Hi, a bit late but now I realize I still have not replied this to the list. This thread has been discused in private e-mails.
> hi folks > i builded glucas 2.8b and then tested. > during test huge always in test # 24 it crashes - segmentation > fault. > when i run test 24 alone it pass it fine. > This was a strange bug not affecting other than openbsd OS, and in some rare circumstances. When Glucas was finishing a job (an exponent or a short test) it does not free the main array in the right moment, so it could be two big arrays allocated in a moment and spend all the allowable memory. Gregory only had managed 64Mb of memory for Glucas and it crashed in huge selftest, but no in alone selftest . The Glucas's users don't have to worry about this problem unless they are checking more than 40M exponents in a limited memory system. I can send a patch to these users by request. I will release the patch, and a new version 2.8c in few days. Of course, like new Mlucas versions, there will be some improvements. The most remarkable feature is for Itanium binaries, It almost doubles the speed from 2.8b to 2.8c. Now, Itanium will be the fastest platform for Glucas (in performance/clock). Regards. Guillermo -- Guillermo Ballester Valor [EMAIL PROTECTED] Granada (Spain) _________________________________________________________________________ Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
