<div dir="ltr"><br><div class="gmail_extra"><br><div class="gmail_quote">On Tue, Jan 27, 2015 at 9:29 AM, Hans Wennborg <span dir="ltr"><<a href="mailto:hans@chromium.org" target="_blank">hans@chromium.org</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="HOEnZb"><div class="h5">On Thu, Jan 22, 2015 at 4:55 PM, Hans Wennborg <<a href="mailto:hans@chromium.org">hans@chromium.org</a>> wrote:<br>
> On Thu, Jan 22, 2015 at 2:43 PM, David Blaikie <<a href="mailto:dblaikie@gmail.com">dblaikie@gmail.com</a>> wrote:<br>
>><br>
>><br>
>> On Thu, Jan 22, 2015 at 2:11 PM, Hans Wennborg <<a href="mailto:hans@hanshq.net">hans@hanshq.net</a>> wrote:<br>
>>><br>
>>> Author: hans<br>
>>> Date: Thu Jan 22 16:11:56 2015<br>
>>> New Revision: 226870<br>
>>><br>
>>> URL: <a href="http://llvm.org/viewvc/llvm-project?rev=226870&view=rev" target="_blank">http://llvm.org/viewvc/llvm-project?rev=226870&view=rev</a><br>
>>> Log:<br>
>>> Make the ?: precedence warning handle pointers to the left of ?<br>
>>><br>
>>> Previously, Clang would fail to warn on:<br>
>>><br>
>>> int n = x + foo ? 1 : 2;<br>
>>><br>
>>> when foo is a pointer.<br>
>>><br>
>>> Modified:<br>
>>> cfe/trunk/lib/Sema/SemaExpr.cpp<br>
>>> cfe/trunk/test/Sema/parentheses.c<br>
>>><br>
>>> Modified: cfe/trunk/lib/Sema/SemaExpr.cpp<br>
>>> URL:<br>
>>> <a href="http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/Sema/SemaExpr.cpp?rev=226870&r1=226869&r2=226870&view=diff" target="_blank">http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/Sema/SemaExpr.cpp?rev=226870&r1=226869&r2=226870&view=diff</a><br>
>>><br>
>>> ==============================================================================<br>
>>> --- cfe/trunk/lib/Sema/SemaExpr.cpp (original)<br>
>>> +++ cfe/trunk/lib/Sema/SemaExpr.cpp Thu Jan 22 16:11:56 2015<br>
>>> @@ -6129,6 +6129,8 @@ static bool ExprLooksBoolean(Expr *E) {<br>
>>> return IsLogicOp(OP->getOpcode());<br>
>>> if (UnaryOperator *OP = dyn_cast<UnaryOperator>(E))<br>
>>> return OP->getOpcode() == UO_LNot;<br>
>>> + if (E->getType()->isPointerType())<br>
>><br>
>><br>
>> Could we generalize this a bit further, somehow? (I haven't looked at the<br>
>> code in question, but it sounds like this should use some more general tool<br>
>> of "try to apply contextual conversion to bool" so that it matches the<br>
>> actual semantic situation we're interested in here)<br>
><br>
> It's tricky, because we don't really want to match the actual<br>
> semantics, we want to figure out if the intention was to use 'foo' as<br>
> a conditional expression. That's what 'ExprLooksBoolean' does, and<br>
> it's erring on the side of caution.<br>
><br>
> For example, we don't want to warn if 'foo' is int, even if that could<br>
> be used as a conditional expression. But we do want to warn if 'foo'<br>
> is 'a==b', even in C where the type of that expression is int.<br>
><br>
> Having said that, I now realised I might have made the warning a<br>
> little too broad.. we want to warn here:<br>
><br>
> int x = x + ptr ? 1 : 0;<br>
><br>
> because 'x + ptr' seems like a pretty unlikely condition expression.<br>
><br>
> But we don't want to warn here:<br>
><br>
> int y = ptr1 - ptr2 ? 1 : 0;<br>
><br>
> I'll think about that.<br>
<br>
</div></div>The last example actually showed up in code yesterday:<br>
<a href="http://git.savannah.gnu.org/cgit/emacs.git/tree/src/emacs.c?id=emacs-24.4#n2307" target="_blank">http://git.savannah.gnu.org/cgit/emacs.git/tree/src/emacs.c?id=emacs-24.4#n2307</a><br>
<br>
The reason I said I didn't want to warn here is because the expression<br>
can only be interpreted one way. For example,<br>
<br>
int y = ptr1 - (ptr2 ? 1 : 0);<br>
<br>
would not compile because the type has changed.<br>
<br>
On the other hand, parentheses would certainly make it more readable,<br>
and that is what the warning suggests:<br>
<br>
int y = (ptr1 - ptr2) ? 1 : 0;<br>
<br>
So I'm now thinking maybe warning here is the right thing, just as<br>
-Wparentheses warns about 'a && b || c'.<br></blockquote><div><br>Perhaps - though &&|| is trickier because it could easily be either way. As you point out with ?: it can't actually be the other way sometimes. So perhaps we could/should have the smarts to detect that case, but I'm not sure where the effort tipping point is.<br> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<span class="HOEnZb"><font color="#888888"><br>
- Hans<br>
</font></span></blockquote></div><br></div></div>