203 lines · plain
1 2.. _local_ops:3 4=================================================5Semantics and Behavior of Local Atomic Operations6=================================================7 8:Author: Mathieu Desnoyers9 10 11This document explains the purpose of the local atomic operations, how12to implement them for any given architecture and shows how they can be used13properly. It also stresses on the precautions that must be taken when reading14those local variables across CPUs when the order of memory writes matters.15 16.. note::17 18 Note that ``local_t`` based operations are not recommended for general19 kernel use. Please use the ``this_cpu`` operations instead unless there is20 really a special purpose. Most uses of ``local_t`` in the kernel have been21 replaced by ``this_cpu`` operations. ``this_cpu`` operations combine the22 relocation with the ``local_t`` like semantics in a single instruction and23 yield more compact and faster executing code.24 25 26Purpose of local atomic operations27==================================28 29Local atomic operations are meant to provide fast and highly reentrant per CPU30counters. They minimize the performance cost of standard atomic operations by31removing the LOCK prefix and memory barriers normally required to synchronize32across CPUs.33 34Having fast per CPU atomic counters is interesting in many cases: it does not35require disabling interrupts to protect from interrupt handlers and it permits36coherent counters in NMI handlers. It is especially useful for tracing purposes37and for various performance monitoring counters.38 39Local atomic operations only guarantee variable modification atomicity wrt the40CPU which owns the data. Therefore, care must taken to make sure that only one41CPU writes to the ``local_t`` data. This is done by using per cpu data and42making sure that we modify it from within a preemption safe context. It is43however permitted to read ``local_t`` data from any CPU: it will then appear to44be written out of order wrt other memory writes by the owner CPU.45 46 47Implementation for a given architecture48=======================================49 50It can be done by slightly modifying the standard atomic operations: only51their UP variant must be kept. It typically means removing LOCK prefix (on52i386 and x86_64) and any SMP synchronization barrier. If the architecture does53not have a different behavior between SMP and UP, including54``asm-generic/local.h`` in your architecture's ``local.h`` is sufficient.55 56The ``local_t`` type is defined as an opaque ``signed long`` by embedding an57``atomic_long_t`` inside a structure. This is made so a cast from this type to58a ``long`` fails. The definition looks like::59 60 typedef struct { atomic_long_t a; } local_t;61 62 63Rules to follow when using local atomic operations64==================================================65 66* Variables touched by local ops must be per cpu variables.67* *Only* the CPU owner of these variables must write to them.68* This CPU can use local ops from any context (process, irq, softirq, nmi, ...)69 to update its ``local_t`` variables.70* Preemption (or interrupts) must be disabled when using local ops in71 process context to make sure the process won't be migrated to a72 different CPU between getting the per-cpu variable and doing the73 actual local op.74* When using local ops in interrupt context, no special care must be75 taken on a mainline kernel, since they will run on the local CPU with76 preemption already disabled. I suggest, however, to explicitly77 disable preemption anyway to make sure it will still work correctly on78 -rt kernels.79* Reading the local cpu variable will provide the current copy of the80 variable.81* Reads of these variables can be done from any CPU, because updates to82 "``long``", aligned, variables are always atomic. Since no memory83 synchronization is done by the writer CPU, an outdated copy of the84 variable can be read when reading some *other* cpu's variables.85 86 87How to use local atomic operations88==================================89 90::91 92 #include <linux/percpu.h>93 #include <asm/local.h>94 95 static DEFINE_PER_CPU(local_t, counters) = LOCAL_INIT(0);96 97 98Counting99========100 101Counting is done on all the bits of a signed long.102 103In preemptible context, use ``get_cpu_var()`` and ``put_cpu_var()`` around104local atomic operations: it makes sure that preemption is disabled around write105access to the per cpu variable. For instance::106 107 local_inc(&get_cpu_var(counters));108 put_cpu_var(counters);109 110If you are already in a preemption-safe context, you can use111``this_cpu_ptr()`` instead::112 113 local_inc(this_cpu_ptr(&counters));114 115 116 117Reading the counters118====================119 120Those local counters can be read from foreign CPUs to sum the count. Note that121the data seen by local_read across CPUs must be considered to be out of order122relatively to other memory writes happening on the CPU that owns the data::123 124 long sum = 0;125 for_each_online_cpu(cpu)126 sum += local_read(&per_cpu(counters, cpu));127 128If you want to use a remote local_read to synchronize access to a resource129between CPUs, explicit ``smp_wmb()`` and ``smp_rmb()`` memory barriers must be used130respectively on the writer and the reader CPUs. It would be the case if you use131the ``local_t`` variable as a counter of bytes written in a buffer: there should132be a ``smp_wmb()`` between the buffer write and the counter increment and also a133``smp_rmb()`` between the counter read and the buffer read.134 135 136Here is a sample module which implements a basic per cpu counter using137``local.h``::138 139 /* test-local.c140 *141 * Sample module for local.h usage.142 */143 144 145 #include <asm/local.h>146 #include <linux/module.h>147 #include <linux/timer.h>148 149 static DEFINE_PER_CPU(local_t, counters) = LOCAL_INIT(0);150 151 static struct timer_list test_timer;152 153 /* IPI called on each CPU. */154 static void test_each(void *info)155 {156 /* Increment the counter from a non preemptible context */157 printk("Increment on cpu %d\n", smp_processor_id());158 local_inc(this_cpu_ptr(&counters));159 160 /* This is what incrementing the variable would look like within a161 * preemptible context (it disables preemption) :162 *163 * local_inc(&get_cpu_var(counters));164 * put_cpu_var(counters);165 */166 }167 168 static void do_test_timer(unsigned long data)169 {170 int cpu;171 172 /* Increment the counters */173 on_each_cpu(test_each, NULL, 1);174 /* Read all the counters */175 printk("Counters read from CPU %d\n", smp_processor_id());176 for_each_online_cpu(cpu) {177 printk("Read : CPU %d, count %ld\n", cpu,178 local_read(&per_cpu(counters, cpu)));179 }180 mod_timer(&test_timer, jiffies + 1000);181 }182 183 static int __init test_init(void)184 {185 /* initialize the timer that will increment the counter */186 timer_setup(&test_timer, do_test_timer, 0);187 mod_timer(&test_timer, jiffies + 1);188 189 return 0;190 }191 192 static void __exit test_exit(void)193 {194 timer_shutdown_sync(&test_timer);195 }196 197 module_init(test_init);198 module_exit(test_exit);199 200 MODULE_LICENSE("GPL");201 MODULE_AUTHOR("Mathieu Desnoyers");202 MODULE_DESCRIPTION("Local Atomic Ops");203