binutils-gdb.git
32 years ago * elf.c (section_from_elf_index): Return bfd_abs_section, not 0,
Brendan Kehoe [Fri, 25 Sep 1992 22:05:15 +0000 (22:05 +0000)]
* elf.c (section_from_elf_index): Return bfd_abs_section, not 0,
since we should never have a NULL section.
(elf_slurp_symbol_table): If st_shndx doesn't match any of our
tests, set the section to bfd_abs_section.

32 years agodefined(__STDC__) / __STDC__==1 changes recommended by gnu
Ken Raeburn [Fri, 25 Sep 1992 21:14:46 +0000 (21:14 +0000)]
defined(__STDC__) / __STDC__==1 changes recommended by gnu

32 years ago* Makefile.in (ldexp.o, ldctor.o, ldlang.o, ldmain.o, ldwrite.o, lexsup.o,
Ken Raeburn [Fri, 25 Sep 1992 21:05:20 +0000 (21:05 +0000)]
* Makefile.in (ldexp.o, ldctor.o, ldlang.o, ldmain.o, ldwrite.o, lexsup.o,
mri.o, relax.o): Indicate dependence on ldgram.h.

* ld.h (strip_symbols_type): Add value STRIP_SOME.
* ldgram.y (OPTION_RETAIN_SYMBOLS_FILE): New terminal token.
(lang_add_keepsyms_file): New function.
* ldlex.l: Handle "-retain-symbols-file".
* ldsym.c (keepsyms_file, kept_syms): New vars.
(process_keepsyms): New functihon; reads file, marks symbols for saving.
(write_file_locals): File symbols should always be kept.
(ldsym_write): Warn about "-retain-symbols-file" overriding "-S" and "-s".
Process retain-symbols file before setting symtab.
* ldsym.h (SYM_KEEP): New flag for ldsym_type flags.
(keepsyms_file, kept_syms): Declare them.

* ldmain.c (main): Non-fatal errors should still cause non-zero exit status
even with -r.

32 years ago* configure.in: Add ncr* configuration.
Stu Grossman [Fri, 25 Sep 1992 19:54:06 +0000 (19:54 +0000)]
* configure.in:  Add ncr* configuration.
* z8k-dis.c (struct instr_data_s):  Make instr_asmsrc char to make
picayune ANSI compilers happy.

32 years ago* readline.c (clear_to_eol, rl_generic_bind): Make static.
Stu Grossman [Fri, 25 Sep 1992 19:48:50 +0000 (19:48 +0000)]
* readline.c (clear_to_eol, rl_generic_bind):  Make static.
(rl_digit_loop):  Add arg to call to rl_message().
* vi_mode.c (rl_vi_first_print):  Add arg to call to
rl_back_to_indent().

32 years ago * inftarg.c (child_create_inferior, child_attach,
K. Richard Pixley [Fri, 25 Sep 1992 19:19:08 +0000 (19:19 +0000)]
* inftarg.c (child_create_inferior, child_attach,
  child_mourn_inferior): collect unix child process stratum
  functions which live below the target vector into this file to
  facilitate host/target/native split.  Also, make them static.
* inflow.c (child_mourn_inferior): removed.
* infrun.c (child_create_inferior, child_attach): removed.
  (resume): becomes global so that functions below the
  target vector can find it.
* inferior.h (resume): add prototype.
  (child_mourn_inferior, child_create_inferior, child_attach):
  remove prototypes.

32 years ago * xcoffexec.c (exec_ops): child_attach and child_create_inferior
K. Richard Pixley [Fri, 25 Sep 1992 19:11:25 +0000 (19:11 +0000)]
* xcoffexec.c (exec_ops): child_attach and child_create_inferior
  replaced with find_default_attach and
  find_default_create_inferior.

32 years agoFri Sep 25 11:11:57 1992 Steve Chamberlain (sac@thepub.cygnus.com)
Steve Chamberlain [Fri, 25 Sep 1992 18:14:29 +0000 (18:14 +0000)]
Fri Sep 25 11:11:57 1992  Steve Chamberlain  (sac@thepub.cygnus.com)

* coff-z8k.c: getting closer
* coffcode.h (coff_get_relocated_section_contents): allow
EXTRA_CASES hooks for different arches to provide different reloc
types.
* seclet.h: #ifndef around it to allow multiple inclusion
* srec.c: minor doc fix

32 years agoFri Sep 25 11:08:01 1992 Steve Chamberlain (sac@thepub.cygnus.com)
Steve Chamberlain [Fri, 25 Sep 1992 18:11:04 +0000 (18:11 +0000)]
Fri Sep 25 11:08:01 1992  Steve Chamberlain  (sac@thepub.cygnus.com)

Added initial support for the z8k
* z8ksim.em, z8ksim.sc-sh, z8ksim.sh: new files
* configure.in, Makefile.in: modified to reflect above

* ldlang.c (lang_check): when linking conflicting architectures,
make the output file reflect at least one of the bad inputs.

32 years agouse _filtered routines for printing so symbolic addresses show up in the right place
Ken Raeburn [Fri, 25 Sep 1992 17:23:59 +0000 (17:23 +0000)]
use _filtered routines for printing so symbolic addresses show up in the right place

32 years agokeep corelow.c
K. Richard Pixley [Fri, 25 Sep 1992 17:04:36 +0000 (17:04 +0000)]
keep corelow.c

32 years ago Separate core functions along target vector in preparation for
K. Richard Pixley [Fri, 25 Sep 1992 17:03:55 +0000 (17:03 +0000)]
    Separate core functions along target vector in preparation for
        native support.  Functions above vector now live in core.c.  Those
        below in corelow.c.

* core.c (solib_add_stub, core_close, core_open, core_detach,
  get_core_registers, core_files_info, core_ops): moved to corelow.c
  (_initialize_core): removed addition of core_ops target.
* corelow.c: new file.
(solib_add_stub, core_close, core_open, core_detach,
get_core_registers, core_files_info, core_ops): moved from core.c
(_initialize_corelow): new function.
* gdbcore.h (core_open, core_detach): added prototypes.
(core_ops): add forward declaration.
* Makefile.in (SFILES_MAINDIR): add core.c
(OBS): add core.o
(TSOBS): change core.o to corelow.o

32 years agoz8kgen: temp file to generate z8k-opc.h from
Steve Chamberlain [Thu, 24 Sep 1992 20:28:14 +0000 (20:28 +0000)]
z8kgen: temp file to generate z8k-opc.h from

z8k-dis.c: fixed various bugs

32 years ago* listing.c (debugging_pseudo): Add stabs and stabn as things to ignore.
Brendan Kehoe [Thu, 24 Sep 1992 19:43:45 +0000 (19:43 +0000)]
* listing.c (debugging_pseudo): Add stabs and stabn as things to ignore.

32 years ago* m68k-tdep.c (m68k_saved_pc_after_call): Use 'GDB_TARGET_IS_SUN3'
Stu Grossman [Wed, 23 Sep 1992 18:19:19 +0000 (18:19 +0000)]
* m68k-tdep.c (m68k_saved_pc_after_call):  Use 'GDB_TARGET_IS_SUN3'
instead of 'sun' predefined symbol so that trap analysis code is
enabled only when the TARGET is a sun3.

32 years agofix repeated ext_format_i960 local decls so sun4 cc accepts it
Ken Raeburn [Tue, 22 Sep 1992 22:02:54 +0000 (22:02 +0000)]
fix repeated ext_format_i960 local decls so sun4 cc accepts it

32 years agofix repeated ext_format_i960 local decls so sun4 cc accepts it
Ken Raeburn [Tue, 22 Sep 1992 21:55:21 +0000 (21:55 +0000)]
fix repeated ext_format_i960 local decls so sun4 cc accepts it

32 years agoTue Sep 22 13:02:07 1992 Sean Eric Fagan (sef@cygnus.com)
Sean Eric Fagan [Tue, 22 Sep 1992 20:07:55 +0000 (20:07 +0000)]
Tue Sep 22 13:02:07 1992  Sean Eric Fagan  (sef@cygnus.com)

* obj-coffbfd.c (do_relocs_for,fill_section): now allocate all
sections starting from zero, rather than making them consecutive.
This makes subsequent reloc calculations easier, esp if the
object format doesn't understand addends. (obj_coff_lcomm): (maybe
temporarily) allocate lcomm in .data rather than in .bss.  It
seems that some tools can't cope with a non-zero sized bss
before linkage.

32 years agocpu32 and -mno-688?1 support from p3
Ken Raeburn [Tue, 22 Sep 1992 19:24:33 +0000 (19:24 +0000)]
cpu32 and -mno-688?1 support from p3

32 years agocpu32 and -mno-688?1 support from p3
Ken Raeburn [Tue, 22 Sep 1992 19:19:02 +0000 (19:19 +0000)]
cpu32 and -mno-688?1 support from p3

32 years agoRoll bask the WRS change, since it had already been applied, and the
Ian Lance Taylor [Tue, 22 Sep 1992 15:32:03 +0000 (15:32 +0000)]
Roll bask the WRS change, since it had already been applied, and the
reversed patch they sent blew my mind.

32 years ago* mips-pinsn.c (print_insn_arg, case 'B'): Disassemble `break'
John Gilmore [Tue, 22 Sep 1992 07:46:14 +0000 (07:46 +0000)]
* mips-pinsn.c (print_insn_arg, case 'B'):  Disassemble `break'
instruction's argument.  Patch from jonathan@cs.stanford.edu
(Jonathan Stone).

32 years agoDescribe coding style of GDB.
John Gilmore [Tue, 22 Sep 1992 07:34:45 +0000 (07:34 +0000)]
Describe coding style of GDB.

32 years agoBreak the thread of control that implies that a unix child
John Gilmore [Tue, 22 Sep 1992 07:23:35 +0000 (07:23 +0000)]
Break the thread of control that implies that a unix child
process will be the default target.

* target.c (find_default_run_target, find_default_attach,
find_default_create_inferior, return_zero):  new functions.
(cleanup_target):  Make return_zero the default for to_can_run.

* exec.c (exec_ops), core.c (core_ops):  Replace child_attach and
child_create_inferior references with find_default_XXX instead.

* target.h (struct target_ops): new field, to_can_run.
(find_default_attach, find_default_create_inferior):  new prototypes.
(target_can_run): new macro.

* Also added a zero (default) to_can_run element to all static
struct target_ops initializations throughout GDB, except:
* inftarg.c (child_ops): Use new child_can_run() to enable child runs.

* infrun.c (child_create_inferior):  Clean up error handling when
no exec file is specified.
(child_attach):  Don't require exec file.

32 years agoRemove kill_inferior_fast, in favor of target_kill, which goes
John Gilmore [Tue, 22 Sep 1992 05:23:57 +0000 (05:23 +0000)]
Remove kill_inferior_fast, in favor of target_kill, which goes
through the target vector.

* inferior.h (kill_inferior_fast): remove declaration.

* main.c (disconnect): call quit_cover using catch_errors rather
than calling kill_inferior_fast directly.  New way goes through
the target vector, handles attached processes, and writes
command history if appropriate.
(quit_cover): new function, wrapper for quit_command.

* convex-xdep.c, go32-xdep.c, hppabsd-xdep.c, hppahpux-xdep.c,
infptrace.c, procfs.c:  Removed all instances of kill_inferior_fast,
inlining them into the local kill_inferior when needed.

32 years ago* infrun.c (_initialize_infrun): Alias `i handle' == `i signals'.
John Gilmore [Tue, 22 Sep 1992 02:33:11 +0000 (02:33 +0000)]
* infrun.c (_initialize_infrun):  Alias `i handle' == `i signals'.
* stabsread.c (read_struct_type): Simplify complicated expression
for dumb DECstation compiler.

32 years agoMinor wording changes.
John Gilmore [Tue, 22 Sep 1992 02:32:39 +0000 (02:32 +0000)]
Minor wording changes.

32 years agoMinor wording changes.
John Gilmore [Tue, 22 Sep 1992 02:26:50 +0000 (02:26 +0000)]
Minor wording changes.

32 years agoPatches from WRS:
Ian Lance Taylor [Mon, 21 Sep 1992 22:36:04 +0000 (22:36 +0000)]
Patches from WRS:

Mon Sep 21 14:54:35 1992  Ian Lance Taylor  (ian@cygnus.com)

* m68k-pinsn (print_insn_arg, fetch_arg): added support for
operands to memory management instructions, from WRS.

32 years agoPatches from WRS:
Ian Lance Taylor [Mon, 21 Sep 1992 22:31:41 +0000 (22:31 +0000)]
Patches from WRS:

Mon Sep 21 14:44:17 1992  Ian Lance Taylor  (ian@cygnus.com)

* tc-m68k.c (m68k_ip): fixed move16 operand handling, from WRS.

32 years agoPatches from WRS:
Ian Lance Taylor [Mon, 21 Sep 1992 22:30:36 +0000 (22:30 +0000)]
Patches from WRS:

Mon Sep 21 14:41:30 1992  Ian Lance Taylor  (ian@cygnus.com)

* mh-irix4: put in patches from WRS.

32 years agoPatches from WRS:
Ian Lance Taylor [Mon, 21 Sep 1992 22:29:51 +0000 (22:29 +0000)]
Patches from WRS:

Mon Sep 21 14:39:56 1992  Ian Lance Taylor  (ian@cygnus.com)

* m68k-pinsn (print_insn_arg, fetch_arg): added support for
operands to memory management instructions, from WRS.

32 years agoPatches from WRS:
Ian Lance Taylor [Mon, 21 Sep 1992 22:28:42 +0000 (22:28 +0000)]
Patches from WRS:

Mon Sep 21 14:33:58 1992  Ian Lance Taylor  (ian@cygnus.com)

* hosts/hp9000.h, hosts/irix3.h: changes from WRS.

32 years agotense test typo
K. Richard Pixley [Mon, 21 Sep 1992 20:01:00 +0000 (20:01 +0000)]
tense test typo

32 years agofix typo
K. Richard Pixley [Mon, 21 Sep 1992 19:47:39 +0000 (19:47 +0000)]
fix typo

32 years agoMore conditional markup so HMS (Hitachi) can have the configuration
Roland Pesch [Mon, 21 Sep 1992 17:27:34 +0000 (17:27 +0000)]
More conditional markup so HMS (Hitachi) can have the configuration
of their dreams.

32 years ago * configure.in: Use i386-elf for all i386/i486 sysv4 hosts, not
Fred Fish [Sun, 20 Sep 1992 15:50:12 +0000 (15:50 +0000)]
* configure.in:  Use i386-elf for all i386/i486 sysv4 hosts, not
just ncr.

32 years ago * main.c (main): Back out previous language setting changes.
Fred Fish [Sun, 20 Sep 1992 15:45:52 +0000 (15:45 +0000)]
* main.c (main):  Back out previous language setting changes.
Replace with simple default to C before processing any init
files.  There MUST be a language set, even in the absence of
init files or executables, or expression parsing fails.

32 years ago * main.c (main): Move code that sets initial language to
Fred Fish [Sat, 19 Sep 1992 16:55:34 +0000 (16:55 +0000)]
* main.c (main):  Move code that sets initial language to
symfile.c.  Fixup places where command files are processed to
be consistent in setting a default language if none has been
previously set.
* symfile.c (set_initial_language):  Add code moved from
main() that sets an initial default language when a new
symbol file is read.
* symfile.c (symbol_file_command):  Call set_initial_language.
* symtab.c (find_main_psymtab):  Add FIXME comment.

32 years ago* rs6000-tdep.c (push_dummy_frame, pop_dummy_frame): replace
John Gilmore [Sat, 19 Sep 1992 12:08:23 +0000 (12:08 +0000)]
* rs6000-tdep.c (push_dummy_frame, pop_dummy_frame): replace
calls to {fetch|store}_inferior_registers with calls to
target_{fetch|store}_registers, for remote-target independence.

32 years ago* configure.in: removed target_dependent line.
John Gilmore [Sat, 19 Sep 1992 11:55:20 +0000 (11:55 +0000)]
* configure.in: removed target_dependent line.
* remote-mm.c (mm_attach): change printf to error to prevent
fallthrough bug.
* remote-udi.c (udi_attach): change printf to error to prevent
fallthrough bug.

32 years agocomment changes
John Gilmore [Sat, 19 Sep 1992 11:47:19 +0000 (11:47 +0000)]
comment changes

32 years agoLeft out ChangeLog, oops
John Gilmore [Sat, 19 Sep 1992 11:45:20 +0000 (11:45 +0000)]
Left out ChangeLog, oops

32 years ago* Makefile.in: Fix bug I introduced in merging Rich's change.
John Gilmore [Sat, 19 Sep 1992 11:25:52 +0000 (11:25 +0000)]
* Makefile.in:  Fix bug I introduced in merging Rich's change.
* infrun.c (child_create_inferior):  Use proceed() rather than
doing all the same stuff by hand.

32 years ago* Makefile.in, configure.in: add support for files used only when
John Gilmore [Sat, 19 Sep 1992 11:09:01 +0000 (11:09 +0000)]
* Makefile.in, configure.in: add support for files used only when
configured native, that is, when host == target.

32 years ago* tm-sparc.h, xm-sparc.h: externs and macros relating to deferred
John Gilmore [Sat, 19 Sep 1992 10:56:01 +0000 (10:56 +0000)]
* tm-sparc.h, xm-sparc.h: externs and macros relating to deferred
stores are target dependent and were moved from xm to tm.

32 years ago* Makefile.in (gdb-all.texi): Build in objdir, not $srcdir.
John Gilmore [Sat, 19 Sep 1992 10:47:09 +0000 (10:47 +0000)]
* Makefile.in (gdb-all.texi):  Build in objdir, not $srcdir.

32 years ago* config/decstation.mh: Add MMALLOC_LIB and MMALLOC_DISABLE
John Gilmore [Sat, 19 Sep 1992 10:21:04 +0000 (10:21 +0000)]
* config/decstation.mh:  Add MMALLOC_LIB and MMALLOC_DISABLE
overrides, to avoid problem in xdr_bytes with malloc(0) => 0.

32 years agofix typo in install rule
Ken Raeburn [Fri, 18 Sep 1992 21:02:18 +0000 (21:02 +0000)]
fix typo in install rule

32 years agoPatched a couple of minor bugs in UDI support. These have been sent
Ian Lance Taylor [Fri, 18 Sep 1992 19:37:55 +0000 (19:37 +0000)]
Patched a couple of minor bugs in UDI support.  These have been sent
to AMD.

32 years agofix uninitialized pointer when reading relocs
Ken Raeburn [Fri, 18 Sep 1992 16:33:16 +0000 (16:33 +0000)]
fix uninitialized pointer when reading relocs

32 years ago * tc-m68k.c (m68k_ip): An(disp) is not pc relative.
Steve Chamberlain [Fri, 18 Sep 1992 15:03:59 +0000 (15:03 +0000)]
* tc-m68k.c (m68k_ip): An(disp) is not pc relative.

32 years agoRemove convert_{to,from}_virtual members from target structs.
John Gilmore [Fri, 18 Sep 1992 09:56:28 +0000 (09:56 +0000)]
Remove convert_{to,from}_virtual members from target structs.

32 years ago* target.h (struct target_ops): removed to_convert_to_virtual and
John Gilmore [Fri, 18 Sep 1992 09:53:47 +0000 (09:53 +0000)]
* target.h (struct target_ops):  removed to_convert_to_virtual and
to_convert_from_virtual elements.  Initializations removed from
all static initializations.
(target_convert_to_virtual, target_convert_from_virtual):
removed.
(host_convert_to_virtual, host_convert_from_virtual):  Removed
forward declarations.
* target.c (cleanup_target): removed default assignments for
to_convert_to_virtual and to_convert_from_virtual.
* inftarg.c (host_convert_to_virtual, host_convert_from_virtual):
removed.
* findvar.c (value_of_register, value_from_register):
target_convert_to_virtual inlined.
* infcmd.c (do_registers_info):  target_convert_to_virtual inlined.
* valops.c (value_assign):  target_convert_from_virtual inlined.

32 years ago* main.c (gdb_readline): Avoid printf_filtered, which sets char
John Gilmore [Fri, 18 Sep 1992 09:21:44 +0000 (09:21 +0000)]
* main.c (gdb_readline):  Avoid printf_filtered, which sets char
position wrong if used for the prompt.
* utils.c (puts_filtered):  Comment:  NOT a puts() replacement!

32 years agoSupport for accessing arbitrary MIPS stack frames in memory.
John Gilmore [Fri, 18 Sep 1992 09:20:15 +0000 (09:20 +0000)]
Support for accessing arbitrary MIPS stack frames in memory.

* blockframe.c (get_prev_frame_info):  If INIT_FRAME_PC_FIRST is
set, run it before INIT_EXTRA_FRAME_INFO.
* stack.c (frame_info):  If PRINT_EXTRA_FRAME_INFO defined, call it.

* mips-tdep.c (init_extra_frame_info): Only clobber the `frame'
(FP) value in the frame_info struct if it is zero (as from top of
execution stack).
(setup_arbitrary_frame):  Implement FRAME_SPECIFICATION_DYADIC.

* mips-xdep.c (fetch_inferior_registers):  ZERO_REGNUM always
comes back as zero.  So does FP_REGNUM, as a trigger for
init_extra_frame_info.

* tm-mips.h (INIT_FRAME_PC_FIRST):  Kludge, FIXME, defined to get
the program counter set before INIT_EXTRA_FRAME_INFO is run.
(INIT_FRAME_PC):  Defined to null.
(PRINT_EXTRA_FRAME_INFO): print frame pointer location via symtab.
(FRAME_SPECIFICATION_DYADIC):  Ask for two args in frame command.
Briefly explain MIPS stacks in GDB.

32 years ago* Makefile.in (make): Remove obsolete crud.
John Gilmore [Fri, 18 Sep 1992 09:05:54 +0000 (09:05 +0000)]
* Makefile.in (make):  Remove obsolete crud.
(sparc-opc.o):  Avoid Sun Make VPATH bug.

32 years ago* tm-rs6000.h (AIX_BUGGY_PTRACE_CALL): Zap, we think we fixed it.
John Gilmore [Fri, 18 Sep 1992 08:09:25 +0000 (08:09 +0000)]
* tm-rs6000.h (AIX_BUGGY_PTRACE_CALL): Zap, we think we fixed it.
Rich and I believe the "real problem" was that both single_step
and target_resume were issuing PT_CONTINUE calls.  This would
cause the second PT_CONTINUE to sometimes fail because the process
was already running.
* infptrace.c (child_resume):  Remove AIX_BUGGY_PTRACE_CALL kludge.

32 years agoPreliminary cleanup for splitting host/native/target.
John Gilmore [Fri, 18 Sep 1992 07:40:15 +0000 (07:40 +0000)]
Preliminary cleanup for splitting host/native/target.
* infptrace.c (child_resume):  Don't deal with NO_SINGLE_STEP
here; it is dealt with at a gdb-target-independent level.
* rs6000-tdep.c (single_step):  Don't call ptrace, we are a
high toned routine.  Fix return type to void.

32 years agoonly set PWD if it is already set (and likely wrong). from p3.
Mark Eichin [Fri, 18 Sep 1992 07:37:38 +0000 (07:37 +0000)]
only set PWD if it is already set (and likely wrong). from p3.

32 years ago* Makefile.in (update-depend, update-alldeps): Split out of
John Gilmore [Fri, 18 Sep 1992 07:06:38 +0000 (07:06 +0000)]
* Makefile.in (update-depend, update-alldeps): Split out of
setup-to-dist, for convenience in rebuilding the depend and
alldeps.mak files.
(assorted):  Update to catch straggler files when building gdb.tar.Z

32 years agoChanges to make the gdb.tar.Z rule work better.
John Gilmore [Fri, 18 Sep 1992 06:59:58 +0000 (06:59 +0000)]
Changes to make the gdb.tar.Z rule work better.
* Makefile.in (GDB_SUPPORT_DIRS):  Add opcodes.
(DEVO_SUPPORT):  Add configure.texi.
(bfd-ilrt.tar.Z):  Remove ancient rule.

32 years agofrom p3. Some shells leave PWD in the environment (old tcsh, ksh at least)
Mark Eichin [Fri, 18 Sep 1992 06:16:37 +0000 (06:16 +0000)]
from p3. Some shells leave PWD in the environment (old tcsh, ksh at least)
screws nested configures. "unset PWD" is the right answer, except that
Decstation sh doesn't *have* unset.
Note also that the send-pr build will fail if you have PWD in your
environment, because emacs also (foolishly) trusts it.

* configure: just set PWD=`pwd` at the top, since Ultrix sh
doesn't have unset and all success paths (and most error paths)
out set it anyway. (Note: should change all uses of ${PWD=`pwd`}
to just ${PWD} to avoid confusion.)

32 years agoImprove release doc slightly.
John Gilmore [Thu, 17 Sep 1992 11:22:54 +0000 (11:22 +0000)]
Improve release doc slightly.

32 years ago* copying.awk, copying.c (show_copying_command,
John Gilmore [Thu, 17 Sep 1992 11:02:44 +0000 (11:02 +0000)]
* copying.awk, copying.c (show_copying_command,
show_warranty_command):  Rename from copying_info, warranty_info,
to match command function conventions.

32 years ago* utils.c (prompt_for_continue): Reinitialize more-counts
John Gilmore [Thu, 17 Sep 1992 10:58:53 +0000 (10:58 +0000)]
* utils.c (prompt_for_continue):  Reinitialize more-counts
before printing anything, and again afterward.  Fix comments.
(vfprintf_filtered):  Eliminate static buffer; use auto buffer,
or alloca() if needed.

* rs6000-xdep.c:  Use correct conditional (IBM6000_TARGET) to
detect native versus cross-host.

32 years ago* m68k-tdep.c (sun3_saved_pc_after_call): Only do trap check for
Stu Grossman [Thu, 17 Sep 1992 05:01:51 +0000 (05:01 +0000)]
* m68k-tdep.c (sun3_saved_pc_after_call):  Only do trap check for
Suns.  Rename to m68k_saved_pc_after_call.
* tm-68k-noun.h, tm-sun3.h (SAVED_PC_AFTER_CALL):  Use
m68k_saved... instead of sun3_saved...

32 years agoMake as.new depend on LIBS not the never set LIBDEPS, so that we get an
Jim Wilson [Wed, 16 Sep 1992 00:26:24 +0000 (00:26 +0000)]
Make as.new depend on LIBS not the never set LIBDEPS, so that we get an
opcodes directory dependence.

32 years agoUpdate to SPARC V9 draft 1.1. Put brfcx before brfc to avoid superfluous
Jim Wilson [Wed, 16 Sep 1992 00:23:52 +0000 (00:23 +0000)]
Update to SPARC V9 draft 1.1.  Put brfcx before brfc to avoid superfluous
symbol table entries.  Document unimplemented synthetic instructions.

32 years agoTue Sep 15 15:35:38 1992 Ian Lance Taylor (ian@cygnus.com)
Ian Lance Taylor [Tue, 15 Sep 1992 23:30:30 +0000 (23:30 +0000)]
Tue Sep 15 15:35:38 1992  Ian Lance Taylor  (ian@cygnus.com)

* Makefile.in (install): if $(tooldir) exists, install ld in
$(tooldir)/bin.

32 years agoChanges to make an installed cross-compiler find everything correctly.
Ian Lance Taylor [Tue, 15 Sep 1992 23:27:10 +0000 (23:27 +0000)]
Changes to make an installed cross-compiler find everything correctly.

Tue Sep 15 15:32:02 1992  Ian Lance Taylor  (ian@cygnus.com)

* Makefile.in (install): if $(tooldir) exists, install as in
$(tooldir)/bin.

32 years agoChanges to make an installed cross-compiler find everything correctly.
Ian Lance Taylor [Tue, 15 Sep 1992 23:26:41 +0000 (23:26 +0000)]
Changes to make an installed cross-compiler find everything correctly.

Tue Sep 15 15:26:38 1992  Ian Lance Taylor  (ian@cygnus.com)

* Makefile.in (install): if $(tooldir) exists, install nm and
strip in $(tooldir)/bin.

32 years agoChanges to make an installed cross-compiler find everything correctly.
Ian Lance Taylor [Tue, 15 Sep 1992 23:26:16 +0000 (23:26 +0000)]
Changes to make an installed cross-compiler find everything correctly.

Tue Sep 15 16:00:54 1992  Ian Lance Taylor  (ian@cygnus.com)

* configure: always set $(tooldir) to $(libdir)/$(target_alias),
even for a native compilation.

32 years agoIn brx macro, check for XCC bit set in xcc insns, and not set in icc insns.
Jim Wilson [Tue, 15 Sep 1992 19:55:26 +0000 (19:55 +0000)]
In brx macro, check for XCC bit set in xcc insns, and not set in icc insns.
In cond macro, do brx before br, to avoid getting extra symbol table entries
due to failed pattern matches.

32 years agoKeep h8.m4 now.
Roland Pesch [Tue, 15 Sep 1992 17:51:14 +0000 (17:51 +0000)]
Keep h8.m4 now.

32 years agoAdd missing utils.c patches.
Stu Grossman [Tue, 15 Sep 1992 09:19:53 +0000 (09:19 +0000)]
Add missing utils.c patches.

32 years agoDon't use prototypes for vfprintf_filtered(). Someday we'll get prototypes of
Stu Grossman [Tue, 15 Sep 1992 08:22:29 +0000 (08:22 +0000)]
Don't use prototypes for vfprintf_filtered().  Someday we'll get prototypes of
varargs routines working...

32 years agoAll new patches for Energize!
Stu Grossman [Tue, 15 Sep 1992 07:28:05 +0000 (07:28 +0000)]
All new patches for Energize!

32 years ago* energize.c (pty_to_kernel): Must check for EAGAIN as
Stu Grossman [Tue, 15 Sep 1992 07:12:52 +0000 (07:12 +0000)]
* energize.c (pty_to_kernel):  Must check for EAGAIN as
termination condition.
(full_filename):  Don't prepend dirname if filename begins with /.
(breakpoint_notify, kernel_dispatch):  Deal with DynamicLoad
messages properly!!
(getpty (NCR)):  Lotsa bug fixes...
(kernel_dispatch (case StopRType)):  Use kill() instead of
killpg() so we can deal with losing systems...
(wait_for_events):  Work around NCR select() lossage.  It doesn't
zero out readfds when select() returns zero.
(energize_initialize):  Use I_SETSIG to get SIGIO interrupts.  Use
getcwd() instead of getwd() (NCR getwd() seg faults).
(energize_wait):  Use sigaction() for NCR.  Their version of signal()
is unreliable by default.

32 years ago* procfs.c (proc_wait): Call print_sys_errmsg() if we get an
Stu Grossman [Tue, 15 Sep 1992 06:42:19 +0000 (06:42 +0000)]
* procfs.c (proc_wait):  Call print_sys_errmsg() if we get an
error from PIOCSWSTOP.

32 years ago* breakpoint.c, core.c, exec.c, language.c, main.c, printcmd.c,
Stu Grossman [Tue, 15 Sep 1992 06:27:18 +0000 (06:27 +0000)]
* breakpoint.c, core.c, exec.c, language.c, main.c, printcmd.c,
symfile.c, target.c, valprint.c:  Use _filtered form of *printf.
defs.h, utils.c:  Make vfprintf_filtered global.

32 years agoOops! Retract last change. Didn't mean to pollute things with energize just
Stu Grossman [Tue, 15 Sep 1992 05:11:52 +0000 (05:11 +0000)]
Oops!  Retract last change.  Didn't mean to pollute things with energize just
yet!

32 years ago* utils.c (query): Remove test for energize from top. Now
Stu Grossman [Tue, 15 Sep 1992 05:04:08 +0000 (05:04 +0000)]
* utils.c (query):  Remove test for energize from top.  Now
handled in input_from_terminal_p().

32 years ago* energize.c (send_location): New routine to consolidate all
Stu Grossman [Tue, 15 Sep 1992 04:50:46 +0000 (04:50 +0000)]
* energize.c (send_location):  New routine to consolidate all
places where we must notify kernel of where the given pc is.
(cplus_demangle):  Put single-quotes around demangled names.
(energize_call_command):  Call send_location() after doing up,
down, and frame commands.

32 years agoFix typos and thinkos reported by Sun Ming (and one reported by John).
Roland Pesch [Mon, 14 Sep 1992 18:53:57 +0000 (18:53 +0000)]
Fix typos and thinkos reported by Sun Ming (and one reported by John).

32 years agoSun Sep 13 20:29:39 1992 Ian Lance Taylor (ian@cygnus.com)
Ian Lance Taylor [Mon, 14 Sep 1992 05:25:56 +0000 (05:25 +0000)]
Sun Sep 13 20:29:39 1992  Ian Lance Taylor  (ian@cygnus.com)

* obj-coffbfd.h: define WARN_SIGNED_OVERFLOW_WORD.
obj-coffbfd.c (fixup_segment): check for signed .word overflow if
WARN_SIGNED_OVERFLOW_WORD.

32 years agoSun Sep 13 20:30:10 1992 Ian Lance Taylor (ian@cygnus.com)
Ian Lance Taylor [Mon, 14 Sep 1992 05:25:24 +0000 (05:25 +0000)]
Sun Sep 13 20:30:10 1992  Ian Lance Taylor  (ian@cygnus.com)

* Added WARN_SIGNED_OVERFLOW_WORD define to give an error if any
.word is < -32768 or > 32767.  The -J flag causes the error to be
ignored.  This is to catch over-sized switches generated by gcc on
systems which don't support the broken .word hack.
as.c (main): permit -J if WARN_SIGNED_OVERFLOW_WORD.
write.c (fixup_segment): check for signed .word overflow if
WARN_SIGNED_OVERFLOW_WORD.

32 years agoSun Sep 13 20:29:39 1992 Ian Lance Taylor (ian@cygnus.com)
Ian Lance Taylor [Mon, 14 Sep 1992 03:34:44 +0000 (03:34 +0000)]
Sun Sep 13 20:29:39 1992  Ian Lance Taylor  (ian@cygnus.com)

* obj-coffbfd.c (fixup_segment): fixed missing parens in
expression checking for byte or word overflow.

32 years agoSun Sep 13 20:30:10 1992 Ian Lance Taylor (ian@cygnus.com)
Ian Lance Taylor [Mon, 14 Sep 1992 03:34:19 +0000 (03:34 +0000)]
Sun Sep 13 20:30:10 1992  Ian Lance Taylor  (ian@cygnus.com)

* write.c (fixup_segment): fixed missing parens in expression
checking for byte or word overflow.

32 years ago...plus c'est la meme chose.
Roland Pesch [Sat, 12 Sep 1992 05:28:20 +0000 (05:28 +0000)]
...plus c'est la meme chose.

32 years ago * doc/gdb.texinfo, doc/gdbinv-m.m4.in, doc/gdbinv-s.m4.in: first
Roland Pesch [Sat, 12 Sep 1992 05:27:40 +0000 (05:27 +0000)]
* doc/gdb.texinfo, doc/gdbinv-m.m4.in, doc/gdbinv-s.m4.in: first
pass at doc for two new remote targets---29K using UDI, and Tandem
ST2000 using STDBUG.

* doc/all.m4, doc/none.m4: new m4 switch, _ST2000__

* doc/all.m4: turn on H8/300 stuff for generic manual

32 years ago*** empty log message ***
Roland Pesch [Sat, 12 Sep 1992 01:29:46 +0000 (01:29 +0000)]
*** empty log message ***

32 years agodoc/gdbinv-s.m4.in: remove text on special procedures to continue after
Roland Pesch [Sat, 12 Sep 1992 01:14:57 +0000 (01:14 +0000)]
doc/gdbinv-s.m4.in: remove text on special procedures to continue after
                    explicit call to breakpoint() in serial stubs.

32 years agoFri Sep 11 15:37:06 1992 Ian Lance Taylor (ian@cygnus.com)
Ian Lance Taylor [Fri, 11 Sep 1992 22:43:53 +0000 (22:43 +0000)]
Fri Sep 11 15:37:06 1992  Ian Lance Taylor  (ian@cygnus.com)

* coffcode.h (coff_renumber_symbols): set the last renumbered
symbol pointer to NULL, as expected by coff_write_linenumbers.
(coff_write_relocs): apparently a non-zero addend is OK for reloc
type R_IHCONST used on the 29k.

32 years agoWhoops, forgot this one
Steve Chamberlain [Fri, 11 Sep 1992 22:02:52 +0000 (22:02 +0000)]
Whoops, forgot this one

32 years agoSanitize slip
Steve Chamberlain [Fri, 11 Sep 1992 20:43:33 +0000 (20:43 +0000)]
Sanitize slip

32 years agoAdd i386coff.sh
Steve Chamberlain [Fri, 11 Sep 1992 17:47:22 +0000 (17:47 +0000)]
Add i386coff.sh

32 years agoNew files
Steve Chamberlain [Fri, 11 Sep 1992 17:47:01 +0000 (17:47 +0000)]
New files

32 years agoSupport for i386-coff
Steve Chamberlain [Fri, 11 Sep 1992 17:25:43 +0000 (17:25 +0000)]
Support for i386-coff

32 years agoSupport for 386 coff
Steve Chamberlain [Fri, 11 Sep 1992 17:23:31 +0000 (17:23 +0000)]
Support for 386 coff