| Message ID | 20230228-topic-qos-v7-2-815606092fff@linaro.org (mailing list archive) |
|---|---|
| State | Superseded |
| Headers | show |
| Series | The great interconnecification fixation | expand |
On 08/03/2023 23:40, Konrad Dybcio wrote: > Some nodes, like EBI0 (DDR) or L3/LLCC, may be connected over more than > one channel. This should be taken into account in bandwidth calcualtion, > as we're supposed to feed msmbus with the per-channel bandwidth. Add > support for specifying that and use it during bandwidth aggregation. > > Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org> > Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org> > --- > drivers/interconnect/qcom/icc-rpm.c | 7 ++++++- > drivers/interconnect/qcom/icc-rpm.h | 2 ++ > 2 files changed, 8 insertions(+), 1 deletion(-) Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Hi Konrad, Thanks for the patch! On 8.03.23 23:40, Konrad Dybcio wrote: > Some nodes, like EBI0 (DDR) or L3/LLCC, may be connected over more than > one channel. This should be taken into account in bandwidth calcualtion, > as we're supposed to feed msmbus with the per-channel bandwidth. Add > support for specifying that and use it during bandwidth aggregation. > This looks good, but do you have any follow-up patch to use this and set the channels in some driver? BR, Georgi > Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org> > Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org> > --- > drivers/interconnect/qcom/icc-rpm.c | 7 ++++++- > drivers/interconnect/qcom/icc-rpm.h | 2 ++ > 2 files changed, 8 insertions(+), 1 deletion(-) > > diff --git a/drivers/interconnect/qcom/icc-rpm.c b/drivers/interconnect/qcom/icc-rpm.c > index 35fd75ae70e3..27c4c6497994 100644 > --- a/drivers/interconnect/qcom/icc-rpm.c > +++ b/drivers/interconnect/qcom/icc-rpm.c > @@ -317,6 +317,7 @@ static void qcom_icc_bus_aggregate(struct icc_provider *provider, > { > struct icc_node *node; > struct qcom_icc_node *qn; > + u64 sum_avg[QCOM_ICC_NUM_BUCKETS]; > int i; > > /* Initialise aggregate values */ > @@ -334,7 +335,11 @@ static void qcom_icc_bus_aggregate(struct icc_provider *provider, > list_for_each_entry(node, &provider->nodes, node_list) { > qn = node->data; > for (i = 0; i < QCOM_ICC_NUM_BUCKETS; i++) { > - agg_avg[i] += qn->sum_avg[i]; > + if (qn->channels) > + sum_avg[i] = div_u64(qn->sum_avg[i], qn->channels); > + else > + sum_avg[i] = qn->sum_avg[i]; > + agg_avg[i] += sum_avg[i]; > agg_peak[i] = max_t(u64, agg_peak[i], qn->max_peak[i]); > } > } > diff --git a/drivers/interconnect/qcom/icc-rpm.h b/drivers/interconnect/qcom/icc-rpm.h > index 8ba1918d7997..8aed5400afda 100644 > --- a/drivers/interconnect/qcom/icc-rpm.h > +++ b/drivers/interconnect/qcom/icc-rpm.h > @@ -66,6 +66,7 @@ struct qcom_icc_qos { > * @id: a unique node identifier > * @links: an array of nodes where we can go next while traversing > * @num_links: the total number of @links > + * @channels: number of channels at this node (e.g. DDR channels) > * @buswidth: width of the interconnect between a node and the bus (bytes) > * @sum_avg: current sum aggregate value of all avg bw requests > * @max_peak: current max aggregate value of all peak bw requests > @@ -78,6 +79,7 @@ struct qcom_icc_node { > u16 id; > const u16 *links; > u16 num_links; > + u16 channels; > u16 buswidth; > u64 sum_avg[QCOM_ICC_NUM_BUCKETS]; > u64 max_peak[QCOM_ICC_NUM_BUCKETS]; >
On 21.03.2023 15:06, Georgi Djakov wrote: > Hi Konrad, > > Thanks for the patch! > > On 8.03.23 23:40, Konrad Dybcio wrote: >> Some nodes, like EBI0 (DDR) or L3/LLCC, may be connected over more than >> one channel. This should be taken into account in bandwidth calcualtion, >> as we're supposed to feed msmbus with the per-channel bandwidth. Add >> support for specifying that and use it during bandwidth aggregation. >> > > This looks good, but do you have any follow-up patch to use this and set > the channels in some driver? Yes, I have a couple of OOT drivers that are gonna make use of it. TBF it should have been sent separately from the QoS mess, but I don't think it's much of an issue to take it as-is. The aforementioned OOT drivers for MSM8998 and SM6375 will be submitted after we reach a consensus on how we want to ensure that each node is guaranteed to have its clocks enabled before access, among some other minor things. Konrad > > BR, > Georgi > >> Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org> >> Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org> >> --- >> drivers/interconnect/qcom/icc-rpm.c | 7 ++++++- >> drivers/interconnect/qcom/icc-rpm.h | 2 ++ >> 2 files changed, 8 insertions(+), 1 deletion(-) >> >> diff --git a/drivers/interconnect/qcom/icc-rpm.c b/drivers/interconnect/qcom/icc-rpm.c >> index 35fd75ae70e3..27c4c6497994 100644 >> --- a/drivers/interconnect/qcom/icc-rpm.c >> +++ b/drivers/interconnect/qcom/icc-rpm.c >> @@ -317,6 +317,7 @@ static void qcom_icc_bus_aggregate(struct icc_provider *provider, >> { >> struct icc_node *node; >> struct qcom_icc_node *qn; >> + u64 sum_avg[QCOM_ICC_NUM_BUCKETS]; >> int i; >> /* Initialise aggregate values */ >> @@ -334,7 +335,11 @@ static void qcom_icc_bus_aggregate(struct icc_provider *provider, >> list_for_each_entry(node, &provider->nodes, node_list) { >> qn = node->data; >> for (i = 0; i < QCOM_ICC_NUM_BUCKETS; i++) { >> - agg_avg[i] += qn->sum_avg[i]; >> + if (qn->channels) >> + sum_avg[i] = div_u64(qn->sum_avg[i], qn->channels); >> + else >> + sum_avg[i] = qn->sum_avg[i]; >> + agg_avg[i] += sum_avg[i]; >> agg_peak[i] = max_t(u64, agg_peak[i], qn->max_peak[i]); >> } >> } >> diff --git a/drivers/interconnect/qcom/icc-rpm.h b/drivers/interconnect/qcom/icc-rpm.h >> index 8ba1918d7997..8aed5400afda 100644 >> --- a/drivers/interconnect/qcom/icc-rpm.h >> +++ b/drivers/interconnect/qcom/icc-rpm.h >> @@ -66,6 +66,7 @@ struct qcom_icc_qos { >> * @id: a unique node identifier >> * @links: an array of nodes where we can go next while traversing >> * @num_links: the total number of @links >> + * @channels: number of channels at this node (e.g. DDR channels) >> * @buswidth: width of the interconnect between a node and the bus (bytes) >> * @sum_avg: current sum aggregate value of all avg bw requests >> * @max_peak: current max aggregate value of all peak bw requests >> @@ -78,6 +79,7 @@ struct qcom_icc_node { >> u16 id; >> const u16 *links; >> u16 num_links; >> + u16 channels; >> u16 buswidth; >> u64 sum_avg[QCOM_ICC_NUM_BUCKETS]; >> u64 max_peak[QCOM_ICC_NUM_BUCKETS]; >> >
On 21.03.23 16:09, Konrad Dybcio wrote: > > On 21.03.2023 15:06, Georgi Djakov wrote: >> Hi Konrad, >> >> Thanks for the patch! >> >> On 8.03.23 23:40, Konrad Dybcio wrote: >>> Some nodes, like EBI0 (DDR) or L3/LLCC, may be connected over more than >>> one channel. This should be taken into account in bandwidth calcualtion, >>> as we're supposed to feed msmbus with the per-channel bandwidth. Add >>> support for specifying that and use it during bandwidth aggregation. >>> >> >> This looks good, but do you have any follow-up patch to use this and set >> the channels in some driver? > Yes, I have a couple of OOT drivers that are gonna make use of it. > TBF it should have been sent separately from the QoS mess, but I > don't think it's much of an issue to take it as-is. > > The aforementioned OOT drivers for MSM8998 and SM6375 will be > submitted after we reach a consensus on how we want to ensure > that each node is guaranteed to have its clocks enabled before > access, among some other minor things. Yes, these QoS clocks are confusing. Maybe you can even submit them without configuring any QoS stuff in first place? Does enabling QoS actually show any benefits on these devices? Thanks, Georgi > Konrad >> >> BR, >> Georgi >> >>> Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org> >>> Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org> >>> --- >>> drivers/interconnect/qcom/icc-rpm.c | 7 ++++++- >>> drivers/interconnect/qcom/icc-rpm.h | 2 ++ >>> 2 files changed, 8 insertions(+), 1 deletion(-) >>> >>> diff --git a/drivers/interconnect/qcom/icc-rpm.c b/drivers/interconnect/qcom/icc-rpm.c >>> index 35fd75ae70e3..27c4c6497994 100644 >>> --- a/drivers/interconnect/qcom/icc-rpm.c >>> +++ b/drivers/interconnect/qcom/icc-rpm.c >>> @@ -317,6 +317,7 @@ static void qcom_icc_bus_aggregate(struct icc_provider *provider, >>> { >>> struct icc_node *node; >>> struct qcom_icc_node *qn; >>> + u64 sum_avg[QCOM_ICC_NUM_BUCKETS]; >>> int i; >>> /* Initialise aggregate values */ >>> @@ -334,7 +335,11 @@ static void qcom_icc_bus_aggregate(struct icc_provider *provider, >>> list_for_each_entry(node, &provider->nodes, node_list) { >>> qn = node->data; >>> for (i = 0; i < QCOM_ICC_NUM_BUCKETS; i++) { >>> - agg_avg[i] += qn->sum_avg[i]; >>> + if (qn->channels) >>> + sum_avg[i] = div_u64(qn->sum_avg[i], qn->channels); >>> + else >>> + sum_avg[i] = qn->sum_avg[i]; >>> + agg_avg[i] += sum_avg[i]; >>> agg_peak[i] = max_t(u64, agg_peak[i], qn->max_peak[i]); >>> } >>> } >>> diff --git a/drivers/interconnect/qcom/icc-rpm.h b/drivers/interconnect/qcom/icc-rpm.h >>> index 8ba1918d7997..8aed5400afda 100644 >>> --- a/drivers/interconnect/qcom/icc-rpm.h >>> +++ b/drivers/interconnect/qcom/icc-rpm.h >>> @@ -66,6 +66,7 @@ struct qcom_icc_qos { >>> * @id: a unique node identifier >>> * @links: an array of nodes where we can go next while traversing >>> * @num_links: the total number of @links >>> + * @channels: number of channels at this node (e.g. DDR channels) >>> * @buswidth: width of the interconnect between a node and the bus (bytes) >>> * @sum_avg: current sum aggregate value of all avg bw requests >>> * @max_peak: current max aggregate value of all peak bw requests >>> @@ -78,6 +79,7 @@ struct qcom_icc_node { >>> u16 id; >>> const u16 *links; >>> u16 num_links; >>> + u16 channels; >>> u16 buswidth; >>> u64 sum_avg[QCOM_ICC_NUM_BUCKETS]; >>> u64 max_peak[QCOM_ICC_NUM_BUCKETS]; >>> >>
On 21.03.2023 15:21, Georgi Djakov wrote: > On 21.03.23 16:09, Konrad Dybcio wrote: >> >> On 21.03.2023 15:06, Georgi Djakov wrote: >>> Hi Konrad, >>> >>> Thanks for the patch! >>> >>> On 8.03.23 23:40, Konrad Dybcio wrote: >>>> Some nodes, like EBI0 (DDR) or L3/LLCC, may be connected over more than >>>> one channel. This should be taken into account in bandwidth calcualtion, >>>> as we're supposed to feed msmbus with the per-channel bandwidth. Add >>>> support for specifying that and use it during bandwidth aggregation. >>>> >>> >>> This looks good, but do you have any follow-up patch to use this and set >>> the channels in some driver? >> Yes, I have a couple of OOT drivers that are gonna make use of it. >> TBF it should have been sent separately from the QoS mess, but I >> don't think it's much of an issue to take it as-is. >> >> The aforementioned OOT drivers for MSM8998 and SM6375 will be >> submitted after we reach a consensus on how we want to ensure >> that each node is guaranteed to have its clocks enabled before >> access, among some other minor things. > > Yes, these QoS clocks are confusing. Maybe you can even submit them > without configuring any QoS stuff in first place? Does enabling QoS > actually show any benefits on these devices? Haven't tested that thoroughly to be honest. But I'll try to get some numbers. Konrad > > Thanks, > Georgi > >> Konrad >>> >>> BR, >>> Georgi >>> >>>> Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org> >>>> Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org> >>>> --- >>>> drivers/interconnect/qcom/icc-rpm.c | 7 ++++++- >>>> drivers/interconnect/qcom/icc-rpm.h | 2 ++ >>>> 2 files changed, 8 insertions(+), 1 deletion(-) >>>> >>>> diff --git a/drivers/interconnect/qcom/icc-rpm.c b/drivers/interconnect/qcom/icc-rpm.c >>>> index 35fd75ae70e3..27c4c6497994 100644 >>>> --- a/drivers/interconnect/qcom/icc-rpm.c >>>> +++ b/drivers/interconnect/qcom/icc-rpm.c >>>> @@ -317,6 +317,7 @@ static void qcom_icc_bus_aggregate(struct icc_provider *provider, >>>> { >>>> struct icc_node *node; >>>> struct qcom_icc_node *qn; >>>> + u64 sum_avg[QCOM_ICC_NUM_BUCKETS]; >>>> int i; >>>> /* Initialise aggregate values */ >>>> @@ -334,7 +335,11 @@ static void qcom_icc_bus_aggregate(struct icc_provider *provider, >>>> list_for_each_entry(node, &provider->nodes, node_list) { >>>> qn = node->data; >>>> for (i = 0; i < QCOM_ICC_NUM_BUCKETS; i++) { >>>> - agg_avg[i] += qn->sum_avg[i]; >>>> + if (qn->channels) >>>> + sum_avg[i] = div_u64(qn->sum_avg[i], qn->channels); >>>> + else >>>> + sum_avg[i] = qn->sum_avg[i]; >>>> + agg_avg[i] += sum_avg[i]; >>>> agg_peak[i] = max_t(u64, agg_peak[i], qn->max_peak[i]); >>>> } >>>> } >>>> diff --git a/drivers/interconnect/qcom/icc-rpm.h b/drivers/interconnect/qcom/icc-rpm.h >>>> index 8ba1918d7997..8aed5400afda 100644 >>>> --- a/drivers/interconnect/qcom/icc-rpm.h >>>> +++ b/drivers/interconnect/qcom/icc-rpm.h >>>> @@ -66,6 +66,7 @@ struct qcom_icc_qos { >>>> * @id: a unique node identifier >>>> * @links: an array of nodes where we can go next while traversing >>>> * @num_links: the total number of @links >>>> + * @channels: number of channels at this node (e.g. DDR channels) >>>> * @buswidth: width of the interconnect between a node and the bus (bytes) >>>> * @sum_avg: current sum aggregate value of all avg bw requests >>>> * @max_peak: current max aggregate value of all peak bw requests >>>> @@ -78,6 +79,7 @@ struct qcom_icc_node { >>>> u16 id; >>>> const u16 *links; >>>> u16 num_links; >>>> + u16 channels; >>>> u16 buswidth; >>>> u64 sum_avg[QCOM_ICC_NUM_BUCKETS]; >>>> u64 max_peak[QCOM_ICC_NUM_BUCKETS]; >>>> >>> >
On 21.03.23 16:23, Konrad Dybcio wrote: > > > On 21.03.2023 15:21, Georgi Djakov wrote: >> On 21.03.23 16:09, Konrad Dybcio wrote: >>> >>> On 21.03.2023 15:06, Georgi Djakov wrote: >>>> Hi Konrad, >>>> >>>> Thanks for the patch! >>>> >>>> On 8.03.23 23:40, Konrad Dybcio wrote: >>>>> Some nodes, like EBI0 (DDR) or L3/LLCC, may be connected over more than >>>>> one channel. This should be taken into account in bandwidth calcualtion, >>>>> as we're supposed to feed msmbus with the per-channel bandwidth. Add >>>>> support for specifying that and use it during bandwidth aggregation. >>>>> >>>> >>>> This looks good, but do you have any follow-up patch to use this and set >>>> the channels in some driver? >>> Yes, I have a couple of OOT drivers that are gonna make use of it. >>> TBF it should have been sent separately from the QoS mess, but I >>> don't think it's much of an issue to take it as-is. >>> >>> The aforementioned OOT drivers for MSM8998 and SM6375 will be >>> submitted after we reach a consensus on how we want to ensure >>> that each node is guaranteed to have its clocks enabled before >>> access, among some other minor things. >> >> Yes, these QoS clocks are confusing. Maybe you can even submit them >> without configuring any QoS stuff in first place? Does enabling QoS >> actually show any benefits on these devices? > Haven't tested that thoroughly to be honest. But I'll try to get > some numbers. I expect this to have impact only on some latency sensitive stuff like modem or when there is heavy traffic flows. Maybe we can start without QoS first and then add it on top as a next step? BR, Georgi > > Konrad >> >> Thanks, >> Georgi >> >>> Konrad >>>> >>>> BR, >>>> Georgi >>>> >>>>> Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org> >>>>> Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org> >>>>> --- >>>>> drivers/interconnect/qcom/icc-rpm.c | 7 ++++++- >>>>> drivers/interconnect/qcom/icc-rpm.h | 2 ++ >>>>> 2 files changed, 8 insertions(+), 1 deletion(-) >>>>> >>>>> diff --git a/drivers/interconnect/qcom/icc-rpm.c b/drivers/interconnect/qcom/icc-rpm.c >>>>> index 35fd75ae70e3..27c4c6497994 100644 >>>>> --- a/drivers/interconnect/qcom/icc-rpm.c >>>>> +++ b/drivers/interconnect/qcom/icc-rpm.c >>>>> @@ -317,6 +317,7 @@ static void qcom_icc_bus_aggregate(struct icc_provider *provider, >>>>> { >>>>> struct icc_node *node; >>>>> struct qcom_icc_node *qn; >>>>> + u64 sum_avg[QCOM_ICC_NUM_BUCKETS]; >>>>> int i; >>>>> /* Initialise aggregate values */ >>>>> @@ -334,7 +335,11 @@ static void qcom_icc_bus_aggregate(struct icc_provider *provider, >>>>> list_for_each_entry(node, &provider->nodes, node_list) { >>>>> qn = node->data; >>>>> for (i = 0; i < QCOM_ICC_NUM_BUCKETS; i++) { >>>>> - agg_avg[i] += qn->sum_avg[i]; >>>>> + if (qn->channels) >>>>> + sum_avg[i] = div_u64(qn->sum_avg[i], qn->channels); >>>>> + else >>>>> + sum_avg[i] = qn->sum_avg[i]; >>>>> + agg_avg[i] += sum_avg[i]; >>>>> agg_peak[i] = max_t(u64, agg_peak[i], qn->max_peak[i]); >>>>> } >>>>> } >>>>> diff --git a/drivers/interconnect/qcom/icc-rpm.h b/drivers/interconnect/qcom/icc-rpm.h >>>>> index 8ba1918d7997..8aed5400afda 100644 >>>>> --- a/drivers/interconnect/qcom/icc-rpm.h >>>>> +++ b/drivers/interconnect/qcom/icc-rpm.h >>>>> @@ -66,6 +66,7 @@ struct qcom_icc_qos { >>>>> * @id: a unique node identifier >>>>> * @links: an array of nodes where we can go next while traversing >>>>> * @num_links: the total number of @links >>>>> + * @channels: number of channels at this node (e.g. DDR channels) >>>>> * @buswidth: width of the interconnect between a node and the bus (bytes) >>>>> * @sum_avg: current sum aggregate value of all avg bw requests >>>>> * @max_peak: current max aggregate value of all peak bw requests >>>>> @@ -78,6 +79,7 @@ struct qcom_icc_node { >>>>> u16 id; >>>>> const u16 *links; >>>>> u16 num_links; >>>>> + u16 channels; >>>>> u16 buswidth; >>>>> u64 sum_avg[QCOM_ICC_NUM_BUCKETS]; >>>>> u64 max_peak[QCOM_ICC_NUM_BUCKETS]; >>>>> >>>> >>
On 21.03.2023 15:49, Georgi Djakov wrote: > On 21.03.23 16:23, Konrad Dybcio wrote: >> >> >> On 21.03.2023 15:21, Georgi Djakov wrote: >>> On 21.03.23 16:09, Konrad Dybcio wrote: >>>> >>>> On 21.03.2023 15:06, Georgi Djakov wrote: >>>>> Hi Konrad, >>>>> >>>>> Thanks for the patch! >>>>> >>>>> On 8.03.23 23:40, Konrad Dybcio wrote: >>>>>> Some nodes, like EBI0 (DDR) or L3/LLCC, may be connected over more than >>>>>> one channel. This should be taken into account in bandwidth calcualtion, >>>>>> as we're supposed to feed msmbus with the per-channel bandwidth. Add >>>>>> support for specifying that and use it during bandwidth aggregation. >>>>>> >>>>> >>>>> This looks good, but do you have any follow-up patch to use this and set >>>>> the channels in some driver? >>>> Yes, I have a couple of OOT drivers that are gonna make use of it. >>>> TBF it should have been sent separately from the QoS mess, but I >>>> don't think it's much of an issue to take it as-is. >>>> >>>> The aforementioned OOT drivers for MSM8998 and SM6375 will be >>>> submitted after we reach a consensus on how we want to ensure >>>> that each node is guaranteed to have its clocks enabled before >>>> access, among some other minor things. >>> >>> Yes, these QoS clocks are confusing. Maybe you can even submit them >>> without configuring any QoS stuff in first place? Does enabling QoS >>> actually show any benefits on these devices? >> Haven't tested that thoroughly to be honest. But I'll try to get >> some numbers. > > I expect this to have impact only on some latency sensitive stuff like > modem or when there is heavy traffic flows. Maybe we can start without > QoS first and then add it on top as a next step? I only now remembered why I didn't do that.. Adding QoS at a later time will break older DTs, as with QoS we need to pass some clocks to the driver. Konrad > > BR, > Georgi > >> >> Konrad >>> >>> Thanks, >>> Georgi >>> >>>> Konrad >>>>> >>>>> BR, >>>>> Georgi >>>>> >>>>>> Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org> >>>>>> Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org> >>>>>> --- >>>>>> drivers/interconnect/qcom/icc-rpm.c | 7 ++++++- >>>>>> drivers/interconnect/qcom/icc-rpm.h | 2 ++ >>>>>> 2 files changed, 8 insertions(+), 1 deletion(-) >>>>>> >>>>>> diff --git a/drivers/interconnect/qcom/icc-rpm.c b/drivers/interconnect/qcom/icc-rpm.c >>>>>> index 35fd75ae70e3..27c4c6497994 100644 >>>>>> --- a/drivers/interconnect/qcom/icc-rpm.c >>>>>> +++ b/drivers/interconnect/qcom/icc-rpm.c >>>>>> @@ -317,6 +317,7 @@ static void qcom_icc_bus_aggregate(struct icc_provider *provider, >>>>>> { >>>>>> struct icc_node *node; >>>>>> struct qcom_icc_node *qn; >>>>>> + u64 sum_avg[QCOM_ICC_NUM_BUCKETS]; >>>>>> int i; >>>>>> /* Initialise aggregate values */ >>>>>> @@ -334,7 +335,11 @@ static void qcom_icc_bus_aggregate(struct icc_provider *provider, >>>>>> list_for_each_entry(node, &provider->nodes, node_list) { >>>>>> qn = node->data; >>>>>> for (i = 0; i < QCOM_ICC_NUM_BUCKETS; i++) { >>>>>> - agg_avg[i] += qn->sum_avg[i]; >>>>>> + if (qn->channels) >>>>>> + sum_avg[i] = div_u64(qn->sum_avg[i], qn->channels); >>>>>> + else >>>>>> + sum_avg[i] = qn->sum_avg[i]; >>>>>> + agg_avg[i] += sum_avg[i]; >>>>>> agg_peak[i] = max_t(u64, agg_peak[i], qn->max_peak[i]); >>>>>> } >>>>>> } >>>>>> diff --git a/drivers/interconnect/qcom/icc-rpm.h b/drivers/interconnect/qcom/icc-rpm.h >>>>>> index 8ba1918d7997..8aed5400afda 100644 >>>>>> --- a/drivers/interconnect/qcom/icc-rpm.h >>>>>> +++ b/drivers/interconnect/qcom/icc-rpm.h >>>>>> @@ -66,6 +66,7 @@ struct qcom_icc_qos { >>>>>> * @id: a unique node identifier >>>>>> * @links: an array of nodes where we can go next while traversing >>>>>> * @num_links: the total number of @links >>>>>> + * @channels: number of channels at this node (e.g. DDR channels) >>>>>> * @buswidth: width of the interconnect between a node and the bus (bytes) >>>>>> * @sum_avg: current sum aggregate value of all avg bw requests >>>>>> * @max_peak: current max aggregate value of all peak bw requests >>>>>> @@ -78,6 +79,7 @@ struct qcom_icc_node { >>>>>> u16 id; >>>>>> const u16 *links; >>>>>> u16 num_links; >>>>>> + u16 channels; >>>>>> u16 buswidth; >>>>>> u64 sum_avg[QCOM_ICC_NUM_BUCKETS]; >>>>>> u64 max_peak[QCOM_ICC_NUM_BUCKETS]; >>>>>> >>>>> >>> >
diff --git a/drivers/interconnect/qcom/icc-rpm.c b/drivers/interconnect/qcom/icc-rpm.c index 35fd75ae70e3..27c4c6497994 100644 --- a/drivers/interconnect/qcom/icc-rpm.c +++ b/drivers/interconnect/qcom/icc-rpm.c @@ -317,6 +317,7 @@ static void qcom_icc_bus_aggregate(struct icc_provider *provider, { struct icc_node *node; struct qcom_icc_node *qn; + u64 sum_avg[QCOM_ICC_NUM_BUCKETS]; int i; /* Initialise aggregate values */ @@ -334,7 +335,11 @@ static void qcom_icc_bus_aggregate(struct icc_provider *provider, list_for_each_entry(node, &provider->nodes, node_list) { qn = node->data; for (i = 0; i < QCOM_ICC_NUM_BUCKETS; i++) { - agg_avg[i] += qn->sum_avg[i]; + if (qn->channels) + sum_avg[i] = div_u64(qn->sum_avg[i], qn->channels); + else + sum_avg[i] = qn->sum_avg[i]; + agg_avg[i] += sum_avg[i]; agg_peak[i] = max_t(u64, agg_peak[i], qn->max_peak[i]); } } diff --git a/drivers/interconnect/qcom/icc-rpm.h b/drivers/interconnect/qcom/icc-rpm.h index 8ba1918d7997..8aed5400afda 100644 --- a/drivers/interconnect/qcom/icc-rpm.h +++ b/drivers/interconnect/qcom/icc-rpm.h @@ -66,6 +66,7 @@ struct qcom_icc_qos { * @id: a unique node identifier * @links: an array of nodes where we can go next while traversing * @num_links: the total number of @links + * @channels: number of channels at this node (e.g. DDR channels) * @buswidth: width of the interconnect between a node and the bus (bytes) * @sum_avg: current sum aggregate value of all avg bw requests * @max_peak: current max aggregate value of all peak bw requests @@ -78,6 +79,7 @@ struct qcom_icc_node { u16 id; const u16 *links; u16 num_links; + u16 channels; u16 buswidth; u64 sum_avg[QCOM_ICC_NUM_BUCKETS]; u64 max_peak[QCOM_ICC_NUM_BUCKETS];