py/compile: Allow new qstrs to be allocated at all compiler passes.

Prior to this commit, all qstrs were required to be allocated (by calling
mp_emit_common_use_qstr) in the MP_PASS_SCOPE pass (the first one).  But
this is an unnecessary restriction, which is lifted by this commit.
Lifting the restriction simplifies the compiler because it can allocate
qstrs in later passes.

This also generates better code, because in some cases (eg when a variable
is closed over) the scope of an identifier is not known until a bit later
and then the identifier no longer needs its qstr allocated in the global
table.

Code size is reduced for all ports with this commit.

Signed-off-by: Damien George <damien@micropython.org>

4 files changed, 22 insertions, 33 deletions

diff --git a/py/compile.c b/py/compile.c
index 2760d8bfb..75cd562a1 100644
--- a/py/compile.c
+++ b/py/compile.c
@@ -188,7 +188,6 @@ typedef struct _compiler_t {
     mp_emit_common_t emit_common;
 
     emit_t *emit;                                   // current emitter
-    emit_t *emit_bc;
     #if NEED_METHOD_TABLE
     const emit_method_table_t *emit_method_table;   // current emit method table
     #endif
@@ -359,8 +358,7 @@ STATIC void compile_generic_all_nodes(compiler_t *comp, mp_parse_node_struct_t *
 STATIC void compile_load_id(compiler_t *comp, qstr qst) {
     if (comp->pass == MP_PASS_SCOPE) {
         mp_emit_common_get_id_for_load(comp->scope_cur, qst);
-    }
-    {
+    } else {
         #if NEED_METHOD_TABLE
         mp_emit_common_id_op(comp->emit, &comp->emit_method_table->load_id, comp->scope_cur, qst);
         #else
@@ -372,8 +370,7 @@ STATIC void compile_load_id(compiler_t *comp, qstr qst) {
 STATIC void compile_store_id(compiler_t *comp, qstr qst) {
     if (comp->pass == MP_PASS_SCOPE) {
         mp_emit_common_get_id_for_modification(comp->scope_cur, qst);
-    }
-    {
+    } else {
         #if NEED_METHOD_TABLE
         mp_emit_common_id_op(comp->emit, &comp->emit_method_table->store_id, comp->scope_cur, qst);
         #else
@@ -385,8 +382,7 @@ STATIC void compile_store_id(compiler_t *comp, qstr qst) {
 STATIC void compile_delete_id(compiler_t *comp, qstr qst) {
     if (comp->pass == MP_PASS_SCOPE) {
         mp_emit_common_get_id_for_modification(comp->scope_cur, qst);
-    }
-    {
+    } else {
         #if NEED_METHOD_TABLE
         mp_emit_common_id_op(comp->emit, &comp->emit_method_table->delete_id, comp->scope_cur, qst);
         #else
@@ -2154,7 +2150,6 @@ STATIC void compile_lambdef(compiler_t *comp, mp_parse_node_struct_t *pns) {
 STATIC void compile_namedexpr_helper(compiler_t *comp, mp_parse_node_t pn_name, mp_parse_node_t pn_expr) {
     if (!MP_PARSE_NODE_IS_ID(pn_name)) {
         compile_syntax_error(comp, (mp_parse_node_t)pn_name, MP_ERROR_TEXT("can't assign to expression"));
-        return; // because pn_name is not a valid qstr (in compile_store_id below)
     }
     compile_node(comp, pn_expr);
     EMIT(dup_top);
@@ -2858,7 +2853,6 @@ STATIC void compile_scope_func_lambda_param(compiler_t *comp, mp_parse_node_t pn
             // comes before a star, so counts as a positional parameter
             comp->scope_cur->num_pos_args += 1;
         }
-        mp_emit_common_use_qstr(&comp->emit_common, param_name);
     } else {
         assert(MP_PARSE_NODE_IS_STRUCT(pn));
         pns = (mp_parse_node_struct_t *)pn;
@@ -2872,7 +2866,6 @@ STATIC void compile_scope_func_lambda_param(compiler_t *comp, mp_parse_node_t pn
                 // comes before a star, so counts as a positional parameter
                 comp->scope_cur->num_pos_args += 1;
             }
-            mp_emit_common_use_qstr(&comp->emit_common, param_name);
         } else if (MP_PARSE_NODE_STRUCT_KIND(pns) == pn_star) {
             if (comp->have_star) {
                 // more than one star
@@ -3111,7 +3104,6 @@ STATIC bool compile_scope(compiler_t *comp, scope_t *scope, pass_kind_t pass) {
         if (comp->pass == MP_PASS_SCOPE) {
             scope_find_or_add_id(comp->scope_cur, qstr_arg, ID_INFO_KIND_LOCAL);
             scope->num_pos_args = 1;
-            mp_emit_common_use_qstr(&comp->emit_common, MP_QSTR__star_);
         }
 
         // Set the source line number for the start of the comprehension
@@ -3361,7 +3353,7 @@ STATIC void compile_scope_inline_asm(compiler_t *comp, scope_t *scope, pass_kind
 }
 #endif
 
-STATIC void scope_compute_things(scope_t *scope, mp_emit_common_t *emit_common) {
+STATIC void scope_compute_things(scope_t *scope) {
     // in MicroPython we put the *x parameter after all other parameters (except **y)
     if (scope->scope_flags & MP_SCOPE_FLAG_VARARGS) {
         id_info_t *id_param = NULL;
@@ -3450,7 +3442,6 @@ STATIC void scope_compute_things(scope_t *scope, mp_emit_common_t *emit_common)
             }
             scope->num_pos_args += num_free; // free vars are counted as params for passing them into the function
             scope->num_locals += num_free;
-            mp_emit_common_use_qstr(emit_common, MP_QSTR__star_);
         }
     }
 }
@@ -3481,7 +3472,6 @@ mp_compiled_module_t mp_compile_to_raw_code(mp_parse_tree_t *parse_tree, qstr so
 
     // compile MP_PASS_SCOPE
     comp->emit = emit_bc;
-    comp->emit_bc = emit_bc;
     #if MICROPY_EMIT_NATIVE
     comp->emit_method_table = &emit_bc_method_table;
     #endif
@@ -3519,7 +3509,7 @@ mp_compiled_module_t mp_compile_to_raw_code(mp_parse_tree_t *parse_tree, qstr so
         }
         #endif
 
-        scope_compute_things(s, &comp->emit_common);
+        scope_compute_things(s);
     }
 
     // set max number of labels now that it's calculated
diff --git a/py/emitcommon.c b/py/emitcommon.c
index 647430232..02f96f111 100644
--- a/py/emitcommon.c
+++ b/py/emitcommon.c
@@ -34,7 +34,6 @@
 qstr_short_t mp_emit_common_use_qstr(mp_emit_common_t *emit, qstr qst) {
     mp_map_elem_t *elem = mp_map_lookup(&emit->qstr_map, MP_OBJ_NEW_QSTR(qst), MP_MAP_LOOKUP_ADD_IF_NOT_FOUND);
     if (elem->value == MP_OBJ_NULL) {
-        assert(emit->pass == MP_PASS_SCOPE);
         elem->value = MP_OBJ_NEW_SMALL_INT(emit->qstr_map.used - 1);
     }
     return MP_OBJ_SMALL_INT_VALUE(elem->value);
diff --git a/tests/cmdline/cmd_parsetree.py.exp b/tests/cmdline/cmd_parsetree.py.exp
index 7f9aa047e..3049267c0 100644
--- a/tests/cmdline/cmd_parsetree.py.exp
+++ b/tests/cmdline/cmd_parsetree.py.exp
@@ -42,9 +42,9 @@
 File cmdline/cmd_parsetree.py, code block '<module>' (descriptor: \.\+, bytecode @\.\+ 62 bytes)
 Raw bytecode (code_info_size=13, bytecode_size=49):
  20 16 01 60 27 22 23 24 24 24 24 24 25 2a 00 5f
- 4b 04 16 02 42 3a 51 16 03 10 04 16 05 23 00 16
- 06 23 01 16 07 23 02 16 08 23 03 16 09 22 80 7b
- 16 0a 23 04 14 0b 11 05 36 01 16 0c 51 63
+ 4b 04 16 04 42 3a 51 16 05 10 02 16 06 23 00 16
+ 07 23 01 16 08 23 02 16 09 23 03 16 0a 22 80 7b
+ 16 0b 23 04 14 03 11 06 36 01 16 0c 51 63
 arg names:
 (N_STATE 5)
 (N_EXC_STACK 0)
diff --git a/tests/cmdline/cmd_showbc.py.exp b/tests/cmdline/cmd_showbc.py.exp
index f3ca2eb63..45a1d169b 100644
--- a/tests/cmdline/cmd_showbc.py.exp
+++ b/tests/cmdline/cmd_showbc.py.exp
@@ -1,9 +1,9 @@
 File cmdline/cmd_showbc.py, code block '<module>' (descriptor: \.\+, bytecode @\.\+ 63 bytes)
 Raw bytecode (code_info_size=18, bytecode_size=45):
  10 20 01 60 20 84 7d 64 60 88 07 64 60 69 20 62
- 64 20 32 00 16 02 32 01 16 02 81 2a 01 53 33 02
- 16 02 32 03 16 02 54 32 04 10 03 34 02 16 03 19
- 03 32 05 16 02 80 10 04 2a 01 1b 05 69 51 63
+ 64 20 32 00 16 05 32 01 16 05 81 2a 01 53 33 02
+ 16 05 32 03 16 05 54 32 04 10 02 34 02 16 02 19
+ 02 32 05 16 05 80 10 03 2a 01 1b 04 69 51 63
 arg names:
 (N_STATE 3)
 (N_EXC_STACK 0)
@@ -49,7 +49,7 @@ arg names:
 44 RETURN_VALUE
 File cmdline/cmd_showbc.py, code block 'f' (descriptor: \.\+, bytecode @\.\+ 45\[46\] bytes)
 Raw bytecode (code_info_size=8\[46\], bytecode_size=370):
- a8 12 9\[bf\] 03 02 60 60 26 22 24 64 22 24 25 25 24
+ a8 12 9\[bf\] 03 05 60 60 26 22 24 64 22 24 25 25 24
  26 23 63 22 22 25 23 23 2f 6c 25 65 25 25 69 68
  26 65 27 6a 62 20 23 62 2a 29 69 24 25 28 67 26
 ########
@@ -405,7 +405,7 @@ arg names:
 369 RETURN_VALUE
 File cmdline/cmd_showbc.py, code block 'f' (descriptor: \.\+, bytecode @\.\+ 59 bytes)
 Raw bytecode (code_info_size=8, bytecode_size=51):
- a8 10 0a 02 80 82 34 38 81 57 c0 57 c1 57 c2 57
+ a8 10 0a 05 80 82 34 38 81 57 c0 57 c1 57 c2 57
  c3 57 c4 57 c5 57 c6 57 c7 57 c8 c9 82 57 ca 57
  cb 57 cc 57 cd 57 ce 57 cf 57 26 10 57 26 11 57
  26 12 26 13 b9 24 13 f2 59 51 63
@@ -464,7 +464,7 @@ arg names:
 50 RETURN_VALUE
 File cmdline/cmd_showbc.py, code block 'f' (descriptor: \.\+, bytecode @\.\+ 20 bytes)
 Raw bytecode (code_info_size=9, bytecode_size=11):
- a1 01 0b 02 06 80 88 40 00 82 2a 01 53 b0 21 00
+ a1 01 0b 05 06 80 88 40 00 82 2a 01 53 b0 21 00
  01 c1 51 63
 arg names: a
 (N_STATE 5)
@@ -483,7 +483,7 @@ arg names: a
 10 RETURN_VALUE
 File cmdline/cmd_showbc.py, code block 'f' (descriptor: \.\+, bytecode @\.\+ 21 bytes)
 Raw bytecode (code_info_size=8, bytecode_size=13):
- 88 40 0a 02 80 8f 23 23 51 67 59 81 67 59 81 5e
+ 88 40 0a 05 80 8f 23 23 51 67 59 81 67 59 81 5e
  51 68 59 51 63
 arg names:
 (N_STATE 2)
@@ -507,7 +507,7 @@ arg names:
 12 RETURN_VALUE
 File cmdline/cmd_showbc.py, code block 'Class' (descriptor: \.\+, bytecode @\.\+ 1\[56\] bytes)
 Raw bytecode (code_info_size=\[56\], bytecode_size=10):
- 00 \.\+ 11 0c 16 0d 10 03 16 0e 51 63
+ 00 \.\+ 11 0f 16 10 10 02 16 11 51 63
 arg names:
 (N_STATE 1)
 (N_EXC_STACK 0)
@@ -522,7 +522,7 @@ arg names:
 09 RETURN_VALUE
 File cmdline/cmd_showbc.py, code block 'f' (descriptor: \.\+, bytecode @\.\+ 18 bytes)
 Raw bytecode (code_info_size=6, bytecode_size=12):
- 19 08 02 0f 80 9c 12 10 12 11 b0 15 02 36 00 59
+ 19 08 05 12 80 9c 12 13 12 14 b0 15 05 36 00 59
  51 63
 arg names: self
 (N_STATE 4)
@@ -539,7 +539,7 @@ arg names: self
 11 RETURN_VALUE
 File cmdline/cmd_showbc.py, code block '<genexpr>' (descriptor: \.\+, bytecode @\.\+ 28 bytes)
 Raw bytecode (code_info_size=9, bytecode_size=19):
- c3 40 0c 12 04 04 04 80 3b 53 b2 53 53 4b 0b c3
+ c3 40 0c 09 03 03 03 80 3b 53 b2 53 53 4b 0b c3
  25 01 44 39 25 00 67 59 42 33 51 63
 arg names: * * *
 (N_STATE 9)
@@ -562,7 +562,7 @@ arg names: * * *
 18 RETURN_VALUE
 File cmdline/cmd_showbc.py, code block '<listcomp>' (descriptor: \.\+, bytecode @\.\+ 26 bytes)
 Raw bytecode (code_info_size=8, bytecode_size=18):
- 4b 0c 14 04 04 04 80 3c 2b 00 b2 5f 4b 0b c3 25
+ 4b 0c 0a 03 03 03 80 3c 2b 00 b2 5f 4b 0b c3 25
  01 44 39 25 00 2f 14 42 33 63
 arg names: * * *
 (N_STATE 10)
@@ -582,7 +582,7 @@ arg names: * * *
 17 RETURN_VALUE
 File cmdline/cmd_showbc.py, code block '<dictcomp>' (descriptor: \.\+, bytecode @\.\+ 28 bytes)
 Raw bytecode (code_info_size=8, bytecode_size=20):
- 53 0c 15 04 04 04 80 3d 2c 00 b2 5f 4b 0d c3 25
+ 53 0c 0b 03 03 03 80 3d 2c 00 b2 5f 4b 0d c3 25
  01 44 39 25 00 25 00 2f 19 42 31 63
 arg names: * * *
 (N_STATE 11)
@@ -603,7 +603,7 @@ arg names: * * *
 19 RETURN_VALUE
 File cmdline/cmd_showbc.py, code block 'closure' (descriptor: \.\+, bytecode @\.\+ 20 bytes)
 Raw bytecode (code_info_size=8, bytecode_size=12):
- 19 0c 16 04 80 6f 25 23 25 00 81 f2 c1 81 27 00
+ 19 0c 0c 03 80 6f 25 23 25 00 81 f2 c1 81 27 00
  29 00 51 63
 arg names: *
 (N_STATE 4)
@@ -623,7 +623,7 @@ arg names: *
 11 RETURN_VALUE
 File cmdline/cmd_showbc.py, code block 'f' (descriptor: \.\+, bytecode @\.\+ 13 bytes)
 Raw bytecode (code_info_size=8, bytecode_size=5):
- 9a 01 0a 02 04 09 80 8b b1 25 00 f2 63
+ 9a 01 0a 05 03 08 80 8b b1 25 00 f2 63
 arg names: * b
 (N_STATE 4)
 (N_EXC_STACK 0)