[Insight-users] Is the bug tracking page dead?

[Bill Lorensen]

Hans,
We still use the bug tracker, but the activity ebbs and flows with our=
 quarterly contractor meetings. We have one coming up so there will be more=
 activity in the next week or so.

As a user, you are better off for now to post to the list.

Bill

At 03:00 PM 6/7/02 -0500, Hans J. Johnson wrote:
>Hello,
>
>I have submitted a bug to the http://www.itk.org/cgi-bin/gnatsweb.pl bug=
 tracking page, but it looks like that is not being used any longer (Last=
 resolved issue was almost 4 months ago, yet this news group is quite=
 active).
>
>In case the bug page is not being maintained, I have replicated the bug=
 report here.
>
>Regards,
>Hans J. Johnson
>hans-johnson@uiowa.edu
>
>PS: I am just getting started with ITK, so if there is any else I can do to=
 help resolve this issue, please let me know.
>PPS: ITK is GREAT!
>
>
>*Category:* Source
>*Synopsis:* MetaImageIO::CanReadFile Segfaults when checking non-MetaImage=
 File
>*Confidential:* no
>*Severity:* serious
>*Priority:* high
>*Responsible:*
>*State:* open
>*Quarter:*
>*Keywords:*
>*Date-Required:*
>*Class:* bug
>*Submitter-Id:* developers
>*Arrival-Date:* Fri Jun 07 15:50:05 -0400 2002
>*Closed-Date:*
>*Cases:*
>*Last-Modified:*
>*Originator:* Hans J. Johnson
>*Release:* 1.19
>*Organization:* The University of Iowa
>Mental Health Clinical Research Center
>Image Processing Lab
>*Environment:* Athlon, Redhat 7.3, ITK IO
>*Description:* Module: $RCSfile: itkMetaImageIO.cxx,v $
>Version: $Revision: 1.19 $
>
>I have written an Analyze v7.5 file reader. When=
 MetaImageIOFactory::RegisterOne() is not included, the reader works just=
 fine, but when MetaImageIOFactory::RegisterOne() is included, a=
 segmentation fault occurs in MetaImageIO::CanReadFile. The problem is a=
 buffer overrun of the variable key[8000] when the 80MB analyze file is=
 streamed into that variable.
>
>The fix that I propose below has been tested, and works, but if a binary=
 file with more than 8000 bytes and a .mh[da] extension were passed to this=
 file, it would likely segfault, instead of throwing an exception.
>
>Just random thoughts on better fixes:
>A more thorough solution would be to ensure that no more than the number of=
 bytes were streamed into a variable, or to change key[8000] into a variable=
 length structure (i.e. std::string).
>
>maxLineLength and restOfTheLine will have similar problems if the first=
 line is OK, but the rest is broken.
>*File Attachments:*
>*How-To-Repeat:* Try to determine the type of a binary file greater than=
 8000 bytes long when MetaImageIOFactory is the first one registered.
>*Fix:* Replace:
>bool MetaImageIO::CanReadFile( const char* filename )
>{
>WITH:
>bool MetaImageIO::CanReadFile( const char* filename )=B7
>{
>//MetaImage files must have an mha or mhd extension!
>{
>const std::string tempfilename=3Dfilename;
>if( tempfilename =3D=3D "" ||
>!( tempfilename.find(".mha") < tempfilename.length() ||
>tempfilename.find(".mhd") < tempfilename.length() ) )
>{
>return false;
>}
>}
>
>
>
>
>
>
>_______________________________________________
>Insight-users mailing list
>Insight-users@public.kitware.com
>http://public.kitware.com/mailman/listinfo/insight-users