[PATCH for 4.6] configure: don't silently disable systemd support

[PATCH for 4.6] configure: don't silently disable systemd support

[Wei Liu]

Originally when user runs ./configure --enable-systemd and systemd
development library is not available the build system silently disables
systemd support. This is not in line with normal expectation.

Instead, configure should error out when user has asked for systemd
support but development libraries can't be found.

Reported-by: George Dunlap <george.dunlap@eu.citrix.com>
Signed-off-by: Wei Liu <wei.liu2@citrix.com>
---
Please rerun ./autogen.sh.
---
 m4/systemd.m4 | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/m4/systemd.m4 b/m4/systemd.m4
index 8284993..84508a1 100644
--- a/m4/systemd.m4
+++ b/m4/systemd.m4
@@ -85,7 +85,11 @@ AC_DEFUN([AX_CHECK_SYSTEMD], [
 		AC_DEFINE([HAVE_SYSTEMD], [1], [Systemd available and enabled])
 			systemd=y
 			AX_CHECK_SYSTEMD_LIBS()
-	    ],[systemd=n])
+	    ],[
+		AS_IF([test "x$enable_systemd" != "x"],
+			[AC_MSG_ERROR([Unable to find systemd development library])],
+			[systemd=n])
+	    ])
 	],[systemd=n])
 ])
 
-- 
2.1.4

Re: [PATCH for 4.6] configure: don't silently disable systemd support

[Ian Campbell]

On Wed, 2015-09-09 at 23:35 +0100, Wei Liu wrote:
> Originally when user runs ./configure --enable-systemd and systemd
> development library is not available the build system silently disables
> systemd support. This is not in line with normal expectation.
> 
> Instead, configure should error out when user has asked for systemd
> support but development libraries can't be found.
> 
> Reported-by: George Dunlap <george.dunlap@eu.citrix.com>
> Signed-off-by: Wei Liu <wei.liu2@citrix.com>
> ---
> Please rerun ./autogen.sh.
> ---
>  m4/systemd.m4 | 6 +++++-
>  1 file changed, 5 insertions(+), 1 deletion(-)
> 
> diff --git a/m4/systemd.m4 b/m4/systemd.m4
> index 8284993..84508a1 100644
> --- a/m4/systemd.m4
> +++ b/m4/systemd.m4
> @@ -85,7 +85,11 @@ AC_DEFUN([AX_CHECK_SYSTEMD], [
>  		AC_DEFINE([HAVE_SYSTEMD], [1], [Systemd available and
> enabled])
>  			systemd=y
>  			AX_CHECK_SYSTEMD_LIBS()
> -	    ],[systemd=n])
> +	    ],[
> +> 	> 	> AS_IF([test "x$enable_systemd" != "x"],

At this point the value of enable_systemd is either empty or ...what...? I
think "yes"?, with "no" having been ruled out earlier?

Or can it be just "y"?

I could live with this version, but I think it would be less of a brain ben
der if the check was inverted to be "x$enable_systemd" = "xyes" i.e. more
explicit.

Ian.

Re: [PATCH for 4.6] configure: don't silently disable systemd support

[Wei Liu]

On Thu, Sep 10, 2015 at 11:54:16AM +0100, Ian Campbell wrote:
> On Wed, 2015-09-09 at 23:35 +0100, Wei Liu wrote:
> > Originally when user runs ./configure --enable-systemd and systemd
> > development library is not available the build system silently disables
> > systemd support. This is not in line with normal expectation.
> > 
> > Instead, configure should error out when user has asked for systemd
> > support but development libraries can't be found.
> > 
> > Reported-by: George Dunlap <george.dunlap@eu.citrix.com>
> > Signed-off-by: Wei Liu <wei.liu2@citrix.com>
> > ---
> > Please rerun ./autogen.sh.
> > ---
> >  m4/systemd.m4 | 6 +++++-
> >  1 file changed, 5 insertions(+), 1 deletion(-)
> > 
> > diff --git a/m4/systemd.m4 b/m4/systemd.m4
> > index 8284993..84508a1 100644
> > --- a/m4/systemd.m4
> > +++ b/m4/systemd.m4
> > @@ -85,7 +85,11 @@ AC_DEFUN([AX_CHECK_SYSTEMD], [
> >  		AC_DEFINE([HAVE_SYSTEMD], [1], [Systemd available and
> > enabled])
> >  			systemd=y
> >  			AX_CHECK_SYSTEMD_LIBS()
> > -	    ],[systemd=n])
> > +	    ],[
> > +> 	> 	> AS_IF([test "x$enable_systemd" != "x"],
> 
> At this point the value of enable_systemd is either empty or ...what...? I
> think "yes"?, with "no" having been ruled out earlier?
> 
> Or can it be just "y"?
> 

At this point, it can't be no. It's either empty or "yes" (or "y"?). I
printed it out when I wrote this patch but could remember exactly what
it was.

> I could live with this version, but I think it would be less of a brain ben
> der if the check was inverted to be "x$enable_systemd" = "xyes" i.e. more
> explicit.
> 

I can do that as well. V2 coming soon.

Wei.

> Ian.

Re: [PATCH for 4.6] configure: don't silently disable systemd support

[Wei Liu]

On Thu, Sep 10, 2015 at 12:04:10PM +0100, Wei Liu wrote:
> On Thu, Sep 10, 2015 at 11:54:16AM +0100, Ian Campbell wrote:
> > On Wed, 2015-09-09 at 23:35 +0100, Wei Liu wrote:
> > > Originally when user runs ./configure --enable-systemd and systemd
> > > development library is not available the build system silently disables
> > > systemd support. This is not in line with normal expectation.
> > > 
> > > Instead, configure should error out when user has asked for systemd
> > > support but development libraries can't be found.
> > > 
> > > Reported-by: George Dunlap <george.dunlap@eu.citrix.com>
> > > Signed-off-by: Wei Liu <wei.liu2@citrix.com>
> > > ---
> > > Please rerun ./autogen.sh.
> > > ---
> > >  m4/systemd.m4 | 6 +++++-
> > >  1 file changed, 5 insertions(+), 1 deletion(-)
> > > 
> > > diff --git a/m4/systemd.m4 b/m4/systemd.m4
> > > index 8284993..84508a1 100644
> > > --- a/m4/systemd.m4
> > > +++ b/m4/systemd.m4
> > > @@ -85,7 +85,11 @@ AC_DEFUN([AX_CHECK_SYSTEMD], [
> > >  		AC_DEFINE([HAVE_SYSTEMD], [1], [Systemd available and
> > > enabled])
> > >  			systemd=y
> > >  			AX_CHECK_SYSTEMD_LIBS()
> > > -	    ],[systemd=n])
> > > +	    ],[
> > > +> 	> 	> AS_IF([test "x$enable_systemd" != "x"],
> > 
> > At this point the value of enable_systemd is either empty or ...what...? I
> > think "yes"?, with "no" having been ruled out earlier?
> > 
> > Or can it be just "y"?
> > 
> 
> At this point, it can't be no. It's either empty or "yes" (or "y"?). I

For the record, it was "yes".