[lld] r227132 - Use parallel_sort in the LayoutPass.

[Michael Spencer]

On Mon, Jan 26, 2015 at 12:29 PM, Shankar Easwaran
<shankare at codeaurora.org> wrote:
> parallel_sort had some issue with building few tests in the LLVM test
> project. Bigcheese will have more information on this.
>
> Shankar Easwaran

I don't recall any issues with this, but I also don't remember why it
wasn't parallel before now.

- Michael Spencer

>
>
> On 1/26/2015 2:18 PM, Rui Ueyama wrote:
>>
>> Author: ruiu
>> Date: Mon Jan 26 14:18:37 2015
>> New Revision: 227132
>>
>> URL: http://llvm.org/viewvc/llvm-project?rev=227132&view=rev
>> Log:
>> Use parallel_sort in the LayoutPass.
>>
>> Time to link lld using lld improved from 5.7s to 5.4s on Windows.
>> It's not a significant improvement but not bad for one-line change.
>>
>> This patch includes a bug fix for Parallel.h as the original code
>> uses operator< instead of a compare function there.
>>
>> Modified:
>>      lld/trunk/include/lld/Core/Parallel.h
>>      lld/trunk/lib/Passes/LayoutPass.cpp
>>
>> Modified: lld/trunk/include/lld/Core/Parallel.h
>> URL:
>> http://llvm.org/viewvc/llvm-project/lld/trunk/include/lld/Core/Parallel.h?rev=227132&r1=227131&r2=227132&view=diff
>>
>> ==============================================================================
>> --- lld/trunk/include/lld/Core/Parallel.h (original)
>> +++ lld/trunk/include/lld/Core/Parallel.h Mon Jan 26 14:18:37 2015
>> @@ -221,8 +221,8 @@ void parallel_quick_sort(RandomAccessIte
>>     auto pivot = medianOf3(start, end, comp);
>>     // Move pivot to end.
>>     std::swap(*(end - 1), *pivot);
>> -  pivot = std::partition(start, end - 1, [end](decltype(*start) v) {
>> -    return v < *(end - 1);
>> +  pivot = std::partition(start, end - 1, [&comp, end](decltype(*start) v)
>> {
>> +    return comp(v, *(end - 1));
>>     });
>>     // Move pivot to middle of partition.
>>     std::swap(*pivot, *(end - 1));
>>
>> Modified: lld/trunk/lib/Passes/LayoutPass.cpp
>> URL:
>> http://llvm.org/viewvc/llvm-project/lld/trunk/lib/Passes/LayoutPass.cpp?rev=227132&r1=227131&r2=227132&view=diff
>>
>> ==============================================================================
>> --- lld/trunk/lib/Passes/LayoutPass.cpp (original)
>> +++ lld/trunk/lib/Passes/LayoutPass.cpp Mon Jan 26 14:18:37 2015
>> @@ -9,6 +9,7 @@
>>     #include "lld/Passes/LayoutPass.h"
>>   #include "lld/Core/Instrumentation.h"
>> +#include "lld/Core/Parallel.h"
>>   #include "llvm/ADT/Twine.h"
>>   #include "llvm/Support/Debug.h"
>>   #include <algorithm>
>> @@ -534,7 +535,7 @@ void LayoutPass::perform(std::unique_ptr
>>     });
>>       std::vector<LayoutPass::SortKey> vec = decorate(atomRange);
>> -  std::sort(vec.begin(), vec.end(),
>> +  parallel_sort(vec.begin(), vec.end(),
>>         [&](const LayoutPass::SortKey &l, const LayoutPass::SortKey &r) ->
>> bool {
>>           return compareAtoms(l, r, _customSorter);
>>         });
>>
>>
>> _______________________________________________
>> llvm-commits mailing list
>> llvm-commits at cs.uiuc.edu
>> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
>>
>>
>
>
> --
> Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, hosted by
> the Linux Foundation
>
>
> _______________________________________________
> llvm-commits mailing list
> llvm-commits at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits