Re: [PATCH V4] Add warning options -W[no-]compare-distinct-pointer-types

["Jose E. Marchesi" <jose.marchesi@oracle.com>]

Hi Marek.

> On Thu, Aug 17, 2023 at 05:37:03PM +0200, Jose E. Marchesi via Gcc-patches wrote:
>> 
>> > On Thu, 17 Aug 2023, Jose E. Marchesi via Gcc-patches wrote:
>> >
>> >> +@opindex Wcompare-distinct-pointer-types
>> >> +@item -Wcompare-distinct-pointer-types
>> >
>> > This @item should say @r{(C and Objective-C only)}, since the option isn't 
>> > implemented for C++.  OK with that change.
>> 
>> Pushed with that change.
>> Thanks for the prompt review!
>
> I see the following failures:
>
> FAIL: gcc.c-torture/compile/pr106537-1.c   -Os   (test for warnings, line 28)
> FAIL: gcc.c-torture/compile/pr106537-1.c   -Os   (test for warnings, line 30)
> FAIL: gcc.c-torture/compile/pr106537-1.c -O2 -flto
> -fno-use-linker-plugin -flto-partition=none (test for warnings, line
> 28)
> FAIL: gcc.c-torture/compile/pr106537-1.c -O2 -flto
> -fno-use-linker-plugin -flto-partition=none (test for warnings, line
> 30)
> FAIL: gcc.c-torture/compile/pr106537-1.c -O2 -flto -fuse-linker-plugin
> -fno-fat-lto-objects (test for warnings, line 28)
> FAIL: gcc.c-torture/compile/pr106537-1.c -O2 -flto -fuse-linker-plugin
> -fno-fat-lto-objects (test for warnings, line 30)
> FAIL: gcc.c-torture/compile/pr106537-2.c   -O0   (test for warnings, line 26)
> FAIL: gcc.c-torture/compile/pr106537-2.c   -O0   (test for warnings, line 28)
> FAIL: gcc.c-torture/compile/pr106537-2.c   -O1   (test for warnings, line 26)
> FAIL: gcc.c-torture/compile/pr106537-2.c   -O1   (test for warnings, line 28)
> FAIL: gcc.c-torture/compile/pr106537-2.c   -O2   (test for warnings, line 26)
> FAIL: gcc.c-torture/compile/pr106537-2.c   -O2   (test for warnings, line 28)
> FAIL: gcc.c-torture/compile/pr106537-2.c   -O3 -g   (test for warnings, line 26)
> FAIL: gcc.c-torture/compile/pr106537-2.c   -O3 -g   (test for warnings, line 28)
> FAIL: gcc.c-torture/compile/pr106537-2.c   -Os   (test for warnings, line 26)
> FAIL: gcc.c-torture/compile/pr106537-2.c   -Os   (test for warnings, line 28)
> FAIL: gcc.c-torture/compile/pr106537-2.c -O2 -flto
> -fno-use-linker-plugin -flto-partition=none (test for warnings, line
> 26)
> FAIL: gcc.c-torture/compile/pr106537-2.c -O2 -flto
> -fno-use-linker-plugin -flto-partition=none (test for warnings, line
> 28)
> FAIL: gcc.c-torture/compile/pr106537-2.c -O2 -flto -fuse-linker-plugin
> -fno-fat-lto-objects (test for warnings, line 26)
> FAIL: gcc.c-torture/compile/pr106537-2.c -O2 -flto -fuse-linker-plugin
> -fno-fat-lto-objects (test for warnings, line 28)
>
> The problem is that for ==/!=, when one of the types is void*,
> build_binary_op goes to the branch attempting to warn about
> comparing void* with a function pointer, and never gets to the 
> -Wcompare-distinct-pointer-types warning.

Oof I wonder what happened with my regtesting.

I just pushed the patch below as obvious, which adjusts the tests to
conform to GCC's behavior of not emitting that pedwarn for
equality/inequality of void pointers with non-function pointers.

Sorry about this.  And thanks for reporting.

From 721f7e2c4e5eed645593258624dd91e6c39f3bd2 Mon Sep 17 00:00:00 2001
From: "Jose E. Marchesi" <jose.marchesi@oracle.com>
Date: Thu, 24 Aug 2023 17:10:52 +0200
Subject: [PATCH] Fix tests for PR 106537.

This patch fixes the tests for PR 106537 (support for
-W[no]-compare-distinct-pointer-types) which were expecting the
warning when checking for equality/inequality of void pointers with
non-function pointers.

gcc/testsuite/ChangeLog:

	PR c/106537
	* gcc.c-torture/compile/pr106537-1.c: Comparing void pointers to
	non-function pointers is legit.
	* gcc.c-torture/compile/pr106537-2.c: Likewise.
---
 gcc/testsuite/gcc.c-torture/compile/pr106537-1.c | 6 ++++--
 gcc/testsuite/gcc.c-torture/compile/pr106537-2.c | 6 ++++--
 2 files changed, 8 insertions(+), 4 deletions(-)

diff --git a/gcc/testsuite/gcc.c-torture/compile/pr106537-1.c b/gcc/testsuite/gcc.c-torture/compile/pr106537-1.c
index 3f3b06577d5..b67b6090dc3 100644
--- a/gcc/testsuite/gcc.c-torture/compile/pr106537-1.c
+++ b/gcc/testsuite/gcc.c-torture/compile/pr106537-1.c
@@ -25,9 +25,11 @@ int xdp_context (struct xdp_md *xdp)
     return 3;
   if (metadata + 1 <= data) /* { dg-warning "comparison of distinct pointer types" } */
     return 4;
-  if (metadata + 1 == data) /* { dg-warning "comparison of distinct pointer types" } */
+  /* Note that it is ok to check for equality or inequality betewen void
+     pointers and any other non-function pointers.  */
+  if ((int*) (metadata + 1) == (long*) data) /* { dg-warning "comparison of distinct pointer types" } */
     return 5;
-  if (metadata + 1 != data) /* { dg-warning "comparison of distinct pointer types" } */
+  if ((int*) metadata + 1 != (long*) data) /* { dg-warning "comparison of distinct pointer types" } */
     return 5;
 
   return 1;
diff --git a/gcc/testsuite/gcc.c-torture/compile/pr106537-2.c b/gcc/testsuite/gcc.c-torture/compile/pr106537-2.c
index 6876adf3aab..d4223c25c94 100644
--- a/gcc/testsuite/gcc.c-torture/compile/pr106537-2.c
+++ b/gcc/testsuite/gcc.c-torture/compile/pr106537-2.c
@@ -23,9 +23,11 @@ int xdp_context (struct xdp_md *xdp)
     return 3;
   if (metadata + 1 <= data) /* { dg-warning "comparison of distinct pointer types" } */
     return 4;
-  if (metadata + 1 == data) /* { dg-warning "comparison of distinct pointer types" } */
+  /* Note that it is ok to check for equality or inequality betewen void
+     pointers and any other non-function pointers.  */
+  if ((int*) (metadata + 1) == (long*) data) /* { dg-warning "comparison of distinct pointer types" } */
     return 5;
-  if (metadata + 1 != data) /* { dg-warning "comparison of distinct pointer types" } */
+  if ((int*) metadata + 1 != (long*) data) /* { dg-warning "comparison of distinct pointer types" } */
     return 5;
 
   return 1;
-- 
2.30.2