05.02.2013 Views

ARM Architecture Reference Manual ARMv7-A and ARMv7-R edition

ARM Architecture Reference Manual ARMv7-A and ARMv7-R edition

ARM Architecture Reference Manual ARMv7-A and ARMv7-R edition

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

exponent == 0x7FF<br />

fraction != 0<br />

The value is a denormalized number <strong>and</strong> is equal to:<br />

1 –S × 2 –1022 × (0.fraction)<br />

Application Level Programmers’ Model<br />

The minimum positive denormalized number is 2 –1074 , or approximately 4.941 × 10 –324 .<br />

Optionally, denormalized numbers are flushed to zero in the VFP extension. For details see<br />

Flush-to-zero on page A2-39.<br />

The value is either an infinity or a NaN, depending on the fraction bits:<br />

fraction == 0<br />

the value is an infinity. As for single-precision, there are two infinities:<br />

+∞ Plus infinity, when S==0<br />

-∞ Minus infinity, when S==1.<br />

fraction != 0<br />

The value is a NaN, <strong>and</strong> is either a quiet NaN or a signaling NaN.<br />

In the VFP architecture, the two types of NaN are distinguished on the basis of<br />

their most significant fraction bit, bit [19] of the most significant word:<br />

bit [19] == 0<br />

The NaN is a signaling NaN. The sign bit can take any value, <strong>and</strong><br />

the remaining fraction bits can take any value except all zeros.<br />

bit [19] == 1<br />

The NaN is a quiet NaN. The sign bit <strong>and</strong> the remaining fraction bits<br />

can take any value.<br />

For details of the default NaN see NaN h<strong>and</strong>ling <strong>and</strong> the Default NaN on page A2-41.<br />

<strong>ARM</strong> DDI 0406B Copyright © 1996-1998, 2000, 2004-2008 <strong>ARM</strong> Limited. All rights reserved. A2-37

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!