On 05/18/2018 03:34 AM, Peter Maydell wrote: > On 15 May 2018 at 15:46, Richard Henderson <r...@twiddle.net> wrote: >> On 05/15/2018 06:25 AM, Peter Maydell wrote: >>> That said, I think we're planning to deprecate vixl now we have the >>> capstone support. Richard, what's the status here? Could we just remove >>> the vixl code now? >> >> We could just remove vixl, yes. I'd like to see updates to capstone to >> support >> instructions post v8.0, but it's not like we have those with vixl either... > > Just to check my understanding: with QEMU at the moment, you > always get the capstone disassembler unless you specifically > turn it off by passing --disable-capstone to configure, right > (since we provide it as a submodule)?
Correct. > We put that in in September last year, and we haven't had a > pile of complaints about the disassembly (or indeed any > complaints that I can recall), so I think we can consider it > a success, and remove both vixl and the ancient binutils arm > disassembler. Yep. > It would also be interesting to try interacting with capstone > upstream about adding support for newer instructions (for > instance they don't do the v8M insns). Do you know if capstone > deals with new insns via resync from LLVM or if they've > entirely forked and just make changes locally by hand? They seem to have entirely forked, but I'm not completely sure. r~
