Valdemar Moreira Pavesi wrote: > Hello all, > > Do you think that the explanation is reasonable ? > > Do you think that just a re-compile can fix it ?
most likely not. if someone's misinterpreting a pointer, that code needs to be fixed. No amount of compiling will change that. > +++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Response > from Vendor: > > We were able to determine that the reason the core files occur when > using libumem is that the utility is misinterpreting a pointer to a > shared memory structure. The structure is allocated as a 1 element > array with additional space dynamically allocated after it. So when > the reference to the additional structures after the original one is > made in the get_next_condition() routine, it cores thinking that the > memory boundary has been past, when in fact, it is not the case at > all. > > > get_next_condition+0x30() > delete_all_conditions+0xa9() > > +++++++++++++++++++++++++++++++++++++++++++++++++++++ is this code by the vendor, or ours? if it's Sun's, you need to proceed with the Service organisation, as has been pointed out. If it's the vendor's, then it's his call. Michael -- Michael Schuster http://blogs.sun.com/recursion Recursion, n.: see 'Recursion'
