linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Arnd Bergmann <arnd@arndb.de>
To: stable@vger.kernel.org, Will Deacon <will.deacon@arm.com>,
	Florian La Roche <florian.laroche@googlemail.com>
Cc: Peter Zijlstra <peterz@infradead.org>,
	Anshul Garg <aksgarg1989@gmail.com>,
	Linus Torvalds <torvalds@linux-foundation.org>,
	Davidlohr Bueso <dave@stgolabs.net>,
	Thomas Gleixner <tglx@linutronix.de>,
	Ingo Molnar <mingo@kernel.org>, Joe Perches <joe@perches.com>,
	David Miller <davem@davemloft.net>,
	Matthew Wilcox <mawilcox@microsoft.com>,
	Kees Cook <keescook@chromium.org>,
	Michael Davidson <md@google.com>,
	Andrew Morton <akpm@linux-foundation.org>,
	Arnd Bergmann <arnd@arndb.de>,
	linux-kernel@vger.kernel.org
Subject: [BACKPORT 4.4.y 24/25] lib/int_sqrt: optimize small argument
Date: Fri, 22 Mar 2019 16:44:15 +0100	[thread overview]
Message-ID: <20190322154425.3852517-25-arnd@arndb.de> (raw)
In-Reply-To: <20190322154425.3852517-1-arnd@arndb.de>

From: Peter Zijlstra <peterz@infradead.org>

The current int_sqrt() computation is sub-optimal for the case of small
@x.  Which is the interesting case when we're going to do cumulative
distribution functions on idle times, which we assume to be a random
variable, where the target residency of the deepest idle state gives an
upper bound on the variable (5e6ns on recent Intel chips).

In the case of small @x, the compute loop:

	while (m != 0) {
		b = y + m;
		y >>= 1;

		if (x >= b) {
			x -= b;
			y += m;
		}
		m >>= 2;
	}

can be reduced to:

	while (m > x)
		m >>= 2;

Because y==0, b==m and until x>=m y will remain 0.

And while this is computationally equivalent, it runs much faster
because there's less code, in particular less branches.

      cycles:                 branches:              branch-misses:

OLD:

hot:   45.109444 +- 0.044117  44.333392 +- 0.002254  0.018723 +- 0.000593
cold: 187.737379 +- 0.156678  44.333407 +- 0.002254  6.272844 +- 0.004305

PRE:

hot:   67.937492 +- 0.064124  66.999535 +- 0.000488  0.066720 +- 0.001113
cold: 232.004379 +- 0.332811  66.999527 +- 0.000488  6.914634 +- 0.006568

POST:

hot:   43.633557 +- 0.034373  45.333132 +- 0.002277  0.023529 +- 0.000681
cold: 207.438411 +- 0.125840  45.333132 +- 0.002277  6.976486 +- 0.004219

Averages computed over all values <128k using a LFSR to generate order.
Cold numbers have a LFSR based branch trace buffer 'confuser' ran between
each int_sqrt() invocation.

Link: http://lkml.kernel.org/r/20171020164644.876503355@infradead.org
Fixes: 30493cc9dddb ("lib/int_sqrt.c: optimize square root algorithm")
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Suggested-by: Anshul Garg <aksgarg1989@gmail.com>
Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Davidlohr Bueso <dave@stgolabs.net>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Joe Perches <joe@perches.com>
Cc: David Miller <davem@davemloft.net>
Cc: Matthew Wilcox <mawilcox@microsoft.com>
Cc: Kees Cook <keescook@chromium.org>
Cc: Michael Davidson <md@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
(cherry picked from commit 3f3295709edea6268ff1609855f498035286af73)
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
---
 lib/int_sqrt.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/lib/int_sqrt.c b/lib/int_sqrt.c
index 1ef4cc344977..1afb545a37c5 100644
--- a/lib/int_sqrt.c
+++ b/lib/int_sqrt.c
@@ -22,6 +22,9 @@ unsigned long int_sqrt(unsigned long x)
 		return x;
 
 	m = 1UL << (BITS_PER_LONG - 2);
+	while (m > x)
+		m >>= 2;
+
 	while (m != 0) {
 		b = y + m;
 		y >>= 1;
-- 
2.20.0


  parent reply	other threads:[~2019-03-22 15:50 UTC|newest]

Thread overview: 46+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-03-22 15:43 [BACKPORT 4.4.y 00/25] candidates from spreadtrum 4.4 product kernel Arnd Bergmann
2019-03-22 15:43 ` [BACKPORT 4.4.y 01/25] mmc: pwrseq: constify mmc_pwrseq_ops structures Arnd Bergmann
2019-03-26  1:08   ` Greg KH
2019-03-26  6:44     ` Julia Lawall
2019-03-26  8:11     ` Arnd Bergmann
2019-03-22 15:43 ` [BACKPORT 4.4.y 02/25] ALSA: compress: add support for 32bit calls in a 64bit kernel Arnd Bergmann
2019-03-26  1:09   ` Greg KH
2019-03-26  7:55     ` Arnd Bergmann
2019-03-30  9:40       ` Greg KH
2019-03-22 15:43 ` [BACKPORT 4.4.y 03/25] mmc: pwrseq_simple: Make reset-gpios optional to match doc Arnd Bergmann
2019-03-22 15:43 ` [BACKPORT 4.4.y 04/25] USB: iowarrior: fix oops with malicious USB descriptors Arnd Bergmann
2019-03-26  1:13   ` Greg Kroah-Hartman
2019-03-26  8:20     ` Arnd Bergmann
2019-03-26  9:35       ` Baolin Wang
2019-03-26  9:47         ` 翟京 (Orson Zhai)
2019-03-22 15:43 ` [BACKPORT 4.4.y 05/25] mmc: debugfs: Add a restriction to mmc debugfs clock setting Arnd Bergmann
2019-03-22 15:43 ` [BACKPORT 4.4.y 06/25] mmc: make MAN_BKOPS_EN message a debug Arnd Bergmann
2019-03-22 15:43 ` [BACKPORT 4.4.y 07/25] mmc: sanitize 'bus width' in debug output Arnd Bergmann
2019-03-22 15:43 ` [BACKPORT 4.4.y 08/25] mmc: core: shut up "voltage-ranges unspecified" pr_info() Arnd Bergmann
2019-03-22 15:44 ` [BACKPORT 4.4.y 09/25] usb: dwc3: gadget: Fix suspend/resume during device mode Arnd Bergmann
2019-03-22 15:44 ` [BACKPORT 4.4.y 10/25] arm64: mm: Add trace_irqflags annotations to do_debug_exception() Arnd Bergmann
2019-03-22 15:44 ` [BACKPORT 4.4.y 11/25] mmc: core: fix using wrong io voltage if mmc_select_hs200 fails Arnd Bergmann
2019-03-22 15:44 ` [BACKPORT 4.4.y 12/25] mm/rmap: replace BUG_ON(anon_vma->degree) with VM_WARN_ON Arnd Bergmann
2019-03-22 15:44 ` [BACKPORT 4.4.y 13/25] extcon: usb-gpio: Don't miss event during suspend/resume Arnd Bergmann
2019-03-22 15:44 ` [BACKPORT 4.4.y 14/25] kbuild: setlocalversion: print error to STDERR Arnd Bergmann
2019-03-22 15:44 ` [BACKPORT 4.4.y 15/25] usb: gadget: composite: fix dereference after null check coverify warning Arnd Bergmann
2019-03-22 15:44 ` [BACKPORT 4.4.y 16/25] usb: gadget: Add the gserial port checking in gs_start_tx() Arnd Bergmann
2019-03-22 15:44 ` [BACKPORT 4.4.y 17/25] mmc: core: don't try to switch block size for dual rate mode Arnd Bergmann
2019-03-26  1:27   ` Greg KH
2019-03-26  8:14     ` Arnd Bergmann
2019-03-22 15:44 ` [BACKPORT 4.4.y 18/25] tcp/dccp: drop SYN packets if accept queue is full Arnd Bergmann
2019-03-26  1:21   ` Greg KH
2019-03-22 15:44 ` [BACKPORT 4.4.y 19/25] serial: sprd: adjust TIMEOUT to a big value Arnd Bergmann
2019-03-26  1:21   ` Greg KH
2019-03-22 15:44 ` [BACKPORT 4.4.y 20/25] Hang/soft lockup in d_invalidate with simultaneous calls Arnd Bergmann
2019-03-26  1:30   ` Greg KH
2019-03-22 15:44 ` [BACKPORT 4.4.y 21/25] arm64: traps: disable irq in die() Arnd Bergmann
2019-03-26  1:31   ` Greg KH
2019-03-22 15:44 ` [BACKPORT 4.4.y 22/25] usb: renesas_usbhs: gadget: fix unused-but-set-variable warning Arnd Bergmann
2019-03-22 15:44 ` [BACKPORT 4.4.y 23/25] serial: sprd: clear timeout interrupt only rather than all interrupts Arnd Bergmann
2019-03-26  1:34   ` Greg KH
2019-03-22 15:44 ` Arnd Bergmann [this message]
2019-03-26  1:36   ` [BACKPORT 4.4.y 24/25] lib/int_sqrt: optimize small argument Greg KH
2019-03-22 15:44 ` [BACKPORT 4.4.y 25/25] USB: core: only clean up what we allocated Arnd Bergmann
2019-03-26  1:36   ` Greg Kroah-Hartman
2019-03-26  2:18 ` [BACKPORT 4.4.y 00/25] candidates from spreadtrum 4.4 product kernel Greg KH

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=20190322154425.3852517-25-arnd@arndb.de \
    --to=arnd@arndb.de \
    --cc=akpm@linux-foundation.org \
    --cc=aksgarg1989@gmail.com \
    --cc=dave@stgolabs.net \
    --cc=davem@davemloft.net \
    --cc=florian.laroche@googlemail.com \
    --cc=joe@perches.com \
    --cc=keescook@chromium.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mawilcox@microsoft.com \
    --cc=md@google.com \
    --cc=mingo@kernel.org \
    --cc=peterz@infradead.org \
    --cc=stable@vger.kernel.org \
    --cc=tglx@linutronix.de \
    --cc=torvalds@linux-foundation.org \
    --cc=will.deacon@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 NNTP newsgroup(s).