I used a refactoring script to add target_ops arguments to all the
target methods. In order to make this script work a little better,
this patch adds a new "remote_load" function; this eliminates the need
to later change the signature of generic_load.
2014-02-19 Tom Tromey <tromey@redhat.com>
* remote.c (remote_load): New function.
(init_remote_ops): Use it.
This adds a "self" argument to to_supports_btrace. Due to how one
implementation of this method is shared with gdbserver this required a
small change to gdbserver as well.
2014-02-19 Tom Tromey <tromey@redhat.com>
* common/linux-btrace.c (linux_supports_btrace): Add "ops"
argument.
* common/linux-btrace.h (linux_supports_btrace): Update.
* remote.c (remote_supports_btrace): Add "self" argument.
* target-delegates.c: Rebuild.
* target.c (target_supports_btrace): Remove.
* target.h (struct target_ops) <to_supports_btrace>: Add
target_ops argument.
(target_supports_btrace): New define.
2014-02-19 Tom Tromey <tromey@redhat.com>
* target.h (struct target_ops) <supports_btrace>: Add target_ops
argument.
(target_supports_btrace): Update.
This patch replaces some code in the record targets with target method
delegation.
record-full.c stores pointers to many target methods when the record
target is pushed. Then it later delegates some calls via these. This
is wrong because it violates the target stack contract. In particular
it is ok to unpush a target at any stratum, but record-full does not
keep track of this, so it could potentially call into an unpushed
target.
This patch fixes the problem by using the newly-introduced generic
approach to target delegation for the methods in question.
2014-02-19 Tom Tromey <tromey@redhat.com>
* record-full.c (record_full_beneath_to_resume_ops)
(record_full_beneath_to_resume, record_full_beneath_to_wait_ops)
(record_full_beneath_to_wait)
(record_full_beneath_to_store_registers_ops)
(record_full_beneath_to_store_registers)
(record_full_beneath_to_xfer_partial_ops)
(record_full_beneath_to_xfer_partial)
(record_full_beneath_to_insert_breakpoint_ops)
(record_full_beneath_to_insert_breakpoint)
(record_full_beneath_to_remove_breakpoint_ops)
(record_full_beneath_to_remove_breakpoint)
(record_full_beneath_to_stopped_by_watchpoint)
(record_full_beneath_to_stopped_data_address)
(record_full_beneath_to_async, tmp_to_resume_ops, tmp_to_resume)
(tmp_to_wait_ops, tmp_to_wait, tmp_to_store_registers_ops)
(tmp_to_store_registers, tmp_to_xfer_partial_ops)
(tmp_to_xfer_partial, tmp_to_instmp_to_insert_breakpoint_ops)
(tmp_to_insert_breakpoint, tmp_to_remove_breakpoint_ops)
(tmp_to_remove_breakpoint, tmp_to_stopped_by_watchpoint)
(tmp_to_stopped_data_address, tmp_to_async): Remove.
(record_full_open_1, record_full_open): Update. Use RECORD_IS_USED.
(record_full_resume, record_full_wait_1)
(record_full_stopped_by_watchpoint, record_full_stopped_data_address)
(record_full_store_registers, record_full_xfer_partial)
(record_full_insert_breakpoint, record_full_remove_breakpoint)
(record_full_async, record_full_core_xfer_partial): Use target
delegation.
* target-delegates.c: Rebuild.
* target.c (current_xfer_partial): Remove.
(update_current_target): Do not INHERIT or de_fault
to_insert_breakpoint, to_remove_breakpoint,
to_stopped_data_address, to_stopped_by_watchpoint, to_can_async_p,
to_is_async_p, to_async. Do not set to_xfer_partial field.
(default_xfer_partial): Simplify.
(current_xfer_partial): Remove.
(target_wait, target_resume): Simplify.
(find_default_can_async_p, find_default_is_async_p): Update.
(init_dummy_target): Don't set to_can_async_p, to_is_async_p,
to_xfer_partial, to_stopped_by_watchpoint,
to_stopped_data_address.
(target_store_registers): Simplify.
(forward_target_remove_breakpoint)
(forward_target_insert_breakpoint): Remove.
(target_remove_breakpoint, target_insert_breakpoint)
(debug_to_insert_breakpoint, debug_to_remove_breakpoint): Update.
* target.h (struct target_ops) <to_resume, to_wait,
to_store_registers, to_insert_breakpoint, to_remove_breakpoint,
to_stopped_by_watchpoint, to_stopped_data_address, to_can_async_p,
to_is_async_p, to_async, to_xfer_partial>: Add TARGET_DEFAULT
markup.
(forward_target_remove_breakpoint)
(forward_target_insert_breakpoint): Remove.
* record-btrace.c (record_btrace_remove_breakpoint): Delegate
directly.
(record_btrace_insert_breakpoint): Delegate directly.
This patch adds a new script, call make-target-delegates, which
auto-generates some target delegation code based on annotations in
target.h. This adds the new delegation macros, the new generated
file, and adds the necessary calls to the new generated functions to
target.c. It doesn't, however, add any actual annotations to the
target methods, leaving these for separate patches.
2014-02-19 Tom Tromey <tromey@redhat.com>
PR build/7701:
* target-delegates.c: New file.
* target.c: Include target-delegates.c.
(init_dummy_target): Call install_dummy_methods.
(complete_target_initialization): Call install_delegators.
* target.h (TARGET_DEFAULT_IGNORE, TARGET_DEFAULT_NORETURN)
(TARGET_DEFAULT_RETURN, TARGET_DEFAULT_FUNC): New defines.
* make-target-delegates: New file.
This patch adds find_target_at to determine whether a target appears
at a given stratum. This new function lets us clean up
find_record_target a bit, and is generally useful.
2014-02-19 Tom Tromey <tromey@redhat.com>
* record.c (find_record_target): Use find_target_at.
* target.c (find_target_at): New function.
* target.h (find_target_at): Declare.
This patch removes startup_gdbarch because it is not used anymore.
After this patch,
[PATCH] Multi-process + multi-arch: GDB
https://sourceware.org/ml/gdb-patches/2012-11/msg00228.html
global variable target_gdbarch is removed and gdbarch becomes
per-inferior. Then, we don't have to keep global variable
startup_gdbarch.
Rebuild GDB with all targets. Is it OK?
gdb:
2014-02-18 Yao Qi <yao@codesourcery.com>
PR gdb/16220
* gdbarch.sh: Remove startup_gdbarch.
* gdbarch.c: Regenerated.
* gdbarch.h: Likewise.
* Makefile.in (TESTS): New variable.
(expanded_tests, expanded_tests_or_none): New variables
(check-single): Pass $(expanded_tests_or_none) to runtest.
(check-parallel): Only run tests in $(TESTS) if non-empty.
(check/no-matching-tests-found): New rule.
* README: Document TESTS makefile variable.
The g10 is an rl78 variant which has fewer registers. Aside from the
obvious addition of a new register_name() function which omits
registers which the g10 lacks, this change also updates the
return_value() function with the conventions specified by the ABI for
finding and setting return values.
* rl78-tdep.c (rl78_g10_register_name): New function.
(rl78_return_value): Add g10 support.
(rl78_gdbarch_init): Register rl78_g10_register_name for the
g10.
A recent change (commit 3398af6aa3)
in gnu-nat.c causes the some missing-prototypes warnings,
../../../git/gdb/gnu-nat.c:1864:1: error: no previous prototype for 'S_proc_pid2task_reply' [-Werror=missing-prototypes]
../../../git/gdb/gnu-nat.c:1866:1: error: no previous prototype for 'S_proc_task2pid_reply' [-Werror=missing-prototypes]
../../../git/gdb/gnu-nat.c:1868:1: error: no previous prototype for 'S_proc_task2proc_reply' [-Werror=missing-prototypes]
A new macro ILL_RPC was added recently, which defines some external
functions. However, they are not declared and GCC complains about this.
This patch is to add the declarations of these external function in
macro ILL_RPC.
gdb:
2014-02-17 Yao Qi <yao@codesourcery.com>
* gnu-nat.c (ILL_RPC): Declare defined function.
../../../git/gdb/gnu-nat.c: In function 'gnu_read_inferior':
../../../git/gdb/gnu-nat.c:2282:3: error: pointer targets in passing argument 5 of 'vm_read' differ in signedness [-Werror=pointer-sign]
In file included from /home/yao/Software/hurd-toolchain/bin/../i686-pc-gnu/libc/usr/include/mach.h:38:0,
from ./nm.h:23,
from ../../../git/gdb/defs.h:500,
from ../../../git/gdb/gnu-nat.c:23:
/home/yao/Software/hurd-toolchain/bin/../i686-pc-gnu/libc/usr/include/mach/mach_interface.h:843:15: note: expected 'mach_msg_type_number_t *' but argument is of type 'int *'
../../../git/gdb/gnu-nat.c: In function 'gnu_write_inferior':
../../../git/gdb/gnu-nat.c:2339:4: error: pointer targets in passing argument 5 of 'vm_read' differ in signedness [-Werror=pointer-sign]
In file included from /home/yao/Software/hurd-toolchain/bin/../i686-pc-gnu/libc/usr/include/mach.h:38:0,
from ./nm.h:23,
from ../../../git/gdb/defs.h:500,
from ../../../git/gdb/gnu-nat.c:23:
/home/yao/Software/hurd-toolchain/bin/../i686-pc-gnu/libc/usr/include/mach/mach_interface.h:843:15: note: expected 'mach_msg_type_number_t *' but argument is of type 'int *'
This patch fixes these warnings.
gdb:
2014-02-17 Yao Qi <yao@codesourcery.com>
* gnu-nat.c (gnu_read_inferior): Change 'copy_count' type to
mach_msg_type_number_t.
(gnu_write_inferior): Likewise.
I've seen some -Wformat warnings when build native gdb for hurd.
../../../git/gdb/gnu-nat.c:2384:8: error: format '%x' expects argument of type 'unsigned int', but argument 2 has type 'vm_address_t' [-Werror=format]
../../../git/gdb/gnu-nat.c:2394:8: error: format '%x' expects argument of type 'unsigned int', but argument 2 has type 'vm_address_t' [-Werror=format]
../../../git/gdb/gnu-nat.c: In function 'steal_exc_port':
../../../git/gdb/gnu-nat.c:2898:5: error: format '%d' expects argument of type 'int', but argument 2 has type 'mach_port_t' [-Werror=format]
in usr/include/mach/port.h, we have:
typedef vm_offset_t mach_port_t;
and in usr/include/mach/machine/vm_types.h, we have:
typedef unsigned long vm_offset_t;
so this patch changes '%d' to '%lu' in format string for mach_port_t
variables.
Similarly, in usr/include/mach/std_types.h, we have,
typedef vm_offset_t vm_address_t;
this patch also changes '%x' to '%lx' in gnu_write_inferior.
gdb:
2014-02-17 Yao Qi <yao@codesourcery.com>
* gnu-nat.c (proc_get_exception_port): Use 'lu' insetad of 'd'
in format string.
(proc_steal_exc_port, make_proc, inf_set_pid): Likewise.
(inf_validate_procs, inf_signal): Likewise.
(S_exception_raise_request): Likewise.
(do_mach_notify_dead_name): Likewise.
(steal_exc_port): Likewise.
(gnu_read_inferior): Change 'copy_count''s type to
mach_msg_type_number_t.
(gnu_write_inferior): Likewise. Use 'lx' instead of 'x' in
format string.
If GDB has crashed then gdb_spawn_id still exists (although it does not work).
So my patch does not change anything. And also currently it will leave the
stale gdbserver running anyway.
In general if gdb_spawn_id does not exist then send_gdb + gdb_expect just do
not make sense anyway. So this patch just prevents the error in such case.
The killing of stale gdbserver could be improved multiple ways (also as
suggested by Pedro in the original thread) but that is IMO outside of the
scope of this patch. Apparently if there is no good response from GDB then
gdb_finish() should try to call gdb_start just to kill that gdbserver, IIUC.
gdb/testsuite/
2014-02-16 Jan Kratochvil <jan.kratochvil@redhat.com>
Fix "ERROR: no fileid for" in the testsuite.
* lib/gdb.exp (gdb_finish): Check gdb_spawn_id.
Message-ID: <20140206205814.GA18495@host2.jankratochvil.net>
In commit 98882a2651, STARTUP_WITH_SHELL was made
a runtime toggle, startup-with-shell. The Hurd code was missed to be adjusted;
it had a value hard-coded instead of using START_INFERIOR_TRAPS_EXPECTED. Fix
that, and also simplify gnu-nat's pending_execs handling from counting to just
a flag.
gdb/
* gnu-nat.c (struct inf): Change pending_execs member to a 1-bit
flag. Adjust all users; in particular...
(gnu_wait): ..., don't decrement its value in here...
(gnu_create_inferior): ..., and instead set the flag in here,
around the startup_inferior call, and call that one with
START_INFERIOR_TRAPS_EXPECTED.
gdb/
* reply_mig_hack.awk: In phase 5, keep going if we have not yet
collected the type check structures.
Based on a patch by David Michael <fedora.dm0@gmail.com>.
The %fsr register is not being properly accessed from gdb in
sparc64. This is because sparc64_supply_fpregset and
sparc64_collect_fpregset are using the offsets from the
sparc32_bsd_fpregset constant instead of sparc64_bsd_fpregset.
This patch fixes this by registering the proper offsets for
sparc64-linux targets.
2014-02-14 Jose E. Marchesi <jose.marchesi@oracle.com>
* sparc64-linux-nat.c (_initialize_sparc64_linux_nat): Register
the proper offsets to access fpregset_t.
* gdb.dwarf2/Makefile.in (EXECUTABLES): Add dwp-symlink.
(MISCELLANEOUS): New variable.
(clean): rm -rf $(MISCELLANEOUS).
* gdb.dwarf2/dwp-symlink.exp: Test the case where the executable and
dwp live in the same directory as symlinks, with each symlink pointed
to a differently named file in a different directory.
This updates all the comments in rsp-low.[ch], now that the
unification has been completed.
2014-02-12 Tom Tromey <tromey@redhat.com>
* common/rsp-low.c: Update comments.
* common/rsp-low.h: Update comments.
unhexify and hex2bin are identical, so this removes unhexify. The
particular choice of which to keep was made on the basis of
parallelism with the earlier patch that removed hexify.
2014-02-12 Tom Tromey <tromey@redhat.com>
* common/rsp-low.h (unhexify): Don't declare.
* common/rsp-low.c (unhexify): Remove.
2014-02-12 Tom Tromey <tromey@redhat.com>
* server.c (handle_query, handle_v_run): Use hex2bin, not
unhexify.
* tracepoint.c (cmd_qtdpsrc, cmd_qtdv, cmd_qtnotes): Likewise.
convert_int_to_ascii is identical to bin2hex. This removes the
former. In this case I made the choice of which to keep on the basis
that I consider the name bin2hex to be superior to
convert_int_to_ascii.
2014-02-12 Tom Tromey <tromey@redhat.com>
* common/rsp-low.h (convert_int_to_ascii): Don't declare.
* common/rsp-low.c (convert_int_to_ascii): Remove.
2014-02-12 Tom Tromey <tromey@redhat.com>
* ax.c (gdb_unparse_agent_expr): Use bin2hex, not
convert_int_to_ascii.
* regcache.c (registers_to_string, collect_register_as_string):
Likewise.
* remote-utils.c (look_up_one_symbol, relocate_instruction):
Likewise.
* server.c (process_serial_event): Likewise.
* tracepoint.c (cmd_qtstatus, response_source, response_tsv)
(cmd_qtbuffer, cstr_to_hexstr): Likewise.
This removes hexify in favor of bin2hex.
The choice of which to keep was arbitrary.
2014-02-12 Tom Tromey <tromey@redhat.com>
* common/rsp-low.h (hexify): Don't declare.
* common/rsp-low.c (hexify): Remove.
2014-02-12 Tom Tromey <tromey@redhat.com>
* remote-utils.c (look_up_one_symbol, monitor_output): Use
bin2hex, not hexify.
* tracepoint.c (cmd_qtstatus): Likewise.
hexify had the same issue as bin2hex; and the fix is the same.
2014-02-12 Tom Tromey <tromey@redhat.com>
* common/rsp-low.c (hexify): Never take strlen of argument.
2014-02-12 Tom Tromey <tromey@redhat.com>
* remote-utils.c (monitor_output): Pass explicit length to
hexify.
Currently bin2hex may call strlen if the length argument is zero.
This prevents some function unification; and also it seems cleaner to
me not to have a special meaning for a zero length.
2014-02-12 Tom Tromey <tromey@redhat.com>
* common/rsp-low.c (bin2hex): Never take strlen of argument.
* remote.c (extended_remote_run, remote_rcmd)
(remote_download_trace_state_variable, remote_save_trace_data)
(remote_set_trace_notes): Update.
* tracepoint.c (encode_source_string, tfile_write_status)
(tfile_write_uploaded_tsv): Update.
This moves various low-level remote serial protocol bits into
common/rsp-low.[ch].
This is as close to a pure move as possible. There are some
redundancies remaining but those will be dealt with in a subsequent
patch.
Note that the two variants of remote_escape_output disagreed on the
treatment of "*". On the theory that quoting cannot hurt but the
absence possibly can, I chose the gdbserver variant to be the
canonical one.
2014-02-12 Tom Tromey <tromey@redhat.com>
* tracepoint.c: Include rsp-low.h.
* remote.h (hex2bin, bin2hex, unpack_varlen_hex): Don't declare.
* remote.c: Include rsp-low.h.
(hexchars, ishex, unpack_varlen_hex, pack_nibble, pack_hex_byte)
(fromhex, hex2bin, tohex, bin2hex, remote_escape_output)
(remote_unescape_input): Move to common/rsp-low.c.
* common/rsp-low.h: New file.
* common/rsp-low.c: New file.
* Makefile.in (SFILES): Add common/rsp-low.c.
(HFILES_NO_SRCDIR): Add common/rsp-low.h.
(COMMON_OBS): Add rsp-low.o.
(rsp-low.o): New target.
2014-02-12 Tom Tromey <tromey@redhat.com>
* tracepoint.c: Include rsp-low.h.
* server.c: Include rsp-low.h.
* remote-utils.h (convert_ascii_to_int, convert_int_to_ascii)
(unhexify, hexify, remote_escape_output, unpack_varlen_hex): Don't
declare.
* remote-utils.c: Include rsp-low.h.
(fromhex, hexchars, ishex, unhexify, tohex, hexify)
(remote_escape_output, remote_unescape_input, unpack_varlen_hex)
(convert_int_to_ascii, convert_ascii_to_int): Move to
common/rsp-low.c.
* regcache.c: Include rsp-low.h.
* ax.c: Include rsp-low.h.
* Makefile.in (SFILES): Add common/rsp-low.c.
(OBS): Add rsp-low.o.
(rsp-low.o): New target.
Since this change:
2014-02-12 Sanimir Agovic <sanimir.agovic@intel.com>
* nios2-tdep.c (nios2_stub_frame_base): Remove global.
nios2-tdep hasn't built:
../../binutils-gdb/gdb/nios2-tdep.c:1337:1: error: ‘nios2_stub_frame_base_address’ defined but not used [-Werror=unused-function]
This patch removes the offending function.
2014-02-12 Tom Tromey <tromey@redhat.com>
* nios2-tdep.c (nios2_stub_frame_base_address): Remove.
At least on OpenBSD PT_IO/PIOD_READ_AUXV can return sucessfully without
transferring any bytes. Arguably a kernel bug, but interpreting this as EOF
seems sensible.
gdb/ChangeLog:
* inf-ptrace.c (inf_ptrace_xfer_partial): Return TARGET_XFER_EOF
if a PT_IO ptrace request returns sucessfully but indicates that 0
bytes were transferred.
Ports for Hardvard architectures will typically have in their
pointer_to_address hook a check for TYPE_CODE_SPACE in their
"pointer_to_address" gdbarch method. E.g., rl78's:
/* Is it a code address? */
if (TYPE_CODE (TYPE_TARGET_TYPE (type)) == TYPE_CODE_FUNC
|| TYPE_CODE (TYPE_TARGET_TYPE (type)) == TYPE_CODE_METHOD
|| TYPE_CODE_SPACE (TYPE_TARGET_TYPE (type))
|| TYPE_LENGTH (type) == 4)
return rl78_make_instruction_address (addr);
else
return rl78_make_data_address (addr);
The avr port is similar.
The vtable type is a struct type gdb itself bakes. Being neither a
function, nor a method, and absent explicit flagging as residing in
code space, ends up being considered data.
This patch marks the type as code when it is created, on the theory
that this is needed for all Hardvard architectures. I believe this
should make no difference on archs with flat address space. Testing
on x86-64 GNU/Linux shows no changes.
gdb/
2014-02-12 Pedro Alves <palves@redhat.com>
Kevin Buettner <kevinb@redhat.com>
* gnu-v3-abi.c (build_gdb_vtable_type): Return a type marked with
TYPE_INSTANCE_FLAG_CODE_SPACE.
Kevin Buettner, at
<https://sourceware.org/ml/gdb-patches/2014-02/msg00338.html>, writes,
re. rl78:
This patch, for rl78 using the default multilib, fixes 5 failures in
gdb.cp/casts.exp, 2 failures in gdb.cp/class2.exp, 115 failures in
gdb.mi/mi-var-rtti.exp, and 2 failures in gdb.python/py-value.exp.
It introduces 9 failures (regressions) in gdb.mi/mi-var-rtti.exp.
One of the regressions is:
FAIL: gdb.mi/mi-var-rtti.exp: list children of s.public in type_update_when_use_rtti
The relevant lines from the log file from a pre-patch test run are as
follows:
-var-list-children S.public
^done,numchild="1",children=[child={name="S.public.ptr",exp="ptr",numchild="1",type="Base *",thread-id="1"}],has_more="0"
(gdb)
PASS: gdb.mi/mi-var-rtti.exp: list children of s.public in type_update_when_use_rtti
Expecting: \^done,numchild=".*",children=\[child={name="S.public.ptr.public",exp="public",numchild="1"(,thread-id="[0-9]+")?}.*\],has_more="0"
Expecting: ^(-var-list-children S\.public\.ptr [
]+)?(\^done,numchild=".*",children=\[child={name="S.public.ptr.public",exp="public",numchild="1"(,thread-id="[0-9]+")?}.*\],has_more="0"[
]+[(]gdb[)]
[ ]*)
The same set of lines for the failing (post-patch) run are instead:
-var-list-children S.public
&"warning: can't find linker symbol for virtual table for `Base' value\n"
&"warning: found `typeinfo for __cxxabiv1::__vmi_class_type_info' instead\n"
&"warning: can't find linker symbol for virtual table for `Base' value\n"
&"warning: found `typeinfo for __cxxabiv1::__vmi_class_type_info' instead\n"
&"warning: can't find linker symbol for virtual table for `Base' value\n"
&"warning: found `typeinfo for __cxxabiv1::__vmi_class_type_info' instead\n"
^done,numchild="1",children=[child={name="S.public.ptr",exp="ptr",numchild="1",type="Base *",thread-id="1"}],has_more="0"
(gdb)
FAIL: gdb.mi/mi-var-rtti.exp: list children of s.public in type_update_when_use_rtti
Expecting: \^done,numchild=".*",children=\[child={name="S.public.ptr.public",exp="public",numchild="1"(,thread-id="[0-9]+")?}.*\],has_more="0"
Expecting: ^(-var-list-children S\.public\.ptr [
]+)?(\^done,numchild=".*",children=\[child={name="S.public.ptr.public",exp="public",numchild="1"(,thread-id="[0-9]+")?}.*\],has_more="0"[
]+[(]gdb[)]
[ ]*)
Note that the difference between these are the warnings regarding
linker symbols. Aside from the warnings, the result is the same.
I.e. gdb produces the correct answer despite the warnings. The
reason for the other 8 failures is the same: the test harness is not
expecting these warnings.
I spent some time (a while ago) looking at the reason for these
warnings. As I recall, we are now getting further along in the type
resolution process than we were without my patch. I.e. for the
example above, the code in question never got to the point where it
was looking for the specified linker symbol. So it seems to me that,
at worst, my patch exposes some other problem, but is not directly the
cause of the problem.
'info registers ccr' corrupts memory.
Debugging gdb under Valgrind, we see:
(gdb) info registers ccr
==23225== Invalid write of size 1
==23225== at 0x4A0A308: memcpy@@GLIBC_2.14 (mc_replace_strmem.c:881)
==23225== by 0x52D334: regcache_raw_read (regcache.c:625)
==23225== by 0x45E4D8: h8300_pseudo_register_read (h8300-tdep.c:1171)
==23225== by 0x5B694B: gdbarch_pseudo_register_read (gdbarch.c:1926)
==23225== by 0x52DADB: regcache_cooked_read (regcache.c:740)
==23225== by 0x52DC10: regcache_cooked_read_value (regcache.c:765)
==23225== by 0x68CA41: sentinel_frame_prev_register (sentinel-frame.c:52)
==23225== by 0x6B80CB: frame_unwind_register_value (frame.c:1105)
==23225== by 0x6B7C97: frame_register_unwind (frame.c:1010)
==23225== by 0x6B7F73: frame_unwind_register (frame.c:1064)
==23225== by 0x6B8359: frame_unwind_register_signed (frame.c:1162)
==23225== by 0x6B8396: get_frame_register_signed (frame.c:1169)
==23225== Address 0x4f7b031 is 0 bytes after a block of size 1 alloc'd
==23225== at 0x4A06B0F: calloc (vg_replace_malloc.c:593)
==23225== by 0x6EB754: xcalloc (common-utils.c:91)
==23225== by 0x6EB793: xzalloc (common-utils.c:101)
==23225== by 0x53A782: allocate_value_contents (value.c:854)
==23225== by 0x53A7B4: allocate_value (value.c:864)
==23225== by 0x52DBC8: regcache_cooked_read_value (regcache.c:757)
==23225== by 0x68CA41: sentinel_frame_prev_register (sentinel-frame.c:52)
==23225== by 0x6B80CB: frame_unwind_register_value (frame.c:1105)
==23225== by 0x6B7C97: frame_register_unwind (frame.c:1010)
==23225== by 0x6B7F73: frame_unwind_register (frame.c:1064)
==23225== by 0x6B8359: frame_unwind_register_signed (frame.c:1162)
==23225== by 0x6B8396: get_frame_register_signed (frame.c:1169)
==23225==
ccr 0x00 0 I-0 UI-0 H-0 U-0 N-0 Z-0 V-0 C-0 u> u>= != >= >
(gdb)
This bit:
==23225== Invalid write of size 1
==23225== at 0x4A0A308: memcpy@@GLIBC_2.14 (mc_replace_strmem.c:881)
==23225== by 0x52D334: regcache_raw_read (regcache.c:625)
==23225== by 0x45E4D8: h8300_pseudo_register_read (h8300-tdep.c:1171)
shows the problem. The CCR pseudo register has type length of 1,
while the corresponding CCR raw register has a length of 2 or 4
(depending on mode). In
sim/h8300/compile.c:sim_{fetch|store}_register we see that the sim
also treats those raw registers (CCR/EXR) as 2 or 4 bytes length.
gdb/
2014-02-12 Pedro Alves <palves@redhat.com>
* h8300-tdep.c (pseudo_from_raw_register)
(raw_from_pseudo_register): New functions.
(h8300_pseudo_register_read, h8300_pseudo_register_write): Use
them.
Currently, printing the H8/300 ccr register when debugging with the
sim is broken:
(gdb) target sim
...
(gdb) load
...
(gdb) start
...
Breakpoint 1, foo (i=0x0 <foo>) at main.c:4
4 {
(gdb) info registers ccr
Register 13 is not available
'13' is the ccr pseudo-register. This pseudo-register provides an
8-bit view into the raw ccr register (regno=8).
The problem is that the H8/300 port does not define a
register_sim_regno gdbarch hook, and thus when fetching the raw
register off of the sim, we end up in legacy_register_sim_regno trying
to figure out the sim register number for the raw CCR register:
int
legacy_register_sim_regno (struct gdbarch *gdbarch, int regnum)
{
/* Only makes sense to supply raw registers. */
gdb_assert (regnum >= 0 && regnum < gdbarch_num_regs (gdbarch));
/* NOTE: cagney/2002-05-13: The old code did it this way and it is
suspected that some GDB/SIM combinations may rely on this
behavour. The default should be one2one_register_sim_regno
(below). */
if (gdbarch_register_name (gdbarch, regnum) != NULL
&& gdbarch_register_name (gdbarch, regnum)[0] != '\0')
return regnum;
else
return LEGACY_SIM_REGNO_IGNORE;
}
Because the raw ccr register does not have a name (so that it is
hidden from the user), that returns LEGACY_SIM_REGNO_IGNORE. That
means that we never actually read the value of the raw ccr register.
Before the <unavailable> support, this must have meant that ccr was
_always_ read as 0... At least, I'm not seeing how this ever worked.
The fix for that is adding a gdbarch_register_sim_regno hook that maps
all raw registers. Looking at sim/h8300/sim-main.h, I believe the
sim's register numbers are compatible with gdb's, so no actual
convertion is necessary.
gdb/
2014-02-12 Pedro Alves <palves@redhat.com>
* h8300-tdep.c (h8300_register_sim_regno): New function.
(h8300_gdbarch_init): Install h8300_register_sim_regno as
gdbarch_register_sim_regno hook.
Double float complex objects are not 16-byte aligned in either
gcc or solaris studio. This patch makes gdb to not align double
float complex arguments in the dummy frame when calling a
function.
2014-02-11 Jose E. Marchesi <jose.marchesi@oracle.com>
* sparc64-tdep.c (sparc64_store_arguments): Do not align complex
double float arguments to 16-byte in the argument slots.
* configure.ac: Don't crash if pkg-config is not found and guile
wasn't explicitly requested. Use AC_MSG_ERROR instead of AC_ERROR
in guile checks.
* configure: Regenerate.
This patch does the conversion of to_xfer_partial from
LONGEST (*to_xfer_partial) (struct target_ops *ops,
enum target_object object, const char *annex,
gdb_byte *readbuf, const gdb_byte *writebuf,
ULONGEST offset, ULONGEST len);
to
enum target_xfer_status (*to_xfer_partial) (struct target_ops *ops,
enum target_object object, const char *annex,
gdb_byte *readbuf, const gdb_byte *writebuf,
ULONGEST offset, ULONGEST len, ULONGEST *xfered_len);
It changes to_xfer_partial return the transfer status and the transfered
length by *XFERED_LEN. Generally, the return status has three stats,
- TARGET_XFER_OK,
- TARGET_XFER_EOF,
- TARGET_XFER_E_XXXX,
See the comments to them in 'enum target_xfer_status'. Note that
Pedro suggested not name TARGET_XFER_DONE, as it is confusing,
compared with "TARGET_XFER_OK". We finally name it TARGET_XFER_EOF.
With this change, GDB core can handle unavailable data in a convenient
way.
The rationale behind this change was mentioned here
https://sourceware.org/ml/gdb-patches/2013-10/msg00761.html
Consider an object/value like this:
0 100 150 200 512
DDDDDDDDDDDxxxxxxxxxDDDDDD...DDIIIIIIIIIIII..III
where D is valid data, and xxx is unavailable data, and I is beyond
the end of the object (Invalid). Currently, if we start the
xfer at 0, requesting, say 512 bytes, we'll first get back 100 bytes.
The xfer machinery then retries fetching [100,512), and gets back
TARGET_XFER_E_UNAVAILABLE. That's sufficient when you're either
interested in either having the whole of the 512 bytes available,
or erroring out. But, in this scenario, we're interested in
the data at [150,512). The problem is that the last
TARGET_XFER_E_UNAVAILABLE gives us no indication where to
start the read next. We'd need something like:
get me [0,512) >>>
<<< here's [0,100), *xfered_len is 100, returns TARGET_XFER_OK
get me [100,512) >>> (**1)
<<< [100,150) is unavailable, *xfered_len is 50, return TARGET_XFER_E_UNAVAILABLE.
get me [150,512) >>>
<<< here's [150,200), *xfered_len is 50, return TARGET_XFER_OK.
get me [200,512) >>>
<<< no more data, return TARGET_XFER_EOF.
This naturally implies pushing down the decision of whether
to return TARGET_XFER_E_UNAVAILABLE or something else
down to the target. (Which kinds of leads back to tfile
itself reading from RO memory from file (though we could
export a function in exec.c for that that tfile delegates to,
instead of re-adding the old code).
Beside this change, we also add a macro TARGET_XFER_STATUS_ERROR_P to
check whether a status is an error or not, to stop using "status < 0".
This patch also eliminates the comparison between status and 0.
No target implementations to to_xfer_partial adapts this new
interface. The interface still behaves as before.
gdb:
2014-02-11 Yao Qi <yao@codesourcery.com>
* target.h (enum target_xfer_error): Rename to ...
(enum target_xfer_status): ... it. New. All users updated.
(enum target_xfer_status) <TARGET_XFER_OK>, <TARGET_XFER_EOF>:
New.
(TARGET_XFER_STATUS_ERROR_P): New macro.
(target_xfer_error_to_string): Remove declaration.
(target_xfer_status_to_string): Declare.
(target_xfer_partial_ftype): Adjust it.
(struct target_ops) <to_xfer_partial>: Return
target_xfer_status. Add argument xfered_len. Update
comments.
* target.c (target_xfer_error_to_string): Rename to ...
(target_xfer_status_to_string): ... it. New. All callers
updated.
(target_read_live_memory): Likewise. Call target_xfer_partial
instead of target_read.
(memory_xfer_live_readonly_partial): Return
target_xfer_status. Add argument xfered_len.
(raw_memory_xfer_partial): Likewise.
(memory_xfer_partial_1): Likewise.
(memory_xfer_partial): Likewise.
(target_xfer_partial): Likewise. Check *XFERED_LEN is set
properly. Update debug message.
(default_xfer_partial, current_xfer_partial): Likewise.
(target_write_partial): Likewise.
(target_read_partial): Likewise. All callers updated.
(read_whatever_is_readable): Likewise.
(target_write_with_progress): Likewise.
(target_read_alloc_1): Likewise.
* aix-thread.c (aix_thread_xfer_partial): Likewise.
* auxv.c (procfs_xfer_auxv): Likewise.
(ld_so_xfer_auxv, memory_xfer_auxv): Likewise.
* bfd-target.c (target_bfd_xfer_partial): Likewise.
* bsd-kvm.c (bsd_kvm_xfer_partial): Likewise.
* bsd-uthread.c (bsd_uthread_xfer_partia): Likewise.
* corefile.c (read_memory): Adjust.
* corelow.c (core_xfer_partial): Likewise.
* ctf.c (ctf_xfer_partial): Likewise.
* darwin-nat.c (darwin_read_dyld_info): Likewise. All callers
updated.
(darwin_xfer_partial): Likewise.
* exec.c (section_table_xfer_memory_partial): Likewise. All
callers updated.
(exec_xfer_partial): Likewise.
* exec.h (section_table_xfer_memory_partial): Update
declaration.
* gnu-nat.c (gnu_xfer_memory): Likewise. Assert 'res' is not
negative.
(gnu_xfer_partial): Likewise.
* ia64-hpux-nat.c (ia64_hpux_xfer_memory_no_bs): Likewise.
(ia64_hpux_xfer_memory, ia64_hpux_xfer_uregs): Likewise.
(ia64_hpux_xfer_solib_got): Likewise.
* inf-ptrace.c (inf_ptrace_xfer_partial): Likewise. Change
type of 'partial_len' to ULONGEST.
* inf-ttrace.c (inf_ttrace_xfer_partial): Likewise.
* linux-nat.c (linux_xfer_siginfo ): Likewise.
(linux_nat_xfer_partial): Likewise.
(linux_proc_xfer_partial, linux_xfer_partial): Likewise.
(linux_proc_xfer_spu, linux_nat_xfer_osdata): Likewise.
* monitor.c (monitor_xfer_memory): Likewise.
(monitor_xfer_partial): Likewise.
* procfs.c (procfs_xfer_partial): Likewise.
* record-btrace.c (record_btrace_xfer_partial): Likewise.
* record-full.c (record_full_xfer_partial): Likewise.
(record_full_core_xfer_partial): Likewise.
* remote-sim.c (gdbsim_xfer_memory): Likewise.
(gdbsim_xfer_partial): Likewise.
* remote.c (remote_write_bytes_aux): Likewise. All callers
updated.
(remote_write_bytes, remote_read_bytes): Likewise. All
callers updated.
(remote_flash_erase): Likewise. All callers updated.
(remote_write_qxfer): Likewise. All callers updated.
(remote_read_qxfer): Likewise. All callers updated.
(remote_xfer_partial): Likewise.
* rs6000-nat.c (rs6000_xfer_partial): Likewise.
(rs6000_xfer_shared_libraries): Likewise.
* sol-thread.c (sol_thread_xfer_partial): Likewise.
(sol_thread_xfer_partial): Likewise.
* sparc-nat.c (sparc_xfer_wcookie): Likewise.
(sparc_xfer_partial): Likewise.
* spu-linux-nat.c (spu_proc_xfer_spu): Likewise. All callers
updated.
(spu_xfer_partial): Likewise.
* spu-multiarch.c (spu_xfer_partial): Likewise.
* tracepoint.c (tfile_xfer_partial): Likewise.
* windows-nat.c (windows_xfer_memory): Likewise.
(windows_xfer_shared_libraries): Likewise.
(windows_xfer_partial): Likewise.
* valprint.c: Replace 'target_xfer_error' with
'target_xfer_status' in comments.
gdb/ChangeLog:
2014-02-11 Simon Marchi <simon.marchi@ericsson.com> (tiny patch)
Checked in by Joel Brobecker <brobecker@adacore.com>.
* mi/mi-main.c (mi_cmd_data_write_memory_bytes): Fix comment.
There are two failures in the gnu-ifunc.exp test on ARM. These are
due to the failure to resolve the correct target function when
attempting to breakpoint a GNU ifunc resolved function:
(gdb) break gnu_ifunc
Breakpoint 4 at gnu-indirect-function resolver at 0x2aacb5a2
when gnu_ifunc has been resolved this should actually be:
(gdb) break gnu_ifunc
Breakpoint 4 at 0x868c
There are two reasons for this. The first is that ARM does not have a
separate .got.plt section so looking this up will always fail. The second
is that the Thumb bit needs to be stripped from the address to allow
it to be reliably compared when inserting into the ifunc cache.
Tested with no regressions on arm-linux-gnueabihf and
x86_64-unknown-linux-gnu.
gdb/ChangeLog:
2014-02-10 Will Newton <will.newton@linaro.org>
* elfread.c (elf_rel_plt_read): Look for a .got section if
looking up .got.plt fails.
(elf_gnu_ifunc_resolve_by_got): Call gdbarch_addr_bits_remove
on address passed to elf_gnu_ifunc_record_cache.
(elf_gnu_ifunc_resolve_addr): Likewise.
(elf_gnu_ifunc_resolver_return_stop): Likewise.
watchpoint_update and watchpoint_cond avoid checking for
watchpoints when we are located at a function epilogue in the
current frame. This is done in order to avoid using corrupted
local registers and unwinding a corrupted/destroyed stack.
The code determining whether we are in a function epilogue is
provided by the backends via the gdbarch_in_function_epilogue_p
hook. This commit adds such a hook for sparc64 targets.
2014-02-10 Jose E. Marchesi <jose.marchesi@oracle.com>
* sparc-tdep.c (sparc_in_function_epilogue_p): New function.
(X_RETTURN): New macro.
* sparc-tdep.h: sparc_in_function_epilogue_p prototype.
* sparc64-tdep.c (sparc64_init_abi): Hook
sparc_in_function_epilogue_p.
This commit fixes a compile error introduced by my previous commit.
Checked in as obvious.
2014-02-10 Gary Benson <gbenson@redhat.com>
* symfile-debug.c (debug_qf_expand_symtabs_matching):
Rename name_matcher to symbol_matcher.
Commit 206f2a5777 added two new
typedefs, but did not update debug_qf_expand_symtabs_matching
to use them. This patch fixes this.
Checked in as obvious.
2014-02-10 Gary Benson <gbenson@redhat.com>
* symfile-debug.c (debug_qf_expand_symtabs_matching):
Use expand_symtabs_file_matcher_ftype and
expand_symtabs_symbol_matcher_ftype.
This patch moves the Ada symbol cache to per-program-space data.
gdb/ChangeLog:
* ada-lang.c (struct cache_entry, HASH_SIZE): Move definition up.
(struct ada_symbol_cache): New.
(ada_free_symbol_cache): Forward declare.
(struct ada_pspace_data): New.
(ada_pspace_data_handle): New static global.
(get_ada_pspace_data, ada_pspace_data_cleanup)
(ada_init_symbol_cache, ada_free_symbol_cache): New functions.
(cache_space, cache): Delete, now folded inside struct
ada_pspace_data.
(ada_get_symbol_cache): New function.
(ada_clear_symbol_cache, find_entry, cache_symbol): Adjust
implementation.
(_initialize_ada_language): Remove initialization of cache_space.
Move call to observer_attach_inferior_exit up, grouping it
with the other observer registrations inside this function.
Rename command to be more general. Add call to
register_program_space_data_with_cleanup.
This patch is mostly cosmetic, avoiding us to use the same callback
names as in ada-lang.c.
gdb/ChangeLog:
* ada-tasks.c (ada_tasks_new_objfile_observer): Renames
ada_new_objfile_observer.
(ada_tasks_normal_stop_observer): Renames ada_normal_stop_observer.
(_initialize_tasks): Update uses of ada_new_objfile_observer
and ada_tasks_normal_stop_observer.
Consider the following code:
type Color is (Black, Red, Green, Blue, White);
type Primary_Table is array (Color range Red .. Blue) of Boolean;
Prim : Primary_Table := (True, False, False);
GDB prints the length of arrays in a fairly odd way:
(gdb) p prim'length
$2 = blue
The length returned should be an integer, not the array index type,
and this patch fixes this.
gdb/ChangeLog:
* ada-lang.c (ada_evaluate_subexp): Set the type of the value
returned by the 'Length attribute to integer.
testsuite/ChangeLog:
* gdb.ada/tick_length_array_enum_idx: New testcase.
This bit was meant to be merged with the following patch:
commit 3d9434b5dd
Subject: [Ada] Add a symbol lookup cache
... but I forgot :-(. This causes the cache to be undefined, and
with a bit of (bad) luck:
% gdb
(gdb) set lang ada
(gdb) set $xxx := 1
[SEGV]
gdb/ChangeLog:
* ada-lang.c (_initialize_ada_language): Initialize
cache_space obstack.
This patch implements the caching mechanism alluded to in a comment
next to some stubbed functions.
gdb/ChangeLog:
* ada-lang.c (HASH_SIZE): New macro.
(struct cache_entry): New type.
(cache_space, cache): New static globals.
(ada_clear_symbol_cache, find_entry): New functions.
(lookup_cached_symbol, cache_symbol): Implement.
(ada_new_objfile_observer, ada_free_objfile_observer): New.
(_initialize_ada_language): Attach ada_new_objfile_observer
and ada_free_objfile_observer.
This patch series constifies a number of struct block * parameters.
gdb/ChangeLog:
* ada-lang.c (ada_add_block_symbols, add_defn_to_vec)
(lookup_cached_symbol, ada_add_local_symbols): Add "const" to
struct block * parameter.
(ada_lookup_symbol_list_worker): Constify local variable "block".
Remove cast which is no longer necessary.
Tests in gdb.gdb fail because directory gdb/testsuite/gdb.gdb doesn't
exist in build tree. This patch appends gdb.gdb/Makefile in AC_OUTPUT,
and adds new Makefile.in in gdb.gdb, so that directory gdb.gdb can be
created during configure.
With this patch applied, tests under gdb.gdb can be run,
$ make check RUNTESTFLAGS='--directory=gdb.gdb'
Using /usr/share/dejagnu/baseboards/unix.exp as board description file for target.
Using /usr/share/dejagnu/config/unix.exp as generic interface file for target.
Using ../../../../git/gdb/testsuite/config/unix.exp as tool-and-target-specific interface file.
Running ../../../../git/gdb/testsuite/gdb.gdb/complaints.exp ...
Running ../../../../git/gdb/testsuite/gdb.gdb/observer.exp ...
Running ../../../../git/gdb/testsuite/gdb.gdb/python-interrupts.exp ...
FAIL: gdb.gdb/python-interrupts.exp: signal SIGINT
Running ../../../../git/gdb/testsuite/gdb.gdb/python-selftest.exp ...
FAIL: gdb.gdb/python-selftest.exp: call catch_command_errors(execute_command, "python print 5", 0, RETURN_MASK_ALL)
Running ../../../../git/gdb/testsuite/gdb.gdb/selftest.exp ...
Running ../../../../git/gdb/testsuite/gdb.gdb/xfullpath.exp ...
=== gdb Summary ===
gdb/testsuite:
2014-02-10 Yao Qi <yao@codesourcery.com>
PR testsuite/16543
* configure.ac: Append gdb.gdb/Makefile in AC_OUTPUT.
* configure: Regenerated.
* Makefile.in: New file.
Revert this patch (which I approved, mea culpa).
2014-02-08 Mark Kettenis <kettenis@gnu.org>
* Makefile.in (all-lib): Remove.
($(LIBGNU) $(GNULIB_H)): Replace with gits of remove all-lib target.
As design, =breakpoint-modified isn't emitted when breakpoints are modified
by MI commands. This patch is to add tests for this.
gdb/testsuite:
2014-02-08 Yao Qi <yao@codesourcery.com>
* gdb.mi/mi-breakpoint-changed.exp (test_insert_delete_modify): Test
that no =breakpoint-modified is emitted when breakpoints are
modified through MI commands.
Say:
<stopped at a breakpoint in thread 2>
(gdb) thread 3
(gdb) step
The above triggers the prepare_to_proceed/deferred_step_ptid process,
which switches back to thread 2, to step over its breakpoint before
getting back to thread 3 and "step" it.
If while stepping over the breakpoint in thread 2, a signal arrives,
and it is set to pass/nostop, we'll set a step-resume breakpoint at
the supposed signal-handler resume address, and call keep_going. The
problem is that we were supposedly stepping thread 3, and that
keep_going delivers a signal to thread 2, and due to scheduler-locking
off, resumes everything else, _including_ thread 3, the thread we want
stepping. This means that we lose control of thread 3 until the next
event, when we stop everything. The end result for the user, is that
GDB lost control of the "step".
Here's the current infrun debug output of the above, with the testcase
in the patch below:
infrun: clear_proceed_status_thread (Thread 0x2aaaab8f5700 (LWP 11663))
infrun: clear_proceed_status_thread (Thread 0x2aaaab6f4700 (LWP 11662))
infrun: clear_proceed_status_thread (Thread 0x2aaaab4f2b20 (LWP 11659))
infrun: proceed (addr=0xffffffffffffffff, signal=144, step=1)
infrun: prepare_to_proceed (step=1), switched to [Thread 0x2aaaab6f4700 (LWP 11662)]
infrun: resume (step=1, signal=0), trap_expected=1, current thread [Thread 0x2aaaab6f4700 (LWP 11662)] at 0x40098f
infrun: wait_for_inferior ()
infrun: target_wait (-1, status) =
infrun: 11659 [Thread 0x2aaaab6f4700 (LWP 11662)],
infrun: status->kind = stopped, signal = SIGUSR1
infrun: infwait_normal_state
infrun: TARGET_WAITKIND_STOPPED
infrun: stop_pc = 0x40098f
infrun: random signal 30
Program received signal SIGUSR1, User defined signal 1.
infrun: signal arrived while stepping over breakpoint
infrun: inserting step-resume breakpoint at 0x40098f
infrun: resume (step=0, signal=30), trap_expected=0, current thread [Thread 0x2aaaab6f4700 (LWP 11662)] at 0x40098f
^^^ this is a wildcard resume.
infrun: prepare_to_wait
infrun: target_wait (-1, status) =
infrun: 11659 [Thread 0x2aaaab6f4700 (LWP 11662)],
infrun: status->kind = stopped, signal = SIGTRAP
infrun: infwait_normal_state
infrun: TARGET_WAITKIND_STOPPED
infrun: stop_pc = 0x40098f
infrun: BPSTAT_WHAT_STEP_RESUME
infrun: resume (step=1, signal=0), trap_expected=1, current thread [Thread 0x2aaaab6f4700 (LWP 11662)] at 0x40098f
^^^ step-resume hit, meaning the handler returned, so we go back to stepping thread 3.
infrun: prepare_to_wait
infrun: target_wait (-1, status) =
infrun: 11659 [Thread 0x2aaaab6f4700 (LWP 11662)],
infrun: status->kind = stopped, signal = SIGTRAP
infrun: infwait_normal_state
infrun: TARGET_WAITKIND_STOPPED
infrun: stop_pc = 0x40088b
infrun: switching back to stepped thread
infrun: Switching context from Thread 0x2aaaab6f4700 (LWP 11662) to Thread 0x2aaaab8f5700 (LWP 11663)
infrun: resume (step=1, signal=0), trap_expected=0, current thread [Thread 0x2aaaab8f5700 (LWP 11663)] at 0x400938
infrun: prepare_to_wait
infrun: target_wait (-1, status) =
infrun: 11659 [Thread 0x2aaaab8f5700 (LWP 11663)],
infrun: status->kind = stopped, signal = SIGTRAP
infrun: infwait_normal_state
infrun: TARGET_WAITKIND_STOPPED
infrun: stop_pc = 0x40093a
infrun: keep going
infrun: resume (step=1, signal=0), trap_expected=0, current thread [Thread 0x2aaaab8f5700 (LWP 11663)] at 0x40093a
infrun: prepare_to_wait
infrun: target_wait (-1, status) =
infrun: 11659 [Thread 0x2aaaab8f5700 (LWP 11663)],
infrun: status->kind = stopped, signal = SIGTRAP
infrun: infwait_normal_state
infrun: TARGET_WAITKIND_STOPPED
infrun: stop_pc = 0x40091e
infrun: stepped to a different line
infrun: stop_stepping
[Switching to Thread 0x2aaaab8f5700 (LWP 11663)]
69 (*myp) ++; /* set breakpoint child_two here */
^^^ we stopped at the wrong line. We still stepped a bit because the
test is running in a loop, and when we got back to stepping thread 3,
it happened to be in the stepping range. (The loop increments a
counter, and the test makes sure it increments exactly once. Without
the fix, the counter increments a bunch, since the user-stepped thread
runs free without GDB noticing.)
The fix is to switch to the stepping thread before continuing for the
step-resume breakpoint.
gdb/
2014-02-07 Pedro Alves <palves@redhat.com>
* infrun.c (handle_signal_stop) <signal arrives while stepping
over a breakpoint>: Switch back to the stepping thread.
gdb/testsuite/
2014-02-07 Pedro Alves <pedro@codesourcery.com>
Pedro Alves <palves@redhat.com>
* gdb.threads/step-after-sr-lock.c: New file.
* gdb.threads/step-after-sr-lock.exp: New file.
Currently on software single-step Linux targets we get:
(gdb) PASS: gdb.threads/stepi-random-signal.exp: before stepi: get hexadecimal valueof "$pc"
stepi
infrun: clear_proceed_status_thread (Thread 0x7ffff7fca700 (LWP 7073))
infrun: clear_proceed_status_thread (Thread 0x7ffff7fcb740 (LWP 7069))
infrun: proceed (addr=0xffffffffffffffff, signal=GDB_SIGNAL_DEFAULT, step=1)
infrun: resume (step=1, signal=GDB_SIGNAL_0), trap_expected=0, current thread [Thread 0x7ffff7fcb740 (LWP 7069)] at 0x400700
infrun: wait_for_inferior ()
infrun: target_wait (-1, status) =
infrun: 7069 [Thread 0x7ffff7fcb740 (LWP 7069)],
infrun: status->kind = stopped, signal = GDB_SIGNAL_TRAP
infrun: infwait_normal_state
infrun: TARGET_WAITKIND_STOPPED
infrun: stop_pc = 0x400704
infrun: software single step trap for Thread 0x7ffff7fcb740 (LWP 7069)
infrun: stepi/nexti
infrun: stop_stepping
44 while (counter != 0)
(gdb) FAIL: gdb.threads/stepi-random-signal.exp: stepi (no random signal)
Vs hardware-step targets:
(gdb) PASS: gdb.threads/stepi-random-signal.exp: before stepi: get hexadecimal valueof "$pc"
stepi
infrun: clear_proceed_status_thread (Thread 0x7ffff7fca700 (LWP 9565))
infrun: clear_proceed_status_thread (Thread 0x7ffff7fcb740 (LWP 9561))
infrun: proceed (addr=0xffffffffffffffff, signal=GDB_SIGNAL_DEFAULT, step=1)
infrun: resume (step=1, signal=GDB_SIGNAL_0), trap_expected=0, current thread [Thread 0x7ffff7fcb740 (LWP 9561)] at 0x400700
infrun: wait_for_inferior ()
infrun: target_wait (-1, status) =
infrun: 9561 [Thread 0x7ffff7fcb740 (LWP 9561)],
infrun: status->kind = stopped, signal = GDB_SIGNAL_CHLD
infrun: infwait_normal_state
infrun: TARGET_WAITKIND_STOPPED
infrun: stop_pc = 0x400700
infrun: random signal (GDB_SIGNAL_CHLD)
infrun: random signal, keep going
infrun: resume (step=1, signal=GDB_SIGNAL_CHLD), trap_expected=0, current thread [Thread 0x7ffff7fcb740 (LWP 9561)] at 0x400700
infrun: prepare_to_wait
infrun: target_wait (-1, status) =
infrun: 9561 [Thread 0x7ffff7fcb740 (LWP 9561)],
infrun: status->kind = stopped, signal = GDB_SIGNAL_TRAP
infrun: infwait_normal_state
infrun: TARGET_WAITKIND_STOPPED
infrun: stop_pc = 0x400704
infrun: stepi/nexti
infrun: stop_stepping
44 while (counter != 0)
(gdb) PASS: gdb.threads/stepi-random-signal.exp: stepi
The test turns on infrun debug, does a stepi while a SIGCHLD is
pending, and checks whether the "random signal" paths in infrun.c are
taken.
On the software single-step variant above, those paths were not taken.
This is a test bug.
The Linux backend short-circuits reporting signals that are set to
pass/nostop/noprint. But _only_ if the thread is _not_
single-stepping. So on hardware-step targets, even though the signal
is set to pass/nostop/noprint by default, the thread is indeed told to
single-step, and so the core sees the signal. On the other hand, on
software single-step architectures, the backend never actually gets a
single-step request (steps are emulated by setting a breakpoint at the
next pc, and then the target told to continue, not step). So the
short-circuiting code triggers and the core doesn't see the signal.
The fix is to make the test be sure the target doesn't bypass
reporting the signal to the core.
Tested on x86_64 Fedora 17, both with and without a series that
implements software single-step for x86_64.
gdb/testsuite/
2014-02-07 Pedro Alves <palves@redhat.com>
* gdb.threads/stepi-random-signal.exp: Set SIGCHLD to print.
Nowadays, argument LEN of to_xfer_partial can be zero in some cases,
and each implementation may do nothing and return zero, indicating
transfer is done. That is fine. However, when we change
to_xfer_partial to return target_xfer_status, we have to check every
return value of most of to_xfer_partial implementations, return
TARGET_XFER_DONE if return value is zero.
This patch simplifies this by checking LEN in target_xfer_partial, and
return 0 if LEN is zero. Regression tested on x86_84-linux. Is it
OK?
gdb:
2014-02-07 Yao Qi <yao@codesourcery.com>
* target.c (target_xfer_partial): Return zero if LEN is zero.
This patch documents the return value of core_xfer_shared_libraries_aix
and core_xfer_shared_libraries gdbarch methods, and changes return
type to ULONGEST from LONGEST.
In a following patch, core_xfer_partial. is changed to check their
return values and return an appropriate target_xfer_status.
gdb:
2014-02-07 Yao Qi <yao@codesourcery.com>
* gdbarch.sh (core_xfer_shared_libraries): Returns ULONGEST. Add
comments.
(core_xfer_shared_libraries_aix): Likewise.
* gdbarch.c, gdbarch.h: Regenerated.
* i386-cygwin-tdep.c (windows_core_xfer_shared_libraries): Return
ULONGEST. Change 'len_avail' type to ULONGEST.
* rs6000-aix-tdep.c (rs6000_aix_ld_info_to_xml): Likewise.
* rs6000-aix-tdep.h (rs6000_aix_ld_info_to_xml): Update
declaration.
(rs6000_aix_core_xfer_shared_libraries_aix): Return ULONGEST.
This patch adds a local variable exception of type 'enum errors' and
pass it to throw_error, because 'err' is of type
'enum target_xfer_error', and we're abusing it to store an 'enum errors'.
gdb:
2014-02-07 Yao Qi <yao@codesourcery.com>
* corefile.c (memory_error): Get 'exception' from ERR and pass
'exception' to throw_error.
gdb/testsuite/
2014-02-06 Jan Kratochvil <jan.kratochvil@redhat.com>
Fix i386-sse-stack-align.exp regression since GDB_PARALLEL.
* gdb.arch/i386-sse-stack-align.exp: Use standard_output_file.
Tom Tromey
Yao Qi
Doug Evans
Kevin Buettner
Doug Evans
Ludovic Courtès
Jan Kratochvil
Thomas Schwinge
Jose E. Marchesi
Sanimir Agovic
Mark Kettenis
Pedro Alves
Joel Brobecker
Stan Shebs
Will Newton
Gary Benson
Andreas Schwab