4.6.4. NEON in AArch32 execution state.

In AArch32, the smaller registers are packed into larger ones (D0 and D1 are combined to form Q1, for instance). This introduces some tricky loop-carried dependencies which can reduce the ability of the compiler to vectorize loop structures.

Figure 4.13. Arrangement of ARMv7 SIMD registers

To view this graphic, your browser must support the SVG format. Either install a browser with native support, or install an appropriate plugin such as Adobe SVG Viewer.

The floating-point and Advanced SIMD registers in AArch32 are mapped into the AArch64 FP and SIMD registers. This is done to allow the floating-point and NEON registers of an application or a virtual machine to be interpreted (and, as necessary, modified) by a higher level of system software, for example, the OS or the Hypervisor.

The AArch64 V16-V31 FP and NEON registers are not accessible from AArch32. As with the general-purpose registers, during execution in an Exception level using AArch32 these registers retain their state from the previous execution using AArch64.

Copyright © 2015 ARM. All rights reserved.ARM DEN0024A