[gcc r13-226] [Ada] Failure compiling "for ... of" loop over a slice

[gcc r13-226] [Ada] Failure compiling "for ... of" loop over a slice

[Pierre-Marie de Rodat]

https://gcc.gnu.org/g:a445a8692c43d14c494e1edf505ee4f33cb98bb6

commit r13-226-ga445a8692c43d14c494e1edf505ee4f33cb98bb6
Author: Steve Baird <baird@adacore.com>
Date:   Fri Jan 14 14:10:25 2022 -0800

    [Ada] Failure compiling "for ... of" loop over a slice
    
    In some cases involving a "for ... of" loop (not to be confused with the
    more common "for ... in" loop) iterating over a slice, compilation would
    fail with an internal compiler error.
    
    gcc/ada/
    
            * sem_util.adb (Get_Actual_Subtype): If a new subtype is built,
            do not freeze it if Expander_Active is False. The idea here is
            to avoid generating an unwanted Freeze_Node for a subtype that
            has been conjured up solely for purposes of preanalysis.

Diff:
---
 gcc/ada/sem_util.adb | 26 ++++++++++++++++++--------
 1 file changed, 18 insertions(+), 8 deletions(-)

diff --git a/gcc/ada/sem_util.adb b/gcc/ada/sem_util.adb
index d76b5d93511..c58b63db924 100644
--- a/gcc/ada/sem_util.adb
+++ b/gcc/ada/sem_util.adb
@@ -10724,14 +10724,24 @@ package body Sem_Util is
                Set_Is_Itype (Atyp);
                Analyze (Decl, Suppress => All_Checks);
                Set_Associated_Node_For_Itype (Atyp, N);
-               Set_Has_Delayed_Freeze (Atyp, False);
-
-               --  We need to freeze the actual subtype immediately. This is
-               --  needed, because otherwise this Itype will not get frozen
-               --  at all, and it is always safe to freeze on creation because
-               --  any associated types must be frozen at this point.
-
-               Freeze_Itype (Atyp, N);
+               if Expander_Active then
+                  Set_Has_Delayed_Freeze (Atyp, False);
+
+                  --  We need to freeze the actual subtype immediately. This is
+                  --  needed because otherwise this Itype will not get frozen
+                  --  at all; it is always safe to freeze on creation because
+                  --  any associated types must be frozen at this point.
+
+                  --  On the other hand, if we are performing preanalysis on
+                  --  a conjured-up copy of a name (see calls to
+                  --  Preanalyze_Range in sem_ch5.adb) then we don't want
+                  --  to freeze Atyp, now or ever. In this case, the tree
+                  --  we eventually pass to the back end should contain no
+                  --  references to Atyp (and a freeze node would contain
+                  --  such a reference). That's why Expander_Active is tested.
+
+                  Freeze_Itype (Atyp, N);
+               end if;
                return Atyp;
 
             --  Otherwise we did not build a declaration, so return original