PING: [PATCH] X86-64: Remove the previous SSE2/AVX2 memsets

[H.J. Lu]

On Thu, Apr 14, 2016 at 5:03 AM, H.J. Lu <hjl.tools@gmail.com> wrote:
> On Tue, Apr 12, 2016 at 8:34 AM, H.J. Lu <hjl.tools@gmail.com> wrote:
>> On Thu, Mar 31, 2016 at 12:42 PM, H.J. Lu <hongjiu.lu@intel.com> wrote:
>>> Since the new SSE2/AVX2 memsets are faster than the previous ones, we
>>> can remove the previous SSE2/AVX2 memsets and replace them with the
>>> new ones.
>>>
>>> No change in IFUNC selection if SSE2 and AVX2 memsets weren't used
>>> before.  If SSE2 or AVX2 memset was used, the new SSE2 or AVX2 memset
>>> optimized with Enhanced REP STOSB will be used for processors with
>>> ERMS.
>>>
>>> Tested on Penryn, Westmere, Ivy Bridge and Haswell with and without
>>> --disable-multi-arch.  OK for master?
>>>
>>> H.J.
>>> ---
>>>         [BZ #19881]
>>>         * sysdeps/x86_64/multiarch/memset-sse2-unaligned-erms.S: Folded
>>>         into ...
>>>         * sysdeps/x86_64/memset.S: This.
>>>         (__bzero): Removed.
>>>         (__memset_tail): Likewise.
>>>         (__memset_chk): Likewise.
>>>         (memset): Likewise.
>>>         (MEMSET_CHK_SYMBOL): New. Define only if MEMSET_SYMBOL isn't
>>>         defined.
>>>         (MEMSET_SYMBOL): Define only if MEMSET_SYMBOL isn't defined.
>>>         * sysdeps/x86_64/multiarch/memset-avx2.S: Removed.
>>>         (__memset_zero_constant_len_parameter): Check SHARED instead of
>>>         PIC.
>>>         * sysdeps/x86_64/multiarch/Makefile (sysdep_routines): Remove
>>>         memset-avx2 and memset-sse2-unaligned-erms.
>>>         * sysdeps/x86_64/multiarch/ifunc-impl-list.c
>>>         (__libc_ifunc_impl_list): Remove __memset_chk_sse2,
>>>         __memset_chk_avx2, __memset_sse2 and __memset_avx2_unaligned.
>>>         * sysdeps/x86_64/multiarch/memset-avx2-unaligned-erms.S: Skip
>>>         if not in libc.
>>>         * sysdeps/x86_64/multiarch/memset-avx512-unaligned-erms.S:
>>>         Likewise.
>>>         * sysdeps/x86_64/multiarch/memset-vec-unaligned-erms.S
>>>         (MEMSET_CHK_SYMBOL): New.  Define if not defined.
>>>         (__bzero): Check VEC_SIZE == 16 instead of USE_MULTIARCH.
>>>         Replace MEMSET_SYMBOL with MEMSET_CHK_SYMBOL on __memset_chk
>>>         symbols.
>>>         Properly check USE_MULTIARCH on __memset symbols.
>>>         * sysdeps/x86_64/multiarch/memset.S (memset): Replace
>>>         __memset_sse2 and __memset_avx2 with __memset_sse2_unaligned
>>>         and __memset_avx2_unaligned.  Use __memset_sse2_unaligned_erms
>>>         or __memset_avx2_unaligned_erms if processor has ERMS.
>>>         (memset): Removed.
>>>         (__memset_chk): Likewise.
>>>         (MEMSET_SYMBOL): New.
>>>         (libc_hidden_builtin_def): Replace __memset_sse2 with
>>>         __memset_sse2_unaligned.
>>>         * sysdeps/x86_64/multiarch/memset_chk.S (__memset_chk): Replace
>>>         __memset_chk_sse2 and __memset_chk_avx2 with
>>>         __memset_chk_sse2_unaligned and __memset_chk_avx2_unaligned_erms.
>>>         Use __memset_chk_sse2_unaligned_erms or
>>>         __memset_chk_avx2_unaligned_erms if processor has ERMS.
>>> ---
>>>  sysdeps/x86_64/memset.S                            | 121 +++------------
>>>  sysdeps/x86_64/multiarch/Makefile                  |   3 +-
>>>  sysdeps/x86_64/multiarch/ifunc-impl-list.c         |   9 --
>>>  .../x86_64/multiarch/memset-avx2-unaligned-erms.S  |  18 ++-
>>>  sysdeps/x86_64/multiarch/memset-avx2.S             | 168 ---------------------
>>>  .../multiarch/memset-avx512-unaligned-erms.S       |   2 +-
>>>  .../x86_64/multiarch/memset-sse2-unaligned-erms.S  |  16 --
>>>  .../x86_64/multiarch/memset-vec-unaligned-erms.S   |  32 ++--
>>>  sysdeps/x86_64/multiarch/memset.S                  |  26 ++--
>>>  sysdeps/x86_64/multiarch/memset_chk.S              |  14 +-
>>>  10 files changed, 76 insertions(+), 333 deletions(-)
>>>  delete mode 100644 sysdeps/x86_64/multiarch/memset-avx2.S
>>>  delete mode 100644 sysdeps/x86_64/multiarch/memset-sse2-unaligned-erms.S
>>>
>>
>> PING.
>>
>
> Any comments, feedbacks?  Objections?
>

The performance data shows the new one is better than the
old one on various Intel and AMD machines:

https://sourceware.org/bugzilla/show_bug.cgi?id=19881

I am enclosing the patch and I will check it in shortly.


-- 
H.J.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0001-X86-64-Remove-the-previous-SSE2-AVX2-memsets.patch
Type: text/x-patch
Size: 17020 bytes
Desc: not available
URL: <http://sourceware.org/pipermail/libc-alpha/attachments/20160608/ddff335b/attachment.bin>