Re: [PATCH v35 01/13] target/avr: Add outward facing interfaces and core CPU logic

[Philippe Mathieu-Daudé]

On 11/22/19 11:54 AM, Michael Rolnik wrote:
> Hi all.
>
> Once implemented, how to invoke this function?

I'm testing your series with:

qemu-system-avr -kernel ~/Downloads/ATmegaBOOT.elf -d in_asm,unimp

The '-d in_asm' enable dumping the assembler instructions translated.

> Sent from my cell phone, please ignore typos
>
> On Fri, Nov 22, 2019, 12:46 PM Philippe Mathieu-Daudé <address@hidden 
> <mailto:address@hidden>> wrote:
>
>     Hi Michael,
>
>     On 11/22/19 9:43 AM, Aleksandar Markovic wrote:
>      > On Friday, November 22, 2019, Richard Henderson
>      > <address@hidden
>     <mailto:address@hidden>
>     <mailto:address@hidden
>     <mailto:address@hidden>>> wrote:
>      >
>      >     On 11/21/19 8:53 PM, Michael Rolnik wrote:
>      >      > It seems to be a huge investment. this function should
>     parse the
>      >      > binary data as `decode_insn` does, so I suggest to modify
>     decodetree
>      >      > tool to make decoding information available to the
>     instruction print
>      >      > function.
>      >      > what do you think?
>      >
>      >     See target/openrisc/disas.c, which makes use of decodetree.
>      >     It shouldn't be difficult to do something slimiar for avr.
>      >
>      >
>      > I support Richard's suggested direction. Since you already
>     implemenred
>      > core AVR decoder using decodetree, and you have this openrisc QEMU
>      > disassembler example, could you perhaps give a try to what
>     Richard said,
>      > Michael?
>
>     As Aleksandar commented elsewhere in this thread, the port is useful as
>     it, and having the disas output is a feature request.
>
>     If you have time to look at it before 5.0 development window opens,
>     I'll
>     be pleased to review and test it. Anyway I won't object we merge your
>     work because this is missing.
>
>     Regards,
>
>     Phil.