Re: [Patch] Re: Crash when closing document

[Stefan Schimanski]

Am 29.05.2007 um 22:03 schrieb Andre Poenitz:

On Tue, May 29, 2007 at 07:29:56PM +0200, Abdelrazak Younes wrote:

Abdelrazak Younes wrote:

Richard Heck wrote:

Can you send me the full backtrace? What you sent suggests that the

the CursorSlice, at lines 45 and 64 of CursorSlice.cpp.

crash is being caused when the destroyed() signal is sent by the new buffer's text inset, and the only thing that seems to connect to it is

and commit if it does.

If that is the problem, the attached patch should solve it. Please test

The patch...

Index: CursorSlice.cpp
===================================================================
--- CursorSlice.cpp     (revision 18563)
+++ CursorSlice.cpp     (working copy)
@@ -42,7 +42,7 @@
        : inset_(&p), idx_(0), pit_(0), pos_(0)
 {
        BOOST_ASSERT(inset_);
-       inset_->destroyed.connect(
+       inset_connection_ = inset_->destroyed.connect(
                        boost::bind(&CursorSlice::invalidate, this));
 }

Erm. I seem to have missed the discussion of using slot/signals
in individual cursor slices.

I hope this is only a short term solution to get 1.5 out of the
door.

Stefan

I enabled stdlib-debug and then were surprised about the slowness. Some profiling was full of signal/slot stuff in copying cursors....

PGP.sig
Description: Signierter Teil der Nachricht