Re: Serious GORM bug

[Jamie Ramone]

OK, here it is:

(gdb) file /SystemApps/Gorm.app/Gorm
Reading symbols from /SystemApps/Gorm.app/Gorm...done.
(gdb) r
Starting program: /SystemApps/Gorm.app/Gorm
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".

Program received signal SIGSEGV, Segmentation fault.
-[NSWindow sendEvent:] (self=0x0, _cmd=<optimized out>, theEvent=0xe121e0)
    at NSWindow.m:4414
4414    NSWindow.m: No such file or directory.
(gdb) bt
#0  -[NSWindow sendEvent:] (self=0x0, _cmd=<optimized out>, theEvent=0xe121e0)
    at NSWindow.m:4414
#1  0x00007ffff71ef09c in -[GSDragView(Private) _handleDrag:slidePoint:] (
    self=0xc55d00, _cmd=<optimized out>, theEvent=0x1061780, slidePoint=...)
    at GSDragView.m:720
#2  0x00007ffff71ed20e in -[GSDragView dragImage:at:offset:event:pasteboard:source:slideBack:] (self=0xc55d00, _cmd=<optimized out>, anImage=0x9dc150,
    screenLocation=..., initialOffset=..., event=0x11073a0,
    pboard=<optimized out>, sourceObject=0xf2e5f0, slideFlag=1 '\001')
    at GSDragView.m:290
#3  0x00007ffff045e344 in -[XGDragView dragImage:at:offset:event:pasteboard:source:slideBack:] (self=0xc55d00, _cmd=<optimized out>, anImage=0x9dc150,
    screenLocation=..., initialOffset=..., event=0x11073a0, pboard=0x9d9470,
    sourceObject=0xf2e5f0, slideFlag=1 '\001') at XGDragView.m:228
#4  0x00007ffff71bebda in -[NSWindow dragImage:at:offset:event:pasteboard:source:slideBack:] (self=0xb140b0, _cmd=<optimized out>, anImage=0x9dc150,
    baseLocation=..., initialOffset=..., event=0x11073a0,
    pboard=<optimized out>, sourceObject=<optimized out>, slideFlag=1 '\001')
    at NSWindow.m:4674
#5  0x00007ffff71a8c08 in -[NSView dragImage:at:offset:event:pasteboard:source:slideBack:] (self=0xf2e5f0, _cmd=<optimized out>, anImage=0x9dc150,
---Type <return> to continue, or q <return> to quit---
    viewLocation=..., initialOffset=..., event=0x11073a0,
    pboard=<optimized out>, sourceObject=<optimized out>, slideFlag=1 '\001')
    at NSView.m:3860
#6  0x00007ffff7b2983c in -[GormObjectEditor mouseDown:] (self=0xf2e5f0,
    _cmd=<optimized out>, theEvent=0x11073a0) at GormObjectEditor.m:481
#7  0x00007ffff71ca953 in -[NSWindow sendEvent:] (self=0xb140b0,
    _cmd=<optimized out>, theEvent=0x11073a0) at NSWindow.m:3896
#8  0x00007ffff70343e3 in -[NSApplication run] (self=0x8c8450,
    _cmd=<optimized out>) at NSApplication.m:1562
#9  0x00007ffff70130d5 in NSApplicationMain (argc=<optimized out>,
    argv=<optimized out>) at Functions.m:91
#10 0x00007ffff5eab76d in __libc_start_main ()
   from /lib/x86_64-linux-gnu/libc.so.6
#11 0x0000000000401965 in _start ()
(gdb)

On Sun, Dec 29, 2013 at 6:50 PM, Fred Kiefer <address@hidden> wrote:

Yes! That was what I was asking for.

Thank you,
Fred

On 29.12.2013 22:28, Jamie Ramone wrote:
> I'm not sure how much more details you need. I stated that dragging a
> connection to any of the objects in the document window (i.e. the main
> project window) caused a segfault. I later discovered that dragging from
> these objects toward any other one, outside that window, worked OK. And it
> seems to be present in the current version of Gorm and updating the GNUstep
> libs didn't alleviate the problem, which says to me that it's
> Gorm-specific. Oh and the GDB backtrace is from the old code. Would it help
> to make another one with the new code? If so just let me know and I'll
> produce one.
>
>
> On Sun, Dec 29, 2013 at 6:20 PM, Fred Kiefer <address@hidden> wrote:
>
>> On 29.12.2013 21:58, Gregory Casamento wrote:
>>> Between windows or between applications.  I believe you’re hitting
>>> the same bug Fred may be hitting, I’m working on a potential fix
>>> now.
>>
>> You might be wrong here. German's bug, that I investigated, was Gorm
>> specific. If you advice Jamie to drag between another application and
>> GWorkspace it may never trigger the same bug. It might trigger a similar
>> bug in GWorkspace if it has similar code to Gorm, but how likely is that?
>>
>> And I wasn't able to reproduce Jamie's bug with Gorm. But then he never
>> gave enough details to be sure.
>>
>> As you may remember German's bug did not include a segmentation fault,
>> which is the symptom Jamie is getting. What I would like to see is a
>> stack trace from Jamie with the current GNUstep code. With his old one I
>> was never sure whether I was looking at the same line of code. For me
>> NSWindow.m:4288 is in the middle of the GSPerformVoidDragSelector macro,
>> which isn't very likely.
>>
>> I really would like to establish the facts before coming to conclusions
>> about the nature of the bug.