| Message ID | 1347637215-4830-4-git-send-email-l.majewski@samsung.com |
|---|---|
| State | Superseded |
| Delegated to: | Minkyu Kang |
| Headers | show |
On 14/09/2012 17:40, Lukasz Majewski wrote: > PMIC MAX8997 is now ready to work with single and multibus soft I2C > implementation. > > Signed-off-by: Lukasz Majewski <l.majewski@samsung.com> > Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com> > --- > drivers/misc/pmic_max8997.c | 3 ++- > 1 files changed, 2 insertions(+), 1 deletions(-) > > diff --git a/drivers/misc/pmic_max8997.c b/drivers/misc/pmic_max8997.c > index 62dbc05..4943f66 100644 > --- a/drivers/misc/pmic_max8997.c > +++ b/drivers/misc/pmic_max8997.c > @@ -24,6 +24,7 @@ > #include <common.h> > #include <pmic.h> > #include <max8997_pmic.h> > +#include <i2c.h> > > int pmic_init(void) > { > @@ -37,7 +38,7 @@ int pmic_init(void) > p->number_of_regs = PMIC_NUM_OF_REGS; > p->hw.i2c.addr = MAX8997_I2C_ADDR; > p->hw.i2c.tx_num = 1; > - p->bus = I2C_PMIC; > + p->bus = I2C_0; > I do not see so useful to add an enum for each instance of the I2C bus. And we have to add it if the number of i2c busses grows. IMHO it is better to use directly the constant, so later in another patch pmic_init(5) instead of pmic(I2C_5). Regards, Stefano Babic
Hi Stefano, > On 14/09/2012 17:40, Lukasz Majewski wrote: > > PMIC MAX8997 is now ready to work with single and multibus soft I2C > > implementation. > > > > Signed-off-by: Lukasz Majewski <l.majewski@samsung.com> > > Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com> > > --- > > drivers/misc/pmic_max8997.c | 3 ++- > > 1 files changed, 2 insertions(+), 1 deletions(-) > > > > diff --git a/drivers/misc/pmic_max8997.c > > b/drivers/misc/pmic_max8997.c index 62dbc05..4943f66 100644 > > --- a/drivers/misc/pmic_max8997.c > > +++ b/drivers/misc/pmic_max8997.c > > @@ -24,6 +24,7 @@ > > #include <common.h> > > #include <pmic.h> > > #include <max8997_pmic.h> > > +#include <i2c.h> > > > > int pmic_init(void) > > { > > @@ -37,7 +38,7 @@ int pmic_init(void) > > p->number_of_regs = PMIC_NUM_OF_REGS; > > p->hw.i2c.addr = MAX8997_I2C_ADDR; > > p->hw.i2c.tx_num = 1; > > - p->bus = I2C_PMIC; > > + p->bus = I2C_0; > > > > I do not see so useful to add an enum for each instance of the I2C > bus. And we have to add it if the number of i2c busses grows. IMHO it > is better to use directly the constant, so later in another patch > pmic_init(5) instead of pmic(I2C_5). > This problem has been already discussed with Heiko: http://patchwork.ozlabs.org/patch/181789/ I think, that we will have to have an enum of available I2C_x busses at <i2c.h> header file. In the case of Trats and PMIC framework, the I2C_0 is done on purpose to keep the trats board working (since the PMIC is in reality connected to I2C_5). This numbering (I2C_0) needs to be there until prerequisite patches aren't accepted (the multibus I2C support on trats board) Regards, Lukasz Majewski
diff --git a/drivers/misc/pmic_max8997.c b/drivers/misc/pmic_max8997.c index 62dbc05..4943f66 100644 --- a/drivers/misc/pmic_max8997.c +++ b/drivers/misc/pmic_max8997.c @@ -24,6 +24,7 @@ #include <common.h> #include <pmic.h> #include <max8997_pmic.h> +#include <i2c.h> int pmic_init(void) { @@ -37,7 +38,7 @@ int pmic_init(void) p->number_of_regs = PMIC_NUM_OF_REGS; p->hw.i2c.addr = MAX8997_I2C_ADDR; p->hw.i2c.tx_num = 1; - p->bus = I2C_PMIC; + p->bus = I2C_0; return 0; }