public inbox for gcc-cvs@sourceware.org
help / color / mirror / Atom feed
* [gcc r12-5727] [Ada] Refactor nested loops in warning on unassigned out parameter
@ 2021-12-02 16:28 Pierre-Marie de Rodat
0 siblings, 0 replies; only message in thread
From: Pierre-Marie de Rodat @ 2021-12-02 16:28 UTC (permalink / raw)
To: gcc-cvs
https://gcc.gnu.org/g:1fde9333dc31007083ac792896820ca29837c342
commit r12-5727-g1fde9333dc31007083ac792896820ca29837c342
Author: Piotr Trojanek <trojanek@adacore.com>
Date: Mon Nov 22 12:15:29 2021 +0100
[Ada] Refactor nested loops in warning on unassigned out parameter
gcc/ada/
* sem_warn.adb (Warn_On_Unassigned_Out_Parameter): Move inner
loop at the beginning of subprogram, so it is executed only
once; fix order in the "add an ad hoc" phrase.
Diff:
---
gcc/ada/sem_warn.adb | 44 +++++++++++++++++++++-----------------------
1 file changed, 21 insertions(+), 23 deletions(-)
diff --git a/gcc/ada/sem_warn.adb b/gcc/ada/sem_warn.adb
index 3d7e4124e0a..2842a2ccb40 100644
--- a/gcc/ada/sem_warn.adb
+++ b/gcc/ada/sem_warn.adb
@@ -4237,8 +4237,7 @@ package body Sem_Warn is
(Return_Node : Node_Id;
Scope_Id : Entity_Id)
is
- Form : Entity_Id;
- Form2 : Entity_Id;
+ Form : Entity_Id;
begin
-- Ignore if procedure or return statement does not come from source
@@ -4249,6 +4248,26 @@ package body Sem_Warn is
return;
end if;
+ -- Before we issue the warning, add an ad hoc defence against the most
+ -- common case of false positives with this warning which is the case
+ -- where there is a Boolean OUT parameter that has been set, and whose
+ -- meaning is "ignore the values of the other parameters". We can't of
+ -- course reliably tell this case at compile time, but the following
+ -- test kills a lot of false positives, without generating a significant
+ -- number of false negatives (missed real warnings).
+
+ Form := First_Formal (Scope_Id);
+ while Present (Form) loop
+ if Ekind (Form) = E_Out_Parameter
+ and then Root_Type (Etype (Form)) = Standard_Boolean
+ and then not Never_Set_In_Source_Check_Spec (Form)
+ then
+ return;
+ end if;
+
+ Next_Formal (Form);
+ end loop;
+
-- Loop through formals
Form := First_Formal (Scope_Id);
@@ -4263,27 +4282,6 @@ package body Sem_Warn is
and then Is_Scalar_Type (Etype (Form))
and then not Present (Unset_Reference (Form))
then
- -- Before we issue the warning, an add ad hoc defence against the
- -- most common case of false positives with this warning which is
- -- the case where there is a Boolean OUT parameter that has been
- -- set, and whose meaning is "ignore the values of the other
- -- parameters". We can't of course reliably tell this case at
- -- compile time, but the following test kills a lot of false
- -- positives, without generating a significant number of false
- -- negatives (missed real warnings).
-
- Form2 := First_Formal (Scope_Id);
- while Present (Form2) loop
- if Ekind (Form2) = E_Out_Parameter
- and then Root_Type (Etype (Form2)) = Standard_Boolean
- and then not Never_Set_In_Source_Check_Spec (Form2)
- then
- return;
- end if;
-
- Next_Formal (Form2);
- end loop;
-
-- Here all conditions are met, record possible unset reference
Set_Unset_Reference (Form, Return_Node);
^ permalink raw reply [flat|nested] only message in thread
only message in thread, other threads:[~2021-12-02 16:28 UTC | newest]
Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-12-02 16:28 [gcc r12-5727] [Ada] Refactor nested loops in warning on unassigned out parameter Pierre-Marie de Rodat
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).