The following pragmas are specific to the Intel® C++ Compiler:
| Pragma | Description |
|---|---|
| alloc_section | allocate variable in specified section |
| force_align | specifies the alignment of a class type |
| intel_omp_task | specifies a unit of work, potentially executed by a different thread |
| intel_omp_taskq | specifies an environment for the while loop in which to enqueue the units of work specified by the enclosed task pragma |
| ivdep | instructs the compiler to ignore assumed vector dependencies |
| loop count | indicates the loop count is likely to be an integer. |
| memref_control | provides a method for controlling load latency at the variable level |
| noprefetch | asserts that the data prefetches are not generated for some memory references |
| nounroll | instructs the compiler not to unroll a loop |
| novector | specifies that the loop should never be vectorized |
| optimization_level | enables control of optimization for a specific function |
| prefetch | asserts that the data prefetches are generated for some memory references |
| unused | describes variables that are unused (warnings not generated) |
| unroll | tells the compiler how many times to unroll a loop |
| vector | specifies how to vectorize the loop and indicates that efficiency heuristics should be ignored |