Re: [libvirt] [PATCH v3] util: systemd: Define MSG_NOSIGNAL if needed

On 20 Jul 2016, at 14:46, Andrea Bolognani <abologna(a)redhat.com&gt; wrote: > The symbol being missing has been reported as causing build > failures on OS X. If it's not already defined, define it to > zero so that it won't have any effect. > --- > Changes from v2: > > * define MSG_NOSIGNAL to zero if not already defined > instead of conditionally compiling the code depending > on it > > src/util/virsystemd.c | 4 ++++ > 1 file changed, 4 insertions(+) > > diff --git a/src/util/virsystemd.c b/src/util/virsystemd.c > index 969cd68..7d6985b 100644 > --- a/src/util/virsystemd.c > +++ b/src/util/virsystemd.c > @@ -41,6 +41,10 @@ > > VIR_LOG_INIT("util.systemd"); > > +#ifndef MSG_NOSIGNAL > +# define MSG_NOSIGNAL 0 > +#endif > + > static void virSystemdEscapeName(virBufferPtr buf, > const char *name) > { > -- > 2.7.4 This compiles fine on OSX too. :) One niggle with it though... this is what is looks like in context: ******************************************************************* ... #include "virerror.h" #include "virfile.h" #define VIR_FROM_THIS VIR_FROM_SYSTEMD VIR_LOG_INIT("util.systemd"); #ifndef MSG_NOSIGNAL # define MSG_NOSIGNAL 0 #endif static void virSystemdEscapeName(virBufferPtr buf, const char *name) { static const char hextable[16] = "0123456789abcdef"; #define ESCAPE(c) ... ******************************************************************* Isn't it kind of crying out for a useful comment, so the next person looking through this code has a good chance to understand why it's there? ;) Maybe something like: /* OSX is missing MSG_NOSIGNAL, so we define it here to avoid * failure during building. */