Re: [PATCH 11/21] test/avocado/machine_aspeed.py: Add an I2C RTC test

[Cédric Le Goater]

On 6/7/22 01:16, Joel Stanley wrote:
> On Mon, 6 Jun 2022 at 15:08, Cédric Le Goater <clg@kaod.org> wrote:
> > Add a RTC device on bus 15 and check that the ouput of the hwclock
>
> spelling: output
>
> > command matches the current year.
> >
> > Signed-off-by: Cédric Le Goater <clg@kaod.org>
>
> Reviewed-by: Joel Stanley <joel@jms.id.au>
>
> > ---
> >   tests/avocado/machine_aspeed.py | 8 ++++++++
> >   1 file changed, 8 insertions(+)
> >
> > diff --git a/tests/avocado/machine_aspeed.py b/tests/avocado/machine_aspeed.py
> > index a3b4b9e5093c..28b8a4c8124b 100644
> > --- a/tests/avocado/machine_aspeed.py
> > +++ b/tests/avocado/machine_aspeed.py
> > @@ -136,10 +136,18 @@ def test_arm_ast2600_evb_builroot(self):
> >
> >           self.vm.add_args('-device',
> >                            'tmp423,bus=aspeed.i2c.bus.15,address=0x4c');
> > +        self.vm.add_args('-device',
> > +                         'ds1338,bus=aspeed.i2c.bus.15,address=0x32');
>
> Is there any value running this on the 2400 and 2500 machine types
> too?

We could do that, yes. Send patches !

> They all use the same model so perhaps not?

Currently, all models are exercised more or less in the same way by
the upstream Linux driver. Things are different for the AST1030 using
Zephir and for the AST2600 when using newer drivers from the SDK.

These images seem to be using the new AST2600 register mode:

  https://github.com/AspeedTech-BMC/openbmc/releases/

  root@ast2600-default:~# dmesg | grep i2c
  [    0.211289] i2c global registered
  [    1.442027] i2c_dev: i2c /dev entries driver
  [    1.447944] i2c_new_aspeed 1e78a280.i2c-bus: NEW-I2C: i2c-bus [4]: adapter 
[100 khz] mode [2]
  [    1.451158] ipmb-dev 5-0010: i2c_slave_register: client slave flag not 
set. You might see address collisions
  [    1.451660] i2c_new_aspeed 1e78a300.i2c-bus: NEW-I2C: i2c-bus [5]: adapter 
[100 khz] mode [2]
  [    1.454567] ipmb-dev 6-0012: i2c_slave_register: client slave flag not 
set. You might see address collisions
  [    1.454938] i2c_new_aspeed 1e78a380.i2c-bus: NEW-I2C: i2c-bus [6]: adapter 
[100 khz] mode [2]
  [    1.462953] i2c_new_aspeed 1e78a400.i2c-bus: NEW-I2C: i2c-bus [7]: adapter 
[95 khz] mode [2]
  [    1.466394] i2c_new_aspeed 1e78a480.i2c-bus: NEW-I2C: i2c-bus [8]: adapter 
[100 khz] mode [2]
  [    1.468394] i2c_new_aspeed 1e78a500.i2c-bus: NEW-I2C: i2c-bus [9]: adapter 
[100 khz] mode [2]

It could be an additional avocado test.


C.


> >           self.do_test_arm_aspeed_buidroot_start(image_path, '0xf00')
> >           exec_command_and_wait_for_pattern(self,
> >                                             'i2cget -y 15 0x4c 0xff', '0x23');
> >           exec_command_and_wait_for_pattern(self,
> >                                             'i2cget -y 15 0x4c 0xfe', '0x55');
> >
> > +        exec_command_and_wait_for_pattern(self,
> > +             'echo ds1307 0x32 > /sys/class/i2c-dev/i2c-15/device/new_device',
> > +             'i2c i2c-15: new_device: Instantiated device ds1307 at 0x32');
> > +        year = time.strftime("%Y")
> > +        exec_command_and_wait_for_pattern(self, 'hwclock -f /dev/rtc1', year);
> > +
> >           self.do_test_arm_aspeed_buidroot_poweroff()
> > --
> > 2.35.3