OK. Sorry for miss-reading the message.
In that case, linking in libatomic becomes a separate issue. We don't
need to touch gcc.c in this patch.

Thanks,

-Rong

On Wed, Nov 20, 2013 at 2:19 PM, Andrew Pinski <pins...@gmail.com> wrote:
> On Wed, Nov 20, 2013 at 2:07 PM, Rong Xu <x...@google.com> wrote:
>> Joseph and Andrew, thanks for the suggestion. That's really helpful.
>>
>> Here is the new patch for gcc.c.
>> Basically, it's just what you have suggested: enclosing -latomic with
>> --as-needed, and using macros.
>> For the case of no --as-needed support, I use static link. (just found
>> that some code already using this in the SPEC).
>> I'm flexible on this part -- if you think this is unnecessary, I can remove.
>
>
> I think Joseph's suggestion was also to include -latomic even when not
> generating atomic profiling due to the C11 code requiring it.
>
> Thanks,
> Andrew
>
>>
>> Thanks,
>>
>> -Rong
>>
>> Index: gcc.c
>> ===================================================================
>> --- gcc.c       (revision 205053)
>> +++ gcc.c       (working copy)
>> @@ -748,6 +748,23 @@ proper position among the other output files.  */
>>      %{fvtable-verify=preinit: -lvtv -u_vtable_map_vars_start
>> -u_vtable_map_vars_end}}"
>>  #endif
>>
>> +/* This spec is for linking in libatomic in gcov atomic counter update.
>> +   We will use the atomic functions defined in libatomic, only when the 
>> builtin
>> +   versions are not available. In the case of no LD_AS_NEEDED support, we
>> +   link libatomic statically.  */
>> +
>> +#ifndef GCOV_ATOMIC_SPEC
>> +#if USE_LD_AS_NEEDED
>> +#define GCOV_ATOMIC_SPEC "%{fprofile-generate-atomic=*:" 
>> LD_AS_NEEDED_OPTION \
>> +  " -latomic} " LD_NO_AS_NEEDED_OPTION
>> +#elif defined(HAVE_LD_STATIC_DYNAMIC)
>> +#define GCOV_ATOMIC_SPEC "%{fprofile-generate-atomic=*:" LD_STATIC_OPTION \
>> +                    " -latomic " LD_DYNAMIC_OPTION "}"
>> +#else /* !USE_LD_AS_NEEDED && !HAVE_LD_STATIC_DYNAMIC  */
>> +#define GCOV_ATOMIC_SPEC "%{fprofile-generate-atomic=*:-latomic}"
>> +#endif
>> +#endif
>> +
>>  /* -u* was put back because both BSD and SysV seem to support it.  */
>>  /* %{static:} simply prevents an error message if the target machine
>>     doesn't handle -static.  */
>> @@ -771,7 +788,8 @@ proper position among the other output files.  */
>>      
>> %{fopenmp|ftree-parallelize-loops=*:%:include(libgomp.spec)%(link_gomp)}\
>>      %{fgnu-tm:%:include(libitm.spec)%(link_itm)}\
>>      %(mflib) " STACK_SPLIT_SPEC "\
>> -    %{fprofile-arcs|fprofile-generate*|coverage:-lgcov} " SANITIZER_SPEC " \
>> +    %{fprofile-arcs|fprofile-generate*|coverage:-lgcov\
>> +      " GCOV_ATOMIC_SPEC "} " SANITIZER_SPEC " \
>>      %{!nostdlib:%{!nodefaultlibs:%(link_ssp) %(link_gcc_c_sequence)}}\
>>      %{!nostdlib:%{!nostartfiles:%E}} %{T*} }}}}}}"
>>
>>
>>
>> On Wed, Nov 20, 2013 at 1:04 PM, Joseph S. Myers
>> <jos...@codesourcery.com> wrote:
>>> On Wed, 20 Nov 2013, Rong Xu wrote:
>>>
>>>> I could do this in the SPEC
>>>>   -Wl,-Bstatic -latomic -Wl,-Bdynamic
>>>> which would link libatomic statically.
>>>> I works for me. But it looks a little weird in gcc driver.
>>>
>>> I think we should generally link libatomic with --as-needed by default on
>>> platforms supporting --as-needed, in line with the general principle that
>>> C code just using language not library facilities (_Atomic in this case)
>>> shouldn't need any special options to link it (libatomic is like libgcc,
>>> which is linked in automatically); the trickier question is what to do
>>> with it on any systems supporting shared libraries but not --as-needed.
>>>
>>> --
>>> Joseph S. Myers
>>> jos...@codesourcery.com

Reply via email to