linear Clause

Clause for several omp pragmas: Specifies that all variables in a list are private to a SIMD chunk and that they have a linear relationship within the iteration space of a loop.

linear (linear-list[:linear-step][, linear-list:[linear-step]...])

linear-step

Is a compile-time positive, integer, scalar constant expression. If linear-step is not specified, it is assumed to be 1.

linear-list

Takes one of the following forms:

list-item
modifier (list-item)

Where list is a formal parameter of the specified function or the C++ keyword this.

modifier-list

Determines how a list item is initialized. Possible values are:

ref	Specifies that the storage location of each list item on each lane corresponds to an array at the storage location upon entry to the function indexed by the logical number of the lane times linear-step.
val	Specifies that the value of each list item on each lane corresponds to the value of the list item upon entry to the function plus the logical number of the lane times linear-step.
uval	Similar to val, but each invocation uses the same storage location for each SIMD lane. This storage location is updated with the final value of the logically last lane.

The behavior of the linear clause is dependent on when the clause is specified and whether or not a modifier is specified.

When a linear clause is specified on a construct:

The value of the new list item on each iteration of the associated loop(s) is the sum of the original list item before entering the construct, added to the logical number of the iteration, multiplied by linear-step. The value of the last iteration of the associated loop(s) in the sequence is assigned to the original list item.

When a linear clause is specified on a declarative directive:

All list items must be formal parameters of a function that will be invoked concurrently on each SIMD lane.

When the ref modifier is specified:

The storage location of the list items on each lane corresponds to an array at the storage location upon entry, to the function indexed by the logical number of the lane, multiplied by linear-step.

When the uval modifier is specified:

Each invocation uses the same storage location for each SIMD lane, and the storage location is updated with the final value of the logically last lane.

When no modifier is specified or the val or uval modifier is specified:

The value of the list items on each lane is the sum of the list item upon entry to the function added to the logical number of the lane, multiplied by linear-step.

If modifier is not specified, or if the val or uval modifier is specified, the value of each list item on each lane is the sum of the list item upon entry to the function added to the logical number of the lane, multiplied by linear-step.

You can only use ref, val, or uval if the list item is of a reference type.

If a list item appears without the ref modifier in a linear clause, the list item must be of integral or pointer type, or must be a reference to an integral or pointer type. Additionally, when a list item is of a reference type and appears without the ref modifier, if any write to the list item occurs before any read of the list item, the result is unspecified

A list item in a linear clause on a worksharing construct with a reference type must bind to the same object for all threads of the team. This restriction does not apply to pragma omp declare simd.