From 85d7dda5a9f665ea579741ec873a8841f37e8943 Mon Sep 17 00:00:00 2001 From: "Mario Limonciello (AMD)" Date: Tue, 23 Sep 2025 10:29:29 -0500 Subject: [PATCH 1/2] cpufreq/amd-pstate: Fix a regression leading to EPP 0 after hibernate After resuming from S4, all CPUs except the boot CPU have the wrong EPP hint programmed. This is because when the CPUs were offlined the EPP value was reset to 0. This is a similar problem as fixed by commit ba3319e590571 ("cpufreq/amd-pstate: Fix a regression leading to EPP 0 after resume") and the solution is also similar. When offlining rather than reset the values to zero, reset them to match those chosen by the policy. When the CPUs are onlined again these values will be restored. Closes: https://community.frame.work/t/increased-power-usage-after-resuming-from-suspend-on-ryzen-7040-kernel-6-15-regression/74531/20?u=mario_limonciello Fixes: 608a76b65288 ("cpufreq/amd-pstate: Add support for the "Requested CPU Min frequency" BIOS option") Reviewed-by: Gautham R. Shenoy Signed-off-by: Mario Limonciello (AMD) --- drivers/cpufreq/amd-pstate.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/drivers/cpufreq/amd-pstate.c b/drivers/cpufreq/amd-pstate.c index 298e92d8cc03..b44f0f7a5ba1 100644 --- a/drivers/cpufreq/amd-pstate.c +++ b/drivers/cpufreq/amd-pstate.c @@ -1614,7 +1614,11 @@ static int amd_pstate_cpu_offline(struct cpufreq_policy *policy) * min_perf value across kexec reboots. If this CPU is just onlined normally after this, the * limits, epp and desired perf will get reset to the cached values in cpudata struct */ - return amd_pstate_update_perf(policy, perf.bios_min_perf, 0U, 0U, 0U, false); + return amd_pstate_update_perf(policy, perf.bios_min_perf, + FIELD_GET(AMD_CPPC_DES_PERF_MASK, cpudata->cppc_req_cached), + FIELD_GET(AMD_CPPC_MAX_PERF_MASK, cpudata->cppc_req_cached), + FIELD_GET(AMD_CPPC_EPP_PERF_MASK, cpudata->cppc_req_cached), + false); } static int amd_pstate_suspend(struct cpufreq_policy *policy) From 10fad4012234a7dea621ae17c0c9486824f645a0 Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Sat, 18 Oct 2025 14:27:15 +0200 Subject: [PATCH 2/2] Revert "cpuidle: menu: Avoid discarding useful information" It is reported that commit 85975daeaa4d ("cpuidle: menu: Avoid discarding useful information") led to a performance regression on Intel Jasper Lake systems because it reduced the time spent by CPUs in idle state C7 which is correlated to the maximum frequency the CPUs can get to because of an average running power limit [1]. Before that commit, get_typical_interval() would have returned UINT_MAX whenever it had been unable to make a high-confidence prediction which had led to selecting the deepest available idle state too often and both power and performance had been inadequate as a result of that on some systems. However, this had not been a problem on systems with relatively aggressive average running power limits, like the Jasper Lake systems in question, because on those systems it was compensated by the ability to run CPUs faster. It was addressed by causing get_typical_interval() to return a number based on the recent idle duration information available to it even if it could not make a high-confidence prediction, but that clearly did not take the possible correlation between idle power and available CPU capacity into account. For this reason, revert most of the changes made by commit 85975daeaa4d, except for one cosmetic cleanup, and add a comment explaining the rationale for returning UINT_MAX from get_typical_interval() when it is unable to make a high-confidence prediction. Fixes: 85975daeaa4d ("cpuidle: menu: Avoid discarding useful information") Closes: https://lore.kernel.org/linux-pm/36iykr223vmcfsoysexug6s274nq2oimcu55ybn6ww4il3g3cv@cohflgdbpnq7/ [1] Reported-by: Sergey Senozhatsky Cc: All applicable Signed-off-by: Rafael J. Wysocki Link: https://patch.msgid.link/3663603.iIbC2pHGDl@rafael.j.wysocki --- drivers/cpuidle/governors/menu.c | 21 +++++++++------------ 1 file changed, 9 insertions(+), 12 deletions(-) diff --git a/drivers/cpuidle/governors/menu.c b/drivers/cpuidle/governors/menu.c index 4d9aa5ce31f0..7d21fb5a72f4 100644 --- a/drivers/cpuidle/governors/menu.c +++ b/drivers/cpuidle/governors/menu.c @@ -188,20 +188,17 @@ static unsigned int get_typical_interval(struct menu_device *data) * * This can deal with workloads that have long pauses interspersed * with sporadic activity with a bunch of short pauses. + * + * However, if the number of remaining samples is too small to exclude + * any more outliers, allow the deepest available idle state to be + * selected because there are systems where the time spent by CPUs in + * deep idle states is correlated to the maximum frequency the CPUs + * can get to. On those systems, shallow idle states should be avoided + * unless there is a clear indication that the given CPU is most likley + * going to be woken up shortly. */ - if (divisor * 4 <= INTERVALS * 3) { - /* - * If there are sufficiently many data points still under - * consideration after the outliers have been eliminated, - * returning without a prediction would be a mistake because it - * is likely that the next interval will not exceed the current - * maximum, so return the latter in that case. - */ - if (divisor >= INTERVALS / 2) - return max; - + if (divisor * 4 <= INTERVALS * 3) return UINT_MAX; - } /* Update the thresholds for the next round. */ if (avg - min > max - avg)