| Message ID | 20221221020520.1326964-8-dmitry.baryshkov@linaro.org |
|---|---|
| State | New |
| Headers | show |
| Series | thermal/drivers/tsens: specify nvmem cells in DT rather than parsing them manually | expand |
On 21/12/2022 03:05, Dmitry Baryshkov wrote: > According to the vendor kernels (msm-3.10, 3.14 and 3.18), msm8939 > supports only 9 sensors. Remove the rogue sensor's hw_id. > > Fixes: 332bc8ebab2c ("thermal: qcom: tsens-v0_1: Add support for MSM8939") > Cc: Bryan O'Donoghue <bryan.odonoghue@linaro.org> > Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> > --- > drivers/thermal/qcom/tsens-v0_1.c | 4 ++-- > 1 file changed, 2 insertions(+), 2 deletions(-) > > diff --git a/drivers/thermal/qcom/tsens-v0_1.c b/drivers/thermal/qcom/tsens-v0_1.c > index 0bc4e5cec184..57ac23f9d9b7 100644 > --- a/drivers/thermal/qcom/tsens-v0_1.c > +++ b/drivers/thermal/qcom/tsens-v0_1.c > @@ -605,9 +605,9 @@ static const struct tsens_ops ops_8939 = { > }; > > struct tsens_plat_data data_8939 = { > - .num_sensors = 10, > + .num_sensors = 9, > .ops = &ops_8939, > - .hw_ids = (unsigned int []){ 0, 1, 2, 3, 5, 6, 7, 8, 9, 10 }, > + .hw_ids = (unsigned int []){ 0, 1, 2, 3, 5, 6, 7, 8, 9 }, Does not patch 4 says we can drop those hw_ids ? > > .feat = &tsens_v0_1_feat, > .fields = tsens_v0_1_regfields,
On 21/12/2022 19:05, Dmitry Baryshkov wrote: > On 21/12/2022 17:45, Daniel Lezcano wrote: >> On 21/12/2022 03:05, Dmitry Baryshkov wrote: >>> According to the vendor kernels (msm-3.10, 3.14 and 3.18), msm8939 >>> supports only 9 sensors. Remove the rogue sensor's hw_id. >>> >>> Fixes: 332bc8ebab2c ("thermal: qcom: tsens-v0_1: Add support for >>> MSM8939") >>> Cc: Bryan O'Donoghue <bryan.odonoghue@linaro.org> >>> Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> >>> --- >>> drivers/thermal/qcom/tsens-v0_1.c | 4 ++-- >>> 1 file changed, 2 insertions(+), 2 deletions(-) >>> >>> diff --git a/drivers/thermal/qcom/tsens-v0_1.c >>> b/drivers/thermal/qcom/tsens-v0_1.c >>> index 0bc4e5cec184..57ac23f9d9b7 100644 >>> --- a/drivers/thermal/qcom/tsens-v0_1.c >>> +++ b/drivers/thermal/qcom/tsens-v0_1.c >>> @@ -605,9 +605,9 @@ static const struct tsens_ops ops_8939 = { >>> }; >>> struct tsens_plat_data data_8939 = { >>> - .num_sensors = 10, >>> + .num_sensors = 9, >>> .ops = &ops_8939, >>> - .hw_ids = (unsigned int []){ 0, 1, 2, 3, 5, 6, 7, 8, 9, >>> 10 }, >>> + .hw_ids = (unsigned int []){ 0, 1, 2, 3, 5, 6, 7, 8, 9 }, >> >> Does not patch 4 says we can drop those hw_ids ? > > No. In patch 4 we drop contiguous IDs. For the msm8939 sensor 4 is > omitted from hw_ids, so we can not drop the array. Ah, yes, indeed. .num_sensor = ARRAY_SIZE(data_8939.hw_ids); should work If the hw_ids are ordered, you may consider a bitmask instead of an array
On 21/12/2022 20:15, Daniel Lezcano wrote: > On 21/12/2022 19:05, Dmitry Baryshkov wrote: >> On 21/12/2022 17:45, Daniel Lezcano wrote: >>> On 21/12/2022 03:05, Dmitry Baryshkov wrote: >>>> According to the vendor kernels (msm-3.10, 3.14 and 3.18), msm8939 >>>> supports only 9 sensors. Remove the rogue sensor's hw_id. >>>> >>>> Fixes: 332bc8ebab2c ("thermal: qcom: tsens-v0_1: Add support for >>>> MSM8939") >>>> Cc: Bryan O'Donoghue <bryan.odonoghue@linaro.org> >>>> Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> >>>> --- >>>> drivers/thermal/qcom/tsens-v0_1.c | 4 ++-- >>>> 1 file changed, 2 insertions(+), 2 deletions(-) >>>> >>>> diff --git a/drivers/thermal/qcom/tsens-v0_1.c >>>> b/drivers/thermal/qcom/tsens-v0_1.c >>>> index 0bc4e5cec184..57ac23f9d9b7 100644 >>>> --- a/drivers/thermal/qcom/tsens-v0_1.c >>>> +++ b/drivers/thermal/qcom/tsens-v0_1.c >>>> @@ -605,9 +605,9 @@ static const struct tsens_ops ops_8939 = { >>>> }; >>>> struct tsens_plat_data data_8939 = { >>>> - .num_sensors = 10, >>>> + .num_sensors = 9, >>>> .ops = &ops_8939, >>>> - .hw_ids = (unsigned int []){ 0, 1, 2, 3, 5, 6, 7, 8, 9, >>>> 10 }, >>>> + .hw_ids = (unsigned int []){ 0, 1, 2, 3, 5, 6, 7, 8, 9 }, >>> >>> Does not patch 4 says we can drop those hw_ids ? >> >> No. In patch 4 we drop contiguous IDs. For the msm8939 sensor 4 is >> omitted from hw_ids, so we can not drop the array. > > Ah, yes, indeed. > > .num_sensor = ARRAY_SIZE(data_8939.hw_ids); should work Interesting suggestion, I should give it a thought. > > If the hw_ids are ordered, you may consider a bitmask instead of an array I'm not sure if it's worth doing that. With the current approach it is easy to get hw_id corresponding to the sensor #i, If we switch to the bitmask, it wouldn't be that easy. And the saved space isn't that big.
diff --git a/drivers/thermal/qcom/tsens-v0_1.c b/drivers/thermal/qcom/tsens-v0_1.c index 0bc4e5cec184..57ac23f9d9b7 100644 --- a/drivers/thermal/qcom/tsens-v0_1.c +++ b/drivers/thermal/qcom/tsens-v0_1.c @@ -605,9 +605,9 @@ static const struct tsens_ops ops_8939 = { }; struct tsens_plat_data data_8939 = { - .num_sensors = 10, + .num_sensors = 9, .ops = &ops_8939, - .hw_ids = (unsigned int []){ 0, 1, 2, 3, 5, 6, 7, 8, 9, 10 }, + .hw_ids = (unsigned int []){ 0, 1, 2, 3, 5, 6, 7, 8, 9 }, .feat = &tsens_v0_1_feat, .fields = tsens_v0_1_regfields,
According to the vendor kernels (msm-3.10, 3.14 and 3.18), msm8939 supports only 9 sensors. Remove the rogue sensor's hw_id. Fixes: 332bc8ebab2c ("thermal: qcom: tsens-v0_1: Add support for MSM8939") Cc: Bryan O'Donoghue <bryan.odonoghue@linaro.org> Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> --- drivers/thermal/qcom/tsens-v0_1.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)