It used to work, but now the expading/collapsing test isn't flowing around the floating div on the right. btw, the :first-line improvement is cool, seems much faster!‰

Setting all current Open/Normal to M4.

per leger, assigning QA contacts to all open bugs without QA contacts according to list at http://bugzilla.mozilla.org/describecomponents.cgi?product=Browser

Currently the above test crashes on me after a short amount of time with: Assertion failure: root_points_to_gcArenaPool, at jsgc.c:599 Here is the stack trace: #3 0x407abbf7 in JS_Assert (s=0x407b2ae8 "root_points_to_gcArenaPool", file=0x407b2ac6 "jsgc.c", ln=599) at jsutil.c:130 #4 0x4075ee74 in gc_root_marker (he=0x82bd110, i=6, arg=0x81ef960) at jsgc.c:599 #5 0x40760385 in JS_HashTableEnumerateEntries (ht=0x81f1b48, f=0x4075edcc <gc_root_marker at jsgc.c:583>, arg=0x81ef960) at jshash.c:347 #6 0x4075f0cf in js_GC (cx=0x81f2140) at jsgc.c:717 #7 0x4075eedc in js_ForceGC (cx=0x81f2140) at jsgc.c:614 #8 0x40737cc3 in JS_GC (cx=0x81f2140) at jsapi.c:865 #9 0x404e16b1 in nsJSContext::GC (this=0x81f2108) at nsJSEnvironment.cpp:295 #10 0x404e16ef in nsJSContext::ScriptEvaluated (this=0x81f2108) at nsJSEnvironment.cpp:306 #11 0x404e866c in GlobalWindowImpl::RunTimeout (this=0x8201b10, aTimeout=0x82bcfe8) at nsGlobalWindow.cpp:1133 #12 0x404e844d in nsGlobalWindow_RunTimeout (aTimer=0x8285e38, aClosure=0x82bcfe8) at nsGlobalWindow.cpp:1037 #13 0x4046cdf9 in TimerImpl::FireTimeout (this=0x8285e38) at nsTimer.cpp:73 #14 0x4046d2e2 in nsTimerExpired (aCallData=0x8285e38) at nsTimer.cpp:189 #15 0x40992990 in g_timeout_dispatch (source_data=0x8289488, current_time=0xbffff6d0, user_data=0x8285e38) at gmain.c:1147 #16 0x40991c83 in g_main_dispatch (current_time=0xbffff6d0) at gmain.c:647 #17 0x4099220f in g_main_iterate (block=1, dispatch=1) at gmain.c:854 #18 0x40992391 in g_main_run (loop=0x81cf6f8) at gmain.c:912 #19 0x408bf44b in gtk_main () at gtkmain.c:475 #20 0x400475e0 in nsAppShell::Run (this=0x80998e8) at nsAppShell.cpp:182 #21 0x8053832 in nsNativeViewerApp::Run (this=0x8092a00) at nsGTKMain.cpp:42 #22 0x8053a46 in main (argc=2, argv=0xbffff77c) at nsGTKMain.cpp:97

Scott, this is the test case for the JS_GC() problem we discussed earlier. The timeout code is adding named roots for each of the arguments to a timeout. Some of these arguments could be scalar values. The GC code isn't taking that into account.

Fix in revision 3.5.12.10 of jsgc.c (on SpiderMonkey140_BRANCH). Will be checked into trunk when branch lands in next week or so.

Now fixed in the trunk too. Reassigning to Kipp, as per Vidur's request.

*** This bug has been marked as a duplicate of 1264 ***

Agreed. This is a duplicate of 1264.

revised delivery date - 07/22

Sorry again the revised delivery date is 07/21. Richard