Commit | Line | Data |
---|---|---|
355afbcd KR |
1 | Mon Nov 23 11:50:00 1992 Ken Raeburn (raeburn@cygnus.com) |
2 | ||
3 | * obj-coff.c (obj_coff_endef): Use as_warn, not fprintf. | |
4 | ||
5 | * tc-m68k.c (md_assemble): Don't complain about 68000 with 68881; | |
6 | could be doing emulation. | |
7 | ||
587c4264 ILT |
8 | Tue Nov 10 09:50:25 1992 Ian Lance Taylor (ian@cygnus.com) |
9 | ||
0069b1f6 ILT |
10 | * tc-m68k.c (m68k_reg_parse): If REGISTER_PREFIX isn't defined, |
11 | still accept (but don't require) OPTIONAL_REGISTER_PREFIX before | |
12 | the register name. | |
b62a08d6 ILT |
13 | (insert_reg): put REGISTER_PREFIX before register names before |
14 | putting them in the symbol table. | |
0069b1f6 ILT |
15 | * tc-m68k.h (OPTIONAL_REGISTER_PREFIX): Define to be "%", if not |
16 | M68KCOFF. | |
17 | ||
587c4264 ILT |
18 | * obj-coffbfd.c (fill_section): set STYP_NOLOAD bit for .bss |
19 | section. | |
20 | ||
21 | * atof-ieee.c, atof-ns32k.c, tc-*.c: made EXP_CHARS, FLT_CHARS, | |
22 | comment_chars, line_comment_chars and line_seperator_chars | |
23 | consistently const, and always initialized them. Included read.h. | |
24 | ||
0cef0e20 JW |
25 | Thu Nov 5 17:55:41 1992 Jim Wilson (wilson@sphagnum.cygnus.com) |
26 | ||
27 | * tc-sparc.c (sparc_ip): Add code to flag error if an absolute | |
28 | constant will not fit in an immediate field. | |
29 | (md_apply_fix, RELOC_BASE13 case): Check for relocation overflow. | |
30 | ||
f3f45b10 KR |
31 | Wed Nov 4 07:50:46 1992 Ken Raeburn (raeburn@cygnus.com) |
32 | ||
33 | * obj-coff.c (callj_table): Delete global variable. | |
34 | (obj_emit_relocations): Define it locally here, and only if | |
35 | TC_I960 is defined. | |
36 | ||
37 | * tc-m68k.c (m68k_reg_parse): Underscore is part of a symbol name. | |
38 | (m68k_ip): Don't warn about bignum used as float bit-pattern. | |
39 | ||
40 | * obj-coff.c: Replaced ANSI and non-ANSI function declarations | |
41 | with a single set using PARAMS macro. | |
42 | ||
43 | * tc-i960.c (tc_bout_fix_to_chars): Bit-field fixups want a length | |
44 | of 2. | |
45 | ||
46 | * tc-i960.c: Missed a couple of 0->NO_RELOC conversions. | |
47 | ||
48 | * tc-i960.h (N_BALNAME, N_CALLNAME): Define as char-type values, | |
49 | so widening works consistently. | |
50 | ||
1045c20b ILT |
51 | Fri Oct 23 14:40:38 1992 Ian Lance Taylor (ian@cygnus.com) |
52 | ||
53 | * obj-coffbfd.c (write_object_file): check return value of | |
54 | bfd_close_all_done. | |
55 | ||
6142210d ILT |
56 | Tue Oct 20 13:02:25 1992 Ian Lance Taylor (ian@cygnus.com) |
57 | ||
58 | * Support for i386-sysv. | |
59 | obj-coffbfd.c (do_relocs_for, write_object_file): set segment | |
60 | addresses to reasonable sizes. New define ZERO_BASED_SEGMENTS can | |
61 | be used to set them all to zero as was done before. | |
62 | (fill_section): segment addresses now set in write_object_file. | |
63 | (fill_section): Don't set STYP_NOLOAD for .bss section. | |
64 | (fixup_segment): 386 uses strange common symbol format. | |
65 | tc-i386.c (tc_coff_fix2rtype): use R_DIR32, not R_RELLONG, for | |
66 | compatibility with SVR3.2 linker. | |
67 | ||
4f447678 SC |
68 | Mon Oct 5 09:28:57 1992 Steve Chamberlain (sac@thepub.cygnus.com) |
69 | ||
c116445e SC |
70 | fix i960+non-bfd coff bit rot. |
71 | * obj-coff.c (c_dot_file_symbol, obj_coff_ln, obj_coff_line): | |
72 | support for C source listings. (obj_coff_endef): look in the right | |
73 | part of the symbol for the symbol name | |
74 | ||
4f447678 SC |
75 | * tc-m68k.c (get_num): make it work for all segments, not just the |
76 | first three. | |
77 | ||
2f68cb54 SC |
78 | Tue Sep 29 12:22:52 1992 Steve Chamberlain (sac@thepub.cygnus.com) |
79 | ||
80 | * obj-coffbfd.c (write_object_file): don't fixup for the z8k | |
81 | * tc-z8k.c: lots of bug fixes | |
82 | ||
7c42ad77 ILT |
83 | Tue Sep 29 10:51:55 1992 Ian Lance Taylor (ian@cygnus.com) |
84 | ||
85 | * tc-i960.h, tc-i960.c: avoid the ANSI preprocessor addition | |
86 | #elif, since it is not supported by old compilers. | |
87 | ho-rs6000.h, tc-m68k.c: the native RS/6000 compiler miscompiles a | |
88 | couple of expressions in tc-m68k.c. | |
89 | ||
fdc6ba64 KR |
90 | Mon Sep 28 21:16:51 1992 Ken Raeburn (raeburn@cambridge.cygnus.com) |
91 | ||
92 | * tc-i960.c (get_cdisp): Use NO_RELOC, not 0, in call to fix_new. | |
93 | ||
fe86a817 ILT |
94 | Fri Sep 25 18:18:52 1992 Ian Lance Taylor (ian@cygnus.com) |
95 | ||
96 | * tc-m68k.h: if M68KCOFF, define DOT_LABEL_PREFIX (to require | |
97 | local labels to start with a .) and set REGISTER_PREFIX to %. | |
98 | tc-m68k.c (m68k_reg_parse): accept REGISTER_PREFIX if defined. | |
99 | ||
b066f445 SEF |
100 | Tue Sep 22 13:02:07 1992 Sean Eric Fagan (sef@cygnus.com) |
101 | ||
102 | * obj-coffbfd.c (do_relocs_for,fill_section): now allocate all | |
103 | sections starting from zero, rather than making them consecutive. | |
104 | This makes subsequent reloc calculations easier, esp if the object | |
105 | format doesn't understand addends. (obj_coff_lcomm): (maybe temporarily) | |
106 | allocate lcomm in .data rather than in .bss. It seems that some | |
107 | tools can't cope with a non-zero sized bss before linkage. | |
108 | ||
ae9c717e KR |
109 | Tue Sep 22 15:10:51 1992 Ken Raeburn (raeburn@cambridge.cygnus.com) |
110 | ||
111 | * tc-m68k.c: Replace "enum m68k_architecture" with "int" | |
112 | throughout. That enum no longer means what we thought it meant. | |
113 | ||
114 | * tc-m68k.c (md_assemble, md_parse_option): Handle new | |
115 | "-mno-688[58]1" options. | |
116 | ||
117 | * tc-m68k.c: Added CPU32 support. | |
118 | ||
04b07d16 SC |
119 | Fri Sep 18 08:02:18 1992 Steve Chamberlain (sac@thepub.cygnus.com) |
120 | ||
121 | * tc-m68k.c (m68k_ip): An(disp) is not pc relative. | |
122 | ||
76bf2d5d ILT |
123 | Sun Sep 13 20:29:39 1992 Ian Lance Taylor (ian@cygnus.com) |
124 | ||
e41474b7 ILT |
125 | * obj-coffbfd.h: define WARN_SIGNED_OVERFLOW_WORD. |
126 | obj-coffbfd.c (fixup_segment): check for signed .word overflow if | |
127 | WARN_SIGNED_OVERFLOW_WORD. | |
128 | ||
76bf2d5d ILT |
129 | * obj-coffbfd.c (fixup_segment): fixed missing parens in |
130 | expression checking for byte or word overflow. | |
131 | ||
28c8c50b SC |
132 | Fri Sep 11 10:21:04 1992 Steve Chamberlain (sac@thepub.cygnus.com) |
133 | ||
134 | Support for i386 coff | |
135 | * obj-coffbfd.h : added stuff | |
136 | * tc-i386.c (tc_coff_fix2rtype): new function | |
137 | * tc-i386.h : new coff defines | |
138 | ||
3ad9ec6a ILT |
139 | Wed Sep 9 11:10:24 1992 Ian Lance Taylor (ian@cygnus.com) |
140 | ||
141 | * tc-m68k.c (m68k_ip, m68k_ip_op, get_num, try_moto_index): merge | |
142 | Motorola and MIT syntax; gas can now assemble either type of | |
143 | file. | |
144 | tc-m68kmote.c, tc-m68kmote.h: removed now superfluous files. | |
145 | From Steve Chamberlain: | |
146 | m68kcoff.mt: for m68k COFF. | |
147 | obj-coffbfd.c: (fixup_mdeps) added | |
148 | (size_section) removed bad sanity check | |
149 | (fill_section) added rs_machine_dependent case | |
150 | (write_object_file) call fixup_mdeps | |
151 | (fixup_segment) set fx_subsy to 0. | |
152 | obj-coffbfd.h: define WORKING_DOT_WORD (too hard to support) and | |
153 | handle m68k. | |
154 | tc-m68k.c, config/tc-m68k.h: added m68k COFF support and Motorala | |
155 | pseudo ops. | |
156 | ||
8ae35e59 ILT |
157 | Fri Aug 28 16:25:22 1992 Ian Lance Taylor (ian@cygnus.com) |
158 | ||
159 | * obj-bout.h, obj-bout.c (obj_header_append, obj_symbol_to_chars), | |
160 | tc-i960.c (md_ri_to_chars): Always output bout object file in | |
161 | little endian byte order (used to use endianness of host). | |
162 | ||
163107a1 SC |
163 | Tue Aug 25 15:11:08 1992 Steve Chamberlain (sac@thepub.cygnus.com) |
164 | ||
165 | * z8k.c, z8k.h, z8k.mt: z8000 support stuff | |
166 | ||
148eb5dd ILT |
167 | Mon Aug 24 12:45:12 1992 Ian Lance Taylor (ian@cygnus.com) |
168 | ||
169 | * tc-m68k.c: use TARGET_CPU to choose default cpu type. | |
170 | ||
171 | * te-generic.h: default to LOCAL_LABELS_DOLLAR and LOCAL_LABELS_FB | |
172 | so that we can assemble hand-written libgcc code. | |
173 | ||
c50140c8 ILT |
174 | Wed Aug 19 11:20:59 1992 Ian Lance Taylor (ian@cygnus.com) |
175 | ||
176 | * tc-m68k.c, tc-m68kmote.c: the cas2 instruction is supposed to be | |
177 | written with indirection on the last two operands, which can be | |
178 | either data or address registers. Added a new operand type 'r' | |
179 | which accepts either register type. Added '(' to notend stuff in | |
180 | tc-m68kmote.c to accept (a0):(a2) in cas2 instruction. | |
181 | ||
0e7d3df7 | 182 | Tue Aug 11 12:58:14 1992 Ken Raeburn (raeburn@cygnus.com) |
f101ac57 KR |
183 | |
184 | * sparc.mt: New file. | |
185 | ||
bec66218 PB |
186 | Mon Aug 10 14:37:08 1992 Per Bothner (bothner@cygnus.com) |
187 | ||
188 | * tc-m68k.c: ".align N" means align to N-byte boundary *only* | |
189 | if TN_SUN3; otherwise align to 2**N-byte bounary. | |
190 | ||
1eaf4dd6 SC |
191 | Thu Aug 6 12:08:42 1992 Steve Chamberlain (sac@thepub.cygnus.com) |
192 | ||
193 | * config/tc-h8300.c: if a :8 is seen after an operand, fill top | |
194 | two bytes of any constant with 0xff: | |
195 | ||
934afcd4 JG |
196 | Wed Aug 5 01:54:34 1992 John Gilmore (gnu at cygnus.com) |
197 | ||
198 | * tc-m68k.c (try_index): Error if index scaling specified and | |
199 | assembling for an older CPU than a 68020. | |
200 | ||
f264b1f5 SC |
201 | Mon Jul 6 17:09:32 1992 Steve Chamberlain (sac@cygnus.com) |
202 | ||
203 | * obj-coffbfd.c (fill_section): mark .lit sections as STYP_LIT | |
204 | ||
a1765cf0 SC |
205 | Tue Jun 9 07:50:46 1992 Steve Chamberlain (sac@thepub.cygnus.com) |
206 | ||
207 | * tc-m68k.c (m68kip): Fix typo so that only arch's >=68020 do | |
208 | pcrel data stuff. (md_estimate_size_before_relax): when relaxing a | |
209 | 68010 bxx into a bra+6 jmpxx, put the bytes of the jmp opcode into | |
210 | the right place. (s_bss): Don't put .bss stuff into SEG_DATA, put | |
211 | it into SEG_BSS | |
212 | ||
a0fffd0d JW |
213 | Thu May 28 11:20:27 1992 Jim Wilson (wilson@sphagnum.cygnus.com) |
214 | ||
215 | * tc-sparc.c: Use new ARCHITECTURES_CONFLICT_P macro. Mention new | |
216 | -Asparclite flag. | |
217 | ||
c58dbabf SC |
218 | Mon May 4 18:56:19 1992 Steve Chamberlain (sac@thepub.cygnus.com) |
219 | ||
220 | * obj-coffbfd.c: use is a synonym for section, (do_relocs_for): | |
221 | calc the base of relocs correctly. | |
222 | * tc-a29k.c (parse_operand): allow expressions to be in any section. | |
c58dbabf | 223 | |
df7204ad RP |
224 | Thu Mar 12 04:42:38 1992 K. Richard Pixley (rich@cygnus.com) |
225 | ||
226 | * tc-m68k.h, te-sun3.h: moved LOCAL_LABELS_FB definition from | |
227 | tc-m68k.h to te-sun3.h. | |
228 | ||
229 | Wed Mar 11 23:34:46 1992 K. Richard Pixley (rich@cygnus.com) | |
230 | ||
231 | * te-sun3.h: add LOCAL_LABELS_DOLLAR. | |
232 | ||
1a9e2cf8 RP |
233 | Fri Feb 21 06:22:15 1992 K. Richard Pixley (rich@rtl.cygnus.com) |
234 | ||
235 | * obj-aout.c: do not include stab.gnu.h if NO_LISTING. | |
236 | ||
237 | * tc-i860.c, a.out.gnu.h: move i860 relocs to a proper place. | |
238 | ||
239 | * a.out.h: removed. | |
240 | ||
6d5460ab RP |
241 | Fri Feb 21 02:17:22 1992 Minh Tran-Le (TRANLE@INTELLICORP.COM) |
242 | ||
243 | * tc-i386.c: tc-i386.c: added handling of the following opcodes: | |
244 | i/o opcodes - inb, inw, outb and outw. string manipulation with | |
245 | att syntax - scmp, slod, smov, ssca, ssto. | |
246 | ||
247 | Fri Feb 21 01:53:50 1992 Minh Tran-Le (TRANLE@INTELLICORP.COM) | |
248 | ||
249 | * obj-coff.c: (for aix386) Moved the symbols .text, .data and .bss | |
250 | to just after .file . | |
251 | ||
252 | In obj_crawl_symbol_chain() where it tries to put the external | |
253 | symbols apart, with the condition: | |
254 | (!S_IS_DEFINED(symbolP) && | |
255 | !S_IS_DEBUG(symbolP) && | |
256 | !SF_GET_STATICS(symbolP)) | |
257 | it was moving too many symbols out. So I switch it back to the | |
258 | condition: | |
259 | (S_GET_STORAGE_CLASS(symbolP) == C_EXT && !SF_GET_FUNCTION(symbolP)) | |
260 | ||
261 | In obj_emit_relocations() added the conditional on KEEP_RELOC_INFO | |
262 | so that we don't use the F_RELFLG which make the linker complain | |
263 | that somebody has stripped the relocation info. | |
264 | ||
265 | Also, the AIX ld program require that the relocation table | |
266 | is sorted by r_vaddr like the standard ATT assembler does. | |
267 | ||
268 | [he also changed the sizeof(struct ...)'s into the coff | |
269 | style FOOSZ macros. I'm not sure this is right, but I can't | |
270 | remember why. xoxorich.] | |
271 | ||
e6a125b1 RP |
272 | Tue Feb 18 02:11:10 1992 K. Richard Pixley (rich at cygnus.com) |
273 | ||
90d85bc4 RP |
274 | * tc-m88k.[hc]: pulled in from hack's unfinished work. These |
275 | aren't yet integrated. | |
276 | ||
4b83ef1c RP |
277 | * tc-i860.[hc]: blew off the dust. Something must still be done |
278 | about conflicting relocation types. | |
279 | ||
e6a125b1 RP |
280 | * tc-ns32k.c: Replaced previous tc_aout_fix_to_chars stub with the |
281 | real thing. | |
282 | ||
283 | * tc-i960.c, tc-sparc.c: white space and comments only. | |
284 | ||
285 | * tc-a29k.h: delete duplicate macro definition. | |
286 | ||
287 | * new file atof-ns32k.c copied from hack's last unreleased gas. | |
288 | ||
289 |