[llvm-bugs] [Bug 24845] New: C++ says shift-overflow should produce 0

[via llvm-bugs]

https://llvm.org/bugs/show_bug.cgi?id=24845

            Bug ID: 24845
           Summary: C++ says shift-overflow should produce 0
           Product: clang
           Version: 3.4
          Hardware: PC
                OS: FreeBSD
            Status: NEW
          Severity: normal
          Priority: P
         Component: C++
          Assignee: unassignedclangbugs at nondot.org
          Reporter: huntting at glarp.com
                CC: dgregor at apple.com, llvm-bugs at lists.llvm.org
    Classification: Unclassified

Created attachment 14887
  --> https://llvm.org/bugs/attachment.cgi?id=14887&action=edit
example.cc prints expressions with shift-count-overflow

This may be new to C++14, but section 5.8 of draft standard
(http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2014/n4296.pdf) allows for
shifting unsigned integers by more than their size.

Hence at least in C++14, with 64bit long longs:
    assert(~0ULL << 64 == 0)


Note that the C11 standard by contrast allows for "undefined behavior" in this
case.

I have not check previous draft versions of the C++ standard or the actual
standard to see if this clause has recently changed.

The attached example.cc produces this output on my x86_64 workstation:

                  sizeof(0ULL) = 8
                         ~0ULL = 0xffffffffffffffff
     ~0ULL << (8*sizeof(0ULL)) = 0x1
  ~(~0ULL << (8*sizeof(0ULL))) = 0x1
                    sizeof(0U) = 4
                           ~0U = 0xffffffff
         ~0U << (8*sizeof(0U)) = 0x1
      ~(~0U << (8*sizeof(0U))) = 0x1

-- 
You are receiving this mail because:
You are on the CC list for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-bugs/attachments/20150916/b0357460/attachment.html>