Handling floating-point operations

A method of operating a data processing apparatus comprising:

• performing a first comparison and a second comparison in dependence on a more significant portion of a first floating-point operand and of a second floating-point operand, wherein performing the first comparison comprises comparing the more significant portion of the first floating-point operand and of the second floating-point operand, and wherein performing the second comparison comprises comparing a test value given by the more significant portion of the first floating-point operand minus a least significant bit in the more significant portion with the more significant portion of the second floating-point operand;
• performing a third comparison of a less significant portion of the first floating-point operand and of the second floating-point operand;
• generating a subnormal flag signal in response to either the first comparison or the second comparison revealing no difference and the second comparison revealing a difference in the less significant portion;
• calculating a difference between the first floating-point operand and the second floating-point operand;
• and
• performing a 1-bit right shift on the difference calculated when the subnormal flag signal is generated.

A data processing apparatus comprising:

• means for performing a first comparison and a second comparison in dependence on a more significant portion of a first floating-point operand and of a second floating-point operand, wherein performing the first comparison comprises comparing the more significant portion of the first floating-point operand and of the second floating-point operand, and wherein performing the second comparison comprises comparing a test value given by the more significant portion of the first floating-point operand minus a least significant bit in the more significant portion with the more significant portion of the second floating-point operand;
• means for performing a third comparison of a less significant portion of the first floating-point operand and of the second floating-point operand;
• means for generating a subnormal flag signal in response to either the first comparison or the second comparison revealing no difference and the second comparison revealing a difference in the less significant portion;
• means for calculating a difference between the first floating-point operand and the second floating-point operand;
• and
• means for performing a 1-bit right shift on the difference calculated in response to the means for generating the subnormal flag signal generating the subnormal flag.