From: Noah Goldstein <goldstein.w.n@gmail.com>
To: libc-alpha@sourceware.org
Subject: [PATCH v1 2/2] x86_64: Add {u}int_fast defs ideal for x86_64
Date: Sun, 10 Apr 2022 20:09:13 -0500 [thread overview]
Message-ID: <20220411010913.1061799-2-goldstein.w.n@gmail.com> (raw)
In-Reply-To: <20220411010913.1061799-1-goldstein.w.n@gmail.com>
{u}int_fast{16|32} should never use 64-bit types. 64-bit instructions
often use more code size (REX prefix) and a variety of instructions are
slower on many implementations (div, bswap, etc...).
Full xcheck passes on x86_64.
---
sysdeps/x86_64/int-fast.h | 50 +++++++++++++++++++++++++++++++++++++++
1 file changed, 50 insertions(+)
create mode 100644 sysdeps/x86_64/int-fast.h
diff --git a/sysdeps/x86_64/int-fast.h b/sysdeps/x86_64/int-fast.h
new file mode 100644
index 0000000000..8f341de646
--- /dev/null
+++ b/sysdeps/x86_64/int-fast.h
@@ -0,0 +1,50 @@
+/* Copyright (C) 2022 Free Software Foundation, Inc.
+ This file is part of the GNU C Library.
+
+ The GNU C Library is free software; you can redistribute it and/or
+ modify it under the terms of the GNU Lesser General Public
+ License as published by the Free Software Foundation; either
+ version 2.1 of the License, or (at your option) any later version.
+
+ The GNU C Library is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public
+ License along with the GNU C Library; if not, see
+ <https://www.gnu.org/licenses/>. */
+
+#ifndef _INT_FAST_X86_64_H
+#define _INT_FAST_X86_64_H 1
+
+#define __HAS_INT_FAST_DEFS 1
+
+/* Signed. */
+typedef signed char int_fast8_t;
+/* On x86_64 32-bit instructions are almost always fastest. */
+typedef int int_fast16_t;
+typedef int int_fast32_t;
+typedef long int int_fast64_t;
+
+/* Unsigned. */
+typedef unsigned char uint_fast8_t;
+/* On x86_64 32-bit instructions are almost always fastest. */
+typedef unsigned int uint_fast16_t;
+typedef unsigned int uint_fast32_t;
+typedef unsigned long int uint_fast64_t;
+
+
+# define __INT_FAST8_BASE INT8
+# define __INT_FAST16_BASE INT32
+# define __INT_FAST32_BASE INT32
+# define __INT_FAST64_BASE INT64
+
+# define __UINT_FAST8_BASE UINT8
+# define __UINT_FAST16_BASE UINT32
+# define __UINT_FAST32_BASE UINT32
+# define __UINT_FAST64_BASE UINT64
+
+# include <sysdeps/generic/int-fast.h>
+
+#endif
--
2.25.1
next prev parent reply other threads:[~2022-04-11 1:09 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-04-11 1:09 [PATCH v1 1/2] stdlib: Refactor {u}int_fast defs in stdint Noah Goldstein
2022-04-11 1:09 ` Noah Goldstein [this message]
2022-04-11 2:15 ` [PATCH v1 2/2] x86_64: Add {u}int_fast defs ideal for x86_64 Adam Sampson
2022-04-11 6:34 ` Florian Weimer
2022-04-14 22:24 ` Joseph Myers
2022-04-14 22:44 ` Noah Goldstein
2022-04-14 22:21 ` [PATCH v1 1/2] stdlib: Refactor {u}int_fast defs in stdint Joseph Myers
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=20220411010913.1061799-2-goldstein.w.n@gmail.com \
--to=goldstein.w.n@gmail.com \
--cc=libc-alpha@sourceware.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).