comiss

Compare Scalar Ordered Single-Precision

COMISS xmm1, xmm2/m32

Compares low float and sets EFLAGS (Signaling NaN raises exception).

Details

Performs an ordered comparison of the low single-precision floating-point values in two XMM registers or memory, setting EFLAGS (ZF, PF, CF) accordingly. Signals invalid operation exception if either operand is a signaling NaN; quiet NaNs set ZF=1, PF=1, CF=1. Does not modify the XMM registers.

Pseudocode Operation

src1_val ← src1[0:31] (as float); src2_val ← src2[0:31] (as float); if (IsNaN(src1_val) || IsNaN(src2_val)) { ZF ← 1; PF ← 1; CF ← 1; OF ← 0; AF ← 0; SF ← 0; } else if (src1_val < src2_val) { CF ← 1; ZF ← 0; PF ← 0; OF ← 0; AF ← 0; SF ← 0; } else if (src1_val == src2_val) { ZF ← 1; CF ← 0; PF ← 0; OF ← 0; AF ← 0; SF ← 0; } else { ZF ← 0; CF ← 0; PF ← 0; OF ← 0; AF ← 0; SF ← 0; }

Example

COMISS xmm1, xmm2/m32

Encoding

Binary Layout

Format SSE

Opcode NP 0F 2F /r

Extension SSE

Operands

src1
XMM
src2
XMM/Mem

Reference (Intel® SDM)

Instruction Forms

Opcode	Instruction	Op/En	64/32-bit Mode	CPUID	Description
NP 0F 2F /r	COMISS xmm1, xmm2/m32	A	V/V	SSE	Compare low single precision floating-point values in xmm1 and xmm2/mem32 and set the EFLAGS flags accordingly.
VEX.LIG.0F.WIG 2F /r	VCOMISS xmm1, xmm2/m32	A	V/V	AVX	Compare low single precision floating-point values in xmm1 and xmm2/mem32 and set the EFLAGS flags accordingly.
EVEX.LLIG.0F.W0 2F /r	VCOMISS xmm1, xmm2/m32{sae}	B	V/V	AVX512F OR AVX10.1	Compare low single precision floating-point values in xmm1 and xmm2/mem32 and set the EFLAGS flags accordingly.

Instruction Operand Encoding

Op/En	Tuple Type	Operand 1	Operand 2	Operand 3	Operand 4
A	N/A	ModRM:reg (w)	ModRM:r/m (r)	N/A	N/A
B	Tuple1 Scalar	ModRM:reg (w)	ModRM:r/m (r)	N/A	N/A

Description

Compares the single precision floating-point values in the low quadwords of operand 1 (first operand) and operand 2 (second operand), and sets the ZF, PF, and CF flags in the EFLAGS register according to the result (unordered, greater than, less than, or equal). The OF, SF, and AF flags in the EFLAGS register are set to 0. The unordered result is returned if either source operand is a NaN (QNaN or SNaN). Operand 1 is an XMM register; operand 2 can be an XMM register or a 32 bit memory location. The COMISS instruction differs from the UCOMISS instruction in that it signals a SIMD floating-point invalid operation exception (#I) when a source operand is either a QNaN or SNaN. The UCOMISS instruction signals an invalid operation exception only if a source operand is an SNaN. The EFLAGS register is not updated if an unmasked SIMD floating-point exception is generated. VEX.vvvv and EVEX.vvvv are reserved and must be 1111b, otherwise instructions will #UD. Software should ensure VCOMISS is encoded with VEX.L=0. Encoding VCOMISS with VEX.L=1 may encounter unpredictable behavior across different processor generations.

Operation

COMISS (All Versions)
RESULT :=OrderedCompare(DEST[31:0] <> SRC[31:0]) {
(* Set EFLAGS *) CASE (RESULT) OF
UNORDERED: ZF,PF,CF := 111;
GREATER_THAN: ZF,PF,CF := 000;
LESS_THAN: ZF,PF,CF := 001;
EQUAL: ZF,PF,CF := 100;
ESAC;
OF, AF, SF := 0; }





COMISS—Compare Scalar Ordered Single Precision Floating-Point Values and Set EFLAGS                                                     Vol. 2A 3-201

Intel C/C++ Compiler Intrinsic Equivalent

VCOMISS int _mm_comi_round_ss(__m128 a, __m128 b, int imm, int sae);
VCOMISS int _mm_comieq_ss (__m128 a, __m128 b)
VCOMISS int _mm_comilt_ss (__m128 a, __m128 b)
VCOMISS int _mm_comile_ss (__m128 a, __m128 b)
VCOMISS int _mm_comigt_ss (__m128 a, __m128 b)
VCOMISS int _mm_comige_ss (__m128 a, __m128 b)
VCOMISS int _mm_comineq_ss (__m128 a, __m128 b)

Exceptions

SIMD Floating-Point Exceptions

Invalid (if SNaN or QNaN operands), Denormal.

Other Exceptions

VEX-encoded instructions, see Table 2-20, “Type 3 Class Exception Conditions.” EVEX-encoded instructions, see Table 2-50, “Type E3NF Class Exception Conditions.” Additionally: #UD If VEX.vvvv != 1111B or EVEX.vvvv != 1111B. COMISS—Compare Scalar Ordered Single Precision Floating-Point Values and Set EFLAGS Vol. 2A 3-202