This commit adds support for source files that contain non utf-8
characters when performing source styling using the Python pygments
package. This does not change the behaviour of GDB when the GNU
Source Highlight library is used.
For the following problem description, assume that either GDB is built
without GNU Source Highlight support, of that this has been disabled
using 'maintenance set gnu-source-highlight enabled off'.
The initial problem reported was that a source file containing non
utf-8 characters would cause GDB to print a Python exception, and then
display the source without styling, e.g.:
Python Exception <class 'UnicodeDecodeError'>: 'utf-8' codec can't decode byte 0xc0 in position 142: invalid start byte
/* Source code here, without styling... */
Further, as the user steps through different source files, each time
the problematic source file was evicted from the source cache, and
then later reloaded, the exception would be printed again.
Finally, this problem is only present when using Python 3, this issue
is not present for Python 2.
What makes this especially frustrating is that GDB can clearly print
the source file contents, they're right there... If we disable
styling completely, or make use of the GNU Source Highlight library,
then everything is fine. So why is there an error when we try to
apply styling using Python?
The problem is the use of PyString_FromString (which is an alias for
PyUnicode_FromString in Python 3), this function converts a C string
into a either a Unicode object (Py3) or a str object (Py2). For
Python 2 there is no unicode encoding performed during this function
call, but for Python 3 the input is assumed to be a uft-8 encoding
string for the purpose of the conversion. And here of course, is the
problem, if the source file contains non utf-8 characters, then it
should not be treated as utf-8, but that's what we do, and that's why
we get an error.
My first thought when looking at this was to spot when the
PyString_FromString call failed with a UnicodeDecodeError and silently
ignore the error. This would mean that GDB would print the source
without styling, but would also avoid the annoying exception message.
However, I also make use of `pygmentize`, a command line wrapper
around the Python pygments module, which I use to apply syntax
highlighting in the output of `less`. And this command line wrapper
is quite happy to syntax highlight my source file that contains non
utf-8 characters, so it feels like the problem should be solvable.
It turns out that inside the pygments module there is already support
for guessing the encoding of the incoming file content, if the
incoming content is not already a Unicode string. This is what
happens for Python 2 where the incoming content is of `str` type.
We could try and make GDB smarter when it comes to converting C
strings into Python Unicode objects; this would probably require us to
just try a couple of different encoding schemes rather than just
giving up after utf-8.
However, I figure, why bother? The pygments module already does this
for us, and the colorize API is not part of the documented external
API of GDB. So, why not just change the colorize API, instead of the
content being a Unicode string (for Python 3), lets just make the
content be a bytes object. The pygments module can then take
responsibility for guessing the encoding.
So, currently, the colorize API receives a unicode object, and returns
a unicode object. I propose that the colorize API receive a bytes
object, and return a bytes object.
Add a new function gdb.history_count to the Python api, this function
returns an integer, the number of items in GDB's value history.
This is useful if you want to pull items from the history by their
absolute number, for example, if you wanted to show a complete history
list. Previously we could figure out how many items are in the
history list by trying to fetch the items, and then catching the
exception when the item is not available, but having this function
seems nicer.
Currently, gdb's Python layer captures the current architecture and
language when "entering" Python code. This has some undesirable
effects, and so this series changes how this is handled.
First, there is code like this:
gdbpy_enter enter_py (python_gdbarch, python_language);
This is incorrect, because both of these are NULL when not otherwise
assigned. This can cause crashes in some cases -- I've added one to
the test suite. (Note that this crasher is just an example, other
ones along the same lines are possible.)
Second, when the language is captured in this way, it means that
Python code cannot affect the current language for its own purposes.
It's reasonable to want to write code like this:
gdb.execute('set language mumble')
... stuff using the current language
gdb.execute('set language previous-value')
However, this won't actually work, because the language is captured on
entry. I've added a test to show this as well.
This patch changes gdb to try to avoid capturing the current values.
The Python concept of the current gdbarch is only set in those few
cases where a non-default value is computed or needed; and the
language is not captured at all -- instead, in the cases where it's
required, the current language is temporarily changed.
This commit brings all the changes made by running gdb/copyright.py
as per GDB's Start of New Year Procedure.
For the avoidance of doubt, all changes in this commits were
performed by the script.
This commit adds a new object type gdb.TargetConnection. This new
type represents a connection within GDB (a connection as displayed by
'info connections').
There's three ways to find a gdb.TargetConnection, there's a new
'gdb.connections()' function, which returns a list of all currently
active connections.
Or you can read the new 'connection' property on the gdb.Inferior
object type, this contains the connection for that inferior (or None
if the inferior has no connection, for example, it is exited).
Finally, there's a new gdb.events.connection_removed event registry,
this emits a new gdb.ConnectionEvent whenever a connection is removed
from GDB (this can happen when all inferiors using a connection exit,
though this is not always the case, depending on the connection type).
The gdb.ConnectionEvent has a 'connection' property, which is the
gdb.TargetConnection being removed from GDB.
The gdb.TargetConnection has an 'is_valid()' method. A connection
object becomes invalid when the underlying connection is removed from
GDB (as discussed above, this might be when all inferiors using a
connection exit, or it might be when the user explicitly replaces a
connection in GDB by issuing another 'target' command).
The gdb.TargetConnection has the following read-only properties:
'num': The number for this connection,
'type': e.g. 'native', 'remote', 'sim', etc
'description': The longer description as seen in the 'info
connections' command output.
'details': A string or None. Extra details for the connection, for
example, a remote connection's details might be
'hostname:port'.
In this commit:
commit c6a6aad52d
Date: Mon Oct 25 17:25:45 2021 +0100
gdb/python: make some global variables static
building without Python was broken. The extension_language_python
global was moved from being always defined, to only being defined when
the HAVE_PYTHON macro was defined. As a consequence, building without
Python support would result in errors like:
/usr/bin/ld: extension.o:(.rodata+0x120): undefined reference to `extension_language_python'
This commit fixes the problem by moving the definition of
extension_language_python outside of the HAVE_PYTHON macro protection.
Make a couple of global variables static in python/python.c. To do
this I had to move the definition of extension_language_python to
later in the file.
There should be no user visible changes after this commit.
There's a common pattern to call add_basic_prefix_cmd and
add_show_prefix_cmd to add matching set and show commands. Add the
add_setshow_prefix_cmd function to factor that out and use it at a few
places.
Change-Id: I6e9e90a30e9efb7b255bf839cac27b85d7069cfd
In a future commit I'm going to be creating gdb.Membuf objects from a
new file within gdb/python/py*.c. Currently all gdb.Membuf objects
are created directly within infpy_read_memory (as a result of calling
gdb.Inferior.read_memory()).
Initially I split out the Membuf creation code into a new function,
and left the new function in gdb/python/py-inferior.c, however, it
felt a little random that the Membuf creation code should live with
the inferior handling code.
So, then I moved all of the Membuf related code out into a new file,
gdb/python/py-membuf.c, the interface is gdbpy_buffer_to_membuf, which
wraps an array of bytes into a gdb.Membuf object.
Most of the code is moved directly from py-inferior.c with only minor
tweaks to layout and replacing NULL with nullptr, hence, I've left the
copyright date on py-membuf.c as 2009-2021 to match py-inferior.c.
Currently, the only user of this code is still py-inferior.c, but in
later commits this will change.
There should be no user visible changes after this commit.
Add a new function to the Python API, gdb.architecture_names(). This
function returns a list containing all of the supported architecture
names within the current build of GDB.
The values returned in this list are all of the possible values that
can be returned from gdb.Architecture.name().
The test-case gdb.gdb/python-interrupts.exp:
- runs to captured_command_loop
- sets a breakpoint at set_active_ext_lang
- calls a python command
- verifies the command triggers the breakpoint
- sends a signal and verifies the result
The test-case is fragile, because (f.i. with -flto) it cannot be guaranteed
that captured_command_loop and set_active_ext_lang are available for setting
breakpoints.
Reimplement the test-case as unittest, using:
- execute_command_to_string to capture the output
- try/catch to catch the "Error while executing Python code" exception
- a new hook selftests::hook_set_active_ext_lang to raise the signal
Tested on x86_64-linux.
The pattern for using execute_command_to_string is:
...
std::string output;
output = execute_fn_to_string (fn, term_out);
...
This results in a problem when using it in a try/catch:
...
try
{
output = execute_fn_to_string (fn, term_out)
}
catch (const gdb_exception &e)
{
/* Use output. */
}
...
If an expection was thrown during execute_fn_to_string, then the output
remains unassigned, while it could be worthwhile to known what output was
generated by gdb before the expection was thrown.
Fix this by returning the string using a parameter instead:
...
execute_fn_to_string (output, fn, term_out)
...
Also add a variant without string parameter, to support places where the
function is used while ignoring the result:
...
execute_fn_to_string (fn, term_out)
...
Tested on x86_64-linux.
Add a new event, gdb.events.gdb_exiting, which is called once GDB
decides it is going to exit.
This event is not triggered in the case that GDB performs a hard
abort, for example, when handling an internal error and the user
decides to quit the debug session, or if GDB hits an unexpected,
fatal, signal.
This event is triggered if the user just types 'quit' at the command
prompt, or if GDB is run with '-batch' and has processed all of the
required commands.
The new event type is gdb.GdbExitingEvent, and it has a single
attribute exit_code, which is the value that GDB is about to exit
with.
The event is triggered before GDB starts dismantling any of its own
internal state, so, my expectation is that most Python calls should
work just fine at this point.
When considering this functionality I wondered about using the
'atexit' Python module. However, this is triggered when the Python
environment is shut down, which is done from a final cleanup. At
this point we don't know for sure what other GDB state has already
been cleaned up.
String-like settings (var_string, var_filename, var_optional_filename,
var_string_noescape) currently take a pointer to a `char *` storage
variable (typically global) that holds the setting's value. I'd like to
"mordernize" this by changing them to use an std::string for storage.
An obvious reason is that string operations on std::string are often
easier to write than with C strings. And they avoid having to do any
manual memory management.
Another interesting reason is that, with `char *`, nullptr and an empty
string often both have the same meaning of "no value". String settings
are initially nullptr (unless initialized otherwise). But when doing
"set foo" (where `foo` is a string setting), the setting now points to
an empty string. For example, solib_search_path is nullptr at startup,
but points to an empty string after doing "set solib-search-path". This
leads to some code that needs to check for both to check for "no value".
Or some code that converts back and forth between NULL and "" when
getting or setting the value. I find this very error-prone, because it
is very easy to forget one or the other. With std::string, we at least
know that the variable is not "NULL". There is only one way of
representing an empty string setting, that is with an empty string.
I was wondering whether the distinction between NULL and "" would be
important for some setting, but it doesn't seem so. If that ever
happens, it would be more C++-y and self-descriptive to use
optional<string> anyway.
Actually, there's one spot where this distinction mattered, it's in
init_history, for the test gdb.base/gdbinit-history.exp. init_history
sets the history filename to the default ".gdb_history" if it sees that
the setting was never set - if history_filename is nullptr. If
history_filename is an empty string, it means the setting was explicitly
cleared, so it leaves it as-is. With the change to std::string, this
distinction doesn't exist anymore. This can be fixed by moving the code
that chooses a good default value for history_filename to
_initialize_top. This is ran before -ex commands are processed, so an
-ex command can then clear that value if needed (what
gdb.base/gdbinit-history.exp tests).
Another small improvement, in my opinion is that we can now easily
give string parameters initial values, by simply initializing the global
variables, instead of xstrdup-ing it in the _initialize function.
In Python and Guile, when registering a string-like parameter, we
allocate (with new) an std::string that is owned by the param_smob (in
Guile) and the parmpy_object (in Python) objects.
This patch started by changing all relevant add_setshow_* commands to
take an `std::string *` instead of a `char **` and fixing everything
that failed to build. That includes of course all string setting
variable and their uses.
string_option_def now uses an std::string also, because there's a
connection between options and settings (see
add_setshow_cmds_for_options).
The add_path function in source.c is really complex and twisted, I'd
rather not try to change it to work on an std::string right now.
Instead, I added an overload that copies the std:string to a `char *`
and back. This means more copying, but this is not used in a hot path
at all, so I think it is acceptable.
Change-Id: I92c50a1bdd8307141cdbacb388248e4e4fc08c93
Co-authored-by: Lancelot SIX <lsix@lancelotsix.com>
cmd_list_element can contain a pointer to data that can be set and / or
shown. This is achieved with the void* VAR member which points to the
data that can be accessed, while the VAR_TYPE member (of type enum
var_types) indicates how to interpret the data pointed to.
With this pattern, the user of the cmd_list_element needs to know what
is the storage type associated with a given VAR_TYPES in order to do
the proper casting. No automatic safeguard is available to prevent
miss-use of the pointer. Client code typically looks something like:
switch (c->var_type)
{
case var_zuinteger:
unsigned int v = *(unsigned int*) c->var;
...
break;
case var_boolean:
bool v = *(bool *) c->var;
...
break;
...
}
This patch proposes to add an abstraction around the var_types and void*
pointer pair. The abstraction is meant to prevent the user from having
to handle the cast and verify that the data is read or written as a type
that is coherent with the setting's var_type. This is achieved by
introducing the struct setting which exposes a set of templated get /
set member functions. The template parameter is the type of the
variable that holds the referred variable.
Using those accessors allows runtime checks to be inserted in order to
ensure that the data pointed to has the expected type. For example,
instantiating the member functions with bool will yield something
similar to:
const bool &get<bool> () const
{
gdb_assert (m_var_type == var_boolean);
gdb_assert (m_var != nullptr);
return *static_cast<bool *> (m_var);
}
void set<bool> (const bool &var)
{
gdb_assert (m_var_type == var_boolean);
gdb_assert (m_var != nullptr);
*static_cast<bool *> (m_var) = var;
}
Using the new abstraction, our initial example becomes:
switch (c->var_type)
{
case var_zuinteger:
unsigned int v = c->var->get<unsigned int> ();
...
break;
case var_boolean:
bool v = c->var->get<bool> ();
...
break;
...
}
While the call site is still similar, the introduction of runtime checks
help ensure correct usage of the data.
In order to avoid turning the bulk of add_setshow_cmd_full into a
templated function, and following a suggestion from Pedro Alves, a
setting can be constructed from a pre validated type erased reference to
a variable. This is what setting::erased_args is used for.
Introducing an opaque abstraction to describe a setting will also make
it possible to use callbacks to retrieve or set the value of the setting
on the fly instead of pointing to a static chunk of memory. This will
be done added in a later commit.
Given that a cmd_list_element may or may not reference a setting, the
VAR and VAR_TYPES members of the struct are replaced with a
gdb::optional<setting> named VAR.
Few internal function signatures have been modified to take into account
this new abstraction:
-The functions value_from_setting, str_value_from_setting and
get_setshow_command_value_string used to have a 'cmd_list_element *'
parameter but only used it for the VAR and VAR_TYPE member. They now
take a 'const setting &' parameter instead.
- Similarly, the 'void *' and a 'enum var_types' parameters of
pascm_param_value and gdbpy_parameter_value have been replaced with a
'const setting &' parameter.
No user visible change is expected after this patch.
Tested on GNU/Linux x86_64, with no regression noticed.
Co-authored-by: Simon Marchi <simon.marchi@polymtl.ca>
Change-Id: Ie1d08c3ceb8b30b3d7bf1efe036eb8acffcd2f34
A build without Python will currently fail, because
selftests::test_python uses gdb_python_initialized, which is only
conditionally defined.
This patch fixes the build by making test_python also be conditionally
defined. I chose this approach because the selftest will fail if
Python is not enabled, so it didn't seem useful to leave it defined.
The test-case gdb.gdb/python-selftest.exp:
- patches the gdb_python_initialized variable in gdb to 0
- checks that the output of a python command is "Python not initialized"
Reimplement gdb.gdb/python-selftest.exp as unittest, using:
- execute_command_to_string to capture the output
- try/catch to catch the "Python not initialized" exception.
Tested on x86_64-linux.
The guile API has (history-append! <value>) to add values into GDB's
history list. There is currently no equivalent in the Python API.
This commit adds gdb.add_history(<value>) to the Python API, this
function takes <value> a gdb.Value (or anything that can be passed to
the constructor of gdb.Value), and adds the value it represents to
GDB's history list. The index of the newly added value is returned.
The alias creation functions currently accept a name to specify the
target command. They pass this to add_alias_cmd, which needs to lookup
the target command by name.
Given that:
- We don't support creating an alias for a command before that command
exists.
- We always use add_info_alias just after creating that target command,
and therefore have access to the target command's cmd_list_element.
... change add_com_alias to accept the target command as a
cmd_list_element (other functions are done in subsequent patches). This
ensures we don't create the alias before the target command, because you
need to get the cmd_list_element from somewhere when you call the alias
creation function. And it avoids an unecessary command lookup. So it
seems better to me in every aspect.
gdb/ChangeLog:
* command.h (add_com_alias): Accept target as
cmd_list_element. Update callers.
Change-Id: I24bed7da57221cc77606034de3023fedac015150
Previously, the prefixname field of struct cmd_list_element was manually
set for prefix commands. This seems verbose and error prone as it
required every single call to functions adding prefix commands to
specify the prefix name while the same information can be easily
generated.
Historically, this was not possible as the prefix field was null for
many commands, but this was fixed in commit
3f4d92ebdf by Philippe Waroquiers, so
we can rely on the prefix field being set when generating the prefix
name.
This commit also fixes a use after free in this scenario:
* A command gets created via Python (using the gdb.Command class).
The prefix name member is dynamically allocated.
* An alias to the new command is created. The alias's prefixname is set
to point to the prefixname for the original command with a direct
assignment.
* A new command with the same name as the Python command is created.
* The object for the original Python command gets freed and its
prefixname gets freed as well.
* The alias is updated to point to the new command, but its prefixname
is not updated so it keeps pointing to the freed one.
gdb/ChangeLog:
* command.h (add_prefix_cmd): Remove the prefixname argument as
it can now be generated automatically. Update all callers.
(add_basic_prefix_cmd): Ditto.
(add_show_prefix_cmd): Ditto.
(add_prefix_cmd_suppress_notification): Ditto.
(add_abbrev_prefix_cmd): Ditto.
* cli/cli-decode.c (add_prefix_cmd): Ditto.
(add_basic_prefix_cmd): Ditto.
(add_show_prefix_cmd): Ditto.
(add_prefix_cmd_suppress_notification): Ditto.
(add_prefix_cmd_suppress_notification): Ditto.
(add_abbrev_prefix_cmd): Ditto.
* cli/cli-decode.h (struct cmd_list_element): Replace the
prefixname member variable with a method which generates the
prefix name at runtime. Update all code reading the prefix
name to use the method, and remove all code setting it.
* python/py-cmd.c (cmdpy_destroyer): Remove code to free the
prefixname member as it's now a method.
(cmdpy_function): Determine if the command is a prefix by
looking at prefixlist, not prefixname.
Add two new commands to GDB that can be placed into the early
initialization to control how Python starts up. The new options are:
set python ignore-environment on|off
set python dont-write-bytecode auto|on|off
show python ignore-environment
show python dont-write-bytecode
These can be used from GDB's startup file to control how the Python
extension language behaves. These options are equivalent to the -E
and -B flags to python respectively, their descriptions from the
Python man page:
-E Ignore environment variables like PYTHONPATH and PYTHONHOME
that modify the behavior of the interpreter.
-B Don't write .pyc files on import.
gdb/ChangeLog:
* NEWS: Mention new commands.
* python/python.c (python_ignore_environment): New static global.
(show_python_ignore_environment): New function.
(set_python_ignore_environment): New function.
(python_dont_write_bytecode): New static global.
(show_python_dont_write_bytecode): New function.
(set_python_dont_write_bytecode): New function.
(_initialize_python): Register new commands.
gdb/doc/ChangeLog:
* python.texinfo (Python Commands): Mention new commands.
gdb/testsuite/ChangeLog:
* gdb.python/py-startup-opt.exp: New file.
Now that both Python and Guile are fully initialized from their
respective finish_initialization methods, the "finish" in the method
name doesn't really make sense; initialization starts _and_ finishes
with that method.
As such, this commit renames finish_initialization to just initialize.
There should be no user visible changes after this commit.
gdb/ChangeLog:
* extension-priv.h (struct extension_language_ops): Rename
'finish_initialization' to 'initialize'.
* extension.c (finish_ext_lang_initialization): Renamed to...
(ext_lang_initialization): ...this, update comment, and updated
the calls to reflect the change in struct extension_language_ops.
* extension.h (finish_ext_lang_initialization): Renamed to...
(ext_lang_initialization): ...this.
* guile/guile.c (gdbscm_finish_initialization): Renamed to...
(gdbscm_initialize): ...this, update comment at definition.
(guile_extension_ops): Update.
* main.c (captured_main_1): Update call to
finish_ext_lang_initialization.
* python/python.c (gdbpy_finish_initialization): Rename to...
(gdbpy_initialize): ...this, update comment at definition, and
update call to do_finish_initialization.
(python_extension_ops): Update.
(do_finish_initialization): Rename to...
(do_initialize): ...this, and update comment.
Delay Python initialisation until gdbpy_finish_initialization.
This is mostly about splitting the existing gdbpy_initialize_*
functions in two, all the calls to register_objfile_data_with_cleanup,
gdbarch_data_register_post_init, etc are moved into new _initialize_*
functions, but everything else is left in the gdbpy_initialize_*
functions.
Then the call to do_start_initialization (in python/python.c) is moved
from the _initialize_python function into gdbpy_finish_initialization.
There should be no user visible changes after this commit.
gdb/ChangeLog:
* python/py-arch.c (_initialize_py_arch): New function.
(gdbpy_initialize_arch): Move code to _initialize_py_arch.
* python/py-block.c (_initialize_py_block): New function.
(gdbpy_initialize_blocks): Move code to _initialize_py_block.
* python/py-inferior.c (_initialize_py_inferior): New function.
(gdbpy_initialize_inferior): Move code to _initialize_py_inferior.
* python/py-objfile.c (_initialize_py_objfile): New function.
(gdbpy_initialize_objfile): Move code to _initialize_py_objfile.
* python/py-progspace.c (_initialize_py_progspace): New function.
(gdbpy_initialize_pspace): Move code to _initialize_py_progspace.
* python/py-registers.c (_initialize_py_registers): New function.
(gdbpy_initialize_registers): Move code to
_initialize_py_registers.
* python/py-symbol.c (_initialize_py_symbol): New function.
(gdbpy_initialize_symbols): Move code to _initialize_py_symbol.
* python/py-symtab.c (_initialize_py_symtab): New function.
(gdbpy_initialize_symtabs): Move code to _initialize_py_symtab.
* python/py-type.c (_initialize_py_type): New function.
(gdbpy_initialize_types): Move code to _initialize_py_type.
* python/py-unwind.c (_initialize_py_unwind): New function.
(gdbpy_initialize_unwind): Move code to _initialize_py_unwind.
* python/python.c (_initialize_python): Move call to
do_start_initialization to gdbpy_finish_initialization.
(gdbpy_finish_initialization): Add call to
do_start_initialization.
The current mechanism by which the Python gdb.current_objfile is
maintained does not allow for nested auto-load events. It is assumed
that once an auto-load script has finished loading then the current
objfile should be set back to NULL. In a nested situation, we should
be restoring the previous value.
We already have an RAII class to handle save/restore type behaviour,
so lets just switch to use that.
The test is a little contrived, but is simple enough, and triggers the
bug. The real use case might involve the auto-load script calling
functions (either in the just-loaded object file, or in the main
executable), which in turn trigger further auto-loads to occur.
gdb/ChangeLog:
* python/python.c (gdbpy_source_objfile_script): Use
make_scoped_restore to restore gdbpy_current_objfile.
(gdbpy_execute_objfile_script): Likewise.
gdb/testsuite/ChangeLog:
* gdb.python/py-auto-load-chaining-f1.c: New file.
* gdb.python/py-auto-load-chaining-f1.o-gdb.py: New file.
* gdb.python/py-auto-load-chaining-f2.c: New file.
* gdb.python/py-auto-load-chaining-f2.o-gdb.py: New file.
* gdb.python/py-auto-load-chaining.c: New file.
* gdb.python/py-auto-load-chaining.exp: New file.
While working on another patch I noticed an oddly formatted error
message in the Python code.
When 'set python print-stack message' is in effect then consider this
Python script:
class TestCommand (gdb.Command):
def __init__ (self):
gdb.Command.__init__ (self, "test-cmd", gdb.COMMAND_DATA)
def invoke(self, args, from_tty):
raise RuntimeError ("bad")
TestCommand ()
And this GDB session:
(gdb) source path/to/python/script.py
(gdb) test-cmd
Python Exception <class 'RuntimeError'> bad:
Error occurred in Python: bad
The line 'Python Exception <class 'RuntimeError'> bad:' doesn't look
terrible in this situation, the colon at the end of the first line
makes sense given the second line.
However, there are places in GDB where there is no second line
printed, for example consider this python script:
def stop_listener (e):
raise RuntimeError ("bad")
gdb.events.stop.connect (stop_listener)
Then this GDB session:
(gdb) file helloworld.exe
(gdb) start
Temporary breakpoint 1 at 0x40112a: file hello.c, line 6.
Starting program: helloworld.exe
Temporary breakpoint 1, main () at hello.c:6
6 printf ("Hello World\n");
Python Exception <class 'RuntimeError'> bad:
(gdb) si
0x000000000040112f 6 printf ("Hello World\n");
Python Exception <class 'RuntimeError'> bad:
In this case there is no auxiliary information displayed after the
warning, and the line ending in the colon looks weird to me.
A quick survey of the code seems to indicate that it is not uncommon
for there to be no auxiliary information line printed, its not just
the one case I found above.
I propose that the line that currently looks like this:
Python Exception <class 'RuntimeError'> bad:
Be reformatted like this:
Python Exception <class 'RuntimeError'>: bad
I think this looks fine then in either situation. The first now looks
like this:
(gdb) test-cmd
Python Exception <class 'RuntimeError'>: bad
Error occurred in Python: bad
And the second like this:
(gdb) si
0x000000000040112f 6 printf ("Hello World\n");
Python Exception <class 'RuntimeError'>: bad
There's just two tests that needed updating. Errors are checked for
in many more tests, but most of the time the pattern doesn't care
about the colon.
gdb/ChangeLog:
* python/python.c (gdbpy_print_stack): Reformat an error message.
gdb/testsuite/ChangeLog:
* gdb.python/py-framefilter.exp: Update expected results.
* gdb.python/python.exp: Update expected results.
This commits the result of running gdb/copyright.py as per our Start
of New Year procedure...
gdb/ChangeLog
Update copyright year range in copyright header of all GDB files.
I noticed that a few "#if"s could be removed from the Python code.
This patch is the result.
gdb/ChangeLog
2020-11-02 Tom Tromey <tromey@adacore.com>
* python/python.c: Consolidate two HAVE_PYTHON blocks.
(python_GdbModuleDef): Move earlier. Now static.
(do_start_initialization): Consolidate some IS_PY3K blocks.
This changes gdb to avoid PyLong_FromUnsignedLong, preferring
gdb_py_object_from_ulongest instead.
gdb/ChangeLog
2020-09-15 Tom Tromey <tromey@adacore.com>
* python/python.c (gdbpy_parameter_value): Use
gdb_py_object_from_ulongest.
This changes gdb to avoid PyLong_FromLong, preferring to
gdb_py_object_from_longest instead.
gdb/ChangeLog
2020-09-15 Tom Tromey <tromey@adacore.com>
* python/python.c (gdbpy_parameter_value): Use
gdb_py_object_from_longest.
* python/py-type.c (convert_field, typy_range): Use
gdb_py_object_from_longest.
* python/py-tui.c (gdbpy_tui_width, gdbpy_tui_height): Use
gdb_py_object_from_longest.
* python/py-lazy-string.c (stpy_get_length): Use
gdb_py_object_from_longest.
* python/py-infthread.c (thpy_get_num, thpy_get_global_num): Use
gdb_py_object_from_longest.
* python/py-infevents.c (create_memory_changed_event_object): Use
gdb_py_object_from_longest.
* python/py-inferior.c (infpy_get_num): Use
gdb_py_object_from_longest.
(infpy_get_pid): Likewise.
This commit adds a new method gdb.Architecture.registers that returns
an object of the new type gdb.RegisterDescriptorIterator. This
iterator returns objects of the new type gdb.RegisterDescriptor.
A RegisterDescriptor is not a way to read the value of a register,
this is already covered by Frame.read_register, a RegisterDescriptor
is simply a way to discover from Python, which registers are
available for a given architecture.
I did consider just returning a string, the name of each register,
instead of a RegisterDescriptor, however, I'm aware that it we don't
want to break the existing Python API in any way, so if I return just
a string now, but in the future we want more information about a
register then we would have to add a second API to get that
information. By going straight to a descriptor object now, it is easy
to add additional properties in the future should we wish to.
Right now the only property of a register that a user can access is
the name of the register.
In future we might want to be able to ask the register about is
register groups, or its type.
gdb/ChangeLog:
* Makefile.in (SUBDIR_PYTHON_SRCS): Add py-registers.c
* python/py-arch.c (archpy_registers): New function.
(arch_object_methods): Add 'registers' method.
* python/py-registers.c: New file.
* python/python-internal.h
(gdbpy_new_register_descriptor_iterator): Declare.
(gdbpy_initialize_registers): Declare.
* python/python.c (do_start_initialization): Call
gdbpy_initialize_registers.
* NEWS: Mention additions to the Python API.
gdb/testsuite/ChangeLog:
* gdb.python/py-arch-reg-names.exp: New file.
gdb/doc/ChangeLog:
* python.texi (Python API): Add new section the menu.
(Frames In Python): Add new @anchor.
(Architectures In Python): Document new registers method.
(Registers In Python): New section.
Python3.9b1 is now available on Rawhide. GDB w/ Python 3.9 support
can be built using the configure switch -with-python=/usr/bin/python3.9.
Attempting to run gdb/Python3.9 segfaults on startup:
#0 0x00007ffff7b0582c in PyEval_ReleaseLock () from /lib64/libpython3.9.so.1.0
#1 0x000000000069ccbf in do_start_initialization ()
at worktree-test1/gdb/python/python.c:1789
#2 _initialize_python ()
at worktree-test1/gdb/python/python.c:1877
#3 0x00000000007afb0a in initialize_all_files () at init.c:237
...
Consulting the the documentation...
https://docs.python.org/3/c-api/init.html
...we find that PyEval_ReleaseLock() has been deprecated since version
3.2. It recommends using PyEval_SaveThread or PyEval_ReleaseThread()
instead. In do_start_initialization, in gdb/python/python.c, we
can replace the calls to PyThreadState_Swap() and PyEval_ReleaseLock()
with a single call to PyEval_SaveThread. (Thanks to Keith Seitz
for working this out.)
With that in place, GDB gets a little bit further. It still dies
on startup, but the backtrace is different:
#0 0x00007ffff7b04306 in PyOS_InterruptOccurred ()
from /lib64/libpython3.9.so.1.0
#1 0x0000000000576e86 in check_quit_flag ()
at worktree-test1/gdb/extension.c:776
#2 0x0000000000576f8a in set_active_ext_lang (now_active=now_active@entry=0x983c00 <extension_language_python>)
at worktree-test1/gdb/extension.c:705
#3 0x000000000069d399 in gdbpy_enter::gdbpy_enter (this=0x7fffffffd2d0,
gdbarch=0x0, language=0x0)
at worktree-test1/gdb/python/python.c:211
#4 0x0000000000686e00 in python_new_inferior (inf=0xddeb10)
at worktree-test1/gdb/python/py-inferior.c:251
#5 0x00000000005d9fb9 in std::function<void (inferior*)>::operator()(inferior*) const (__args#0=<optimized out>, this=0xccad20)
at /usr/include/c++/10/bits/std_function.h:617
#6 gdb::observers::observable<inferior*>::notify (args#0=0xddeb10,
this=<optimized out>)
at worktree-test1/gdb/../gdbsupport/observable.h:106
#7 add_inferior_silent (pid=0)
at worktree-test1/gdb/inferior.c:113
#8 0x00000000005dbcb8 in initialize_inferiors ()
at worktree-test1/gdb/inferior.c:947
...
We checked with some Python Developers and were told that we should
acquire the GIL prior to calling any Python C API function. We
definitely don't have the GIL for calls of PyOS_InterruptOccurred().
I moved class_gdbpy_gil earlier in the file and use it in
gdbpy_check_quit_flag() to acquire (and automatically release) the
GIL.
With those changes in place, I was able to run to a GDB prompt. But,
when trying to quit, it segfaulted again due to due to some other
problems with gdbpy_check_quit_flag():
Thread 1 "gdb" received signal SIGSEGV, Segmentation fault.
0x00007ffff7bbab0c in new_threadstate () from /lib64/libpython3.9.so.1.0
(top-gdb) bt 8
#0 0x00007ffff7bbab0c in new_threadstate () from /lib64/libpython3.9.so.1.0
#1 0x00007ffff7afa5ea in PyGILState_Ensure.cold ()
from /lib64/libpython3.9.so.1.0
#2 0x000000000069b58c in gdbpy_gil::gdbpy_gil (this=<synthetic pointer>)
at worktree-test1/gdb/python/python.c:278
#3 gdbpy_check_quit_flag (extlang=<optimized out>)
at worktree-test1/gdb/python/python.c:278
#4 0x0000000000576e96 in check_quit_flag ()
at worktree-test1/gdb/extension.c:776
#5 0x000000000057700c in restore_active_ext_lang (previous=0xe9c050)
at worktree-test1/gdb/extension.c:729
#6 0x000000000088913a in do_my_cleanups (
pmy_chain=0xc31870 <final_cleanup_chain>,
old_chain=0xae5720 <sentinel_cleanup>)
at worktree-test1/gdbsupport/cleanups.cc:131
#7 do_final_cleanups ()
at worktree-test1/gdbsupport/cleanups.cc:143
In this case, we're trying to call a Python C API function after
Py_Finalize() has been called from finalize_python(). I made
finalize_python set gdb_python_initialized to false and then cause
check_quit_flag() to return early when it's false.
With these changes in place, GDB seems to be working again with
Python3.9b1. I think it likely that there are other problems lurking.
I wouldn't be surprised to find that there are other calls into Python
where we don't first make sure that we have the GIL. Further changes
may well be needed.
I see no regressions testing on Rawhide using a GDB built with the
default Python version (3.8.3) versus one built using Python 3.9b1.
I've also tested on Fedora 28, 29, 30, 31, and 32 (all x86_64) using
the default (though updated) system installed versions of Python on
those OSes. This means that I've tested against Python versions
2.7.15, 2.7.17, 2.7.18, 3.7.7, 3.8.2, and 3.8.3. In each case GDB
still builds without problem and shows no regressions after applying
this patch.
gdb/ChangeLog:
2020-MM-DD Kevin Buettner <kevinb@redhat.com>
Keith Seitz <keiths@redhat.com>
* python/python.c (do_start_initialization): For Python 3.9 and
later, call PyEval_SaveThread instead of PyEval_ReleaseLock.
(class gdbpy_gil): Move to earlier in file.
(finalize_python): Set gdb_python_initialized.
(gdbpy_check_quit_flag): Acquire GIL via gdbpy_gil. Return early
when not initialized.
I recently learned that move constructors generally should be marked
"noexcept". This ensures that standard containers will move objects
when possible, rather than copy them.
This patch fixes the cases I could find. Note that implicitly-defined
or defaulted move constructors will automatically do what you'd
expect; that is, they are noexcept if all the members have noexcept
move constructors.
While doing this, I noticed a couple of odd cases where the move
constructor seemed to assume that the object being constructed could
have state requiring destruction. I've fixed these as well. See
completion_result and scoped_mmap.
gdb/ChangeLog
2020-04-20 Tom Tromey <tromey@adacore.com>
* python/python.c (struct gdbpy_event): Mark move constructor as
noexcept.
* python/py-tui.c (class gdbpy_tui_window_maker): Mark move
constructor as noexcept.
* completer.h (struct completion_result): Mark move constructor as
noexcept.
* completer.c (completion_result::completion_result): Use
initialization style. Don't call reset_match_list.
gdbsupport/ChangeLog
2020-04-20 Tom Tromey <tromey@adacore.com>
* scoped_mmap.h (scoped_mmap): Mark move constructor as noexcept.
Use initialization style. Don't call destroy.
* scoped_fd.h (class scoped_fd): Mark move constructor as
noexcept.
* gdb_ref_ptr.h (class ref_ptr): Mark move constructor as
noexcept.
Currently there are many prefix commands that do nothing but call
either help_list or cmd_show_list. I happened to notice that one such
call, for "set print type", used the wrong command list parameter,
causing incorrect output.
Rather than fix this bug in isolation, I decided to eliminate this
possibility by adding two new ways to add prefix commands, which
simply route the call to help_list or cmd_show_list, as appropriate.
This makes it impossible for a mismatch to occur.
In some cases, a bit of output was removed; however, I don't think
this output in general was very useful. It seemed redundant with
what's already printed by help_list. A representative example is this
hunk, removed from ada-lang.c:
- printf_unfiltered (_(\
-"\"set ada\" must be followed by the name of a setting.\n"));
This simplified the CLI style set/show commands quite a bit, and
allowed the deletion of a macro.
This also cleans up some unusual code in windows-tdep.c.
Tested on x86-64 Fedora 30. Note that I have no way to build the
go32-nat.c change.
gdb/ChangeLog
2020-04-17 Tom Tromey <tromey@adacore.com>
* auto-load.c (show_auto_load_cmd): Remove.
(auto_load_show_cmdlist_get): Use add_show_prefix_cmd.
* arc-tdep.c (_initialize_arc_tdep): Use add_show_prefix_cmd.
(maintenance_print_arc_command): Remove.
* tui/tui-win.c (tui_command): Remove.
(tui_get_cmd_list): Use add_basic_prefix_cmd.
* tui/tui-layout.c (tui_layout_command): Remove.
(_initialize_tui_layout): Use add_basic_prefix_cmd.
* python/python.c (user_set_python, user_show_python): Remove.
(_initialize_python): Use add_basic_prefix_cmd,
add_show_prefix_cmd.
* guile/guile.c (set_guile_command, show_guile_command): Remove.
(install_gdb_commands): Use add_basic_prefix_cmd,
add_show_prefix_cmd.
(info_guile_command): Remove.
* dwarf2/read.c (set_dwarf_cmd, show_dwarf_cmd): Remove.
(_initialize_dwarf2_read): Use add_basic_prefix_cmd,
add_show_prefix_cmd.
* cli/cli-style.h (class cli_style_option) <add_setshow_commands>:
Remove do_set and do_show parameters.
* cli/cli-style.c (set_style, show_style): Remove.
(_initialize_cli_style): Use add_basic_prefix_cmd,
add_show_prefix_cmd.
(cli_style_option::add_setshow_commands): Remove do_set and
do_show parameters.
(cli_style_option::add_setshow_commands): Use
add_basic_prefix_cmd, add_show_prefix_cmd.
(STYLE_ADD_SETSHOW_COMMANDS): Remove macro.
(set_style_name): Remove.
* cli/cli-dump.c (dump_command, append_command): Remove.
(srec_dump_command, ihex_dump_command, verilog_dump_command)
(tekhex_dump_command, binary_dump_command)
(binary_append_command): Remove.
(_initialize_cli_dump): Use add_basic_prefix_cmd.
* windows-tdep.c (w32_prefix_command_valid): Remove global.
(init_w32_command_list): Remove; move into ...
(_initialize_windows_tdep): ... here. Use add_basic_prefix_cmd.
* valprint.c (set_print, show_print, set_print_raw)
(show_print_raw): Remove.
(_initialize_valprint): Use add_basic_prefix_cmd,
add_show_prefix_cmd.
* typeprint.c (set_print_type, show_print_type): Remove.
(_initialize_typeprint): Use add_basic_prefix_cmd,
add_show_prefix_cmd.
* record.c (set_record_command, show_record_command): Remove.
(_initialize_record): Use add_basic_prefix_cmd,
add_show_prefix_cmd.
* cli/cli-cmds.c (_initialize_cli_cmds): Use add_basic_prefix_cmd,
add_show_prefix_cmd.
(info_command, show_command, set_debug, show_debug): Remove.
* top.h (set_history, show_history): Don't declare.
* top.c (set_history, show_history): Remove.
* target-descriptions.c (set_tdesc_cmd, show_tdesc_cmd)
(unset_tdesc_cmd): Remove.
(_initialize_target_descriptions): Use add_basic_prefix_cmd,
add_show_prefix_cmd.
* symtab.c (info_module_command): Remove.
(_initialize_symtab): Use add_basic_prefix_cmd.
* symfile.c (overlay_command): Remove.
(_initialize_symfile): Use add_basic_prefix_cmd.
* sparc64-tdep.c (info_adi_command): Remove.
(_initialize_sparc64_adi_tdep): Use add_basic_prefix_cmd.
* sh-tdep.c (show_sh_command, set_sh_command): Remove.
(_initialize_sh_tdep): Use add_basic_prefix_cmd,
add_show_prefix_cmd.
* serial.c (serial_set_cmd, serial_show_cmd): Remove.
(_initialize_serial): Use add_basic_prefix_cmd,
add_show_prefix_cmd.
* ser-tcp.c (set_tcp_cmd, show_tcp_cmd): Remove.
(_initialize_ser_tcp): Use add_basic_prefix_cmd,
add_show_prefix_cmd.
* rs6000-tdep.c (set_powerpc_command, show_powerpc_command)
(_initialize_rs6000_tdep): Use add_basic_prefix_cmd,
add_show_prefix_cmd.
* riscv-tdep.c (show_riscv_command, set_riscv_command)
(show_debug_riscv_command, set_debug_riscv_command): Remove.
(_initialize_riscv_tdep): Use add_basic_prefix_cmd,
add_show_prefix_cmd.
* remote.c (remote_command, set_remote_cmd): Remove.
(_initialize_remote): Use add_basic_prefix_cmd.
* record-full.c (set_record_full_command)
(show_record_full_command): Remove.
(_initialize_record_full): Use add_basic_prefix_cmd,
add_show_prefix_cmd.
* record-btrace.c (cmd_set_record_btrace)
(cmd_show_record_btrace, cmd_set_record_btrace_bts)
(cmd_show_record_btrace_bts, cmd_set_record_btrace_pt)
(cmd_show_record_btrace_pt): Remove.
(_initialize_record_btrace): Use add_basic_prefix_cmd,
add_show_prefix_cmd.
* ravenscar-thread.c (set_ravenscar_command)
(show_ravenscar_command): Remove.
(_initialize_ravenscar): Use add_basic_prefix_cmd,
add_show_prefix_cmd.
* mips-tdep.c (show_mips_command, set_mips_command)
(_initialize_mips_tdep): Use add_basic_prefix_cmd,
add_show_prefix_cmd.
* maint.c (maintenance_command, maintenance_info_command)
(maintenance_check_command, maintenance_print_command)
(maintenance_set_cmd, maintenance_show_cmd): Remove.
(_initialize_maint_cmds): Use add_basic_prefix_cmd,
add_show_prefix_cmd.
(show_per_command_cmd): Remove.
* maint-test-settings.c (maintenance_set_test_settings_cmd):
Remove.
(maintenance_show_test_settings_cmd): Remove.
(_initialize_maint_test_settings): Use add_basic_prefix_cmd,
add_show_prefix_cmd.
* maint-test-options.c (maintenance_test_options_command):
Remove.
(_initialize_maint_test_options): Use add_basic_prefix_cmd.
* macrocmd.c (macro_command): Remove
(_initialize_macrocmd): Use add_basic_prefix_cmd.
* language.c (set_check, show_check): Remove.
(_initialize_language): Use add_basic_prefix_cmd,
add_show_prefix_cmd.
* infcmd.c (unset_command): Remove.
(_initialize_infcmd): Use add_basic_prefix_cmd.
* i386-tdep.c (set_mpx_cmd, show_mpx_cmd): Remove.
(_initialize_i386_tdep): Use add_basic_prefix_cmd,
add_show_prefix_cmd.
* go32-nat.c (go32_info_dos_command): Remove.
(_initialize_go32_nat): Use add_basic_prefix_cmd.
* cli/cli-decode.c (do_prefix_cmd, add_basic_prefix_cmd)
(do_show_prefix_cmd, add_show_prefix_cmd): New functions.
* frame.c (set_backtrace_cmd, show_backtrace_cmd): Remove.
(_initialize_frame): Use add_basic_prefix_cmd,
add_show_prefix_cmd.
* dcache.c (set_dcache_command, show_dcache_command): Remove.
(_initialize_dcache): Use add_basic_prefix_cmd,
add_show_prefix_cmd.
* cp-support.c (maint_cplus_command): Remove.
(_initialize_cp_support): Use add_basic_prefix_cmd.
* btrace.c (maint_btrace_cmd, maint_btrace_set_cmd)
(maint_btrace_show_cmd, maint_btrace_pt_set_cmd)
(maint_btrace_pt_show_cmd, _initialize_btrace): Use
add_basic_prefix_cmd, add_show_prefix_cmd.
* breakpoint.c (save_command): Remove.
(_initialize_breakpoint): Use add_basic_prefix_cmd.
* arm-tdep.c (set_arm_command, show_arm_command): Remove.
(_initialize_arm_tdep): Use add_basic_prefix_cmd,
add_show_prefix_cmd.
* ada-lang.c (maint_set_ada_cmd, maint_show_ada_cmd)
(set_ada_command, show_ada_command): Remove.
(_initialize_ada_language): Use add_basic_prefix_cmd,
add_show_prefix_cmd.
* command.h (add_basic_prefix_cmd, add_show_prefix_cmd): Declare.
gdb/testsuite/ChangeLog
2020-04-17 Tom Tromey <tromey@adacore.com>
* gdb.cp/maint.exp (test_help): Simplify multiple_help_body.
Update tests.
* gdb.btrace/cpu.exp: Update tests.
* gdb.base/maint.exp: Update tests.
* gdb.base/default.exp: Update tests.
* gdb.base/completion.exp: Update tests.
This commit fixes a compilation warning/error when building GDB
with Python 3.9:
g++ -x c++ -O2 -g -pipe -Wall -Werror=format-security -Wp,-D_FORTIFY_SOURCE=2 -Wp,-D_GLIBCXX_ASSERTIONS -fexceptions -fstack-protector-strong -grecord-gcc-switches -specs=/usr/lib/rpm/redhat/redhat-hardened-cc1 -specs=/usr/lib/rpm/redhat/redhat-annobin-cc1 -m64 -mtune=generic -fasynchronous-unwind-tables -fstack-clash-protection -fcf-protection -DDNF_DEBUGINFO_INSTALL -I. -I../../gdb -I../../gdb/config -DLOCALEDIR="\"/usr/share/locale\"" -DHAVE_CONFIG_H -I../../gdb/../include/opcode -I../bfd -I../../gdb/../bfd -I../../gdb/../include -I../libdecnumber -I../../gdb/../libdecnumber -I../../gdb/../gnulib/import -I../gnulib/import -DTUI=1 -I/usr/include/guile/2.0 -pthread -I/usr/include/python3.9 -I/usr/include/python3.9 -I../../gdb/.. -pthread -Wall -Wpointer-arith -Wno-unused -Wunused-value -Wunused-variable -Wunused-function -Wno-switch -Wno-char-subscripts -Wempty-body -Wunused-but-set-parameter -Wunused-but-set-variable -Wno-sign-compare -Wno-error=maybe-uninitialized -Wno-mismatched-tags -Wsuggest-override -Wimplicit-fallthrough=3 -Wduplicated-cond -Wshadow=local -Wdeprecated-copy -Wdeprecated-copy-dtor -Wredundant-move -Wformat -Wformat-nonliteral -Wno-unused -Werror -c -o ser-tcp.o -MT ser-tcp.o -MMD -MP -MF ./.deps/ser-tcp.Tpo ../../gdb/ser-tcp.c
../../gdb/python/python.c: In function 'bool do_start_initialization()':
../../gdb/python/python.c:1621:23: error: 'void PyEval_InitThreads()' is deprecated [-Werror=deprecated-declarations]
1621 | PyEval_InitThreads ();
| ^
In file included from /usr/include/python3.9/Python.h:141,
from ../../gdb/python/python-internal.h:86,
from ../../gdb/python/python.c:92:
/usr/include/python3.9/ceval.h:132:37: note: declared here
132 | Py_DEPRECATED(3.9) PyAPI_FUNC(void) PyEval_InitThreads(void);
| ^~~~~~~~~~~~~~~~~~
Information about the deprecated function can be found here:
https://docs.python.org/3.9/whatsnew/3.9.html#deprecated
Specifically, with regard to PyEval_InitThreads(), it says:
The PyEval_InitThreads() and PyEval_ThreadsInitialized() functions
are now deprecated and will be removed in Python 3.11. Calling
PyEval_InitThreads() now does nothing. The GIL is initialized by
Py_Initialize() since Python 3.7. (Contributed by Victor Stinner
in bpo-39877.)
I chose to disable the call with a #if test using PY_VERSION_HEX.
There is precedent for use of PY_VERSION_HEX; it's used in two places
in python-internal.h. I noticed that under certain circumstances
python-internal.h defines PyEval_InitThreads to be nothing, which
accomplishes the same thing. I considered doing something similar for
this case, but decided against it because, at some point in the future,
the presence of PyEval_InitThreads() without some explanation will be
confusing to a reader who won't be able to find PyEval_InitThreads in
the current (future for us) Python API. IMO, use of the #if along
with an accompanying comment seemed more straightforward.
gdb/ChangeLog:
* python/python.c (do_start_initialization): Don't call
PyEval_InitThreads for Python 3.9 and beyond.
Change-Id: I0679fc10b6b76761a99538568f13188c6d8014e0
This patch adds support for writing new TUI windows in Python.
2020-02-22 Tom Tromey <tom@tromey.com>
* NEWS: Add entry for gdb.register_window_type.
* tui/tui-layout.h (window_factory): New typedef.
(tui_register_window): Declare.
* tui/tui-layout.c (saved_tui_windows): New global.
(tui_apply_current_layout): Use it.
(tui_register_window): New function.
* python/python.c (do_start_initialization): Call
gdbpy_initialize_tui.
(python_GdbMethods): Add "register_window_type" function.
* python/python-internal.h (gdbpy_register_tui_window)
(gdbpy_initialize_tui): Declare.
* python/py-tui.c: New file.
* Makefile.in (SUBDIR_PYTHON_SRCS): Add py-tui.c.
gdb/doc/ChangeLog
2020-02-22 Tom Tromey <tom@tromey.com>
* python.texi (Python API): Add menu item.
(TUI Windows In Python): New node.
gdb/testsuite/ChangeLog
2020-02-22 Tom Tromey <tom@tromey.com>
* gdb.python/tui-window.exp: New file.
* gdb.python/tui-window.py: New file.
Change-Id: I85fbfb923a1840450a00a7dce113a05d7f048baa
I noticed that do_start_initialization, in python.c, checks the result
of xmalloc. However, xmalloc cannot fail, so this check is useless.
This patch also changes the code to use XNEWVEC.
gdb/ChangeLog
2020-02-19 Tom Tromey <tromey@adacore.com>
* python/python.c (do_start_initialization): Use XNEWVEC. Remove
NULL check.
This is an update of this patch:
https://sourceware.org/ml/gdb-patches/2018-09/msg00884.html
This patch attempts to address PR gdb/23718 by re-enabling stdin
whenever an exception is caught during gdb.execute().
When Python gdb.execute() is called, an exception could occur (e.g. the
target disappearing), which is then converted into a Python exception. If
stdin was disabled before the exception is caught, it is not re-enabled,
because the exception doesn't propagate to the top level of the event loop,
whose catch block would otherwise enable it.
The result is that when execution of a Python script completes, GDB does
not prompt or accept input, and is effectively hung.
This change rectifies the issue by re-enabling stdin in the catch block of
execute_gdb_command, prior to converting the exception to a Python
exception.
Since this patch was originally posted I've added a test, and also I
converted the code to re-enable stdin from this:
SWITCH_THRU_ALL_UIS ()
{
async_enable_stdin ();
}
to simply this:
async_enable_stdin ();
My reasoning is that we only need the SWITCH_THRU_ALL_UIS if, at the time
the exception is caught, the current_ui might be different than at the time
we called async_disable_stdin. Within python's execute_gdb_command I think
it should be impossible to switch current_ui, so the SWITCH_THRU_ALL_UIS
isn't needed.
gdb/ChangeLog:
PR gdb/23718
* gdb/python/python.c (execute_gdb_command): Call
async_enable_stdin in catch block.
gdb/testsuite/ChangeLog:
PR gdb/23718
* gdb.server/server-kill-python.exp: New file.
Change-Id: I1cfc36ee9f8484cc1ed82be9be338353db6bc080
While GNU Source Highlight is good, it's also difficult to build and
distribute. For one thing, it needs Boost. For another, it has an
unusual configuration and installation setup.
Pygments, a Python library, doesn't suffer from these issues, and so I
thought it would be a reasonable fallback.
This patch implements this idea. GNU Source Highlight is preferred,
but if it is unavailable (or fails), the extension languages are
tried. This patch also implements support for Pygments.
Something similar could be done for Guile, using:
https://dthompson.us/projects/guile-syntax-highlight.html
However, I don't know enough about Guile internals to make this
happen, so I have not done it here.
gdb/ChangeLog
2020-01-21 Tom Tromey <tromey@adacore.com>
* source-cache.c (source_cache::ensure): Call ext_lang_colorize.
* python/python.c (python_extension_ops): Update.
(gdbpy_colorize): New function.
* python/lib/gdb/__init__.py (colorize): New function.
* extension.h (ext_lang_colorize): Declare.
* extension.c (ext_lang_colorize): New function.
* extension-priv.h (struct extension_language_ops) <colorize>: New
member.
* cli/cli-style.c (_initialize_cli_style): Update help text.
Change-Id: I5e21623ee05f1f66baaa6deaeca78b578c031bf4
PR symtab/12535 points out that gdb.decode_line("") will cause a
valgrind report.
I think the empty linespec does not really make sense. So, this patch
changes gdb.decode_line to treat a whitespace-only linespec the same
as a non-existing argument.
gdb/ChangeLog
2020-01-14 Tom Tromey <tom@tromey.com>
PR symtab/12535:
* python/python.c (gdbpy_decode_line): Treat empty string the same
as no argument.
gdb/testsuite/ChangeLog
2020-01-14 Tom Tromey <tom@tromey.com>
PR symtab/12535:
* gdb.python/python.exp: Test decode_line with empty string
argument.
Change-Id: I1d95812b4b7a21d69a3e9afd05b9e3141a931897
When I try to enable -Wmissing-declarations, I get this error:
CXX python/python.o
/home/smarchi/src/binutils-gdb/gdb/python/python.c: In function ‘PyObject* init__gdb_module()’:
/home/smarchi/src/binutils-gdb/gdb/python/python.c:1582:1: error: no previous declaration for ‘PyObject* init__gdb_module()’ [-Werror=missing-declarations]
init__gdb_module (void)
^~~~~~~~~~~~~~~~
Prevent it by providing a declaration just before the definition.
gdb/ChangeLog:
* python/python.c (init__gdb_module): Add declaration.
Change-Id: I394bc691b7db624708cc4cb2cda28a56ab85a82b
Introduce a new class to wrap up the parameters needed for the
function search_symbols, which has now become a member function of
this new class.
The motivation is that search_symbols already takes a lot of
parameters, and a future commit is going to add even more. This
commit hopefully makes collecting the state required for a search
easier.
As part of this conversion the list of filenames in which to search
has been converted to a std::vector.
There should be no user visible changes after this commit.
gdb/ChangeLog:
* python/python.c (gdbpy_rbreak): Convert to using
global_symbol_searcher.
* symtab.c (file_matches): Convert return type to bool, change
file list to std::vector, update header comment.
(search_symbols): Rename to...
(global_symbol_searcher::search): ...this and update now its
a member function of global_symbol_searcher. Take account of the
changes to file_matches.
(symtab_symbol_info): Convert to using global_symbol_searcher.
(rbreak_command): Likewise.
(search_module_symbols): Likewise.
* symtab.h (enum symbol_search): Update comment.
(search_symbols): Remove declaration.
(class global_symbol_searcher): New class.
Change-Id: I488ab292a892d9e9e84775c632c5f198b6ad3710
This changes gdb.post_event to use the new run_on_main_thread
function. This is somewhat tricky because the Python GIL must be held
while manipulating reference counts.
2019-11-26 Tom Tromey <tom@tromey.com>
* python/python.c (class gdbpy_gil): New.
(struct gdbpy_event): Add constructor, destructor, operator().
(gdbpy_post_event): Use run_on_main_thread.
(gdbpy_initialize_events): Remove.
(do_start_initialization): Update.
Change-Id: Ie4431e60f328dae48bd96b6c6a8e875e70bda1de
When running GDB tests under Valgrind, various tests are failing due
to invalid memory access.
Here is the stack trace reported by Valgrind, for gdb.base/freebpcmd.exp :
==18658== Invalid read of size 8
==18658== at 0x7F9107: is_main (signalmodule.c:195)
==18658== by 0x7F9107: PyOS_InterruptOccurred (signalmodule.c:1730)
==18658== by 0x3696E2: check_quit_flag() (extension.c:829)
==18658== by 0x36980B: restore_active_ext_lang(active_ext_lang_state*) (extension.c:782)
==18658== by 0x48F617: gdbpy_enter::~gdbpy_enter() (python.c:235)
==18658== by 0x47BB71: add_thread_object(thread_info*) (object.h:470)
==18658== by 0x53A84D: operator() (std_function.h:687)
==18658== by 0x53A84D: notify (observable.h:106)
==18658== by 0x53A84D: add_thread_silent(ptid_t) (thread.c:311)
==18658== by 0x3CD954: inf_ptrace_target::create_inferior(char const*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&
, char**, int) (inf-ptrace.c:139)
==18658== by 0x3FE644: linux_nat_target::create_inferior(char const*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&,
char**, int) (linux-nat.c:1094)
==18658== by 0x3D5727: run_command_1(char const*, int, run_how) (infcmd.c:633)
==18658== by 0x2C05D1: cmd_func(cmd_list_element*, char const*, int) (cli-decode.c:1948)
==18658== by 0x53F29F: execute_command(char const*, int) (top.c:639)
==18658== by 0x3638EB: command_handler(char const*) (event-top.c:586)
==18658== by 0x36468C: command_line_handler(std::unique_ptr<char, gdb::xfree_deleter<char> >&&) (event-top.c:771)
==18658== by 0x36407C: gdb_rl_callback_handler(char*) (event-top.c:217)
==18658== by 0x5B2A1F: rl_callback_read_char (callback.c:281)
==18658== by 0x36346D: gdb_rl_callback_read_char_wrapper_noexcept() (event-top.c:175)
==18658== by 0x363F70: gdb_rl_callback_read_char_wrapper(void*) (event-top.c:192)
==18658== by 0x3633AF: stdin_event_handler(int, void*) (event-top.c:514)
==18658== by 0x362504: gdb_wait_for_event (event-loop.c:857)
==18658== by 0x362504: gdb_wait_for_event(int) (event-loop.c:744)
==18658== by 0x362676: gdb_do_one_event() [clone .part.11] (event-loop.c:321)
==18658== by 0x3627AD: gdb_do_one_event (event-loop.c:303)
==18658== by 0x3627AD: start_event_loop() (event-loop.c:370)
==18658== by 0x41D35A: captured_command_loop() (main.c:381)
==18658== by 0x41F2A4: captured_main (main.c:1224)
==18658== by 0x41F2A4: gdb_main(captured_main_args*) (main.c:1239)
==18658== by 0x227D0A: main (gdb.c:32)
==18658== Address 0x10 is not stack'd, malloc'd or (recently) free'd
The problem seems to be created by gdbpy_enter::~gdbpy_enter () releasing the GIL lock
too early:
~gdbpy_enter () does:
...
PyGILState_Release (m_state);
python_gdbarch = m_gdbarch;
python_language = m_language;
restore_active_ext_lang (m_previous_active);
}
So, it releases the GIL lock, does 2 assignments and then leads to the following
call sequence:
restore_active_ext_lang => check_quit_flag => python.c gdbpy_check_quit_flag
=> PyOS_InterruptOccurred => is_main.
is_main code is:
static int
is_main(_PyRuntimeState *runtime)
{
unsigned long thread = PyThread_get_thread_ident();
PyInterpreterState *interp = _PyRuntimeState_GetThreadState(runtime)->interp;
return (thread == runtime->main_thread
&& interp == runtime->interpreters.main);
}
The macros and functions to access the thread state are documented as:
/* Variable and macro for in-line access to current thread
and interpreter state */
#define _PyRuntimeState_GetThreadState(runtime) \
((PyThreadState*)_Py_atomic_load_relaxed(&(runtime)->gilstate.tstate_current))
/* Get the current Python thread state.
Efficient macro reading directly the 'gilstate.tstate_current' atomic
variable. The macro is unsafe: it does not check for error and it can
return NULL.
The caller must hold the GIL.
See also PyThreadState_Get() and PyThreadState_GET(). */
#define _PyThreadState_GET() _PyRuntimeState_GetThreadState(&_PyRuntime)
So, we see that GDB releases the GIL and then potentially calls
_PyRuntimeState_GetThreadState that needs the GIL.
It is not very clear why the problem is only observed when running under
Valgrind. Probably caused by the slowdown due to Valgrind and/or to the 'single
thread' scheduling by Valgrind.
This patch fixes the crashes by releasing the GIT lock later.
2019-11-26 Philippe Waroquiers <philippe.waroquiers@skynet.be>
* python/python.c (gdbpy_enter::~gdbpy_enter): Release GIL after
restore_active_ext_lang, as GIL is needed for (indirectly)
called PyOS_InterruptOccurred.
Andrew Burgess