https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104258
Bug ID: 104258
Summary: [11/12 regression] Generic formal ">" used incorrectly
Product: gcc
Version: 12.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: ada
Assignee: unassigned at gcc dot gnu.org
Reporter: simon at pushface dot org
Target Milestone: ---
Host: x86_64-apple-darwin15
Target: x86_64-apple-darwin15
Build: x86_64-apple-darwin15
Created attachment 52301
--> https://gcc.gnu.org/bugzilla/attachment.cgi?id=52301&action=edit
Reproducer
I have a generic with a formal private type Element_Type and a formal
function ">" (L, R : Element_Type) return Boolean. This gives me
puzzling and compiler-dependent behaviour when I instantiate with
Element_Type => Integer and ">" => "<" (or, as in the code below, ">"
=> Less_Than, where Less_Than outputs a message if it's called).
The output from the code attached is
GCC 10.1.0:
Data'Length: 2
Data'Length > 1: TRUE
Integer'(Data'Length) > 1: TRUE
Standard.">" (Data'Length, 1): TRUE
Standard.">" (Integer'(Data'Length), 1): TRUE
GCC 11.2.0, 12.0.0 of 2022-01-11:
Data'Length: 2
Data'Length > 1: **less_than ( 2, 1) returns FALSE** FALSE
Integer'(Data'Length) > 1: TRUE
Standard.">" (Data'Length, 1): **less_than ( 2, 1) returns FALSE** FALSE
Standard.">" (Integer'(Data'Length), 1): TRUE
