create a 0-length array type, and set BOUND_CANNOT_BE_DETERMINED.
* valops.c (value_cast): If a cast like (TYPE[])VALUE (i.e. array
of unknown length) use sizeof(VALUE)/sizeof(TYPE) as the length.
* c-typeprint.c (c_type_print_varspec_suffix): If array length
is 0, print it, but not if upper_bound is BOUND_CANNOT_BE_DETERMINED.
enable/disable_breakpoint_hook with modify_breakpoint_hook.
* gdbtk.c: General cleanups, get rid of unused variables. Redo
handling of stdout/stderr to just return output as the result of
the tcl command that caused the output. Cleanup -Wall stuff.
* (breakpoint_notify): Now returns just action and breakpoint
number.
* (gdb_get_breakpoint_list): New routine. Does the obvious.
* (gdb_get_breakpoint_info): Mostly derived from the old
breakpoint_notify, but returns lots more info.
* (dsprintf_append_element): Helper routine, works like printf,
but appends a tcl element onto the specified DString. Good for
building up lists as return values.
* (gdbtk_enable/disable_breakpoint): Go away. Replaced with
gdbtk_modify_breakpoint.
* (*many routines*): Use new result protocol.
* (call_wrapper): Make sure that recursive calls don't trash results.
* gdbtk.tcl: New windows, autocmd, and breakpoints.
* (gdbtk_tcl_fputs): Don't use $current_output_win redirection
anymore. It's not needed (in fact, this routine may not be needed
anymore).
* (gdbtk_tcl_breakpoint): Change to reflect new breakpoint
notification protocol.
* (gdbtk_tcl_busy gdbtk_tcl_idle): Straighten out buttons, remove
catches.
* (interactive_cmd): Use this wrapper around button invocations
of many commands. This will catch errors and put the results into
the command window. It also updates all the other windows.
* Also, change reliefs of most things to sunken. This actually
looks better.
* (create_file_win): Fix margin binding to allow breakpoints to
work again.
* (create_asm_win): Use return value of gdb_disassemble instead
of implicit I/O to the command window.
* (create_command_window): Use new result protocol to get output
from commands.
simplify code, declare things, format to standards.
(inferior.h, value.h, command.h, remote-utils.h): Include.
(e7000_login): Rename to e7000_login_command.
(e7000_ftp): Rename to e7000_ftp_command.
(e7000_drain): Rename to e7000_drain_command.
arg, it is a value not a symbol.
gcc -Wall lint:
* eval.c: Move declaration of evaluate_subexp_with_coercion from here..
* expression.h: ..to here.
* expression.h: Include value.h.
* ch-lang.c (evaluate_subexp_chill): Add default case in switch.
* c-lang.c (c_language_defn, cplus_language_defn, asm_langauge_defn),
f-lang.c (f_language_defn), language.c (unknown_language_defn,
auto_language_defn, local_language_defn), m2-lang.c (m2_language_defn):
Set evaluate_exp to evaluate_subexp_standard.
* ch-lang.c (evaluate_subexp_chill): New function. Chill-specific
support for MULTI_SUBSCRIPT.
(chill_language_defn): Set evaluate_exp to evaluate_subexp_chill.
* eval.c (enum noside): Move from here ....
* expression.h (enum noside): ... to here.
(evaluate_subexp_standard): New prototype.
* eval.c (evaluate_subexp): Renamed to evaluate_subexp_standard.
Removed lo-longer-needed test for chill_varying_type.
(evaluate_subexp): New. Calls exp->language_defn->evaluate_exp.
since that functionality has been re-written.
* coffread.c: Remove commented-out add_param_to_type support.
* mdebugread.c (parse_symbol): Likewise.
* stabsread.c (define_symbol): Likewise.
parameter type info, set that from parameter symbols.
* c-typeprint.c (c_type_print_varspec_suffix): For TYPE_CODE_FUNC,
print parameter types, if available.
* ch-typeprint.c (chill_type_print_base): Likewise.
* gdbtypes.h (struct type): Remove function type field.
(TYPE_FUNCTION_TYPE): Remove macro. We can't as simply re-use
function types now that we're also storing parameter types.
And the payoff is much less.
* gdbtypes.c (make_function_type): Don't use/set TYPE_FUNCTION_TYPE.
(recursive_dump_type): Don't print TYPE_FUNCTION_TYPE.
* dwarfread.c (read_subroutine_type): Don't set TYPE_FUNCTION_TYPE.
* partial-stab.h: Call it.
* stabsread.c (define_symbol) [STATIC_TRANSFORM_NAME]: Call
STATIC_TRANSFORM_NAME to get the name and use minimal symbols to
get the address.
* sparc-tdep.c (solaris_static_transform_name): New function.
text and data symbols.
(som_symfile_offsets): If objfile is a shared library, then get
text and data offsets from the shared library structures.
* somsolib.c (som_solib_add): Copy the bfd pointer from the
objfile rather than reopening the file again.
(som_solib_section_offsets): New function.
* somsolib.h (som_solib_section_offsets): Declare.
top.c : Get rid of dis_asm_read_memory_hook. We can now call the
disassemblers directly and have no need for this hook anymore.
* defs.h, printcmd.c: Make print_insn be static.
* ser-go32.c (dos_comisr): Make this 8 bit clean.
* (dos_open dos_close): Allow multiple opens to the same device.
Use a ref count to prevent unwanted deallocations.
* sparcl-tdep.c: Put #ifdefs around all socket stuff to make GO32
happy.
* (sparclite_ops): Switch to download_stratum.
* target.h (enum strata): Move download_stratum before
process_stratum so that executable targets get pushed on top of
download targets.
* remote-arc.c: new file for arc-specific protocol through
parallel line.
* ser-go32-para.c: new file for go32 parallel port communication.
* .Sanitize: sanitize arc specific files out.
* config/arc/arc.mt: new target makefile for arc processor.
* config/arc/tm-arc.h: new target header for arc processor.
* config/arc/go32.mh: new go32 host makefile for arc processor.
* config/arc/xm-go32.h: new go32 host header for arc processor.
OP_UNDETERMINED_ARGLIST (no backtracking, more general).
* f-valprint.c (f_val_print): Print TYPE_CODE_STRING using
LA_PRINT_STRING, and not val_print_string (which reads from inferior).
* ch-lang.c (chill_is_varying_struct), ch-lang.h: Remve function
duplicate function made redundant by chill_varying_type.
Re-write of f77 string and complex number support:
* language.h (struct language_defn): New fields string_lower_bound
and string_char_type.
* c-lang.c (c_language_defn, cplus_language_defn, asm_language_defn),
language.c (unknown_language_defn, auto_language_defn,
local_language_defn), m2-lang.c (m2_language_defn), f-lang.c
(f_language_defn), ch-lang.c (chill_language_defn): Set new fields.
* gdbtypes.c (create_string_type): Use new string_char_type field.
* valops.c (value_string): Use new string_lower_bound field.
* defs.h (TARGET_COMPLEX_BIT, TARGET_DOUBLE_COMPLEX_BIT): Removed.
* f-lang.c (f_create_fundamental_type, _initialize_f_language),
m2-lang.c (m2_create_fundamental_type),
gdbtypes.c (_initialize_gdbtypes): Set TYPE_TARGET_TYPE of complex
types. Set their TYPE_CODEs to TYPE_CODE_COMPLEX.
* mdebugread.c (mdebug_type_complex, mdebug_type_double_complex):
Removed. Use builtin_type_complex and builtin_type_double_complex.
* gdbtypes.h (enum type_code): Removed TYPE_CODE_LITERAL_STRING
and TYPE_CODE_LITERAL_COMPLEX.
* c-typeprint.c, f-typeprint.c, f-valprint.c, eval.c: Removed uses of
TYPE_CODE_LITERAL_STRING and TYPE_CODE_LITERAL_COMPLEX.
* gdbtypes.c, gdbtypes.h (f77_create_literal_complex_type,
f77_create_literal_string_type): Removed.
* value.h (VALUE_LITERAL_DATA, VALUE_SUBSTRING_MEMADDR,
VALUE_SUBSTRING_MYADDR): Removed.
* expression.h (enum exp_opcode): Rename OP_F77_LITERAL_COMPLEX to
OP_COMPLEX.
* parse.c: Update accordingly.
* f-valprint.c (f77_print_cmplx): Removed.
(f_val_print case TYPE_CODE_COMPLEX): Re-write to use print_floating.
* f-exp.y (STRING_LITERAL): Use OP_STRING instead of OP_ARRAY.
* eval.c (evaluate_subexp): For case OP_ARRAY, don't call
f77_value_literal_string.
* valops.c, value.h (f77_value_literal_string, f77_value_substring,
f77_assign_from_literal_string, f77_assign_from_literal_complex):
Removed.
(value_assign): No longer need to handle literal types.
* valops.c (f77_value_literal_complex), value.h: Re-written and
renamed to value_literal_complex. Last arg is now a (complex) type.
* valops.c (f77_cast_into_complex): Re-written and renamed to
cast_into_complex.
* eval.c (evaluate_subexp): Update accordingly.
start_element, left_element, right_element, slice_size,
lower_element, upper_element, first_element): Removed.
(value_string_slice, value_array_slice): Replaced by ...
(slice): New non-terminal, with working slice support.
(primitive_value_lparen, rparen): New non-terminals.
(maybe_tuple_elements): New non-terminal, to allow empty tuples.
(idtokentab): Added "up".
* value.h (COERCE_VARYING_ARRAY): New macro.
* valarith.c (value_subscript): Use it.
* valops.c (value_cast): Likewise. Also, do nothing if already
correct type, and allow converting from/to range to/from scalar.
* valops.c, value.h (varying_to_slice, value_slice): New functions.
* eval.c (OP_ARRAY): Add cast for array element.
* expression.h (TERNOP_SLICE, TERNOP_SLICE_COUNT): New exp_opcodes.
* valops.c (chill_varying_type): Moved function frp, here ...
* gdbtypes.c (chill_varying_type), gdbtypes.h: ... to here.
* parse.c (length_of_subexp, prefixify_subexp): Add support
for TERNOP_SLICE, TERNOP_SLICE_COUNT.
* expprint.c (print_subexp, dump_expression): Likewise.
* eval.c (evaluate_subexp): Likewise.
* eval.c (evaluate_subexp case MULTI_SUBSCRIPT): Don't call
value_x_binop on a Chill varying string.
remove them.
* gdb.base/nodebug.exp: Accept "function" in addition to "text
variable and "variable" in addition to "data variable".
* gdb.base/nodebug.exp: xfail datalocal and bsslocal tests for AIX.
* ser-go32.c (dos_close): Don't crash if scb null.
(dos_sendbreak): New function.
(dos_ops): Point to dos_sendbreak.
(dos_info): Calculate COM number correctly.
func_symbol_type and var_symbol_type as type of functions and
variables which don't have any stabs associated with them.
Reindent most of function.
(_initialize_xcoffread): Initialize *_symbol_type.
* xcoffread.c (read_xcoff_symtab): Reindent most of function.
Put C_HIDEXT symbols in the minimal symbols, rather than ignoring
them (this part commented out as I didn't quite get it to work).
(cs_to_section, find_targ_sec): New functions, to support above code.
* xcoffread.c (RECORD_MINIMAL_SYMBOL): Only skip '.' if it is
actually present.
2.
* gdb.base/list1.c (long_line): Add additional statement at start
of function.
* gdb.base/list.exp (test_forward_search): Update line number to
reflect new statement.
no hope of working as is for a long time (in particular, shebs' 27
Jan 95 change confuses the issue further--target_read_memory and
xfer_core_file do *not* do the same thing in this context).
Revise comment.
(exec_file_command, xfer_core_file): Comment out.
(arm_print_insn): Remove, now in libopcodes.
(skip_prologue): Comment out most of body.
(arm_frame_find_saved_regs): Move here from tm-arm.h.
(_initialize_arm_tdep): Set tm_print_insn.
* config/arm/tm-arm.h: Remove old refs to first_object_file_end.
(XFER_CORE_FILE): Remove.
(FRAME_FIND_SAVED_REGS): Call arm_frame_find_saved_regs.
problems with those changes were (a) the file didn't compile, (b)
they changed memset to bzero--memset is correct, (c) they took out
code to deal with boards lacking floating point, (d) who knows
what I didn't discover in a quick read.
* language.c (unknown_language_defn, auto_language_defn,
local_language_defn), c-lang.c (c_language_defn, cplus_language_defn,
asm_language_defn): Set c_style_arrays to true.
* m2-lang.c (m2_language_defn), ch-lang.c (chill_language_defn),
f-lang.c (f_language_defn): Set c_style_arrays to false.
* valops.c (value_string): If c_style_array is not set,
allocate string in gdb (not inferior) using allocate_value.
* value.h (COERCE_ARRAY), valops.c (value_addr, value_arg_coerce):
Only call value_coerce_array if current_language->c_style_arrays.
* values.c: Add #include "language.h". (Needed for COERCE_ARRAY.)
* valops.c (chill_varying_type): New predicate.
* valops.c (value_cast): Support assigning a fixed string or array
to a variable string/array structure.
* valarith.c (value_subscripted_rvalue): Extra parameter lowerbound.
Check index>=lowerbound, and then add lowerbound to index here,
instead of in caller. Generalize to arbitrary lval_types.
(value_subscript): Use enhanced value_subscripted_rvalue if
c_style_arrays is false (and index is in range).
restart GDB for h8300 after a continue.
(return): Add pattern to match a29k-udi behavior.
* gdb.base/nodebug.exp: Remove obsolete comment.
* gdb.base/regs.exp: New file, testing of register displays.
* gdb.base/return.exp: Fix typo in reference to $objdir.
* gdb.base/whatis.exp: Always start with a fresh GDB.
code following the catch silently ignored some errors; without the
catch they should be thrown like any other tcl error. Also, the
catch used "" instead of {} which meant all the callers who wanted
to include one of the characters "[]$ had to quote it an extra time.
* gdb.base/{callfuncs.exp,commands.exp,gdbvars.exp,printcmds.exp,
ptype.exp,signals.exp,watchpoint.exp}, gdb.c++/{cplusfuncs.exp,
demangle.exp}, gdb.chill/chexp.exp, gdb.fortran/exprs.exp:
Remove extra quoting.
* ch-exp.y (match_bitstring_literal): Fix for proper endianness.
* expprint.c (print_subexp): Don't call error on OP_BITSTRING,
just print B'<unimlemented>'.
* gdbtypes.c (create_set_type): Fix bug in length calculation.
* valops.c, value.h (value_bitstring): New function.
* eval.c (evaluate_subexp): Implement support for OP_BITSTRING.
* ch-typeprint.c (chill_type_print_base): For TYPE_CODE_FUNC,
check that return type is non-void, and print in proper Chill syntax.
configuration didn't use the ".if alpha" code in weird.def.
* gdb.stabs/weird.def (v_comb): Use type attribute to specify 32
bit pointer, so the 32 bit version works even on the alpha.
* gdb.stabs/{alpha.mt,ecoff.mt,aout.sed,hppa.sed,xcoff.sed,weird.def}:
Remove the stuff which allowed us to have separate versions for
alpha and everything else.
regexp which "info func operator\[\](" is supposed to match. Fix
string to send to GDB. Note that GDB still doesn't work right. Add a
new test for "info func .perator\[\](" which GDB does handle correctly.
labeled structre tuples.
* ch-exp.y (tuple_element, named_record_element, tuple_elements):
New non-terminals, to handle labeled structure tuples.
(tuple): Re-define using tuple_elements.
* eval.c (evaluate_labeled_field_init): New function, to handle
initialization of structure fields, possibly using OP_LABELED.
(evaluate_subexp): Use it.
* expprint.c (print_subexp case): For OP_ARRAY, use Chill syntax
for Chill. Handled OP_LABELED.
* parse.c (length_of_subexp, prefixify_subexp): Handle OP_LABELED.
* eval.c (evaluate_subexp): Handle Chill Powerset tuples.
* valarith.c (value_bit_index): Just treat bitstring as represented
by an array of bytes. Alignment is handled by compiler.
gdb.c++/inherit.exp, gdb.c++/templates.exp, gdb.c++/virtfunc.exp,
gdb.fortran/exprs.exp, gdb.fortran/types.exp, gdb.chill/chexp.exp,
gdb.base/printcmds.exp: Remove passcount, failcount, etc., stuff;
it makes the tests harder to understand and confuses test-o-matic.
The preferred style is that each test provides a PASS or a FAIL,
and has a unique message (e.g. "continue #54" not just "continue")
which is the same for the pass and the fail.
* gdb.fortran/exprs.exp, gdb.fortran/types.exp,
gdb.chill/chexp.exp: Move test_print_accept and test_print_reject
to lib/gdb.exp.
* gdb.base/printcmds.exp: Use test_print_accept. Remove
prt_accept which was basically the same thing. Likewise for
test_print_reject and prt_reject.
* lib/gdb.exp (test_print_reject): Add some more error message
patterns to match from the former printcmds.exp (prt_reject).
* gdb.c++/classes.exp, gdb.base/scope.exp: Remove spurious xfails.
One defect of the passcount stuff is that some of it failed to
report XPASS where appropriate.
* gdb.c++/cplusfuncs.exp (print_addr_of): No longer accept extra
stuff before and after arg in expected pattern.
(test_paddr_operator_functions): Re-do test without print_addr_of;
this is the only test which seems to want extra stuff there.