vcomplaint now has a single caller, so merge it with that caller.
gdb/ChangeLog
2018-05-23 Tom Tromey <tom@tromey.com>
* complaints.c (vcomplaint): Remove.
(complaint_internal) Merge in contents of vcomplaint.
Now that there's only a single reason for a complaint to be emitted,
this removes "struct explanation" and changes vcomplaint to emit the
desired messages directly.
gdb/ChangeLog
2018-05-23 Tom Tromey <tom@tromey.com>
* complaints.c (struct complaints) <explanation>: Remove.
(symfile_explanations): Remove.
(symfile_complaint_book): Update.
(vcomplaint): Update.
(struct explanation): Remove.
After the previous patch, the "noisy" parameter to clear_complaints is
no longer used, so this patch removes it.
gdb/ChangeLog
2018-05-23 Tom Tromey <tom@tromey.com>
* complaints.c (clear_complaints): Remove "noisy" parameter.
* complaints.h (clear_complaints): Update.
* symfile.c (syms_from_objfile_1, finish_new_objfile)
(reread_symbols): Update.
gdb/testsuite/ChangeLog
2018-05-23 Tom Tromey <tom@tromey.com>
* gdb.gdb/complaints.exp (test_empty_complaints): Update.
I couldn't find a way to get complaints to use a couple of cases, and
the difference between the actual printed output for these cases was
minimal anyway. So, this patch removes a couple of constants from
complaint_series, plus the associated code.
gdb/ChangeLog
2018-05-23 Tom Tromey <tom@tromey.com>
* complaints.c (enum complaint_series): Remove FIRST_MESSAGE,
SUBSEQUENT_MESSAGE.
(vcomplaint, clear_complaints): Update.
(symfile_explanations): Remove some messages.
gdb/testsuite/ChangeLog
2018-05-23 Tom Tromey <tom@tromey.com>
* gdb.gdb/complaints.exp (test_serial_complaints): Remove.
(test_short_complaints): Update.
I happened to notice that gdb has both complaint_internal and
internal_complaint. The latter is unused, so this patch removes it.
gdb/ChangeLog
2018-05-23 Tom Tromey <tom@tromey.com>
* complaints.c (internal_complaint): Remove.
* complaints.h (internal_complaint): Remove.
Currently, the function used by gdbserver to parse integers from
received File I/O commands will detect overflow and fail for any value
over 0xfffffff. Among other things, this has the effect of limiting
the file offsets for reading or writing to about 268MB which can be
insufficient for particularly large libraries.
This change allows the parsing of integers up to the true maximum
positive value of 0x7fffffff, increasing the file size limit to about
2GB.
gdb/gdbserver/ChangeLog:
2018-05-23 Erik Kurzinger <ekurzinger@nvidia.com>
PR server/23198
* hostio.c (require_int): Do not report overflow for integers
between 0xfffffff and 0x7fffffff.
This fixes gdb.base/remote.exp regressions caused by the previous
commit to the testcase, when tested with
--target_board=native-extended-gdbserver. For example:
...
show remote memory-write-packet-size
The memory-write-packet-size is 0 (default). Packets are limited to 16383 bytes.
(gdb) FAIL: gdb.base/remote.exp: write-packet default
...
With that board, GDB connects to GDBserver at gdb_start time, so GDB
is showing the actual remote/gdbserver packet size limits.
Fix it using the usual "disconnect" pattern. While at it, there's no
need to start GDB before compiling the testcase.
gdb/testsuite/ChangeLog:
2018-05-22 Pedro Alves <palves@redhat.com>
* gdb.base/remote.exp: Only gdb_start after compiling the
testcase. Issue "disconnect" before testing "set remote" command
defaults. Issue clean_restart before running to main.
Make `i' defined within `i387_collect_xsave' unsigned, removing a
`-Werror=strict-overflow' compilation error:
.../gdb/i387-tdep.c: In function 'void i387_collect_xsave(const regcache*, int, void*, int)':
.../gdb/i387-tdep.c:1348:1: error: assuming signed overflow does not occur when assuming that (X + c) < X is always false [-Werror=strict-overflow]
i387_collect_xsave (const struct regcache *regcache, int regnum,
^
cc1plus: all warnings being treated as errors
Makefile:1610: recipe for target 'i387-tdep.o' failed
make: *** [i387-tdep.o] Error 1
seen with GCC 5.4.0, a commit 8ee22052f6 ("gdb/x86: Handle kernels
using compact xsave format") regression. While `regnum' can be -1 on
entry to the function, to mean all registers, `i' is only used with
non-negative register numbers.
gdb/
* i387-tdep.c (i387_collect_xsave): Make `i' unsigned.
Fix an issue with `gdbserver' on big-endian n64 MIPS targets, where
$dspctl is 32-bit while the `ptrace' transfer data type is 64-bit.
Such register data is held in the low order 32 bits of the 64-bit data
quantity exchanged with the buffer used by `fetch_register' and
`store_register', however `supply_register' and `collect_register'
access the same data as a 32-bit quantity. Consequently the register is
presented and written as all-zeros held in the most-significant part of
the big-endian 64-bit data quantity represented in the buffer:
(gdb) info registers
zero at v0 v1
R0 0000000000000000 0000000000000001 0000000000000001 0000000000000000
a0 a1 a2 a3
R4 00000001200212b0 0000000000000000 0000000000000021 000000012001a260
a4 a5 a6 a7
R8 000000012001a260 0000000000000004 800000010c60c000 fffffffffffffff8
t0 t1 t2 t3
R12 0000000000000000 000000fff7edab68 0000000000000001 0000000000000000
s0 s1 s2 s3
R16 000000fff7ee2068 0000000120008b80 0000000000000000 0000000000000000
s4 s5 s6 s7
R20 000000000052e5c8 000000000052f008 0000000000000000 0000000000000000
t8 t9 k0 k1
R24 0000000000000000 00000001200027c0 0000000000000000 0000000000000000
gp sp s8 ra
R28 00000001200212b0 000000ffffffc850 000000ffffffc850 0000000120005ee8
status lo hi badvaddr
0000000000109cf3 0000000000943efe 000000000000000e 000000012001a008
cause pc
0000000000800024 0000000120005ee8
fcsr fir hi1 lo1
0e800000 00f30000 0000000000000000 0101010101010101
hi2 lo2 hi3 lo3
0202020202020202 0303030303030303 0404040404040404 0505050505050505
dspctl restart
00000000 0000000000000000
(gdb)
Correct this problem then by using the `mips_supply_register'
`mips_collect_register' accessors for 32-bit registers where the
`ptrace' data type is 64-bit. These accessors already operate on 32-bit
data quantities held in 64-bit containers:
(gdb) info registers
zero at v0 v1
R0 0000000000000000 0000000000000001 0000000000000001 0000000000000000
a0 a1 a2 a3
R4 00000001200212b0 0000000000000000 0000000000000021 000000012001a260
a4 a5 a6 a7
R8 000000012001a260 0000000000000004 800000010d82e900 fffffffffffffff8
t0 t1 t2 t3
R12 0000000000000000 000000fff7edab68 0000000000000001 0000000000000000
s0 s1 s2 s3
R16 000000fff7ee2068 0000000120008b80 0000000000000000 0000000000000000
s4 s5 s6 s7
R20 000000000052e5c8 000000000052f008 0000000000000000 0000000000000000
t8 t9 k0 k1
R24 0000000000000000 00000001200027c0 0000000000000000 0000000000000000
gp sp s8 ra
R28 00000001200212b0 000000ffffffc850 000000ffffffc850 0000000120005ee8
status lo hi badvaddr
0000000000109cf3 0000000000943efe 000000000000000e 000000012001a008
cause pc
0000000000800024 0000000120005ee8
fcsr fir hi1 lo1
0e800000 00f30000 0000000000000000 0101010101010101
hi2 lo2 hi3 lo3
0202020202020202 0303030303030303 0404040404040404 0505050505050505
dspctl restart
55aa33cc 0000000000000000
(gdb)
gdb/gdbserver/
* linux-mips-low.c [HAVE_PTRACE_GETREGS] (mips_collect_register)
(mips_supply_register): Move outside HAVE_PTRACE_GETREGS.
(mips_collect_ptrace_register, mips_supply_ptrace_register): New
functions.
(the_low_target): Wire them.
'struct remote_state' today contains per-connection state, however
there's only a single global instance of that type. In order to
support multiple connections, we must have one such object per
connection.
Thus this patch eliminates the 'remote_state' global in favor of
having a remote_state instance per remote_target instance.
The get_remote_state free function is eliminated as well, by making it
a remote_target method instead.
The patch then fixes the fallout by making all free functions that
refer to get_remote_state() directly or indirectly be methods of
remote_target too.
Likewise, remote-fileio.c and remote-notif.c routines are
parameterized with a remote_target pointer too, so they can call into
the right remote_target instance.
References to the global 'get_remote_state ()->remote_desc' to tell
whether the remote target is open (!= nullptr) must be replaced with
something else:
- Command implementations use a new get_current_remote_target free
function.
- remote_target::open_1 checks the exception type instead.
Finally, remote_target and extended_remote_target are made
heap-allocated targets. As with the earlier core target patches, it
still won't be possible to have more than one remote_target instance
in practice, but this puts us closer.
gdb/ChangeLog:
2018-05-22 Pedro Alves <palves@redhat.com>
* remote-fileio.c (remote_fileio_reply, remote_fileio_ioerror)
(remote_fileio_badfd, remote_fileio_return_errno)
(remote_fileio_return_success, remote_fileio_func_open)
(remote_fileio_func_open, remote_fileio_func_close)
(remote_fileio_func_read, remote_fileio_func_write)
(remote_fileio_func_lseek, remote_fileio_func_rename)
(remote_fileio_func_unlink, remote_fileio_func_stat)
(remote_fileio_func_fstat, remote_fileio_func_gettimeofday)
(remote_fileio_func_isatty, remote_fileio_func_system): Add
remote_target parameter.
(remote_fio_func_map) <func>: Add remote_target parameter.
(do_remote_fileio_request, remote_fileio_request):
* remote-fileio.h (remote_fileio_request):
* remote-notif.c (remote_notif_ack, remote_notif_parse, ): Add
remote_target parameter.
(remote_notif_process, handle_notification): Adjust to pass down
the remote.
(remote_notif_state_allocate): Add remote_target parameter. Save
it.
* remote-notif.h (struct remote_target): Forward declare.
(struct notif_client) <parse, ack, can_get_pending_events>: Add
remote_target parameter.
(struct remote_notif_state) <remote>: New field.
(remote_notif_ack, remote_notif_parse): Add remote_target
parameter.
(remote_notif_state_allocate, remote_notif_state_allocate): Add
remote_target parameter.
* remote.c (OPAQUETHREADBYTES, threadref, gdb_ext_thread_info)
(threads_listing_context, rmt_thread_action, protocol_feature)
(packet_reg, stop_reply, stop_reply_p, enum packet_support)
(packet_result, struct threads_listing_context, remote_state):
Move definitions and declarations higher up.
(remote_target) <~remote_target>: Declare.
(remote_download_command_source, remote_file_put, remote_file_get)
(remote_file_delete, remote_hostio_pread, remote_hostio_pwrite)
(remote_hostio_pread_vFile, remote_hostio_send_command)
(remote_hostio_set_filesystem, remote_hostio_open)
(remote_hostio_close, remote_hostio_unlink, remote_state)
(get_remote_state, get_remote_packet_size, get_memory_packet_size)
(get_memory_write_packet_size, get_memory_read_packet_size)
(append_pending_thread_resumptions, remote_detach_1)
(append_resumption, remote_resume_with_vcont)
(add_current_inferior_and_thread, wait_ns, wait_as)
(process_stop_reply, remote_notice_new_inferior)
(process_initial_stop_replies, remote_add_thread)
(btrace_sync_conf, remote_btrace_maybe_reopen)
(remove_new_fork_children, kill_new_fork_children)
(discard_pending_stop_replies, stop_reply_queue_length)
(check_pending_events_prevent_wildcard_vcont)
(discard_pending_stop_replies_in_queue, stop_reply)
(remote_notif_remove_queued_reply, stop_reply *queued_stop_reply)
(peek_stop_reply, remote_parse_stop_reply, remote_stop_ns)
(remote_interrupt_as, remote_interrupt_ns)
(remote_get_noisy_reply, remote_query_attached)
(remote_add_inferior, remote_current_thread, get_current_thread)
(set_thread, set_general_thread, set_continue_thread)
(set_general_process, write_ptid)
(remote_unpack_thread_info_response, remote_get_threadinfo)
(parse_threadlist_response, remote_get_threadlist)
(remote_threadlist_iterator, remote_get_threads_with_ql)
(remote_get_threads_with_qxfer)
(remote_get_threads_with_qthreadinfo, extended_remote_restart)
(get_offsets, remote_check_symbols, remote_supported_packet)
(remote_query_supported, remote_packet_size)
(remote_serial_quit_handler, remote_detach_pid)
(remote_vcont_probe, remote_resume_with_hc)
(send_interrupt_sequence, interrupt_query)
(remote_notif_get_pending_events, fetch_register_using_p)
(send_g_packet, process_g_packet, fetch_registers_using_g)
(store_register_using_P, store_registers_using_G)
(set_remote_traceframe, check_binary_download)
(remote_write_bytes_aux, remote_write_bytes, remote_read_bytes_1)
(remote_xfer_live_readonly_partial, remote_read_bytes)
(remote_send_printf, remote_flash_write, readchar)
(remote_serial_write, putpkt, putpkt_binary, skip_frame)
(read_frame, getpkt, getpkt_or_notif_sane_1, getpkt_sane)
(getpkt_or_notif_sane, remote_vkill, remote_kill_k)
(extended_remote_disable_randomization, extended_remote_run)
(send_environment_packet, extended_remote_environment_support)
(extended_remote_set_inferior_cwd, remote_write_qxfer)
(remote_read_qxfer, push_stop_reply, vcont_r_supported)
(packet_command): Now methods of ...
(remote_target): ... this class.
(m_remote_state) <remote_target>: New field.
(struct remote_state) <stop_reply_queue,
remote_async_inferior_event_token, wait_forever_enabled_p>: New
fields.
(remote_state::remote_state): Allocate stop_reply_queue.
(remote_state): Delete global.
(get_remote_state_raw): Delete.
(remote_target::get_remote_state): Allocate m_remote_state on
demand.
(get_current_remote_target): New.
(remote_ops, extended_remote_ops): Delete.
(wait_forever_enabled_p, remote_async_inferior_event_token):
Delete, moved to struct remote_state.
(remote_target::close): Delete self. Destruction bits split to
...
(remote_target::~remote_target): ... this.
(show_memory_packet_size): Adjust to use
get_current_remote_target.
(struct protocol_feature) <func>: Add remote_target parameter.
All callers adjusted.
(curr_quit_handler_target): New.
(remote_serial_quit_handler): Reimplement.
(remote_target::open_1): Adjust to use get_current_remote_target.
Heap-allocate remote_target/extended_remote_target instances.
(vcont_builder::vcont_builder): Add remote_target parameter, and
save it in m_remote. All callers adjusted.
(vcont_builder::m_remote): New field.
(vcont_builder::restart, vcont_builder::flush)
(vcont_builder::push_action): Use it.
(remote_target::commit_resume): Use it.
(struct queue_iter_param) <remote>: New field.
(remote_target::remove_new_fork_children): Fill in 'remote' field.
(check_pending_event_prevents_wildcard_vcont_callback_data): New.
(check_pending_event_prevents_wildcard_vcont_callback)
(remote_target::check_pending_events_prevent_wildcard_vcont)
(remote_target::discard_pending_stop_replies)
(remote_target::discard_pending_stop_replies_in_queue)
(remote_target::remote_notif_remove_queued_reply): Fill in
'remote' field.
(remote_notif_get_pending_events): New.
(remote_target::readchar, remote_target::remote_serial_write):
Save/restore curr_quit_handler_target.
(putpkt): New.
(kill_new_fork_children): Fill in 'remote' field.
(packet_command): Use get_current_remote_target, defer to
remote_target method of same name.
(scoped_remote_fd::scoped_remote_fd): Add 'remote_target'
parameter, and save it in m_remote. All callers adjusted.
(scoped_remote_fd::release): Use m_remote.
(scoped_remote_fd::m_remote): New field.
(remote_file_put, remote_file_get, remote_file_delete): Use
get_current_remote_target, defer to remote_target method of same
name.
(remote_btrace_reset): Add remote_state paremeter. Update all
callers.
(remote_async_inferior_event_handler). Pass down 'data'.
(remote_new_objfile): Use get_current_remote_target.
(remote_target::vcont_r_supported): New.
(set_range_stepping): Use get_current_remote_target and
remote_target::vcont_r_supported.
(_initialize_remote): Don't allocate 'remote_state' and
'stop_reply_queue' globals.
* remote.h (struct remote_target): Forward declare.
(getpkt, putpkt, remote_notif_get_pending_events): Add
'remote_target' parameter.
gdb/ChangeLog:
2018-05-22 Pedro Alves <palves@redhat.com>
* remote.c (vcont_builder): Now a class. Make all data members
private.
(vcont_builder) <vcont_builder, restart, flush, push_action>:
Declare methods.
(vcont_builder_restart): Rename to ...
(vcont_builder::restart): ... this.
(vcont_builder_flush): Rename to ...
(vcont_builder::flush): ... this.
(vcont_builder_push_action): Rename to ...
(vcont_builder::push_action): ... this.
(remote_target::commit_resume): Adjust.
Currently "show remote memory-write-packet-size" says that the packet
size is limited to whatever is stored in the remote_state global, even
if not connected to a target.
When we get to support multiple instances of remote targets, there
won't be a remote_state global anymore, so that must be replaced by
something else.
Since it doesn't make sense to print the limit of the packet size of a
non-existing connection, this patch makes us say that the limit will
be further reduced when we connect.
The text is taken from the command's online help, which says:
"The actual limit is further reduced dependent on the target."
Note that a value of "0" is special, as per "help set remote
memory-write-packet-size":
~~~
Specify the number of bytes in a packet or 0 (zero) for the
default packet size.
~~~
I've tweaked "show remote memory-write-packet-size" to include
"(default)" in the output in that case, like this:
(gdb) show remote memory-write-packet-size
The memory-write-packet-size is 0 (default). The actual limit will be further reduced dependent on the target.
While working on this, I noticed that an explicit "set remote
write-packet-size 0" does not makes GDB go back to the exact same
state as the default state when GDB starts up:
(gdb) show remote memory-write-packet-size
The memory-write-packet-size is 0. [...]
^^
(gdb) set remote memory-write-packet-size 0
(gdb) show remote memory-write-packet-size
The memory-write-packet-size is 16384. [...]
^^^^^
The "16384" number comes from DEFAULT_MAX_MEMORY_PACKET_SIZE.
This happens because git commit a5c0808e22 ("gdb: remove packet size
limit") at
<https://sourceware.org/ml/gdb-patches/2015-08/msg00743.html>, added
this:
/* So that the query shows the correct value. */
if (size <= 0)
size = DEFAULT_MAX_MEMORY_PACKET_SIZE;
to set_memory_packet_size, but despite what the comment suggests, that
also has the side-effect of recording DEFAULT_MAX_MEMORY_PACKET_SIZE
in config->size.
Finally, DEFAULT_MAX_MEMORY_PACKET_SIZE only makes sense for "set
remote memory-write-packet-size fixed", so I've renamed it
accordingly, to make it a little bit clearer.
gdb/ChangeLog:
2018-05-22 Pedro Alves <palves@redhat.com>
* remote.c (DEFAULT_MAX_MEMORY_PACKET_SIZE): Rename to ...
(DEFAULT_MAX_MEMORY_PACKET_SIZE_FIXED): ... this.
(get_fixed_memory_packet_size): New.
(get_memory_packet_size): Use it.
(set_memory_packet_size): Don't override the config size with
DEFAULT_MAX_MEMORY_PACKET_SIZE.
(show_memory_packet_size): Use get_fixed_memory_packet_size.
Don't refer to get_memory_packet_size if not connected to a remote
target. Show "(default)" if configured size is 0.
gdb/testsuite/ChangeLog:
2018-05-22 Pedro Alves <palves@redhat.com>
* gdb.base/remote.exp: Adjust expected output of "show remote
memory-write-packet-size". Add tests for "set remote
memory-write-packet-size 0" and "set remote
memory-write-packet-size fixed/limit".
This helps because discard_pending_stop_replies will later become a
method of remote_target. Otherwise, when we have multiple instances
of remote_target, we'd have to make discard_pending_stop_replies find
the inferior's target_ops, and upcast it to remote_target (if indeed a
remote) to find the pending stop replies queue to clear.
gdb/ChangeLog:
2018-05-22 Pedro Alves <palves@redhat.com>
* remote.c (remote_target::mourn_inferior): Move
discard_pending_stop_replies call here from ...
(_initialize_remote): ... here.
The set_general_process call in compare_sections_command isn't
actually needed. remote_target::verify_memory and
remote_target::xfer_partial already handle making sure the remote is
pointing at the right process or thread.
Getting this out of the way helps a bit with further elimination of
the remote_state global, because we have to handle the case of a user
invoking the command even if not connect to a remote target.
gdb/ChangeLog:
2018-05-22 Pedro Alves <palves@redhat.com>
* remote.c (compare_section_command): Remove set_general_process
call.
The previous patch made the map store pointers to remote_arch_state
instead of objects directly, simply because struct remote_arch_state
is still incomplete where struct remote_state is declared. This patch
thus moves the remote_arch_state declaration higher up in the file,
and makes the map store remote_arch_state objects directly instead of
pointers to objects.
gdb/ChangeLog:
2018-05-22 Pedro Alves <palves@redhat.com>
* remote.c (struct packet_reg, struct remote_arch_state):
Move higher up in the file.
(remote_state) <m_arch_states>: Store remote_arch_state values
instead of remote_arch_state pointers.
(remote_state::get_remote_arch_state): Adjust.
Currently, we associate gdbarch-related remote protocol state on a
per-gdbarch data object. Things like the size of the g/G packet, and
the max remote packet size. If we'll support being connected to
different remote servers at the same time, then we need to cope with
each having their own packet sizes, even if they are each debugging
programs of the same architecture. I.e., a single instance of
remote_arch_state per arch is not sufficient.
This patch moves the remote_arch_state object to a map of
gdbarch-to-remote_arch_state saved in the remote_state structure.
Usually there will only be one entry in the map, though we may see
more with stubs that support multi-process and/or archs with multiple
ABIs (e.g, one remote_arch_state for 64-bit inferiors and another for
32-bit inferiors).
gdb/ChangeLog:
2018-05-22 Pedro Alves <palves@redhat.com>
* remote.c: Include <unordered_map>.
(remote_state): Now a class.
(remote_state) <get_remote_arch_state>: Declare method.
<get_remote_arch_state>: New field.
(remote_arch_state) <remote_arch_state>: Declare ctor.
<regs>: Now a unique_ptr.
(remote_gdbarch_data_handle): Delete.
(get_remote_arch_state): Delete.
(remote_state::get_remote_arch_state): New.
(get_remote_state): Adjust to call remote_state's
get_remote_arch_state method.
(init_remote_state): Delete, bits factored out to ...
(remote_arch_state::remote_arch_state): ... this new method.
(get_remote_packet_size, get_memory_packet_size)
(process_g_packet, remote_target::fetch_registers)
(remote_target::prepare_to_store, store_registers_using_G)
(remote_target::store_registers, remote_target::get_trace_status):
Adjust to call remote_state's method.
(_initialize_remote): Remove reference to
remote_gdbarch_data_handle.
The idea here is eliminate the get_remote_state calls from within
readahead_cache_invalidate, readahead_cache_invalidate_fd,
remote_hostio_pread_from_cache by making those functions be class
methods instead. Later on we'll have one readahead_cache instance per
remote connection, and this change makes that easier.
gdb/ChangeLog:
2018-05-22 Pedro Alves <palves@redhat.com>
* remote.c (struct readahead_cache) <invalidate, invalidate_fd,
pread>: New method declarations.
(remote_target::open_1): Adjust.
(readahead_cache_invalidate): Rename to ...
(readahead_cache::invalidate): ... this, and adjust to be a class
method.
(readahead_cache_invalidate_fd): Rename to ...
(readahead_cache::invalidate_fd): ... this, and adjust to be a
class method.
(remote_hostio_pwrite): Adjust.
(remote_hostio_pread_from_cache): Rename to ...
(readahead_cache::pread): ... this, and adjust to be a class
method.
(remote_hostio_close): Adjust.
A bit of C++ification. Also plugs a few leaks, since currently
nothing is freeing a few fields of remote_state. They're now freed in
the added destructor.
gdb/ChangeLog:
2018-05-22 Pedro Alves <palves@redhat.com>
(struct vCont_action_support): Use bool and initialize all fields.
(struct readahead_cache): Initialize all fields.
(remote_state): Use bool and initialize all fields.
(remote_state::remote_state, remote_state::~remote_state): New.
(new_remote_state): Delete.
(_initialize_remote): Use new to allocate remote_state.
Factor out cp_ident_is_alpha/cp_ident_is_alnum out of
gdb/cp-name-parser.y and use it in the C/C++ expression parser too.
New test included.
gdb/ChangeLog:
2018-05-22 Pedro Alves <palves@redhat.com>
張俊芝 <zjz@zjz.name>
PR gdb/22973
* c-exp.y: Include "c-support.h".
(parse_number, c_parse_escape, lex_one_token): Use TOLOWER instead
of tolower. Use c_ident_is_alpha to scan names.
* c-lang.c: Include "c-support.h".
(convert_ucn, convert_octal, convert_hex, convert_escape): Use
ISXDIGIT instead of isxdigit and ISDIGIT instead of isdigit.
* c-support.h: New file, with bits factored out from ...
* cp-name-parser.y: ... this file.
Include "c-support.h".
(cp_ident_is_alpha, cp_ident_is_alnum): Deleted, moved to
c-support.h and renamed.
(symbol_end, yylex): Adjust.
gdb/testsuite/ChangeLog:
2018-05-22 Pedro Alves <palves@redhat.com>
PR gdb/22973
* gdb.base/utf8-identifiers.c: New file.
* gdb.base/utf8-identifiers.exp: New file.
Currently the ppc linux core file target doesn't return target
descriptions with the lager FPSCR introduced in isa205.
This patch changes the core file target so that the auxv is read from
the core file to determine the size of FPSCR, so that the appropriate
target description is selected.
gdb/ChangeLog:
2018-05-22 Pedro Franco de Carvalho <pedromfc@linux.vnet.ibm.com>
* arch/ppc-linux-common.c (ppc_linux_has_isa205): Change the
parameter type to CORE_ADDR.
* arch/ppc-linux-common.h (ppc_linux_has_isa205): Change the
parameter type in declaration to CORE_ADDR.
* ppc-linux-tdep.c (ppc_linux_core_read_description): Call
target_auxv_search to get AT_HWCAP and use the result to get the
target description.
* ppc-linux-nat.c (ppc_linux_get_hwcap): Change the return type
to CORE_ADDR. Remove the cast of the return value to unsigned
long. Fix error predicate of target_auxv_search.
(ppc_linux_nat_target::read_description): Change the type of the
hwcap variable to CORE_ADDR.
gdb/testsuite/ChangeLog:
2018-05-22 Pedro Franco de Carvalho <pedromfc@linux.vnet.ibm.com>
* gdb.arch/powerpc-fpscr-gcore.exp: New file.
Previously, decimal floating point pseudoregisters were always included
in the target if it had a floating point unit.
This patch changes this to only include them if the target description
indicates that they are present, i.e. if the FPSCR register has more
than 32 bits.
gdb/ChangeLog:
2018-05-22 Pedro Franco de Carvalho <pedromfc@linux.vnet.ibm.com>
* rs6000-tdep.c (rs6000_gdbarch_init): Assign 1 to have_dfp only
if the size of fpscr is larger than 32 bits.
The functions used by the VSX regset to collect and supply registers
from core files where incorrect. This patch changes the regset to use
the standard regset collect/supply functions to fix this. The native
target is also changed to use the same regset.
gdb/ChangeLog:
2018-05-22 Pedro Franco de Carvalho <pedromfc@linux.vnet.ibm.com>
* ppc-linux-tdep.c (ppc_linux_vsxregset): New function.
(ppc32_linux_vsxregmap): New global.
(ppc32_linux_vsxregset): Initialize with ppc32_linux_vsxregmap,
regcache_supply_regset, and regcache_collect_regset.
* ppc-linux-tdep.h (ppc_linux_vsxregset): Declare.
* ppc-linux-nat.c (supply_vsxregset, fill_vsxregset): Remove.
(fetch_vsx_register, store_vsx_register): Remove.
(fetch_vsx_registers): Add regno parameter. Get regset using
ppc_linux_vsxregset. Use regset to supply registers.
(store_vsx_registers): Add regno parameter. Get regset using
ppc_linux_vsxregset. Use regset to collect registers.
(fetch_register): Call fetch_vsx_registers instead of
fetch_vsx_register.
(store_register): Call store_vsx_registers instead of
store_vsx_register.
(fetch_ppc_registers): Call fetch_vsx_registers with -1 for the
new regno parameter.
(store_ppc_registers): Call store_vsx_registers with -1 for the
new regno parameter.
* rs6000-tdep.c (ppc_vsx_support_p, ppc_supply_vsxreget)
(ppc_collect_vsxregset): Remove.
gdb/testsuite/ChangeLog:
2018-05-22 Pedro Franco de Carvalho <pedromfc@linux.vnet.ibm.com>
* gdb.arch/powerpc-vsx-gcore.exp: New file.
The 4-byte VSCR register is found inside a 16-byte field in the regset
returned by ptrace and in core files. The position of VSCR depends on
the endianess of the target, which was previously assumed to be
big-endian for the purpose of getting VSCR. This patch removes this
assumption to fix access to VSCR in little-endian mode.
gdb/ChangeLog:
2018-05-22 Pedro Franco de Carvalho <pedromfc@linux.vnet.ibm.com>
* ppc-tdep.h (struct ppc_reg_offsets): Remove vector register
offset fields.
* ppc-fbsd-tdep.c (ppc32_fbsd_reg_offsets): Remove initializers
for vector register offset fields.
(ppc64_fbsd_reg_offsets): Likewise.
* ppc-nbsd-tdep.c (_initialize_ppcnbsd_tdep): Remove assignment
to vector register offset fields.
* ppc-obsd-tdep.c (_initialize_ppcnbsd_tdep): Remove assignment
to vector register offset fields.
* ppc-obsd-nat.c (_initialize_ppcobsd_nat): Remove assignment to
vector register offset fields.
* rs6000-aix-tdep.c (rs6000_aix32_reg_offsets): Remove
initializers for vector register offset fields.
(rs6000_aix64_reg_offsets): Likewise.
* rs6000-tdep.c (ppc_vrreg_offset): Remove.
(ppc_supply_vrregset): Remove.
(ppc_collect_vrregset): Remove.
* ppc-linux-tdep.c (ppc_linux_collect_vrregset): New function.
(ppc_linux_vrregset) : New function.
(ppc32_le_linux_vrregmap, ppc32_be_linux_vrregmap)
(ppc32_le_linux_vrregset, ppc32_be_linux_vrregset): New globals.
(ppc32_linux_vrregset): Remove.
(ppc_linux_iterate_over_regset_sections): Call ppc_linux_vrregset
and use result instead of ppc32_linux_vrregset.
(ppc32_linux_reg_offsets): Remove initializers for vector register
offset fields.
(ppc64_linux_reg_offsets): Likewise.
* ppc-linux-tdep.h (ppc_linux_vrregset): New declaration.
* ppc-linux-nat.c: Include regset.h.
(gdb_vrregset_t): Adjust comment to account for little-endian
mode.
(supply_vrregset, fill_vrregset): Remove.
(fetch_altivec_register, store_altivec_register): Remove.
(fetch_altivec_registers): Add regno parameter. Get regset using
ppc_linux_vrregset. Use regset to supply registers.
(store_altivec_registers): Add regno parameter. Get regset using
ppc_linux_vrregset. Use regset to collect registers.
(fetch_register): Call fetch_altivec_registers instead of
fetch_altivec_register.
(store_register): Call store_altivec_registers instead of
store_altivec_register.
(fetch_ppc_registers): Call fetch_altivec_registers with -1 for
the new regno parameter.
(store_ppc_registers): Call store_altivec_registers with -1 for
the new regno parameter.
gdb/gdbserver/ChangeLog:
2018-05-22 Pedro Franco de Carvalho <pedromfc@linux.vnet.ibm.com>
* linux-ppc-low.c (ppc_fill_vrregset): Add vscr_offset variable.
Set vscr_offset to 0 in little-endian mode and 12 in big-endian
mode. Call collect_register_by_name with vscr using
vscr_offset. Zero-pad vscr and vrsave fields in collector buffer.
(ppc_store_vrregset): Add and set vscr_offset variable as in
ppc_fill_vrregset. Call supply_register_by_name with vscr using
vscr_offset.
This patch defines constants for the sizes of the two vector
regsets (vector-scalar registers and regular vector registers).
The native, gdbserver and core file targets are changed to use these
constants.
The Linux ptrace calls return (or read) a smaller regset than the one
found in core files for vector registers, because ptrace uses a single
4-byte quantity for vrsave at the end of the regset, while the
core-file regset uses a full 16-byte field for vrsave. For simplicity,
the larger size is used in both cases, and so a buffer with 12 unused
additional bytes is passed to ptrace in the native target.
gdb/ChangeLog:
2018-05-22 Pedro Franco de Carvalho <pedromfc@linux.vnet.ibm.com>
* arch/ppc-linux-common.h (PPC_LINUX_SIZEOF_VRREGSET)
(PPC_LINUX_SIZEOF_VSXREGSET): Define.
* ppc-linux-nat.c (SIZEOF_VSXREGS, SIZEOF_VRREGS): Remove.
(gdb_vrregset_t): Change array type size to
PPC_LINUX_SIZEOF_VRREGSET.
(gdb_vsxregset_t): Change array type size to
PPC_LINUX_SIZEOF_VSXREGSET.
* ppc-linux-tdep.c (ppc_linux_iterate_over_regset_sections):
Change integer literals to PPC_LINUX_SIZEOF_VRREGSET and
PPC_LINUX_SIZEOF_VSXREGSET.
gdb/gdbserver/ChangeLog:
2018-05-22 Pedro Franco de Carvalho <pedromfc@linux.vnet.ibm.com>
* linux-ppc-low.c (SIZEOF_VSXREGS, SIZEOF_VRREGS): Remove.
(ppc_arch_setup): Change SIZEOF_VRREGS and SIZEOF_VSXREGS to
PPC_LINUX_SIZEOF_VRREGSET and PPC_LINUX_SIZEOF_VSXREGSET.
Currently the linux-ppc-low.c fill/store functions for extended
regsets check whether they should execute by using the global hwcap
variable.
This patch explicitly sets the regset sizes to zero when needed to
disable them instead, so that the fill/store functions are not called
in the first place by regsets_fetch_inferior_registers in linux-low.c.
gdb/gdbserver/ChangeLog:
2018-05-22 Pedro Franco de Carvalho <pedromfc@linux.vnet.ibm.com>
* linux-ppc-low.c (ppc_fill_vsxregset): Remove ppc_hwcap check.
(ppc_store_vsxregset): Likewise.
(ppc_fill_vrregset): Likewise.
(ppc_store_vrregset): Likewise.
(ppc_fill_evrregset): Likewise.
(ppc_store_evrregset): Likewise.
(ppc_regsets): Set VSX/VR/EVR regset sizes to 0.
(ppc_arch_setup): Iterate through ppc_regsets and set sizes when
needed.
This patch moves the native target wordsize getter for ppc linux to
nat/ so that it can be used to simplify ppc_arch_setup in
gdbserver. The ptrace call used to get MSR for this is ultimately the
same as before, but it is no longer necessary to create a temporary
regcache to call fetch_inferior_registers.
gdb/ChangeLog:
2018-05-22 Pedro Franco de Carvalho <pedromfc@linux.vnet.ibm.com>
* configure.nat <linux powerpc>: Add ppc-linux.o to NATDEPFILES.
* ppc-linux-nat.c (ppc_linux_target_wordsize): Move to
nat/ppc-linux.c.
(ppc_linux_nat_target::auxv_parse): Get thread id tid. Call
ppc_linux_target_wordsize with tid.
(ppc_linux_nat_target::read_description): Call ppc_linux_target
wordsize with tid.
* nat/ppc-linux.c: Include nat/gdb_ptrace.h.
(ppc64_64bit_inferior_p): Add static and inline specifiers.
(ppc_linux_target_wordsize): Move here from ppc-linux-nat.c. Add
tid parameter. Remove static specifier.
* nat/ppc-linux.h (ppc64_64bit_inferior_p): Remove declaration.
(ppc_linux_target_wordsize): New declaration.
gdb/gdbserver/ChangeLog:
2018-05-22 Pedro Franco de Carvalho <pedromfc@linux.vnet.ibm.com>
* linux-ppc-low.c (ppc_arch_setup): Remove code for getting the
wordsize of the inferior. Call ppc_linux_target_wordsize.
Share target description declarations and selection among ppc linux
native targets, core files, gdbserver and IPA.
To avoid complicated define guards, gdbserver and IPA now have
declarations for all descriptions, including 64-bit generated
descriptions when compiled in 32-bit mode. These have always been
linked into the gdbserver and IPA binaries. Because they might be
uninitialized, the selection function checks that the selected
description is initialized.
gdb/ChangeLog:
2018-05-22 Pedro Franco de Carvalho <pedromfc@linux.vnet.ibm.com>
* arch/ppc-linux-common.c: New file.
* arch/ppc-linux-common.h: New file.
* arch/ppc-linux-tdesc.h: New file.
* configure.tgt (powerpc*-*-linux*): Add arch/ppc-linux-common.o.
* Makefile.in (ALL_TARGET_OBS): Add arch/ppc-linux-common.o.
(HFILES_NO_SRCDIR): Add arch/ppc-linux-common.h and
arch/ppc-linux-tdesc.h.
* ppc-linux-nat.c: Include arch/ppc-linux-common.h and
arch/ppc-linux-tdesc.h.
(ppc_linux_nat_target::read_description): Remove target
description matching code. Fill a ppc_linux_features struct and
call ppc_linux_match_description with it. Move comment about ISA
2.05 to ppc-linux-common.c.
* ppc-linux-tdep.c: Include arch/ppc-linux-common.h and
arch/ppc-linux-tdesc.h.
(ppc_linux_core_read_description): Remove target description
matching code. Fill a ppc_linux_features struct and call
ppc_linux_match_description with it.
* ppc-linux-tdep.h (tdesc_powerpc_32l, tdesc_powerpc_64l)
(tdesc_powerpc_altivec32l, tdesc_powerpc_altivec64l)
(tdesc_powerpc_cell32l, tdesc_powerpc_cell64l)
(tdesc_powerpc_vsx32l, tdesc_powerpc_vsx64l)
(tdesc_powerpc_isa205_32l, tdesc_powerpc_isa205_64l)
(tdesc_powerpc_isa205_altivec32l, tdesc_powerpc_isa205_altivec64l)
(tdesc_powerpc_isa205_vsx32l, tdesc_powerpc_isa205_vsx64l)
(tdesc_powerpc_e500l): Remove.
gdb/gdbserver/ChangeLog:
2018-05-22 Pedro Franco de Carvalho <pedromfc@linux.vnet.ibm.com>
* configure.srv (srv_tgtobj): Add arch/ppc-linux-common.o.
* Makefile.in (SFILES): Add arch/ppc-linux-common.c.
* linux-ppc-tdesc.h: Rename to linux-ppc-tdesc-init.h.
* linux-ppc-tdesc-init.h (tdesc_powerpc_32l, tdesc_powerpc_64l)
(tdesc_powerpc_altivec32l, tdesc_powerpc_altivec64l)
(tdesc_powerpc_cell32l, tdesc_powerpc_cell64l)
(tdesc_powerpc_vsx32l, tdesc_powerpc_vsx64l)
(tdesc_powerpc_isa205_32l, tdesc_powerpc_isa205_64l)
(tdesc_powerpc_isa205_altivec32l, tdesc_powerpc_isa205_altivec64l)
(tdesc_powerpc_isa205_vsx32l, tdesc_powerpc_isa205_vsx64l)
(tdesc_powerpc_e500l): Remove.
* linux-ppc-ipa.c: Include arch/ppc-linux-tdesc.h and
linux-ppc-tdesc-init.h. Don't include linux-ppc-tdesc.h.
* linux-ppc-low.c: Include arch/ppc-linux-common.h,
arch/ppc-linux-tdesc.h, and linux-ppc-tdesc-init.h. Don't include
linux-ppc-tdesc.h.
(ppc_arch_setup): Remove target description matching code. Fill a
ppc_linux_features struct and call ppc_linux_match_description
with it.
Trying to insert a catchpoint on all Ada assertions now triggers
the following internal warning regardless of the situation. For
instance, not even debugging any program:
(gdb) catch assert
/[...]/gdb/common/cleanups.c:264: internal-warning:
restore_my_cleanups has found a stale cleanup
This is due to a small bug in the following C++-ification commit:
commit bc18fbb575
Author: Tom Tromey <tom@tromey.com>
Date: Fri May 18 15:58:50 2018 -0600
Subject: Change ada_catchpoint::excep_string to be a std::string
The stale cleanup in question is the following one in top.c:execute_command:
cleanup_if_error = make_bpstat_clear_actions_cleanup ();
This cleanup is expected to be discarded if there are no exception.
There were no GDB exception; however, a C++ exception was triggered,
because we passed NULL as the excep_string argument when calling
create_ada_exception_catchpoint, which is a reference to a const
string. So we get a C++ exception during the std::string constructor,
which propagates up, causing the cleanup to unexpectedly remain
in the cleanup chain.
This patch fixes the immediate issue of the incorrect call to
create_ada_exception_catchpoint.
gdb/ChangeLog:
* ada-lang.c (catch_assert_command): Pass empty string instead
of NULL for excep_string argument.
Tested on x86_64-linux, fixes the following failures:
* catch_assert_if.exp: insert catchpoint on failed assertions with condition
* catch_ex.exp: insert catchpoint on failed assertions
This also fixes about a dozen UNRESOLVED tests that are a consequence
of the two tests above failing and crashing GDB.
On the MIPS target DSP ASE registers can only be accessed with the
PTRACE_PEEKUSR and PTRACE_POKEUSR `ptrace' requests. With the n32 ABI
these requests only pass 32-bit data quantities, which are narrower than
the width of DSP accumulator registers, which are 64-bit.
Generic code is prepared to transfer registers wider than the `ptrace'
data type by offsetting into the USR address space, by the data width
transferred. That however does not work with the MIPS target, because
of how the API has been defined, where USR register addresses are
actually indices rather than offsets. Consequently given address `a'
using `a + 4' accesses the fourth next register rather than the upper
half of the original register.
With native debugging this causes clobbered register contents, as well
as access failures as locations beyond the available USR space are
addressed:
(gdb) info registers
zero at v0 v1
R0 0000000000000000 0000000000000001 0000000000000001 0000000000000000
a0 a1 a2 a3
R4 0000000010019158 0000000000000000 0000000000000011 0000000010019160
a4 a5 a6 a7
R8 0000000010019160 fffffffffff00000 fffffffffffffff8 0000000000000000
t0 t1 t2 t3
R12 0000000010019150 0000000000000001 0000000000000001 000000000000000f
s0 s1 s2 s3
R16 0000000077ee6f20 0000000010007bb0 0000000000000000 0000000000000000
s4 s5 s6 s7
R20 000000000052e668 000000000052f008 0000000000000000 0000000000000000
t8 t9 k0 k1
R24 0000000000000001 0000000010019010 0000000000000000 0000000000000000
gp sp s8 ra
R28 0000000010020280 000000007fff4c10 000000007fff4c10 0000000010004f48
status lo hi badvaddr
0000000000109cf3 0000000000943efe 000000000000000e 000000001001900c
cause pc
0000000000800024 0000000010004f48
fcsr fir hi1 lo1
0e800000 00f30000 0000000004040404 0101010105050505
hi2 lo2 hi3 lo3
0202020255aa33cc Couldn't read register (#75): Input/output error.
(gdb)
With `gdbserver' this makes debugging impossible due to a fatal failure:
(gdb) target remote :2346
Remote debugging using :2346
Reading symbols from .../sysroot/mips-r2-hard/lib32/ld.so.1...done.
0x77fc3d50 in __start () from .../sysroot/mips-r2-hard/lib32/ld.so.1
(gdb) continue
Continuing.
warning: Remote failure reply: E01
Remote communication error. Target disconnected.: Connection reset by peer.
(gdb)
Correct the problem by marking any register in the MIPS backend whose
width exceeds the width of the `ptrace' data type unavailable for the
purpose of PTRACE_PEEKUSR and PTRACE_POKEUSR requests:
(gdb) info registers
zero at v0 v1
R0 0000000000000000 0000000000000001 0000000000000001 0000000000000000
a0 a1 a2 a3
R4 0000000010019158 0000000000000000 0000000000000011 0000000010019160
a4 a5 a6 a7
R8 0000000010019160 fffffffffff00000 fffffffffffffff8 0000000000000000
t0 t1 t2 t3
R12 0000000010019150 0000000000000001 0000000000000001 000000000000000f
s0 s1 s2 s3
R16 0000000077ee6f20 0000000010007bb0 0000000000000000 0000000000000000
s4 s5 s6 s7
R20 000000000052e5c8 000000000052f008 0000000000000000 0000000000000000
t8 t9 k0 k1
R24 0000000000000001 0000000010019010 0000000000000000 0000000000000000
gp sp s8 ra
R28 0000000010020280 000000007fff4be0 000000007fff4be0 0000000010004f48
status lo hi badvaddr
0000000000109cf3 0000000000943efe 000000000000000e 000000001001900c
cause pc
0000000000800024 0000000010004f48
fcsr fir hi1 lo1
0e800000 00f30000 <unavailable> <unavailable>
hi2 lo2 hi3 lo3
<unavailable> <unavailable> <unavailable> <unavailable>
dspctl restart
55aa33cc 0000000000000000
(gdb)
as there is no way to access full contents of these registers with the
limited API available anyway.
This obviously does not affect general-purpose registers (which use the
PTRACE_GETREGS and PTRACE_SETREGS requests for access) or floating-point
general registers (which use PTRACE_GETFPREGS and PTRACE_SETFPREGS).
And $dspctl, being 32-bit, remains accessible too, which is important
for BPOSGE32 branch decoding in single-stepping.
For DSP accumulator access with the n32 ABI a new `ptrace' API is required
on the kernel side.
gdb/
* mips-linux-nat.c (mips64_linux_register_addr): Return -1 if
the width of the requested register exceeds the width of the
`ptrace' data type.
gdb/gdbserver/
* linux-mips-low.c (mips_cannot_fetch_register): Return 1 if the
width of the requested register exceeds the width of the
`ptrace' data type.
(mips_cannot_store_register): Likewise.
Consistently supply hardwired $zero as a zeroed register, correcting
issues with the PTRACE_GETREGS path that currently copies the value of
$restart into $zero as illustrated by this program:
$ cat read.c
int
main (void)
{
char buf[1024];
ssize_t size;
size = read (0, buf, sizeof (buf));
return size;
}
$
and this corresponding debug session:
(gdb) break main
Breakpoint 1 at 0x120000970: file read.c, line 9.
(gdb) target remote :2346
Remote debugging using :2346
Reading symbols from .../sysroot/mips-r2-hard/lib64/ld.so.1...done.
0x000000fff7fca5a0 in __start ()
from .../sysroot/mips-r2-hard/lib64/ld.so.1
(gdb) continue
Continuing.
Breakpoint 1, main () at read.c:9
9 size = read (0, buf, sizeof (buf));
(gdb) info registers
zero at v0 v1
R0 0000000000000000 0000000000000001 000000fff7ffe710 0000000000000000
a0 a1 a2 a3
R4 0000000000000001 000000ffffffeb88 000000ffffffeb98 0000000000000000
a4 a5 a6 a7
R8 000000fff7fc8800 000000fff7fc38f0 000000ffffffeb80 2f2f2f2f2f2f2f2f
t0 t1 t2 t3
R12 0000000000000437 0000000000000002 000000fff7ffd000 0000000120000a00
s0 s1 s2 s3
R16 000000fff7fc7068 0000000120000b90 0000000000000000 0000000000000000
s4 s5 s6 s7
R20 0000000000521d88 0000000000522608 0000000000000000 0000000000000000
t8 t9 k0 k1
R24 0000000000000000 0000000120000970 0000000000000000 0000000000000000
gp sp s8 ra
R28 000000fff7fc8800 000000ffffffea50 0000000000000000 000000fff7e4088c
status lo hi badvaddr
0000000000109cf3 0000000000005ea5 0000000000000211 000000fff7eadf00
cause pc
0000000000800024 0000000120000970
fcsr fir restart
00000000 00f30000 0000000000000000
(gdb) continue
Continuing.
^C
Program received signal SIGINT, Interrupt.
0x000000fff7f084ac in __GI___libc_read (fd=0, buf=0xffffffe640, nbytes=1024)
at ../sysdeps/unix/sysv/linux/read.c:27
27 return SYSCALL_CANCEL (read, fd, buf, nbytes);
(gdb) info registers
zero at v0 v1
R0 0000000000001388 0000000000000001 0000000000000200 000000fff7ffe710
a0 a1 a2 a3
R4 0000000000000000 000000ffffffe640 0000000000000400 0000000000000001
a4 a5 a6 a7
R8 000000fff7fc8800 000000fff7fc38f0 000000ffffffeb80 2f2f2f2f2f2f2f2f
t0 t1 t2 t3
R12 00000000000005e3 0000000000000002 000000fff7ffd000 000000012000099c
s0 s1 s2 s3
R16 000000fff7fc7068 0000000120000b90 0000000000000000 0000000000000000
s4 s5 s6 s7
R20 0000000000521d88 0000000000522608 0000000000000000 0000000000000000
t8 t9 k0 k1
R24 0000000000000000 000000fff7f2da20 0000000000000000 0000000000000000
gp sp s8 ra
R28 000000fff7fc8800 000000ffffffe600 0000000000000000 000000012000099c
status lo hi badvaddr
0000000000109cf3 00000000000001e6 00000000000000be 000000fff7f08470
cause pc
0000000000800020 000000fff7f084ac
fcsr fir restart
00000000 00f30000 0000000000001388
(gdb)
and with the PTRACE_PEEKUSR path that does not supply this register at
all, causing issues analogous to ones addressed for the native MIPS
backend with commit 4e6ff0e1b8 ("MIPS/Linux/native: Supply $zero for
the !PTRACE_GETREGS case"):
(gdb) info registers
zero at v0 v1
R0 <unavailable> 0000000000000001 0000000000000001 0000000000000000
a0 a1 a2 a3
R4 00000001200212b0 0000000000000000 0000000000000021 000000012001a260
a4 a5 a6 a7
R8 000000012001a260 0000000000000004 800000010cab1680 fffffffffffffff8
t0 t1 t2 t3
R12 0000000000000000 000000fff7edab68 0000000000000001 0000000000000000
s0 s1 s2 s3
R16 000000fff7ee2068 0000000120008b80 0000000000000000 0000000000000000
s4 s5 s6 s7
R20 000000000052e5c8 000000000052f008 0000000000000000 0000000000000000
t8 t9 k0 k1
R24 0000000000000000 00000001200027c0 0000000000000000 0000000000000000
gp sp s8 ra
R28 00000001200212b0 000000ffffffc880 000000ffffffc880 0000000120005ee8
status lo hi badvaddr
<unavailable> 0000000000943efe 000000000000000e 000000012001a008
cause pc
0000000000800024 0000000120005ee8
fcsr fir restart
0e800000 00f30000 0000000000000000
(gdb)
and (under certain circumstances):
(gdb) next
Register 0 is not available
(gdb)
The problem with PTRACE_GETREGS happens because `mips_store_gregset'
supplies the contents of register slot #0, occupied by $restart, to
$zero. The problem with PTRACE_PEEKUSR happens because for $zero
`mips_cannot_fetch_register' returns one, and no alternative way to
supply that register has been defined.
Correct `mips_store_gregset' then for the PTRACE_GETREGS case and add
`mips_fetch_register' for the PTRACE_PEEKUSR case.
gdb/gdbserver/
* linux-mips-low.c (mips_fetch_register): New function. Update
preceding comment.
(mips_store_gregset): Supply 0 rather than $restart for $zero.
(the_low_target): Wire `mips_fetch_register'.
I happened to notice that output_command_const still exists, but is
not needed any more -- commands are always const-correct now. This
patch removes this leftover.
2018-05-21 Tom Tromey <tom@tromey.com>
* printcmd.c (output_command): Remove.
(output_command_const): Rename to output_command.
* valprint.h (output_command): Rename from output_command_const.
* tracepoint.c (trace_dump_actions): Call output_command.
This changes ada_catchpoint::excep_string to be a std::string and then
fixes up all t he users.
This found a memory leak in catch_ada_exception_command_split, where
"cond" was copied but never freed.
I changed the type of the "cond_string" argument to
catch_ada_exception_command_split to follow the rule that out
parameters should be pointers and not references.
This patch enables the removal of some cleanups and also the function
ada_get_next_arg.
ChangeLog
2018-05-21 Tom Tromey <tom@tromey.com>
* mi/mi-cmd-catch.c (mi_cmd_catch_assert)
(mi_cmd_catch_exception, mi_cmd_catch_handlers): Update.
* ada-lang.h (create_ada_exception_catchpoint): Update.
* ada-lang.c (struct ada_catchpoint) <excep_string>: Now a
std::string.
(create_excep_cond_exprs, ~ada_catchpoint)
(should_stop_exception, print_one_exception)
(print_mention_exception, print_recreate_exception): Update.
(ada_get_next_arg): Remove.
(catch_ada_exception_command_split): Use std::string. Change type
of "excep_string", "cond_string".
(catch_ada_exception_command): Update.
(create_ada_exception_catchpoint): Change type of excep_string.
(ada_exception_sal): Remove excep_string parameter.
(~ada_catchpoint): Remove.
ada_collect_symbol_completion_matches installs a null_cleanup but not
any other cleanups. This patch removes it.
ChangeLog
2018-05-21 Tom Tromey <tom@tromey.com>
* ada-lang.c (ada_collect_symbol_completion_matches): Remove
cleanup.
This removes a cleanup from ada-lang.c by having
ada_exception_message_1 return a unique_xmalloc_ptr.
ChangeLog
2018-05-21 Tom Tromey <tom@tromey.com>
* ada-lang.c (ada_exception_message_1, ada_exception_message):
Return unique_xmalloc_ptr.
(print_it_exception): Update.
This changes trace_dump_actions to use std::string, removing a
cleanup.
Tested by the buildbot.
ChangeLog
2018-05-21 Tom Tromey <tom@tromey.com>
* tracepoint.c (trace_dump_actions): Use std::string.
This removes a cleanup from reread_symbols by using std::string. This
fixes a memory leak, because this cleanup is ordinarily discarded, not
run.
Tested by the buildbot.
ChangeLog
2018-05-21 Tom Tromey <tom@tromey.com>
* symfile.c (reread_symbols): Use std::string for original_name.
This changes dwarf2_read_debug_names to use std::unique_ptr from the
outset. This simplifies the code that installs the resulting map into
dwarf2_per_objfile.
Tested by the buildbot.
ChangeLog
2018-05-21 Tom Tromey <tom@tromey.com>
* dwarf2read.c (dwarf2_read_debug_names): Use std::unique_ptr.
(mapped_index_base): Use DISABLE_COPY_AND_ASSIGN. Default
constructor.
I realized after pushing that I made a copy-pasto, I had:
# define HAVE_IS_TRIVIALLY_COPYABLE 1
instead of
# define HAVE_IS_TRIVIALLY_CONSTRUCTIBLE 1
with the consequence that IsMallocable was always std::true_type (and
was therefore not enforcing anything). Fixing that mistake triggered a
build failure:
/home/simark/src/binutils-gdb/gdb/objfiles.c:150:12: required from here
/home/simark/src/binutils-gdb/gdb/common/poison.h:228:3: error: static assertion failed: Trying to use XOBNEW with a non-POD data type.
I am not sure why I did not see this when I originally wrote the patch
(but I saw and fixed other failures). In any case, I swapped XOBNEW
with obstack_new to get rid of it.
Regtested on the buildbot.
gdb/ChangeLog:
* common/traits.h (HAVE_IS_TRIVIALLY_COPYABLE): Rename the wrong
instance to...
(HAVE_IS_TRIVIALLY_CONSTRUCTIBLE): ... this.
* objfiles.c (get_objfile_bfd_data): Allocate
objfile_per_bfd_storage with obstack_new when allocating on
obstack.
Since XOBNEW/XOBNEWVEC/OBSTACK_ZALLOC are now poisoned to prevent using
them with non-trivially-constructible objects, it is worth using them
over plain obstack_alloc. This patch changes the locations I could find
where we can do that change easily.
gdb/ChangeLog:
* ada-lang.c (cache_symbol): Use XOBNEW and/or XOBNEWVEC and/or
OBSTACK_ZALLOC.
* dwarf2-frame.c (dwarf2_build_frame_info): Likewise.
* hppa-tdep.c (hppa_init_objfile_priv_data): Likewise.
* mdebugread.c (mdebug_build_psymtabs): Likewise.
(add_pending): Likewise.
(parse_symbol): Likewise.
(parse_partial_symbols): Likewise.
(psymtab_to_symtab_1): Likewise.
(new_psymtab): Likewise.
(elfmdebug_build_psymtabs): Likewise.
* minsyms.c (terminate_minimal_symbol_table): Likewise.
* objfiles.c (get_objfile_bfd_data): Likewise.
(objfile_register_static_link): Likewise.
* psymtab.c (allocate_psymtab): Likewise.
* stabsread.c (read_member_functions): Likewise.
* xcoffread.c (xcoff_end_psymtab): Likewise.
Since we use obstacks with objects that are not default constructible,
we sometimes need to manually call the constructor by hand using
placement new:
foo *f = obstack_alloc (obstack, sizeof (foo));
f = new (f) foo;
It's possible to use allocate_on_obstack instead, but there are types
that we sometimes want to allocate on an obstack, and sometimes on the
regular heap. This patch introduces a utility to make this pattern
simpler if allocate_on_obstack is not an option:
foo *f = obstack_new<foo> (obstack);
Right now there's only one usage (in tdesc_data_init).
To help catch places where we would forget to call new when allocating
such an object on an obstack, this patch also poisons some other methods
of allocating an instance of a type on an obstack:
- OBSTACK_ZALLOC/OBSTACK_CALLOC
- XOBNEW/XOBNEW
- GDBARCH_OBSTACK_ZALLOC/GDBARCH_OBSTACK_CALLOC
Unfortunately, there's no way to catch wrong usages of obstack_alloc.
By pulling on that string though, it tripped on allocating struct
template_symbol using OBSTACK_ZALLOC. The criterion currently used to
know whether it's safe to "malloc" an instance of a struct is whether it
is a POD. Because it inherits from struct symbol, template_symbol is
not a POD. This criterion is a bit too strict however, it should still
safe to allocate memory for a template_symbol and memset it to 0. We
didn't use is_trivially_constructible as the criterion in the first
place only because it is not available in gcc < 5. So here I considered
two alternatives:
1. Relax that criterion to use std::is_trivially_constructible and add a
bit more glue code to make it work with gcc < 5
2. Continue pulling on the string and change how the symbol structures
are allocated and initialized
I managed to do both, but I decided to go with #1 to keep this patch
simpler and more focused. When building with a compiler that does not
have is_trivially_constructible, the check will just not be enforced.
gdb/ChangeLog:
* common/traits.h (HAVE_IS_TRIVIALLY_COPYABLE): Define if
compiler supports std::is_trivially_constructible.
* common/poison.h: Include obstack.h.
(IsMallocable): Define to is_trivially_constructible if the
compiler supports it, define to true_type otherwise.
(xobnew): New.
(XOBNEW): Redefine.
(xobnewvec): New.
(XOBNEWVEC): Redefine.
* gdb_obstack.h (obstack_zalloc): New.
(OBSTACK_ZALLOC): Redefine.
(obstack_calloc): New.
(OBSTACK_CALLOC): Redefine.
(obstack_new): New.
* gdbarch.sh: Include gdb_obstack in gdbarch.h.
(gdbarch_obstack): New declaration in gdbarch.h, definition in
gdbarch.c.
(GDBARCH_OBSTACK_CALLOC, GDBARCH_OBSTACK_ZALLOC): Use
obstack_calloc/obstack_zalloc.
(gdbarch_obstack_zalloc): Remove.
* target-descriptions.c (tdesc_data_init): Use obstack_new.
This adds a constructor to struct dwz_file and arranges for it to be
allocated with "new" and wrapped in a unique_ptr. This cuts down on
the amount of manual memory management that must be done.
Regression tested by the buildbot.
gdb/ChangeLog
2018-05-18 Tom Tromey <tom@tromey.com>
* dwarf2read.c (struct dwz_file): Add constructor, initializers.
<dwz_bfd>: Now a gdb_bfd_ref_ptr.
(~dwarf2_per_objfile): Update
(dwarf2_get_dwz_file): Use new.
* dwarf2read.h (struct dwarf2_per_objfile) <dwz_file>: Now a
unique_ptr.
This adds a constructor and initializer to dwp_file and changes it to
be allocated with "new". This removes a bit of manual refcount
management.
Tested by the buildbot.
gdb/ChangeLog
2018-05-18 Tom Tromey <tom@tromey.com>
* dwarf2read.h (struct dwarf2_per_objfile) <dwp_file>: Now a
unique_ptr.
* dwarf2read.c (struct dwp_file): Add constructor and
initializers.
(open_and_init_dwp_file): Return a unique_ptr.
(dwarf2_per_objfile, create_dwp_hash_table)
(create_dwo_unit_in_dwp_v1, create_dwo_unit_in_dwp_v2)
(lookup_dwo_unit_in_dwp): Update.
(open_and_init_dwp_file, get_dwp_file): Update.
This changes struct mapped_index to be allocated with new. This
simplifies the creation a bit (see dwarf2_read_index) and also removes
a somewhat ugly explicit destructor call from ~dwarf2_per_objfile.
Tested by the buildbot.
gdb/ChangeLog
2018-05-18 Tom Tromey <tom@tromey.com>
* dwarf2read.c (dwarf2_per_objfile): Update.
(struct mapped_index): Add initializers.
(dwarf2_read_index): Use new.
(dw2_symtab_iter_init): Update.
* dwarf2read.h (struct dwarf2_per_objfile) <index_table>: Now a
unique_ptr.
Silence this:
unittests/format_pieces-selftests.c:51: warning: code: Do not use printf("%ll"), instead use printf("%s",phex()) to dump a `long long' value
unittests/format_pieces-selftests.c:56: warning: code: Do not use printf("%ll"), instead use printf("%s",phex()) to dump a `long long' value
gdb/ChangeLog:
* unittests/format_pieces-selftests.c (test_format_specifier):
Add ARI comments.
I was recently using ptype/o to look at the layout of some objects in
gdb. I noticed that trailing padding was not shown -- but I wanted to
be able to look at that, too.
This patch changes ptype/o to also print trailing holes.
Tested on x86-64 Fedora 26.
gdb/ChangeLog
2018-05-18 Tom Tromey <tom@tromey.com>
* c-typeprint.c (maybe_print_hole): New function.
(c_print_type_struct_field_offset): Update.
(c_type_print_base_struct_union): Call maybe_print_hole.
gdb/testsuite/ChangeLog
2018-05-18 Tom Tromey <tom@tromey.com>
* gdb.base/ptype-offsets.exp: Update.
This patch essentially causes GDB to treat inlined frames like "normal"
frames from the user's perspective. This means, for example, that when a
user sets a breakpoint in an inlined function, GDB will now actually stop
"in" that function.
Using the test case from breakpoints/17534,
3 static inline void NVIC_EnableIRQ(int IRQn)
4 {
5 volatile int y;
6 y = IRQn;
7 }
8
9 __attribute__( ( always_inline ) ) static inline void __WFI(void)
10 {
11 __asm volatile ("nop");
12 }
13
14 int main(void) {
15
16 x= 42;
17
18 if (x)
19 NVIC_EnableIRQ(16);
20 else
21 NVIC_EnableIRQ(18);
(gdb) b NVIC_EnableIRQ
Breakpoint 1 at 0x4003e4: NVIC_EnableIRQ. (2 locations)
(gdb) r
Starting program: 17534
Breakpoint 1, main () at 17534.c:19
19 NVIC_EnableIRQ(16);
Because skip_inline_frames currently skips every inlined frame, GDB "stops"
in the caller. This patch adds a new parameter to skip_inline_frames
that allows us to pass in a bpstat stop chain. The breakpoint locations
on the stop chain can be used to determine if we've stopped inside an inline
function (due to a user breakpoint). If we have, we do not elide the frame.
With this patch, GDB now reports that the inferior has stopped inside the
inlined function:
(gdb) r
Starting program: 17534
Breakpoint 1, NVIC_EnableIRQ (IRQn=16) at 17534.c:6
6 y = IRQn;
Many thanks to Jan and Pedro for guidance on this.
gdb/ChangeLog:
* breakpoint.c (build_bpstat_chain): New function, moved from
bpstat_stop_status.
(bpstat_stop_status): Add optional parameter, `stop_chain'.
If no stop chain is passed, call build_bpstat_chain to build it.
* breakpoint.h (build_bpstat_chain): Declare.
(bpstat_stop_status): Move documentation here from breakpoint.c.
* infrun.c (handle_signal_stop): Before eliding inlined frames,
build the stop chain and pass it to skip_inline_frames.
Pass this stop chain to bpstat_stop_status.
* inline-frame.c: Include breakpoint.h.
(stopped_by_user_bp_inline_frame): New function.
(skip_inline_frames): Add parameter `stop_chain'.
Move documention to inline-frame.h.
If non-NULL, use stopped_by_user_bp_inline_frame to determine
whether the frame should be elided.
* inline-frame.h (skip_inline_frames): Add parameter `stop_chain'.
Add moved documentation and update for new parameter.
gdb/testsuite/ChangeLog:
* gdb.ada/bp_inlined_func.exp: Update inlined frame locations
in expected breakpoint stop locations.
* gdb.dwarf2/implptr.exp (implptr_test_baz): Use up/down to
move to proper scope to test variable values.
* gdb.opt/inline-break.c (inline_func1, not_inline_func1)
(inline_func2, not_inline_func2, inline_func3, not_inline_func3):
New functions.
(main): Call not_inline_func3.
* gdb.opt/inline-break.exp: Start inferior and set breakpoints at
inline_func1, inline_func2, and inline_func3. Test that when each
breakpoint is hit, GDB properly reports both the stop location
and the backtrace. Repeat tests for temporary breakpoints.
I noticed that the printf command did not recognize the \e escape
sequence, used amongst other things to use colors:
(gdb) printf "This is \e[32mgreen\e[m!\n"
Unrecognized escape character \e in format string.
This patch makes format_pieces recognize it, which makes that command
print the expected result in glorious color.
I wrote a really simple unit test for format_pieces.
format_pieces::operator[] is unused so I removed it. I added
format_piece::operator==, which is needed to compare vectors of
format_piece.
gdb/ChangeLog:
PR cli/14975
* Makefile.in (SUBDIR_UNITTESTS_SRCS): Add
unittests/format_pieces-selftests.c.
* common/format.h (format_piece) <operator==>: New.
(format_pieces) <operator[]>: Remove.
* common/format.c (format_pieces::format_pieces): Handle \e.
* unittests/format_pieces-selftests.c: New.
PR symtab/23010 reports a crash that occurs when using -readnow
on a dwz-generated debuginfo file.
The crash occurs because the DWARF has a partial CU with no language
set, and then a full CU that references this partial CU using
DW_AT_abstract_origin.
In this case, the partial CU is read by dw2_expand_all_symtabs using
language_minimal; but then this conflicts with the creation of the
block's symbol table in the C++ CU.
This patch fixes the problem by arranging for partial CUs not to be
read by -readnow. I tend to think that it doesn't make sense to read
a partial CU in isolation -- they should only be read when imported
into some other CU.
In conjunction with some other patches I am going to post, this also
fixes the Rust -readnow crash that Jan reported.
There are two problems with this patch:
1. It is difficult to reason about. There are many cases where I've
patched the code to call init_cutu_and_read_dies with the flag set
to "please do read partial units" -- but I find it difficult to be
sure that this is always correct.
2. It is still missing a standalone test case. This seemed hard.
2018-05-17 Tom Tromey <tom@tromey.com>
PR symtab/23010:
* dwarf2read.c (load_cu, dw2_do_instantiate_symtab)
(dw2_instantiate_symtab): Add skip_partial parameter.
(dw2_find_last_source_symtab, dw2_map_expand_apply)
(dw2_lookup_symbol, dw2_expand_symtabs_for_function)
(dw2_expand_all_symtabs, dw2_expand_symtabs_with_fullname)
(dw2_expand_symtabs_matching_one)
(dw2_find_pc_sect_compunit_symtab)
(dw2_debug_names_lookup_symbol)
(dw2_debug_names_expand_symtabs_for_function): Update.
(init_cutu_and_read_dies): Add skip_partial parameter.
(process_psymtab_comp_unit, build_type_psymtabs_1)
(process_skeletonless_type_unit, load_partial_comp_unit)
(psymtab_to_symtab_1): Update.
(load_full_comp_unit): Add skip_partial parameter.
(process_imported_unit_die, dwarf2_read_addr_index)
(follow_die_offset, dwarf2_fetch_die_loc_sect_off)
(dwarf2_fetch_constant_bytes, dwarf2_fetch_die_type_sect_off)
(read_signatured_type): Update.
Obvious patch to remove unused local variables (found by adding
-Wunused). I didn't touch this one in value_fetch_lazy, because
check_typedef could have a desired side-effect.
3743 struct type *type = check_typedef (value_type (val));
gdb/ChangeLog:
* value.c (release_value): Remove unused variable.
(record_latest_value): Likewise.
(access_value_history): Likewise.
(preserve_values): Likewise.
When running gdb in the build directory without passing
--data-directory, I noticed I could provoke a crash by:
$ ./gdb -nx ./gdb
(gdb) ptype/o struct dwarf2_per_objfile
... and then trying to "q" out at the pagination prompt.
valgrind complained about an uninitialized use of py_type_printers.
Initializing this member fixes the bug.
I believe this bug can occur even when the gdb Python libraries are
available, for example if get_type_recognizers fails.
Tested by hand on x86-64 Fedora 26. No test case because it seemed
difficult to guarantee failures.
gdb/ChangeLog
2018-05-17 Tom Tromey <tom@tromey.com>
* extension.h (struct ext_lang_type_printers) <py_type_printers>:
Initialize.
Update `fetch_register' and `store_register' code to support arbitrary
register widths rather than only ones that are a multiply of the size of
the `ptrace' data type used with PTRACE_PEEKUSR and PTRACE_POKEUSR
requests to access registers. Remove associated assertions, correcting
an issue with accessing the DSPControl (`$dspctl') register on n64 MIPS
native targets:
(gdb) print /x $dspctl
.../gdb/linux-nat-trad.c:50: internal-error: void linux_nat_trad_target::fetch_register(regcache*, int): Assertion `(size % sizeof (PTRACE_TYPE_RET)) == 0' failed.
A problem internal to GDB has been detected,
further debugging may prove unreliable.
Quit this debugging session? (y or n) n
This is a bug, please report it. For instructions, see:
<http://www.gnu.org/software/gdb/bugs/>.
.../gdb/linux-nat-trad.c:50: internal-error: void linux_nat_trad_target::fetch_register(regcache*, int): Assertion `(size % sizeof (PTRACE_TYPE_RET)) == 0' failed.
A problem internal to GDB has been detected,
further debugging may prove unreliable.
Create a core file of GDB? (y or n) n
Command aborted.
(gdb)
All registers are now reported correctly and their architectural
hardware widths respected:
(gdb) print /x $dspctl
$1 = 0x55aa33cc
(gdb) info registers
zero at v0 v1
R0 0000000000000000 0000000000000001 000000fff7ffeb20 0000000000000000
a0 a1 a2 a3
R4 0000000000000001 000000ffffffeaf8 000000ffffffeb08 0000000000000000
a4 a5 a6 a7
R8 000000fff7ee3800 000000fff7ede8f0 000000ffffffeaf0 2f2f2f2f2f2f2f2f
t0 t1 t2 t3
R12 0000000000000437 0000000000000002 000000fff7ffd000 0000000120000ad0
s0 s1 s2 s3
R16 000000fff7ee2068 0000000120000e60 0000000000000000 0000000000000000
s4 s5 s6 s7
R20 0000000000521ec8 0000000000522608 0000000000000000 0000000000000000
t8 t9 k0 k1
R24 0000000000000000 0000000120000d9c 0000000000000000 0000000000000000
gp sp s8 ra
R28 0000000120019030 000000ffffffe990 000000ffffffe990 000000fff7d5b88c
status lo hi badvaddr
0000000000109cf3 0000000000005ea5 0000000000000211 000000fff7fc6fe0
cause pc
0000000000800024 0000000120000dbc
fcsr fir hi1 lo1
00000000 00f30000 0000000000000000 0101010101010101
hi2 lo2 hi3 lo3
0202020202020202 0303030303030303 0404040404040404 0505050505050505
dspctl restart
55aa33cc 0000000000000000
(gdb)
NB due to the lack of access to 64-bit DSP hardware all DSP register
values in the dumps are artificial and have been created with a debug
change applied to the kernel handler of the `ptrace' syscall.
The use of `store_unsigned_integer' and `extract_unsigned_integer'
unconditionally in all cases rather than when actual data occupies a
part of the data quantity exchanged with `ptrace' makes code perhaps
marginally slower, however I think avoiding it is not worth code
obfuscation it would cause. If this turns out unfounded, then there
should be no problem with optimizing this code later.
gdb/
PR gdb/22286
* linux-nat-trad.c (linux_nat_trad_target::fetch_register):
Also handle registers whose width is not a multiple of
PTRACE_TYPE_RET.
(linux_nat_trad_target::store_register): Likewise.
This changes program_space::cbfd to be a gdb_bfd_ref_ptr. This makes
it somewhat less error-prone to use, because now it manages the
reference counting automatically.
Tested by the buildbot.
2018-05-16 Tom Tromey <tom@tromey.com>
* gdbcore.h (core_bfd): Redefine.
* corelow.c (core_target::close): Update.
(core_target_open): Update.
* progspace.h (struct program_space) <cbfd>: Now a
gdb_bfd_ref_ptr.
One part of PR cli/19551 is that the mini debug info objfile reuses the
name of the main objfile from which it comes. This can be seen because
gdb claims to be reading symbols from the same file two times, like:
Reading symbols from /bin/gdb...Reading symbols from /bin/gdb...(no debugging symbols found)...done.
I think this would be less confusing if the minidebug objfile were given
a different name. That is what this patch implements. It also arranges
for the minidebug objfile to be marked OBJF_NOT_FILENAME.
After this patch the output looks like:
Reading symbols from /bin/gdb...Reading symbols from .gnu_debugdata for /usr/libexec/gdb...(no debugging symbols found)...done.
Tested by the buildbot.
gdb/ChangeLog
2018-05-16 Tom Tromey <tom@tromey.com>
PR cli/19551:
* symfile-add-flags.h (enum symfile_add_flags)
<SYMFILE_NOT_FILENAME>: New constant.
* symfile.c (read_symbols): Use SYMFILE_NOT_FILENAME. Get
objfile name from BFD.
(symbol_file_add_with_addrs): Check SYMFILE_NOT_FILENAME.
* minidebug.c (find_separate_debug_file_in_section): Put
".gnu_debugdata" into BFD's file name.
This patch if the first patch in a series to add the ability to add constraints
to system registers that an instruction must adhere to in order for the register
to be usable with that instruction.
These constraints can also be used to disambiguate between registers with the
same encoding during disassembly.
This patch adds a new flags entry in the sysreg structures and ensures it is
filled in and read out during assembly/disassembly. It also adds the ability for
the assemble and disassemble functions to be able to gracefully fail and re-use
the existing error reporting infrastructure.
The return type of these functions are changed to a boolean to denote success or
failure and the error structure is passed around to them. This requires
aarch64-gen changes so a lot of the changes here are just mechanical.
gas/
PR binutils/21446
* config/tc-aarch64.c (parse_sys_reg): Return register flags.
(parse_operands): Fill in register flags.
gdb/
PR binutils/21446
* aarch64-tdep.c (aarch64_analyze_prologue,
aarch64_software_single_step, aarch64_displaced_step_copy_insn):
Indicate not interested in errors.
include/
PR binutils/21446
* opcode/aarch64.h (aarch64_opnd_info): Change sysreg to struct.
(aarch64_decode_insn): Accept error struct.
opcodes/
PR binutils/21446
* aarch64-asm.h (aarch64_insert_operand, aarch64_##x): Return boolean
and take error struct.
* aarch64-asm.c (aarch64_ext_regno, aarch64_ins_reglane,
aarch64_ins_reglist, aarch64_ins_ldst_reglist,
aarch64_ins_ldst_reglist_r, aarch64_ins_ldst_elemlist,
aarch64_ins_advsimd_imm_shift, aarch64_ins_imm, aarch64_ins_imm_half,
aarch64_ins_advsimd_imm_modified, aarch64_ins_fpimm,
aarch64_ins_imm_rotate1, aarch64_ins_imm_rotate2, aarch64_ins_fbits,
aarch64_ins_aimm, aarch64_ins_limm_1, aarch64_ins_limm,
aarch64_ins_inv_limm, aarch64_ins_ft, aarch64_ins_addr_simple,
aarch64_ins_addr_regoff, aarch64_ins_addr_offset, aarch64_ins_addr_simm,
aarch64_ins_addr_simm10, aarch64_ins_addr_uimm12,
aarch64_ins_simd_addr_post, aarch64_ins_cond, aarch64_ins_sysreg,
aarch64_ins_pstatefield, aarch64_ins_sysins_op, aarch64_ins_barrier,
aarch64_ins_prfop, aarch64_ins_hint, aarch64_ins_reg_extended,
aarch64_ins_reg_shifted, aarch64_ins_sve_addr_ri_s4xvl,
aarch64_ins_sve_addr_ri_s6xvl, aarch64_ins_sve_addr_ri_s9xvl,
aarch64_ins_sve_addr_ri_s4, aarch64_ins_sve_addr_ri_u6,
aarch64_ins_sve_addr_rr_lsl, aarch64_ins_sve_addr_rz_xtw,
aarch64_ins_sve_addr_zi_u5, aarch64_ext_sve_addr_zz,
aarch64_ins_sve_addr_zz_lsl, aarch64_ins_sve_addr_zz_sxtw,
aarch64_ins_sve_addr_zz_uxtw, aarch64_ins_sve_aimm,
aarch64_ins_sve_asimm, aarch64_ins_sve_index, aarch64_ins_sve_limm_mov,
aarch64_ins_sve_quad_index, aarch64_ins_sve_reglist,
aarch64_ins_sve_scale, aarch64_ins_sve_shlimm, aarch64_ins_sve_shrimm,
aarch64_ins_sve_float_half_one, aarch64_ins_sve_float_half_two,
aarch64_ins_sve_float_zero_one, aarch64_opcode_encode): Likewise.
* aarch64-dis.h (aarch64_extract_operand, aarch64_##x): Likewise.
* aarch64-dis.c (aarch64_ext_regno, aarch64_ext_reglane,
aarch64_ext_reglist, aarch64_ext_ldst_reglist,
aarch64_ext_ldst_reglist_r, aarch64_ext_ldst_elemlist,
aarch64_ext_advsimd_imm_shift, aarch64_ext_imm, aarch64_ext_imm_half,
aarch64_ext_advsimd_imm_modified, aarch64_ext_fpimm,
aarch64_ext_imm_rotate1, aarch64_ext_imm_rotate2, aarch64_ext_fbits,
aarch64_ext_aimm, aarch64_ext_limm_1, aarch64_ext_limm, decode_limm,
aarch64_ext_inv_limm, aarch64_ext_ft, aarch64_ext_addr_simple,
aarch64_ext_addr_regoff, aarch64_ext_addr_offset, aarch64_ext_addr_simm,
aarch64_ext_addr_simm10, aarch64_ext_addr_uimm12,
aarch64_ext_simd_addr_post, aarch64_ext_cond, aarch64_ext_sysreg,
aarch64_ext_pstatefield, aarch64_ext_sysins_op, aarch64_ext_barrier,
aarch64_ext_prfop, aarch64_ext_hint, aarch64_ext_reg_extended,
aarch64_ext_reg_shifted, aarch64_ext_sve_addr_ri_s4xvl,
aarch64_ext_sve_addr_ri_s6xvl, aarch64_ext_sve_addr_ri_s9xvl,
aarch64_ext_sve_addr_ri_s4, aarch64_ext_sve_addr_ri_u6,
aarch64_ext_sve_addr_rr_lsl, aarch64_ext_sve_addr_rz_xtw,
aarch64_ext_sve_addr_zi_u5, aarch64_ext_sve_addr_zz,
aarch64_ext_sve_addr_zz_lsl, aarch64_ext_sve_addr_zz_sxtw,
aarch64_ext_sve_addr_zz_uxtw, aarch64_ext_sve_aimm,
aarch64_ext_sve_asimm, aarch64_ext_sve_index, aarch64_ext_sve_limm_mov,
aarch64_ext_sve_quad_index, aarch64_ext_sve_reglist,
aarch64_ext_sve_scale, aarch64_ext_sve_shlimm, aarch64_ext_sve_shrimm,
aarch64_ext_sve_float_half_one, aarch64_ext_sve_float_half_two,
aarch64_ext_sve_float_zero_one, aarch64_opcode_decode): Likewise.
(determine_disassembling_preference, aarch64_decode_insn,
print_insn_aarch64_word, print_insn_data): Take errors struct.
(print_insn_aarch64): Use errors.
* aarch64-asm-2.c: Regenerate.
* aarch64-dis-2.c: Regenerate.
* aarch64-gen.c (print_operand_inserter): Use errors and change type to
boolean in aarch64_insert_operan.
(print_operand_extractor): Likewise.
* aarch64-opc.c (aarch64_print_operand): Use sysreg struct.
With native MIPS/Linux targets the $zero register is inaccessible, with
its supposed context slot provided by the OS occupied by the $restart
register. The PTRACE_GETREGS path takes care of it by artificially
supplying the hardwired contents of $zero in `mips_supply_gregset' or
`mips64_supply_gregset', as applicable, however the PTRACE_PEEKUSER
fallback does not, making the register unavailable, e.g.:
(gdb) info registers
zero at v0 v1 a0 a1 a2 a3
R0 <unavl> 00000001 00000001 d2f1a9fc 00000000 00000000 00417158 00417150
t0 t1 t2 t3 t4 t5 t6 t7
R8 00000004 00000000 fffffff8 00000000 00000000 00000000 00000001 00000007
s0 s1 s2 s3 s4 s5 s6 s7
R16 00000000 00405e30 00000000 00500000 00000000 0052ec08 00000000 00000000
t8 t9 k0 k1 gp sp s8 ra
R24 00000000 00417008 00000000 00000000 0041e220 7fff4ce0 7fff4ce0 00405d0c
status lo hi badvaddr cause pc
<unavl> 00441cf1 00000017 00417004 00800024 00405d10
fcsr fir restart
00800000 00f30000 00000000
(gdb)
or (under certain circumstances):
(gdb) stepi
Register 0 is not available
(gdb)
This is specifically because `mips_linux_register_addr' and
`mips64_linux_register_addr', both correctly return -1 for
MIPS_ZERO_REGNUM, and therefore `linux_nat_trad_target::fetch_registers'
faithfully marks this register as unavailable.
Supply this register artificially then in the PTRACE_PEEKUSER case as
well, correcting this issue.
gdb/
* mips-linux-nat.c (mips_linux_nat_target::fetch_registers):
Supply the MIPS_ZERO_REGNUM register.
Make the `mask_address_var' variable static, it is not used outside
mips-tdep.c and having no target name embedded within it causes a risk
of a namespace clash.
gdb/
* mips-tdep.c (mask_address_var): Make variable static.
Fix a commit f90183d7e3 ("Get GDBserver pid on remote target") bug and
correctly handle the case where the PID of `gdbserver' could not have
been retrieved. If that happens, $server_pid is unset causing:
FAIL: gdb.server/server-kill.exp: p server_pid
ERROR: tcl error sourcing .../gdb/testsuite/gdb.server/server-kill.exp.
ERROR: can't read "server_pid": no such variable
while executing
"if {$server_pid == "" } {
return -1
}"
(file ".../gdb/testsuite/gdb.server/server-kill.exp" line 49)
invoked from within
"source .../gdb/testsuite/gdb.server/server-kill.exp"
("uplevel" body line 1)
invoked from within
"uplevel #0 source .../gdb/testsuite/gdb.server/server-kill.exp"
invoked from within
"catch "uplevel #0 source $test_file_name""
Verify that the variable exists then rather than trying to access it.
gdb/testsuite/
* gdb.server/server-kill.exp: Verify whether `server_pid' exists
rather then trying to access it in determining whether the PID
of `gdbserver' could have been retrieved.
It turns out that a dwarf2_cu can remain allocated after psymtab
expansion is done, and so it makes sense to clear rust_unions when
done processing it.
Tested on x86-64 Fedora 27.
2018-05-14 Tom Tromey <tom@tromey.com>
* dwarf2read.c (rust_union_quirks): Clear rust_unions.
In commit:
commit 8ee22052f6
Author: Andrew Burgess <andrew.burgess@embecosm.com>
Date: Thu May 3 17:46:14 2018 +0100
gdb/x86: Handle kernels using compact xsave format
in two places FXSAVE_ADDR was used instead of FXSAVE_MXCSR_ADDR to get
the address of the mxcsr register within the xsave buffer. This will
mean we are potentially accessing the wrong location within the xsave
buffer.
There are no tests included with this patch. The first mistake would
only trigger an issue if/when the user tries to manually set the mxcsr
register to a value that matches the random (value off stack) value
that is in the xsave buffer, in this case the change by the user will
go unnoticed by GDB, and the default value of mxcsr will be preserved.
The second mistake only happens on the code path where all x87
registers are being written out of the register cache. I'm not sure
how to trigger that code path.
gdb/ChangeLog:
* i387-tdep.c (i387_collect_xsave): Use FXSAVE_MXCSR_ADDR not
FXSAVE_ADDR for the mxcsr register.
gdb_target definitions were removed from configure.tgt in 2007, before
xtensa port was merged. Remove it from the xtensa target as well.
gdb/
2018-05-11 Max Filippov <jcmvbkbc@gmail.com>
* configure.tgt (xtensa*-*-linux*): Drop gdb_target definition.
This gets rid of the core_ops global, and replaces it with
heap-allocated core_target instances. In practice, there will only be
one such instance, though that will change further ahead as more
pieces of multi-target support are merged.
Notice that this replaces one heap-allocated object for another, the
number of allocations is the same. Specifically, currently we
heap-allocate the 'core_data' object, which holds the core's section
table. With this patch, that object is made a field of the
core_target class, and no longer allocated separately.
Note that this bit:
- /* Looks semi-reasonable. Toss the old core file and work on the
- new. */
-
- unpush_target (&core_ops);
does not need a replacement, because by the time we get here, the
target_preopen call at the top of core_target_open has already
unpushed any previous target.
gdb/ChangeLog:
2018-05-11 Pedro Alves <palves@redhat.com>
* corelow.c (core_target) <core_target>: No longer inline.
Initialize m_core_gdbarch, m_core_vec and build the section table
here.
<~core_target>: New.
<core_gdbarch, get_core_register_section>: New methods.
<m_core_section_table, m_core_vec, m_core_gdbarch>: New fields,
factored out from ...
<core_data, core_vec, core_gdbarch>: ... these deleted globals.
(core_ops): Delete.
(sniff_core_bfd): Add gdbarch parameter.
(core_close): Delete, merged into ...
(core_target::close): ... here. Delete self.
(core_close_cleanup): Delete.
(core_target_open): Allocate a core_target on the heap. Use a
unique_ptr instead of a cleanup. Bits moved into the core_target
ctor. Adjust to use core_target methods instead of globals.
(get_core_register_section): Rename to ...
(core_target::get_core_register_section): ... this and adjust.
(struct get_core_registers_cb_data): New.
(get_core_registers_cb): Use it. Use bool.
(core_target::fetch_registers, core_target::files_info)
(core_target::xfer_partial, core_target::read_description)
(core_target::pid_to, core_target::thread_name): Adjust to
reference class fields instead of globals.
* target.h (struct target_ops_deleter, target_ops_up): New.
(previously called 'core_target', but since renamed because
'core_target' is the name of the target_ops class now.)
This eliminates the "the_core_target" global, as preparation for being
able to have more than one core loaded. When we get there, we will
instantiate one core_target object per core instead.
Essentially, this replaces the reference to the_core_target in
core_file_command by a reference to core_bfd, which is per
program_space.
Currently, core_file_command calls 'the_core_target->detach()' even if
the core target is not open and pushed on the target stack. If it is
indeed not open, then the practical effect is that
core_target::detach() prints "No core file now.". That is preserved
by printing that directly from within core_file_command if not
debugging a core.
gdb/ChangeLog:
2018-05-11 Pedro Alves <palves@redhat.com>
* corefile.c (core_file_command): Move to corelow.c.
* corelow.c (the_core_target): Delete.
(core_file_command): Moved from corefile.c. Check exec_bfd
instead of the_core_target. Use target_detach instead of calling
into the_core_target directly.
(maybe_say_no_core_file_now): New.
(core_target::detach): Use it.
(_initialize_corelow): Remove references to the_core_target.
* gdbcore.h (the_core_target): Delete.
This moves the core_bfd global to be a field of the program space. It
then replaces core_bfd with a macro to avoid a massive patch -- the
same approach taken for various other program space fields.
This is a basic transformation for multi-target work.
2018-05-11 Tom Tromey <tromey@redhat.com>
Pedro Alves <tromey@redhat.com>
* corefile.c (core_bfd): Remove.
* gdbcore.h (core_bfd): Now a macro.
* progspace.h (struct program_space) <cbfd>: New field.
This removes the remaining cleanups from mdebugread.c, replacing them
with gdb::def_vector.
Tested by the buildbot, though I doubt this exercises mdebugread.
gdb/ChangeLog
2018-05-11 Tom Tromey <tom@tromey.com>
* mdebugread.c (parse_partial_symbols, psymtab_to_symtab_1): Use
gdb::def_vector.
The floating point context structure on x86 LynxOS-178 is not
the same as on LynxOS 5.x. As a consequence, trying to print
the return value of a function returning a float, for instance,
yields incorrect results.
This patch fixes the issue by providing an updated definition
for LynxOS-178 (the reason why we cannot access the actual definition
provided by the system still remains true).
gdb/gdbserver/ChangeLog:
* lynx-i386-low.c (LYNXOS_178): New macro.
[LYNXOS_178] (usr_fcontext_t): Provide a definition that matches
the layout on LynxOS-178.
(lynx_i386_fill_fpregset, lynx_i386_store_fpregset): Do not
handle floating point registers that are not supported by
LynxOS-178.
Simon pointed out that gdb would not build with clang, due to the
addition of -Wimplicit-fallthrough. This patch fixes the problem by
using -Wimplicit-fallthrough=3 -- this does not work with clang,
bypassing the issue.
Tested by rebuilding with both gcc and clang; and also by verifying
that -Wimplicit-fallthrough=3 is used in the gcc build.
I will file a follow-up bug to convert the fall-through comments to a
form that can be used by both clang and gcc.
gdb/ChangeLog
2018-05-10 Tom Tromey <tom@tromey.com>
* configure: Rebuild.
* warning.m4 (AM_GDB_WARNINGS): Use -Wimplicit-fallthrough=3.
gdb/gdbserver/ChangeLog
2018-05-10 Tom Tromey <tom@tromey.com>
* configure: Rebuild.
On Windows, starting a new process with GDBserver seems to work,
in the sense that the program does get started, and GDBserver
confirms that it is listening for GDB to connect. However, as soon as
GDB establishes the connection with GDBserver, and starts discussing
with it, GDBserver crashes, with a SEGV.
This SEGV occurs in remote-utils.c::prepare_resume_reply...
| regp = current_target_desc ()->expedite_regs;
| [...]
| while (*regp)
... because, in our case, REGP is NULL.
This patches fixes the issues by adding a parameter to init_target_desc,
in order to make sure that we always provide the list of registers when
we initialize a target description.
gdb/ChangeLog:
PR server/23158:
* regformats/regdat.sh: Adjust script, following the addition
of the new expedite_regs parameter to init_target_desc.
gdb/gdbserver/ChangeLog:
PR server/23158:
* tdesc.h (init_target_desc) <expedite_regs>: New parameter.
* tdesc.c (init_target_desc) <expedite_regs>: New parameter.
Use it to set the expedite_regs field in the given tdesc.
* x86-tdesc.h: New file.
* linux-aarch64-tdesc.c (aarch64_linux_read_description):
Adjust following the addition of the new expedite_regs parameter
to init_target_desc.
* linux-tic6x-low.c (tic6x_read_description): Likewise.
* linux-x86-tdesc.c: #include "x86-tdesc.h".
(i386_linux_read_description, amd64_linux_read_description):
Adjust following the addition of the new expedite_regs parameter
to init_target_desc.
* lynx-i386-low.c: #include "x86-tdesc.h".
(lynx_i386_arch_setup): Adjust following the addition of the new
expedite_regs parameter to init_target_desc.
* nto-x86-low.c: #include "x86-tdesc.h".
(nto_x86_arch_setup): Adjust following the addition of the new
expedite_regs parameter to init_target_desc.
* win32-i386-low.c: #include "x86-tdesc.h".
(i386_arch_setup): Adjust following the addition of the new
expedite_regs parameter to init_target_desc.
Trying to start a program with GDBserver on Windows yields
the following error:
$ gdbserver.exe --once :4444 simple_main.exe
Killing process(es): 5008
No program to debug
Exiting
The error itself comes from the following code shortly after
create_inferior gets called (in server.c::main):
/* Wait till we are at first instruction in program. */
create_inferior (program_path.get (), program_args);
[...]
if (last_status.kind == TARGET_WAITKIND_EXITED
|| last_status.kind == TARGET_WAITKIND_SIGNALLED)
was_running = 0;
else
was_running = 1;
if (!was_running && !multi_mode)
error ("No program to debug");
What happens is that the "last_status" global starts initialized
as zeroes, which means last_status.kind == TARGET_WAITKIND_EXITED,
and we expect create_inferior to be waiting for the inferior to
start until reaching the SIGTRAP, and to set the "last_status"
global to match that last event we received.
I suspect this is an unintended side-effect of the following change...
commit 2090129c36
Date: Thu Dec 22 21:11:11 2016 -0500
Subject: Share fork_inferior et al with gdbserver
... which removes some code in server.c that was responsible for
starting the inferior in a functin that was named start_inferior,
and looked like this:
signal_pid = create_inferior (new_argv[0], &new_argv[0]);
[...]
/* Wait till we are at 1st instruction in program, return new pid
(assuming success). */
last_ptid = mywait (pid_to_ptid (signal_pid), &last_status, 0, 0);
The code has been transitioned to using fork_inferior, but sadly,
only for the targets that support it. On Windows, the calls to wait
setting "last_status" simply disappeared.
This patch adds it back in the Windows-specific implementation of
create_inferior.
gdb/gdbserver/ChangeLog:
PR server/23158:
* win32-low.c (win32_create_inferior): Add call to my_wait
setting last_status global.
Trying to start GDBserver on Windows currently yields the following
error...
$ gdbserver.exe --once :4444 simple_main.exe
glob could not process pattern '(null)'.
Exiting
... after which GDB terminates with a nonzero status.
This is because create_process in win32-low.c calls gdb_tilde_expand
with the result of a call to get_inferior_cwd without verifying that
the returned directory is not NULL:
| static BOOL
| create_process (const char *program, char *args,
| DWORD flags, PROCESS_INFORMATION *pi)
| {
| const char *inferior_cwd = get_inferior_cwd ();
| std::string expanded_infcwd = gdb_tilde_expand (inferior_cwd);
This patch avoids this by only calling gdb_tilde_expand when
INFERIOR_CWD is not NULL, which is similar to what is done on
GNU/Linux for instance.
gdb/gdbserver/ChangeLog:
PR server/23158:
* win32-low.c (create_process): Only call gdb_tilde_expand if
inferior_cwd is not NULL.
This patch fixes tagged pointer support for AArch64 GDB. Linux kernel
debugging failure was reported after tagged pointer support was committed.
After a discussion around best path forward to manage tagged pointers
on GDB side we are going to disable tagged pointers support for
aarch64-none-elf-gdb because for non-linux applications we cant be
sure if tagged pointers will be used by MMU or not.
Also for aarch64-linux-gdb we are going to sign extend user-space
address after clearing tag bits. This will help debug both kernel
and user-space addresses based on information from linux kernel
documentation given below:
According to AArch64 memory map:
https://www.kernel.org/doc/Documentation/arm64/memory.txt
"User addresses have bits 63:48 set to 0 while the kernel addresses have
the same bits set to 1."
According to AArch64 tagged pointers document:
https://www.kernel.org/doc/Documentation/arm64/tagged-pointers.txt
The kernel configures the translation tables so that translations made
via TTBR0 (i.e. userspace mappings) have the top byte (bits 63:56) of
the virtual address ignored by the translation hardware. This frees up
this byte for application use.
Running gdb testsuite after applying this patch introduces no regressions
and tagged pointer test cases still pass.
gdb/ChangeLog:
2018-05-10 Omair Javaid <omair.javaid@linaro.org>
PR gdb/23127
* aarch64-linux-tdep.c (aarch64_linux_init_abi): Add call to
set_gdbarch_significant_addr_bit.
* aarch64-tdep.c (aarch64_gdbarch_init): Remove call to
set_gdbarch_significant_addr_bit.
* utils.c (address_significant): Update to sign extend addr.
xtensa GDB may be used with both bare-metal and linux-based
applications. In case of bare-metal application gdbserver is able to
provide information about all CPU registers: both unprivileged and
privileged. In case of linux-based application only a small subset of
privileged state is available. Currently xtensa GDB only expects
unprivileged registers in 'g' packets and it fails to communicate with
server that sends both privileged and unprivileged registers.
Allow bare-metal xtensa GDB to deal with both privileged and
unprivileged registers by initializing tdep->num_regs with the total
number of target CPU registers. Keep linux-based xtensa GDB
functionality as is by copying tdep->num_nopriv_regs to tdep->num_regs.
gdb/
2018-05-09 Max Filippov <jcmvbkbc@gmail.com>
* xtensa-linux-tdep.c (xtensa-tdep.h): New include.
(xtensa_linux_init_abi): Limit tdep->num_regs by
tdep->num_nopriv_regs.
* xtensa-tdep.c (xtensa_derive_tdep): Calculate
tdep->num_nopriv_regs and only copy it to tdep->num_regs if it's
not initialized.
When building with x86_64-w64-mingw32-g++ (to test cross-compiling for
Windows), I get this error:
unittests/string_view-selftests.o: In function `selftests::string_view::inserters_2::test05(unsigned long long)':
/home/emaisin/src/binutils-gdb/gdb/unittests/basic_string_view/inserters/char/2.cc:60: undefined reference to `std::basic_ofstream<char, std::char_traits<char> >::rpl_close()'
This is caused by gnulib redefining "close" as "rpl_close", and
therefore messing up the declaration of basic_ofstream in the libstdc++
header. The solution would be to use gnulib namespaces [1]. Until we
use them across GDB, we can use them locally in files that are
problematic, like this one.
gdb/ChangeLog:
* unittests/string_view-selftests.c: Define GNULIB_NAMESPACE.
For GNU/Linux on x86-64, if the target is using the xsave format for
passing the floating-point information from the inferior then there
currently exists a bug relating to the x87 control registers, and the
mxcsr register.
The xsave format allows different floating-point features to be lazily
enabled, a bit in the xsave format tells GDB which floating-point
features have been enabled, and which have not.
Currently in GDB, when reading the floating point state, we check the
xsave bit flags, if the feature is enabled then we read the feature
from the xsave buffer, and if the feature is not enabled, then we
supply the default value from within GDB.
Within GDB, when writing the floating point state, we first fetch the
xsave state from the target and then, for any feature that is not yet
enabled, we write the default values into the xsave buffer. Next we
compare the regcache value with the value in the xsave buffer, and, if
the value has changed we update the value in the xsave buffer, and
mark the feature enabled in the xsave bit flags.
The problem then, is that the x87 control registers were not following
this pattern. We assumed that these registers were always written out
by the kernel, and we always wrote them out to the xsave buffer (but
didn't enabled the feature). The result of this is that if the kernel
had not yet enabled the x87 feature then within GDB we would see
random values for the x87 floating point control registers, and if the
user tried to modify one of these register, that modification would be
lost.
Finally, the mxcsr register was also broken in the same way as the x87
control registers. The added complexity with this case is that the
mxcsr register is part of both the avx and sse floating point feature
set. When reading or writing this register we need to check that at
least one of these features is enabled.
This bug was present in native GDB, and within gdbserver. Both are
fixed with this commit.
gdb/ChangeLog:
* common/x86-xstate.h (I387_FCTRL_INIT_VAL): New constant.
(I387_MXCSR_INIT_VAL): New constant.
* amd64-tdep.c (amd64_supply_xsave): Only read state from xsave
buffer if it was supplied by the inferior.
* i387-tdep.c (i387_supply_fsave): Use I387_MXCSR_INIT_VAL.
(i387_xsave_get_clear_bv): New function.
(i387_supply_xsave): Only read x87 control registers from the
xsave buffer if the feature is enabled, and the state will have
been written, otherwise, provide a suitable default.
(i387_collect_xsave): Pre-clear all registers in xsave buffer,
including x87 control registers. Update control registers if they
have changed from the default value, and mark features as enabled
as required.
* i387-tdep.h (i387_xsave_get_clear_bv): Declare.
gdb/gdbserver/ChangeLog:
* i387-fp.c (i387_cache_to_xsave): Only write x87 control
registers to the cache if their values have changed.
(i387_xsave_to_cache): Provide default values for x87 control
registers when these features are available, but disabled.
* regcache.c (supply_register_by_name_zeroed): New function.
* regcache.h (supply_register_by_name_zeroed): Declare new
function.
gdb/testsuite/ChangeLog:
* gdb.arch/amd64-init-x87-values.S: New file.
* gdb.arch/amd64-init-x87-values.exp: New file.
The formatting of the output of the "info spu event" command changed, causing
spurious test suite failures. Use phex instead of phex_nz to get back the
expected format, and fix emission of new line characters.
gdb/ChangeLog:
2018-05-08 Ulrich Weigand <uweigand@de.ibm.com>
* spu-tdep.c (info_spu_event_command): Fix output formatting.
This adds -Wduplicated-cond to warnings.m4. This caught one bug.
I tried adding -Wduplicated-branches as well, but it results in some
spurious failures from code like this in cgen.h:
#define CGEN_ATTR_TYPE(n) \
struct { unsigned int bool_; \
CGEN_ATTR_VALUE_TYPE nonbool[(n) ? (n) : 1]; }
This will trigger a warning if passed n==1, which seems like a
perfectly valid thing to do; and there were other issues like this as
well.
ChangeLog
2018-05-07 Tom Tromey <tom@tromey.com>
* configure: Rebuild.
* warning.m4 (AM_GDB_WARNINGS): Add -Wduplicated-cond.
gdbserver/ChangeLog
2018-05-07 Tom Tromey <tom@tromey.com>
* configure: Rebuild.
-Wduplicated-cond pointed out that arm_record_vfp_data_proc_insn
checks "opc1 == 0x0b" twice. I filed this a while ago as
PR tdep/20362.
Based on the ARM instruction manual at
https://www.scss.tcd.ie/~waldroj/3d1/arm_arm.pdf, I think the
instruction decoding in this function has two bugs.
First, opc1 is computed as:
opc1 = bits (arm_insn_r->arm_insn, 20, 23);
[...]
opc1 = opc1 & 0x04;
This means that tests like:
else if (opc1 == 0x01)
can never be true.
In the ARM manual, "opc1" corresponds to these bits:
name bit
r 20
q 21
D 22
p 23
... where the D bit is not used for VFP instruction decoding.
So, I believe this code should use ~0x04 instead.
Second, VDIV is recognized by the bits "pqrs" being equal to "1000".
This tranlates to opc1 == 0x08 -- not 0x0b. Note that pqrs==1001 is
an undefined encoding, which is probably why opc2 is not checked here;
this code doesn't seem to really deal with undefined encodings in
general, so I've left that as is.
I don't have an ARM machine or any reasonable way to test this.
ChangeLog
2018-05-07 Tom Tromey <tom@tromey.com>
PR tdep/20362:
* arm-tdep.c (arm_record_vfp_data_proc_insn): Properly mask off D
bit. Use correct value for VDIV.
This adds -Wimplicit-fallthrough to the set of default warnings.
2018-05-04 Tom Tromey <tom@tromey.com>
* configure: Rebuild.
* warning.m4 (AM_GDB_WARNINGS): Add -Wimplicit-fallthrough.
gdbserver/ChangeLog
2018-05-04 Tom Tromey <tom@tromey.com>
* configure: Rebuild.
This adds a "break" at the end of the RECORD_SYS_RECVFROM case in
record_linux_system_call. This seemed correct to me.
2018-05-04 Tom Tromey <tom@tromey.com>
* linux-record.c (record_linux_system_call) <case
RECORD_SYS_RECVFROM>: Add "break".
This adds a "break" to a couple of spots where it was erroneously
omitted. I think these are the two (potential) real bugs caught by
this series.
2018-05-04 Tom Tromey <tom@tromey.com>
* mi/mi-main.c (mi_cmd_trace_frame_collected) <REGISTERS_FORMAT>:
Add missing "break".
* mi/mi-cmd-stack.c (mi_cmd_stack_list_locals) <NO_FRAME_FILTERS>:
Add missing "break".
This adds two fall-through comments in rs6000-tdep.c. I looked at the
PPC instruction manual and convinced myself that this was correct.
And, this isn't a semantic change. However, close review would still
be good.
2018-05-04 Tom Tromey <tom@tromey.com>
* rs6000-tdep.c (ppc_process_record_op4)
(ppc_process_record_op63): Add fall-through comment.
This adds a fall-through comment in i386-tdep.c. I was not sure what
to do here, so I elected to preserve the status quo. In review, John
Baldwin pointed out that: "I believe this is correct based on the diff
that added the special cases for xgetbv and xsetbv as previously ldgt
and lidt were treated the same".
gdb/ChangeLog
2018-05-04 Tom Tromey <tom@tromey.com>
* i386-tdep.c (i386_process_record): Add fall-through comment.
This adds a fall-through comment to stabsread.c. I skimmed the stabs
manual a bit and it seems that 'p' and 'P' are similar enough that
this makes sense. Also, stabs is mostly deprecated, and the code has
been this way for a long time, so it seemed safest to keep the status
quo.
ChangeLog
2018-05-04 Tom Tromey <tom@tromey.com>
* stabsread.c (define_symbol) <case 'p'>: Add fall-through
comment.
This patch fixes the subset of -Wimplicit-fallthrough warnings that I
considered obvious. In most cases it was obvious from context that
falling through was desired; here I added the appropriate comment. In
a couple of cases it seemed clear that a "break" was missing.
ChangeLog
2018-05-04 Tom Tromey <tom@tromey.com>
* riscv-tdep.c (riscv_isa_xlen): Add fall-through comment.
* utils.c (can_dump_core) <LIMIT_CUR>: Add fall-through comment.
* eval.c (fetch_subexp_value) <MEMORY_ERROR>: Add fall-through
comment.
* d-valprint.c (d_val_print) <TYPE_CODE_STRUCT>: Add fall-through
comment.
* coffread.c (coff_symtab_read) <C_LABEL>: Add fall-through
comment.
This patch adds a missing ATTRIBUTE_NORETURN. This lets
-Wimplicit-fallthrough recognize that a given case does not fall
through.
ChangeLog
2018-05-04 Tom Tromey <tom@tromey.com>
* dwarf2loc.c (unimplemented): Add ATTRIBUTE_NORETURN.
This changes the Python API so that breakpoint commands can be set by
writing to the "commands" attribute.
ChangeLog
2018-05-04 Tom Tromey <tom@tromey.com>
PR python/22731:
* NEWS: Mention that breakpoint commands are writable.
* python/py-breakpoint.c (bppy_set_commands): New function.
(breakpoint_object_getset) <"commands">: Use it.
doc/ChangeLog
2018-05-04 Tom Tromey <tom@tromey.com>
PR python/22731:
* python.texi (Breakpoints In Python): Mention that "commands" is
writable.
testsuite/ChangeLog
2018-05-04 Tom Tromey <tom@tromey.com>
PR python/22731:
* gdb.python/py-breakpoint.exp: Test setting breakpoint commands.
This changes some functions in cli-script.c to use function_view
rather than a function pointer and closure argument. This simplifies
the code a bit and is useful in a subsequent patch.
ChangeLog
2018-05-04 Tom Tromey <tom@tromey.com>
* tracepoint.c (actions_command): Update.
* mi/mi-cmd-break.c (mi_command_line_array)
(mi_command_line_array_cnt, mi_command_line_array_ptr)
(mi_read_next_line): Remove.
(mi_cmd_break_commands): Update.
* cli/cli-script.h (read_command_lines, read_command_lines_1): Use
function_view.
* cli/cli-script.c (get_command_line): Update.
(process_next_line): Use function_view. Constify.
(recurse_read_control_structure, read_command_lines)
(read_command_lines_1): Change argument types to function_view.
(do_define_command, document_command): Update.
* breakpoint.h (check_tracepoint_command): Don't declare.
* breakpoint.c (check_tracepoint_command): Remove.
(commands_command_1, create_tracepoint_from_upload): Update.
PR gdb/11750 concerns defining a command inside a user commnad, like:
define outer
define inner
echo hi\n
end
end
This patch adds this capability to gdb.
ChangeLog
2018-05-04 Tom Tromey <tom@tromey.com>
PR gdb/11750:
* cli/cli-script.h (enum command_control_type) <define_control>:
New constant.
* cli/cli-script.c (multi_line_command_p): Handle define_control.
(build_command_line, execute_control_command_1)
(process_next_line): Likewise.
(do_define_command): New function, extracted from define_command.
(define_command): Use it.
testsuite/ChangeLog
2018-05-04 Tom Tromey <tom@tromey.com>
PR gdb/11750:
* gdb.base/define.exp: Test defining a user command inside a user
command.
* gdb.base/commands.exp (define_if_without_arg_test): Test "define".
The prompt argument to read_command_lines can be const. This patch
makes this change, and also removes some fixed-sized buffers in favor
of using string_printf.
ChangeLog
2018-05-04 Tom Tromey <tom@tromey.com>
* tracepoint.c (actions_command): Update.
* cli/cli-script.h (read_command_lines): Update.
* cli/cli-script.c (read_command_lines): Constify prompt_arg.
(MAX_TMPBUF): Remove define.
(define_command): Use string_printf.
(document_command): Likewise.
* breakpoint.c (commands_command_1): Update.
I noticed some code in execute_control_command_1 that could be
simplified by making print_command_trace a printf-like function. This
patch makes this change.
ChangeLog
2018-05-04 Tom Tromey <tom@tromey.com>
* top.c (execute_command): Update.
* cli/cli-script.h (print_command_lines): Now varargs.
* cli/cli-script.c (print_command_lines): Now varargs.
(execute_control_command_1) <case while_control, case if_control>:
Update.
Currently command lines are reference counted using shared_ptr only
when attached to breakpoints. This patch changes gdb to use
shared_ptr in commands as well. This allows for the removal of
copy_command_lines.
Note that the change to execute_user_command explicitly makes a new
reference to the command line. This will be used in a later patch.
This simplifies struct command_line based on the observation that a
given command can have at most two child bodies: an "if" can have both
"then" and "else" parts. Perhaps the names I've chosen for the
replacements here are not very good -- your input requested.
ChangeLog
2018-05-04 Tom Tromey <tom@tromey.com>
* tracepoint.c (all_tracepoint_actions): Rename from
all_tracepoint_actions_and_cleanup. Change return type.
(actions_command, encode_actions_1, encode_actions)
(trace_dump_actions, tdump_command): Update.
* remote.c (remote_download_command_source): Update.
* python/python.c (gdbpy_eval_from_control_command)
(python_command, python_interactive_command): Update.
* mi/mi-cmd-break.c (mi_cmd_break_commands): Update.
* guile/guile.c (guile_command)
(gdbscm_eval_from_control_command, guile_command): Update.
* compile/compile.c (compile_code_command)
(compile_print_command, compile_to_object): Update.
* cli/cli-script.h (struct command_lines_deleter): New.
(counted_command_line): New typedef.
(struct command_line): Add constructor, destructor.
<body_list>: Remove.
<body_list_0, body_list_1>: New members.
(command_line_up): Remove typedef.
(read_command_lines, read_command_lines_1, get_command_line):
Update.
(copy_command_lines): Don't declare.
* cli/cli-script.c (build_command_line): Use "new".
(get_command_line): Return counted_command_line.
(print_command_lines, execute_user_command)
(execute_control_command_1, while_command, if_command): Update.
(realloc_body_list): Remove.
(process_next_line, recurse_read_control_structure): Update.
(read_command_lines, read_command_lines_1): Return counted_command_line.
(free_command_lines): Use "delete".
(copy_command_lines): Remove.
(define_command, document_command, show_user_1): Update.
* cli/cli-decode.h (struct cmd_list_element) <user_commands>: Now
a counted_command_line.
* breakpoint.h (counted_command_line): Remove typedef.
(breakpoint_set_commands): Update.
* breakpoint.c (check_no_tracepoint_commands)
(validate_commands_for_breakpoint): Update.
(breakpoint_set_commands): Change commands to be a
counted_command_line.
(commands_command_1, update_dprintf_command_list)
(create_tracepoint_from_upload): Update.
Tom Tromey
Erik Kurzinger
Pedro Alves
Maciej W. Rozycki
Pedro Franco de Carvalho
Joel Brobecker
Simon Marchi
Simon Marchi
Philippe Waroquiers
Keith Seitz
Tamar Christina
Andrew Burgess
Max Filippov
Tom Tromey
Omair Javaid
Jan Kratochvil
Ulrich Weigand