Hi Alain,
> great that you published your slave-unittest driver.
> I gave it a try rapidly and can confirm it works as expected on the stm32
> platform. I tested with a loopback between 2 i2c controllers.
Thanks, much appreciated! I read this as Tested-by :)
> > +config I2C_SLAVE_TESTUNIT
> > + tristate "I2C eeprom testunit driver"
>
> Should be tristate "I2C testunit driver" I believe (no eeprom)
Oops, yes.
> > + switch (tu->regs[TU_REG_CMD]) {
> > + case TU_CMD_READ_BYTES:
> > + msg.addr = tu->regs[TU_REG_DATAL];
> > + msg.flags = I2C_M_RD;
> > + msg.len = tu->regs[TU_REG_DATAH];
>
> Some protection might be needed here since the size is coming straight from
> the i2cset command while the buffer is limited to 256 bytes.
Well, the size is not a problem because all the regs are u8. But for the
address, we should mask the lower 7 bits, in deed. We could use the free
8th bit to double the length, but we probably better keep it reserved for
something yet unknown.
Thanks for the review!
Wolfram
Attachment:
signature.asc
Description: PGP signature
