llvm dev - Sep 2018 - New warnings when building trunk with GCC 9

On 13/09/2018 18:22, David Blaikie via llvm-dev wrote:
> On Thu, Sep 13, 2018 at 12:13 AM Dávid Bolvanský via llvm-dev 
> <llvm-dev at lists.llvm.org <mailto:llvm-dev at
lists.llvm.org>> wrote:
>    
/home/davidbolvansky/trunk/llvm/unittests/ExecutionEngine/Orc/CompileOnDemandLayerTest.cpp:79:40:
>       required from here
>    
/home/davidbolvansky/trunk/llvm/include/llvm/ExecutionEngine/Orc/CompileOnDemandLayer.h:314:29:
>     warning: redundant move in return statement [-Wredundant-move]
>     314 |         return std::move(Err);
Note that the move (into the implicit JITSymbol(Error) ctor) is only 
redundant if the compiler implements a fix for 
<http://www.open-std.org/jtc1/sc22/wg21/docs/cwg_defects.html#1579> 
"Return by converting move constructor".  (But not sure whether the
LLVM
compiler baselines imply that, anyway.  In LibreOffice it forced me to 
introduce ugly #ifs, to not have to disable that warning outright, 
<https://cgit.freedesktop.org/libreoffice/core/commit/?id=dc06c8f4989fc28d0c31ebd333e53dfe0e0f5f66>
"-Werror=redundant-move (GCC 9), take two".)

On Fri, Sep 14, 2018 at 12:48 AM Stephan Bergmann <sbergman at redhat.com>
wrote:
> On 13/09/2018 18:22, David Blaikie via llvm-dev wrote:
> > On Thu, Sep 13, 2018 at 12:13 AM Dávid Bolvanský via llvm-dev
> > <llvm-dev at lists.llvm.org <mailto:llvm-dev at
lists.llvm.org>> wrote:
> >
> 
/home/davidbolvansky/trunk/llvm/unittests/ExecutionEngine/Orc/CompileOnDemandLayerTest.cpp:79:40:
>
> >       required from here
> >
> 
/home/davidbolvansky/trunk/llvm/include/llvm/ExecutionEngine/Orc/CompileOnDemandLayer.h:314:29:
> >     warning: redundant move in return statement [-Wredundant-move]
> >     314 |         return std::move(Err);
>
> Note that the move (into the implicit JITSymbol(Error) ctor) is only
> redundant if the compiler implements a fix for
> <http://www.open-std.org/jtc1/sc22/wg21/docs/cwg_defects.html#1579>
> "Return by converting move constructor".  (But not sure whether
the LLVM
> compiler baselines imply that, anyway.

Looks like that was accepted for C++14, not 11. So I don't believe it's
valid to remove the std::move in C++11 code. (& I believe Clang's
warning
correctly diagnoses this only in the right language versions) - so we
probably want to disable the buggy GCC warning here, then.

> In LibreOffice it forced me to
> introduce ugly #ifs, to not have to disable that warning outright,
> <
>
https://cgit.freedesktop.org/libreoffice/core/commit/?id=dc06c8f4989fc28d0c31ebd333e53dfe0e0f5f66>
>
> "-Werror=redundant-move (GCC 9), take two".)
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20180914/cf991a8c/attachment.html>

There is a new discussion related to -Wredundant-move warning on GCC
bugzilla.

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=87300

pi 14. 9. 2018 o 9:53 David Blaikie <dblaikie at gmail.com> napísal(a):
>
>
> On Fri, Sep 14, 2018 at 12:48 AM Stephan Bergmann <sbergman at
redhat.com>
> wrote:
>
>> On 13/09/2018 18:22, David Blaikie via llvm-dev wrote:
>> > On Thu, Sep 13, 2018 at 12:13 AM Dávid Bolvanský via llvm-dev
>> > <llvm-dev at lists.llvm.org <mailto:llvm-dev at
lists.llvm.org>> wrote:
>> >
>> 
/home/davidbolvansky/trunk/llvm/unittests/ExecutionEngine/Orc/CompileOnDemandLayerTest.cpp:79:40:
>>
>> >       required from here
>> >
>> 
/home/davidbolvansky/trunk/llvm/include/llvm/ExecutionEngine/Orc/CompileOnDemandLayer.h:314:29:
>> >     warning: redundant move in return statement [-Wredundant-move]
>> >     314 |         return std::move(Err);
>>
>> Note that the move (into the implicit JITSymbol(Error) ctor) is only
>> redundant if the compiler implements a fix for
>>
<http://www.open-std.org/jtc1/sc22/wg21/docs/cwg_defects.html#1579>
>> "Return by converting move constructor".  (But not sure
whether the LLVM
>> compiler baselines imply that, anyway.
>
>
> Looks like that was accepted for C++14, not 11. So I don't believe
it's
> valid to remove the std::move in C++11 code. (& I believe Clang's
warning
> correctly diagnoses this only in the right language versions) - so we
> probably want to disable the buggy GCC warning here, then.
>
>
>> In LibreOffice it forced me to
>> introduce ugly #ifs, to not have to disable that warning outright,
>> <
>>
https://cgit.freedesktop.org/libreoffice/core/commit/?id=dc06c8f4989fc28d0c31ebd333e53dfe0e0f5f66>
>>
>> "-Werror=redundant-move (GCC 9), take two".)
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20180915/95cffd4c/attachment.html>