Non-ConfidentialPDF file icon PDF versionDUI0801J
Arm® Compiler armasm User GuideVersion 6.6
Home > Floating-point Programming > Floating-point operations not affected by flush-to-zero mode

10.16 Floating-point operations not affected by flush-to-zero mode

Some floating-point instructions can be carried out on denormalized numbers even in flush-to-zero mode, without flushing the results to zero.

These instructions are as follows:

  • Absolute value and negate (VABS and VNEG).
  • Load and store (VLDR and VSTR).
  • Load multiple and store multiple (VLDM and VSTM).
  • Transfer between extension registers and Arm® general-purpose registers (VMOV).
Related concepts
10.13 Flush-to-zero mode in floating-point
Related references
15.2 VABS (floating-point)
15.14 VLDM (floating-point)
15.15 VLDR (floating-point)
15.39 VSTM (floating-point)
15.40 VSTR (floating-point)
14.51 VLDM
14.52 VLDR
14.126 VSTM
14.129 VSTR
15.24 VMOV (between one general-purpose register and single precision floating-point register)
14.67 VMOV (between two general-purpose registers and a 64-bit extension register)
15.30 VNEG (floating-point)
14.80 VNEG
Non-ConfidentialPDF file icon PDF versionDUI0801J
Copyright © 2014–2017, 2019 Arm Limited or its affiliates. All rights reserved. 
Home > Floating-point Programming > Floating-point operations not affected by flush-to-zero mode