[issue31834] BLAKE2: the (pure) SSE2 impl forced on x86_64 is slower than reference

2017-11-19 Thread Christian Heimes
Change by Christian Heimes : -- resolution: -> fixed stage: -> resolved status: open -> closed ___ Python tracker ___

[issue31834] BLAKE2: the (pure) SSE2 impl forced on x86_64 is slower than reference

2017-11-07 Thread Benjamin Peterson
Benjamin Peterson added the comment: @tiran, can we close this again? -- ___ Python tracker ___

[issue31834] BLAKE2: the (pure) SSE2 impl forced on x86_64 is slower than reference

2017-10-24 Thread Benjamin Peterson
Benjamin Peterson added the comment: On Tue, Oct 24, 2017, at 00:25, Christian Heimes wrote: > > Christian Heimes added the comment: > > I'm pretty sure that your PR has disabled all SSE optimizations. AFAIK > gcc does not enable SSE3 and SSE4 on X86_64

[issue31834] BLAKE2: the (pure) SSE2 impl forced on x86_64 is slower than reference

2017-10-24 Thread STINNER Victor
STINNER Victor added the comment: > AFAIK gcc does not enable SSE3 and SSE4 on X86_64 by default. Linux now supports multiple variants of the same function, one variant per CPU type, the binding is done when a library is loaded. But I don't know how to implement

[issue31834] BLAKE2: the (pure) SSE2 impl forced on x86_64 is slower than reference

2017-10-24 Thread Christian Heimes
Christian Heimes added the comment: I'm pretty sure that your PR has disabled all SSE optimizations. AFAIK gcc does not enable SSE3 and SSE4 on X86_64 by default. $ gcc -dM -E - < /dev/null | grep SSE #define __SSE2_MATH__ 1 #define __SSE_MATH__ 1 #define __SSE2__ 1 #define

[issue31834] BLAKE2: the (pure) SSE2 impl forced on x86_64 is slower than reference

2017-10-24 Thread Benjamin Peterson
Benjamin Peterson added the comment: New changeset 1aa00ff383c43335e4a5044274617dbf59bc839e by Benjamin Peterson (Michał Górny) in branch 'master': fixes bpo-31834: Use optimized code for BLAKE2 only with SSSE3+ (#4066)

[issue31834] BLAKE2: the (pure) SSE2 impl forced on x86_64 is slower than reference

2017-10-21 Thread Michał Górny
Change by Michał Górny : -- keywords: +patch pull_requests: +4036 stage: -> patch review ___ Python tracker ___

[issue31834] BLAKE2: the (pure) SSE2 impl forced on x86_64 is slower than reference

2017-10-21 Thread Michał Górny
New submission from Michał Górny : The setup.py file for Python states: if (not cross_compiling and os.uname().machine == "x86_64" and sys.maxsize > 2**32): # Every x86_64 machine has at least SSE2. Check for sys.maxsize