Re: [RFA] Remove ignoring leading exec events code.

[Michael Snyder <msnyder at specifix dot com>]

On Fri, 2008-03-14 at 14:06 -0400, Daniel Jacobowitz wrote:
> On Fri, Mar 14, 2008 at 12:12:31PM +0300, Vladimir Prus wrote:
> > 
> > GDB has some code to handle a case where target reports more
> > that one exec event when execing. However, it turns out that
> > all target say that exactly one event is reported, and 
> > therefore that code is of no use. The comments for that code
> > is dated 2002-12-05 and mentions HP-UX, which I presume had this
> > bug fixed. Even if some new target appears with the same bug,
> > as the comment say, it's better to handle this in target_wait,
> > not in the generic infrun code.
> > 
> > OK?
> 
> OK.

Hmm, just a cautionary note.  The code / practice that you're
talking about stretches WAAAAY back in history.  I'm pretty
sure it was not only for HP-UX.  There may be other (possibly
historical) targets that depend on this.

I can remember that code being there (and being necessary)
in the mid-1990's.  I don't necessarily remember on what
targets it came into play, but I might guess that SunOS 
(perhaps solaris or perhaps before) was one of them.