1 Sat Jan 4 16:34:16 1992 John Gilmore (gnu at cygnus.com)
3 * as.h: Move redefine of xfree.
4 * config/ho-generic.h: Declare free(), since it's used without
6 * read.c: Use xmalloc, not malloc.
7 * expr.c, input-scrub.c, write.c: Avoid "know" macros
8 that cover more than one line, since this causes char strings
9 with embedded newlines.
11 Wed Dec 11 18:13:07 1991 Steve Chamberlain (sac at rtl.cygnus.com)
12 * configure.in: handle rename of obj-coffbfd
13 * config/obj-coffbfd.c: lint, don't fixup relocs if H8300, use
14 tc_reloc_mangle to prepare relocation if one available
15 * config/tc-h8300.h: added tc_reloc_mangle routine
17 Tue Dec 10 04:07:28 1991 K. Richard Pixley (rich at rtl.cygnus.com)
19 * input-file.c: don't close a NULL file pointer.
21 * Makefile.in: infodir belongs in datadir.
23 Sat Dec 7 17:13:42 1991 K. Richard Pixley (rich at rtl.cygnus.com)
25 * Makefile.in: fix fake-as target to also create as.new.
27 Fri Dec 6 23:16:43 1991 K. Richard Pixley (rich at rtl.cygnus.com)
29 * Makefile.in: install using INSTALL_DATA and INSTALL_PROGRAM.
30 correct include directory locations. added standards.text
33 * configure.in: mark this directory as target dependent.
35 Thu Dec 5 22:46:23 1991 K. Richard Pixley (rich at rtl.cygnus.com)
37 * Makefile.in: idestdir and ddestdir go away. Added
38 copyrights and shift gpl to v2. Added ChangeLog if it
39 didn't exist. docdir and mandir now keyed off datadir by
42 Wed Dec 4 10:28:48 1991 Steve Chamberlain (sac at cygnus.com)
44 * config/obj-coff-bfd.c (crawl_symbol_chain): only move a
45 symbol to the end of a file if it really is external.
48 Sat Nov 30 23:03:59 1991 Steve Chamberlain (sac at rtl.cygnus.com)
50 * as.h: reloc.h->aout/reloc.h.
51 * a.out.gnu.h: reloc.h->aout/reloc.h
53 Thu Nov 28 13:21:41 1991 Steve Chamberlain (sac at rtl.cygnus.com)
55 * config/obj-coff-bfd: made multiple sections with subsections
57 * config/tc-h8300.c: made $ statement sep, and int alloc 2 bytes.
60 Mon Nov 25 17:26:22 1991 Steve Chamberlain (sac at cygnus.com)
62 * as.c: (perform_an_assembly_pass): moved declaration of i to
63 avoid a syntax error. Now always create the three default
64 sections when in MANY_SEGMENTS mode.
65 * as.h: New improved way of specifying MANY_SEGMENTS
66 * configure.in: hds object format is now coff-bfd
67 * subsegs.h: add line numbers and section stuff to the
68 segment_info structure. (But go back later and fix.. this is
72 Thu Oct 24 17:25:59 1991 Steve Chamberlain (steve at rtl.cygnus.com)
74 * config/obj-coff.c (obj_coff_dim): Allows newline to end a
77 Mon Oct 21 09:25:33 1991 Steve Chamberlain (steve at rtl.cygnus.com)
79 * configure.in: added ebmon entry.
80 * output-file.c (output_file_create, output_file_close) added
82 * write.c: more bfd support
83 * config/obj-coff.c config/obj-coff.h use bfd for swapping
85 * tc-a29k: added the reloc types we support now
88 Mon Oct 14 11:44:03 1991 Roland H. Pesch (pesch at cygnus.com)
90 * doc/Makefile (as.ms): one more sed workaround for texi2roff
91 (as.mm, as.me): new targets
93 Fri Oct 11 18:18:37 1991 Roland H. Pesch (pesch at cygnus.com)
95 * doc/Makefile: new targets as.info, as.ms
96 doc/as.texinfo: minor changes for texi2roff-2
97 Makefile.in: revise 'doc' target.
99 Fri Jan 4 12:48:22 EST 1991 Jay Fenlason (hack@ai.mit.edu)
101 * messages.c Moved as_perror from input-scrub.c Modified the
102 error messages to look better.
104 * output-file.c Don't call as_fatal, just call exit()
106 expr.c Slightly improve checking for foo-foo case in
107 clean_up_expression(). Detect foo: bar: ... foo-bar...
109 Tue Dec 4 16:29:20 EST 1990 Jay Fenlason (hack@ai.mit.edu)
111 * m68k.c Fixed an obscure bug involving AOFF mode with a
112 large constant displacement (Was forgetting to output the
115 make-gas.com Added a three line patch from Eric Youngdale that
116 makes it possible to submit make-gas.com to a batch queue.
118 Wed Nov 21 15:07:51 EST 1990 Jay Fenlason (hack@ai.mit.edu)
120 * vms.c (VMS_TBT_Routine_END) Add a four line patch from
123 Tue Nov 13 14:02:15 EST 1990 Jay Fenlason (hack@ai.mti.edu)
125 * vms-dbg.c (VMS_DBG_record()) Another one character patch from
128 Mon Oct 29 15:49:21 EST 1990 Jay Fenlason (hack@ai.mit.edu)
130 * read.c Replace some as_warn calls with as_bad.
132 Fri Oct 26 15:21:15 EDT 1990 Jay Fenlason (hack@ai.mit.edu)
134 * i386.c, i860.c, ns32k.c Add const changes.
136 Mon Oct 22 14:04:26 EDT 1990 Jay Fenlason (hack@ai.mit.edu)
138 * sparc.c Add const changes.
140 * make-gas.com define const= for VMS, since many older versions of
141 GCC don't work correctly with const under VMS.
143 Thu Oct 18 12:44:11 EDT 1990 Jay Fenlason (hack@ai.mit.edu)
145 * i860.c i860-opcode.h Added patches from rgb@mcc.com
147 * read.c, symbols.c, vms.c, + new_file vms-dbg-module.c
148 Added Eric Youngdale's <YOUNGDALE@v6550c.nrl.navy.mil> VMS debugging
149 patches, so debugging GCC output now works.
151 * hash.c (hash_grow) Remember to blank out the wall entry in the new
152 hash table. This is important on systems where malloc() returns
153 non-zero storage. . .
155 Tue Oct 16 10:11:35 EDT 1990 Jay Fenlason (hack@ai.mit.edu)
157 * output-file.c (output_file_create) if output filename is given as
158 '-', write to stdout.
160 * m68k.c Finally get the PCREL code to work right. Add relaxation of
161 PCREL stuff This small fix from Ken Woodland
162 (kenny%datacube.uucp@uunet.uu.net).
164 * m68k.c Added some const declarations to constants. (md_relax_table,
165 md_pseudo_table, etc. . .)
167 Thu Oct 11 11:15:10 EDT 1990 Jay Fenlason (hack@ai.mit.edu)
169 * Makefile, read.c, write.c Include the i860 port.
170 (New files i860.c i860-opcode.h i860.h)
172 * m68k.c Fix some addressing modes, (AOFF, AINDEX, etc) to work in
175 * (all over) Raeburn's const hacking. This reduces the data-space size by
176 declaring many tables, etc, as 'const'.
178 Mon Oct 22 22:48:22 1990 John Gilmore (gnu at cygint)
180 Make gas work if you turn on the know() checks.
182 * app.c: Only pass a single space through: the one after
183 the opcode. All other whitespace is removed, to match the
184 expectations of the parser in read.c.
186 * as.h: Remove obsolete comments. Remove JF's NDEBUG so
187 that know() can really work if you turn it on. Make
188 SEG_MAXIMUM_ORDINAL == SEG_REGISTER.
190 * expr.c (operand): Change BITS_PER_INT to 8*sizeof(int).
192 * input-scrub.c: strlen("\0") doesn't return 1...
193 (as_where): Add space after line number in errors, like gcc.
195 * m68k.c (s_bss): Fake .bss into data section 255.
196 We can't cope with a real "BSS section" yet, but we can at
197 least do the right thing less efficiently (with lots of
200 * read.c: Turn lots of as_warn()'s into as_bad()'s.
202 * read.h (SKIP_WHITESPACE): Replace last instance of ASSERT()
205 * sparc.c (s_seg): We can't put frags into the BSS segment
206 yet, so just fake bss seg as 255th data segment.
208 * vax.c: Remove \'s from continued macro *parameters*. These
209 must have been added after the last time someone turned on
212 * write.c (relax_segment): Refine what we know() about the
213 symbols referenced during relaxation.
215 * Makefile (OTHER_ALIGN): Remove, handled in tables now.
216 Flip options a bit. These options really ought to go
219 Sun Oct 21 03:57:21 1990 John Gilmore (gnu at cygint)
223 * expr.c, write.c: Missing semicolon after know().
225 * write.c (fixup_segment): Allow pc-relative accesses to undefined
226 external symbols. Previously this would turn off pc-rel calc
227 of displacement, while leaving pc-rel opcode alone, botching.
229 * m68k.c (m68k_ip): Allow pc-relative effective addresses
230 for source operands. "pea" instructions are a good example
231 where we can shorten from abs long to pc+16bit.
232 (md_convert_frag): Fix "JBSR" comments to refer to "JSR", the
233 actual instruction. Insert comments about bug in 68000 bcc
234 and dbcc long code (that doesn't get exercised much). Add
235 comments about long pcrel displacements for -pic. Remove
236 "this code has not been tested" comment.
237 (md_estimate_size_before_relax): Now that fixup_segment
238 doesn't shortcircuit pc-relative fixups for undefined symbols,
239 only output them if -pic; else turn them absolute, which is
240 slightly faster. More JBSR->JSR comments.
241 (md_parse_options): Parse -pic.
243 Fri Oct 19 14:35:20 1990 John Gilmore (gnu at cygint)
245 * Make sparc assembler more compatible with Sun assembler.
246 sparc.c: reformat pseudo-op table to match main table.
247 (md_assemble): Add SPECIAL_CASE_FDIV to assemble FDIV*
248 instructions as fdiv followed by fmovs to get around chip bug.
249 (s_common, s_seg): Accept "bss" section name.
250 (md_assemble): Handle "set" instructions with absolute
251 operands, that only take one instruction rather than two.
252 sparc-opcode.h (fdiv*): Mark instructions "S"pecial.
253 subsegs.c (subseg_change): Move tail pointer too.
254 symbols.c (colon): Allow new definitions to override .comm symbols,
255 as in VMS. Sun CC depends on this.
256 write.c (new_fix): Always take r_type argument, not just on sparc.
257 Chain fixP's in order, using tail pointer, so relocation
258 records come out in forward order like Sun as. Remove SPARC
260 write.h: Add seg_fix_tailP, data_fix_tailP, text_fix_tailP.
262 * am29k.c: Use s_align_bytes rather than a local copy.
264 * read.c (s_align): Rather than ifdef it, make two functions,
265 s_align_bytes and s_align_ptwo. Individual pseudo-op tables
266 can call whichever one they like.
268 * write.c (append): Move from append.c to here.
269 append.c: Remove file.
271 * Makefile (MDSRC, mdsrc): Easy way to edit all md.c's.
272 Fix options. Add option for -DDEBUG for know() and assert().
273 Remove append.c, am29k.h. Don't build special read-sparc.o.
274 Remove sparc.h. "make clean" removes am29k .o's. Add
275 dependencies on reloc.h.
277 Thu Oct 18 17:56:34 1990 John Gilmore (gnu at cygint)
279 * Generalize sparc extensions to relocation info. Gas now
280 keeps relocation information internally in a different format
281 than how it is stored in the resulting .o. md_ri_to_bytes()
282 converts to external format. md_reloc_size says how large
283 each relocation record is in external format.
284 sparc.h: Remove this file. Rename to reloc.h. Rename struct
285 to reloc_info_generic.
286 reloc.h: Add relocation types for AMD 29000.
287 read.c, write.c: Always call fix_new with reloc-type argument.
288 write.c (emit_relocations): Make md_ri_to_bytes write directly
289 to output area rather than overwriting its argument then
291 md.h: Declare md_reloc_size and md_ri_to_bytes.
292 i386.c, am29k.c, vax.c, ms32k.c, m68k.c, sparc.c: include reloc.h.
293 (md_reloc_size): Specify correct value.
294 (md_ri_to_bytes): Convert format from internal to external.
296 write.c (write_object_file): Call md_section_align() which
297 rounds section sizes up if desired.
298 sparc.c (md_section_align): Round to 8 byte boundary.
299 i386.c, am29k.c, vax.c, ns32k.c, m68k.c (md_section_align): Nop.
301 Mon Oct 15 22:06:11 1990 John Gilmore (gnu at cygint)
303 Changes in support of the AMD 29000 version of gas.
305 * am29k-opcode.h: Add dummy entry to end so we can examine
306 item N+1 without exceeding table.
308 * am29k.h: New include file, derived from sparc.h. Kludged
309 together, still needs major work to get relocation working.
311 * am29k.c: New file, derived from sparc.c.
312 Put 29k-specific ASM29K pseudo-ops into table.
313 Change comment_chars to ASM29K.
314 Change s_seg to s_use.
315 Change s_sparc_align to s_29k_align, default operand to 4.
316 (define_some_regs): Define special register names.
317 (md_begin): Preprocess opcode table to mash together all
318 the variants of single opcodes. This simplifies later handling.
319 Call define_some_regs to preset special reg names.
320 (parse_operand): Add, parses out an operand from a stmt.
321 (machine_ip): Simplify, since 29K is simpler asm language.
322 Handle the various keyletters in the opcode table.
325 Handle include files in the assembler, with a .include
327 * as.h (input_scrub_include_file): declare.
328 * as.c (perform_an_assembly_pass): Avoid buffer hacking.
329 Start us off in text segment.
330 * read.c (read_a_source_file): Take a name as argument,
331 internalize all buffer handling. Don't start a new text
332 subsegment on each entry. Actually use the start and end
333 pointers returned by input_scrub_next_buffer.
334 (s_include): Call input_scrub_include_file for .include.
335 * input-scrub.c: Fix comments.
336 (struct input_save): Add, for saving state at .include.
337 (input_scrub_push, input_scrub_pop): Add, push & pop state.
338 (input_scrub_begin): Initialize next_saved_file.
339 (input_scrub_end): Free buffer.
340 (input_scrub_include_file): Add, to include a file.
341 (input_scrub_close): Add, to close a file.
342 (input_scrub_next_buffer): Set buffer-start address for
343 caller. If we hit EOF and were included, pop to previous file.
344 * input-file.c: Remove old includes. Remove old file-descriptor
345 hacking code, that was commented out.
346 (struct saved_file): Add, for saving state at .include.
347 (input_file_push, input_file_pop): Add, push & pop state.
348 (input_file_open): Don't buffer all files in one place.
349 (input_file_close): Add, close input file.
350 * input-file.h: Declare new functions.
351 * app.c: (struct app_save): Add, for saving state at .include.
352 (app_push, app_pop): Add, push and pop state.
353 (do_scrub_next_char): Move its static state variables out so
354 they can be saved and restored.
357 * app.c: Allow overriding of character meanings by machine
358 dependent strings. Avoid wiring character constants into app.c.
359 (do_scrub_begin): New meanings override old ones, not "OR" them.
360 Only handle /* comments if / is not already in use.
361 (do_scrub_next_char): Reorganize mass of nested if's
362 into a switch for speed. Don't assume ';' is line terminator.
363 Reorganize switch on characters, into a switch on their (machine-
364 dependent) lexer table meanings.
367 Encapsulate knowledge of segment types in fewer places.
368 This allows us to add the "SEG_REGISTER" type, as well as
369 providing flexibility for eventual COFF/ELF support.
370 * struc-symbol.h (symbol_to_segment, symbol_to_segment_type,
371 set_symbol_segment, set_symbol_segment_keep_ext,
372 segment_name): Define macros to encapsulate this info.
373 * as.h: Remove externs for seg_name, seg_N_TYPE, N_TYPE_seg.
374 * symbols.c (symbol_new): Change 'type' arg to 'seg'.
375 * expr.c, i386.c, m68k.c, ns32k.c, read.c, symbols.c, vax.c,
377 * i386.c, m68k.c, ns32k.c, vax.c, write.c: Change 2nd arg type of
378 md_estimate_size_before_relax.
379 * expr.c, read.c: Change 'type' arg to symbol_new.
380 * read.c, symbols.c, vax.c, write.c: Move md.h to end of includes.
383 Allow expressions to evaluate to registers.
384 * as.h: Add SEG_REGISTER.
385 * struc-symbol.h: Add fake N_REGISTER type.
386 * subseg.c: Add types to tables.
387 * expr.c (operand): Symbols of SEG_REGISTER type are
388 immediately evaluated like those of SEG_ABSOLUTE.
389 (clean_up_expression): Clean up SEG_REGISTER exprs.
392 Allow machine descriptions to cleanly extend the set of
394 * expr.c (operand): Call md_operand before rejecting an
395 operand as unacceptable.
397 * i386.c, ns32k.c, m68k.c, sparc.c, vax.c: Define null function.
398 * am29k.c (md_operand): Use this for %% and & prefix operators.
401 Allow machine descriptions to cleanly permit symbols to be
402 predefined upon first usage.
403 * symbols.c (symbol_find_or_make): Call md_undefined_symbol
404 before making an undefined symbol.
406 * i386.c, ns32k.c, m68k.c, sparc.c, vax.c: Define null function.
407 * am29k.c (md_undefined_symbol): use this for the local and
408 global register names; since there are hundreds of them, it
409 only defines them upon their first use.
410 * expr.c (operand): Call symbol_find_or_make rather than
411 roll our own undefined symbols.
414 Miscellaneous changes:
416 * as.h (know): Make this useful if DEBUG defined.
418 * write.h: Support SPARC-like relocation throughout all
421 * read.c (read_a_source_file): Report the name of invalid
422 pseudo-ops. Don't double-report junk characters. Close the
423 input file, which gas never used to do.
424 (ignore_rest_of_line): Report junk chars in ascii if
426 (s_ignore): Ignore entire statement; used for 'listing
427 control' statements from ASM29K, e.g. .eject.
429 * read.c (s_lsym): Handle register equates too.
431 * read.c: Add most ASM29K pseudo-ops to the master table.
432 Not all are implemented yet.
434 * cond.c: New file, for functions implementing conditional
435 assembly pseudo-ops: .ifdef, .ifndef, .else, .endif, .ifseq,
436 .ifsne, .end. So far, they are just stubbed out.
438 * read.c (pobegin): Allow the machine dependent pseudo-op
439 table to override the generic one. Remove #ifdef SPARC
440 on .word, since it can now be overridden.
442 * expr.c (operand): Support radix-2 constants. Kill off
443 support for octals with digits '8' and '9'. Initial steps
444 toward more general support for local-labels.
446 * symbols.h (symbol_table_lookup): Remove macro, change all
447 occurrences (in read.c, expr.c, symbols.c) to symbol_find.
449 * read.h (is_end_of_line): Define for external use.
451 * i386.c (alloca): Use builtin_alloca or include or extern.
453 * Makefile: Add ALL and all: entries. Add asm29k entries.
454 Add cond.c and cond.o. Remove special handling for messages.o.
457 Thu Sep 27 13:43:49 EDT 1990 Jay Fenlason (hack@ai.mit.edu)
459 * m68k.c (get_num) Fix so that 1:w is treated properly.
461 * Makefile Replace references to a.out.h with a.out.gnu.h
463 Tue Sep 25 15:50:36 EDT 1990 Jay Fenlason (hack@ai.mit.edu)
465 * sparc.c (md_number_to_imm) Fix so that RELOC_32 locations contain
466 zero, so that it will work with some sparc loaders which don't assume
467 that the locations in question do. A xix line patch from Michael Bloom
468 (usc!srhqla!quad1!psivax!ttidca!mb@zaphod.mps.ohio-state.edu)
470 Mon Sep 24 11:43:15 EDT 1990 Jay Fenlason (hack@ai.mit.edu)
472 * as.c #include <sys/types.h> if _POSIX_SOURCE defined. This because
473 <signal.h> uses pid_t that's defined in it.
475 * m68k.c reverse the sense of -l option, and allow :w and :l to
476 override the default size of AOFF indexes.
478 Also, allow -l to shorten branches to unknown labels from LONG to WORD.
480 Thu Sep 13 17:05:09 EDT 1990 Jay Fenlason (hack@ai.mit.edu)
482 * vax.c (md_parse_option) Don't print a warning msg if given -J
484 Wed Sep 5 14:26:14 EDT 1990 Jay Fenlason
486 * expr.c (operand) Don't get garbaged high-order bits when given a
487 lot of leading zeroes.
489 Tue Sep 4 11:40:21 EDT 1990 Jay Fenlason
491 * read.c (pseudo_set) Compain if we're setting the symbol to the
492 difference of two symbols which are in different frags. (We can't
493 find out how far apart they are.)
495 Wed Aug 15 12:18:58 EDT 1990 Jay Fenlason
497 * m68k.c (m68k_ip_op) Dyke out the code that deals with parsing
498 :[wl] at the end of expressions since it is handled in get_num()
499 and this was putting the result in the wrong place anyway.
500 Corrected a couple of other references to ->isiz instead of con?->e_siz
502 Mon Aug 13 15:53:46 EDT 1990 Jay Fenlason
504 * read.c Handle .align specially on the sparc, or any other machine
505 where OTHER_ALIGN is defined. Added option and comments about it
508 Fri Aug 10 12:24:33 EDT 1990 Jay Fenlason
510 * m68k.c (get_num) Handle SEG_PASS1 expressions.
512 Mon Aug 6 16:32:29 EDT 1990 Jay Fenlason
514 * write.c (fixup_segment) Added two patches for the NS32k
515 and SEQUENT A six line patch from Ian Dall
516 (asgard!aegir!hugin!augean!sibyl!ian@munnari.oz.au)
518 Wed Aug 1 13:30:48 EDT 1990 Jay Fenlason (hack@ai.mit.edu)
520 * m68k.c Include REGISTER_PREFIX ifdefs.
522 * write.c Include LOCAL_LABEL() and DOT_LABEL_PREFIX feature.
524 * vax.c (md_parse_option) Accept -H option.
526 * vms.c New version of case hasher, etc. These from Eric Youngdale
527 <YOUNGDALE@v6550c.nrl.navy.mil>
529 Fri Jul 20 13:39:02 EDT 1990 Jay Fenlason (hack@ai.mit.edu)
531 * README Added README.APOLLO and README.COFF stuff
533 Wed Jul 18 16:29:22 EDT 1990 Jay Fenlason (hack@ai.mit.edu)
535 * Makefile Added option for SEQUENT_COMPATABILITY
537 * ns32k.c Add configurable syntax feature from
538 ian@sibyl.eleceng.ua.oz@augean.ua.oz.au
539 and SEQUENT_COMPATABILITY
541 * ns32k-opcode.h Add configurable syntax feature.
543 Mon Jul 16 11:44:14 EDT 1990 Jay Fenlason (hack@ai.mit.edu)
545 * write.c (relax_segment) On ns32k, add fragP->fr_pcrel_adjust to
547 (fixup_segment) On ns32k, don't subtract size from
548 add_number on pcrel external symbols.
550 * ns32k.c (md_relax_table) Use correct max displacements.
551 This is a six-line patch from ian@sibyl.eleceng.ua.oz.au
553 * ns32k.c (md_atof, convert_iif) Emit floating point numbers in
554 the correct byte order. A seven line patch from
555 ian@sibyl.eleceng.ua.oz.au
557 * ns32k.c (all over) Some lint fixes.
559 Mon Jul 9 13:17:00 EDT 1990 Jay Fenlason (hack@ai.mit.edu)
561 * app.c (do_scrub_next_char) If a comment is #{whitespace}
562 don't treat the next line as comment also.
564 * m68k.c (...) Accept apc@(num:[wl]), etc.
566 * i386.c (md_assemble) Get bitfields correct when doing cross
567 assembly to 386. A two line patch from Michael Bloom.
568 (usc!srhqla!quad1!ttidca!mb@zaphod.mps.ohio-state.edu).
570 * README.APOLLO a new file with vasta@apollo's name, address
573 * make-gas.com Deleted references to gdb source files.
575 Fri Jul 6 14:34:27 EDT 1990 Jay Fenlason (hack@ai.mit.edu)
577 * i386.c Ignore the .optim directive
579 * input-file.c Change from _IOLBF to _IOFBF in setbuffer emulation
582 Mon Jun 18 15:36:49 EDT 1990 Jay Fenlason (hack@ai.mit.edu)
584 * sparc.c #ifdef DONTDEF s_sparc_align, since it isn't called from
587 Fri Jun 15 15:53:30 EDT 1990 Jay Fenlason (hack@ai.mit.edu)
589 * vax.c (md_parse_option) make the code agree with the documentation
590 on the behaviour of the -d option.
592 Thu Jun 7 14:23:54 EDT 1990 Jay Fenlason (hack@ai.mit.edu)
594 * atof-ieee.c (gen_to_words) Assemble 0r-0 correctly.
596 * Makefile Remove last references to gdb*.c files.
598 * version.c New version 1.36
600 Tue May 22 13:22:26 EDT 1990 Jay Fenlason (hack@ai.mit.edu)
602 * Makefile Mention a work-around for a possible problem with HPUX7.0
604 Mon May 21 14:06:04 1990 Jim Kingdon (kingdon at pogo.ai.mit.edu)
606 * sparc.c (sparc_ip): Change error message from "not in hash table"
609 Wed May 16 15:33:14 EDT 1990 hack@wookumz
611 * i386.c (i386_operand) Print error msg if given an operand like
612 4(mumble) which we can't parse.
614 * m68k.c (md_assemble) Add '3' to the list of operand-places that
615 can be found in 'symbol-dependent info'. Also change
616 'confusing width' diagnostic to something more meaningful.
618 Fri May 11 12:09:21 EDT 1990 hack@wookumz
620 app.c (do_scrub_next_char) Don't flush the line after a line
621 consisting of a single '/' A one-line patch from Mike Kupfer
622 (kupfer@orc.olivetti.com)
624 * i386.c (md_assemble) Call frag_wane() before calling frag_new()
625 A one line patch from Steve Bleazard (steve@robobar.co.uk
627 Tue May 8 12:56:25 EDT 1990 hack@wookumz
629 * atof-generic.c (atof-generic) Modified the Infinity detection code
630 to accept 0rinfinity and 0r-infinity as well as 0rinf and 0r-inf
632 Thu Apr 26 15:17:31 EDT 1990 hack@wookumz
634 * atof-ieee.c Change value of NaNs to 0x7fff ffff (float) and
635 0x7fff ffff ffff ffff (double) If you want some other value for
636 NaN, use .long and spell it out yourself.
638 atof-generic.c (atof_generic) Cleaned up code that detects NaN
641 vax.c (md_assemble) print a useful error message if expression()
642 returns SEG_PASS1 or SEG_DIFFERENCE and we can't deal with those.
644 Thu Apr 19 10:30:47 EDT 1990 hack@wookumz
646 * input-scrub.c (AFTER_STRING) Make AFTER_STRING contain a null
647 so that the strstr() call when looking for "#NO_APP" after a "#APP"
648 will work. A two character patch from Bruce Robertson
649 (bruce@heather.pooh.com)
651 * Makefile, i386.c Use atof-ieee.c instead of atof-i386.c
653 Mon Apr 16 16:20:55 EDT 1990 hack@wookumz
655 * m68k.c (md_relax_table) Many of the offsets were off by two.
656 Fixed some generic spacing problems thoughout the file.
658 Thu Apr 12 12:22:35 EDT 1990 hack@wookumz
660 * sparc.c (md_ri_to_chars) Handle little-endian cross assembly.
662 * write.c (relax_segment) Compare addresses correctly to avoid
663 accidentally relaxing a branch that we don't have to.
664 These small changes from John Gilmore (gnu@toad.com)
666 Fri Apr 6 12:52:15 EDT 1990 hack@wookumz
668 * Makefile, expr.c symbols.c Correctly document the SUN_ASM_SYNTAX
669 option, and make it work.
671 Tue Mar 20 12:46:59 EST 1990
673 * as.c (main) Only trap SIGHUP, SIGINT, SIGPIPE, and SIGTERM,
674 and only if they aren't being ignored. A three line patch
675 from Paul Eggert (eggert@twinsun.com)
677 * write.c (relax_segment) Correct typo 'growth - ' should have been
680 * atof-vax.c (next_bits, flonum_gen2vax) Clean up by sharing some
681 variables. While we're at it, fix next_bits so that it
682 doesn't use littlenums that don't exist. . .
684 Tue Mar 13 16:23:21 EST 1990 hack@wookumz
686 * Rename atof-m68k.c atof-ieee.c
688 * Delete atof-ns32k.c
690 * m68k.c sparc.c ns32k.c atof-ieee.c Call atof-ieee instead of
691 atof-m68k or atof-ns32k
693 * Makefile Compile with atof-ieee.c instead of atof-ns32k.c or
696 Mon Mar 12 14:06:55 EST 1990 hack@wookumz
698 * as.c If the signal handler gets called twice, exit immediatly.
700 * ns32k.c Call gen_to_words with a pointer of the proper type, and
701 call md_number_to_chars to put the results in the proper byte-order.
702 Whoever wrote this code was *sloppy*!
704 * Makefile ns32k.o depends on ns32k.c
706 * vax.c (md_parse_option) If VMS, accept -+ and -h options.
708 * vms.c (VMS_Case_Hack_Symbol) Replace #if NO_CASE_HACKING
709 with references to the -h option. These small VMS patches
710 from Angel Li (angel@flipper.miami.edu).
712 Thu Mar 8 19:18:59 EST 1990 hack@wookumz
713 * vms.c Some trivial patches from Eric Youngdale
714 (YOUNGDALE@v6550c.nrl.navy.mil)
716 Wed Mar 7 17:12:09 EST 1990 hack@wookumz
717 * make-gas.com (Define error as as_fatal when compiling vax.c and vms.c
718 A two line patch from Eric Youngdale
719 (YOUNGDALE@v6550c.nrl.navy.mil)
721 Tue Mar 6 16:01:09 EST 1990 hack@wookumz
723 * Makefile Include ns32k options in makefile. A small patch from
724 David Taylor (taylor@think.com).
726 * as.c read.c write.c Makefile #ifdef DONTDEF out all the gdb
727 symbol stuff, since it isn't used anymore and it doesn't work.
729 Mon Mar 5 14:51:04 EST 1990 hack@wookumz
731 * i386.c (md_assemble) Replace memchr() with index().
733 * as.c Trap signals 1 through NSIG, print an error msg, and don't
734 produce an object file.
736 * m68k.c Added a hack so that fsincosx fpx,fpy:fpz works.
738 * messages.c New function: as_bad This is like as_warn, except
739 -W doesn't disable it, and calling it inhibits production of an
740 object file and causes a non-zero exit code.
742 Tue Feb 13 14:25:53 EST 1990 hack@wookumz
743 * Makefile Include G0 and LOADLIBES for Sequent Symmetry.
744 Based on a small patch from Johan Widen (jw@sics.se)
746 Thu Feb 1 14:08:58 EST 1990 hack@wookumz
747 * m68k.c Replace 'abort' with 'abort()' which will work.
749 Wed Jan 24 17:15:08 EST 1990 hack@ai.mit.edu
751 * read.c (ignore_rest_of_line) Have it print the first junk char
752 in both decimal and %c form.
754 (read_a_source_file) On bad pseudo-op, print out the unknown
757 Tue Jan 23 13:12:48 EST 1990 hack@ai.mit.edu
759 * read.c (pseudo_set) If the symbol is external, have it remain
762 * i386-opcode.h Allow jc as a synonym for jb and jnc as a syn for jnb.
765 Wed Jan 3 09:35:31 EST 1990 hack@ai.mit.edu
767 * ns32k.c [cpureg_032] Change register id of psr from 0x0b to 0x0d
768 * ns32k-opcode.h Change shift-counts for lsh and lshd
769 to one byte instead of 2 and 4.
770 A Trivial patch from John F. Peters (think!ames!practic.com!jfp@eddie)
772 Tue Dec 5 16:37:44 EST 1989 hack@ai.mit.edu
774 * ns32k.c (md_create_{long,short}_jump) Six line patch from
775 John F Peters (think!ames!vine!practice.com!jfp) to use the
776 correct addressing mode and byte-order for broken-word stuff.
778 * write.c (write_object_file) One line patch to call fix_new_ns32k
779 with the correct # of args.
781 Fri Dec 1 16:44:21 EST 1989 hack@ai.mit.edu
783 * atof-generic.c, flonum-mult.c A real fix for the trailing-zeroes
784 problem from Georg Feil (ghfeil@white.toronto.edu) (two line change)
786 Mon Nov 27 15:30:46 EST 1989 hack@ai.mit.edu
788 * i386-opcode.h Fixed opcode-table entry for ljmp. A one char
789 patch from eliot@mgm.mit.edu
791 Mon Nov 20 12:41:28 EST 1989 hack@ai.mit.edu
793 * expr.c Replace the generic_buffer hack with a more portable one */
795 * atof-generic.c (atof_generic) Ignore trailing zeroes after a decimal
796 point. For some reason trailing zeroes (but not trailing nonzeroes) were
797 causing loss of precision. I don't know why. . .
799 * vms.c Change copyright notice. Install changes from Kenneth Adelman
800 (adelman@tgv.com) for c++? (A dozen lines or so)
802 Mon Nov 13 11:48:44 EST 1989 hack@ai.mit.edu
804 * Makefile Add BINDIR and use it to control where the executable is
807 * i386.c Use __builtin_alloca if possible (trivial patch from
808 Marco S. Hyman pacbell!dumbcat!marc)
810 Mon Nov 6 18:24:47 EST 1989 hack@ai.mit.edu
812 * version.c New version: 1.35 will be distributed with the
815 Mon Oct 30 10:38:11 EST 1989 hack@ai.mit.edu
817 * atof-m68k.c (atof_m68k) Don't put the bits[] array on the stack,
818 since it may be pointed to after atof-m68k exits.
820 Tue Oct 24 11:15:57 EDT 1989 hack@ai.mit.edu
822 * atof-m68k.c Added #define for bcopy on USG systems.
823 #ifdef TEST the print_gen() function.
825 * a.out.h if USE_HP_INC_HDR then use ../binutils/hp-include/a.out.h
827 Fri Oct 13 14:36:48 EDT 1989 hack@ai.mit.edu
829 * vax.c (all) Ran vax through indent -gnu to make it readable.
831 vax.c (vip_op) Correctly assemble code like jmp $*0x11223344
832 by setting vip_nbytes to 4 when using an immediate address.
835 m68k.c (s_proc (new)) Added s_proc no-op pseudo-op.
837 Makefile Added instructions for compiling on Sequent Symmetry
840 a.out.h Modified to compile on Sequent and HP above. (HP port
841 based on a msg from asjl@comp.vuw.ac.nz (real name unknown)).
843 Tue Oct 10 14:39:44 EDT 1989 hack@ai.mit.edu
844 * vax.c (vip_op) Fixed a typo in an error msg and cleaned
845 up some spacing stuff.
847 Wed Sep 27 19:07:12 EDT 1989 hack@ai.mit.edu
849 * app.c (do_scrub_next_char) Fixed parsing of
850 # <line> "file" garbage
851 text so that it'll work again? (8 line patch from Mike Hibler
854 Mon Sep 18 16:26:01 EDT 1989 hack@ai.mit.edu
856 * app.c (do_scrub_next_char): Modify parsing of /* ... */ to work
859 * sparc.c (sparc_ip): Don't abort on insns that use the Alternate
860 Spaces. Try to assemble them correctly.
862 Thu Sep 14 11:42:44 EDT 1989 hack@ai.mit.edu
864 * sparc.c (md_number_to_imm) Dozen line patch from jkp@sauna.hut.fi
865 (Jyrki Kuoppala) so that gas output will work with shared libraries.
867 * ns32k.c Include <string.h> instead of <strings.h> if USG defined.
869 (md_end) free(freeptr_static) instead of free(freeptr) .
871 * atof-ns32k.c Include as.h so that sysV stuff (bzero) will be
872 defined if needed. These ns32k changes from
873 nixbur!mollers.pad@seismo.css.gov (Josef Moellers)
875 Fri Sep 1 11:39:52 EDT 1989 hack@ai.mit.edu
877 * atof-m68k.c (gen_to_words) Get the sign right on negative
878 floating-point numbers.
880 Wed Aug 30 13:59:57 EDT 1989 hack@ai.mit.edu
882 * Makefile Remove the rest of the $< entries that kill sun make
884 Fri Aug 25 15:00:30 EDT 1989 Nobody You Know (hack@ai.mit.edu)
886 * atof-m68k.c (gen_to_words) deal with denormalized floating-point
889 Tue Aug 22 02:03:05 1989 Roland McGrath (roland at hobbes.ai.mit.edu)
891 * Makefile (gas-dist.tar): Put ChangeLog in the tar file.
893 * version.c: Added comment telling Jay Fenl--I mean people--not to put
894 changes in version.c, but to use ChangeLog instead.
896 * version.c (version_string): Put "GNU" in all-caps.
898 * version.c: Moved all comments about changes to ChangeLog (this file).
899 Many anonymous entries have been attributed to Jay Fenlason (hack).
901 Thu Aug 17 15:53:57 1989 Jay Fenlason (hack at apple-gunkies.ai.mit.edu)
903 * Makefile: Removed $< references that seem
904 to choke some versions of make.
906 * frags.c (frag_grow): Fixed to deal with requests for very
907 large frags (larger than frags.chunk_size).
909 * app.c (do_scrub_next_char): Have it ignore any characters
910 after the filename in a # line "filename".
912 * sparc.c (s_common): On an error, don't print out
913 input_line_pointer past the end of the line where the error is.
915 * atof-generic.c (atof_generic): Accept any case for
918 * m68k.c (m68_ip): Don't use PC-relative mode for alterable
921 Tue Aug 15 04:58:36 1989 Roland McGrath (roland at apple-gunkies.ai.mit.edu)
923 * sparc.c (md_begin): Rewrote this function to perform consistency
924 checks with the new opcode table.
926 Fri Aug 11 16:01:16 1989 Roland McGrath (roland at apple-gunkies.ai.mit.edu)
928 * sparc-opcode.h (struct sparc_opcode): Replaced `mask' field with
929 `lose'; removed `last' field. Updated all opcodes accordingly.
930 Fixed several opcodes that generated the wrong instructions.
931 sparc.c (md_begin, sparc_ip): Changed to use new struct sparc_opcode.
933 Thu Aug 3 14:44:24 1989 Jay Fenlason (hack at apple-gunkies.ai.mit.edu)
935 * Makefile (a32k): Use read- and write-ns32k.o
936 * ns32k.c (encode_operand): Make sure pcrel_adjust starts out zeroed.
937 * read.c (cons): Call fix_new_ns32k() if NS32K is defined.
938 * write.c (write_object_file): Ditto.
939 These so that .word sym-sym (etc) will produce values with
940 the proper byte-order.
942 Wed Aug 2 12:55:?? 1989 Jay Fenlason (hack at apple-gunkies.ai.mit.edu)
944 * sparc.c (comment_chars[]): Removed '|' because it was causing
945 problems. Probably not the best fix, since I suspect other
946 assemblers (68020) may get | in .stabs also, and the 68020 needs
947 the '|' comment character.
949 Mon Jul 31 09:22:28 1989 Roland McGrath (roland at apple-gunkies.ai.mit.edu)
951 * sparc.c (sparc_ip): Allow the characters [0123] in opcodes.
953 Tue Jul 25 16:32:12 1989 Jay Fenlason (hack)
955 * atof-generic.c (atof_generic): Tried to keep
956 size_of_digits_in_littlenum from going negative.
958 * sparc-opcode.h: Added duplicate [i+1] entries to go with
959 the [1+i] entries already there. A kludgy fix, but it works.
961 Mon Jul 24 17:20:03 1989 Jay Fenlason (hack)
963 * write.c (relax_segment): Modified rs_org code so it won't
964 occasionally dump core.
966 * write.c (pseudo_set): Modified SEG_DIFFERENCE to (perhaps)
967 allow one to set a symbol to the difference of two other symbols.
969 * ns32k.c (convert_iif): Moved size_so_far+=size and size=0 inside
970 the check for a valid type.
972 * sparc-opcode.h: Modified the entries for std "q,[1+i]", "D,[1+i]",
975 (In version 1.34) Jay Fenlason (hack)
977 * Makefile: Reorganized, added stuff to make asparc.
979 * sparc.c, sparc-opcode.h, sparc.h: Sparc port.
981 * write.c: Set the size of text and bss segments to a multiple of eight
984 * m68k.c: Moved .single pseudo-op to machine independent part.
986 * atof-generic.c: Fixed type in #ifdef __GNUC__.
988 * sparc-opcode.h: Handle "mov REG, %y".
990 * make-gas.com: Know that error.c no longer exists.
992 * sparc.c: Handle [expr+reg].
993 Don't call getExpression when looking for an immediate and getting
994 something that starts with % and isn't %hi or %lo.
996 * Teach the 68k about long conditional branches.
998 (In version 1.33) Jay Fenlason (hack)
1000 * Use __builtin_alloca if available.
1002 * README: Added more instructions for reporting bugs.
1004 * ns32k-opcode.h: Changed the acbb, acbw, and acbd insns.
1006 * vax.c: Replaced instances of LENGTH[STRING] with STRING[LENGTH].
1008 * ns32k.c (encode_operand): Increased max size of bit field for exts
1009 and inss instructions from 31 to 32 bits.
1011 * flonum-mult.c (flonum_multip): Fixed typo.
1013 * m68kc.: Allow #32 to be the same as #0 for bit-field ops.
1015 * make-gas.com, version.c, hex-value.c, flonum-const.c: VMS fixes.
1017 * ns32k.c, ns32k-opcode.h: More fixes from taylor@think.com.
1018 Mostly typos in comments, etc.
1020 * ns32k-opcode.h: Fixed size of immediate operands to andw and andd
1023 (In version 1.32) Jay Fenlason (hack)
1025 * read.c (s_set): Fixed misnamed variable.
1027 * as.c: Don't hang if given an invalid option.
1029 * m68k.c: Fixed bug in creating absolute long addresses for branches.
1031 * ns3k*: Some small ns32k patches.
1033 * m68k.c: Recognize 0rnan, 0rinf, 0r-inf.
1035 * app.c: Don't dump core on unterminated strings.
1037 * symbols.c: Give reasonable error messages.
1039 * ns32k*: Allow -m32032 and -m32532 options.
1041 * atof-*.c: Added support for NaN, Inf, and -Inf in atof_generic and
1042 the various descriptions.
1044 * m68k.c (add_fix): Replace occurrences of "width==" with
1045 "(width)==". This correct a precedence problem.
1047 * write.c, struc-symbol.h, m68k-opcode.h, m-hpux.h, Makefile: Changes
1048 for HP-UX from Chris Hanson (cph@kleph.ai.mit.edu).
1050 * m68k-opcode.h: Reorder movem insns so gdb will see the ones using the
1051 register list syntax first.
1053 * symbols.c (colon): Give more useful error messages when something was
1054 defined as a .comm and is now trying to be defined locally.
1055 Also, redefining a symbol is a fatal, not a warning.
1057 * m68k.c: Fixed a bug in using bignums as literal bit patterns for
1058 floating-point numbers.
1060 (In version 1.31) Jay Fenlason (hack)
1062 * i386*: More patches.
1064 * Moved machine-dependent option parsing into the machine-dependent
1067 (In version 1.30) Jay Fenlason (hack)
1069 * i386*: New new version.
1071 * atof-m68k.c: Changed to be smaller, with somewhat better modularity.
1072 Also fixed an obscure bug wherein next_bits would return random bits.
1074 * m68k.c: Be more careful about creating PC-relative addressing modes
1075 on the 68000 and 68010.
1077 * frags.c (frag_new): Zero out the new frag.
1079 * Don't choke on "foo= bar" or on formfeeds.
1081 * read.c: Allow Sun-syntax local labels #ifdef SUN_ASM_SYNTAX.
1082 * m-sun3.h: Defined SUN_ASM_SYNTAX.
1084 (In version 1.29) Jay Fenlason (hack)
1086 * i386.c: Newer version that fixes a bug wherein a jump instruction
1087 would be split between two frags.
1089 * i386*: New version.
1091 * m68k.c: #ifdef M_SUN and -m68010, produce Sun-2 executables.
1093 (In version 1.28) Jay Fenlason (hack)
1095 * m68k.c: Added .single pseudo-op.
1097 * Made ". = X" and ".set .,X" equivalent to ".org X".
1098 The pseudo-symbol "." has the value of the location the assembler is
1099 currently assembling to.
1101 (In version 1.27) Jay Fenlason (hack)
1103 * Merged ns32k and i386 support.
1105 (In version 1.26) Jay Fenlason (hack)
1107 * Added partial ns32k support.
1109 * Added RMS's evil .word misfeature. Invented the -k (kludge) option
1110 to warn that this misfeature was used.
1112 * Modified some files to get rid of warnings from GCC.
1114 * Added fix so that / can also be a comment character by itself.
1116 (In version 1.25) Jay Fenlason (hack)
1118 * Installed patches for VMS.
1120 * as.h (SIZEOF_STRUCT_FRAG): Added space before backslash-newline.
1122 * messages.c: Fixed typo.
1124 * app.c: Handle : correctly.
1126 * error.c: Removed; no longer used.
1128 * m68k-opcode.h: Added fnop.
1129 Fixed to correctly handle fmovem with a register list and
1130 non-predecriment addressing mode.
1132 * m68k-opcode.h: Fixed to know about long form of FBcc insns.
1134 * write.c: Warn if a fixup ended up being wider than its field width.
1136 (In version 1.24) Jay Fenlason (hack)
1138 * Accept and ignore -mc68010 and -m68010 switches.
1140 * Correctly assemble long subroutine calls on the 68000 without using a
1141 68020-specific instruction.
1143 * When calling with no filenames, read stdin.
1145 (In version 1.23) Jay Fenlason (hack)
1149 * xmalloc.c, xrealloc.c: Replaced to work with GCC.
1151 (In version 1.22) Jay Fenlason (hack)
1153 * write.c: Fixed a VMS bug.
1155 * m68k.c: Fixed a bug having to do with turning absolute into
1158 * atof-m68k.c (atof_m68k, gen_to_words): Try to avoid a problem with
1159 running off the end of the LITTLENUMS.
1161 * vax.c: Fixed so parenthesized expressions work.
1163 * atof-generic.c: Added a cast that fixes problems with some C
1168 * Changes for VMS support and correct bitfield order for
1173 * m68k*: Fixed "fmovel #N, fpcr". Added fpcr and fpsr to the list of
1178 * m68k.c? (md_convert_frag): Don't put the fixups for absolute long to
1179 PC-relative in the data segment.
1181 * atof-generic.c: #include <alloca.h> #ifdef sparc.
1185 * Re-fixed _vfprintf stuff (?).
1187 * Made "movem REG, ADDR" work.
1189 * Improved preprocessing, without temporary files.
1193 * Don't produce an undefined empty symbol for ".globl foo," (a line
1194 ending with a comma).
1196 * Fixed a bug wherein ".long X" became ".long 0" on the Sparc.
1198 * Fixed a bug which caused many "#APP" "#NO_APP" pairs to dump core.
1200 * Fixed calls to _doprnt to call _vfprintf #ifndef NO_VARARGS.
1204 * Merged HP-UX changes from Chris Hanson (cph@zurich.ai.mit.edu).
1206 * flonum-multip.c: Renamed to flonum-mult.c.
1208 * m-hpux.h: Created.
1210 * m68k.c (bcopy): Fixed.
1214 * struct-symbol.h: Renamed to struc-symbol.h.
1218 * vax.c: Added a quick fix for the offset of fixed-width branches not
1219 fitting in the field given.
1221 * gdb-lines.c, read.c: Added support for .gdline and .gdbline
1226 * read.c, atof-generic.c: Fixed bugs in reading in floating-point
1229 * m68k-opcode.h: Made "fmovep a0@, fp0" work.
1233 * write.c: Fixed an obscure bug in relaction that would occasionally
1234 cause the assembler to stop relaxing when it really had at least one
1239 * m68k*: Allow register lists in fmovem.
1241 * Added more floating-point exponents.
1243 * Print an error message on exponent overflow.
1247 * Fixed floating point bugs that made it generate incorrect numbers for
1248 values over 10^16 or so.
1252 * Fixed bug wherein you couldn't forward reference local label 0.
1256 * m68k.c, m68k-opcode.h: Added support for fmovem with register lists.
1258 * Fixed an obscure bug having to do with generating PC-relative
1259 addressing mode for things in the middle of the instruction instead of
1262 Wed Mar 1 15:29:24 1989 Randall Smith (randy at apple-gunkies.ai.mit.edu)
1264 * *.*: Modified copyright notices to reflect new General Public
1267 * Makefile: Added copyright notice.
1269 Fri Feb 17 09:42:01 1989 Jay Fenlason (hack at spiff)
1271 * Patched frags.c so that new frags start out bzero()ed.
1273 Thu Jan 26 14:23:44 1989 Jay Fenlason (hack at apple-gunkies.ai.mit.edu)
1275 * Added patches from pace to files as.h i386.c i386-opcode.h
1276 imull foo,%eax no longer gets assembled into the 32-64 bit
1277 multiply, which clobbers %edx behind gcc's back
1279 jcxz/jecxz were backwards
1281 There was a bug when using %ebp as a base register with no
1284 Instructions like andb $0xffffff, %al used to put out too many
1287 The splitting jump instructions across frags could happen when
1288 obstack_room()==6 too.
1293 version-control: never