All of lore.kernel.org
 help / color / mirror / Atom feed
From: Guenter Roeck <linux@roeck-us.net>
To: Marc Titinger <mtitinger@baylibre.com>, jdelvare@suse.com
Cc: lm-sensors@lm-sensors.org, linux-kernel@vger.kernel.org,
	bcousson@baylibre.com, mturquette@baylibre.com
Subject: Re: [PATCH v2] hwmon: ina2xx: allow for actual measurement bandwidth above 160 Hz
Date: Tue, 20 Oct 2015 05:54:11 -0700	[thread overview]
Message-ID: <56263973.300@roeck-us.net> (raw)
In-Reply-To: <1445329244-21627-1-git-send-email-mtitinger@baylibre.com>

On 10/20/2015 01:20 AM, Marc Titinger wrote:
> With the current implementation, the driver will prevent a readout at a
> pace faster than the default conversion time (2ms) times the averaging
> setting, min AVG being 1:1.
>
> Any sysfs "show" read access from the client app faster than 500 Hz will be
> 'cached' by the driver, but actually since do_update reads all 8 registers,
> the best achievable measurement rate is roughly 8*800 us (for the time
> spent in i2c-core) i.e. <= 156Hz with Beagle Bone Black.
>
> This change set uses a register mask to allow for the readout of a single
> i2c register at a time. Furthermore, performing subsequent reads on the
> same register will make use of the ability of the i2c chip to retain the
> last reg offset, hence use a shorter i2c message (roughly 400us instead of
> 800us spent in i2c-core.c).
>
That doesn't work. There could be accesses from other sources (such as through
i2c-dev, or in multi-master systems) between two reads.

> The best readout rate for a single measurement is now around 2kHz. And for
> four measurements around (1/(4*800us) = 312 Hz. Since for any readout rate
> faster than 160 Hz the interval is set by the i2c transactions completion,
> the 'last-update' anti-flooding code will not have a limiting effect in
> practice. Hence I also remove the elapsed time checking in the hwmon driver
> for ina2xx.
>
> To summarize, the patch provides a max bandwidth improvement with hwmon
> client apps from ~160 Hz to ~320 Hz, and better in single-channel
> polling mode.
>
Overall your patch pretty much re-implements regmap. Since you drop caching,
it is also unnecessary to read all registers at a time, so you can just use
a function to read _one_ register and returns its value (with retries).
Or use regmap. Either case, do_update() and ina2xx_update_device() are no
longer needed.

If you want to convert the driver to regmap, just look for 'regmap' in
drivers/hwmon for examples.

Guenter


WARNING: multiple messages have this Message-ID (diff)
From: Guenter Roeck <linux@roeck-us.net>
To: Marc Titinger <mtitinger@baylibre.com>, jdelvare@suse.com
Cc: lm-sensors@lm-sensors.org, linux-kernel@vger.kernel.org,
	bcousson@baylibre.com, mturquette@baylibre.com
Subject: Re: [lm-sensors] [PATCH v2] hwmon: ina2xx: allow for actual measurement bandwidth above 160 Hz
Date: Tue, 20 Oct 2015 12:54:11 +0000	[thread overview]
Message-ID: <56263973.300@roeck-us.net> (raw)
In-Reply-To: <1445329244-21627-1-git-send-email-mtitinger@baylibre.com>

On 10/20/2015 01:20 AM, Marc Titinger wrote:
> With the current implementation, the driver will prevent a readout at a
> pace faster than the default conversion time (2ms) times the averaging
> setting, min AVG being 1:1.
>
> Any sysfs "show" read access from the client app faster than 500 Hz will be
> 'cached' by the driver, but actually since do_update reads all 8 registers,
> the best achievable measurement rate is roughly 8*800 us (for the time
> spent in i2c-core) i.e. <= 156Hz with Beagle Bone Black.
>
> This change set uses a register mask to allow for the readout of a single
> i2c register at a time. Furthermore, performing subsequent reads on the
> same register will make use of the ability of the i2c chip to retain the
> last reg offset, hence use a shorter i2c message (roughly 400us instead of
> 800us spent in i2c-core.c).
>
That doesn't work. There could be accesses from other sources (such as through
i2c-dev, or in multi-master systems) between two reads.

> The best readout rate for a single measurement is now around 2kHz. And for
> four measurements around (1/(4*800us) = 312 Hz. Since for any readout rate
> faster than 160 Hz the interval is set by the i2c transactions completion,
> the 'last-update' anti-flooding code will not have a limiting effect in
> practice. Hence I also remove the elapsed time checking in the hwmon driver
> for ina2xx.
>
> To summarize, the patch provides a max bandwidth improvement with hwmon
> client apps from ~160 Hz to ~320 Hz, and better in single-channel
> polling mode.
>
Overall your patch pretty much re-implements regmap. Since you drop caching,
it is also unnecessary to read all registers at a time, so you can just use
a function to read _one_ register and returns its value (with retries).
Or use regmap. Either case, do_update() and ina2xx_update_device() are no
longer needed.

If you want to convert the driver to regmap, just look for 'regmap' in
drivers/hwmon for examples.

Guenter


_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

  reply	other threads:[~2015-10-20 12:54 UTC|newest]

Thread overview: 58+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-10-19 16:21 [RFC] hwmon: ina2xx: allow for actual measurement bandwidth above 160 Hz Marc Titinger
2015-10-19 16:21 ` [lm-sensors] " Marc Titinger
2015-10-20  1:32 ` Guenter Roeck
2015-10-20  1:32   ` [lm-sensors] " Guenter Roeck
2015-10-20  7:58   ` Marc Titinger
2015-10-20  7:58     ` [lm-sensors] " Marc Titinger
2015-10-20  8:20   ` [PATCH v2] " Marc Titinger
2015-10-20  8:20     ` [lm-sensors] " Marc Titinger
2015-10-20 12:54     ` Guenter Roeck [this message]
2015-10-20 12:54       ` Guenter Roeck
2015-10-20 13:17       ` Marc Titinger
2015-10-20 13:17         ` [lm-sensors] " Marc Titinger
2015-10-20 13:30         ` Guenter Roeck
2015-10-20 13:30           ` [lm-sensors] " Guenter Roeck
2015-10-20 13:46           ` Marc Titinger
2015-10-20 13:46             ` [lm-sensors] " Marc Titinger
2015-10-20 17:00             ` Guenter Roeck
2015-10-20 17:00               ` [lm-sensors] " Guenter Roeck
2015-10-21  7:46               ` Marc Titinger
2015-10-21  7:46                 ` [lm-sensors] " Marc Titinger
2015-10-20 13:52           ` Michael Turquette
2015-10-20 13:52             ` [lm-sensors] " Michael Turquette
2015-10-23 16:13       ` [RFC] hwmon: ina2xx: port to using remap, improve bandwidth Marc Titinger
2015-10-23 16:13         ` [lm-sensors] " Marc Titinger
2015-10-23 16:49         ` kbuild test robot
2015-10-23 16:49           ` [lm-sensors] " kbuild test robot
2015-10-23 16:52         ` Guenter Roeck
2015-10-23 16:52           ` [lm-sensors] " Guenter Roeck
2015-10-23 20:35           ` Marc Titinger
2015-10-23 20:35             ` [lm-sensors] " Marc Titinger
2015-10-24  2:21             ` Guenter Roeck
2015-10-24  2:21               ` [lm-sensors] " Guenter Roeck
2015-10-24 12:45             ` Guenter Roeck
2015-10-24 12:45               ` [lm-sensors] " Guenter Roeck
2015-10-26 16:24               ` [PATCH 0/2] hwmon: ina2xx: convert driver to using regmap Marc Titinger
2015-10-26 16:24                 ` [lm-sensors] " Marc Titinger
2015-10-26 16:24               ` [PATCH 1/2] " Marc Titinger
2015-10-26 16:24                 ` [lm-sensors] " Marc Titinger
2015-10-27  1:02                 ` Guenter Roeck
2015-10-27  1:02                   ` [lm-sensors] " Guenter Roeck
2015-10-27  1:12                 ` Guenter Roeck
2015-10-27  1:12                   ` [lm-sensors] " Guenter Roeck
2015-10-27  9:51                   ` [PATCH v2 " Marc Titinger
2015-10-27  9:51                     ` [lm-sensors] " Marc Titinger
2015-10-28  2:47                     ` Guenter Roeck
2015-10-28  2:47                       ` [lm-sensors] " Guenter Roeck
2015-10-28  9:23                       ` Marc Titinger
2015-10-28  9:23                         ` [lm-sensors] " Marc Titinger
2015-10-26 16:24               ` [PATCH 2/2] hwmon: ina2xx: give precedence to DT over checking for platform data Marc Titinger
2015-10-26 16:24                 ` [lm-sensors] " Marc Titinger
2015-10-27  1:10                 ` Guenter Roeck
2015-10-27  1:10                   ` [lm-sensors] " Guenter Roeck
2015-10-27  9:51                   ` [PATCH v2 " Marc Titinger
2015-10-27  9:51                     ` [lm-sensors] " Marc Titinger
2015-10-23 16:55         ` [RFC] hwmon: ina2xx: port to using remap, improve bandwidth kbuild test robot
2015-10-23 16:55           ` [lm-sensors] " kbuild test robot
2015-10-23 20:10         ` kbuild test robot
2015-10-23 20:10           ` [lm-sensors] " kbuild test robot

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=56263973.300@roeck-us.net \
    --to=linux@roeck-us.net \
    --cc=bcousson@baylibre.com \
    --cc=jdelvare@suse.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=lm-sensors@lm-sensors.org \
    --cc=mtitinger@baylibre.com \
    --cc=mturquette@baylibre.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.