On Fri, 2016-02-12 at 14:17 +0100, David Henningsson wrote: > > On 2016-02-12 01:20, Ahmed S. Darwish wrote: > > Now that all layers in the stack support memfd blocks, use memfd > > pools for client context and audio data by default. > > > > Signed-off-by: Ahmed S. Darwish <darwish.07 at gmail.com> > > --- > >  man/pulse-client.conf.5.xml.in |  5 +++++ > >  src/pulse/client-conf.c        |  1 + > >  src/pulse/client-conf.h        |  2 +- > >  src/pulse/context.c            | 11 ++++++++++- > >  4 files changed, 17 insertions(+), 2 deletions(-) > > > > diff --git a/man/pulse-client.conf.5.xml.in b/man/pulse-client.conf.5.xml.in > > index cca2219..a04dedc 100644 > > --- a/man/pulse-client.conf.5.xml.in > > +++ b/man/pulse-client.conf.5.xml.in > > @@ -107,6 +107,11 @@ License along with PulseAudio; if not, see . > >       > > > >       > > +      disable-memfd=. Disable memfd shared memory. Takes > > +      a boolean argument, defaults to no. > > Good idea. The option should be "enable-memfd" rather than "disable-memfd", though. "disable-memfd=no" is a double negative, and it's good to avoid double negatives. (In the past we have had negative options that have later been superseded by positive ones, which is why we support both versions for some options, such as disable-shm/enable-shm.) -- Tanu
