The porting part of this patch fixes bugs on non-IEEE platforms
with frexp, ldexp, logb.
* admin/CPP-DEFINES (HAVE_CBRT, HAVE_LOGB, logb): Remove.
* configure.ac (logb, cbrt): Do not check for these functions,
as they are not being used.
* doc/lispref/numbers.texi (Float Basics, Arithmetic Operations, Math Functions):
Document that / and mod (with floating point arguments), along
with asin, acos, log, log10, expt and sqrt, return special values
instead of signaling exceptions.
(Float Basics): Document that logb operates on the absolute value
of its argument.
(Math Functions): Document that (log ARG BASE) also returns NaN if
BASE is negative. Document that (expt X Y) returns NaN if X is a
finite negative number and Y a finite non-integer.
* etc/NEWS: Document NaNs versus signaling-error change.
* src/data.c, src/lisp.h (Qdomain_error, Qsingularity_error, Qunderflow_error):
Now static.
* src/floatfns.c: Simplify discussion of functions that Emacs doesn't
support, by removing commented-out code and briefly listing the
C89 functions excluded. The commented-out stuff was confusing
maintenance, e.g., we thought we needed cbrt but it was commented out.
(logb): Remove decl; no longer needed.
(isfinite): New macro, if not already supplied.
(isnan): Don't replace any existing macro.
(Ffrexp, Fldexp): Define even if !HAVE_COPYSIGN, as frexp and ldexp
are present on all C89 platforms.
(Ffrexp): Do not special-case zero, as frexp does the right thing
for that case.
(Flogb): Do not use logb, as it doesn't have the desired meaning
on hosts that use non-base-2 floating point. Instead, stick with
frexp, which is C89 anyway. Do not pass an infinity or a NaN to
frexp, to avoid getting an unspecified result.
This simplifies the code, and makes it a bit smaller and faster,
and (most important) makes it easier to clean up signal handling
since we can stop worring about floating-point exceptions in
library code. That was a problem before C89, but the problem
went away many years ago on all practical Emacs targets.
* configure.ac (frexp, fmod): Remove checks for these functions,
as we now assume them.
(FLOAT_CHECK_DOMAIN, HAVE_INVERSE_HYPERBOLIC, NO_MATHERR)
(HAVE_EXCEPTION):
Remove; no longer needed.
* admin/CPP-DEFINES (HAVE_FMOD, HAVE_FREXP, FLOAT_CHECK_DOMAIN)
(HAVE_INVERSE_HYPERBOLIC, NO_MATHERR): Remove.
* src/data.c, src/image.c, src/lread.c, src/print.c:
Don't include <math.h>; no longer needed.
* src/data.c, src/floatfns.c (IEEE_FLOATING_POINT): Don't worry that it
might be autoconfigured, as that never happens.
* src/data.c (fmod):
* src/doprnt.c (DBL_MAX_10_EXP):
* src/print.c (DBL_DIG):
Remove. C89 or later always defines these.
* src/floatfns.c (HAVE_MATHERR, FLOAT_CHECK_ERRNO, FLOAT_CHECK_DOMAIN)
(in_float, float_error_arg, float_error_arg2, float_error_fn_name)
(arith_error, domain_error, domain_error2):
Remove all this pre-C89 cruft. Do not include <errno.h> as that's
no longer needed -- we simply return what C returns. All uses removed.
(IN_FLOAT, IN_FLOAT2): Remove. All uses replaced with
the wrapped code.
(FLOAT_TO_INT, FLOAT_TO_INT2, range_error, range_error2):
Remove. All uses expanded, as these macros are no longer used
more than once and are now more trouble than they're worth.
(Ftan): Use tan, not sin / cos.
(Flogb): Assume C89 frexp.
(fmod_float): Assume C89 fmod.
(matherr) [HAVE_MATHERR]: Remove; no longer needed.
(init_floatfns): Remove. All uses removed.
Do not try to redefine the 'abort' function. Instead, redo
the code so that it calls 'emacs_abort' rather than 'abort'.
This removes the need for the NO_ABORT configure-time macro
and makes it easier to change the abort code to do a backtrace.
* configure.ac (NO_ABRT): Remove.
* admin/CPP-DEFINES (NO_ABORT): Remove.
* nt/inc/ms-w32.h (w32_abort) [HAVE_NTGUI]: Remove.
* src/.gdbinit: Just stop at emacs_abort, not at w32_abort or abort.
* src/emacs.c (abort) [!DOS_NT && !NO_ABORT]:
Remove; sysdep.c's emacs_abort now takes its place.
* src/lisp.h (emacs_abort): New decl. All calls from Emacs code to
'abort' changed to use 'emacs_abort'.
* src/msdos.c (dos_abort) [defined abort]: Remove; not used.
(abort) [!defined abort]: Rename to ...
(emacs_abort): ... new name.
* src/sysdep.c (emacs_abort) [!HAVE_NTGUI]: New function, taking
the place of the old 'abort' in emacs.c.
* src/w32.c, src/w32fns.c (abort): Do not #undef.
* src/w32.c (emacs_abort): Rename from w32_abort.
Merge from gnulib, incorporating:
2012-08-24 execinfo: port to FreeBSD
2012-08-22 execinfo: new module
* admin/merge-gnulib (GNULIB_MODULES): Add execinfo.
* lib/gnulib.mk, m4/gnulib-comp.m4: Regenerate.
* lib/execinfo.c, lib/execinfo.in.h, m4/execinfo.m4: New files.
* src/alloc.c [ENABLE_CHECKING]: Include <execinfo.h>.
(die) [ENABLE_CHECKING]: Print a backtrace if available.
* src/Makefile.in (LIB_EXECINFO): New macro.
(LIBES): Use it.
This is more natural, and on my platform (GCC 4.7.1 x86-64) it
makes Emacs's text size .03% smaller and presumably a bit faster.
* admin/merge-gnulib (GNULIB_MODULES): Add stdbool. This documents a
new direct dependency; stdbool was already being used indirectly
via other gnulib modules.
* lib-src/make-docfile.c (enum global_type): Sort values roughly in
decreasing alignment, except put functions last.
(compare_globals): Use this new property of enum global_type.
(write_globals): Use bool, not int, for booleans.
* src/lisp.h: Include <stdbool.h>.
(struct Lisp_Boolfwd, defvar_bool):
* src/lread.c (defvar_bool): Use bool, not int, for Lisp booleans.
* src/regex.c [!emacs]: Include <stdbool.h>.
(false, true): Remove; <stdbool.h> does this for us now.
* src/lisp.h (struct Lisp_Symbol): Change xname to meaningful
name since all xname users are fixed long time ago. Do not use
INTERNAL_FIELD.
(set_symbol_name, set_symbol_function, set_symbol_plist):
(set_symbol_next, set_overlay_plist): New function.
(struct Lisp_Cons): Do not use INTERNAL_FIELD.
(struct Lisp_Overlay): Likewise.
(CVAR, MVAR, SVAR): Remove.
* src/alloc.c, src/buffer.c, src/buffer.h, src/bytecode.c:
* src/cmds.c, src/data.c, src/doc.c, src/eval.c, src/fns.c:
* src/keyboard.c, src/lread.c, src/nsselect.m, src/xterm.c:
Adjust users.
* src/.gdbinit: Change to use name field of struct Lisp_Symbol
where appropriate.
* admin/coccinelle/overlay.cocci, admin/coccinelle/symbol.cocci:
Remove.
* configure.ac (GNU_LIBRARY_PENDING_OUTPUT_COUNT): Remove.
(PENDING_OUTPUT_COUNT): Absorb GNU_LIBRARY_PENDING_OUTPUT_COUNT.
(DISPNEW_NEEDS_STDIO_EXT): New define.
* src/dispnew.c (GNU_LIBRARY_PENDING_OUTPUT_COUNT, PENDING_OUTPUT_COUNT):
Let configure handle it.
(stdio_ext.h) [DISPNEW_NEEDS_STDIO_EXT]: Include it.
* nt/config.nt (GNU_LIBRARY_PENDING_OUTPUT_COUNT): Remove.
(PENDING_OUTPUT_COUNT): Define it as dispnew.c used to.
* admin/CPP-DEFINES (GNU_LIBRARY_PENDING_OUTPUT_COUNT): Remove.
Without this change, 'configure' fails because the recently-added
wait3 prototype in config.h messes up later 'configure' tests.
Fix this problem by droping wait3 and WRETCODE, as they're
no longer needed on hosts that are current porting targets.
* configure.ac (wait3, WRETCODE): Remove, fixing a FIXME.
All uses changed to waitpid and WEXITSTATUS.
* src/syswait.h (WRETCODE): Remove, consistently with ../configure.ac.
* src/lisp.h (struct Lisp_Cons): Use INTERNAL_FIELD.
Remove obsolete comment.
(MVAR): New macro.
(struct Lisp_Overlay): Use INTERNAL_FIELD.
* src/alloc.c, src/buffer.c, src/buffer.h, src/fns.c: Adjust users.
* admin/coccinelle/overlay.cocci: Semantic patch to replace direct
access to Lisp_Object members of struct Lisp_Overlay to MVAR.
* src/process.h (PVAR): New macro. Adjust style.
(struct Lisp_Process): Change Lisp_Object members to INTERNAL_FIELD.
* src/print.c, src/process.c, src/sysdep.c, src/w32.c:
* src/xdisp.c: Users changed.
* admin/coccinelle/process.cocci: Semantic patch to replace direct
access to Lisp_Object members of struct Lisp_Process to PVAR.
* .bzrignore: Add lib/stdalign.h.
* config.bat: Do not set NO_DECL_ALIGN; no longer needed.
Copy lib/stdalign.in.h to lib/stdalign.in-h as needed.
* configure.ac (HAVE_ATTRIBUTE_ALIGNED): Remove the code that
fiddles with this, as gnulib now does this for us.
* admin/merge-gnulib: Add environ, stdalign.
* m4/environ.m4: New file, from gnulib.
* lib/gnulib.mk, m4/gnulib-comp.m4: Regenerate.
* lib/stdalign.in.h, m4/stdalign.m4: New files, from gnulib.
* sed2v2.inp (HAVE_ATTRIBUTE_ALIGNED): Remove edit.
* sedlibmk.inp (STDALIGN_H, @GL_GENERATE_STDALIGN_H_TRUE@)
(GL_GENERATE_STDALIGN_H_FALSE): New edits.
* nt/config.nt (HAVE_ATTRIBUTE_ALIGNED): Remove.
* src/alloc.c (XMALLOC_BASE_ALIGNMENT, GC_POINTER_ALIGNMENT, pure_alloc):
Simplify by using alignof.
(pure_alloc) [! USE_LSB_TAG]: Don't over-align EMACS_INT values.
* src/lisp.h: Include <stdalign.h>.
(GCALIGNMENT): New macro and constant.
(DECL_ALIGN): Remove. All uses replaced by alignas (GCALIGNMENT).
(USE_LSB_TAG): ifdef on alignas, not on DECL_ALIGN.
(stdalign): New macro, if not already defined.
* src/lisp.h (build_unibyte_string): New function.
* src/dosfns.c, src/fileio.c, src/fns.c, src/ftfont.c, src/process.c:
* src/sysdep.c, src/w32fns.c, src/xfns.c: Use it.
* src/font.c (font_open_by_name): Change 2nd and 3rd args to the only arg
of type Lisp_Object to avoid redundant calls to make_unibyte_string.
Adjust users accordingly.
* src/font.h (font_open_by_name): Adjust prototype.
* admin/coccinelle/unibyte_string.cocci: Semantic patch to convert from
make_unibyte_string to build_unibyte_string where appropriate.
src/s/msdos.h (FILE_SYSTEM_CASE): Don't define.
src/fileio.c (FILE_SYSTEM_CASE): Don't define.
(Ffile_name_directory, Fexpand_file_name): Don't use FILE_SYSTEM_CASE.
Fixes problems on MS-DOS with Vtemp_file_name_pattern when
call-process-region passes it through expand-file-name.
src/dired.c (file_name_completion): Don't use FILE_SYSTEM_CASE.
admin/CPP-DEFINES: Remove FILE_SYSTEM_CASE.
The old name is also the name of a Mach system call.
* src/process.c (init_process_emacs): Rename from init_process.
* src/lisp.h, src/emacs.c: Update for this name change.
* src/nsgui.h, src/sysselect.h, src/s/darwin.h:
Remove workaround that is no longer needed.
* admin/coccinelle/list_loop.cocci: Semantic patch to convert from Fcdr
to XCDR and consistently use CONSP in the list iteration loops.
* admin/coccinelle/vector_contents.cocci: Fix indentation.
* src/callint.c, src/coding.c, src/doc.c, src/editfns.c, src/eval.c,
* src/font.c, src/fontset.c, src/frame.c, src/gnutls.c, src/minibuf.c,
* src/msdos.c, src/textprop.c, src/w32fns.c, src/w32menu.c, src/window.c,
* src/xmenu.c: Changed to use XCAR and XCDR where argument type is known
to be a Lisp_Cons.
* src/lisp.h: Define build_string as static inline. This provides
a better opportunity to optimize away calls to strlen when the
function is called with compile-time constant argument.
* src/image.c (imagemagick_error): Convert to build_string.
* src/w32proc.c (sys_spawnve): Likewise.
* src/xterm.c (x_term_init): Likewise.
* admin/coccinelle/build_string.cocci: Semantic patch
to convert from make_string to build_string.
* coccinelle: New subdirectory
* coccinelle/README: Documentation stub.
* coccinelle/vector_contents.cocci: Semantic patch to replace direct
access to `contents' member of Lisp_Vector objects with AREF and ASET
where appropriate.
