Andrea Parri | 48d44d4 | 2018-02-20 15:25:01 -0800 | [diff] [blame] | 1 | ===================================== |
| 2 | LINUX KERNEL MEMORY CONSISTENCY MODEL |
| 3 | ===================================== |
Paul E. McKenney | 1c27b64 | 2018-01-18 19:58:55 -0800 | [diff] [blame] | 4 | |
| 5 | ============ |
| 6 | INTRODUCTION |
| 7 | ============ |
| 8 | |
Andrea Parri | 48d44d4 | 2018-02-20 15:25:01 -0800 | [diff] [blame] | 9 | This directory contains the memory consistency model (memory model, for |
| 10 | short) of the Linux kernel, written in the "cat" language and executable |
| 11 | by the externally provided "herd7" simulator, which exhaustively explores |
| 12 | the state space of small litmus tests. |
Paul E. McKenney | 1c27b64 | 2018-01-18 19:58:55 -0800 | [diff] [blame] | 13 | |
| 14 | In addition, the "klitmus7" tool (also externally provided) may be used |
| 15 | to convert a litmus test to a Linux kernel module, which in turn allows |
| 16 | that litmus test to be exercised within the Linux kernel. |
| 17 | |
| 18 | |
| 19 | ============ |
| 20 | REQUIREMENTS |
| 21 | ============ |
| 22 | |
Akira Yokosawa | 5b62832c | 2018-05-14 16:33:41 -0700 | [diff] [blame^] | 23 | Version 7.49 of the "herd7" and "klitmus7" tools must be downloaded |
Paul E. McKenney | 8f7f2fb | 2018-02-20 15:25:09 -0800 | [diff] [blame] | 24 | separately: |
Paul E. McKenney | 1c27b64 | 2018-01-18 19:58:55 -0800 | [diff] [blame] | 25 | |
| 26 | https://github.com/herd/herdtools7 |
| 27 | |
| 28 | See "herdtools7/INSTALL.md" for installation instructions. |
| 29 | |
Paul E. McKenney | 1c27b64 | 2018-01-18 19:58:55 -0800 | [diff] [blame] | 30 | |
| 31 | ================== |
| 32 | BASIC USAGE: HERD7 |
| 33 | ================== |
| 34 | |
| 35 | The memory model is used, in conjunction with "herd7", to exhaustively |
| 36 | explore the state space of small litmus tests. |
| 37 | |
| 38 | For example, to run SB+mbonceonces.litmus against the memory model: |
| 39 | |
| 40 | $ herd7 -conf linux-kernel.cfg litmus-tests/SB+mbonceonces.litmus |
| 41 | |
| 42 | Here is the corresponding output: |
| 43 | |
| 44 | Test SB+mbonceonces Allowed |
| 45 | States 3 |
| 46 | 0:r0=0; 1:r0=1; |
| 47 | 0:r0=1; 1:r0=0; |
| 48 | 0:r0=1; 1:r0=1; |
| 49 | No |
| 50 | Witnesses |
| 51 | Positive: 0 Negative: 3 |
| 52 | Condition exists (0:r0=0 /\ 1:r0=0) |
| 53 | Observation SB+mbonceonces Never 0 3 |
| 54 | Time SB+mbonceonces 0.01 |
| 55 | Hash=d66d99523e2cac6b06e66f4c995ebb48 |
| 56 | |
| 57 | The "Positive: 0 Negative: 3" and the "Never 0 3" each indicate that |
| 58 | this litmus test's "exists" clause can not be satisfied. |
| 59 | |
| 60 | See "herd7 -help" or "herdtools7/doc/" for more information. |
| 61 | |
| 62 | |
| 63 | ===================== |
| 64 | BASIC USAGE: KLITMUS7 |
| 65 | ===================== |
| 66 | |
| 67 | The "klitmus7" tool converts a litmus test into a Linux kernel module, |
| 68 | which may then be loaded and run. |
| 69 | |
| 70 | For example, to run SB+mbonceonces.litmus against hardware: |
| 71 | |
| 72 | $ mkdir mymodules |
| 73 | $ klitmus7 -o mymodules litmus-tests/SB+mbonceonces.litmus |
| 74 | $ cd mymodules ; make |
| 75 | $ sudo sh run.sh |
| 76 | |
| 77 | The corresponding output includes: |
| 78 | |
| 79 | Test SB+mbonceonces Allowed |
| 80 | Histogram (3 states) |
| 81 | 644580 :>0:r0=1; 1:r0=0; |
| 82 | 644328 :>0:r0=0; 1:r0=1; |
| 83 | 711092 :>0:r0=1; 1:r0=1; |
| 84 | No |
| 85 | Witnesses |
| 86 | Positive: 0, Negative: 2000000 |
| 87 | Condition exists (0:r0=0 /\ 1:r0=0) is NOT validated |
| 88 | Hash=d66d99523e2cac6b06e66f4c995ebb48 |
| 89 | Observation SB+mbonceonces Never 0 2000000 |
| 90 | Time SB+mbonceonces 0.16 |
| 91 | |
| 92 | The "Positive: 0 Negative: 2000000" and the "Never 0 2000000" indicate |
| 93 | that during two million trials, the state specified in this litmus |
| 94 | test's "exists" clause was not reached. |
| 95 | |
| 96 | And, as with "herd7", please see "klitmus7 -help" or "herdtools7/doc/" |
| 97 | for more information. |
| 98 | |
| 99 | |
| 100 | ==================== |
| 101 | DESCRIPTION OF FILES |
| 102 | ==================== |
| 103 | |
| 104 | Documentation/cheatsheet.txt |
| 105 | Quick-reference guide to the Linux-kernel memory model. |
| 106 | |
| 107 | Documentation/explanation.txt |
| 108 | Describes the memory model in detail. |
| 109 | |
| 110 | Documentation/recipes.txt |
| 111 | Lists common memory-ordering patterns. |
| 112 | |
| 113 | Documentation/references.txt |
| 114 | Provides background reading. |
| 115 | |
| 116 | linux-kernel.bell |
| 117 | Categorizes the relevant instructions, including memory |
| 118 | references, memory barriers, atomic read-modify-write operations, |
| 119 | lock acquisition/release, and RCU operations. |
| 120 | |
| 121 | More formally, this file (1) lists the subtypes of the various |
| 122 | event types used by the memory model and (2) performs RCU |
| 123 | read-side critical section nesting analysis. |
| 124 | |
| 125 | linux-kernel.cat |
| 126 | Specifies what reorderings are forbidden by memory references, |
| 127 | memory barriers, atomic read-modify-write operations, and RCU. |
| 128 | |
| 129 | More formally, this file specifies what executions are forbidden |
| 130 | by the memory model. Allowed executions are those which |
| 131 | satisfy the model's "coherence", "atomic", "happens-before", |
| 132 | "propagation", and "rcu" axioms, which are defined in the file. |
| 133 | |
| 134 | linux-kernel.cfg |
| 135 | Convenience file that gathers the common-case herd7 command-line |
| 136 | arguments. |
| 137 | |
| 138 | linux-kernel.def |
| 139 | Maps from C-like syntax to herd7's internal litmus-test |
| 140 | instruction-set architecture. |
| 141 | |
| 142 | litmus-tests |
| 143 | Directory containing a few representative litmus tests, which |
| 144 | are listed in litmus-tests/README. A great deal more litmus |
| 145 | tests are available at https://github.com/paulmckrcu/litmus. |
| 146 | |
| 147 | lock.cat |
| 148 | Provides a front-end analysis of lock acquisition and release, |
| 149 | for example, associating a lock acquisition with the preceding |
| 150 | and following releases and checking for self-deadlock. |
| 151 | |
| 152 | More formally, this file defines a performance-enhanced scheme |
| 153 | for generation of the possible reads-from and coherence order |
| 154 | relations on the locking primitives. |
| 155 | |
| 156 | README |
| 157 | This file. |
| 158 | |
| 159 | |
| 160 | =========== |
| 161 | LIMITATIONS |
| 162 | =========== |
| 163 | |
| 164 | The Linux-kernel memory model has the following limitations: |
| 165 | |
| 166 | 1. Compiler optimizations are not modeled. Of course, the use |
| 167 | of READ_ONCE() and WRITE_ONCE() limits the compiler's ability |
| 168 | to optimize, but there is Linux-kernel code that uses bare C |
| 169 | memory accesses. Handling this code is on the to-do list. |
| 170 | For more information, see Documentation/explanation.txt (in |
| 171 | particular, the "THE PROGRAM ORDER RELATION: po AND po-loc" |
| 172 | and "A WARNING" sections). |
| 173 | |
| 174 | 2. Multiple access sizes for a single variable are not supported, |
| 175 | and neither are misaligned or partially overlapping accesses. |
| 176 | |
| 177 | 3. Exceptions and interrupts are not modeled. In some cases, |
| 178 | this limitation can be overcome by modeling the interrupt or |
| 179 | exception with an additional process. |
| 180 | |
| 181 | 4. I/O such as MMIO or DMA is not supported. |
| 182 | |
| 183 | 5. Self-modifying code (such as that found in the kernel's |
| 184 | alternatives mechanism, function tracer, Berkeley Packet Filter |
| 185 | JIT compiler, and module loader) is not supported. |
| 186 | |
| 187 | 6. Complete modeling of all variants of atomic read-modify-write |
| 188 | operations, locking primitives, and RCU is not provided. |
| 189 | For example, call_rcu() and rcu_barrier() are not supported. |
| 190 | However, a substantial amount of support is provided for these |
| 191 | operations, as shown in the linux-kernel.def file. |
| 192 | |
| 193 | The "herd7" tool has some additional limitations of its own, apart from |
| 194 | the memory model: |
| 195 | |
| 196 | 1. Non-trivial data structures such as arrays or structures are |
| 197 | not supported. However, pointers are supported, allowing trivial |
| 198 | linked lists to be constructed. |
| 199 | |
| 200 | 2. Dynamic memory allocation is not supported, although this can |
| 201 | be worked around in some cases by supplying multiple statically |
| 202 | allocated variables. |
| 203 | |
| 204 | Some of these limitations may be overcome in the future, but others are |
| 205 | more likely to be addressed by incorporating the Linux-kernel memory model |
| 206 | into other tools. |