Hello all,
I have been posting quite often recently regarding a possible failure of the RTX operating system. I am working on a product that should run for years without reset and have been experiencing frequent failures (see here:http://www.keil.com/forum/docs/thread15346.asp and here: http://www.keil.com/forum/docs/thread15337.asp). I feel compelled to inform you in a separate thread that the LPC2478 revision C most likely contains a bug in its MAM unit, which results in sporadic program failures (the exact conditions are described in the links; I will post more once NXP responds to Keil's Franc Urbanc report. I have been testing on his behalf today and we are convinced that this is indeed a hardware failure) without any prior warning. This is not documented in the errata sheet yet. It may be that this only occurs when operating at 72[MHz] - but one thing is for sure: if you can afford turning off the MAM - do it! It will prevent the failure, just as running your entire program in SRAM. It is absolutely essential that you take proper measures if you have such a chip in use and you are working on a mission critical system. I will add further update on this matter as I learn more about this problem.