public inbox for glibc-cvs@sourceware.org help / color / mirror / Atom feed
From: H.J. Lu <hjl@sourceware.org> To: glibc-cvs@sourceware.org Subject: [glibc] Deleted branch hjl/x86/optimize Date: Wed, 22 Apr 2020 17:40:42 +0000 (GMT) [thread overview] Message-ID: <20200422174042.88B4C3858D33@sourceware.org> (raw) The branch 'hjl/x86/optimize' was deleted. It previously pointed to: 3e31bc4a93... Add x86_cache.non_temporal_threshold to GLIBC_TUNABLES Diff: !!! WARNING: THE FOLLOWING COMMITS ARE NO LONGER ACCESSIBLE (LOST): ------------------------------------------------------------------- 3e31bc4... Add x86_cache.non_temporal_threshold to GLIBC_TUNABLES bfb716e... x86: Update __x86_shared_non_temporal_threshold 3b4afb9... Test only a subset of memcpy f47a8ff... Integrate memcpy_benchmark.cc with glibc benchtests c91f98a... Build memcpy_benchmark in benchtests 84158fd... Import memcpy_benchmark.cc 006859c... x86-64: Restore memcpy-sse2-unaligned.S from glibc 2.19 41ab67b... x86-64: Restore the old SSE4 strlen cbb5ed7... Include bench-timing.h only if _ISOMAC isn't defined 2b04ad4... Add __BEGIN_DECLS and __END_DECLS for C++ commit 3e31bc4a930e7b32924befe762014f85d5408692 Author: H.J. Lu <hjl.tools@gmail.com> Date: Mon May 22 12:00:43 2017 -0700 Add x86_cache.non_temporal_threshold to GLIBC_TUNABLES Add support for "glibc.x86_cache.non_temporal_threshold=number" to GLIBC_TUNABLES. * elf/dl-tunables.list (x86_cache): New name space. * sysdeps/x86/cacheinfo.c [HAVE_TUNABLES] (TUNABLE_NAMESPACE): New. [HAVE_TUNABLES]: Include <elf/dl-tunables.h>. [HAVE_TUNABLES] (DL_TUNABLE_CALLBACK (set_non_temporal_threshold)): New. [HAVE_TUNABLES] (init_cacheinfo): Call TUNABLE_SET_VAL_WITH_CALLBACK with set_non_temporal_threshold. commit bfb716e07b77f0ed8e0c2689d5cd01e2c8251fc5 Author: H.J. Lu <hjl.tools@gmail.com> Date: Fri May 12 13:38:04 2017 -0700 x86: Update __x86_shared_non_temporal_threshold __x86_shared_non_temporal_threshold was set to 6 times of per-core shared cache size, based on the large memcpy micro benchmark in glibc on a 8-core processor. For a processor with more than 8 cores, the threshold is too low. Set __x86_shared_non_temporal_threshold to the 3/4 of the total shared cache size so that it is unchanged on 8-core processors. On processors with less than 8 cores, the threshold is lower. * sysdeps/x86/cacheinfo.c (__x86_shared_non_temporal_threshold): Set to the 3/4 of the total shared cache size. commit 3b4afb94b84bba83751f861d5267a74b698caa6c Author: H.J. Lu <hjl.tools@gmail.com> Date: Fri May 19 11:14:19 2017 -0700 Test only a subset of memcpy commit f47a8ff13ba2f4e039f028453573c1489ab33186 Author: H.J. Lu <hjl.tools@gmail.com> Date: Wed May 10 16:05:09 2017 -0700 Integrate memcpy_benchmark.cc with glibc benchtests commit c91f98af397a5f07f48c464178e140f41e74a8e7 Author: H.J. Lu <hjl.tools@gmail.com> Date: Wed May 10 15:25:54 2017 -0700 Build memcpy_benchmark in benchtests Compile memcpy_benchmark.cc with -fpermissive -Wno-error -std=c++11 to silence GCC. commit 84158fda529ff223af2b3ba5bb293863034233e8 Author: H.J. Lu <hjl.tools@gmail.com> Date: Wed May 10 14:30:08 2017 -0700 Import memcpy_benchmark.cc From https://gist.github.com/ekelsen/b66cc085eb39f0495b57679cdb1874fa commit 006859c5be0cfc7797aaa91399ed32efc335df23 Author: H.J. Lu <hjl.tools@gmail.com> Date: Wed May 10 10:21:08 2017 -0700 x86-64: Restore memcpy-sse2-unaligned.S from glibc 2.19 commit 41ab67bceb481e4721fcbfe4c3787315303fc6a2 Author: H.J. Lu <hjl.tools@gmail.com> Date: Mon May 1 08:32:22 2017 -0700 x86-64: Restore the old SSE4 strlen commit cbb5ed76f4899c3052ec533cbc5879bdc60af8eb Author: H.J. Lu <hjl.tools@gmail.com> Date: Fri May 19 10:59:53 2017 -0700 Include bench-timing.h only if _ISOMAC isn't defined bench-timing.h includes hp-timing.h for high precision timing functions. But hp-timing.h is an internal header file, which can't be included when _ISOMAC is defined. But _ISOMAC is defined for C++ benchmark programs via libc-symbols.h. We don't include bench-timing.h if _ISOMAC is defined. If high precision timing functions are needed in C++ benchmark programs, we need to revisit this issue. * benchtests/bench-string.h: Include bench-timing.h only if _ISOMAC isn't defined. commit 2b04ad441adde9f5299ec69df8815542deab1add Author: H.J. Lu <hjl.tools@gmail.com> Date: Wed May 10 16:02:56 2017 -0700 Add __BEGIN_DECLS and __END_DECLS for C++ Add __BEGIN_DECLS and __END_DECLS to support C++. IFUNC_IMPL_ADD and IFUNC_IMPL are used internally in libc. They shouldn't be used in any programs. * include/ifunc-impl-list.h: Add __BEGIN_DECLS and __END_DECLS. (IFUNC_IMPL_ADD, IFUNC_IMPL): Define only if __cplusplus isn't defined.
reply other threads:[~2020-04-22 17:40 UTC|newest] Thread overview: [no followups] expand[flat|nested] mbox.gz Atom feed
Reply instructions: You may reply publicly to this message via plain-text email using any one of the following methods: * Save the following mbox file, import it into your mail client, and reply-to-all from there: mbox Avoid top-posting and favor interleaved quoting: https://en.wikipedia.org/wiki/Posting_style#Interleaved_style * Reply using the --to, --cc, and --in-reply-to switches of git-send-email(1): git send-email \ --in-reply-to=20200422174042.88B4C3858D33@sourceware.org \ --to=hjl@sourceware.org \ --cc=glibc-cvs@sourceware.org \ /path/to/YOUR_REPLY https://kernel.org/pub/software/scm/git/docs/git-send-email.html * If your mail client supports setting the In-Reply-To header via mailto: links, try the mailto: linkBe sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).