- In the special opcode case (bmaski:
001011000000rrrr, bgeni: 001100100111rrrr), the immediate field is missing in disassembly.
e.g.
test.o: file format elf32-mcore-little
Disassembly of section .text:
00000000 <.text>:
0: 3272 bgeni r2, 7
2: 2c02 bmaski r2, 0
(An IMM5 value of 32 is encoded in the instruction as an IMM5 field of 00000 by the assembler. Here objdump decodes bmaski r2, 32 to bmaski r2, 0, which seems to be a bug of binutils as well...)
- In the other opcode case, IMM5 field seems to be missing some bits.
e.g.:
test.o: file format elf32-mcore-little
Disassembly of section .text:
00000000 <.text>:
0: 3282 bgeni r2, 8
2: 3292 bgeni r2, 9
4: 32f2 bgeni r2, 15
6: 3302 bgeni r2, 16
8: 33f2 bgeni r2, 31
a: 2c82 bmaski r2, 8
c: 2c92 bmaski r2, 9
e: 2cf2 bmaski r2, 15
10: 2d02 bmaski r2, 16
12: 2df2 bmaski r2, 31
001011000000rrrr, bgeni:001100100111rrrr), the immediate field is missing in disassembly.e.g.
(
An IMM5 value of 32 is encoded in the instruction as an IMM5 field of 00000 by the assembler.Here objdump decodesbmaski r2, 32tobmaski r2, 0, which seems to be a bug of binutils as well...)e.g.: