Which register classes do rd and rs1 use in fcvt.d.wu?
This is an integer-to-floating-point conversion: rs1 is an integer register and rd is a floating-point register.
FCVT.D.WU
RISC-V FCVT.D.WU Instruction Details
Instruction ManualR-typeConvert unsigned int32 to double: f[rd] = (double)(uint32_t)rs1. Always exact.
Instruction Syntax
fcvt.d.wu rd, rs1
Operand Breakdown
Destination rd: register receiving the operation result.
Source rs1: register holding the first operand.
Source rs2: register holding the second operand.
DFloating-Point Convert
Instruction Behavior
fcvt.d.wu converts an unsigned 32-bit integer source to a double-precision floating-point result. This conversion is exact, so the encoded rounding mode does not change the numeric result.
FCVT.D.WU Decode And Execute Animation
Shows the D-extension FCVT flow: decode OP-FP fields, fmt/rs2 conversion selection, an rm=000 encoding example, read the integer source, then write the floating-point destination.
rd
rs1
fcvt.d.wu
,
Execution Context
fmt=D (01)
rs2=WU
rounding: example rm=000; conversion never rounds
fflags: 0
31..27
26..25
24..20
19..15
14..12
11..7
6..0
11010
FCVT.D
01
fmt=D
00001
rs2=WU
01011
rs1
000
rm=000 example
01010
rd
1010011
OP-FP
Execution Data Path
instruction
0xD2158553
opcode
1010011 -> OP-FP
funct5/fmt
11010 + 01 -> FCVT.D.WU
rs2
00001 -> WU source
rm
000 -> example encoding; conversion never rounds
rd/rs1
fa0(f10) / a1(x11)
read
a1(x11) = 0xFFFFFFFF (4294967295)
convert
uint32 0xFFFFFFFF (4294967295) -> D 4294967295.0
write
fa0(f10) = 4294967295.0; 0
Current Step
Fetch: show the 32-bit OP-FP encoding
The word is split as an R-type OP-FP instruction; FCVT uses opcode 1010011.
encoding: 0xD2158553
syntax : fcvt.d.wu fa0(f10), a1(x11)
result : fa0(f10) = 4294967295.0; 0
A 32-bit integer is exactly representable as double precision, so example fflags remains 0.
FCVT Conversion View
WU source
0xFFFFFFFF (4294967295)
conversion
uint32 0xFFFFFFFF (4294967295) -> D 4294967295.0
writeback
fa0(f10) = 4294967295.0
A 32-bit integer is exactly representable as double precision, so example fflags remains 0.
This animation shows only D-extension FCVT ISA-visible encoding, source/destination register direction, rounding field, and example fflags; it does not model FPU pipelines, exception handling, NaN payloads, or microarchitecture.
Quick Understanding & Search Notes
fcvt.d.wu reads an unsigned 32-bit integer from integer register rs1, converts it to double precision, and writes f[rd]; uint32 is exactly representable as double, so the numeric result does not depend on rm.
OP-FP opcode is 1010011; funct5=11010 denotes integer-to-FP FCVT and fmt=01 denotes D destination format.
rs2=00001 selects the WU unsigned 32-bit integer source; rs1 is an integer register.
rd is an FP register and the result is written in D format; uint32 to double conversion is exact.
The encoding still has an rm field, but assembly has no rm operand; legal rm values do not affect this uint32-to-double numeric result, while reserved rm encodings retain reserved behavior.
The instruction does not read an FP source register, so there is no source NaN-boxing requirement.
Common Usage Scenarios
Type Conversion
Understand this scenario with real code like «fcvt.d.wu f0, x10 # f0 = (double)(uint32_t)x10».
Pre-Use Checklist
Syntax Check
- Verify rd, rs1, rs2 (and rs3) are valid GPRs.
- Confirm funct3 and funct7 encoding is correct.
Semantic Check
- Check if the result affects subsequent branches or address calculations.
- Ensure the rd register is not overwritten by another instruction.
Pitfalls / Common Confusions
W/WU forms read a 32-bit signed or unsigned integer source value.
The integer source operand comes from x[rs1], not a floating-point register, so there is no source NaN-boxing requirement.
The floating-point result is NaN-boxed in wider FLEN registers according to the destination format.
FAQ
How is the rm field handled by fcvt.d.wu?
Assembly has no rm operand. The encoding still contains the field; legal rm values do not change the exact uint32-to-double result, while reserved rm encodings retain reserved behavior.
What does the rs2 field select in fcvt.d.wu?
rs2 is not an ordinary source register; in the FCVT encoding it selects the WU unsigned 32-bit integer format.