[Bug c++/113565] New: __builtin_clz(0) is undefined behavior, but not detected in constant expressions

[Bug c++/113565] New: __builtin_clz(0) is undefined behavior, but not detected in constant expressions

[janschultke at googlemail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113565

            Bug ID: 113565
           Summary: __builtin_clz(0) is undefined behavior, but not
                    detected in constant expressions
           Product: gcc
           Version: 14.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c++
          Assignee: unassigned at gcc dot gnu.org
          Reporter: janschultke at googlemail dot com
  Target Milestone: ---

Code to reproduce
=================

static_assert(__builtin_clz(0));


Actual output (https://godbolt.org/z/v6v5nGxv8)
===============================================

None.


Expected output (Clang)
=======================

<source>:1:15: error: static assertion expression is not an integral constant
expression
    1 | static_assert(__builtin_clz(0));
      |               ^~~~~~~~~~~~~~~~

[Bug c++/113565] __builtin_clz(0) is undefined behavior, but not detected in constant expressions

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113565

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Last reconfirmed|                            |2024-01-23
             Status|UNCONFIRMED                 |NEW
     Ever confirmed|0                           |1

--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
This should fix it (but I am have not tested it either and I am not 100% sure
we want/need this):

```
[apinski@xeond2 gcc]$ git diff fold-const-call.cc
diff --git a/gcc/fold-const-call.cc b/gcc/fold-const-call.cc
index 47bf8d64391..c9fa51e35ee 100644
--- a/gcc/fold-const-call.cc
+++ b/gcc/fold-const-call.cc
@@ -1031,7 +1031,7 @@ fold_const_call_ss (wide_int *result, combined_fn fn,
const wide_int_ref &arg,
          tmp = TYPE_PRECISION (arg_type);
        else if (!CLZ_DEFINED_VALUE_AT_ZERO (SCALAR_INT_TYPE_MODE (arg_type),
                                             tmp))
-         tmp = TYPE_PRECISION (arg_type);
+         return false;
        *result = wi::shwi (tmp, precision);
        return true;
       }
@@ -1046,7 +1046,7 @@ fold_const_call_ss (wide_int *result, combined_fn fn,
const wide_int_ref &arg,
          tmp = TYPE_PRECISION (arg_type);
        else if (!CTZ_DEFINED_VALUE_AT_ZERO (SCALAR_INT_TYPE_MODE (arg_type),
                                             tmp))
-         tmp = TYPE_PRECISION (arg_type);
+         return false;
        *result = wi::shwi (tmp, precision);
        return true;
       }

```

Confirmed.

[Bug c++/113565] __builtin_clz(0) is undefined behavior, but not detected in constant expressions

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113565

--- Comment #2 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Note starting in GCC 14, it is better to use __builtin_clzg with the 2
arguments anyways so you can control exactly what value you want/need for 0.

[Bug c++/113565] __builtin_clz(0) is undefined behavior, but not detected in constant expressions

[redi at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113565

Jonathan Wakely <redi at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Severity|normal                      |enhancement

--- Comment #3 from Jonathan Wakely <redi at gcc dot gnu.org> ---
I'm not sure it's reasonable to expect an error here. The C++ standard says
nothing about whether __builtin_clz is a constant expression, obviously.

So I'm changing this to severity=enhancement.

[Bug c++/113565] __builtin_clz(0) is undefined behavior, but not detected in constant expressions

[janschultke at googlemail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113565

--- Comment #4 from Jan Schultke <janschultke at googlemail dot com> ---
I would expect an error here because things that are undefined behavior are
generally supposed to fail in constant expressions. I don't see a good reason
why builtins should be exempt from that rule.

The lack of diagnostic has cost me a few minutes of debugging yesterday. I had
a static_assert:

> static_assert(my_function(0u) == 32);

This succeeded at compile time, but my_function(0) would return 0 at run-time
as a result of passing through to __builtin_clz. UBSan may have caught it, but
regardless, it's not sane to have different results inside/outside constant
expressions like that.

[Bug c++/113565] __builtin_clz(0) is undefined behavior, but not detected in constant expressions

[janschultke at googlemail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113565

--- Comment #5 from Jan Schultke <janschultke at googlemail dot com> ---
You can reproduce this as follows:

> static_assert(__builtin_clz(0u) == 32);
> 
> unsigned x = 0;
> 
> int main() {
>     return __builtin_clz(x);
> }


For base x86_64, GCC emits: (https://godbolt.org/z/nqzYrTWd1)

> main:
>         bsr     eax, DWORD PTR x[rip]
>         xor     eax, 31
>         ret
> x:
>         .zero   4

Even though __builtin_clz(0u) == 32 passes, this program returns 63. This is
obviously not in the interest of the developer, regardless of what the standard
mandates.