180741 - nsNativeThemeGTK::nsNativeThemeGTK crashes if it can't get a display

Status: RESOLVED WORKSFORME

(Core :: Widget: Gtk, defect)

Description

[timeless]

Gtk-WARNING **: gtk_type_create(): unknown parent type `21'.

Gtk-CRITICAL **: file gtkwidget.c: line 1641 (gtk_widget_realize): assertion
`GTK_IS_WIDGET (widget)' failed.

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 1024 (LWP 26366)]
0x41b3d896 in nsNativeThemeGTK::nsNativeThemeGTK (this=0x850c760) at
nsNativeThemeGTK.cpp:104
104       GtkThemeEngine* gtkThemeEng = gButtonWidget->style->engine;
(gdb) where
#0  0x41b3d896 in nsNativeThemeGTK::nsNativeThemeGTK (this=0x850c760) at
nsNativeThemeGTK.cpp:104
#1  0x41b1795b in nsNativeThemeGTKConstructor (aOuter=0x0, aIID=@0xbfffd7f8,
aResult=0xbfffd7e8)
    at nsGfxFactoryGTK.cpp:89
#2  0x4020dc9a in nsGenericFactory::CreateInstance (this=0x850c740, aOuter=0x0,
aIID=@0xbfffd7f8,
    aResult=0xbfffd7e8) at nsGenericFactory.cpp:84
#3  0x401bb709 in nsComponentManagerImpl::CreateInstance (this=0x8058158,
aClass=@0x850e89c, aDelegate=0x0,
    aIID=@0xbfffd7f8, aResult=0xbfffd7e8) at nsComponentManager.cpp:1817
#4  0x401c06d6 in nsComponentManager::CreateInstance (aClass=@0x850e89c,
aDelegate=0x0, aIID=@0xbfffd7f8,
    aResult=0xbfffd7e8) at nsComponentManagerObsolete.cpp:102
#5  0x4050e062 in nsJSCID::CreateInstance (this=0x850e880, _retval=0xbfffda58)
at xpcjsid.cpp:793
#6  0x401f1ca8 in XPTC_InvokeByIndex () at xptcinvoke_gcc_x86_unix.cpp:86
#7  0x40524c39 in XPCWrappedNative::CallMethod (ccx=@0xbfffdb10,
mode=CALL_METHOD) at xpcwrappednative.cpp:2016
#8  0x4052dc2e in XPC_WN_CallMethod (cx=0x80b80b0, obj=0x867d9d8, argc=0,
argv=0x80d47e8, vp=0xbfffdc50)
    at xpcwrappednativejsops.cpp:1283
#9  0x4006291b in js_Invoke (cx=0x80b80b0, argc=0, flags=0) at jsinterp.c:839
#10 0x40071a62 in js_Interpret (cx=0x80b80b0, result=0xbffff6dc) at jsinterp.c:2803
#11 0x4006306a in js_Execute (cx=0x80b80b0, chain=0x808dcc0, script=0x80d4640,
down=0x0, special=0,
    result=0xbffff6dc) at jsinterp.c:1020
#12 0x40030677 in JS_ExecuteScript (cx=0x80b80b0, obj=0x808dcc0,
script=0x80d4640, rval=0xbffff6dc)
    at jsapi.c:3277
#13 0x804badb in Process (cx=0x80b80b0, obj=0x808dcc0, filename=0xbffff9d5
"driver.js", filehandle=0x0)
    at xpcshell.cpp:479
#14 0x804c1d5 in ProcessArgs (cx=0x80b80b0, obj=0x808dcc0, argv=0xbffff878,
argc=1) at xpcshell.cpp:655
#15 0x804cd8e in main (argc=1, argv=0xbffff878) at xpcshell.cpp:912
#16 0x403852eb in __libc_start_main (main=0x804c544 <main>, argc=2,
ubp_av=0xbffff874, init=0x804a7a4 <_init>,
    fini=0x804f030 <_fini>, rtld_fini=0x4000c130 <_dl_fini>, stack_end=0xbffff86c)
    at ../sysdeps/generic/libc-start.c:129

(gdb) p gButtonWidget->style
$2 = (GtkStyle *) 0x0

Testcase:
Components.classes["@mozilla.org/chrome/chrome-native-theme;1"].createInstance()

There are lots and lots of asserts that you get from this single line, i don't
know how to get GLib/Gtk to break/crash so that i can get an earlier stack.

I can attach a fix which will make the testcase not crash.

The constructor shouldn't crash, and any initialization that can fail should be
attached to an Init method.

Comment 1

[timeless]

Attachment: patch

someone will at somepoint want to refactor this code to use an Init() function.

Comment 2

[Serge Gautherie (:sgautherie)]

Comment on attachment 106691 [details] [diff] [review]
patch


This patch was made against r1.32;
r1.50 is current, and this patch seems to have fully bit-rotted now...

Comment 3

[Reed Loden [:reed]]

timeless, is this still a problem after all these years?

Comment 4

[timeless]

seems to work w/ ff2?

timeless@swift:/usr/lib/firefox% ./run-mozilla.sh ./xpcshell
js> build()
built on Nov 13 2006 at 15:50:51
js> Components.classes["@mozilla.org/chrome/chrome-native-theme;1"]. createInstance()

[xpconnect wrapped nsISupports]

  CATEGORY:  FIREFOX,application,JDS4
   VERSION:  2.0,REV=110.0.4.2006.11.13.15.42