Hi Dave,
Thank you for the feedback! I've incorporated the changes and sent a
revised version of the patch.
On Tue, Aug 1, 2023 at 1:02 PM David Malcolm <dmalc...@redhat.com> wrote:
>
> On Tue, 2023-08-01 at 09:52 -0400, Eric Feng wrote:
> > Hi all,
> >
> > This patch adds a hook to the end of ana::on_finish_translation_unit
> > which calls relevant stashing-related callbacks registered during
> > plugin
> > initialization. This feature is used to stash named types and global
> > variables for a CPython analyzer plugin [PR107646].
> >
> > Bootstrapped and tested on aarch64-unknown-linux-gnu. Does it look
> > okay?
>
> Hi Eric, thanks for the patch.
>
> The patch touches the C frontend, so those parts would need approval
> from the C FE maintainers/reviewers; I've CCed them.
>
> Overall, I like the patch, but it's not ready for trunk yet; various
> comments inline below...
>
> >
> > ---
> >
> > gcc/analyzer/ChangeLog:
>
> You could add: PR analyzer/107646 to these ChangeLog entries; have a
> look at how other ChangeLog entries refer to such bugzilla entries.
>
> >
> > * analyzer-language.cc (run_callbacks): New function.
> > (on_finish_translation_unit): New function.
> > * analyzer-language.h (GCC_ANALYZER_LANGUAGE_H): New include.
> > (class translation_unit): New vfuncs.
> >
> > gcc/c/ChangeLog:
> >
> > * c-parser.cc: New functions.
>
> I think this ChangeLog entry needs more detail.
Added in revised version of the patch.
> >
> > gcc/testsuite/ChangeLog:
> >
> > * gcc.dg/plugin/analyzer_cpython_plugin.c: New test.
> >
> > Signed-off-by: Eric Feng <ef2...@columbia.edu>
> > ---
> > gcc/analyzer/analyzer-language.cc | 22 ++
> > gcc/analyzer/analyzer-language.h | 9 +
> > gcc/c/c-parser.cc | 26 ++
> > .../gcc.dg/plugin/analyzer_cpython_plugin.c | 224
> > ++++++++++++++++++
> > 4 files changed, 281 insertions(+)
> > create mode 100644
> > gcc/testsuite/gcc.dg/plugin/analyzer_cpython_plugin.c
> >
> > diff --git a/gcc/analyzer/analyzer-language.cc
> > b/gcc/analyzer/analyzer-language.cc
> > index 2c8910906ee..fc41b9c17b8 100644
> > --- a/gcc/analyzer/analyzer-language.cc
> > +++ b/gcc/analyzer/analyzer-language.cc
> > @@ -35,6 +35,26 @@ static GTY (()) hash_map <tree, tree>
> > *analyzer_stashed_constants;
> > #if ENABLE_ANALYZER
> >
> > namespace ana {
> > +static vec<finish_translation_unit_callback>
> > + *finish_translation_unit_callbacks;
> > +
> > +void
> > +register_finish_translation_unit_callback (
> > + finish_translation_unit_callback callback)
> > +{
> > + if (!finish_translation_unit_callbacks)
> > + vec_alloc (finish_translation_unit_callbacks, 1);
> > + finish_translation_unit_callbacks->safe_push (callback);
> > +}
> > +
> > +void
> > +run_callbacks (logger *logger, const translation_unit &tu)
>
> This function could be "static" since it's not needed outside of
> analyzer-language.cc
>
> > +{
> > + for (auto const &cb : finish_translation_unit_callbacks)
> > + {
> > + cb (logger, tu);
> > + }
> > +}
> >
> > /* Call into TU to try to find a value for NAME.
> > If found, stash its value within analyzer_stashed_constants. */
> > @@ -102,6 +122,8 @@ on_finish_translation_unit (const
> > translation_unit &tu)
> > the_logger.set_logger (new logger (logfile, 0, 0,
> > *global_dc->printer));
> > stash_named_constants (the_logger.get_logger (), tu);
> > +
> > + run_callbacks (the_logger.get_logger (), tu);
> > }
> >
> > /* Lookup NAME in the named constants stashed when the frontend TU
> > finished.
> > diff --git a/gcc/analyzer/analyzer-language.h
> > b/gcc/analyzer/analyzer-language.h
> > index 00f85aba041..8deea52d627 100644
> > --- a/gcc/analyzer/analyzer-language.h
> > +++ b/gcc/analyzer/analyzer-language.h
> > @@ -21,6 +21,8 @@ along with GCC; see the file COPYING3. If not see
> > #ifndef GCC_ANALYZER_LANGUAGE_H
> > #define GCC_ANALYZER_LANGUAGE_H
> >
> > +#include "analyzer/analyzer-logging.h"
> > +
> > #if ENABLE_ANALYZER
> >
> > namespace ana {
> > @@ -35,8 +37,15 @@ class translation_unit
> > have been seen). If it is defined and an integer (e.g. either
> > as a
> > macro or enum), return the INTEGER_CST value, otherwise return
> > NULL. */
> > virtual tree lookup_constant_by_id (tree id) const = 0;
> > + virtual tree lookup_type_by_id (tree id) const = 0;
> > + virtual tree lookup_global_var_by_id (tree id) const = 0;
> > };
> >
> > +typedef void (*finish_translation_unit_callback)
> > + (logger *, const translation_unit &);
> > +void register_finish_translation_unit_callback (
> > + finish_translation_unit_callback callback);
> > +
> > /* Analyzer hook for frontends to call at the end of the TU. */
> >
> > void on_finish_translation_unit (const translation_unit &tu);
> > diff --git a/gcc/c/c-parser.cc b/gcc/c/c-parser.cc
> > index 80920b31f83..f0ee55e416b 100644
> > --- a/gcc/c/c-parser.cc
> > +++ b/gcc/c/c-parser.cc
> > @@ -1695,6 +1695,32 @@ public:
> > return NULL_TREE;
> > }
> >
> > + tree
> > + lookup_type_by_id (tree id) const final override
> > + {
> > + if (tree type_decl = lookup_name (id))
> > + {
> > + if (TREE_CODE (type_decl) == TYPE_DECL)
> > + {
> > + tree record_type = TREE_TYPE (type_decl);
> > + if (TREE_CODE (record_type) == RECORD_TYPE)
> > + return record_type;
> > + }
>
> It looks like something's wrong with the indentation here, but the idea
> seems OK to me (but needs C FE reviewer approval).
>
> > + }
> > +
> > + return NULL_TREE;
> > + }
> > +
> > + tree
> > + lookup_global_var_by_id (tree id) const final override
> > + {
> > + if (tree var_decl = lookup_name (id))
> > + if (TREE_CODE (var_decl) == VAR_DECL)
> > + return var_decl;
>
> Likewise, is this indentation correct?
Sorry about that; looks like my formatter messed this up and neither I
nor check_GNU_style noticed! Should be fixed in the revised patch.
>
> > +
> > + return NULL_TREE;
> > + }
> > +
> > private:
> > /* Attempt to get an INTEGER_CST from MACRO.
> > Only handle the simplest cases: where MACRO's definition is a
> > single
> > diff --git a/gcc/testsuite/gcc.dg/plugin/analyzer_cpython_plugin.c
> > b/gcc/testsuite/gcc.dg/plugin/analyzer_cpython_plugin.c
> > new file mode 100644
> > index 00000000000..285da102edb
> > --- /dev/null
> > +++ b/gcc/testsuite/gcc.dg/plugin/analyzer_cpython_plugin.c
> > @@ -0,0 +1,224 @@
> > +/* -fanalyzer plugin for CPython extension modules */
> > +/* { dg-options "-g" } */
> > +
> > +#define INCLUDE_MEMORY
> > +#include "gcc-plugin.h"
> > +#include "config.h"
> > +#include "system.h"
> > +#include "coretypes.h"
> > +#include "tree.h"
> > +#include "function.h"
> > +#include "basic-block.h"
> > +#include "gimple.h"
> > +#include "gimple-iterator.h"
> > +#include "diagnostic-core.h"
> > +#include "graphviz.h"
> > +#include "options.h"
> > +#include "cgraph.h"
> > +#include "tree-dfa.h"
> > +#include "stringpool.h"
> > +#include "convert.h"
> > +#include "target.h"
> > +#include "fold-const.h"
> > +#include "tree-pretty-print.h"
> > +#include "diagnostic-color.h"
> > +#include "diagnostic-metadata.h"
> > +#include "tristate.h"
> > +#include "bitmap.h"
> > +#include "selftest.h"
> > +#include "function.h"
> > +#include "json.h"
> > +#include "analyzer/analyzer.h"
> > +#include "analyzer/analyzer-language.h"
> > +#include "analyzer/analyzer-logging.h"
> > +#include "ordered-hash-map.h"
> > +#include "options.h"
> > +#include "cgraph.h"
> > +#include "cfg.h"
> > +#include "digraph.h"
> > +#include "analyzer/supergraph.h"
> > +#include "sbitmap.h"
> > +#include "analyzer/call-string.h"
> > +#include "analyzer/program-point.h"
> > +#include "analyzer/store.h"
> > +#include "analyzer/region-model.h"
> > +#include "analyzer/call-details.h"
> > +#include "analyzer/call-info.h"
> > +#include "make-unique.h"
> > +
> > +int plugin_is_GPL_compatible;
> > +
> > +#if ENABLE_ANALYZER
> > +static GTY (()) hash_map<tree, tree> *analyzer_stashed_types;
> > +static GTY (()) hash_map<tree, tree> *analyzer_stashed_globals;
> > +
> > +namespace ana
> > +{
> > +static tree pyobj_record = NULL_TREE;
> > +static tree varobj_record = NULL_TREE;
> > +static tree pylistobj_record = NULL_TREE;
> > +static tree pylongobj_record = NULL_TREE;
> > +static tree pylongtype_vardecl = NULL_TREE;
> > +static tree pylisttype_vardecl = NULL_TREE;
> > +
> > +tree
> > +get_field_by_name (tree type, const char *name)
> > +{
> > + for (tree field = TYPE_FIELDS (type); field; field = TREE_CHAIN
> > (field))
> > + {
> > + if (TREE_CODE (field) == FIELD_DECL)
> > + {
> > + const char *field_name = IDENTIFIER_POINTER (DECL_NAME
> > (field));
> > + if (strcmp (field_name, name) == 0)
> > + return field;
> > + }
> > + }
> > + return NULL_TREE;
> > +}
> > +
> > +static void
> > +maybe_stash_named_type (logger *logger, const translation_unit &tu,
> > + const char *name)
> > +{
> > + LOG_FUNC_1 (logger, "name: %qs", name);
> > + if (!analyzer_stashed_types)
> > + analyzer_stashed_types = hash_map<tree, tree>::create_ggc ();
> > +
> > + tree id = get_identifier (name);
> > + if (tree t = tu.lookup_type_by_id (id))
> > + {
> > + gcc_assert (TREE_CODE (t) == RECORD_TYPE);
> > + analyzer_stashed_types->put (id, t);
> > + if (logger)
> > + logger->log ("found %qs: %qE", name, t);
> > + }
> > + else
> > + {
> > + if (logger)
> > + logger->log ("%qs: not found", name);
> > + }
> > +}
> > +
> > +static void
> > +maybe_stash_global_var (logger *logger, const translation_unit &tu,
> > + const char *name)
> > +{
> > + LOG_FUNC_1 (logger, "name: %qs", name);
> > + if (!analyzer_stashed_globals)
> > + analyzer_stashed_globals = hash_map<tree, tree>::create_ggc ();
> > +
> > + tree id = get_identifier (name);
> > + if (tree t = tu.lookup_global_var_by_id (id))
> > + {
> > + gcc_assert (TREE_CODE (t) == VAR_DECL);
> > + analyzer_stashed_globals->put (id, t);
> > + if (logger)
> > + logger->log ("found %qs: %qE", name, t);
> > + }
> > + else
> > + {
> > + if (logger)
> > + logger->log ("%qs: not found", name);
> > + }
> > +}
> > +
> > +static void
> > +stash_named_types (logger *logger, const translation_unit &tu)
> > +{
> > + LOG_SCOPE (logger);
> > +
> > + maybe_stash_named_type (logger, tu, "PyObject");
> > + maybe_stash_named_type (logger, tu, "PyListObject");
> > + maybe_stash_named_type (logger, tu, "PyVarObject");
> > + maybe_stash_named_type (logger, tu, "PyLongObject");
> > +}
> > +
> > +static void
> > +stash_global_vars (logger *logger, const translation_unit &tu)
> > +{
> > + LOG_SCOPE (logger);
> > +
> > + maybe_stash_global_var (logger, tu, "PyLong_Type");
> > + maybe_stash_global_var (logger, tu, "PyList_Type");
> > +}
> > +
> > +tree
> > +get_stashed_type_by_name (const char *name)
> > +{
> > + if (!analyzer_stashed_types)
> > + return NULL_TREE;
> > + tree id = get_identifier (name);
> > + if (tree *slot = analyzer_stashed_types->get (id))
> > + {
> > + gcc_assert (TREE_CODE (*slot) == RECORD_TYPE);
> > + return *slot;
> > + }
> > + return NULL_TREE;
> > +}
> > +
> > +tree
> > +get_stashed_global_var_by_name (const char *name)
> > +{
> > + if (!analyzer_stashed_globals)
> > + return NULL_TREE;
> > + tree id = get_identifier (name);
> > + if (tree *slot = analyzer_stashed_globals->get (id))
> > + {
> > + gcc_assert (TREE_CODE (*slot) == VAR_DECL);
> > + return *slot;
> > + }
> > + return NULL_TREE;
> > +}
> > +
> > +static void
> > +init_py_structs ()
> > +{
> > + if (pyobj_record == NULL_TREE)
> > + pyobj_record = get_stashed_type_by_name ("PyObject");
>
> This function is called once as the plugin starts up, so I don't think
> we need the various comparisons against NULL_TREE here. Or am I
> missing something?
No, you are right — thank you for mentioning that!
>
> > + if (varobj_record == NULL_TREE)
> > + varobj_record = get_stashed_type_by_name ("PyVarObject");
> > + if (pylistobj_record == NULL_TREE)
> > + pylistobj_record = get_stashed_type_by_name ("PyListObject");
> > + if (pylongobj_record == NULL_TREE)
> > + pylongobj_record = get_stashed_type_by_name ("PyLongObject");
> > + if (pylongtype_vardecl == NULL_TREE)
> > + pylongtype_vardecl = get_stashed_global_var_by_name
> > ("PyLong_Type");
> > + if (pylisttype_vardecl == NULL_TREE)
> > + pylisttype_vardecl = get_stashed_global_var_by_name
> > ("PyList_Type");
> > +}
> > +
> > +static void
> > +cpython_analyzer_init_cb (void *gcc_data, void * /*user_data */)
> > +{
> > + ana::plugin_analyzer_init_iface *iface
> > + = (ana::plugin_analyzer_init_iface *)gcc_data;
> > + LOG_SCOPE (iface->get_logger ());
> > + if (1)
> > + inform (input_location, "got here: cpython_analyzer_init_cb");
>
> We don't want that debugging "inform", or, at least, not on by default.
Fixed.
>
> > +
> > + init_py_structs ();
> > + if (pyobj_record == NULL_TREE)
> > + return;
>
> This conditional doesn't make much sense as-is, but presumably you're
> going to use it to eventually conditionalize the rest of the plugin
> startup, right?
Yes, that's correct.
>
> > +}
> > +} // namespace ana
> > +
> > +#endif /* #if ENABLE_ANALYZER */
> > +
> > +int
> > +plugin_init (struct plugin_name_args *plugin_info,
> > + struct plugin_gcc_version *version)
> > +{
> > +#if ENABLE_ANALYZER
> > + const char *plugin_name = plugin_info->base_name;
> > + if (1)
> > + inform (input_location, "got here; %qs", plugin_name);
>
> Another debugging "inform" that we don't want on by default when
> running the tests.
>
> > + ana::register_finish_translation_unit_callback
> > (&stash_named_types);
> > + ana::register_finish_translation_unit_callback
> > (&stash_global_vars);
> > + register_callback (plugin_info->base_name, PLUGIN_ANALYZER_INIT,
> > + ana::cpython_analyzer_init_cb,
> > + NULL); /* void *user_data */
> > +#else
> > + sorry_no_analyzer ();
> > +#endif
> > + return 0;
> > +}
> > \ No newline at end of file
>
> The patch doesn't touch plugin.exp, so this plugin won't actually get
> compiled yet as part of the testsuite; I think we need a change to
> plugin.exp. You could add a minimal nonempty source file to be
> compiled with the plugin (to test the "gracefully handling non-CPython
> code" case).
Changes made in the revised patch. I have also since added a "sorry"
diagnostic to be emitted in the case of handling non-CPython code, as
I believe it may be helpful for users. When Python/C API definitions
are not found, the user will receive the following message: 'Python/C
API' definitions not found. Please ensure to '#include <Python.h>'.
Exiting. Let me know if you disagree with this addition.
>
> Hope this is constructive; sorry if it comes across as nitpicky in
> places (patch review tends to)
Not at all! Thanks for the feedback. Let me know what you think of the
revised patch.
>
> Dave
>
Best,
Eric
