Re: [Qemu-devel] [PULL 08/16] tcg/i386: Support vector comparison select value

[Aleksandar Markovic]

On May 30, 2019 7:45 PM, "Richard Henderson" <address@hidden>
wrote:
>
> On 5/30/19 9:54 AM, Aleksandar Markovic wrote:
> >
> > On May 30, 2019 2:50 PM, "Richard Henderson" <
address@hidden
> > <mailto:address@hidden>> wrote:
> >>
> >> On 5/30/19 6:26 AM, Peter Maydell wrote:
> >> >> -#define TCG_TARGET_HAS_cmpsel_vec       0
> >> >> +#define TCG_TARGET_HAS_cmpsel_vec       -1
> >> >
> >> > This is the only place where we define a TCG_TARGET_HAS_* macro
> >> > to something other than 0 or 1, which means that Coverity
> >> > complains (CID 1401702) when we use it in a logical boolean
expression
> >> >   return have_vec && TCG_TARGET_HAS_cmpsel_vec;
> >> > later on.
> >> >
> >> > Should it really be -1, or is this a typo for 1 ?
> >>
> >> It really should be -1.
> >> See commit 25c012b4009256505be3430480954a0233de343e,
> >> which contains the rationale.
> >>
> >
> > How about extending commit message so that it contains explanation for
-1
> > introduced in this very patch allowing future developers not to need to
reverse
> > engineer whole git history to (maybe) find the explanation?
>
> No.
>
> There seems to be no point at which you would stop, and not include the
entire
> git history of the project into each and every commit message.
>
> I will not be drawn into such a discussion further.

Your commit messages are often disconnected with the content of the code
change, sometimes even look like cryptic puzzles. You can do much better
job there, and not look for what is good and clear for you, but what is
good and clear for others.

Regards,
Aleksandar

>
> r~