M6809 `JSR [1,x]` is assembled to `AD 11` but should be `AD 98 01`

[ig3]

test.a09:

        jsr [1,x]

test.lst:

0000   AD 11                  JSR   [1,x] 

[maly]

Thank you for your message, I'll check it. Thanks  Martin Dne 15.04.2025 0:22:51, Ian Goodacre ***@***.***> napsal: test.a09: jsr [1,x] test.lst: 0000 AD 11 JSR [1,x] — Reply to this email directly, view it on GitHub [#17], or unsubscribe [https://github.com/notifications/unsubscribe-auth/AABFKPVDZPVZ5PYVCOCE7IL2ZQYLXAVCNFSM6AAAAAB3ECU5AOVHI2DSMVQWIX3LMV43ASLTON2WKOZSHE4TIMZZGY3TENA]. You are receiving this because you are subscribed to this thread. Message ID: ***@***.***> ig3 created an issue (asm80/asm80-node#17) [#17] test.a09: jsr [1,x] test.lst: 0000 AD 11 JSR [1,x] — Reply to this email directly, view it on GitHub [#17], or unsubscribe [https://github.com/notifications/unsubscribe-auth/AABFKPVDZPVZ5PYVCOCE7IL2ZQYLXAVCNFSM6AAAAAB3ECU5AOVHI2DSMVQWIX3LMV43ASLTON2WKOZSHE4TIMZZGY3TENA]. You are receiving this because you are subscribed to this thread. Message ID: ***@***.***> [adaec2e4-c5eb-4833-b8df-bf42701ac469]

[ig3]

I had a look at the code of parseOpcode for M6809. Because the offset is 1, it is executing the block that emits opcodes for 5-bit offset but this offset size is not available for indirect addressing. It should be emitting 8-bit offset.

Changing

return t > 65519 && 4 != b(a) ? (r.lens[c] = P(a) | d | 32 - (65536 - t) & 31, r) : t < 16 && t > -17 && 4 != b(a) ? (r.lens[c] = P(a) | d | 31 & t, r) : t < 128 && t > -129 && null !== t ? (t < 0 && (t = 256 + t), r.lens[c] = b(a) | d | 136, r.bytes++, "PC" == a.toUpperCase() ? r.lens[c + 1] = d ? function(e)

to

return t > 65519 && 4 != b(a) ? (r.lens[c] = P(a) | d | 32 - (65536 - t) & 31, r) : t < 16 && t > -17 && 4 != b(a) && !d ? (r.lens[c] = P(a) | d | 31 & t, r) : t < 128 && t > -129 && null !== t ? (t < 0 && (t = 256 + t), r.lens[c] = b(a) | d | 136, r.bytes++, "PC" == a.toUpperCase() ? r.lens[c + 1] = d ? function(e)

would be a start to fixing the error. This would avoid emitting 5-bit offset for the indirect address. Further changes might be necessary to emit the correct 8-bit offset. It's a bit difficult working with the minified code :-(