Concurrency Algorithms and Theories 03 Operational

Basic operational semantics for concurrency

Basic domains:

Expressions and commands

Programs

P : Tid $\to$ Cmd, written as P = $c_1$ || … || $c_N$

Thread subsystem

• Thread-local steps: c,s $\stackrel{l}{\to }$ c’, s’
• Interpret sequential programs
• Lift them to program steps: P, S $\stackrel{i:l}{\to }$ P’, S’

Storage subsystem (defined by the memory model)

• Describe the effect of memory accesses and fences
• M $\stackrel{i:l}{\to }$ M’ is the state of storage subsystem

Linking the two

• Either the thread or the storage subsystem make an internal step, $\epsilon$; or they make matching i : l steps
• P,S,M $\Rightarrow$ P’, S’, M’

The thread subsystem

Store: s : Reg $\to$ Val (Initial store: $s_0\stackrel{△}{=}{\lambda }_r.0$ )

State: <c,s> $\in$ Command x Store

Transitions:

The thread subsystem: RMW and fence commands

Lifting to concurrent programs

State: $⟨P,S⟩\in Program\times (Tid\to Store)$

• Initial stores: $S_0\stackrel{△}{=}{\lambda }_i.s_0$
• Initial state: $⟨P,S_0⟩$

Transition:

$\frac{P(i),S(i)\stackrel{l}{\to }c,s}{P,S\stackrel{i:l}{\to }P[i\mapsto c],S[i\mapsto s]}$

SC storage subsystem

Machine state: M : Loc $\to$ Val

• Maps each location to its value

• Initial state: $M_0\stackrel{△}{=}{\lambda }_X.0$

  i.e., the memory that maps every location to 0

TSO storage subsystem

The state consists of:

• A memory M: Loc $\to$ Val

• A function B: Tid $\to$ (Loc x Val)*

  assigning a store buffer to every thread.

**Initial state: ** $⟨M_0,B_0⟩$ where

• $M_0$ $={\lambda }_X.0$ (the memory maps 0 to every location)
• $B_0={\lambda }_i.ϵ$ (all store buffers are empty)

TSO storage subsystem transitions

TSO: linking thread and storage subsystems

Definition (Allowed outcome)

An outcome O is allowed for a program P under TSO if there exist M such that $P$, $S_0$, $M_0$, $B_0{\Rightarrow }^{\ast }$ skip || … || skip , $O,M,B_0$

Exercise: PSO storage subsystemd