[LLVMbugs] [Bug 14026] With large tuple size(60) clang much slower than gcc

[bugzilla-daemon at llvm.org]

http://llvm.org/bugs/show_bug.cgi?id=14026

Larry Evans <cppljevans at gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |INVALID

--- Comment #2 from Larry Evans <cppljevans at gmail.com> 2012-10-06 20:07:18 CDT ---
(In reply to comment #1)
> Created attachment 9312 [details]
> test run shown gcc faster than clang no matter TUPLE_SIZE value
> 
> The attachment shows timing for TUPLE_SIZE in 10,20,...,60.
> In each case gcc runs faster.
> 
>   TUPLE_SIZE  clang    gcc  clang/gcc
>   ----------  -----  -----  ---------
>   10           2.68   0.80       3.35
>   30           7.65   2.37       3.22
>   60          16.03   5.74       2.79

OOPS.  The above results are for the Debug+Asserts clang.
When the Release+Asserts clang is used, clang beats gcc,
as shown in the new attachment.

Sorry for noise.

-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.