Can FLI.H load any floating-point constant?
No. It can select only the 32 table entries defined by Zfa.
FLI.H
RISC-V FLI.H Instruction Details
Instruction ManualR-typeLoad one half-precision floating-point constant from the 32-entry Zfa constant table.
Instruction Syntax
fli.h rd, constant
Operand Breakdown
rd: destination floating-point register.
constant: assembler-level Zfa FLI table constant; the machine encoding uses rs1 to select entries 0-31.
ZfaZfhZvfhFloating-Point Convert
Instruction Behavior
FLI.H uses the OP-FP R-type field layout: rs1 encodes an index into the 32-entry Zfa floating-point constant table, funct5=11110, fmt=H, rs2=1, and funct3=000; it writes the selected half-precision constant to f[rd]. It requires Zfa plus Zfh or Zvfh and is not an arbitrary immediate encoding. Entry 1 is the half-precision minimum positive normal number, entries 29 and 30 are both +inf, and entry 31 is canonical NaN. Loading the constant sets no floating-point exception flags.
FLI.H Decode And Execute Animation
Shows the Zfa FLI OP-FP fixed fields, rs1 constant-table index, and FP-register writeback; it has neither an arbitrary FP immediate nor rounding.
rd
constant
fli.h
,
31..27
26..25
24..20
19..15
14..12
11..7
6..0
11110
FLI
10
fmt
00001
fixed 1
10000
table index
000
fixed
01010
FP rd
1010011
OP-FP
Execution Data Path
instruction
0xF4180553
opcode
1010011 -> OP-FP
funct5 / fmt
11110 / 10 -> FLI.H
fixed fields
rs2=00001, funct3=000
constant index
rs1=16 -> 1.0
destination
rd=fa0(f10)
writeback
fa0(f10) = 1.0; fflags unchanged
Current Step
Fetch: show the 32-bit OP-FP FLI encoding
FLI uses the OP-FP field layout; it does not place an arbitrary FP literal in a normal I-type immediate field.
Awaiting FP-register writeback
This animation shows only Zfa FLI ISA-visible field selection, official constant-table selection, and register result; it does not model assembler implementation, FPU pipelines, exception handling, or microarchitectural timing.
Quick Understanding & Search Notes
FLI.H selects one half-precision constant from the official 32-entry table through the instruction rs1 field and writes it to rd; it cannot express an arbitrary binary floating literal.
Entry 1 varies by target format and is the minimum positive normal number for H/S/D/Q.
Entries 30 and 31 are +inf and canonical NaN; loading them does not raise exception flags.
Common Usage Scenarios
Floating-Point Constants
Understand this scenario with real code like «fli.h f0, 1.0 # f0 = 1.0».
Pre-Use Checklist
Syntax Check
- rd: destination floating-point register.
- constant: assembler-level Zfa FLI table constant; the machine encoding uses rs1 to select entries 0-31.
Semantic Check
- Confirm the constant is one of the official 32 Zfa FLI table entries.
- Confirm this instruction does not load arbitrary floating immediates and does not set floating-point exception flags.
Pitfalls / Common Confusions
The constant must be representable by the official FLI table; it is not an arbitrary floating-point immediate.
The rs1 field actually encodes a table index from 0 to 31.
Loading +inf or canonical NaN still sets no floating-point exception flags.
FAQ
How is the constant encoded for FLI.H?
The assembler maps an accepted constant name or value to a 0-31 table index in the rs1 field.