OSDN Git Service

pf3gnuchains/pf3gnuchains4x.git
23 years ago* i386-tdep.c (i386_extract_return_value): Undo 2001-07-11 changes
Mark Kettenis [Thu, 12 Jul 2001 18:50:01 +0000 (18:50 +0000)]
* i386-tdep.c (i386_extract_return_value): Undo 2001-07-11 changes
to comment.
(i386_store_return_value): Improve comments about storing
floating-point return values.

23 years ago* config/arm/xm-linux.h, config/i386/xm-linux.h,
Mark Kettenis [Thu, 12 Jul 2001 18:22:37 +0000 (18:22 +0000)]
* config/arm/xm-linux.h, config/i386/xm-linux.h,
config/m68k/xm-linux.h, config/powerpc/xm-linux.h,
config/sparc/xm-linux.h, config/ia64/xm-linux.h (HAVE_TERMIOS):
Removed.  Taken care of by autoconf and terminal.h.

23 years ago * mips-tdep.c (mips_type_needs_double_align): New function.
Daniel Jacobowitz [Thu, 12 Jul 2001 17:34:33 +0000 (17:34 +0000)]
    * mips-tdep.c (mips_type_needs_double_align): New function.
        (mips_push_arguments): Align o32 structs to even argument
        registers if necessary.

23 years agofix mips/linux 8.3 errors
Daniel Jacobowitz [Thu, 12 Jul 2001 17:23:32 +0000 (17:23 +0000)]
fix mips/linux 8.3 errors

23 years ago* config/powerpc/xm-linux.h: Fix corrupted file.
Mark Kettenis [Thu, 12 Jul 2001 14:02:23 +0000 (14:02 +0000)]
* config/powerpc/xm-linux.h: Fix corrupted file.

23 years ago* configure.in: Add check for sigsetjmp.
Mark Kettenis [Thu, 12 Jul 2001 13:48:46 +0000 (13:48 +0000)]
* configure.in: Add check for sigsetjmp.
* config/m68k/xm-linux.h (HAVE_SIGSETJMP): Removed.
* config/xm-sysv4.h (HAVE_SIGSETJMP): Removed.
* configure, config.in: Regenerated.

23 years ago* gdb.base/so-impl-ld.exp: Remove stray space that prevented
Mark Kettenis [Thu, 12 Jul 2001 13:07:32 +0000 (13:07 +0000)]
* gdb.base/so-impl-ld.exp: Remove stray space that prevented
running this test on Linux.

23 years ago* mi-main.c (mi_execute_command): Flush output after ``(gdb)''
Andrew Cagney [Thu, 12 Jul 2001 05:54:48 +0000 (05:54 +0000)]
* mi-main.c (mi_execute_command): Flush output after ``(gdb)''
prompt.  Bug reported by David Whedon.
(mi_execute_async_cli_command): Ditto.
(mi_exec_async_cli_cmd_continuation): Ditto.
(mi_command_loop): Ditto.

23 years ago2001-07-11 H.J. Lu <hjl@gnu.org>
H.J. Lu [Thu, 12 Jul 2001 04:24:44 +0000 (04:24 +0000)]
2001-07-11  H.J. Lu  <hjl@gnu.org>

* elf64-alpha.c (elf64_alpha_check_relocs): Set the ALLOC|LOAD
flags when creating the reloc section if the ALLOC flag in the
source section is set.

23 years ago* some support for funny-endian 16/32-bit insn sets
Frank Ch. Eigler [Thu, 12 Jul 2001 02:32:25 +0000 (02:32 +0000)]
* some support for funny-endian 16/32-bit insn sets

[cgen/ChangeLog]
2001-07-11  Frank Ch. Eigler  <fche@redhat.com>

        * desc-cpu.scm (-gen-mach-table-defns): Emit fourth field: the
        mach->cpu insn-chunk-bitsize.
        (-gen-cpu-open): In @arch@_cgen_rebuild_tables, process above new
        field toward CGEN_CPU_TABLE->insn_chunk_bitsize.
        * mach.scm (<cpu>): New field insn-chunk-bitsize.
        (-cpu-parse, -cpu-read): Parse/initialize it.
        * doc/rtl.texi (define-cpu): Document it.

[opcodes/ChangeLog]
2001-07-11  Frank Ch. Eigler  <fche@redhat.com>

        * cgen-dis.in (print_insn): Use cgen_get_insn_value instead of
        bfd_get_bits.
        * cgen-opc.c (cgen_get_insn_value, cgen_put_insn_value): Respect
        non-zero CGEN_CPU_DESC->insn_chunk_bitsize.

[include/opcode/ChangeLog]
2001-07-11  Frank Ch. Eigler  <fche@redhat.com>

        * cgen.h (CGEN_MACH): Add insn_chunk_bitsize field.
        (cgen_cpu_desc): Ditto.

23 years ago2001-07-11 Elena Zannoni <ezannoni@redhat.com>
Elena Zannoni [Thu, 12 Jul 2001 02:29:33 +0000 (02:29 +0000)]
2001-07-11  Elena Zannoni  <ezannoni@redhat.com>

* config/sh/tm-sh.h (struct gdbarch_tdep): Add
FLOAT_ARGLAST_REGNUM, RETURN_REGNUM, ARG0_REGNUM, ARGLAST_REGNUM,
PR_REGNUM fields.

* sh-tdep.c (sh_skip_prologue_hard_way): Rename from
skip_prologue_hard_way().
(skip_prologue_hard_way): Make it a function pointer.
        (sh_print_register): Use function pointer do_pseudo_register,
instead of sh_do_pseudo_register.
        (sh_do_registers_info): Ditto.
        (sh_gdbarch_init): Initialize do_pseudo_register.
(IS_MOV_TO_R14): Rename from IS_MOV_R14.
(skip_prologue_hard_way): Update.
(sh_gdbarch_init): Move setting of frame_chain,
get_saved_register, init_extra_frame_info, push_arguments,
extract_return_value, pop_frame, store_struct_return,
extract_struct_value_address, use_struct_convention,
init_extra_frame_info to before target specific settings.
(sh_gdbarch_init): Initialize new tdep fields
FLOAT_ARGLAST_REGNUM, RETURN_REGNUM, ARG0_REGNUM, ARGLAST_REGNUM,
PR_REGNUM.
(sh_push_arguments): Use new gdbarch_tdep fields
{ARG0,ARGLAST}_REGNUM.
(sh_saved_pc_after_call, sh_nofp_frame_init_saved_regs,
sh_init_extra_frame_info, sh_push_return_address,
sh_generic_show_regs, sh3_show_regs, sh3e_show_regs,
sh3_dsp_show_regs, sh4_show_regs, sh_dsp_show_regs): Update to use
gdbarch_tdep->PR_REGNUM.
(dr_reg_base_num, fv_reg_base_num): Move to earlier in the file.

23 years agosje patch for 32-bit IA-64 ELF relocs for HPUX
Jim Wilson [Thu, 12 Jul 2001 01:24:14 +0000 (01:24 +0000)]
sje patch for 32-bit IA-64 ELF relocs for HPUX
* reloc.c (bfd_reloc_code_type): Add IA64 relocs
BFD_RELOC_IA64_LTOFF_FPTR32MSB and BFD_RELOC_IA64_LTOFF_FPTR32LSB
* bfd-in2.h: Regenerate.
* libbfd.h: Regenerate.
* elfxx-ia64.c (ia64_howto_table): Define how to handle
new relocations.
(elfNN_ia64_reloc_type_lookup): Handle new relocations.
(elfNN_ia64_check_relocs): Likewise.
(elfNN_ia64_install_value): Likewise.
(elfNN_ia64_relocate_section): Likewise.

23 years ago*** empty log message ***
gdbadmin [Thu, 12 Jul 2001 00:00:03 +0000 (00:00 +0000)]
*** empty log message ***

23 years agoDon't duplicate tm-linux.h
Daniel Jacobowitz [Wed, 11 Jul 2001 21:49:41 +0000 (21:49 +0000)]
Don't duplicate tm-linux.h

23 years agoAdd myself to MAINTAINERS for mips/linux
Daniel Jacobowitz [Wed, 11 Jul 2001 20:44:51 +0000 (20:44 +0000)]
Add myself to MAINTAINERS for mips/linux

23 years ago * elf32-sparc.c (_bfd_sparc_elf_howto_table): Remove support for
Jakub Jelinek [Wed, 11 Jul 2001 19:40:32 +0000 (19:40 +0000)]
* elf32-sparc.c (_bfd_sparc_elf_howto_table): Remove support for
R_SPARC_UA64.
(elf32_sparc_check_relocs): Likewise.
Only create .rela section for alloced sections in shared libraries.
(elf32_sparc_relocate_section): Likewise.
Remove redundant check.
Optimize unaligned reloc usage.
* elf64-sparc.c (sparc64_elf_relocate_section): Remove redundant
check.

* testsuite/gas/sparc/unalign.s: Remove .uaxword test.
* testsuite/gas/sparc/unalign.d: Adjust accordingly.

23 years agominor F_SETOWN nit
Daniel Jacobowitz [Wed, 11 Jul 2001 19:11:07 +0000 (19:11 +0000)]
minor F_SETOWN nit

23 years ago2001-07-11 H.J. Lu <hjl@gnu.org>
H.J. Lu [Wed, 11 Jul 2001 19:03:10 +0000 (19:03 +0000)]
2001-07-11  H.J. Lu  <hjl@gnu.org>

* elf64-alpha.c (elf64_alpha_check_relocs): Only use the same
ALLOC|LOAD flags as the source section for debugging sections
when creating the reloc section.

23 years agofix m68k-linux build
Daniel Jacobowitz [Wed, 11 Jul 2001 18:39:12 +0000 (18:39 +0000)]
fix m68k-linux build

23 years ago* configure.tgt: Mark d10v as pure multi-arch.
Andrew Cagney [Wed, 11 Jul 2001 17:58:15 +0000 (17:58 +0000)]
* configure.tgt: Mark d10v as pure multi-arch.
* config/d10v/tm-d10v.h: Delete file.
* config/d10v/d10v.mt (TM_FILE): Delete.

23 years agos/typedef serial_t/struct serial */
Andrew Cagney [Wed, 11 Jul 2001 17:52:32 +0000 (17:52 +0000)]
s/typedef serial_t/struct serial */

23 years agoFix Credit.
Andrew Cagney [Wed, 11 Jul 2001 17:43:01 +0000 (17:43 +0000)]
Fix Credit.

23 years ago* gdbserver/remote-utils.c (remote_open): Set gdbserver as "owner"
J.T. Conklin [Wed, 11 Jul 2001 17:33:24 +0000 (17:33 +0000)]
* gdbserver/remote-utils.c (remote_open): Set gdbserver as "owner"
of SIGIO.
(input_interrupt): Don't block on read, in case we got redundant
SIGIO.  Don't gripe about redundant SIGIO.
* gdbserver/low-hppabsd.c (mywait): Use waitpid().  Enable SIGIO
handler while waiting.
* gdbserver/low-linux.c (mywait): Likewise.
* gdbserver/low-nbsd.c (mywait): Likewise.
* gdbserver/low-sparc.c (mywait): Likewise.

23 years ago * infrun.c (print_stop_reason): Add missing uiout field
Keith Seitz [Wed, 11 Jul 2001 16:46:57 +0000 (16:46 +0000)]
* infrun.c (print_stop_reason): Add missing uiout field
"reason" for SIGNAL_RECEIVED case.

23 years ago* config/alpha/nm-linux.h (TARGET_ELF64, PSIGNAL_IN_SIGNAL_H):
Mark Kettenis [Wed, 11 Jul 2001 10:33:29 +0000 (10:33 +0000)]
* config/alpha/nm-linux.h (TARGET_ELF64, PSIGNAL_IN_SIGNAL_H):
Remove unused macro's.

23 years ago* config/powerpc/nm-linux.h (NO_SYS_REG_H): Remove, it's no longer
Mark Kettenis [Wed, 11 Jul 2001 10:24:35 +0000 (10:24 +0000)]
* config/powerpc/nm-linux.h (NO_SYS_REG_H): Remove, it's no longer
used.

23 years ago * go32-nat.c (info_dos_cmdlist): New variable.
Eli Zaretskii [Wed, 11 Jul 2001 10:11:31 +0000 (10:11 +0000)]
* go32-nat.c (info_dos_cmdlist): New variable.
(go32_info_dos_command): New function.
(_initialize_go32_nat): Convert all DOS-specific commands into
subcommands of "info dos".  Suggested by Andrew Cagney
<ac131313@redhat.com>.

23 years ago* config/nm-linux.h: Fix comments.
Mark Kettenis [Wed, 11 Jul 2001 10:03:32 +0000 (10:03 +0000)]
* config/nm-linux.h: Fix comments.

23 years ago* thread-db.c (enable_thread_event_reporting): Correct warning
Mark Kettenis [Wed, 11 Jul 2001 09:14:51 +0000 (09:14 +0000)]
* thread-db.c (enable_thread_event_reporting): Correct warning
message about getting thread death breakpoint.
Reported by John S Kallal <jskallal@home.com>.

23 years ago* i386-tdep.c (i386_extract_return_value): "Fix" comment.
Mark Kettenis [Wed, 11 Jul 2001 08:43:41 +0000 (08:43 +0000)]
* i386-tdep.c (i386_extract_return_value): "Fix" comment.
(i386_store_return_value): Frob FPU status and tag word to make
sure the return value is the only value on the FPU stack.

23 years ago* config/tm-linux.h: Do not include <signal.h>. Instead provide
Mark Kettenis [Wed, 11 Jul 2001 08:15:36 +0000 (08:15 +0000)]
* config/tm-linux.h: Do not include <signal.h>.  Instead provide
reasonable defaults for REALTIME_LO and REALTIME_HI if they're not
already defined.
* config/nm-linux.h: Include <signal.h>.
[__SIGRTMIN] (REALTIME_LO, REALTIME_HI): Define to __SIGRTMIN and
(__SIGRTMAX + 1) respectively.

23 years agomerge from gcc
DJ Delorie [Wed, 11 Jul 2001 00:15:54 +0000 (00:15 +0000)]
merge from gcc

23 years ago*** empty log message ***
gdbadmin [Wed, 11 Jul 2001 00:00:03 +0000 (00:00 +0000)]
*** empty log message ***

23 years ago* mcore-rom.c: Include "serial.h".
Andrew Cagney [Tue, 10 Jul 2001 23:08:12 +0000 (23:08 +0000)]
* mcore-rom.c: Include "serial.h".
* Makefile.in (mcore-rom.o): Depends on "serial.h".

23 years agoZap botched merge.
Andrew Cagney [Tue, 10 Jul 2001 23:02:29 +0000 (23:02 +0000)]
Zap botched merge.

23 years ago* Makefile.in (gencode): Provide explicit path to gencode.c.
Andrew Cagney [Tue, 10 Jul 2001 22:46:59 +0000 (22:46 +0000)]
* Makefile.in (gencode): Provide explicit path to gencode.c.

23 years agom88k builds. Tweek remote-bug.c to include "serial.h".
Andrew Cagney [Tue, 10 Jul 2001 22:38:38 +0000 (22:38 +0000)]
m88k builds.  Tweek remote-bug.c to include "serial.h".

23 years agoClean up the D10V port so that GDB and the target program no
Jim Blandy [Tue, 10 Jul 2001 21:24:48 +0000 (21:24 +0000)]
Clean up the D10V port so that GDB and the target program no
longer disagree on how big pointers are.
* findvar.c (value_from_register): Remove special case code for D10V.
* printcmd.c (print_frame_args): Same.
* valops.c (value_at, value_fetch_lazy): Same.
* values.c (unpack_long): Same.
* gdbarch.sh: Changes to effect the following:
* gdbarch.h (GDB_TARGET_IS_D10V, D10V_MAKE_DADDR,
gdbarch_d10v_make_daddr_ftype, gdbarch_d10v_make_daddr,
set_gdbarch_d10v_make_daddr, D10V_MAKE_IADDR,
gdbarch_d10v_make_iaddr_ftype, gdbarch_d10v_make_iaddr,
set_gdbarch_d10v_make_iaddr, D10V_DADDR_P,
gdbarch_d10v_daddr_p_ftype, gdbarch_d10v_daddr_p,
set_gdbarch_d10v_daddr_p, D10V_IADDR_P,
gdbarch_d10v_iaddr_p_ftype, gdbarch_d10v_iaddr_p,
set_gdbarch_d10v_iaddr_p, D10V_CONVERT_DADDR_TO_RAW,
gdbarch_d10v_convert_daddr_to_raw_ftype,
gdbarch_d10v_convert_daddr_to_raw,
set_gdbarch_d10v_convert_daddr_to_raw, D10V_CONVERT_IADDR_TO_RAW,
gdbarch_d10v_convert_iaddr_to_raw_ftype,
gdbarch_d10v_convert_iaddr_to_raw,
set_gdbarch_d10v_convert_iaddr_to_raw): Delete declarations.
* gdbarch.c: Delete the corresponding definitions.
(struct gdbarch): Delete members d10v_make_daddr,
d10v_make_iaddr, d10v_daddr_p, d10v_iaddr_p,
d10v_convert_daddr_to_raw, and d10v_convert_iaddr_to_raw.
(startup_gdbarch): Remove initializers for the above.
(verify_gdbarch, gdbarch_dump): Don't verify or dump them any
more.
* d10v-tdep.c (d10v_register_virtual_type): Rather that
claiming the stack pointer and PC are 32 bits long (which they
aren't), say that the stack pointer is an int16_t, and the
program counter is a function pointer.  This allows the rest
of GDB to make the appropriate conversions between the code
pointer format and real addresses.
(d10v_register_convertible, d10v_register_convert_to_virtual,
d10v_register_convert_to_raw): Delete function; no registers
are convertible now, so we use
generic_register_convertible_not instead.
(d10v_address_to_pointer, d10v_pointer_to_address): New gdbarch
methods.
(d10v_push_arguments, d10v_extract_return_value): Remove special
cases for code and data pointers.
(d10v_gdbarch_init): Set gdbarch_ptr_bit to 16, so that GDB and
the target agree on how large pointers are.  Say that addresses
are 32 bits long.  Register the address_to_pointer and
pointer_to_address conversion functions.  Since no registers are
convertible now, register generic_register_convertible_not as the
gdbarch_register_convertible method instead of
d10v_register_convertible.  Remove registrations for
d10v_register_convert_to_virtual,
d10v_register_convert_to_raw, gdbarch_d10v_make_daddr,
gdbarch_d10v_make_iaddr, gdbarch_d10v_daddr_p,
gdbarch_d10v_iaddr_p, gdbarch_d10v_convert_daddr_to_raw, and
gdbarch_d10v_convert_iaddr_to_raw.

23 years ago* printcmd.c (print_scalar_formatted): If we are printing an
Jim Blandy [Tue, 10 Jul 2001 21:17:49 +0000 (21:17 +0000)]
* printcmd.c (print_scalar_formatted): If we are printing an
address, remember that TARGET_ADDR_BIT is not always equal to
TARGET_PTR_BIT.

23 years ago* valops.c (value_cast): When casting a pointer to an integer,
Jim Blandy [Tue, 10 Jul 2001 21:15:28 +0000 (21:15 +0000)]
* valops.c (value_cast): When casting a pointer to an integer,
don't convert it to an address.

23 years ago* remote-utils.h (struct serial): Declare as opaque. Remove
Andrew Cagney [Tue, 10 Jul 2001 21:06:34 +0000 (21:06 +0000)]
* remote-utils.h (struct serial): Declare as opaque.  Remove
include of "serial.h".
* Makefile.in (remote_utils_h): Update.
* monitor.h (struct serial): Declare as opaque.  Remove include of
"serial.h".
(struct monitor_ops): Replace serial_t with `struct serial *'.
* monitor.c (monitor_desc): Ditto.

23 years agoMIPS/Linux support
Daniel Jacobowitz [Tue, 10 Jul 2001 20:41:54 +0000 (20:41 +0000)]
MIPS/Linux support

23 years ago*** empty log message ***
gdbadmin [Tue, 10 Jul 2001 00:00:08 +0000 (00:00 +0000)]
*** empty log message ***

23 years ago* mi-out.c (mi_out_new): Initialize suppress_ouput field of newly
Mark Kettenis [Mon, 9 Jul 2001 22:19:50 +0000 (22:19 +0000)]
* mi-out.c (mi_out_new): Initialize suppress_ouput field of newly
created `struct ui_out_data'.

23 years ago2001-06-24 Michael Chastain <chastain@redhat.com>
Michael Chastain [Mon, 9 Jul 2001 21:42:52 +0000 (21:42 +0000)]
2001-06-24  Michael Chastain  <chastain@redhat.com>

* gdb.base/arithmet.exp: Remove some tests to make all test names
unique.

23 years agoFix memory leak and compilation problems using AIX 5 native compiler.
Kevin Buettner [Mon, 9 Jul 2001 16:53:03 +0000 (16:53 +0000)]
Fix memory leak and compilation problems using AIX 5 native compiler.

23 years ago* mi-main.c (mi_cmd_data_evaluate_expression): Replace value_ptr
Andrew Cagney [Mon, 9 Jul 2001 14:50:37 +0000 (14:50 +0000)]
* mi-main.c (mi_cmd_data_evaluate_expression): Replace value_ptr
with `struct value *'.

23 years agos/struct _serial_t/struct serial/
Andrew Cagney [Mon, 9 Jul 2001 14:38:49 +0000 (14:38 +0000)]
s/struct _serial_t/struct serial/

23 years ago2001-07-09 Andreas Jaeger <aj@suse.de>, Karsten Keil <kkeil@suse.de>
Andreas Jaeger [Mon, 9 Jul 2001 14:22:11 +0000 (14:22 +0000)]
2001-07-09  Andreas Jaeger  <aj@suse.de>, Karsten Keil <kkeil@suse.de>

* i386-dis.c (set_op): Handle 64 bit and 32 bit mode.
(OP_J): Use bfd_vma for mask to work properly with 64 bits.
(op_address,op_riprel): Use bfd_vma to handle 64 bits.

23 years agoMake function declarators match function declarations.
Kevin Buettner [Mon, 9 Jul 2001 05:53:44 +0000 (05:53 +0000)]
Make function declarators match function declarations.

23 years ago*** empty log message ***
gdbadmin [Mon, 9 Jul 2001 00:00:03 +0000 (00:00 +0000)]
*** empty log message ***

23 years ago* gdbtypes.h (builtin_type_void_data_ptr): Rename
Andrew Cagney [Sun, 8 Jul 2001 20:42:15 +0000 (20:42 +0000)]
* gdbtypes.h (builtin_type_void_data_ptr): Rename
builtin_type_ptr.
* gdbtypes.c (builtin_type_void_data_ptr): Update.
(build_gdbtypes): Update.
(_initialize_gdbtypes): Update.
* values.c (value_as_pointer): Update.
* utils.c (host_pointer_to_address): Update.
(address_to_host_pointer): Update.

23 years ago* remote-udi.c (udi_wait): Make type, instead of name, of first
Andrew Cagney [Sun, 8 Jul 2001 20:21:36 +0000 (20:21 +0000)]
* remote-udi.c (udi_wait): Make type, instead of name, of first
parameter a ptid_t.

23 years ago* ser-mac.c: Make obsolete.
Andrew Cagney [Sun, 8 Jul 2001 20:11:12 +0000 (20:11 +0000)]
* ser-mac.c: Make obsolete.
* Makefile.in (ser-mac.o): Ditto.

23 years ago * go32-nat.c (go32_get_windows_version, print_mem, go32_sysinfo)
Eli Zaretskii [Sun, 8 Jul 2001 10:28:20 +0000 (10:28 +0000)]
* go32-nat.c (go32_get_windows_version, print_mem, go32_sysinfo)
(read_memory_region, get_descriptor, display_descriptor)
(go32_sldt, go32_sgdt, go32_sidt): New functions.
(top-level): Include ctype.h, utsname.h, dos.h, and go32.h.  Ifdef
away `disable' from dos.h, since breakpoint.h defines an enum
member of the same name, and GCC 2.7.2 barfs.
(_initialize_go32_nat): Provide new commands dos-sysinfo, dos-ldt,
dos-gdt, and dos-idt, all of them in the "info" class

23 years ago*** empty log message ***
gdbadmin [Sun, 8 Jul 2001 00:00:01 +0000 (00:00 +0000)]
*** empty log message ***

23 years agoAllocate/free space associated with saved_entryset and saved_exitset.
Kevin Buettner [Sat, 7 Jul 2001 22:02:51 +0000 (22:02 +0000)]
Allocate/free space associated with saved_entryset and saved_exitset.

23 years agoUpdates.
Andrew Cagney [Sat, 7 Jul 2001 17:39:35 +0000 (17:39 +0000)]
Updates.

23 years ago* symtab.c (main_name): New function.
Andrew Cagney [Sat, 7 Jul 2001 17:19:50 +0000 (17:19 +0000)]
* symtab.c (main_name): New function.
(set_main_name): New function.
* symtab.h: Declare.
* TODO: Update

From 2000-03-05 Anthony Green <green@redhat.com>:
* dbxread.c (process_one_symbol): Handle the N_MAIN stab by
setting main_name.
* blockframe.c (inside_main_func): Use main_name instead of
"main".
* symtab.c (find_main_psymtab): Ditto.
* source.c (select_source_symtab): Ditto.
* nlmread.c (nlm_symfile_read): Ditto.
* rs6000-tdep.c (skip_prologue): Ditto.

23 years ago* gdbint.texinfo (User Interface): Update ui-out documentation to
Andrew Cagney [Sat, 7 Jul 2001 16:20:57 +0000 (16:20 +0000)]
* gdbint.texinfo (User Interface): Update ui-out documentation to
refelect recent UI/MI updates.

23 years agoTODO: Convert most items into PRs.
Andrew Cagney [Sat, 7 Jul 2001 16:12:19 +0000 (16:12 +0000)]
TODO: Convert most items into PRs.

23 years ago* lin-lwp.c (status_to_str): New function.
Mark Kettenis [Sat, 7 Jul 2001 10:58:37 +0000 (10:58 +0000)]
* lin-lwp.c (status_to_str): New function.
(lin_lwp_wait): Use it to print debug messages where appropriate.

23 years agoI'm using the obvious fix rule to commit this.
Michael Chastain [Sat, 7 Jul 2001 00:14:06 +0000 (00:14 +0000)]
I'm using the obvious fix rule to commit this.

Testing: I tested on native Red Hat Linux 7 using gcc 3.0.
The "info float" command exercises print_i387_value.

2001-07-06  Michael Chastain  <chastain@redhat.com>

* i387-tdep.c (print_i387_value): Fix pointer glitch.

===

Index: i387-tdep.c
===================================================================
RCS file: /cvs/src/src/gdb/i387-tdep.c,v
retrieving revision 1.11
diff -c -1 -0 -p -r1.11 i387-tdep.c
*** gdb/i387-tdep.c 2001/07/04 21:14:05 1.11
--- gdb/i387-tdep.c 2001/07/06 12:47:00
*************** print_i387_value (char *raw)
*** 163,184 ****
    DOUBLEST value;
    int len = TARGET_LONG_DOUBLE_BIT / TARGET_CHAR_BIT;
    char *tmp = alloca (len);

    /* This code only works on targets where ... */
    gdb_assert (TARGET_LONG_DOUBLE_FORMAT == &floatformat_i387_ext);

    /* Take care of the padding.  FP reg is 80 bits.  The same value in
       memory is 96 bits.  */
    gdb_assert (FPU_REG_RAW_SIZE < len);
!   memcpy (&tmp, raw, FPU_REG_RAW_SIZE);
!   memset (&tmp + FPU_REG_RAW_SIZE, 0, len - FPU_REG_RAW_SIZE);

    /* Extract the value as a DOUBLEST.  */
    /* Use extract_floating() rather than floatformat_to_doublest().
       The latter is lossy in nature.  Once GDB gets a host/target
       independent and non-lossy FP it will become possible to bypass
       extract_floating() and call floatformat*() directly.  Note also
       the assumptions about TARGET_LONG_DOUBLE above.  */
    value = extract_floating (tmp, len);

    /* We try to print 19 digits.  The last digit may or may not contain
--- 163,184 ----
    DOUBLEST value;
    int len = TARGET_LONG_DOUBLE_BIT / TARGET_CHAR_BIT;
    char *tmp = alloca (len);

    /* This code only works on targets where ... */
    gdb_assert (TARGET_LONG_DOUBLE_FORMAT == &floatformat_i387_ext);

    /* Take care of the padding.  FP reg is 80 bits.  The same value in
       memory is 96 bits.  */
    gdb_assert (FPU_REG_RAW_SIZE < len);
!   memcpy (tmp, raw, FPU_REG_RAW_SIZE);
!   memset (tmp + FPU_REG_RAW_SIZE, 0, len - FPU_REG_RAW_SIZE);

    /* Extract the value as a DOUBLEST.  */
    /* Use extract_floating() rather than floatformat_to_doublest().
       The latter is lossy in nature.  Once GDB gets a host/target
       independent and non-lossy FP it will become possible to bypass
       extract_floating() and call floatformat*() directly.  Note also
       the assumptions about TARGET_LONG_DOUBLE above.  */
    value = extract_floating (tmp, len);

    /* We try to print 19 digits.  The last digit may or may not contain

23 years ago*** empty log message ***
gdbadmin [Sat, 7 Jul 2001 00:00:02 +0000 (00:00 +0000)]
*** empty log message ***

23 years ago2001-07-02 Michael Snyder <msnyder@redhat.com>
Michael Snyder [Fri, 6 Jul 2001 22:51:21 +0000 (22:51 +0000)]
2001-07-02  Michael Snyder  <msnyder@redhat.com>

        * gdb.base/completion.exp: Don't assume that break.c is the only
source file that may contain functions named "marker".

23 years ago* lin-lwp.c (count_events_callback): Fix formatting. Turn check
Mark Kettenis [Fri, 6 Jul 2001 22:31:47 +0000 (22:31 +0000)]
* lin-lwp.c (count_events_callback): Fix formatting.  Turn check
commented with "paranoia" into gdb_assert.
(select_event_lwp_callback): Likewise.
(cancel_breakpoints_callback): Bail out early if LP is the event
LWP.  Add comment about backup up breakpoints.  Fix formatting and
debug message.
(select_event_lwp): Make solely repsonsible for switching event
LWP.  Fix formatting and remove bogus "ERROR" debug message.
Don't backup breakpoints from here.
(lin_lwp_wait): Don't touch LP->status, let select_event_lwp
handle that.  Only call select_event_lwp if we're not waiting for
a specific LWP, i.e. when PID == -1.  Backup breakpoints from here.

23 years ago2001-07-07 Ben Elliston <bje@redhat.com>
Ben Elliston [Fri, 6 Jul 2001 22:14:07 +0000 (22:14 +0000)]
2001-07-07  Ben Elliston  <bje@redhat.com>

* m88k.h: Clean up and reformat. Remove unused code.

23 years ago2001-07-06 Michael Snyder <msnyder@redhat.com>
Michael Snyder [Fri, 6 Jul 2001 21:31:04 +0000 (21:31 +0000)]
2001-07-06  Michael Snyder  <msnyder@redhat.com>

* procfs.c (procfs_resume): Silence noisy warning.

23 years ago2001-06-12 Michael Snyder <msnyder@redhat.com>
Michael Snyder [Fri, 6 Jul 2001 19:06:24 +0000 (19:06 +0000)]
2001-06-12  Michael Snyder  <msnyder@redhat.com>

* lin-lwp.c: Prevent thread starvation by using a monte carlo
method to choose which of several event threads to handle next.

(stop_wait_callback): Defer pushback of breakpoint events until
later; add SIGTRAP events to the queue of unhandled events.
Keep calling waitpid until SIGSTOP retrieved.  If more than one
non-SIGSTOP event is retrieved, push them back onto the process
queue using kill.
(count_events_callback, select_singlestep_lwp_callback,
select_event_lwp_callback, cancel_breakpoints_callback,
select_event_lwp): New functions.  Implement monte carlo method
for selecting which of several SIGTRAP threads to handle next.
Push back the breakpoint event for all threads other than the
selected one.
(lin_lwp_wait): Call select_event_lwp to decide which of several
sigtrapped lwps to handle next.
(resume_callback): Disable code that attempts to handle
step_resume breakpoints.  Let core gdb handle this.

23 years ago* gdbtypes.h (builtin_type_void_func_ptr): New builtin type. dberlin-typesystem-branchpoint
Jim Blandy [Fri, 6 Jul 2001 14:44:19 +0000 (14:44 +0000)]
* gdbtypes.h (builtin_type_void_func_ptr): New builtin type.
* gdbtypes.c (builtin_type_void_func_ptr): Define the variable.
(build_gdbtypes): Initialize it.
(_initialize_gdbtypes): Swap it.

23 years agoBranch prediction code cleanup
Daniel Jacobowitz [Fri, 6 Jul 2001 05:35:17 +0000 (05:35 +0000)]
Branch prediction code cleanup

23 years agoMention mi0 and mi1 as --interpreter options.
Andrew Cagney [Fri, 6 Jul 2001 04:07:29 +0000 (04:07 +0000)]
Mention mi0 and mi1 as --interpreter options.

23 years agoReplace:
Andrew Cagney [Fri, 6 Jul 2001 03:53:11 +0000 (03:53 +0000)]
Replace:
  if (interpreter_p && strncmp (interpreter_p, "mi", 2) == 0)
with:
  if (ui_out_is_mi_like_p (uiout))

23 years ago*** empty log message ***
gdbadmin [Fri, 6 Jul 2001 00:00:03 +0000 (00:00 +0000)]
*** empty log message ***

23 years ago * mips-tdep.c (mips_software_single_step): New function.
Daniel Jacobowitz [Thu, 5 Jul 2001 23:30:43 +0000 (23:30 +0000)]
    * mips-tdep.c (mips_software_single_step): New function.
        * config/mips/tm-mips.h: Add prototype for
        mips_software_single_step.

23 years ago * library/managedwin.itb (_create): Don't pass "-over"
Keith Seitz [Thu, 5 Jul 2001 23:24:10 +0000 (23:24 +0000)]
* library/managedwin.itb (_create): Don't pass "-over"
to center_window unless we really are asked to put the
window over another!

23 years ago * ppc-linux-nat.c (supply_gregset): Use elf_greg_t instead
Daniel Jacobowitz [Thu, 5 Jul 2001 23:22:04 +0000 (23:22 +0000)]
    * ppc-linux-nat.c (supply_gregset): Use elf_greg_t instead
        of greg_t.
        (fill_gregset): Likewise.

23 years agoFix ia64-linux fortran common linking problem.
Jim Wilson [Thu, 5 Jul 2001 22:40:16 +0000 (22:40 +0000)]
Fix ia64-linux fortran common linking problem.
* linker.c (_bfd_generic_link_add_one_symbol, case BIG): Use
the section of the bigger symbol.

23 years agos/basename/lbasename/
Andrew Cagney [Thu, 5 Jul 2001 21:32:39 +0000 (21:32 +0000)]
s/basename/lbasename/

23 years ago* d10v-tdep.c (d10v_frame_chain, d10v_frame_init_saved_regs,
Jim Blandy [Thu, 5 Jul 2001 21:27:08 +0000 (21:27 +0000)]
* d10v-tdep.c (d10v_frame_chain, d10v_frame_init_saved_regs,
show_regs, d10v_read_pc, d10v_write_pc, d10v_read_sp,
d10v_write_sp, d10v_write_fp, d10v_read_fp,
d10v_push_return_address): Call the functions d10v_make_daddr,
d10v_make_iaddr, d10v_convert_iaddr_to_raw, and
d10v_convert_daddr_to_raw, not the global macros D10V_MAKE_DADDR,
D10V_MAKE_IADDR, D10V_CONVERT_IADDR_TO_RAW, and
D10V_CONVERT_DADDR_TO_RAW.

23 years agoFix ia64-linux binutils readelf -s testsuite failure.
Jim Wilson [Thu, 5 Jul 2001 18:27:55 +0000 (18:27 +0000)]
Fix ia64-linux binutils readelf -s testsuite failure.
* syms.c (bfd_is_local_label): Return false if BSF_SECTION_SYM.

23 years agomerge from gcc
DJ Delorie [Thu, 5 Jul 2001 17:29:17 +0000 (17:29 +0000)]
merge from gcc

23 years ago* dwarf2read (dwarf2_build_psymtabs_hard): Doc fix.
Jim Blandy [Thu, 5 Jul 2001 16:45:48 +0000 (16:45 +0000)]
* dwarf2read (dwarf2_build_psymtabs_hard): Doc fix.

23 years ago* config/i386/xm-go32.h (HOST_I386): Removed.
Mark Kettenis [Thu, 5 Jul 2001 14:40:51 +0000 (14:40 +0000)]
* config/i386/xm-go32.h (HOST_I386): Removed.
* config/i386/xm-linux.h (HOST_I386): Removed.

23 years ago2001-07-05 Ben Elliston <bje@redhat.com>
Ben Elliston [Thu, 5 Jul 2001 13:51:26 +0000 (13:51 +0000)]
2001-07-05  Ben Elliston  <bje@redhat.com>

* Make-common.in (srccgen): Remove.
(CGEN_CPU_DIR): Define.
(CGEN_READ_SCM): Redefine without $(srccgen).
(CGEN_ARCH_SCM): Ditto.
(CGEN_CPU_SCM): Ditto.
(CGEN_DECODE_SCM): Ditto.
(CGEN_DESC_SCM): Ditto.

* $arch/Makefile.in: Use $(CGEN_CPU_DIR) where applicable.

23 years ago2001-07-05 Ben Elliston <bje@redhat.com>
Ben Elliston [Thu, 5 Jul 2001 13:12:23 +0000 (13:12 +0000)]
2001-07-05  Ben Elliston  <bje@redhat.com>

* Makefile.am (CPUDIR): Define.
(stamp-m32r): Update dependencies.
(stamp-fr30): Ditto.
(stamp-openrisc): Ditto.
* Makefile.in: Regenerate.

23 years agoFor sections containing uninitialised data, only set their size to s_paddr
Nick Clifton [Thu, 5 Jul 2001 10:52:25 +0000 (10:52 +0000)]
For sections containing uninitialised data, only set their size to s_paddr
if this does not reset their size to zero.  MS's latest compilers will set
s_paddr to zero.

23 years ago*** empty log message ***
gdbadmin [Thu, 5 Jul 2001 00:00:04 +0000 (00:00 +0000)]
*** empty log message ***

23 years ago* i387-tdep.c (print_i387_value): Add extra space after final full
Mark Kettenis [Wed, 4 Jul 2001 21:14:05 +0000 (21:14 +0000)]
* i387-tdep.c (print_i387_value): Add extra space after final full
stop in comment.

23 years ago* gdbint.texinfo (Target Architecture Definition): GDBARCH is a C
Andrew Cagney [Wed, 4 Jul 2001 20:52:14 +0000 (20:52 +0000)]
* gdbint.texinfo (Target Architecture Definition): GDBARCH is a C
structure and not macros.
(Host Definition): Document that much of this chapter is obsolete.
(Target Architecture Definition): Update list of files that make
up a target architecture.
(Coding): Update.

23 years ago* config/arm/tm-arm.h: Include "floatformat.h".
Andrew Cagney [Wed, 4 Jul 2001 19:57:53 +0000 (19:57 +0000)]
* config/arm/tm-arm.h: Include "floatformat.h".

23 years ago* i387-tdep.c: Include "gdb_assert.h".
Andrew Cagney [Wed, 4 Jul 2001 19:19:58 +0000 (19:19 +0000)]
* i387-tdep.c: Include "gdb_assert.h".
(print_i387_value): Use extract_floating to extract the FP value
from a zero padded local buffer.

23 years ago* TODO: Delete all thread items. The thread code was overhauled.
Andrew Cagney [Wed, 4 Jul 2001 18:51:51 +0000 (18:51 +0000)]
* TODO: Delete all thread items.  The thread code was overhauled.

23 years ago2001-07-04 Elena Zannoni <ezannoni@redhat.com>
Elena Zannoni [Wed, 4 Jul 2001 18:06:02 +0000 (18:06 +0000)]
2001-07-04  Elena Zannoni  <ezannoni@redhat.com>

* memattr.c (create_mem_region): Move n to next memory region,
to avoid infinite loop.

23 years ago2001-07-04 Elena Zannoni <ezannoni@redhat.com>
Elena Zannoni [Wed, 4 Jul 2001 17:59:02 +0000 (17:59 +0000)]
2001-07-04  Elena Zannoni  <ezannoni@redhat.com>

* memattr.h: Add copyright statement.
* memattr.c: Ditto.

23 years ago* dwarf2read.c (struct partial_die_info): New member: has_pc_info.
Jim Blandy [Wed, 4 Jul 2001 16:08:20 +0000 (16:08 +0000)]
* dwarf2read.c (struct partial_die_info): New member: has_pc_info.
(read_partial_die): Delete fourth argument; we return this info in
the struct partial_die_info object itself now.
(dwarf2_build_psymtabs_hard, scan_partial_symbols): Use the
has_pc_info field of the partial die struct, rather than passing a
variable by reference to read_partial_die.

23 years agoDoc fix.
Jim Blandy [Wed, 4 Jul 2001 16:00:35 +0000 (16:00 +0000)]
Doc fix.

23 years ago* dwarf2read.c (dwarf2_build_psymtabs_hard): Remove extraneous
Jim Blandy [Wed, 4 Jul 2001 16:00:08 +0000 (16:00 +0000)]
* dwarf2read.c (dwarf2_build_psymtabs_hard): Remove extraneous
code in loop condition.  This seemed to be trying to round
info_ptr up to the next four-byte boundary, but that's not what it
actually did.  If we discover the problem the old code was really
trying to address, we can fix it properly.

23 years ago2001-07-04 H.J. Lu <hjl@gnu.org>
H.J. Lu [Wed, 4 Jul 2001 07:15:20 +0000 (07:15 +0000)]
2001-07-04  H.J. Lu  <hjl@gnu.org>

* bfd/elf32-mips.c (_bfd_mips_elf_final_write_processing):
Handle bfd_mach_mips4400, bfd_mach_mips4600 and
bfd_mach_mips5000.

23 years ago*** empty log message ***
gdbadmin [Wed, 4 Jul 2001 00:00:02 +0000 (00:00 +0000)]
*** empty log message ***