gnustep-dev
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Problem with Gorm


From: Eric Wasylishen
Subject: Re: Problem with Gorm
Date: Mon, 11 Jul 2011 10:51:24 -0600

Thanks for finding that Fred, I bet that's it. 

Germán, you could try a minimal test program like:

#import <Foundation/NSValue.h>
int main(int argc, const char **argv)
{
        NSSize s = [((NSValue*)nil) sizeValue];
        return 0;
}

and see if that segfaults on sparc.

Eric

On 2011-07-11, at 3:42 AM, Fred Kiefer wrote:

> Here we got the size from rep before the test for nil. On some machines 
> (sparc) such a call may lead to a segmentation fault. Not sure, whether this 
> was your problem, but you should try again with current SVN to see if 
> something changed.
> 
> Fred
> 
> On 09.07.2011 02:38, Germán Arias wrote:
>> Well, Gorm crash at line 959 in NSImage.m (when rep is equal to nil)
>> 
>>   if (rep == nil)
>>       return;
>> 
>> after call [self nativeDrawInRect: dstRect fromRect: .... in line 1293.
>> But definitely this don't have sense. Fortunately debug=yes solve the
>> problem.
>> 
>> 
>> On lun, 2011-07-04 at 12:23 -0600, Eric Wasylishen wrote:
>>> That's really strange. I don't see any #ifdef's in NSImage.m.
>>> 
>>> Maybe you can try to compile without "debug=yes" (so you get the crash), 
>>> but somehow add "-g" to the compile flags so debug info is generated. Do 
>>> you get any info if you select stack frame 0 and type "l" (for listing 
>>> source code)?
>>> 
>>> 
>>> On 2011-07-04, at 12:15 PM, Germán Arias wrote:
>>> 
>>>> If the problem disappears with "debug=yes". Could be the cause a code
>>>> line inside a #ifdef DEBUG? A code line that should be outside of that
>>>> ifdef?
>>>> 
>>>> On lun, 2011-07-04 at 10:41 +0200, Fred Kiefer wrote:
>>>>> Which backend are you using? That will result in different methods being
>>>>> called next and most likely the problem is one level deeper.
>>>>> 
>>>>> 
>>>>> I had a quick look into these new methods and didn't like too much what
>>>>> I saw there. We have a lot of code duplication between these two
>>>>> methods, but we also have differences there which I don't understand.
>>>>> Why do we use scaling for the native drawing but not for the gui drawing?
>>>>> Anyway, the code in native drawing should not check for
>>>>> isDrawingToScreen. When drawing to a PDF or PostScript surface we should
>>>>> be able to composite or dissolve, we just need to package that up
>>>>> correctly as backend operations. And of course need to have the PS and
>>>>> PDF context report back that they are able to support this.
>>>>> 
>>>>> I also noticed that the new code wont fill the cached image rep with the
>>>>> image background colour. Why isn't this done and does this match the
>>>>> Cocoa behaviour?
>>>>> Why does the old code (guiDrawInRect:...) use PSgsave and PSgrestore
>>>>> around the cache drawing? We wont do any other operations on the cache
>>>>> and don't need to restore the old state.
>>>>> 
>>>>> I think that we should start a discussion about this whole drawing code.
>>>>> Hopefully we end up with a deeper understanding of what should happen
>>>>> here and will then be able to package it up with a nice and clean call
>>>>> to the backend and three or four (almost?) correct implementations for
>>>>> this in GSContext, GSStreamContext and GSCairoContext. But then I may be
>>>>> completely wrong here, as I really don't understand much of this new code.
>>>>> 
>>>>> On 03.07.2011 01:49, Germán Arias wrote:
>>>>>> Well, after compiling gui with "debug=yes" all works fine. But then,
>>>>>> recompiling gui just with "make" the problem is there again.
>>>>>> 
>>>>>> 
>>>>>> On sáb, 2011-07-02 at 11:36 -0600, Eric Wasylishen wrote:
>>>>>>> Hey German,
>>>>>>> I wasn't able to reproduce that in Gorm or Gemas, but it looks like 
>>>>>>> it's coming from code I've worked on lately...
>>>>>>> Frame 1 of the backtrace is just the line which redirects -drawInRect: 
>>>>>>> to -guiDrawInRect:, but frame 0 has no debug info. Would you be able to 
>>>>>>> recompile -gui (and maybe -back as well) with debugging on (just make 
>>>>>>> clean and make debug=yes), and see if it you can get a better 
>>>>>>> backtrace? Thanks.
>>>>>>> Eric
>>>>>>> 
>>>>>>> 
>>>>>>> On 2011-07-01, at 4:20 PM, Germán Arias wrote:
>>>>>>> 
>>>>>>>> When I select "Containers Palette" the app crash. The backtrace is:
>>>>>>>> 
>>>>>>>> Program received signal SIGSEGV, Segmentation fault.
>>>>>>>> [Switching to Thread 0xb669fb30 (LWP 6463)]
>>>>>>>> 0x0847ac03 in ?? ()
>>>>>>>> (gdb) backtrace
>>>>>>>> #0  0x0847ac03 in ?? ()
>>>>>>>> #1  0xb7344734 in -[NSImage drawInRect:fromRect:operation:fraction:] (
>>>>>>>>    self=0x86385f8, _cmd=0xb759a358, dstRect=
>>>>>>>>        {origin = {x = 0, y = 0}, size = {width = 0, height = 0}},
>>>>>>>> srcRect=
>>>>>>>>        {origin = {x = 0, y = 0}, size = {width = 0, height = 0}},
>>>>>>>>    op=NSCompositeSourceOver, delta=1) at NSImage.m:1297
>>>>>>>> #2  0xb73405dd in -[NSImage drawAtPoint:fromRect:operation:fraction:] (
>>>>>>>>    self=0x86385f8, _cmd=0xb759a320, point={x = 0, y = 0}, srcRect=
>>>>>>>>        {origin = {x = 0, y = 0}, size = {width = 0, height = 0}},
>>>>>>>>    op=NSCompositeSourceOver, delta=1) at NSImage.m:925
>>>>>>>> #3  0xb7344650 in -[NSImage
>>>>>>>> compositeToPoint:fromRect:operation:fraction:] (
>>>>>>>>    self=0x86385f8, _cmd=0xb759a2d0, aPoint={x = 0, y = 9}, srcRect=
>>>>>>>>        {origin = {x = 0, y = 0}, size = {width = 0, height = 0}},
>>>>>>>>    op=NSCompositeSourceOver, delta=1) at NSImage.m:864
>>>>>>>> #4  0xb73402a1 in -[NSImage compositeToPoint:operation:]
>>>>>>>> (self=0x86385f8,
>>>>>>>>    _cmd=0xb756f658, aPoint={x = 0, y = 9}, op=NSCompositeSourceOver)
>>>>>>>>    at NSImage.m:811
>>>>>>>> #5  0xb72df7dc in -[NSCell drawInteriorWithFrame:inView:]
>>>>>>>> (self=0x8600bf0,
>>>>>>>>    _cmd=0xb756f668, cellFrame=
>>>>>>>>        {origin = {x = 0, y = 0}, size = {width = 9, height = 9}},
>>>>>>>>    controlView=0x86393f0) at NSCell.m:2031
>>>>>>>> #6  0x086385f8 in ?? ()
>>>>>>>> #7  0xb756f658 in _OBJC_SELECTOR_TABLE ()
>>>>>>>>   from /usr/GNUstep/Local/Library/Libraries/libgnustep-gui.so.0.21
>>>>>>>> ---Type<return>   to continue, or q<return>   to quit---
>>>>>>>> #8  0x00000000 in ?? ()
>>>>>>>> (gdb)
>>>>>>>> 
>>>>>>>> FisicaLab also crash, when you open the help panel, with a similar
>>>>>>>> backtrace.
> 
> _______________________________________________
> Gnustep-dev mailing list
> address@hidden
> https://lists.gnu.org/mailman/listinfo/gnustep-dev




reply via email to

[Prev in Thread] Current Thread [Next in Thread]