Re: [PATCH] libgccjit: add some reflection functions in the jit C api

["Martin Liška" <mliska@suse.cz>]

@David: PING

On 11/3/20 11:13 PM, Antoni Boucher via Gcc-patches wrote:
> I was missing a check in gcc_jit_struct_get_field, I added it in this new patch.
> 
> On Thu, Oct 15, 2020 at 05:52:33PM -0400, David Malcolm wrote:
>> On Thu, 2020-10-15 at 13:39 -0400, Antoni Boucher wrote:
>>> Thanks. I updated the patch with these changes.
>>
>> Thanks for patch; review below.  Sorry if it seems excessively nitpicky
>> in places.
>>
>>> 2020-09-1  Antoni Boucher  <bouanto@zoho.com>
>>>
>>>     gcc/jit/
>>>             PR target/96889
>>>             * docs/topics/compatibility.rst (LIBGCCJIT_ABI_14): New ABI tag.
>>
>> 15 now.
>>
>>>             * docs/topics/functions.rst: Add documentation for the
>>>             functions gcc_jit_function_get_return_type and
>>>             gcc_jit_function_get_param_count
>>>             * libgccjit.c: New functions:
>>>               * gcc_jit_function_get_return_type;
>>>               * gcc_jit_function_get_param_count;
>>>               * gcc_jit_function_type_get_return_type;
>>>               * gcc_jit_function_type_get_param_count;
>>>               * gcc_jit_function_type_get_param_type;
>>>               * gcc_jit_type_unqualified;
>>>               * gcc_jit_type_is_array;
>>>               * gcc_jit_type_is_bool;
>>>               * gcc_jit_type_is_function_ptr_type;
>>>               * gcc_jit_type_is_int;
>>>               * gcc_jit_type_is_pointer;
>>>               * gcc_jit_type_is_vector;
>>>               * gcc_jit_vector_type_get_element_type;
>>>               * gcc_jit_vector_type_get_num_units;
>>>               * gcc_jit_struct_get_field;
>>>               * gcc_jit_type_is_struct;
>>>               * gcc_jit_struct_get_field_count;
>>
>> This isn't valid ChangeLog format; it will fail the git hooks.
>>
>>>             * libgccjit.h
>>
>> Likewise.
>>
>>>             * jit-recording.h: New functions (is_struct and is_vector)
>>>             * libgccjit.map (LIBGCCJIT_ABI_14): New ABI tag.
>>
>> 15 now.
>>
>>>
>>>     gcc/testsuite/
>>>             PR target/96889
>>>             * jit.dg/all-non-failing-tests.h: Add test-reflection.c.
>>>             * jit.dg/test-reflection.c: New test.
>>
>> [...]
>>
>>
>>> diff --git a/gcc/jit/docs/topics/functions.rst b/gcc/jit/docs/topics/functions.rst
>>> index eb40d64010e..9819c28cda2 100644
>>> --- a/gcc/jit/docs/topics/functions.rst
>>> +++ b/gcc/jit/docs/topics/functions.rst
>>> @@ -171,6 +171,16 @@ Functions
>>>     underlying string, so it is valid to pass in a pointer to an on-stack
>>>     buffer.
>>>
>>> +.. function::  size_t \
>>> +               gcc_jit_function_get_param_count (gcc_jit_function *func)
>>> +
>>> +   Get the number of parameters of the function.
>>> +
>>> +.. function::  gcc_jit_type \*
>>> +               gcc_jit_function_get_return_type (gcc_jit_function *func)
>>> +
>>> +   Get the return type of the function.
>>> +
>>
>> The documentation part of the patch is incomplete: it hasn't been
>> updated to add all the new entrypoints.
>> Also, the return type of gcc_jit_function_get_param_count is
>> inconsistent (size_t above, but ssize_t below).
>>
>>
>>> diff --git a/gcc/jit/jit-recording.h b/gcc/jit/jit-recording.h
>>> index 30e37aff387..525b8bc921d 100644
>>> --- a/gcc/jit/jit-recording.h
>>> +++ b/gcc/jit/jit-recording.h
>>> @@ -538,7 +538,9 @@ public:
>>>    virtual bool is_bool () const = 0;
>>>    virtual type *is_pointer () = 0;
>>>    virtual type *is_array () = 0;
>>> +  virtual struct_ *is_struct () { return NULL; }
>>
>> Can't you use dyn_cast_struct for this?
>> Or is this about looking through decorated_type? e.g. for const and
>> volatile variants?
>>
>> I guess my question is, what is the purpose of gcc_jit_type_is_struct?
>>
>>>    virtual bool is_void () const { return false; }
>>> +  virtual vector_type *is_vector () { return NULL; }
>>
>> Likewise, can't you use dyn_cast_vector_type for this?
>>
>>>    virtual bool has_known_size () const { return true; }
>>>
>>>    bool is_numeric () const
>>> @@ -595,6 +597,8 @@ public:
>>>    bool is_bool () const FINAL OVERRIDE;
>>>    type *is_pointer () FINAL OVERRIDE { return dereference (); }
>>>    type *is_array () FINAL OVERRIDE { return NULL; }
>>> +  vector_type *is_vector () FINAL OVERRIDE { return NULL; }
>>> +  struct_ *is_struct () FINAL OVERRIDE { return NULL; }
>>
>> Likewise, and this is redundant, as it's merely copying the base class
>> implementation.
>>
>>>    bool is_void () const FINAL OVERRIDE { return m_kind == GCC_JIT_TYPE_VOID; }
>>>
>>>  public:
>>> @@ -629,6 +633,8 @@ public:
>>>    bool is_bool () const FINAL OVERRIDE { return false; }
>>>    type *is_pointer () FINAL OVERRIDE { return m_other_type; }
>>>    type *is_array () FINAL OVERRIDE { return NULL; }
>>> +  vector_type *is_vector () FINAL OVERRIDE { return NULL; }
>>> +  struct_ *is_struct () FINAL OVERRIDE { return NULL; }
>>
>> Likewise.
>>
>>
>>> @@ -655,6 +661,7 @@ public:
>>>    bool is_bool () const FINAL OVERRIDE { return m_other_type->is_bool (); }
>>>    type *is_pointer () FINAL OVERRIDE { return m_other_type->is_pointer (); }
>>>    type *is_array () FINAL OVERRIDE { return m_other_type->is_array (); }
>>> +  struct_ *is_struct () FINAL OVERRIDE { return m_other_type->is_struct (); }
>>
>> Aha: with a decorated type you look through the decoration.
>>
>>>  protected:
>>>    type *m_other_type;
>>> @@ -737,6 +744,8 @@ public:
>>>
>>>    void replay_into (replayer *) FINAL OVERRIDE;
>>>
>>> +  vector_type *is_vector () FINAL OVERRIDE { return this; }
>>> +
>>>  private:
>>>    string * make_debug_string () FINAL OVERRIDE;
>>>    void write_reproducer (reproducer &r) FINAL OVERRIDE;
>>> @@ -765,6 +774,8 @@ class array_type : public type
>>>    bool is_bool () const FINAL OVERRIDE { return false; }
>>>    type *is_pointer () FINAL OVERRIDE { return NULL; }
>>>    type *is_array () FINAL OVERRIDE { return m_element_type; }
>>> +  vector_type *is_vector () FINAL OVERRIDE { return NULL; }
>>> +  struct_ *is_struct () FINAL OVERRIDE { return NULL; }
>>
>> Are these redundant?
>>
>>>    int num_elements () { return m_num_elements; }
>>>
>>>    void replay_into (replayer *) FINAL OVERRIDE;
>>> @@ -799,6 +810,8 @@ public:
>>>    bool is_bool () const FINAL OVERRIDE { return false; }
>>>    type *is_pointer () FINAL OVERRIDE { return NULL; }
>>>    type *is_array () FINAL OVERRIDE { return NULL; }
>>> +  vector_type *is_vector () FINAL OVERRIDE { return NULL; }
>>> +  struct_ *is_struct () FINAL OVERRIDE { return NULL; }
>>
>> Likewise.
>>
>>>    void replay_into (replayer *) FINAL OVERRIDE;
>>>
>>> @@ -912,6 +925,7 @@ public:
>>>    bool is_bool () const FINAL OVERRIDE { return false; }
>>>    type *is_pointer () FINAL OVERRIDE { return NULL; }
>>>    type *is_array () FINAL OVERRIDE { return NULL; }
>>> +  vector_type *is_vector () FINAL OVERRIDE { return NULL; }
>>
>> Likewise.
>>
>>>    bool has_known_size () const FINAL OVERRIDE { return m_fields != NULL; }
>>>
>>> @@ -943,6 +957,8 @@ public:
>>>
>>>    const char *access_as_type (reproducer &r) FINAL OVERRIDE;
>>>
>>> +  struct_ *is_struct () FINAL OVERRIDE { return this; }
>>> +
>>>  private:
>>>    string * make_debug_string () FINAL OVERRIDE;
>>>    void write_reproducer (reproducer &r) FINAL OVERRIDE;
>>> diff --git a/gcc/jit/libgccjit.c b/gcc/jit/libgccjit.c
>>> index a00aefc7108..8e6a9e85ff0 100644
>>> --- a/gcc/jit/libgccjit.c
>>> +++ b/gcc/jit/libgccjit.c
>>> @@ -21,6 +21,7 @@ along with GCC; see the file COPYING3.  If not see
>>>  #include "config.h"
>>>  #include "system.h"
>>>  #include "coretypes.h"
>>> +#include "tm.h"
>>
>> Why is this included?  It's not mentioned in the ChangeLog.
>>
>>>  #include "timevar.h"
>>>  #include "typed-splay-tree.h"
>>>  #include "cppbuiltin.h"
>>> @@ -60,6 +61,14 @@ struct gcc_jit_struct : public gcc::jit::recording::struct_
>>>  {
>>>  };
>>>
>>> +struct gcc_jit_function_type : public gcc::jit::recording::function_type
>>> +{
>>> +};
>>> +
>>> +struct gcc_jit_vector_type : public gcc::jit::recording::vector_type
>>> +{
>>> +};
>>> +
>>
>> FWIW these aren't mentioned in the ChangeLog either.
>>
>>>  struct gcc_jit_field : public gcc::jit::recording::field
>>>  {
>>>  };
>>> @@ -510,6 +519,197 @@ gcc_jit_type_get_volatile (gcc_jit_type *type)
>>
>> [...]
>>
>>> +/* Public entrypoint.  See description in libgccjit.h.
>>> +
>>> +   After error-checking, the real work is done by the
>>> +   gcc::jit::recording::function_type::get_param_types method, in
>>> +   jit-recording.c.  */
>>> +
>>> +gcc_jit_type *
>>> +gcc_jit_function_type_get_param_type (gcc_jit_function_type *function_type,
>>> +                int index)
>>> +{
>>> +  RETURN_NULL_IF_FAIL (function_type, NULL, NULL, "NULL struct_type");
>>> +  int num_params = function_type->get_param_types ().length ();
>>> +  gcc::jit::recording::context *ctxt = function_type->m_ctxt;
>>
>> Should check that index >= 0 as well (or make it unsigned).
>>
>> Given that gcc_jit_function_get_param_count returns ssize_t, should
>> that be the type of "index"?
>>
>>> +  RETURN_NULL_IF_FAIL_PRINTF3 (index < num_params,
>>> +                   ctxt, NULL,
>>> +                   "index of %d is too large (%s has %d params)",
>>> +                   index,
>>> +                   function_type->get_debug_string (),
>>> +                   num_params);
>>> +  return (gcc_jit_type *)function_type->get_param_types ()[index];
>>> +}
>>
>> [...]
>>
>>> +extern gcc_jit_field *
>>> +gcc_jit_struct_get_field (gcc_jit_struct *struct_type,
>>> +               int index)
>>> +{
>>> +  RETURN_NULL_IF_FAIL (struct_type, NULL, NULL, "NULL struct type");
>>
>> Similar comments as per gcc_jit_function_type_get_param_type above.
>> Should check here that index is in range (>=0 and < num fields).
>>
>>> +  return (gcc_jit_field *)struct_type->get_fields ()->get_field (index);
>>> +}
>>> +
>>> +/* Public entrypoint.  See description in libgccjit.h.
>>> +
>>> +   After error-checking, this calls the trivial
>>> +   gcc::jit::recording::struct_::get_fields method in
>>> +   jit-recording.h.  */
>>> +
>>> +ssize_t
>>> +gcc_jit_struct_get_field_count (gcc_jit_struct *struct_type)
>>> +{
>>> +  RETURN_VAL_IF_FAIL (struct_type, -1, NULL, NULL, "NULL struct type");
>>> +  return struct_type->get_fields ()->length ();
>>> +}
>>> +
>>
>> [...]
>>
>>> diff --git a/gcc/jit/libgccjit.h b/gcc/jit/libgccjit.h
>>> index 7134841bb07..6056d00a3ea 100644
>>> --- a/gcc/jit/libgccjit.h
>>> +++ b/gcc/jit/libgccjit.h
>>> @@ -96,6 +96,12 @@ typedef struct gcc_jit_field gcc_jit_field;
>>>     the layout for, or an opaque type.  */
>>>  typedef struct gcc_jit_struct gcc_jit_struct;
>>>
>>> +/* A gcc_jit_function_type encapsulates a function type.  */
>>> +typedef struct gcc_jit_function_type gcc_jit_function_type;
>>> +
>>> +/* A gcc_jit_vector_type encapsulates a vector type.  */
>>> +typedef struct gcc_jit_vector_type gcc_jit_vector_type;
>>> +
>>>  /* A gcc_jit_function encapsulates a function: either one that you're
>>>     creating yourself, or a reference to one that you're dynamically
>>>     linking to within the rest of the process.  */
>>> @@ -586,6 +592,61 @@ gcc_jit_type_get_const (gcc_jit_type *type);
>>>  extern gcc_jit_type *
>>>  gcc_jit_type_get_volatile (gcc_jit_type *type);
>>>
>>> +/* Given type "T", return TRUE if the type is an array.  */
>>> +extern gcc_jit_type *
>>> +gcc_jit_type_is_array (gcc_jit_type *type);
>>
>> The comment is unclear; does this actually return the type vs NULL?
>>
>> As noted above, I think we need to think about what happens on
>> qualified types; is stripping them the responsibility of the caller?
>> Should it be?  (I'm thinking "yes", but presumably all this is
>> motivated by your rust hacking).
>>
>> Also, there isn't a type "T" here; the param name is "type", and there
>> isn't a supporting example using "T" as a placeholder.
>> Maybe just lose the 'Given type "T", ' prefix here and below (unless
>> giving a example).
>>
>>> +/* Given type "T", return TRUE if the type is a bool.  */
>>> +extern int
>>> +gcc_jit_type_is_bool (gcc_jit_type *type);
>>
>> I'd prefer "return non-zero" over "return TRUE" (here, and in various
>> places below).
>>
>> Does this check for the "bool" type i.e.:
>>  /* C++'s bool type; also C99's "_Bool" type, aka "bool" if using
>>     stdbool.h.  */
>>  GCC_JIT_TYPE_BOOL,
>> ?
>>
>>> +/* Given type "T", return the function type if it is one.  */
>>
>> ..."or NULL".
>>
>>> +extern gcc_jit_function_type *
>>> +gcc_jit_type_is_function_ptr_type (gcc_jit_type *type);
>>> +
>>> +/* Given a function type, return its return type.  */
>>> +extern gcc_jit_type *
>>> +gcc_jit_function_type_get_return_type (gcc_jit_function_type *function_type);
>>> +
>>> +/* Given a function type, return its number of parameters.  */
>>> +extern ssize_t
>>> +gcc_jit_function_type_get_param_count (gcc_jit_function_type *function_type);
>>> +
>>> +/* Given a function type, return the type of the specified parameter.  */
>>> +extern gcc_jit_type *
>>> +gcc_jit_function_type_get_param_type (gcc_jit_function_type *function_type,
>>> +                int index);
>>> +
>>> +/* Given type "T", return the type pointed by the pointer type
>>> + * or NULL if it's not a pointer.  */
>>> +extern gcc_jit_type *
>>> +gcc_jit_type_is_pointer (gcc_jit_type *type);
>>> +
>>> +/* Given type "T", return TRUE if the type is an int.  */
>>> +extern int
>>> +gcc_jit_type_is_int (gcc_jit_type *type);
>>
>> I'm not sure I understand what this function does, which suggests it's
>> misnamed.  The current name suggests that it checks for the C "int"
>> type, whereas I believe it actually checks for an integral
>> type.  Should it be:
>>
>> extern int
>> gcc_jit_type_is_integral (gcc_jit_type *type);
>>
>> ?
>>
>> What's the output?  Do you merely need nonzero/zero for boolean, or
>> information on the width of the type/signedness, or whatnot?  (perhaps
>> have some out params for extracting width/signedness for when it
>> returns nonzero?)
>>
>> Looking at the testcase, it looks like it currently returns false on
>> the _Bool type.  Is that correct?
>>
>>
>>> +/* Given type "T", return the unqualified type, i.e. for a
>>> + * decorated type, return the type it wraps.  */
>>> +extern gcc_jit_type *
>>> +gcc_jit_type_unqualified (gcc_jit_type *type);
>>
>> "decorated" is a term used inside the implementation, but not in the
>> user-facing docs.
>>
>> How about:
>>
>>  /* Given a type, return the unqualified type, removing "const",
>> "volatile"
>>     and alignment qualifiers.  */
>>
>> or somesuch.
>>
>>> +/* Given type "T", return TRUE if the type is a vector type.  */
>>> +extern gcc_jit_vector_type *
>>> +gcc_jit_type_is_vector (gcc_jit_type *type);
>>
>> As above, this doesn't return TRUE, it's more of a dynamic cast.
>> Likewise about responsibility for qualifiers.
>>
>>> +/* Given type "T", return the struct type or NULL.  */
>>> +extern gcc_jit_struct *
>>> +gcc_jit_type_is_struct (gcc_jit_type *type);
>>
>> Wording could be improved as noted above.
>>
>>> +/* Given a vector type, return the number of units it contains.  */
>>> +extern ssize_t
>>> +gcc_jit_vector_type_get_num_units (gcc_jit_vector_type *vector_type);
>>> +
>>> +/* Given a vector type, return the type of its elements.  */
>>> +extern gcc_jit_type *
>>> +gcc_jit_vector_type_get_element_type (gcc_jit_vector_type *vector_type);
>>> +
>>
>> [...]
>>
>>> @@ -647,6 +708,15 @@ gcc_jit_struct_set_fields (gcc_jit_struct *struct_type,
>>>                 int num_fields,
>>>                 gcc_jit_field **fields);
>>>
>>> +/* Get a field by index.  */
>>> +extern gcc_jit_field *
>>> +gcc_jit_struct_get_field (gcc_jit_struct *struct_type,
>>> +               int index);
>>> +
>>> +/* Get the number of fields.  */
>>> +extern ssize_t
>>> +gcc_jit_struct_get_field_count (gcc_jit_struct *struct_type);
>>
>> Do we need these for unions?  Or skip them for now?
>>
>>> @@ -740,6 +810,14 @@ gcc_jit_function_as_object (gcc_jit_function *func);
>>>  extern gcc_jit_param *
>>>  gcc_jit_function_get_param (gcc_jit_function *func, int index);
>>>
>>> +/* Get the number of params of a function.  */
>>> +extern ssize_t
>>> +gcc_jit_function_get_param_count (gcc_jit_function *func);
>>> +
>>> +/* Get the return type of a function.  */
>>> +extern gcc_jit_type *
>>> +gcc_jit_function_get_return_type (gcc_jit_function *func);
>>
>> Are these redundant if there's a way to ask a gcc_jit_function for its
>> type, and then to use the gcc_jit_function_type_ functions?
>> (I don't mind if they are)
>>
>>> @@ -1518,6 +1596,47 @@ gcc_jit_version_minor (void);
>>>  extern int
>>>  gcc_jit_version_patchlevel (void);
>>>
>>> +#define LIBGCCJIT_HAVE_gcc_jit_function_reflection
>>> +
>>> +/* Reflection functions to get the number of parameters, return type of
>>> +   a function and whether a type is a bool from the C API.
>>> +
>>> +   This API entrypoint was added in LIBGCCJIT_ABI_15; you can test for its
>>> +   presence using
>>> +     #ifdef LIBGCCJIT_HAVE_gcc_jit_function_reflection
>>> +*/
>>
>> This has become a big mixture of things beyond just function
>> reflection, so I think I'd prefer it if each entrypoint "funcname" had
>> its own "LIBGCCJIT_HAVE_funcname" define.  Or just "REFLECTION" rather
>> than "function_reflection", that's probably simpler.
>> Note that the existing "HAVE_foo" defines are lowercase "foo" when it's
>> an actual entrypoint named "foo", and are uppercase otherwise e.g.
>> LIBGCCJIT_HAVE_SWITCH_STATEMENTS and LIBGCCJIT_HAVE_TIMING_API.
>>
>>> +extern gcc_jit_type *
>>> +gcc_jit_function_get_return_type (gcc_jit_function *func);
>>> +extern ssize_t
>>> +gcc_jit_function_get_param_count (gcc_jit_function *func);
>>> +extern gcc_jit_type *
>>> +gcc_jit_type_is_array (gcc_jit_type *type);
>>> +extern int
>>> +gcc_jit_type_is_bool (gcc_jit_type *type);
>>> +extern gcc_jit_function_type *
>>> +gcc_jit_type_is_function_ptr_type (gcc_jit_type *type);
>>> +extern gcc_jit_type *
>>> +gcc_jit_function_type_get_return_type (gcc_jit_function_type *function_type);
>>> +extern ssize_t
>>> +gcc_jit_function_type_get_param_count (gcc_jit_function_type *function_type);
>>> +extern gcc_jit_type *
>>> +gcc_jit_function_type_get_param_type (gcc_jit_function_type *function_type,
>>> +                int index);
>>> +extern int
>>> +gcc_jit_type_is_int (gcc_jit_type *type);
>>> +extern gcc_jit_type *
>>> +gcc_jit_type_is_pointer (gcc_jit_type *type);
>>> +extern gcc_jit_vector_type *
>>> +gcc_jit_type_is_vector (gcc_jit_type *type);
>>> +extern gcc_jit_struct *
>>> +gcc_jit_type_is_struct (gcc_jit_type *type);
>>> +extern ssize_t
>>> +gcc_jit_vector_type_get_num_units (gcc_jit_vector_type *vector_type);
>>> +extern gcc_jit_type *
>>> +gcc_jit_vector_type_get_element_type (gcc_jit_vector_type *vector_type);
>>> +extern gcc_jit_type *
>>> +gcc_jit_type_unqualified (gcc_jit_type *type);
>>
>> Is this part of the patch a duplicate copy of the various things added
>> above?  Perhaps copy-and-paste error, or am I misreading this?
>>
>> [...]
>>
>> Hope this is constructive
>> Dave
>>