* src/thread.c (unmark_main_thread): New function.
* src/lisp.h (unmark_main_thread): Prototype it.
* src/alloc.c (garbage_collect_1): Call it after sweeping.
(Bug#33073)
* test/src/thread-tests.el (threads-test-bug33073): New test.
d28d54c (origin/emacs-26) More accurate docs for 'text-char-description'
b3baf99 Document synchronous behavior of eshell/make (Bug#32513)
98544ea Fix bs-show with wide characters (Bug#17822)
85af51b Improve Custom menu labels for 2 options
72a2a36 Improve wording of last change in dired-x.texi
d4fa83b Fix GnuTLS test suite with GnuTLS versions 3.4.x
b5bee6b Fix build with gnutls versions 3.0 to 3.2 (Bug#32446)
67eb80e ; * etc/enriched.txt (hanging-indents): Remove extra indent.
c71cfb7 Fix the Bubbles game on TTY frames
3bbf21b Add choice to reshow certificate information (Bug#31877)
6f2c471 * src/alloc.c (Fbool_vector, Flist, Fvector): Doc tweak.
39eecb3 * src/alloc.c (vector): Fix grammatical error in doc string: ...
This better corresponds to what emacs-26 did in the
rare cases where it checked for integer overflow.
* src/alloc.c (range_error): Remove.
All uses changed to overflow_error.
* src/eval.c (overflow_error): New function.
sizeof (struct Lisp_Marker) was 32 on x86, where 24 would do.
Problem noted by Stefan Monnier in:
https://lists.gnu.org/r/emacs-devel/2018-09/msg00165.html
* src/bignum.h (struct Lisp_Bignum):
* src/frame.h (struct frame):
* src/lisp.h (struct Lisp_Vector, struct Lisp_Bool_Vector)
(struct Lisp_Char_Table, struct Lisp_Hash_Table)
(struct Lisp_Marker, struct Lisp_Overlay)
(struct Lisp_Misc_Ptr, struct Lisp_User_Ptr)
(struct Lisp_Finalizer, struct Lisp_Float)
(struct Lisp_Module_Function):
* src/process.h (struct Lisp_Process):
* src/termhooks.h (struct terminal):
* src/thread.h (struct thread_state, struct Lisp_Mutex)
(struct Lisp_CondVar):
* src/window.c (struct save_window_data):
* src/window.h (struct window):
* src/xterm.h (struct scroll_bar):
* src/xwidget.h (struct xwidget, struct xwidget_view):
Add GCALIGNED_STRUCT attribute.
* src/lisp.h (GCALIGNED_UNION_MEMBER): Renamed from GCALIGNED_UNION.
All uses changed.
(GCALIGNED_STRUCT_MEMBER, GCALIGNED_STRUCT, GCALIGNED): New macros.
All uses of open-coded GCALIGNED changed to use GCALIGNED.
(union vectorlike_header): No longer GC-aligned.
(PSEUDOVECSIZE): Yield 0 for pseudovectors without Lisp
objects that place a member before where the first Lisp object
member would be.
Problem reported by Andy Moreton in:
https://lists.gnu.org/r/emacs-devel/2018-09/msg00072.html
and crystal-ball diagnosis by Eli Zaretskii in:
https://lists.gnu.org/r/emacs-devel/2018-09/msg00075.html
* src/alloc.c (xrealloc_for_gmp, xfree_for_gmp): Move to bignum.c.
(init_alloc): Move bignum initialization to init_bignum.
* src/bignum.c (init_bignum): Rename from init_bignum_once.
All users changed.
* src/emacs.c (main): Call init_bignum after init_alloc,
instead of calling init_bignum_once after init_bignum.
* src/bignum.c, src/bignum.h: New files. Only modules that
need to know how bignums are implemented should include
bignum.h. Currently these are alloc.c, bignum.c (of course),
data.c, emacs.c, emacs-module.c, floatfns.c, fns.c, print.c.
* src/Makefile.in (base_obj): Add bignum.o.
* src/alloc.c (make_bignum_str): Move to bignum.c.
(make_number): Remove; replaced by bignum.c’s make_integer.
All callers changed.
* src/conf_post.h (ARG_NONNULL): New macro.
* src/json.c (json_to_lisp): Use it.
* src/data.c (Fnatnump):
Move NATNUMP’s implementation here from lisp.h.
* src/data.c (Fnumber_to_string):
* src/editfns.c (styled_format):
Move conversion of string to bignum to bignum_to_string, and
call it here.
* src/emacs-module.c (module_make_integer):
* src/floatfns.c (Fabs):
Simplify by using make_int.
* src/emacs.c: Include bignum.h, to expand its inline fns.
* src/floatfns.c (Ffloat): Simplify by using XFLOATINT.
(rounding_driver): Simplify by using double_to_bignum.
(rounddiv_q): Clarify use of temporaries.
* src/lisp.h: Move decls that need to know bignum internals to
bignum.h. Do not include gmp.h or mini-gmp.h; that is now
bignum.h’s job.
(GMP_NUM_BITS, struct Lisp_Bignum, XBIGNUM, mpz_set_intmax):
Move to bignum.h.
(make_int): New function.
(NATNUMP): Remove; all callers changed to use Fnatnump.
(XFLOATINT): If arg is a bignum, use bignum_to_double, so that
bignum internals are not exposed here.
* src/print.c (print_vectorlike): Use SAFE_ALLOCA to avoid the
need for a record_unwind_protect_ptr.
Also, since Emacs historically reported a range error when
rounding operations overflowed, do that consistently for all
bignum overflows.
* doc/lispref/errors.texi (Standard Errors):
* doc/lispref/numbers.texi (Integer Basics): Document range errors.
* src/alloc.c (range_error): Rename from integer_overflow.
All uses changed.
* src/floatfns.c (rounding_driver): When the result of a floating
point rounding operation does not fit into a fixnum, put it
into a bignum instead of always signaling an range error.
* test/src/floatfns-tests.el (divide-extreme-sign):
These tests now return the mathematically-correct answer
instead of signaling an error.
(bignum-round): Check that integers round to themselves.
libgmp calls ‘abort’ when given numbers too big for its
internal data structures. The numeric limit is large and
platform-dependent; with 64-bit GMP 6.1.2 it is around
2**2**37. Work around the problem by refusing to call libgmp
functions with arguments that would cause an abort. With luck
libgmp will have a better way to do this in the future.
Also, introduce a variable integer-width that lets the user
control how large bignums can be. This currently defaults
to 2**16, i.e., it allows bignums up to 2**2**16. This
should be enough for ordinary computation, and should
help Emacs to avoid thrashing or hanging.
Problem noted by Pip Cet (Bug#32463#71).
* doc/lispref/numbers.texi, etc/NEWS:
Document recent bignum changes, including this one.
Improve documentation for bitwise operations, in the light
of bignums.
* src/alloc.c (make_number): Enforce integer-width.
(integer_overflow): New function.
(xrealloc_for_gmp, xfree_for_gmp):
Move here from emacs.c, as it's memory allocation.
(init_alloc): Initialize GMP here, rather than in emacs.c.
(integer_width): New var.
* src/data.c (GMP_NLIMBS_MAX, NLIMBS_LIMIT): New constants.
(emacs_mpz_size, emacs_mpz_mul)
(emacs_mpz_mul_2exp, emacs_mpz_pow_ui): New functions.
(arith_driver, Fash, expt_integer): Use them.
(expt_integer): New function, containing integer code
that was out of place in floatfns.c.
(check_bignum_size, xmalloc_for_gmp): Remove.
* src/emacs.c (main): Do not initialize GMP here.
* src/floatfns.c (Fexpt): Use expt_integer, which
now contains integer code moved from here.
* src/lisp.h (GMP_NUMB_BITS): Define if gmp.h doesn’t.
* src/alloc.c (mpz_set_intmax_slow): Tighten assertion.
Work even in the unlikely case where libgmp uses nails.
* src/data.c (FIXNUMS_FIT_IN_LONG): New constant.
(arith_driver): Use it to tighten compile-time checks.
* src/lisp.h (mpz_set_intmax): Do not assume that converting
an out-of-range value to ‘long’ is harmless, as it might raise
a signal. Use simpler expression; compiler can optimize.
These tuneups and minor simplifications should affect only
platforms with EMACS_INT wider than ‘long’.
* src/alloc.c (make_number): If the number fits in long but
not in fixnum, do not attempt to convert to fixnum again.
Tighten the compile-time check for whether the second attempt
is worth trying, from sizeof (long) < sizeof (EMACS_INT) to
LONG_WIDTH < FIXNUM_BITS. Do not bother computing the sign of
the value to tighten the bounds for whether to try the second
attempt, as it’s not worth the effort. Do not call mpz_size,
which is unnecessary since the number of bits is already known
and the loop can iterate over a shift count instead. Avoid
unnecessary casts. Use + instead of | where either will do,
as + is typically better for optimization.
Improve mpz_t to fixnum when --with-wide-int
* src/alloc.c (make_number): Avoid undefined behavior
when shifting an EMACS_UINT by more than EMACS_UINT_WIDTH bits.
Check for integer overflow when shifting.
These operations incorrectly treated negative fixnums as
bignums greater than most-positive-fixnum.
* src/alloc.c (mpz_set_intmax_slow): Avoid undefined
behavior if signed unary negation overflows, while
we’re in the neighborhood.
(mpz_set_uintmax_slow): Remove. All uses removed.
* src/data.c (arith_driver): Treat fixnums as signed, not
unsigned, even for logical operations.
* src/lisp.h (mpz_set_uintmax): Remove. All uses removed.
* test/src/data-tests.el (data-tests-logand)
(data-tests-logior, data-tests-logxor): New tests.
Remove misc-objects-consed and the misc component of
memory-use-count, since misc objects no longer exist.
* doc/lispref/internals.texi, etc/NEWS: Mention this,
and adjust better to recent removal of misc objects.
* src/alloc.c (MEM_TYPE_MISC): Remove; no longer used.
(Fmemory_use_counts): Omit misc count, since miscs
no longer exist.
(misc-objects-consed): Remove.
614cc65 ; * lisp/simple.el (line-move-visual): Fix typo.
d2ad4ba Do not consider external packages to be removable (Bug#27822)
ec0995c * src/alloc.c: Remove obsolete comments.
ec6f588 Better support utf-8-with-signature and utf-8-hfs in HTML
eb026a8 Don't use -Wabi compiler option
Suggested by Pip Cet (Bug#32405#14).
* src/alloc.c (mark_object): Remove unnecessary special cases for
PVEC_MARKER, PVEC_BOOL_VECTOR, PVEC_MISC_PTR, PVEC_USER_PTR, and
PVEC_FINALIZER.
change is to free up an enum Lisp_Type tag value, a scarce
Eliminate the category of miscellaneous objects, and turn all
such objects into pseudovectors. The immediate motivation
for this change is to free up an enum Lisp_Type tag value, a
scarce resource that can be better used elsewhere. However,
this change is worthwhile in its own right, as it improves
performance slightly on my platform, 0.3% faster for 'make
compile-always' on Fedora 28, and it simplifies the garbage
collector and interpreter (Bug#32405).
* doc/lispref/internals.texi (Garbage Collection):
* etc/NEWS:
Document change to garbage-collect return value.
* src/alloc.c (total_markers, total_free_markers):
(union aligned_Lisp_Misc, MARKER_BLOCK_SIZE)
(struct marker_block, marker_block, marker_block_index)
(misc_free_list, allocate_misc, live_misc_holding)
(live_misc_p, sweep_misc):
* src/lisp.h (lisp_h_MARKERP, lisp_h_MISCP, MARKERP, MISCP)
(Lisp_Misc, enum Lisp_Misc_Type, Lisp_Misc_Free)
(Lisp_Misc_Marker, Lisp_Misc_Overlay, Lisp_Misc_Finalizer)
(Lisp_Misc_Ptr, Lisp_Misc_User_Ptr, Lisp_Misc_Limit)
(Lisp_Misc_Bignum)
(XSETMISC, struct Lisp_Misc_Any, XMISCANY, XMISCTYPE)
(struct Lisp_Free, union Lisp_Misc, XMISC):
Remove. All uses removed.
(cleanup_vector): Clean up objects that were formerly misc
and are now pseudovectors.
(make_misc_ptr, build_overlay, Fmake_marker, build_marker)
(make_bignum_str, make_number, make_pure_bignum)
(make_user_ptr, Fmake_finalizer):
Build as pseudovectors, not as misc objects.
(mark_finalizer_list, queue_doomed_finalizers)
(compact_undo_list, mark_overlay, mark_object)
(unchain_dead_markers):
Mark as vector-like objects, not as misc objects.
(mark_maybe_object, mark_maybe_pointer, valid_lisp_object_p)
(total_bytes_of_live_objects, survives_gc_p):
* src/fns.c (sxhash):
No need to worry about misc objects.
(garbage_collect_1): Do not generate a 'misc' component.
(syms_of_alloc): No need for 'misc' symbol.
* src/buffer.c (overlays_at, overlays_in, overlay_touches_p)
(overlay_strings, recenter_overlay_lists)
(fix_start_end_in_overlays, fix_overlays_before)
(Foverlay_lists, report_overlay_modification)
(evaporate_overlays):
* src/editfns.c (overlays_around):
* src/data.c (Ftype_of):
* src/fns.c (internal_equal):
* src/lisp.h (mint_ptrp, xmint_pointer, FINALIZERP)
(XFINALIZER, MARKERP, XMARKER, OVERLAYP, XOVERLAY, USER_PTRP)
(XUSER_PTR, BIGNUMP, XBIGNUM):
* src/print.c (print_vectorlike, print_object):
* src/undo.c (record_marker_adjustments):
* src/xdisp.c (load_overlay_strings):
Formerly misc objects are now pseudovectors.
* src/lisp.h (PVEC_MARKER, PVEC_OVERLAY, PVEC_FINALIZER)
(PVEC_BIGNUM, PVEC_MISC_PTR, PVEC_USER_PTR):
New constants, replacing their misc versions. All uses changed.
(struct Lisp_Marker, struct Lisp_Overlay, struct Lisp_Misc_Ptr)
(struct Lisp_Bignum, struct Lisp_User_Ptr, struct Lisp_Finalizer):
Make usable as a pseudovector by using a pseudovector header,
replacing any DIY components, and putting Lisp_Object members
first. All uses changed.
* src/alloc.c (VECTOR_BLOCK_SIZE, VECTOR_BLOCK_BYTES)
(VBLOCK_BYTES_MIN, VBLOCK_BYTES_MAX, VECTOR_MAX_FREE_LIST_INDEX):
Prefer enums to macros where either will do.
(allocate_vector_from_block): Arg is ptrdiff_t, not size_t.
Use eassume instead of eassert.
(PSEUDOVEC_STRUCT): New macro, which verifies the already-existing
assumption that the vector-like objects are small.
(cleanup_vector): Use it. Use if-then-else systematically;
this lets GCC do a bit better job.
2018-08-08 Paul Eggert <eggert@cs.ucla.edu>
* src/alloc.c (VBLOCK_BYTES_MAX): Use vroundup_ct, not
vroundup, so that can be used in static assertions.
* lisp/international/ccl.el (ccl-fixnum): New function.
(ccl-embed-data, ccl-embed-current-address, ccl-dump): Use it.
* src/alloc.c (make_number): Handle case where EMACS_INT is
larger than long.
* src/data.c (bignumcompare): Handle case where EMACS_INT is
larger than long.
(arith_driver): Likewise. Coerce markers.
(float_arith_driver): Coerce markers.
(Flogcount): Use mpz_sgn.
(ash_lsh_impl): Fix bugs.
(Fsub1): Fix underflow check.
* src/lisp.h (NUMBERP): Don't check BIGNUMP.
(CHECK_FIXNUM_OR_FLOAT_COERCE_MARKER): Fix indentation.
* test/lisp/international/ccl-tests.el: New file.
Have it return Emacs virtual memory size, not the sbrk value
which is often useless newadays.
* doc/lispref/internals.texi (Garbage Collection):
* etc/NEWS: Document this.
* lisp/subr.el (memory-limit): New implementation in Lisp,
written in terms of process-attributes, and which returns
virtual memory size.
* src/alloc.c (Fmemory_limit): Remove C implementation.
Mostly, this avoids munging executables when CANNOT_DUMP = yes,
as the munging is needed only for unexec.
* configure.ac (PAXCTL_dumped, PAXCTL_notdumped) [CANNOT_DUMP]:
Leave these empty.
(LD_SWITCH_SYSTEM_TEMACS) [CANNOT_DUMP]:
Do not append -no-pie or -nopie.
* src/alloc.c (my_heap_start) [CANNOT_DUMP]: Omit; not used.
This is a streamlined version of Lisp_Save_Value, which contains just
a pointer, as that is all Lisp_Save_Values are used for any more.
With the previous changes, these objects are not primarily used as
save values, so just call them "Misc" rather than "Save".
* src/alloc.c (make_misc_ptr): New function.
(mark_object): Mark Lisp_Misc_Ptr too.
* src/lisp.h (Lisp_Misc_Ptr): New constant.
(struct Lisp_Misc_Ptr): New type.
(make_mint_ptr, mint_ptrp, xmint_pointer):
Use Lisp_Misc_Ptr, not Lisp_Save_Value.
(union Lisp_Misc): Add Lisp_Misc_Ptr.
* src/print.c (print_object): Print Lisp_Misc_Ptr.
* src/alloc.c (mark_maybe_objects): New function.
* src/eval.c (default_toplevel_binding)
(backtrace_eval_unrewind, Fbacktrace__locals):
Treat array unwindings like other miscellaneous pdl types.
(record_unwind_protect_array): New function.
(do_one_unbind): Free the array while unwinding.
(mark_specpdl): Mark arrays directly.
* src/lisp.h (SPECPDL_UNWIND_ARRAY): New constant.
(union specbinding): New member unwind_array.
(SAFE_ALLOCA_LISP_EXTRA): Use record_unwind_protect_array
instead of make_save_memory + record_unwind_protect.
Do not assume that 8-byte alignment suffices for all C objects,
as some platforms require 16-byte alignment for some objects,
and this will start to bite us as time goes on (e.g., if an
Emacs module ever uses an object containing a long
double, which requires 16-byte alignment on x86-64).
Conversely, on !USE_LSB_TAG platforms, do not insist on
aligning Lisp objects to a multiple of 8, as this is not
needed for high-order tag bits.
* src/alloc.c (LISP_ALIGNMENT, MALLOC_IS_LISP_ALIGNED):
New constants.
(XMALLOC_BASE_ALIGNMENT, XMALLOC_HEADER_ALIGNMENT):
Removed. All uses replaced by LISP_ALIGNMENT.
(aligned_alloc, laligned, lmalloc, lrealloc, union aligned_Lisp_Misc)
(maybe_lisp_pointer, pure_alloc):
Use LISP_ALIGNMENT rather than GCALIGNMENT.
(aligned_alloc): Do not worry about an alignment of
LISP_ALIGNMENT when MALLOC_IS_LISP_ALIGNED, as the code never
uses aligned_alloc with alignment == LISP_ALIGNMENT in that case.
(__alignof__): Remove. All uses removed.
(MALLOC_IS_GC_ALIGNED): Remove.
All uses replaced with MALLOC_IS_LISP_ALIGNED.
(vector_alignment): Remove.
All uses replaced with LISP_ALIGNMENT.
* src/alloc.c (mark_maybe_pointer):
* src/emacs-module.c (value_to_lisp_bits):
Do not assume GCALIGNMENT == 1 << GCTYPEBITS, as GCALIGNMENT
is 1 on !USE_LSB_TAG platforms now.
* src/lisp.h (GCALIGNMENT) [!USE_LSB_TAG]: Now 1.
(struct Lisp_Symbol, union vectorlike_header, struct Lisp_Cons)
(struct Lisp_String): Simplify test for verifying alignment.
Instead of using __builtin_assume_aligned (P, GCALIGNMENT) to
tell GCC that P has alignment 8, use (T *) P where T is the
type of the pointed-to object, to tell GCC that P has native
alignment. This is simpler, matches the intent better, and
should help simplify future improvements. Some of these
changes are to pacify gcc -Wnull-dereference, since GCC is
smarter about pointers now that Emacs no longer uses
__builtin_assume_aligned; these minor changes should improve
code efficiency slightly. On Fedora 28 x86-64 with default
optimization this patch shrinks the size of the Emacs text
segment by 0.36%.
* src/conf_post.h (__has_builtin, __builtin_assume_aligned):
Remove; no longer used.
* src/dbusbind.c (XD_OBJECT_TO_DBUS_TYPE):
Pacify -Wnull-dereference by using XCAR instead of CAR_SAFE
and XCDR instead of CDR_SAFE when this is safe.
* src/fileio.c (Fexpand_file_name):
* src/font.c (clear_font_cache):
Pacify -Wnull-dereference by removing unnecessary NILP test.
* src/keyboard.c (xevent_start): New function.
(read_char, read_key_sequence): Pacify -Wnull-dereference by
using xevent_start instead of EVENT_START.
* src/lisp.h (lisp_h_XUNTAG): Remove; XUNTAG is always a macro
now, since it can no longer be implemented as a function.
(XUNTAG): New third argument CTYPE. All uses changed.
Cast result to CTYPE * instead of using __builtin_assume_aligned.
Simplify by using LISP_WORD_TAG.
(LISP_WORD_TAG): New macro.
(TAG_PTR): Use it.
* src/menu.c (x_popup_menu_1):
Pacify -Wnull-dereference by using XCAR instead of Fcar and
XCDR instead of Fcdr where this is safe.
This workaround no longer appears to be needed.
* src/alloc.c (USE_ALIGNED_ALLOC): Don’t leave undefined
merely because ADDRESS_SANITIZER is defined, as that bug
in -fsanitize=address appears to have been fixed. See:
https://github.com/google/sanitizers/issues/337
* src/conf_post.h (vfork): Improve comment.
* test/src/data-tests.el (data-tests-kill-all-local-variables): New test.
* src/buffer.c (swap_out_buffer_local_variables): Remove.
Fuse the body of its loop into that of reset_buffer_local_variables.
(Fkill_buffer, Fkill_all_local_variables): Don't call it any more.
(reset_buffer_local_variables): Make sure the buffer's local binding
is swapped out before removing it from the alist (bug#30846).
Call watchers before actually killing the var.
* src/data.c (Fmake_local_variable): Simplify.
Use swap_in_global_binding to swap out any local binding, instead of
a mix of find_symbol_value followed by messing with where&found.
Don't call swap_in_symval_forwarding since the currently swapped
binding is never one we've modified.
(Fkill_local_variable): Use swap_in_global_binding rather than messing
with where&found to try and trick find_symbol_value into doing the same.
* src/alloc.c (mark_localized_symbol): 'where' can't be a frame any more.
(cherry picked from commit 3ddff08034)
* src/alloc.c (free_marker): Remove.
* src/editfns.c (save_restriction_restore):
* src/insdel.c (signal_before_change): Detach the markers from the
buffer when we're done with them instead of calling free_marker on
them.
* test/src/editfns-tests.el (delete-region-undo-markers-1)
(delete-region-undo-markers-2): New tests.
(cherry picked from commit 96b8747d5c)
In GCC 8, gcc -fsanitize=undefined flags the undefined behavior
that Emacs relies on in its XPNTR and XSYMBOL low-level functions.
Disable undefined sanitization in these functions. Although this
disabling doesn’t suffice if DEFINE_KEY_OPS_AS_MACROS is true, it
works for -fsanitize=undefined -DINLINING=0, which is good enough.
* src/alloc.c (macro_PNTR_ADD): New macro.
(PNTR_ADD): New function and macro.
The function disables -fsanitize=undefined.
(macro_XPNTR): Use it.
* src/conf_post.h (ATTRIBUTE_NO_SANITIZE_UNDEFINED): New macro.
* src/lisp.h (XSYMBOL): Disable -fsanitize=undefined.
* src/alloc.c (free_marker): Remove.
* src/editfns.c (save_restriction_restore):
* src/insdel.c (signal_before_change): Detach the markers from the
buffer when we're done with them instead of calling free_marker on
them.
* test/src/editfns-tests.el (delete-region-undo-markers-1)
(delete-region-undo-markers-2): New tests.
* test/src/data-tests.el (data-tests-kill-all-local-variables): New test.
* src/buffer.c (swap_out_buffer_local_variables): Remove.
Fuse the body of its loop into that of reset_buffer_local_variables.
(Fkill_buffer, Fkill_all_local_variables): Don't call it any more.
(reset_buffer_local_variables): Make sure the buffer's local binding
is swapped out before removing it from the alist (bug#30846).
Call watchers before actually killing the var.
* src/data.c (Fmake_local_variable): Simplify.
Use swap_in_global_binding to swap out any local binding, instead of
a mix of find_symbol_value followed by messing with where&found.
Don't call swap_in_symval_forwarding since the currently swapped
binding is never one we've modified.
(Fkill_local_variable): Use swap_in_global_binding rather than messing
with where&found to try and trick find_symbol_value into doing the same.
* src/alloc.c (mark_localized_symbol): 'where' can't be a frame any more.
Unchain all dead markers with a single scan of the markers list,
instead of calling the O(N) 'unchain_marker' N times.
(unchain_dead_markers): New function.
(sweep_buffers): Use it.
(gc_sweep): Sweep buffers before markers.
(sweep_misc): Check that markers have been unchained when reclaiming them.
