Another problem with MC + subshell on cygwin

[Christopher Faylor]

On Sat, Aug 17, 2002 at 09:51:04PM -0400, Christopher Faylor wrote:
>Probably the right way to fix this is for the parent process not to
>deliver the SIGCHLD to itself until the child has actually been suspended.
>I don't know of any way of figuring out if the child is really suspended
>though.

I tried to fix this by not using SuspendThread at all.  Instead, there
is YA event (called sigCONT) which the process waits on.  Essentially we
were trying to synchronize using SuspendThread/ResumeThread and
Microsoft tells you that you shouldn't do that...  It shure would be
convenient if it was possible, though.  It would mean one less handle.

Oh well.

Anyway, I've checked in a change and am generating a new snapshot.

cgf