From: Vincent Mailhol <mailhol.vincent@wanadoo.fr>
To: Nick Desaulniers <ndesaulniers@google.com>,
Thomas Gleixner <tglx@linutronix.de>,
Ingo Molnar <mingo@redhat.com>, Borislav Petkov <bp@alien8.de>
Cc: Dave Hansen <dave.hansen@linux.intel.com>,
x86@kernel.org, "H . Peter Anvin" <hpa@zytor.com>,
Nathan Chancellor <nathan@kernel.org>, Tom Rix <trix@redhat.com>,
linux-kernel@vger.kernel.org, llvm@lists.linux.dev,
David Howells <dhowells@redhat.com>,
Jan Beulich <JBeulich@suse.com>,
Vincent Mailhol <mailhol.vincent@wanadoo.fr>
Subject: [PATCH v3 2/2] x86/asm/bitops: __ffs,ffz: use __builtin_ctzl to evaluate constant expressions
Date: Thu, 12 May 2022 09:03:40 +0900 [thread overview]
Message-ID: <20220512000340.1171988-3-mailhol.vincent@wanadoo.fr> (raw)
In-Reply-To: <20220512000340.1171988-1-mailhol.vincent@wanadoo.fr>
__ffs(x) is equivalent to (unsigned long)__builtin_ctzl(x) and ffz(x)
is equivalent to (unsigned long)__builtin_ctzl(~x). Because
__builting_ctzl() returns an int, a cast to (unsigned long) is
necessary to avoid potential warnings on implicit casts.
For x86_64, the current __ffs() and ffz() implementations do not
produce optimized code when called with a constant expression. On the
contrary, the __builtin_ctzl() gets simplified into a single
instruction.
However, for non constant expressions, the __ffs() and ffz() asm
versions of the kernel remains slightly better than the code produced
by GCC (it produces a useless instruction to clear eax).
This patch uses the __builtin_constant_p() to select between the
kernel's __ffs()/ffz() and the __builtin_ctzl() depending on whether
the argument is constant or not.
** Statistics **
On a allyesconfig, before applying this patch...:
| $ objdump -d vmlinux.o | grep tzcnt | wc -l
| 3607
...and after:
| $ objdump -d vmlinux.o | grep tzcnt | wc -l
| 2600
So, roughly 27.9% of the calls to either __ffs() or ffz() were using
constant expressions and could be optimized out.
(tests done on linux v5.18-rc5 x86_64 using GCC 11.2.1)
Note: on x86_64, the asm bsf instruction produces tzcnt when used with
the ret prefix (which is why we grep tzcnt instead of bsf in above
benchmark). c.f. [1]
[1] commit e26a44a2d618 ("x86: Use REP BSF unconditionally")
http://lkml.kernel.org/r/5058741E020000780009C014@nat28.tlf.novell.com
CC: Nick Desaulniers <ndesaulniers@google.com>
Signed-off-by: Vincent Mailhol <mailhol.vincent@wanadoo.fr>
---
arch/x86/include/asm/bitops.h | 38 ++++++++++++++++++++++-------------
1 file changed, 24 insertions(+), 14 deletions(-)
diff --git a/arch/x86/include/asm/bitops.h b/arch/x86/include/asm/bitops.h
index 6ed979547086..fb0d7cd9f957 100644
--- a/arch/x86/include/asm/bitops.h
+++ b/arch/x86/include/asm/bitops.h
@@ -224,13 +224,7 @@ static __always_inline bool variable_test_bit(long nr, volatile const unsigned l
? constant_test_bit((nr), (addr)) \
: variable_test_bit((nr), (addr)))
-/**
- * __ffs - find first set bit in word
- * @word: The word to search
- *
- * Undefined if no bit exists, so code should check against 0 first.
- */
-static __always_inline unsigned long __ffs(unsigned long word)
+static __always_inline unsigned long variable___ffs(unsigned long word)
{
asm("rep; bsf %1,%0"
: "=r" (word)
@@ -238,13 +232,18 @@ static __always_inline unsigned long __ffs(unsigned long word)
return word;
}
-/**
- * ffz - find first zero bit in word
- * @word: The word to search
- *
- * Undefined if no zero exists, so code should check against ~0UL first.
- */
-static __always_inline unsigned long ffz(unsigned long word)
+/**
+ * __ffs - find first set bit in word
+ * @word: The word to search
+ *
+ * Undefined if no bit exists, so code should check against 0 first.
+ */
+#define __ffs(word) \
+ (__builtin_constant_p(word) ? \
+ (unsigned long)__builtin_ctzl(word) : \
+ variable___ffs(word))
+
+static __always_inline unsigned long variable_ffz(unsigned long word)
{
asm("rep; bsf %1,%0"
: "=r" (word)
@@ -252,6 +251,17 @@ static __always_inline unsigned long ffz(unsigned long word)
return word;
}
+/**
+ * ffz - find first zero bit in word
+ * @word: The word to search
+ *
+ * Undefined if no zero exists, so code should check against ~0UL first.
+ */
+#define ffz(word) \
+ (__builtin_constant_p(word) ? \
+ (unsigned long)__builtin_ctzl(~word) : \
+ variable_ffz(word))
+
/*
* __fls: find last set bit in word
* @word: The word to search
--
2.35.1
next prev parent reply other threads:[~2022-05-12 0:04 UTC|newest]
Thread overview: 68+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-05-11 16:03 [PATCH v2 0/2] x86/asm/bitops: optimize ff{s,z} functions for constant expressions Vincent Mailhol
2022-05-11 16:03 ` [PATCH v2 1/2] x86/asm/bitops: ffs: use __builtin_ffs to evaluate " Vincent Mailhol
2022-05-11 20:56 ` Christophe JAILLET
2022-05-11 23:30 ` Vincent MAILHOL
2022-05-11 21:35 ` Nick Desaulniers
2022-05-11 23:48 ` Vincent MAILHOL
2022-05-11 16:03 ` [PATCH v2 2/2] x86/asm/bitops: __ffs,ffz: use __builtin_ctzl " Vincent Mailhol
2022-05-11 22:20 ` Nick Desaulniers
2022-05-11 23:23 ` Vincent MAILHOL
2022-05-12 0:03 ` [PATCH v3 0/2] x86/asm/bitops: optimize ff{s,z} functions for " Vincent Mailhol
2022-05-12 0:03 ` [PATCH v3 1/2] x86/asm/bitops: ffs: use __builtin_ffs to evaluate " Vincent Mailhol
2022-05-12 0:28 ` Nick Desaulniers
2022-05-12 1:18 ` Vincent MAILHOL
2022-05-12 0:03 ` Vincent Mailhol [this message]
2022-05-12 0:19 ` [PATCH v3 2/2] x86/asm/bitops: __ffs,ffz: use __builtin_ctzl " Nick Desaulniers
2022-05-12 1:18 ` [PATCH v4 0/2] x86/asm/bitops: optimize ff{s,z} functions for " Vincent Mailhol
2022-05-12 1:18 ` [PATCH v4 1/2] x86/asm/bitops: ffs: use __builtin_ffs to evaluate " Vincent Mailhol
2022-05-12 3:02 ` Joe Perches
2022-05-12 4:29 ` Vincent MAILHOL
2022-05-12 1:18 ` [PATCH v4 2/2] x86/asm/bitops: __ffs,ffz: use __builtin_ctzl " Vincent Mailhol
2022-05-23 9:22 ` [PATCH v4 0/2] x86/asm/bitops: optimize ff{s,z} functions for " Vincent MAILHOL
2022-06-25 7:26 ` [RESEND PATCH " Vincent Mailhol
2022-06-25 7:26 ` [RESEND PATCH v4 1/2] x86/asm/bitops: ffs: use __builtin_ffs to evaluate " Vincent Mailhol
2022-06-25 7:26 ` [RESEND PATCH v4 2/2] x86/asm/bitops: __ffs,ffz: use __builtin_ctzl " Vincent Mailhol
2022-07-23 15:15 ` [RESEND PATCH v4 0/2] x86/asm/bitops: optimize ff{s,z} functions for " Vincent Mailhol
2022-07-23 15:15 ` [RESEND PATCH v4 1/2] x86/asm/bitops: ffs: use __builtin_ffs to evaluate " Vincent Mailhol
2022-08-11 14:59 ` Borislav Petkov
2022-08-12 11:55 ` Vincent MAILHOL
2022-07-23 15:15 ` [RESEND PATCH v4 2/2] x86/asm/bitops: __ffs,ffz: use __builtin_ctzl " Vincent Mailhol
2022-07-29 11:24 ` [RESEND PATCH v4 0/2] x86/asm/bitops: optimize ff{s,z} functions for " Vincent MAILHOL
2022-07-29 12:22 ` Borislav Petkov
2022-07-29 13:50 ` Vincent MAILHOL
2022-08-12 11:44 ` [PATCH v5 " Vincent Mailhol
2022-08-12 11:44 ` [PATCH v5 1/2] x86/asm/bitops: ffs: use __builtin_ffs to evaluate " Vincent Mailhol
2022-08-12 11:44 ` [PATCH v5 2/2] x86/asm/bitops: __ffs,ffz: use __builtin_ctzl " Vincent Mailhol
2022-08-23 16:23 ` Borislav Petkov
2022-08-23 17:12 ` Nick Desaulniers
2022-08-23 17:43 ` Borislav Petkov
2022-08-23 20:31 ` Vincent MAILHOL
2022-08-24 8:43 ` Borislav Petkov
2022-08-24 12:10 ` Vincent MAILHOL
2022-08-24 13:24 ` Borislav Petkov
2022-08-26 21:32 ` Vincent MAILHOL
2022-09-07 4:06 ` Borislav Petkov
2022-09-07 5:35 ` Vincent MAILHOL
2022-09-07 8:50 ` Borislav Petkov
2022-08-31 7:57 ` [PATCH v6 0/2] x86/asm/bitops: optimize ff{s,z} functions for " Vincent Mailhol
2022-08-31 7:57 ` [PATCH v6 1/2] x86/asm/bitops: ffs: use __builtin_ffs to evaluate " Vincent Mailhol
2022-08-31 7:57 ` [PATCH v6 2/2] x86/asm/bitops: __ffs,ffz: use __builtin_ctzl " Vincent Mailhol
2022-08-31 8:51 ` [PATCH v6 0/2] x86/asm/bitops: optimize ff{s,z} functions for " Yury Norov
2022-09-01 3:49 ` Yury Norov
2022-09-01 10:30 ` Vincent MAILHOL
2022-09-01 14:19 ` Yury Norov
2022-09-01 17:06 ` Nick Desaulniers
2022-09-02 5:34 ` Borislav Petkov
2022-09-02 0:41 ` Vincent MAILHOL
2022-09-02 1:19 ` Vincent MAILHOL
2022-09-05 0:37 ` [PATCH v7 " Vincent Mailhol
2022-09-05 0:37 ` [PATCH v7 1/2] x86/asm/bitops: ffs: use __builtin_ffs to evaluate " Vincent Mailhol
2022-09-05 0:37 ` [PATCH v7 2/2] x86/asm/bitops: __ffs,ffz: use __builtin_ctzl " Vincent Mailhol
2022-09-06 18:26 ` [PATCH v7 0/2] x86/asm/bitops: optimize ff{s,z} functions for " Nick Desaulniers
2022-09-07 7:04 ` Nick Desaulniers
2022-09-07 7:49 ` Vincent MAILHOL
2022-09-07 9:09 ` [PATCH v8 " Vincent Mailhol
2022-09-07 9:09 ` [PATCH v8 1/2] x86/asm/bitops: ffs: use __builtin_ffs to evaluate " Vincent Mailhol
2022-09-07 9:09 ` [PATCH v8 2/2] x86/asm/bitops: __ffs,ffz: use __builtin_ctzl " Vincent Mailhol
2022-09-20 20:00 ` [tip: x86/asm] x86/asm/bitops: Use __builtin_ctzl() " tip-bot2 for Vincent Mailhol
2022-09-20 20:00 ` [tip: x86/asm] x86/asm/bitops: Use __builtin_ffs() " tip-bot2 for Vincent Mailhol
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=20220512000340.1171988-3-mailhol.vincent@wanadoo.fr \
--to=mailhol.vincent@wanadoo.fr \
--cc=JBeulich@suse.com \
--cc=bp@alien8.de \
--cc=dave.hansen@linux.intel.com \
--cc=dhowells@redhat.com \
--cc=hpa@zytor.com \
--cc=linux-kernel@vger.kernel.org \
--cc=llvm@lists.linux.dev \
--cc=mingo@redhat.com \
--cc=nathan@kernel.org \
--cc=ndesaulniers@google.com \
--cc=tglx@linutronix.de \
--cc=trix@redhat.com \
--cc=x86@kernel.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 NNTP newsgroup(s).