On Tue, 18 Dec 2012, Jean Delvare wrote: > Hi Julia, > > On Thu, 11 Oct 2012 08:45:43 +0200 (CEST), Julia Lawall wrote: > > I found 6 cases where there are more than 2 messages in the array. I > > didn't check how many cases where there are two messages but there is > > something other than one read and one write. > > > > Perhaps a reasonable option would be to use > > > > I2C_MSG_READ > > I2C_MSG_WRITE > > I2C_MSG_READ_OP > > I2C_MSG_WRITE_OP > > > > The last two are for the few cases where more flags are specified. As > > compared to the original proposal of I2C_MSG_OP, these keep the READ or > > WRITE idea in the macro name. The additional argument to the OP macros > > would be or'd with the read or write (nothing to do in this case) flags as > > appropriate. > > > > Mauro proposed INIT_I2C_READ_SUBADDR for the very common case where a > > message array has one read and one write. I think that putting one > > I2C_MSG_READ and one I2C_MSG_WRITE in this case is readable enough, and > > avoids the need to do something special for the cases that don't match the > > expectations of INIT_I2C_READ_SUBADDR. > > > > I propose not to do anything for the moment either for sizes or for > > message or buffer arrays that contain only one element. > > Please note that I resigned from my position of i2c subsystem > maintainer, so I will not handle this. If you think this is important, > you'll have to resubmit and Wolfram will decide what he wants to do > about it. OK, I had the impression that the conclusion was that the danger was greater than the benefit. If there is interest in it, since I think it does make the code more readable, I can pick it up again. julia -- To unsubscribe from this list: send the line "unsubscribe linux-i2c" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
