public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
From: "peter at cordes dot ca" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug target/104688] gcc and libatomic can use SSE for 128-bit atomic loads on Intel and AMD CPUs with AVX
Date: Mon, 28 Nov 2022 19:03:07 +0000 [thread overview]
Message-ID: <bug-104688-4-Wd5SbGJ7rD@http.gcc.gnu.org/bugzilla/> (raw)
In-Reply-To: <bug-104688-4@http.gcc.gnu.org/bugzilla/>
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104688
--- Comment #25 from Peter Cordes <peter at cordes dot ca> ---
(In reply to Alexander Monakov from comment #24)
>
> I think it's possible to get UC/WC mappings via a graphics/compute API (e.g.
> OpenGL, Vulkan, OpenCL, CUDA) on any OS if you get a mapping to device
> memory (and then CPU vendor cannot guarantee that 128b access won't tear
> because it might depend on downstream devices).
Even atomic_int doesn't work properly if you deref a pointer to WC memory. WC
doesn't have the same ordering guarantees, so it would break acquire/release
semantics.
So we already don't support WC for this.
We do at least de-facto support atomics on UC memory because the ordering
guarantees are a superset of cacheable memory, and 8-byte atomicity for aligned
load/store is guaranteed even for non-cacheable memory types since P5 Pentium
(and on AMD). (And lock cmpxchg16b is always atomic even on UC memory.)
But you're right that only Intel guarantees that 16-byte VMOVDQA loads/stores
would be atomic on UC memory. So this change could break that very unwise
corner-case on AMD which only guarantees that for cacheable loads/stores, and
Zhaoxin only for WB.
But was anyone previously using 16-byte atomics on UC device memory? Do we
actually care about supporting that? I'd guess no and no, so it's just a
matter of documenting that somewhere.
Since GCC7 we've reported 16-byte atomics as being non-lock-free, so I *hope*
people weren't using __atomic_store_n on device memory. The underlying
implementation was never guaranteed.
next prev parent reply other threads:[~2022-11-28 19:03 UTC|newest]
Thread overview: 35+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-02-25 14:22 [Bug target/104688] New: gcc and libatomic can use SSE for 128-bit atomic loads on Intel " xry111 at mengyan1223 dot wang
2022-02-25 14:30 ` [Bug target/104688] " jakub at gcc dot gnu.org
2022-02-25 14:33 ` xry111 at mengyan1223 dot wang
2022-02-25 14:34 ` fw at gcc dot gnu.org
2022-02-25 16:36 ` jakub at gcc dot gnu.org
2022-02-25 16:37 ` jakub at gcc dot gnu.org
2022-03-17 17:50 ` cvs-commit at gcc dot gnu.org
2022-03-29 5:54 ` cvs-commit at gcc dot gnu.org
2022-04-05 12:30 ` xry111 at mengyan1223 dot wang
2022-04-05 12:35 ` jakub at gcc dot gnu.org
2022-11-14 3:31 ` Ganesh.Gopalasubramanian at amd dot com
2022-11-14 4:58 ` sam at gentoo dot org
2022-11-14 5:10 ` [Bug target/104688] gcc and libatomic can use SSE for 128-bit atomic loads on Intel and AMD " xry111 at gcc dot gnu.org
2022-11-14 7:54 ` jakub at gcc dot gnu.org
2022-11-14 9:24 ` amonakov at gcc dot gnu.org
2022-11-14 9:29 ` jakub at gcc dot gnu.org
2022-11-14 9:52 ` amonakov at gcc dot gnu.org
2022-11-15 7:18 ` cvs-commit at gcc dot gnu.org
2022-11-15 7:20 ` jakub at gcc dot gnu.org
2022-11-20 23:11 ` cvs-commit at gcc dot gnu.org
2022-11-21 9:23 ` cvs-commit at gcc dot gnu.org
2022-11-23 9:18 ` xry111 at gcc dot gnu.org
2022-11-23 9:51 ` jakub at gcc dot gnu.org
2022-11-23 10:23 ` xry111 at gcc dot gnu.org
2022-11-28 18:35 ` peter at cordes dot ca
2022-11-28 18:46 ` amonakov at gcc dot gnu.org
2022-11-28 19:03 ` peter at cordes dot ca [this message]
2022-11-28 20:11 ` amonakov at gcc dot gnu.org
2022-11-28 20:47 ` peter at cordes dot ca
2022-11-29 8:11 ` fw at gcc dot gnu.org
2023-02-15 12:27 ` segher at gcc dot gnu.org
2023-02-15 12:46 ` fw at gcc dot gnu.org
2023-02-15 16:03 ` segher at gcc dot gnu.org
2023-02-15 16:07 ` pinskia at gcc dot gnu.org
2023-02-15 16:09 ` segher 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-104688-4-Wd5SbGJ7rD@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: link
Be 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).