The Linux Kernel Mailing List
 help / color / mirror / Atom feed
From: John Stultz <johnstul@us.ibm.com>
To: Thomas Gleixner <tglx@linutronix.de>
Cc: Michal Simek <monstr@monstr.eu>,
	LKML <linux-kernel@vger.kernel.org>,
	john.williams@petalogix.com
Subject: Re: [PATCH 08/57] microblaze_v7: Interrupt handling, timer support, selfmod code
Date: Thu, 19 Mar 2009 19:24:21 -0700	[thread overview]
Message-ID: <1237515861.7106.215.camel@jstultz-laptop> (raw)
In-Reply-To: <alpine.LFD.2.00.0903192219500.29264@localhost.localdomain>

On Thu, 2009-03-19 at 22:47 +0100, Thomas Gleixner wrote:
> On Thu, 19 Mar 2009, Michal Simek wrote:
> > And the second question is about shift and rating values.
> > I wrote one message in past http://lkml.org/lkml/2009/1/11/291
> > Here is the important of part of that message.
> > 
> > ...
> > 
> > And the second part is about shift and rating values. Rating is
> > describe(linux/clocksource.h) and seems to me that should be
> > corresponded with CONFIG_HZ value,right?

Not sure where the idea of correspondence w/ CONFIG_HZ came from. The
rating value just provides a relative ordering of preferences between
possible clocksources. Since different hardware may have a number of
different clocksources available, we just need to have a method of
selecting a preferred clocksource, and the rating value is used for
that.

The guide in linux/clocksource.h is just a guide. Most arches, which
only have one or two clocksource options probably won't need much care,
and a rating of 200 or 300 will probably suffice. Or if there really
isn't any option about it and there is only one which is a must-use
clocksource, 400.


> > And I found any explanation of shift value -> max value for equation
> > (2-5) * freq << shift / NSEC_PER_SEC should be for my case still 32bit
> > number, where (2-5s) are because of NTP
> 
> @John, can you explain the shift vlaue please ?

The shift value is a bit more difficult to explain. The algorithm you
describe above is used by sparc to generate shift, and I think it will
work, but may not be optimal.

This question comes up over and over, so I figured I should sit down and
really solve it. 

Basically the constraint is you want to calculate a mult value using the
highest shift possible. However we have to be careful not to overflow
64bits when we multiply ~5second worth of cycles times the mult value.

So I finally put this down into code and here it is. No promises that it
is 100% right, but from my simple test examples it worked ok.

Let me know if it helps.

thanks
-john


Add helper functions to calculate ideal clocksource shift values given
its frequency (either hz or khz). Lightly tested, use with care. 
Also fixes a few old-old-old timesource (well, timsource) references my
find and replace attempts failed on.

Signed-off-by: John Stultz <johnstul@us.ibm.com>

diff --git a/include/linux/clocksource.h b/include/linux/clocksource.h
index f88d32f..9ef062d 100644
--- a/include/linux/clocksource.h
+++ b/include/linux/clocksource.h
@@ -116,7 +116,7 @@ extern struct clocksource *clock;	/* current clocksource */
  * @khz:		Clocksource frequency in KHz
  * @shift_constant:	Clocksource shift factor
  *
- * Helper functions that converts a khz counter frequency to a timsource
+ * Helper functions that converts a khz counter frequency to a clocksource
  * multiplier, given the clocksource shift value
  */
 static inline u32 clocksource_khz2mult(u32 khz, u32 shift_constant)
@@ -142,7 +142,7 @@ static inline u32 clocksource_khz2mult(u32 khz, u32 shift_constant)
  * @shift_constant:	Clocksource shift factor
  *
  * Helper functions that converts a hz counter
- * frequency to a timsource multiplier, given the
+ * frequency to a clocksource multiplier, given the
  * clocksource shift value
  */
 static inline u32 clocksource_hz2mult(u32 hz, u32 shift_constant)
@@ -162,6 +162,10 @@ static inline u32 clocksource_hz2mult(u32 hz, u32 shift_constant)
 	return (u32)tmp;
 }
 
+extern u32 clocksource_khz2shift(u32 freq_khz);
+extern u32 clocksource_hz2shift(u32 freq_hz);
+
+
 /**
  * clocksource_read: - Access the clocksource's current cycle value
  * @cs:		pointer to clocksource being read
diff --git a/kernel/time/clocksource.c b/kernel/time/clocksource.c
index ca89e15..0a2538d 100644
--- a/kernel/time/clocksource.c
+++ b/kernel/time/clocksource.c
@@ -212,6 +212,83 @@ static void clocksource_check_watchdog(struct clocksource *cs)
 static inline void clocksource_resume_watchdog(void) { }
 #endif
 
+
+/* Maximum seconds worth of cycles we consider valid */
+#define MAX_SEC_OF_CYCLES 5ULL
+
+/**
+ * clocksource_khz2shift - calculates shift from khz
+ * @freq_khz:		Clocksource frequency in KHz
+ *
+ * Helper functions that converts a khz counter
+ * frequency to a clocksource shift.
+ */
+u32 clocksource_khz2shift(u32 freq_khz)
+{
+	s32 shift = 32;
+	u32 mult;
+
+	while (1) {
+		s64 tmp;
+		u64 cycles;
+		mult = clocksource_khz2mult(freq_khz, shift);
+
+		cycles = (u64)freq_khz*1000*MAX_SEC_OF_CYCLES;
+		tmp = (cycles * mult) >> shift;
+
+		tmp = (MAX_SEC_OF_CYCLES*NSEC_PER_SEC) - tmp;
+		if (tmp < 0)
+			tmp = -tmp;
+
+		/* if we are witin one mult unit, we're good */
+		if (tmp < cycles>>shift)
+			break;
+
+		shift--;
+		if (shift < 1)
+			BUG();
+	}
+	return shift;
+}
+
+/**
+ * clocksource_hz2shift - calculates shift from hz
+ * @freq_hz:		Clocksource frequency in Hz
+ *
+ * Helper functions that converts a hz counter
+ * frequency to a clocksource shift.
+ */
+
+u32 clocksource_hz2shift(u32 freq_hz)
+{
+	s32 shift = 32;
+	u32 mult;
+
+	while (1) {
+		s64 tmp;
+		u64 cycles;
+		mult = clocksource_hz2mult(freq_hz, shift);
+
+		cycles = (u64)freq_hz*MAX_SEC_OF_CYCLES;
+		tmp = (cycles * mult) >> shift;
+
+		tmp = MAX_SEC_OF_CYCLES*NSEC_PER_SEC - tmp;
+		if (tmp < 0)
+			tmp = -tmp;
+
+		/* if we are witin one mult unit, we're good */
+		if (tmp < cycles>>shift)
+			break;
+
+		shift--;
+		if (shift < 1)
+			BUG();
+	}
+	return shift;
+}
+
+
+
 /**
  * clocksource_resume - resume the clocksource(s)
  */



  reply	other threads:[~2009-03-20  2:30 UTC|newest]

Thread overview: 113+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-03-18 20:30 Microblaze linux support monstr
2009-03-18 20:30 ` [PATCH 01/57] microblaze_v7: Kconfig patches monstr
2009-03-18 20:30   ` [PATCH 02/57] microblaze_v7: Makefiles for Microblaze cpu monstr
2009-03-18 20:30   ` [PATCH 03/57] microblaze_v7: Cpuinfo handling monstr
2009-03-18 20:30   ` [PATCH 04/57] microblaze_v7: Open firmware files monstr
2009-03-23 18:51     ` Arnd Bergmann
2009-03-23 20:53       ` Anton Vorontsov
2009-03-26  0:01         ` Benjamin Herrenschmidt
2009-03-24 11:17       ` Michal Simek
2009-03-18 20:30   ` [PATCH 05/57] microblaze_v7: Platorm bus registration monstr
2009-03-18 20:30   ` [PATCH 06/57] microblaze_v7: exception handling monstr
2009-03-18 20:30   ` [PATCH 07/57] microblaze_v7: Signal support monstr
2009-03-23 19:35     ` Arnd Bergmann
2009-03-24 11:14       ` Michal Simek
2009-03-18 20:30   ` [PATCH 08/57] microblaze_v7: Interrupt handling, timer support, selfmod code monstr
2009-03-19 15:49     ` Thomas Gleixner
2009-03-19 20:28       ` Michal Simek
2009-03-19 21:47         ` Thomas Gleixner
2009-03-20  2:24           ` John Stultz [this message]
2009-03-20  7:27             ` Michal Simek
2009-03-20 20:40               ` john stultz
2009-03-21 10:38                 ` Michal Simek
2009-03-21 11:14                   ` Thomas Gleixner
2009-03-21 11:57                     ` Michal Simek
2009-03-21 12:05                       ` Thomas Gleixner
2009-03-21 12:07                         ` Michal Simek
2009-03-20  6:38           ` Michal Simek
2009-03-20 10:07             ` Thomas Gleixner
2009-03-20  9:26           ` Michal Simek
2009-03-20  9:58             ` Thomas Gleixner
2009-03-20 10:37               ` Michal Simek
2009-03-20 10:49                 ` Thomas Gleixner
2009-03-20 11:28                   ` Michal Simek
2009-03-20 11:41                     ` Thomas Gleixner
2009-03-20 11:59                       ` Michal Simek
2009-03-20 14:08           ` Michal Simek
2009-03-20 14:12             ` Thomas Gleixner
2009-03-20 14:27               ` Michal Simek
2009-03-18 20:30   ` [PATCH 09/57] microblaze_v7: cache support monstr
2009-03-18 20:30   ` [PATCH 10/57] microblaze_v7: Generic dts file for platforms monstr
2009-03-18 20:30   ` [PATCH 11/57] microblaze_v7: kernel modules support monstr
2009-03-18 20:30   ` [PATCH 12/57] microblaze_v7: lmb include file monstr
2009-03-18 20:30   ` [PATCH 13/57] microblaze_v7: PVR support, cpuinfo support monstr
2009-03-18 20:30   ` [PATCH 14/57] microblaze_v7: defconfig file monstr
2009-03-18 20:30   ` [PATCH 15/57] microblaze_v7: assembler files head.S, entry-nommu.S, syscall_table.S monstr
2009-03-18 20:30   ` [PATCH 16/57] microblaze_v7: vmlinux.lds.S - linker script monstr
2009-03-18 20:30   ` [PATCH 17/57] microblaze_v7: supported function for memory - kernel/lib monstr
2009-03-18 20:30   ` [PATCH 18/57] microblaze_v7: checksum support monstr
2009-03-18 20:30   ` [PATCH 19/57] microblaze_v7: early_printk support monstr
2009-03-18 20:30   ` [PATCH 20/57] microblaze_v7: uaccess files monstr
2009-03-18 20:30   ` [PATCH 21/57] microblaze_v7: heartbeat file monstr
2009-03-18 20:30   ` [PATCH 22/57] microblaze_v7: setup.c, setup.h - system setting monstr
2009-03-18 20:30   ` [PATCH 23/57] microblaze_v7: asm-offsets monstr
2009-03-18 20:30   ` [PATCH 24/57] microblaze_v7: process and init task function monstr
2009-03-18 20:30   ` [PATCH 25/57] microblaze_v7: delay.h, timex.h monstr
2009-03-18 20:30   ` [PATCH 26/57] microblaze_v7: ptrace support monstr
2009-03-18 20:30   ` [PATCH 27/57] microblaze_v7: IPC support monstr
2009-03-18 20:30   ` [PATCH 28/57] microblaze_v7: traps support monstr
2009-03-18 20:30   ` [PATCH 29/57] microblaze_v7: memory inicialization, MMU, TLB monstr
2009-03-18 20:30   ` [PATCH 30/57] microblaze_v7: page.h, segment.h, unaligned.h monstr
2009-03-18 20:30   ` [PATCH 31/57] microblaze_v7: includes SHM*, msgbuf monstr
2009-03-18 20:30   ` [PATCH 32/57] microblaze_v7: bug headers files monstr
2009-03-18 20:31   ` [PATCH 33/57] microblaze_v7: definitions of types monstr
2009-03-18 20:31   ` [PATCH 34/57] microblaze_v7: ioctl support monstr
2009-03-18 20:31   ` [PATCH 35/57] microblaze_v7: io.h IO operations monstr
2009-03-18 20:31   ` [PATCH 36/57] microblaze_v7: headers for executables format FLAT, ELF monstr
2009-03-18 20:31   ` [PATCH 37/57] microblaze_v7: dma support monstr
2009-03-18 20:31   ` [PATCH 38/57] microblaze_v7: headers for irq monstr
2009-03-18 20:31   ` [PATCH 39/57] microblaze_v7: atomic.h bitops.h swab.h byteorder.h monstr
2009-03-18 20:31   ` [PATCH 40/57] microblaze_v7: headers pgalloc.h pgtable.h monstr
2009-03-18 20:31   ` [PATCH 41/57] microblaze_v7: system.h processor.h monstr
2009-03-18 20:31   ` [PATCH 42/57] microblaze_v7: clinkage.h linkage.h sections.h kmap_types.h monstr
2009-03-18 20:31   ` [PATCH 43/57] microblaze_v7: stats headers monstr
2009-03-18 20:31   ` [PATCH 44/57] microblaze_v7: termbits.h termios.h monstr
2009-03-23 19:37     ` Arnd Bergmann
2009-03-24 11:20       ` Michal Simek
2009-03-24 13:57         ` Arnd Bergmann
2009-03-24 14:06           ` John Williams
2009-03-24 14:44           ` Michal Simek
2009-03-18 20:31   ` [PATCH 45/57] microblaze_v7: sigcontext.h siginfo.h monstr
2009-03-18 20:31   ` [PATCH 46/57] microblaze_v7: headers simple files - empty or redirect to asm-generic monstr
2009-03-18 20:31   ` [PATCH 47/57] microblaze_v7: namei.h monstr
2009-03-18 20:31   ` [PATCH 48/57] microblaze_v7: headers files entry.h current.h mman.h registers.h sembuf.h monstr
2009-03-18 20:31   ` [PATCH 49/57] microblaze_v7: device.h param.h topology.h monstr
2009-03-18 20:31   ` [PATCH 50/57] microblaze_v7: pool.h socket.h monstr
2009-03-18 20:31   ` [PATCH 51/57] microblaze_v7: fcntl.h sockios.h ucontext.h monstr
2009-03-18 20:31   ` [PATCH 52/57] microblaze_v7: unistd.h monstr
2009-03-18 20:31   ` [PATCH 53/57] microblaze_v7: string.h thread_info.h monstr
2009-03-18 20:31   ` [PATCH 54/57] microblaze_v7: Kbuild file monstr
2009-03-18 20:31   ` [PATCH 55/57] microblaze_v7: pci headers monstr
2009-03-18 20:31   ` [PATCH 56/57] microblaze_v7: syscalls.h monstr
2009-03-18 20:31   ` [PATCH 57/57] microblaze_v7: Uartlite for Microblaze monstr
2009-03-24 16:03     ` Michal Simek
2009-03-24 16:22       ` John Williams
2009-03-24 16:47       ` Peter Korsgaard
2009-03-25 15:42   ` [PATCH 01/57] microblaze_v7: Kconfig patches Arnd Bergmann
2009-03-25 16:10     ` Michal Simek
2009-03-19  7:22 ` Microblaze linux support Ingo Molnar
2009-03-19  9:42   ` Michal Simek
2009-03-19 10:21     ` Ingo Molnar
2009-03-19 10:26       ` Michal Simek
2009-03-19 10:47         ` Jaswinder Singh Rajput
2009-03-19 11:10           ` Michal Simek
2009-03-19 10:50         ` Ingo Molnar
2009-03-19 10:52           ` Michal Simek
2009-03-19 11:00             ` Ingo Molnar
2009-03-19 11:04               ` Michal Simek
2009-03-19 20:35 ` Randy Dunlap
2009-03-19 20:41   ` Michal Simek
2009-03-24 15:26 ` Michal Simek
2009-03-24 15:33   ` John Linn
2009-03-24 15:42     ` Michal Simek
2009-03-24 17:46   ` [microblaze-uclinux] " Stephen Neuendorffer

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=1237515861.7106.215.camel@jstultz-laptop \
    --to=johnstul@us.ibm.com \
    --cc=john.williams@petalogix.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=monstr@monstr.eu \
    --cc=tglx@linutronix.de \
    /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