public inbox for gcc-bugs@sourceware.org help / color / mirror / Atom feed
From: "amonakov at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org> To: gcc-bugs@gcc.gnu.org Subject: [Bug tree-optimization/100363] gcc generating wider load/store than warranted at -O3 Date: Sat, 01 May 2021 22:54:42 +0000 [thread overview] Message-ID: <bug-100363-4-9LlWxyjMDC@http.gcc.gnu.org/bugzilla/> (raw) In-Reply-To: <bug-100363-4@http.gcc.gnu.org/bugzilla/> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100363 Alexander Monakov <amonakov at gcc dot gnu.org> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |amonakov at gcc dot gnu.org --- Comment #7 from Alexander Monakov <amonakov at gcc dot gnu.org> --- The github issue has a more relevant code quote: #ifdef CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS <-- this is enabled for ARCv2 279: PUP(sout) = PUP(sfrom); #else PUP(sout) = UP_UNALIGNED(sfrom); #endif Most likely the issue is that sout/sfrom are misaligned at runtime, while the vectorized code somewhere relies on them being sufficiently aligned for a 'short'. It is unsafe to dereference a misaligned pointer. The pointed-to-type must have reduced alignment: typedef unsigned short u16_u __attribute__((aligned(1))); u16_u *sout = ... u16_u *sfrom = (void *)(from - OFF); (without -ffreestanding, memcpy/memmove is a portable way to express a misaligned access) https://trust-in-soft.com/blog/2020/04/06/gcc-always-assumes-aligned-pointer-accesses/
next prev parent reply other threads:[~2021-05-01 22:54 UTC|newest] Thread overview: 22+ messages / expand[flat|nested] mbox.gz Atom feed top 2021-04-30 20:09 [Bug c/100363] New: " vgupta at synopsys dot com 2021-04-30 21:43 ` [Bug middle-end/100363] " pinskia at gcc dot gnu.org 2021-04-30 21:44 ` pinskia at gcc dot gnu.org 2021-04-30 21:56 ` vgupta at synopsys dot com 2021-04-30 22:02 ` torvalds@linux-foundation.org 2021-04-30 22:19 ` [Bug tree-optimization/100363] " pinskia at gcc dot gnu.org 2021-04-30 22:20 ` pinskia at gcc dot gnu.org 2021-04-30 22:33 ` vgupta at synopsys dot com 2021-05-01 22:54 ` amonakov at gcc dot gnu.org [this message] 2021-05-01 23:09 ` torvalds@linux-foundation.org 2021-05-03 7:41 ` rguenth at gcc dot gnu.org 2021-05-03 16:03 ` torvalds@linux-foundation.org 2021-05-03 16:18 ` torvalds@linux-foundation.org 2021-05-03 17:25 ` vgupta at synopsys dot com 2021-05-03 17:28 ` vgupta at synopsys dot com 2021-05-03 17:47 ` torvalds@linux-foundation.org 2021-05-03 18:45 ` vgupta at synopsys dot com 2021-05-03 19:43 ` pinskia at gcc dot gnu.org 2021-05-04 6:24 ` rguenth at gcc dot gnu.org 2021-05-04 19:33 ` vgupta at synopsys dot com 2021-05-05 6:32 ` rguenther at suse dot de 2021-05-05 19:59 ` ndesaulniers at google dot com
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-100363-4-9LlWxyjMDC@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).