another possible mem leak?

[olafvdspek]

another possible mem leak?

[GoogleCodeExporter]

sorry, "global template_root_directory_" should be "static 
template_root_directory_"

Original comment by [email protected] on 10 Feb 2009 at 3:40

[GoogleCodeExporter]

Just to make sure I understand: valgrind is declaring that a global pointer is 
leaked
because it's not deleted at exit?  That seems wrong to me: it's not leaked, 
it's just
cleaned up by the O/S instead of the application. :-)  Nothing wrong with that.

Does valgrind have a mode where it will declare memory 'live' if it's 
accessible from
global variables at program-exit, and thus not report the leak?  If so, I think 
the
best solution is to use valgrind in that mode.  I don't plan to add 
housekeeping code
to ctemplate that performs no useful purpose except to quiet a memory checker.

Original comment by [email protected] on 10 Feb 2009 at 10:43

• Changed state: WontFix
• Added labels: Type-Defect, Priority-Medium