[PATCH] use the right conversion warning option (PR c/80892)

[PATCH] use the right conversion warning option (PR c/80892)

[Martin Sebor]

[-- Attachment #1: Type: text/plain, Size: 189 bytes --]

The conversion enhancements I committed in r248431 introduced
an unintended change in which warning option is used to issue
certain integer conversion warnings.  Attached is a fix.

Martin

[-- Attachment #2: gcc-80892.diff --]
[-- Type: text/x-patch, Size: 1897 bytes --]

PR c/80892 - -Wfloat-conversion now warns about non-floats

gcc/c-family/ChangeLog:

	PR c/80892
	* c-warn.c (conversion_warning): Use -Wconversion for integer
	conversion and -Wfloat-conversion for floating one.

gcc/testsuite/ChangeLog:

	PR c/80892
	* c-c++-common/Wfloat-conversion-2.c: New test.

diff --git a/gcc/c-family/c-warn.c b/gcc/c-family/c-warn.c
index 012675b..35321a6 100644
--- a/gcc/c-family/c-warn.c
+++ b/gcc/c-family/c-warn.c
@@ -1043,10 +1043,19 @@ conversion_warning (location_t loc, tree type, tree expr, tree result)
 		    "conversion from %qT to to %qT discards imaginary "
 		    "component",
 		    expr_type, type);
-      else if (conversion_kind == UNSAFE_REAL || conversion_kind)
-	warning_at (loc, OPT_Wfloat_conversion,
-		    "conversion from %qT to %qT may change value",
-		    expr_type, type);
+      else
+	{
+	  int warnopt;
+	  if (conversion_kind == UNSAFE_REAL)
+	    warnopt = OPT_Wfloat_conversion;
+	  else if (conversion_kind)
+	    warnopt = OPT_Wconversion;
+	  else
+	    break;
+	  warning_at (loc, warnopt,
+		      "conversion from %qT to %qT may change value",
+		      expr_type, type);
+	}
     }
 }
 
diff --git a/gcc/testsuite/c-c++-common/Wfloat-conversion-2.c b/gcc/testsuite/c-c++-common/Wfloat-conversion-2.c
new file mode 100644
index 0000000..626e35b
--- /dev/null
+++ b/gcc/testsuite/c-c++-common/Wfloat-conversion-2.c
@@ -0,0 +1,19 @@
+/* PR c/80892 - [8 regression] -Wfloat-conversion now warns about non-floats
+   { dg-do compile }
+   { dg-options "-Wconversion -Wfloat-conversion" }
+   { dg-require-effective-target int32plus }
+   { dg-require-effective-target double64plus } */
+
+void fschar (char);
+
+void fulong (unsigned long x)
+{
+  fschar (x);   /* { dg-warning "\\[-Wconversion\\]" } */
+}
+
+void ffloat (float);
+
+void fdouble (double x)
+{
+  ffloat (x);   /* { dg-warning "\\[-Wfloat-conversion\\]" } */
+}

Re: [PATCH] use the right conversion warning option (PR c/80892)

[Eric Gallager]

I tested this patch; it fixes the warnings that caused me to open the
bug in the first place. Thank you!

Eric


On 5/30/17, Martin Sebor <msebor@gmail.com> wrote:
> The conversion enhancements I committed in r248431 introduced
> an unintended change in which warning option is used to issue
> certain integer conversion warnings.  Attached is a fix.
>
> Martin
>

Re: [PATCH] use the right conversion warning option (PR c/80892)

[Martin Sebor]

On 06/01/2017 06:38 PM, Eric Gallager wrote:
> I tested this patch; it fixes the warnings that caused me to open the
> bug in the first place. Thank you!

Thanks for validating it.  I'll go ahead and commit the fix
as obvious if there are no objections in the next few hours.

Martin

>
> Eric
>
>
> On 5/30/17, Martin Sebor <msebor@gmail.com> wrote:
>> The conversion enhancements I committed in r248431 introduced
>> an unintended change in which warning option is used to issue
>> certain integer conversion warnings.  Attached is a fix.
>>
>> Martin
>>

[committed] use the right conversion warning option (PR c/80892)

[Martin Sebor]

I have committed the patch in r248852.

Martin

On 06/02/2017 08:33 AM, Martin Sebor wrote:
> On 06/01/2017 06:38 PM, Eric Gallager wrote:
>> I tested this patch; it fixes the warnings that caused me to open the
>> bug in the first place. Thank you!
>
> Thanks for validating it.  I'll go ahead and commit the fix
> as obvious if there are no objections in the next few hours.
>
> Martin
>
>>
>> Eric
>>
>>
>> On 5/30/17, Martin Sebor <msebor@gmail.com> wrote:
>>> The conversion enhancements I committed in r248431 introduced
>>> an unintended change in which warning option is used to issue
>>> certain integer conversion warnings.  Attached is a fix.
>>>
>>> Martin
>>>
>