Re: [hatari-devel] VME access

[Eero Tamminen <oak@xxxxxxxxxxxxxx>]

Hi,

On 1/7/21 9:35 AM, Thomas Huth wrote:
> Am Thu, 7 Jan 2021 01:35:56 +0200
> basically ok for me, but:
>
> - "--vme empty" is not implemented yet, so I'd rather not mention it
>    in the manual
>
> - Please use the main Hatari coding style for new files in the main
>    src directory (i.e. curly braces should be opened on new lines, not
>    at the end of a line)

I pushed the changes with above fixed and few other improvements.


> - does it also compile without warning when running configure with
>    --disable-tracing ?

I didn't get them, but to be sure I added ifdef guard.


Btw. I don't really see what's the point of
--disable-tracing option though.  Tracing is very
useful feature and register accesses & OS calls
happen so rarely that few extra shouldn't be any
kind of performance problem, and while tracing
calls increase Hatari size a bit, it's hard to
imagine that being a problem either.

Overhead of extra if(s) in instruction tracing
(when not tracing) might be measurable, but if
that's a problem, why building of tracing support
couldn't be made conditional only for that
(i.e. have "--disable-instr-tracing" instead of
"--disable-tracing")?


	- Eero