[PATCH v4 4/7] Add SYS_FUNC(tkill)

[Dmitry V. Levin]

On Sat, Aug 01, 2020 at 03:50:20PM +0200, Ákos Uzonyi wrote:
> On Sat, 1 Aug 2020 at 14:52, Dmitry V. Levin <ldv at altlinux.org> wrote:
> > On Mon, Jul 27, 2020 at 08:54:07PM +0200, Ákos Uzonyi wrote:
> > > syscallent*.h files are updated automatically by:
> > > find linux -name syscallent*.h -exec \
> > >         sed -i '/"tkill"/ s/SEN(kill)/SEN(tkill)/' {} +
> > >
> > > * signal.c (tkill): New syscall decoder.
> > > * linux/32/syscallent.h: Use SEN(tkill) for "tkill" syscall.
> >
> > I think it would be better if this commit was split into two parts:
> > the first one that introduces a tkill syscall decoder, and the second part
> > that makes the tkill syscall decoder use printpid.  The second part then
> > should be merged into "Use printpid in decoders" commit that updates
> > decoders of kill and tgkill syscalls.
> >
> > Does this make sense?
> 
> Yes, that makes sense, I will do it.
> 
> Also I'm not sure if "Use printpid in decoders" and "Use printpid in
> print_sockaddr and printsiginfo" should be two different commits. I
> didn't squash them together, because the latter one adds a tcp
> parameter to a lot of functions so I thought it could be cleaner to
> keep them separate. But since logically these two commits should be
> one, I think I should squash them. Do you agree?

I think those changes that just add "struct tcb *" argument here and
there can now be moved to a separate commit (I'm currently testing this
in ldv/next branch), and both "Use printpid" commits would indeed
be better merged together.


-- 
ldv