[prev in list] [next in list] [prev in thread] [next in thread] 

List:       binutils
Subject:    Re: [PATCH] RISC-V: don't disassemble unrecognized insns as .byte
From:       Nelson Chu <nelson () rivosinc ! com>
Date:       2023-01-31 1:51:04
Message-ID: CAPpQWtB=MvYLUJpx34XUfPzFqD6dPJ-pFs_TuiLDFGEj0Fgmog () mail ! gmail ! com
[Download RAW message or body]

On Mon, Jan 30, 2023 at 11:33 AM Maciej W. Rozycki <macro@orcam.me.uk> wrote:
>
> Hi Nelson,
>
> > 0000000000000000 <.text>:
> >    0: 0001                .short 0x0001
> >    2: 0001                .2byte 0x1
> >    4: 0001                .short 0x0001
> > ...
> >
> > The .<N>byte in assembly will also be treated as data to dump, so
> > ideally, dump these unrecognized instructions as .insn is probably
> > more reasonable and won't be confused.
>
>  I wasn't aware about the RISC-V `.insn' pseudo-op.  I think it will make
> a lot of sense to use it here then, so I'd be happier with either:
>
>    2: 0001                .insn 0x0001
>
> or:
>
>    2: 0001                .insn 0x2, 0x1
>
> and analogously for unrecognised instructions comprised of a higher number
> of parcels according to the opcode allocation, e.g.:
>
> Disassembly of section .text:
> 0000000000000000 <foo> .insn    0x00000000001f
>
> or:
>
> Disassembly of section .text:
> 0000000000000000 <foo> .insn    0x6, 0x1f
>
> (here in the `--prefix-addresses' format).  The single-argument format may
> by clearer to the casual eye at the cost of extra leading zeros with some
> encodings.

Yeah, dump these unrecognized instructions into the .insn directives
with the <bytes> field looks better.  Thanks for pointing this out :-)

Nelson
[prev in list] [next in list] [prev in thread] [next in thread]