John Gilmore [Wed, 24 Jun 1992 19:58:57 +0000 (19:58 +0000)]
* dbxread.c (read_ofile_symtab): Remove N_CATCH special case,
since it is no longer used and burns time for every symbol read.
(process_one_symbol): Treat N_CATCH like default (complain).
* app.c (process_escape): new function to handle escapes the right
way, (do_scrub_next_char): use new function
* cond.c (s_ifdef): do ifdef/ifndef right
* read.c (s_fill): make the , expressions optional like the doc
says
* config/tc-h8300.[ch]: better warnings
Stu Grossman [Wed, 24 Jun 1992 07:31:55 +0000 (07:31 +0000)]
* partial-stab.h (case N_TEXT): Fix fencepost error when
detecting start of new symbol info after reading symtab for a
module with a zero length TEXT segment.
(Note that this fixes the BNR problem with printing vxWorksVersion. That ought
to make Ken happy).
Fred Fish [Wed, 24 Jun 1992 04:56:01 +0000 (04:56 +0000)]
* libbfd.h, libbfd-in.h (bfd_seek): Make prototype match definition.
* reloc.c (bfd_default_reloc_type_lookup): Make cast on return
type match actual return type.
* elf_print_symbol (section_name): Make CONST.
Fred Fish [Wed, 24 Jun 1992 04:49:48 +0000 (04:49 +0000)]
* Makefile.in (VERSION): Bump to 4.5.6.
* coffread.c (coff_end_symtab): Cast 2nd arg of complain() to
correct type.
* defs.h (NORETURN): Define away for Lucid compiler.
* remote.c (remote_timer, remote_interrupt): Signal handlers
take one int arg.
* ser-termios.c (serial_write, serial_close): Return whatever
value the write/close call returns, rather than falling off end.
* inferior.h (PTRACE_ARG3_TYPE): Third arg to ptrace is int on
more systems than it is "char *". Define PTRACE_ARG3_TYPE to
default to int.
* infptrace.c, hppabsd-xdep.c, hppahpux-xdep.c, i386-xdep.c,
inferior.h (call_ptrace): Use PTRACE_ARG3_TYPE to declare type
of third arg.
* a68v-xdep.c, arm-xdep.c, convex-xdep.c, hp300ux-xdep.c, infrun.c,
m88k-xdep.c, mach386-xdep.c, mips-xdep.c, os68k-xdep.c, pyr-tdep.c,
pyr-xdep.c, rs6000-xdep.c, sparc-xdep.c, sun3-xdep.c, sun386-xdep.c,
symm-xdep.c, ultra3-xdep.c: Use PTRACE_ARG3_TYPE to cast ptrace
argument 3.
* sparc-xdep.c, a68v-xdep.c (fetch_inferior_registers,
store_inferior_registers): Supply missing fourth argument to
ptrace().
Roland Pesch [Wed, 24 Jun 1992 00:15:44 +0000 (00:15 +0000)]
(1) removed three-part table of config name components from Installing
appendix (too hard to maintain, esp w/recent versions of config.sub).
Also corrected xref to this section accordingly.
(2) updated sample output from config.sub to match latest variation of
that script
(3) minor editing to gnu contrib text on set/show remotedebug
Stu Grossman [Tue, 23 Jun 1992 04:31:56 +0000 (04:31 +0000)]
* Centralize all Cadillac interface defs into defs.h.
* Use prototypes for all interface routines (fix bugs found as a result of
doing so).
* Make full_filename more robust in the face of null dirname/filenames.
* Get pty at startup instead of at child creation time. Put GDB descriptors
0, 1 and 2 on it so that we can see the output of shell commands.
* Get rid of prompt(), use print_prompt() instead. Fixes loads of stupid
prompting problems.
Fred Fish [Tue, 23 Jun 1992 03:33:47 +0000 (03:33 +0000)]
* command.c (add_abbrev_prefix_cmd): Fix misspelling in comment.
* dwarfread.c (enum_type): Fix misspelling in comment.
* valprint.c (val_print_fields, cplus_val_print): Minor
adjustment to printing of C++ class structures to more closely
match format for printing C structures, with and without setting
pretty print.
Per Bothner [Tue, 23 Jun 1992 01:05:47 +0000 (01:05 +0000)]
* i386linux.c, hosts/i386linux.h, config/i386-linux.mt:
New files, for Linux (a free Unix clone for 386 machines).
* Makefile.in, configure.in, targets.c: Update accordingly.
Per Bothner [Tue, 23 Jun 1992 00:25:11 +0000 (00:25 +0000)]
* infrun.c (wait_for_inferior): Don't test for SIGEMT
(which is not a Posix signal) if it isn't defined.
* tm-linux.h, xm-linux.h, config/linux.m[ht], configure.in:
New port to Linux (a free Unix clone for 386 machines).
Fri Jun 19 20:23:21 1992 Steve Chamberlain ([email protected])
Lints for bfd_reloc_code_type->bfd_reloc_code_real_type, and
correct calling of howto special functions and fixes from Raeburn
for gas<>bfdness
* archures.c: nuke bfd_reloc_code_type
* bout.c, cpu-h8300.c, coff-i960.c, coff-m88k.c: special function lint
* coffcode.h (styp_to_sec_flags): STYP_INFO is marked as
SEC_NEVER_LOAD, various other bfdgas newness
* reloc.c->libbfd.h: change protype of bfd_default_reloc_type_lookup
* targets.c: change jump table vector to above
John Gilmore [Sun, 21 Jun 1992 02:20:30 +0000 (02:20 +0000)]
COFF changes for dealing better with EPI 29K C compiler output.
* coffread.c (record_minimal_symbol): Pass the minsym type to it.
Callers changed.
(coff_end_symtab): Sort blocks if needed. Complain if misordered.
(read_coff_symtab): Move patch_opaque_types call from
coff_symfile_read. Restrict it to symtabs from this objfile.
(process_coff_symbol: C_TPDEF): Don't put ordinary foward
references on opaque type chain; just let coff_lookup_type handle 'em.
(decode_type): Complain about tagndx values on
non-struct/union/enum types, which the EPI compiler tends to produce.
* symtab.c (list_symbols): Make minimal symbol variable-finding work.
* tm-68k.h (FIX_CALL_DUMMY): Avoid alignment and byte order
dependency.
* elfread.c (elf_symfile_read): Update bfd_elf_find_section
usage to match new prototype. Include libbfd.h to get prototype.
John Gilmore [Sat, 20 Jun 1992 02:04:20 +0000 (02:04 +0000)]
* elf.c (bfd_elf_find_section): Mark as INTERNAL_FUNCTION so
its prototype will be included in libbfd.h. Change result type
to struct * (rather than equivalent typedef) so it can be used in
the prototype, where they typedef won't be known.
* libbfd.h: Updated version.
Stu Grossman [Fri, 19 Jun 1992 22:39:44 +0000 (22:39 +0000)]
* Makefile.in (c-exp.tab.c, m2-exp.tab.c): Filter out bogus extern
declarations of malloc/realloc/free that are inserted by some
versions of yacc.
* m2-exp.y: Prevent conflicts with TRUE and FALSE tokens by
#undeffing them.
* xm-rs6000.h: Declare malloc/realloc/free appropriately. Yet
another decl of strdup (this really ought to come from libiberty.h!).
* dbxread.c (process_one_symbol:N_FUN): GCC now produces relative
N_SLINE's, etc, just like Sun cc on Solaris2.
* am29k-tdep.c (read_register_stack, write_register_stack):
Change RSTACK_HIGH_ADDR to rstack_high_address, a user-settable
variable. Add `set' and `show' commands for it.
* doc/gdb.texinfo: Document it.
* eval.c: Avoid residue-by-zero when evaluating without side effects.
(Bug and fix found by Pierre Willard.)
Fred Fish [Fri, 19 Jun 1992 02:43:22 +0000 (02:43 +0000)]
* valprint.c (type_print_1): Plug memory leak. Print all
C++ syms as demangled, not just functions.
* buildsym.c (read_range_type): When we find a signed char
type, do a lookup of signed char, not plain char. Plain char's
still get looked up as plain char's elsewhere.
Roland Pesch [Thu, 18 Jun 1992 20:03:38 +0000 (20:03 +0000)]
1) Better fix for section "Readline Killing Commands": delete entire
first paragraph. It looks like it's left over from previous
nomenclature; term "cutting" introduced there is not used in rest of
document. ("killing" is used instead, as in the section title and in
the following paragraph.)
2) More consistent use of @key for things to type in; in a few cases
not having it actually produced wrong doc on paper---notably "M--"
comes out as "M" followed by emdash, if in the plain Roman body
text...
Roland Pesch [Thu, 18 Jun 1992 19:04:16 +0000 (19:04 +0000)]
Delete dangling "There is a" at end of first paragraph under "Readline
Killing Commands". None of our saved versions has the rest of this sentence,
whatever it may have been...
* h8300hms.em, h8300hms.sc-sh: do the right thing for -r
* ldexp.c: lint
* ldlang.c(open_output): set the target arch and machine as soon
as we can. (lang_size_sections): use new macros for setting vma
* ldwrite.c: lint
Fred Fish [Wed, 17 Jun 1992 20:58:46 +0000 (20:58 +0000)]
* elf.c (bfd_section_from_shdr, elf_slurp_symbol_table):
Correct misconception that there can be only one symbol table.
Only call elf_slurp_symbol_table on the full symbol table, not
the dynamic one which is only a subset of the full one.
(The bfd library was slurping in the first ELF symbol table it
found and then ignoring any others. For dynamically linked
executables, this happened to be the dynamic symbol table, which
is a subset of the full symbol table and only contains the
symbols that need to be used in dynamic linking.)
Stu Grossman [Wed, 17 Jun 1992 20:15:03 +0000 (20:15 +0000)]
* xcoffexec.c: Remove ' from comment.
* xm-sun3os4.h: Define MALLOC_INCOMPATIBLE to avoid conflicts
with decls of malloc in c-exp.tab.c (as produced by yacc).
There's got to be a better way to do this...
Fred Fish [Wed, 17 Jun 1992 18:14:22 +0000 (18:14 +0000)]
* partial-stab.h: Convert single rindex use to strrchr.
* mipsread.c, dbxread.c: Remove troublesome inclusion of non-
standard <strings.h> file, now that the only single use of
rindex in the gdb source files is gone.
Fred Fish [Wed, 17 Jun 1992 05:23:00 +0000 (05:23 +0000)]
* mipsread.c: Undo ill effects from including <strings.h>,
which #defines index to be strchr. Unfortunately, index is
a member of a symbol table structure that can't be changed.
* mipsread.c: tm-mips.h includes coff/symconst.h and coff/sym.h,
remove redundant #include's.
Ken Raeburn [Tue, 16 Jun 1992 12:04:03 +0000 (12:04 +0000)]
[time to file a PR on cvs...]
Various changes to get linker working again for a.out:
* don't set/adjust section file positions or vmas more than once
* use correct page size and segment size when calculating them
* deal with some variations in a.out implementations
Tested on sun4 and sun4->sun3 so far, will be testing further but
needed to get wider exposure&testing. See ChangeLog for details.
Also:
* coffcode.h (coff_write_relocs): Write out swapped reloc, not
pre-swapped version.
* hosts/sparc.h (abort, exit): Hide these names if compiling with
gcc version 2, to avoid warnings.