public inbox for gcc-bugs@sourceware.org help / color / mirror / Atom feed
From: "munroesj at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org> To: gcc-bugs@gcc.gnu.org Subject: [Bug target/100085] Bad code for union transfer from __float128 to vector types Date: Fri, 16 Apr 2021 20:30:08 +0000 [thread overview] Message-ID: <bug-100085-4-1KIHR9yeDz@http.gcc.gnu.org/bugzilla/> (raw) In-Reply-To: <bug-100085-4@http.gcc.gnu.org/bugzilla/> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100085 --- Comment #4 from Steven Munroe <munroesj at gcc dot gnu.org> --- I am seeing this a similar problem with union transfers from __float128 to __int128. static inline unsigned __int128 vec_xfer_bin128_2_int128t (__binary128 f128) { __VF_128 vunion; vunion.vf1 = f128; return (vunion.ui1); } and unsigned __int128 test_xfer_bin128_2_int128 (__binary128 f128) { return vec_xfer_bin128_2_int128t (f128); } generates: 0000000000000030 <test_xfer_bin128_2_int128>: 30: 57 12 42 f0 xxswapd vs34,vs34 34: 20 00 20 39 li r9,32 38: d0 ff 41 39 addi r10,r1,-48 3c: 99 4f 4a 7c stxvd2x vs34,r10,r9 40: f0 ff 61 e8 ld r3,-16(r1) 44: f8 ff 81 e8 ld r4,-8(r1) 48: 20 00 80 4e blr For POWER8 should use mfvsrd/xxpermdi/mfvsrd. This looks like the root cause of poor performance for __float128 soft-float on POWER8. A simple benchmark using __float128 in C code calling libgcc for -mcpu=power8 and then hardware instructions for -mcpu=power9. P8 target P8AT14, Uses libgcc __addkf3_sw and __mulkf3_sw: test_time_f128 f128 CC tb delta = 52589, sec = 0.000102713 P9 Target P8AT14, Uses libgcc __addkf3_hw and __mulkf3_hw: test_time_f128 f128 CC tb delta = 18762, sec = 3.66445e-05 P9 Target P9AT14, inline hardware binary128 float: test_time_f128 f128 CC tb delta = 3809, sec = 7.43945e-06 I used Valgrind Itrace and Sim-ppc and perfstat analysis. Every call to libgcc __add/sub/mul/divkf3 takes a load-hit-store flush every call. This explains why __float128 is so 13.8 X slower on P8 then P9.
next prev parent reply other threads:[~2021-04-16 20:30 UTC|newest] Thread overview: 24+ messages / expand[flat|nested] mbox.gz Atom feed top 2021-04-14 18:19 [Bug rtl-optimization/100085] New: " munroesj at gcc dot gnu.org 2021-04-14 18:22 ` [Bug rtl-optimization/100085] " munroesj at gcc dot gnu.org 2021-04-15 6:59 ` [Bug target/100085] " rguenth at gcc dot gnu.org 2021-04-15 18:41 ` segher at gcc dot gnu.org 2021-04-16 20:30 ` munroesj at gcc dot gnu.org [this message] 2021-04-29 15:04 ` munroesj at gcc dot gnu.org 2021-04-30 19:52 ` bergner at gcc dot gnu.org 2021-05-24 6:41 ` luoxhu at gcc dot gnu.org 2021-05-24 21:49 ` segher at gcc dot gnu.org 2021-06-02 8:27 ` luoxhu at gcc dot gnu.org 2021-06-09 5:13 ` luoxhu at gcc dot gnu.org 2021-06-09 21:35 ` bergner at gcc dot gnu.org 2021-06-09 22:08 ` segher at gcc dot gnu.org 2021-06-10 15:00 ` munroesj at gcc dot gnu.org 2021-06-11 20:28 ` segher at gcc dot gnu.org 2022-01-14 17:17 ` wschmidt at gcc dot gnu.org 2022-02-24 20:48 ` munroesj at gcc dot gnu.org 2022-02-24 20:53 ` munroesj at gcc dot gnu.org 2022-02-24 21:17 ` segher at gcc dot gnu.org 2022-02-24 21:22 ` segher at gcc dot gnu.org 2022-02-24 21:26 ` segher at gcc dot gnu.org 2022-02-25 15:31 ` munroesj at gcc dot gnu.org 2022-02-25 22:57 ` segher at gcc dot gnu.org 2022-02-26 16:22 ` munroesj at gcc dot gnu.org
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=bug-100085-4-1KIHR9yeDz@http.gcc.gnu.org/bugzilla/ \ --to=gcc-bugzilla@gcc.gnu.org \ --cc=gcc-bugs@gcc.gnu.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).