Re: [hatari-devel] WinUAE CPU core disassembler output options

[ Thread Index | Date Index | More Archives ]


On 25.10.2022 16.25, Nicolas Pomarède wrote:
Le 20/10/2022 à 11:33, Nicolas Pomarède a écrit :
Le 19/10/2022 à 23:34, Eero Tamminen a écrit :
I've pushed it now, with doc updates.

Please check out the new WinUAE CPU disassembly options!

 >>> I hope there are eventually suitable checks in the above listed
 >>> places, but somebody more familiar with WinUAE CPU needs first to
 >>> explain why they're currently missing from it.

regarding the recent DISASM_FLAG_xxx that were added to WinUAE for more control on the output, I think I just forgot to look at what those flags did when I merged the changes, so our 68kDisass.c was not updated to use these flags. thanks for doing it.

I saw a small regression since these changes : opcodes rte/rts that can be followed by the effective address are now printing garbage after "rts ".

The code I added, just sets the WinUAE disassembler 'disasm_flags' option variable flag bits.

WinUAE disassember option flags it sets by default, differ from WinUAE defaults though. Toni probably just has not tested the flags enough.


00fc054a 4e75                     rts  == $00fc03bc

now :

00FC054A 4E75                     rts šZòºÌ^?

Bcc od DBcc are correct, they still show "== EA" after.

Can you have a look ? (maybe it's a regression in WinUae, I didn't check)

I think *both* of above behaviors are WinUAE disassembler bugs.

First bug is disassembler not respecting its DISASM_FLAG_CC / DISASM_FLAG_EA / DISASM_FLAG_VAL options being off.

Second bug is it showing garbage when it skips showing EA values for short enough instruction names. E.g. not padding things properly (like Hatari ext disassembler does).

Just enable "uae: show EA + CC values after instruction" disassembly option until Toni fixes those bugs.

	- Eero

Mail converted by MHonArc 2.6.19+