Copy stable/9 to releng/9.3 as part of the 9.3-RELEASE cycle.

Approved by: re (implicit)
Sponsored by: The FreeBSD Foundation

MFC r258663:
Use sysctl KERN_PROC_SIGTRAMP to retrieve the signal trampoline
location for the native amd64 ABI. This fixes unwinding over the
signal frame after trampoline was moved to the shared page.

MFC r251844: Include die tag in error message

MFC r251667: Handle DW_TAG_unspecified_type

2011-11-14 Jim Ingham <jingham@apple.com>

* dwarf2read.c (read_type_die): Handle DW_TAG_unspecified_type.
(read_tag_unspecified_type): New function, add a type for the
DW_TAG_unspecified_type die.

Obtained from: Apple, gdb-1752

MFC r251665: Handle DW_TAG_rvalue_reference_type

2012-05-21 Jim Ingham <jingham@apple.com>

* dwarf2read.c (process_die): Handle DW_TAG_rvalue_reference_type.
(read_type_die): Ditto.
(dwarf_tag_name): Ditto.

* elf/dwarf2.h: Add DW_TAG_rvalue_reference_type.

Obtained from: Apple, gdb-1820

MFC r244437: Support restrict qualifier in dwarf debug info

Newer clang/llvm emit DW_TAG_restrict_type, which wasn't handled by gdb.
Import support from Apple's gdb-1822:

| 2009-03-24 Jason Molenda (jmolenda@apple.com)
|
| * gdbtypes.c (make_cv_type): Rename this function to make_cvr_type to
| also handle restrict qualifiers.
| (check_typedef): Handle TYPE_RESTRICT.
| * gdbtypes.h (TYPE_FLAG_RESTRICT, TYPE_RESTRICT): New.
| * hpread.c (hpread_type_lookup): Update to use make_cvr_type.
| * stabsread.c (read_type): Pass the restrict qualifiers along.
| * parse.c (follow_types): Pass the restrict qualifiers along.
| * dwarf2read.c (read_tag_const_type): Call make_cvr_type.
| (read_tag_volatile_type): Same.
| (read_tag_restrict_type): New function.
| (read_type_die): Handle DW_TAG_restrict_type.

Obtained from: Apple

Clean some 'svn:executable' properties in the tree.

Submitted by: Christoph Mallon

While here, merge some other mergeinfo properties that
were left behind from my commits

/head/include:r241008,241141,241181
/head/contrib/gcc:r244776,244792
/head/cddl:r238457,238509,238558

MFC r242943: Do not unwind past a zero PC frame.

This improves GDB usability when debugging code compiled with
optimization.

Upstream GDB revision f0031b6d3ae9b164b3747986ab898190bd4dcf8c (prior
to GDB's switch to GPLv3), with frame_debug_got_null_frame expanded
inline.

| 2004-12-12 Andrew Cagney <cagney@gnu.org>
|
| * frame.c (get_prev_frame): When unwinding normal frames, check
| that the PC isn't zero.

Sponsored by: ADARA Networks

MFC r242936: Set optimized_out instead of reporting an error.

This provides a better display when debugging code compiled with
optimization on.

GDB git revision e8395b4efd184f745070afb953f451f99e922be7 (prior to
GDB's switch to GPLv3), modified for the interfaces provided by
GDB 6.1.1.

2005-02-28 Daniel Jacobowitz <d...@codesourcery.com>

* dwarf2loc.c (loclist_read_variable): Set optimized_out
instead of reporting an error.
* valprint.c (value_check_printable): New function.
(common_val_print): New function. Use value_check_printable.
(value_print): Use value_check_printable.
* value.h (common_val_print): Add prototype.
* c-valprint.c (c_val_print): Use common_val_print.
* cp-valprint.c (cp_print_value_fields): Likewise.
(cp_print_hpacc_virtual_table_entries): Likewise.
* f-valprint.c (f_val_print): Likewise.
* jv-valprint.c (java_value_print, java_print_value_fields):
Likewise.
* scm-valprint.c (scm_value_print): Likewise.
* stack.c (print_frame_args): Likewise.
* varobj.c (c_value_of_variable): Likewise.
* p-valprint.c (pascal_val_print, pascal_value_print): Likewise.
(pascal_object_print_value_fields): Likewise. Update call to
pascal_object_print_static_field.
(pascal_object_print_static_field): Remove TYPE argument. Use
common_val_print.

Sponsored by: ADARA Networks

MFC r240163: Fix "Corrupted DWARF expression" from (k)gdb.

Google turned up Debian bug 405116, which describes the problem in
sufficient detail to identify the overflowing variables.

MFC r239870:

Teach gdb about the DW_FORM_flag_present dwarf attribute, so it doesn't
error out on files that contain it. (This attribute can be emitted by
newer versions of clang.)

Copy head to stable/9 as part of 9.0-RELEASE release cycle.

Approved by: re (implicit)

Implement functions necessary for compiling fbsd-threads.c.

Reviewed by: marcel
Approved by: re (kib)
MFC after: 1 week

- Check on target wordsize instead of compile time define if we build on
64-bit PowerPC or 32-bit PowerPC.
- Make gdb work on powerpc64, the code for this is obtained from
ppc-linux-tdep.c.
- Remove non-elf core read functionality. Implement core read functionality
similar like other FreeBSD targets.
- Set long double limitations.

Remove stray semicolon. This fix appears in later upstream gdb versions.

Submitted by: eadler

Merge binutils 2.17.50 to head. This brings a number of improvements to
x86 CPU support, better support for powerpc64, some new directives, and
many other things. Bump __FreeBSD_version, and add a note to UPDATING.

Thanks to the many people that have helped to test this.

Obtained from: projects/binutils-2.17

Add the ability for GDB to printout the thread name along with other
thread specific informations.

In order to do that, and in order to avoid KBI breakage with existing
infrastructure the following semantic is implemented:
- For live programs, a new member to the PT_LWPINFO is added (pl_tdname)
- For cores, a new ELF note is added (NT_THRMISC) that can be used for
storing thread specific, miscellaneous, informations. Right now it is
just popluated with a thread name.

GDB, then, retrieves the correct informations from the corefile via the
BFD interface, as it groks the ELF notes and create appropriate
pseudo-sections.

Sponsored by: Sandvine Incorporated
Tested by: gianni
Discussed with: dim, kan, kib
MFC after: 2 weeks

Change register numbers according to current struct reg and
struct fpreg in src/sys/mips/include/reg.h

Calculate offset from frame top for registers saved on a stack frame.

- Add ELF core for FreeBSD/mips
- Register all known cores in init function

Fix reading of .debug_line on MIPS64, big-endian:
read_initial_length detects pointer size by checking first
4 bytes of .debug_line and stores it in struct comp_unit_head *
passed to it as second argument. By passing NULL to a
read_initial_length we ignore actual pointer size (8 bytes) and
use default (4 bytes) which results in wrong offsets of header fields.

Teach our toolchain how to generate 64-bit PowerPC binaries. This fixes
a variety of bugs in binutils related to handling of 64-bit PPC ELF,
provides a GCC configuration for 64-bit PowerPC on FreeBSD, and
associated build systems tweaks.

Obtained from: projects/ppc64

Adds the missing mips gdb files that I
foobar'd on my last commit.. Sorry about
that JC.. my fault.

Obtained from: JC

These contain JC's patch to get gdb sort of working
on mips. Its not fully done yet but its a start.

Obtained from: JC - c.jayachandran@gmail.com

M gnu/usr.bin/gdb/kgdb/trgt_mips.c
M gnu/usr.bin/gdb/arch/mips/init.c
M gnu/usr.bin/gdb/arch/mips/Makefile
M gnu/usr.bin/Makefile
M contrib/gdb/gdb/mips-tdep.h

gdb: make 'thread apply all bt' always work on all threads

even if some appear to have (partially) corrupted stack traces.
E.g. kernel crashdumps typically have stack weirdness at
userland-kernel boundary.

Obtained from: vendor/upstream (CVS rev 1.118 of stack.c)
Reviewed by: emaste
Approved by: jhb

Remove empty directories from the HEAD.

Discussed with: developers, imp

gdb: Remove arm_pc_is_thumb_dummy() and related code.

This is basically an import of the following gdb change:
http://sourceware.org/ml/gdb-cvs/2005-03/msg00143.html (which in effect fixes
problems with gracefully closing down the non-Thumb program being debugged).

This commit was generated by cvs2svn to compensate for changes in r173619,
which included commits to RCS files with non-trunk default branches.

This commit was generated by cvs2svn to compensate for changes in r169185,
which included commits to RCS files with non-trunk default branches.

This commit was generated by cvs2svn to compensate for changes in r161561,
which included commits to RCS files with non-trunk default branches.

This commit was generated by cvs2svn to compensate for changes in r161537,
which included commits to RCS files with non-trunk default branches.

Abstract the handling of dirty stacked registers in ia64_read_reg() and
ia64_write_reg() and call these whenever we read or write registers from
and to memory. This way we correctly read registers from addresses in
the register cache and thus fixes backtraces when the saved registers
(such as rp and ar.pfs) are in dirty stacked registers.

Disable surplus detection of breakpoint event in signal stack, the check
causes a normal breakpoint event in new thread to be ignored unexpectly,
result is process SIGSEGV.

Test on: AMD64, i386

Various improvements for ia64:
o s/TARGET_XFER_DIRTY/NATIVE_XFER_DIRTY/g to be conformant,
o Handle TARGET_OBJECT_DIRTY for core files. Required on Linux,
o Replace ia64_fbsd_supply_fpregs() and ia64_fbsd_supply_gregs()
with supply_fpregset() and supply_gregset() resp. Needed to
support libthread_db.
o For target_read_partial(), pass the address of bspstore as the
annex instead of bspstore itself. This fixes crossbuilds. The
annex is a pointer, while bspstore is a ULONGEST.
o In ia64_frame_prev_register(), handle dirty stacked registers.

Remove fbsd-thread.o, it's not for consumption yet.

Update HEAD with support for ia64.

This commit was generated by cvs2svn to compensate for changes in r131082,
which included commits to RCS files with non-trunk default branches.

Remove from HEAD, that what is not on the vendor branch anymore.

Revive files which were pulled from the vendor branch only to be
removed on HEAD, but which are part of the GDB 6.1.1 import. This
makes HEAD identical to the vendor branch, for better or for
worse.

Update to GDB 6.1.1. These files are identical to the vendor branch.

This commit was generated by cvs2svn to compensate for changes in r130805,
which included commits to RCS files with non-trunk default branches.

This commit was generated by cvs2svn to compensate for changes in r130803,
which included commits to RCS files with non-trunk default branches.

This file is not part of GDB 5.2.1, but was off the vendor branch.

Revert local changes.

This file is not part of GDB 5.2.1 and is also off the vendor branch.
It was removed on 2001-11-07, according to ChangeLog.

Track changed definition of the debug registers.

Sorry for missing this the first time.

Resolve conflicts with gdb-5.2.1 import.

This commit was generated by cvs2svn to compensate for changes in r104990,
which included commits to RCS files with non-trunk default branches.

Make our i386 hacks in a less vendor file disturbing manner.

Quiet warnings by making this ANSI-C clean.

This commit was generated by cvs2svn to compensate for changes in r99005,
which included commits to RCS files with non-trunk default branches.

- kernel core debugging for i386 (Mark Peek)
- thread debugging for i386 (Mark Peek) and alpha (Doug Rabson)

Submitted by: mp,dfr

Note that this is a modified version of GDB.

Use the stock source.

Best that I can tell, we've never used these files.

Actually rev 1.3 (don't declare tilde_expand()) is needed with GDB 5.2.

Our FreeBSD custom changes are OBE. Return to using the stock vendor sources.

This commit was generated by cvs2svn to compensate for changes in r98944,
which included commits to RCS files with non-trunk default branches.

Use these malloc-like bits from libiberty now.

Remove the filename_completion_function() extern that conflicts with the
latest version of this fuction in Readline 4.2

The internal declaration for tilde_expand() now conflicts with the one
that was finally added to the readline headers in readline 4.2.

Submitted by: ache

Remove these MD files that we've never needed.

Remove this auto-generated file (and isn't part of 5.0 as it shouldn't be).

Remove these GDB 4.18 docs that did not make it into 5.0.

Remove this GDB 4.18 file that is no longer part of GDB [5.0].

This commit was generated by cvs2svn to compensate for changes in r64234,
which included commits to RCS files with non-trunk default branches.

These files are for an arch we don't care about.

These files are for an arch we don't care about.

These files were part of GDB 4.16, but not version 4.18 (that we
currently use).

Fix (I believe) Bill Paul's core dump when using gdb on shared executables.
This is an addition to rev 1.5. If lm == NULL, new->lm is uninitialized.

Reduce default timeout on remote serial debugging sessions from 20
seconds to 1 second. This fixes a problem where gdb would appear to
hang on flaky serial connections. There's a theoretical problem that
the relatively short timeout could cause problems on slow links, but
you can override the default value with the 'set remotetimeout'
command.

Approved-by: dfr

Fix a coredump when debugging corefiles from static programs.

Resurrect a file for alpha support in GDB and add some minor FreeBSD tweaks.

Magic wand waved by: Peter Wemm <peter@netplex.com.au>

Merge local changes into gdb-4.18 and add changes for FreeBSD/alpha.

This commit was generated by cvs2svn to compensate for changes in r46283,
which included commits to RCS files with non-trunk default branches.

Fix bug: If gdb couldn't find the dynamic linker's "r_debug_state"
symbol, it aborted with "virtual memory exhausted." Now it warns
about the problem and proceeds.

Rewrite the bits of the solib code to handle implentations where dlopen
may insert into the list of loaded libraries (ours is one) instead of
appending to the end of the list. Also cope with dlclose() removing
libraries from the list.

Bring core-regset.c back to life. It is needed for handling ELF
core dumps.

Don't assume that time_t is long. Don't assume that time_t's can be
represented by ints. Keep assuming that time_t's can be represented
by longs.

Add support for baudrates higher than 37800.

Use the dynamic symtab of the runtime linker when looking for places to set
breakpoints. It may have been installed stripped.

Backed out all FreeBSD changes. They are unnecessary now that bfd
understands kernels.

Reviewed by: jdp

Close some file descriptors on exec so that the inferior process gets
a normal environment. Fd's 3 and 4 were usually left open.

The previous revision (objfiles.c:1.2) fixed backtracing in the
absence of full debugging symbols for the kernel, but broke it for
application programs. This commit disables that change except when
kernel debugging mode is in effect.

This needs to go into -2.2 as well, after a suitable burn-in period.

Removed most unused files (about 8M total).

Make all the info dir entries line up neatly.

Base work to make gdbserver work with FreeBSD

Don't use these that were generated on a foreign system, generate our own.

Quick fix for section limits not being relocated. This has always been
broken for gdb -k, but the section limits weren't used much in previous
versions of gdb in FreeBSD. Now they are used for backtracing when full
symbols aren't available, and in some other new cases.

This should be fixed properly by someone who knows bfd.

This should be fixed in 2.2.

Removed binary file.

Removed unnecessary #include of obsolete <sys/dir.h> again.

Restored my fixes for printing the FP registers. Their
history was lost with FreeBSD-1.x and they were blown away
by the gdb-4.16 merge. I needed this to debug e_exp.S ...

Restored even older code (from 386BSD-0.0) for converting the
FP registers to doubles. floatformat_to_double() and/or
valprint() still don't understand NaNs.

Removed unnecessary #include of obsolete <sys/dir.h> again.

FreeBSD specific modifications.

Obtained from /usr/ports/devel/gdb.

2.2. candidate ?

Should I put $FreeBSD$ into these files ?

This commit was generated by cvs2svn to compensate for changes in r19370,
which included commits to RCS files with non-trunk default branches.