From: naohirot--- via Libc-alpha <libc-alpha@sourceware.org>
To: Szabolcs Nagy <szabolcs.nagy@arm.com>,
Wilco Dijkstra <Wilco.Dijkstra@arm.com>
Cc: 'GNU C Library' <libc-alpha@sourceware.org>
Subject: RE: [PATCH v3 1/5] AArch64: Improve A64FX memset
Date: Tue, 3 Aug 2021 02:57:35 +0000 [thread overview]
Message-ID: <TYAPR01MB6025217DC110E2C4296D1163DFF09@TYAPR01MB6025.jpnprd01.prod.outlook.com> (raw)
In-Reply-To: <20210802145003.GH14854@arm.com>
Hi Szabolcs, Wilco,
> From: Szabolcs Nagy <szabolcs.nagy@arm.com>
> Sent: Monday, August 2, 2021 11:50 PM
> The 08/02/2021 14:38, Wilco Dijkstra via Libc-alpha wrote:
> > > We discussed how should be defined BTI_C macro before, at that time conclusion
> > > was "NOP" rather than empty unless HAVE_AARCH64_BTI.
> > > Now the above code defines BTI_C as empty unconditionally.
> > > A64FX doesn't support BTI, so this code is OK.
> > > But I'm just interested in the reason why it is changed.
> >
> > We changed to NOP in the generic code, so that works for all string functions.
> > In this specific case removing the initial NOP as well allows all performance critical
> > code for <= 512 bytes to be perfectly aligned to 16-byte fetch blocks.
>
> yes, this makes sense:
>
> originally BTI_C was always hint 34, but since that can be
> slow it was changed for !HAVE_AARCH64_BTI. We don't want the
> layout of asm code to change based on toolchain configuration
> so BTI_C is defined as a place holder nop then.
Now I understood the difference between nop and empty is the layout.
When we discussed BTI_C before, I didn't ask the difference so as not
to prolong the discussion because there is no performance difference.
> but in a64fx specific code bti is never needed so we also
> don't need the place holder nop, BTI_C can be unconditionally
> empty.
Yes, I'd like to change __memcpy_a64fx and __memmove_a64fx to the same
way too.
Thanks.
Naohiro
next prev parent reply other threads:[~2021-08-03 2:58 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-07-22 15:59 [PATCH v3 1/5] AArch64: Improve A64FX memset Wilco Dijkstra via Libc-alpha
2021-07-28 8:10 ` naohirot--- via Libc-alpha
2021-08-02 13:53 ` naohirot--- via Libc-alpha
2021-08-02 14:38 ` Wilco Dijkstra via Libc-alpha
2021-08-02 14:50 ` Szabolcs Nagy via Libc-alpha
2021-08-03 2:57 ` naohirot--- via Libc-alpha [this message]
2021-08-03 8:01 ` Szabolcs Nagy via Libc-alpha
2021-09-24 7:56 ` naohirot--- via Libc-alpha
2021-08-03 2:56 ` naohirot--- via Libc-alpha
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
List information: https://www.gnu.org/software/libc/involved.html
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=TYAPR01MB6025217DC110E2C4296D1163DFF09@TYAPR01MB6025.jpnprd01.prod.outlook.com \
--to=libc-alpha@sourceware.org \
--cc=Wilco.Dijkstra@arm.com \
--cc=naohirot@fujitsu.com \
--cc=szabolcs.nagy@arm.com \
/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).