1 2008-01-30 H.J. Lu <hongjiu.lu@intel.com>
4 * src/x86/sysv.S (ffi_closure_SYSV): Pop 4 byte from stack when
7 * testsuite/libffi.call/call.exp: Add "-O2 -fomit-frame-pointer"
10 2008-01-24 David Edelsohn <edelsohn@gnu.org>
12 * configure: Regenerate.
14 2008-01-06 Andreas Tobler <a.tobler@schweiz.org>
16 * src/x86/ffi.c (ffi_prep_cif_machdep): Fix thinko.
18 2008-01-05 Andreas Tobler <a.tobler@schweiz.org>
21 * src/x86/ffi.c (ffi_prep_cif_machdep): Add code for
22 signed/unsigned int8/16 for X86_DARWIN.
23 Updated copyright info.
24 Handle one and two byte structs with special cif->flags.
25 * src/x86/ffitarget.h: Add special types for one and two byte structs.
26 Updated copyright info.
27 * src/x86/darwin.S (ffi_call_SYSV): Rewrite to use a jump table like
29 Remove code to pop args from the stack after call.
30 Special-case signed/unsigned for int8/16, one and two byte structs.
31 (ffi_closure_raw_SYSV): Handle FFI_TYPE_UINT8,
32 FFI_TYPE_SINT8, FFI_TYPE_UINT16, FFI_TYPE_SINT16, FFI_TYPE_UINT32,
34 Updated copyright info.
36 2007-12-08 David Daney <ddaney@avtrex.com>
38 * src/mips/n32.S (ffi_call_N32): Replace dadd with ADDU, dsub with
39 SUBU, add with ADDU and use smaller code sequences.
41 2007-12-07 David Daney <ddaney@avtrex.com>
43 * src/mips/ffi.c (ffi_prep_cif_machdep): Handle long double return
46 2007-12-06 David Daney <ddaney@avtrex.com>
48 * include/ffi.h.in (FFI_SIZEOF_JAVA_RAW): Define if not already
50 (ffi_java_raw): New typedef.
51 (ffi_java_raw_call, ffi_java_ptrarray_to_raw,
52 ffi_java_raw_to_ptrarray): Change parameter types from ffi_raw to
54 (ffi_java_raw_closure) : Same.
55 (ffi_prep_java_raw_closure, ffi_prep_java_raw_closure_loc): Change
57 * src/java_raw_api.c (ffi_java_raw_size): Replace FFI_SIZEOF_ARG with
59 (ffi_java_raw_to_ptrarray): Change type of raw to ffi_java_raw.
60 Replace FFI_SIZEOF_ARG with FFI_SIZEOF_JAVA_RAW. Use
61 sizeof(ffi_java_raw) for alignment calculations.
62 (ffi_java_ptrarray_to_raw): Same.
63 (ffi_java_rvalue_to_raw): Add special handling for FFI_TYPE_POINTER
64 if FFI_SIZEOF_JAVA_RAW == 4.
65 (ffi_java_raw_to_rvalue): Same.
66 (ffi_java_raw_call): Change type of raw to ffi_java_raw.
67 (ffi_java_translate_args): Same.
68 (ffi_prep_java_raw_closure_loc, ffi_prep_java_raw_closure): Change
70 * src/mips/ffitarget.h (FFI_SIZEOF_JAVA_RAW): Define for N32 ABI.
72 2007-12-06 David Daney <ddaney@avtrex.com>
74 * src/mips/n32.S (ffi_closure_N32): Use 64-bit add instruction on
77 2007-12-01 Andreas Tobler <a.tobler@schweiz.org>
80 * src/powerpc/ffitarget.h: Introduce new ABI FFI_LINUX_SOFT_FLOAT.
81 Add local FFI_TYPE_UINT128 to handle soft-float long-double-128.
82 * src/powerpc/ffi.c: Distinguish between __NO_FPRS__ and not and
83 set the NUM_FPR_ARG_REGISTERS according to.
84 Add support for potential soft-float support under hard-float
86 (ffi_prep_args_SYSV): Set NUM_FPR_ARG_REGISTERS to 0 in case of
87 FFI_LINUX_SOFT_FLOAT, handle float, doubles and long-doubles according
88 to the FFI_LINUX_SOFT_FLOAT ABI.
89 (ffi_prep_cif_machdep): Likewise.
90 (ffi_closure_helper_SYSV): Likewise.
91 * src/powerpc/ppc_closure.S: Make sure not to store float/double
92 on archs where __NO_FPRS__ is true.
93 Add FFI_TYPE_UINT128 support.
94 * src/powerpc/sysv.S: Add support for soft-float long-double-128.
95 Adjust copyright notice.
97 2007-11-25 Andreas Tobler <a.tobler@schweiz.org>
99 * src/closures.c: Move defintion of MAYBE_UNUSED from here to ...
100 * include/ffi_common.h: ... here.
103 2007-11-17 Andreas Tobler <a.tobler@schweiz.org>
105 * src/powerpc/sysv.S: Load correct cr to compare if we have long double.
106 * src/powerpc/linux64.S: Likewise.
107 * src/powerpc/ffi.c: Add a comment to show which part goes into cr6.
108 * testsuite/libffi.call/return_ldl.c: New test.
110 2007-09-04 <aph@redhat.com>
112 * src/arm/sysv.S (UNWIND): New.
113 (Whole file): Conditionally compile unwinder directives.
114 * src/arm/sysv.S: Add unwinder directives.
116 * src/arm/ffi.c (ffi_prep_args): Align structs by at least 4 bytes.
117 Only treat r0 as a struct address if we're actually returning a
119 Only copy the bytes that are actually within a struct.
120 (ffi_prep_cif_machdep): A Composite Type not larger than 4 bytes
121 is returned in r0, not passed by address.
122 (ffi_call): Allocate a word-sized temporary for the case where
123 a composite is returned in r0.
124 (ffi_prep_incoming_args_SYSV): Align as necessary.
126 2007-08-05 Steven Newbury <s_j_newbury@yahoo.co.uk>
128 * src/arm/ffi.c (FFI_INIT_TRAMPOLINE): Use __clear_cache instead of
129 directly using the sys_cacheflush syscall.
131 2007-07-27 Andrew Haley <aph@redhat.com>
133 * src/arm/sysv.S (ffi_closure_SYSV): Add soft-float.
135 2007-09-03 Maciej W. Rozycki <macro@linux-mips.org>
137 * Makefile.am: Unify MIPS_IRIX and MIPS_LINUX into MIPS.
138 * configure.ac: Likewise.
139 * Makefile.in: Regenerate.
140 * include/Makefile.in: Likewise.
141 * testsuite/Makefile.in: Likewise.
142 * configure: Likewise.
144 2007-08-24 David Daney <ddaney@avtrex.com>
146 * testsuite/libffi.call/return_sl.c: New test.
148 2007-08-10 David Daney <ddaney@avtrex.com>
150 * testsuite/libffi.call/cls_multi_ushort.c,
151 testsuite/libffi.call/cls_align_uint16.c,
152 testsuite/libffi.call/nested_struct1.c,
153 testsuite/libffi.call/nested_struct3.c,
154 testsuite/libffi.call/cls_7_1_byte.c,
155 testsuite/libffi.call/nested_struct5.c,
156 testsuite/libffi.call/cls_double.c,
157 testsuite/libffi.call/nested_struct7.c,
158 testsuite/libffi.call/cls_sint.c,
159 testsuite/libffi.call/nested_struct9.c,
160 testsuite/libffi.call/cls_20byte1.c,
161 testsuite/libffi.call/cls_multi_sshortchar.c,
162 testsuite/libffi.call/cls_align_sint64.c,
163 testsuite/libffi.call/cls_3byte2.c,
164 testsuite/libffi.call/cls_multi_schar.c,
165 testsuite/libffi.call/cls_multi_uchar.c,
166 testsuite/libffi.call/cls_19byte.c,
167 testsuite/libffi.call/cls_9byte1.c,
168 testsuite/libffi.call/cls_align_float.c,
169 testsuite/libffi.call/closure_fn1.c,
170 testsuite/libffi.call/problem1.c,
171 testsuite/libffi.call/closure_fn3.c,
172 testsuite/libffi.call/cls_sshort.c,
173 testsuite/libffi.call/closure_fn5.c,
174 testsuite/libffi.call/cls_align_double.c,
175 testsuite/libffi.call/nested_struct.c,
176 testsuite/libffi.call/cls_2byte.c,
177 testsuite/libffi.call/nested_struct10.c,
178 testsuite/libffi.call/cls_4byte.c,
179 testsuite/libffi.call/cls_6byte.c,
180 testsuite/libffi.call/cls_8byte.c,
181 testsuite/libffi.call/cls_multi_sshort.c,
182 testsuite/libffi.call/cls_align_sint16.c,
183 testsuite/libffi.call/cls_align_uint32.c,
184 testsuite/libffi.call/cls_20byte.c,
185 testsuite/libffi.call/cls_float.c,
186 testsuite/libffi.call/nested_struct2.c,
187 testsuite/libffi.call/cls_5_1_byte.c,
188 testsuite/libffi.call/nested_struct4.c,
189 testsuite/libffi.call/cls_24byte.c,
190 testsuite/libffi.call/nested_struct6.c,
191 testsuite/libffi.call/cls_64byte.c,
192 testsuite/libffi.call/nested_struct8.c,
193 testsuite/libffi.call/cls_uint.c,
194 testsuite/libffi.call/cls_multi_ushortchar.c,
195 testsuite/libffi.call/cls_schar.c,
196 testsuite/libffi.call/cls_uchar.c,
197 testsuite/libffi.call/cls_align_uint64.c,
198 testsuite/libffi.call/cls_ulonglong.c,
199 testsuite/libffi.call/cls_align_longdouble.c,
200 testsuite/libffi.call/cls_1_1byte.c,
201 testsuite/libffi.call/cls_12byte.c,
202 testsuite/libffi.call/cls_3_1byte.c,
203 testsuite/libffi.call/cls_3byte1.c,
204 testsuite/libffi.call/cls_4_1byte.c,
205 testsuite/libffi.call/cls_6_1_byte.c,
206 testsuite/libffi.call/cls_16byte.c,
207 testsuite/libffi.call/cls_18byte.c,
208 testsuite/libffi.call/closure_fn0.c,
209 testsuite/libffi.call/cls_9byte2.c,
210 testsuite/libffi.call/closure_fn2.c,
211 testsuite/libffi.call/closure_fn4.c,
212 testsuite/libffi.call/cls_ushort.c,
213 testsuite/libffi.call/closure_fn6.c,
214 testsuite/libffi.call/cls_5byte.c,
215 testsuite/libffi.call/cls_align_pointer.c,
216 testsuite/libffi.call/cls_7byte.c,
217 testsuite/libffi.call/cls_align_sint32.c,
218 testsuite/libffi.special/unwindtest_ffi_call.cc,
219 testsuite/libffi.special/unwindtest.cc: Remove xfail for mips64*-*-*.
221 2007-08-10 David Daney <ddaney@avtrex.com>
224 * configure.ac: Don't treat mips64 as a special case.
225 * Makefile.am (nodist_libffi_la_SOURCES): Add n32.S.
226 * configure: Regenerate
227 * Makefile.in: Ditto.
228 * fficonfig.h.in: Ditto.
229 * src/mips/ffitarget.h (REG_L, REG_S, SUBU, ADDU, SRL, LI): Indent.
230 (LA, EH_FRAME_ALIGN, FDE_ADDR_BYTES): New preprocessor macros.
231 (FFI_DEFAULT_ABI): Set for n64 case.
232 (FFI_CLOSURES, FFI_TRAMPOLINE_SIZE): Define for n32 and n64 cases.
233 * src/mips/n32.S (ffi_call_N32): Add debug macros and labels for FDE.
234 (ffi_closure_N32): New function.
235 (.eh_frame): New section
236 * src/mips/o32.S: Clean up comments.
237 (ffi_closure_O32): Pass ffi_closure parameter in $12.
238 * src/mips/ffi.c: Use FFI_MIPS_N32 instead of
239 _MIPS_SIM == _ABIN32 throughout.
240 (FFI_MIPS_STOP_HERE): New, use in place of
242 (ffi_prep_args): Use unsigned long to hold pointer values. Rewrite
243 to support n32/n64 ABIs.
244 (calc_n32_struct_flags): Rewrite.
245 (calc_n32_return_struct_flags): Remove unused variable. Reverse
246 position of flag bits.
247 (ffi_prep_cif_machdep): Rewrite n32 portion.
248 (ffi_call): Enable for n64. Add special handling for small structure
250 (ffi_prep_closure_loc): Add n32 and n64 support.
251 (ffi_closure_mips_inner_O32): Add cast to silence warning.
252 (copy_struct_N32, ffi_closure_mips_inner_N32): New functions.
254 2007-08-08 David Daney <ddaney@avtrex.com>
256 * testsuite/libffi.call/ffitest.h (ffi_type_mylong): Remove definition.
257 * testsuite/libffi.call/cls_align_uint16.c (main): Use correct type
259 * testsuite/libffi.call/nested_struct1.c (main): Ditto.
260 * testsuite/libffi.call/cls_sint.c (main): Ditto.
261 * testsuite/libffi.call/nested_struct9.c (main): Ditto.
262 * testsuite/libffi.call/cls_20byte1.c (main): Ditto.
263 * testsuite/libffi.call/cls_9byte1.c (main): Ditto.
264 * testsuite/libffi.call/closure_fn1.c (main): Ditto.
265 * testsuite/libffi.call/closure_fn3.c (main): Ditto.
266 * testsuite/libffi.call/return_dbl2.c (main): Ditto.
267 * testsuite/libffi.call/cls_sshort.c (main): Ditto.
268 * testsuite/libffi.call/return_fl3.c (main): Ditto.
269 * testsuite/libffi.call/closure_fn5.c (main): Ditto.
270 * testsuite/libffi.call/nested_struct.c (main): Ditto.
271 * testsuite/libffi.call/nested_struct10.c (main): Ditto.
272 * testsuite/libffi.call/return_ll1.c (main): Ditto.
273 * testsuite/libffi.call/cls_8byte.c (main): Ditto.
274 * testsuite/libffi.call/cls_align_uint32.c (main): Ditto.
275 * testsuite/libffi.call/cls_align_sint16.c (main): Ditto.
276 * testsuite/libffi.call/cls_20byte.c (main): Ditto.
277 * testsuite/libffi.call/nested_struct2.c (main): Ditto.
278 * testsuite/libffi.call/cls_24byte.c (main): Ditto.
279 * testsuite/libffi.call/nested_struct6.c (main): Ditto.
280 * testsuite/libffi.call/cls_uint.c (main): Ditto.
281 * testsuite/libffi.call/cls_12byte.c (main): Ditto.
282 * testsuite/libffi.call/cls_16byte.c (main): Ditto.
283 * testsuite/libffi.call/closure_fn0.c (main): Ditto.
284 * testsuite/libffi.call/cls_9byte2.c (main): Ditto.
285 * testsuite/libffi.call/closure_fn2.c (main): Ditto.
286 * testsuite/libffi.call/return_dbl1.c (main): Ditto.
287 * testsuite/libffi.call/closure_fn4.c (main): Ditto.
288 * testsuite/libffi.call/closure_fn6.c (main): Ditto.
289 * testsuite/libffi.call/cls_align_sint32.c (main): Ditto.
291 2007-08-07 Andrew Haley <aph@redhat.com>
293 * src/x86/sysv.S (ffi_closure_raw_SYSV): Fix typo in previous
296 2007-08-06 Andrew Haley <aph@redhat.com>
299 * src/x86/sysv.S (ffi_closure_raw_SYSV): Handle FFI_TYPE_UINT8,
300 FFI_TYPE_SINT8, FFI_TYPE_UINT16, FFI_TYPE_SINT16, FFI_TYPE_UINT32,
303 2007-08-02 David Daney <ddaney@avtrex.com>
305 * testsuite/libffi.call/return_ul.c (main): Define return type as
306 ffi_arg. Use proper printf conversion specifier.
308 2007-07-30 Andrew Haley <aph@redhat.com>
311 * src/x86/ffi.c (ffi_prep_cif_machdep): in x86 case, add code for
312 signed/unsigned int8/16.
313 * src/x86/sysv.S (ffi_call_SYSV): Rewrite to:
315 Remove code to pop args from the stack after call.
316 Special-case signed/unsigned int8/16.
317 * testsuite/libffi.call/return_sc.c (main): Revert.
319 2007-07-26 Richard Guenther <rguenther@suse.de>
322 * testsuite/libffi.call/return_sc.c (main): Verify call
323 result as signed char, not ffi_arg.
325 2007-07-16 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
327 * configure.ac (i?86-*-solaris2.1[0-9]): Set TARGET to X86_64.
328 * configure: Regenerate.
330 2007-07-11 David Daney <ddaney@avtrex.com>
332 * src/mips/ffi.c: Don't include sys/cachectl.h.
333 (ffi_prep_closure_loc): Use __builtin___clear_cache() instead of
336 2007-05-18 Aurelien Jarno <aurelien@aurel32.net>
338 * src/arm/ffi.c (ffi_prep_closure_loc): Renamed and ajusted
339 from (ffi_prep_closure): ... this.
340 (FFI_INIT_TRAMPOLINE): Adjust.
342 2005-12-31 Phil Blundell <pb@reciva.com>
344 * src/arm/ffi.c (ffi_prep_incoming_args_SYSV,
345 ffi_closure_SYSV_inner, ffi_prep_closure): New, add closure support.
346 * src/arm/sysv.S(ffi_closure_SYSV): Likewise.
347 * src/arm/ffitarget.h (FFI_TRAMPOLINE_SIZE): Likewise.
348 (FFI_CLOSURES): Enable closure support.
350 2007-07-03 Andrew Haley <aph@hedges.billgatliff.com>
352 * testsuite/libffi.call/cls_multi_ushort.c,
353 testsuite/libffi.call/cls_align_uint16.c,
354 testsuite/libffi.call/nested_struct1.c,
355 testsuite/libffi.call/nested_struct3.c,
356 testsuite/libffi.call/cls_7_1_byte.c,
357 testsuite/libffi.call/cls_double.c,
358 testsuite/libffi.call/nested_struct5.c,
359 testsuite/libffi.call/nested_struct7.c,
360 testsuite/libffi.call/cls_sint.c,
361 testsuite/libffi.call/nested_struct9.c,
362 testsuite/libffi.call/cls_20byte1.c,
363 testsuite/libffi.call/cls_multi_sshortchar.c,
364 testsuite/libffi.call/cls_align_sint64.c,
365 testsuite/libffi.call/cls_3byte2.c,
366 testsuite/libffi.call/cls_multi_schar.c,
367 testsuite/libffi.call/cls_multi_uchar.c,
368 testsuite/libffi.call/cls_19byte.c,
369 testsuite/libffi.call/cls_9byte1.c,
370 testsuite/libffi.call/cls_align_float.c,
371 testsuite/libffi.call/closure_fn1.c,
372 testsuite/libffi.call/problem1.c,
373 testsuite/libffi.call/closure_fn3.c,
374 testsuite/libffi.call/cls_sshort.c,
375 testsuite/libffi.call/closure_fn5.c,
376 testsuite/libffi.call/cls_align_double.c,
377 testsuite/libffi.call/cls_2byte.c,
378 testsuite/libffi.call/nested_struct.c,
379 testsuite/libffi.call/nested_struct10.c,
380 testsuite/libffi.call/cls_4byte.c,
381 testsuite/libffi.call/cls_6byte.c,
382 testsuite/libffi.call/cls_8byte.c,
383 testsuite/libffi.call/cls_multi_sshort.c,
384 testsuite/libffi.call/cls_align_uint32.c,
385 testsuite/libffi.call/cls_align_sint16.c,
386 testsuite/libffi.call/cls_float.c,
387 testsuite/libffi.call/cls_20byte.c,
388 testsuite/libffi.call/cls_5_1_byte.c,
389 testsuite/libffi.call/nested_struct2.c,
390 testsuite/libffi.call/cls_24byte.c,
391 testsuite/libffi.call/nested_struct4.c,
392 testsuite/libffi.call/nested_struct6.c,
393 testsuite/libffi.call/cls_64byte.c,
394 testsuite/libffi.call/nested_struct8.c,
395 testsuite/libffi.call/cls_uint.c,
396 testsuite/libffi.call/cls_multi_ushortchar.c,
397 testsuite/libffi.call/cls_schar.c,
398 testsuite/libffi.call/cls_uchar.c,
399 testsuite/libffi.call/cls_align_uint64.c,
400 testsuite/libffi.call/cls_ulonglong.c,
401 testsuite/libffi.call/cls_align_longdouble.c,
402 testsuite/libffi.call/cls_1_1byte.c,
403 testsuite/libffi.call/cls_12byte.c,
404 testsuite/libffi.call/cls_3_1byte.c,
405 testsuite/libffi.call/cls_3byte1.c,
406 testsuite/libffi.call/cls_4_1byte.c,
407 testsuite/libffi.call/cls_6_1_byte.c,
408 testsuite/libffi.call/cls_16byte.c,
409 testsuite/libffi.call/cls_18byte.c,
410 testsuite/libffi.call/closure_fn0.c,
411 testsuite/libffi.call/cls_9byte2.c,
412 testsuite/libffi.call/closure_fn2.c,
413 testsuite/libffi.call/closure_fn4.c,
414 testsuite/libffi.call/cls_ushort.c,
415 testsuite/libffi.call/closure_fn6.c,
416 testsuite/libffi.call/cls_5byte.c,
417 testsuite/libffi.call/cls_align_pointer.c,
418 testsuite/libffi.call/cls_7byte.c,
419 testsuite/libffi.call/cls_align_sint32.c,
420 testsuite/libffi.special/unwindtest_ffi_call.cc,
421 testsuite/libffi.special/unwindtest.cc: Enable for ARM.
423 2007-07-05 H.J. Lu <hongjiu.lu@intel.com>
425 * aclocal.m4: Regenerated.
427 2007-06-02 Paolo Bonzini <bonzini@gnu.org>
429 * configure: Regenerate.
431 2007-05-23 Steve Ellcey <sje@cup.hp.com>
433 * Makefile.in: Regenerate.
434 * configure: Regenerate.
435 * aclocal.m4: Regenerate.
436 * include/Makefile.in: Regenerate.
437 * testsuite/Makefile.in: Regenerate.
439 2007-05-10 Roman Zippel <zippel@linux-m68k.org>
441 * src/m68k/ffi.c (ffi_prep_incoming_args_SYSV,
442 ffi_closure_SYSV_inner,ffi_prep_closure): New, add closure support.
443 * src/m68k/sysv.S(ffi_closure_SYSV,ffi_closure_struct_SYSV): Likewise.
444 * src/m68k/ffitarget.h (FFI_TRAMPOLINE_SIZE): Likewise.
445 (FFI_CLOSURES): Enable closure support.
447 2007-05-10 Roman Zippel <zippel@linux-m68k.org>
449 * configure.ac (HAVE_AS_CFI_PSEUDO_OP): New test.
450 * configure: Regenerate.
451 * fficonfig.h.in: Regenerate.
452 * src/m68k/sysv.S (CFI_STARTPROC,CFI_ENDPROC,
453 CFI_OFFSET,CFI_DEF_CFA): New macros.
454 (ffi_call_SYSV): Add callframe annotation.
456 2007-05-10 Roman Zippel <zippel@linux-m68k.org>
458 * src/m68k/ffi.c (ffi_prep_args,ffi_prep_cif_machdep): Fix
459 numerous test suite failures.
460 * src/m68k/sysv.S (ffi_call_SYSV): Likewise.
462 2007-04-11 Paolo Bonzini <bonzini@gnu.org>
464 * Makefile.am (EXTRA_DIST): Bring up to date.
465 * Makefile.in: Regenerate.
466 * src/frv/eabi.S: Remove RCS keyword.
468 2007-04-06 Richard Henderson <rth@redhat.com>
470 * configure.ac: Tidy target case.
471 (HAVE_LONG_DOUBLE): Allow the target to override.
472 * configure: Regenerate.
473 * include/ffi.h.in: Don't define ffi_type_foo if
474 LIBFFI_HIDE_BASIC_TYPES is defined.
475 (ffi_type_longdouble): If not HAVE_LONG_DOUBLE, define
477 * types.c (LIBFFI_HIDE_BASIC_TYPES): Define.
478 (FFI_TYPEDEF, ffi_type_void): Mark the data const.
479 (ffi_type_longdouble): Special case for Alpha. Don't define
480 if long double == double.
482 * src/alpha/ffi.c (FFI_TYPE_LONGDOUBLE): Assert unique value.
483 (ffi_prep_cif_machdep): Handle it as the 128-bit type.
484 (ffi_call, ffi_closure_osf_inner): Likewise.
485 (ffi_closure_osf_inner): Likewise. Mark hidden.
486 (ffi_call_osf, ffi_closure_osf): Mark hidden.
487 * src/alpha/ffitarget.h (FFI_LAST_ABI): Tidy definition.
488 * src/alpha/osf.S (ffi_call_osf, ffi_closure_osf): Mark hidden.
489 (load_table): Handle 128-bit long double.
491 * testsuite/libffi.call/float4.c: Add -mieee for alpha.
493 2007-04-06 Tom Tromey <tromey@redhat.com>
496 * README: Fixed bug in example.
498 2007-04-03 Jakub Jelinek <jakub@redhat.com>
500 * src/closures.c: Include sys/statfs.h.
501 (_GNU_SOURCE): Define on Linux.
502 (FFI_MMAP_EXEC_SELINUX): Define.
503 (selinux_enabled): New variable.
504 (selinux_enabled_check): New function.
505 (is_selinux_enabled): Define.
508 2007-03-24 Uros Bizjak <ubizjak@gmail.com>
510 * testsuite/libffi.call/return_fl2.c (return_fl): Mark as static.
511 Use 'volatile float sum' to create sum of floats to avoid false
512 negative due to excess precision on ix86 targets.
515 2007-03-08 Alexandre Oliva <aoliva@redhat.com>
517 * src/powerpc/ffi.c (flush_icache): Fix left-over from previous
519 (ffi_prep_closure_loc): Remove unneeded casts. Add needed ones.
521 2007-03-07 Alexandre Oliva <aoliva@redhat.com>
523 * include/ffi.h.in (ffi_closure_alloc, ffi_closure_free): New.
524 (ffi_prep_closure_loc): New.
525 (ffi_prep_raw_closure_loc): New.
526 (ffi_prep_java_raw_closure_loc): New.
527 * src/closures.c: New file.
528 * src/dlmalloc.c [FFI_MMAP_EXEC_WRIT] (struct malloc_segment):
529 Replace sflags with exec_offset.
530 [FFI_MMAP_EXEC_WRIT] (mmap_exec_offset, add_segment_exec_offset,
531 sub_segment_exec_offset): New macros.
532 (get_segment_flags, set_segment_flags, check_segment_merge): New
534 (is_mmapped_segment, is_extern_segment): Use get_segment_flags.
535 (add_segment, sys_alloc, create_mspace, create_mspace_with_base,
536 destroy_mspace): Use new macros.
537 (sys_alloc): Silence warning.
538 * Makefile.am (libffi_la_SOURCES): Add src/closures.c.
539 * Makefile.in: Rebuilt.
540 * src/prep_cif [FFI_CLOSURES] (ffi_prep_closure): Implement in
541 terms of ffi_prep_closure_loc.
542 * src/raw_api.c (ffi_prep_raw_closure_loc): Renamed and adjusted
544 (ffi_prep_raw_closure): ... this. Re-implement in terms of the
546 * src/java_raw_api (ffi_prep_java_raw_closure_loc): Renamed and
548 (ffi_prep_java_raw_closure): ... this. Re-implement in terms of
550 * src/alpha/ffi.c (ffi_prep_closure_loc): Renamed from
551 (ffi_prep_closure): ... this.
552 * src/pa/ffi.c: Likewise.
553 * src/cris/ffi.c: Likewise. Adjust.
554 * src/frv/ffi.c: Likewise.
555 * src/ia64/ffi.c: Likewise.
556 * src/mips/ffi.c: Likewise.
557 * src/powerpc/ffi_darwin.c: Likewise.
558 * src/s390/ffi.c: Likewise.
559 * src/sh/ffi.c: Likewise.
560 * src/sh64/ffi.c: Likewise.
561 * src/sparc/ffi.c: Likewise.
562 * src/x86/ffi64.c: Likewise.
563 * src/x86/ffi.c: Likewise.
564 (FFI_INIT_TRAMPOLINE): Adjust.
565 (ffi_prep_raw_closure_loc): Renamed and adjusted from...
566 (ffi_prep_raw_closure): ... this.
567 * src/powerpc/ffi.c (ffi_prep_closure_loc): Renamed from
568 (ffi_prep_closure): ... this.
569 (flush_icache): Adjust.
571 2007-03-07 Alexandre Oliva <aoliva@redhat.com>
573 * src/dlmalloc.c: New file, imported version 2.8.3 of Doug
576 2007-03-01 Brooks Moses <brooks.moses@codesourcery.com>
578 * Makefile.am: Add dummy install-pdf target.
579 * Makefile.in: Regenerate
581 2007-02-13 Andreas Krebbel <krebbel1@de.ibm.com>
583 * src/s390/ffi.c (ffi_prep_args, ffi_prep_cif_machdep,
584 ffi_closure_helper_SYSV): Add long double handling.
586 2007-02-02 Jakub Jelinek <jakub@redhat.com>
588 * src/powerpc/linux64.S (ffi_call_LINUX64): Move restore of r2
589 immediately after bctrl instruction.
591 2007-01-18 Alexandre Oliva <aoliva@redhat.com>
593 * Makefile.am (all-recursive, install-recursive,
594 mostlyclean-recursive, clean-recursive, distclean-recursive,
595 maintainer-clean-recursive): Add missing targets.
596 * Makefile.in: Rebuilt.
598 2006-12-14 Andreas Tobler <a.tobler@schweiz.org>
600 * configure.ac: Add TARGET for x86_64-*-darwin*.
601 * Makefile.am (nodist_libffi_la_SOURCES): Add rules for 64-bit sources
603 * src/x86/ffitarget.h: Set trampoline size for x86_64-*-darwin*.
604 * src/x86/darwin64.S: New file for x86_64-*-darwin* support.
605 * configure: Regenerate.
606 * Makefile.in: Regenerate.
607 * include/Makefile.in: Regenerate.
608 * testsuite/Makefile.in: Regenerate.
609 * testsuite/libffi.special/unwindtest_ffi_call.cc: New test case for
612 2006-12-13 Andreas Tobler <a.tobler@schweiz.org>
614 * aclocal.m4: Regenerate with aclocal -I .. as written in the
617 2006-10-31 Geoffrey Keating <geoffk@apple.com>
619 * src/powerpc/ffi_darwin.c (darwin_adjust_aggregate_sizes): New.
620 (ffi_prep_cif_machdep): Call darwin_adjust_aggregate_sizes for
622 * testsuite/libffi.call/nested_struct4.c: Remove Darwin XFAIL.
623 * testsuite/libffi.call/nested_struct6.c: Remove Darwin XFAIL.
625 2006-10-10 Paolo Bonzini <bonzini@gnu.org>
626 Sandro Tolaini <tolaini@libero.it>
628 * configure.ac [i*86-*-darwin*]: Set X86_DARWIN symbol and
630 * configure: Regenerated.
631 * Makefile.am (nodist_libffi_la_SOURCES) [X86_DARWIN]: New case.
632 (EXTRA_DIST): Add src/x86/darwin.S.
633 * Makefile.in: Regenerated.
634 * include/Makefile.in: Regenerated.
635 * testsuite/Makefile.in: Regenerated.
637 * src/x86/ffi.c (ffi_prep_cif_machdep) [X86_DARWIN]: Treat like
638 X86_WIN32, and additionally align stack to 16 bytes.
639 * src/x86/darwin.S: New, based on sysv.S.
640 * src/prep_cif.c (ffi_prep_cif) [X86_DARWIN]: Align > 8-byte structs.
642 2006-09-12 David Daney <ddaney@avtrex.com>
645 * include/Makefile.am: Install both ffi.h and ffitarget.h in
646 $(libdir)/gcc/$(target_alias)/$(gcc_version)/include.
647 * aclocal.m4: Regenerated for automake 1.9.6.
648 * Makefile.in: Regenerated.
649 * include/Makefile.in: Regenerated.
650 * testsuite/Makefile.in: Regenerated.
652 2006-08-17 Andreas Tobler <a.tobler@schweiz.ch>
654 * include/ffi_common.h (struct): Revert accidental commit.
656 2006-08-15 Andreas Tobler <a.tobler@schweiz.ch>
658 * include/ffi_common.h: Remove lint directives.
659 * include/ffi.h.in: Likewise.
661 2006-07-25 Torsten Schoenfeld <kaffeetisch@gmx.de>
663 * include/ffi.h.in (ffi_type_ulong, ffi_type_slong): Define correctly
664 for 32-bit architectures.
665 * testsuite/libffi.call/return_ul.c: New test case.
667 2006-07-19 David Daney <ddaney@avtrex.com>
669 * testsuite/libffi.call/closure_fn6.c: Remove xfail for mips,
670 xfail remains for mips64.
672 2006-05-23 Carlos O'Donell <carlos@codesourcery.com>
674 * Makefile.am: Add install-html target. Add install-html to .PHONY
675 * Makefile.in: Regenerate.
676 * aclocal.m4: Regenerate.
677 * include/Makefile.in: Regenerate.
678 * testsuite/Makefile.in: Regenerate.
680 2006-05-18 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
682 * pa/ffi.c (ffi_prep_args_pa32): Load floating point arguments from
685 2006-04-22 Andreas Tobler <a.tobler@schweiz.ch>
687 * README: Remove notice about 'Crazy Comments'.
688 * src/debug.c: Remove lint directives. Cleanup white spaces.
689 * src/java_raw_api.c: Likewise.
690 * src/prep_cif.c: Likewise.
691 * src/raw_api.c: Likewise.
692 * src/ffitest.c: Delete. No longer needed, all test cases migrated
694 * src/arm/ffi.c: Remove lint directives.
695 * src/m32r/ffi.c: Likewise.
696 * src/pa/ffi.c: Likewise.
697 * src/powerpc/ffi.c: Likewise.
698 * src/powerpc/ffi_darwin.c: Likewise.
699 * src/sh/ffi.c: Likewise.
700 * src/sh64/ffi.c: Likewise.
701 * src/x86/ffi.c: Likewise.
702 * testsuite/libffi.call/float2.c: Likewise.
703 * testsuite/libffi.call/promotion.c: Likewise.
704 * testsuite/libffi.call/struct1.c: Likewise.
706 2006-04-13 Andreas Tobler <a.tobler@schweiz.ch>
708 * src/pa/hpux32.S: Correct unwind offset calculation for
710 * src/pa/linux.S: Likewise.
712 2006-04-12 James E Wilson <wilson@specifix.com>
715 * src/ia64/ffi.c (stf_spill, ldf_fill): Rewrite as macros.
716 (hfa_type_load): Call stf_spill.
717 (hfa_type_store): Call ldf_fill.
718 (ffi_call): Adjust calls to above routines. Add local temps for
721 2006-04-10 Matthias Klose <doko@debian.org>
723 * testsuite/lib/libffi-dg.exp (libffi-init): Recognize multilib
724 directory names containing underscores.
726 2006-04-07 James E Wilson <wilson@specifix.com>
728 * testsuite/libffi.call/float4.c: New testcase.
730 2006-04-05 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
731 Andreas Tobler <a.tobler@schweiz.ch>
733 * Makefile.am: Add PA_HPUX port.
734 * Makefile.in: Regenerate.
735 * include/Makefile.in: Likewise.
736 * testsuite/Makefile.in: Likewise.
737 * configure.ac: Add PA_HPUX rules.
738 * configure: Regenerate.
739 * src/pa/ffitarget.h: Rename linux target to PA_LINUX.
740 Add PA_HPUX and PA64_HPUX.
741 Rename FFI_LINUX ABI to FFI_PA32 ABI.
742 (FFI_TRAMPOLINE_SIZE): Define for 32-bit HP-UX targets.
743 (FFI_TYPE_SMALL_STRUCT2): Define.
744 (FFI_TYPE_SMALL_STRUCT4): Likewise.
745 (FFI_TYPE_SMALL_STRUCT8): Likewise.
746 (FFI_TYPE_SMALL_STRUCT3): Redefine.
747 (FFI_TYPE_SMALL_STRUCT5): Likewise.
748 (FFI_TYPE_SMALL_STRUCT6): Likewise.
749 (FFI_TYPE_SMALL_STRUCT7): Likewise.
750 * src/pa/ffi.c (ROUND_DOWN): Delete.
751 (fldw, fstw, fldd, fstd): Use '__asm__'.
752 (ffi_struct_type): Add support for FFI_TYPE_SMALL_STRUCT2,
753 FFI_TYPE_SMALL_STRUCT4 and FFI_TYPE_SMALL_STRUCT8.
754 (ffi_prep_args_LINUX): Rename to ffi_prep_args_pa32. Update comment.
755 Simplify incrementing of stack slot variable. Change type of local
757 (ffi_size_stack_LINUX): Rename to ffi_size_stack_pa32. Handle long
759 (ffi_prep_cif_machdep): Likewise.
760 (ffi_call): Likewise.
761 (ffi_closure_inner_LINUX): Rename to ffi_closure_inner_pa32. Change
762 return type to ffi_status. Simplify incrementing of stack slot
763 variable. Only copy floating point argument registers when PA_LINUX
764 is true. Reformat debug statement.
765 Add support for FFI_TYPE_SMALL_STRUCT2, FFI_TYPE_SMALL_STRUCT4 and
766 FFI_TYPE_SMALL_STRUCT8.
767 (ffi_closure_LINUX): Rename to ffi_closure_pa32. Add 'extern' to
769 (ffi_prep_closure): Make linux trampoline conditional on PA_LINUX.
770 Add nops to cache flush. Add trampoline for PA_HPUX.
771 * src/pa/hpux32.S: New file.
772 * src/pa/linux.S (ffi_call_LINUX): Rename to ffi_call_pa32. Rename
773 ffi_prep_args_LINUX to ffi_prep_args_pa32.
774 Localize labels. Add support for 2, 4 and 8-byte small structs. Handle
775 unaligned destinations in 3, 5, 6 and 7-byte small structs. Order
776 argument type checks so that common argument types appear first.
777 (ffi_closure_LINUX): Rename to ffi_closure_pa32. Rename
778 ffi_closure_inner_LINUX to ffi_closure_inner_pa32.
780 2006-03-24 Alan Modra <amodra@bigpond.net.au>
782 * src/powerpc/ffitarget.h (enum ffi_abi): Add FFI_LINUX. Default
783 for 32-bit using IBM extended double format. Fix FFI_LAST_ABI.
784 * src/powerpc/ffi.c (ffi_prep_args_SYSV): Handle linux variant of
786 (ffi_prep_args64): Assert using IBM extended double.
787 (ffi_prep_cif_machdep): Don't munge FFI_TYPE_LONGDOUBLE type.
788 Handle FFI_LINUX FFI_TYPE_LONGDOUBLE return and args.
789 (ffi_call): Handle FFI_LINUX.
790 (ffi_closure_helper_SYSV): Non FFI_LINUX long double return needs
791 gpr3 return pointer as for struct return. Handle FFI_LINUX
792 FFI_TYPE_LONGDOUBLE return and args. Don't increment "nf"
794 * src/powerpc/ppc_closure.S (ffi_closure_SYSV): Load both f1 and f2
795 for FFI_TYPE_LONGDOUBLE. Move epilogue insns into case table.
796 Don't use r6 as pointer to results, instead use sp offset. Don't
797 make a special call to load lr with case table address, instead
798 use offset from previous call.
799 * src/powerpc/sysv.S (ffi_call_SYSV): Save long double return.
800 * src/powerpc/linux64.S (ffi_call_LINUX64): Simplify long double
803 2006-03-15 Kaz Kojima <kkojima@gcc.gnu.org>
805 * src/sh64/ffi.c (ffi_prep_cif_machdep): Handle float arguments
806 passed with FP registers correctly.
807 (ffi_closure_helper_SYSV): Likewise.
808 * src/sh64/sysv.S: Likewise.
810 2006-03-01 Andreas Tobler <a.tobler@schweiz.ch>
812 * testsuite/libffi.special/unwindtest.cc (closure_test_fn): Mark cif,
813 args and userdata unused.
814 (closure_test_fn1): Mark cif and userdata unused.
815 (main): Remove unused res.
817 2006-02-28 Andreas Tobler <a.tobler@schweiz.ch>
819 * testsuite/libffi.call/call.exp: Adjust FSF address. Add test runs for
820 -O2, -O3, -Os and the warning flags -W -Wall.
821 * testsuite/libffi.special/special.exp: Likewise.
822 * testsuite/libffi.call/ffitest.h: Add an __UNUSED__ macro to mark
823 unused parameter unused for gcc or else do nothing.
824 * testsuite/libffi.special/ffitestcxx.h: Likewise.
825 * testsuite/libffi.call/cls_12byte.c (cls_struct_12byte_gn): Mark cif
827 * testsuite/libffi.call/cls_16byte.c (cls_struct_16byte_gn): Likewise.
828 * testsuite/libffi.call/cls_18byte.c (cls_struct_18byte_gn): Likewise.
829 * testsuite/libffi.call/cls_19byte.c (cls_struct_19byte_gn): Likewise.
830 * testsuite/libffi.call/cls_1_1byte.c (cls_struct_1_1byte_gn): Likewise.
831 * testsuite/libffi.call/cls_20byte.c (cls_struct_20byte_gn): Likewise.
832 * testsuite/libffi.call/cls_20byte1.c (cls_struct_20byte_gn): Likewise.
833 * testsuite/libffi.call/cls_24byte.c (cls_struct_24byte_gn): Likewise.
834 * testsuite/libffi.call/cls_2byte.c (cls_struct_2byte_gn): Likewise.
835 * testsuite/libffi.call/cls_3_1byte.c (cls_struct_3_1byte_gn): Likewise.
836 * testsuite/libffi.call/cls_3byte1.c (cls_struct_3byte_gn): Likewise.
837 * testsuite/libffi.call/cls_3byte2.c (cls_struct_3byte_gn1): Likewise.
838 * testsuite/libffi.call/cls_4_1byte.c (cls_struct_4_1byte_gn): Likewise.
839 * testsuite/libffi.call/cls_4byte.c (cls_struct_4byte_gn): Likewise.
840 * testsuite/libffi.call/cls_5_1_byte.c (cls_struct_5byte_gn): Likewise.
841 * testsuite/libffi.call/cls_5byte.c (cls_struct_5byte_gn): Likewise.
842 * testsuite/libffi.call/cls_64byte.c (cls_struct_64byte_gn): Likewise.
843 * testsuite/libffi.call/cls_6_1_byte.c (cls_struct_6byte_gn): Likewise.
844 * testsuite/libffi.call/cls_6byte.c (cls_struct_6byte_gn): Likewise.
845 * testsuite/libffi.call/cls_7_1_byte.c (cls_struct_7byte_gn): Likewise.
846 * testsuite/libffi.call/cls_7byte.c (cls_struct_7byte_gn): Likewise.
847 * testsuite/libffi.call/cls_8byte.c (cls_struct_8byte_gn): Likewise.
848 * testsuite/libffi.call/cls_9byte1.c (cls_struct_9byte_gn): Likewise.
849 * testsuite/libffi.call/cls_9byte2.c (cls_struct_9byte_gn): Likewise.
850 * testsuite/libffi.call/cls_align_double.c (cls_struct_align_gn):
852 * testsuite/libffi.call/cls_align_float.c (cls_struct_align_gn):
854 * testsuite/libffi.call/cls_align_longdouble.c (cls_struct_align_gn):
856 * testsuite/libffi.call/cls_align_pointer.c (cls_struct_align_fn): Cast
857 void* to avoid compiler warning.
859 (cls_struct_align_gn): Mark cif and userdata unused.
860 * testsuite/libffi.call/cls_align_sint16.c (cls_struct_align_gn):
862 * testsuite/libffi.call/cls_align_sint32.c (cls_struct_align_gn):
864 * testsuite/libffi.call/cls_align_sint64.c (cls_struct_align_gn):
866 * testsuite/libffi.call/cls_align_uint16.c (cls_struct_align_gn):
868 * testsuite/libffi.call/cls_align_uint32.c (cls_struct_align_gn):
870 * testsuite/libffi.call/cls_double.c (cls_ret_double_fn): Likewise.
871 * testsuite/libffi.call/cls_float.c (cls_ret_float_fn): Likewise.
872 * testsuite/libffi.call/cls_multi_schar.c (test_func_gn): Mark cif and
874 (main): Cast res_call to silence gcc.
875 * testsuite/libffi.call/cls_multi_sshort.c (test_func_gn): Mark cif and
877 (main): Cast res_call to silence gcc.
878 * testsuite/libffi.call/cls_multi_sshortchar.c (test_func_gn): Mark cif
880 (main): Cast res_call to silence gcc.
881 * testsuite/libffi.call/cls_multi_uchar.c (test_func_gn): Mark cif and
883 (main): Cast res_call to silence gcc.
884 * testsuite/libffi.call/cls_multi_ushort.c (test_func_gn): Mark cif and
886 (main): Cast res_call to silence gcc.
887 * testsuite/libffi.call/cls_multi_ushortchar.c (test_func_gn): Mark cif
889 (main): Cast res_call to silence gcc.
890 * testsuite/libffi.call/cls_schar.c (cls_ret_schar_fn): Mark cif and
892 (cls_ret_schar_fn): Cast printf parameter to silence gcc.
893 * testsuite/libffi.call/cls_sint.c (cls_ret_sint_fn): Mark cif and
895 (cls_ret_sint_fn): Cast printf parameter to silence gcc.
896 * testsuite/libffi.call/cls_sshort.c (cls_ret_sshort_fn): Mark cif and
898 (cls_ret_sshort_fn): Cast printf parameter to silence gcc.
899 * testsuite/libffi.call/cls_uchar.c (cls_ret_uchar_fn): Mark cif and
901 (cls_ret_uchar_fn): Cast printf parameter to silence gcc.
902 * testsuite/libffi.call/cls_uint.c (cls_ret_uint_fn): Mark cif and
904 (cls_ret_uint_fn): Cast printf parameter to silence gcc.
905 * testsuite/libffi.call/cls_ulonglong.c (cls_ret_ulonglong_fn): Mark cif
907 * testsuite/libffi.call/cls_ushort.c (cls_ret_ushort_fn): Mark cif and
909 (cls_ret_ushort_fn): Cast printf parameter to silence gcc.
910 * testsuite/libffi.call/float.c (floating): Remove unused parameter e.
911 * testsuite/libffi.call/float1.c (main): Remove unused variable i.
912 Cleanup white spaces.
913 * testsuite/libffi.call/negint.c (checking): Remove unused variable i.
914 * testsuite/libffi.call/nested_struct.c (cls_struct_combined_gn): Mark
915 cif and userdata unused.
916 * testsuite/libffi.call/nested_struct1.c (cls_struct_combined_gn):
918 * testsuite/libffi.call/nested_struct10.c (B_gn): Likewise.
919 * testsuite/libffi.call/nested_struct2.c (B_fn): Adjust printf
920 formatters to silence gcc.
921 (B_gn): Mark cif and userdata unused.
922 * testsuite/libffi.call/nested_struct3.c (B_gn): Mark cif and userdata
924 * testsuite/libffi.call/nested_struct4.c: Mention related PR.
925 (B_gn): Mark cif and userdata unused.
926 * testsuite/libffi.call/nested_struct5.c (B_gn): Mark cif and userdata
928 * testsuite/libffi.call/nested_struct6.c: Mention related PR.
929 (B_gn): Mark cif and userdata unused.
930 * testsuite/libffi.call/nested_struct7.c (B_gn): Mark cif and userdata
932 * testsuite/libffi.call/nested_struct8.c (B_gn): Likewise.
933 * testsuite/libffi.call/nested_struct9.c (B_gn): Likewise.
934 * testsuite/libffi.call/problem1.c (stub): Likewise.
935 * testsuite/libffi.call/pyobjc-tc.c (main): Cast the result to silence
937 * testsuite/libffi.call/return_fl2.c (return_fl): Add the note mentioned
938 in the last commit for this test case in the test case itself.
939 * testsuite/libffi.call/closure_fn0.c (closure_test_fn0): Mark cif as
941 * testsuite/libffi.call/closure_fn1.c (closure_test_fn1): Likewise.
942 * testsuite/libffi.call/closure_fn2.c (closure_test_fn2): Likewise.
943 * testsuite/libffi.call/closure_fn3.c (closure_test_fn3): Likewise.
944 * testsuite/libffi.call/closure_fn4.c (closure_test_fn0): Likewise.
945 * testsuite/libffi.call/closure_fn5.c (closure_test_fn5): Likewise.
946 * testsuite/libffi.call/closure_fn6.c (closure_test_fn0): Likewise.
948 2006-02-22 Kaz Kojima <kkojima@gcc.gnu.org>
950 * src/sh/sysv.S: Fix register numbers in the FDE for
953 2006-02-20 Andreas Tobler <a.tobler@schweiz.ch>
955 * testsuite/libffi.call/return_fl2.c (return_fl): Remove static
956 declaration to avoid a false negative on ix86. See PR323.
958 2006-02-18 Kaz Kojima <kkojima@gcc.gnu.org>
960 * src/sh/ffi.c (ffi_closure_helper_SYSV): Remove unused variable
961 and cast integer to void * if needed. Update the pointer to
962 the FP register saved area correctly.
964 2006-02-17 Andreas Tobler <a.tobler@schweiz.ch>
966 * testsuite/libffi.call/nested_struct6.c: XFAIL this test until PR25630
968 * testsuite/libffi.call/nested_struct4.c: Likewise.
970 2006-02-16 Andreas Tobler <a.tobler@schweiz.ch>
972 * testsuite/libffi.call/return_dbl.c: New test case.
973 * testsuite/libffi.call/return_dbl1.c: Likewise.
974 * testsuite/libffi.call/return_dbl2.c: Likewise.
975 * testsuite/libffi.call/return_fl.c: Likewise.
976 * testsuite/libffi.call/return_fl1.c: Likewise.
977 * testsuite/libffi.call/return_fl2.c: Likewise.
978 * testsuite/libffi.call/return_fl3.c: Likewise.
979 * testsuite/libffi.call/closure_fn6.c: Likewise.
981 * testsuite/libffi.call/nested_struct2.c: Remove ffi_type_mylong
983 * testsuite/libffi.call/ffitest.h: Add ffi_type_mylong definition
984 here to be used by other test cases too.
986 * testsuite/libffi.call/nested_struct10.c: New test case.
987 * testsuite/libffi.call/nested_struct9.c: Likewise.
988 * testsuite/libffi.call/nested_struct8.c: Likewise.
989 * testsuite/libffi.call/nested_struct7.c: Likewise.
990 * testsuite/libffi.call/nested_struct6.c: Likewise.
991 * testsuite/libffi.call/nested_struct5.c: Likewise.
992 * testsuite/libffi.call/nested_struct4.c: Likewise.
994 2006-01-21 Andreas Tobler <a.tobler@schweiz.ch>
996 * configure.ac: Enable libffi for sparc64-*-freebsd*.
997 * configure: Rebuilt.
999 2006-01-18 Jakub Jelinek <jakub@redhat.com>
1001 * src/powerpc/sysv.S (smst_two_register): Don't call __ashldi3,
1002 instead do the shifting inline.
1003 * src/powerpc/ppc_closure.S (ffi_closure_SYSV): Don't compute %r5
1004 shift count unconditionally. Simplify load sequences for 1, 2, 3, 4
1005 and 8 byte structs, for the remaining struct sizes don't call
1006 __lshrdi3, instead do the shifting inline.
1008 2005-12-07 Thiemo Seufer <ths@networkno.de>
1010 * src/mips/ffitarget.h: Remove obsolete sgidefs.h include. Add
1011 missing parentheses.
1012 * src/mips/o32.S (ffi_call_O32): Code formatting. Define
1013 and use A3_OFF, FP_OFF, RA_OFF. Micro-optimizations.
1014 (ffi_closure_O32): Likewise, but with newly defined A3_OFF2,
1015 A2_OFF2, A1_OFF2, A0_OFF2, RA_OFF2, FP_OFF2, S0_OFF2, GP_OFF2,
1016 V1_OFF2, V0_OFF2, FA_1_1_OFF2, FA_1_0_OFF2, FA_0_1_OFF2,
1018 * src/mips/ffi.c (ffi_prep_args): Code formatting. Fix
1020 (ffi_prep_closure): Improve trampoline instruction scheduling.
1021 (ffi_closure_mips_inner_O32): Fix endianness bugs.
1023 2005-12-03 Alan Modra <amodra@bigpond.net.au>
1025 * src/powerpc/ffi.c: Formatting.
1026 (ffi_prep_args_SYSV): Avoid possible aliasing problems by using unions.
1027 (ffi_prep_args64): Likewise.
1029 2005-09-30 Geoffrey Keating <geoffk@apple.com>
1031 * testsuite/lib/libffi-dg.exp (libffi_target_compile): For
1032 darwin, use -shared-libgcc not -lgcc_s, and explain why.
1034 2005-09-26 Tom Tromey <tromey@redhat.com>
1036 * testsuite/libffi.call/float1.c (value_type): New typedef.
1037 (CANARY): New define.
1038 (main): Check for result buffer overflow.
1039 * src/powerpc/linux64.S: Handle linux64 long double returns.
1040 * src/powerpc/ffi.c (FLAG_RETURNS_128BITS): New constant.
1041 (ffi_prep_cif_machdep): Handle linux64 long double returns.
1043 2005-08-25 Alan Modra <amodra@bigpond.net.au>
1046 * src/powerpc/ffi.c (ffi_prep_args_SYSV): Correct placement of stack
1048 (ffi_status ffi_prep_cif_machdep): Correct stack sizing for same.
1050 2005-08-11 Jakub Jelinek <jakub@redhat.com>
1052 * configure.ac (HAVE_HIDDEN_VISIBILITY_ATTRIBUTE): New test.
1053 (AH_BOTTOM): Add FFI_HIDDEN definition.
1054 * configure: Rebuilt.
1055 * fficonfig.h.in: Rebuilt.
1056 * src/powerpc/ffi.c (hidden): Remove.
1057 (ffi_closure_LINUX64, ffi_prep_args64, ffi_call_LINUX64,
1058 ffi_closure_helper_LINUX64): Use FFI_HIDDEN instead of hidden.
1059 * src/powerpc/linux64_closure.S (ffi_closure_LINUX64,
1060 .ffi_closure_LINUX64): Use FFI_HIDDEN instead of .hidden.
1061 * src/x86/ffi.c (ffi_closure_SYSV, ffi_closure_raw_SYSV): Remove,
1062 add FFI_HIDDEN to its prototype.
1063 (ffi_closure_SYSV_inner): New.
1064 * src/x86/sysv.S (ffi_closure_SYSV, ffi_closure_raw_SYSV): New.
1065 * src/x86/win32.S (ffi_closure_SYSV, ffi_closure_raw_SYSV): New.
1067 2005-08-10 Alfred M. Szmidt <ams@gnu.org>
1070 * configure: Rebuilt.
1071 * configure.ac: Handle i*86-*-gnu*.
1073 2005-08-09 Jakub Jelinek <jakub@redhat.com>
1075 * src/powerpc/ppc_closure.S (ffi_closure_SYSV): Use
1076 DW_CFA_offset_extended_sf rather than
1077 DW_CFA_GNU_negative_offset_extended.
1078 * src/powerpc/sysv.S (ffi_call_SYSV): Likewise.
1080 2005-07-22 SUGIOKA Toshinobu <sugioka@itonet.co.jp>
1082 * src/sh/sysv.S (ffi_call_SYSV): Stop argument popping correctly
1084 (ffi_closure_SYSV): Change the stack layout for sh3 struct argument.
1085 * src/sh/ffi.c (ffi_prep_args): Fix sh3 argument copy, when it is
1086 partially on register.
1087 (ffi_closure_helper_SYSV): Likewise.
1088 (ffi_prep_cif_machdep): Don't set too many cif->flags.
1090 2005-07-20 Kaz Kojima <kkojima@gcc.gnu.org>
1092 * src/sh/ffi.c (ffi_call): Handle small structures correctly.
1094 * src/sh64/ffi.c (simple_type): Remove.
1095 (return_type): Handle small structures correctly.
1096 (ffi_prep_args): Likewise.
1097 (ffi_call): Likewise.
1098 (ffi_closure_helper_SYSV): Likewise.
1099 * src/sh64/sysv.S (ffi_call_SYSV): Handle 1, 2 and 4-byte return.
1100 Emit position independent code if PIC and remove wrong datalabel
1101 prefixes from EH data.
1103 2005-07-19 Andreas Tobler <a.tobler@schweiz.ch>
1105 * Makefile.am (nodist_libffi_la_SOURCES): Add POWERPC_FREEBSD.
1106 * Makefile.in: Regenerate.
1107 * include/Makefile.in: Likewise.
1108 * testsuite/Makefile.in: Likewise.
1109 * configure.ac: Add POWERPC_FREEBSD rules.
1110 * configure: Regenerate.
1111 * src/powerpc/ffitarget.h: Add POWERPC_FREEBSD rules.
1112 (FFI_SYSV_TYPE_SMALL_STRUCT): Define.
1113 * src/powerpc/ffi.c: Add flags to handle small structure returns
1115 (ffi_prep_cif_machdep): Handle small structures for SYSV 4 ABI.
1117 (ffi_closure_helper_SYSV): Likewise.
1118 * src/powerpc/ppc_closure.S: Add return types for small structures.
1119 * src/powerpc/sysv.S: Add bits to handle small structures for
1122 2005-07-10 Andreas Tobler <a.tobler@schweiz.ch>
1124 * testsuite/libffi.call/cls_5_1_byte.c: New test file.
1125 * testsuite/libffi.call/cls_6_1_byte.c: Likewise.
1126 * testsuite/libffi.call/cls_7_1_byte.c: Likewise.
1128 2005-07-05 Randolph Chung <tausq@debian.org>
1130 * src/pa/ffi.c (ffi_struct_type): Rename FFI_TYPE_SMALL_STRUCT1
1131 as FFI_TYPE_SMALL_STRUCT3. Break out handling for 5-7 byte
1132 structures. Kill compilation warnings.
1133 (ffi_closure_inner_LINUX): Print return values as hex in debug
1134 message. Rename FFI_TYPE_SMALL_STRUCT1 as FFI_TYPE_SMALL_STRUCT3.
1135 Properly handle 5-7 byte structure returns.
1136 * src/pa/ffitarget.h (FFI_TYPE_SMALL_STRUCT1)
1137 (FFI_TYPE_SMALL_STRUCT2): Remove.
1138 (FFI_TYPE_SMALL_STRUCT3, FFI_TYPE_SMALL_STRUCT5)
1139 (FFI_TYPE_SMALL_STRUCT6, FFI_TYPE_SMALL_STRUCT7): Define.
1140 * src/pa/linux.S: Mark source file as using PA1.1 assembly.
1141 (checksmst1, checksmst2): Remove.
1142 (checksmst3): Optimize handling of 3-byte struct returns.
1143 (checksmst567): Properly handle 5-7 byte struct returns.
1145 2005-06-15 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
1148 * src/mips/n32.S: Enforce PIC code.
1149 * src/mips/o32.S: Likewise.
1151 2005-06-15 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
1153 * configure.ac: Treat i*86-*-solaris2.10 and up as X86_64.
1154 * configure: Regenerate.
1156 2005-06-01 Alan Modra <amodra@bigpond.net.au>
1158 * src/powerpc/ppc_closure.S (ffi_closure_SYSV): Don't use JUMPTARGET
1159 to call ffi_closure_helper_SYSV. Append @local instead.
1160 * src/powerpc/sysv.S (ffi_call_SYSV): Likewise for ffi_prep_args_SYSV.
1162 2005-05-17 Kelley Cook <kcook@gcc.gnu.org>
1164 * configure.ac: Use AC_C_BIGENDIAN instead of AC_C_BIGENDIAN_CROSS.
1165 Use AC_CHECK_SIZEOF instead of AC_COMPILE_CHECK_SIZEOF.
1166 * Makefile.am (ACLOCAL_AMFLAGS): Remove -I ../config.
1167 * aclocal.m4, configure, fficonfig.h.in, Makefile.in,
1168 include/Makefile.in, testsuite/Makefile.in: Regenerate.
1170 2005-05-09 Mike Stump <mrs@apple.com>
1172 * configure: Regenerate.
1174 2005-05-08 Richard Henderson <rth@redhat.com>
1177 * src/alpha/osf.S: Update unwind into to match code.
1179 2005-05-04 Andreas Degert <ad@papyrus-gmbh.de>
1180 Richard Henderson <rth@redhat.com>
1182 * src/x86/ffi64.c (ffi_prep_cif_machdep): Save sse-used flag in
1184 (ffi_call): Mask return type field. Pass ssecount to ffi_call_unix64.
1185 (ffi_prep_closure): Set carry bit if sse-used flag set.
1186 * src/x86/unix64.S (ffi_call_unix64): Add ssecount argument.
1187 Only load sse registers if ssecount non-zero.
1188 (ffi_closure_unix64): Only save sse registers if carry set on entry.
1190 2005-04-29 Ralf Corsepius <ralf.corsepius@rtems.org>
1192 * configure.ac: Add i*86-*-rtems*, sparc*-*-rtems*,
1193 powerpc-*rtems*, arm*-*-rtems*, sh-*-rtems*.
1194 * configure: Regenerate.
1196 2005-04-20 Hans-Peter Nilsson <hp@axis.com>
1198 * testsuite/lib/libffi-dg.exp (libffi-dg-test-1): In regsub use,
1199 have Tcl8.3-compatible intermediate variable.
1201 2005-04-18 Simon Posnjak <simon.posnjak@siol.net>
1202 Hans-Peter Nilsson <hp@axis.com>
1204 * Makefile.am: Add CRIS support.
1205 * configure.ac: Likewise.
1206 * Makefile.in, configure, testsuite/Makefile.in,
1207 include/Makefile.in: Regenerate.
1208 * src/cris: New directory.
1209 * src/cris/ffi.c, src/cris/sysv.S, src/cris/ffitarget.h: New files.
1210 * src/prep_cif.c (ffi_prep_cif): Wrap in #ifndef __CRIS__.
1212 * testsuite/lib/libffi-dg.exp (libffi-dg-test-1): Replace \n with
1213 \r?\n in output tests.
1215 2005-04-12 Mike Stump <mrs@apple.com>
1217 * configure: Regenerate.
1219 2005-03-30 Hans Boehm <Hans.Boehm@hp.com>
1221 * src/ia64/ffitarget.h (ffi_arg): Use long long instead of DI.
1223 2005-03-30 Steve Ellcey <sje@cup.hp.com>
1225 * src/ia64/ffitarget.h (ffi_arg) ADD DI attribute.
1227 * src/ia64/unix.S (ffi_closure_unix): Extend gp
1228 to 64 bits in ILP32 mode.
1229 Load 64 bits even for short data.
1231 2005-03-23 Mike Stump <mrs@apple.com>
1233 * src/powerpc/darwin.S: Update for -m64 multilib.
1234 * src/powerpc/darwin_closure.S: Likewise.
1236 2005-03-21 Zack Weinberg <zack@codesourcery.com>
1238 * configure.ac: Do not invoke TL_AC_GCC_VERSION.
1239 Do not set tool_include_dir.
1240 * aclocal.m4, configure, Makefile.in, testsuite/Makefile.in:
1242 * include/Makefile.am: Set gcc_version and toollibffidir.
1243 * include/Makefile.in: Regenerate.
1245 2005-02-22 Andrew Haley <aph@redhat.com>
1247 * src/powerpc/ffi.c (ffi_prep_cif_machdep): Bump alignment to
1248 odd-numbered register pairs for 64-bit integer types.
1250 2005-02-23 Andreas Tobler <a.tobler@schweiz.ch>
1253 * testsuite/libffi.call/return_ll1.c: New test case.
1255 2005-02-11 Janis Johnson <janis187@us.ibm.com>
1257 * testsuite/libffi.call/cls_align_longdouble.c: Remove dg-options.
1258 * testsuite/libffi.call/float.c: Ditto.
1259 * testsuite/libffi.call/float2.c: Ditto.
1260 * testsuite/libffi.call/float3.c: Ditto.
1262 2005-02-08 Andreas Tobler <a.tobler@schweiz.ch>
1264 * src/frv/ffitarget.h: Remove PPC stuff which does not belong to frv.
1266 2005-01-12 Eric Botcazou <ebotcazou@libertysurf.fr>
1268 * testsuite/libffi.special/special.exp (cxx_options): Add
1271 2004-12-31 Richard Henderson <rth@redhat.com>
1273 * src/types.c (FFI_AGGREGATE_TYPEDEF): Remove.
1274 (FFI_TYPEDEF): Rename from FFI_INTEGRAL_TYPEDEF. Replace size and
1275 offset parameters with a type parameter; deduce size and structure
1276 alignment. Update all users.
1278 2004-12-31 Richard Henderson <rth@redhat.com>
1280 * src/types.c (FFI_TYPE_POINTER): Define with sizeof.
1281 (FFI_TYPE_LONGDOUBLE): Fix for ia64.
1282 * src/ia64/ffitarget.h (struct ffi_ia64_trampoline_struct): Move
1283 into ffi_prep_closure.
1284 * src/ia64/ia64_flags.h, src/ia64/ffi.c, src/ia64/unix.S: Rewrite
1287 2004-12-27 Richard Henderson <rth@redhat.com>
1289 * src/x86/unix64.S: Fix typo in unwind info.
1291 2004-12-25 Richard Henderson <rth@redhat.com>
1293 * src/x86/ffi64.c (struct register_args): Rename from stackLayout.
1294 (enum x86_64_reg_class): Add X86_64_COMPLEX_X87_CLASS.
1295 (merge_classes): Check for it.
1297 (classify_argument): Pass byte_offset by value; perform all updates
1299 (examine_argument): Add classes argument; handle
1300 X86_64_COMPLEX_X87_CLASS.
1301 (ffi_prep_args): Merge into ...
1302 (ffi_call): ... here. Share stack frame with ffi_call_unix64.
1303 (ffi_prep_cif_machdep): Setup cif->flags for proper structure return.
1304 (ffi_fill_return_value): Remove.
1305 (ffi_prep_closure): Remove dead assert.
1306 (ffi_closure_unix64_inner): Rename from ffi_closure_UNIX64_inner.
1307 Rewrite to use struct register_args instead of va_list. Create
1308 flags for handling structure returns.
1309 * src/x86/unix64.S: Remove dead strings.
1310 (ffi_call_unix64): Rename from ffi_call_UNIX64. Rewrite to share
1311 stack frame with ffi_call. Handle structure returns properly.
1312 (float2sse, floatfloat2sse, double2sse): Remove.
1313 (sse2float, sse2double, sse2floatfloat): Remove.
1314 (ffi_closure_unix64): Rename from ffi_closure_UNIX64. Rewrite
1315 to handle structure returns properly.
1317 2004-12-08 David Edelsohn <edelsohn@gnu.org>
1319 * Makefile.am (AM_MAKEFLAGS): Remove duplicate LIBCFLAGS and
1321 * Makefile.in: Regenerated.
1323 2004-12-02 Richard Sandiford <rsandifo@redhat.com>
1325 * configure.ac: Use TL_AC_GCC_VERSION to set gcc_version.
1326 * configure, aclocal.m4, Makefile.in: Regenerate.
1327 * include/Makefile.in, testsuite/Makefile.in: Regenerate.
1329 2004-11-29 Kelley Cook <kcook@gcc.gnu.org>
1331 * configure: Regenerate for libtool change.
1333 2004-11-25 Kelley Cook <kcook@gcc.gnu.org>
1335 * configure: Regenerate for libtool reversion.
1337 2004-11-24 Kelley Cook <kcook@gcc.gnu.org>
1339 * configure: Regenerate for libtool change.
1341 2004-11-23 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
1343 * testsuite/lib/libffi-dg.exp: Use new procs in target-libpath.exp.
1345 2004-11-23 Richard Sandiford <rsandifo@redhat.com>
1347 * src/mips/o32.S (ffi_call_O32, ffi_closure_O32): Use jalr instead
1348 of jal. Use an absolute encoding for the frame information.
1350 2004-11-23 Kelley Cook <kcook@gcc.gnu.org>
1352 * Makefile.am: Remove no-dependencies. Add ACLOCAL_AMFLAGS.
1353 * acinclude.m4: Delete logic for sincludes.
1354 * aclocal.m4, Makefile.in, configure: Regenerate.
1355 * include/Makefile: Likewise.
1356 * testsuite/Makefile: Likewise.
1358 2004-11-22 Eric Botcazou <ebotcazou@libertysurf.fr>
1360 * src/sparc/ffi.c (ffi_prep_closure): Align doubles and 64-bit integers
1361 on a 8-byte boundary.
1362 * src/sparc/v8.S (ffi_closure_v8): Reserve frame space for arguments.
1364 2004-10-27 Richard Earnshaw <rearnsha@arm.com>
1366 * src/arm/ffi.c (ffi_prep_cif_machdep): Handle functions that return
1367 long long values. Round stack allocation to a multiple of 8 bytes
1368 for ATPCS compatibility.
1369 * src/arm/sysv.S (ffi_call_SYSV): Rework to avoid use of APCS register
1370 names. Handle returning long long types. Add Thumb and interworking
1371 support. Improve soft-float code.
1373 2004-10-27 Richard Earnshaw <rearnsha@arm.com>
1375 * testsuite/lib/libffi-db.exp (load_gcc_lib): New function.
1376 (libffi_exit): New function.
1377 (libffi_init): Build the testglue wrapper if needed.
1379 2004-10-25 Eric Botcazou <ebotcazou@libertysurf.fr>
1382 * testsuite/lib/libffi-dg.exp: Accept more than one multilib libgcc.
1384 2004-10-25 Kazuhiro Inaoka <inaoka.kazuhiro@renesas.com>
1386 * src/m32r/libffitarget.h (FFI_CLOSURES): Set to 0.
1388 2004-10-20 Kaz Kojima <kkojima@gcc.gnu.org>
1390 * src/sh/sysv.S (ffi_call_SYSV): Don't align for double data.
1391 * testsuite/libffi.call/float3.c: New test case.
1393 2004-10-18 Kaz Kojima <kkojima@gcc.gnu.org>
1395 * src/sh/ffi.c (ffi_prep_closure): Set T bit in trampoline for
1396 the function returning a structure pointed with R2.
1397 * src/sh/sysv.S (ffi_closure_SYSV): Use R2 as the pointer to
1398 the structure return value if T bit set. Emit position
1399 independent code and EH data if PIC.
1401 2004-10-13 Kazuhiro Inaoka <inaoka.kazuhiro@renesas.com>
1403 * Makefile.am: Add m32r support.
1404 * configure.ac: Likewise.
1405 * Makefile.in: Regenerate.
1406 * confiugre: Regenerate.
1407 * src/types.c: Add m32r port to FFI_INTERNAL_TYPEDEF
1408 (uint64, sint64, double, longdouble)
1409 * src/m32r: New directory.
1410 * src/m32r/ffi.c: New file.
1411 * src/m32r/sysv.S: Likewise.
1412 * src/m32r/ffitarget.h: Likewise.
1414 2004-10-02 Kaz Kojima <kkojima@gcc.gnu.org>
1416 * testsuite/libffi.call/negint.c: New test case.
1418 2004-09-14 H.J. Lu <hongjiu.lu@intel.com>
1421 * testsuite/lib/libffi-dg.exp: Don't use global ld_library_path.
1422 Set up LD_LIBRARY_PATH, SHLIB_PATH, LD_LIBRARYN32_PATH,
1423 LD_LIBRARY64_PATH, LD_LIBRARY_PATH_32, LD_LIBRARY_PATH_64 and
1426 2004-09-05 Andreas Tobler <a.tobler@schweiz.ch>
1428 * testsuite/libffi.call/many_win32.c: Remove whitespaces.
1429 * testsuite/libffi.call/promotion.c: Likewise.
1430 * testsuite/libffi.call/return_ll.c: Remove unused var. Cleanup
1432 * testsuite/libffi.call/return_sc.c: Likewise.
1433 * testsuite/libffi.call/return_uc.c: Likewise.
1435 2004-09-05 Andreas Tobler <a.tobler@schweiz.ch>
1437 * src/powerpc/darwin.S: Fix comments and identation.
1438 * src/powerpc/darwin_closure.S: Likewise.
1440 2004-09-02 Andreas Tobler <a.tobler@schweiz.ch>
1442 * src/powerpc/ffi_darwin.c: Add flag for longdouble return values.
1443 (ffi_prep_args): Handle longdouble arguments.
1444 (ffi_prep_cif_machdep): Set flags for longdouble. Calculate space for
1446 (ffi_closure_helper_DARWIN): Add closure handling for longdouble.
1447 * src/powerpc/darwin.S (_ffi_call_DARWIN): Add handling of longdouble
1449 * src/powerpc/darwin_closure.S (_ffi_closure_ASM): Likewise.
1450 * src/types.c: Defined longdouble size and alignment for darwin.
1452 2004-09-02 Andreas Tobler <a.tobler@schweiz.ch>
1454 * src/powerpc/aix.S: Remove whitespaces.
1455 * src/powerpc/aix_closure.S: Likewise.
1456 * src/powerpc/asm.h: Likewise.
1457 * src/powerpc/ffi.c: Likewise.
1458 * src/powerpc/ffitarget.h: Likewise.
1459 * src/powerpc/linux64.S: Likewise.
1460 * src/powerpc/linux64_closure.S: Likewise.
1461 * src/powerpc/ppc_closure.S: Likewise.
1462 * src/powerpc/sysv.S: Likewise.
1464 2004-08-30 Anthony Green <green@redhat.com>
1466 * Makefile.am: Add frv support.
1467 * Makefile.in, testsuite/Makefile.in: Rebuilt.
1468 * configure.ac: Read configure.host.
1469 * configure.in: Read configure.host.
1470 * configure.host: New file. frv-elf needs libgloss.
1471 * include/ffi.h.in: Force ffi_closure to have a nice big (8)
1472 alignment. This is needed to frv and shouldn't harm the others.
1473 * include/ffi_common.h (ALIGN_DOWN): New macro.
1474 * src/frv/ffi.c, src/frv/ffitarget.h, src/frv/eabi.S: New files.
1476 2004-08-24 David Daney <daney@avtrex.com>
1478 * testsuite/libffi.call/closure_fn0.c: Xfail mips64* instead of mips*.
1479 * testsuite/libffi.call/closure_fn1.c: Likewise.
1480 * testsuite/libffi.call/closure_fn2.c Likewise.
1481 * testsuite/libffi.call/closure_fn3.c: Likewise.
1482 * testsuite/libffi.call/closure_fn4.c: Likewise.
1483 * testsuite/libffi.call/closure_fn5.c: Likewise.
1484 * testsuite/libffi.call/cls_18byte.c: Likewise.
1485 * testsuite/libffi.call/cls_19byte.c: Likewise.
1486 * testsuite/libffi.call/cls_1_1byte.c: Likewise.
1487 * testsuite/libffi.call/cls_20byte.c: Likewise.
1488 * testsuite/libffi.call/cls_20byte1.c: Likewise.
1489 * testsuite/libffi.call/cls_24byte.c: Likewise.
1490 * testsuite/libffi.call/cls_2byte.c: Likewise.
1491 * testsuite/libffi.call/cls_3_1byte.c: Likewise.
1492 * testsuite/libffi.call/cls_3byte1.c: Likewise.
1493 * testsuite/libffi.call/cls_3byte2.c: Likewise.
1494 * testsuite/libffi.call/cls_4_1byte.c: Likewise.
1495 * testsuite/libffi.call/cls_4byte.c: Likewise.
1496 * testsuite/libffi.call/cls_64byte.c: Likewise.
1497 * testsuite/libffi.call/cls_6byte.c: Likewise.
1498 * testsuite/libffi.call/cls_7byte.c: Likewise.
1499 * testsuite/libffi.call/cls_8byte.c: Likewise.
1500 * testsuite/libffi.call/cls_9byte1.c: Likewise.
1501 * testsuite/libffi.call/cls_9byte2.c: Likewise.
1502 * testsuite/libffi.call/cls_align_double.c: Likewise.
1503 * testsuite/libffi.call/cls_align_float.c: Likewise.
1504 * testsuite/libffi.call/cls_align_longdouble.c: Likewise.
1505 * testsuite/libffi.call/cls_align_pointer.c: Likewise.
1506 * testsuite/libffi.call/cls_align_sint16.c: Likewise.
1507 * testsuite/libffi.call/cls_align_sint32.c: Likewise.
1508 * testsuite/libffi.call/cls_align_sint64.c: Likewise.
1509 * testsuite/libffi.call/cls_align_uint16.c: Likewise.
1510 * testsuite/libffi.call/cls_align_uint32.c: Likewise.
1511 * testsuite/libffi.call/cls_align_uint64.c: Likewise.
1512 * testsuite/libffi.call/cls_double.c: Likewise.
1513 * testsuite/libffi.call/cls_float.c: Likewise.
1514 * testsuite/libffi.call/cls_multi_schar.c: Likewise.
1515 * testsuite/libffi.call/cls_multi_sshort.c: Likewise.
1516 * testsuite/libffi.call/cls_multi_sshortchar.c: Likewise.
1517 * testsuite/libffi.call/cls_multi_uchar.c: Likewise.
1518 * testsuite/libffi.call/cls_multi_ushort.c: Likewise.
1519 * testsuite/libffi.call/cls_multi_ushortchar.c: Likewise.
1520 * testsuite/libffi.call/cls_schar.c: Likewise.
1521 * testsuite/libffi.call/cls_sint.c: Likewise.
1522 * testsuite/libffi.call/cls_sshort.c: Likewise.
1523 * testsuite/libffi.call/cls_uchar.c: Likewise.
1524 * testsuite/libffi.call/cls_uint.c: Likewise.
1525 * testsuite/libffi.call/cls_ulonglong.c: Likewise.
1526 * testsuite/libffi.call/cls_ushort.c: Likewise.
1527 * testsuite/libffi.call/nested_struct.c: Likewise.
1528 * testsuite/libffi.call/nested_struct1.c: Likewise.
1529 * testsuite/libffi.call/nested_struct2.c: Likewise.
1530 * testsuite/libffi.call/nested_struct3.c: Likewise.
1531 * testsuite/libffi.call/problem1.c: Likewise.
1532 * testsuite/libffi.special/unwindtest.cc: Likewise.
1533 * testsuite/libffi.call/cls_12byte.c: Likewise and set return value
1535 * testsuite/libffi.call/cls_16byte.c: Likewise.
1536 * testsuite/libffi.call/cls_5byte.c: Likewise.
1538 2004-08-23 David Daney <daney@avtrex.com>
1541 * src/mips/ffitarget.h (FFI_O32_SOFT_FLOAT): New ABI.
1542 * src/mips/ffi.c (ffi_prep_args): Fix alignment calculation.
1543 (ffi_prep_cif_machdep): Handle FFI_O32_SOFT_FLOAT floating point
1544 parameters and return types.
1545 (ffi_call): Handle FFI_O32_SOFT_FLOAT ABI.
1546 (ffi_prep_closure): Ditto.
1547 (ffi_closure_mips_inner_O32): Handle FFI_O32_SOFT_FLOAT ABI, fix
1548 alignment calculations.
1549 * src/mips/o32.S (ffi_closure_O32): Don't use floating point
1550 instructions if FFI_O32_SOFT_FLOAT, make stack frame ABI compliant.
1552 2004-08-14 Casey Marshall <csm@gnu.org>
1554 * src/mips/ffi.c (ffi_pref_cif_machdep): set `cif->flags' to
1555 contain `FFI_TYPE_UINT64' as return type for any 64-bit
1556 integer (O32 ABI only).
1557 (ffi_prep_closure): new function.
1558 (ffi_closure_mips_inner_O32): new function.
1559 * src/mips/ffitarget.h: Define `FFI_CLOSURES' and
1560 `FFI_TRAMPOLINE_SIZE' appropriately if the ABI is o32.
1561 * src/mips/o32.S (ffi_call_O32): add labels for .eh_frame. Return
1562 64 bit integers correctly.
1563 (ffi_closure_O32): new function.
1564 Added DWARF-2 unwind info for both functions.
1566 2004-08-10 Andrew Haley <aph@redhat.com>
1568 * src/x86/ffi64.c (ffi_prep_args ): 8-align all stack arguments.
1570 2004-08-01 Robert Millan <robertmh@gnu.org>
1572 * configure.ac: Detect knetbsd-gnu and kfreebsd-gnu.
1573 * configure: Regenerate.
1575 2004-07-30 Maciej W. Rozycki <macro@linux-mips.org>
1577 * acinclude.m4 (AC_FUNC_MMAP_BLACKLIST): Check for <sys/mman.h>
1578 and mmap() explicitly instead of relying on preset autoconf cache
1580 * aclocal.m4: Regenerate.
1581 * configure: Regenerate.
1583 2004-07-11 Ulrich Weigand <uweigand@de.ibm.com>
1585 * src/s390/ffi.c (ffi_prep_args): Fix C aliasing violation.
1586 (ffi_check_float_struct): Remove unused prototype.
1588 2004-06-30 Geoffrey Keating <geoffk@apple.com>
1590 * src/powerpc/ffi_darwin.c (flush_icache): ';' is a comment
1591 character on Darwin, use '\n\t' instead.
1593 2004-06-26 Matthias Klose <doko@debian.org>
1595 * libtool-version: Fix typo in revision/age.
1597 2004-06-17 Matthias Klose <doko@debian.org>
1599 * libtool-version: New.
1600 * Makefile.am (libffi_la_LDFLAGS): Use -version-info for soname.
1601 * Makefile.in: Regenerate.
1603 2004-06-15 Paolo Bonzini <bonzini@gnu.org>
1605 * Makefile.am: Remove useless multilib rules.
1606 * Makefile.in: Regenerate.
1607 * aclocal.m4: Regenerate with automake 1.8.5.
1608 * configure.ac: Remove useless multilib configury.
1609 * configure: Regenerate.
1611 2004-06-15 Paolo Bonzini <bonzini@gnu.org>
1613 * .cvsignore: New file.
1615 2004-06-10 Jakub Jelinek <jakub@redhat.com>
1617 * src/ia64/unix.S (ffi_call_unix): Insert group barrier break
1619 (ffi_closure_UNIX): Fix f14/f15 adjustment if FLOAT_SZ is ever
1622 2004-06-06 Sean McNeil <sean@mcneil.com>
1624 * configure.ac: Add x86_64-*-freebsd* support.
1625 * configure: Regenerate.
1627 2004-04-26 Joe Buck <jbuck@welsh-buck.org>
1630 * configure.ac: Test for existence of mmap and sys/mman.h before
1631 checking blacklist. Fix suggested by Jim Wilson.
1632 * configure: Regenerate.
1634 2004-04-26 Matt Austern <austern@apple.com>
1636 * src/powerpc/darwin.S: Go through a non-lazy pointer for initial
1638 * src/powerpc/darwin_closure.S: Likewise.
1640 2004-04-24 Andreas Tobler <a.tobler@schweiz.ch>
1642 * testsuite/libffi.call/cls_multi_schar.c (main): Fix initialization
1643 error. Reported by Thomas Heller <theller@python.net>.
1644 * testsuite/libffi.call/cls_multi_sshort.c (main): Likewise.
1645 * testsuite/libffi.call/cls_multi_ushort.c (main): Likewise.
1647 2004-03-20 Matthias Klose <doko@debian.org>
1649 * src/pa/linux.S: Fix typo.
1651 2004-03-19 Matthias Klose <doko@debian.org>
1653 * Makefile.am: Update.
1654 * Makefile.in: Regenerate.
1655 * src/pa/ffi.h.in: Remove.
1656 * src/pa/ffitarget.h: New file.
1658 2004-02-10 Randolph Chung <tausq@debian.org>
1660 * Makefile.am: Add PA support.
1661 * Makefile.in: Regenerate.
1662 * include/Makefile.in: Regenerate.
1663 * configure.ac: Add PA target.
1664 * configure: Regenerate.
1665 * src/pa/ffi.c: New file.
1666 * src/pa/ffi.h.in: Add PA support.
1667 * src/pa/linux.S: New file.
1668 * prep_cif.c: Add PA support.
1670 2004-03-16 Hosaka Yuji <hos@tamanegi.org>
1672 * src/types.c: Fix alignment size of X86_WIN32 case int64 and
1674 * src/x86/ffi.c (ffi_prep_args): Replace ecif->cif->rtype->type
1675 with ecif->cif->flags.
1676 (ffi_call, ffi_prep_incoming_args_SYSV): Replace cif->rtype->type
1678 (ffi_prep_cif_machdep): Add X86_WIN32 struct case.
1679 (ffi_closure_SYSV): Add 1 or 2-bytes struct case for X86_WIN32.
1680 * src/x86/win32.S (retstruct1b, retstruct2b, sc_retstruct1b,
1681 sc_retstruct2b): Add for 1 or 2-bytes struct case.
1683 2004-03-15 Kelley Cook <kcook@gcc.gnu.org>
1685 * configure.in: Rename file to ...
1686 * configure.ac: ... this.
1687 * fficonfig.h.in: Regenerate.
1688 * Makefile.in: Regenerate.
1689 * include/Makefile.in: Regenerate.
1690 * testsuite/Makefile.in: Regenerate.
1692 2004-03-12 Matt Austern <austern@apple.com>
1694 * src/powerpc/darwin.S: Fix EH information so it corresponds to
1695 changes in EH format resulting from addition of linkonce support.
1696 * src/powerpc/darwin_closure.S: Likewise.
1698 2004-03-11 Andreas Tobler <a.tobler@schweiz.ch>
1699 Paolo Bonzini <bonzini@gnu.org>
1701 * Makefile.am (AUTOMAKE_OPTIONS): Set them.
1702 Remove VPATH. Remove rules for object files. Remove multilib support.
1703 (AM_CCASFLAGS): Add.
1704 * configure.in (AC_CONFIG_HEADERS): Relace AM_CONFIG_HEADER.
1705 (AC_PREREQ): Bump version to 2.59.
1706 (AC_INIT): Fill with version info and bug address.
1707 (ORIGINAL_LD_FOR_MULTILIBS): Remove.
1708 (AM_ENABLE_MULTILIB): Use this instead of AC_ARG_ENABLE.
1709 De-precious CC so that the right flags are passed down to multilibs.
1710 (AC_MSG_ERROR): Replace obsolete macro AC_ERROR.
1711 (AC_CONFIG_FILES): Replace obsolete macro AC_LINK_FILES.
1712 (AC_OUTPUT): Reorganize the output with AC_CONFIG_COMMANDS.
1713 * configure: Rebuilt.
1714 * aclocal.m4: Likewise.
1715 * Makefile.in, include/Makefile.in, testsuite/Makefile.in: Likewise.
1716 * fficonfig.h.in: Likewise.
1718 2004-03-11 Andreas Schwab <schwab@suse.de>
1720 * src/ia64/ffi.c (ffi_prep_incoming_args_UNIX): Get floating point
1721 arguments from fp registers only for the first 8 parameter slots.
1722 Don't convert a float parameter when passed in memory.
1724 2004-03-09 Hans-Peter Nilsson <hp@axis.com>
1726 * configure: Regenerate for config/accross.m4 correction.
1728 2004-02-25 Matt Kraai <kraai@alumni.cmu.edu>
1730 * src/powerpc/ffi.c (ffi_prep_args_SYSV): Change
1731 ecif->cif->bytes to bytes.
1732 (ffi_prep_cif_machdep): Add braces around nested if statement.
1734 2004-02-09 Alan Modra <amodra@bigpond.net.au>
1736 * src/types.c (pointer): POWERPC64 has 8 byte pointers.
1738 * src/powerpc/ffi.c (ffi_prep_args64): Correct long double handling.
1739 (ffi_closure_helper_LINUX64): Fix typo.
1740 * testsuite/libffi.call/cls_align_longdouble.c: Pass -mlong-double-128
1742 * testsuite/libffi.call/float.c: Likewise.
1743 * testsuite/libffi.call/float2.c: Likewise.
1745 2004-02-08 Alan Modra <amodra@bigpond.net.au>
1747 * src/powerpc/ffi.c (ffi_prep_cif_machdep <FFI_LINUX64>): Correct
1748 long double function return and long double arg handling.
1749 (ffi_closure_helper_LINUX64): Formatting. Delete unused "ng" var.
1750 Use "end_pfr" instead of "nf". Correct long double handling.
1752 * src/powerpc/linux64.S (ffi_call_LINUX64): Save f2 long double
1754 * src/powerpc/linux64_closure.S (ffi_closure_LINUX64): Allocate
1755 space for long double return value. Adjust stack frame and offsets.
1756 Load f2 long double return.
1758 2004-02-07 Alan Modra <amodra@bigpond.net.au>
1760 * src/types.c: Use 16 byte long double for POWERPC64.
1762 2004-01-25 Eric Botcazou <ebotcazou@libertysurf.fr>
1764 * src/sparc/ffi.c (ffi_prep_args_v9): Shift the parameter array
1765 when the structure return address is passed in %o0.
1766 (ffi_V9_return_struct): Rename into ffi_v9_layout_struct.
1767 (ffi_v9_layout_struct): Align the field following a nested structure
1768 on a word boundary. Use memmove instead of memcpy.
1769 (ffi_call): Update call to ffi_V9_return_struct.
1770 (ffi_prep_closure): Define 'ctx' only for V8.
1771 (ffi_closure_sparc_inner): Clone into ffi_closure_sparc_inner_v8
1772 and ffi_closure_sparc_inner_v9.
1773 (ffi_closure_sparc_inner_v8): Return long doubles by reference.
1774 Always skip the structure return address. For structures and long
1775 doubles, copy the argument directly.
1776 (ffi_closure_sparc_inner_v9): Skip the structure return address only
1777 if required. Shift the maximum floating-point slot accordingly. For
1778 big structures, copy the argument directly; otherwise, left-justify the
1779 argument and call ffi_v9_layout_struct to lay out the structure on
1781 * src/sparc/v8.S: Undef STACKFRAME before defining it.
1782 (ffi_closure_v8): Pass the structure return address. Update call to
1783 ffi_closure_sparc_inner_v8. Short-circuit FFI_TYPE_INT handling.
1784 Skip the 'unimp' insn when returning long doubles and structures.
1785 * src/sparc/v9.S: Undef STACKFRAME before defining it.
1786 (ffi_closure_v9): Increase the frame size by 2 words. Short-circuit
1787 FFI_TYPE_INT handling. Load structures both in integers and
1788 floating-point registers on return.
1789 * README: Update status of the SPARC port.
1791 2004-01-24 Andreas Tobler <a.tobler@schweiz.ch>
1793 * testsuite/libffi.call/pyobjc-tc.c (main): Treat result value
1795 * testsuite/libffi.call/struct3.c (main): Fix CHECK.
1797 2004-01-22 Ulrich Weigand <uweigand@de.ibm.com>
1799 * testsuite/libffi.call/cls_uint.c (cls_ret_uint_fn): Treat result
1800 value as of type ffi_arg, not unsigned int.
1802 2004-01-21 Michael Ritzert <ritzert@t-online.de>
1804 * ffi64.c (ffi_prep_args): Cast the RHS of an assignment instead
1807 2004-01-12 Andreas Tobler <a.tobler@schweiz.ch>
1809 * testsuite/lib/libffi-dg.exp: Set LD_LIBRARY_PATH_32 for
1812 2004-01-08 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
1814 * testsuite/libffi.call/ffitest.h (allocate_mmap): Cast MAP_FAILED
1817 2003-12-10 Richard Henderson <rth@redhat.com>
1819 * testsuite/libffi.call/cls_align_pointer.c: Cast pointers to
1820 size_t instead of int.
1822 2003-12-04 Hosaka Yuji <hos@tamanegi.org>
1824 * testsuite/libffi.call/many_win32.c: Include <float.h>.
1825 * testsuite/libffi.call/many_win32.c (main): Replace variable
1826 int i with unsigned long ul.
1828 * testsuite/libffi.call/cls_align_uint64.c: New test case.
1829 * testsuite/libffi.call/cls_align_sint64.c: Likewise.
1830 * testsuite/libffi.call/cls_align_uint32.c: Likewise.
1831 * testsuite/libffi.call/cls_align_sint32.c: Likewise.
1832 * testsuite/libffi.call/cls_align_uint16.c: Likewise.
1833 * testsuite/libffi.call/cls_align_sint16.c: Likewise.
1834 * testsuite/libffi.call/cls_align_float.c: Likewise.
1835 * testsuite/libffi.call/cls_align_double.c: Likewise.
1836 * testsuite/libffi.call/cls_align_longdouble.c: Likewise.
1837 * testsuite/libffi.call/cls_align_pointer.c: Likewise.
1839 2003-12-02 Hosaka Yuji <hos@tamanegi.org>
1842 * src/x86/ffi.c (ffi_prep_args, ffi_prep_incoming_args_SYSV):
1843 Align arguments to 32 bits.
1845 2003-12-01 Andreas Tobler <a.tobler@schweiz.ch>
1848 * testsuite/libffi.call/cls_multi_sshort.c: New test case.
1849 * testsuite/libffi.call/cls_multi_sshortchar.c: Likewise.
1850 * testsuite/libffi.call/cls_multi_uchar.c: Likewise.
1851 * testsuite/libffi.call/cls_multi_schar.c: Likewise.
1852 * testsuite/libffi.call/cls_multi_ushortchar.c: Likewise.
1853 * testsuite/libffi.call/cls_multi_ushort.c: Likewise.
1855 * testsuite/libffi.special/unwindtest.cc: Cosmetics.
1857 2003-11-26 Kaveh R. Ghazi <ghazi@caip.rutgers.edu>
1859 * testsuite/libffi.call/ffitest.h: Include <fcntl.h>.
1860 * testsuite/libffi.special/ffitestcxx.h: Likewise.
1862 2003-11-22 Andreas Tobler <a.tobler@schweiz.ch>
1864 * Makefile.in: Rebuilt.
1865 * configure: Likewise.
1866 * testsuite/libffi.special/unwindtest.cc: Convert the mmap to
1869 2003-11-21 Andreas Jaeger <aj@suse.de>
1870 Andreas Tobler <a.tobler@schweiz.ch>
1872 * acinclude.m4: Add AC_FUNC_MMAP_BLACKLIST.
1873 * configure.in: Call AC_FUNC_MMAP_BLACKLIST.
1874 * Makefile.in: Rebuilt.
1875 * aclocal.m4: Likewise.
1876 * configure: Likewise.
1877 * fficonfig.h.in: Likewise.
1878 * testsuite/lib/libffi-dg.exp: Add include dir.
1879 * testsuite/libffi.call/ffitest.h: Add MMAP definitions.
1880 * testsuite/libffi.special/ffitestcxx.h: Likewise.
1881 * testsuite/libffi.call/closure_fn0.c: Use MMAP functionality
1882 for ffi_closure if available.
1883 * testsuite/libffi.call/closure_fn1.c: Likewise.
1884 * testsuite/libffi.call/closure_fn2.c: Likewise.
1885 * testsuite/libffi.call/closure_fn3.c: Likewise.
1886 * testsuite/libffi.call/closure_fn4.c: Likewise.
1887 * testsuite/libffi.call/closure_fn5.c: Likewise.
1888 * testsuite/libffi.call/cls_12byte.c: Likewise.
1889 * testsuite/libffi.call/cls_16byte.c: Likewise.
1890 * testsuite/libffi.call/cls_18byte.c: Likewise.
1891 * testsuite/libffi.call/cls_19byte.c: Likewise.
1892 * testsuite/libffi.call/cls_1_1byte.c: Likewise.
1893 * testsuite/libffi.call/cls_20byte.c: Likewise.
1894 * testsuite/libffi.call/cls_20byte1.c: Likewise.
1895 * testsuite/libffi.call/cls_24byte.c: Likewise.
1896 * testsuite/libffi.call/cls_2byte.c: Likewise.
1897 * testsuite/libffi.call/cls_3_1byte.c: Likewise.
1898 * testsuite/libffi.call/cls_3byte1.c: Likewise.
1899 * testsuite/libffi.call/cls_3byte2.c: Likewise.
1900 * testsuite/libffi.call/cls_4_1byte.c: Likewise.
1901 * testsuite/libffi.call/cls_4byte.c: Likewise.
1902 * testsuite/libffi.call/cls_5byte.c: Likewise.
1903 * testsuite/libffi.call/cls_64byte.c: Likewise.
1904 * testsuite/libffi.call/cls_6byte.c: Likewise.
1905 * testsuite/libffi.call/cls_7byte.c: Likewise.
1906 * testsuite/libffi.call/cls_8byte.c: Likewise.
1907 * testsuite/libffi.call/cls_9byte1.c: Likewise.
1908 * testsuite/libffi.call/cls_9byte2.c: Likewise.
1909 * testsuite/libffi.call/cls_double.c: Likewise.
1910 * testsuite/libffi.call/cls_float.c: Likewise.
1911 * testsuite/libffi.call/cls_schar.c: Likewise.
1912 * testsuite/libffi.call/cls_sint.c: Likewise.
1913 * testsuite/libffi.call/cls_sshort.c: Likewise.
1914 * testsuite/libffi.call/cls_uchar.c: Likewise.
1915 * testsuite/libffi.call/cls_uint.c: Likewise.
1916 * testsuite/libffi.call/cls_ulonglong.c: Likewise.
1917 * testsuite/libffi.call/cls_ushort.c: Likewise.
1918 * testsuite/libffi.call/nested_struct.c: Likewise.
1919 * testsuite/libffi.call/nested_struct1.c: Likewise.
1920 * testsuite/libffi.call/nested_struct2.c: Likewise.
1921 * testsuite/libffi.call/nested_struct3.c: Likewise.
1922 * testsuite/libffi.call/problem1.c: Likewise.
1923 * testsuite/libffi.special/unwindtest.cc: Likewise.
1925 2003-11-20 Andreas Tobler <a.tobler@schweiz.ch>
1927 * testsuite/lib/libffi-dg.exp: Make the -lgcc_s conditional.
1929 2003-11-19 Andreas Tobler <a.tobler@schweiz.ch>
1931 * testsuite/lib/libffi-dg.exp: Add DYLD_LIBRARY_PATH for darwin.
1932 Add -lgcc_s to additional flags.
1934 2003-11-12 Andreas Tobler <a.tobler@schweiz.ch>
1936 * configure.in, include/Makefile.am: PR libgcj/11147, install
1937 the ffitarget.h header file in a gcc versioned and target
1939 * configure: Regenerated.
1940 * Makefile.in, include/Makefile.in: Likewise.
1941 * testsuite/Makefile.in: Likewise.
1943 2003-11-09 Andreas Tobler <a.tobler@schweiz.ch>
1945 * testsuite/libffi.call/closure_fn0.c: Print result and check
1946 with dg-output to make debugging easier.
1947 * testsuite/libffi.call/closure_fn1.c: Likewise.
1948 * testsuite/libffi.call/closure_fn2.c: Likewise.
1949 * testsuite/libffi.call/closure_fn3.c: Likewise.
1950 * testsuite/libffi.call/closure_fn4.c: Likewise.
1951 * testsuite/libffi.call/closure_fn5.c: Likewise.
1952 * testsuite/libffi.call/cls_12byte.c: Likewise.
1953 * testsuite/libffi.call/cls_16byte.c: Likewise.
1954 * testsuite/libffi.call/cls_18byte.c: Likewise.
1955 * testsuite/libffi.call/cls_19byte.c: Likewise.
1956 * testsuite/libffi.call/cls_1_1byte.c: Likewise.
1957 * testsuite/libffi.call/cls_20byte.c: Likewise.
1958 * testsuite/libffi.call/cls_20byte1.c: Likewise.
1959 * testsuite/libffi.call/cls_24byte.c: Likewise.
1960 * testsuite/libffi.call/cls_2byte.c: Likewise.
1961 * testsuite/libffi.call/cls_3_1byte.c: Likewise.
1962 * testsuite/libffi.call/cls_3byte1.c: Likewise.
1963 * testsuite/libffi.call/cls_3byte2.c: Likewise.
1964 * testsuite/libffi.call/cls_4_1byte.c: Likewise.
1965 * testsuite/libffi.call/cls_4byte.c: Likewise.
1966 * testsuite/libffi.call/cls_5byte.c: Likewise.
1967 * testsuite/libffi.call/cls_64byte.c: Likewise.
1968 * testsuite/libffi.call/cls_6byte.c: Likewise.
1969 * testsuite/libffi.call/cls_7byte.c: Likewise.
1970 * testsuite/libffi.call/cls_8byte.c: Likewise.
1971 * testsuite/libffi.call/cls_9byte1.c: Likewise.
1972 * testsuite/libffi.call/cls_9byte2.c: Likewise.
1973 * testsuite/libffi.call/cls_double.c: Likewise.
1974 * testsuite/libffi.call/cls_float.c: Likewise.
1975 * testsuite/libffi.call/cls_schar.c: Likewise.
1976 * testsuite/libffi.call/cls_sint.c: Likewise.
1977 * testsuite/libffi.call/cls_sshort.c: Likewise.
1978 * testsuite/libffi.call/cls_uchar.c: Likewise.
1979 * testsuite/libffi.call/cls_uint.c: Likewise.
1980 * testsuite/libffi.call/cls_ulonglong.c: Likewise.
1981 * testsuite/libffi.call/cls_ushort.c: Likewise.
1982 * testsuite/libffi.call/problem1.c: Likewise.
1984 * testsuite/libffi.special/unwindtest.cc: Make ffi_closure
1987 2003-11-08 Andreas Tobler <a.tobler@schweiz.ch>
1989 * testsuite/libffi.call/cls_9byte2.c: New test case.
1990 * testsuite/libffi.call/cls_9byte1.c: Likewise.
1991 * testsuite/libffi.call/cls_64byte.c: Likewise.
1992 * testsuite/libffi.call/cls_20byte1.c: Likewise.
1993 * testsuite/libffi.call/cls_19byte.c: Likewise.
1994 * testsuite/libffi.call/cls_18byte.c: Likewise.
1995 * testsuite/libffi.call/closure_fn4.c: Likewise.
1996 * testsuite/libffi.call/closure_fn5.c: Likewise.
1997 * testsuite/libffi.call/cls_schar.c: Likewise.
1998 * testsuite/libffi.call/cls_sint.c: Likewise.
1999 * testsuite/libffi.call/cls_sshort.c: Likewise.
2000 * testsuite/libffi.call/nested_struct2.c: Likewise.
2001 * testsuite/libffi.call/nested_struct3.c: Likewise.
2003 2003-11-08 Andreas Tobler <a.tobler@schweiz.ch>
2005 * testsuite/libffi.call/cls_double.c: Do a check on the result.
2006 * testsuite/libffi.call/cls_uchar.c: Likewise.
2007 * testsuite/libffi.call/cls_uint.c: Likewise.
2008 * testsuite/libffi.call/cls_ulonglong.c: Likewise.
2009 * testsuite/libffi.call/cls_ushort.c: Likewise.
2010 * testsuite/libffi.call/return_sc.c: Cleanup whitespaces.
2012 2003-11-06 Andreas Tobler <a.tobler@schweiz.ch>
2014 * src/prep_cif.c (ffi_prep_cif): Move the validity check after
2017 2003-10-23 Andreas Tobler <a.tobler@schweiz.ch>
2019 * src/java_raw_api.c (ffi_java_ptrarray_to_raw): Replace
2020 FFI_ASSERT(FALSE) with FFI_ASSERT(0).
2022 2003-10-22 David Daney <ddaney@avtrex.com>
2024 * src/mips/ffitarget.h: Replace undefined UINT32 and friends with
2025 __attribute__((__mode__(__SI__))) and friends.
2027 2003-10-22 Andreas Schwab <schwab@suse.de>
2029 * src/ia64/ffi.c: Replace FALSE/TRUE with false/true.
2031 2003-10-21 Andreas Tobler <a.tobler@schweiz.ch>
2033 * configure.in: AC_LINK_FILES(ffitarget.h).
2034 * configure: Regenerate.
2035 * Makefile.in: Likewise.
2036 * include/Makefile.in: Likewise.
2037 * testsuite/Makefile.in: Likewise.
2038 * fficonfig.h.in: Likewise.
2040 2003-10-21 Paolo Bonzini <bonzini@gnu.org>
2041 Richard Henderson <rth@redhat.com>
2043 Avoid that ffi.h includes fficonfig.h.
2045 * Makefile.am (EXTRA_DIST): Include ffitarget.h files
2046 (TARGET_SRC_MIPS_GCC): Renamed to TARGET_SRC_MIPS_IRIX.
2047 (TARGET_SRC_MIPS_SGI): Removed.
2048 (MIPS_GCC): Renamed to TARGET_SRC_MIPS_IRIX.
2049 (MIPS_SGI): Removed.
2050 (CLEANFILES): Removed.
2051 (mostlyclean-am, clean-am, mostlyclean-sub, clean-sub): New
2053 * acconfig.h: Removed.
2054 * configure.in: Compute sizeofs only for double and long double.
2055 Use them to define and subst HAVE_LONG_DOUBLE. Include comments
2056 into AC_DEFINE instead of using acconfig.h. Create
2057 include/ffitarget.h instead of include/fficonfig.h. Rename
2058 MIPS_GCC to MIPS_IRIX, drop MIPS_SGI since we are in gcc's tree.
2059 AC_DEFINE EH_FRAME_FLAGS.
2060 * include/Makefile.am (DISTCLEANFILES): New automake macro.
2061 (hack_DATA): Add ffitarget.h.
2062 * include/ffi.h.in: Remove all system specific definitions.
2063 Declare raw API even if it is not installed, why bother?
2064 Use limits.h instead of SIZEOF_* to define ffi_type_*. Do
2065 not define EH_FRAME_FLAGS, it is in fficonfig.h now. Include
2066 ffitarget.h instead of fficonfig.h. Remove ALIGN macro.
2067 (UINT_ARG, INT_ARG): Removed, use ffi_arg and ffi_sarg instead.
2068 * include/ffi_common.h (bool): Do not define.
2069 (ffi_assert): Accept failed assertion.
2070 (ffi_type_test): Return void and accept file/line.
2071 (FFI_ASSERT): Pass stringized failed assertion.
2072 (FFI_ASSERT_AT): New macro.
2073 (FFI_ASSERT_VALID_TYPE): New macro.
2074 (UINT8, SINT8, UINT16, SINT16, UINT32, SINT32,
2075 UINT64, SINT64): Define here with gcc's __attribute__ macro
2077 (FLOAT32, ALIGN): Define here instead of in ffi.h
2078 * include/ffi-mips.h: Removed. Its content moved to
2079 src/mips/ffitarget.h after separating assembly and C sections.
2080 * src/alpha/ffi.c, src/alpha/ffi.c, src/java_raw_api.c
2081 src/prep_cif.c, src/raw_api.c, src/ia64/ffi.c,
2082 src/mips/ffi.c, src/mips/n32.S, src/mips/o32.S,
2083 src/mips/ffitarget.h, src/sparc/ffi.c, src/x86/ffi64.c:
2084 SIZEOF_ARG -> FFI_SIZEOF_ARG.
2085 * src/ia64/ffi.c: Include stdbool.h (provided by GCC 2.95+).
2086 * src/debug.c (ffi_assert): Accept stringized failed assertion.
2087 (ffi_type_test): Rewritten.
2088 * src/prep-cif.c (initialize_aggregate, ffi_prep_cif): Call
2089 FFI_ASSERT_VALID_TYPE.
2090 * src/alpha/ffitarget.h, src/arm/ffitarget.h,
2091 src/ia64/ffitarget.h, src/m68k/ffitarget.h,
2092 src/mips/ffitarget.h, src/powerpc/ffitarget.h,
2093 src/s390/ffitarget.h, src/sh/ffitarget.h,
2094 src/sh64/ffitarget.h, src/sparc/ffitarget.h,
2095 src/x86/ffitarget.h: New files.
2096 * src/alpha/osf.S, src/arm/sysv.S, src/ia64/unix.S,
2097 src/m68k/sysv.S, src/mips/n32.S, src/mips/o32.S,
2098 src/powerpc/aix.S, src/powerpc/darwin.S,
2099 src/powerpc/ffi_darwin.c, src/powerpc/linux64.S,
2100 src/powerpc/linux64_closure.S, src/powerpc/ppc_closure.S,
2101 src/powerpc/sysv.S, src/s390/sysv.S, src/sh/sysv.S,
2102 src/sh64/sysv.S, src/sparc/v8.S, src/sparc/v9.S,
2103 src/x86/sysv.S, src/x86/unix64.S, src/x86/win32.S:
2106 2003-10-20 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
2108 * src/mips/ffi.c: Use _ABIN32, _ABIO32 instead of external
2109 _MIPS_SIM_NABI32, _MIPS_SIM_ABI32.
2111 2003-10-19 Andreas Tobler <a.tobler@schweiz.ch>
2113 * src/powerpc/ffi_darwin.c (ffi_prep_args): Declare bytes again.
2114 Used when FFI_DEBUG = 1.
2116 2003-10-14 Alan Modra <amodra@bigpond.net.au>
2118 * src/types.c (double, longdouble): Default POWERPC64 to 8 byte size
2121 2003-10-06 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
2123 * include/ffi_mips.h: Define FFI_MIPS_N32 for N32/N64 ABIs,
2124 FFI_MIPS_O32 for O32 ABI.
2126 2003-10-01 Andreas Tobler <a.tobler@schweiz.ch>
2128 * testsuite/lib/libffi-dg.exp: Set LD_LIBRARY_PATH_64 for
2129 SPARC64. Cleanup whitespaces.
2131 2003-09-19 Andreas Tobler <a.tobler@schweiz.ch>
2133 * testsuite/libffi.call/closure_fn0.c: Xfail mips, arm,
2134 strongarm, xscale. Cleanup whitespaces.
2135 * testsuite/libffi.call/closure_fn1.c: Likewise.
2136 * testsuite/libffi.call/closure_fn2.c: Likewise.
2137 * testsuite/libffi.call/closure_fn3.c: Likewise.
2138 * testsuite/libffi.call/cls_12byte.c: Likewise.
2139 * testsuite/libffi.call/cls_16byte.c: Likewise.
2140 * testsuite/libffi.call/cls_1_1byte.c: Likewise.
2141 * testsuite/libffi.call/cls_20byte.c: Likewise.
2142 * testsuite/libffi.call/cls_24byte.c: Likewise.
2143 * testsuite/libffi.call/cls_2byte.c: Likewise.
2144 * testsuite/libffi.call/cls_3_1byte.c: Likewise.
2145 * testsuite/libffi.call/cls_3byte1.c: Likewise.
2146 * testsuite/libffi.call/cls_3byte2.c: Likewise.
2147 * testsuite/libffi.call/cls_4_1byte.c: Likewise.
2148 * testsuite/libffi.call/cls_4byte.c: Likewise.
2149 * testsuite/libffi.call/cls_5byte.c: Likewise.
2150 * testsuite/libffi.call/cls_6byte.c: Likewise.
2151 * testsuite/libffi.call/cls_7byte.c: Likewise.
2152 * testsuite/libffi.call/cls_8byte.c: Likewise.
2153 * testsuite/libffi.call/cls_double.c: Likewise.
2154 * testsuite/libffi.call/cls_float.c: Likewise.
2155 * testsuite/libffi.call/cls_uchar.c: Likewise.
2156 * testsuite/libffi.call/cls_uint.c: Likewise.
2157 * testsuite/libffi.call/cls_ulonglong.c: Likewise.
2158 * testsuite/libffi.call/cls_ushort.c: Likewise.
2159 * testsuite/libffi.call/nested_struct.c: Likewise.
2160 * testsuite/libffi.call/nested_struct1.c: Likewise.
2161 * testsuite/libffi.call/problem1.c: Likewise.
2162 * testsuite/libffi.special/unwindtest.cc: Likewise.
2163 * testsuite/libffi.call/pyobjc-tc.c: Cleanup whitespaces.
2165 2003-09-18 David Edelsohn <edelsohn@gnu.org>
2167 * src/powerpc/aix.S: Cleanup whitespaces.
2168 * src/powerpc/aix_closure.S: Likewise.
2170 2003-09-18 Andreas Tobler <a.tobler@schweiz.ch>
2172 * src/powerpc/darwin.S: Cleanup whitespaces, comment formatting.
2173 * src/powerpc/darwin_closure.S: Likewise.
2174 * src/powerpc/ffi_darwin.c: Likewise.
2176 2003-09-18 Andreas Tobler <a.tobler@schweiz.ch>
2177 David Edelsohn <edelsohn@gnu.org>
2179 * src/types.c (double): Add AIX and Darwin to the right TYPEDEF.
2180 * src/powerpc/aix_closure.S: Remove the pointer to the outgoing
2182 * src/powerpc/darwin_closure.S: Likewise.
2183 * src/powerpc/ffi_darwin.c (ffi_prep_args): Handle structures
2184 according to the Darwin/AIX ABI.
2185 (ffi_prep_cif_machdep): Likewise.
2186 (ffi_closure_helper_DARWIN): Likewise.
2187 Remove the outgoing parameter stack logic. Simplify the evaluation
2188 of the different CASE types.
2189 (ffi_prep_clousure): Avoid the casts on lvalues. Change the branch
2190 statement in the trampoline code.
2192 2003-09-18 Kaz Kojima <kkojima@gcc.gnu.org>
2194 * src/sh/ffi.c (ffi_prep_args): Take account into the alignement
2195 for the register size.
2196 (ffi_closure_helper_SYSV): Handle the structure return value
2198 (ffi_closure_helper_SYSV): Return the appropriate type when
2199 the registers are used for the structure return value.
2200 * src/sh/sysv.S (ffi_closure_SYSV): Fix the stack layout for
2201 the 64-bit return value. Update copyright years.
2203 2003-09-17 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
2205 * testsuite/lib/libffi-dg.exp (libffi_target_compile): Search in
2206 srcdir for ffi_mips.h.
2208 2003-09-12 Alan Modra <amodra@bigpond.net.au>
2210 * src/prep_cif.c (initialize_aggregate): Include tail padding in
2212 * src/powerpc/linux64_closure.S (ffi_closure_LINUX64): Correct
2213 placement of float result.
2214 * testsuite/libffi.special/unwindtest.cc (closure_test_fn1): Correct
2215 cast of "resp" for big-endian 64 bit machines.
2217 2003-09-11 Alan Modra <amodra@bigpond.net.au>
2219 * src/types.c (double, longdouble): Merge identical SH and ARM
2220 typedefs, and add POWERPC64.
2221 * src/powerpc/ffi.c (ffi_prep_args64): Correct next_arg calc for
2222 struct split over gpr and rest.
2223 (ffi_prep_cif_machdep): Correct intarg_count for structures.
2224 * src/powerpc/linux64.S (ffi_call_LINUX64): Fix gpr offsets.
2226 2003-09-09 Andreas Tobler <a.tobler@schweiz.ch>
2228 * src/powerpc/ffi.c (ffi_closure_helper_SYSV) Handle struct
2231 2003-09-09 Alan Modra <amodra@bigpond.net.au>
2233 * configure: Regenerate.
2235 2003-09-04 Andreas Tobler <a.tobler@schweiz.ch>
2237 * Makefile.am: Remove build rules for ffitest.
2238 * Makefile.in: Rebuilt.
2240 2003-09-04 Andreas Tobler <a.tobler@schweiz.ch>
2242 * src/java_raw_api.c: Include <stdlib.h> to fix compiler warning
2243 about implicit declaration of abort().
2245 2003-09-04 Andreas Tobler <a.tobler@schweiz.ch>
2247 * Makefile.am: Add dejagnu test framework. Fixes PR other/11411.
2248 * Makefile.in: Rebuilt.
2249 * configure.in: Add dejagnu test framework.
2250 * configure: Rebuilt.
2252 * testsuite/Makefile.am: New file.
2253 * testsuite/Makefile.in: Built
2254 * testsuite/lib/libffi-dg.exp: New file.
2255 * testsuite/config/default.exp: Likewise.
2256 * testsuite/libffi.call/call.exp: Likewise.
2257 * testsuite/libffi.call/ffitest.h: Likewise.
2258 * testsuite/libffi.call/closure_fn0.c: Likewise.
2259 * testsuite/libffi.call/closure_fn1.c: Likewise.
2260 * testsuite/libffi.call/closure_fn2.c: Likewise.
2261 * testsuite/libffi.call/closure_fn3.c: Likewise.
2262 * testsuite/libffi.call/cls_1_1byte.c: Likewise.
2263 * testsuite/libffi.call/cls_3_1byte.c: Likewise.
2264 * testsuite/libffi.call/cls_4_1byte.c: Likewise.
2265 * testsuite/libffi.call/cls_2byte.c: Likewise.
2266 * testsuite/libffi.call/cls_3byte1.c: Likewise.
2267 * testsuite/libffi.call/cls_3byte2.c: Likewise.
2268 * testsuite/libffi.call/cls_4byte.c: Likewise.
2269 * testsuite/libffi.call/cls_5byte.c: Likewise.
2270 * testsuite/libffi.call/cls_6byte.c: Likewise.
2271 * testsuite/libffi.call/cls_7byte.c: Likewise.
2272 * testsuite/libffi.call/cls_8byte.c: Likewise.
2273 * testsuite/libffi.call/cls_12byte.c: Likewise.
2274 * testsuite/libffi.call/cls_16byte.c: Likewise.
2275 * testsuite/libffi.call/cls_20byte.c: Likewise.
2276 * testsuite/libffi.call/cls_24byte.c: Likewise.
2277 * testsuite/libffi.call/cls_double.c: Likewise.
2278 * testsuite/libffi.call/cls_float.c: Likewise.
2279 * testsuite/libffi.call/cls_uchar.c: Likewise.
2280 * testsuite/libffi.call/cls_uint.c: Likewise.
2281 * testsuite/libffi.call/cls_ulonglong.c: Likewise.
2282 * testsuite/libffi.call/cls_ushort.c: Likewise.
2283 * testsuite/libffi.call/float.c: Likewise.
2284 * testsuite/libffi.call/float1.c: Likewise.
2285 * testsuite/libffi.call/float2.c: Likewise.
2286 * testsuite/libffi.call/many.c: Likewise.
2287 * testsuite/libffi.call/many_win32.c: Likewise.
2288 * testsuite/libffi.call/nested_struct.c: Likewise.
2289 * testsuite/libffi.call/nested_struct1.c: Likewise.
2290 * testsuite/libffi.call/pyobjc-tc.c: Likewise.
2291 * testsuite/libffi.call/problem1.c: Likewise.
2292 * testsuite/libffi.call/promotion.c: Likewise.
2293 * testsuite/libffi.call/return_ll.c: Likewise.
2294 * testsuite/libffi.call/return_sc.c: Likewise.
2295 * testsuite/libffi.call/return_uc.c: Likewise.
2296 * testsuite/libffi.call/strlen.c: Likewise.
2297 * testsuite/libffi.call/strlen_win32.c: Likewise.
2298 * testsuite/libffi.call/struct1.c: Likewise.
2299 * testsuite/libffi.call/struct2.c: Likewise.
2300 * testsuite/libffi.call/struct3.c: Likewise.
2301 * testsuite/libffi.call/struct4.c: Likewise.
2302 * testsuite/libffi.call/struct5.c: Likewise.
2303 * testsuite/libffi.call/struct6.c: Likewise.
2304 * testsuite/libffi.call/struct7.c: Likewise.
2305 * testsuite/libffi.call/struct8.c: Likewise.
2306 * testsuite/libffi.call/struct9.c: Likewise.
2307 * testsuite/libffi.special/special.exp: New file.
2308 * testsuite/libffi.special/ffitestcxx.h: Likewise.
2309 * testsuite/libffi.special/unwindtest.cc: Likewise.
2312 2003-08-13 Kaz Kojima <kkojima@gcc.gnu.org>
2314 * src/sh/ffi.c (OFS_INT16): Set 0 for little endian case. Update
2317 2003-08-02 Alan Modra <amodra@bigpond.net.au>
2319 * src/powerpc/ffi.c (ffi_prep_args64): Modify for changed gcc
2321 (ffi_closure_helper_LINUX64): Likewise.
2322 * src/powerpc/linux64.S: Remove code writing to parm save area.
2323 * src/powerpc/linux64_closure.S (ffi_closure_LINUX64): Use return
2324 address in lr from ffi_closure_helper_LINUX64 call to calculate
2325 table address. Optimize function tail.
2327 2003-07-28 Andreas Tobler <a.tobler@schweiz.ch>
2329 * src/sparc/ffi.c: Handle all floating point registers.
2330 * src/sparc/v9.S: Likewise. Fixes second part of PR target/11410.
2332 2003-07-11 Gerald Pfeifer <pfeifer@dbai.tuwien.ac.at>
2334 * README: Note that libffi is not part of GCC. Update the project
2337 2003-06-19 Franz Sirl <Franz.Sirl-kernel@lauterbach.com>
2339 * src/powerpc/ppc_closure.S: Include ffi.h.
2341 2003-06-13 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
2343 * src/x86/sysv.S: Avoid gas-only .uleb128/.sleb128 directives.
2344 Use C style comments.
2346 2003-06-13 Kaz Kojima <kkojima@rr.iij4u.or.jp>
2348 * Makefile.am: Add SHmedia support. Fix a typo of SH support.
2349 * Makefile.in: Regenerate.
2350 * configure.in (sh64-*-linux*, sh5*-*-linux*): Add target.
2351 * configure: Regenerate.
2352 * include/ffi.h.in: Add SHmedia support.
2353 * src/sh64/ffi.c: New file.
2354 * src/sh64/sysv.S: New file.
2356 2003-05-16 Jakub Jelinek <jakub@redhat.com>
2358 * configure.in (HAVE_RO_EH_FRAME): Check whether .eh_frame section
2359 should be read-only.
2360 * configure: Rebuilt.
2361 * fficonfig.h.in: Rebuilt.
2362 * include/ffi.h.in (EH_FRAME_FLAGS): Define.
2363 * src/alpha/osf.S: Use EH_FRAME_FLAGS.
2364 * src/powerpc/linux64.S: Likewise.
2365 * src/powerpc/linux64_closure.S: Likewise. Include ffi.h.
2366 * src/powerpc/sysv.S: Use EH_FRAME_FLAGS. Use pcrel encoding
2367 if -fpic/-fPIC/-mrelocatable.
2368 * src/powerpc/powerpc_closure.S: Likewise.
2369 * src/sparc/v8.S: If HAVE_RO_EH_FRAME is defined, don't include
2370 #write in .eh_frame flags.
2371 * src/sparc/v9.S: Likewise.
2372 * src/x86/unix64.S: Use EH_FRAME_FLAGS.
2373 * src/x86/sysv.S: Likewise. Use pcrel encoding if -fpic/-fPIC.
2374 * src/s390/sysv.S: Use EH_FRAME_FLAGS. Include ffi.h.
2376 2003-05-07 Jeff Sturm <jsturm@one-point.com>
2378 Fixes PR bootstrap/10656
2379 * configure.in (HAVE_AS_REGISTER_PSEUDO_OP): Test assembler
2380 support for .register pseudo-op.
2381 * src/sparc/v8.S: Use it.
2382 * fficonfig.h.in: Rebuilt.
2383 * configure: Rebuilt.
2385 2003-04-18 Jakub Jelinek <jakub@redhat.com>
2387 * include/ffi.h.in (POWERPC64): Define if 64-bit.
2388 (enum ffi_abi): Add FFI_LINUX64 on POWERPC.
2389 Make it the default on POWERPC64.
2390 (FFI_TRAMPOLINE_SIZE): Define to 24 on POWERPC64.
2391 * configure.in: Change powerpc-*-linux* into powerpc*-*-linux*.
2392 * configure: Rebuilt.
2393 * src/powerpc/ffi.c (hidden): Define.
2394 (ffi_prep_args_SYSV): Renamed from
2395 ffi_prep_args. Cast pointers to unsigned long to shut up warnings.
2396 (NUM_GPR_ARG_REGISTERS64, NUM_FPR_ARG_REGISTERS64,
2397 ASM_NEEDS_REGISTERS64): New.
2398 (ffi_prep_args64): New function.
2399 (ffi_prep_cif_machdep): Handle FFI_LINUX64 ABI.
2400 (ffi_call): Likewise.
2401 (ffi_prep_closure): Likewise.
2402 (flush_icache): Surround by #ifndef POWERPC64.
2403 (ffi_dblfl): New union type.
2404 (ffi_closure_helper_SYSV): Use it to avoid aliasing problems.
2405 (ffi_closure_helper_LINUX64): New function.
2406 * src/powerpc/ppc_closure.S: Surround whole file by #ifndef
2408 * src/powerpc/sysv.S: Likewise.
2409 (ffi_call_SYSV): Rename ffi_prep_args to ffi_prep_args_SYSV.
2410 * src/powerpc/linux64.S: New file.
2411 * src/powerpc/linux64_closure.S: New file.
2412 * Makefile.am (EXTRA_DIST): Add src/powerpc/linux64.S and
2413 src/powerpc/linux64_closure.S.
2414 (TARGET_SRC_POWERPC): Likewise.
2416 * src/ffitest.c (closure_test_fn, closure_test_fn1, closure_test_fn2,
2417 closure_test_fn3): Fix result printing on big-endian 64-bit
2419 (main): Print tst2_arg instead of uninitialized tst2_result.
2421 * src/ffitest.c (main): Hide what closure pointer really points to
2424 2003-04-16 Richard Earnshaw <rearnsha@arm.com>
2426 * configure.in (arm-*-netbsdelf*): Add configuration.
2427 (configure): Regenerated.
2429 2003-04-04 Loren J. Rittle <ljrittle@acm.org>
2431 * include/Makefile.in: Regenerate.
2433 2003-03-21 Zdenek Dvorak <rakdver@atrey.karlin.mff.cuni.cz>
2435 * libffi/include/ffi.h.in: Define X86 instead of X86_64 in 32
2437 * libffi/src/x86/ffi.c (ffi_closure_SYSV, ffi_closure_raw_SYSV):
2438 Receive closure pointer through parameter, read args using
2439 __builtin_dwarf_cfa.
2440 (FFI_INIT_TRAMPOLINE): Send closure reference through eax.
2442 2003-03-12 Andreas Schwab <schwab@suse.de>
2444 * configure.in: Avoid trailing /. in toolexeclibdir.
2445 * configure: Rebuilt.
2447 2003-03-03 Andreas Tobler <a.tobler@schweiz.ch>
2449 * src/powerpc/darwin_closure.S: Recode to fit dynamic libraries.
2451 2003-02-06 Andreas Tobler <a.tobler@schweiz.ch>
2453 * libffi/src/powerpc/darwin_closure.S:
2454 Fix alignement bug, allocate 8 bytes for the result.
2455 * libffi/src/powerpc/aix_closure.S:
2457 * libffi/src/powerpc/ffi_darwin.c:
2458 Update stackframe description for aix/darwin_closure.S.
2460 2003-02-06 Jakub Jelinek <jakub@redhat.com>
2462 * src/s390/ffi.c (ffi_closure_helper_SYSV): Add hidden visibility
2465 2003-01-31 Christian Cornelssen <ccorn@cs.tu-berlin.de>,
2466 Andreas Schwab <schwab@suse.de>
2468 * configure.in: Adjust command to source config-ml.in to account
2469 for changes to the libffi_basedir definition.
2470 (libffi_basedir): Remove ${srcdir} from value and include trailing
2473 * configure: Regenerate.
2475 2003-01-29 Franz Sirl <Franz.Sirl-kernel@lauterbach.com>
2477 * src/powerpc/ppc_closure.S: Recode to fit shared libs.
2479 2003-01-28 Andrew Haley <aph@redhat.com>
2481 * include/ffi.h.in: Enable FFI_CLOSURES for x86_64.
2482 * src/x86/ffi64.c (ffi_prep_closure): New.
2483 (ffi_closure_UNIX64_inner): New.
2484 * src/x86/unix64.S (ffi_closure_UNIX64): New.
2486 2003-01-27 Alexandre Oliva <aoliva@redhat.com>
2488 * configure.in (toolexecdir, toolexeclibdir): Set and AC_SUBST.
2489 Remove USE_LIBDIR conditional.
2490 * Makefile.am (toolexecdir, toolexeclibdir): Don't override.
2491 * Makefile.in, configure: Rebuilt.
2493 2003-01027 David Edelsohn <edelsohn@gnu.org>
2495 * Makefile.am (TARGET_SRC_POWERPC_AIX): Fix typo.
2496 * Makefile.in: Regenerate.
2498 2003-01-22 Andrew Haley <aph@redhat.com>
2500 * src/powerpc/darwin.S (_ffi_call_AIX): Add Augmentation size to
2503 2003-01-21 Andreas Tobler <a.tobler@schweiz.ch>
2505 * src/powerpc/darwin.S: Add unwind info.
2506 * src/powerpc/darwin_closure.S: Likewise.
2508 2003-01-14 Andrew Haley <aph@redhat.com>
2510 * src/x86/ffi64.c (ffi_prep_args): Check for void retval.
2511 (ffi_prep_cif_machdep): Likewise.
2512 * src/x86/unix64.S: Add unwind info.
2514 2003-01-14 Andreas Jaeger <aj@suse.de>
2516 * src/ffitest.c (main): Only use ffi_closures if those are
2519 2003-01-13 Andreas Tobler <a.tobler@schweiz.ch>
2521 * libffi/src/ffitest.c
2522 add closure testcases
2524 2003-01-13 Kevin B. Hendricks <khendricks@ivey.uwo.ca>
2526 * libffi/src/powerpc/ffi.c
2527 fix alignment bug for float (4 byte aligned iso 8 byte)
2529 2003-01-09 Geoffrey Keating <geoffk@apple.com>
2531 * src/powerpc/ffi_darwin.c: Remove RCS version string.
2532 * src/powerpc/darwin.S: Remove RCS version string.
2534 2003-01-03 Jeff Sturm <jsturm@one-point.com>
2536 * include/ffi.h.in: Add closure defines for SPARC, SPARC64.
2537 * src/ffitest.c (main): Use static storage for closure.
2538 * src/sparc/ffi.c (ffi_prep_closure, ffi_closure_sparc_inner): New.
2539 * src/sparc/v8.S (ffi_closure_v8): New.
2540 * src/sparc/v9.S (ffi_closure_v9): New.
2542 2002-11-10 Ranjit Mathew <rmathew@hotmail.com>
2544 * include/ffi.h.in: Added FFI_STDCALL ffi_type
2545 enumeration for X86_WIN32.
2546 * src/x86/win32.S: Added ffi_call_STDCALL function
2548 * src/x86/ffi.c (ffi_call/ffi_raw_call): Added
2549 switch cases for recognising FFI_STDCALL and
2550 calling ffi_call_STDCALL if target is X86_WIN32.
2551 * src/ffitest.c (my_stdcall_strlen/stdcall_many):
2552 stdcall versions of the "my_strlen" and "many"
2553 test functions (for X86_WIN32).
2554 Added test cases to test stdcall invocation using
2557 2002-12-02 Kaz Kojima <kkojima@gcc.gnu.org>
2559 * src/sh/sysv.S: Add DWARF2 unwind info.
2561 2002-11-27 Ulrich Weigand <uweigand@de.ibm.com>
2563 * src/s390/sysv.S (.eh_frame section): Make section read-only.
2565 2002-11-26 Jim Wilson <wilson@redhat.com>
2567 * src/types.c (FFI_TYPE_POINTER): Has size 8 on IA64.
2569 2002-11-23 H.J. Lu <hjl@gnu.org>
2571 * acinclude.m4: Add dummy AM_PROG_LIBTOOL.
2572 Include ../config/accross.m4.
2573 * aclocal.m4; Rebuild.
2574 * configure: Likewise.
2576 2002-11-15 Ulrich Weigand <uweigand@de.ibm.com>
2578 * src/s390/sysv.S (.eh_frame section): Adapt to pcrel FDE encoding.
2580 2002-11-11 DJ Delorie <dj@redhat.com>
2582 * configure.in: Look for common files in the right place.
2584 2002-10-08 Ulrich Weigand <uweigand@de.ibm.com>
2586 * src/java_raw_api.c (ffi_java_raw_to_ptrarray): Interpret
2587 raw data as _Jv_word values, not ffi_raw.
2588 (ffi_java_ptrarray_to_raw): Likewise.
2589 (ffi_java_rvalue_to_raw): New function.
2590 (ffi_java_raw_call): Call it.
2591 (ffi_java_raw_to_rvalue): New function.
2592 (ffi_java_translate_args): Call it.
2593 * src/ffitest.c (closure_test_fn): Interpret return value
2594 as ffi_arg, not int.
2595 * src/s390/ffi.c (ffi_prep_cif_machdep): Add missing
2596 FFI_TYPE_POINTER case.
2597 (ffi_closure_helper_SYSV): Likewise. Also, assume return
2598 values extended to word size.
2600 2002-10-02 Andreas Jaeger <aj@suse.de>
2602 * src/x86/ffi64.c (ffi_prep_cif_machdep): Remove debug output.
2604 2002-10-01 Bo Thorsen <bo@smetana.suse.de>
2606 * include/ffi.h.in: Fix i386 win32 compilation.
2608 2002-09-30 Ulrich Weigand <uweigand@de.ibm.com>
2610 * configure.in: Add s390x-*-linux-* target.
2611 * configure: Regenerate.
2612 * include/ffi.h.in: Define S390X for s390x targets.
2613 (FFI_CLOSURES): Define for s390/s390x.
2614 (FFI_TRAMPOLINE_SIZE): Likewise.
2615 (FFI_NATIVE_RAW_API): Likewise.
2616 * src/prep_cif.c (ffi_prep_cif): Do not compute stack space for s390.
2617 * src/types.c (FFI_TYPE_POINTER): Use 8-byte pointers on s390x.
2618 * src/s390/ffi.c: Major rework of existing code. Add support for
2619 s390x targets. Add closure support.
2620 * src/s390/sysv.S: Likewise.
2622 2002-09-29 Richard Earnshaw <rearnsha@arm.com>
2624 * src/arm/sysv.S: Fix typo.
2626 2002-09-28 Richard Earnshaw <rearnsha@arm.com>
2628 * src/arm/sysv.S: If we don't have machine/asm.h and the pre-processor
2629 has defined __USER_LABEL_PREFIX__, then use it in CNAME.
2630 (ffi_call_SYSV): Handle soft-float.
2632 2002-09-27 Bo Thorsen <bo@suse.de>
2634 * include/ffi.h.in: Fix multilib x86-64 support.
2636 2002-09-22 Kaveh R. Ghazi <ghazi@caip.rutgers.edu>
2638 * Makefile.am (all-multi): Fix multilib parallel build.
2640 2002-07-19 Kaz Kojima <kkojima@gcc.gnu.org>
2642 * configure.in (sh[34]*-*-linux*): Add brackets.
2643 * configure: Regenerate.
2645 2002-07-18 Kaz Kojima <kkojima@gcc.gnu.org>
2647 * Makefile.am: Add SH support.
2648 * Makefile.in: Regenerate.
2649 * configure.in (sh-*-linux*, sh[34]*-*-linux*): Add target.
2650 * configure: Regenerate.
2651 * include/ffi.h.in: Add SH support.
2652 * src/sh/ffi.c: New file.
2653 * src/sh/sysv.S: New file.
2654 * src/types.c: Add SH support.
2656 2002-07-16 Bo Thorsen <bo@suse.de>
2658 * src/x86/ffi64.c: New file that adds x86-64 support.
2659 * src/x86/unix64.S: New file that handles argument setup for
2661 * src/x86/sysv.S: Don't use this on x86-64.
2662 * src/x86/ffi.c: Don't use this on x86-64.
2664 * src/prep_cif.c (ffi_prep_cif): Don't do stack size calculation
2666 * src/ffitest.c (struct6): New test that tests a special case in
2668 (struct7): Likewise.
2669 (struct8): Likewise.
2670 (struct9): Likewise.
2671 (closure_test_fn): Silence warning about this when it's not used.
2672 (main): Add the new tests.
2673 (main): Fix a couple of wrong casts and silence some compiler warnings.
2674 * include/ffi.h.in: Add x86-64 ABI definition.
2675 * fficonfig.h.in: Regenerate.
2676 * Makefile.am: Add x86-64 support.
2677 * configure.in: Likewise.
2678 * Makefile.in: Regenerate.
2679 * configure: Likewise.
2681 2002-06-24 Bo Thorsen <bo@suse.de>
2683 * src/types.c: Merge settings for similar architectures.
2684 Add x86-64 sizes and alignments.
2686 2002-06-23 Bo Thorsen <bo@suse.de>
2688 * src/arm/ffi.c (ffi_prep_args): Remove unused vars.
2689 * src/sparc/ffi.c (ffi_prep_args_v8): Likewise.
2690 * src/mips/ffi.c (ffi_prep_args): Likewise.
2691 * src/m68k/ffi.c (ffi_prep_args): Likewise.
2693 2002-07-18 H.J. Lu (hjl@gnu.org)
2695 * Makefile.am (TARGET_SRC_MIPS_LINUX): New.
2696 (libffi_la_SOURCES): Support MIPS_LINUX.
2697 (libffi_convenience_la_SOURCES): Likewise.
2698 * Makefile.in: Regenerated.
2700 * configure.in (mips64*-*): Skip.
2701 (mips*-*-linux*): New.
2702 * configure: Regenerated.
2704 * src/mips/ffi.c: Include <sgidefs.h>.
2706 2002-06-06 Ulrich Weigand <uweigand@de.ibm.com>
2708 * src/s390/sysv.S: Save/restore %r6. Add DWARF-2 unwind info.
2710 2002-05-27 Roger Sayle <roger@eyesopen.com>
2712 * src/x86/ffi.c (ffi_prep_args): Remove reference to avn.
2714 2002-05-27 Bo Thorsen <bo@suse.de>
2716 * src/x86/ffi.c (ffi_prep_args): Remove unused variable and
2719 2002-05-13 Andreas Tobler <a.tobler@schweiz.ch>
2721 * src/powerpc/ffi_darwin.c (ffi_prep_closure): Declare fd at
2722 beginning of function (for older apple cc).
2724 2002-05-08 Alexandre Oliva <aoliva@redhat.com>
2726 * configure.in (ORIGINAL_LD_FOR_MULTILIBS): Preserve LD at
2727 script entry, and set LD to it when configuring multilibs.
2728 * configure: Rebuilt.
2730 2002-05-05 Jason Thorpe <thorpej@wasabisystems.com>
2732 * configure.in (sparc64-*-netbsd*): Add target.
2733 (sparc-*-netbsdelf*): Likewise.
2734 * configure: Regenerate.
2736 2002-04-28 David S. Miller <davem@redhat.com>
2738 * configure.in, configure: Fix SPARC test in previous change.
2740 2002-04-29 Gerhard Tonn <GerhardTonn@swol.de>
2742 * Makefile.am: Add Linux for S/390 support.
2743 * Makefile.in: Regenerate.
2744 * configure.in: Add Linux for S/390 support.
2745 * configure: Regenerate.
2746 * include/ffi.h.in: Add Linux for S/390 support.
2747 * src/s390/ffi.c: New file from libffi CVS tree.
2748 * src/s390/sysv.S: New file from libffi CVS tree.
2750 2002-04-28 Jakub Jelinek <jakub@redhat.com>
2752 * configure.in (HAVE_AS_SPARC_UA_PCREL): Check for working
2754 * src/sparc/v8.S: Use it.
2755 * src/sparc/v9.S: Likewise.
2756 * fficonfig.h.in: Rebuilt.
2757 * configure: Rebuilt.
2759 2002-04-08 Hans Boehm <Hans_Boehm@hp.com>
2761 * src/java_raw_api.c (ffi_java_raw_size): Handle FFI_TYPE_DOUBLE
2763 * src/ia64/unix.S: Add unwind information. Fix comments.
2764 Save sp in a way that's compatible with unwind info.
2765 (ffi_call_unix): Correctly restore sp in all cases.
2766 * src/ia64/ffi.c: Add, fix comments.
2768 2002-04-08 Jakub Jelinek <jakub@redhat.com>
2770 * src/sparc/v8.S: Make .eh_frame dependent on target word size.
2772 2002-04-06 Jason Thorpe <thorpej@wasabisystems.com>
2774 * configure.in (alpha*-*-netbsd*): Add target.
2775 * configure: Regenerate.
2777 2002-04-04 Jeff Sturm <jsturm@one-point.com>
2779 * src/sparc/v8.S: Add unwind info.
2780 * src/sparc/v9.S: Likewise.
2782 2002-03-30 Krister Walfridsson <cato@df.lth.se>
2784 * configure.in: Enable i*86-*-netbsdelf*.
2785 * configure: Rebuilt.
2787 2002-03-29 David Billinghurst <David.Billinghurst@riotinto.com>
2790 * src/mips/n32.s: Delete
2791 * src/mips/o32.s: Delete
2793 2002-03-21 Loren J. Rittle <ljrittle@acm.org>
2795 * configure.in: Enable alpha*-*-freebsd*.
2796 * configure: Rebuilt.
2798 2002-03-17 Bryce McKinlay <bryce@waitaki.otago.ac.nz>
2800 * Makefile.am: libfficonvenience -> libffi_convenience.
2801 * Makefile.in: Rebuilt.
2803 * Makefile.am: Define ffitest_OBJECTS.
2804 * Makefile.in: Rebuilt.
2806 2002-03-07 Andreas Tobler <toa@pop.agri.ch>
2807 David Edelsohn <edelsohn@gnu.org>
2809 * Makefile.am (EXTRA_DIST): Add Darwin and AIX closure files.
2810 (TARGET_SRC_POWERPC_AIX): Add aix_closure.S.
2811 (TARGET_SRC_POWERPC_DARWIN): Add darwin_closure.S.
2812 * Makefile.in: Regenerate.
2813 * include/ffi.h.in: Add AIX and Darwin closure definitions.
2814 * src/powerpc/ffi_darwin.c (ffi_prep_closure): New function.
2815 (flush_icache, flush_range): New functions.
2816 (ffi_closure_helper_DARWIN): New function.
2817 * src/powerpc/aix_closure.S: New file.
2818 * src/powerpc/darwin_closure.S: New file.
2820 2002-02-24 Jeff Sturm <jsturm@one-point.com>
2822 * include/ffi.h.in: Add typedef for ffi_arg.
2823 * src/ffitest.c (main): Declare rint with ffi_arg.
2825 2002-02-21 Andreas Tobler <toa@pop.agri.ch>
2827 * src/powerpc/ffi_darwin.c (ffi_prep_args): Skip appropriate
2828 number of GPRs for floating-point arguments.
2830 2002-01-31 Anthony Green <green@redhat.com>
2832 * configure: Rebuilt.
2833 * configure.in: Replace CHECK_SIZEOF and endian tests with
2834 cross-compiler friendly macros.
2835 * aclocal.m4 (AC_COMPILE_CHECK_SIZEOF, AC_C_BIGENDIAN_CROSS): New
2838 2002-01-18 David Edelsohn <edelsohn@gnu.org>
2840 * src/powerpc/darwin.S (_ffi_call_AIX): New.
2841 * src/powerpc/aix.S (ffi_call_DARWIN): New.
2843 2002-01-17 David Edelsohn <edelsohn@gnu.org>
2845 * Makefile.am (EXTRA_DIST): Add Darwin and AIX files.
2846 (TARGET_SRC_POWERPC_AIX): New.
2847 (POWERPC_AIX): New stanza.
2848 * Makefile.in: Regenerate.
2849 * configure.in: Add AIX case.
2850 * configure: Regenerate.
2851 * include/ffi.h.in (ffi_abi): Add FFI_AIX.
2852 * src/powerpc/ffi_darwin.c (ffi_status): Use "long" to scale frame
2853 size. Fix "long double" support.
2854 (ffi_call): Add FFI_AIX case.
2855 * src/powerpc/aix.S: New.
2857 2001-10-09 John Hornkvist <john@toastedmarshmallow.com>
2859 Implement Darwin PowerPC ABI.
2860 * configure.in: Handle powerpc-*-darwin*.
2861 * Makefile.am: Set source files for POWERPC_DARWIN.
2862 * configure: Rebuilt.
2863 * Makefile.in: Rebuilt.
2864 * include/ffi.h.in: Define FFI_DARWIN and FFI_DEFAULT_ABI for
2866 * src/powerpc/darwin.S: New file.
2867 * src/powerpc/ffi_darwin.c: New file.
2869 2001-10-07 Joseph S. Myers <jsm28@cam.ac.uk>
2871 * src/x86/ffi.c: Fix spelling error of "separate" as "seperate".
2873 2001-07-16 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
2875 * src/x86/sysv.S: Avoid gas-only .balign directive.
2876 Use C style comments.
2878 2001-07-16 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
2880 * src/alpha/ffi.c (ffi_prep_closure): Avoid gas-only mnemonic.
2881 Fixes PR bootstrap/3563.
2883 2001-06-26 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
2885 * src/alpha/osf.S (ffi_closure_osf): Use .rdata for ECOFF.
2887 2001-06-25 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
2889 * configure.in: Recognize sparc*-sun-* host.
2890 * configure: Regenerate.
2892 2001-06-06 Andrew Haley <aph@redhat.com>
2894 * src/alpha/osf.S (__FRAME_BEGIN__): Conditionalize for ELF.
2896 2001-06-03 Andrew Haley <aph@redhat.com>
2898 * src/alpha/osf.S: Add unwind info.
2899 * src/powerpc/sysv.S: Add unwind info.
2900 * src/powerpc/ppc_closure.S: Likewise.
2902 2000-05-31 Jeff Sturm <jsturm@one-point.com>
2904 * configure.in: Fix AC_ARG_ENABLE usage.
2905 * configure: Rebuilt.
2907 2001-05-06 Bryce McKinlay <bryce@waitaki.otago.ac.nz>
2909 * configure.in: Remove warning about beta code.
2910 * configure: Rebuilt.
2912 2001-04-25 Hans Boehm <Hans_Boehm@hp.com>
2914 * src/ia64/unix.S: Restore stack pointer when returning from
2916 * src/ia64/ffi.c: Fix typo in comment.
2918 2001-04-18 Jim Wilson <wilson@redhat.com>
2920 * src/ia64/unix.S: Delete unnecessary increment and decrement of loc2
2921 to eliminate RAW DV.
2923 2001-04-12 Bryce McKinlay <bryce@albatross.co.nz>
2925 * Makefile.am: Make a libtool convenience library.
2926 * Makefile.in: Rebuilt.
2928 2001-03-29 Bryce McKinlay <bryce@albatross.co.nz>
2930 * configure.in: Use different syntax for subdirectory creation.
2931 * configure: Rebuilt.
2933 2001-03-27 Jon Beniston <jon@beniston.com>
2935 * configure.in: Added X86_WIN32 target (Win32, CygWin, MingW).
2936 * configure: Rebuilt.
2937 * Makefile.am: Added X86_WIN32 target support.
2938 * Makefile.in: Rebuilt.
2940 * include/ffi.h.in: Added X86_WIN32 target support.
2942 * src/ffitest.c: Doesn't run structure tests for X86_WIN32 targets.
2943 * src/types.c: Added X86_WIN32 target support.
2945 * src/x86/win32.S: New file. Based on sysv.S, but with EH
2946 stuff removed and made to work with CygWin's gas.
2948 2001-03-26 Bryce McKinlay <bryce@albatross.co.nz>
2950 * configure.in: Make target subdirectory in build dir.
2951 * Makefile.am: Override suffix based rules to specify correct output
2953 * Makefile.in: Rebuilt.
2954 * configure: Rebuilt.
2956 2001-03-23 Kevin B Hendricks <khendricks@ivey.uwo.ca>
2958 * src/powerpc/ppc_closure.S: New file.
2959 * src/powerpc/ffi.c (ffi_prep_args): Fixed ABI compatibility bug
2960 involving long long and register pairs.
2961 (ffi_prep_closure): New function.
2962 (flush_icache): Likewise.
2963 (ffi_closure_helper_SYSV): Likewise.
2964 * include/ffi.h.in (FFI_CLOSURES): Define on PPC.
2965 (FFI_TRAMPOLINE_SIZE): Likewise.
2966 (FFI_NATIVE_RAW_API): Likewise.
2967 * Makefile.in: Rebuilt.
2968 * Makefile.am (EXTRA_DIST): Added src/powerpc/ppc_closure.S.
2969 (TARGET_SRC_POWERPC): Likewise.
2971 2001-03-19 Tom Tromey <tromey@redhat.com>
2973 * Makefile.in: Rebuilt.
2974 * Makefile.am (ffitest_LDFLAGS): New macro.
2976 2001-03-02 Nick Clifton <nickc@redhat.com>
2978 * include/ffi.h.in: Remove RCS ident string.
2979 * include/ffi_mips.h: Remove RCS ident string.
2980 * src/debug.c: Remove RCS ident string.
2981 * src/ffitest.c: Remove RCS ident string.
2982 * src/prep_cif.c: Remove RCS ident string.
2983 * src/types.c: Remove RCS ident string.
2984 * src/alpha/ffi.c: Remove RCS ident string.
2985 * src/alpha/osf.S: Remove RCS ident string.
2986 * src/arm/ffi.c: Remove RCS ident string.
2987 * src/arm/sysv.S: Remove RCS ident string.
2988 * src/mips/ffi.c: Remove RCS ident string.
2989 * src/mips/n32.S: Remove RCS ident string.
2990 * src/mips/o32.S: Remove RCS ident string.
2991 * src/sparc/ffi.c: Remove RCS ident string.
2992 * src/sparc/v8.S: Remove RCS ident string.
2993 * src/sparc/v9.S: Remove RCS ident string.
2994 * src/x86/ffi.c: Remove RCS ident string.
2995 * src/x86/sysv.S: Remove RCS ident string.
2997 2001-02-08 Joseph S. Myers <jsm28@cam.ac.uk>
2999 * include/ffi.h.in: Change sourceware.cygnus.com references to
3002 2000-12-09 Richard Henderson <rth@redhat.com>
3004 * src/alpha/ffi.c (ffi_call): Simplify struct return test.
3005 (ffi_closure_osf_inner): Index rather than increment avalue
3006 and arg_types. Give ffi_closure_osf the raw return value type.
3007 * src/alpha/osf.S (ffi_closure_osf): Handle return value type
3010 2000-12-07 Richard Henderson <rth@redhat.com>
3012 * src/raw_api.c (ffi_translate_args): Fix typo.
3013 (ffi_prep_closure): Likewise.
3015 * include/ffi.h.in [ALPHA]: Define FFI_CLOSURES and
3016 FFI_TRAMPOLINE_SIZE.
3017 * src/alpha/ffi.c (ffi_prep_cif_machdep): Adjust minimal
3018 cif->bytes for new ffi_call_osf implementation.
3019 (ffi_prep_args): Absorb into ...
3020 (ffi_call): ... here. Do all stack allocation here and
3021 avoid a callback function.
3022 (ffi_prep_closure, ffi_closure_osf_inner): New.
3023 * src/alpha/osf.S (ffi_call_osf): Reimplement with no callback.
3024 (ffi_closure_osf): New.
3026 2000-09-10 Alexandre Oliva <aoliva@redhat.com>
3028 * config.guess, config.sub, install-sh: Removed.
3029 * ltconfig, ltmain.sh, missing, mkinstalldirs: Likewise.
3030 * Makefile.in: Rebuilt.
3032 * acinclude.m4: Include libtool macros from the top level.
3033 * aclocal.m4, configure: Rebuilt.
3035 2000-08-22 Alexandre Oliva <aoliva@redhat.com>
3037 * configure.in [i*86-*-freebsd*] (TARGET, TARGETDIR): Set.
3038 * configure: Rebuilt.
3040 2000-05-11 Scott Bambrough <scottb@netwinder.org>
3042 * libffi/src/arm/sysv.S (ffi_call_SYSV): Doubles are not saved to
3043 memory correctly. Use conditional instructions, not branches where
3046 2000-05-04 Tom Tromey <tromey@cygnus.com>
3048 * configure: Rebuilt.
3049 * configure.in: Match `arm*-*-linux-*'.
3050 From Chris Dornan <cdornan@arm.com>.
3052 2000-04-28 Jakub Jelinek <jakub@redhat.com>
3054 * Makefile.am (SUBDIRS): Define.
3055 (AM_MAKEFLAGS): Likewise.
3056 (Multilib support.): Add section.
3057 * Makefile.in: Rebuilt.
3058 * ltconfig (extra_compiler_flags, extra_compiler_flags_value):
3059 New variables. Set for gcc using -print-multi-lib. Export them
3061 (sparc64-*-linux-gnu*): Use libsuff 64 for search paths.
3062 * ltmain.sh (B|b|V): Don't throw away gcc's -B, -b and -V options
3064 (extra_compiler_flags_value, extra_compiler_flags): Check these
3065 for extra compiler options which need to be passed down in
3068 2000-04-16 Anthony Green <green@redhat.com>
3070 * configure: Rebuilt.
3071 * configure.in: Change i*86-pc-linux* to i*86-*-linux*.
3073 2000-04-14 Jakub Jelinek <jakub@redhat.com>
3075 * include/ffi.h.in (SPARC64): Define for 64bit SPARC builds.
3076 Set SPARC FFI_DEFAULT_ABI based on SPARC64 define.
3077 * src/sparc/ffi.c (ffi_prep_args_v8): Renamed from ffi_prep_args.
3078 Replace all void * sizeofs with sizeof(int).
3079 Only compare type with FFI_TYPE_LONGDOUBLE if LONGDOUBLE is
3080 different than DOUBLE.
3081 Remove FFI_TYPE_SINT32 and FFI_TYPE_UINT32 cases (handled elsewhere).
3082 (ffi_prep_args_v9): New function.
3083 (ffi_prep_cif_machdep): Handle V9 ABI and long long on V8.
3084 (ffi_V9_return_struct): New function.
3085 (ffi_call): Handle FFI_V9 ABI from 64bit code and FFI_V8 ABI from
3086 32bit code (not yet cross-arch calls).
3087 * src/sparc/v8.S: Add struct return delay nop.
3089 * src/sparc/v9.S: New file.
3090 * src/prep_cif.c (ffi_prep_cif): Return structure pointer
3091 is used on sparc64 only for structures larger than 32 bytes.
3092 Pass by reference for structures is done for structure arguments
3093 larger than 16 bytes.
3094 * src/ffitest.c (main): Use 64bit rint on sparc64.
3095 Run long long tests on sparc.
3096 * src/types.c (FFI_TYPE_POINTER): Pointer is 64bit on alpha and
3098 (FFI_TYPE_LONGDOUBLE): long double is 128 bit aligned to 128 bits
3100 * configure.in (sparc-*-linux*): New supported target.
3101 (sparc64-*-linux*): Likewise.
3102 * configure: Rebuilt.
3103 * Makefile.am: Add v9.S to SPARC files.
3104 * Makefile.in: Likewise.
3105 (LINK): Surround $(CCLD) into double quotes, so that multilib
3106 compiles work correctly.
3108 2000-04-04 Alexandre Petit-Bianco <apbianco@cygnus.com>
3110 * configure: Rebuilt.
3111 * configure.in: (i*86-*-solaris*): New libffi target. Patch
3112 proposed by Bryce McKinlay.
3114 2000-03-20 Tom Tromey <tromey@cygnus.com>
3116 * Makefile.in: Hand edit for java_raw_api.lo.
3118 2000-03-08 Bryce McKinlay <bryce@albatross.co.nz>
3120 * config.guess, config.sub: Update from the gcc tree.
3121 Fix for PR libgcj/168.
3123 2000-03-03 Tom Tromey <tromey@cygnus.com>
3125 * Makefile.in: Fixed ia64 by hand.
3127 * configure: Rebuilt.
3128 * configure.in (--enable-multilib): New option.
3129 (libffi_basedir): New subst.
3130 (AC_OUTPUT): Added multilib code.
3132 2000-03-02 Tom Tromey <tromey@cygnus.com>
3134 * Makefile.in: Rebuilt.
3135 * Makefile.am (TARGET_SRC_IA64): Use `ia64', not `alpha', as
3138 2000-02-25 Hans Boehm <boehm@acm.org>
3140 * src/ia64/ffi.c, src/ia64/ia64_flags.h, src/ia64/unix.S: New
3142 * src/raw_api.c (ffi_translate_args): Fixed typo in argument
3144 (ffi_prep_raw_closure): Use ffi_translate_args, not
3145 ffi_closure_translate.
3146 * src/java_raw_api.c: New file.
3147 * src/ffitest.c (closure_test_fn): New function.
3148 (main): Define `rint' as long long on IA64. Added new test when
3149 FFI_CLOSURES is defined.
3150 * include/ffi.h.in (ALIGN): Use size_t, not unsigned.
3151 (ffi_abi): Recognize IA64.
3152 (ffi_raw): Added `flt' field.
3153 Added "Java raw API" code.
3154 * configure.in: Recognize ia64.
3155 * Makefile.am (TARGET_SRC_IA64): New macro.
3156 (libffi_la_common_SOURCES): Added java_raw_api.c.
3157 (libffi_la_SOURCES): Define in IA64 case.
3159 2000-01-04 Tom Tromey <tromey@cygnus.com>
3161 * Makefile.in: Rebuilt with newer automake.
3163 1999-12-31 Tom Tromey <tromey@cygnus.com>
3165 * Makefile.am (INCLUDES): Added -I$(top_srcdir)/src.
3167 1999-09-01 Tom Tromey <tromey@cygnus.com>
3169 * include/ffi.h.in: Removed PACKAGE and VERSION defines and
3171 * fficonfig.h.in: Rebuilt.
3172 * configure: Rebuilt.
3173 * configure.in: Pass 3rd argument to AM_INIT_AUTOMAKE.
3174 Use AM_PROG_LIBTOOL (automake 1.4 compatibility).
3175 * acconfig.h: Don't #undef PACKAGE or VERSION.
3177 1999-08-09 Anthony Green <green@cygnus.com>
3179 * include/ffi.h.in: Try to work around messy header problem
3180 with PACKAGE and VERSION.
3182 * configure: Rebuilt.
3183 * configure.in: Change version to 2.00-beta.
3185 * fficonfig.h.in: Rebuilt.
3186 * acconfig.h (FFI_NO_STRUCTS, FFI_NO_RAW_API): Define.
3188 * src/x86/ffi.c (ffi_raw_call): Rename.
3190 1999-08-02 Kresten Krab Thorup <krab@dominiq.is.s.u-tokyo.ac.jp>
3192 * src/x86/ffi.c (ffi_closure_SYSV): New function.
3193 (ffi_prep_incoming_args_SYSV): Ditto.
3194 (ffi_prep_closure): Ditto.
3195 (ffi_closure_raw_SYSV): Ditto.
3196 (ffi_prep_raw_closure): More ditto.
3197 (ffi_call_raw): Final ditto.
3199 * include/ffi.h.in: Add definitions for closure and raw API.
3201 * src/x86/ffi.c (ffi_prep_cif_machdep): Added case for
3204 * Makefile.am (libffi_la_common_SOURCES): Added raw_api.c
3206 * src/raw_api.c: New file.
3208 * include/ffi.h.in (ffi_raw): New type.
3209 (UINT_ARG, SINT_ARG): New defines.
3210 (ffi_closure, ffi_raw_closure): New types.
3211 (ffi_prep_closure, ffi_prep_raw_closure): New declarations.
3213 * configure.in: Add check for endianness and sizeof void*.
3215 * src/x86/sysv.S (ffi_call_SYSV): Call fixup routine via argument,
3216 instead of directly.
3218 * configure: Rebuilt.
3220 Thu Jul 8 14:28:42 1999 Anthony Green <green@cygnus.com>
3222 * configure.in: Add x86 and powerpc BeOS configurations.
3223 From Makoto Kato <m_kato@ga2.so-net.ne.jp>.
3225 1999-05-09 Anthony Green <green@cygnus.com>
3227 * configure.in: Add warning about this being beta code.
3228 Remove src/Makefile.am from the picture.
3229 * configure: Rebuilt.
3231 * Makefile.am: Move logic from src/Makefile.am. Add changes
3232 to support libffi as a target library.
3233 * Makefile.in: Rebuilt.
3235 * aclocal.m4, config.guess, config.sub, ltconfig, ltmain.sh:
3236 Upgraded to new autoconf, automake, libtool.
3240 * LICENSE: Update copyright date.
3242 * src/Makefile.am, src/Makefile.in: Removed.
3244 1998-11-29 Anthony Green <green@cygnus.com>
3246 * include/ChangeLog: Removed.
3247 * src/ChangeLog: Removed.
3248 * src/mips/ChangeLog: Removed.
3249 * src/sparc/ChangeLog: Remboved.
3250 * src/x86/ChangeLog: Removed.
3252 * ChangeLog.v1: Created.