable alpha*-*-freebsd*.
2576 * configure: Rebuilt.
2577
9d0a2dd5 25782002-03-17 Bryce McKinlay <bryce@waitaki.otago.ac.nz>
2579
2580 * Makefile.am: libfficonvenience -> libffi_convenience.
2581 * Makefile.in: Rebuilt.
805e22b2 2582
ee7750c2 2583 * Makefile.am: Define ffitest_OBJECTS.
2584 * Makefile.in: Rebuilt.
9d0a2dd5 2585
f6a16bd8 25862002-03-07 Andreas Tobler <toa@pop.agri.ch>
2587 David Edelsohn <edelsohn@gnu.org>
2588
2589 * Makefile.am (EXTRA_DIST): Add Darwin and AIX closure files.
2590 (TARGET_SRC_POWERPC_AIX): Add aix_closure.S.
2591 (TARGET_SRC_POWERPC_DARWIN): Add darwin_closure.S.
2592 * Makefile.in: Regenerate.
2593 * include/ffi.h.in: Add AIX and Darwin closure definitions.
2594 * src/powerpc/ffi_darwin.c (ffi_prep_closure): New function.
2595 (flush_icache, flush_range): New functions.
2596 (ffi_closure_helper_DARWIN): New function.
2597 * src/powerpc/aix_closure.S: New file.
2598 * src/powerpc/darwin_closure.S: New file.
805e22b2 2599
fed8d18d 26002002-02-24 Jeff Sturm <jsturm@one-point.com>
2601
2602 * include/ffi.h.in: Add typedef for ffi_arg.
2603 * src/ffitest.c (main): Declare rint with ffi_arg.
805e22b2 2604
ddac7e16 26052002-02-21 Andreas Tobler <toa@pop.agri.ch>
2606
2607 * src/powerpc/ffi_darwin.c (ffi_prep_args): Skip appropriate
2608 number of GPRs for floating-point arguments.
2609
ed35b400 26102002-01-31 Anthony Green <green@redhat.com>
2611
2612 * configure: Rebuilt.
2613 * configure.in: Replace CHECK_SIZEOF and endian tests with
2614 cross-compiler friendly macros.
2615 * aclocal.m4 (AC_COMPILE_CHECK_SIZEOF, AC_C_BIGENDIAN_CROSS): New
2616 macros.
2617
bf270768 26182002-01-18 David Edelsohn <edelsohn@gnu.org>
2619
2620 * src/powerpc/darwin.S (_ffi_call_AIX): New.
dc155f85 2621 * src/powerpc/aix.S (ffi_call_DARWIN): New.
bf270768 2622
faae36ce 26232002-01-17 David Edelsohn <edelsohn@gnu.org>
2624
2625 * Makefile.am (EXTRA_DIST): Add Darwin and AIX files.
2626 (TARGET_SRC_POWERPC_AIX): New.
2627 (POWERPC_AIX): New stanza.
2628 * Makefile.in: Regenerate.
2629 * configure.in: Add AIX case.
2630 * configure: Regenerate.
2631 * include/ffi.h.in (ffi_abi): Add FFI_AIX.
2632 * src/powerpc/ffi_darwin.c (ffi_status): Use "long" to scale frame
2633 size. Fix "long double" support.
2634 (ffi_call): Add FFI_AIX case.
2635 * src/powerpc/aix.S: New.
2636
b81d6fe1 26372001-10-09 John Hornkvist <john@toastedmarshmallow.com>
2638
2639 Implement Darwin PowerPC ABI.
2640 * configure.in: Handle powerpc-*-darwin*.
2641 * Makefile.am: Set source files for POWERPC_DARWIN.
2642 * configure: Rebuilt.
2643 * Makefile.in: Rebuilt.
2644 * include/ffi.h.in: Define FFI_DARWIN and FFI_DEFAULT_ABI for
2645 POWERPC_DARWIN.
2646 * src/powerpc/darwin.S: New file.
2647 * src/powerpc/ffi_darwin.c: New file.
2648
9bccb7a3 26492001-10-07 Joseph S. Myers <jsm28@cam.ac.uk>
2650
2651 * src/x86/ffi.c: Fix spelling error of "separate" as "seperate".
2652
ac1f371e 26532001-07-16 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
2654
2655 * src/x86/sysv.S: Avoid gas-only .balign directive.
2656 Use C style comments.
2657
e1e36390 26582001-07-16 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
2659
2660 * src/alpha/ffi.c (ffi_prep_closure): Avoid gas-only mnemonic.
2661 Fixes PR bootstrap/3563.
2662
deb29e7a 26632001-06-26 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
2664
2665 * src/alpha/osf.S (ffi_closure_osf): Use .rdata for ECOFF.
2666
8008de24 26672001-06-25 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
2668
2669 * configure.in: Recognize sparc*-sun-* host.
2670 * configure: Regenerate.
805e22b2 2671
5fc46419 26722001-06-06 Andrew Haley <aph@redhat.com>
2673
968b0d1e 2674 * src/alpha/osf.S (__FRAME_BEGIN__): Conditionalize for ELF.
5fc46419 2675
49ff2ded 26762001-06-03 Andrew Haley <aph@redhat.com>
2677
968b0d1e 2678 * src/alpha/osf.S: Add unwind info.
2679 * src/powerpc/sysv.S: Add unwind info.
2680 * src/powerpc/ppc_closure.S: Likewise.
49ff2ded 2681
070883fe 26822000-05-31 Jeff Sturm <jsturm@one-point.com>
2683
2684 * configure.in: Fix AC_ARG_ENABLE usage.
2685 * configure: Rebuilt.
2686
09da9506 26872001-05-06 Bryce McKinlay <bryce@waitaki.otago.ac.nz>
2688
2689 * configure.in: Remove warning about beta code.
2690 * configure: Rebuilt.
2691
26922001-04-25 Hans Boehm <Hans_Boehm@hp.com>
2693
6c1b002f 2694 * src/ia64/unix.S: Restore stack pointer when returning from
2695 ffi_closure_UNIX.
2696 * src/ia64/ffi.c: Fix typo in comment.
2697
c4312026 26982001-04-18 Jim Wilson <wilson@redhat.com>
2699
2700 * src/ia64/unix.S: Delete unnecessary increment and decrement of loc2
2701 to eliminate RAW DV.
2702
db6ca0af 27032001-04-12 Bryce McKinlay <bryce@albatross.co.nz>
2704
2705 * Makefile.am: Make a libtool convenience library.
2706 * Makefile.in: Rebuilt.
2707
9461b350 27082001-03-29 Bryce McKinlay <bryce@albatross.co.nz>
2709
2710 * configure.in: Use different syntax for subdirectory creation.
2711 * configure: Rebuilt.
2712
a4cc90ca 27132001-03-27 Jon Beniston <jon@beniston.com>
2714
2715 * configure.in: Added X86_WIN32 target (Win32, CygWin, MingW).
2716 * configure: Rebuilt.
2717 * Makefile.am: Added X86_WIN32 target support.
2718 * Makefile.in: Rebuilt.
2719
2720 * include/ffi.h.in: Added X86_WIN32 target support.
2721
2722 * src/ffitest.c: Doesn't run structure tests for X86_WIN32 targets.
2723 * src/types.c: Added X86_WIN32 target support.
2724
2725 * src/x86/win32.S: New file. Based on sysv.S, but with EH
2726 stuff removed and made to work with CygWin's gas.
2727
21f4579d 27282001-03-26 Bryce McKinlay <bryce@albatross.co.nz>
2729
2730 * configure.in: Make target subdirectory in build dir.
2731 * Makefile.am: Override suffix based rules to specify correct output
2732 subdirectory.
2733 * Makefile.in: Rebuilt.
2734 * configure: Rebuilt.
2735
29183c57 27362001-03-23 Kevin B Hendricks <khendricks@ivey.uwo.ca>
384eea1f 2737
2738 * src/powerpc/ppc_closure.S: New file.
2739 * src/powerpc/ffi.c (ffi_prep_args): Fixed ABI compatibility bug
2740 involving long long and register pairs.
2741 (ffi_prep_closure): New function.
2742 (flush_icache): Likewise.
2743 (ffi_closure_helper_SYSV): Likewise.
2744 * include/ffi.h.in (FFI_CLOSURES): Define on PPC.
2745 (FFI_TRAMPOLINE_SIZE): Likewise.
2746 (FFI_NATIVE_RAW_API): Likewise.
2747 * Makefile.in: Rebuilt.
2748 * Makefile.am (EXTRA_DIST): Added src/powerpc/ppc_closure.S.
2749 (TARGET_SRC_POWERPC): Likewise.
2750
49604b5b 27512001-03-19 Tom Tromey <tromey@redhat.com>
2752
2753 * Makefile.in: Rebuilt.
2754 * Makefile.am (ffitest_LDFLAGS): New macro.
2755
c8fdcc07 27562001-03-02 Nick Clifton <nickc@redhat.com>
2757
2758 * include/ffi.h.in: Remove RCS ident string.
2759 * include/ffi_mips.h: Remove RCS ident string.
2760 * src/debug.c: Remove RCS ident string.
2761 * src/ffitest.c: Remove RCS ident string.
2762 * src/prep_cif.c: Remove RCS ident string.
2763 * src/types.c: Remove RCS ident string.
2764 * src/alpha/ffi.c: Remove RCS ident string.
2765 * src/alpha/osf.S: Remove RCS ident string.
2766 * src/arm/ffi.c: Remove RCS ident string.
2767 * src/arm/sysv.S: Remove RCS ident string.
2768 * src/mips/ffi.c: Remove RCS ident string.
2769 * src/mips/n32.S: Remove RCS ident string.
2770 * src/mips/o32.S: Remove RCS ident string.
2771 * src/sparc/ffi.c: Remove RCS ident string.
2772 * src/sparc/v8.S: Remove RCS ident string.
2773 * src/sparc/v9.S: Remove RCS ident string.
2774 * src/x86/ffi.c: Remove RCS ident string.
2775 * src/x86/sysv.S: Remove RCS ident string.
2776
8cb0f8ea 27772001-02-08 Joseph S. Myers <jsm28@cam.ac.uk>
2778
2779 * include/ffi.h.in: Change sourceware.cygnus.com references to
2780 gcc.gnu.org.
2781
18f704a9 27822000-12-09 Richard Henderson <rth@redhat.com>
2783
2784 * src/alpha/ffi.c (ffi_call): Simplify struct return test.
2785 (ffi_closure_osf_inner): Index rather than increment avalue
2786 and arg_types. Give ffi_closure_osf the raw return value type.
2787 * src/alpha/osf.S (ffi_closure_osf): Handle return value type
2788 promotion.
2789
27902000-12-07 Richard Henderson <rth@redhat.com>
fdf04128 2791
2792 * src/raw_api.c (ffi_translate_args): Fix typo.
2793 (ffi_prep_closure): Likewise.
2794
2795 * include/ffi.h.in [ALPHA]: Define FFI_CLOSURES and
2796 FFI_TRAMPOLINE_SIZE.
2797 * src/alpha/ffi.c (ffi_prep_cif_machdep): Adjust minimal
2798 cif->bytes for new ffi_call_osf implementation.
2799 (ffi_prep_args): Absorb into ...
2800 (ffi_call): ... here. Do all stack allocation here and
2801 avoid a callback function.
2802 (ffi_prep_closure, ffi_closure_osf_inner): New.
2803 * src/alpha/osf.S (ffi_call_osf): Reimplement with no callback.
2804 (ffi_closure_osf): New.
2805
652f285b 28062000-09-10 Alexandre Oliva <aoliva@redhat.com>
2807
4bfa769b 2808 * config.guess, config.sub, install-sh: Removed.
2809 * ltconfig, ltmain.sh, missing, mkinstalldirs: Likewise.
2810 * Makefile.in: Rebuilt.
2811
652f285b 2812 * acinclude.m4: Include libtool macros from the top level.
2813 * aclocal.m4, configure: Rebuilt.
2814
6f52caf7 28152000-08-22 Alexandre Oliva <aoliva@redhat.com>
2816
2817 * configure.in [i*86-*-freebsd*] (TARGET, TARGETDIR): Set.
2818 * configure: Rebuilt.
2819
d03e800c 28202000-05-11 Scott Bambrough <scottb@netwinder.org>
2821
805e22b2 2822 * libffi/src/arm/sysv.S (ffi_call_SYSV): Doubles are not saved to
2823 memory correctly. Use conditional instructions, not branches where
d03e800c 2824 possible.
2825
f3e32e0b 28262000-05-04 Tom Tromey <tromey@cygnus.com>
2827
2828 * configure: Rebuilt.
2829 * configure.in: Match `arm*-*-linux-*'.
2830 From Chris Dornan <cdornan@arm.com>.
2831
b23e0d76 28322000-04-28 Jakub Jelinek <jakub@redhat.com>
2833
2834 * Makefile.am (SUBDIRS): Define.
2835 (AM_MAKEFLAGS): Likewise.
2836 (Multilib support.): Add section.
2837 * Makefile.in: Rebuilt.
2838 * ltconfig (extra_compiler_flags, extra_compiler_flags_value):
968b0d1e 2839 New variables. Set for gcc using -print-multi-lib. Export them
2840 to libtool.
2841 (sparc64-*-linux-gnu*): Use libsuff 64 for search paths.
2842 * ltmain.sh (B|b|V): Don't throw away gcc's -B, -b and -V options
2843 for -shared links.
2844 (extra_compiler_flags_value, extra_compiler_flags): Check these
2845 for extra compiler options which need to be passed down in
2846 compiler_flags.
b23e0d76 2847
2c9b297c 28482000-04-16 Anthony Green <green@redhat.com>
2849
2850 * configure: Rebuilt.
2851 * configure.in: Change i*86-pc-linux* to i*86-*-linux*.
2852
d7a35c6c 28532000-04-14 Jakub Jelinek <jakub@redhat.com>
2854
968b0d1e 2855 * include/ffi.h.in (SPARC64): Define for 64bit SPARC builds.
d7a35c6c 2856 Set SPARC FFI_DEFAULT_ABI based on SPARC64 define.
2857 * src/sparc/ffi.c (ffi_prep_args_v8): Renamed from ffi_prep_args.
2858 Replace all void * sizeofs with sizeof(int).
2859 Only compare type with FFI_TYPE_LONGDOUBLE if LONGDOUBLE is
2860 different than DOUBLE.
2861 Remove FFI_TYPE_SINT32 and FFI_TYPE_UINT32 cases (handled elsewhere).
2862 (ffi_prep_args_v9): New function.
2863 (ffi_prep_cif_machdep): Handle V9 ABI and long long on V8.
2864 (ffi_V9_return_struct): New function.
2865 (ffi_call): Handle FFI_V9 ABI from 64bit code and FFI_V8 ABI from
2866 32bit code (not yet cross-arch calls).
2867 * src/sparc/v8.S: Add struct return delay nop.
2868 Handle long long.
2869 * src/sparc/v9.S: New file.
2870 * src/prep_cif.c (ffi_prep_cif): Return structure pointer
2871 is used on sparc64 only for structures larger than 32 bytes.
2872 Pass by reference for structures is done for structure arguments
2873 larger than 16 bytes.
2874 * src/ffitest.c (main): Use 64bit rint on sparc64.
2875 Run long long tests on sparc.
2876 * src/types.c (FFI_TYPE_POINTER): Pointer is 64bit on alpha and
2877 sparc64.
2878 (FFI_TYPE_LONGDOUBLE): long double is 128 bit aligned to 128 bits
2879 on sparc64.
2880 * configure.in (sparc-*-linux*): New supported target.
2881 (sparc64-*-linux*): Likewise.
2882 * configure: Rebuilt.
2883 * Makefile.am: Add v9.S to SPARC files.
2884 * Makefile.in: Likewise.
2885 (LINK): Surround $(CCLD) into double quotes, so that multilib
2886 compiles work correctly.
2887
800d12c7 28882000-04-04 Alexandre Petit-Bianco <apbianco@cygnus.com>
2889
2890 * configure: Rebuilt.
2891 * configure.in: (i*86-*-solaris*): New libffi target. Patch
2892 proposed by Bryce McKinlay.
2893
c44e8808 28942000-03-20 Tom Tromey <tromey@cygnus.com>
2895
2896 * Makefile.in: Hand edit for java_raw_api.lo.
2897
f764cb60 28982000-03-08 Bryce McKinlay <bryce@albatross.co.nz>
2899
2900 * config.guess, config.sub: Update from the gcc tree.
2901 Fix for PR libgcj/168.
2902
3b43a26e 29032000-03-03 Tom Tromey <tromey@cygnus.com>
2904
96215ab3 2905 * Makefile.in: Fixed ia64 by hand.
2906
3b43a26e 2907 * configure: Rebuilt.
2908 * configure.in (--enable-multilib): New option.
2909 (libffi_basedir): New subst.
2910 (AC_OUTPUT): Added multilib code.
2911
03ae52bb 29122000-03-02 Tom Tromey <tromey@cygnus.com>
2913
2914 * Makefile.in: Rebuilt.
2915 * Makefile.am (TARGET_SRC_IA64): Use `ia64', not `alpha', as
2916 directory name.
2917
28c07da5 29182000-02-25 Hans Boehm <boehm@acm.org>
2919
2920 * src/ia64/ffi.c, src/ia64/ia64_flags.h, src/ia64/unix.S: New
2921 files.
2922 * src/raw_api.c (ffi_translate_args): Fixed typo in argument
2923 list.
2924 (ffi_prep_raw_closure): Use ffi_translate_args, not
2925 ffi_closure_translate.
2926 * src/java_raw_api.c: New file.
2927 * src/ffitest.c (closure_test_fn): New function.
2928 (main): Define `rint' as long long on IA64. Added new test when
2929 FFI_CLOSURES is defined.
2930 * include/ffi.h.in (ALIGN): Use size_t, not unsigned.
2931 (ffi_abi): Recognize IA64.
2932 (ffi_raw): Added `flt' field.
2933 Added "Java raw API" code.
2934 * configure.in: Recognize ia64.
2935 * Makefile.am (TARGET_SRC_IA64): New macro.
2936 (libffi_la_common_SOURCES): Added java_raw_api.c.
2937 (libffi_la_SOURCES): Define in IA64 case.
2938
0bcdf5c5 29392000-01-04 Tom Tromey <tromey@cygnus.com>
2940
2941 * Makefile.in: Rebuilt with newer automake.
2942
e9de8c8b 29431999-12-31 Tom Tromey <tromey@cygnus.com>
2944
2945 * Makefile.am (INCLUDES): Added -I$(top_srcdir)/src.
2946
20e58c02 29471999-09-01 Tom Tromey <tromey@cygnus.com>
2948
2949 * include/ffi.h.in: Removed PACKAGE and VERSION defines and
2950 undefs.
2951 * fficonfig.h.in: Rebuilt.
2952 * configure: Rebuilt.
2953 * configure.in: Pass 3rd argument to AM_INIT_AUTOMAKE.
2954 Use AM_PROG_LIBTOOL (automake 1.4 compatibility).
2955 * acconfig.h: Don't #undef PACKAGE or VERSION.
2956
47c5fab6 29571999-08-09 Anthony Green <green@cygnus.com>
2958
2959 * include/ffi.h.in: Try to work around messy header problem
2960 with PACKAGE and VERSION.
2961
2962 * configure: Rebuilt.
2963 * configure.in: Change version to 2.00-beta.
2964
968b0d1e 2965 * fficonfig.h.in: Rebuilt.
47c5fab6 2966 * acconfig.h (FFI_NO_STRUCTS, FFI_NO_RAW_API): Define.
2967
2968 * src/x86/ffi.c (ffi_raw_call): Rename.
2969
29701999-08-02 Kresten Krab Thorup <krab@dominiq.is.s.u-tokyo.ac.jp>
2971
2972 * src/x86/ffi.c (ffi_closure_SYSV): New function.
2973 (ffi_prep_incoming_args_SYSV): Ditto.
2974 (ffi_prep_closure): Ditto.
2975 (ffi_closure_raw_SYSV): Ditto.
2976 (ffi_prep_raw_closure): More ditto.
2977 (ffi_call_raw): Final ditto.
2978
2979 * include/ffi.h.in: Add definitions for closure and raw API.
2980
2981 * src/x86/ffi.c (ffi_prep_cif_machdep): Added case for
805e22b2 2982 FFI_TYPE_UINT64.
47c5fab6 2983
2984 * Makefile.am (libffi_la_common_SOURCES): Added raw_api.c
2985
2986 * src/raw_api.c: New file.
2987
2988 * include/ffi.h.in (ffi_raw): New type.
2989 (UINT_ARG, SINT_ARG): New defines.
2990 (ffi_closure, ffi_raw_closure): New types.
2991 (ffi_prep_closure, ffi_prep_raw_closure): New declarations.
805e22b2 2992
47c5fab6 2993 * configure.in: Add check for endianness and sizeof void*.
2994
2995 * src/x86/sysv.S (ffi_call_SYSV): Call fixup routine via argument,
2996 instead of directly.
2997
2998 * configure: Rebuilt.
2999
3000Thu Jul 8 14:28:42 1999 Anthony Green <green@cygnus.com>
3001
3002 * configure.in: Add x86 and powerpc BeOS configurations.
3003 From Makoto Kato <m_kato@ga2.so-net.ne.jp>.
3004
30051999-05-09 Anthony Green <green@cygnus.com>
3006
3007 * configure.in: Add warning about this being beta code.
3008 Remove src/Makefile.am from the picture.
3009 * configure: Rebuilt.
3010
3011 * Makefile.am: Move logic from src/Makefile.am. Add changes
3012 to support libffi as a target library.
3013 * Makefile.in: Rebuilt.
3014
3015 * aclocal.m4, config.guess, config.sub, ltconfig, ltmain.sh:
3016 Upgraded to new autoconf, automake, libtool.
3017
3018 * README: Tweaks.
3019
3020 * LICENSE: Update copyright date.
3021
3022 * src/Makefile.am, src/Makefile.in: Removed.
3023
30241998-11-29 Anthony Green <green@cygnus.com>
3025
3026 * include/ChangeLog: Removed.
3027 * src/ChangeLog: Removed.
3028 * src/mips/ChangeLog: Removed.
3029 * src/sparc/ChangeLog: Remboved.
3030 * src/x86/ChangeLog: Removed.
3031
3032 * ChangeLog.v1: Created.
Mirror of https://gcc.gnu.org/git/gcc.git