[R-SIG-Mac] rgl crashes after one successful draw on R.app

[Bryan Hanson]

Unfortunately, I already have XQuartz 2.7.4 and it gives the problem I reported. :-(  Bryan

On May 29, 2013, at 7:16 PM, Robert J Goedman <goedman at icloud.com> wrote:

> Great, like with Rcmdr, example(rgl) seems to work well.
> 
> Repeatedly just 'example(plot3d)' and each time removing all rgl windows, will lead to the same problem Bryan mentioned this morning.
> 
> Rob J. Goedman
> goedman at icloud.com
> 
> 
> 
> 
> On May 29, 2013, at 3:57 PM, Roger Koenker <rkoenker at illinois.edu> wrote:
> 
>> Rob,
>> 
>> Well, that was easy....  I upgraded to XQuartz 2.7.4  and now can run example(rgl) 
>> and also close the windows that are created. ;-).  Many thanks!
>> 
>> Roger
>> 
>> url:    www.econ.uiuc.edu/~roger            Roger Koenker
>> email    rkoenker at uiuc.edu            Department of Economics
>> vox:     217-333-4558                University of Illinois
>> fax:       217-244-6678                Urbana, IL 61801
>> 
>> On May 29, 2013, at 5:41 PM, Robert J Goedman wrote:
>> 
>>> Roger,
>>> 
>>> Not sure if that would fix it, but my system is the patched R.3.0.1, using XQuartz 2.7.4 and a slightly newer version of rgl (rgl_0.93.940).
>>> 
>>> Rob J. Goedman
>>> goedman at icloud.com
>>> 
>>> 
>>> 
>>> 
>>> On May 29, 2013, at 3:19 PM, Roger Koenker <rkoenker at illinois.edu> wrote:
>>> 
>>>> no,  I never get a window....  I get an immediate segfault as soon as
>>>> rgl.open() is called.
>>>> 
>>>> 
>>>> url:    www.econ.uiuc.edu/~roger            Roger Koenker
>>>> email    rkoenker at uiuc.edu            Department of Economics
>>>> vox:     217-333-4558                University of Illinois
>>>> fax:       217-244-6678                Urbana, IL 61801
>>>> 
>>>> On May 29, 2013, at 5:17 PM, Robert J Goedman wrote:
>>>> 
>>>>> Roger,
>>>>> 
>>>>> Do you ever see an rgl window pop-up? Just wondering if this is the same issue.
>>>>> 
>>>>> Rob J. Goedman
>>>>> goedman at icloud.com
>>>>> 
>>>>> 
>>>>> 
>>>>> 
>>>>> On May 29, 2013, at 2:49 PM, Roger Koenker <rkoenker at illinois.edu> wrote:
>>>>> 
>>>>>> I doubt that this is helpful, but I get:
>>>>>> 
>>>>>>> rgl.open()
>>>>>> Reading symbols for shared libraries . done
>>>>>> 
>>>>>> Program received signal EXC_BAD_ACCESS, Could not access memory.
>>>>>> Reason: KERN_INVALID_ADDRESS at address: 0x00000000000000e8
>>>>>> 0x0000000102e31a90 in gui::X11GUIFactory::createWindowImpl (this=0x1006871d0, 
>>>>>> window=0x7fff5fbfd330) at x11gui.cpp:612
>>>>>> 612  ::Window xparent = RootWindow(xdisplay, DefaultScreen(xdisplay));
>>>>>> (gdb) Quit
>>>>>> 
>>>>>> url:    www.econ.uiuc.edu/~roger            Roger Koenker
>>>>>> email    rkoenker at uiuc.edu            Department of Economics
>>>>>> vox:     217-333-4558                University of Illinois
>>>>>> fax:       217-244-6678                Urbana, IL 61801
>>>>>> 
>>>>>> On May 29, 2013, at 4:37 PM, Duncan Murdoch wrote:
>>>>>> 
>>>>>>> On 13-05-29 11:19 AM, Bryan Hanson wrote:
>>>>>>>> [originally on R-help, reposted here as the most relevent place]
>>>>>>>> 
>>>>>>>> NOTE: problem is with R.app, looks OK on command line R
>>>>>>>> 
>>>>>>>> This is really odd, and probably 100% local to me, but I'm at a loss as to a next step.
>>>>>>>> 
>>>>>>>> After narrowing things down, here's how to reproduce:
>>>>>>> 
>>>>>>> I don't see it, but it looks to me like it may be a timing problem.  I'm 
>>>>>>> guessing the destruction of the old window triggers events in the wrong 
>>>>>>> order and leaves rgl in an unstable state.  Unfortunately, without being 
>>>>>>> able to reproduce this, it looks really hard to track down.  Can anyone 
>>>>>>> reproduce it while running in gdb or some other debugger?
>>>>>>> 
>>>>>>> Duncan Murdoch
>>>>>>> 
>>>>>>>> 
>>>>>>>> library("rgl")
>>>>>>>> showSomething <- function() {
>>>>>>>> open3d()
>>>>>>>> points3d(rnorm(10),rnorm(10),rnorm(10))
>>>>>>>> axes3d()
>>>>>>>> title3d('main','sub','xlab','ylab','zlab')
>>>>>>>> }
>>>>>>>> showSomething() # works as expected.
>>>>>>>> # Close the current rgl window
>>>>>>>> showSomething() # crashes R with the following report:
>>>>>>>> 
>>>>>>>> *** caught segfault ***
>>>>>>>> address 0x0, cause 'unknown'
>>>>>>>> 
>>>>>>>> Traceback:
>>>>>>>> 1: .C(rgl_dev_open, success = FALSE)
>>>>>>>> 2: rgl.open()
>>>>>>>> 3: open3d()
>>>>>>>> 4: showSomething()
>>>>>>>> 
>>>>>>>> This will probably work fine for any of you, it's so simple and fundamental.  If I don't close the open rgl windows, it doesn't crash, and I can execute the function over and over successfully.  And the same thing happens if I just do the commands inside the function individually.  SessionInfo() below.
>>>>>>>> 
>>>>>>>> Perhaps something is corrupt with my X11 window system?  Thanks, Bryan
>>>>>>>> 
>>>>>>>> R version 3.0.1 (2013-05-16)
>>>>>>>> Platform: x86_64-apple-darwin10.8.0 (64-bit)
>>>>>>>> 
>>>>>>>> locale:
>>>>>>>> [1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8
>>>>>>>> 
>>>>>>>> attached base packages:
>>>>>>>> [1] stats     graphics  grDevices utils     datasets  methods   base
>>>>>>>> 
>>>>>>>> other attached packages:
>>>>>>>> [1] rgl_0.93.940 sos_1.3-5    brew_1.0-6
>>>>>>>> _______________________________________________
>>>>>>>> R-SIG-Mac mailing list
>>>>>>>> R-SIG-Mac at r-project.org
>>>>>>>> https://stat.ethz.ch/mailman/listinfo/r-sig-mac
>>>>>>>> 
>>>>>>> 
>>>>>>> _______________________________________________
>>>>>>> R-SIG-Mac mailing list
>>>>>>> R-SIG-Mac at r-project.org
>>>>>>> https://stat.ethz.ch/mailman/listinfo/r-sig-mac
>>>>>> 
>>>>>> _______________________________________________
>>>>>> R-SIG-Mac mailing list
>>>>>> R-SIG-Mac at r-project.org
>>>>>> https://stat.ethz.ch/mailman/listinfo/r-sig-mac
>>>>> 
>>>> 
>>> 
>> 
> 
> 
> 	[[alternative HTML version deleted]]
> 
> _______________________________________________
> R-SIG-Mac mailing list
> R-SIG-Mac at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-mac