public inbox for gcc-bugs@sourceware.org help / color / mirror / Atom feed
From: "rguenth at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org> To: gcc-bugs@gcc.gnu.org Subject: [Bug tree-optimization/107096] Fully masking vectorization with AVX512 ICEs gcc.dg/vect/vect-over-widen-*.c Date: Fri, 30 Sep 2022 12:30:45 +0000 [thread overview] Message-ID: <bug-107096-4-WqeH33qQqB@http.gcc.gnu.org/bugzilla/> (raw) In-Reply-To: <bug-107096-4@http.gcc.gnu.org/bugzilla/> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107096 Richard Biener <rguenth at gcc dot gnu.org> changed: What |Removed |Added ---------------------------------------------------------------------------- Target| |x86_64-*-* CC| |rsandifo at gcc dot gnu.org --- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> --- Take the simplified void foo (int * __restrict dst, short *src, int n) { for (int i = 0; i < n; ++i) dst[i] = src[2*i] + src[2*i+1]; } here we get vect_record_loop_mask twice with nvectors == 2 and V16HImode for the load which populates masks[2]. Then we once get V8SImode but also with nvectors == 2 which leaves the data unadjusted since it looks at masks[2] as well but if it were to come first we'd have recorded a different mask vector type. The masks seem to be constructed in a way to produce two bits per lane (but we still apply it naively?!) and the V_C_E does actually look wrong to me. Huh. With SVE I seem to get (besides a .LOAD_LANE version) permutes of the mask vector: loop_mask_116 = VEC_PERM_EXPR <loop_mask_131, loop_mask_131, { 0, POLY_INT_CST [4, 4], 1, POLY_INT_CST [5, 4], 2, POLY_INT_CST [6, 4], ... }>; loop_mask_117 = VEC_PERM_EXPR <loop_mask_131, loop_mask_131, { POLY_INT_CST [2, 2], POLY_INT_CST [6, 6], POLY_INT_CST [3, 2], POLY_INT_CST [7, 6], POLY_INT_CST [4, 2], POLY_INT_CST [8, 6], ... }>; ... vect__69.20_118 = .MASK_LOAD (vectp_src.18_114, 16B, loop_mask_116); vectp_src.18_119 = vectp_src.18_114 + POLY_INT_CST [8, 8]; vect__69.21_120 = .MASK_LOAD (vectp_src.18_119, 16B, loop_mask_117); ... .MASK_STORE (vectp_dst.25_129, 32B, loop_mask_131, vect__79.24_125); so the original mask provider here is the larger element vector type and the smaller element masks are produced from that. That's something I'd expect for AVX512 as well, not sure where it goes "wrong". See PR107093 for a patch implementing WHILE_ULT for AVX512.
next prev parent reply other threads:[~2022-09-30 12:30 UTC|newest] Thread overview: 15+ messages / expand[flat|nested] mbox.gz Atom feed top 2022-09-30 11:41 [Bug tree-optimization/107096] New: " rguenth at gcc dot gnu.org 2022-09-30 12:30 ` rguenth at gcc dot gnu.org [this message] 2022-09-30 13:07 ` [Bug tree-optimization/107096] " rsandifo at gcc dot gnu.org 2022-10-10 9:49 ` rguenth at gcc dot gnu.org 2022-10-10 11:01 ` ams at gcc dot gnu.org 2022-10-10 11:33 ` rguenth at gcc dot gnu.org 2022-10-10 12:38 ` rsandifo at gcc dot gnu.org 2022-10-10 12:55 ` rguenther at suse dot de 2022-10-10 14:20 ` rsandifo at gcc dot gnu.org 2022-10-11 12:43 ` rguenth at gcc dot gnu.org 2023-02-15 7:08 ` crazylht at gmail dot com 2023-02-15 7:09 ` crazylht at gmail dot com 2023-02-15 7:26 ` rguenther at suse dot de 2023-02-16 2:09 ` crazylht at gmail dot com 2023-06-14 12:55 ` rguenth 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-107096-4-WqeH33qQqB@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).