Re: [PATCH] Add libgsystem as a git submodule
On Tuesday 21 of May 2013 11:07:02 Colin Walters wrote: On Tue, 2013-05-21 at 09:08 +0200, Jirka Klimes wrote: CC libgsystem_la-gsystem-file-utils.lo libgsystem/gsystem-file-utils.c: In function 'linkcopy_internal_attempt': libgsystem/gsystem-file-utils.c:358:3: error: implicit declaration of function 'g_clear_pointer' [-Werror=implicit-function-declaration] cc1: all warnings being treated as errors Can you try with https://git.gnome.org/browse/libgsystem/commit/?id=759384420c1578bf399473440 ab34af18fa46b8d The patch misses terminating #endif: diff --git a/gsystem-glib-compat.h b/gsystem-glib-compat.h index edbdaab..70a9c1b 100644 --- a/gsystem-glib-compat.h +++ b/gsystem-glib-compat.h @@ -51,3 +51,4 @@ g_type_ensure (GType type) #endif +#endif /* __LIBGSYSTEM_GLIB_COMPAT__ */ Other than that it is good. Thanks, Jirka ___ networkmanager-list mailing list networkmanager-list@gnome.org https://mail.gnome.org/mailman/listinfo/networkmanager-list
Re: [PATCH] Add libgsystem as a git submodule
On Wed, 2013-05-22 at 08:55 +0200, Jirka Klimes wrote: +#endif /* __LIBGSYSTEM_GLIB_COMPAT__ */ Hum, I must have failed to actually apply the patch to NM's copy of libgsystem. Regardless, pushed, thanks! ___ networkmanager-list mailing list networkmanager-list@gnome.org https://mail.gnome.org/mailman/listinfo/networkmanager-list
Re: [PATCH] Add libgsystem as a git submodule
On Friday 17 of May 2013 14:10:56 Colin Walters wrote: On Fri, 2013-05-17 at 08:04 -0400, Pavel Simerda wrote: Colin, do you have any ideas for the NetworkManager codebase apart from the cleanup functions? Possibly; the structured logging for example. NM could tag e.g. DHCP log messages with the interface ID, and system management interfaces could then show log messages related to that interface. If not, why don't we remove the submodule and just copy in the header with the relevant macros for now? Or maybe we could make the build system only choose relevant parts from libgsystem. Let me try a patch. Colin, thanks for the Build with GLib 2.32 as well patch. But still, libgsystem doesn't compile due to g_clear_pointer(): CC libgsystem_la-gsystem-file-utils.lo libgsystem/gsystem-file-utils.c: In function 'linkcopy_internal_attempt': libgsystem/gsystem-file-utils.c:358:3: error: implicit declaration of function 'g_clear_pointer' [-Werror=implicit-function-declaration] cc1: all warnings being treated as errors We now use g_clear_pointer() as well, but define it for glib 2.34 in include/nm-glib-compat.h Jirka ___ networkmanager-list mailing list networkmanager-list@gnome.org https://mail.gnome.org/mailman/listinfo/networkmanager-list
Re: [PATCH] Add libgsystem as a git submodule
On Tue, 2013-05-21 at 09:08 +0200, Jirka Klimes wrote: CC libgsystem_la-gsystem-file-utils.lo libgsystem/gsystem-file-utils.c: In function 'linkcopy_internal_attempt': libgsystem/gsystem-file-utils.c:358:3: error: implicit declaration of function 'g_clear_pointer' [-Werror=implicit-function-declaration] cc1: all warnings being treated as errors Can you try with https://git.gnome.org/browse/libgsystem/commit/?id=759384420c1578bf399473440ab34af18fa46b8d ___ networkmanager-list mailing list networkmanager-list@gnome.org https://mail.gnome.org/mailman/listinfo/networkmanager-list
Re: [PATCH] Add libgsystem as a git submodule
On Thursday 16 of May 2013 14:09:32 Dan Williams wrote: On Thu, 2013-05-16 at 04:53 -0400, Pavel Simerda wrote: From: Colin Walters walt...@verbum.org On Mon, 2013-05-06 at 12:50 -0500, Dan Williams wrote: On Fri, 2013-04-26 at 16:42 -0400, Colin Walters wrote: And change src/main.c to use the local allocation macros. This results in much cleaner code, as one can see from the diff. Because libgsystem is designed for nonrecursive make, it fits best in the current recursive setup if we build . first. This will be a lot nicer when we switch NM to a nonrecursive setup. --- .gitmodules |3 +++ Makefile.am |7 +++ libgsystem |1 + src/Makefile.am |2 ++ src/main.c | 50 -- 5 files changed, 25 insertions(+), 38 deletions(-) create mode 100644 .gitmodules create mode 16 libgsystem Patch looks good to me; Pavel, I'll assume you acked it too, given you had objections last time? Any thoughts, Pavel? Patch looks good to me. It might be good to have Dan Winship's (Cc) ack too, as he recently modified the build system files and might have some comments or suggestions. libgsystem pushed; everyone re-run autogen.sh to get the new module bits. Otherwise it'll fail pretty quick. Dan autogen.sh properly gets libgsystem. However it seems libgsystem simply uses latest GLib bits. I guess it should not if it can be avoided, because it spoils its usability. And that is the case for NetworkManager (that is just happy with glib 2.32). libgsystem now uses these symbols that require glib=2.34: * g_mapped_file_get_bytes (GMappedFile *file) * g_spawn_check_exit_status() * and G_SPAWN_EXIT_ERROR, G_SPAWN_SEARCH_PATH_FROM_ENVP Making a compatibility definition for them will allow compilation. Jirka ___ networkmanager-list mailing list networkmanager-list@gnome.org https://mail.gnome.org/mailman/listinfo/networkmanager-list
Re: [PATCH] Add libgsystem as a git submodule
- Original Message - From: Jirka Klimes jkli...@redhat.com To: Dan Williams d...@redhat.com Cc: Pavel Simerda psime...@redhat.com, Colin Walters walt...@verbum.org, networkmanager-list@gnome.org, danw d...@gnome.org Sent: Friday, May 17, 2013 9:37:29 AM Subject: Re: [PATCH] Add libgsystem as a git submodule On Thursday 16 of May 2013 14:09:32 Dan Williams wrote: On Thu, 2013-05-16 at 04:53 -0400, Pavel Simerda wrote: From: Colin Walters walt...@verbum.org On Mon, 2013-05-06 at 12:50 -0500, Dan Williams wrote: On Fri, 2013-04-26 at 16:42 -0400, Colin Walters wrote: And change src/main.c to use the local allocation macros. This results in much cleaner code, as one can see from the diff. Because libgsystem is designed for nonrecursive make, it fits best in the current recursive setup if we build . first. This will be a lot nicer when we switch NM to a nonrecursive setup. --- .gitmodules |3 +++ Makefile.am |7 +++ libgsystem |1 + src/Makefile.am |2 ++ src/main.c | 50 -- 5 files changed, 25 insertions(+), 38 deletions(-) create mode 100644 .gitmodules create mode 16 libgsystem Patch looks good to me; Pavel, I'll assume you acked it too, given you had objections last time? Any thoughts, Pavel? Patch looks good to me. It might be good to have Dan Winship's (Cc) ack too, as he recently modified the build system files and might have some comments or suggestions. libgsystem pushed; everyone re-run autogen.sh to get the new module bits. Otherwise it'll fail pretty quick. Dan autogen.sh properly gets libgsystem. However it seems libgsystem simply uses latest GLib bits. I guess it should not if it can be avoided, because it spoils its usability. And that is the case for NetworkManager (that is just happy with glib 2.32). libgsystem now uses these symbols that require glib=2.34: * g_mapped_file_get_bytes (GMappedFile *file) * g_spawn_check_exit_status() * and G_SPAWN_EXIT_ERROR, G_SPAWN_SEARCH_PATH_FROM_ENVP Making a compatibility definition for them will allow compilation. Would it be a problem to bump the requirement to 2.34? Pavel ___ networkmanager-list mailing list networkmanager-list@gnome.org https://mail.gnome.org/mailman/listinfo/networkmanager-list
Re: [PATCH] Add libgsystem as a git submodule
autogen.sh properly gets libgsystem. However it seems libgsystem simply uses latest GLib bits. I guess it should not if it can be avoided, because it spoils its usability. And that is the case for NetworkManager (that is just happy with glib 2.32). libgsystem now uses these symbols that require glib=2.34: * g_mapped_file_get_bytes (GMappedFile *file) * g_spawn_check_exit_status() * and G_SPAWN_EXIT_ERROR, G_SPAWN_SEARCH_PATH_FROM_ENVP Making a compatibility definition for them will allow compilation. Would it be a problem to bump the requirement to 2.34? Yes, we picked the current versions to work well with various distros. Are we planning to use anything in libgsystem besides the auto-free stuff? If not, why don't we remove the submodule and just copy in the header with the relevant macros for now? -- Dan ___ networkmanager-list mailing list networkmanager-list@gnome.org https://mail.gnome.org/mailman/listinfo/networkmanager-list
Re: [PATCH] Add libgsystem as a git submodule
From: Dan Winship d...@gnome.org autogen.sh properly gets libgsystem. However it seems libgsystem simply uses latest GLib bits. I guess it should not if it can be avoided, because it spoils its usability. And that is the case for NetworkManager (that is just happy with glib 2.32). libgsystem now uses these symbols that require glib=2.34: * g_mapped_file_get_bytes (GMappedFile *file) * g_spawn_check_exit_status() * and G_SPAWN_EXIT_ERROR, G_SPAWN_SEARCH_PATH_FROM_ENVP Making a compatibility definition for them will allow compilation. Would it be a problem to bump the requirement to 2.34? Yes, we picked the current versions to work well with various distros. Thanks. Are we planning to use anything in libgsystem besides the auto-free stuff? No idea. Colin, do you have any ideas for the NetworkManager codebase apart from the cleanup functions? If not, why don't we remove the submodule and just copy in the header with the relevant macros for now? Or maybe we could make the build system only choose relevant parts from libgsystem. I hope we don't have to revert it the second time and would rather have the bumped dependency for a couple of days. Maybe Colin comes up with some neat solution. Cheers, Pavel ___ networkmanager-list mailing list networkmanager-list@gnome.org https://mail.gnome.org/mailman/listinfo/networkmanager-list
Re: [PATCH] Add libgsystem as a git submodule
- Original Message - From: Colin Walters walt...@verbum.org To: Pavel Simerda psime...@redhat.com Cc: Dan Winship d...@gnome.org, jkli...@redhat.com, Dan Williams d...@redhat.com, networkmanager-list@gnome.org Sent: Friday, May 17, 2013 8:10:56 PM Subject: Re: [PATCH] Add libgsystem as a git submodule On Fri, 2013-05-17 at 08:04 -0400, Pavel Simerda wrote: Colin, do you have any ideas for the NetworkManager codebase apart from the cleanup functions? Possibly; the structured logging for example. NM could tag e.g. DHCP log messages with the interface ID, and system management interfaces could then show log messages related to that interface. Any more resources/documentation/posts for that? Sounds nice but I just can't imagine how it works, how NetworkManager could use it, whether it's somehow related to journald, etc... If not, why don't we remove the submodule and just copy in the header with the relevant macros for now? Or maybe we could make the build system only choose relevant parts from libgsystem. Let me try a patch. Thanks! As a side note, in future, I'll be adding a dependency on Jiří Pírko's libndp and we would most probably like to temporarily bundle it for distributions without libndp packages. As libgsystem is a dependency imported as a git submodule, it looks like you already did part of my work ;). Pavel ___ networkmanager-list mailing list networkmanager-list@gnome.org https://mail.gnome.org/mailman/listinfo/networkmanager-list
Re: [PATCH] Add libgsystem as a git submodule
On Fri, 2013-05-17 at 15:05 -0400, Pavel Simerda wrote:
Any more resources/documentation/posts for that?
Look at the attached patch as a demo. Results:
bash-4.2# journalctl NM_IFACE=eth0
-- Logs begin at Mon 2013-03-25 18:35:24 EDT, end at Fri 2013-05-17 16:31:52
EDT. --
May 17 16:31:18 qemux86-64 NetworkManager[327]: Bound interface eth0
bash-4.2#
As a side note, in future, I'll be adding a dependency on Jiří Pírko's libndp
Link?
From f5554e5371b27c564b0e248d6ccefd9da4176916 Mon Sep 17 00:00:00 2001
From: Colin Walters walt...@verbum.org
Date: Fri, 17 May 2013 16:35:14 -0400
Subject: [PATCH] Demo structured logging
---
configure.ac | 11 +++
src/Makefile.am |3 +++
src/dhcp-manager/nm-dhcp-client.c | 10 ++
3 files changed, 24 insertions(+), 0 deletions(-)
diff --git a/configure.ac b/configure.ac
index 43e420f..d7641fa 100644
--- a/configure.ac
+++ b/configure.ac
@@ -291,6 +291,17 @@ if test $with_session_tracking = consolekit; then
fi
AC_MSG_RESULT($with_session_tracking)
+AC_ARG_WITH(systemd-journal,
+ AS_HELP_STRING([--without-systemd-journal], [Use systemd @:@default=auto@:@]),
+ [], [with_systemd_journal=auto])
+AS_IF([test x$with_systemd_journal != xno], [
+ PKG_CHECK_MODULES([SYSTEMD_JOURNAL], [libsystemd-journal = 200], have_systemd_journal=yes, have_systemd_journal=no)
+ ])
+AM_CONDITIONAL(ENABLE_SYSTEMD_JOURNAL, test x$have_systemd_journal = xyes)
+AS_IF([test x$have_systemd_journal = xyes], [
+ AC_DEFINE([ENABLE_SYSTEMD_JOURNAL],[1],[Define if you want to build with systemd journal support])
+])
+
AC_ARG_WITH(suspend-resume, AS_HELP_STRING([--with-suspend-resume=upower|systemd], [Build NetworkManager with specific suspend/resume support]))
if test z$with_suspend_resume = z; then
PKG_CHECK_EXISTS([libsystemd-login = 183], [have_systemd_inhibit=yes], [have_systemd_inhibit=no])
diff --git a/src/Makefile.am b/src/Makefile.am
index d34e911..b7b49ef 100644
--- a/src/Makefile.am
+++ b/src/Makefile.am
@@ -345,6 +345,7 @@ AM_CPPFLAGS = \
$(MM_GLIB_CFLAGS) \
$(POLKIT_CFLAGS) \
$(SYSTEMD_LOGIN_CFLAGS) \
+ $(SYSTEMD_JOURNAL_CFLAGS) \
\
-DBINDIR=\$(bindir)\ \
-DDATADIR=\$(datadir)\ \
@@ -375,6 +376,7 @@ libNetworkManager_la_SOURCES = \
libNetworkManager_la_LIBADD = \
$(top_builddir)/libnm-util/libnm-util.la \
+ $(top_builddir)/libgsystem.la \
$(BLUEZ_LIBS) \
$(DBUS_LIBS) \
$(GLIB_LIBS) \
@@ -383,6 +385,7 @@ libNetworkManager_la_LIBADD = \
$(MM_GLIB_LIBS) \
$(POLKIT_LIBS) \
$(SYSTEMD_LOGIN_LIBS) \
+ $(SYSTEMD_JOURNAL_LIBS) \
$(LIBDL) \
$(LIBM)
diff --git a/src/dhcp-manager/nm-dhcp-client.c b/src/dhcp-manager/nm-dhcp-client.c
index e6c36e9..17d70e7 100644
--- a/src/dhcp-manager/nm-dhcp-client.c
+++ b/src/dhcp-manager/nm-dhcp-client.c
@@ -32,6 +32,7 @@
#include nm-logging.h
#include nm-dbus-glib-types.h
#include nm-dhcp-client.h
+#include libgsystem.h
typedef struct {
char * iface;
@@ -728,6 +729,15 @@ nm_dhcp_client_new_options (NMDHCPClient *self,
if (state_is_bound (new_state)) {
/* Cancel the timeout if the DHCP client is now bound */
timeout_cleanup (self);
+ {
+ #define BOUND_IFACE_MSGID bc498d9d08dfdda9432d74a6737ac73e
+ gs_free char *iface_key = g_strconcat (NM_IFACE=, priv-iface);
+ gs_free char *msg = g_strdup_printf (Bound interface %s, priv-iface);
+ char *keys[] = { MESSAGE_ID= BOUND_IFACE_MSGID,
+ iface_key,
+ NULL };
+ gs_log_structured (msg, keys);
+ }
}
if (priv-ipv6) {
--
1.7.1
___
networkmanager-list mailing list
networkmanager-list@gnome.org
https://mail.gnome.org/mailman/listinfo/networkmanager-list
Re: [PATCH] Add libgsystem as a git submodule
From: Colin Walters walt...@verbum.org On Fri, 2013-05-17 at 15:05 -0400, Pavel Simerda wrote: Any more resources/documentation/posts for that? Look at the attached patch as a demo. Ah, interesting. Is it capable of working without systemd at least to the level of using syslog and/org stderr? Results: bash-4.2# journalctl NM_IFACE=eth0 -- Logs begin at Mon 2013-03-25 18:35:24 EDT, end at Fri 2013-05-17 16:31:52 EDT. -- May 17 16:31:18 qemux86-64 NetworkManager[327]: Bound interface eth0 bash-4.2# As a side note, in future, I'll be adding a dependency on Jiří Pírko's libndp Link? https://github.com/jpirko/libndp https://bugzilla.gnome.org/show_bug.cgi?id=699772#c1 Cheers, Pavel ___ networkmanager-list mailing list networkmanager-list@gnome.org https://mail.gnome.org/mailman/listinfo/networkmanager-list
Re: [PATCH] Add libgsystem as a git submodule
On Fri, 2013-05-17 at 19:44 -0400, Pavel Simerda wrote: From: Colin Walters walt...@verbum.org On Fri, 2013-05-17 at 15:05 -0400, Pavel Simerda wrote: Any more resources/documentation/posts for that? Look at the attached patch as a demo. Ah, interesting. Is it capable of working without systemd at least to the level of using syslog and/org stderr? Yes, it just prints to stdout on non-systemd, though I'd definitely consider adding the ability to do syslog instead of stdout. ___ networkmanager-list mailing list networkmanager-list@gnome.org https://mail.gnome.org/mailman/listinfo/networkmanager-list
Re: [PATCH] Add libgsystem as a git submodule
From: Colin Walters walt...@verbum.org On Fri, 2013-05-17 at 19:44 -0400, Pavel Simerda wrote: From: Colin Walters walt...@verbum.org On Fri, 2013-05-17 at 15:05 -0400, Pavel Simerda wrote: Any more resources/documentation/posts for that? Look at the attached patch as a demo. Ah, interesting. Is it capable of working without systemd at least to the level of using syslog and/org stderr? Yes, it just prints to stdout on non-systemd, though I'd definitely consider adding the ability to do syslog instead of stdout. Works for me then. Pavel ___ networkmanager-list mailing list networkmanager-list@gnome.org https://mail.gnome.org/mailman/listinfo/networkmanager-list
Re: [PATCH] Add libgsystem as a git submodule
From: Colin Walters walt...@verbum.org On Mon, 2013-05-06 at 12:50 -0500, Dan Williams wrote: On Fri, 2013-04-26 at 16:42 -0400, Colin Walters wrote: And change src/main.c to use the local allocation macros. This results in much cleaner code, as one can see from the diff. Because libgsystem is designed for nonrecursive make, it fits best in the current recursive setup if we build . first. This will be a lot nicer when we switch NM to a nonrecursive setup. --- .gitmodules |3 +++ Makefile.am |7 +++ libgsystem |1 + src/Makefile.am |2 ++ src/main.c | 50 -- 5 files changed, 25 insertions(+), 38 deletions(-) create mode 100644 .gitmodules create mode 16 libgsystem Patch looks good to me; Pavel, I'll assume you acked it too, given you had objections last time? Any thoughts, Pavel? Patch looks good to me. It might be good to have Dan Winship's (Cc) ack too, as he recently modified the build system files and might have some comments or suggestions. Thanks! Pavel ___ networkmanager-list mailing list networkmanager-list@gnome.org https://mail.gnome.org/mailman/listinfo/networkmanager-list
Re: [PATCH] Add libgsystem as a git submodule
On Mon, 2013-05-06 at 12:50 -0500, Dan Williams wrote: On Fri, 2013-04-26 at 16:42 -0400, Colin Walters wrote: And change src/main.c to use the local allocation macros. This results in much cleaner code, as one can see from the diff. Because libgsystem is designed for nonrecursive make, it fits best in the current recursive setup if we build . first. This will be a lot nicer when we switch NM to a nonrecursive setup. --- .gitmodules |3 +++ Makefile.am |7 +++ libgsystem |1 + src/Makefile.am |2 ++ src/main.c | 50 -- 5 files changed, 25 insertions(+), 38 deletions(-) create mode 100644 .gitmodules create mode 16 libgsystem Patch looks good to me; Pavel, I'll assume you acked it too, given you had objections last time? Any thoughts, Pavel? ___ networkmanager-list mailing list networkmanager-list@gnome.org https://mail.gnome.org/mailman/listinfo/networkmanager-list
Re: [PATCH] Add libgsystem as a git submodule
On Fri, 2013-04-26 at 16:42 -0400, Colin Walters wrote: And change src/main.c to use the local allocation macros. This results in much cleaner code, as one can see from the diff. Because libgsystem is designed for nonrecursive make, it fits best in the current recursive setup if we build . first. This will be a lot nicer when we switch NM to a nonrecursive setup. --- .gitmodules |3 +++ Makefile.am |7 +++ libgsystem |1 + src/Makefile.am |2 ++ src/main.c | 50 -- 5 files changed, 25 insertions(+), 38 deletions(-) create mode 100644 .gitmodules create mode 16 libgsystem Patch looks good to me; Pavel, I'll assume you acked it too, given you had objections last time? Dan ___ networkmanager-list mailing list networkmanager-list@gnome.org https://mail.gnome.org/mailman/listinfo/networkmanager-list
Re: [PATCH] Add libgsystem as a git submodule
On Fri, 2013-04-26 at 16:42 -0400, Colin Walters wrote: And change src/main.c to use the local allocation macros. This results in much cleaner code, as one can see from the diff. Because libgsystem is designed for nonrecursive make, it fits best in the current recursive setup if we build . first. This will be a lot nicer when we switch NM to a nonrecursive setup. Any opinions on this? ___ networkmanager-list mailing list networkmanager-list@gnome.org https://mail.gnome.org/mailman/listinfo/networkmanager-list
[PATCH] Add libgsystem as a git submodule
And change src/main.c to use the local allocation macros. This
results in much cleaner code, as one can see from the diff.
Because libgsystem is designed for nonrecursive make, it fits best in
the current recursive setup if we build . first. This will be a lot
nicer when we switch NM to a nonrecursive setup.
---
.gitmodules |3 +++
Makefile.am |7 +++
libgsystem |1 +
src/Makefile.am |2 ++
src/main.c | 50 --
5 files changed, 25 insertions(+), 38 deletions(-)
create mode 100644 .gitmodules
create mode 16 libgsystem
From 1bf944af8a321cebe2ddc45d0649f744c8828bab Mon Sep 17 00:00:00 2001
From: Colin Walters walt...@verbum.org
Date: Fri, 26 Apr 2013 16:20:22 -0400
Subject: [PATCH] Add libgsystem as a git submodule
And change src/main.c to use the local allocation macros. This
results in much cleaner code, as one can see from the diff.
Because libgsystem is designed for nonrecursive make, it fits best in
the current recursive setup if we build . first. This will be a lot
nicer when we switch NM to a nonrecursive setup.
---
.gitmodules |3 +++
Makefile.am |7 +++
libgsystem |1 +
src/Makefile.am |2 ++
src/main.c | 50 --
5 files changed, 25 insertions(+), 38 deletions(-)
create mode 100644 .gitmodules
create mode 16 libgsystem
diff --git a/.gitmodules b/.gitmodules
new file mode 100644
index 000..e93bbea
--- /dev/null
+++ b/.gitmodules
@@ -0,0 +1,3 @@
+[submodule libgsystem]
+ path = libgsystem
+ url = git://git.gnome.org/libgsystem
diff --git a/Makefile.am b/Makefile.am
index 2847616..9419d35 100644
--- a/Makefile.am
+++ b/Makefile.am
@@ -1,6 +1,7 @@
include $(GLIB_MAKEFILE)
SUBDIRS = \
+ . \
include \
libnm-util \
libnm-glib \
@@ -53,3 +54,9 @@ CLEANFILES = cscope.in.out cscope.out cscope.po.out
.PHONY: cscope
cscope:
cscope -b -q -R -Iinclude -ssrc -slibnm-glib -slibnm-util -scli/src;
+
+libgsystem_srcpath := libgsystem
+libgsystem_cflags := $(GLIB_CFLAGS) -I$(srcdir)/libgsystem
+libgsystem_libs = $(GLIB_LIBS)
+include libgsystem/Makefile-libgsystem.am
+noinst_LTLIBRARIES = libgsystem.la
diff --git a/libgsystem b/libgsystem
new file mode 16
index 000..51736db
--- /dev/null
+++ b/libgsystem
@@ -0,0 +1 @@
+Subproject commit 51736db5f49b4b798fd2f9d0ccb8eff453105162
diff --git a/src/Makefile.am b/src/Makefile.am
index 113b527..f81c37a 100644
--- a/src/Makefile.am
+++ b/src/Makefile.am
@@ -26,6 +26,7 @@ endif
SUBDIRS += . tests
INCLUDES = -I${top_srcdir} \
+ -I${top_srcdir}/libgsystem \
-I${top_builddir}/include \
-I${top_srcdir}/include \
-I${top_builddir}/src/generated \
@@ -324,6 +325,7 @@ NetworkManager_CPPFLAGS += -DCKDB_PATH=\${CKDB_PATH}\
endif
NetworkManager_LDADD = \
+ $(top_builddir)/libgsystem.la \
./generated/libnm-generated.la \
./logging/libnm-logging.la \
./config/libnm-config.la \
diff --git a/src/main.c b/src/main.c
index ed71bb6..37e05c4 100644
--- a/src/main.c
+++ b/src/main.c
@@ -38,6 +38,7 @@
#include gmodule.h
#include string.h
+#include gsystem-local-alloc.h
#include NetworkManager.h
#include NetworkManagerUtils.h
#include nm-manager.h
@@ -305,18 +306,19 @@ main (int argc, char *argv[])
GOptionContext *opt_ctx = NULL;
gboolean become_daemon = FALSE;
gboolean g_fatal_warnings = FALSE;
- char *pidfile = NULL, *state_file = NULL;
+ gs_free char *pidfile = NULL;
+ gs_free char *state_file = NULL;
gboolean wifi_enabled = TRUE, net_enabled = TRUE, wwan_enabled = TRUE, wimax_enabled = TRUE;
gboolean success, show_version = FALSE;
NMPolicy *policy = NULL;
- NMVPNManager *vpn_manager = NULL;
- NMDnsManager *dns_mgr = NULL;
- NMDBusManager *dbus_mgr = NULL;
- NMSupplicantManager *sup_mgr = NULL;
- NMDHCPManager *dhcp_mgr = NULL;
- NMFirewallManager *fw_mgr = NULL;
- NMSettings *settings = NULL;
- NMConfig *config;
+ gs_unref_object NMVPNManager *vpn_manager = NULL;
+ gs_unref_object NMDnsManager *dns_mgr = NULL;
+ gs_unref_object NMDBusManager *dbus_mgr = NULL;
+ gs_unref_object NMSupplicantManager *sup_mgr = NULL;
+ gs_unref_object NMDHCPManager *dhcp_mgr = NULL;
+ gs_unref_object NMFirewallManager *fw_mgr = NULL;
+ gs_unref_object NMSettings *settings = NULL;
+ gs_unref_object NMConfig *config = NULL;
GError *error = NULL;
gboolean wrote_pidfile = FALSE;
@@ -579,41 +581,13 @@ done:
if (policy)
nm_policy_destroy (policy);
- if (manager)
- g_object_unref (manager);
-
- if (settings)
- g_object_unref (settings);
-
- if (vpn_manager)
- g_object_unref (vpn_manager);
-
- if (dns_mgr)
- g_object_unref (dns_mgr);
-
- if (dhcp_mgr)
- g_object_unref (dhcp_mgr);
-
- if (sup_mgr)
- g_object_unref (sup_mgr);
-
- if (fw_mgr)
- g_object_unref (fw_mgr);
-
- if (dbus_mgr)
- g_object_unref (dbus_mgr);
+ g_clear_object (manager);
nm_logging_shutdown ();
if (pidfile wrote_pidfile
