ia-32_volume1_basic-arch

Precision format if snan is in single precision and

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: omputation exception occurs. Unmasked Response and Exception Code src1, src2 unchanged; #DE = 1 Note that SQRT, CVTPS2PD, CVTSS2SD, CVTPD2PS, CVTSD2SS have only 1 src. E-16 Vol. 1 GUIDELINES FOR WRITING SIMD FLOATING-POINT EXCEPTION HANDLERS Table E-15. #D - Denormal Operand Instruction CVTSS2SD CVTPD2PS CVTSD2SS NOTE: 1. For denormal encodings, see Section 4.8.3.2, "Normalized and Denormalized Finite Numbers." Condition Masked Response Unmasked Response and Exception Code Table E-16. #O - Numeric Overflow Instruction ADDPS ADDSUBPS HADDPS SUBPS HSUBPS MULPS DIVPS ADDSS SUBSS MULSS DIVSS CVTPD2PS CVTSD2SS Condition Rounded result > largest single precision finite normal value Masked Response Roundi ng To nearest Toward Sign Unmasked Response and Exception Code res = (result calculated with unbounded exponent and rounded to the destination precision) / 2192 #OE = 1 #PE = 1 if the result is inexact Result & Status Flags #OE = 1, #PE = 1 res = + res = #OE = 1, #PE = 1 res = 1.11...1 * 2127 res = #OE = 1, #PE = 1 res = + res = -1.11...1 * 2127 #OE = 1, #PE = 1 res = 1.11...1 * 2127 res = -1.11...1 * 2127 + + + + - Toward + Toward 0 Vol. 1 E-17 GUIDELINES FOR WRITING SIMD FLOATING-POINT EXCEPTION HANDLERS Table E-16. #O - Numeric Overflow (Contd.) Instruction ADDPD ADDSUBPD HADDPD SUBPD HSUBPD MULPD DIVPD ADDSD SUBSD MULSD DIVSD Condition Rounded result > largest double precision finite normal value Masked Response Roundi ng To nearest Toward Sign Unmasked Response and Exception Code res = (result calculated with unbounded exponent and rounded to the destination precision) / 21536 #OE = 1 #PE = 1 if the result is inexact Result & Status Flags #OE = 1, #PE = 1 res = + res = #OE = 1, #PE = 1 res = 1.11...1 * 21023 res = #OE = 1, #PE = 1 res = + res = -1.11...1 * 21023 #OE = 1, #PE = 1 res = 1.11...1 * 21023 res = -1.11...1 * 21023 + + - Toward + + - Toward 0 + - E-18 Vol. 1 GUIDELINES FOR WRITING SIMD FLOATING-POINT EXCEPTION HANDLERS Table E-17. #U - Numeric Underflow Instruction ADDPS ADDSUBPS HADDPS SUBPS HSUBPS MULPS DIVPS ADDSS SUBSS MULSS DIVSS CVTPD2PS CVTSD2SS ADDPD ADDSUBPD HADDPD SUBPD HSUBPD MULPD DIVPD ADDSD SUBSD MULSD DIVSD Condition Result calculated with unbounded exponent and rounded to the destination precision < smallest single precision finite normal value. Masked Response res = 0, denormal, or normal #UE = 1 and #PE = 1, but only if the result is inexact Unmasked Response and Exception Code res = (result calculated with unbounded exponent and rounded to the destination precision) * 2192 #UE = 1 #PE = 1 if the result is inexact Result calculated with unbounded exponent and rounded to the destination precision < smallest double precision finite normal value. res = 0, denormal or normal #UE = 1 and #PE = 1, but only if the result is inexact res = (result calculated with unbounded exponent and rounded to the destination precision) * 21536 #UE = 1 #PE = 1 if the result is inexact Vol. 1 E-19 GUIDELINES FOR WR...
View Full Document

Ask a homework question - tutors are online