]> git.ipfire.org Git - thirdparty/gcc.git/blame_incremental - libffi/ChangeLog
projects / thirdparty / gcc.git / blame_incremental
? search:
summary | shortlog | log | commit | commitdiff | tree
blob | blame (non-incremental) | history | HEAD
New tests
[thirdparty/gcc.git] / libffi / ChangeLog
... / ...
CommitLineData
12008-03-16 Ralf Wildenhues <Ralf.Wildenhues@gmx.de>
2
3 * aclocal.m4: Regenerate.
4 * configure: Likewise.
5 * Makefile.in: Likewise.
6 * include/Makefile.in: Likewise.
7 * testsuite/Makefile.in: Likewise.
8
92008-02-12 Bjoern Koenig <bkoenig@alpha-tierchen.de>
10 Andreas Tobler <a.tobler@schweiz.org>
11
12 * configure.ac: Add amd64-*-freebsd* target.
13 * configure: Regenerate.
14
152008-01-30 H.J. Lu <hongjiu.lu@intel.com>
16
17 PR libffi/34612
18 * src/x86/sysv.S (ffi_closure_SYSV): Pop 4 byte from stack when
19 returning struct.
20
21 * testsuite/libffi.call/call.exp: Add "-O2 -fomit-frame-pointer"
22 tests.
23
242008-01-24 David Edelsohn <edelsohn@gnu.org>
25
26 * configure: Regenerate.
27
282008-01-06 Andreas Tobler <a.tobler@schweiz.org>
29
30 * src/x86/ffi.c (ffi_prep_cif_machdep): Fix thinko.
31
322008-01-05 Andreas Tobler <a.tobler@schweiz.org>
33
34 PR testsuite/32843
35 * src/x86/ffi.c (ffi_prep_cif_machdep): Add code for
36 signed/unsigned int8/16 for X86_DARWIN.
37 Updated copyright info.
38 Handle one and two byte structs with special cif->flags.
39 * src/x86/ffitarget.h: Add special types for one and two byte structs.
40 Updated copyright info.
41 * src/x86/darwin.S (ffi_call_SYSV): Rewrite to use a jump table like
42 sysv.S
43 Remove code to pop args from the stack after call.
44 Special-case signed/unsigned for int8/16, one and two byte structs.
45 (ffi_closure_raw_SYSV): Handle FFI_TYPE_UINT8,
46 FFI_TYPE_SINT8, FFI_TYPE_UINT16, FFI_TYPE_SINT16, FFI_TYPE_UINT32,
47 FFI_TYPE_SINT32.
48 Updated copyright info.
49
502007-12-08 David Daney <ddaney@avtrex.com>
51
52 * src/mips/n32.S (ffi_call_N32): Replace dadd with ADDU, dsub with
53 SUBU, add with ADDU and use smaller code sequences.
54
552007-12-07 David Daney <ddaney@avtrex.com>
56
57 * src/mips/ffi.c (ffi_prep_cif_machdep): Handle long double return
58 type.
59
602007-12-06 David Daney <ddaney@avtrex.com>
61
62 * include/ffi.h.in (FFI_SIZEOF_JAVA_RAW): Define if not already
63 defined.
64 (ffi_java_raw): New typedef.
65 (ffi_java_raw_call, ffi_java_ptrarray_to_raw,
66 ffi_java_raw_to_ptrarray): Change parameter types from ffi_raw to
67 ffi_java_raw.
68 (ffi_java_raw_closure) : Same.
69 (ffi_prep_java_raw_closure, ffi_prep_java_raw_closure_loc): Change
70 parameter types.
71 * src/java_raw_api.c (ffi_java_raw_size): Replace FFI_SIZEOF_ARG with
72 FFI_SIZEOF_JAVA_RAW.
73 (ffi_java_raw_to_ptrarray): Change type of raw to ffi_java_raw.
74 Replace FFI_SIZEOF_ARG with FFI_SIZEOF_JAVA_RAW. Use
75 sizeof(ffi_java_raw) for alignment calculations.
76 (ffi_java_ptrarray_to_raw): Same.
77 (ffi_java_rvalue_to_raw): Add special handling for FFI_TYPE_POINTER
78 if FFI_SIZEOF_JAVA_RAW == 4.
79 (ffi_java_raw_to_rvalue): Same.
80 (ffi_java_raw_call): Change type of raw to ffi_java_raw.
81 (ffi_java_translate_args): Same.
82 (ffi_prep_java_raw_closure_loc, ffi_prep_java_raw_closure): Change
83 parameter types.
84 * src/mips/ffitarget.h (FFI_SIZEOF_JAVA_RAW): Define for N32 ABI.
85
862007-12-06 David Daney <ddaney@avtrex.com>
87
88 * src/mips/n32.S (ffi_closure_N32): Use 64-bit add instruction on
89 pointer values.
90
912007-12-01 Andreas Tobler <a.tobler@schweiz.org>
92
93 PR libffi/31937
94 * src/powerpc/ffitarget.h: Introduce new ABI FFI_LINUX_SOFT_FLOAT.
95 Add local FFI_TYPE_UINT128 to handle soft-float long-double-128.
96 * src/powerpc/ffi.c: Distinguish between __NO_FPRS__ and not and
97 set the NUM_FPR_ARG_REGISTERS according to.
98 Add support for potential soft-float support under hard-float
99 architecture.
100 (ffi_prep_args_SYSV): Set NUM_FPR_ARG_REGISTERS to 0 in case of
101 FFI_LINUX_SOFT_FLOAT, handle float, doubles and long-doubles according
102 to the FFI_LINUX_SOFT_FLOAT ABI.
103 (ffi_prep_cif_machdep): Likewise.
104 (ffi_closure_helper_SYSV): Likewise.
105 * src/powerpc/ppc_closure.S: Make sure not to store float/double
106 on archs where __NO_FPRS__ is true.
107 Add FFI_TYPE_UINT128 support.
108 * src/powerpc/sysv.S: Add support for soft-float long-double-128.
109 Adjust copyright notice.
110
1112007-11-25 Andreas Tobler <a.tobler@schweiz.org>
112
113 * src/closures.c: Move defintion of MAYBE_UNUSED from here to ...
114 * include/ffi_common.h: ... here.
115 Update copyright.
116
1172007-11-17 Andreas Tobler <a.tobler@schweiz.org>
118
119 * src/powerpc/sysv.S: Load correct cr to compare if we have long double.
120 * src/powerpc/linux64.S: Likewise.
121 * src/powerpc/ffi.c: Add a comment to show which part goes into cr6.
122 * testsuite/libffi.call/return_ldl.c: New test.
123
1242007-09-04 <aph@redhat.com>
125
126 * src/arm/sysv.S (UNWIND): New.
127 (Whole file): Conditionally compile unwinder directives.
128 * src/arm/sysv.S: Add unwinder directives.
129
130 * src/arm/ffi.c (ffi_prep_args): Align structs by at least 4 bytes.
131 Only treat r0 as a struct address if we're actually returning a
132 struct by address.
133 Only copy the bytes that are actually within a struct.
134 (ffi_prep_cif_machdep): A Composite Type not larger than 4 bytes
135 is returned in r0, not passed by address.
136 (ffi_call): Allocate a word-sized temporary for the case where
137 a composite is returned in r0.
138 (ffi_prep_incoming_args_SYSV): Align as necessary.
139
1402007-08-05 Steven Newbury <s_j_newbury@yahoo.co.uk>
141
142 * src/arm/ffi.c (FFI_INIT_TRAMPOLINE): Use __clear_cache instead of
143 directly using the sys_cacheflush syscall.
144
1452007-07-27 Andrew Haley <aph@redhat.com>
146
147 * src/arm/sysv.S (ffi_closure_SYSV): Add soft-float.
148
1492007-09-03 Maciej W. Rozycki <macro@linux-mips.org>
150
151 * Makefile.am: Unify MIPS_IRIX and MIPS_LINUX into MIPS.
152 * configure.ac: Likewise.
153 * Makefile.in: Regenerate.
154 * include/Makefile.in: Likewise.
155 * testsuite/Makefile.in: Likewise.
156 * configure: Likewise.
157
1582007-08-24 David Daney <ddaney@avtrex.com>
159
160 * testsuite/libffi.call/return_sl.c: New test.
161
1622007-08-10 David Daney <ddaney@avtrex.com>
163
164 * testsuite/libffi.call/cls_multi_ushort.c,
165 testsuite/libffi.call/cls_align_uint16.c,
166 testsuite/libffi.call/nested_struct1.c,
167 testsuite/libffi.call/nested_struct3.c,
168 testsuite/libffi.call/cls_7_1_byte.c,
169 testsuite/libffi.call/nested_struct5.c,
170 testsuite/libffi.call/cls_double.c,
171 testsuite/libffi.call/nested_struct7.c,
172 testsuite/libffi.call/cls_sint.c,
173 testsuite/libffi.call/nested_struct9.c,
174 testsuite/libffi.call/cls_20byte1.c,
175 testsuite/libffi.call/cls_multi_sshortchar.c,
176 testsuite/libffi.call/cls_align_sint64.c,
177 testsuite/libffi.call/cls_3byte2.c,
178 testsuite/libffi.call/cls_multi_schar.c,
179 testsuite/libffi.call/cls_multi_uchar.c,
180 testsuite/libffi.call/cls_19byte.c,
181 testsuite/libffi.call/cls_9byte1.c,
182 testsuite/libffi.call/cls_align_float.c,
183 testsuite/libffi.call/closure_fn1.c,
184 testsuite/libffi.call/problem1.c,
185 testsuite/libffi.call/closure_fn3.c,
186 testsuite/libffi.call/cls_sshort.c,
187 testsuite/libffi.call/closure_fn5.c,
188 testsuite/libffi.call/cls_align_double.c,
189 testsuite/libffi.call/nested_struct.c,
190 testsuite/libffi.call/cls_2byte.c,
191 testsuite/libffi.call/nested_struct10.c,
192 testsuite/libffi.call/cls_4byte.c,
193 testsuite/libffi.call/cls_6byte.c,
194 testsuite/libffi.call/cls_8byte.c,
195 testsuite/libffi.call/cls_multi_sshort.c,
196 testsuite/libffi.call/cls_align_sint16.c,
197 testsuite/libffi.call/cls_align_uint32.c,
198 testsuite/libffi.call/cls_20byte.c,
199 testsuite/libffi.call/cls_float.c,
200 testsuite/libffi.call/nested_struct2.c,
201 testsuite/libffi.call/cls_5_1_byte.c,
202 testsuite/libffi.call/nested_struct4.c,
203 testsuite/libffi.call/cls_24byte.c,
204 testsuite/libffi.call/nested_struct6.c,
205 testsuite/libffi.call/cls_64byte.c,
206 testsuite/libffi.call/nested_struct8.c,
207 testsuite/libffi.call/cls_uint.c,
208 testsuite/libffi.call/cls_multi_ushortchar.c,
209 testsuite/libffi.call/cls_schar.c,
210 testsuite/libffi.call/cls_uchar.c,
211 testsuite/libffi.call/cls_align_uint64.c,
212 testsuite/libffi.call/cls_ulonglong.c,
213 testsuite/libffi.call/cls_align_longdouble.c,
214 testsuite/libffi.call/cls_1_1byte.c,
215 testsuite/libffi.call/cls_12byte.c,
216 testsuite/libffi.call/cls_3_1byte.c,
217 testsuite/libffi.call/cls_3byte1.c,
218 testsuite/libffi.call/cls_4_1byte.c,
219 testsuite/libffi.call/cls_6_1_byte.c,
220 testsuite/libffi.call/cls_16byte.c,
221 testsuite/libffi.call/cls_18byte.c,
222 testsuite/libffi.call/closure_fn0.c,
223 testsuite/libffi.call/cls_9byte2.c,
224 testsuite/libffi.call/closure_fn2.c,
225 testsuite/libffi.call/closure_fn4.c,
226 testsuite/libffi.call/cls_ushort.c,
227 testsuite/libffi.call/closure_fn6.c,
228 testsuite/libffi.call/cls_5byte.c,
229 testsuite/libffi.call/cls_align_pointer.c,
230 testsuite/libffi.call/cls_7byte.c,
231 testsuite/libffi.call/cls_align_sint32.c,
232 testsuite/libffi.special/unwindtest_ffi_call.cc,
233 testsuite/libffi.special/unwindtest.cc: Remove xfail for mips64*-*-*.
234
2352007-08-10 David Daney <ddaney@avtrex.com>
236
237 PR libffi/28313
238 * configure.ac: Don't treat mips64 as a special case.
239 * Makefile.am (nodist_libffi_la_SOURCES): Add n32.S.
240 * configure: Regenerate
241 * Makefile.in: Ditto.
242 * fficonfig.h.in: Ditto.
243 * src/mips/ffitarget.h (REG_L, REG_S, SUBU, ADDU, SRL, LI): Indent.
244 (LA, EH_FRAME_ALIGN, FDE_ADDR_BYTES): New preprocessor macros.
245 (FFI_DEFAULT_ABI): Set for n64 case.
246 (FFI_CLOSURES, FFI_TRAMPOLINE_SIZE): Define for n32 and n64 cases.
247 * src/mips/n32.S (ffi_call_N32): Add debug macros and labels for FDE.
248 (ffi_closure_N32): New function.
249 (.eh_frame): New section
250 * src/mips/o32.S: Clean up comments.
251 (ffi_closure_O32): Pass ffi_closure parameter in $12.
252 * src/mips/ffi.c: Use FFI_MIPS_N32 instead of
253 _MIPS_SIM == _ABIN32 throughout.
254 (FFI_MIPS_STOP_HERE): New, use in place of
255 ffi_stop_here.
256 (ffi_prep_args): Use unsigned long to hold pointer values. Rewrite
257 to support n32/n64 ABIs.
258 (calc_n32_struct_flags): Rewrite.
259 (calc_n32_return_struct_flags): Remove unused variable. Reverse
260 position of flag bits.
261 (ffi_prep_cif_machdep): Rewrite n32 portion.
262 (ffi_call): Enable for n64. Add special handling for small structure
263 return values.
264 (ffi_prep_closure_loc): Add n32 and n64 support.
265 (ffi_closure_mips_inner_O32): Add cast to silence warning.
266 (copy_struct_N32, ffi_closure_mips_inner_N32): New functions.
267
2682007-08-08 David Daney <ddaney@avtrex.com>
269
270 * testsuite/libffi.call/ffitest.h (ffi_type_mylong): Remove definition.
271 * testsuite/libffi.call/cls_align_uint16.c (main): Use correct type
272 specifiers.
273 * testsuite/libffi.call/nested_struct1.c (main): Ditto.
274 * testsuite/libffi.call/cls_sint.c (main): Ditto.
275 * testsuite/libffi.call/nested_struct9.c (main): Ditto.
276 * testsuite/libffi.call/cls_20byte1.c (main): Ditto.
277 * testsuite/libffi.call/cls_9byte1.c (main): Ditto.
278 * testsuite/libffi.call/closure_fn1.c (main): Ditto.
279 * testsuite/libffi.call/closure_fn3.c (main): Ditto.
280 * testsuite/libffi.call/return_dbl2.c (main): Ditto.
281 * testsuite/libffi.call/cls_sshort.c (main): Ditto.
282 * testsuite/libffi.call/return_fl3.c (main): Ditto.
283 * testsuite/libffi.call/closure_fn5.c (main): Ditto.
284 * testsuite/libffi.call/nested_struct.c (main): Ditto.
285 * testsuite/libffi.call/nested_struct10.c (main): Ditto.
286 * testsuite/libffi.call/return_ll1.c (main): Ditto.
287 * testsuite/libffi.call/cls_8byte.c (main): Ditto.
288 * testsuite/libffi.call/cls_align_uint32.c (main): Ditto.
289 * testsuite/libffi.call/cls_align_sint16.c (main): Ditto.
290 * testsuite/libffi.call/cls_20byte.c (main): Ditto.
291 * testsuite/libffi.call/nested_struct2.c (main): Ditto.
292 * testsuite/libffi.call/cls_24byte.c (main): Ditto.
293 * testsuite/libffi.call/nested_struct6.c (main): Ditto.
294 * testsuite/libffi.call/cls_uint.c (main): Ditto.
295 * testsuite/libffi.call/cls_12byte.c (main): Ditto.
296 * testsuite/libffi.call/cls_16byte.c (main): Ditto.
297 * testsuite/libffi.call/closure_fn0.c (main): Ditto.
298 * testsuite/libffi.call/cls_9byte2.c (main): Ditto.
299 * testsuite/libffi.call/closure_fn2.c (main): Ditto.
300 * testsuite/libffi.call/return_dbl1.c (main): Ditto.
301 * testsuite/libffi.call/closure_fn4.c (main): Ditto.
302 * testsuite/libffi.call/closure_fn6.c (main): Ditto.
303 * testsuite/libffi.call/cls_align_sint32.c (main): Ditto.
304
3052007-08-07 Andrew Haley <aph@redhat.com>
306
307 * src/x86/sysv.S (ffi_closure_raw_SYSV): Fix typo in previous
308 checkin.
309
3102007-08-06 Andrew Haley <aph@redhat.com>
311
312 PR testsuite/32843
313 * src/x86/sysv.S (ffi_closure_raw_SYSV): Handle FFI_TYPE_UINT8,
314 FFI_TYPE_SINT8, FFI_TYPE_UINT16, FFI_TYPE_SINT16, FFI_TYPE_UINT32,
315 FFI_TYPE_SINT32.
316
3172007-08-02 David Daney <ddaney@avtrex.com>
318
319 * testsuite/libffi.call/return_ul.c (main): Define return type as
320 ffi_arg. Use proper printf conversion specifier.
321
3222007-07-30 Andrew Haley <aph@redhat.com>
323
324 PR testsuite/32843
325 * src/x86/ffi.c (ffi_prep_cif_machdep): in x86 case, add code for
326 signed/unsigned int8/16.
327 * src/x86/sysv.S (ffi_call_SYSV): Rewrite to:
328 Use a jump table.
329 Remove code to pop args from the stack after call.
330 Special-case signed/unsigned int8/16.
331 * testsuite/libffi.call/return_sc.c (main): Revert.
332
3332007-07-26 Richard Guenther <rguenther@suse.de>
334
335 PR testsuite/32843
336 * testsuite/libffi.call/return_sc.c (main): Verify call
337 result as signed char, not ffi_arg.
338
3392007-07-16 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
340
341 * configure.ac (i?86-*-solaris2.1[0-9]): Set TARGET to X86_64.
342 * configure: Regenerate.
343
3442007-07-11 David Daney <ddaney@avtrex.com>
345
346 * src/mips/ffi.c: Don't include sys/cachectl.h.
347 (ffi_prep_closure_loc): Use __builtin___clear_cache() instead of
348 cacheflush().
349
3502007-05-18 Aurelien Jarno <aurelien@aurel32.net>
351
352 * src/arm/ffi.c (ffi_prep_closure_loc): Renamed and ajusted
353 from (ffi_prep_closure): ... this.
354 (FFI_INIT_TRAMPOLINE): Adjust.
355
3562005-12-31 Phil Blundell <pb@reciva.com>
357
358 * src/arm/ffi.c (ffi_prep_incoming_args_SYSV,
359 ffi_closure_SYSV_inner, ffi_prep_closure): New, add closure support.
360 * src/arm/sysv.S(ffi_closure_SYSV): Likewise.
361 * src/arm/ffitarget.h (FFI_TRAMPOLINE_SIZE): Likewise.
362 (FFI_CLOSURES): Enable closure support.
363
3642007-07-03 Andrew Haley <aph@hedges.billgatliff.com>
365
366 * testsuite/libffi.call/cls_multi_ushort.c,
367 testsuite/libffi.call/cls_align_uint16.c,
368 testsuite/libffi.call/nested_struct1.c,
369 testsuite/libffi.call/nested_struct3.c,
370 testsuite/libffi.call/cls_7_1_byte.c,
371 testsuite/libffi.call/cls_double.c,
372 testsuite/libffi.call/nested_struct5.c,
373 testsuite/libffi.call/nested_struct7.c,
374 testsuite/libffi.call/cls_sint.c,
375 testsuite/libffi.call/nested_struct9.c,
376 testsuite/libffi.call/cls_20byte1.c,
377 testsuite/libffi.call/cls_multi_sshortchar.c,
378 testsuite/libffi.call/cls_align_sint64.c,
379 testsuite/libffi.call/cls_3byte2.c,
380 testsuite/libffi.call/cls_multi_schar.c,
381 testsuite/libffi.call/cls_multi_uchar.c,
382 testsuite/libffi.call/cls_19byte.c,
383 testsuite/libffi.call/cls_9byte1.c,
384 testsuite/libffi.call/cls_align_float.c,
385 testsuite/libffi.call/closure_fn1.c,
386 testsuite/libffi.call/problem1.c,
387 testsuite/libffi.call/closure_fn3.c,
388 testsuite/libffi.call/cls_sshort.c,
389 testsuite/libffi.call/closure_fn5.c,
390 testsuite/libffi.call/cls_align_double.c,
391 testsuite/libffi.call/cls_2byte.c,
392 testsuite/libffi.call/nested_struct.c,
393 testsuite/libffi.call/nested_struct10.c,
394 testsuite/libffi.call/cls_4byte.c,
395 testsuite/libffi.call/cls_6byte.c,
396 testsuite/libffi.call/cls_8byte.c,
397 testsuite/libffi.call/cls_multi_sshort.c,
398 testsuite/libffi.call/cls_align_uint32.c,
399 testsuite/libffi.call/cls_align_sint16.c,
400 testsuite/libffi.call/cls_float.c,
401 testsuite/libffi.call/cls_20byte.c,
402 testsuite/libffi.call/cls_5_1_byte.c,
403 testsuite/libffi.call/nested_struct2.c,
404 testsuite/libffi.call/cls_24byte.c,
405 testsuite/libffi.call/nested_struct4.c,
406 testsuite/libffi.call/nested_struct6.c,
407 testsuite/libffi.call/cls_64byte.c,
408 testsuite/libffi.call/nested_struct8.c,
409 testsuite/libffi.call/cls_uint.c,
410 testsuite/libffi.call/cls_multi_ushortchar.c,
411 testsuite/libffi.call/cls_schar.c,
412 testsuite/libffi.call/cls_uchar.c,
413 testsuite/libffi.call/cls_align_uint64.c,
414 testsuite/libffi.call/cls_ulonglong.c,
415 testsuite/libffi.call/cls_align_longdouble.c,
416 testsuite/libffi.call/cls_1_1byte.c,
417 testsuite/libffi.call/cls_12byte.c,
418 testsuite/libffi.call/cls_3_1byte.c,
419 testsuite/libffi.call/cls_3byte1.c,
420 testsuite/libffi.call/cls_4_1byte.c,
421 testsuite/libffi.call/cls_6_1_byte.c,
422 testsuite/libffi.call/cls_16byte.c,
423 testsuite/libffi.call/cls_18byte.c,
424 testsuite/libffi.call/closure_fn0.c,
425 testsuite/libffi.call/cls_9byte2.c,
426 testsuite/libffi.call/closure_fn2.c,
427 testsuite/libffi.call/closure_fn4.c,
428 testsuite/libffi.call/cls_ushort.c,
429 testsuite/libffi.call/closure_fn6.c,
430 testsuite/libffi.call/cls_5byte.c,
431 testsuite/libffi.call/cls_align_pointer.c,
432 testsuite/libffi.call/cls_7byte.c,
433 testsuite/libffi.call/cls_align_sint32.c,
434 testsuite/libffi.special/unwindtest_ffi_call.cc,
435 testsuite/libffi.special/unwindtest.cc: Enable for ARM.
436
4372007-07-05 H.J. Lu <hongjiu.lu@intel.com>
438
439 * aclocal.m4: Regenerated.
440
4412007-06-02 Paolo Bonzini <bonzini@gnu.org>
442
443 * configure: Regenerate.
444
4452007-05-23 Steve Ellcey <sje@cup.hp.com>
446
447 * Makefile.in: Regenerate.
448 * configure: Regenerate.
449 * aclocal.m4: Regenerate.
450 * include/Makefile.in: Regenerate.
451 * testsuite/Makefile.in: Regenerate.
452
4532007-05-10 Roman Zippel <zippel@linux-m68k.org>
454
455 * src/m68k/ffi.c (ffi_prep_incoming_args_SYSV,
456 ffi_closure_SYSV_inner,ffi_prep_closure): New, add closure support.
457 * src/m68k/sysv.S(ffi_closure_SYSV,ffi_closure_struct_SYSV): Likewise.
458 * src/m68k/ffitarget.h (FFI_TRAMPOLINE_SIZE): Likewise.
459 (FFI_CLOSURES): Enable closure support.
460
4612007-05-10 Roman Zippel <zippel@linux-m68k.org>
462
463 * configure.ac (HAVE_AS_CFI_PSEUDO_OP): New test.
464 * configure: Regenerate.
465 * fficonfig.h.in: Regenerate.
466 * src/m68k/sysv.S (CFI_STARTPROC,CFI_ENDPROC,
467 CFI_OFFSET,CFI_DEF_CFA): New macros.
468 (ffi_call_SYSV): Add callframe annotation.
469
4702007-05-10 Roman Zippel <zippel@linux-m68k.org>
471
472 * src/m68k/ffi.c (ffi_prep_args,ffi_prep_cif_machdep): Fix
473 numerous test suite failures.
474 * src/m68k/sysv.S (ffi_call_SYSV): Likewise.
475
4762007-04-11 Paolo Bonzini <bonzini@gnu.org>
477
478 * Makefile.am (EXTRA_DIST): Bring up to date.
479 * Makefile.in: Regenerate.
480 * src/frv/eabi.S: Remove RCS keyword.
481
4822007-04-06 Richard Henderson <rth@redhat.com>
483
484 * configure.ac: Tidy target case.
485 (HAVE_LONG_DOUBLE): Allow the target to override.
486 * configure: Regenerate.
487 * include/ffi.h.in: Don't define ffi_type_foo if
488 LIBFFI_HIDE_BASIC_TYPES is defined.
489 (ffi_type_longdouble): If not HAVE_LONG_DOUBLE, define
490 to ffi_type_double.
491 * types.c (LIBFFI_HIDE_BASIC_TYPES): Define.
492 (FFI_TYPEDEF, ffi_type_void): Mark the data const.
493 (ffi_type_longdouble): Special case for Alpha. Don't define
494 if long double == double.
495
496 * src/alpha/ffi.c (FFI_TYPE_LONGDOUBLE): Assert unique value.
497 (ffi_prep_cif_machdep): Handle it as the 128-bit type.
498 (ffi_call, ffi_closure_osf_inner): Likewise.
499 (ffi_closure_osf_inner): Likewise. Mark hidden.
500 (ffi_call_osf, ffi_closure_osf): Mark hidden.
501 * src/alpha/ffitarget.h (FFI_LAST_ABI): Tidy definition.
502 * src/alpha/osf.S (ffi_call_osf, ffi_closure_osf): Mark hidden.
503 (load_table): Handle 128-bit long double.
504
505 * testsuite/libffi.call/float4.c: Add -mieee for alpha.
506
5072007-04-06 Tom Tromey <tromey@redhat.com>
508
509 PR libffi/31491:
510 * README: Fixed bug in example.
511
5122007-04-03 Jakub Jelinek <jakub@redhat.com>
513
514 * src/closures.c: Include sys/statfs.h.
515 (_GNU_SOURCE): Define on Linux.
516 (FFI_MMAP_EXEC_SELINUX): Define.
517 (selinux_enabled): New variable.
518 (selinux_enabled_check): New function.
519 (is_selinux_enabled): Define.
520 (dlmmap): Use it.
521
5222007-03-24 Uros Bizjak <ubizjak@gmail.com>
523
524 * testsuite/libffi.call/return_fl2.c (return_fl): Mark as static.
525 Use 'volatile float sum' to create sum of floats to avoid false
526 negative due to excess precision on ix86 targets.
527 (main): Ditto.
528
5292007-03-08 Alexandre Oliva <aoliva@redhat.com>
530
531 * src/powerpc/ffi.c (flush_icache): Fix left-over from previous
532 patch.
533 (ffi_prep_closure_loc): Remove unneeded casts. Add needed ones.
534
5352007-03-07 Alexandre Oliva <aoliva@redhat.com>
536
537 * include/ffi.h.in (ffi_closure_alloc, ffi_closure_free): New.
538 (ffi_prep_closure_loc): New.
539 (ffi_prep_raw_closure_loc): New.
540 (ffi_prep_java_raw_closure_loc): New.
541 * src/closures.c: New file.
542 * src/dlmalloc.c [FFI_MMAP_EXEC_WRIT] (struct malloc_segment):
543 Replace sflags with exec_offset.
544 [FFI_MMAP_EXEC_WRIT] (mmap_exec_offset, add_segment_exec_offset,
545 sub_segment_exec_offset): New macros.
546 (get_segment_flags, set_segment_flags, check_segment_merge): New
547 macros.
548 (is_mmapped_segment, is_extern_segment): Use get_segment_flags.
549 (add_segment, sys_alloc, create_mspace, create_mspace_with_base,
550 destroy_mspace): Use new macros.
551 (sys_alloc): Silence warning.
552 * Makefile.am (libffi_la_SOURCES): Add src/closures.c.
553 * Makefile.in: Rebuilt.
554 * src/prep_cif [FFI_CLOSURES] (ffi_prep_closure): Implement in
555 terms of ffi_prep_closure_loc.
556 * src/raw_api.c (ffi_prep_raw_closure_loc): Renamed and adjusted
557 from...
558 (ffi_prep_raw_closure): ... this. Re-implement in terms of the
559 renamed version.
560 * src/java_raw_api (ffi_prep_java_raw_closure_loc): Renamed and
561 adjusted from...
562 (ffi_prep_java_raw_closure): ... this. Re-implement in terms of
563 the renamed version.
564 * src/alpha/ffi.c (ffi_prep_closure_loc): Renamed from
565 (ffi_prep_closure): ... this.
566 * src/pa/ffi.c: Likewise.
567 * src/cris/ffi.c: Likewise. Adjust.
568 * src/frv/ffi.c: Likewise.
569 * src/ia64/ffi.c: Likewise.
570 * src/mips/ffi.c: Likewise.
571 * src/powerpc/ffi_darwin.c: Likewise.
572 * src/s390/ffi.c: Likewise.
573 * src/sh/ffi.c: Likewise.
574 * src/sh64/ffi.c: Likewise.
575 * src/sparc/ffi.c: Likewise.
576 * src/x86/ffi64.c: Likewise.
577 * src/x86/ffi.c: Likewise.
578 (FFI_INIT_TRAMPOLINE): Adjust.
579 (ffi_prep_raw_closure_loc): Renamed and adjusted from...
580 (ffi_prep_raw_closure): ... this.
581 * src/powerpc/ffi.c (ffi_prep_closure_loc): Renamed from
582 (ffi_prep_closure): ... this.
583 (flush_icache): Adjust.
584
5852007-03-07 Alexandre Oliva <aoliva@redhat.com>
586
587 * src/dlmalloc.c: New file, imported version 2.8.3 of Doug
588 Lea's malloc.
589
5902007-03-01 Brooks Moses <brooks.moses@codesourcery.com>
591
592 * Makefile.am: Add dummy install-pdf target.
593 * Makefile.in: Regenerate
594
5952007-02-13 Andreas Krebbel <krebbel1@de.ibm.com>
596
597 * src/s390/ffi.c (ffi_prep_args, ffi_prep_cif_machdep,
598 ffi_closure_helper_SYSV): Add long double handling.
599
6002007-02-02 Jakub Jelinek <jakub@redhat.com>
601
602 * src/powerpc/linux64.S (ffi_call_LINUX64): Move restore of r2
603 immediately after bctrl instruction.
604
6052007-01-18 Alexandre Oliva <aoliva@redhat.com>
606
607 * Makefile.am (all-recursive, install-recursive,
608 mostlyclean-recursive, clean-recursive, distclean-recursive,
609 maintainer-clean-recursive): Add missing targets.
610 * Makefile.in: Rebuilt.
611
6122006-12-14 Andreas Tobler <a.tobler@schweiz.org>
613
614 * configure.ac: Add TARGET for x86_64-*-darwin*.
615 * Makefile.am (nodist_libffi_la_SOURCES): Add rules for 64-bit sources
616 for X86_DARWIN.
617 * src/x86/ffitarget.h: Set trampoline size for x86_64-*-darwin*.
618 * src/x86/darwin64.S: New file for x86_64-*-darwin* support.
619 * configure: Regenerate.
620 * Makefile.in: Regenerate.
621 * include/Makefile.in: Regenerate.
622 * testsuite/Makefile.in: Regenerate.
623 * testsuite/libffi.special/unwindtest_ffi_call.cc: New test case for
624 ffi_call only.
625
6262006-12-13 Andreas Tobler <a.tobler@schweiz.org>
627
628 * aclocal.m4: Regenerate with aclocal -I .. as written in the
629 Makefile.am.
630
6312006-10-31 Geoffrey Keating <geoffk@apple.com>
632
633 * src/powerpc/ffi_darwin.c (darwin_adjust_aggregate_sizes): New.
634 (ffi_prep_cif_machdep): Call darwin_adjust_aggregate_sizes for
635 Darwin.
636 * testsuite/libffi.call/nested_struct4.c: Remove Darwin XFAIL.
637 * testsuite/libffi.call/nested_struct6.c: Remove Darwin XFAIL.
638
6392006-10-10 Paolo Bonzini <bonzini@gnu.org>
640 Sandro Tolaini <tolaini@libero.it>
641
642 * configure.ac [i*86-*-darwin*]: Set X86_DARWIN symbol and
643 conditional.
644 * configure: Regenerated.
645 * Makefile.am (nodist_libffi_la_SOURCES) [X86_DARWIN]: New case.
646 (EXTRA_DIST): Add src/x86/darwin.S.
647 * Makefile.in: Regenerated.
648 * include/Makefile.in: Regenerated.
649 * testsuite/Makefile.in: Regenerated.
650
651 * src/x86/ffi.c (ffi_prep_cif_machdep) [X86_DARWIN]: Treat like
652 X86_WIN32, and additionally align stack to 16 bytes.
653 * src/x86/darwin.S: New, based on sysv.S.
654 * src/prep_cif.c (ffi_prep_cif) [X86_DARWIN]: Align > 8-byte structs.
655
6562006-09-12 David Daney <ddaney@avtrex.com>
657
658 PR libffi/23935
659 * include/Makefile.am: Install both ffi.h and ffitarget.h in
660 $(libdir)/gcc/$(target_alias)/$(gcc_version)/include.
661 * aclocal.m4: Regenerated for automake 1.9.6.
662 * Makefile.in: Regenerated.
663 * include/Makefile.in: Regenerated.
664 * testsuite/Makefile.in: Regenerated.
665
6662006-08-17 Andreas Tobler <a.tobler@schweiz.ch>
667
668 * include/ffi_common.h (struct): Revert accidental commit.
669
6702006-08-15 Andreas Tobler <a.tobler@schweiz.ch>
671
672 * include/ffi_common.h: Remove lint directives.
673 * include/ffi.h.in: Likewise.
674
6752006-07-25 Torsten Schoenfeld <kaffeetisch@gmx.de>
676
677 * include/ffi.h.in (ffi_type_ulong, ffi_type_slong): Define correctly
678 for 32-bit architectures.
679 * testsuite/libffi.call/return_ul.c: New test case.
680
6812006-07-19 David Daney <ddaney@avtrex.com>
682
683 * testsuite/libffi.call/closure_fn6.c: Remove xfail for mips,
684 xfail remains for mips64.
685
6862006-05-23 Carlos O'Donell <carlos@codesourcery.com>
687
688 * Makefile.am: Add install-html target. Add install-html to .PHONY
689 * Makefile.in: Regenerate.
690 * aclocal.m4: Regenerate.
691 * include/Makefile.in: Regenerate.
692 * testsuite/Makefile.in: Regenerate.
693
6942006-05-18 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
695
696 * pa/ffi.c (ffi_prep_args_pa32): Load floating point arguments from
697 stack slot.
698
6992006-04-22 Andreas Tobler <a.tobler@schweiz.ch>
700
701 * README: Remove notice about 'Crazy Comments'.
702 * src/debug.c: Remove lint directives. Cleanup white spaces.
703 * src/java_raw_api.c: Likewise.
704 * src/prep_cif.c: Likewise.
705 * src/raw_api.c: Likewise.
706 * src/ffitest.c: Delete. No longer needed, all test cases migrated
707 to the testsuite.
708 * src/arm/ffi.c: Remove lint directives.
709 * src/m32r/ffi.c: Likewise.
710 * src/pa/ffi.c: Likewise.
711 * src/powerpc/ffi.c: Likewise.
712 * src/powerpc/ffi_darwin.c: Likewise.
713 * src/sh/ffi.c: Likewise.
714 * src/sh64/ffi.c: Likewise.
715 * src/x86/ffi.c: Likewise.
716 * testsuite/libffi.call/float2.c: Likewise.
717 * testsuite/libffi.call/promotion.c: Likewise.
718 * testsuite/libffi.call/struct1.c: Likewise.
719
7202006-04-13 Andreas Tobler <a.tobler@schweiz.ch>
721
722 * src/pa/hpux32.S: Correct unwind offset calculation for
723 ffi_closure_pa32.
724 * src/pa/linux.S: Likewise.
725
7262006-04-12 James E Wilson <wilson@specifix.com>
727
728 PR libgcj/26483
729 * src/ia64/ffi.c (stf_spill, ldf_fill): Rewrite as macros.
730 (hfa_type_load): Call stf_spill.
731 (hfa_type_store): Call ldf_fill.
732 (ffi_call): Adjust calls to above routines. Add local temps for
733 macro result.
734
7352006-04-10 Matthias Klose <doko@debian.org>
736
737 * testsuite/lib/libffi-dg.exp (libffi-init): Recognize multilib
738 directory names containing underscores.
739
7402006-04-07 James E Wilson <wilson@specifix.com>
741
742 * testsuite/libffi.call/float4.c: New testcase.
743
7442006-04-05 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
745 Andreas Tobler <a.tobler@schweiz.ch>
746
747 * Makefile.am: Add PA_HPUX port.
748 * Makefile.in: Regenerate.
749 * include/Makefile.in: Likewise.
750 * testsuite/Makefile.in: Likewise.
751 * configure.ac: Add PA_HPUX rules.
752 * configure: Regenerate.
753 * src/pa/ffitarget.h: Rename linux target to PA_LINUX.
754 Add PA_HPUX and PA64_HPUX.
755 Rename FFI_LINUX ABI to FFI_PA32 ABI.
756 (FFI_TRAMPOLINE_SIZE): Define for 32-bit HP-UX targets.
757 (FFI_TYPE_SMALL_STRUCT2): Define.
758 (FFI_TYPE_SMALL_STRUCT4): Likewise.
759 (FFI_TYPE_SMALL_STRUCT8): Likewise.
760 (FFI_TYPE_SMALL_STRUCT3): Redefine.
761 (FFI_TYPE_SMALL_STRUCT5): Likewise.
762 (FFI_TYPE_SMALL_STRUCT6): Likewise.
763 (FFI_TYPE_SMALL_STRUCT7): Likewise.
764 * src/pa/ffi.c (ROUND_DOWN): Delete.
765 (fldw, fstw, fldd, fstd): Use '__asm__'.
766 (ffi_struct_type): Add support for FFI_TYPE_SMALL_STRUCT2,
767 FFI_TYPE_SMALL_STRUCT4 and FFI_TYPE_SMALL_STRUCT8.
768 (ffi_prep_args_LINUX): Rename to ffi_prep_args_pa32. Update comment.
769 Simplify incrementing of stack slot variable. Change type of local
770 'n' to unsigned int.
771 (ffi_size_stack_LINUX): Rename to ffi_size_stack_pa32. Handle long
772 double on PA_HPUX.
773 (ffi_prep_cif_machdep): Likewise.
774 (ffi_call): Likewise.
775 (ffi_closure_inner_LINUX): Rename to ffi_closure_inner_pa32. Change
776 return type to ffi_status. Simplify incrementing of stack slot
777 variable. Only copy floating point argument registers when PA_LINUX
778 is true. Reformat debug statement.
779 Add support for FFI_TYPE_SMALL_STRUCT2, FFI_TYPE_SMALL_STRUCT4 and
780 FFI_TYPE_SMALL_STRUCT8.
781 (ffi_closure_LINUX): Rename to ffi_closure_pa32. Add 'extern' to
782 declaration.
783 (ffi_prep_closure): Make linux trampoline conditional on PA_LINUX.
784 Add nops to cache flush. Add trampoline for PA_HPUX.
785 * src/pa/hpux32.S: New file.
786 * src/pa/linux.S (ffi_call_LINUX): Rename to ffi_call_pa32. Rename
787 ffi_prep_args_LINUX to ffi_prep_args_pa32.
788 Localize labels. Add support for 2, 4 and 8-byte small structs. Handle
789 unaligned destinations in 3, 5, 6 and 7-byte small structs. Order
790 argument type checks so that common argument types appear first.
791 (ffi_closure_LINUX): Rename to ffi_closure_pa32. Rename
792 ffi_closure_inner_LINUX to ffi_closure_inner_pa32.
793
7942006-03-24 Alan Modra <amodra@bigpond.net.au>
795
796 * src/powerpc/ffitarget.h (enum ffi_abi): Add FFI_LINUX. Default
797 for 32-bit using IBM extended double format. Fix FFI_LAST_ABI.
798 * src/powerpc/ffi.c (ffi_prep_args_SYSV): Handle linux variant of
799 FFI_TYPE_LONGDOUBLE.
800 (ffi_prep_args64): Assert using IBM extended double.
801 (ffi_prep_cif_machdep): Don't munge FFI_TYPE_LONGDOUBLE type.
802 Handle FFI_LINUX FFI_TYPE_LONGDOUBLE return and args.
803 (ffi_call): Handle FFI_LINUX.
804 (ffi_closure_helper_SYSV): Non FFI_LINUX long double return needs
805 gpr3 return pointer as for struct return. Handle FFI_LINUX
806 FFI_TYPE_LONGDOUBLE return and args. Don't increment "nf"
807 unnecessarily.
808 * src/powerpc/ppc_closure.S (ffi_closure_SYSV): Load both f1 and f2
809 for FFI_TYPE_LONGDOUBLE. Move epilogue insns into case table.
810 Don't use r6 as pointer to results, instead use sp offset. Don't
811 make a special call to load lr with case table address, instead
812 use offset from previous call.
813 * src/powerpc/sysv.S (ffi_call_SYSV): Save long double return.
814 * src/powerpc/linux64.S (ffi_call_LINUX64): Simplify long double
815 return.
816
8172006-03-15 Kaz Kojima <kkojima@gcc.gnu.org>
818
819 * src/sh64/ffi.c (ffi_prep_cif_machdep): Handle float arguments
820 passed with FP registers correctly.
821 (ffi_closure_helper_SYSV): Likewise.
822 * src/sh64/sysv.S: Likewise.
823
8242006-03-01 Andreas Tobler <a.tobler@schweiz.ch>
825
826 * testsuite/libffi.special/unwindtest.cc (closure_test_fn): Mark cif,
827 args and userdata unused.
828 (closure_test_fn1): Mark cif and userdata unused.
829 (main): Remove unused res.
830
8312006-02-28 Andreas Tobler <a.tobler@schweiz.ch>
832
833 * testsuite/libffi.call/call.exp: Adjust FSF address. Add test runs for
834 -O2, -O3, -Os and the warning flags -W -Wall.
835 * testsuite/libffi.special/special.exp: Likewise.
836 * testsuite/libffi.call/ffitest.h: Add an __UNUSED__ macro to mark
837 unused parameter unused for gcc or else do nothing.
838 * testsuite/libffi.special/ffitestcxx.h: Likewise.
839 * testsuite/libffi.call/cls_12byte.c (cls_struct_12byte_gn): Mark cif
840 and userdata unused.
841 * testsuite/libffi.call/cls_16byte.c (cls_struct_16byte_gn): Likewise.
842 * testsuite/libffi.call/cls_18byte.c (cls_struct_18byte_gn): Likewise.
843 * testsuite/libffi.call/cls_19byte.c (cls_struct_19byte_gn): Likewise.
844 * testsuite/libffi.call/cls_1_1byte.c (cls_struct_1_1byte_gn): Likewise.
845 * testsuite/libffi.call/cls_20byte.c (cls_struct_20byte_gn): Likewise.
846 * testsuite/libffi.call/cls_20byte1.c (cls_struct_20byte_gn): Likewise.
847 * testsuite/libffi.call/cls_24byte.c (cls_struct_24byte_gn): Likewise.
848 * testsuite/libffi.call/cls_2byte.c (cls_struct_2byte_gn): Likewise.
849 * testsuite/libffi.call/cls_3_1byte.c (cls_struct_3_1byte_gn): Likewise.
850 * testsuite/libffi.call/cls_3byte1.c (cls_struct_3byte_gn): Likewise.
851 * testsuite/libffi.call/cls_3byte2.c (cls_struct_3byte_gn1): Likewise.
852 * testsuite/libffi.call/cls_4_1byte.c (cls_struct_4_1byte_gn): Likewise.
853 * testsuite/libffi.call/cls_4byte.c (cls_struct_4byte_gn): Likewise.
854 * testsuite/libffi.call/cls_5_1_byte.c (cls_struct_5byte_gn): Likewise.
855 * testsuite/libffi.call/cls_5byte.c (cls_struct_5byte_gn): Likewise.
856 * testsuite/libffi.call/cls_64byte.c (cls_struct_64byte_gn): Likewise.
857 * testsuite/libffi.call/cls_6_1_byte.c (cls_struct_6byte_gn): Likewise.
858 * testsuite/libffi.call/cls_6byte.c (cls_struct_6byte_gn): Likewise.
859 * testsuite/libffi.call/cls_7_1_byte.c (cls_struct_7byte_gn): Likewise.
860 * testsuite/libffi.call/cls_7byte.c (cls_struct_7byte_gn): Likewise.
861 * testsuite/libffi.call/cls_8byte.c (cls_struct_8byte_gn): Likewise.
862 * testsuite/libffi.call/cls_9byte1.c (cls_struct_9byte_gn): Likewise.
863 * testsuite/libffi.call/cls_9byte2.c (cls_struct_9byte_gn): Likewise.
864 * testsuite/libffi.call/cls_align_double.c (cls_struct_align_gn):
865 Likewise.
866 * testsuite/libffi.call/cls_align_float.c (cls_struct_align_gn):
867 Likewise.
868 * testsuite/libffi.call/cls_align_longdouble.c (cls_struct_align_gn):
869 Likewise.
870 * testsuite/libffi.call/cls_align_pointer.c (cls_struct_align_fn): Cast
871 void* to avoid compiler warning.
872 (main): Likewise.
873 (cls_struct_align_gn): Mark cif and userdata unused.
874 * testsuite/libffi.call/cls_align_sint16.c (cls_struct_align_gn):
875 Likewise.
876 * testsuite/libffi.call/cls_align_sint32.c (cls_struct_align_gn):
877 Likewise.
878 * testsuite/libffi.call/cls_align_sint64.c (cls_struct_align_gn):
879 Likewise.
880 * testsuite/libffi.call/cls_align_uint16.c (cls_struct_align_gn):
881 Likewise.
882 * testsuite/libffi.call/cls_align_uint32.c (cls_struct_align_gn):
883 Likewise.
884 * testsuite/libffi.call/cls_double.c (cls_ret_double_fn): Likewise.
885 * testsuite/libffi.call/cls_float.c (cls_ret_float_fn): Likewise.
886 * testsuite/libffi.call/cls_multi_schar.c (test_func_gn): Mark cif and
887 data unused.
888 (main): Cast res_call to silence gcc.
889 * testsuite/libffi.call/cls_multi_sshort.c (test_func_gn): Mark cif and
890 data unused.
891 (main): Cast res_call to silence gcc.
892 * testsuite/libffi.call/cls_multi_sshortchar.c (test_func_gn): Mark cif
893 and data unused.
894 (main): Cast res_call to silence gcc.
895 * testsuite/libffi.call/cls_multi_uchar.c (test_func_gn): Mark cif and
896 data unused.
897 (main): Cast res_call to silence gcc.
898 * testsuite/libffi.call/cls_multi_ushort.c (test_func_gn): Mark cif and
899 data unused.
900 (main): Cast res_call to silence gcc.
901 * testsuite/libffi.call/cls_multi_ushortchar.c (test_func_gn): Mark cif
902 and data unused.
903 (main): Cast res_call to silence gcc.
904 * testsuite/libffi.call/cls_schar.c (cls_ret_schar_fn): Mark cif and
905 userdata unused.
906 (cls_ret_schar_fn): Cast printf parameter to silence gcc.
907 * testsuite/libffi.call/cls_sint.c (cls_ret_sint_fn): Mark cif and
908 userdata unused.
909 (cls_ret_sint_fn): Cast printf parameter to silence gcc.
910 * testsuite/libffi.call/cls_sshort.c (cls_ret_sshort_fn): Mark cif and
911 userdata unused.
912 (cls_ret_sshort_fn): Cast printf parameter to silence gcc.
913 * testsuite/libffi.call/cls_uchar.c (cls_ret_uchar_fn): Mark cif and
914 userdata unused.
915 (cls_ret_uchar_fn): Cast printf parameter to silence gcc.
916 * testsuite/libffi.call/cls_uint.c (cls_ret_uint_fn): Mark cif and
917 userdata unused.
918 (cls_ret_uint_fn): Cast printf parameter to silence gcc.
919 * testsuite/libffi.call/cls_ulonglong.c (cls_ret_ulonglong_fn): Mark cif
920 and userdata unused.
921 * testsuite/libffi.call/cls_ushort.c (cls_ret_ushort_fn): Mark cif and
922 userdata unused.
923 (cls_ret_ushort_fn): Cast printf parameter to silence gcc.
924 * testsuite/libffi.call/float.c (floating): Remove unused parameter e.
925 * testsuite/libffi.call/float1.c (main): Remove unused variable i.
926 Cleanup white spaces.
927 * testsuite/libffi.call/negint.c (checking): Remove unused variable i.
928 * testsuite/libffi.call/nested_struct.c (cls_struct_combined_gn): Mark
929 cif and userdata unused.
930 * testsuite/libffi.call/nested_struct1.c (cls_struct_combined_gn):
931 Likewise.
932 * testsuite/libffi.call/nested_struct10.c (B_gn): Likewise.
933 * testsuite/libffi.call/nested_struct2.c (B_fn): Adjust printf
934 formatters to silence gcc.
935 (B_gn): Mark cif and userdata unused.
936 * testsuite/libffi.call/nested_struct3.c (B_gn): Mark cif and userdata
937 unused.
938 * testsuite/libffi.call/nested_struct4.c: Mention related PR.
939 (B_gn): Mark cif and userdata unused.
940 * testsuite/libffi.call/nested_struct5.c (B_gn): Mark cif and userdata
941 unused.
942 * testsuite/libffi.call/nested_struct6.c: Mention related PR.
943 (B_gn): Mark cif and userdata unused.
944 * testsuite/libffi.call/nested_struct7.c (B_gn): Mark cif and userdata
945 unused.
946 * testsuite/libffi.call/nested_struct8.c (B_gn): Likewise.
947 * testsuite/libffi.call/nested_struct9.c (B_gn): Likewise.
948 * testsuite/libffi.call/problem1.c (stub): Likewise.
949 * testsuite/libffi.call/pyobjc-tc.c (main): Cast the result to silence
950 gcc.
951 * testsuite/libffi.call/return_fl2.c (return_fl): Add the note mentioned
952 in the last commit for this test case in the test case itself.
953 * testsuite/libffi.call/closure_fn0.c (closure_test_fn0): Mark cif as
954 unused.
955 * testsuite/libffi.call/closure_fn1.c (closure_test_fn1): Likewise.
956 * testsuite/libffi.call/closure_fn2.c (closure_test_fn2): Likewise.
957 * testsuite/libffi.call/closure_fn3.c (closure_test_fn3): Likewise.
958 * testsuite/libffi.call/closure_fn4.c (closure_test_fn0): Likewise.
959 * testsuite/libffi.call/closure_fn5.c (closure_test_fn5): Likewise.
960 * testsuite/libffi.call/closure_fn6.c (closure_test_fn0): Likewise.
961
9622006-02-22 Kaz Kojima <kkojima@gcc.gnu.org>
963
964 * src/sh/sysv.S: Fix register numbers in the FDE for
965 ffi_closure_SYSV.
966
9672006-02-20 Andreas Tobler <a.tobler@schweiz.ch>
968
969 * testsuite/libffi.call/return_fl2.c (return_fl): Remove static
970 declaration to avoid a false negative on ix86. See PR323.
971
9722006-02-18 Kaz Kojima <kkojima@gcc.gnu.org>
973
974 * src/sh/ffi.c (ffi_closure_helper_SYSV): Remove unused variable
975 and cast integer to void * if needed. Update the pointer to
976 the FP register saved area correctly.
977
9782006-02-17 Andreas Tobler <a.tobler@schweiz.ch>
979
980 * testsuite/libffi.call/nested_struct6.c: XFAIL this test until PR25630
981 is fixed.
982 * testsuite/libffi.call/nested_struct4.c: Likewise.
983
9842006-02-16 Andreas Tobler <a.tobler@schweiz.ch>
985
986 * testsuite/libffi.call/return_dbl.c: New test case.
987 * testsuite/libffi.call/return_dbl1.c: Likewise.
988 * testsuite/libffi.call/return_dbl2.c: Likewise.
989 * testsuite/libffi.call/return_fl.c: Likewise.
990 * testsuite/libffi.call/return_fl1.c: Likewise.
991 * testsuite/libffi.call/return_fl2.c: Likewise.
992 * testsuite/libffi.call/return_fl3.c: Likewise.
993 * testsuite/libffi.call/closure_fn6.c: Likewise.
994
995 * testsuite/libffi.call/nested_struct2.c: Remove ffi_type_mylong
996 definition.
997 * testsuite/libffi.call/ffitest.h: Add ffi_type_mylong definition
998 here to be used by other test cases too.
999
1000 * testsuite/libffi.call/nested_struct10.c: New test case.
1001 * testsuite/libffi.call/nested_struct9.c: Likewise.
1002 * testsuite/libffi.call/nested_struct8.c: Likewise.
1003 * testsuite/libffi.call/nested_struct7.c: Likewise.
1004 * testsuite/libffi.call/nested_struct6.c: Likewise.
1005 * testsuite/libffi.call/nested_struct5.c: Likewise.
1006 * testsuite/libffi.call/nested_struct4.c: Likewise.
1007
10082006-01-21 Andreas Tobler <a.tobler@schweiz.ch>
1009
1010 * configure.ac: Enable libffi for sparc64-*-freebsd*.
1011 * configure: Rebuilt.
1012
10132006-01-18 Jakub Jelinek <jakub@redhat.com>
1014
1015 * src/powerpc/sysv.S (smst_two_register): Don't call __ashldi3,
1016 instead do the shifting inline.
1017 * src/powerpc/ppc_closure.S (ffi_closure_SYSV): Don't compute %r5
1018 shift count unconditionally. Simplify load sequences for 1, 2, 3, 4
1019 and 8 byte structs, for the remaining struct sizes don't call
1020 __lshrdi3, instead do the shifting inline.
1021
10222005-12-07 Thiemo Seufer <ths@networkno.de>
1023
1024 * src/mips/ffitarget.h: Remove obsolete sgidefs.h include. Add
1025 missing parentheses.
1026 * src/mips/o32.S (ffi_call_O32): Code formatting. Define
1027 and use A3_OFF, FP_OFF, RA_OFF. Micro-optimizations.
1028 (ffi_closure_O32): Likewise, but with newly defined A3_OFF2,
1029 A2_OFF2, A1_OFF2, A0_OFF2, RA_OFF2, FP_OFF2, S0_OFF2, GP_OFF2,
1030 V1_OFF2, V0_OFF2, FA_1_1_OFF2, FA_1_0_OFF2, FA_0_1_OFF2,
1031 FA_0_0_OFF2.
1032 * src/mips/ffi.c (ffi_prep_args): Code formatting. Fix
1033 endianness bugs.
1034 (ffi_prep_closure): Improve trampoline instruction scheduling.
1035 (ffi_closure_mips_inner_O32): Fix endianness bugs.
1036
10372005-12-03 Alan Modra <amodra@bigpond.net.au>
1038
1039 * src/powerpc/ffi.c: Formatting.
1040 (ffi_prep_args_SYSV): Avoid possible aliasing problems by using unions.
1041 (ffi_prep_args64): Likewise.
1042
10432005-09-30 Geoffrey Keating <geoffk@apple.com>
1044
1045 * testsuite/lib/libffi-dg.exp (libffi_target_compile): For
1046 darwin, use -shared-libgcc not -lgcc_s, and explain why.
1047
10482005-09-26 Tom Tromey <tromey@redhat.com>
1049
1050 * testsuite/libffi.call/float1.c (value_type): New typedef.
1051 (CANARY): New define.
1052 (main): Check for result buffer overflow.
1053 * src/powerpc/linux64.S: Handle linux64 long double returns.
1054 * src/powerpc/ffi.c (FLAG_RETURNS_128BITS): New constant.
1055 (ffi_prep_cif_machdep): Handle linux64 long double returns.
1056
10572005-08-25 Alan Modra <amodra@bigpond.net.au>
1058
1059 PR target/23404
1060 * src/powerpc/ffi.c (ffi_prep_args_SYSV): Correct placement of stack
1061 homed fp args.
1062 (ffi_status ffi_prep_cif_machdep): Correct stack sizing for same.
1063
10642005-08-11 Jakub Jelinek <jakub@redhat.com>
1065
1066 * configure.ac (HAVE_HIDDEN_VISIBILITY_ATTRIBUTE): New test.
1067 (AH_BOTTOM): Add FFI_HIDDEN definition.
1068 * configure: Rebuilt.
1069 * fficonfig.h.in: Rebuilt.
1070 * src/powerpc/ffi.c (hidden): Remove.
1071 (ffi_closure_LINUX64, ffi_prep_args64, ffi_call_LINUX64,
1072 ffi_closure_helper_LINUX64): Use FFI_HIDDEN instead of hidden.
1073 * src/powerpc/linux64_closure.S (ffi_closure_LINUX64,
1074 .ffi_closure_LINUX64): Use FFI_HIDDEN instead of .hidden.
1075 * src/x86/ffi.c (ffi_closure_SYSV, ffi_closure_raw_SYSV): Remove,
1076 add FFI_HIDDEN to its prototype.
1077 (ffi_closure_SYSV_inner): New.
1078 * src/x86/sysv.S (ffi_closure_SYSV, ffi_closure_raw_SYSV): New.
1079 * src/x86/win32.S (ffi_closure_SYSV, ffi_closure_raw_SYSV): New.
1080
10812005-08-10 Alfred M. Szmidt <ams@gnu.org>
1082
1083 PR libffi/21819:
1084 * configure: Rebuilt.
1085 * configure.ac: Handle i*86-*-gnu*.
1086
10872005-08-09 Jakub Jelinek <jakub@redhat.com>
1088
1089 * src/powerpc/ppc_closure.S (ffi_closure_SYSV): Use
1090 DW_CFA_offset_extended_sf rather than
1091 DW_CFA_GNU_negative_offset_extended.
1092 * src/powerpc/sysv.S (ffi_call_SYSV): Likewise.
1093
10942005-07-22 SUGIOKA Toshinobu <sugioka@itonet.co.jp>
1095
1096 * src/sh/sysv.S (ffi_call_SYSV): Stop argument popping correctly
1097 on sh3.
1098 (ffi_closure_SYSV): Change the stack layout for sh3 struct argument.
1099 * src/sh/ffi.c (ffi_prep_args): Fix sh3 argument copy, when it is
1100 partially on register.
1101 (ffi_closure_helper_SYSV): Likewise.
1102 (ffi_prep_cif_machdep): Don't set too many cif->flags.
1103
11042005-07-20 Kaz Kojima <kkojima@gcc.gnu.org>
1105
1106 * src/sh/ffi.c (ffi_call): Handle small structures correctly.
1107 Remove empty line.
1108 * src/sh64/ffi.c (simple_type): Remove.
1109 (return_type): Handle small structures correctly.
1110 (ffi_prep_args): Likewise.
1111 (ffi_call): Likewise.
1112 (ffi_closure_helper_SYSV): Likewise.
1113 * src/sh64/sysv.S (ffi_call_SYSV): Handle 1, 2 and 4-byte return.
1114 Emit position independent code if PIC and remove wrong datalabel
1115 prefixes from EH data.
1116
11172005-07-19 Andreas Tobler <a.tobler@schweiz.ch>
1118
1119 * Makefile.am (nodist_libffi_la_SOURCES): Add POWERPC_FREEBSD.
1120 * Makefile.in: Regenerate.
1121 * include/Makefile.in: Likewise.
1122 * testsuite/Makefile.in: Likewise.
1123 * configure.ac: Add POWERPC_FREEBSD rules.
1124 * configure: Regenerate.
1125 * src/powerpc/ffitarget.h: Add POWERPC_FREEBSD rules.
1126 (FFI_SYSV_TYPE_SMALL_STRUCT): Define.
1127 * src/powerpc/ffi.c: Add flags to handle small structure returns
1128 in ffi_call_SYSV.
1129 (ffi_prep_cif_machdep): Handle small structures for SYSV 4 ABI.
1130 Aka FFI_SYSV.
1131 (ffi_closure_helper_SYSV): Likewise.
1132 * src/powerpc/ppc_closure.S: Add return types for small structures.
1133 * src/powerpc/sysv.S: Add bits to handle small structures for
1134 final SYSV 4 ABI.
1135
11362005-07-10 Andreas Tobler <a.tobler@schweiz.ch>
1137
1138 * testsuite/libffi.call/cls_5_1_byte.c: New test file.
1139 * testsuite/libffi.call/cls_6_1_byte.c: Likewise.
1140 * testsuite/libffi.call/cls_7_1_byte.c: Likewise.
1141
11422005-07-05 Randolph Chung <tausq@debian.org>
1143
1144 * src/pa/ffi.c (ffi_struct_type): Rename FFI_TYPE_SMALL_STRUCT1
1145 as FFI_TYPE_SMALL_STRUCT3. Break out handling for 5-7 byte
1146 structures. Kill compilation warnings.
1147 (ffi_closure_inner_LINUX): Print return values as hex in debug
1148 message. Rename FFI_TYPE_SMALL_STRUCT1 as FFI_TYPE_SMALL_STRUCT3.
1149 Properly handle 5-7 byte structure returns.
1150 * src/pa/ffitarget.h (FFI_TYPE_SMALL_STRUCT1)
1151 (FFI_TYPE_SMALL_STRUCT2): Remove.
1152 (FFI_TYPE_SMALL_STRUCT3, FFI_TYPE_SMALL_STRUCT5)
1153 (FFI_TYPE_SMALL_STRUCT6, FFI_TYPE_SMALL_STRUCT7): Define.
1154 * src/pa/linux.S: Mark source file as using PA1.1 assembly.
1155 (checksmst1, checksmst2): Remove.
1156 (checksmst3): Optimize handling of 3-byte struct returns.
1157 (checksmst567): Properly handle 5-7 byte struct returns.
1158
11592005-06-15 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
1160
1161 PR libgcj/21943
1162 * src/mips/n32.S: Enforce PIC code.
1163 * src/mips/o32.S: Likewise.
1164
11652005-06-15 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
1166
1167 * configure.ac: Treat i*86-*-solaris2.10 and up as X86_64.
1168 * configure: Regenerate.
1169
11702005-06-01 Alan Modra <amodra@bigpond.net.au>
1171
1172 * src/powerpc/ppc_closure.S (ffi_closure_SYSV): Don't use JUMPTARGET
1173 to call ffi_closure_helper_SYSV. Append @local instead.
1174 * src/powerpc/sysv.S (ffi_call_SYSV): Likewise for ffi_prep_args_SYSV.
1175
11762005-05-17 Kelley Cook <kcook@gcc.gnu.org>
1177
1178 * configure.ac: Use AC_C_BIGENDIAN instead of AC_C_BIGENDIAN_CROSS.
1179 Use AC_CHECK_SIZEOF instead of AC_COMPILE_CHECK_SIZEOF.
1180 * Makefile.am (ACLOCAL_AMFLAGS): Remove -I ../config.
1181 * aclocal.m4, configure, fficonfig.h.in, Makefile.in,
1182 include/Makefile.in, testsuite/Makefile.in: Regenerate.
1183
11842005-05-09 Mike Stump <mrs@apple.com>
1185
1186 * configure: Regenerate.
1187
11882005-05-08 Richard Henderson <rth@redhat.com>
1189
1190 PR libffi/21285
1191 * src/alpha/osf.S: Update unwind into to match code.
1192
11932005-05-04 Andreas Degert <ad@papyrus-gmbh.de>
1194 Richard Henderson <rth@redhat.com>
1195
1196 * src/x86/ffi64.c (ffi_prep_cif_machdep): Save sse-used flag in
1197 bit 11 of flags.
1198 (ffi_call): Mask return type field. Pass ssecount to ffi_call_unix64.
1199 (ffi_prep_closure): Set carry bit if sse-used flag set.
1200 * src/x86/unix64.S (ffi_call_unix64): Add ssecount argument.
1201 Only load sse registers if ssecount non-zero.
1202 (ffi_closure_unix64): Only save sse registers if carry set on entry.
1203
12042005-04-29 Ralf Corsepius <ralf.corsepius@rtems.org>
1205
1206 * configure.ac: Add i*86-*-rtems*, sparc*-*-rtems*,
1207 powerpc-*rtems*, arm*-*-rtems*, sh-*-rtems*.
1208 * configure: Regenerate.
1209
12102005-04-20 Hans-Peter Nilsson <hp@axis.com>
1211
1212 * testsuite/lib/libffi-dg.exp (libffi-dg-test-1): In regsub use,
1213 have Tcl8.3-compatible intermediate variable.
1214
12152005-04-18 Simon Posnjak <simon.posnjak@siol.net>
1216 Hans-Peter Nilsson <hp@axis.com>
1217
1218 * Makefile.am: Add CRIS support.
1219 * configure.ac: Likewise.
1220 * Makefile.in, configure, testsuite/Makefile.in,
1221 include/Makefile.in: Regenerate.
1222 * src/cris: New directory.
1223 * src/cris/ffi.c, src/cris/sysv.S, src/cris/ffitarget.h: New files.
1224 * src/prep_cif.c (ffi_prep_cif): Wrap in #ifndef __CRIS__.
1225
1226 * testsuite/lib/libffi-dg.exp (libffi-dg-test-1): Replace \n with
1227 \r?\n in output tests.
1228
12292005-04-12 Mike Stump <mrs@apple.com>
1230
1231 * configure: Regenerate.
1232
12332005-03-30 Hans Boehm <Hans.Boehm@hp.com>
1234
1235 * src/ia64/ffitarget.h (ffi_arg): Use long long instead of DI.
1236
12372005-03-30 Steve Ellcey <sje@cup.hp.com>
1238
1239 * src/ia64/ffitarget.h (ffi_arg) ADD DI attribute.
1240 (ffi_sarg) Ditto.
1241 * src/ia64/unix.S (ffi_closure_unix): Extend gp
1242 to 64 bits in ILP32 mode.
1243 Load 64 bits even for short data.
1244
12452005-03-23 Mike Stump <mrs@apple.com>
1246
1247 * src/powerpc/darwin.S: Update for -m64 multilib.
1248 * src/powerpc/darwin_closure.S: Likewise.
1249
12502005-03-21 Zack Weinberg <zack@codesourcery.com>
1251
1252 * configure.ac: Do not invoke TL_AC_GCC_VERSION.
1253 Do not set tool_include_dir.
1254 * aclocal.m4, configure, Makefile.in, testsuite/Makefile.in:
1255 Regenerate.
1256 * include/Makefile.am: Set gcc_version and toollibffidir.
1257 * include/Makefile.in: Regenerate.
1258
12592005-02-22 Andrew Haley <aph@redhat.com>
1260
1261 * src/powerpc/ffi.c (ffi_prep_cif_machdep): Bump alignment to
1262 odd-numbered register pairs for 64-bit integer types.
1263
12642005-02-23 Andreas Tobler <a.tobler@schweiz.ch>
1265
1266 PR libffi/20104
1267 * testsuite/libffi.call/return_ll1.c: New test case.
1268
12692005-02-11 Janis Johnson <janis187@us.ibm.com>
1270
1271 * testsuite/libffi.call/cls_align_longdouble.c: Remove dg-options.
1272 * testsuite/libffi.call/float.c: Ditto.
1273 * testsuite/libffi.call/float2.c: Ditto.
1274 * testsuite/libffi.call/float3.c: Ditto.
1275
12762005-02-08 Andreas Tobler <a.tobler@schweiz.ch>
1277
1278 * src/frv/ffitarget.h: Remove PPC stuff which does not belong to frv.
1279
12802005-01-12 Eric Botcazou <ebotcazou@libertysurf.fr>
1281
1282 * testsuite/libffi.special/special.exp (cxx_options): Add
1283 -shared-libgcc.
1284
12852004-12-31 Richard Henderson <rth@redhat.com>
1286
1287 * src/types.c (FFI_AGGREGATE_TYPEDEF): Remove.
1288 (FFI_TYPEDEF): Rename from FFI_INTEGRAL_TYPEDEF. Replace size and
1289 offset parameters with a type parameter; deduce size and structure
1290 alignment. Update all users.
1291
12922004-12-31 Richard Henderson <rth@redhat.com>
1293
1294 * src/types.c (FFI_TYPE_POINTER): Define with sizeof.
1295 (FFI_TYPE_LONGDOUBLE): Fix for ia64.
1296 * src/ia64/ffitarget.h (struct ffi_ia64_trampoline_struct): Move
1297 into ffi_prep_closure.
1298 * src/ia64/ia64_flags.h, src/ia64/ffi.c, src/ia64/unix.S: Rewrite
1299 from scratch.
1300
13012004-12-27 Richard Henderson <rth@redhat.com>
1302
1303 * src/x86/unix64.S: Fix typo in unwind info.
1304
13052004-12-25 Richard Henderson <rth@redhat.com>
1306
1307 * src/x86/ffi64.c (struct register_args): Rename from stackLayout.
1308 (enum x86_64_reg_class): Add X86_64_COMPLEX_X87_CLASS.
1309 (merge_classes): Check for it.
1310 (SSE_CLASS_P): New.
1311 (classify_argument): Pass byte_offset by value; perform all updates
1312 inside struct case.
1313 (examine_argument): Add classes argument; handle
1314 X86_64_COMPLEX_X87_CLASS.
1315 (ffi_prep_args): Merge into ...
1316 (ffi_call): ... here. Share stack frame with ffi_call_unix64.
1317 (ffi_prep_cif_machdep): Setup cif->flags for proper structure return.
1318 (ffi_fill_return_value): Remove.
1319 (ffi_prep_closure): Remove dead assert.
1320 (ffi_closure_unix64_inner): Rename from ffi_closure_UNIX64_inner.
1321 Rewrite to use struct register_args instead of va_list. Create
1322 flags for handling structure returns.
1323 * src/x86/unix64.S: Remove dead strings.
1324 (ffi_call_unix64): Rename from ffi_call_UNIX64. Rewrite to share
1325 stack frame with ffi_call. Handle structure returns properly.
1326 (float2sse, floatfloat2sse, double2sse): Remove.
1327 (sse2float, sse2double, sse2floatfloat): Remove.
1328 (ffi_closure_unix64): Rename from ffi_closure_UNIX64. Rewrite
1329 to handle structure returns properly.
1330
13312004-12-08 David Edelsohn <edelsohn@gnu.org>
1332
1333 * Makefile.am (AM_MAKEFLAGS): Remove duplicate LIBCFLAGS and
1334 PICFLAG.
1335 * Makefile.in: Regenerated.
1336
13372004-12-02 Richard Sandiford <rsandifo@redhat.com>
1338
1339 * configure.ac: Use TL_AC_GCC_VERSION to set gcc_version.
1340 * configure, aclocal.m4, Makefile.in: Regenerate.
1341 * include/Makefile.in, testsuite/Makefile.in: Regenerate.
1342
13432004-11-29 Kelley Cook <kcook@gcc.gnu.org>
1344
1345 * configure: Regenerate for libtool change.
1346
13472004-11-25 Kelley Cook <kcook@gcc.gnu.org>
1348
1349 * configure: Regenerate for libtool reversion.
1350
13512004-11-24 Kelley Cook <kcook@gcc.gnu.org>
1352
1353 * configure: Regenerate for libtool change.
1354
13552004-11-23 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
1356
1357 * testsuite/lib/libffi-dg.exp: Use new procs in target-libpath.exp.
1358
13592004-11-23 Richard Sandiford <rsandifo@redhat.com>
1360
1361 * src/mips/o32.S (ffi_call_O32, ffi_closure_O32): Use jalr instead
1362 of jal. Use an absolute encoding for the frame information.
1363
13642004-11-23 Kelley Cook <kcook@gcc.gnu.org>
1365
1366 * Makefile.am: Remove no-dependencies. Add ACLOCAL_AMFLAGS.
1367 * acinclude.m4: Delete logic for sincludes.
1368 * aclocal.m4, Makefile.in, configure: Regenerate.
1369 * include/Makefile: Likewise.
1370 * testsuite/Makefile: Likewise.
1371
13722004-11-22 Eric Botcazou <ebotcazou@libertysurf.fr>
1373
1374 * src/sparc/ffi.c (ffi_prep_closure): Align doubles and 64-bit integers
1375 on a 8-byte boundary.
1376 * src/sparc/v8.S (ffi_closure_v8): Reserve frame space for arguments.
1377
13782004-10-27 Richard Earnshaw <rearnsha@arm.com>
1379
1380 * src/arm/ffi.c (ffi_prep_cif_machdep): Handle functions that return
1381 long long values. Round stack allocation to a multiple of 8 bytes
1382 for ATPCS compatibility.
1383 * src/arm/sysv.S (ffi_call_SYSV): Rework to avoid use of APCS register
1384 names. Handle returning long long types. Add Thumb and interworking
1385 support. Improve soft-float code.
1386
13872004-10-27 Richard Earnshaw <rearnsha@arm.com>
1388
1389 * testsuite/lib/libffi-db.exp (load_gcc_lib): New function.
1390 (libffi_exit): New function.
1391 (libffi_init): Build the testglue wrapper if needed.
1392
13932004-10-25 Eric Botcazou <ebotcazou@libertysurf.fr>
1394
1395 PR other/18138
1396 * testsuite/lib/libffi-dg.exp: Accept more than one multilib libgcc.
1397
13982004-10-25 Kazuhiro Inaoka <inaoka.kazuhiro@renesas.com>
1399
1400 * src/m32r/libffitarget.h (FFI_CLOSURES): Set to 0.
1401
14022004-10-20 Kaz Kojima <kkojima@gcc.gnu.org>
1403
1404 * src/sh/sysv.S (ffi_call_SYSV): Don't align for double data.
1405 * testsuite/libffi.call/float3.c: New test case.
1406
14072004-10-18 Kaz Kojima <kkojima@gcc.gnu.org>
1408
1409 * src/sh/ffi.c (ffi_prep_closure): Set T bit in trampoline for
1410 the function returning a structure pointed with R2.
1411 * src/sh/sysv.S (ffi_closure_SYSV): Use R2 as the pointer to
1412 the structure return value if T bit set. Emit position
1413 independent code and EH data if PIC.
1414
14152004-10-13 Kazuhiro Inaoka <inaoka.kazuhiro@renesas.com>
1416
1417 * Makefile.am: Add m32r support.
1418 * configure.ac: Likewise.
1419 * Makefile.in: Regenerate.
1420 * confiugre: Regenerate.
1421 * src/types.c: Add m32r port to FFI_INTERNAL_TYPEDEF
1422 (uint64, sint64, double, longdouble)
1423 * src/m32r: New directory.
1424 * src/m32r/ffi.c: New file.
1425 * src/m32r/sysv.S: Likewise.
1426 * src/m32r/ffitarget.h: Likewise.
1427
14282004-10-02 Kaz Kojima <kkojima@gcc.gnu.org>
1429
1430 * testsuite/libffi.call/negint.c: New test case.
1431
14322004-09-14 H.J. Lu <hongjiu.lu@intel.com>
1433
1434 PR libgcj/17465
1435 * testsuite/lib/libffi-dg.exp: Don't use global ld_library_path.
1436 Set up LD_LIBRARY_PATH, SHLIB_PATH, LD_LIBRARYN32_PATH,
1437 LD_LIBRARY64_PATH, LD_LIBRARY_PATH_32, LD_LIBRARY_PATH_64 and
1438 DYLD_LIBRARY_PATH.
1439
14402004-09-05 Andreas Tobler <a.tobler@schweiz.ch>
1441
1442 * testsuite/libffi.call/many_win32.c: Remove whitespaces.
1443 * testsuite/libffi.call/promotion.c: Likewise.
1444 * testsuite/libffi.call/return_ll.c: Remove unused var. Cleanup
1445 whitespaces.
1446 * testsuite/libffi.call/return_sc.c: Likewise.
1447 * testsuite/libffi.call/return_uc.c: Likewise.
1448
14492004-09-05 Andreas Tobler <a.tobler@schweiz.ch>
1450
1451 * src/powerpc/darwin.S: Fix comments and identation.
1452 * src/powerpc/darwin_closure.S: Likewise.
1453
14542004-09-02 Andreas Tobler <a.tobler@schweiz.ch>
1455
1456 * src/powerpc/ffi_darwin.c: Add flag for longdouble return values.
1457 (ffi_prep_args): Handle longdouble arguments.
1458 (ffi_prep_cif_machdep): Set flags for longdouble. Calculate space for
1459 longdouble.
1460 (ffi_closure_helper_DARWIN): Add closure handling for longdouble.
1461 * src/powerpc/darwin.S (_ffi_call_DARWIN): Add handling of longdouble
1462 values.
1463 * src/powerpc/darwin_closure.S (_ffi_closure_ASM): Likewise.
1464 * src/types.c: Defined longdouble size and alignment for darwin.
1465
14662004-09-02 Andreas Tobler <a.tobler@schweiz.ch>
1467
1468 * src/powerpc/aix.S: Remove whitespaces.
1469 * src/powerpc/aix_closure.S: Likewise.
1470 * src/powerpc/asm.h: Likewise.
1471 * src/powerpc/ffi.c: Likewise.
1472 * src/powerpc/ffitarget.h: Likewise.
1473 * src/powerpc/linux64.S: Likewise.
1474 * src/powerpc/linux64_closure.S: Likewise.
1475 * src/powerpc/ppc_closure.S: Likewise.
1476 * src/powerpc/sysv.S: Likewise.
1477
14782004-08-30 Anthony Green <green@redhat.com>
1479
1480 * Makefile.am: Add frv support.
1481 * Makefile.in, testsuite/Makefile.in: Rebuilt.
1482 * configure.ac: Read configure.host.
1483 * configure.in: Read configure.host.
1484 * configure.host: New file. frv-elf needs libgloss.
1485 * include/ffi.h.in: Force ffi_closure to have a nice big (8)
1486 alignment. This is needed to frv and shouldn't harm the others.
1487 * include/ffi_common.h (ALIGN_DOWN): New macro.
1488 * src/frv/ffi.c, src/frv/ffitarget.h, src/frv/eabi.S: New files.
1489
14902004-08-24 David Daney <daney@avtrex.com>
1491
1492 * testsuite/libffi.call/closure_fn0.c: Xfail mips64* instead of mips*.
1493 * testsuite/libffi.call/closure_fn1.c: Likewise.
1494 * testsuite/libffi.call/closure_fn2.c Likewise.
1495 * testsuite/libffi.call/closure_fn3.c: Likewise.
1496 * testsuite/libffi.call/closure_fn4.c: Likewise.
1497 * testsuite/libffi.call/closure_fn5.c: Likewise.
1498 * testsuite/libffi.call/cls_18byte.c: Likewise.
1499 * testsuite/libffi.call/cls_19byte.c: Likewise.
1500 * testsuite/libffi.call/cls_1_1byte.c: Likewise.
1501 * testsuite/libffi.call/cls_20byte.c: Likewise.
1502 * testsuite/libffi.call/cls_20byte1.c: Likewise.
1503 * testsuite/libffi.call/cls_24byte.c: Likewise.
1504 * testsuite/libffi.call/cls_2byte.c: Likewise.
1505 * testsuite/libffi.call/cls_3_1byte.c: Likewise.
1506 * testsuite/libffi.call/cls_3byte1.c: Likewise.
1507 * testsuite/libffi.call/cls_3byte2.c: Likewise.
1508 * testsuite/libffi.call/cls_4_1byte.c: Likewise.
1509 * testsuite/libffi.call/cls_4byte.c: Likewise.
1510 * testsuite/libffi.call/cls_64byte.c: Likewise.
1511 * testsuite/libffi.call/cls_6byte.c: Likewise.
1512 * testsuite/libffi.call/cls_7byte.c: Likewise.
1513 * testsuite/libffi.call/cls_8byte.c: Likewise.
1514 * testsuite/libffi.call/cls_9byte1.c: Likewise.
1515 * testsuite/libffi.call/cls_9byte2.c: Likewise.
1516 * testsuite/libffi.call/cls_align_double.c: Likewise.
1517 * testsuite/libffi.call/cls_align_float.c: Likewise.
1518 * testsuite/libffi.call/cls_align_longdouble.c: Likewise.
1519 * testsuite/libffi.call/cls_align_pointer.c: Likewise.
1520 * testsuite/libffi.call/cls_align_sint16.c: Likewise.
1521 * testsuite/libffi.call/cls_align_sint32.c: Likewise.
1522 * testsuite/libffi.call/cls_align_sint64.c: Likewise.
1523 * testsuite/libffi.call/cls_align_uint16.c: Likewise.
1524 * testsuite/libffi.call/cls_align_uint32.c: Likewise.
1525 * testsuite/libffi.call/cls_align_uint64.c: Likewise.
1526 * testsuite/libffi.call/cls_double.c: Likewise.
1527 * testsuite/libffi.call/cls_float.c: Likewise.
1528 * testsuite/libffi.call/cls_multi_schar.c: Likewise.
1529 * testsuite/libffi.call/cls_multi_sshort.c: Likewise.
1530 * testsuite/libffi.call/cls_multi_sshortchar.c: Likewise.
1531 * testsuite/libffi.call/cls_multi_uchar.c: Likewise.
1532 * testsuite/libffi.call/cls_multi_ushort.c: Likewise.
1533 * testsuite/libffi.call/cls_multi_ushortchar.c: Likewise.
1534 * testsuite/libffi.call/cls_schar.c: Likewise.
1535 * testsuite/libffi.call/cls_sint.c: Likewise.
1536 * testsuite/libffi.call/cls_sshort.c: Likewise.
1537 * testsuite/libffi.call/cls_uchar.c: Likewise.
1538 * testsuite/libffi.call/cls_uint.c: Likewise.
1539 * testsuite/libffi.call/cls_ulonglong.c: Likewise.
1540 * testsuite/libffi.call/cls_ushort.c: Likewise.
1541 * testsuite/libffi.call/nested_struct.c: Likewise.
1542 * testsuite/libffi.call/nested_struct1.c: Likewise.
1543 * testsuite/libffi.call/nested_struct2.c: Likewise.
1544 * testsuite/libffi.call/nested_struct3.c: Likewise.
1545 * testsuite/libffi.call/problem1.c: Likewise.
1546 * testsuite/libffi.special/unwindtest.cc: Likewise.
1547 * testsuite/libffi.call/cls_12byte.c: Likewise and set return value
1548 to zero.
1549 * testsuite/libffi.call/cls_16byte.c: Likewise.
1550 * testsuite/libffi.call/cls_5byte.c: Likewise.
1551
15522004-08-23 David Daney <daney@avtrex.com>
1553
1554 PR libgcj/13141
1555 * src/mips/ffitarget.h (FFI_O32_SOFT_FLOAT): New ABI.
1556 * src/mips/ffi.c (ffi_prep_args): Fix alignment calculation.
1557 (ffi_prep_cif_machdep): Handle FFI_O32_SOFT_FLOAT floating point
1558 parameters and return types.
1559 (ffi_call): Handle FFI_O32_SOFT_FLOAT ABI.
1560 (ffi_prep_closure): Ditto.
1561 (ffi_closure_mips_inner_O32): Handle FFI_O32_SOFT_FLOAT ABI, fix
1562 alignment calculations.
1563 * src/mips/o32.S (ffi_closure_O32): Don't use floating point
1564 instructions if FFI_O32_SOFT_FLOAT, make stack frame ABI compliant.
1565
15662004-08-14 Casey Marshall <csm@gnu.org>
1567
1568 * src/mips/ffi.c (ffi_pref_cif_machdep): set `cif->flags' to
1569 contain `FFI_TYPE_UINT64' as return type for any 64-bit
1570 integer (O32 ABI only).
1571 (ffi_prep_closure): new function.
1572 (ffi_closure_mips_inner_O32): new function.
1573 * src/mips/ffitarget.h: Define `FFI_CLOSURES' and
1574 `FFI_TRAMPOLINE_SIZE' appropriately if the ABI is o32.
1575 * src/mips/o32.S (ffi_call_O32): add labels for .eh_frame. Return
1576 64 bit integers correctly.
1577 (ffi_closure_O32): new function.
1578 Added DWARF-2 unwind info for both functions.
1579
15802004-08-10 Andrew Haley <aph@redhat.com>
1581
1582 * src/x86/ffi64.c (ffi_prep_args ): 8-align all stack arguments.
1583
15842004-08-01 Robert Millan <robertmh@gnu.org>
1585
1586 * configure.ac: Detect knetbsd-gnu and kfreebsd-gnu.
1587 * configure: Regenerate.
1588
15892004-07-30 Maciej W. Rozycki <macro@linux-mips.org>
1590
1591 * acinclude.m4 (AC_FUNC_MMAP_BLACKLIST): Check for <sys/mman.h>
1592 and mmap() explicitly instead of relying on preset autoconf cache
1593 variables.
1594 * aclocal.m4: Regenerate.
1595 * configure: Regenerate.
1596
15972004-07-11 Ulrich Weigand <uweigand@de.ibm.com>
1598
1599 * src/s390/ffi.c (ffi_prep_args): Fix C aliasing violation.
1600 (ffi_check_float_struct): Remove unused prototype.
1601
16022004-06-30 Geoffrey Keating <geoffk@apple.com>
1603
1604 * src/powerpc/ffi_darwin.c (flush_icache): ';' is a comment
1605 character on Darwin, use '\n\t' instead.
1606
16072004-06-26 Matthias Klose <doko@debian.org>
1608
1609 * libtool-version: Fix typo in revision/age.
1610
16112004-06-17 Matthias Klose <doko@debian.org>
1612
1613 * libtool-version: New.
1614 * Makefile.am (libffi_la_LDFLAGS): Use -version-info for soname.
1615 * Makefile.in: Regenerate.
1616
16172004-06-15 Paolo Bonzini <bonzini@gnu.org>
1618
1619 * Makefile.am: Remove useless multilib rules.
1620 * Makefile.in: Regenerate.
1621 * aclocal.m4: Regenerate with automake 1.8.5.
1622 * configure.ac: Remove useless multilib configury.
1623 * configure: Regenerate.
1624
16252004-06-15 Paolo Bonzini <bonzini@gnu.org>
1626
1627 * .cvsignore: New file.
1628
16292004-06-10 Jakub Jelinek <jakub@redhat.com>
1630
1631 * src/ia64/unix.S (ffi_call_unix): Insert group barrier break
1632 fp_done.
1633 (ffi_closure_UNIX): Fix f14/f15 adjustment if FLOAT_SZ is ever
1634 changed from 8.
1635
16362004-06-06 Sean McNeil <sean@mcneil.com>
1637
1638 * configure.ac: Add x86_64-*-freebsd* support.
1639 * configure: Regenerate.
1640
16412004-04-26 Joe Buck <jbuck@welsh-buck.org>
1642
1643 Bug 15093
1644 * configure.ac: Test for existence of mmap and sys/mman.h before
1645 checking blacklist. Fix suggested by Jim Wilson.
1646 * configure: Regenerate.
1647
16482004-04-26 Matt Austern <austern@apple.com>
1649
1650 * src/powerpc/darwin.S: Go through a non-lazy pointer for initial
1651 FDE location.
1652 * src/powerpc/darwin_closure.S: Likewise.
1653
16542004-04-24 Andreas Tobler <a.tobler@schweiz.ch>
1655
1656 * testsuite/libffi.call/cls_multi_schar.c (main): Fix initialization
1657 error. Reported by Thomas Heller <theller@python.net>.
1658 * testsuite/libffi.call/cls_multi_sshort.c (main): Likewise.
1659 * testsuite/libffi.call/cls_multi_ushort.c (main): Likewise.
1660
16612004-03-20 Matthias Klose <doko@debian.org>
1662
1663 * src/pa/linux.S: Fix typo.
1664
16652004-03-19 Matthias Klose <doko@debian.org>
1666
1667 * Makefile.am: Update.
1668 * Makefile.in: Regenerate.
1669 * src/pa/ffi.h.in: Remove.
1670 * src/pa/ffitarget.h: New file.
1671
16722004-02-10 Randolph Chung <tausq@debian.org>
1673
1674 * Makefile.am: Add PA support.
1675 * Makefile.in: Regenerate.
1676 * include/Makefile.in: Regenerate.
1677 * configure.ac: Add PA target.
1678 * configure: Regenerate.
1679 * src/pa/ffi.c: New file.
1680 * src/pa/ffi.h.in: Add PA support.
1681 * src/pa/linux.S: New file.
1682 * prep_cif.c: Add PA support.
1683
16842004-03-16 Hosaka Yuji <hos@tamanegi.org>
1685
1686 * src/types.c: Fix alignment size of X86_WIN32 case int64 and
1687 double.
1688 * src/x86/ffi.c (ffi_prep_args): Replace ecif->cif->rtype->type
1689 with ecif->cif->flags.
1690 (ffi_call, ffi_prep_incoming_args_SYSV): Replace cif->rtype->type
1691 with cif->flags.
1692 (ffi_prep_cif_machdep): Add X86_WIN32 struct case.
1693 (ffi_closure_SYSV): Add 1 or 2-bytes struct case for X86_WIN32.
1694 * src/x86/win32.S (retstruct1b, retstruct2b, sc_retstruct1b,
1695 sc_retstruct2b): Add for 1 or 2-bytes struct case.
1696
16972004-03-15 Kelley Cook <kcook@gcc.gnu.org>
1698
1699 * configure.in: Rename file to ...
1700 * configure.ac: ... this.
1701 * fficonfig.h.in: Regenerate.
1702 * Makefile.in: Regenerate.
1703 * include/Makefile.in: Regenerate.
1704 * testsuite/Makefile.in: Regenerate.
1705
17062004-03-12 Matt Austern <austern@apple.com>
1707
1708 * src/powerpc/darwin.S: Fix EH information so it corresponds to
1709 changes in EH format resulting from addition of linkonce support.
1710 * src/powerpc/darwin_closure.S: Likewise.
1711
17122004-03-11 Andreas Tobler <a.tobler@schweiz.ch>
1713 Paolo Bonzini <bonzini@gnu.org>
1714
1715 * Makefile.am (AUTOMAKE_OPTIONS): Set them.
1716 Remove VPATH. Remove rules for object files. Remove multilib support.
1717 (AM_CCASFLAGS): Add.
1718 * configure.in (AC_CONFIG_HEADERS): Relace AM_CONFIG_HEADER.
1719 (AC_PREREQ): Bump version to 2.59.
1720 (AC_INIT): Fill with version info and bug address.
1721 (ORIGINAL_LD_FOR_MULTILIBS): Remove.
1722 (AM_ENABLE_MULTILIB): Use this instead of AC_ARG_ENABLE.
1723 De-precious CC so that the right flags are passed down to multilibs.
1724 (AC_MSG_ERROR): Replace obsolete macro AC_ERROR.
1725 (AC_CONFIG_FILES): Replace obsolete macro AC_LINK_FILES.
1726 (AC_OUTPUT): Reorganize the output with AC_CONFIG_COMMANDS.
1727 * configure: Rebuilt.
1728 * aclocal.m4: Likewise.
1729 * Makefile.in, include/Makefile.in, testsuite/Makefile.in: Likewise.
1730 * fficonfig.h.in: Likewise.
1731
17322004-03-11 Andreas Schwab <schwab@suse.de>
1733
1734 * src/ia64/ffi.c (ffi_prep_incoming_args_UNIX): Get floating point
1735 arguments from fp registers only for the first 8 parameter slots.
1736 Don't convert a float parameter when passed in memory.
1737
17382004-03-09 Hans-Peter Nilsson <hp@axis.com>
1739
1740 * configure: Regenerate for config/accross.m4 correction.
1741
17422004-02-25 Matt Kraai <kraai@alumni.cmu.edu>
1743
1744 * src/powerpc/ffi.c (ffi_prep_args_SYSV): Change
1745 ecif->cif->bytes to bytes.
1746 (ffi_prep_cif_machdep): Add braces around nested if statement.
1747
17482004-02-09 Alan Modra <amodra@bigpond.net.au>
1749
1750 * src/types.c (pointer): POWERPC64 has 8 byte pointers.
1751
1752 * src/powerpc/ffi.c (ffi_prep_args64): Correct long double handling.
1753 (ffi_closure_helper_LINUX64): Fix typo.
1754 * testsuite/libffi.call/cls_align_longdouble.c: Pass -mlong-double-128
1755 for powerpc64-*-*.
1756 * testsuite/libffi.call/float.c: Likewise.
1757 * testsuite/libffi.call/float2.c: Likewise.
1758
17592004-02-08 Alan Modra <amodra@bigpond.net.au>
1760
1761 * src/powerpc/ffi.c (ffi_prep_cif_machdep <FFI_LINUX64>): Correct
1762 long double function return and long double arg handling.
1763 (ffi_closure_helper_LINUX64): Formatting. Delete unused "ng" var.
1764 Use "end_pfr" instead of "nf". Correct long double handling.
1765 Localise "temp".
1766 * src/powerpc/linux64.S (ffi_call_LINUX64): Save f2 long double
1767 return value.
1768 * src/powerpc/linux64_closure.S (ffi_closure_LINUX64): Allocate
1769 space for long double return value. Adjust stack frame and offsets.
1770 Load f2 long double return.
1771
17722004-02-07 Alan Modra <amodra@bigpond.net.au>
1773
1774 * src/types.c: Use 16 byte long double for POWERPC64.
1775
17762004-01-25 Eric Botcazou <ebotcazou@libertysurf.fr>
1777
1778 * src/sparc/ffi.c (ffi_prep_args_v9): Shift the parameter array
1779 when the structure return address is passed in %o0.
1780 (ffi_V9_return_struct): Rename into ffi_v9_layout_struct.
1781 (ffi_v9_layout_struct): Align the field following a nested structure
1782 on a word boundary. Use memmove instead of memcpy.
1783 (ffi_call): Update call to ffi_V9_return_struct.
1784 (ffi_prep_closure): Define 'ctx' only for V8.
1785 (ffi_closure_sparc_inner): Clone into ffi_closure_sparc_inner_v8
1786 and ffi_closure_sparc_inner_v9.
1787 (ffi_closure_sparc_inner_v8): Return long doubles by reference.
1788 Always skip the structure return address. For structures and long
1789 doubles, copy the argument directly.
1790 (ffi_closure_sparc_inner_v9): Skip the structure return address only
1791 if required. Shift the maximum floating-point slot accordingly. For
1792 big structures, copy the argument directly; otherwise, left-justify the
1793 argument and call ffi_v9_layout_struct to lay out the structure on
1794 the stack.
1795 * src/sparc/v8.S: Undef STACKFRAME before defining it.
1796 (ffi_closure_v8): Pass the structure return address. Update call to
1797 ffi_closure_sparc_inner_v8. Short-circuit FFI_TYPE_INT handling.
1798 Skip the 'unimp' insn when returning long doubles and structures.
1799 * src/sparc/v9.S: Undef STACKFRAME before defining it.
1800 (ffi_closure_v9): Increase the frame size by 2 words. Short-circuit
1801 FFI_TYPE_INT handling. Load structures both in integers and
1802 floating-point registers on return.
1803 * README: Update status of the SPARC port.
1804
18052004-01-24 Andreas Tobler <a.tobler@schweiz.ch>
1806
1807 * testsuite/libffi.call/pyobjc-tc.c (main): Treat result value
1808 as of type ffi_arg.
1809 * testsuite/libffi.call/struct3.c (main): Fix CHECK.
1810
18112004-01-22 Ulrich Weigand <uweigand@de.ibm.com>
1812
1813 * testsuite/libffi.call/cls_uint.c (cls_ret_uint_fn): Treat result
1814 value as of type ffi_arg, not unsigned int.
1815
18162004-01-21 Michael Ritzert <ritzert@t-online.de>
1817
1818 * ffi64.c (ffi_prep_args): Cast the RHS of an assignment instead
1819 of the LHS.
1820
18212004-01-12 Andreas Tobler <a.tobler@schweiz.ch>
1822
1823 * testsuite/lib/libffi-dg.exp: Set LD_LIBRARY_PATH_32 for
1824 Solaris.
1825
18262004-01-08 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
1827
1828 * testsuite/libffi.call/ffitest.h (allocate_mmap): Cast MAP_FAILED
1829 to void *.
1830
18312003-12-10 Richard Henderson <rth@redhat.com>
1832
1833 * testsuite/libffi.call/cls_align_pointer.c: Cast pointers to
1834 size_t instead of int.
1835
18362003-12-04 Hosaka Yuji <hos@tamanegi.org>
1837
1838 * testsuite/libffi.call/many_win32.c: Include <float.h>.
1839 * testsuite/libffi.call/many_win32.c (main): Replace variable
1840 int i with unsigned long ul.
1841
1842 * testsuite/libffi.call/cls_align_uint64.c: New test case.
1843 * testsuite/libffi.call/cls_align_sint64.c: Likewise.
1844 * testsuite/libffi.call/cls_align_uint32.c: Likewise.
1845 * testsuite/libffi.call/cls_align_sint32.c: Likewise.
1846 * testsuite/libffi.call/cls_align_uint16.c: Likewise.
1847 * testsuite/libffi.call/cls_align_sint16.c: Likewise.
1848 * testsuite/libffi.call/cls_align_float.c: Likewise.
1849 * testsuite/libffi.call/cls_align_double.c: Likewise.
1850 * testsuite/libffi.call/cls_align_longdouble.c: Likewise.
1851 * testsuite/libffi.call/cls_align_pointer.c: Likewise.
1852
18532003-12-02 Hosaka Yuji <hos@tamanegi.org>
1854
1855 PR other/13221
1856 * src/x86/ffi.c (ffi_prep_args, ffi_prep_incoming_args_SYSV):
1857 Align arguments to 32 bits.
1858
18592003-12-01 Andreas Tobler <a.tobler@schweiz.ch>
1860
1861 PR other/13221
1862 * testsuite/libffi.call/cls_multi_sshort.c: New test case.
1863 * testsuite/libffi.call/cls_multi_sshortchar.c: Likewise.
1864 * testsuite/libffi.call/cls_multi_uchar.c: Likewise.
1865 * testsuite/libffi.call/cls_multi_schar.c: Likewise.
1866 * testsuite/libffi.call/cls_multi_ushortchar.c: Likewise.
1867 * testsuite/libffi.call/cls_multi_ushort.c: Likewise.
1868
1869 * testsuite/libffi.special/unwindtest.cc: Cosmetics.
1870
18712003-11-26 Kaveh R. Ghazi <ghazi@caip.rutgers.edu>
1872
1873 * testsuite/libffi.call/ffitest.h: Include <fcntl.h>.
1874 * testsuite/libffi.special/ffitestcxx.h: Likewise.
1875
18762003-11-22 Andreas Tobler <a.tobler@schweiz.ch>
1877
1878 * Makefile.in: Rebuilt.
1879 * configure: Likewise.
1880 * testsuite/libffi.special/unwindtest.cc: Convert the mmap to
1881 the right type.
1882
18832003-11-21 Andreas Jaeger <aj@suse.de>
1884 Andreas Tobler <a.tobler@schweiz.ch>
1885
1886 * acinclude.m4: Add AC_FUNC_MMAP_BLACKLIST.
1887 * configure.in: Call AC_FUNC_MMAP_BLACKLIST.
1888 * Makefile.in: Rebuilt.
1889 * aclocal.m4: Likewise.
1890 * configure: Likewise.
1891 * fficonfig.h.in: Likewise.
1892 * testsuite/lib/libffi-dg.exp: Add include dir.
1893 * testsuite/libffi.call/ffitest.h: Add MMAP definitions.
1894 * testsuite/libffi.special/ffitestcxx.h: Likewise.
1895 * testsuite/libffi.call/closure_fn0.c: Use MMAP functionality
1896 for ffi_closure if available.
1897 * testsuite/libffi.call/closure_fn1.c: Likewise.
1898 * testsuite/libffi.call/closure_fn2.c: Likewise.
1899 * testsuite/libffi.call/closure_fn3.c: Likewise.
1900 * testsuite/libffi.call/closure_fn4.c: Likewise.
1901 * testsuite/libffi.call/closure_fn5.c: Likewise.
1902 * testsuite/libffi.call/cls_12byte.c: Likewise.
1903 * testsuite/libffi.call/cls_16byte.c: Likewise.
1904 * testsuite/libffi.call/cls_18byte.c: Likewise.
1905 * testsuite/libffi.call/cls_19byte.c: Likewise.
1906 * testsuite/libffi.call/cls_1_1byte.c: Likewise.
1907 * testsuite/libffi.call/cls_20byte.c: Likewise.
1908 * testsuite/libffi.call/cls_20byte1.c: Likewise.
1909 * testsuite/libffi.call/cls_24byte.c: Likewise.
1910 * testsuite/libffi.call/cls_2byte.c: Likewise.
1911 * testsuite/libffi.call/cls_3_1byte.c: Likewise.
1912 * testsuite/libffi.call/cls_3byte1.c: Likewise.
1913 * testsuite/libffi.call/cls_3byte2.c: Likewise.
1914 * testsuite/libffi.call/cls_4_1byte.c: Likewise.
1915 * testsuite/libffi.call/cls_4byte.c: Likewise.
1916 * testsuite/libffi.call/cls_5byte.c: Likewise.
1917 * testsuite/libffi.call/cls_64byte.c: Likewise.
1918 * testsuite/libffi.call/cls_6byte.c: Likewise.
1919 * testsuite/libffi.call/cls_7byte.c: Likewise.
1920 * testsuite/libffi.call/cls_8byte.c: Likewise.
1921 * testsuite/libffi.call/cls_9byte1.c: Likewise.
1922 * testsuite/libffi.call/cls_9byte2.c: Likewise.
1923 * testsuite/libffi.call/cls_double.c: Likewise.
1924 * testsuite/libffi.call/cls_float.c: Likewise.
1925 * testsuite/libffi.call/cls_schar.c: Likewise.
1926 * testsuite/libffi.call/cls_sint.c: Likewise.
1927 * testsuite/libffi.call/cls_sshort.c: Likewise.
1928 * testsuite/libffi.call/cls_uchar.c: Likewise.
1929 * testsuite/libffi.call/cls_uint.c: Likewise.
1930 * testsuite/libffi.call/cls_ulonglong.c: Likewise.
1931 * testsuite/libffi.call/cls_ushort.c: Likewise.
1932 * testsuite/libffi.call/nested_struct.c: Likewise.
1933 * testsuite/libffi.call/nested_struct1.c: Likewise.
1934 * testsuite/libffi.call/nested_struct2.c: Likewise.
1935 * testsuite/libffi.call/nested_struct3.c: Likewise.
1936 * testsuite/libffi.call/problem1.c: Likewise.
1937 * testsuite/libffi.special/unwindtest.cc: Likewise.
1938
19392003-11-20 Andreas Tobler <a.tobler@schweiz.ch>
1940
1941 * testsuite/lib/libffi-dg.exp: Make the -lgcc_s conditional.
1942
19432003-11-19 Andreas Tobler <a.tobler@schweiz.ch>
1944
1945 * testsuite/lib/libffi-dg.exp: Add DYLD_LIBRARY_PATH for darwin.
1946 Add -lgcc_s to additional flags.
1947
19482003-11-12 Andreas Tobler <a.tobler@schweiz.ch>
1949
1950 * configure.in, include/Makefile.am: PR libgcj/11147, install
1951 the ffitarget.h header file in a gcc versioned and target
1952 dependent place.
1953 * configure: Regenerated.
1954 * Makefile.in, include/Makefile.in: Likewise.
1955 * testsuite/Makefile.in: Likewise.
1956
19572003-11-09 Andreas Tobler <a.tobler@schweiz.ch>
1958
1959 * testsuite/libffi.call/closure_fn0.c: Print result and check
1960 with dg-output to make debugging easier.
1961 * testsuite/libffi.call/closure_fn1.c: Likewise.
1962 * testsuite/libffi.call/closure_fn2.c: Likewise.
1963 * testsuite/libffi.call/closure_fn3.c: Likewise.
1964 * testsuite/libffi.call/closure_fn4.c: Likewise.
1965 * testsuite/libffi.call/closure_fn5.c: Likewise.
1966 * testsuite/libffi.call/cls_12byte.c: Likewise.
1967 * testsuite/libffi.call/cls_16byte.c: Likewise.
1968 * testsuite/libffi.call/cls_18byte.c: Likewise.
1969 * testsuite/libffi.call/cls_19byte.c: Likewise.
1970 * testsuite/libffi.call/cls_1_1byte.c: Likewise.
1971 * testsuite/libffi.call/cls_20byte.c: Likewise.
1972 * testsuite/libffi.call/cls_20byte1.c: Likewise.
1973 * testsuite/libffi.call/cls_24byte.c: Likewise.
1974 * testsuite/libffi.call/cls_2byte.c: Likewise.
1975 * testsuite/libffi.call/cls_3_1byte.c: Likewise.
1976 * testsuite/libffi.call/cls_3byte1.c: Likewise.
1977 * testsuite/libffi.call/cls_3byte2.c: Likewise.
1978 * testsuite/libffi.call/cls_4_1byte.c: Likewise.
1979 * testsuite/libffi.call/cls_4byte.c: Likewise.
1980 * testsuite/libffi.call/cls_5byte.c: Likewise.
1981 * testsuite/libffi.call/cls_64byte.c: Likewise.
1982 * testsuite/libffi.call/cls_6byte.c: Likewise.
1983 * testsuite/libffi.call/cls_7byte.c: Likewise.
1984 * testsuite/libffi.call/cls_8byte.c: Likewise.
1985 * testsuite/libffi.call/cls_9byte1.c: Likewise.
1986 * testsuite/libffi.call/cls_9byte2.c: Likewise.
1987 * testsuite/libffi.call/cls_double.c: Likewise.
1988 * testsuite/libffi.call/cls_float.c: Likewise.
1989 * testsuite/libffi.call/cls_schar.c: Likewise.
1990 * testsuite/libffi.call/cls_sint.c: Likewise.
1991 * testsuite/libffi.call/cls_sshort.c: Likewise.
1992 * testsuite/libffi.call/cls_uchar.c: Likewise.
1993 * testsuite/libffi.call/cls_uint.c: Likewise.
1994 * testsuite/libffi.call/cls_ulonglong.c: Likewise.
1995 * testsuite/libffi.call/cls_ushort.c: Likewise.
1996 * testsuite/libffi.call/problem1.c: Likewise.
1997
1998 * testsuite/libffi.special/unwindtest.cc: Make ffi_closure
1999 static.
2000
20012003-11-08 Andreas Tobler <a.tobler@schweiz.ch>
2002
2003 * testsuite/libffi.call/cls_9byte2.c: New test case.
2004 * testsuite/libffi.call/cls_9byte1.c: Likewise.
2005 * testsuite/libffi.call/cls_64byte.c: Likewise.
2006 * testsuite/libffi.call/cls_20byte1.c: Likewise.
2007 * testsuite/libffi.call/cls_19byte.c: Likewise.
2008 * testsuite/libffi.call/cls_18byte.c: Likewise.
2009 * testsuite/libffi.call/closure_fn4.c: Likewise.
2010 * testsuite/libffi.call/closure_fn5.c: Likewise.
2011 * testsuite/libffi.call/cls_schar.c: Likewise.
2012 * testsuite/libffi.call/cls_sint.c: Likewise.
2013 * testsuite/libffi.call/cls_sshort.c: Likewise.
2014 * testsuite/libffi.call/nested_struct2.c: Likewise.
2015 * testsuite/libffi.call/nested_struct3.c: Likewise.
2016
20172003-11-08 Andreas Tobler <a.tobler@schweiz.ch>
2018
2019 * testsuite/libffi.call/cls_double.c: Do a check on the result.
2020 * testsuite/libffi.call/cls_uchar.c: Likewise.
2021 * testsuite/libffi.call/cls_uint.c: Likewise.
2022 * testsuite/libffi.call/cls_ulonglong.c: Likewise.
2023 * testsuite/libffi.call/cls_ushort.c: Likewise.
2024 * testsuite/libffi.call/return_sc.c: Cleanup whitespaces.
2025
20262003-11-06 Andreas Tobler <a.tobler@schweiz.ch>
2027
2028 * src/prep_cif.c (ffi_prep_cif): Move the validity check after
2029 the initialization.
2030
20312003-10-23 Andreas Tobler <a.tobler@schweiz.ch>
2032
2033 * src/java_raw_api.c (ffi_java_ptrarray_to_raw): Replace
2034 FFI_ASSERT(FALSE) with FFI_ASSERT(0).
2035
20362003-10-22 David Daney <ddaney@avtrex.com>
2037
2038 * src/mips/ffitarget.h: Replace undefined UINT32 and friends with
2039 __attribute__((__mode__(__SI__))) and friends.
2040
20412003-10-22 Andreas Schwab <schwab@suse.de>
2042
2043 * src/ia64/ffi.c: Replace FALSE/TRUE with false/true.
2044
20452003-10-21 Andreas Tobler <a.tobler@schweiz.ch>
2046
2047 * configure.in: AC_LINK_FILES(ffitarget.h).
2048 * configure: Regenerate.
2049 * Makefile.in: Likewise.
2050 * include/Makefile.in: Likewise.
2051 * testsuite/Makefile.in: Likewise.
2052 * fficonfig.h.in: Likewise.
2053
20542003-10-21 Paolo Bonzini <bonzini@gnu.org>
2055 Richard Henderson <rth@redhat.com>
2056
2057 Avoid that ffi.h includes fficonfig.h.
2058
2059 * Makefile.am (EXTRA_DIST): Include ffitarget.h files
2060 (TARGET_SRC_MIPS_GCC): Renamed to TARGET_SRC_MIPS_IRIX.
2061 (TARGET_SRC_MIPS_SGI): Removed.
2062 (MIPS_GCC): Renamed to TARGET_SRC_MIPS_IRIX.
2063 (MIPS_SGI): Removed.
2064 (CLEANFILES): Removed.
2065 (mostlyclean-am, clean-am, mostlyclean-sub, clean-sub): New
2066 targets.
2067 * acconfig.h: Removed.
2068 * configure.in: Compute sizeofs only for double and long double.
2069 Use them to define and subst HAVE_LONG_DOUBLE. Include comments
2070 into AC_DEFINE instead of using acconfig.h. Create
2071 include/ffitarget.h instead of include/fficonfig.h. Rename
2072 MIPS_GCC to MIPS_IRIX, drop MIPS_SGI since we are in gcc's tree.
2073 AC_DEFINE EH_FRAME_FLAGS.
2074 * include/Makefile.am (DISTCLEANFILES): New automake macro.
2075 (hack_DATA): Add ffitarget.h.
2076 * include/ffi.h.in: Remove all system specific definitions.
2077 Declare raw API even if it is not installed, why bother?
2078 Use limits.h instead of SIZEOF_* to define ffi_type_*. Do
2079 not define EH_FRAME_FLAGS, it is in fficonfig.h now. Include
2080 ffitarget.h instead of fficonfig.h. Remove ALIGN macro.
2081 (UINT_ARG, INT_ARG): Removed, use ffi_arg and ffi_sarg instead.
2082 * include/ffi_common.h (bool): Do not define.
2083 (ffi_assert): Accept failed assertion.
2084 (ffi_type_test): Return void and accept file/line.
2085 (FFI_ASSERT): Pass stringized failed assertion.
2086 (FFI_ASSERT_AT): New macro.
2087 (FFI_ASSERT_VALID_TYPE): New macro.
2088 (UINT8, SINT8, UINT16, SINT16, UINT32, SINT32,
2089 UINT64, SINT64): Define here with gcc's __attribute__ macro
2090 instead of in ffi.h
2091 (FLOAT32, ALIGN): Define here instead of in ffi.h
2092 * include/ffi-mips.h: Removed. Its content moved to
2093 src/mips/ffitarget.h after separating assembly and C sections.
2094 * src/alpha/ffi.c, src/alpha/ffi.c, src/java_raw_api.c
2095 src/prep_cif.c, src/raw_api.c, src/ia64/ffi.c,
2096 src/mips/ffi.c, src/mips/n32.S, src/mips/o32.S,
2097 src/mips/ffitarget.h, src/sparc/ffi.c, src/x86/ffi64.c:
2098 SIZEOF_ARG -> FFI_SIZEOF_ARG.
2099 * src/ia64/ffi.c: Include stdbool.h (provided by GCC 2.95+).
2100 * src/debug.c (ffi_assert): Accept stringized failed assertion.
2101 (ffi_type_test): Rewritten.
2102 * src/prep-cif.c (initialize_aggregate, ffi_prep_cif): Call
2103 FFI_ASSERT_VALID_TYPE.
2104 * src/alpha/ffitarget.h, src/arm/ffitarget.h,
2105 src/ia64/ffitarget.h, src/m68k/ffitarget.h,
2106 src/mips/ffitarget.h, src/powerpc/ffitarget.h,
2107 src/s390/ffitarget.h, src/sh/ffitarget.h,
2108 src/sh64/ffitarget.h, src/sparc/ffitarget.h,
2109 src/x86/ffitarget.h: New files.
2110 * src/alpha/osf.S, src/arm/sysv.S, src/ia64/unix.S,
2111 src/m68k/sysv.S, src/mips/n32.S, src/mips/o32.S,
2112 src/powerpc/aix.S, src/powerpc/darwin.S,
2113 src/powerpc/ffi_darwin.c, src/powerpc/linux64.S,
2114 src/powerpc/linux64_closure.S, src/powerpc/ppc_closure.S,
2115 src/powerpc/sysv.S, src/s390/sysv.S, src/sh/sysv.S,
2116 src/sh64/sysv.S, src/sparc/v8.S, src/sparc/v9.S,
2117 src/x86/sysv.S, src/x86/unix64.S, src/x86/win32.S:
2118 include fficonfig.h
2119
21202003-10-20 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
2121
2122 * src/mips/ffi.c: Use _ABIN32, _ABIO32 instead of external
2123 _MIPS_SIM_NABI32, _MIPS_SIM_ABI32.
2124
21252003-10-19 Andreas Tobler <a.tobler@schweiz.ch>
2126
2127 * src/powerpc/ffi_darwin.c (ffi_prep_args): Declare bytes again.
2128 Used when FFI_DEBUG = 1.
2129
21302003-10-14 Alan Modra <amodra@bigpond.net.au>
2131
2132 * src/types.c (double, longdouble): Default POWERPC64 to 8 byte size
2133 and align.
2134
21352003-10-06 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
2136
2137 * include/ffi_mips.h: Define FFI_MIPS_N32 for N32/N64 ABIs,
2138 FFI_MIPS_O32 for O32 ABI.
2139
21402003-10-01 Andreas Tobler <a.tobler@schweiz.ch>
2141
2142 * testsuite/lib/libffi-dg.exp: Set LD_LIBRARY_PATH_64 for
2143 SPARC64. Cleanup whitespaces.
2144
21452003-09-19 Andreas Tobler <a.tobler@schweiz.ch>
2146
2147 * testsuite/libffi.call/closure_fn0.c: Xfail mips, arm,
2148 strongarm, xscale. Cleanup whitespaces.
2149 * testsuite/libffi.call/closure_fn1.c: Likewise.
2150 * testsuite/libffi.call/closure_fn2.c: Likewise.
2151 * testsuite/libffi.call/closure_fn3.c: Likewise.
2152 * testsuite/libffi.call/cls_12byte.c: Likewise.
2153 * testsuite/libffi.call/cls_16byte.c: Likewise.
2154 * testsuite/libffi.call/cls_1_1byte.c: Likewise.
2155 * testsuite/libffi.call/cls_20byte.c: Likewise.
2156 * testsuite/libffi.call/cls_24byte.c: Likewise.
2157 * testsuite/libffi.call/cls_2byte.c: Likewise.
2158 * testsuite/libffi.call/cls_3_1byte.c: Likewise.
2159 * testsuite/libffi.call/cls_3byte1.c: Likewise.
2160 * testsuite/libffi.call/cls_3byte2.c: Likewise.
2161 * testsuite/libffi.call/cls_4_1byte.c: Likewise.
2162 * testsuite/libffi.call/cls_4byte.c: Likewise.
2163 * testsuite/libffi.call/cls_5byte.c: Likewise.
2164 * testsuite/libffi.call/cls_6byte.c: Likewise.
2165 * testsuite/libffi.call/cls_7byte.c: Likewise.
2166 * testsuite/libffi.call/cls_8byte.c: Likewise.
2167 * testsuite/libffi.call/cls_double.c: Likewise.
2168 * testsuite/libffi.call/cls_float.c: Likewise.
2169 * testsuite/libffi.call/cls_uchar.c: Likewise.
2170 * testsuite/libffi.call/cls_uint.c: Likewise.
2171 * testsuite/libffi.call/cls_ulonglong.c: Likewise.
2172 * testsuite/libffi.call/cls_ushort.c: Likewise.
2173 * testsuite/libffi.call/nested_struct.c: Likewise.
2174 * testsuite/libffi.call/nested_struct1.c: Likewise.
2175 * testsuite/libffi.call/problem1.c: Likewise.
2176 * testsuite/libffi.special/unwindtest.cc: Likewise.
2177 * testsuite/libffi.call/pyobjc-tc.c: Cleanup whitespaces.
2178
21792003-09-18 David Edelsohn <edelsohn@gnu.org>
2180
2181 * src/powerpc/aix.S: Cleanup whitespaces.
2182 * src/powerpc/aix_closure.S: Likewise.
2183
21842003-09-18 Andreas Tobler <a.tobler@schweiz.ch>
2185
2186 * src/powerpc/darwin.S: Cleanup whitespaces, comment formatting.
2187 * src/powerpc/darwin_closure.S: Likewise.
2188 * src/powerpc/ffi_darwin.c: Likewise.
2189
21902003-09-18 Andreas Tobler <a.tobler@schweiz.ch>
2191 David Edelsohn <edelsohn@gnu.org>
2192
2193 * src/types.c (double): Add AIX and Darwin to the right TYPEDEF.
2194 * src/powerpc/aix_closure.S: Remove the pointer to the outgoing
2195 parameter stack.
2196 * src/powerpc/darwin_closure.S: Likewise.
2197 * src/powerpc/ffi_darwin.c (ffi_prep_args): Handle structures
2198 according to the Darwin/AIX ABI.
2199 (ffi_prep_cif_machdep): Likewise.
2200 (ffi_closure_helper_DARWIN): Likewise.
2201 Remove the outgoing parameter stack logic. Simplify the evaluation
2202 of the different CASE types.
2203 (ffi_prep_clousure): Avoid the casts on lvalues. Change the branch
2204 statement in the trampoline code.
2205
22062003-09-18 Kaz Kojima <kkojima@gcc.gnu.org>
2207
2208 * src/sh/ffi.c (ffi_prep_args): Take account into the alignement
2209 for the register size.
2210 (ffi_closure_helper_SYSV): Handle the structure return value
2211 address correctly.
2212 (ffi_closure_helper_SYSV): Return the appropriate type when
2213 the registers are used for the structure return value.
2214 * src/sh/sysv.S (ffi_closure_SYSV): Fix the stack layout for
2215 the 64-bit return value. Update copyright years.
2216
22172003-09-17 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
2218
2219 * testsuite/lib/libffi-dg.exp (libffi_target_compile): Search in
2220 srcdir for ffi_mips.h.
2221
22222003-09-12 Alan Modra <amodra@bigpond.net.au>
2223
2224 * src/prep_cif.c (initialize_aggregate): Include tail padding in
2225 structure size.
2226 * src/powerpc/linux64_closure.S (ffi_closure_LINUX64): Correct
2227 placement of float result.
2228 * testsuite/libffi.special/unwindtest.cc (closure_test_fn1): Correct
2229 cast of "resp" for big-endian 64 bit machines.
2230
22312003-09-11 Alan Modra <amodra@bigpond.net.au>
2232
2233 * src/types.c (double, longdouble): Merge identical SH and ARM
2234 typedefs, and add POWERPC64.
2235 * src/powerpc/ffi.c (ffi_prep_args64): Correct next_arg calc for
2236 struct split over gpr and rest.
2237 (ffi_prep_cif_machdep): Correct intarg_count for structures.
2238 * src/powerpc/linux64.S (ffi_call_LINUX64): Fix gpr offsets.
2239
22402003-09-09 Andreas Tobler <a.tobler@schweiz.ch>
2241
2242 * src/powerpc/ffi.c (ffi_closure_helper_SYSV) Handle struct
2243 passing correctly.
2244
22452003-09-09 Alan Modra <amodra@bigpond.net.au>
2246
2247 * configure: Regenerate.
2248
22492003-09-04 Andreas Tobler <a.tobler@schweiz.ch>
2250
2251 * Makefile.am: Remove build rules for ffitest.
2252 * Makefile.in: Rebuilt.
2253
22542003-09-04 Andreas Tobler <a.tobler@schweiz.ch>
2255
2256 * src/java_raw_api.c: Include <stdlib.h> to fix compiler warning
2257 about implicit declaration of abort().
2258
22592003-09-04 Andreas Tobler <a.tobler@schweiz.ch>
2260
2261 * Makefile.am: Add dejagnu test framework. Fixes PR other/11411.
2262 * Makefile.in: Rebuilt.
2263 * configure.in: Add dejagnu test framework.
2264 * configure: Rebuilt.
2265
2266 * testsuite/Makefile.am: New file.
2267 * testsuite/Makefile.in: Built
2268 * testsuite/lib/libffi-dg.exp: New file.
2269 * testsuite/config/default.exp: Likewise.
2270 * testsuite/libffi.call/call.exp: Likewise.
2271 * testsuite/libffi.call/ffitest.h: Likewise.
2272 * testsuite/libffi.call/closure_fn0.c: Likewise.
2273 * testsuite/libffi.call/closure_fn1.c: Likewise.
2274 * testsuite/libffi.call/closure_fn2.c: Likewise.
2275 * testsuite/libffi.call/closure_fn3.c: Likewise.
2276 * testsuite/libffi.call/cls_1_1byte.c: Likewise.
2277 * testsuite/libffi.call/cls_3_1byte.c: Likewise.
2278 * testsuite/libffi.call/cls_4_1byte.c: Likewise.
2279 * testsuite/libffi.call/cls_2byte.c: Likewise.
2280 * testsuite/libffi.call/cls_3byte1.c: Likewise.
2281 * testsuite/libffi.call/cls_3byte2.c: Likewise.
2282 * testsuite/libffi.call/cls_4byte.c: Likewise.
2283 * testsuite/libffi.call/cls_5byte.c: Likewise.
2284 * testsuite/libffi.call/cls_6byte.c: Likewise.
2285 * testsuite/libffi.call/cls_7byte.c: Likewise.
2286 * testsuite/libffi.call/cls_8byte.c: Likewise.
2287 * testsuite/libffi.call/cls_12byte.c: Likewise.
2288 * testsuite/libffi.call/cls_16byte.c: Likewise.
2289 * testsuite/libffi.call/cls_20byte.c: Likewise.
2290 * testsuite/libffi.call/cls_24byte.c: Likewise.
2291 * testsuite/libffi.call/cls_double.c: Likewise.
2292 * testsuite/libffi.call/cls_float.c: Likewise.
2293 * testsuite/libffi.call/cls_uchar.c: Likewise.
2294 * testsuite/libffi.call/cls_uint.c: Likewise.
2295 * testsuite/libffi.call/cls_ulonglong.c: Likewise.
2296 * testsuite/libffi.call/cls_ushort.c: Likewise.
2297 * testsuite/libffi.call/float.c: Likewise.
2298 * testsuite/libffi.call/float1.c: Likewise.
2299 * testsuite/libffi.call/float2.c: Likewise.
2300 * testsuite/libffi.call/many.c: Likewise.
2301 * testsuite/libffi.call/many_win32.c: Likewise.
2302 * testsuite/libffi.call/nested_struct.c: Likewise.
2303 * testsuite/libffi.call/nested_struct1.c: Likewise.
2304 * testsuite/libffi.call/pyobjc-tc.c: Likewise.
2305 * testsuite/libffi.call/problem1.c: Likewise.
2306 * testsuite/libffi.call/promotion.c: Likewise.
2307 * testsuite/libffi.call/return_ll.c: Likewise.
2308 * testsuite/libffi.call/return_sc.c: Likewise.
2309 * testsuite/libffi.call/return_uc.c: Likewise.
2310 * testsuite/libffi.call/strlen.c: Likewise.
2311 * testsuite/libffi.call/strlen_win32.c: Likewise.
2312 * testsuite/libffi.call/struct1.c: Likewise.
2313 * testsuite/libffi.call/struct2.c: Likewise.
2314 * testsuite/libffi.call/struct3.c: Likewise.
2315 * testsuite/libffi.call/struct4.c: Likewise.
2316 * testsuite/libffi.call/struct5.c: Likewise.
2317 * testsuite/libffi.call/struct6.c: Likewise.
2318 * testsuite/libffi.call/struct7.c: Likewise.
2319 * testsuite/libffi.call/struct8.c: Likewise.
2320 * testsuite/libffi.call/struct9.c: Likewise.
2321 * testsuite/libffi.special/special.exp: New file.
2322 * testsuite/libffi.special/ffitestcxx.h: Likewise.
2323 * testsuite/libffi.special/unwindtest.cc: Likewise.
2324
2325
23262003-08-13 Kaz Kojima <kkojima@gcc.gnu.org>
2327
2328 * src/sh/ffi.c (OFS_INT16): Set 0 for little endian case. Update
2329 copyright years.
2330
23312003-08-02 Alan Modra <amodra@bigpond.net.au>
2332
2333 * src/powerpc/ffi.c (ffi_prep_args64): Modify for changed gcc
2334 structure passing.
2335 (ffi_closure_helper_LINUX64): Likewise.
2336 * src/powerpc/linux64.S: Remove code writing to parm save area.
2337 * src/powerpc/linux64_closure.S (ffi_closure_LINUX64): Use return
2338 address in lr from ffi_closure_helper_LINUX64 call to calculate
2339 table address. Optimize function tail.
2340
23412003-07-28 Andreas Tobler <a.tobler@schweiz.ch>
2342
2343 * src/sparc/ffi.c: Handle all floating point registers.
2344 * src/sparc/v9.S: Likewise. Fixes second part of PR target/11410.
2345
23462003-07-11 Gerald Pfeifer <pfeifer@dbai.tuwien.ac.at>
2347
2348 * README: Note that libffi is not part of GCC. Update the project
2349 URL and status.
2350
23512003-06-19 Franz Sirl <Franz.Sirl-kernel@lauterbach.com>
2352
2353 * src/powerpc/ppc_closure.S: Include ffi.h.
2354
23552003-06-13 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
2356
2357 * src/x86/sysv.S: Avoid gas-only .uleb128/.sleb128 directives.
2358 Use C style comments.
2359
23602003-06-13 Kaz Kojima <kkojima@rr.iij4u.or.jp>
2361
2362 * Makefile.am: Add SHmedia support. Fix a typo of SH support.
2363 * Makefile.in: Regenerate.
2364 * configure.in (sh64-*-linux*, sh5*-*-linux*): Add target.
2365 * configure: Regenerate.
2366 * include/ffi.h.in: Add SHmedia support.
2367 * src/sh64/ffi.c: New file.
2368 * src/sh64/sysv.S: New file.
2369
23702003-05-16 Jakub Jelinek <jakub@redhat.com>
2371
2372 * configure.in (HAVE_RO_EH_FRAME): Check whether .eh_frame section
2373 should be read-only.
2374 * configure: Rebuilt.
2375 * fficonfig.h.in: Rebuilt.
2376 * include/ffi.h.in (EH_FRAME_FLAGS): Define.
2377 * src/alpha/osf.S: Use EH_FRAME_FLAGS.
2378 * src/powerpc/linux64.S: Likewise.
2379 * src/powerpc/linux64_closure.S: Likewise. Include ffi.h.
2380 * src/powerpc/sysv.S: Use EH_FRAME_FLAGS. Use pcrel encoding
2381 if -fpic/-fPIC/-mrelocatable.
2382 * src/powerpc/powerpc_closure.S: Likewise.
2383 * src/sparc/v8.S: If HAVE_RO_EH_FRAME is defined, don't include
2384 #write in .eh_frame flags.
2385 * src/sparc/v9.S: Likewise.
2386 * src/x86/unix64.S: Use EH_FRAME_FLAGS.
2387 * src/x86/sysv.S: Likewise. Use pcrel encoding if -fpic/-fPIC.
2388 * src/s390/sysv.S: Use EH_FRAME_FLAGS. Include ffi.h.
2389
23902003-05-07 Jeff Sturm <jsturm@one-point.com>
2391
2392 Fixes PR bootstrap/10656
2393 * configure.in (HAVE_AS_REGISTER_PSEUDO_OP): Test assembler
2394 support for .register pseudo-op.
2395 * src/sparc/v8.S: Use it.
2396 * fficonfig.h.in: Rebuilt.
2397 * configure: Rebuilt.
2398
23992003-04-18 Jakub Jelinek <jakub@redhat.com>
2400
2401 * include/ffi.h.in (POWERPC64): Define if 64-bit.
2402 (enum ffi_abi): Add FFI_LINUX64 on POWERPC.
2403 Make it the default on POWERPC64.
2404 (FFI_TRAMPOLINE_SIZE): Define to 24 on POWERPC64.
2405 * configure.in: Change powerpc-*-linux* into powerpc*-*-linux*.
2406 * configure: Rebuilt.
2407 * src/powerpc/ffi.c (hidden): Define.
2408 (ffi_prep_args_SYSV): Renamed from
2409 ffi_prep_args. Cast pointers to unsigned long to shut up warnings.
2410 (NUM_GPR_ARG_REGISTERS64, NUM_FPR_ARG_REGISTERS64,
2411 ASM_NEEDS_REGISTERS64): New.
2412 (ffi_prep_args64): New function.
2413 (ffi_prep_cif_machdep): Handle FFI_LINUX64 ABI.
2414 (ffi_call): Likewise.
2415 (ffi_prep_closure): Likewise.
2416 (flush_icache): Surround by #ifndef POWERPC64.
2417 (ffi_dblfl): New union type.
2418 (ffi_closure_helper_SYSV): Use it to avoid aliasing problems.
2419 (ffi_closure_helper_LINUX64): New function.
2420 * src/powerpc/ppc_closure.S: Surround whole file by #ifndef
2421 __powerpc64__.
2422 * src/powerpc/sysv.S: Likewise.
2423 (ffi_call_SYSV): Rename ffi_prep_args to ffi_prep_args_SYSV.
2424 * src/powerpc/linux64.S: New file.
2425 * src/powerpc/linux64_closure.S: New file.
2426 * Makefile.am (EXTRA_DIST): Add src/powerpc/linux64.S and
2427 src/powerpc/linux64_closure.S.
2428 (TARGET_SRC_POWERPC): Likewise.
2429
2430 * src/ffitest.c (closure_test_fn, closure_test_fn1, closure_test_fn2,
2431 closure_test_fn3): Fix result printing on big-endian 64-bit
2432 machines.
2433 (main): Print tst2_arg instead of uninitialized tst2_result.
2434
2435 * src/ffitest.c (main): Hide what closure pointer really points to
2436 from the compiler.
2437
24382003-04-16 Richard Earnshaw <rearnsha@arm.com>
2439
2440 * configure.in (arm-*-netbsdelf*): Add configuration.
2441 (configure): Regenerated.
2442
24432003-04-04 Loren J. Rittle <ljrittle@acm.org>
2444
2445 * include/Makefile.in: Regenerate.
2446
24472003-03-21 Zdenek Dvorak <rakdver@atrey.karlin.mff.cuni.cz>
2448
2449 * libffi/include/ffi.h.in: Define X86 instead of X86_64 in 32
2450 bit mode.
2451 * libffi/src/x86/ffi.c (ffi_closure_SYSV, ffi_closure_raw_SYSV):
2452 Receive closure pointer through parameter, read args using
2453 __builtin_dwarf_cfa.
2454 (FFI_INIT_TRAMPOLINE): Send closure reference through eax.
2455
24562003-03-12 Andreas Schwab <schwab@suse.de>
2457
2458 * configure.in: Avoid trailing /. in toolexeclibdir.
2459 * configure: Rebuilt.
2460
24612003-03-03 Andreas Tobler <a.tobler@schweiz.ch>
2462
2463 * src/powerpc/darwin_closure.S: Recode to fit dynamic libraries.
2464
24652003-02-06 Andreas Tobler <a.tobler@schweiz.ch>
2466
2467 * libffi/src/powerpc/darwin_closure.S:
2468 Fix alignement bug, allocate 8 bytes for the result.
2469 * libffi/src/powerpc/aix_closure.S:
2470 Likewise.
2471 * libffi/src/powerpc/ffi_darwin.c:
2472 Update stackframe description for aix/darwin_closure.S.
2473
24742003-02-06 Jakub Jelinek <jakub@redhat.com>
2475
2476 * src/s390/ffi.c (ffi_closure_helper_SYSV): Add hidden visibility
2477 attribute.
2478
24792003-01-31 Christian Cornelssen <ccorn@cs.tu-berlin.de>,
2480 Andreas Schwab <schwab@suse.de>
2481
2482 * configure.in: Adjust command to source config-ml.in to account
2483 for changes to the libffi_basedir definition.
2484 (libffi_basedir): Remove ${srcdir} from value and include trailing
2485 slash if nonempty.
2486
2487 * configure: Regenerate.
2488
24892003-01-29 Franz Sirl <Franz.Sirl-kernel@lauterbach.com>
2490
2491 * src/powerpc/ppc_closure.S: Recode to fit shared libs.
2492
24932003-01-28 Andrew Haley <aph@redhat.com>
2494
2495 * include/ffi.h.in: Enable FFI_CLOSURES for x86_64.
2496 * src/x86/ffi64.c (ffi_prep_closure): New.
2497 (ffi_closure_UNIX64_inner): New.
2498 * src/x86/unix64.S (ffi_closure_UNIX64): New.
2499
25002003-01-27 Alexandre Oliva <aoliva@redhat.com>
2501
2502 * configure.in (toolexecdir, toolexeclibdir): Set and AC_SUBST.
2503 Remove USE_LIBDIR conditional.
2504 * Makefile.am (toolexecdir, toolexeclibdir): Don't override.
2505 * Makefile.in, configure: Rebuilt.
2506
25072003-01027 David Edelsohn <edelsohn@gnu.org>
2508
2509 * Makefile.am (TARGET_SRC_POWERPC_AIX): Fix typo.
2510 * Makefile.in: Regenerate.
2511
25122003-01-22 Andrew Haley <aph@redhat.com>
2513
2514 * src/powerpc/darwin.S (_ffi_call_AIX): Add Augmentation size to
2515 unwind info.
2516
25172003-01-21 Andreas Tobler <a.tobler@schweiz.ch>
2518
2519 * src/powerpc/darwin.S: Add unwind info.
2520 * src/powerpc/darwin_closure.S: Likewise.
2521
25222003-01-14 Andrew Haley <aph@redhat.com>
2523
2524 * src/x86/ffi64.c (ffi_prep_args): Check for void retval.
2525 (ffi_prep_cif_machdep): Likewise.
2526 * src/x86/unix64.S: Add unwind info.
2527
25282003-01-14 Andreas Jaeger <aj@suse.de>
2529
2530 * src/ffitest.c (main): Only use ffi_closures if those are
2531 supported.
2532
25332003-01-13 Andreas Tobler <a.tobler@schweiz.ch>
2534
2535 * libffi/src/ffitest.c
2536 add closure testcases
2537
25382003-01-13 Kevin B. Hendricks <khendricks@ivey.uwo.ca>
2539
2540 * libffi/src/powerpc/ffi.c
2541 fix alignment bug for float (4 byte aligned iso 8 byte)
2542
25432003-01-09 Geoffrey Keating <geoffk@apple.com>
2544
2545 * src/powerpc/ffi_darwin.c: Remove RCS version string.
2546 * src/powerpc/darwin.S: Remove RCS version string.
2547
25482003-01-03 Jeff Sturm <jsturm@one-point.com>
2549
2550 * include/ffi.h.in: Add closure defines for SPARC, SPARC64.
2551 * src/ffitest.c (main): Use static storage for closure.
2552 * src/sparc/ffi.c (ffi_prep_closure, ffi_closure_sparc_inner): New.
2553 * src/sparc/v8.S (ffi_closure_v8): New.
2554 * src/sparc/v9.S (ffi_closure_v9): New.
2555
25562002-11-10 Ranjit Mathew <rmathew@hotmail.com>
2557
2558 * include/ffi.h.in: Added FFI_STDCALL ffi_type
2559 enumeration for X86_WIN32.
2560 * src/x86/win32.S: Added ffi_call_STDCALL function
2561 definition.
2562 * src/x86/ffi.c (ffi_call/ffi_raw_call): Added
2563 switch cases for recognising FFI_STDCALL and
2564 calling ffi_call_STDCALL if target is X86_WIN32.
2565 * src/ffitest.c (my_stdcall_strlen/stdcall_many):
2566 stdcall versions of the "my_strlen" and "many"
2567 test functions (for X86_WIN32).
2568 Added test cases to test stdcall invocation using
2569 these functions.
2570
25712002-12-02 Kaz Kojima <kkojima@gcc.gnu.org>
2572
2573 * src/sh/sysv.S: Add DWARF2 unwind info.
2574
25752002-11-27 Ulrich Weigand <uweigand@de.ibm.com>
2576
2577 * src/s390/sysv.S (.eh_frame section): Make section read-only.
2578
25792002-11-26 Jim Wilson <wilson@redhat.com>
2580
2581 * src/types.c (FFI_TYPE_POINTER): Has size 8 on IA64.
2582
25832002-11-23 H.J. Lu <hjl@gnu.org>
2584
2585 * acinclude.m4: Add dummy AM_PROG_LIBTOOL.
2586 Include ../config/accross.m4.
2587 * aclocal.m4; Rebuild.
2588 * configure: Likewise.
2589
25902002-11-15 Ulrich Weigand <uweigand@de.ibm.com>
2591
2592 * src/s390/sysv.S (.eh_frame section): Adapt to pcrel FDE encoding.
2593
25942002-11-11 DJ Delorie <dj@redhat.com>
2595
2596 * configure.in: Look for common files in the right place.
2597
25982002-10-08 Ulrich Weigand <uweigand@de.ibm.com>
2599
2600 * src/java_raw_api.c (ffi_java_raw_to_ptrarray): Interpret
2601 raw data as _Jv_word values, not ffi_raw.
2602 (ffi_java_ptrarray_to_raw): Likewise.
2603 (ffi_java_rvalue_to_raw): New function.
2604 (ffi_java_raw_call): Call it.
2605 (ffi_java_raw_to_rvalue): New function.
2606 (ffi_java_translate_args): Call it.
2607 * src/ffitest.c (closure_test_fn): Interpret return value
2608 as ffi_arg, not int.
2609 * src/s390/ffi.c (ffi_prep_cif_machdep): Add missing
2610 FFI_TYPE_POINTER case.
2611 (ffi_closure_helper_SYSV): Likewise. Also, assume return
2612 values extended to word size.
2613
26142002-10-02 Andreas Jaeger <aj@suse.de>
2615
2616 * src/x86/ffi64.c (ffi_prep_cif_machdep): Remove debug output.
2617
26182002-10-01 Bo Thorsen <bo@smetana.suse.de>
2619
2620 * include/ffi.h.in: Fix i386 win32 compilation.
2621
26222002-09-30 Ulrich Weigand <uweigand@de.ibm.com>
2623
2624 * configure.in: Add s390x-*-linux-* target.
2625 * configure: Regenerate.
2626 * include/ffi.h.in: Define S390X for s390x targets.
2627 (FFI_CLOSURES): Define for s390/s390x.
2628 (FFI_TRAMPOLINE_SIZE): Likewise.
2629 (FFI_NATIVE_RAW_API): Likewise.
2630 * src/prep_cif.c (ffi_prep_cif): Do not compute stack space for s390.
2631 * src/types.c (FFI_TYPE_POINTER): Use 8-byte pointers on s390x.
2632 * src/s390/ffi.c: Major rework of existing code. Add support for
2633 s390x targets. Add closure support.
2634 * src/s390/sysv.S: Likewise.
2635
26362002-09-29 Richard Earnshaw <rearnsha@arm.com>
2637
2638 * src/arm/sysv.S: Fix typo.
2639
26402002-09-28 Richard Earnshaw <rearnsha@arm.com>
2641
2642 * src/arm/sysv.S: If we don't have machine/asm.h and the pre-processor
2643 has defined __USER_LABEL_PREFIX__, then use it in CNAME.
2644 (ffi_call_SYSV): Handle soft-float.
2645
26462002-09-27 Bo Thorsen <bo@suse.de>
2647
2648 * include/ffi.h.in: Fix multilib x86-64 support.
2649
26502002-09-22 Kaveh R. Ghazi <ghazi@caip.rutgers.edu>
2651
2652 * Makefile.am (all-multi): Fix multilib parallel build.
2653
26542002-07-19 Kaz Kojima <kkojima@gcc.gnu.org>
2655
2656 * configure.in (sh[34]*-*-linux*): Add brackets.
2657 * configure: Regenerate.
2658
26592002-07-18 Kaz Kojima <kkojima@gcc.gnu.org>
2660
2661 * Makefile.am: Add SH support.
2662 * Makefile.in: Regenerate.
2663 * configure.in (sh-*-linux*, sh[34]*-*-linux*): Add target.
2664 * configure: Regenerate.
2665 * include/ffi.h.in: Add SH support.
2666 * src/sh/ffi.c: New file.
2667 * src/sh/sysv.S: New file.
2668 * src/types.c: Add SH support.
2669
26702002-07-16 Bo Thorsen <bo@suse.de>
2671
2672 * src/x86/ffi64.c: New file that adds x86-64 support.
2673 * src/x86/unix64.S: New file that handles argument setup for
2674 x86-64.
2675 * src/x86/sysv.S: Don't use this on x86-64.
2676 * src/x86/ffi.c: Don't use this on x86-64.
2677 Remove unused vars.
2678 * src/prep_cif.c (ffi_prep_cif): Don't do stack size calculation
2679 for x86-64.
2680 * src/ffitest.c (struct6): New test that tests a special case in
2681 the x86-64 ABI.
2682 (struct7): Likewise.
2683 (struct8): Likewise.
2684 (struct9): Likewise.
2685 (closure_test_fn): Silence warning about this when it's not used.
2686 (main): Add the new tests.
2687 (main): Fix a couple of wrong casts and silence some compiler warnings.
2688 * include/ffi.h.in: Add x86-64 ABI definition.
2689 * fficonfig.h.in: Regenerate.
2690 * Makefile.am: Add x86-64 support.
2691 * configure.in: Likewise.
2692 * Makefile.in: Regenerate.
2693 * configure: Likewise.
2694
26952002-06-24 Bo Thorsen <bo@suse.de>
2696
2697 * src/types.c: Merge settings for similar architectures.
2698 Add x86-64 sizes and alignments.
2699
27002002-06-23 Bo Thorsen <bo@suse.de>
2701
2702 * src/arm/ffi.c (ffi_prep_args): Remove unused vars.
2703 * src/sparc/ffi.c (ffi_prep_args_v8): Likewise.
2704 * src/mips/ffi.c (ffi_prep_args): Likewise.
2705 * src/m68k/ffi.c (ffi_prep_args): Likewise.
2706
27072002-07-18 H.J. Lu (hjl@gnu.org)
2708
2709 * Makefile.am (TARGET_SRC_MIPS_LINUX): New.
2710 (libffi_la_SOURCES): Support MIPS_LINUX.
2711 (libffi_convenience_la_SOURCES): Likewise.
2712 * Makefile.in: Regenerated.
2713
2714 * configure.in (mips64*-*): Skip.
2715 (mips*-*-linux*): New.
2716 * configure: Regenerated.
2717
2718 * src/mips/ffi.c: Include <sgidefs.h>.
2719
27202002-06-06 Ulrich Weigand <uweigand@de.ibm.com>
2721
2722 * src/s390/sysv.S: Save/restore %r6. Add DWARF-2 unwind info.
2723
27242002-05-27 Roger Sayle <roger@eyesopen.com>
2725
2726 * src/x86/ffi.c (ffi_prep_args): Remove reference to avn.
2727
27282002-05-27 Bo Thorsen <bo@suse.de>
2729
2730 * src/x86/ffi.c (ffi_prep_args): Remove unused variable and
2731 fix formatting.
2732
27332002-05-13 Andreas Tobler <a.tobler@schweiz.ch>
2734
2735 * src/powerpc/ffi_darwin.c (ffi_prep_closure): Declare fd at
2736 beginning of function (for older apple cc).
2737
27382002-05-08 Alexandre Oliva <aoliva@redhat.com>
2739
2740 * configure.in (ORIGINAL_LD_FOR_MULTILIBS): Preserve LD at
2741 script entry, and set LD to it when configuring multilibs.
2742 * configure: Rebuilt.
2743
27442002-05-05 Jason Thorpe <thorpej@wasabisystems.com>
2745
2746 * configure.in (sparc64-*-netbsd*): Add target.
2747 (sparc-*-netbsdelf*): Likewise.
2748 * configure: Regenerate.
2749
27502002-04-28 David S. Miller <davem@redhat.com>
2751
2752 * configure.in, configure: Fix SPARC test in previous change.
2753
27542002-04-29 Gerhard Tonn <GerhardTonn@swol.de>
2755
2756 * Makefile.am: Add Linux for S/390 support.
2757 * Makefile.in: Regenerate.
2758 * configure.in: Add Linux for S/390 support.
2759 * configure: Regenerate.
2760 * include/ffi.h.in: Add Linux for S/390 support.
2761 * src/s390/ffi.c: New file from libffi CVS tree.
2762 * src/s390/sysv.S: New file from libffi CVS tree.
2763
27642002-04-28 Jakub Jelinek <jakub@redhat.com>
2765
2766 * configure.in (HAVE_AS_SPARC_UA_PCREL): Check for working
2767 %r_disp32().
2768 * src/sparc/v8.S: Use it.
2769 * src/sparc/v9.S: Likewise.
2770 * fficonfig.h.in: Rebuilt.
2771 * configure: Rebuilt.
2772
27732002-04-08 Hans Boehm <Hans_Boehm@hp.com>
2774
2775 * src/java_raw_api.c (ffi_java_raw_size): Handle FFI_TYPE_DOUBLE
2776 correctly.
2777 * src/ia64/unix.S: Add unwind information. Fix comments.
2778 Save sp in a way that's compatible with unwind info.
2779 (ffi_call_unix): Correctly restore sp in all cases.
2780 * src/ia64/ffi.c: Add, fix comments.
2781
27822002-04-08 Jakub Jelinek <jakub@redhat.com>
2783
2784 * src/sparc/v8.S: Make .eh_frame dependent on target word size.
2785
27862002-04-06 Jason Thorpe <thorpej@wasabisystems.com>
2787
2788 * configure.in (alpha*-*-netbsd*): Add target.
2789 * configure: Regenerate.
2790
27912002-04-04 Jeff Sturm <jsturm@one-point.com>
2792
2793 * src/sparc/v8.S: Add unwind info.
2794 * src/sparc/v9.S: Likewise.
2795
27962002-03-30 Krister Walfridsson <cato@df.lth.se>
2797
2798 * configure.in: Enable i*86-*-netbsdelf*.
2799 * configure: Rebuilt.
2800
28012002-03-29 David Billinghurst <David.Billinghurst@riotinto.com>
2802
2803 PR other/2620
2804 * src/mips/n32.s: Delete
2805 * src/mips/o32.s: Delete
2806
28072002-03-21 Loren J. Rittle <ljrittle@acm.org>
2808
2809 * configure.in: Enable alpha*-*-freebsd*.
2810 * configure: Rebuilt.
2811
28122002-03-17 Bryce McKinlay <bryce@waitaki.otago.ac.nz>
2813
2814 * Makefile.am: libfficonvenience -> libffi_convenience.
2815 * Makefile.in: Rebuilt.
2816
2817 * Makefile.am: Define ffitest_OBJECTS.
2818 * Makefile.in: Rebuilt.
2819
28202002-03-07 Andreas Tobler <toa@pop.agri.ch>
2821 David Edelsohn <edelsohn@gnu.org>
2822
2823 * Makefile.am (EXTRA_DIST): Add Darwin and AIX closure files.
2824 (TARGET_SRC_POWERPC_AIX): Add aix_closure.S.
2825 (TARGET_SRC_POWERPC_DARWIN): Add darwin_closure.S.
2826 * Makefile.in: Regenerate.
2827 * include/ffi.h.in: Add AIX and Darwin closure definitions.
2828 * src/powerpc/ffi_darwin.c (ffi_prep_closure): New function.
2829 (flush_icache, flush_range): New functions.
2830 (ffi_closure_helper_DARWIN): New function.
2831 * src/powerpc/aix_closure.S: New file.
2832 * src/powerpc/darwin_closure.S: New file.
2833
28342002-02-24 Jeff Sturm <jsturm@one-point.com>
2835
2836 * include/ffi.h.in: Add typedef for ffi_arg.
2837 * src/ffitest.c (main): Declare rint with ffi_arg.
2838
28392002-02-21 Andreas Tobler <toa@pop.agri.ch>
2840
2841 * src/powerpc/ffi_darwin.c (ffi_prep_args): Skip appropriate
2842 number of GPRs for floating-point arguments.
2843
28442002-01-31 Anthony Green <green@redhat.com>
2845
2846 * configure: Rebuilt.
2847 * configure.in: Replace CHECK_SIZEOF and endian tests with
2848 cross-compiler friendly macros.
2849 * aclocal.m4 (AC_COMPILE_CHECK_SIZEOF, AC_C_BIGENDIAN_CROSS): New
2850 macros.
2851
28522002-01-18 David Edelsohn <edelsohn@gnu.org>
2853
2854 * src/powerpc/darwin.S (_ffi_call_AIX): New.
2855 * src/powerpc/aix.S (ffi_call_DARWIN): New.
2856
28572002-01-17 David Edelsohn <edelsohn@gnu.org>
2858
2859 * Makefile.am (EXTRA_DIST): Add Darwin and AIX files.
2860 (TARGET_SRC_POWERPC_AIX): New.
2861 (POWERPC_AIX): New stanza.
2862 * Makefile.in: Regenerate.
2863 * configure.in: Add AIX case.
2864 * configure: Regenerate.
2865 * include/ffi.h.in (ffi_abi): Add FFI_AIX.
2866 * src/powerpc/ffi_darwin.c (ffi_status): Use "long" to scale frame
2867 size. Fix "long double" support.
2868 (ffi_call): Add FFI_AIX case.
2869 * src/powerpc/aix.S: New.
2870
28712001-10-09 John Hornkvist <john@toastedmarshmallow.com>
2872
2873 Implement Darwin PowerPC ABI.
2874 * configure.in: Handle powerpc-*-darwin*.
2875 * Makefile.am: Set source files for POWERPC_DARWIN.
2876 * configure: Rebuilt.
2877 * Makefile.in: Rebuilt.
2878 * include/ffi.h.in: Define FFI_DARWIN and FFI_DEFAULT_ABI for
2879 POWERPC_DARWIN.
2880 * src/powerpc/darwin.S: New file.
2881 * src/powerpc/ffi_darwin.c: New file.
2882
28832001-10-07 Joseph S. Myers <jsm28@cam.ac.uk>
2884
2885 * src/x86/ffi.c: Fix spelling error of "separate" as "seperate".
2886
28872001-07-16 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
2888
2889 * src/x86/sysv.S: Avoid gas-only .balign directive.
2890 Use C style comments.
2891
28922001-07-16 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
2893
2894 * src/alpha/ffi.c (ffi_prep_closure): Avoid gas-only mnemonic.
2895 Fixes PR bootstrap/3563.
2896
28972001-06-26 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
2898
2899 * src/alpha/osf.S (ffi_closure_osf): Use .rdata for ECOFF.
2900
29012001-06-25 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
2902
2903 * configure.in: Recognize sparc*-sun-* host.
2904 * configure: Regenerate.
2905
29062001-06-06 Andrew Haley <aph@redhat.com>
2907
2908 * src/alpha/osf.S (__FRAME_BEGIN__): Conditionalize for ELF.
2909
29102001-06-03 Andrew Haley <aph@redhat.com>
2911
2912 * src/alpha/osf.S: Add unwind info.
2913 * src/powerpc/sysv.S: Add unwind info.
2914 * src/powerpc/ppc_closure.S: Likewise.
2915
29162000-05-31 Jeff Sturm <jsturm@one-point.com>
2917
2918 * configure.in: Fix AC_ARG_ENABLE usage.
2919 * configure: Rebuilt.
2920
29212001-05-06 Bryce McKinlay <bryce@waitaki.otago.ac.nz>
2922
2923 * configure.in: Remove warning about beta code.
2924 * configure: Rebuilt.
2925
29262001-04-25 Hans Boehm <Hans_Boehm@hp.com>
2927
2928 * src/ia64/unix.S: Restore stack pointer when returning from
2929 ffi_closure_UNIX.
2930 * src/ia64/ffi.c: Fix typo in comment.
2931
29322001-04-18 Jim Wilson <wilson@redhat.com>
2933
2934 * src/ia64/unix.S: Delete unnecessary increment and decrement of loc2
2935 to eliminate RAW DV.
2936
29372001-04-12 Bryce McKinlay <bryce@albatross.co.nz>
2938
2939 * Makefile.am: Make a libtool convenience library.
2940 * Makefile.in: Rebuilt.
2941
29422001-03-29 Bryce McKinlay <bryce@albatross.co.nz>
2943
2944 * configure.in: Use different syntax for subdirectory creation.
2945 * configure: Rebuilt.
2946
29472001-03-27 Jon Beniston <jon@beniston.com>
2948
2949 * configure.in: Added X86_WIN32 target (Win32, CygWin, MingW).
2950 * configure: Rebuilt.
2951 * Makefile.am: Added X86_WIN32 target support.
2952 * Makefile.in: Rebuilt.
2953
2954 * include/ffi.h.in: Added X86_WIN32 target support.
2955
2956 * src/ffitest.c: Doesn't run structure tests for X86_WIN32 targets.
2957 * src/types.c: Added X86_WIN32 target support.
2958
2959 * src/x86/win32.S: New file. Based on sysv.S, but with EH
2960 stuff removed and made to work with CygWin's gas.
2961
29622001-03-26 Bryce McKinlay <bryce@albatross.co.nz>
2963
2964 * configure.in: Make target subdirectory in build dir.
2965 * Makefile.am: Override suffix based rules to specify correct output
2966 subdirectory.
2967 * Makefile.in: Rebuilt.
2968 * configure: Rebuilt.
2969
29702001-03-23 Kevin B Hendricks <khendricks@ivey.uwo.ca>
2971
2972 * src/powerpc/ppc_closure.S: New file.
2973 * src/powerpc/ffi.c (ffi_prep_args): Fixed ABI compatibility bug
2974 involving long long and register pairs.
2975 (ffi_prep_closure): New function.
2976 (flush_icache): Likewise.
2977 (ffi_closure_helper_SYSV): Likewise.
2978 * include/ffi.h.in (FFI_CLOSURES): Define on PPC.
2979 (FFI_TRAMPOLINE_SIZE): Likewise.
2980 (FFI_NATIVE_RAW_API): Likewise.
2981 * Makefile.in: Rebuilt.
2982 * Makefile.am (EXTRA_DIST): Added src/powerpc/ppc_closure.S.
2983 (TARGET_SRC_POWERPC): Likewise.
2984
29852001-03-19 Tom Tromey <tromey@redhat.com>
2986
2987 * Makefile.in: Rebuilt.
2988 * Makefile.am (ffitest_LDFLAGS): New macro.
2989
29902001-03-02 Nick Clifton <nickc@redhat.com>
2991
2992 * include/ffi.h.in: Remove RCS ident string.
2993 * include/ffi_mips.h: Remove RCS ident string.
2994 * src/debug.c: Remove RCS ident string.
2995 * src/ffitest.c: Remove RCS ident string.
2996 * src/prep_cif.c: Remove RCS ident string.
2997 * src/types.c: Remove RCS ident string.
2998 * src/alpha/ffi.c: Remove RCS ident string.
2999 * src/alpha/osf.S: Remove RCS ident string.
3000 * src/arm/ffi.c: Remove RCS ident string.
3001 * src/arm/sysv.S: Remove RCS ident string.
3002 * src/mips/ffi.c: Remove RCS ident string.
3003 * src/mips/n32.S: Remove RCS ident string.
3004 * src/mips/o32.S: Remove RCS ident string.
3005 * src/sparc/ffi.c: Remove RCS ident string.
3006 * src/sparc/v8.S: Remove RCS ident string.
3007 * src/sparc/v9.S: Remove RCS ident string.
3008 * src/x86/ffi.c: Remove RCS ident string.
3009 * src/x86/sysv.S: Remove RCS ident string.
3010
30112001-02-08 Joseph S. Myers <jsm28@cam.ac.uk>
3012
3013 * include/ffi.h.in: Change sourceware.cygnus.com references to
3014 gcc.gnu.org.
3015
30162000-12-09 Richard Henderson <rth@redhat.com>
3017
3018 * src/alpha/ffi.c (ffi_call): Simplify struct return test.
3019 (ffi_closure_osf_inner): Index rather than increment avalue
3020 and arg_types. Give ffi_closure_osf the raw return value type.
3021 * src/alpha/osf.S (ffi_closure_osf): Handle return value type
3022 promotion.
3023
30242000-12-07 Richard Henderson <rth@redhat.com>
3025
3026 * src/raw_api.c (ffi_translate_args): Fix typo.
3027 (ffi_prep_closure): Likewise.
3028
3029 * include/ffi.h.in [ALPHA]: Define FFI_CLOSURES and
3030 FFI_TRAMPOLINE_SIZE.
3031 * src/alpha/ffi.c (ffi_prep_cif_machdep): Adjust minimal
3032 cif->bytes for new ffi_call_osf implementation.
3033 (ffi_prep_args): Absorb into ...
3034 (ffi_call): ... here. Do all stack allocation here and
3035 avoid a callback function.
3036 (ffi_prep_closure, ffi_closure_osf_inner): New.
3037 * src/alpha/osf.S (ffi_call_osf): Reimplement with no callback.
3038 (ffi_closure_osf): New.
3039
30402000-09-10 Alexandre Oliva <aoliva@redhat.com>
3041
3042 * config.guess, config.sub, install-sh: Removed.
3043 * ltconfig, ltmain.sh, missing, mkinstalldirs: Likewise.
3044 * Makefile.in: Rebuilt.
3045
3046 * acinclude.m4: Include libtool macros from the top level.
3047 * aclocal.m4, configure: Rebuilt.
3048
30492000-08-22 Alexandre Oliva <aoliva@redhat.com>
3050
3051 * configure.in [i*86-*-freebsd*] (TARGET, TARGETDIR): Set.
3052 * configure: Rebuilt.
3053
30542000-05-11 Scott Bambrough <scottb@netwinder.org>
3055
3056 * libffi/src/arm/sysv.S (ffi_call_SYSV): Doubles are not saved to
3057 memory correctly. Use conditional instructions, not branches where
3058 possible.
3059
30602000-05-04 Tom Tromey <tromey@cygnus.com>
3061
3062 * configure: Rebuilt.
3063 * configure.in: Match `arm*-*-linux-*'.
3064 From Chris Dornan <cdornan@arm.com>.
3065
30662000-04-28 Jakub Jelinek <jakub@redhat.com>
3067
3068 * Makefile.am (SUBDIRS): Define.
3069 (AM_MAKEFLAGS): Likewise.
3070 (Multilib support.): Add section.
3071 * Makefile.in: Rebuilt.
3072 * ltconfig (extra_compiler_flags, extra_compiler_flags_value):
3073 New variables. Set for gcc using -print-multi-lib. Export them
3074 to libtool.
3075 (sparc64-*-linux-gnu*): Use libsuff 64 for search paths.
3076 * ltmain.sh (B|b|V): Don't throw away gcc's -B, -b and -V options
3077 for -shared links.
3078 (extra_compiler_flags_value, extra_compiler_flags): Check these
3079 for extra compiler options which need to be passed down in
3080 compiler_flags.
3081
30822000-04-16 Anthony Green <green@redhat.com>
3083
3084 * configure: Rebuilt.
3085 * configure.in: Change i*86-pc-linux* to i*86-*-linux*.
3086
30872000-04-14 Jakub Jelinek <jakub@redhat.com>
3088
3089 * include/ffi.h.in (SPARC64): Define for 64bit SPARC builds.
3090 Set SPARC FFI_DEFAULT_ABI based on SPARC64 define.
3091 * src/sparc/ffi.c (ffi_prep_args_v8): Renamed from ffi_prep_args.
3092 Replace all void * sizeofs with sizeof(int).
3093 Only compare type with FFI_TYPE_LONGDOUBLE if LONGDOUBLE is
3094 different than DOUBLE.
3095 Remove FFI_TYPE_SINT32 and FFI_TYPE_UINT32 cases (handled elsewhere).
3096 (ffi_prep_args_v9): New function.
3097 (ffi_prep_cif_machdep): Handle V9 ABI and long long on V8.
3098 (ffi_V9_return_struct): New function.
3099 (ffi_call): Handle FFI_V9 ABI from 64bit code and FFI_V8 ABI from
3100 32bit code (not yet cross-arch calls).
3101 * src/sparc/v8.S: Add struct return delay nop.
3102 Handle long long.
3103 * src/sparc/v9.S: New file.
3104 * src/prep_cif.c (ffi_prep_cif): Return structure pointer
3105 is used on sparc64 only for structures larger than 32 bytes.
3106 Pass by reference for structures is done for structure arguments
3107 larger than 16 bytes.
3108 * src/ffitest.c (main): Use 64bit rint on sparc64.
3109 Run long long tests on sparc.
3110 * src/types.c (FFI_TYPE_POINTER): Pointer is 64bit on alpha and
3111 sparc64.
3112 (FFI_TYPE_LONGDOUBLE): long double is 128 bit aligned to 128 bits
3113 on sparc64.
3114 * configure.in (sparc-*-linux*): New supported target.
3115 (sparc64-*-linux*): Likewise.
3116 * configure: Rebuilt.
3117 * Makefile.am: Add v9.S to SPARC files.
3118 * Makefile.in: Likewise.
3119 (LINK): Surround $(CCLD) into double quotes, so that multilib
3120 compiles work correctly.
3121
31222000-04-04 Alexandre Petit-Bianco <apbianco@cygnus.com>
3123
3124 * configure: Rebuilt.
3125 * configure.in: (i*86-*-solaris*): New libffi target. Patch
3126 proposed by Bryce McKinlay.
3127
31282000-03-20 Tom Tromey <tromey@cygnus.com>
3129
3130 * Makefile.in: Hand edit for java_raw_api.lo.
3131
31322000-03-08 Bryce McKinlay <bryce@albatross.co.nz>
3133
3134 * config.guess, config.sub: Update from the gcc tree.
3135 Fix for PR libgcj/168.
3136
31372000-03-03 Tom Tromey <tromey@cygnus.com>
3138
3139 * Makefile.in: Fixed ia64 by hand.
3140
3141 * configure: Rebuilt.
3142 * configure.in (--enable-multilib): New option.
3143 (libffi_basedir): New subst.
3144 (AC_OUTPUT): Added multilib code.
3145
31462000-03-02 Tom Tromey <tromey@cygnus.com>
3147
3148 * Makefile.in: Rebuilt.
3149 * Makefile.am (TARGET_SRC_IA64): Use `ia64', not `alpha', as
3150 directory name.
3151
31522000-02-25 Hans Boehm <boehm@acm.org>
3153
3154 * src/ia64/ffi.c, src/ia64/ia64_flags.h, src/ia64/unix.S: New
3155 files.
3156 * src/raw_api.c (ffi_translate_args): Fixed typo in argument
3157 list.
3158 (ffi_prep_raw_closure): Use ffi_translate_args, not
3159 ffi_closure_translate.
3160 * src/java_raw_api.c: New file.
3161 * src/ffitest.c (closure_test_fn): New function.
3162 (main): Define `rint' as long long on IA64. Added new test when
3163 FFI_CLOSURES is defined.
3164 * include/ffi.h.in (ALIGN): Use size_t, not unsigned.
3165 (ffi_abi): Recognize IA64.
3166 (ffi_raw): Added `flt' field.
3167 Added "Java raw API" code.
3168 * configure.in: Recognize ia64.
3169 * Makefile.am (TARGET_SRC_IA64): New macro.
3170 (libffi_la_common_SOURCES): Added java_raw_api.c.
3171 (libffi_la_SOURCES): Define in IA64 case.
3172
31732000-01-04 Tom Tromey <tromey@cygnus.com>
3174
3175 * Makefile.in: Rebuilt with newer automake.
3176
31771999-12-31 Tom Tromey <tromey@cygnus.com>
3178
3179 * Makefile.am (INCLUDES): Added -I$(top_srcdir)/src.
3180
31811999-09-01 Tom Tromey <tromey@cygnus.com>
3182
3183 * include/ffi.h.in: Removed PACKAGE and VERSION defines and
3184 undefs.
3185 * fficonfig.h.in: Rebuilt.
3186 * configure: Rebuilt.
3187 * configure.in: Pass 3rd argument to AM_INIT_AUTOMAKE.
3188 Use AM_PROG_LIBTOOL (automake 1.4 compatibility).
3189 * acconfig.h: Don't #undef PACKAGE or VERSION.
3190
31911999-08-09 Anthony Green <green@cygnus.com>
3192
3193 * include/ffi.h.in: Try to work around messy header problem
3194 with PACKAGE and VERSION.
3195
3196 * configure: Rebuilt.
3197 * configure.in: Change version to 2.00-beta.
3198
3199 * fficonfig.h.in: Rebuilt.
3200 * acconfig.h (FFI_NO_STRUCTS, FFI_NO_RAW_API): Define.
3201
3202 * src/x86/ffi.c (ffi_raw_call): Rename.
3203
32041999-08-02 Kresten Krab Thorup <krab@dominiq.is.s.u-tokyo.ac.jp>
3205
3206 * src/x86/ffi.c (ffi_closure_SYSV): New function.
3207 (ffi_prep_incoming_args_SYSV): Ditto.
3208 (ffi_prep_closure): Ditto.
3209 (ffi_closure_raw_SYSV): Ditto.
3210 (ffi_prep_raw_closure): More ditto.
3211 (ffi_call_raw): Final ditto.
3212
3213 * include/ffi.h.in: Add definitions for closure and raw API.
3214
3215 * src/x86/ffi.c (ffi_prep_cif_machdep): Added case for
3216 FFI_TYPE_UINT64.
3217
3218 * Makefile.am (libffi_la_common_SOURCES): Added raw_api.c
3219
3220 * src/raw_api.c: New file.
3221
3222 * include/ffi.h.in (ffi_raw): New type.
3223 (UINT_ARG, SINT_ARG): New defines.
3224 (ffi_closure, ffi_raw_closure): New types.
3225 (ffi_prep_closure, ffi_prep_raw_closure): New declarations.
3226
3227 * configure.in: Add check for endianness and sizeof void*.
3228
3229 * src/x86/sysv.S (ffi_call_SYSV): Call fixup routine via argument,
3230 instead of directly.
3231
3232 * configure: Rebuilt.
3233
3234Thu Jul 8 14:28:42 1999 Anthony Green <green@cygnus.com>
3235
3236 * configure.in: Add x86 and powerpc BeOS configurations.
3237 From Makoto Kato <m_kato@ga2.so-net.ne.jp>.
3238
32391999-05-09 Anthony Green <green@cygnus.com>
3240
3241 * configure.in: Add warning about this being beta code.
3242 Remove src/Makefile.am from the picture.
3243 * configure: Rebuilt.
3244
3245 * Makefile.am: Move logic from src/Makefile.am. Add changes
3246 to support libffi as a target library.
3247 * Makefile.in: Rebuilt.
3248
3249 * aclocal.m4, config.guess, config.sub, ltconfig, ltmain.sh:
3250 Upgraded to new autoconf, automake, libtool.
3251
3252 * README: Tweaks.
3253
3254 * LICENSE: Update copyright date.
3255
3256 * src/Makefile.am, src/Makefile.in: Removed.
3257
32581998-11-29 Anthony Green <green@cygnus.com>
3259
3260 * include/ChangeLog: Removed.
3261 * src/ChangeLog: Removed.
3262 * src/mips/ChangeLog: Removed.
3263 * src/sparc/ChangeLog: Remboved.
3264 * src/x86/ChangeLog: Removed.
3265
3266 * ChangeLog.v1: Created.
Mirror of https://gcc.gnu.org/git/gcc.git