Scaling min/max freq values were being cached and lagging a setting each time. Fix the ordering of the clamp call to ensure they work.
Closes: https://bugzilla.kernel.org/show_bug.cgi?id=217931 Cc: stable@vger.kernel.org Cc: wkarny@gmail.com Fixes: febab20caeba ("cpufreq/amd-pstate: Fix scaling_min_freq and scaling_max_freq update") Signed-off-by: Mario Limonciello mario.limonciello@amd.com --- drivers/cpufreq/amd-pstate.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-)
diff --git a/drivers/cpufreq/amd-pstate.c b/drivers/cpufreq/amd-pstate.c index 1f6186475715..1791d37fbc53 100644 --- a/drivers/cpufreq/amd-pstate.c +++ b/drivers/cpufreq/amd-pstate.c @@ -1232,14 +1232,13 @@ static void amd_pstate_epp_update_limit(struct cpufreq_policy *policy) max_limit_perf = div_u64(policy->max * cpudata->highest_perf, cpudata->max_freq); min_limit_perf = div_u64(policy->min * cpudata->highest_perf, cpudata->max_freq);
+ WRITE_ONCE(cpudata->max_limit_perf, max_limit_perf); + WRITE_ONCE(cpudata->min_limit_perf, min_limit_perf); + max_perf = clamp_t(unsigned long, max_perf, cpudata->min_limit_perf, cpudata->max_limit_perf); min_perf = clamp_t(unsigned long, min_perf, cpudata->min_limit_perf, cpudata->max_limit_perf); - - WRITE_ONCE(cpudata->max_limit_perf, max_limit_perf); - WRITE_ONCE(cpudata->min_limit_perf, min_limit_perf); - value = READ_ONCE(cpudata->cppc_req_cached);
if (cpudata->policy == CPUFREQ_POLICY_PERFORMANCE)
On Sat, Jan 20, 2024 at 5:49 AM Mario Limonciello mario.limonciello@amd.com wrote:
Scaling min/max freq values were being cached and lagging a setting each time. Fix the ordering of the clamp call to ensure they work.
Closes: https://bugzilla.kernel.org/show_bug.cgi?id=217931 Cc: stable@vger.kernel.org Cc: wkarny@gmail.com Fixes: febab20caeba ("cpufreq/amd-pstate: Fix scaling_min_freq and scaling_max_freq update") Signed-off-by: Mario Limonciello mario.limonciello@amd.com
Reviewed-by: Wyes Karny wkarny@gmail.com
drivers/cpufreq/amd-pstate.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-)
diff --git a/drivers/cpufreq/amd-pstate.c b/drivers/cpufreq/amd-pstate.c index 1f6186475715..1791d37fbc53 100644 --- a/drivers/cpufreq/amd-pstate.c +++ b/drivers/cpufreq/amd-pstate.c @@ -1232,14 +1232,13 @@ static void amd_pstate_epp_update_limit(struct cpufreq_policy *policy) max_limit_perf = div_u64(policy->max * cpudata->highest_perf, cpudata->max_freq); min_limit_perf = div_u64(policy->min * cpudata->highest_perf, cpudata->max_freq);
WRITE_ONCE(cpudata->max_limit_perf, max_limit_perf);WRITE_ONCE(cpudata->min_limit_perf, min_limit_perf);max_perf = clamp_t(unsigned long, max_perf, cpudata->min_limit_perf, cpudata->max_limit_perf); min_perf = clamp_t(unsigned long, min_perf, cpudata->min_limit_perf, cpudata->max_limit_perf);
WRITE_ONCE(cpudata->max_limit_perf, max_limit_perf);WRITE_ONCE(cpudata->min_limit_perf, min_limit_perf);value = READ_ONCE(cpudata->cppc_req_cached); if (cpudata->policy == CPUFREQ_POLICY_PERFORMANCE)-- 2.34.1
On Mon, Jan 22, 2024 at 3:57 PM Wyes Karny wkarny@gmail.com wrote:
On Sat, Jan 20, 2024 at 5:49 AM Mario Limonciello mario.limonciello@amd.com wrote:
Scaling min/max freq values were being cached and lagging a setting each time. Fix the ordering of the clamp call to ensure they work.
Closes: https://bugzilla.kernel.org/show_bug.cgi?id=217931 Cc: stable@vger.kernel.org Cc: wkarny@gmail.com Fixes: febab20caeba ("cpufreq/amd-pstate: Fix scaling_min_freq and scaling_max_freq update") Signed-off-by: Mario Limonciello mario.limonciello@amd.com
Reviewed-by: Wyes Karny wkarny@gmail.com
drivers/cpufreq/amd-pstate.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-)
diff --git a/drivers/cpufreq/amd-pstate.c b/drivers/cpufreq/amd-pstate.c index 1f6186475715..1791d37fbc53 100644 --- a/drivers/cpufreq/amd-pstate.c +++ b/drivers/cpufreq/amd-pstate.c @@ -1232,14 +1232,13 @@ static void amd_pstate_epp_update_limit(struct cpufreq_policy *policy) max_limit_perf = div_u64(policy->max * cpudata->highest_perf, cpudata->max_freq); min_limit_perf = div_u64(policy->min * cpudata->highest_perf, cpudata->max_freq);
WRITE_ONCE(cpudata->max_limit_perf, max_limit_perf);WRITE_ONCE(cpudata->min_limit_perf, min_limit_perf);max_perf = clamp_t(unsigned long, max_perf, cpudata->min_limit_perf, cpudata->max_limit_perf); min_perf = clamp_t(unsigned long, min_perf, cpudata->min_limit_perf, cpudata->max_limit_perf);
WRITE_ONCE(cpudata->max_limit_perf, max_limit_perf);WRITE_ONCE(cpudata->min_limit_perf, min_limit_perf);value = READ_ONCE(cpudata->cppc_req_cached); if (cpudata->policy == CPUFREQ_POLICY_PERFORMANCE)--
Applied as 6.8-rc material, thanks!
On Friday, January 19, 2024 6:33:19 A.M. EST Mario Limonciello wrote:
Scaling min/max freq values were being cached and lagging a setting each time. Fix the ordering of the clamp call to ensure they work.
Closes: https://bugzilla.kernel.org/show_bug.cgi?id=217931 Cc: stable@vger.kernel.org Cc: wkarny@gmail.com Fixes: febab20caeba ("cpufreq/amd-pstate: Fix scaling_min_freq and scaling_max_freq update") Signed-off-by: Mario Limonciello
mario.limonciello@amd.com
drivers/cpufreq/amd-pstate.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-)
diff --git a/drivers/cpufreq/amd-pstate.c b/drivers/cpufreq/amd-pstate.c index 1f6186475715..1791d37fbc53 100644 --- a/drivers/cpufreq/amd-pstate.c +++ b/drivers/cpufreq/amd-pstate.c @@ -1232,14 +1232,13 @@ static void amd_pstate_epp_update_limit(struct cpufreq_policy *policy) max_limit_perf = div_u64(policy->max * cpudata->highest_perf, cpudata->max_freq); min_limit_perf = div_u64(policy->min * cpudata->highest_perf, cpudata->max_freq);
- WRITE_ONCE(cpudata->max_limit_perf, max_limit_perf);
- WRITE_ONCE(cpudata->min_limit_perf, min_limit_perf);
- max_perf = clamp_t(unsigned long, max_perf, cpudata-
min_limit_perf, cpudata->max_limit_perf); min_perf = clamp_t(unsigned long, min_perf, cpudata- min_limit_perf, cpudata->max_limit_perf);
WRITE_ONCE(cpudata->max_limit_perf, max_limit_perf);
WRITE_ONCE(cpudata->min_limit_perf, min_limit_perf);
value = READ_ONCE(cpudata->cppc_req_cached);
if (cpudata->policy == CPUFREQ_POLICY_PERFORMANCE)
Tested-by: Dan Martins dan.martins@zoho.com
linux-stable-mirror@lists.linaro.org
-
Dan Martins -
Mario Limonciello -
Rafael J. Wysocki -
Wyes Karny