branches to labels work, branches to absolute addresses do not

[nealcrook]

Consider the input code:

thecode = """
                ORG  $1234
start:
                BPL loop
                LDA #$10
                LDX #$00
loop:
                STA $1000,x
                INX
                SBC #$01
                BPL loop
                BPL $1234 ; backwards
                BPL $1250 ; forwards
                RTS
"""

Generated assembly/code is:

1    0000 :                                
2    1234 :                  org $1234     
3    1234 : start:                         
4    1234 :         10 04    bpl loop      
5    1236 :         A9 10    lda #$10      
6    1238 :         A2 00    ldx #$00      
7    123A : loop:                          
8    123A :         9D 00 10 sta $1000,x   
9    123D :         E8       inx           
10   123E :         E9 01    sbc #$01      
11   1240 :         10 F8    bpl loop      
12   1242 :         10 34    bpl $1234     ; backwards BUG!!
13   1244 :         10 50    bpl $1250     ; forwards BUG!!
14   1246 :         60       rts           

Observe that the two branches to absolute addresses generate incorrect code: instead of an offset, the value is the l_ow byte of the target address_!!

thanks for sharing your code. Even if I end up fixing this bug myself it will still be quicker than starting from scratch..

Neal.

[nealcrook]

Oh, I guess you think the argument is an offset whereas I'm expecting it to be an absolute address. But, but..

• if the argument is a label, it will not be an offset it will be an absolute address..
• I can see situations where you might want to specify an offset and situations where you might want to specify an absolute address..

In my particular example the code came from disassembly elsewhere

[dj-on-github]

Thank you. Feel free to send a pull request if you get it working. I made the mistake of starting two companies so I've not got a lot of time to work on it.

…

On Mon, Feb 9, 2026, 11:45 PM Neal Crook ***@***.***> wrote: *nealcrook* left a comment (dj-on-github/py6502#11) <#11 (comment)> Oh, I guess you think the argument is an offset whereas I'm expecting it to be an absolute address. But, but.. - if the argument is a *label*, it will not be an offset it will be an absolute address.. - I can see situations where you might want to specify an offset and situations where you might want to specify an absolute address.. In my particular example the code came from disassembly elsewhere — Reply to this email directly, view it on GitHub <#11 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABYMCGIGOEDDE5YJEK4FURD4LEL3NAVCNFSM6AAAAACURKLECSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTQNZUGQ3TQMBQGY> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>