[gjs] object: Fix comments about toggle refs
- From: Philip Chimento <pchimento src gnome org>
- To: commits-list gnome org
- Cc:
- Subject: [gjs] object: Fix comments about toggle refs
- Date: Sun, 12 Mar 2017 05:13:34 +0000 (UTC)
commit d55d10a7f485ea4d56bd744f399c0835b1f64584
Author: Philip Chimento <philip endlessm com>
Date: Thu Mar 9 18:43:17 2017 -0800
object: Fix comments about toggle refs
These comments were lagging behind the code since we switched from
KeepAlive to GjsMaybeOwned.
https://bugzilla.gnome.org/show_bug.cgi?id=778862
gi/object.cpp | 6 +++---
1 files changed, 3 insertions(+), 3 deletions(-)
---
diff --git a/gi/object.cpp b/gi/object.cpp
index 3ca662b..c68ad4c 100644
--- a/gi/object.cpp
+++ b/gi/object.cpp
@@ -1175,7 +1175,7 @@ wrapped_gobj_toggle_notify(gpointer data,
* to it.
* In case we're toggling up (and thus rooting the JS object) we
* also need to take care if GC is running. The marking side
- * of it is taken care by JS::Heap, which we use in KeepAlive,
+ * of it is taken care by JS::Heap, which we use in GjsMaybeOwned,
* so we're safe. As for sweeping, it is too late: the JS object
* is dead, and attempting to keep it alive would soon crash
* the process. Plus, if we touch the JSAPI, libmozjs aborts in
@@ -1189,8 +1189,8 @@ wrapped_gobj_toggle_notify(gpointer data,
* and C code needs to clean after itself before it returns
* from dispose()/finalize().
* On the other hand, toggling down is a lot simpler, because
- * we're creating more garbage. So we just add the object to
- * the keep alive and wait for the next GC cycle.
+ * we're creating more garbage. So we just unroot the object, make it a
+ * weak pointer, and wait for the next GC cycle.
*
* Note that one would think that toggling up only happens
* in the main thread (because toggling up is the result of
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]