svn commit: r289072 - in head/contrib: libc++/src llvm/lib/Transforms/Vectorize

[Ed Maste]

On 10 October 2015 at 06:08, Konstantin Belousov <kostikbel at gmail.com> wrote:
> On Fri, Oct 09, 2015 at 06:21:45PM +0000, Dimitry Andric wrote:
>> Author: dim
>> Date: Fri Oct  9 18:21:45 2015
>> New Revision: 289072
>> URL: https://svnweb.freebsd.org/changeset/base/289072
>>
>> Log:
>>   Temporarily revert upstream llvm trunk r240144 (by Michael Zolotukhin):
>
> This might be indeed a clang bug, but probably not in the revision which
> you reverted.  Or it might be a libc++ bug.

Yes, the bug is almost certainly not r240144; reverting it is a
short-term workaround until the underlying issue is fixed.

> The story, from what I understand from the discussion at some other
> place, is that the fault occurs on access to the common unaligned
> symbol. Apparently, ELF does allow to specify alignment of the common
> symbols, and static linker, when finally allocating .bss space for the
> object, must obey the requirement. The symbol value for the common
> symbol must be interpreted as the desired alignment.
>
> Could you look up the symbol reference in the .o files (again) and see
> which alignment is requested ?  I think that it is probably clang which
> should set it to be at least 16 bytes to generate ABI-compliant code.

The declaration is extern ostream cout; which correctly has 8 byte alignment.

In the implementation (contrib/libc++/src/iostream.cpp) it is defined
as a char array which is where the 16-byte ABI alignment would arise,
but explicitly specifies 8-byte alignment:
_ALIGNAS_TYPE (ostream)  _LIBCPP_FUNC_VIS char cout[sizeof(ostream)];

Compiling with -emit-llvm --save-temps shows the desired 8-byte
alignment for cout in the LLVM IR but 16-byte in the assembly file.