GH-146527: Add more data to GC statistics and add it to PyDebugOffsets (#146532)

Author: sergey-miryanov

Include/internal/pycore_debug_offsets.h

@@ -222,6 +222,8 @@ typedef struct _Py_DebugOffsets {
         uint64_t size;
         uint64_t collecting;
         uint64_t frame;
+        uint64_t generation_stats_size;
+        uint64_t generation_stats;
     } gc;
 
     // Generator object offset;
@@ -373,6 +375,8 @@ typedef struct _Py_DebugOffsets {
         .size = sizeof(struct _gc_runtime_state), \
         .collecting = offsetof(struct _gc_runtime_state, collecting), \
         .frame = offsetof(struct _gc_runtime_state, frame), \
+        .generation_stats_size = sizeof(struct gc_stats), \
+        .generation_stats = offsetof(struct _gc_runtime_state, generation_stats), \
     }, \
     .gen_object = { \
         .size = sizeof(PyGenObject), \

Include/internal/pycore_interp_structs.h

@@ -177,31 +177,54 @@ struct gc_generation {
                   generations */
 };
 
-struct gc_collection_stats {
-    /* number of collected objects */
-    Py_ssize_t collected;
-    /* total number of uncollectable objects (put into gc.garbage) */
-    Py_ssize_t uncollectable;
-    // Total number of objects considered for collection and traversed:
-    Py_ssize_t candidates;
-    // Duration of the collection in seconds:
-    double duration;
-};
-
 /* Running stats per generation */
 struct gc_generation_stats {
+    PyTime_t ts_start;
+    PyTime_t ts_stop;
+
+    /* heap_size on the start of the collection */
+    Py_ssize_t heap_size;
+
+    /* work_to_do on the start of the collection */
+    Py_ssize_t work_to_do;
+
     /* total number of collections */
     Py_ssize_t collections;
+
+    /* total number of visited objects */
+    Py_ssize_t object_visits;
+
     /* total number of collected objects */
     Py_ssize_t collected;
     /* total number of uncollectable objects (put into gc.garbage) */
     Py_ssize_t uncollectable;
     // Total number of objects considered for collection and traversed:
     Py_ssize_t candidates;
-    // Duration of the collection in seconds:
+
+    Py_ssize_t objects_transitively_reachable;
+    Py_ssize_t objects_not_transitively_reachable;
+
+    // Total duration of the collection in seconds:
     double duration;
 };
 
+#ifdef Py_GIL_DISABLED
+#define GC_YOUNG_STATS_SIZE 1
+#define GC_OLD_STATS_SIZE 1
+#else
+#define GC_YOUNG_STATS_SIZE 11
+#define GC_OLD_STATS_SIZE 3
+#endif
+struct gc_young_stats_buffer {
+    struct gc_generation_stats items[GC_YOUNG_STATS_SIZE];
+    int8_t index;
+};
+
+struct gc_old_stats_buffer {
+    struct gc_generation_stats items[GC_OLD_STATS_SIZE];
+    int8_t index;
+};
+
 enum _GCPhase {
     GC_PHASE_MARK = 0,
     GC_PHASE_COLLECT = 1
@@ -211,6 +234,11 @@ enum _GCPhase {
    signature of gc.collect and change the size of PyStats.gc_stats */
 #define NUM_GENERATIONS 3
 
+struct gc_stats {
+    struct gc_young_stats_buffer young;
+    struct gc_old_stats_buffer old[2];
+};
+
 struct _gc_runtime_state {
     /* Is automatic collection enabled? */
     int enabled;
@@ -220,7 +248,7 @@ struct _gc_runtime_state {
     struct gc_generation old[2];
     /* a permanent generation which won't be collected */
     struct gc_generation permanent_generation;
-    struct gc_generation_stats generation_stats[NUM_GENERATIONS];
+    struct gc_stats generation_stats;
     /* true if we are currently running the collector */
     int collecting;
     // The frame that started the current collection. It might be NULL even when

Modules/gcmodule.c

@@ -347,9 +347,9 @@ gc_get_stats_impl(PyObject *module)
     /* To get consistent values despite allocations while constructing
        the result list, we use a snapshot of the running stats. */
     GCState *gcstate = get_gc_state();
-    for (i = 0; i < NUM_GENERATIONS; i++) {
-        stats[i] = gcstate->generation_stats[i];
-    }
+    stats[0] = gcstate->generation_stats.young.items[gcstate->generation_stats.young.index];
+    stats[1] = gcstate->generation_stats.old[0].items[gcstate->generation_stats.old[0].index];
+    stats[2] = gcstate->generation_stats.old[1].items[gcstate->generation_stats.old[1].index];
 
     PyObject *result = PyList_New(0);
     if (result == NULL)