On Monday 21 January 2008, Mateo Obregon wrote: > > >> > We did manage to get it to load a Debian Sarge Netinst CD, but > > >> > this kernel only uses 1 CPU. > > UPDATE: I installed the SARGE packages (because the ETCH GCC 4 compiler > would not compile the 2.6.8 kernel) and recompiled 2.6.8 with SMP > (Pentium III XEON) and it is working fine so far. > > But, what could be the difference with the 2.6.18 kernels that make this > machine reboot in the booting of 2.6.18? I'm afraid that we are not the right people to answer that question. The differences between the 2.6.8 kernel and 2.6.18 are _huge_. The most likely reason is that more recent kernels just haven't been tested the hardware you have and that a regression has been introduced at some point that has not been noticed. Your best option is to report your issue in the kernel bugzilla [1] so that the upstream kernel developers, but that only makes sense if you're willing to follow up on it. They will most likely ask you to try a current kernel (2.6.23 or 2.6.24) and to try to pinpoint when exactly the regression was introduced. I'm afraid there are no easy solutions for issues like this and there is little we in Debian can do about them. Cheers, FJP [1] http://bugzilla.kernel.org/
Attachment:
signature.asc
Description: This is a digitally signed message part.