Hi Dave, On Tuesday 10 September 2013 19:59:16 Dave Festing wrote: > I should not have mentioned 1-wire. I was trying to indicate that there > was some USB activity as well as few other processes running in the > background. > > The circumstances stated for having this problem is "when using threads in > their application". As far as I aware my PHP script is not using threads, > so I am confused whether or not my failures could be blamed on this "cache > issue" or on something else not working properly in my system.
It depends on the failures you see. The cache issue happens when you make massive use of threads (that is our experience, and it would also explain why this bug exists since Linux-3.5, nobody uses this massive amount of threads and the bug never bites). So, once again: what kind of failures do you see? With this information we may have a chance to give you some ideas how to solve your problem. > I read through: > http://git-public.pengutronix.de/?p=OSELAS.BSP-Pengutronix-Mini2440.git;a=b >lob;f=configs/platform-friendlyarm-mini2440/patches/linux-3.7/VIVT_processor >s.patch;h=0c67f8184579ba8558884b6ad5027d134c1e3111;hb=af7fa90298181e3419d8da >fa94c6ce7793db7d6d but unfortunately am none the wiser as too how it might > apply to my situation. We have a small test program which triggers this bug in the kernel. If you want take a look into it, I can send it to you. Regards, Juergen -- Pengutronix e.K. | Juergen Beisert | Linux Solutions for Science and Industry | http://www.pengutronix.de/ |
