"l" 32 bit floating point constant in .lit4
MDMX instruction operands (note that while these use the FP register
- fields, they accept both $fN and $vN names for the registers):
+ fields, they accept both $fN and $vN names for the registers):
"O" MDMX alignment offset (OP_*_ALN)
"Q" MDMX vector/scalar/immediate source (OP_*_VSEL and OP_*_FT)
- "X" MDMX destination register (OP_*_FD)
+ "X" MDMX destination register (OP_*_FD)
"Y" MDMX source register (OP_*_FS)
"Z" MDMX source register (OP_*_FT)
Because of the lookup algorithm used, entries with the same opcode
name must be contiguous.
-
+
Many instructions are short hand for other instructions (i.e., The
jal <register> instruction is short for jalr <register>). */
lsb = ((l >> OP_SH_SHAMT) & OP_MASK_SHAMT) + 32;
(*info->fprintf_func) (info->stream, "0x%x", lsb);
break;
-
+
case 'F':
msb = ((l >> OP_SH_INSMSB) & OP_MASK_INSMSB) + 32;
(*info->fprintf_func) (info->stream, "0x%x", msb - lsb + 1);