[dpdk-dev] [PATCH] net/tap: avoid using SIGIO

[Morten Brørup]

> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Stephen Hemminger
> Sent: Monday, July 27, 2020 9:44 PM
> 
> On Mon, 27 Jul 2020 15:19:32 +0200
> Morten Brørup <mb at smartsharesystems.com> wrote:
> 
> > > From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Stephen
> Hemminger
> > > Sent: Wednesday, July 15, 2020 1:58 AM
> > >
> > > SIGIO maybe used by application, instead choose another rt-signal.
> > > Linux allows any signal to be used for signal based IO.
> > > Search for an unused signal in the available rt-signal range.
> >
> > Just an observation. Feel free to ignore at your convenience:
> >
> > The problem is the same as for SIGIO if the application sets up its
> own signal handler after this, and uses some hardcoded rt-signal that
> happens to be the one found to be free.
> >
> > Unless the application doesn't use a hardcoded rt-signal, but also
> searches for an unused one.
> >
> > So perhaps the "search for unused rt-signal" should be exposed as a
> generic support function for the application (and this driver) to use.
> 
> There is no safe way to use a signal deep inside DPDK in a driver.
> 
> This is not the kind of thing that should be exposed to the
> application.
> 
> The algorithm for finding an RT signal conforms to the recommended
> policy on the signal(7)
> manual page.
> 
>        programs should never refer to real-time signals using hard-
>        coded numbers, but instead should always refer to real-time
> signals
>        using the notation SIGRTMIN+n, and include suitable (run-time)
> checks
>        that SIGRTMIN+n does not exceed SIGRTMAX.
> 
> The application should be following the proscribed policy on the man
> page.
> If it doesn't it is broken.

Great. That fully addresses my concern.

Acked-by: Morten Brørup <mb at smartsharesystems.com>