After thread support over core files was added for GNU/Linux was added,
we started noticing the following type of crash when trying to perform
task switches (this is a bit accademic, since task switching is not
supported when debugging core files - this is what our testcase was
verifying).
(please check out the comment inside ada-tasks.c:task_command for
more details on this topic)
The reason for the crash comes from the fact that the GNU/Linux thread
layer now gets pushed on the target stack, causing the associated
to_get_ada_task_ptid target method to be activated. This routine
makes the assumption that, for all threads, the private area is not
NULL. This is incorrect in the case of core files, as the core layer
creates some threads with no private data.
But, taking a step back, we don't need to try to compute the task ptid,
as we'll never be using it anyways (we only use it for task switching).
So the fix is to avoid the ptid computation altogether when debugging
a core file.
gdb/ChangeLog:
* ada-tasks.c (read_atcb): Do not compute the task ptid when
debugging a core file.
* dwarf2read.c (die_needs_namespace) <DW_TAG_variable>: Add also
DW_TAG_constant.
gdb/testsuite/
* gdb.fortran/module.exp (fully qualified name of DW_TAG_constant):
New test.
* gdb.fortran/module.f90 (mod1) <var_const>: New constant.
The testcase was failing on various 64bit platforms, because the
debugging info said that the target type for a const type is a
4-byte reference. In the asm file, I used a .long directive, but
the size of .long is of course dependent on the platform. This
replaces this .long by a .4byte directive.
gdb/testsuite/ChangeLog:
* gdb.dwarf2/dw2-const.S: Use .4byte to reference the target
type of our const type.
This patch adds a new --enable-gdbserver=yes/no/auto command-line switch
in gdb/configure. The primary purpose is to allow a user to disable the
build & install of gdbserver when not desired. It also allows the user
to request gdbserver in which case the configure script will abort if
automatic building of gdbserver is not supported for that configuration.
The default keeps things as is: We automatically build gdbserver if
building for a native configuration and if gdbserver is supported for
that configuration.
gdb/ChangeLog:
* configure.ac: Add support for --enable-gdbserver.
* configure: Regenerate.
At AdaCore, this is mostly intended for ia64-linux, where we currently
do not want to distribute gdbserver. This will be useful in the context
of I417-033, where we are now using "make install" to determine what
needs to be packaged inside the GNAT Pro package.
PR mi/11407
* mi/mi-cmd-stack.c (list_args_or_locals): Catch exceptions from
read_var_value and common_val_print and print a warning.
2010-09-16 Phil Muldoon <pmuldoon@redhat.com>
Jan Kratochvil <jan.kratochvil@redhat.com>
PR mi/11407
* gdb.mi/dw2-ref-missing-frame-func.c: New File.
* gdb.mi/dw2-ref-missing-frame-main.c New File.
* gdb.mi/dw2-ref-missing-frame.S New File.
* gdb.mi/dw2-ref-missing-frame.exp New File.
PR exp/11803:
* value.c (value_static_field): Use value_of_variable.
gdb/testsuite
PR exp/11803:
* gdb.threads/tls.exp: Use C++.
(check_thread_local): Use K::another_thread_local.
* gdb.threads/tls.c (class K): New.
(another_thread_local): Now a member of K.
(spin): Update. No longer K&R C.
2010-09-13 Sami Wagiaalla <swagiaal@redhat.com>
PR symtab/11992:
* c-exp.y (classify_name): Check is_a_member_of_this before returning
UNKNOWN_CPP_NAME.
2010-09-13 Sami Wagiaalla <swagiaal@redhat.com>
* gdb.cp/koenig.cc: created class for testing member lookup.
* gdb.cp/koenig.exp: Added test for member lookup.
This patch adjusts the NEWS entry for the LynxOS gdbserver to include
x86-lynxos to the list of supported platforms.
gdb/ChangeLog:
* NEWS: Add x86-lynxos to the list of platforms supported by
gdbserver.
This patch adds support for x86 LynxOS (tested on LynxOS 5.0).
gdb/gdbserver/ChangeLog:
* lynx-i386-low.c: New file.
* configure.srv: Add handling of i[34567]86-*-lynxos* targets.
This adjusts the news entry about ppc-lynxos to add version 5.x
of the OS to the list of supported versions...
gdb/ChangeLog:
* NEWS: Announce GDBserver support for version 5.x of ppc-lynxos.
Some of the PTRACE_ macros/values were removed in LynxOS 5.x.
Since we have not use them so far, the chances that we will use them
one day are very small. So I decided to delete them.
gdb/gdbserver/ChangeLog:
* lynx-low.c (ptrace_request_to_str): Remove handling for
request values that have been removed in LynxOS 5.x.
This is one of the changes needed in order to build gdbserver on
LynxOS 5.x.
Really interesting: On LynxOS 4.x, there is a #warning when sys/ptrace.h
is used (explaining that ptrace.h will be used instead), whereas this
file was removed from LynxOS 5.x. The contents of sys/ptrace.h on 4.x
(or at least the meat of it):
#if defined(__GNUC__) || defined(__GNUG__)
#if !defined(__NO_INCLUDE_WARN__)
#warning Using <ptrace.h> instead of <sys/ptrace.h>
#endif /* defined(__NO_INCLUDE_WARN__) */
#endif /* defined(__GNUC__) || defined(__GNUG__) */
#include <ptrace.h>
The fix I went for, for now, is to just include <sys/ptrace.h>
unconditionally. I could have done some configury, but we already
have to build with -D__NO_INCLUDE_WARN__ to avoid the warnings
anyway, and that's unvoidable, due to system includes themselves
including the "wrong" header file.
Since <sys/ptrace.h> seems to be the choice that was made for LynxOS,
and since it works to include it on LynxOS 4.x, I think that's the simplest
solution.
gdb/gdbserver/ChangeLog:
* lynx-low.c, lynx-ppc-low.c: Include <sys/ptrace.h> instead of
<ptrace.h>
