288 lines · plain
1====================================================================2Interaction of Suspend code (S3) with the CPU hotplug infrastructure3====================================================================4 5(C) 2011 - 2014 Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>6 7 8I. Differences between CPU hotplug and Suspend-to-RAM9======================================================10 11How does the regular CPU hotplug code differ from how the Suspend-to-RAM12infrastructure uses it internally? And where do they share common code?13 14Well, a picture is worth a thousand words... So ASCII art follows :-)15 16[This depicts the current design in the kernel, and focusses only on the17interactions involving the freezer and CPU hotplug and also tries to explain18the locking involved. It outlines the notifications involved as well.19But please note that here, only the call paths are illustrated, with the aim20of describing where they take different paths and where they share code.21What happens when regular CPU hotplug and Suspend-to-RAM race with each other22is not depicted here.]23 24On a high level, the suspend-resume cycle goes like this::25 26 |Freeze| -> |Disable nonboot| -> |Do suspend| -> |Enable nonboot| -> |Thaw |27 |tasks | | cpus | | | | cpus | |tasks|28 29 30More details follow::31 32 Suspend call path33 -----------------34 35 Write 'mem' to36 /sys/power/state37 sysfs file38 |39 v40 Acquire system_transition_mutex lock41 |42 v43 Send PM_SUSPEND_PREPARE44 notifications45 |46 v47 Freeze tasks48 |49 |50 v51 freeze_secondary_cpus()52 /* start */53 |54 v55 Acquire cpu_add_remove_lock56 |57 v58 Iterate over CURRENTLY59 online CPUs60 |61 |62 | ----------63 v | L64 ======> _cpu_down() |65 | [This takes cpuhotplug.lock |66 Common | before taking down the CPU |67 code | and releases it when done] | O68 | While it is at it, notifications |69 | are sent when notable events occur, |70 ======> by running all registered callbacks. |71 | | O72 | |73 | |74 v |75 Note down these cpus in | P76 frozen_cpus mask ----------77 |78 v79 Disable regular cpu hotplug80 by increasing cpu_hotplug_disabled81 |82 v83 Release cpu_add_remove_lock84 |85 v86 /* freeze_secondary_cpus() complete */87 |88 v89 Do suspend90 91 92 93Resuming back is likewise, with the counterparts being (in the order of94execution during resume):95 96* thaw_secondary_cpus() which involves::97 98 | Acquire cpu_add_remove_lock99 | Decrease cpu_hotplug_disabled, thereby enabling regular cpu hotplug100 | Call _cpu_up() [for all those cpus in the frozen_cpus mask, in a loop]101 | Release cpu_add_remove_lock102 v103 104* thaw tasks105* send PM_POST_SUSPEND notifications106* Release system_transition_mutex lock.107 108 109It is to be noted here that the system_transition_mutex lock is acquired at the110very beginning, when we are just starting out to suspend, and then released only111after the entire cycle is complete (i.e., suspend + resume).112 113::114 115 116 117 Regular CPU hotplug call path118 -----------------------------119 120 Write 0 (or 1) to121 /sys/devices/system/cpu/cpu*/online122 sysfs file123 |124 |125 v126 cpu_down()127 |128 v129 Acquire cpu_add_remove_lock130 |131 v132 If cpu_hotplug_disabled > 0133 return gracefully134 |135 |136 v137 ======> _cpu_down()138 | [This takes cpuhotplug.lock139 Common | before taking down the CPU140 code | and releases it when done]141 | While it is at it, notifications142 | are sent when notable events occur,143 ======> by running all registered callbacks.144 |145 |146 v147 Release cpu_add_remove_lock148 [That's it!, for149 regular CPU hotplug]150 151 152 153So, as can be seen from the two diagrams (the parts marked as "Common code"),154regular CPU hotplug and the suspend code path converge at the _cpu_down() and155_cpu_up() functions. They differ in the arguments passed to these functions,156in that during regular CPU hotplug, 0 is passed for the 'tasks_frozen'157argument. But during suspend, since the tasks are already frozen by the time158the non-boot CPUs are offlined or onlined, the _cpu_*() functions are called159with the 'tasks_frozen' argument set to 1.160[See below for some known issues regarding this.]161 162 163Important files and functions/entry points:164-------------------------------------------165 166- kernel/power/process.c : freeze_processes(), thaw_processes()167- kernel/power/suspend.c : suspend_prepare(), suspend_enter(), suspend_finish()168- kernel/cpu.c: cpu_[up|down](), _cpu_[up|down](),169 [disable|enable]_nonboot_cpus()170 171 172 173II. What are the issues involved in CPU hotplug?174------------------------------------------------175 176There are some interesting situations involving CPU hotplug and microcode177update on the CPUs, as discussed below:178 179[Please bear in mind that the kernel requests the microcode images from180userspace, using the request_firmware() function defined in181drivers/base/firmware_loader/main.c]182 183 184a. When all the CPUs are identical:185 186 This is the most common situation and it is quite straightforward: we want187 to apply the same microcode revision to each of the CPUs.188 To give an example of x86, the collect_cpu_info() function defined in189 arch/x86/kernel/microcode_core.c helps in discovering the type of the CPU190 and thereby in applying the correct microcode revision to it.191 But note that the kernel does not maintain a common microcode image for the192 all CPUs, in order to handle case 'b' described below.193 194 195b. When some of the CPUs are different than the rest:196 197 In this case since we probably need to apply different microcode revisions198 to different CPUs, the kernel maintains a copy of the correct microcode199 image for each CPU (after appropriate CPU type/model discovery using200 functions such as collect_cpu_info()).201 202 203c. When a CPU is physically hot-unplugged and a new (and possibly different204 type of) CPU is hot-plugged into the system:205 206 In the current design of the kernel, whenever a CPU is taken offline during207 a regular CPU hotplug operation, upon receiving the CPU_DEAD notification208 (which is sent by the CPU hotplug code), the microcode update driver's209 callback for that event reacts by freeing the kernel's copy of the210 microcode image for that CPU.211 212 Hence, when a new CPU is brought online, since the kernel finds that it213 doesn't have the microcode image, it does the CPU type/model discovery214 afresh and then requests the userspace for the appropriate microcode image215 for that CPU, which is subsequently applied.216 217 For example, in x86, the mc_cpu_callback() function (which is the microcode218 update driver's callback registered for CPU hotplug events) calls219 microcode_update_cpu() which would call microcode_init_cpu() in this case,220 instead of microcode_resume_cpu() when it finds that the kernel doesn't221 have a valid microcode image. This ensures that the CPU type/model222 discovery is performed and the right microcode is applied to the CPU after223 getting it from userspace.224 225 226d. Handling microcode update during suspend/hibernate:227 228 Strictly speaking, during a CPU hotplug operation which does not involve229 physically removing or inserting CPUs, the CPUs are not actually powered230 off during a CPU offline. They are just put to the lowest C-states possible.231 Hence, in such a case, it is not really necessary to re-apply microcode232 when the CPUs are brought back online, since they wouldn't have lost the233 image during the CPU offline operation.234 235 This is the usual scenario encountered during a resume after a suspend.236 However, in the case of hibernation, since all the CPUs are completely237 powered off, during restore it becomes necessary to apply the microcode238 images to all the CPUs.239 240 [Note that we don't expect someone to physically pull out nodes and insert241 nodes with a different type of CPUs in-between a suspend-resume or a242 hibernate/restore cycle.]243 244 In the current design of the kernel however, during a CPU offline operation245 as part of the suspend/hibernate cycle (cpuhp_tasks_frozen is set),246 the existing copy of microcode image in the kernel is not freed up.247 And during the CPU online operations (during resume/restore), since the248 kernel finds that it already has copies of the microcode images for all the249 CPUs, it just applies them to the CPUs, avoiding any re-discovery of CPU250 type/model and the need for validating whether the microcode revisions are251 right for the CPUs or not (due to the above assumption that physical CPU252 hotplug will not be done in-between suspend/resume or hibernate/restore253 cycles).254 255 256III. Known problems257===================258 259Are there any known problems when regular CPU hotplug and suspend race260with each other?261 262Yes, they are listed below:263 2641. When invoking regular CPU hotplug, the 'tasks_frozen' argument passed to265 the _cpu_down() and _cpu_up() functions is *always* 0.266 This might not reflect the true current state of the system, since the267 tasks could have been frozen by an out-of-band event such as a suspend268 operation in progress. Hence, the cpuhp_tasks_frozen variable will not269 reflect the frozen state and the CPU hotplug callbacks which evaluate270 that variable might execute the wrong code path.271 2722. If a regular CPU hotplug stress test happens to race with the freezer due273 to a suspend operation in progress at the same time, then we could hit the274 situation described below:275 276 * A regular cpu online operation continues its journey from userspace277 into the kernel, since the freezing has not yet begun.278 * Then freezer gets to work and freezes userspace.279 * If cpu online has not yet completed the microcode update stuff by now,280 it will now start waiting on the frozen userspace in the281 TASK_UNINTERRUPTIBLE state, in order to get the microcode image.282 * Now the freezer continues and tries to freeze the remaining tasks. But283 due to this wait mentioned above, the freezer won't be able to freeze284 the cpu online hotplug task and hence freezing of tasks fails.285 286 As a result of this task freezing failure, the suspend operation gets287 aborted.288