Re: [PATCH v2 5/7] hw/nvram: Update at24c EEPROM init function in NPCM7xx boards

[Hao Wu]

Thanks for the comment. I'll upload a new version using IF_OTHER as discussed in that thread.

I don't know if adding the assertion for type != IF_NONE is a good idea now so I didn't add that.

If you think that's good I can add it as well.

On Wed, Nov 3, 2021 at 2:13 AM Thomas Huth <thuth@redhat.com> wrote:

On 01/11/2021 18.47, Hao Wu wrote:
>
>
> On Mon, Nov 1, 2021 at 10:41 AM Peter Maydell <peter.maydell@linaro.org
> <mailto:peter.maydell@linaro.org>> wrote:
>
>     On Thu, 21 Oct 2021 at 19:40, Hao Wu <wuhaotsh@google.com
>     <mailto:wuhaotsh@google.com>> wrote:
>      >
>      > We made 3 changes to the at24c_eeprom_init function in
>      > npcm7xx_boards.c:
>      >
>      > 1. We allow the function to take a I2CBus* as parameter. This allows
>      >    us to attach an EEPROM device behind an I2C mux which is not
>      >    possible with the old method.
>      >
>      > 2. We make at24c EEPROMs are backed by drives so that we can
>      >    specify the content of the EEPROMs.
>      >
>      > 3. Instead of using i2c address as unit number, This patch assigns
>      >    unique unit numbers for each eeproms in each board. This avoids
>      >    conflict in providing multiple eeprom contents with the same address.
>      >    In the old method if we specify two drives with the same unit number,
>      >    the following error will occur: `Device with id 'none85' exists`.
>      >
>      > Signed-off-by: Hao Wu <wuhaotsh@google.com <mailto:wuhaotsh@google.com>>
>      > ---
>      >  hw/arm/npcm7xx_boards.c | 15 ++++++++++-----
>      >  1 file changed, 10 insertions(+), 5 deletions(-)
>      >
>      > diff --git a/hw/arm/npcm7xx_boards.c b/hw/arm/npcm7xx_boards.c
>      > index a656169f61..cdb52b9922 100644
>      > --- a/hw/arm/npcm7xx_boards.c
>      > +++ b/hw/arm/npcm7xx_boards.c
>      > @@ -107,13 +107,18 @@ static I2CBus *npcm7xx_i2c_get_bus(NPCM7xxState
>     *soc, uint32_t num)
>      >      return I2C_BUS(qdev_get_child_bus(DEVICE(&soc->smbus[num]),
>     "i2c-bus"));
>      >  }
>      >
>      > -static void at24c_eeprom_init(NPCM7xxState *soc, int bus, uint8_t addr,
>      > -                              uint32_t rsize)
>      > +static void at24c_eeprom_init(I2CBus *i2c_bus, int bus, uint8_t addr,
>      > +                              uint32_t rsize, int unit_number)
>      >  {
>      > -    I2CBus *i2c_bus = npcm7xx_i2c_get_bus(soc, bus);
>      >      I2CSlave *i2c_dev = i2c_slave_new("at24c-eeprom", addr);
>      >      DeviceState *dev = DEVICE(i2c_dev);
>      > +    BlockInterfaceType type = IF_NONE;
>
>     Why make this a variable? We only use it in one place...
>
> You're right, we can actually inline it.

Actually, please do *not* use IF_NONE for such stuff. See the discussion
here for details:

  https://lists.gnu.org/archive/html/qemu-devel/2021-11/msg00970.html

  Thomas