On Wed, Oct 04, 2006 at 07:17:47PM +0200, Andre Poenitz wrote:
> iiiiiiOn Tue, Oct 03, 2006 at 11:13:03PM +0200, Enrico Forestieri wrote:
> > Program received signal SIGSEGV, Segmentation fault.
> > 0x00b61a38 in comp_func_solid_Source ()
> > (gdb) bt
> > #0  0x00b61a38 in comp_func_solid_Source ()
> > #1  0x00b2d29e in blend_color_generic ()
> > #2  0x00a43c74 in drawLine_midpoint_i ()
> > #3  0x00a4c2ed in QRasterPaintEngine::drawPolygon ()
> 
> Welcome to the joys of the Raster Painter, the best thing
> since the invention of sliced bread...

;-)

> Can you get hold of a backtrace with full debug info?

Unfortunately I am not able to build LyX with debug info anymore
(I remember I could before the unicode hurricane). The final link
stage fails because I have not enough memory (256 Mb here).

> Also, which version of Qt is this?

4.1.4

> For a while it was fairly easy to crash the raster painter with
> coordinates that were "much" larger thatn the screen.
> 
> I'd suspect that this is the case here, too, but to make sure I'd
> have to seen the actual coordinates.

This is what gets passed to drawPolyline just before the crash:

i  xp[i] yp[i]
--------------
0)  109   396
1)  109   397
2)  111   398
3)  113   399
4)  111   400
5)  109   401
6)  109   402

but I also saw some negative values for yp[i] a bit before the crash,
such as (they were in sequence):

0) 368, -12
1) 368, -11
2) 370, -10
3) 372, -9
4) 370, -8
5) 368, -7
6) 368, -6

0) 401, -12
1) 401, -11
2) 403, -10
3) 405, -9
4) 403, -8
5) 401, -7
6) 401, -6

0) 439, -12
1) 439, -11
2) 441, -10
3) 443, -9
4) 441, -8
5) 439, -7
6) 439, -6

0) 586, -12
1) 586, -11
2) 588, -10
3) 590, -9
4) 588, -8
5) 586, -7
6) 586, -6

-- 
Enrico

Reply via email to