kernel/drivers/thermal/intel/intel_powerclamp.c, branch linux-6.9.y

x86/cpu/topology: Rename topology_max_die_per_package()

2024-02-15T21:07:45Z

The plural of die is dies. Signed-off-by: Thomas Gleixner Tested-by: Michael Kelley Tested-by: Sohil Mehta Link: https://lore.kernel.org/r/20240213210253.065874205@linutronix.de

thermal: intel: powerclamp: Remove dead code for target mwait value

2024-01-22T10:59:22Z

After conversion of this driver to use powercap idle_inject core, this driver doesn't use target_mwait value. So remove dead code. Signed-off-by: Srinivas Pandruvada Signed-off-by: Rafael J. Wysocki

thermal: intel: powerclamp: fix mismatch in get function for max_idle

2023-10-05T18:18:03Z

KASAN reported this [ 444.853098] BUG: KASAN: global-out-of-bounds in param_get_int+0x77/0x90 [ 444.853111] Read of size 4 at addr ffffffffc16c9220 by task cat/2105 ... [ 444.853442] The buggy address belongs to the variable: [ 444.853443] max_idle+0x0/0xffffffffffffcde0 [intel_powerclamp] There is a mismatch between the param_get_int and the definition of max_idle. Replacing param_get_int with param_get_byte resolves this issue. Fixes: ebf519710218 ("thermal: intel: powerclamp: Add two module parameters") Cc: 6.3+ # 6.3+ Signed-off-by: David Arcari Reviewed-by: Srinivas Pandruvada Signed-off-by: Rafael J. Wysocki

thermal: intel: powerclamp: Fix NULL pointer access issue

2023-05-04T18:30:18Z

If cur_state for the powerclamp cooling device is set to the default minimum state of 0, without setting first to cur_state > 0, this results in NULL pointer access. This NULL pointer access happens in the powercap core idle-inject function idle_inject_set_duration() as there is no NULL check for idle_inject_device pointer. This pointer must be allocated by calling idle_inject_register() or idle_inject_register_full(). In the function powerclamp_set_cur_state(), idle_inject_device pointer is allocated only when the cur_state > 0. But setting 0 without changing to any other state, idle_inject_set_duration() will be called with a NULL idle_inject_device pointer. To address this, just return from powerclamp_set_cur_state() if the current cooling device state is the same as the last one. Since the power-up default cooling device state is 0, changing the state to 0 again here will return without calling idle_inject_set_duration(). Signed-off-by: Srinivas Pandruvada Fixes: 8526eb7fc75a ("thermal: intel: powerclamp: Use powercap idle-inject feature") Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=217386 Tested-by: Risto A. Paju Cc: 6.3+ # 6.3+ Signed-off-by: Rafael J. Wysocki

thermal: intel: powerclamp: Fix cpumask and max_idle module parameters

2023-03-30T18:04:29Z

When cpumask is specified as a module parameter the value is overwritten by the module init routine. This can easily be fixed by checking to see if the mask has already been allocated in the init routine. When max_idle is specified as a module parameter a panic will occur. The problem is that the idle_injection_cpu_mask is not allocated until the module init routine executes. This can easily be fixed by allocating the cpumask if it's not already allocated. Fixes: ebf519710218 ("thermal: intel: powerclamp: Add two module parameters") Signed-off-by: David Arcari Reviewed-by: Srinivas Pandruvada Signed-off-by: Rafael J. Wysocki

thermal: intel: powerclamp: Add two module parameters

2023-02-09T20:01:06Z

In some use cases, it is desirable to only inject idle on certain set of CPUs. For example on Alder Lake systems, it is possible that we force idle only on P-Cores for thermal reasons. Also the idle percent can be more than 50% if we only choose partial set of CPUs in the system. Introduce 2 new module parameters for this purpose. They can be only changed when the cooling device is inactive. cpumask (Read/Write): A bit mask of CPUs to inject idle. The format of this bitmask is same as used in other subsystems like in /proc/irq/*/smp_affinity. The mask is comma separated 32 bit groups. Each CPU is one bit. For example for 256 CPU system the full mask is: ffffffff,ffffffff,ffffffff,ffffffff,ffffffff,ffffffff,ffffffff,ffffffff The rightmost mask is for CPU 0-32. max_idle (Read/Write): Maximum injected idle time to the total CPU time ratio in percent range from 1 to 100. Even if the cooling device max_state is always 100 (100%), this parameter allows to add a max idle percent limit. The default is 50, to match the current implementation of powerclamp driver. Also doesn't allow value more than 75, if the cpumask includes every CPU present in the system. Also when the cpumask doesn't include every CPU, there is no use of compensation using package C-state idle counters. Hence don't start package C-state polling thread even for a single package or a single die system in this case. Signed-off-by: Srinivas Pandruvada Signed-off-by: Rafael J. Wysocki

thermal: intel: powerclamp: Fix duration module parameter

2023-02-09T18:57:35Z

After the switch to use the powercap/idle-inject framework in the Intel powerclamp driver, the idle duration unit is microsecond. However, the module parameter for idle duration is in milliseconds, so convert it to microseconds in the "set" callback and back to milliseconds in a new "get" callback. While here, also use mutex protection for setting and getting "duration". The other uses of "duration" are already protected by the mutex. Fixes: 8526eb7fc75a ("thermal: intel: powerclamp: Use powercap idle-inject feature") Signed-off-by: Srinivas Pandruvada [ rjw: Subject and changelog edits ] Signed-off-by: Rafael J. Wysocki

thermal: intel: powerclamp: Return last requested state as cur_state

2023-02-07T19:51:15Z

When the user is reading cur_state from the thermal cooling device for Intel powerclamp device: - It returns the idle ratio from Package C-state counters when there is active idle injection session. - -1, when there is no active idle injection session. This information is not very useful as the package C-state counters vary a lot from read to read. Instead just return the last requested cur_state. Signed-off-by: Srinivas Pandruvada Signed-off-by: Rafael J. Wysocki

thermal: intel: powerclamp: Use powercap idle-inject feature

2023-02-03T13:48:34Z

There are two idle injection implementation in the Linux kernel. One via intel_powerclamp and the other using powercap/idle_inject. Both implementation end up in calling play_idle* function from a FIFO priority thread. Both can't be used at the same time. It is better to use one idle injection framework for better maintainability. In this way, there is only one caller for play_idle. Here powercap/idle_inject can be used for both per-core and for system wide idle injection. This framework has a well defined interface which allow registry for per-core or for all CPUs (system wide). This reduces code complexity in the intel powerclamp driver as all the per CPU kthreads, delayed work and calls to play_idle can be removed. The changes include: - Remove unneeded include files - Remove per CPU kthread workers: balancing_work and idle_injection_work. - Reuse the compensation related code by moving from previous worker thread to idle_injection callback. - Adjust the idle_duration and runtime by using powercap/idle_inject interface. - Remove all variables, which are not required once powercap/idle_inject is used. - Add mutex to avoid race during removal of idle injection during module unload and user action to change idle inject percent. Also for protection during dynamic adjustment of run and idle time from update() callback. - Remove online/offline callbacks to designate control CPU - Use cpu_present_mask global variable for CPU mask - Remove hot plug locks Signed-off-by: Srinivas Pandruvada Signed-off-by: Rafael J. Wysocki

thermal: intel: powerclamp: Fix cur_state for multi package system

2023-02-02T20:08:32Z

The powerclamp cooling device cur_state shows actual idle observed by package C-state idle counters. But the implementation is not sufficient for multi package or multi die system. The cur_state value is incorrect. On these systems, these counters must be read from each package/die and somehow aggregate them. But there is no good method for aggregation. It was not a problem when explicit CPU model addition was required to enable intel powerclamp. In this way certain CPU models could have been avoided. But with the removal of CPU model check with the availability of Package C-state counters, the driver is loaded on most of the recent systems. For multi package/die systems, just show the actual target idle state, the system is trying to achieve. In powerclamp this is the user set state minus one. Also there is no use of starting a worker thread for polling package C-state counters and applying any compensation for multiple package or multiple die systems. Fixes: b721ca0d1927 ("thermal/powerclamp: remove cpu whitelist") Signed-off-by: Srinivas Pandruvada Cc: 4.14+ # 4.14+ Signed-off-by: Rafael J. Wysocki