[LLVMbugs] [Bug 19473] New: std::function fails to compile with non-default-constructible custom allocators
bugzilla-daemon at llvm.org
bugzilla-daemon at llvm.org
Thu Apr 17 18:13:17 PDT 2014
http://llvm.org/bugs/show_bug.cgi?id=19473
Bug ID: 19473
Summary: std::function fails to compile with
non-default-constructible custom allocators
Product: libc++
Version: 3.3
Hardware: Macintosh
OS: MacOS X
Status: NEW
Severity: normal
Priority: P
Component: All Bugs
Assignee: unassignedclangbugs at nondot.org
Reporter: applmak at google.com
CC: llvmbugs at cs.uiuc.edu, mclow.lists at gmail.com
Classification: Unclassified
I have a custom allocator that takes a single param in its constructor:
explicit FunctionAllocator(CircularAllocator& backing_store)
: backing_store_(backing_store) {}
Then a use of std::function:
CircularAllocator backing_store;
FunctionAllocator<std::function<void()>> fa(backing_store);
std::function<void()> f(std::allocator_arg, fa,
[]{ std::cerr << "EXEC: No bound vars." << std::endl; });
f();
But in my xcode using clang/llvm/libc++, the definition of f doesn't compile,
because of this:
...
typedef __function::__func<_Fp, _Alloc, _Rp(_ArgTypes...)> _FF;
if (sizeof(_FF) <= sizeof(__buf_) &&
is_nothrow_copy_constructible<_Fp>::value)
{
__f_ = (__base*)&__buf_;
::new (__f_) _FF(_VSTD::move(__f)); // Here is where the compile
is failing.
}
else
...
This is the fast-path mode of std::function that is intended to avoid a heap
allocation from the lambda's closed-over vars to the std::function's. I love
this assignment for a lot of reasons, but the problem here is that it calls the
move constructor of _FF which doesn't take an allocator instance. Since the
allocator is mentioned in the type of _FF, it attempts to default-construct a
FunctionAllocator, which fails because FunctionAllocator has no default
constructor.
--
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/20140418/c0202314/attachment.html>
More information about the llvm-bugs
mailing list