This does several things: - Bump our required gtk version at build time to 2.18. I don't think it's worth supporting older versions at this point (see https://bugzilla.mozilla.org/show_bug.cgi?id=772563#c13 ) - Get rid of gtk2compat.h and replace it with ad-hoc header wrappers. This allows not to care about including gtk2compat.h where needed (and avoid left-overs). - Contrary to gtk2compat.h, the header wrappers lead to builds compatible with gtk 2.18/glib 2.22, whatever version of gtk/glib development files are used.

Comment on attachment 663990 [details] [diff] [review] Force builds to be compatible with gtk 2.18/glib 2.22 Not having to explicitly include gtk2compat.h is nice. gdk_x11_window_get_drawable_impl is in GTK+ 2.18: http://git.gnome.org/browse/gtk+/tree/gdk/x11/gdkx.h?id=2.18.0 I guess it is a nice optimization to skip the function call. Was there another reason for including this function? I'm not happy substituting g_ptr_array_free for g_ptr_array_unref as they mean different things. The code in UPowerClient.cpp wasn't reviewed and should have just used g_ptr_array_free (bug 712442 comment 1), but anyway both symbols are in GLib 2.22, which we now require. For g_new, g_new0, g_renew in gmem.h, can you include the (gsize)num <= G_MAXSIZE / sizeof(type) overflow checks, please? There shouldn't be any need to cast sizeof() to gsize. If it makes things easier, then perhaps instead define static inline functions g_malloc_n, g_malloc0_n, g_realloc_n to perform the overflow check.

(In reply to Karl Tomlinson (:karlt) from comment #2) > Comment on attachment 663990 [details] [diff] [review] > Force builds to be compatible with gtk 2.18/glib 2.22 > > Not having to explicitly include gtk2compat.h is nice. > > gdk_x11_window_get_drawable_impl is in GTK+ 2.18: > http://git.gnome.org/browse/gtk+/tree/gdk/x11/gdkx.h?id=2.18.0 > I guess it is a nice optimization to skip the function call. > Was there another reason for including this function? Looks like leftovers from when I tried to make things binary compatible with older versions. > I'm not happy substituting g_ptr_array_free for g_ptr_array_unref as they > mean > different things. The code in UPowerClient.cpp wasn't reviewed and should > have just used g_ptr_array_free (bug 712442 comment 1), but anyway both > symbols are in GLib 2.22, which we now require. Got there like gdk_x11_window_get_drawable_impl, I guess.

Comment on attachment 664441 [details] [diff] [review] Force builds to be compatible with gtk 2.18/glib 2.22 Is mozilla/Assertions.h used? I wonder why you avoided G_GSIZE_FORMAT? I'm guessing the (gsize)(num) casts are unnecessary, and we'd actually be better off with compiler warnings/errors should anything odd be passed for that argument.

(In reply to Karl Tomlinson (:karlt) from comment #5) > Comment on attachment 664441 [details] [diff] [review] > Force builds to be compatible with gtk 2.18/glib 2.22 > > Is mozilla/Assertions.h used? Gah, a leftover from when i thought about using MOZ_CRASH instead of g_error. > I wonder why you avoided G_GSIZE_FORMAT? Because it's not exported in glib headers :(

Here it's hidden away at /usr/lib64/glib-2.0/include/glibconfig.h but that should be pulled in by gmem.h. http://developer.gnome.org/glib/stable/glib-Basic-Types.html#G-GSIZE-FORMAT:CAPS

Here is a refreshed patch with review comments addressed

Comment on attachment 664465 [details] [diff] [review] Force builds to be compatible with gtk 2.18/glib 2.22. Carrying over r+

This broke comm-central. See bug 794285.

Comment on attachment 664465 [details] [diff] [review] Force builds to be compatible with gtk 2.18/glib 2.22. [Approval Request Comment] Bug caused by (feature/regressing bug #): bug 772446 User impact if declined: ESR17 on ancient build platform Testing completed (on m-c, etc.): already landed on m-c Risk to taking this patch (and alternatives if risky): changes runtime requirements for linux users String or UUID changes made by this patch: none

Comment on attachment 664465 [details] [diff] [review] Force builds to be compatible with gtk 2.18/glib 2.22. yes to not having ESR17 on ancient build platform :)

Pushed to Aurora: http://hg.mozilla.org/releases/mozilla-aurora/rev/9be5b2790273