libguile/ChangeLog

   1 2000-03-22  Dirk Herrmann  <D.Herrmann@tu-bs.de>
   2
   3         * numbers.h (SCM_SETNUMDIGS):  Use SCM_BIGSIZEFIELD macro for
   4         shifting, not constant.  Thanks to Dale P. Smith.
   5
   6         * numbers.c (scm_sum, scm_difference):  Don't test a SCM value
   7         for being less than zero.  Decode it to a C value first.  Again,
   8         thank you Dale.
   9
  10 2000-03-22  Dirk Herrmann  <D.Herrmann@tu-bs.de>
  11
  12         * numbers.h, ramap.c, struct.h, vectors.h:  Don't use SCM2PTR for
  13         non scheme values.  If raw data is stored in SCM variables, it has
  14         to be accessed using SCM_UNPACK until a better solution is found.
  15
  16 2000-03-22  Mikael Djurfeldt  <mdj@mdj.nada.kth.se>
  17
  18         * tags.h (SCM_ECONSP, SCM_NECONSP): More corrections of
  19         pointer-arithmetic induced by the SCM_PACK/UNPACK change.
  20
  21         * print.c (scm_iprin1): SCM_PACK/UNPACK corrections.
  22
  23         * gc.c (scm_gc_sweep): SCM_PACK/UNPACK corrections.
  24
  25         * eval.c (SCM_CEVAL, scm_unmemocar): SCM_PACK/UNPACK corrections.
  26
  27         * dynwind.c (scm_swap_bindings): SCM_PACK/UNPACK corrections.
  28
  29         * async.c, __scm.h: Removed lots of the old async click logic.  It
  30         is possible to reinsert it by defining GUILE_OLD_ASYNC_CLICK in
  31         __scm.h.  Let's try this out and dump the old code after the
  32         threads reorganization.
  33         (set-tick-rate, set-switch-rate): Conditionally removed.
  34
  35 2000-03-21  Mikael Djurfeldt  <mdj@mdj.nada.kth.se>
  36
  37         * gc.c (scm_gc_mark): Bugfix 1: The recent SCM_PACK/UNPACK change
  38         made SCM values into pointers.  This turned an arithmetic
  39         computation of the address of the vcell into a pointer-arithmetic
  40         one, thereby screwing up marking of structs.
  41         Bugfix 2: Removed incompletely introduced loop variable `j' used
  42         when protecting the tail array of a struct.
  43
  44 2000-03-21  Dirk Herrmann  <D.Herrmann@tu-bs.de>
  45
  46         * struct.h (SCM_STRUCT_DATA):  Don't cast SCM values to pointers.
  47
  48 2000-03-21  Dirk Herrmann  <D.Herrmann@tu-bs.de>
  49
  50         * symbols.h, symbols.c (scm_strhash):  Declare the string
  51         parameter as constant, since it is not modified.
  52
  53         * symbols.c (scm_intern_obarray_soft,
  54         scm_sysintern0_no_module_lookup):  Can now pass constant strings
  55         to scm_strhash without need for casting.
  56
  57 2000-03-21  Dirk Herrmann  <D.Herrmann@tu-bs.de>
  58
  59         * vectors.h (SCM_VELTS, SCM_VELTS_AS_STACKITEMS):  Don't cast SCM
  60         values to pointers.  Use SCM2PTR instead.
  61
  62 2000-03-21  Dirk Herrmann  <D.Herrmann@tu-bs.de>
  63
  64         * async.c (scm_set_tick_rate, scm_set_switch_rate):  Don't unpack
  65         results of SCM_INUM.
  66
  67 2000-03-21  Mikael Djurfeldt  <mdj@thalamus.nada.kth.se>
  68
  69         * init.c (scm_boot_guile_1): Renamed GUILE_GC_TRIGGER_1 -->
  70         GUILE_MIN_YIELD_1, GUILE_GC_TRIGGER_2 --> GUILE_MIN_YIELD_2.
  71         GUILE_MIN_YIELD_X now take *positive* fractions of heap size.
  72
  73         * gc.c, gc.h (SCM_MIN_YIELD_1, SCM_MIN_YIELD_2,
  74         min_yield_fraction, min_yield, adjust_min_yield): Renamed from
  75         SCM_GC_TRIGGER_1, SCM_GC_TRIGGER_2, gc_trigger_fraction,
  76         gc_trigger, adjust_gc_trigger.
  77
  78         * gc.c (alloc_some_heap): Further improvement of minimal heap size
  79         prediction.
  80         (SCM_MAX): New macro.
  81         (scm_freelist_t): New field: collected_1. Previous amount of
  82         collected cells.
  83         (gc_sweep_freelist_finish): Trigger based on two last values of
  84         freelist->collected to avoid unnecessary allocation due to
  85         temporary peaks.
  86         (SCM_INIT_HEAP_SIZE_1, SCM_GC_TRIGGER_1, SCM_GC_TRIGGER_2):
  87         Adjusted to 45000 cells, 40% and 40%.  Gives quick startup
  88         without extra heap allocation.
  89
  90 2000-03-20  Dirk Herrmann  <D.Herrmann@tu-bs.de>
  91
  92         * numbers.h (SCM_MAKINUM):  The parameter to SCM_MAKINUM should
  93         already be a C value.  No need to unpack it.
  94
  95         * numbers.h (scm_long_long2num):  Cast the parameter to scm_bits_t
  96         if we know it fits into an inum.
  97
  98         * ramap.c (ramap_rp):  An scm_tc7_[ui]vect object does point to a
  99         field of long values.  In contrast, SCM_VELTS accesses a field of
 100         SCM values.
 101
 102 2000-03-20  Mikael Djurfeldt  <mdj@mdj.nada.kth.se>
 103
 104         * gc.c (scm_gc_stats): Inserted explanation of local_scm_mtrigger
 105         etc.
 106         (scm_gc_yield_1): New variable: Holds previous yield.  Used to
 107         make better judgements.
 108         (gc_sweep_freelist_finish): Inserted explanation of use of
 109         gc_trigger.
 110
 111         * print.h, stacks.h, options.c, options.h: Changed C++
 112         commentaries to C.
 113
 114 2000-03-20  Dirk Herrmann  <D.Herrmann@tu-bs.de>
 115
 116         * tags.h (SCM2PTR, PTR2SCM):  Use SCM_PACK / SCM_UNPACK correctly.
 117
 118         * numbers.h (SCM_INUMP, SCM_MAKINUM, SCM_INUM0, SCM_COMPLEX_REAL,
 119         SCM_COMPLEX_IMAG, SCM_NUMP, SCM_BDIGITS):  Use SCM_PACK /
 120         SCM_UNPACK / SCM2PTR correctly.
 121
 122 2000-03-20  Mikael Djurfeldt  <mdj@mdj.nada.kth.se>
 123
 124         * gc.c (adjust_gc_trigger): Improved documentation.
 125         (alloc_some_heap): Since gc_trigger is used against
 126         freelist->collected, this is the value which should be used to
 127         predict minimum growth.
 128
 129 2000-03-20  Dirk Herrmann  <D.Herrmann@tu-bs.de>
 130
 131         * eval.h:  Fix mixup of packed/unpacked SCM values.  (Thanks
 132         Thien-Thi Nguyen for the patch.)
 133
 134 2000-03-20  Dirk Herrmann  <D.Herrmann@tu-bs.de>
 135
 136         * numbers.c (scm_ash):  Fixed typing problems with the second
 137         parameter and added some documentation.  (Thanks Thien-Thi Nguyen
 138         for indicating the problem.)
 139
 140 2000-03-19  Mikael Djurfeldt  <mdj@mdj.nada.kth.se>
 141
 142         * gc.c, gc.h (scm_gc_yield): New variable.
 143         (adjust_gc_trigger): Use scm_gc_yield.
 144         (alloc_some_heap): Use scm_gc_yield instead of
 145         scm_gc_cells_collected.
 146
 147         * coop-threads.c: Addd #include "root.h", #include "strings.h".
 148
 149         * debug.c: Added #include "root.h".  (Thanks to Thien-Thi Nguyen.)
 150
 151         * gc.c (scm_gc_for_newcell, adjust_gc_trigger): Improved GC
 152         trigger adjustmeant: Take yield (freed cells) for all freelists
 153         into account.
 154         (SCM_INIT_HEAP_SIZE_1, SCM_GC_TRIGGER_1, SCM_GC_TRIGGER_2): Tuned
 155         to 50000 cell heap with 45% trigger.
 156         (scm_gc_cells_collected): Reintroduced.
 157         (SCM_HEAP_SIZE): New macro.
 158         (scm_gc_sweep): Reintroduced correct computation of
 159         scm_cells_allocated.
 160         (scm_freelist_t): Corrected commentary for field `cluster_size':
 161         Clustersize counts objects, not cells;  New member
 162         `clusters_allocated'.
 163
 164 2000-03-19  Michael Livshin  <mlivshin@bigfoot.com>
 165
 166         * *.[hc]: add Emacs magic at the end of file, to ensure GNU
 167         indentation style.
 168
 169 2000-03-19  Mikael Djurfeldt  <mdj@thalamus.nada.kth.se>
 170
 171         * threads.h: Added #include "libguile/throw.h".  (Thanks to
 172         Thien-Thi Nguyen.)
 173
 174 2000-03-18  Michael Livshin  <mlivshin@bigfoot.com>
 175
 176         * tags.h: (SCM_DOUBLE_CELLP, SCM_NDOUBLE_CELLP): new macros.
 177
 178         * gc.h: (typedef struct scm_freelist_t) remove from here.
 179
 180         * gc.c: (CELL_UP, CELL_DN) make these macros take additional
 181         parameter (the span).
 182         (CLUSTER_SIZE_IN_BYTES, ALIGNMENT_SLACK) new macros.
 183         (typedef struct scm_freelist_t) move here from gc.h, it had no
 184         business being externally visible.
 185         (typedef struct scm_heap_seg_data_t) renamed from
 186         scm_heap_seg_data, to be style-compliant.
 187         (scm_mark_locations) if the possible pointer points to a
 188         multy-cell, check that it's properly aligned.
 189         (init_heap_seg) alighn multy-cells properly, work with the
 190         assumption that the segment size divides cleanly by cluster size
 191         (so that there's no spill).
 192         (round_to_cluster_size) new function.
 193         (alloc_some_heap, make_initial_segment) use round_to_cluster_size
 194         to satisfy the new init_heap_seg invariant.
 195
 196 2000-03-18  Dirk Herrmann  <D.Herrmann@tu-bs.de>
 197
 198         * _scm.h:  Don't include async.h everywhere...
 199
 200         * eq.c eval.c iselect.c: ... only include it here.
 201
 202 2000-03-18  Dirk Herrmann  <D.Herrmann@tu-bs.de>
 203
 204         * _scm.h:  Don't include root.h everywhere...
 205
 206         * async.c continuations.c eq.c eval.c evalext.c feature.c gc.c
 207         gdbint.c gsubr.c ioext.c keywords.c lang.c load.c macros.c
 208         numbers.c objprop.c ports.c print.c procprop.c ramap.c read.c
 209         srcprop.c stackchk.c stacks.c strports.c symbols.c unif.c
 210         variable.c vectors.c vports.c: ... only include it here.
 211
 212 2000-03-17  Dirk Herrmann  <D.Herrmann@tu-bs.de>
 213
 214         * _scm.h:  Don't include strings.h everywhere...
 215
 216         * backtrace.c dynl.c error.c feature.c filesys.c fports.c gc.c
 217         gdbint.c ioext.c load.c net_db.c numbers.c objects.c options.c
 218         ports.c posix.c print.c procs.c random.c read.c regex-posix.c
 219         simpos.c socket.c stacks.c stime.c strop.c strports.c struct.c
 220         symbols.c unif.c vectors.c version.c vports.c: ... only include it
 221         here.
 222
 223 2000-03-17  Dirk Herrmann  <D.Herrmann@tu-bs.de>
 224
 225         * _scm.h:  Don't include ports.h everywhere...
 226
 227         * arbiters.c backtrace.c debug.c dynl.c dynwind.c eval.c feature.c
 228         fluids.c gc.c gdbint.c guardians.c hash.c keywords.c mallocs.c
 229         numbers.c objects.c print.c read.c root.c smob.c srcprop.c
 230         stackchk.c strports.c struct.c throw.c variable.c:  ... only
 231         include it here.
 232
 233 2000-03-17  Dirk Herrmann  <D.Herrmann@tu-bs.de>
 234
 235         * _scm.h:  Don't include vectors.h everywhere...
 236
 237         * eq.c eval.c filesys.c gc.c gsubr.c guardians.c hash.c hashtab.c
 238         keywords.c net_db.c numbers.c objects.c posix.c print.c procprop.c
 239         procs.c ramap.c random.c read.c scmsigs.c socket.c sort.c stime.c
 240         strports.c symbols.c unif.c vports.c weaks.c:  ... only include it
 241         here.
 242
 243 2000-03-17  Dirk Herrmann  <D.Herrmann@tu-bs.de>
 244
 245         * genio.h:  removed.  (Only content was '/* delete me */'.)
 246
 247         * Makefile.am arbiters.c backtrace.c debug.c dynl.c dynwind.c
 248         error.c filesys.c fluids.c gc.c gsubr.c guardians.c keywords.c
 249         libguile.h mallocs.c numbers.c print.c random.c read.c root.c
 250         srcprop.c stackchk.c struct.c threads.c throw.c variable.c:
 251         Removed reference to genio.h
 252
 253 2000-03-17  Mikael Djurfeldt  <mdj@thalamus.nada.kth.se>
 254
 255         * gc.c, gc.h: Cleanup of the change of 2000-03-15.
 256         Cluster sizes are now independent of GC trigger values.
 257         GUILE_GC_TRIGGER_n can now specify a relative trigger value:
 258         A negative integer gives fraction of total heap size in percent.
 259         (SCM_GC_TRIGGER_1, SCM_GC_TRIGGER_2): Default values set to -40.
 260
 261         * init.c (scm_boot_guile_1): Introduced new variable
 262         GUILE_MAX_SEGMENT_SIZE; New environment variable names:
 263         GUILE_INIT_SEGMENT_SIZE_1, GUILE_GC_TRIGGER_1,
 264         GUILE_INIT_SEGMENT_SIZE_2, GUILE_GC_TRIGGER_2
 265
 266 2000-03-16  Mikael Djurfeldt  <mdj@mdj.nada.kth.se>
 267
 268         * __scm.h (GC_FREE_SEGMENTS): Disable this until we have made
 269         freeing of segment work with the new GC scheme.  (Thanks to
 270         Michael Livshin.)  Oops, also happened to make GUILE_NEW_GC_SCHEME
 271         the default, but I'll let this change stay in CVS Guile since this
 272         code is not expected to contain serious bugs.
 273
 274 2000-03-16  Mikael Djurfeldt  <mdj@thalamus.nada.kth.se>
 275
 276         * gc.c, gc.h (scm_map_free_list): Define also if GUILE_DEBUG is
 277         defined.
 278         (scm_free_list_length): New procedure (GUILE_DEBUG).
 279         Fixed a small but serious bug introduced by the previous change.
 280
 281         * gc.c (scm_gc_sweep): Moved variable n_objects to inner sweep
 282         loop and declare as register.
 283
 284         * gc.c (scm_gc_sweep): Sigh... forgot to clear private freelists
 285         after GC.
 286
 287 Wed Mar 15 08:27:04 2000  Greg J. Badros  <gjb@cs.washington.edu>
 288
 289         * fluids.c: Docstring patch from Keisuke Nishida.  Some
 290         reindentation, too, and a couple formals renamed.  Should
 291         fluid-set! return UNSPECIFIED instead of a value?
 292
 293 Wed Mar 15 08:24:58 2000  Greg J. Badros  <gjb@cs.washington.edu>
 294
 295         * Makefile.am: Separate out DOT_X_FILES and DOT_DOC_FILES, and
 296         generate the latter from the concrete listing of the former.  Then
 297         make guile-procedures.txt depend on DOT_DOC_FILES instead of
 298         *.doc, so that rebuilding it works.
 299
 300 Wed Mar 15 08:12:14 2000  Greg J. Badros  <gjb@cs.washington.edu>
 301
 302         * libguile.h: Include libguile/validate.h.  Thanks Keisuke Nishida!
 303
 304         * guile-snarf.awk.in: Replace docstring line-ending \n" and \n\n"
 305         with nothing and \n, respectively. Thanks Keisuke Nishida for
 306         noticing this problem.
 307
 308 2000-03-15  Mikael Djurfeldt  <mdj@mdj.nada.kth.se>
 309
 310         * __scm.h (GUILE_NEW_GC_SCHEME): Define this if you want to test a
 311         new way of allocating heap.  It makes Guile fast, but still
 312         contains bugs.
 313
 314         * gc.c, gc.h, pairs.h, init.c: Implementation of a new way of
 315         allocating heap.  The basic idea is to trigger GC every Nth
 316         allocated cell and grow heap when free list runs out.  The scheme
 317         has been extended so that GC isn't triggered until all remaining
 318         cells are used.  The implementation is also prepared for
 319         development in the direction of POSIX threads.
 320
 321         * gc.c (SCM_EXPHEAP): In order to grow by a factor of 1.5,
 322         SCM_EXPHEAP should return half of the heap size.
 323
 324 2000-03-14  Mikael Djurfeldt  <mdj@thalamus.nada.kth.se>
 325
 326         The following change to init.c is only enabled if Guile was
 327         configured with --enable-guile-debug.
 328
 329         * init.c (scm_i_getenv_int): New function.
 330         (scm_boot_guile_1): Use the environment variables
 331         GUILE_INIT_HEAP_SIZE, GUILE_INIT_HEAP_SIZE2 to select heap size if
 332         they exist.  (This may be replaced by a Scheme level mechanism in
 333         the future.)
 334
 335         * objprop.c (s_scm_set_object_property_x): Use scm_assq instead of
 336         scm_assoc.  (Thanks to Keisuke Nishida.)
 337
 338 2000-03-14  Mikael Djurfeldt  <mdj@mdj.nada.kth.se>
 339
 340         * eval.c, lang.c, lang.h (scm_lisp_nil, scm_lisp_t): Renamed from
 341         scm_nil, scm_t.  (Thanks to Keisuke Nishida.)
 342
 343 2000-03-14  Mikael Djurfeldt  <mdj@thalamus.nada.kth.se>
 344
 345         * init.c (scm_boot_guile_1): Use same initial segment size for
 346         1-word and 2-word segments.  Having the smaller size causes Guile
 347         to GC too often.  Obviously something needs to be done to allow
 348         for a smaller 2-word segment without this to happen.  (The amount
 349         of heap for each type should be automatically adapted to the
 350         application somehow.)
 351
 352         [Almost all of these changes should be documented in the NEWS
 353         file.]
 354
 355         * gc.h (scm_freelist_t): New type.
 356
 357         * pairs.h (SCM_NEWCELL, SCM_NEWCELL2): Use new style freelists.
 358
 359         * gc.c (SCM_INIT_HEAP_SIZE): Changed from 32768 --> 40000 so that
 360         all of Guile basics fits into one segment and there suitable room
 361         for work.
 362         (SCM_EXPHEAP): Now takes an argument.  Grow by a factor of 1.5
 363         instead of 2.
 364         (scm_freelist, scm_freelist2): Now of type scm_freelist_t.
 365         Freelists now contains information about object span, cells
 366         collected and amount of cells in heap segments belonging to the
 367         list.
 368         (scm_heap_size, scm_gc_cells_collected): Removed.
 369
 370         * init.c (scm_boot_guile_1): Make 2-word segment 8K (512 cells).
 371
 372         * Makefile.am (libguile_la_LDFLAGS): Bumped library version
 373         number.
 374
 375         * __scm.h eq.c, eval.c, gc.c, gc.h, gh_data, hash.c, numbers.c,
 376         numbers.h, objects.c, ramap.c, random.c, unif.c, unif.h: Extensive
 377         rewrite of handling of real and complex numbers.
 378         (SCM_FLOATS, SCM_SINGLES): These #ifdef conditionals have been
 379         removed along with the support for floats.  (Float vectors are
 380         still supported.)
 381
 382         * tags.h (scm_tcs_bignums): Removed.
 383         (scm_tc16_bigpos, scm_tc16_bigneg): Replaced by scm_tc16_big.
 384         Use SCM_BIGSIGN(x) to test for sign!
 385         (scm_tc16_big): The new bignum type.
 386         (SCM_REAL_PART, SCM_IMAG_PART): Removed.
 387
 388         * numbers.h (SCM_BIGSIGN): Sign moved to bit 16.
 389         (scm_makdbl): Deprecated.
 390         (SCM_NEWREAL, SCM_NEWCOMPLEX): New macros.
 391         (SCM_SINGP): Deprecated.
 392         (SCM_FLO): Removed.
 393         (SCM_INEXP, SCM_CPLXP): Deprecated.
 394         (SCM_INEXACTP, SCM_COMPLEXP): New macros.
 395         (SCM_COMPLEX_REAL, SCM_COMPLEX_IMAG): Renamed from
 396         SCM_REAL, SCM_IMAG (and now only valid for complex numbers).
 397         (SCM_REAL, SCM_IMAG): New, *deprecated*, selectors which work both
 398         for doubles and complex numbers.
 399         (SCM_REAL_VALUE): New selector for doubles.
 400         (scm_double_t, scm_complex_t): New types.
 401         (scm_dbl): Removed.
 402
 403         * numbers.c (scm_floprint, scm_floequal): Removed.
 404         (scm_print_real, scm_print_complex, scm_real_equalp,
 405         scm_complex_equalp): New functions.
 406
 407         * unif.c (scm_makflo): Removed.
 408
 409         * smob.h (SCM_SMOB_PREDICATE): New macro.
 410         (SCM_NEWSMOB2, SCM_RETURN_NEWSMOB2, SCM_NEWSMOB3,
 411         SCM_RETURN_NEWSMOB3): New macros.
 412
 413 1999-11-21  Michael Livshin  <mlivshin@bigfoot.com>
 414
 415         The following changes implement primitive support for double cells
 416         (i.e. four-word cells) and change the representation of some
 417         things to multi-cells instead of cons+malloc.  (Applied and
 418         modified by mdj.)
 419
 420         * pairs.h (SCM_NEWCELL2): double-cell variants of SCM_NEWCELL.
 421         (SCM_CELL_WORD, SCM_CELL_WORDLOC, SCM_SET_CELL_WORD): primitive
 422         multi-cell access macros (used by the ones below).
 423         (SCM_CELL_WORD[0-3], SCM_SET_CELL_WORD[0-3]): multi-cell access
 424         macros.
 425
 426         * gc.c (scm_freelist2): multi-cell freelists.
 427         (inner_map_free_list): map_free_list, parameterized on ncells.
 428         "nn cells in segment mm" was misleading for ncells > 1; changed to
 429         "objects".  still print cells too, though.
 430         (scm_map_free_list): rewritten using inner_map_free_list.
 431         (scm_check_freelist): get freelist as parameter, since now we have
 432         more than one.
 433         (scm_debug_newcell2): multi-cell variants of
 434         scm_debug_newcell.
 435         (scm_gc_for_newcell): take ncells and freelist pointer as
 436         parameters.
 437         (scm_gc_mark): add case for tc7_pws (procedures with setters are
 438         now double cells).
 439         (scm_gc_sweep): don't free the float data, since it's not malloced
 440         anymore.
 441         (init_heap_seg): didn't understand what n_new_objects stood for,
 442         so changed to n_new_cells.
 443         (make_initial_segment): new function, makes an initial segment
 444         according to given ncells.
 445         (scm_init_storage): call make_initial_segment, for ncells={1,2,3}.
 446
 447         * numbers.c (scm_makdbl): no malloc'ing needed, so the
 448         {DEFER,ALLOW}_INTS thing removed.
 449
 450         * numbers.h (struct scm_dbl): changed to represent a double cell,
 451         with the number in the second half.
 452
 453         * dynwind.c: changed the wind-guards representation to double
 454         cell.
 455
 456         * procs.c, procs.h: changed the procedure-with-setter representation
 457         to double cell.
 458
 459         * async.c, async.h: made async representation a double cell.
 460
 461         * dynl.c: made dynamic_obj representation a double cell.
 462
 463 2000-03-13  Gary Houston  <ghouston@arglist.com>
 464
 465         * ports.c (flush_void_port): renamed to flush_port_default.
 466         (end_input_void_port): renamed to end_input_default.
 467
 468         * init.c (scm_standard_stream_to_port): create a void port instead
 469         of opening /dev/null if the standard file descriptors are bad.
 470         advantages: no portability problems, doesn't waste a file descriptor,
 471         simplifies the code (thanks to Marius for the idea).
 472
 473         * vports.c (s_scm_make_soft_port): call scm_port_non_buffer.
 474
 475         * void ports: make reading from a void port give EOF instead of
 476         segv:
 477         * ports.c (s_scm_sys_make_void_port): modified docstring.
 478         (fill_input_void_port): new proc.
 479         (scm_init_ports): set up fill_input_void_port.
 480         * ports.c (scm_port_non_buffer): new proc.
 481         (scm_void_port): call scm_port_non_buffer.
 482
 483         * fports.c (scm_setvbuf): docstring: remove the fcntl documentation
 484         which was incorrectly appended.
 485
 486 2000-03-13  Mikael Djurfeldt  <mdj@mdj.nada.kth.se>
 487
 488         * guile-doc-snarf.in: Don't use absolute path for `sed'.  (Note
 489         that we can't use autoconf for this.  Autoconf itself relies on
 490         the existence of `sed' somewhere on your path.)  (Thanks to Dirk
 491         Herrman.)
 492
 493 2000-03-13  Mikael Djurfeldt  <mdj@thalamus.nada.kth.se>
 494
 495         * Makefile.am (libguile_la_SOURCES): Moved iselect.c here from
 496         EXTRA_libguile_la_SOURCES.
 497
 498         * iselect.h: Always declare scm_internal_select.
 499
 500         * iselect.c (scm_internal_select): Added SCM_ASYNC_TICK at the
 501         end.  Also let scm_internal_select be a real function also when
 502         not using threads.
 503
 504         * __scm.h (SCM_TICK): Oops! Forgot to put SCM_ASYNC_TICK here...
 505
 506 2000-03-13  Mikael Djurfeldt  <mdj@mdj.nada.kth.se>
 507
 508         * __scm.h (SCM_ALLOW_INTS, SCM_REALLOW_INTS): Removed call to
 509         SCM_ASYNC_TICK.  (This is a preparation for POSIX threads support,
 510         and kind of an experiment: Will this cause problems?)
 511
 512 Sun Mar 12 13:26:30 2000  Greg J. Badros  <gjb@cs.washington.edu>
 513
 514         * Makefile.am: Added *.doc to DISTCLEANFILES.
 515
 516 2000-03-12  Gary Houston  <ghouston@arglist.com>
 517
 518         * fports.c (scm_fdes_to_port): call fcntl F_GETFL to test that
 519         the fdes is valid before doing anything else.  check that
 520         the file descriptor supports the modes required.
 521         (scm_fport_buffer_add): don't throw an error if fstat doesn't
 522         work: just use the default buffer size.
 523
 524         * throw.c: change an outdated comment about scm_internal_catch
 525         BODY: it doesn't take a jumpbuf arg.
 526
 527         * init.c (scm_standard_stream_to_port): install a handler in case
 528         scm_fdes_to_port throws an error.  don't check here whether the
 529         file descriptor is valid, since scm_fdes_to_port will do that.
 530         set the revealed count depending on whether the port got the
 531         standard file descriptor.
 532         (stream_body_data): new type.
 533         (stream_body, stream_handler): new procs.
 534
 535 2000-03-12  Mikael Djurfeldt  <mdj@thalamus.nada.kth.se>
 536
 537         * stacks.c, stacks.h, struct.c, tags.h, unif.c (scm_bits_t):
 538         Renamed from SCMWORD.
 539
 540         * tags.h (SCM_NCELLP): Removed (SCMWORD).
 541
 542         * arbiters.c (SCM_ARB_LOCKED): Use SCM_UNPACK_CAR.
 543
 544         * async.c, boolean.h, debug.c, dynl.c, dynwind.c, eval.c, eval.h,
 545           feature.h, filesys.h, fluids.h, fports.c, fports.h, gc.c, gc.h,
 546           hash.c, keywords.h, macros.c, numbers.c, numbers.h, objects.c,
 547           objects.h, options.c, pairs.h, ports.c, ports.h, print.c,
 548           procs.h, ramap.c, read.c, smob.c, smob.h, srcprop.h, stacks.c,
 549           stacks.h, strports.c, struct.c, struct.h, tag.c, tags.h,
 550           throw.c, unif.c, unif.h, variable.h, vectors.h, weaks.c,
 551           weaks.h (SCM_PACK, SCM_UNPACK, SCM_UNPACK_CAR): Renamed from
 552           SCM_ASSCM, SCM_ASWORD, SCM_CARW).
 553
 554         * numbers.h (SCM_SRS, SCM_INUM): Corrected SCM_ASSCM/ASWORD fixes.
 555
 556         * alist.c, eval.c, net_db.c, posix.c, print.c, snarf.h, struct.c,
 557         tags.h: Fixed copyright notices.
 558
 559         * struct.c, coop-threads.c: SCM_ASSCM/ASWORD fixes.
 560
 561 2000-03-12  Marius Vollmer  <mvo@zagadka.ping.de>
 562
 563         * init.c (scm_standard_stream_to_port): Check whether the file
 564         descriptor is valid and substitute "/dev/null" when not.
 565
 566 2000-03-12  Mikael Djurfeldt  <mdj@thalamus.nada.kth.se>
 567
 568         * coop-defs.h (struct timespec): Conditionally defined.
 569
 570         * coop.c (coop_condition_variable_timed_wait_mutex): Use ETIMEDOUT
 571         instead of ETIME.
 572
 573         * readline.c (match_paren): Bugfix: First arg to select is not
 574         number of descriptors but the number of the highest descriptor +
 575         1.
 576
 577         This is a preliminary attempt at a cleanup of the threads support
 578         code.  It moves things to better places, makes arguments more
 579         consistent with the POSIX API (which is used in GNOME's glib), and
 580         adds new functionality.
 581
 582         * readline.c (scm_init_readline): Added new arg to scm_init_mutex.
 583
 584         * coop-defs.h (scm_mutex_trylock): New macro: alias for
 585         coop_mutex_trylock.
 586         (scm_cond_init): Changed definition to
 587         coop_new_condition_variable_init.
 588
 589         * coop.c: #include <errno.h>
 590         (coop_timeout_qinsert): Moved here from iselect.c
 591         (coop_new_mutex_init, coop_new_condition_variable_init): New
 592         functions.  The strange names are temporary.  Use scm_mutex_init
 593         and scm_cond_init instead.
 594         (coop_mutex_trylock): New function.  Uses errno.h:EBUSY.  errno.h
 595         is ANSI C, but should we check for individual error codes in
 596         configure.in?
 597         (coop_condition_variable_timed_wait_mutex): New function.
 598         (coop_key_create, coop_setspecific, coop_getspecific,
 599         coop_key_delete): New functions.
 600
 601         * iselect.c (coop_timout_qinsert): Moved to coop.c
 602
 603 2000-03-11  Mikael Djurfeldt  <mdj@thalamus.nada.kth.se>
 604
 605         * pairs.h (SCM_SETAND_CAR, SCM_SETAND_CDR, SCM_SETOR_CAR,
 606         SCM_SETOR_CDR): Corrected SCM_ASSCM/WORD fixes.
 607
 608         * tags.h (SCM_VOIDP_TEST): Renamed from VOIDP_TEST.
 609         Layout cleanups.
 610
 611         * objects.h (SCM_CLASS_FLAGS, SCM_OBJ_CLASS_FLAGS,
 612         SCM_SET_CLASS_INSTANCE_SIZE), struct.h (SCM_STRUCT_VTABLE_DATA),
 613         proc.h (SCM_CLOSCAR): SCM_ASSCM/WORD fixes.
 614
 615         * eval.c (scm_lookupcar1): Inserted SCM_ASWORD in expressions
 616         dealing with ilocs.
 617
 618 2000-03-11  Dale P. Smith  <dpsm@en.com>, applied by Greg J. Badros, <gjb@cs.washington.edu>
 619
 620         * numbers.c (scm_copy_big_dec, scm_copy_smaller, scm_big_ior,
 621         scm_big_xor, scm_big_and, scm_big_test): Added new lowlevel bignum
 622         logical functions from SCM.
 623
 624         (logand, logior, logxor, logtest, logbit?): Extended scheme
 625         logical functions to use bignums from SCM.
 626
 627         (lognot): Removed call to `SCM_VALIDATE_INUM' that prevented
 628         lognot from using bignums.
 629
 630 Thu Mar  9 11:33:25 2000  Greg J. Badros  <gjb@cs.washington.edu>
 631
 632         * vectors.h (SCM_VELTS_AS_STACKITEMS): Added this macro to help in
 633         eliminating some warnings.
 634
 635         * unif.c, strports.c, print.c, options.c: Fix some warnings on
 636         mis-use of SCM/long
 637
 638         * gc.c, gc.h: Added scm_return_first_int(), and added comment re:
 639         what the scm_return_first* functions do.
 640
 641 2000-03-09  Han-Wen Nienhuys  <hanwen@cs.uu.nl>, applied by Greg J. Badros, <gjb@cs.washington.edu>
 642
 643        * libguile/*.[ch]: make a distinction between SCM as a generic
 644        name for a Scheme object (now a void*), and SCM as 32 bit word for
 645        storing tags and immediates (now a long int).  Introduced
 646        SCM_ASWORD and SCM_ASSCM for conversion. Fixed various dubious
 647        code in the process: arbiter.c (use macros), unif.c (scm_array_p),
 648
 649 Wed Mar  8 10:15:59 2000  Greg J. Badros  <gjb@cs.washington.edu>
 650
 651         * numbers.c:  Use SCM_VALIDATE_LONG_COPY, and longs, not ints, in
 652         various logXXX primitives.  Thanks Eric Moore!
 653
 654 Tue Mar  7 08:05:22 2000  Greg J. Badros  <gjb@cs.washington.edu>
 655
 656         * run-test, remaining-docs-needed:  Added these scripts.  The
 657         second one is only temporary until the docstring additions are
 658         complete.   run-test may best live on, but is here mostly for
 659         convenience and awareness for now.
 660
 661         * hash.c: Docs, minor cleanup patch from Dirk Herrman.
 662
 663 Thu Mar  2 16:06:58 2000  Greg J. Badros  <gjb@cs.washington.edu>
 664
 665         * error.h, error.c: Added `scm_wrong_type_arg_msg' to support
 666         displaying the expected type.  Use SCM_LISTn in a couple places
 667         instead of scm_cons-ing by hand.
 668
 669         * __scm.h: Added SCM_ASSERT_TYPE macro.
 670
 671         * validate.h, scm_validate.h: Added the former, as a renamed
 672         version of the latter with SCM_ASSERT_TYPE used in
 673         SCM_MAKE_VALIDATE (instead of just SCM_ASSERT)
 674
 675         * Makefile.am: Rename scm_validate.h to validate.h.
 676
 677         * *.c, *.h: Include validate.h, not scm_validate.h (old name's
 678         prefix was superfluous).
 679
 680 Thu Mar  2 15:33:12 2000  Greg J. Badros  <gjb@cs.washington.edu>
 681
 682         * hashtab.c: Improved documentation for lots of functions.  Added
 683         handwritten docs for `hash-fold'.
 684
 685 Thu Mar  2 15:13:25 2000  Greg J. Badros  <gjb@cs.washington.edu>
 686
 687         * list.c: Added hand-written docs for `del{q,v,ete}1!'.
 688
 689 Thu Mar  2 12:38:30 2000  Greg J. Badros  <gjb@cs.washington.edu>
 690
 691         * list.c: Moved append docs to append! Thanks Dirk Hermann.  Also,
 692         added append docs from R4RS.
 693
 694         * strings.c: Docstring typo fix, + eliminate unneeded IMP tests.
 695         Thanks Dirk Hermann!
 696
 697         * chars.h: Provide SCM_CHARP, SCM_CHAR, SCM_MAKE_CHAR and
 698         deprecate SCM_ICHRP, SCM_ICHR, SCM_MAKICHR.  Thanks Dirk Hermann!
 699
 700         * *.h, *.c: Use SCM_CHARP, SCM_CHAR, SCM_MAKE_CHAR throughout.
 701         Drop use of SCM_P for function prototypes... assume an ANSI C
 702         compiler.  Thanks Dirk Hermann!
 703
 704 Sat Feb 19 12:20:12 2000  Greg J. Badros  <gjb@cs.washington.edu>
 705
 706         * ports.c: Made `set-port-column!' and `set-port-line!' each
 707         return SCM_UNSPECIFIED instead of a (not-scheme-object) integer
 708         that caused a seg fault.  Also fixed `set-port-column!'s
 709         docstring.  Thanks Han-Wen Nienhuys for finding the bug!
 710
 711 Sun Feb 13 19:11:42 2000  Greg J. Badros  <gjb@cs.washington.edu>
 712
 713         * arbiters.c, eq.c, gc.c, guardians.c, list.c, ports.c, print.c,
 714         regex-posix.c, scmsigs.c, stime.c, strings.c, variable.c, stime.c,
 715         strings.c, variable.c:  Added lots of documentation, cleaned up
 716         some existing documentation.  Occasionally changed formal params
 717         to match docs.  Also folded an #ifdef into the inners of a
 718         primitive instead of having two copies of the primitive
 719         (`get-internal-real-time', from stime.c)
 720
 721 Sun Feb 13 18:12:19 2000  Greg J. Badros  <gjb@cs.washington.edu>
 722
 723         * ports.c: Added docs for primitives missing them.  Written by
 724         hand.
 725
 726 Sun Feb 13 09:40:36 2000  Greg J. Badros  <gjb@cs.washington.edu>
 727
 728         * guile-doc-snarf.in: Use ${AWK} -f guile-func-name-check, not
 729         just execing guile-func-name-check.  Thanks Michael Livshin!
 730
 731 Thu Feb 10 11:43:23 2000  Greg J. Badros  <gjb@cs.washington.edu>
 732
 733         * guile-snarf.awk.in: Tweak to work with Sun/HP awk, removed some
 734         dead code.  Patch from Michael Livshin.
 735
 736         * guile-doc-snarf.in: Tweak to work with Sun/HP sh.  Patch from
 737         Michael Livshin.
 738
 739 2000-02-09  Gary Houston  <ghouston@arglist.com>
 740
 741         * init.c (scm_init_standard_ports): when stdout is a tty, make the
 742         current-output-port unbuffered by default.  this is less confusing
 743         for interactive use.  it was line-buffered because of a
 744         performance problem with unbuffered ports, but I think it will be
 745         OK now.
 746
 747 2000-02-08  Gary Houston  <ghouston@arglist.com>
 748
 749         * __scm.h: don't define long_long or ulong_long if HAVE_LONG_LONGS
 750         is not defined.
 751
 752         * stime.c (scm_localtime, scm_mktime): if neither HAVE_TM_ZONE nor
 753         HAVE_TZNAME are defined, use an empty string instead of giving two
 754         spurious compile-time errors.
 755
 756 Tue Feb  8 13:57:46 2000  Greg J. Badros  <gjb@cs.washington.edu>
 757
 758         * ports.c: Doc patches from Richard Kim.  Pasted from MIT Scheme.
 759         Thanks Richard!
 760
 761 Mon Feb  7 09:07:31 2000  Greg J. Badros  <gjb@cs.washington.edu>
 762
 763         * macros.c: Doc patches from Richard Kim.  Pasted from scm.texi.
 764
 765 Sun Feb  6 20:26:21 2000  Greg J. Badros  <gjb@cs.washington.edu>
 766
 767         * pairs.c: Doc patches from Richard Kim.  Pasted from MIT Scheme
 768         (which is GNU GPL'd).
 769
 770 2000-01-31  Gary Houston  <ghouston@arglist.com>
 771
 772         * strings.h: don't use SCM_P.  don't include <string.h>.
 773         * error.c, gh_data.c, ports.c, script.c, strop.c: include <string.h>.
 774
 775         * strings.c (scm_string_ref): make the 2nd argument compulsory.
 776         previously it defaulted to zero for no good reason that I can see.
 777         use a local variable for SCM_INUM (k).  replace
 778         SCM_VALIDATE_INUM_DEF with SCM_VALIDATE_INUM_COPY.
 779
 780         (scm_makfromstr): cosmetic changes.
 781
 782         (scm_string): Accept only chars in the list, not strings, for
 783         conformance to R5RS (particularly for list->string, which is
 784         supposed to be the inverse of string->list.)  remove
 785         SCM_DEFER_INTS/SCM_ALLOW_INTS, which is unnecessary since
 786         scm_makstr handles the cell allocation.  when reporting wrong-type
 787         arg, don't report the position as 1.
 788
 789         * posix.c (scm_init_posix): intern PIPE_BUF if it's defined.
 790
 791 2000-01-29  Gary Houston  <ghouston@arglist.com>
 792
 793         * posix.c (scm_pipe): rewrote the docstring.
 794
 795         * filesys.c (scm_select, retrieve_select_type, get_element,
 796         fill_select_type, set_element): modified so that Scheme
 797         "select" tests port buffers for the ability to provide input
 798         or accept output.  Previously only the underlying file descriptors
 799         were checked.  Rewrote the docstring.
 800
 801 Thu Jan 27 10:14:25 2000  Greg J. Badros  <gjb@cs.washington.edu>
 802
 803         * vectors.c, symbols.c, strorder.c: Documentation cut and pasted
 804         from Gregg Reynolds.  Thanks Gregg!
 805
 806 Thu Jan 27 09:59:38 2000  Greg J. Badros  <gjb@cs.washington.edu>
 807
 808         * strop.c (scm_i_index): Obfuscated commented-out SCM_DEFINE by
 809         adding "x" prefix to the line so that guile-func-name-check
 810         doesn't complain unnecessarily.
 811
 812 Wed Jan 26 17:33:52 2000  Greg J. Badros  <gjb@cs.washington.edu>
 813
 814         * throw.c: Factor out an #ifdef/#else/#endif choice more finely
 815         for maintainability.
 816
 817         * strop.c: Documentation added by Gregg A. Reynolds.   Pasted in
 818         from qdocs, RnRs.
 819
 820 Wed Jan 26 10:02:11 2000  Greg J. Badros  <gjb@cs.washington.edu>
 821
 822         * tag.c:  Added doc for `tag', but mark as deprecated since Mikael
 823         suggests removing tag.c altogether (and using a new `class-of'
 824         instead).
 825
 826         * strings.c:  Added documentation from Gregg A. Reynolds.  Edited
 827         a bit by me to use FOO instead of @var{foo} and to have the
 828         summary come before preconditions on input.  Also dropped trailing
 829         (rnrs) note.
 830
 831         * gsubr.c: Do not use SCM_DEFINE for `gsubr-apply'. Register the
 832         function with scm_make_subr_opt w/ last arg of 0 so it is not
 833         visible at the Scheme level. Mikael says that this is the right
 834         thing because the first arg to the proc is the guts of a compiled
 835         closure and shouldn't be exposed to the Scheme level.
 836
 837 Tue Jan 25 17:15:47 2000  Greg J. Badros  <gjb@cs.washington.edu>
 838
 839         * sort.c: typo in comment fixed.
 840
 841         * keywords.c: Added documentation.
 842
 843         * guardians.c: Added documentation (could be better).
 844
 845         * gc.c: Added docs for gc-set-debug-check-freelist.
 846
 847         * eq.c: Added docs for eq?, eqv? equal? abridged from R4RS.
 848
 849         * boolean.c: Added docs for `not', `boolean?' (by hand).
 850
 851 Tue Jan 25 13:28:56 2000  Greg J. Badros  <gjb@cs.washington.edu>
 852
 853         * random.c: Added documentation, from SLIB page:
 854         http://angela.ctrl-c.liu.se/~calle/scheme/slib_toc.html
 855
 856 Mon Jan 24 17:50:20 2000  Greg J. Badros  <gjb@cs.washington.edu>
 857
 858         * variable.c, version.c:  Added documentation, written by hand
 859         since I could not find anything already written that was
 860         relevant.
 861
 862 2000-01-23  Gary Houston  <ghouston@arglist.com>
 863
 864         * filesys.c (scm_chown): omit port/fdes support if HAVE_FCHOWN is
 865         not defined (thanks to Richard Y. Kim).
 866
 867 Thu Jan 20 13:00:38 2000  Greg J. Badros  <gjb@cs.washington.edu>
 868
 869         * Makefile.in: Removed, this is auto-generated.
 870
 871 Thu Jan 20 11:33:47 2000  Dirk Hermann <dirk@ida.ing.tu-bs.de> --applied 01/20/00 gjb
 872
 873         * list.c: Put some variable initialization code at the point of
 874         declaration; Added a comment for list*; Formatting changes.
 875
 876         * load.c: use SCM_NNULLP to make sure the end of a list is not
 877         reached yet.
 878
 879 2000-01-18  Mikael Djurfeldt  <mdj@mdj.nada.kth.se>
 880
 881         * backtrace.c (scm_display_error_message): Bugfix: Don't use
 882         result of scm_list_p as C boolean.
 883         (scm_display_error_message, scm_set_print_params_x): Use new
 884         validation macros. (Thanks to Dirk Herrmann.)
 885
 886         * net_db.c (scm_resolv_error): Cast result from hstrerror.
 887
 888         * strports.c (st_end_input): Inserted parenthesis to get operator
 889         grouping correct.
 890
 891         * list.h (scm_init_list): Removed SCM_P around prototypes.
 892
 893         * fports.c, list.c, numbers.c, ports.c, stime.c, symbols.c,
 894         filesys.c, posix.c: Converted docstrings to ANSI C format and
 895         escaped " occurring inside string literals.
 896
 897 Tue Jan 18 13:21:08 2000  Mikael Djurfeldt  <mdj@r11n07-s.pdc.kth.se>
 898
 899         * posix.c (scm_mknod): Escape " occuring inside docstring.
 900
 901 2000-01-18  Mikael Djurfeldt  <mdj@mdj.nada.kth.se>
 902
 903         * alist.c, chars.c, debug.c, dynl.c, dynwind.c, error.c, eval.c,
 904            evalext.c, filesys.c, gc.c, hash.c, hashtab.c, ioext.c,
 905            keywords.c, list.c, load.c, macros.c, net_db.c, numbers.c,
 906            objprop.c, ports.c, posix.c, print.c, procprop.c, procs.c,
 907            ramap.c, regex-posix.c, root.c, scmsigs.c, simpos.c, socket.c,
 908            stacks.c, stime.c, strings.c, strop.c, strports.c, struct.c,
 909            symbols.c, throw.c, unif.c, vectors.c, version.c, vports.c,
 910            weaks.c: Converted docstrings to ANSI C format.
 911
 912         * filesys.c (scm_chmod), simpos.c (scm_system), version
 913         (scm_version), vports (scm_make_soft_port): Escape " occuring
 914         inside docstring.
 915
 916 Mon Jan 17 11:41:22 2000  Greg J. Badros  <gjb@cs.washington.edu>
 917
 918         * scm_validate.h: Added SCM_VALIDATE_ULONG_COPY,
 919         SCM_VALIDATE_LONG_COPY
 920
 921         * numbers.c: Use SCM_VALIDATE_ULONG_COPY, instead of
 922         SCM_VALIDATE_INUM_COPY to let bigger numbers be used.  Rename a
 923         couple of formal arguments (and fix their uses) to make arguments
 924         match the documentation.
 925
 926 2000-01-14    <mstachow@alum.mit.edu>
 927
 928         * Makefile.am: Augment path when running guile-doc-snarf so
 929         guile-func-name-check is found.
 930
 931 Fri Jan 14 09:34:55 2000  Greg J. Badros  <gjb@cs.washington.edu>
 932
 933         * scm_validate.h (SCM_NUM2LONG_DEF): Fix this macro to just use
 934         def, not SCM_MAKINUM(def);  thanks Janis Bzerins!
 935
 936 Wed Jan 12 00:06:53 2000  Greg J. Badros  <gjb@cs.washington.edu>
 937
 938         * net_db.c (s_scm_inet_makeaddr): Use SCM_NUM2ULONG since that's
 939         the way guile-1.3.4 worked, but #if 0 out the version using
 940         SCM_VALIDATE_INUM_COPY for stricter testing.
 941
 942 Tue Jan 11 18:24:18 2000  Greg J. Badros  <gjb@cs.washington.edu>
 943
 944         * guile-doc-snarf.in: Use new $fullfilename for running
 945         guile-func-name-check, and put "$fullfilename" and "$filename" in
 946         quotes at uses to make sure re-splitting on whitespace does not
 947         occur (so filenames w/ embedded whitespace would work okay, though
 948         I sure hope we never have to deal with that! :-) ).  Thanks to
 949         Mikael for pointing out the source_dir != build_dir was broken.
 950
 951 Tue Jan 11 17:42:40 2000  Greg J. Badros  <gjb@cs.washington.edu>
 952
 953         * scm_validate.h: Added SCM_NUM2LONG_DEF macro.  Make
 954         SCM_OUT_OF_RANGE use SCM_MAKINUM, not scm_long2num.  Added
 955         SCM_COERCE_ROSTRING macro.  Added SCM_VALIDATE_NONEMPTYLIST
 956         macro.  Fix SCM_VALIDATE_STRINGORSUBSTR macro to not use SLOPPY
 957         variants.
 958
 959         * ports.c (scm_port_closed_p): Validate that the arg is a PORT,
 960         then return whether it's an open port (was validating that it was
 961         an open port -- this was a bug I introduced back in December, but
 962         my careful reading of diffs caught it).
 963
 964         * numbers.c: Recombine the two conditional-compilation paths for
 965         all the log* primitives -- they were split based on #ifndef
 966         scm_long2num;  factored out a SCM_LOGOP_RETURN macro, and fixed
 967         some bugs and inconsistencies in the two sets of implementations.
 968         (scm_lognot) Fixed *atrocious* re-use of a SCM as an integer!
 969
 970         * ioext.c: Use SCM_ASSERT_RANGE in a couple places, and
 971         SCM_VALIDATE_INUM_COPY once where it should've been used.
 972
 973         * fluids.c (scm_internal_with_fluids): Use
 974         SCM_VALIDATE_LIST_COPYLEN.
 975
 976         * filesys.c: Use SCM_NUM2LONG instead of SCM_VALIDATE_INUM_COPY;
 977         this is questionable as it relaxes type safety, but other changes
 978         were useful and all SCM_NUM2LONG's should probably be
 979         revisited. Use SCM_OUT_OF_RANGE, SCM_WRONG_TYPE_ARG.
 980
 981         * evalext.c: line-break change on 1 line.
 982
 983         * eval.c (nconc2last): Takes a non-empty list as its first
 984         argument, not just a list.
 985
 986         * dynl.c: Use new SCM_COERCE_ROSTRING macro.
 987
 988 Tue Jan 11 15:44:23 2000  Greg J. Badros  <gjb@cs.washington.edu>
 989
 990         * dynl.c, feature.c, filesys.c, fports.c, list.c, load.c,
 991         net_db.c, sort.c, stacks.c, unif.c: Use SCM_WTA, SCM_MISC_ERROR
 992         where possible.
 993
 994         * symbols.c (scm_sysintern0): Fixed the function name in a
 995         scm_misc_error invocation.
 996
 997         * print.c (scm_simple_format): Do not need SCM_COERCE_SUBSTR, and
 998         use scm_return_first to ward off latent GC bug that Mikael caught.
 999
1000         * async.c: Use SCM_VALIDATE_ASYNC_COPY one place where it wasn't
1001         used before but should've been.
1002
1003 2000-01-12  Mikael Djurfeldt  <mdj@thalamus.nada.kth.se>
1004
1005         * snarf.h (SCM_PROC1): Replaced SCM (*) (...) with
1006         SCM_FUNC_CAST_ARBITRARY_ARGS.
1007
1008 Tue Jan 11 13:44:07 2000  Greg J. Badros  <gjb@cs.washington.edu>
1009
1010         * guile-func-name-check.in: Added this script to statically check
1011         #define FUNC_NAME, #undef FUNC_NAME in the source.
1012
1013         * sort.c, posix.c: Fix #undef FUNC_NAME lines to not have trailing
1014         redundant comment, semicolon; caught by new guile-func-name-check
1015         script.
1016
1017         * debug.c: Fix mistaken #define FUNC_NAME for scm_make_iloc.
1018         Caught by new guile-func-name-check-script.
1019
1020         * Makefile.am: Added guile-func-name-check to bin_SCRIPTS
1021
1022         * ramap.c: Fix #if 0'd out code to be syntactically acceptable to
1023         guile-func-name-check.
1024
1025         * guile-doc-snarf.in: Run guile-func-name-check on the file before
1026         doing the snarf.
1027
1028 Tue Jan 11 11:31:10 2000  Greg J. Badros  <gjb@cs.washington.edu>
1029
1030         * fports.c, ports.c, ports.h, strports.c, vports.c: Make write
1031         port function take const void*, not void*.
1032
1033 Tue Jan 11 11:18:07 2000  Greg J. Badros  <gjb@cs.washington.edu>
1034
1035         * scm_validate.h, chars.c, ports.c, print.c, read.c, strings.c,
1036         strop.c: Use SCM_VALIDATE_ICHR, SCM_VALIDATE_ICHR_COPY instead of
1037         SCM_VALIDATE_CHAR, SCM_VALIDATE_CHAR_COPY.  Change made for
1038         consistency with the other macros dealing with immediate
1039         characters.  (Similar to INT -> INUM change a week or so ago).
1040
1041 Tue Jan 11 10:41:46 2000  Greg J. Badros  <gjb@cs.washington.edu>
1042
1043         * dynl.c, error.c, eval.c, feature.c, filesys.c, fports.c, list.c, load.c,
1044         net_db.c, read.c, socket.c: Update error messages to use ~A for
1045         %s, ~S for %S to work with new `simple-format' format and be
1046         standardized better.
1047
1048         * print.h, print.c (scm_simple_format): Added `simple-format'
1049         primitive.  It's the old scm_display_error, with ARGS now a rest
1050         parameter, and the destination first instead of last (and a couple
1051         new capabilities inspired by `format' -- #t as destination means
1052         current-output-port, #f means return the formatted text as a
1053         string.
1054
1055         * gh.h, gh_data.c, ports.h, ports.c: Added some missing const specifications.
1056
1057         * backtrace.c (scm_display_error_message): Rewrote to use
1058         scm_simple_format() procedure.
1059
1060         * __scm.h: Added commented-out #define of GUILE_DEBUG_FREELIST
1061
1062 2000-01-09  Marius Vollmer  <mvo@zagadka.ping.de>
1063
1064         Finally applied the libltdl patch from Thomas Tanner, with slight
1065         modifications.
1066
1067         * DYNAMIC-LINKING: Removed because it is obsolete.
1068         * dynl.c: Use ANSI prototypes.
1069         (sysdep_dynl_link): Use lt_dlopenext instead of lt_dlopen.
1070         * scmconfig.h.in: Do not change, as it is automatically generated.
1071
1072     1999-07-25  Thomas Tanner  <tanner@ffii.org>
1073
1074             * dynl-dl.c, dynl-dld.c, dynl-shl.c, dynl-vms.c: deleted
1075             (obsolete)
1076             * Makefile.am: likewise, add INCLTDL (libltdl headers) to
1077             INCLUDES, set dlpreopened files in LDFLAGS, link libguile
1078             against libltdl
1079             * dynl.c: use libltdl if DYNAMIC_LINKING is enabled,
1080             * guile.c: register preloaded modules
1081             * scmconfig.h.in: remove obsolete symbols
1082
1083 2000-01-09  Gary Houston  <ghouston@arglist.com>
1084
1085         * These changes should make it unnecessary to call tzset from
1086         Scheme after modifying the TZ environment variable, even if the
1087         system date facilities cache the value.
1088         * stime.c (setzone, scm_localtime): added comments.
1089         (tzset): don't define a noop tzset macro if HAVE_TZSET not defined.
1090         (setzone): don't call tzset.
1091         (restorezone): call tzset only if HAVE_TZSET is defined.
1092         (scm_tzset): don't define if HAVE_TZSET not defined.  Change the
1093         doc string to indicate that this procedure isn't likely to do
1094         anything useful.
1095         (scm_localtime, scm_strftime, scm_mktime): call tzset if
1096         LOCALTIME_CACHE is defined.
1097
1098 2000-01-09  Mikael Djurfeldt  <mdj@thalamus.nada.kth.se>
1099
1100         * posix.c (scm_sync): Return SCM_UNSPECIFIED.
1101
1102 2000-01-09  Gary Houston  <ghouston@arglist.com>
1103
1104         * eval.c: define scm_unbound_variable_key ('unbound-variable).
1105         scm_lookupcar1: throw an error with key 'unbound-variable instead
1106         of 'misc-error when an unbound variable is encountered.
1107
1108         * filesys.c (scm_mkdir, scm_rmdir, scm_getcwd, scm_select,
1109         scm_symlink, scm_readlink, scm_lstat),
1110         posix.c (scm_setpgid, scm_setsid, scm_ctermid, scm_tcgetpgrp,
1111         scm_tcsetpgrp, scm_uname, scm_setlocale, scm_mknod, scm_nice,
1112         scm_sync),
1113         simpos.c (scm_system),
1114         stime.c (scm_times, scm_strptime):
1115         move the HAVE_XXX feature tests out of the procedure bodies.
1116         don't use SCM_SYSMISSING.
1117         scm_validate.h (SCM_SYSMISSING): removed.
1118         error.h, error.c (scm_sysmissing): comment that this is deprecated.
1119         see ChangeLog entry for 1999-12-28.
1120
1121 Sat Jan  8 19:52:04 2000  Greg J. Badros  <gjb@cs.washington.edu>
1122
1123         * scm_validate.h (SCM_VALIDATE_BOOL_COPY): Fix typo.
1124
1125 Sat Jan  8 17:06:46 2000  Greg J. Badros  <gjb@cs.washington.edu>
1126
1127         * backtrace.c: Fix spelling typo in a comment.
1128
1129         * snarf.h: Use new SCM_DOCS macro to encapsulate the non SCM_INIT
1130         text.  Reformatted some of the expansions.
1131
1132 Fri Jan  7 15:50:46 2000  Greg J. Badros  <gjb@cs.washington.edu>
1133
1134         * scm_validate.h (SCM_OUT_OF_RANGE): Use scm_out_of_range_pos to
1135         report the position of the argument.
1136
1137         * error.h, error.c (scm_out_of_range_pos):  Added this function to
1138         take extra "pos" argument, the position number of the errant
1139         argument.
1140
1141         * debug.c: Use SCM_OUT_OF_RANGE instead of scm_out_of_range.
1142
1143         * snarf.h: Use SCM_HERE and SCM_INIT as names, not SCM_NOTSNARF
1144         and SCM_SNARFING.  Also put the %%% in the SCM_INIT since Mikael
1145         prefers that and I'm reasonably indifferent.
1146
1147 Fri Jan  7 15:03:32 2000  Greg J. Badros  <gjb@cs.washington.edu>
1148
1149         * snarf.h: Factor out differences between C++ and non-C++ into
1150         SCM_FUNC_CAST_ARBITRARY_ARGS macro.  Modify all the snarf macro
1151         definitions to use SCM_NOTSNARF and SCM_SNARFING macros (like
1152         Mikael's macros, below, but changed names and SCM_SNARFING no
1153         longer expands to include %%% -- that must appear in the argument
1154         so that the token appears at the call-site as a reminder).
1155
1156 2000-01-07  Mikael Djurfeldt  <mdj@thalamus.nada.kth.se>
1157
1158         * snarf.h (SCM_INSITU, SCM_INIT): New snarf macros for use in user
1159         snarf macro definitions.
1160
1161 2000-01-06  Mikael Djurfeldt  <mdj@mdj.nada.kth.se>
1162
1163         * chars.c (scm_integer_to_char): Use Greg's nice
1164         SCM_VALIDATE_INUM_RANGE macro for argument checking for closer
1165         adherence to R5RS.
1166
1167 Thu Jan  6 11:48:49 2000  Greg J. Badros  <gjb@cs.washington.edu>
1168
1169         * *.c, snarf.h: Replace GUILE_PROC1 with SCM_DEFINE1 throughout.
1170
1171 Thu Jan  6 11:22:53 2000  Greg J. Badros  <gjb@cs.washington.edu>
1172
1173         * Makefile.am (ETAGS_ARGS): Switch to SCM_DEFINE, SCM_DEFINE1
1174         instead of GUILE_PROC.
1175
1176 Thu Jan  6 11:21:49 2000  Greg J. Badros  <gjb@cs.washington.edu>
1177
1178         * alist.c: Do not report mismatch errors on some uses of `tmp' (do
1179         this by using SCM_ARG2 instead of `2' in the SCM_VALIDATE_CONS
1180         macro call.
1181
1182 Thu Jan  6 09:54:33 2000  Dirk Herrmann <dirk@ida.ing.tu-bs.de> --gjb applied
1183
1184         * scm_validate.h: Remove some redundant NIMP tests.
1185
1186         * alist.c: minimize scope of the tmp variables, and initialize
1187         them when declared. The strange SCM_NIMP tests are replaced by
1188         SCM_CONSP tests that more closely reflect the intended semantics.
1189         However, we don't get a performance penalty here, because the
1190         SCM_CONSP test was performed by the ALISTCELL test anyway.  * The
1191         extremely ugly use of ASRTGO macros was removed: The calls to
1192         ASRTGO were not encapsulated by "#ifndef SCM_RECKLESS", but got a
1193         label parameter that only exists when SCM_RECKLESS is not defined.
1194         This works, because ASRTGO itself is defined in a way that it only
1195         makes use of the label parameter if SCM_RECKLESS is not defined
1196         (shudder!).  Does guile make at all use of the possibility to
1197         define SCM_RECKLESS?  * Codesize is likely to be reduced, since
1198         instead of two calls to SCM_ASSERT performed by the ALISTCELL test
1199         we now only get one test.
1200
1201         * list.c: Use SCM_NNULLP, not SCM_NIMP as appropriate.  Also use
1202         SCM_NULLP instead of SCM_IMP.  Drop use of "register" keyword on
1203         some variables in `list?'.  Fix `reverse' and `reverse!'
1204         primitives to handle improper lists better.
1205
1206 Wed Jan  5 11:24:53 2000  Greg J. Badros  <gjb@cs.washington.edu>
1207
1208         * *.[ch]: Whitespace changes -- added space after SCM_VALIDATE_*
1209         macros and SCM_DEFINE macros to match GNU coding standards.
1210
1211 Wed Jan  5 11:04:24 2000  Greg J. Badros  <gjb@cs.washington.edu>
1212
1213         * *.[ch]: Replace GUILE_PROC w/ SCM_DEFINE.
1214
1215 Wed Jan  5 10:59:06 2000  Greg J. Badros  <gjb@cs.washington.edu>
1216
1217         * *.[ch]:  Replace SCM_VALIDATE_INT w/ SCM_VALIDATE_INUM for
1218         better consistency with the names of other SCM_VALIDATE_ macros
1219         and better conformance to guile naming policy.
1220
1221 Wed Jan  5 10:50:39 2000  Greg J. Badros  <gjb@cs.washington.edu>
1222
1223         * ports.c (s_scm_close_all_ports_except): Use SCM_ARG1 in a
1224         SCM_VALIDATE instead of 1 to avoid a check on the argument (since
1225         it's not the actual name of the formal).
1226
1227         * guile-snarf.awk.in: Do argument/number mismatch checking and
1228         print warnings in an Emacs compile-mode parseable format.
1229
1230         * struct.c: Use SCM_ASSERT_RANGE instead of SCM_ASSERT w/
1231         SCM_OUTOFRANGE as 3rd argument.
1232
1233         * random.c: Fix argument/number mismatch (that I introduced :-( ).
1234
1235         * __scm.h: Do not #define SCM_ARG* when snarfing;
1236         lets us distinguish between 1 and SCM_ARG1 when snarfing as only
1237         the former (using the number) requires the argument to match the
1238         formal in the current argument snarfing check.
1239
1240         * snarf.h: Give new definition of SCM_ASSERT when in
1241         snarfing mode to output a lexically-identifiable sequence that the
1242         guile-snarf.awk script uses to verify argument/position matching.
1243
1244         * ramap.c: Remove extraneous #undef FUNC_NAME.
1245
1246 Wed Jan  5 08:36:38 2000  Greg J. Badros  <gjb@cs.washington.edu>
1247
1248         * guile-doc-snarf.awk.in: Removed -- guile-snarf.awk.in is the
1249         current version of the same functionality; it writes the .x output
1250         to stdout instead of directly into the file.
1251
1252 Wed Jan  5 08:15:04 2000  Greg J. Badros  <gjb@cs.washington.edu>
1253
1254         * unif.c, symbols.c, strings.c, stacks.c, random.c, print.c,
1255         posix.c:  Eliminated a bunch of SCM_NIMP(..)s that are now
1256         redundant with the safer macros.  Patch from Dirk Hermann applied
1257         by hand. Thanks Dirk!
1258
1259         * scm_validate.h: Added SCM_VALIDATE_VECTOR_OR_DVECTOR for some
1260         uses in random.c.
1261
1262         * ramap.c: whitespace change.
1263
1264 Tue Jan  4 14:21:35 2000  Greg J. Badros  <gjb@cs.washington.edu>
1265
1266         * options.c, objects.c, keywords.c, gc.c: Some redundant SCM_NIMP
1267         removals from Dirk Hermann.
1268
1269         * alist.c: Rename formals to match the parameter names in the
1270         documentation, updates to documentation.  Thanks Dirk Hermann!
1271
1272 2000-01-04  Mikael Djurfeldt  <mdj@mdj.nada.kth.se>
1273
1274         * eval.c (SCM_CEVAL): Reverse order of
1275         scm_stack_checking_enabled_p and SCM_STACK_OVERFLOW_P
1276         (Thanks to Brad Knotwell.)
1277
1278 Mon Jan  3 08:30:02 2000  Greg Harvey  <Greg.Harvey@thezone.net> (applied --01/03/00 gjb)
1279
1280         * gc.c (scm_debug_newcell): Added SCM_SETCAR of the newly
1281         allocated cell.
1282
1283         * pairs.h: Added a comment about the need for the SCM_SETCAR in
1284         SCM_NEWCELL macro.
1285
1286 Mon Jan  3 08:25:19 2000  Greg J. Badros  <gjb@cs.washington.edu>
1287
1288         * dynl-vms.c, debug.c, coop-threads.c, backtrace.c, eval.c: More
1289         SCM_NIMP tests that were redundant are now eliminated.  Patches
1290         from Dirk Hermann applied by hand.
1291
1292 The ChangeLog continues in the file: "ChangeLog-1996-1999"