Hi Akash,
On Wed, Feb 19, 2020 at 07:17:44PM +0530, Akash Asthana wrote:
> Hi Matthias,
>
> On 2/19/2020 4:17 AM, Matthias Kaehlcke wrote:
> > On Mon, Feb 17, 2020 at 07:00:02PM +0530, Akash Asthana wrote:
> > > Get the interconnect paths for I2C based Serial Engine device
> > > and vote according to the bus speed of the driver.
> > >
> > > Signed-off-by: Akash Asthana <akashast@xxxxxxxxxxxxxx>
> > > ---
> > > drivers/i2c/busses/i2c-qcom-geni.c | 84 ++++++++++++++++++++++++++++++++++++--
> > > 1 file changed, 80 insertions(+), 4 deletions(-)
> > >
> > > diff --git a/drivers/i2c/busses/i2c-qcom-geni.c b/drivers/i2c/busses/i2c-qcom-geni.c
> > > index 17abf60c..5de10a1 100644
> > > --- a/drivers/i2c/busses/i2c-qcom-geni.c
> > > +++ b/drivers/i2c/busses/i2c-qcom-geni.c
> > >
> > > static void geni_i2c_err_misc(struct geni_i2c_dev *gi2c)
> > > {
> > > u32 m_cmd = readl_relaxed(gi2c->se.base + SE_GENI_M_CMD0);
> > > @@ -563,17 +601,34 @@ static int geni_i2c_probe(struct platform_device *pdev)
> > > gi2c->adap.dev.of_node = pdev->dev.of_node;
> > > strlcpy(gi2c->adap.name, "Geni-I2C", sizeof(gi2c->adap.name));
> > > + ret = geni_i2c_icc_get(&gi2c->se);
> > > + if (ret)
> > > + return ret;
> > > + /* Set the bus quota to a reasonable value */
> > > + gi2c->se.avg_bw_core = Bps_to_icc(1000);
> > > + gi2c->se.peak_bw_core = Bps_to_icc(CORE_2X_100_MHZ);
> > > + gi2c->se.avg_bw_cpu = Bps_to_icc(1000);
> > > + gi2c->se.peak_bw_cpu = Bps_to_icc(1000);
> > > + gi2c->se.avg_bw_ddr = Bps_to_icc(gi2c->clk_freq_out);
> > > + gi2c->se.peak_bw_ddr = Bps_to_icc(2 * gi2c->clk_freq_out);
> > > +
> > > + /* Vote for core clocks and CPU for register access */
> > > + icc_set_bw(gi2c->se.icc_path[GENI_TO_CORE], gi2c->se.avg_bw_core,
> > > + gi2c->se.peak_bw_core);
> > > + icc_set_bw(gi2c->se.icc_path[CPU_TO_GENI], gi2c->se.avg_bw_cpu,
> > > + gi2c->se.peak_bw_cpu);
> > error handling needed?
>
> I will add error handling for GENI_TO_CORE path in all the drivers. Will it
> be okay if we don't handle errors for CPU_TO_GENI and GENI_TO_DDR path
>
> as CPU and DDR will be running at much higher frequency?
It may still work, but you might never know that there was a problem. I
would be inclined to check the return value of all invocations of icc_set_bw()
- including runtime suspend/resume - and log a message if a problem is
detected. For runtime suspend/resume it would probably be wise to use
dev_err_ratelimited(), to avoid spamming the system log too much in case of
a persistent problem.
If others think that error checking all icc_set_bw() calls is overkill
please speak up :)
Thanks
Matthias
