Created attachment 326810 [details] [review]
pixelcache: reuse existing timeout source when possible

First off, this might seem like a layer violation simply because we
know that it is a timeout source and that the timeout source uses
the ready time to fire. But that couldn't really change without a
break in GLib, so it seems perfectly fine to rely on it.

This avoids the g_source_remove(), g_source_destroy(),
g_timer_source_new(), and g_source_set_name_by_id() in the common case.

Instead, we reuse our previous source and update the ready time to our
new deadline. We lose the coalescing with g_timeout_add_seconds(), but
that is not going to help in the common case anyway (unless you have
three hands and can scroll multiple pixelcached backed widgets at once).

Review of attachment 326810 [details] [review]:

::: gtk/gtkpixelcache.c
@@ +449,2 @@
+      deadline = g_get_monotonic_time () + (BLOW_CACHE_TIMEOUT_SEC * G_USEC_PER_SEC);
+      source = g_main_context_find_source_by_id (NULL, cache->timeout_tag);

if you go to this length, why not go all the way: keep the source pointer instead of the tag, and avoid the find-by-id ?

Created attachment 326823 [details] [review]
pixelcache: reuse existing timeout source when possible

This avoids the g_source_remove(), g_source_destroy(),
g_timer_source_new(), and g_source_set_name_by_id() in the common case.

Instead, we reuse our previous source and update the ready time to our
new deadline. We lose the coalescing with g_timeout_add_seconds(), but
that is not going to help in the common case anyway (unless you have
three hands and can scroll multiple pixelcached backed widgets at once).

I was trying to be a little bit too minimal in the patch! This caches the GSource instead as suggested.

While I was there, I cleaned up a couple other open coded pointer check/frees.

Attachment 326823 [details] pushed as 0763a02 - pixelcache: reuse existing timeout source when possible