diff mbox series

cpufreq: tegra186: Fix get frequency callback

Message ID 20201103115514.547047-1-jonathanh@nvidia.com
State Deferred
Headers show
Series cpufreq: tegra186: Fix get frequency callback | expand

Commit Message

Jon Hunter Nov. 3, 2020, 11:55 a.m. UTC
Commit b89c01c96051 ("cpufreq: tegra186: Fix initial frequency")
implemented the CPUFREQ 'get' callback to determine the current
operating frequency for each CPU. This implementation used a simple
looked up to determine the current operating frequency. The problem
with this is that frequency table for different Tegra186 devices may
vary and so the default boot frequency for Tegra186 device may or may
not be present in the frequency table. If the default boot frequency is
not present in the frequency table, this causes the function
tegra186_cpufreq_get() to return 0 and in turn causes cpufreq_online()
to fail which prevents CPUFREQ from working.

Fix this by always calculating the CPU frequency based upon the current
'ndiv' setting for the CPU. Note that the CPU frequency for Tegra186 is
calculated by reading the current 'ndiv' setting, multiplying by the
CPU reference clock and dividing by a constant divisor.

Fixes: b89c01c96051 ("cpufreq: tegra186: Fix initial frequency")

Signed-off-by: Jon Hunter <jonathanh@nvidia.com>
---
 drivers/cpufreq/tegra186-cpufreq.c | 33 +++++++++++++++++++-----------
 1 file changed, 21 insertions(+), 12 deletions(-)

Comments

Viresh Kumar Nov. 4, 2020, 9:33 a.m. UTC | #1
On 03-11-20, 11:55, Jon Hunter wrote:
> Commit b89c01c96051 ("cpufreq: tegra186: Fix initial frequency")
> implemented the CPUFREQ 'get' callback to determine the current
> operating frequency for each CPU. This implementation used a simple
> looked up to determine the current operating frequency. The problem
> with this is that frequency table for different Tegra186 devices may
> vary and so the default boot frequency for Tegra186 device may or may
> not be present in the frequency table. If the default boot frequency is
> not present in the frequency table, this causes the function
> tegra186_cpufreq_get() to return 0 and in turn causes cpufreq_online()
> to fail which prevents CPUFREQ from working.
> 
> Fix this by always calculating the CPU frequency based upon the current
> 'ndiv' setting for the CPU. Note that the CPU frequency for Tegra186 is
> calculated by reading the current 'ndiv' setting, multiplying by the
> CPU reference clock and dividing by a constant divisor.
> 
> Fixes: b89c01c96051 ("cpufreq: tegra186: Fix initial frequency")
> 
> Signed-off-by: Jon Hunter <jonathanh@nvidia.com>
> ---
>  drivers/cpufreq/tegra186-cpufreq.c | 33 +++++++++++++++++++-----------
>  1 file changed, 21 insertions(+), 12 deletions(-)

Acked-by: Viresh Kumar <viresh.kumar@linaro.org>

Rafael: This needs to go in the next rc and so I am not applying it
in my tree as this is the only fix I have for now.
Jon Hunter Nov. 16, 2020, 3:55 p.m. UTC | #2
Hi Rafael,

On 04/11/2020 09:33, Viresh Kumar wrote:
> On 03-11-20, 11:55, Jon Hunter wrote:
>> Commit b89c01c96051 ("cpufreq: tegra186: Fix initial frequency")
>> implemented the CPUFREQ 'get' callback to determine the current
>> operating frequency for each CPU. This implementation used a simple
>> looked up to determine the current operating frequency. The problem
>> with this is that frequency table for different Tegra186 devices may
>> vary and so the default boot frequency for Tegra186 device may or may
>> not be present in the frequency table. If the default boot frequency is
>> not present in the frequency table, this causes the function
>> tegra186_cpufreq_get() to return 0 and in turn causes cpufreq_online()
>> to fail which prevents CPUFREQ from working.
>>
>> Fix this by always calculating the CPU frequency based upon the current
>> 'ndiv' setting for the CPU. Note that the CPU frequency for Tegra186 is
>> calculated by reading the current 'ndiv' setting, multiplying by the
>> CPU reference clock and dividing by a constant divisor.
>>
>> Fixes: b89c01c96051 ("cpufreq: tegra186: Fix initial frequency")
>>
>> Signed-off-by: Jon Hunter <jonathanh@nvidia.com>
>> ---
>>  drivers/cpufreq/tegra186-cpufreq.c | 33 +++++++++++++++++++-----------
>>  1 file changed, 21 insertions(+), 12 deletions(-)
> 
> Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
> 
> Rafael: This needs to go in the next rc and so I am not applying it
> in my tree as this is the only fix I have for now.


Are you able to pick this up for v5.10 fixes?

Thanks
Jon
Viresh Kumar Nov. 17, 2020, 4:36 a.m. UTC | #3
On 16-11-20, 15:55, Jon Hunter wrote:
> Hi Rafael,
> 
> On 04/11/2020 09:33, Viresh Kumar wrote:
> > On 03-11-20, 11:55, Jon Hunter wrote:
> >> Commit b89c01c96051 ("cpufreq: tegra186: Fix initial frequency")
> >> implemented the CPUFREQ 'get' callback to determine the current
> >> operating frequency for each CPU. This implementation used a simple
> >> looked up to determine the current operating frequency. The problem
> >> with this is that frequency table for different Tegra186 devices may
> >> vary and so the default boot frequency for Tegra186 device may or may
> >> not be present in the frequency table. If the default boot frequency is
> >> not present in the frequency table, this causes the function
> >> tegra186_cpufreq_get() to return 0 and in turn causes cpufreq_online()
> >> to fail which prevents CPUFREQ from working.
> >>
> >> Fix this by always calculating the CPU frequency based upon the current
> >> 'ndiv' setting for the CPU. Note that the CPU frequency for Tegra186 is
> >> calculated by reading the current 'ndiv' setting, multiplying by the
> >> CPU reference clock and dividing by a constant divisor.
> >>
> >> Fixes: b89c01c96051 ("cpufreq: tegra186: Fix initial frequency")
> >>
> >> Signed-off-by: Jon Hunter <jonathanh@nvidia.com>
> >> ---
> >>  drivers/cpufreq/tegra186-cpufreq.c | 33 +++++++++++++++++++-----------
> >>  1 file changed, 21 insertions(+), 12 deletions(-)
> > 
> > Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
> > 
> > Rafael: This needs to go in the next rc and so I am not applying it
> > in my tree as this is the only fix I have for now.
> 
> 
> Are you able to pick this up for v5.10 fixes?

Applied to my tree now, will send a PULL request soon.
diff mbox series

Patch

diff --git a/drivers/cpufreq/tegra186-cpufreq.c b/drivers/cpufreq/tegra186-cpufreq.c
index 4b4079f51559..7eb2c56c65de 100644
--- a/drivers/cpufreq/tegra186-cpufreq.c
+++ b/drivers/cpufreq/tegra186-cpufreq.c
@@ -42,6 +42,8 @@  static const struct tegra186_cpufreq_cluster_info tegra186_clusters[] = {
 struct tegra186_cpufreq_cluster {
 	const struct tegra186_cpufreq_cluster_info *info;
 	struct cpufreq_frequency_table *table;
+	u32 ref_clk_khz;
+	u32 div;
 };
 
 struct tegra186_cpufreq_data {
@@ -94,7 +96,7 @@  static int tegra186_cpufreq_set_target(struct cpufreq_policy *policy,
 
 static unsigned int tegra186_cpufreq_get(unsigned int cpu)
 {
-	struct cpufreq_frequency_table *tbl;
+	struct tegra186_cpufreq_data *data = cpufreq_get_driver_data();
 	struct cpufreq_policy *policy;
 	void __iomem *edvd_reg;
 	unsigned int i, freq = 0;
@@ -104,17 +106,23 @@  static unsigned int tegra186_cpufreq_get(unsigned int cpu)
 	if (!policy)
 		return 0;
 
-	tbl = policy->freq_table;
 	edvd_reg = policy->driver_data;
 	ndiv = readl(edvd_reg) & EDVD_CORE_VOLT_FREQ_F_MASK;
 
-	for (i = 0; tbl[i].frequency != CPUFREQ_TABLE_END; i++) {
-		if ((tbl[i].driver_data & EDVD_CORE_VOLT_FREQ_F_MASK) == ndiv) {
-			freq = tbl[i].frequency;
-			break;
+	for (i = 0; i < data->num_clusters; i++) {
+		struct tegra186_cpufreq_cluster *cluster = &data->clusters[i];
+		int core;
+
+		for (core = 0; core < ARRAY_SIZE(cluster->info->cpus); core++) {
+			if (cluster->info->cpus[core] != policy->cpu)
+				continue;
+
+			freq = (cluster->ref_clk_khz * ndiv) / cluster->div;
+			goto out;
 		}
 	}
 
+out:
 	cpufreq_cpu_put(policy);
 
 	return freq;
@@ -133,7 +141,7 @@  static struct cpufreq_driver tegra186_cpufreq_driver = {
 
 static struct cpufreq_frequency_table *init_vhint_table(
 	struct platform_device *pdev, struct tegra_bpmp *bpmp,
-	unsigned int cluster_id)
+	struct tegra186_cpufreq_cluster *cluster)
 {
 	struct cpufreq_frequency_table *table;
 	struct mrq_cpu_vhint_request req;
@@ -152,7 +160,7 @@  static struct cpufreq_frequency_table *init_vhint_table(
 
 	memset(&req, 0, sizeof(req));
 	req.addr = phys;
-	req.cluster_id = cluster_id;
+	req.cluster_id = cluster->info->bpmp_cluster_id;
 
 	memset(&msg, 0, sizeof(msg));
 	msg.mrq = MRQ_CPU_VHINT;
@@ -185,6 +193,9 @@  static struct cpufreq_frequency_table *init_vhint_table(
 		goto free;
 	}
 
+	cluster->ref_clk_khz = data->ref_clk_hz / 1000;
+	cluster->div = data->pdiv * data->mdiv;
+
 	for (i = data->vfloor, j = 0; i <= data->vceil; i++) {
 		struct cpufreq_frequency_table *point;
 		u16 ndiv = data->ndiv[i];
@@ -202,8 +213,7 @@  static struct cpufreq_frequency_table *init_vhint_table(
 
 		point = &table[j++];
 		point->driver_data = edvd_val;
-		point->frequency = data->ref_clk_hz * ndiv / data->pdiv /
-			data->mdiv / 1000;
+		point->frequency = (cluster->ref_clk_khz * ndiv) / cluster->div;
 	}
 
 	table[j].frequency = CPUFREQ_TABLE_END;
@@ -245,8 +255,7 @@  static int tegra186_cpufreq_probe(struct platform_device *pdev)
 		struct tegra186_cpufreq_cluster *cluster = &data->clusters[i];
 
 		cluster->info = &tegra186_clusters[i];
-		cluster->table = init_vhint_table(
-			pdev, bpmp, cluster->info->bpmp_cluster_id);
+		cluster->table = init_vhint_table(pdev, bpmp, cluster);
 		if (IS_ERR(cluster->table)) {
 			err = PTR_ERR(cluster->table);
 			goto put_bpmp;