| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <YYVNJR0oAGuvOv6t@piout.net>
Date: Fri, 5 Nov 2021 16:26:29 +0100
From: Alexandre Belloni <alexandre.belloni@...tlin.com>
To: Pavel Modilaynen <pavelmn@...s.com>
Cc: Pavel Modilaynen <Pavel.Modilaynen@...s.com>,
"a.zummo@...ertech.it" <a.zummo@...ertech.it>,
"robh+dt@...nel.org" <robh+dt@...nel.org>,
"linux-rtc@...r.kernel.org" <linux-rtc@...r.kernel.org>,
"devicetree@...r.kernel.org" <devicetree@...r.kernel.org>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
kernel <kernel@...s.com>
Subject: Re: [PATCH 1/2] rtc: rs5c372: Add support for trim configuration
On 02/11/2021 00:14:39+0100, Pavel Modilaynen wrote:
> Hi Alexandre,
>
> On 11/1/21 7:23 PM, Alexandre Belloni wrote:
> > On 31/10/2021 11:29:12+0100, Pavel Modilaynen wrote:
> > > On 10/31/21 12:57 AM, Alexandre Belloni wrote:
> > > > Hello,
> > > > > Please use the proper RTC interface by implementing .set_offset
> > > and
> > > > .read_offset.
> > >
> > > I am not sure about .set/read_offset. It looks as runtime adjustment
> > > interface,
> > > however this Xtal trimming parameter is based on schematics and Xtal
> > > capacitance (datasheet parameter).
> > > It is found by calibration procedure based on RTC clock output (the
> > > procedure and calculation of trimming parameter is described in datasheets).
> > > So, I would like to say that this parameter is functionally close to
> > > "quartz-load-femtofarads" for rtc-pcf8523/pcf85063.
> > >
> >
> > quartz-load-femtofarads is for analog trimming which this RTC doesn't
> > have, both CD and CG are set to 10pF. .set/read_offset are for digital
> > trimming which is what you are configuring here. You definitively want
> > to be able to do that at runtime as you need to adjust for temperature
> > and ageing of the crystal (datasheet, page 14: "For those systems that
> > have temperature detection precision of clock function may be increased
> > by correcting clock error according to temperature fluctuations.")
> >
>
> Thank you for reply.
>
> I am not denying the need in runtime adjustment related to
> temperature, aging and precision, which you are referring by excerpt from
> p.14.
>
> I would like to make a point that Xtal trimming is for coarse grained
> adjustment (pages 36-39), primarily related to Xtal capacitance CL (not
> CD/CG). Our goal is to keep a reasonable drift of ,say, <1 second per day
> and for Xtal that we use with 12.5pF RTC manufacturer recommends using 0x23
> value for adjustment. In this case we act according to (A) course from page
> 38:
>
> "Adjustment of clock is not made for IC (no adjustment) and any CL value may
> be used for the crystal oscillator. Precision fluctuations of a crystal
> oscillator may be selected as long as clock precision allows. Obtain the
> central frequency as described in section 2.2 using several crystal
> oscillator and ICs, determine an adjustment value as
> described in “2.4 Time Trimming Circuit” which shall be set to the
> RS5C372A/B."
>
This doesn't prevent you from using the .set/read_offset interface.
Simply have a startup script that forcibly set the hardcoded value if
this is what you want. However, I'm pretty sure your products have NTP
which allows to precisely set a calculated value instead of an hardcoded
one.
>
> > > > > On 31/10/2021 00:50:53+0200, Pavel Modilaynen wrote:
> > > > > From: Pavel Modilaynen <pavelmn@...s.com>
> > > > > > > Add support for oscillation adjustment register
> > > RS5C372_REG_TRIM
> > > > > setting that is needed to accommodate for effective crystal
> > > > > capacitance.
> > > > > > > Use optional property ricoh,trim that should contain
> > > > > raw value to setup this register. According to
> > > > > datasheets for RS5C372, R2025S/D, RV5C38[67] and R222[13]
> > > > > the value will be converted to a number of ticks that
> > > > > is to be subtracted or added when the second digits read
> > > > > 00, 20 or 40 seconds.
> > > > > > > Signed-off-by: Pavel Modilaynen <pavelmn@...s.com>
> > > > > ---
> > > > > drivers/rtc/rtc-rs5c372.c | 18 +++++++++++++++++-
> > > > > 1 file changed, 17 insertions(+), 1 deletion(-)
> > > > > > > diff --git a/drivers/rtc/rtc-rs5c372.c
> > > b/drivers/rtc/rtc-rs5c372.c
> > > > > index 80980414890c..3a2db0326669 100644
> > > > > --- a/drivers/rtc/rtc-rs5c372.c
> > > > > +++ b/drivers/rtc/rtc-rs5c372.c
> > > > > @@ -13,6 +13,7 @@
> > > > > #include <linux/slab.h>
> > > > > #include <linux/module.h>
> > > > > #include <linux/of_device.h>
> > > > > +#include <linux/of.h>
> > > > > /*
> > > > > * Ricoh has a family of I2C based RTCs, which differ only slightly from
> > > > > @@ -560,6 +561,8 @@ static int rs5c_oscillator_setup(struct rs5c372 *rs5c372)
> > > > > {
> > > > > unsigned char buf[2];
> > > > > int addr, i, ret = 0;
> > > > > + struct i2c_client *client = rs5c372->client;
> > > > > + u8 trim = 0;
> > > > > addr = RS5C_ADDR(RS5C_REG_CTRL1);
> > > > > buf[0] = rs5c372->regs[RS5C_REG_CTRL1];
> > > > > @@ -599,9 +602,22 @@ static int rs5c_oscillator_setup(struct rs5c372 *rs5c372)
> > > > > break;
> > > > > }
> > > > > + /* optional setup of xtal trimming */
> > > > > + if (!of_property_read_u8(client->dev.of_node, "ricoh,trim", &trim)) {
> > > > > + if (rs5c372->type != rtc_r2221tl && (trim & ~RS5C372_TRIM_MASK)) {
> > > > > + dev_warn(&client->dev, "Erroneous setting for ricoh,trim in devicetree\n");
> > > > > + } else {
> > > > > + int addr = RS5C_ADDR(RS5C372_REG_TRIM);
> > > > > + int ret = i2c_smbus_write_byte_data(client, addr, trim);
> > > > > +
> > > > > + if (unlikely(ret < 0))
> > > > > + return ret;
> > > > > + }
> > > > > + }
> > > > > +
> > > > > for (i = 0; i < sizeof(buf); i++) {
> > > > > addr = RS5C_ADDR(RS5C_REG_CTRL1 + i);
> > > > > - ret = i2c_smbus_write_byte_data(rs5c372->client, addr, buf[i]);
> > > > > + ret = i2c_smbus_write_byte_data(client, addr, buf[i]);
> > > > > if (unlikely(ret < 0))
> > > > > return ret;
> > > > > }
> > > > > -- > > 2.20.1
> > > > > > > -- > Alexandre Belloni, co-owner and COO, Bootlin
> > > > Embedded Linux and Kernel engineering
> > > > https://bootlin.com <https://bootlin.com> <https://bootlin.com
> > <https://bootlin.com>>
> >
> > --
> > Alexandre Belloni, co-owner and COO, Bootlin
> > Embedded Linux and Kernel engineering
> > https://bootlin.com <https://bootlin.com>
--
Alexandre Belloni, co-owner and COO, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com
Powered by blists - more mailing lists