app_server crash in Screen:Frame().

[bga]

And you thought the BDirectWindow saga was over!

Unfortunately I have no stack crawl as my USB keyboard does not work with the app_server debug console. i think this should not be much of a problem as, at least here, it happens 100% of the time.

Here is how I reproduce it:

1 - Change to some resolution using BScreen::SetMode().
2 - Create a BDirectWindow and set it full screen.
3 - Switch to some higher resolution with BScreen::SetMode().
4 - Crash.

[anevilyak]

I'm seeing this as well when the screen saver kicks in:

vm_soft_fault: va 0x0 not covered by area in address space
vm_page_fault: vm_soft_fault returned error 'Bad address' on fault at 0xc, ip 0x2628a0, write 0, user 1, thread 0x521
vm_page_fault: thread "w:1310:ScreenSaver Window" (1313) in team "app_server" (62) tried to read address 0xc, ip 0x2628a0 ("app_server_seg0ro" +0x628a0)
debug_server: Thread 1313 entered the debugger: Segment violation
stack trace, current PC 0x2628a0  _ZNK6Screen5FrameEv + 0xe:
  (0x705d5ddc)  0x27415a  _ZN12ServerWindow19_ResizeToFullScreenEv + 0x2c
  (0x705d5e1c)  0x27422f  _ZN12ServerWindow26_DirectWindowSetFullScreenEb + 0x6b
  (0x705d5e5c)  0x27b5e6  _ZN12ServerWindow16_DispatchMessageElRN8BPrivate12LinkReceiverE + 0xda2
  (0x705d5f1c)  0x276809  _ZN12ServerWindow14_MessageLooperEv + 0x27f
  (0x705d5f9c)  0x25f1a3  _ZN13MessageLooper15_message_threadEPv + 0xf
  (0x705d5fbc)  0x74e0cd  thread_entry + 0x1a

[jackburton]

Here's a backtrace.

[jackburton]

Should be fixed in hrev32774, can you confirm ?

[axeld]

Damn, I should have accepted that bug before looking into it; thanks Stefano! I've added proper locking in hrev32775, as that was missing, too.

[jackburton]

Replying to axeld:

Damn, I should have accepted that bug before looking into it; thanks Stefano! I've added proper locking in hrev32775, as that was missing, too.

Well I should have, too :-)