On Tue, Feb 1, 2022 at 2:06 PM Richard Henderson <
richard.hender...@linaro.org> wrote:
> On 2/2/22 07:52, Warner Losh wrote:
> > The capsicum signal stuff is new with FreeBSD 14, rev 1400026, so only
> > copy them on a new enough system.
> >
> > Signed-off-by: Warner Losh <i...@bsdimp.com>
> > ---
> > bsd-user/signal.c | 4 ++++
> > 1 file changed, 4 insertions(+)
> >
> > diff --git a/bsd-user/signal.c b/bsd-user/signal.c
> > index ad22ba9d90d..b43266e6e06 100644
> > --- a/bsd-user/signal.c
> > +++ b/bsd-user/signal.c
> > @@ -242,8 +242,10 @@ static inline void
> host_to_target_siginfo_noswap(target_siginfo_t *tinfo,
> > * one, then we know what to save.
> > */
> > if (sig == TARGET_SIGTRAP) {
> > +#if defined(__FreeBSD_version) && __FreeBSD_version >= 1400026
> > tinfo->_reason._capsicum._syscall =
> > info->_reason._capsicum._syscall;
> > +#endif
> > si_type = QEMU_SI_CAPSICUM;
> > }
>
> Do you really want the entire IF inside the ifdef?
> Setting QEMU_SI_CAPSICUM doesn't make sense for earlier versions...
>
True. In fact, not defining it at all on earlier versions would be safer to
ensure
all the places where we might want to touch this we'll get a compiler
warning
as well as to communicate this is only for earlier versions w/o sprinkling
the
version through the code. I'll update with this suggestion.
Warner
>
> r~
>
> > break;
> > @@ -296,8 +298,10 @@ static void tswap_siginfo(target_siginfo_t *tinfo,
> const target_siginfo_t *info)
> > __put_user(info->_reason._poll._band,
> &tinfo->_reason._poll._band);
> > break;
> > case QEMU_SI_CAPSICUM:
> > +#if defined(__FreeBSD_version) && __FreeBSD_version >= 1400026
> > __put_user(info->_reason._capsicum._syscall,
> > &tinfo->_reason._capsicum._syscall);
> > +#endif
> > break;
> > default:
> > g_assert_not_reached();
>
>
