R devel - Jul 2002 - R 1.5.1 (Windows) crashes on plotting density(rnorm(1000)) (PR#1837)

I am experiencing a reproducible crash from the following statement:

plot(density(rnorm(1000)))

on R 1.5.1 on Windows 2000.

This is a dual-processor machine with 512Mb of RAM.  Please advise if 
further details of hardware/OS are needed.

If rgui.exe is in use, the crash occurs before the plot can be 
displayed.  If rterm.exe is in use, the plot appears, then the error (and 
crash).

A screen capture of the error dialog box  can be viewed/downloaded at

http://shrimp.ccfhrb.noaa.gov/~mprager/rcrash.png

I happen to have Compaq Fortran installed, and therefore a debugger appears 
after the crash.  As I know nothing about systems programming, the debugger 
information (assembler) is meaningless to me, but if it would be useful, I 
can reproduce the crash, save the debugger information, and send it along.

This seems similar to crashes experienced in S-Plus that have been 
attributed to that program's assuming certain behavior on illegal math 
operations, behavior reportedly reset by HP printer drivers.  I do have an 
HP printer installed (hardly avoidable in the PC world).  But that 
observation may be totally irrelevant.

Thanks!

-- 
Michael Prager, Ph.D.                <Mike.Prager@noaa.gov>
NOAA Center for Coastal Fisheries and Habitat Research
Beaufort, North Carolina  28516
http://shrimp.ccfhrb.noaa.gov/~mprager/


-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To:
r-devel-request@stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._

On Fri, 26 Jul 2002 16:16:59 +0200 (MET DST), you wrote:
>I am experiencing a reproducible crash from the following statement:
>
>plot(density(rnorm(1000)))
>
>on R 1.5.1 on Windows 2000.
>
>This is a dual-processor machine with 512Mb of RAM.  Please advise if 
>further details of hardware/OS are needed.
I just tried it in a single processor WinXP machine, with no error.
It's conceivable that the value of the random seed has an effect, but
it's more likely something else.

The invalid operation error can definitely be triggered by code that
messes with the floating point control word.  I hadn't heard that
printer drivers do this, but it is common in video drivers.  Could you
check that you have the latest video driver (and printer driver)
installed?

I don't know an easy way to diagnose this error remotely.

Duncan Murdoch
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To:
r-devel-request@stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._