On Mon, 2005-10-17 at 08:33 -0700, Mark Knecht wrote: [cut] > > I'll wait the next rt release (rt7) to see if the problem remains. > > In my experience waiting for -rt7 will almost certianly not fix a > problem like this. > > Specifically, where did it hand? What's on the screen? Does > /var/log/messages hold the last boot log on your machine? If so there > may be some clues there. It hangs on the last 2 or 3 services (few seconds before the login). I've checked /var/log/messages, but the freeze doesn't leave any trace. There's no error... the log simply stops when the system hangs. And the log seem very similar to the 2.6.14-rc4-rt1 one which boots correctly although sometime gives me some errors. > Probably you need to change the kernel config slightly to get beyond this. Well, I simply make an oldconfig upon a well functioning 2.6.14-rc4 vanilla kernel. I've read each new voice and I've ended up answering always no (except for the "Complete Preemption" choice). This morning I've tried the rt8 patch and I can't even compile the kernel! This is the error: CC kernel/ktimers.o kernel/ktimers.c: In function 'check_ktimer_signal': kernel/ktimers.c:1209: warning: passing argument 1 of 'unlock_ktimer_base' from incompatible pointer type kernel/ktimers.c:1225: error: 'struct ktimer_base' has no member named 'getoffset' make[2]: *** [kernel/ktimers.o] Error 1 Maybe today there will be some others releases. > - Mark Cheers, ~ Antonio
