riscv-isa-manual
riscv-isa-manual copied to clipboard
riscv
Do we have a standard format or document for assembly expression?
For example, instruction ADD in the manual: "ADD performs the addition of rs1 and rs2." is it ADD rd, rs1, rs2 or ADD rd, rs2, rs1 or ADD rs1, rs2, rd
Another case, for floating point instruction with rounding mode, FADD.S rd, rs1, rs2 ## DYN FADD.S rd, rs1, rs2, rne ## nearest-even this expression is from binutils, not in the manual.
Lots of assembly code in manual seems like there IS a format, if you know MIPS first. In the vector extension, all instructions are introduced with assembly example.
Also, I have the book "The RISC-V Reader". The appendix A is very helpful, provide very specific information. ADD rd, rs1, rs2 x[rd] = x[rs1] + x[rs2]
AMOADD.W rd, rs2, (rs1) x[rd] = AMO32(M[x[rs1]] + x[rs2])
BLE rs1, rs2. offset if (rs1 <=s rs2) pc += sext(offset) ## correct one should be x[rs1] <=s x[rs2]
It was decided a while ago that assembly syntax was out of scope for the ISA manual (even though the ISA manual uses assembly examples in several places), see https://github.com/riscv/riscv-asm-manual for the (very incomplete) documentation for the mips-derived syntax used by binutils and llvm. See also #540.
My understanding is that this is the goal of https://github.com/riscv/riscv-asm-manual , although it appears there's a lot of work to be done on that project. Your contributions would be appreciated!
