1========================== 2Real-Time group scheduling 3========================== 4 5.. CONTENTS 6 7 0. WARNING 8 1. Overview 9 1.1 The problem 10 1.2 The solution 11 2. The interface 12 2.1 System-wide settings 13 2.2 Default behaviour 14 2.3 Basis for grouping tasks 15 3. Future plans 16 17 180. WARNING 19========== 20 21 Fiddling with these settings can result in an unstable system, the knobs are 22 root only and assumes root knows what he is doing. 23 24Most notable: 25 26 * very small values in sched_rt_period_us can result in an unstable 27 system when the period is smaller than either the available hrtimer 28 resolution, or the time it takes to handle the budget refresh itself. 29 30 * very small values in sched_rt_runtime_us can result in an unstable 31 system when the runtime is so small the system has difficulty making 32 forward progress (NOTE: the migration thread and kstopmachine both 33 are real-time processes). 34 351. Overview 36=========== 37 38 391.1 The problem 40--------------- 41 42Realtime scheduling is all about determinism, a group has to be able to rely on 43the amount of bandwidth (eg. CPU time) being constant. In order to schedule 44multiple groups of realtime tasks, each group must be assigned a fixed portion 45of the CPU time available. Without a minimum guarantee a realtime group can 46obviously fall short. A fuzzy upper limit is of no use since it cannot be 47relied upon. Which leaves us with just the single fixed portion. 48 491.2 The solution 50---------------- 51 52CPU time is divided by means of specifying how much time can be spent running 53in a given period. We allocate this "run time" for each realtime group which 54the other realtime groups will not be permitted to use. 55 56Any time not allocated to a realtime group will be used to run normal priority 57tasks (SCHED_OTHER). Any allocated run time not used will also be picked up by 58SCHED_OTHER. 59 60Let's consider an example: a frame fixed realtime renderer must deliver 25 61frames a second, which yields a period of 0.04s per frame. Now say it will also 62have to play some music and respond to input, leaving it with around 80% CPU 63time dedicated for the graphics. We can then give this group a run time of 0.8 64* 0.04s = 0.032s. 65 66This way the graphics group will have a 0.04s period with a 0.032s run time 67limit. Now if the audio thread needs to refill the DMA buffer every 0.005s, but 68needs only about 3% CPU time to do so, it can do with a 0.03 * 0.005s = 690.00015s. So this group can be scheduled with a period of 0.005s and a run time 70of 0.00015s. 71 72The remaining CPU time will be used for user input and other tasks. Because 73realtime tasks have explicitly allocated the CPU time they need to perform 74their tasks, buffer underruns in the graphics or audio can be eliminated. 75 76NOTE: the above example is not fully implemented yet. We still 77lack an EDF scheduler to make non-uniform periods usable. 78 79 802. The Interface 81================ 82 83 842.1 System wide settings 85------------------------ 86 87The system wide settings are configured under the /proc virtual file system: 88 89/proc/sys/kernel/sched_rt_period_us: 90 The scheduling period that is equivalent to 100% CPU bandwidth 91 92/proc/sys/kernel/sched_rt_runtime_us: 93 A global limit on how much time realtime scheduling may use. Even without 94 CONFIG_RT_GROUP_SCHED enabled, this will limit time reserved to realtime 95 processes. With CONFIG_RT_GROUP_SCHED it signifies the total bandwidth 96 available to all realtime groups. 97 98 * Time is specified in us because the interface is s32. This gives an 99 operating range from 1us to about 35 minutes. 100 * sched_rt_period_us takes values from 1 to INT_MAX. 101 * sched_rt_runtime_us takes values from -1 to (INT_MAX - 1). 102 * A run time of -1 specifies runtime == period, ie. no limit. 103 104 1052.2 Default behaviour 106--------------------- 107 108The default values for sched_rt_period_us (1000000 or 1s) and 109sched_rt_runtime_us (950000 or 0.95s). This gives 0.05s to be used by 110SCHED_OTHER (non-RT tasks). These defaults were chosen so that a run-away 111realtime tasks will not lock up the machine but leave a little time to recover 112it. By setting runtime to -1 you'd get the old behaviour back. 113 114By default all bandwidth is assigned to the root group and new groups get the 115period from /proc/sys/kernel/sched_rt_period_us and a run time of 0. If you 116want to assign bandwidth to another group, reduce the root group's bandwidth 117and assign some or all of the difference to another group. 118 119Realtime group scheduling means you have to assign a portion of total CPU 120bandwidth to the group before it will accept realtime tasks. Therefore you will 121not be able to run realtime tasks as any user other than root until you have 122done that, even if the user has the rights to run processes with realtime 123priority! 124 125 1262.3 Basis for grouping tasks 127---------------------------- 128 129Enabling CONFIG_RT_GROUP_SCHED lets you explicitly allocate real 130CPU bandwidth to task groups. 131 132This uses the cgroup virtual file system and "<cgroup>/cpu.rt_runtime_us" 133to control the CPU time reserved for each control group. 134 135For more information on working with control groups, you should read 136Documentation/admin-guide/cgroup-v1/cgroups.rst as well. 137 138Group settings are checked against the following limits in order to keep the 139configuration schedulable: 140 141 \Sum_{i} runtime_{i} / global_period <= global_runtime / global_period 142 143For now, this can be simplified to just the following (but see Future plans): 144 145 \Sum_{i} runtime_{i} <= global_runtime 146 147 1483. Future plans 149=============== 150 151There is work in progress to make the scheduling period for each group 152("<cgroup>/cpu.rt_period_us") configurable as well. 153 154The constraint on the period is that a subgroup must have a smaller or 155equal period to its parent. But realistically its not very useful _yet_ 156as its prone to starvation without deadline scheduling. 157 158Consider two sibling groups A and B; both have 50% bandwidth, but A's 159period is twice the length of B's. 160 161* group A: period=100000us, runtime=50000us 162 163 - this runs for 0.05s once every 0.1s 164 165* group B: period= 50000us, runtime=25000us 166 167 - this runs for 0.025s twice every 0.1s (or once every 0.05 sec). 168 169This means that currently a while (1) loop in A will run for the full period of 170B and can starve B's tasks (assuming they are of lower priority) for a whole 171period. 172 173The next project will be SCHED_EDF (Earliest Deadline First scheduling) to bring 174full deadline scheduling to the linux kernel. Deadline scheduling the above 175groups and treating end of the period as a deadline will ensure that they both 176get their allocated time. 177 178Implementing SCHED_EDF might take a while to complete. Priority Inheritance is 179the biggest challenge as the current linux PI infrastructure is geared towards 180the limited static priority levels 0-99. With deadline scheduling you need to 181do deadline inheritance (since priority is inversely proportional to the 182deadline delta (deadline - now)). 183 184This means the whole PI machinery will have to be reworked - and that is one of 185the most complex pieces of code we have. 186