|
|||
| Products Download Events Support Videos | |||
Technical Support
On-Line Manuals
Assembler User Guide
Conventions and feedback
Overview of the Assembler
Overview of the ARM Architecture
Structure of Assembly Language Modules
Writing ARM Assembly Language
Condition Codes
Using the Assembler
Symbols, Literals, Expressions, and Operators
VFP Programming
Architecture support for VFP
Half-precision extension
Fused Multiply-Add extension
Extension register bank mapping
VFP views of the extension register bank
Load values to VFP registers
Conditional execution of VFP instructions
Floating-point exceptions
VFP data types
Extended notation
VFP system registers
FPSCR, the floating-point status and control regis
FPEXC, the floating-point exception register
FPSID, the floating-point system ID register
Flush-to-zero mode
When to use flush-to-zero mode
The effects of using flush-to-zero mode
Operations not affected by flush-to-zero mode
VFP vector mode
Vectors in the VFP extension register bank
VFP vector wrap-around
VFP vector stride
Restriction on vector length
Control of scalar, vector, and mixed operations
VFP directives and vector notation
Pre-UAL VFP mnemonics
Vector notation
VFPASSERT SCALAR
VFPASSERT VECTOR
Assembler User Guide
Vectors in the VFP extension register bank
| |||
| Home > VFP Programming > Vectors in the VFP extension register bank | |||
For vector operations, the VFP extension register bank can be viewed as a collection of smaller banks. Each of these smaller banks is treated either as a bank of 8 single-precision registers or as a bank of 4 double-precision registers.
In VFPv2, VFPv3-D16, and VFPv3-D16-FP16 the VFP extension register bank can be viewed as a collection of:
four banks of single-precision registers, s0 to s7, s8 to s15, s16 to s23, and s24 to s31
four banks of double-precision registers, d0 to d3, d4 to d7, d8 to d11, and d12 to d15
any combination of single-precision and double-precision banks.
In VFPv3 and VFPv3-FP16, the VFP extension register bank can be viewed as a collection of:
four banks of single-precision registers, s0 to s7, s8 to s15, s16 to s23, and s24 to s31
Eight banks of double-precision registers, d0 to d3, d4 to d7, d8 to d11, d12 to d15, d16 to d19, d20 to d23, d24 to d27, and d28 to d31
any combination of single-precision and double-precision banks.
See Figure 3 and Figure 4 for further clarification.
A vector, in a VFP instruction, can use up to eight single-precision
registers, or four double-precision registers, from the same bank.
The number of registers used by a vector is controlled by the LEN bits
in the FPSCR.
Note
The value of the LEN bits is not a sufficient
condition to perform vector operations using VFP. Whether a VFP
operation is scalar, vector or mixed depends on which bank the specified operand
and destination registers are in.
A vector can start from any register and wraps around to the beginning of the bank. The first register used by an operand vector is the register that is specified as the operand in the individual VFP instructions. The first register used by the destination vector is the register that is specified as the destination in the individual VFP instructions.
See also| Copyright © 2007-2008, 2011-2012 ARM. All rights reserved. | ARM DUI 0379D |
| Non-Confidential | ID062912 |
 PDF version | |
| Home > VFP Programming > Vectors in the VFP extension register bank |
ProductsDevelopment Tools |
Hardware & Collateral |
Downloads |
Support |
Contact |
|
 |
|||
