Mon Jun 29 09:47:33 1998 Craig Burley <burley@gnu.org>
Fix 980628-*.f:
* bld.h: New `pad' field and accessor macros for
ACCTER, ARRTER, and CONTER ops.
* bld.c (ffebld_new_accter, ffebld_new_arrter,
ffebld_new_conter_with_orig): Initialize `pad' field
to zero.
* com.c (ffecom_transform_common_): Include initial
padding (aka modulo aka offset) in size calculation.
Copy initial padding value into FFE initialization expression
so the GBE transformation of that expression includes it.
Make array low bound 0 instead of 1, for consistency.
(ffecom_transform_equiv_): Include initial
padding (aka modulo aka offset) in size calculation.
Copy initial padding value into FFE initialization expression
so the GBE transformation of that expression includes it.
Make array low bound 0 instead of 1, for consistency.
(ffecom_expr_, case FFEBLD_opACCTER): Delete unused `size'
variable.
Track destination offset separately, allowing for
initial padding.
Don't bother setting initial PURPOSE offset if zero.
Include initial padding in size calculation.
(ffecom_expr_, case FFEBLD_opARRTER): Allow for
initial padding.
Include initial padding in size calculation.
Make array low bound 0 instead of 1, for consistency.
(ffecom_finish_global_): Make array low bound 0 instead
of 1, for consistency.
(ffecom_notify_init_storage): Copy `pad' field from old
ACCTER to new ARRTER.
(ffecom_notify_init_symbol): Ditto.
* data.c (ffedata_gather_): Initialize `pad' field in new
ARRTER to 0.
(ffedata_value_): Ditto.
* equiv.c (ffeequiv_layout_local_): When lowering start
of equiv area, extend lowering to maintain needed alignment.
* target.c (ffetarget_align): Handle negative offset correctly.
* global.c (ffeglobal_pad_common): Warn about non-zero
padding only the first time its seen.
If new padding larger than old, update old.
(ffeglobal_save_common): Use correct type for size throughout.
* global.h: Use correct type for size throughout.
(ffeglobal_common_pad): New macro.
(ffeglobal_pad): Delete this unused and broken macro.
From-SVN: r20817
Thu May 28 21:32:18 1998 Craig Burley <burley@gnu.org>
Restore circa-0.5.22 capabilities of `g77' driver:
* Make-lang.in (g77spec.o): Depend on f/version.h.
(g77version.o): New rule to compile g77 version info.
(g77$(exeext)): Depend on and link in g77version.o.
* g77spec.c: Rewrite to be more like 0.5.22 version
of g77.c, making filtering of command line smarter
so mixed Fortran and C (etc.) can be compiled, verbose
version info can be obtained, etc.
* lang-specs.h (f77-version): New "language" to support
"g77 -v" command under new gcc 2.8 regime.
* lex.c (ffelex_file_fixed): If -fnull-version, just
substitute a "source file" that prints out version info.
* top.c, top.h: Support -fnull-version.
* lang-specs.h: Use "%O" instead of OO macro to specify
object extension. Remove old stringizing cruft.
* Make-lang.in (g77.c, g77spec.o, g77.o, g77$(exeext),
g77-cross$(exeext), f771,
$(srcdir)/f/g77.info, $(srcdir)/f/g77.dvi,
$(srcdir)/f/intdoc.texi,
f77.install-common, f77.install-info, f77.install-man,
f77.uninstall, $(G77STAGESTUFF), f77.stage1, f77.stage2,
f77.stage3, f77.stage4, f77.distdir): Don't do anything
unless user specified "f77" or "F77" in $LANGUAGES either
during configuration or explicitly. For convenience of
various tests and to work around lack of the assignment
"LANGUAGES=$(BOOT_LANGUAGES)" in the "make stage1" command
of "make bootstrap" in gcc, use a touch file named "lang-f77"
to communicate whether this is the case.
* Make-lang.in (F77_FLAGS_TO_PASS): Delete this macro,
replace with minimal expansion of its former self in
each of the two instances where it was used.
* Makefile.in (HOST_CC): Delete this definition.
* com.c (index, rindex): Delete these declarations.
* proj.h: (isascii): Delete this.
* Make-lang.in (f77.install-common): Warn if `f77-install-ok'
flag-file exists, since it no longer triggers any activity.
Rename libf2c.a and f2c.h to libg2c.a and g2c.h,
normalize and simplify g77/libg2c build process:
* Make-lang.in: Remove all support for overwriting
/usr/bin/f77 etc., or whatever the actual names are
via $(prefix) and $(local_prefix). (g++ overwrites
/usr/bin/c++, but then it's often the only C++ compiler
on the system; f77 often exists on systems that are
installing g77.)
(f77.realclean): Remove obsolete target.
(g77.c, g77$(exeext)): Minor changes to look more like g++'s
stuff.
(f771): Now built with srcdir=gcc/f, not srcdir=gcc, to be
more like g++ and such.
(f/Makefile): Removed, as g++ doesn't need this rule.
(f77.install-common): No longer install f77, etc.
(f77.install-man): No longer install f77.1.
(f77.uninstall): No longer uninstall f77, f77.1, etc.
(f77.stage1, f77.stage2, f77.stage3, f77.stage4): Do work
only if "f77" appears in $(LANGUAGES).
(Note: gcc's Makefile.in's bootstrap target should set
LANGUAGES=$(BOOT_LANGUAGES) when making the stage1 target.)
* Makefile.in: Update vis-a-vis gcc/cp/Makefile.in.
(none): Remove.
(g77-only): Relocate.
(all.indirect, f771, *.o): Now assumes current directory
is this dir (gcc/f), not the parent directory.
(TAGS): Remove "echo 'parse.y,0' >> TAGS ;" line.
* config-lang.in: Delete commented-out code.
Fix stagestuff definition. Add more stuff to
diff_excludes definition. Don't create any directories.
Set outputs to f/Makefile, to get variable substition
to happen (what does that really do, anyway?!).
* g77spec.c: Rename libf2c to libg2c.
* com.h: Remove all of the gcc back-end decls,
since egcs should have all of them correct.
* com.c: Include "proj.h" before anything else,
as that's how things are supposed to work.
* ste.c: Ditto.
* bad.c: Include "flags.j" here, since some diagnostics
check flag_pedantic_errors.
* Makefile.in (f/*.o): Rebuild dependencies via
deps-kinda.
* output.j: New source file.
* Make-lang.in (F77_SRCS): Update accordingly.
* Makefile.in (OUTPUT_H): Ditto.
(deps-kinda): Ditto.
* com.c: Include "output.j" here.
* lex.c: Ditto.
From-SVN: r20506
* com.c (ffecom_char_args_x_): Renamed from ffecom_char_args_,
which is now a macro (to avoid lots of changes to other code)
with new arg, ffecom_char_args_with_null_ being another new
macro to call same function with different value for new arg.
This function now appends a null byte to opCONTER expression
if the new arg is TRUE.
(ffecom_arg_ptr_to_expr): Support NULL length pointer.
* ste.c (ffeste_io_cilist_):
(ffeste_io_icilist_): Pass NULL length ptr for
FORMAT expression, so null byte gets appended where
feasible.
* target.c (ffetarget_character1):
(ffetarget_concatenate_character1):
(ffetarget_substr_character1):
(ffetarget_convert_character1_character1):
(ffetarget_convert_character1_hollerith):
(ffetarget_convert_character1_integer4):
(ffetarget_convert_character1_logical4):
(ffetarget_convert_character1_typeless):
(ffetarget_hollerith): Append extra phantom null byte as
part of FFETARGET-NULL-BYTE kludge.
Yes, even more patches from Craig :-)
From-SVN: r18187
