Missing or incorrect alias and id info of `jlar`

[Rot127]

Work environment

Questions	Answers
System Capstone runs on OS/arch/bits	all
Capstone module affected	riscv
Source of Capstone	git clone
Version/git commit	next

Instruction bytes giving faulty results

cstool -dr riscv32 "0x67, 0x80, 0x50, 0x00"
 0  67 80 50 00  jr    5(ra)
    ID: 31 (tail)
    Is alias: 0 ((null)) with REAL operand set
    op_count: 3
        operands[0].type: REG = zero
        operands[0].access: WRITE
        operands[1].type: REG = ra
        operands[1].access: READ
        operands[2].type: IMM = 0x5
        operands[2].access: READ

    Groups: jump

Expected results

It should be:

is_alias: 1
id: jlar
alias_id: jr

Steps to get the wrong result

With cstool: above

Additional Logs, screenshots, source code, configuration dump, ...

Drag and drop zip archives containing the Additional info here, don't use external services or link.
Screenshots can be directly dropped here.

[OBarronCS]

Another alias bug I came across is a discrepency between the alias version in cstool and in the Python binding:

In the cmdline, is_alias is zero here:

🧊 [ubuntu:~]$ cstool -d riscv64 "e780604c"
 0  e7 80 60 4c  jalr	0x4c6(ra)
	ID: 31 (tail)
	Is alias: 0 ((null)) with REAL operand set
	op_count: 2
		operands[0].type: REG = ra
		operands[0].access: READ
		operands[1].type: IMM = 0x4c6
		operands[1].access: READ

	Groups: jump 

But in Python, is_alias is True

# /// script
# requires-python = ">=3.12"
# dependencies = ["capstone==6.0.0a7"]
# ///
from capstone import Cs
from capstone import *

cs = Cs(CS_ARCH_RISCV, CS_MODE_RISCV64 | CS_MODE_RISCV_C | CS_MODE_LITTLE_ENDIAN)

x = list(cs.disasm(b"\xe7\x80\x60\x4c", 0x10000000, 1))[0]

print(x.is_alias)
# prints True. alias_id is 0

[Rot127]

cc @moste00

[moste00]

@OBarronCS @Rot127 Yes indeed, I was just noticing this when I was fixing the compressed flag.

My hypothesis: The is_alias flag is correct, because .yaml tests assert on it successfully. The printing in cstool.c has some problems that make it print a NULL pointer, possibly related to some bad table lookups or something.

Anyway, this is on my mind and I'm going to mention this issue in the PR when I get to it.