casts its result, so doesn't yield an lvalue per ANSI C.
* goops.c (s_scm_sys_set_object_setter_x): use
SCM_SET_ENTITY_SETTER.
(clear_method_cache): use SCM_SET_ENTITY_PROCEDURE.
* gc.h (SCM_GC_SET_CARD_BVEC): new macro. SCM_GC_CARD_BVEC casts
its result, so doesn't yield an lvalue per ANSI C.
(SCM_GC_SET_CARD_FLAGS): ditto for SCM_GC_GET_CARD_FLAGS.
(SCM_GC_CLR_CARD_FLAGS): redefined in terms of
SCM_GC_SET_CARD_FLAGS.
(SCM_GC_SET_CARD_FLAG, SCM_GC_CLR_CARD_FLAGS): ditto.
* gc.c (INIT_CARD): use the explicit setter macro to set the bvec.
`scm_gc_mark', but doesn't mark the argument itself. defined
using an arrangement similar to that in eval.c: `scm_gc_mark' and
`scm_gc_mark_dependencies' are derived from the same "template"
by ugly preprocessor magic.
* gc.h: added prototype for `scm_gc_mark_dependencies'.
* init.c (scm_init_guile_1): call the renamed
`scm_init_guardians'.
* guardians.h: changed prototypes for `scm_make_guardian' and
`scm_init_guardians'.
* guardians.c (guardian_t): added new fields `greedy_p' and
`listed_p'.
(GUARDIAN_P): predicate that says whether its argument is a
guardian.
(GUARDIAN_GREEDY_P, GUARDIAN_LISTED_P): new predicates.
(greedy_guardians, sharing_guardians): new variables. hold the
greedy and sharing live guardian lists, respectively.
(first_live_guardian, current_link_field): removed.
(greedily_guarded_prop): new variable. holds the "is greedily
guarded" object property.
(self_centered_zombies): new variable. stores guarded objects
that are parts of cycles.
(add_to_live_list): new function, introduced to decouple marking a
guardian and adding it to the live list.
(guardian_mark): call `add_to_live_list'.
(guardian_print): print whether the guardian is greedy or not.
also change "live" and "zombie" to "reachable" and "unreachable"
respectively, to be less confusing.
(scm_guard): if the guardian is greedy, test whether the object is
already greedily marked. throw an error if so.
(scm_get_one_zombie): if the guardian is greedy, remove the
"greedily guarded" property from the object.
(scm_make_guardian): add a new optional boolean argument which
says whether the guardian is greedy or sharing.
(guardian_gc_init): init the new live lists.
(mark_dependencies): new function.
(mark_and_zombify): new function.
(guardian_zombify): reworked to support the new guardian
semantics. move some logic to `mark_dependencies' and
`mark_and_zombify'.
(whine_about_self_centered_zombies): new function. installed in
the `after-gc-hook' to complain about guarded objects which are
parts of cycles.
(scm_init_guardians): init the new stuff. renamed from
`scm_init_guardian'.
custom tc7 type. This will make it easier to support R5RS
multiple value continuations, without the use of a Scheme-level
wrapper.
* continuations.c (scm_tc16_continuation, continuation_mark,
continuation_free, continuation_print, continuation_apply):
new SMOB support.
(scm_make_continuation): new procedure, replaces scm_make_cont
with a different interface.
(copy_stack_and_call, scm_dynthrow, scm_init_continuations): rewritten.
(CHEAP_CONTINUATIONS): removed non-working code completely.
(scm_call_continuation): removed.
* continuations.h (struct scm_contregs): add num_stack_items and
stack fields. previously stack was stored following this struct:
use a tail array instead.
(SCM_CONTINUATIONP): new macro.
(SCM_CONTINUATION_LENGTH, SCM_SET_CONTINUATION_LENGTH):
rewritten.
(SCM_SET_CONTREGS): removed.
* tags.h: removed scm_tc7_contin (was tag 61).
* debug.c, gc.c, hash.c, print.c, procprop.c, procs.c:
removed scm_tc7_contin support.
* eval.c: use scm_make_continuation instead of scm_make_cont.
don't set jump buffers here. remove scm_tc7_contin support.
* init.c, root.c: create SMOB continuation for rootcont instead
of scm_tc7_contin. call scm_init_continuations before
scm_init_root.
* root.c: remove support for static jmpbuf. It's not used by
default and I broke it. create SMOB continuation for rootcont.
* stacks.c: use SCM_CONTINUATIONP.
* Fix some string/symbol output problems with regards to substrings.
* Fix error output to prefer procedure name parameters over stack data.
* Use SCM_(SET_)?FILENAME where appropriate.
* Prefer calling scm_remember over scm_protect/unprotect_object calls.
about any possible false positives.
(scm_debug_newcell): don't change cell type to `allocated'.
(scm_debug_newcell2): ditto.
(scm_gc_for_newcell): ditto.
(scm_gc_mark): remove the tc16_allocated case.
account, but keeping more or less the same values as previously.
added some simple helper macros.
(CLUSTER_SIZE_IN_BYTES, ALIGNMENT_SLACK): defined to take cards
into account.
(BVEC_*, scm_mark_space_t, current_mark_space, mark_space_ptr,
current_mark_space_offset, mark_space_head, get_bvec,
clear_mark_space): new functions and supporting variables, types
and macros that implement mark space management.
(scm_igc): clear the mark space (all of it) before beginning the
mark phase.
(scm_gc_mark): changed the tests for rogue cells, much simplified
throughout (no different mark bit locations to worry about now).
(scm_mark_locations): don't consider card header cells.
(scm_cellp): ditto.
(scm_gc_sweep): simplified.
(init_heap_seg): changed to take cards into account.
sweep time, total marked cells, total swept cells, and number of
times GC was invoked.
(gc_start_stats): renamed from scm_gc_start, made static, taught
to init the new stats.
(gc_end_stats): renamed from scm_gc_end, made static, taught to
calculate the new stats.
(scm_igc): don't call gc_start_stats unless we are sure that we
are indeed going to collect. also, added some timekeeping between
the mark and sweep phases.
(scm_gc_sweep): count number of cells we sweep as we go.
* gc.h: removed prototypes for scm_gc_{start,end}.
doesn't yield any new cells. In theory this could happen if all
cells allocated with NEWCELL are either in use or conservatively
marked and all cluster spine cells are conservatively marked.
(Thanks to Dirk.)
scm_default_min_yield_1, scm_default_init_heap_size_2,
scm_default_min_yield_2, scm_default_max_segment_size): New global
variables. Can be customized by the application before booting
Guile. (We might want to be able to control these parameters
dynamically through the "options interface" in the future, but
note that that is additional functionality. Here we're giving
default values which the environment variables can override.)
