1This directory contains the following litmus tests: 2 3CoRR+poonceonce+Once.litmus 4 Test of read-read coherence, that is, whether or not two 5 successive reads from the same variable are ordered. 6 7CoRW+poonceonce+Once.litmus 8 Test of read-write coherence, that is, whether or not a read 9 from a given variable followed by a write to that same variable 10 are ordered. 11 12CoWR+poonceonce+Once.litmus 13 Test of write-read coherence, that is, whether or not a write 14 to a given variable followed by a read from that same variable 15 are ordered. 16 17CoWW+poonceonce.litmus 18 Test of write-write coherence, that is, whether or not two 19 successive writes to the same variable are ordered. 20 21IRIW+mbonceonces+OnceOnce.litmus 22 Test of independent reads from independent writes with smp_mb() 23 between each pairs of reads. In other words, is smp_mb() 24 sufficient to cause two different reading processes to agree on 25 the order of a pair of writes, where each write is to a different 26 variable by a different process? 27 28IRIW+poonceonces+OnceOnce.litmus 29 Test of independent reads from independent writes with nothing 30 between each pairs of reads. In other words, is anything at all 31 needed to cause two different reading processes to agree on the 32 order of a pair of writes, where each write is to a different 33 variable by a different process? 34 35ISA2+pooncelock+pooncelock+pombonce.litmus 36 Tests whether the ordering provided by a lock-protected S 37 litmus test is visible to an external process whose accesses are 38 separated by smp_mb(). This addition of an external process to 39 S is otherwise known as ISA2. 40 41ISA2+poonceonces.litmus 42 As below, but with store-release replaced with WRITE_ONCE() 43 and load-acquire replaced with READ_ONCE(). 44 45ISA2+pooncerelease+poacquirerelease+poacquireonce.litmus 46 Can a release-acquire chain order a prior store against 47 a later load? 48 49LB+ctrlonceonce+mbonceonce.litmus 50 Does a control dependency and an smp_mb() suffice for the 51 load-buffering litmus test, where each process reads from one 52 of two variables then writes to the other? 53 54LB+poacquireonce+pooncerelease.litmus 55 Does a release-acquire pair suffice for the load-buffering 56 litmus test, where each process reads from one of two variables then 57 writes to the other? 58 59LB+poonceonces.litmus 60 As above, but with store-release replaced with WRITE_ONCE() 61 and load-acquire replaced with READ_ONCE(). 62 63MP+onceassign+derefonce.litmus 64 As below, but with rcu_assign_pointer() and an rcu_dereference(). 65 66MP+polockmbonce+poacquiresilsil.litmus 67 Protect the access with a lock and an smp_mb__after_spinlock() 68 in one process, and use an acquire load followed by a pair of 69 spin_is_locked() calls in the other process. 70 71MP+polockonce+poacquiresilsil.litmus 72 Protect the access with a lock in one process, and use an 73 acquire load followed by a pair of spin_is_locked() calls 74 in the other process. 75 76MP+polocks.litmus 77 As below, but with the second access of the writer process 78 and the first access of reader process protected by a lock. 79 80MP+poonceonces.litmus 81 As below, but without the smp_rmb() and smp_wmb(). 82 83MP+pooncerelease+poacquireonce.litmus 84 As below, but with a release-acquire chain. 85 86MP+porevlocks.litmus 87 As below, but with the first access of the writer process 88 and the second access of reader process protected by a lock. 89 90MP+wmbonceonce+rmbonceonce.litmus 91 Does a smp_wmb() (between the stores) and an smp_rmb() (between 92 the loads) suffice for the message-passing litmus test, where one 93 process writes data and then a flag, and the other process reads 94 the flag and then the data. (This is similar to the ISA2 tests, 95 but with two processes instead of three.) 96 97R+mbonceonces.litmus 98 This is the fully ordered (via smp_mb()) version of one of 99 the classic counterintuitive litmus tests that illustrates the 100 effects of store propagation delays. 101 102R+poonceonces.litmus 103 As above, but without the smp_mb() invocations. 104 105SB+mbonceonces.litmus 106 This is the fully ordered (again, via smp_mb() version of store 107 buffering, which forms the core of Dekker's mutual-exclusion 108 algorithm. 109 110SB+poonceonces.litmus 111 As above, but without the smp_mb() invocations. 112 113S+poonceonces.litmus 114 As below, but without the smp_wmb() and acquire load. 115 116S+wmbonceonce+poacquireonce.litmus 117 Can a smp_wmb(), instead of a release, and an acquire order 118 a prior store against a subsequent store? 119 120WRC+poonceonces+Once.litmus 121WRC+pooncerelease+rmbonceonce+Once.litmus 122 These two are members of an extension of the MP litmus-test class 123 in which the first write is moved to a separate process. 124 125Z6.0+pooncelock+pooncelock+pombonce.litmus 126 Is the ordering provided by a spin_unlock() and a subsequent 127 spin_lock() sufficient to make ordering apparent to accesses 128 by a process not holding the lock? 129 130Z6.0+pooncelock+poonceLock+pombonce.litmus 131 As above, but with smp_mb__after_spinlock() immediately 132 following the spin_lock(). 133 134Z6.0+pooncerelease+poacquirerelease+mbonceonce.litmus 135 Is the ordering provided by a release-acquire chain sufficient 136 to make ordering apparent to accesses by a process that does 137 not participate in that release-acquire chain? 138 139A great many more litmus tests are available here: 140 141 https://github.com/paulmckrcu/litmus 142